diff options
| author | Moonchild <moonchild@palemoon.org> | 2021-03-03 22:21:27 +0000 |
|---|---|---|
| committer | Moonchild <moonchild@palemoon.org> | 2021-03-03 22:21:27 +0000 |
| commit | ca92d331fc6eca1a279d85dcf462ac666f721654 (patch) | |
| tree | 13c497100a5fe6882d4fdb42ad974e83309f5e0d | |
| parent | 145f031e9e47656eb2143b3c400f806c563b2392 (diff) | |
| download | uxp-libaom-update.tar.gz | |
Issue #1737 - Update build config and mozbuild files.libaom-update
49 files changed, 2904 insertions, 2122 deletions
diff --git a/media/libaom/README_MCP b/media/libaom/README_MCP index 5ae9f96c42..ad04810a30 100644 --- a/media/libaom/README_MCP +++ b/media/libaom/README_MCP @@ -1,6 +1,5 @@ This directory contains build files for the aom video -codec reference implementation. The actual library -source is in $TOPSRCDIR/third_party/aom/ +codec reference implementation. Any patches or additional configuration to be applied to the upstream source should be kept here in the media/libaom @@ -10,4 +9,4 @@ The upstream aom git repository is: https://aomedia.googlesource.com/aom -The git commit ID used was 1e227d41f0616de9548a673a83a21ef990b62591. +The version used was tagged 2.0.2. diff --git a/media/libaom/aom_version.h b/media/libaom/aom_version.h index b791ef64dd..92de763be8 100644 --- a/media/libaom/aom_version.h +++ b/media/libaom/aom_version.h @@ -1,7 +1,7 @@ -#define VERSION_MAJOR 0 -#define VERSION_MINOR 1 -#define VERSION_PATCH 0 +#define VERSION_MAJOR 2 +#define VERSION_MINOR 0 +#define VERSION_PATCH 2 #define VERSION_EXTRA "" #define VERSION_PACKED ((VERSION_MAJOR<<16)|(VERSION_MINOR<<8)|(VERSION_PATCH)) -#define VERSION_STRING_NOSP "v0.1.0" -#define VERSION_STRING " v0.1.0" +#define VERSION_STRING_NOSP "v2.0.2" +#define VERSION_STRING " v2.0.2" diff --git a/media/libaom/config/aom_version.h b/media/libaom/config/aom_version.h index b791ef64dd..92de763be8 100644 --- a/media/libaom/config/aom_version.h +++ b/media/libaom/config/aom_version.h @@ -1,7 +1,7 @@ -#define VERSION_MAJOR 0 -#define VERSION_MINOR 1 -#define VERSION_PATCH 0 +#define VERSION_MAJOR 2 +#define VERSION_MINOR 0 +#define VERSION_PATCH 2 #define VERSION_EXTRA "" #define VERSION_PACKED ((VERSION_MAJOR<<16)|(VERSION_MINOR<<8)|(VERSION_PATCH)) -#define VERSION_STRING_NOSP "v0.1.0" -#define VERSION_STRING " v0.1.0" +#define VERSION_STRING_NOSP "v2.0.2" +#define VERSION_STRING " v2.0.2" diff --git a/media/libaom/config/generic/config/aom_config.asm b/media/libaom/config/generic/config/aom_config.asm index 897eeb4fd1..af2521fa4c 100644 --- a/media/libaom/config/generic/config/aom_config.asm +++ b/media/libaom/config/generic/config/aom_config.asm @@ -1,5 +1,5 @@ ; -; Copyright (c) 2018, Alliance for Open Media. All rights reserved +; Copyright (c) 2021, 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 @@ -14,49 +14,52 @@ ARCH_MIPS equ 0 ARCH_PPC equ 0 ARCH_X86 equ 0 ARCH_X86_64 equ 0 -CONFIG_2PASS_PARTITION_SEARCH_LVL equ 1 CONFIG_ACCOUNTING equ 0 CONFIG_ANALYZER equ 0 CONFIG_AV1_DECODER equ 1 CONFIG_AV1_ENCODER equ 0 +CONFIG_AV1_HIGHBITDEPTH equ 1 CONFIG_BIG_ENDIAN equ 0 CONFIG_BITSTREAM_DEBUG equ 0 CONFIG_COEFFICIENT_RANGE_CHECKING equ 0 -CONFIG_COLLECT_INTER_MODE_RD_STATS equ 0 +CONFIG_COLLECT_COMPONENT_TIMING equ 0 +CONFIG_COLLECT_PARTITION_STATS equ 0 CONFIG_COLLECT_RD_STATS equ 0 CONFIG_DEBUG equ 0 CONFIG_DENOISE equ 1 +CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 equ 1 CONFIG_DIST_8X8 equ 0 CONFIG_ENTROPY_STATS equ 0 -CONFIG_FILEOPTIONS equ 1 -CONFIG_FIX_GF_LENGTH equ 1 -CONFIG_FP_MB_STATS equ 0 CONFIG_GCC equ 1 CONFIG_GCOV equ 0 -CONFIG_GLOBAL_MOTION_SEARCH equ 1 CONFIG_GPROF equ 0 +CONFIG_HTB_TRELLIS equ 0 CONFIG_INSPECTION equ 0 CONFIG_INTERNAL_STATS equ 0 CONFIG_INTER_STATS_ONLY equ 0 CONFIG_LIBYUV equ 0 -CONFIG_LOWBITDEPTH equ 1 +CONFIG_LPF_MASK equ 0 CONFIG_MAX_DECODE_PROFILE equ 2 CONFIG_MISMATCH_DEBUG equ 0 CONFIG_MULTITHREAD equ 1 +CONFIG_NN_V2 equ 0 CONFIG_NORMAL_TILE_MODE equ 0 CONFIG_OS_SUPPORT equ 1 CONFIG_PIC equ 0 CONFIG_RD_DEBUG equ 0 -CONFIG_REDUCED_ENCODER_BORDER equ 0 +CONFIG_REALTIME_ONLY equ 0 CONFIG_RUNTIME_CPU_DETECT equ 1 CONFIG_SHARED equ 0 CONFIG_SHARP_SETTINGS equ 0 CONFIG_SIZE_LIMIT equ 0 CONFIG_SPATIAL_RESAMPLING equ 1 -CONFIG_STATIC equ 1 +CONFIG_SPEED_STATS equ 0 +CONFIG_SUPERRES_IN_RECODE equ 1 +CONFIG_TUNE_VMAF equ 0 CONFIG_WEBM_IO equ 0 DECODE_HEIGHT_LIMIT equ 0 DECODE_WIDTH_LIMIT equ 0 +FORCE_HIGHBITDEPTH_DECODING equ 0 HAVE_AVX equ 0 HAVE_AVX2 equ 0 HAVE_DSPR2 equ 0 diff --git a/media/libaom/config/generic/config/aom_config.h b/media/libaom/config/generic/config/aom_config.h index 54fb243e11..6e300df375 100644 --- a/media/libaom/config/generic/config/aom_config.h +++ b/media/libaom/config/generic/config/aom_config.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018, Alliance for Open Media. All rights reserved + * Copyright (c) 2021, 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 @@ -16,49 +16,52 @@ #define ARCH_PPC 0 #define ARCH_X86 0 #define ARCH_X86_64 0 -#define CONFIG_2PASS_PARTITION_SEARCH_LVL 1 #define CONFIG_ACCOUNTING 0 #define CONFIG_ANALYZER 0 #define CONFIG_AV1_DECODER 1 #define CONFIG_AV1_ENCODER 0 +#define CONFIG_AV1_HIGHBITDEPTH 1 #define CONFIG_BIG_ENDIAN 0 #define CONFIG_BITSTREAM_DEBUG 0 #define CONFIG_COEFFICIENT_RANGE_CHECKING 0 -#define CONFIG_COLLECT_INTER_MODE_RD_STATS 0 +#define CONFIG_COLLECT_COMPONENT_TIMING 0 +#define CONFIG_COLLECT_PARTITION_STATS 0 #define CONFIG_COLLECT_RD_STATS 0 #define CONFIG_DEBUG 0 #define CONFIG_DENOISE 1 +#define CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 1 #define CONFIG_DIST_8X8 0 #define CONFIG_ENTROPY_STATS 0 -#define CONFIG_FILEOPTIONS 1 -#define CONFIG_FIX_GF_LENGTH 1 -#define CONFIG_FP_MB_STATS 0 #define CONFIG_GCC 1 #define CONFIG_GCOV 0 -#define CONFIG_GLOBAL_MOTION_SEARCH 1 #define CONFIG_GPROF 0 +#define CONFIG_HTB_TRELLIS 0 #define CONFIG_INSPECTION 0 #define CONFIG_INTERNAL_STATS 0 #define CONFIG_INTER_STATS_ONLY 0 #define CONFIG_LIBYUV 0 -#define CONFIG_LOWBITDEPTH 1 +#define CONFIG_LPF_MASK 0 #define CONFIG_MAX_DECODE_PROFILE 2 #define CONFIG_MISMATCH_DEBUG 0 #define CONFIG_MULTITHREAD 1 +#define CONFIG_NN_V2 0 #define CONFIG_NORMAL_TILE_MODE 0 #define CONFIG_OS_SUPPORT 1 #define CONFIG_PIC 0 #define CONFIG_RD_DEBUG 0 -#define CONFIG_REDUCED_ENCODER_BORDER 0 +#define CONFIG_REALTIME_ONLY 0 #define CONFIG_RUNTIME_CPU_DETECT 1 #define CONFIG_SHARED 0 #define CONFIG_SHARP_SETTINGS 0 #define CONFIG_SIZE_LIMIT 0 #define CONFIG_SPATIAL_RESAMPLING 1 -#define CONFIG_STATIC 1 +#define CONFIG_SPEED_STATS 0 +#define CONFIG_SUPERRES_IN_RECODE 1 +#define CONFIG_TUNE_VMAF 0 #define CONFIG_WEBM_IO 0 #define DECODE_HEIGHT_LIMIT 0 #define DECODE_WIDTH_LIMIT 0 +#define FORCE_HIGHBITDEPTH_DECODING 0 #define HAVE_AVX 0 #define HAVE_AVX2 0 #define HAVE_DSPR2 0 @@ -76,7 +79,5 @@ #define HAVE_SSSE3 0 #define HAVE_VSX 0 #define HAVE_WXWIDGETS 0 -#define INCLUDE_INSTALL_DIR INSTALLDIR/include #define INLINE inline -#define LIB_INSTALL_DIR INSTALLDIR/lib #endif /* AOM_CONFIG_H_ */ diff --git a/media/libaom/config/generic/config/aom_dsp_rtcd.h b/media/libaom/config/generic/config/aom_dsp_rtcd.h index c3b61b8a8b..f769d1180d 100644 --- a/media/libaom/config/generic/config/aom_dsp_rtcd.h +++ b/media/libaom/config/generic/config/aom_dsp_rtcd.h @@ -25,7 +25,7 @@ extern "C" { void aom_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); #define aom_blend_a64_hmask aom_blend_a64_hmask_c -void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); +void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); #define aom_blend_a64_mask aom_blend_a64_mask_c void aom_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); @@ -37,7 +37,7 @@ void aom_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *ds void aom_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); #define aom_convolve8_vert aom_convolve8_vert_c -void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); +void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); #define aom_convolve_copy aom_convolve_copy_c void aom_dc_128_predictor_16x16_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); @@ -340,25 +340,25 @@ void aom_h_predictor_8x4_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_h_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_h_predictor_8x8 aom_h_predictor_8x8_c -void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); #define aom_highbd_blend_a64_d16_mask aom_highbd_blend_a64_d16_mask_c void aom_highbd_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); #define aom_highbd_blend_a64_hmask aom_highbd_blend_a64_hmask_c -void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); +void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); #define aom_highbd_blend_a64_mask aom_highbd_blend_a64_mask_c void aom_highbd_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); #define aom_highbd_blend_a64_vmask aom_highbd_blend_a64_vmask_c -void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); #define aom_highbd_convolve8_horiz aom_highbd_convolve8_horiz_c -void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); #define aom_highbd_convolve8_vert aom_highbd_convolve8_vert_c -void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); #define aom_highbd_convolve_copy aom_highbd_convolve_copy_c void aom_highbd_dc_128_predictor_16x16_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); @@ -664,7 +664,7 @@ void aom_highbd_h_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_lpf_horizontal_14_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); #define aom_highbd_lpf_horizontal_14 aom_highbd_lpf_horizontal_14_c -void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); #define aom_highbd_lpf_horizontal_14_dual aom_highbd_lpf_horizontal_14_dual_c void aom_highbd_lpf_horizontal_4_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); @@ -1009,7 +1009,7 @@ void aom_highbd_v_predictor_8x4_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_v_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); #define aom_highbd_v_predictor_8x8 aom_highbd_v_predictor_8x8_c -void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); #define aom_lowbd_blend_a64_d16_mask aom_lowbd_blend_a64_d16_mask_c void aom_lpf_horizontal_14_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -1360,9 +1360,6 @@ void aom_v_predictor_8x4_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_v_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_v_predictor_8x8 aom_v_predictor_8x8_c -void av1_round_shift_array_c(int32_t *arr, int size, int bit); -#define av1_round_shift_array av1_round_shift_array_c - void aom_dsp_rtcd(void); #include "config/aom_config.h" diff --git a/media/libaom/config/generic/config/aom_scale_rtcd.h b/media/libaom/config/generic/config/aom_scale_rtcd.h index 1ee9bef3d7..46b018d86a 100644 --- a/media/libaom/config/generic/config/aom_scale_rtcd.h +++ b/media/libaom/config/generic/config/aom_scale_rtcd.h @@ -59,15 +59,27 @@ void aom_yv12_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buf void aom_yv12_extend_frame_borders_c(struct yv12_buffer_config *ybf, const int num_planes); #define aom_yv12_extend_frame_borders aom_yv12_extend_frame_borders_c -void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_coloc_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_u aom_yv12_partial_coloc_copy_u_c + +void aom_yv12_partial_coloc_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_v aom_yv12_partial_coloc_copy_v_c + +void aom_yv12_partial_coloc_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_y aom_yv12_partial_coloc_copy_y_c + +void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_u aom_yv12_partial_copy_u_c -void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_v aom_yv12_partial_copy_v_c -void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_ybc, int hstart2, int vstart2); #define aom_yv12_partial_copy_y aom_yv12_partial_copy_y_c +int aom_yv12_realloc_with_new_border_c(struct yv12_buffer_config *ybf, int new_border, int byte_alignment, int num_planes); +#define aom_yv12_realloc_with_new_border aom_yv12_realloc_with_new_border_c + void aom_scale_rtcd(void); #include "config/aom_config.h" diff --git a/media/libaom/config/generic/config/av1_rtcd.h b/media/libaom/config/generic/config/av1_rtcd.h index e2fe11bbdc..dd8536b690 100644 --- a/media/libaom/config/generic/config/av1_rtcd.h +++ b/media/libaom/config/generic/config/av1_rtcd.h @@ -31,28 +31,53 @@ struct txfm_param; struct aom_variance_vtable; struct search_site_config; struct yv12_buffer_config; +struct NN_CONFIG; +typedef struct NN_CONFIG NN_CONFIG; + +enum { NONE, RELU, SOFTSIGN, SIGMOID } UENUM1BYTE(ACTIVATION); +#if CONFIG_NN_V2 +enum { SOFTMAX_CROSS_ENTROPY } UENUM1BYTE(LOSS); +struct NN_CONFIG_V2; +typedef struct NN_CONFIG_V2 NN_CONFIG_V2; +struct FC_LAYER; +typedef struct FC_LAYER FC_LAYER; +#endif // CONFIG_NN_V2 + +struct CNN_CONFIG; +typedef struct CNN_CONFIG CNN_CONFIG; +struct CNN_LAYER_CONFIG; +typedef struct CNN_LAYER_CONFIG CNN_LAYER_CONFIG; +struct CNN_THREAD_DATA; +typedef struct CNN_THREAD_DATA CNN_THREAD_DATA; +struct CNN_BRANCH_CONFIG; +typedef struct CNN_BRANCH_CONFIG CNN_BRANCH_CONFIG; +struct CNN_MULTI_OUT; +typedef struct CNN_MULTI_OUT CNN_MULTI_OUT; /* Function pointers return by CfL functions */ typedef void (*cfl_subsample_lbd_fn)(const uint8_t *input, int input_stride, uint16_t *output_q3); +#if CONFIG_AV1_HIGHBITDEPTH typedef void (*cfl_subsample_hbd_fn)(const uint16_t *input, int input_stride, uint16_t *output_q3); +typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, + int dst_stride, int alpha_q3, int bd); +#endif + typedef void (*cfl_subtract_average_fn)(const uint16_t *src, int16_t *dst); typedef void (*cfl_predict_lbd_fn)(const int16_t *src, uint8_t *dst, int dst_stride, int alpha_q3); -typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, - int dst_stride, int alpha_q3, int bd); #ifdef __cplusplus extern "C" { #endif -void apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -#define apply_selfguided_restoration apply_selfguided_restoration_c +void av1_apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +#define av1_apply_selfguided_restoration av1_apply_selfguided_restoration_c void av1_build_compound_diffwtd_mask_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); #define av1_build_compound_diffwtd_mask av1_build_compound_diffwtd_mask_c @@ -63,24 +88,57 @@ void av1_build_compound_diffwtd_mask_d16_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask void av1_build_compound_diffwtd_mask_highbd_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); #define av1_build_compound_diffwtd_mask_highbd av1_build_compound_diffwtd_mask_highbd_c -void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +int64_t av1_calc_frame_error_c(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +#define av1_calc_frame_error av1_calc_frame_error_c + +void av1_cnn_activate_c( float **input, int channels, int width, int height, int stride, ACTIVATION layer_activation); +#define av1_cnn_activate av1_cnn_activate_c + +void av1_cnn_add_c( float **input, int channels, int width, int height, int stride, const float **add); +#define av1_cnn_add av1_cnn_add_c + +void av1_cnn_batchnorm_c(float **image, int channels, int width, int height, int stride, const float *gamma, const float *beta, const float *mean, const float *std); +#define av1_cnn_batchnorm av1_cnn_batchnorm_c + +void av1_cnn_convolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride, int start_idx, int step); +#define av1_cnn_convolve av1_cnn_convolve_c + +void av1_cnn_deconvolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride); +#define av1_cnn_deconvolve av1_cnn_deconvolve_c + +void av1_cnn_predict_c( const float **input, int in_width, int in_height, int in_stride, const CNN_CONFIG *cnn_config, const CNN_THREAD_DATA *thread_data, CNN_MULTI_OUT *output_struct); +#define av1_cnn_predict av1_cnn_predict_c + +void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); #define av1_convolve_2d_copy_sr av1_convolve_2d_copy_sr_c -void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); +void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); #define av1_convolve_2d_scale av1_convolve_2d_scale_c -void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); #define av1_convolve_2d_sr av1_convolve_2d_sr_c void av1_convolve_horiz_rs_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); #define av1_convolve_horiz_rs av1_convolve_horiz_rs_c -void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); #define av1_convolve_x_sr av1_convolve_x_sr_c -void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); #define av1_convolve_y_sr av1_convolve_y_sr_c +void av1_dist_wtd_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +#define av1_dist_wtd_convolve_2d av1_dist_wtd_convolve_2d_c + +void av1_dist_wtd_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +#define av1_dist_wtd_convolve_2d_copy av1_dist_wtd_convolve_2d_copy_c + +void av1_dist_wtd_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +#define av1_dist_wtd_convolve_x av1_dist_wtd_convolve_x_c + +void av1_dist_wtd_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +#define av1_dist_wtd_convolve_y av1_dist_wtd_convolve_y_c + void av1_dr_prediction_z1_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); #define av1_dr_prediction_z1 av1_dr_prediction_z1_c @@ -108,13 +166,13 @@ void av1_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint void av1_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve8_vert av1_highbd_convolve8_vert_c -void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_2d_copy_sr av1_highbd_convolve_2d_copy_sr_c -void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_2d_scale av1_highbd_convolve_2d_scale_c -void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_2d_sr av1_highbd_convolve_2d_sr_c void av1_highbd_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); @@ -126,12 +184,24 @@ void av1_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_ void av1_highbd_convolve_horiz_rs_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); #define av1_highbd_convolve_horiz_rs av1_highbd_convolve_horiz_rs_c -void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_x_sr av1_highbd_convolve_x_sr_c -void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_y_sr av1_highbd_convolve_y_sr_c +void av1_highbd_dist_wtd_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +#define av1_highbd_dist_wtd_convolve_2d av1_highbd_dist_wtd_convolve_2d_c + +void av1_highbd_dist_wtd_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +#define av1_highbd_dist_wtd_convolve_2d_copy av1_highbd_dist_wtd_convolve_2d_copy_c + +void av1_highbd_dist_wtd_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +#define av1_highbd_dist_wtd_convolve_x av1_highbd_dist_wtd_convolve_x_c + +void av1_highbd_dist_wtd_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +#define av1_highbd_dist_wtd_convolve_y av1_highbd_dist_wtd_convolve_y_c + void av1_highbd_dr_prediction_z1_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); #define av1_highbd_dr_prediction_z1 av1_highbd_dr_prediction_z1_c @@ -141,25 +211,25 @@ void av1_highbd_dr_prediction_z2_c(uint16_t *dst, ptrdiff_t stride, int bw, int void av1_highbd_dr_prediction_z3_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); #define av1_highbd_dr_prediction_z3 av1_highbd_dr_prediction_z3_c -void av1_highbd_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); #define av1_highbd_inv_txfm_add av1_highbd_inv_txfm_add_c -void av1_highbd_inv_txfm_add_16x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -#define av1_highbd_inv_txfm_add_16x16 av1_highbd_inv_txfm_add_16x16_c - -void av1_highbd_inv_txfm_add_16x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -#define av1_highbd_inv_txfm_add_16x8 av1_highbd_inv_txfm_add_16x8_c +void av1_highbd_inv_txfm_add_16x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +#define av1_highbd_inv_txfm_add_16x4 av1_highbd_inv_txfm_add_16x4_c -void av1_highbd_inv_txfm_add_32x32_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -#define av1_highbd_inv_txfm_add_32x32 av1_highbd_inv_txfm_add_32x32_c +void av1_highbd_inv_txfm_add_4x16_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +#define av1_highbd_inv_txfm_add_4x16 av1_highbd_inv_txfm_add_4x16_c -void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); #define av1_highbd_inv_txfm_add_4x4 av1_highbd_inv_txfm_add_4x4_c -void av1_highbd_inv_txfm_add_8x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -#define av1_highbd_inv_txfm_add_8x16 av1_highbd_inv_txfm_add_8x16_c +void av1_highbd_inv_txfm_add_4x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +#define av1_highbd_inv_txfm_add_4x8 av1_highbd_inv_txfm_add_4x8_c -void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +#define av1_highbd_inv_txfm_add_8x4 av1_highbd_inv_txfm_add_8x4_c + +void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); #define av1_highbd_inv_txfm_add_8x8 av1_highbd_inv_txfm_add_8x8_c void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); @@ -168,22 +238,10 @@ void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int des void av1_highbd_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_1_add av1_highbd_iwht4x4_1_add_c -void av1_highbd_jnt_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -#define av1_highbd_jnt_convolve_2d av1_highbd_jnt_convolve_2d_c - -void av1_highbd_jnt_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -#define av1_highbd_jnt_convolve_2d_copy av1_highbd_jnt_convolve_2d_copy_c - -void av1_highbd_jnt_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -#define av1_highbd_jnt_convolve_x av1_highbd_jnt_convolve_x_c - -void av1_highbd_jnt_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -#define av1_highbd_jnt_convolve_y av1_highbd_jnt_convolve_y_c - void av1_highbd_warp_affine_c(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); #define av1_highbd_warp_affine av1_highbd_warp_affine_c -void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); +void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); #define av1_highbd_wiener_convolve_add_src av1_highbd_wiener_convolve_add_src_c void av1_inv_txfm2d_add_16x16_c(const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type, int bd); @@ -246,17 +304,8 @@ void av1_inv_txfm2d_add_8x8_c(const int32_t *input, uint16_t *output, int stride void av1_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); #define av1_inv_txfm_add av1_inv_txfm_add_c -void av1_jnt_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -#define av1_jnt_convolve_2d av1_jnt_convolve_2d_c - -void av1_jnt_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -#define av1_jnt_convolve_2d_copy av1_jnt_convolve_2d_copy_c - -void av1_jnt_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -#define av1_jnt_convolve_x av1_jnt_convolve_x_c - -void av1_jnt_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -#define av1_jnt_convolve_y av1_jnt_convolve_y_c +void av1_round_shift_array_c(int32_t *arr, int size, int bit); +#define av1_round_shift_array av1_round_shift_array_c int av1_selfguided_restoration_c(const uint8_t *dgd8, int width, int height, int dgd_stride, int32_t *flt0, int32_t *flt1, int flt_stride, @@ -275,7 +324,13 @@ void av1_warp_affine_c(const int32_t *mat, const uint8_t *ref, int width, int he void av1_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); #define av1_wiener_convolve_add_src av1_wiener_convolve_add_src_c -void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); +void cdef_copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +#define cdef_copy_rect8_16bit_to_16bit cdef_copy_rect8_16bit_to_16bit_c + +void cdef_copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +#define cdef_copy_rect8_8bit_to_16bit cdef_copy_rect8_8bit_to_16bit_c + +void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); #define cdef_filter_block cdef_filter_block_c int cdef_find_dir_c(const uint16_t *img, int stride, int32_t *var, int coeff_shift); @@ -299,20 +354,14 @@ cfl_subsample_hbd_fn cfl_get_luma_subsampling_444_hbd_c(TX_SIZE tx_size); cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_c(TX_SIZE tx_size); #define cfl_get_luma_subsampling_444_lbd cfl_get_luma_subsampling_444_lbd_c -void copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -#define copy_rect8_16bit_to_16bit copy_rect8_16bit_to_16bit_c - -void copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -#define copy_rect8_8bit_to_16bit copy_rect8_8bit_to_16bit_c - -cfl_predict_hbd_fn get_predict_hbd_fn_c(TX_SIZE tx_size); -#define get_predict_hbd_fn get_predict_hbd_fn_c +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_c(TX_SIZE tx_size); +#define cfl_get_predict_hbd_fn cfl_get_predict_hbd_fn_c -cfl_predict_lbd_fn get_predict_lbd_fn_c(TX_SIZE tx_size); -#define get_predict_lbd_fn get_predict_lbd_fn_c +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_c(TX_SIZE tx_size); +#define cfl_get_predict_lbd_fn cfl_get_predict_lbd_fn_c -cfl_subtract_average_fn get_subtract_average_fn_c(TX_SIZE tx_size); -#define get_subtract_average_fn get_subtract_average_fn_c +cfl_subtract_average_fn cfl_get_subtract_average_fn_c(TX_SIZE tx_size); +#define cfl_get_subtract_average_fn cfl_get_subtract_average_fn_c void av1_rtcd(void); diff --git a/media/libaom/config/linux/arm/config/aom_config.asm b/media/libaom/config/linux/arm/config/aom_config.asm index f5db6f6ca6..a5d8cc9b99 100644 --- a/media/libaom/config/linux/arm/config/aom_config.asm +++ b/media/libaom/config/linux/arm/config/aom_config.asm @@ -1,5 +1,5 @@ ; -; Copyright (c) 2018, Alliance for Open Media. All rights reserved +; Copyright (c) 2021, 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 @@ -14,49 +14,52 @@ .equ ARCH_PPC, 0 .equ ARCH_X86, 0 .equ ARCH_X86_64, 0 -.equ CONFIG_2PASS_PARTITION_SEARCH_LVL, 1 .equ CONFIG_ACCOUNTING, 0 .equ CONFIG_ANALYZER, 0 .equ CONFIG_AV1_DECODER, 1 .equ CONFIG_AV1_ENCODER, 0 +.equ CONFIG_AV1_HIGHBITDEPTH, 1 .equ CONFIG_BIG_ENDIAN, 0 .equ CONFIG_BITSTREAM_DEBUG, 0 .equ CONFIG_COEFFICIENT_RANGE_CHECKING, 0 -.equ CONFIG_COLLECT_INTER_MODE_RD_STATS, 0 +.equ CONFIG_COLLECT_COMPONENT_TIMING, 0 +.equ CONFIG_COLLECT_PARTITION_STATS, 0 .equ CONFIG_COLLECT_RD_STATS, 0 .equ CONFIG_DEBUG, 0 .equ CONFIG_DENOISE, 1 +.equ CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8, 1 .equ CONFIG_DIST_8X8, 0 .equ CONFIG_ENTROPY_STATS, 0 -.equ CONFIG_FILEOPTIONS, 1 -.equ CONFIG_FIX_GF_LENGTH, 1 -.equ CONFIG_FP_MB_STATS, 0 .equ CONFIG_GCC, 1 .equ CONFIG_GCOV, 0 -.equ CONFIG_GLOBAL_MOTION_SEARCH, 1 .equ CONFIG_GPROF, 0 +.equ CONFIG_HTB_TRELLIS, 0 .equ CONFIG_INSPECTION, 0 .equ CONFIG_INTERNAL_STATS, 0 .equ CONFIG_INTER_STATS_ONLY, 0 .equ CONFIG_LIBYUV, 0 -.equ CONFIG_LOWBITDEPTH, 1 +.equ CONFIG_LPF_MASK, 0 .equ CONFIG_MAX_DECODE_PROFILE, 2 .equ CONFIG_MISMATCH_DEBUG, 0 .equ CONFIG_MULTITHREAD, 1 +.equ CONFIG_NN_V2, 0 .equ CONFIG_NORMAL_TILE_MODE, 0 .equ CONFIG_OS_SUPPORT, 1 .equ CONFIG_PIC, 1 .equ CONFIG_RD_DEBUG, 0 -.equ CONFIG_REDUCED_ENCODER_BORDER, 0 +.equ CONFIG_REALTIME_ONLY, 0 .equ CONFIG_RUNTIME_CPU_DETECT, 1 .equ CONFIG_SHARED, 0 .equ CONFIG_SHARP_SETTINGS, 0 .equ CONFIG_SIZE_LIMIT, 0 .equ CONFIG_SPATIAL_RESAMPLING, 1 -.equ CONFIG_STATIC, 1 +.equ CONFIG_SPEED_STATS, 0 +.equ CONFIG_SUPERRES_IN_RECODE, 1 +.equ CONFIG_TUNE_VMAF, 0 .equ CONFIG_WEBM_IO, 0 .equ DECODE_HEIGHT_LIMIT, 0 .equ DECODE_WIDTH_LIMIT, 0 +.equ FORCE_HIGHBITDEPTH_DECODING, 0 .equ HAVE_AVX, 0 .equ HAVE_AVX2, 0 .equ HAVE_DSPR2, 0 diff --git a/media/libaom/config/linux/arm/config/aom_config.h b/media/libaom/config/linux/arm/config/aom_config.h index 850eebfbd4..5fc909a91c 100644 --- a/media/libaom/config/linux/arm/config/aom_config.h +++ b/media/libaom/config/linux/arm/config/aom_config.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018, Alliance for Open Media. All rights reserved + * Copyright (c) 2021, 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 @@ -16,49 +16,52 @@ #define ARCH_PPC 0 #define ARCH_X86 0 #define ARCH_X86_64 0 -#define CONFIG_2PASS_PARTITION_SEARCH_LVL 1 #define CONFIG_ACCOUNTING 0 #define CONFIG_ANALYZER 0 #define CONFIG_AV1_DECODER 1 #define CONFIG_AV1_ENCODER 0 +#define CONFIG_AV1_HIGHBITDEPTH 1 #define CONFIG_BIG_ENDIAN 0 #define CONFIG_BITSTREAM_DEBUG 0 #define CONFIG_COEFFICIENT_RANGE_CHECKING 0 -#define CONFIG_COLLECT_INTER_MODE_RD_STATS 0 +#define CONFIG_COLLECT_COMPONENT_TIMING 0 +#define CONFIG_COLLECT_PARTITION_STATS 0 #define CONFIG_COLLECT_RD_STATS 0 #define CONFIG_DEBUG 0 #define CONFIG_DENOISE 1 +#define CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 1 #define CONFIG_DIST_8X8 0 #define CONFIG_ENTROPY_STATS 0 -#define CONFIG_FILEOPTIONS 1 -#define CONFIG_FIX_GF_LENGTH 1 -#define CONFIG_FP_MB_STATS 0 #define CONFIG_GCC 1 #define CONFIG_GCOV 0 -#define CONFIG_GLOBAL_MOTION_SEARCH 1 #define CONFIG_GPROF 0 +#define CONFIG_HTB_TRELLIS 0 #define CONFIG_INSPECTION 0 #define CONFIG_INTERNAL_STATS 0 #define CONFIG_INTER_STATS_ONLY 0 #define CONFIG_LIBYUV 0 -#define CONFIG_LOWBITDEPTH 1 +#define CONFIG_LPF_MASK 0 #define CONFIG_MAX_DECODE_PROFILE 2 #define CONFIG_MISMATCH_DEBUG 0 #define CONFIG_MULTITHREAD 1 +#define CONFIG_NN_V2 0 #define CONFIG_NORMAL_TILE_MODE 0 #define CONFIG_OS_SUPPORT 1 #define CONFIG_PIC 1 #define CONFIG_RD_DEBUG 0 -#define CONFIG_REDUCED_ENCODER_BORDER 0 +#define CONFIG_REALTIME_ONLY 0 #define CONFIG_RUNTIME_CPU_DETECT 1 #define CONFIG_SHARED 0 #define CONFIG_SHARP_SETTINGS 0 #define CONFIG_SIZE_LIMIT 0 #define CONFIG_SPATIAL_RESAMPLING 1 -#define CONFIG_STATIC 1 +#define CONFIG_SPEED_STATS 0 +#define CONFIG_SUPERRES_IN_RECODE 1 +#define CONFIG_TUNE_VMAF 0 #define CONFIG_WEBM_IO 0 #define DECODE_HEIGHT_LIMIT 0 #define DECODE_WIDTH_LIMIT 0 +#define FORCE_HIGHBITDEPTH_DECODING 0 #define HAVE_AVX 0 #define HAVE_AVX2 0 #define HAVE_DSPR2 0 @@ -76,7 +79,5 @@ #define HAVE_SSSE3 0 #define HAVE_VSX 0 #define HAVE_WXWIDGETS 0 -#define INCLUDE_INSTALL_DIR INSTALLDIR/include #define INLINE inline -#define LIB_INSTALL_DIR INSTALLDIR/lib #endif /* AOM_CONFIG_H_ */ diff --git a/media/libaom/config/linux/arm/config/aom_dsp_rtcd.h b/media/libaom/config/linux/arm/config/aom_dsp_rtcd.h index d80f8972b1..c6e420acba 100644 --- a/media/libaom/config/linux/arm/config/aom_dsp_rtcd.h +++ b/media/libaom/config/linux/arm/config/aom_dsp_rtcd.h @@ -26,7 +26,7 @@ void aom_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src void aom_blend_a64_hmask_neon(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); RTCD_EXTERN void (*aom_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); -void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); +void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); #define aom_blend_a64_mask aom_blend_a64_mask_c void aom_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); @@ -39,7 +39,7 @@ void aom_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *ds void aom_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); #define aom_convolve8_vert aom_convolve8_vert_c -void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); +void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); #define aom_convolve_copy aom_convolve_copy_c void aom_dc_128_predictor_16x16_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); @@ -362,25 +362,25 @@ void aom_h_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_h_predictor_8x8_neon(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); RTCD_EXTERN void (*aom_h_predictor_8x8)(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); -void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); #define aom_highbd_blend_a64_d16_mask aom_highbd_blend_a64_d16_mask_c void aom_highbd_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); #define aom_highbd_blend_a64_hmask aom_highbd_blend_a64_hmask_c -void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); +void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); #define aom_highbd_blend_a64_mask aom_highbd_blend_a64_mask_c void aom_highbd_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); #define aom_highbd_blend_a64_vmask aom_highbd_blend_a64_vmask_c -void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); #define aom_highbd_convolve8_horiz aom_highbd_convolve8_horiz_c -void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); #define aom_highbd_convolve8_vert aom_highbd_convolve8_vert_c -void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); #define aom_highbd_convolve_copy aom_highbd_convolve_copy_c void aom_highbd_dc_128_predictor_16x16_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); @@ -691,7 +691,7 @@ void aom_highbd_h_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_lpf_horizontal_14_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); #define aom_highbd_lpf_horizontal_14 aom_highbd_lpf_horizontal_14_c -void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); #define aom_highbd_lpf_horizontal_14_dual aom_highbd_lpf_horizontal_14_dual_c void aom_highbd_lpf_horizontal_4_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); @@ -1036,9 +1036,9 @@ void aom_highbd_v_predictor_8x4_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_v_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); #define aom_highbd_v_predictor_8x8 aom_highbd_v_predictor_8x8_c -void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_neon(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_neon(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); void aom_lpf_horizontal_14_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void aom_lpf_horizontal_14_neon(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -1400,10 +1400,6 @@ void aom_v_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_v_predictor_8x8_neon(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); RTCD_EXTERN void (*aom_v_predictor_8x8)(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); -void av1_round_shift_array_c(int32_t *arr, int size, int bit); -void av1_round_shift_array_neon(int32_t *arr, int size, int bit); -RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); - void aom_dsp_rtcd(void); #include "config/aom_config.h" @@ -1496,8 +1492,6 @@ static void setup_rtcd_internal(void) if (flags & HAS_NEON) aom_v_predictor_4x4 = aom_v_predictor_4x4_neon; aom_v_predictor_8x8 = aom_v_predictor_8x8_c; if (flags & HAS_NEON) aom_v_predictor_8x8 = aom_v_predictor_8x8_neon; - av1_round_shift_array = av1_round_shift_array_c; - if (flags & HAS_NEON) av1_round_shift_array = av1_round_shift_array_neon; } #endif diff --git a/media/libaom/config/linux/arm/config/aom_scale_rtcd.h b/media/libaom/config/linux/arm/config/aom_scale_rtcd.h index 165d730936..067ddb4928 100644 --- a/media/libaom/config/linux/arm/config/aom_scale_rtcd.h +++ b/media/libaom/config/linux/arm/config/aom_scale_rtcd.h @@ -59,15 +59,27 @@ void aom_yv12_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buf void aom_yv12_extend_frame_borders_c(struct yv12_buffer_config *ybf, const int num_planes); #define aom_yv12_extend_frame_borders aom_yv12_extend_frame_borders_c -void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_coloc_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_u aom_yv12_partial_coloc_copy_u_c + +void aom_yv12_partial_coloc_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_v aom_yv12_partial_coloc_copy_v_c + +void aom_yv12_partial_coloc_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_y aom_yv12_partial_coloc_copy_y_c + +void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_u aom_yv12_partial_copy_u_c -void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_v aom_yv12_partial_copy_v_c -void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_ybc, int hstart2, int vstart2); #define aom_yv12_partial_copy_y aom_yv12_partial_copy_y_c +int aom_yv12_realloc_with_new_border_c(struct yv12_buffer_config *ybf, int new_border, int byte_alignment, int num_planes); +#define aom_yv12_realloc_with_new_border aom_yv12_realloc_with_new_border_c + void aom_scale_rtcd(void); #include "config/aom_config.h" diff --git a/media/libaom/config/linux/arm/config/av1_rtcd.h b/media/libaom/config/linux/arm/config/av1_rtcd.h index b54b4514dd..7a346efd26 100644 --- a/media/libaom/config/linux/arm/config/av1_rtcd.h +++ b/media/libaom/config/linux/arm/config/av1_rtcd.h @@ -31,29 +31,54 @@ struct txfm_param; struct aom_variance_vtable; struct search_site_config; struct yv12_buffer_config; +struct NN_CONFIG; +typedef struct NN_CONFIG NN_CONFIG; + +enum { NONE, RELU, SOFTSIGN, SIGMOID } UENUM1BYTE(ACTIVATION); +#if CONFIG_NN_V2 +enum { SOFTMAX_CROSS_ENTROPY } UENUM1BYTE(LOSS); +struct NN_CONFIG_V2; +typedef struct NN_CONFIG_V2 NN_CONFIG_V2; +struct FC_LAYER; +typedef struct FC_LAYER FC_LAYER; +#endif // CONFIG_NN_V2 + +struct CNN_CONFIG; +typedef struct CNN_CONFIG CNN_CONFIG; +struct CNN_LAYER_CONFIG; +typedef struct CNN_LAYER_CONFIG CNN_LAYER_CONFIG; +struct CNN_THREAD_DATA; +typedef struct CNN_THREAD_DATA CNN_THREAD_DATA; +struct CNN_BRANCH_CONFIG; +typedef struct CNN_BRANCH_CONFIG CNN_BRANCH_CONFIG; +struct CNN_MULTI_OUT; +typedef struct CNN_MULTI_OUT CNN_MULTI_OUT; /* Function pointers return by CfL functions */ typedef void (*cfl_subsample_lbd_fn)(const uint8_t *input, int input_stride, uint16_t *output_q3); +#if CONFIG_AV1_HIGHBITDEPTH typedef void (*cfl_subsample_hbd_fn)(const uint16_t *input, int input_stride, uint16_t *output_q3); +typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, + int dst_stride, int alpha_q3, int bd); +#endif + typedef void (*cfl_subtract_average_fn)(const uint16_t *src, int16_t *dst); typedef void (*cfl_predict_lbd_fn)(const int16_t *src, uint8_t *dst, int dst_stride, int alpha_q3); -typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, - int dst_stride, int alpha_q3, int bd); #ifdef __cplusplus extern "C" { #endif -void apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_neon(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -RTCD_EXTERN void (*apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_neon(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +RTCD_EXTERN void (*av1_apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); void av1_build_compound_diffwtd_mask_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); #define av1_build_compound_diffwtd_mask av1_build_compound_diffwtd_mask_c @@ -65,27 +90,64 @@ RTCD_EXTERN void (*av1_build_compound_diffwtd_mask_d16)(uint8_t *mask, DIFFWTD_M void av1_build_compound_diffwtd_mask_highbd_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); #define av1_build_compound_diffwtd_mask_highbd av1_build_compound_diffwtd_mask_highbd_c -void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +int64_t av1_calc_frame_error_c(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +#define av1_calc_frame_error av1_calc_frame_error_c + +void av1_cnn_activate_c( float **input, int channels, int width, int height, int stride, ACTIVATION layer_activation); +#define av1_cnn_activate av1_cnn_activate_c + +void av1_cnn_add_c( float **input, int channels, int width, int height, int stride, const float **add); +#define av1_cnn_add av1_cnn_add_c + +void av1_cnn_batchnorm_c(float **image, int channels, int width, int height, int stride, const float *gamma, const float *beta, const float *mean, const float *std); +#define av1_cnn_batchnorm av1_cnn_batchnorm_c -void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); +void av1_cnn_convolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride, int start_idx, int step); +#define av1_cnn_convolve av1_cnn_convolve_c + +void av1_cnn_deconvolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride); +#define av1_cnn_deconvolve av1_cnn_deconvolve_c + +void av1_cnn_predict_c( const float **input, int in_width, int in_height, int in_stride, const CNN_CONFIG *cnn_config, const CNN_THREAD_DATA *thread_data, CNN_MULTI_OUT *output_struct); +#define av1_cnn_predict av1_cnn_predict_c + +void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); #define av1_convolve_2d_scale av1_convolve_2d_scale_c -void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_convolve_horiz_rs_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); #define av1_convolve_horiz_rs av1_convolve_horiz_rs_c -void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); -void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_dr_prediction_z1_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); #define av1_dr_prediction_z1 av1_dr_prediction_z1_c @@ -114,13 +176,13 @@ void av1_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint void av1_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve8_vert av1_highbd_convolve8_vert_c -void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_2d_copy_sr av1_highbd_convolve_2d_copy_sr_c -void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_2d_scale av1_highbd_convolve_2d_scale_c -void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_2d_sr av1_highbd_convolve_2d_sr_c void av1_highbd_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); @@ -132,12 +194,24 @@ void av1_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_ void av1_highbd_convolve_horiz_rs_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); #define av1_highbd_convolve_horiz_rs av1_highbd_convolve_horiz_rs_c -void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_x_sr av1_highbd_convolve_x_sr_c -void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); #define av1_highbd_convolve_y_sr av1_highbd_convolve_y_sr_c +void av1_highbd_dist_wtd_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +#define av1_highbd_dist_wtd_convolve_2d av1_highbd_dist_wtd_convolve_2d_c + +void av1_highbd_dist_wtd_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +#define av1_highbd_dist_wtd_convolve_2d_copy av1_highbd_dist_wtd_convolve_2d_copy_c + +void av1_highbd_dist_wtd_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +#define av1_highbd_dist_wtd_convolve_x av1_highbd_dist_wtd_convolve_x_c + +void av1_highbd_dist_wtd_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +#define av1_highbd_dist_wtd_convolve_y av1_highbd_dist_wtd_convolve_y_c + void av1_highbd_dr_prediction_z1_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); #define av1_highbd_dr_prediction_z1 av1_highbd_dr_prediction_z1_c @@ -147,25 +221,25 @@ void av1_highbd_dr_prediction_z2_c(uint16_t *dst, ptrdiff_t stride, int bw, int void av1_highbd_dr_prediction_z3_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); #define av1_highbd_dr_prediction_z3 av1_highbd_dr_prediction_z3_c -void av1_highbd_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); #define av1_highbd_inv_txfm_add av1_highbd_inv_txfm_add_c -void av1_highbd_inv_txfm_add_16x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -#define av1_highbd_inv_txfm_add_16x16 av1_highbd_inv_txfm_add_16x16_c - -void av1_highbd_inv_txfm_add_16x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -#define av1_highbd_inv_txfm_add_16x8 av1_highbd_inv_txfm_add_16x8_c +void av1_highbd_inv_txfm_add_16x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +#define av1_highbd_inv_txfm_add_16x4 av1_highbd_inv_txfm_add_16x4_c -void av1_highbd_inv_txfm_add_32x32_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -#define av1_highbd_inv_txfm_add_32x32 av1_highbd_inv_txfm_add_32x32_c +void av1_highbd_inv_txfm_add_4x16_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +#define av1_highbd_inv_txfm_add_4x16 av1_highbd_inv_txfm_add_4x16_c -void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); #define av1_highbd_inv_txfm_add_4x4 av1_highbd_inv_txfm_add_4x4_c -void av1_highbd_inv_txfm_add_8x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -#define av1_highbd_inv_txfm_add_8x16 av1_highbd_inv_txfm_add_8x16_c +void av1_highbd_inv_txfm_add_4x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +#define av1_highbd_inv_txfm_add_4x8 av1_highbd_inv_txfm_add_4x8_c -void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +#define av1_highbd_inv_txfm_add_8x4 av1_highbd_inv_txfm_add_8x4_c + +void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); #define av1_highbd_inv_txfm_add_8x8 av1_highbd_inv_txfm_add_8x8_c void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); @@ -174,22 +248,10 @@ void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int des void av1_highbd_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_1_add av1_highbd_iwht4x4_1_add_c -void av1_highbd_jnt_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -#define av1_highbd_jnt_convolve_2d av1_highbd_jnt_convolve_2d_c - -void av1_highbd_jnt_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -#define av1_highbd_jnt_convolve_2d_copy av1_highbd_jnt_convolve_2d_copy_c - -void av1_highbd_jnt_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -#define av1_highbd_jnt_convolve_x av1_highbd_jnt_convolve_x_c - -void av1_highbd_jnt_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -#define av1_highbd_jnt_convolve_y av1_highbd_jnt_convolve_y_c - void av1_highbd_warp_affine_c(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); #define av1_highbd_warp_affine av1_highbd_warp_affine_c -void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); +void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); #define av1_highbd_wiener_convolve_add_src av1_highbd_wiener_convolve_add_src_c void av1_inv_txfm2d_add_16x16_c(const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type, int bd); @@ -253,21 +315,9 @@ void av1_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, con void av1_inv_txfm_add_neon(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); RTCD_EXTERN void (*av1_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_jnt_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_neon(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_round_shift_array_c(int32_t *arr, int size, int bit); +void av1_round_shift_array_neon(int32_t *arr, int size, int bit); +RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); int av1_selfguided_restoration_c(const uint8_t *dgd8, int width, int height, int dgd_stride, int32_t *flt0, int32_t *flt1, int flt_stride, @@ -293,9 +343,17 @@ void av1_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uin void av1_wiener_convolve_add_src_neon(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); RTCD_EXTERN void (*av1_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); -void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_neon(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); +void cdef_copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_neon(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); + +void cdef_copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_neon(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); + +void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_neon(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); int cdef_find_dir_c(const uint16_t *img, int stride, int32_t *var, int coeff_shift); int cdef_find_dir_neon(const uint16_t *img, int stride, int32_t *var, int coeff_shift); @@ -325,25 +383,17 @@ cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_c(TX_SIZE tx_size); cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_neon(TX_SIZE tx_size); RTCD_EXTERN cfl_subsample_lbd_fn (*cfl_get_luma_subsampling_444_lbd)(TX_SIZE tx_size); -void copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_neon(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); - -void copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_neon(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); - -cfl_predict_hbd_fn get_predict_hbd_fn_c(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_neon(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_hbd_fn (*get_predict_hbd_fn)(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_c(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_neon(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_hbd_fn (*cfl_get_predict_hbd_fn)(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_c(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_neon(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_lbd_fn (*get_predict_lbd_fn)(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_c(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_neon(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_lbd_fn (*cfl_get_predict_lbd_fn)(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_c(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_neon(TX_SIZE tx_size); -RTCD_EXTERN cfl_subtract_average_fn (*get_subtract_average_fn)(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_c(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_neon(TX_SIZE tx_size); +RTCD_EXTERN cfl_subtract_average_fn (*cfl_get_subtract_average_fn)(TX_SIZE tx_size); void av1_rtcd(void); @@ -357,8 +407,8 @@ static void setup_rtcd_internal(void) (void)flags; - apply_selfguided_restoration = apply_selfguided_restoration_c; - if (flags & HAS_NEON) apply_selfguided_restoration = apply_selfguided_restoration_neon; + av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_c; + if (flags & HAS_NEON) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_neon; av1_build_compound_diffwtd_mask_d16 = av1_build_compound_diffwtd_mask_d16_c; if (flags & HAS_NEON) av1_build_compound_diffwtd_mask_d16 = av1_build_compound_diffwtd_mask_d16_neon; av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_c; @@ -369,22 +419,28 @@ static void setup_rtcd_internal(void) if (flags & HAS_NEON) av1_convolve_x_sr = av1_convolve_x_sr_neon; av1_convolve_y_sr = av1_convolve_y_sr_c; if (flags & HAS_NEON) av1_convolve_y_sr = av1_convolve_y_sr_neon; + av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_c; + if (flags & HAS_NEON) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_neon; + av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_NEON) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_neon; + av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_c; + if (flags & HAS_NEON) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_neon; + av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_c; + if (flags & HAS_NEON) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_neon; av1_inv_txfm_add = av1_inv_txfm_add_c; if (flags & HAS_NEON) av1_inv_txfm_add = av1_inv_txfm_add_neon; - av1_jnt_convolve_2d = av1_jnt_convolve_2d_c; - if (flags & HAS_NEON) av1_jnt_convolve_2d = av1_jnt_convolve_2d_neon; - av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_c; - if (flags & HAS_NEON) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_neon; - av1_jnt_convolve_x = av1_jnt_convolve_x_c; - if (flags & HAS_NEON) av1_jnt_convolve_x = av1_jnt_convolve_x_neon; - av1_jnt_convolve_y = av1_jnt_convolve_y_c; - if (flags & HAS_NEON) av1_jnt_convolve_y = av1_jnt_convolve_y_neon; + av1_round_shift_array = av1_round_shift_array_c; + if (flags & HAS_NEON) av1_round_shift_array = av1_round_shift_array_neon; av1_selfguided_restoration = av1_selfguided_restoration_c; if (flags & HAS_NEON) av1_selfguided_restoration = av1_selfguided_restoration_neon; av1_warp_affine = av1_warp_affine_c; if (flags & HAS_NEON) av1_warp_affine = av1_warp_affine_neon; av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_c; if (flags & HAS_NEON) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_neon; + cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_c; + if (flags & HAS_NEON) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_neon; + cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_c; + if (flags & HAS_NEON) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_neon; cdef_filter_block = cdef_filter_block_c; if (flags & HAS_NEON) cdef_filter_block = cdef_filter_block_neon; cdef_find_dir = cdef_find_dir_c; @@ -401,16 +457,12 @@ static void setup_rtcd_internal(void) if (flags & HAS_NEON) cfl_get_luma_subsampling_444_hbd = cfl_get_luma_subsampling_444_hbd_neon; cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_c; if (flags & HAS_NEON) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_neon; - copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_c; - if (flags & HAS_NEON) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_neon; - copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_c; - if (flags & HAS_NEON) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_neon; - get_predict_hbd_fn = get_predict_hbd_fn_c; - if (flags & HAS_NEON) get_predict_hbd_fn = get_predict_hbd_fn_neon; - get_predict_lbd_fn = get_predict_lbd_fn_c; - if (flags & HAS_NEON) get_predict_lbd_fn = get_predict_lbd_fn_neon; - get_subtract_average_fn = get_subtract_average_fn_c; - if (flags & HAS_NEON) get_subtract_average_fn = get_subtract_average_fn_neon; + cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_c; + if (flags & HAS_NEON) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_neon; + cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_c; + if (flags & HAS_NEON) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_neon; + cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_c; + if (flags & HAS_NEON) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_neon; } #endif diff --git a/media/libaom/config/linux/ia32/config/aom_config.asm b/media/libaom/config/linux/ia32/config/aom_config.asm index 128f85ec27..591f86e6c9 100644 --- a/media/libaom/config/linux/ia32/config/aom_config.asm +++ b/media/libaom/config/linux/ia32/config/aom_config.asm @@ -1,5 +1,5 @@ ; -; Copyright (c) 2018, Alliance for Open Media. All rights reserved +; Copyright (c) 2021, 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 @@ -14,49 +14,52 @@ ARCH_MIPS equ 0 ARCH_PPC equ 0 ARCH_X86 equ 1 ARCH_X86_64 equ 0 -CONFIG_2PASS_PARTITION_SEARCH_LVL equ 1 CONFIG_ACCOUNTING equ 0 CONFIG_ANALYZER equ 0 CONFIG_AV1_DECODER equ 1 CONFIG_AV1_ENCODER equ 0 +CONFIG_AV1_HIGHBITDEPTH equ 1 CONFIG_BIG_ENDIAN equ 0 CONFIG_BITSTREAM_DEBUG equ 0 CONFIG_COEFFICIENT_RANGE_CHECKING equ 0 -CONFIG_COLLECT_INTER_MODE_RD_STATS equ 0 +CONFIG_COLLECT_COMPONENT_TIMING equ 0 +CONFIG_COLLECT_PARTITION_STATS equ 0 CONFIG_COLLECT_RD_STATS equ 0 CONFIG_DEBUG equ 0 CONFIG_DENOISE equ 1 +CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 equ 1 CONFIG_DIST_8X8 equ 0 CONFIG_ENTROPY_STATS equ 0 -CONFIG_FILEOPTIONS equ 1 -CONFIG_FIX_GF_LENGTH equ 1 -CONFIG_FP_MB_STATS equ 0 CONFIG_GCC equ 1 CONFIG_GCOV equ 0 -CONFIG_GLOBAL_MOTION_SEARCH equ 1 CONFIG_GPROF equ 0 +CONFIG_HTB_TRELLIS equ 0 CONFIG_INSPECTION equ 0 CONFIG_INTERNAL_STATS equ 0 CONFIG_INTER_STATS_ONLY equ 0 CONFIG_LIBYUV equ 0 -CONFIG_LOWBITDEPTH equ 1 +CONFIG_LPF_MASK equ 0 CONFIG_MAX_DECODE_PROFILE equ 2 CONFIG_MISMATCH_DEBUG equ 0 CONFIG_MULTITHREAD equ 1 +CONFIG_NN_V2 equ 0 CONFIG_NORMAL_TILE_MODE equ 0 CONFIG_OS_SUPPORT equ 1 CONFIG_PIC equ 1 CONFIG_RD_DEBUG equ 0 -CONFIG_REDUCED_ENCODER_BORDER equ 0 +CONFIG_REALTIME_ONLY equ 0 CONFIG_RUNTIME_CPU_DETECT equ 1 CONFIG_SHARED equ 0 CONFIG_SHARP_SETTINGS equ 0 CONFIG_SIZE_LIMIT equ 0 CONFIG_SPATIAL_RESAMPLING equ 1 -CONFIG_STATIC equ 1 +CONFIG_SPEED_STATS equ 0 +CONFIG_SUPERRES_IN_RECODE equ 1 +CONFIG_TUNE_VMAF equ 0 CONFIG_WEBM_IO equ 0 DECODE_HEIGHT_LIMIT equ 0 DECODE_WIDTH_LIMIT equ 0 +FORCE_HIGHBITDEPTH_DECODING equ 0 HAVE_AVX equ 1 HAVE_AVX2 equ 1 HAVE_DSPR2 equ 0 diff --git a/media/libaom/config/linux/ia32/config/aom_config.h b/media/libaom/config/linux/ia32/config/aom_config.h index 02f6f72434..d2e651f19b 100644 --- a/media/libaom/config/linux/ia32/config/aom_config.h +++ b/media/libaom/config/linux/ia32/config/aom_config.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018, Alliance for Open Media. All rights reserved + * Copyright (c) 2021, 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 @@ -16,49 +16,52 @@ #define ARCH_PPC 0 #define ARCH_X86 1 #define ARCH_X86_64 0 -#define CONFIG_2PASS_PARTITION_SEARCH_LVL 1 #define CONFIG_ACCOUNTING 0 #define CONFIG_ANALYZER 0 #define CONFIG_AV1_DECODER 1 #define CONFIG_AV1_ENCODER 0 +#define CONFIG_AV1_HIGHBITDEPTH 1 #define CONFIG_BIG_ENDIAN 0 #define CONFIG_BITSTREAM_DEBUG 0 #define CONFIG_COEFFICIENT_RANGE_CHECKING 0 -#define CONFIG_COLLECT_INTER_MODE_RD_STATS 0 +#define CONFIG_COLLECT_COMPONENT_TIMING 0 +#define CONFIG_COLLECT_PARTITION_STATS 0 #define CONFIG_COLLECT_RD_STATS 0 #define CONFIG_DEBUG 0 #define CONFIG_DENOISE 1 +#define CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 1 #define CONFIG_DIST_8X8 0 #define CONFIG_ENTROPY_STATS 0 -#define CONFIG_FILEOPTIONS 1 -#define CONFIG_FIX_GF_LENGTH 1 -#define CONFIG_FP_MB_STATS 0 #define CONFIG_GCC 1 #define CONFIG_GCOV 0 -#define CONFIG_GLOBAL_MOTION_SEARCH 1 #define CONFIG_GPROF 0 +#define CONFIG_HTB_TRELLIS 0 #define CONFIG_INSPECTION 0 #define CONFIG_INTERNAL_STATS 0 #define CONFIG_INTER_STATS_ONLY 0 #define CONFIG_LIBYUV 0 -#define CONFIG_LOWBITDEPTH 1 +#define CONFIG_LPF_MASK 0 #define CONFIG_MAX_DECODE_PROFILE 2 #define CONFIG_MISMATCH_DEBUG 0 #define CONFIG_MULTITHREAD 1 +#define CONFIG_NN_V2 0 #define CONFIG_NORMAL_TILE_MODE 0 #define CONFIG_OS_SUPPORT 1 #define CONFIG_PIC 1 #define CONFIG_RD_DEBUG 0 -#define CONFIG_REDUCED_ENCODER_BORDER 0 +#define CONFIG_REALTIME_ONLY 0 #define CONFIG_RUNTIME_CPU_DETECT 1 #define CONFIG_SHARED 0 #define CONFIG_SHARP_SETTINGS 0 #define CONFIG_SIZE_LIMIT 0 #define CONFIG_SPATIAL_RESAMPLING 1 -#define CONFIG_STATIC 1 +#define CONFIG_SPEED_STATS 0 +#define CONFIG_SUPERRES_IN_RECODE 1 +#define CONFIG_TUNE_VMAF 0 #define CONFIG_WEBM_IO 0 #define DECODE_HEIGHT_LIMIT 0 #define DECODE_WIDTH_LIMIT 0 +#define FORCE_HIGHBITDEPTH_DECODING 0 #define HAVE_AVX 1 #define HAVE_AVX2 1 #define HAVE_DSPR2 0 @@ -76,7 +79,5 @@ #define HAVE_SSSE3 1 #define HAVE_VSX 0 #define HAVE_WXWIDGETS 0 -#define INCLUDE_INSTALL_DIR INSTALLDIR/include #define INLINE inline -#define LIB_INSTALL_DIR INSTALLDIR/lib #endif /* AOM_CONFIG_H_ */ diff --git a/media/libaom/config/linux/ia32/config/aom_dsp_rtcd.h b/media/libaom/config/linux/ia32/config/aom_dsp_rtcd.h index 5b3762f109..9018213bd6 100644 --- a/media/libaom/config/linux/ia32/config/aom_dsp_rtcd.h +++ b/media/libaom/config/linux/ia32/config/aom_dsp_rtcd.h @@ -26,10 +26,10 @@ void aom_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src void aom_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); RTCD_EXTERN void (*aom_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); -void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); +void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); void aom_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); void aom_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); @@ -47,9 +47,9 @@ void aom_convolve8_vert_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t void aom_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*aom_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -RTCD_EXTERN void (*aom_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); +void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +RTCD_EXTERN void (*aom_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); void aom_dc_128_predictor_16x16_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); void aom_dc_128_predictor_16x16_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); @@ -471,33 +471,37 @@ void aom_h_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_h_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); RTCD_EXTERN void (*aom_h_predictor_8x8)(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); -void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params, const int bd); -#define aom_highbd_blend_a64_d16_mask aom_highbd_blend_a64_d16_mask_c +void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); void aom_highbd_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); +void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); void aom_highbd_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_vmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); void aom_highbd_dc_128_predictor_16x16_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); void aom_highbd_dc_128_predictor_16x16_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); @@ -853,10 +857,10 @@ void aom_highbd_lpf_horizontal_14_c(uint16_t *s, int pitch, const uint8_t *blimi void aom_highbd_lpf_horizontal_14_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14)(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); -void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); void aom_highbd_lpf_horizontal_4_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); void aom_highbd_lpf_horizontal_4_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); @@ -1229,10 +1233,10 @@ void aom_highbd_v_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_v_predictor_8x8_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); RTCD_EXTERN void (*aom_highbd_v_predictor_8x8)(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); -void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); void aom_lpf_horizontal_14_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void aom_lpf_horizontal_14_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -1709,10 +1713,6 @@ void aom_v_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_v_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); RTCD_EXTERN void (*aom_v_predictor_8x8)(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); -void av1_round_shift_array_c(int32_t *arr, int size, int bit); -void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); -RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); - void aom_dsp_rtcd(void); #ifdef RTCD_C @@ -1955,6 +1955,9 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE2) aom_h_predictor_8x4 = aom_h_predictor_8x4_sse2; aom_h_predictor_8x8 = aom_h_predictor_8x8_c; if (flags & HAS_SSE2) aom_h_predictor_8x8 = aom_h_predictor_8x8_sse2; + aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_c; + if (flags & HAS_SSE4_1) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_sse4_1; + if (flags & HAS_AVX2) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_avx2; aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_sse4_1; aom_highbd_blend_a64_mask = aom_highbd_blend_a64_mask_c; @@ -1962,8 +1965,10 @@ static void setup_rtcd_internal(void) aom_highbd_blend_a64_vmask = aom_highbd_blend_a64_vmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_vmask = aom_highbd_blend_a64_vmask_sse4_1; aom_highbd_convolve8_horiz = aom_highbd_convolve8_horiz_c; + if (flags & HAS_SSE2) aom_highbd_convolve8_horiz = aom_highbd_convolve8_horiz_sse2; if (flags & HAS_AVX2) aom_highbd_convolve8_horiz = aom_highbd_convolve8_horiz_avx2; aom_highbd_convolve8_vert = aom_highbd_convolve8_vert_c; + if (flags & HAS_SSE2) aom_highbd_convolve8_vert = aom_highbd_convolve8_vert_sse2; if (flags & HAS_AVX2) aom_highbd_convolve8_vert = aom_highbd_convolve8_vert_avx2; aom_highbd_convolve_copy = aom_highbd_convolve_copy_c; if (flags & HAS_SSE2) aom_highbd_convolve_copy = aom_highbd_convolve_copy_sse2; @@ -2367,8 +2372,6 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE2) aom_v_predictor_8x4 = aom_v_predictor_8x4_sse2; aom_v_predictor_8x8 = aom_v_predictor_8x8_c; if (flags & HAS_SSE2) aom_v_predictor_8x8 = aom_v_predictor_8x8_sse2; - av1_round_shift_array = av1_round_shift_array_c; - if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; } #endif diff --git a/media/libaom/config/linux/ia32/config/aom_scale_rtcd.h b/media/libaom/config/linux/ia32/config/aom_scale_rtcd.h index 6b56795cd4..65c184b83d 100644 --- a/media/libaom/config/linux/ia32/config/aom_scale_rtcd.h +++ b/media/libaom/config/linux/ia32/config/aom_scale_rtcd.h @@ -59,15 +59,27 @@ void aom_yv12_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buf void aom_yv12_extend_frame_borders_c(struct yv12_buffer_config *ybf, const int num_planes); #define aom_yv12_extend_frame_borders aom_yv12_extend_frame_borders_c -void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_coloc_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_u aom_yv12_partial_coloc_copy_u_c + +void aom_yv12_partial_coloc_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_v aom_yv12_partial_coloc_copy_v_c + +void aom_yv12_partial_coloc_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_y aom_yv12_partial_coloc_copy_y_c + +void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_u aom_yv12_partial_copy_u_c -void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_v aom_yv12_partial_copy_v_c -void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_ybc, int hstart2, int vstart2); #define aom_yv12_partial_copy_y aom_yv12_partial_copy_y_c +int aom_yv12_realloc_with_new_border_c(struct yv12_buffer_config *ybf, int new_border, int byte_alignment, int num_planes); +#define aom_yv12_realloc_with_new_border aom_yv12_realloc_with_new_border_c + void aom_scale_rtcd(void); #ifdef RTCD_C diff --git a/media/libaom/config/linux/ia32/config/av1_rtcd.h b/media/libaom/config/linux/ia32/config/av1_rtcd.h index a2a6f14ff5..7ecd05f380 100644 --- a/media/libaom/config/linux/ia32/config/av1_rtcd.h +++ b/media/libaom/config/linux/ia32/config/av1_rtcd.h @@ -31,30 +31,55 @@ struct txfm_param; struct aom_variance_vtable; struct search_site_config; struct yv12_buffer_config; +struct NN_CONFIG; +typedef struct NN_CONFIG NN_CONFIG; + +enum { NONE, RELU, SOFTSIGN, SIGMOID } UENUM1BYTE(ACTIVATION); +#if CONFIG_NN_V2 +enum { SOFTMAX_CROSS_ENTROPY } UENUM1BYTE(LOSS); +struct NN_CONFIG_V2; +typedef struct NN_CONFIG_V2 NN_CONFIG_V2; +struct FC_LAYER; +typedef struct FC_LAYER FC_LAYER; +#endif // CONFIG_NN_V2 + +struct CNN_CONFIG; +typedef struct CNN_CONFIG CNN_CONFIG; +struct CNN_LAYER_CONFIG; +typedef struct CNN_LAYER_CONFIG CNN_LAYER_CONFIG; +struct CNN_THREAD_DATA; +typedef struct CNN_THREAD_DATA CNN_THREAD_DATA; +struct CNN_BRANCH_CONFIG; +typedef struct CNN_BRANCH_CONFIG CNN_BRANCH_CONFIG; +struct CNN_MULTI_OUT; +typedef struct CNN_MULTI_OUT CNN_MULTI_OUT; /* Function pointers return by CfL functions */ typedef void (*cfl_subsample_lbd_fn)(const uint8_t *input, int input_stride, uint16_t *output_q3); +#if CONFIG_AV1_HIGHBITDEPTH typedef void (*cfl_subsample_hbd_fn)(const uint16_t *input, int input_stride, uint16_t *output_q3); +typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, + int dst_stride, int alpha_q3, int bd); +#endif + typedef void (*cfl_subtract_average_fn)(const uint16_t *src, int16_t *dst); typedef void (*cfl_predict_lbd_fn)(const int16_t *src, uint8_t *dst, int dst_stride, int alpha_q3); -typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, - int dst_stride, int alpha_q3, int bd); #ifdef __cplusplus extern "C" { #endif -void apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -RTCD_EXTERN void (*apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +RTCD_EXTERN void (*av1_apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); void av1_build_compound_diffwtd_mask_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); void av1_build_compound_diffwtd_mask_sse4_1(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); @@ -71,42 +96,89 @@ void av1_build_compound_diffwtd_mask_highbd_ssse3(uint8_t *mask, DIFFWTD_MASK_TY void av1_build_compound_diffwtd_mask_highbd_avx2(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); RTCD_EXTERN void (*av1_build_compound_diffwtd_mask_highbd)(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); -void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +int64_t av1_calc_frame_error_c(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_sse2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_avx2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +RTCD_EXTERN int64_t (*av1_calc_frame_error)(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); + +void av1_cnn_activate_c( float **input, int channels, int width, int height, int stride, ACTIVATION layer_activation); +#define av1_cnn_activate av1_cnn_activate_c -void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); +void av1_cnn_add_c( float **input, int channels, int width, int height, int stride, const float **add); +#define av1_cnn_add av1_cnn_add_c -void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_cnn_batchnorm_c(float **image, int channels, int width, int height, int stride, const float *gamma, const float *beta, const float *mean, const float *std); +#define av1_cnn_batchnorm av1_cnn_batchnorm_c + +void av1_cnn_convolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride, int start_idx, int step); +#define av1_cnn_convolve av1_cnn_convolve_c + +void av1_cnn_deconvolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride); +#define av1_cnn_deconvolve av1_cnn_deconvolve_c + +void av1_cnn_predict_c( const float **input, int in_width, int in_height, int in_stride, const CNN_CONFIG *cnn_config, const CNN_THREAD_DATA *thread_data, CNN_MULTI_OUT *output_struct); +#define av1_cnn_predict av1_cnn_predict_c + +void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_convolve_horiz_rs_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); void av1_convolve_horiz_rs_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); RTCD_EXTERN void (*av1_convolve_horiz_rs)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); -void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_dr_prediction_z1_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); -#define av1_dr_prediction_z1 av1_dr_prediction_z1_c +void av1_dr_prediction_z1_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z1)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); void av1_dr_prediction_z2_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z2 av1_dr_prediction_z2_c +void av1_dr_prediction_z2_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z2)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); void av1_dr_prediction_z3_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z3 av1_dr_prediction_z3_c +void av1_dr_prediction_z3_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z3)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); void av1_filter_intra_edge_c(uint8_t *p, int sz, int strength); void av1_filter_intra_edge_sse4_1(uint8_t *p, int sz, int strength); @@ -129,19 +201,19 @@ void av1_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint void av1_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve8_vert av1_highbd_convolve8_vert_c -void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve_avg av1_highbd_convolve_avg_c @@ -153,54 +225,76 @@ void av1_highbd_convolve_horiz_rs_c(const uint16_t *src, int src_stride, uint16_ void av1_highbd_convolve_horiz_rs_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); RTCD_EXTERN void (*av1_highbd_convolve_horiz_rs)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); -void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_dr_prediction_z1_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z1 av1_highbd_dr_prediction_z1_c +void av1_highbd_dr_prediction_z1_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z1)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); void av1_highbd_dr_prediction_z2_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z2 av1_highbd_dr_prediction_z2_c +void av1_highbd_dr_prediction_z2_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z2)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); void av1_highbd_dr_prediction_z3_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z3 av1_highbd_dr_prediction_z3_c +void av1_highbd_dr_prediction_z3_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z3)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -void av1_highbd_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_avx2(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x16)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_32x32)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_16_add av1_highbd_iwht4x4_16_add_c @@ -208,34 +302,14 @@ void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int des void av1_highbd_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_1_add av1_highbd_iwht4x4_1_add_c -void av1_highbd_jnt_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - void av1_highbd_warp_affine_c(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_highbd_warp_affine_sse4_1(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_highbd_warp_affine)(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); -void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); +void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); void av1_inv_txfm2d_add_16x16_c(const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type, int bd); #define av1_inv_txfm2d_add_16x16 av1_inv_txfm2d_add_16x16_c @@ -301,25 +375,9 @@ void av1_inv_txfm_add_ssse3(const tran_low_t *dqcoeff, uint8_t *dst, int stride, void av1_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); RTCD_EXTERN void (*av1_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_jnt_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_round_shift_array_c(int32_t *arr, int size, int bit); +void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); +RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); int av1_selfguided_restoration_c(const uint8_t *dgd8, int width, int height, int dgd_stride, int32_t *flt0, int32_t *flt1, int flt_stride, @@ -344,6 +402,7 @@ RTCD_EXTERN void (*av1_upsample_intra_edge_high)(uint16_t *p, int sz, int bd); void av1_warp_affine_c(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_warp_affine_sse4_1(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); +void av1_warp_affine_avx2(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_warp_affine)(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); @@ -351,12 +410,26 @@ void av1_wiener_convolve_add_src_sse2(const uint8_t *src, ptrdiff_t src_stride, void av1_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); RTCD_EXTERN void (*av1_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); -void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); +void cdef_copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); + +void cdef_copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); + +void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); int cdef_find_dir_c(const uint16_t *img, int stride, int32_t *var, int coeff_shift); int cdef_find_dir_sse2(const uint16_t *img, int stride, int32_t *var, int coeff_shift); @@ -395,34 +468,20 @@ cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_ssse3(TX_SIZE tx_size); cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_avx2(TX_SIZE tx_size); RTCD_EXTERN cfl_subsample_lbd_fn (*cfl_get_luma_subsampling_444_lbd)(TX_SIZE tx_size); -void copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); - -void copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); - -cfl_predict_hbd_fn get_predict_hbd_fn_c(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_hbd_fn (*get_predict_hbd_fn)(TX_SIZE tx_size); - -cfl_predict_lbd_fn get_predict_lbd_fn_c(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_lbd_fn (*get_predict_lbd_fn)(TX_SIZE tx_size); - -cfl_subtract_average_fn get_subtract_average_fn_c(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_sse2(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_subtract_average_fn (*get_subtract_average_fn)(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_c(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_hbd_fn (*cfl_get_predict_hbd_fn)(TX_SIZE tx_size); + +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_c(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_lbd_fn (*cfl_get_predict_lbd_fn)(TX_SIZE tx_size); + +cfl_subtract_average_fn cfl_get_subtract_average_fn_c(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_sse2(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_subtract_average_fn (*cfl_get_subtract_average_fn)(TX_SIZE tx_size); void av1_rtcd(void); @@ -434,9 +493,9 @@ static void setup_rtcd_internal(void) (void)flags; - apply_selfguided_restoration = apply_selfguided_restoration_c; - if (flags & HAS_SSE4_1) apply_selfguided_restoration = apply_selfguided_restoration_sse4_1; - if (flags & HAS_AVX2) apply_selfguided_restoration = apply_selfguided_restoration_avx2; + av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_c; + if (flags & HAS_SSE4_1) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_sse4_1; + if (flags & HAS_AVX2) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_avx2; av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_c; if (flags & HAS_SSE4_1) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_sse4_1; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_avx2; @@ -446,6 +505,9 @@ static void setup_rtcd_internal(void) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_c; if (flags & HAS_SSSE3) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_ssse3; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_avx2; + av1_calc_frame_error = av1_calc_frame_error_c; + if (flags & HAS_SSE2) av1_calc_frame_error = av1_calc_frame_error_sse2; + if (flags & HAS_AVX2) av1_calc_frame_error = av1_calc_frame_error_avx2; av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_c; if (flags & HAS_SSE2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_sse2; if (flags & HAS_AVX2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_avx2; @@ -462,6 +524,25 @@ static void setup_rtcd_internal(void) av1_convolve_y_sr = av1_convolve_y_sr_c; if (flags & HAS_SSE2) av1_convolve_y_sr = av1_convolve_y_sr_sse2; if (flags & HAS_AVX2) av1_convolve_y_sr = av1_convolve_y_sr_avx2; + av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_sse2; + if (flags & HAS_SSSE3) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_ssse3; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_avx2; + av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_avx2; + av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_avx2; + av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_avx2; + av1_dr_prediction_z1 = av1_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_dr_prediction_z1 = av1_dr_prediction_z1_avx2; + av1_dr_prediction_z2 = av1_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_dr_prediction_z2 = av1_dr_prediction_z2_avx2; + av1_dr_prediction_z3 = av1_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_dr_prediction_z3 = av1_dr_prediction_z3_avx2; av1_filter_intra_edge = av1_filter_intra_edge_c; if (flags & HAS_SSE4_1) av1_filter_intra_edge = av1_filter_intra_edge_sse4_1; av1_filter_intra_edge_high = av1_filter_intra_edge_high_c; @@ -484,34 +565,39 @@ static void setup_rtcd_internal(void) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_c; if (flags & HAS_SSSE3) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_ssse3; if (flags & HAS_AVX2) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_avx2; + av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_avx2; + av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_avx2; + av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_avx2; + av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_avx2; + av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_avx2; + av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_avx2; + av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_avx2; av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_sse4_1; if (flags & HAS_AVX2) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_avx2; - av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_sse4_1; - av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_sse4_1; - av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_sse4_1; - if (flags & HAS_AVX2) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_avx2; + av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_sse4_1; + av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_sse4_1; av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_sse4_1; - av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_sse4_1; + av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_sse4_1; + av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_sse4_1; av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_sse4_1; - av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_avx2; - av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_avx2; - av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_avx2; - av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_avx2; av1_highbd_warp_affine = av1_highbd_warp_affine_c; if (flags & HAS_SSE4_1) av1_highbd_warp_affine = av1_highbd_warp_affine_sse4_1; av1_highbd_wiener_convolve_add_src = av1_highbd_wiener_convolve_add_src_c; @@ -524,18 +610,8 @@ static void setup_rtcd_internal(void) av1_inv_txfm_add = av1_inv_txfm_add_c; if (flags & HAS_SSSE3) av1_inv_txfm_add = av1_inv_txfm_add_ssse3; if (flags & HAS_AVX2) av1_inv_txfm_add = av1_inv_txfm_add_avx2; - av1_jnt_convolve_2d = av1_jnt_convolve_2d_c; - if (flags & HAS_SSSE3) av1_jnt_convolve_2d = av1_jnt_convolve_2d_ssse3; - if (flags & HAS_AVX2) av1_jnt_convolve_2d = av1_jnt_convolve_2d_avx2; - av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_avx2; - av1_jnt_convolve_x = av1_jnt_convolve_x_c; - if (flags & HAS_SSE2) av1_jnt_convolve_x = av1_jnt_convolve_x_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_x = av1_jnt_convolve_x_avx2; - av1_jnt_convolve_y = av1_jnt_convolve_y_c; - if (flags & HAS_SSE2) av1_jnt_convolve_y = av1_jnt_convolve_y_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_y = av1_jnt_convolve_y_avx2; + av1_round_shift_array = av1_round_shift_array_c; + if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; av1_selfguided_restoration = av1_selfguided_restoration_c; if (flags & HAS_SSE4_1) av1_selfguided_restoration = av1_selfguided_restoration_sse4_1; if (flags & HAS_AVX2) av1_selfguided_restoration = av1_selfguided_restoration_avx2; @@ -545,9 +621,20 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE4_1) av1_upsample_intra_edge_high = av1_upsample_intra_edge_high_sse4_1; av1_warp_affine = av1_warp_affine_c; if (flags & HAS_SSE4_1) av1_warp_affine = av1_warp_affine_sse4_1; + if (flags & HAS_AVX2) av1_warp_affine = av1_warp_affine_avx2; av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_c; if (flags & HAS_SSE2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_sse2; if (flags & HAS_AVX2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_avx2; + cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_c; + if (flags & HAS_SSE2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_avx2; + cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_c; + if (flags & HAS_SSE2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_avx2; cdef_filter_block = cdef_filter_block_c; if (flags & HAS_SSE2) cdef_filter_block = cdef_filter_block_sse2; if (flags & HAS_SSSE3) cdef_filter_block = cdef_filter_block_ssse3; @@ -576,25 +663,15 @@ static void setup_rtcd_internal(void) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_c; if (flags & HAS_SSSE3) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_ssse3; if (flags & HAS_AVX2) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_avx2; - copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_c; - if (flags & HAS_SSE2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_avx2; - copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_c; - if (flags & HAS_SSE2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_avx2; - get_predict_hbd_fn = get_predict_hbd_fn_c; - if (flags & HAS_SSSE3) get_predict_hbd_fn = get_predict_hbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_hbd_fn = get_predict_hbd_fn_avx2; - get_predict_lbd_fn = get_predict_lbd_fn_c; - if (flags & HAS_SSSE3) get_predict_lbd_fn = get_predict_lbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_lbd_fn = get_predict_lbd_fn_avx2; - get_subtract_average_fn = get_subtract_average_fn_c; - if (flags & HAS_SSE2) get_subtract_average_fn = get_subtract_average_fn_sse2; - if (flags & HAS_AVX2) get_subtract_average_fn = get_subtract_average_fn_avx2; + cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_avx2; + cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_avx2; + cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_c; + if (flags & HAS_SSE2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_sse2; + if (flags & HAS_AVX2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_avx2; } #endif diff --git a/media/libaom/config/linux/x64/config/aom_config.asm b/media/libaom/config/linux/x64/config/aom_config.asm index 4ab2dedb42..9ce9b2090e 100644 --- a/media/libaom/config/linux/x64/config/aom_config.asm +++ b/media/libaom/config/linux/x64/config/aom_config.asm @@ -1,5 +1,5 @@ ; -; Copyright (c) 2018, Alliance for Open Media. All rights reserved +; Copyright (c) 2021, 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 @@ -14,49 +14,52 @@ ARCH_MIPS equ 0 ARCH_PPC equ 0 ARCH_X86 equ 0 ARCH_X86_64 equ 1 -CONFIG_2PASS_PARTITION_SEARCH_LVL equ 1 CONFIG_ACCOUNTING equ 0 CONFIG_ANALYZER equ 0 CONFIG_AV1_DECODER equ 1 CONFIG_AV1_ENCODER equ 0 +CONFIG_AV1_HIGHBITDEPTH equ 1 CONFIG_BIG_ENDIAN equ 0 CONFIG_BITSTREAM_DEBUG equ 0 CONFIG_COEFFICIENT_RANGE_CHECKING equ 0 -CONFIG_COLLECT_INTER_MODE_RD_STATS equ 0 +CONFIG_COLLECT_COMPONENT_TIMING equ 0 +CONFIG_COLLECT_PARTITION_STATS equ 0 CONFIG_COLLECT_RD_STATS equ 0 CONFIG_DEBUG equ 0 CONFIG_DENOISE equ 1 +CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 equ 1 CONFIG_DIST_8X8 equ 0 CONFIG_ENTROPY_STATS equ 0 -CONFIG_FILEOPTIONS equ 1 -CONFIG_FIX_GF_LENGTH equ 1 -CONFIG_FP_MB_STATS equ 0 CONFIG_GCC equ 1 CONFIG_GCOV equ 0 -CONFIG_GLOBAL_MOTION_SEARCH equ 1 CONFIG_GPROF equ 0 +CONFIG_HTB_TRELLIS equ 0 CONFIG_INSPECTION equ 0 CONFIG_INTERNAL_STATS equ 0 CONFIG_INTER_STATS_ONLY equ 0 CONFIG_LIBYUV equ 0 -CONFIG_LOWBITDEPTH equ 1 +CONFIG_LPF_MASK equ 0 CONFIG_MAX_DECODE_PROFILE equ 2 CONFIG_MISMATCH_DEBUG equ 0 CONFIG_MULTITHREAD equ 1 +CONFIG_NN_V2 equ 0 CONFIG_NORMAL_TILE_MODE equ 0 CONFIG_OS_SUPPORT equ 1 CONFIG_PIC equ 0 CONFIG_RD_DEBUG equ 0 -CONFIG_REDUCED_ENCODER_BORDER equ 0 +CONFIG_REALTIME_ONLY equ 0 CONFIG_RUNTIME_CPU_DETECT equ 1 CONFIG_SHARED equ 0 CONFIG_SHARP_SETTINGS equ 0 CONFIG_SIZE_LIMIT equ 0 CONFIG_SPATIAL_RESAMPLING equ 1 -CONFIG_STATIC equ 1 +CONFIG_SPEED_STATS equ 0 +CONFIG_SUPERRES_IN_RECODE equ 1 +CONFIG_TUNE_VMAF equ 0 CONFIG_WEBM_IO equ 0 DECODE_HEIGHT_LIMIT equ 0 DECODE_WIDTH_LIMIT equ 0 +FORCE_HIGHBITDEPTH_DECODING equ 0 HAVE_AVX equ 1 HAVE_AVX2 equ 1 HAVE_DSPR2 equ 0 diff --git a/media/libaom/config/linux/x64/config/aom_config.h b/media/libaom/config/linux/x64/config/aom_config.h index 9a0be7cd11..12997ea4dd 100644 --- a/media/libaom/config/linux/x64/config/aom_config.h +++ b/media/libaom/config/linux/x64/config/aom_config.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018, Alliance for Open Media. All rights reserved + * Copyright (c) 2021, 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 @@ -16,49 +16,52 @@ #define ARCH_PPC 0 #define ARCH_X86 0 #define ARCH_X86_64 1 -#define CONFIG_2PASS_PARTITION_SEARCH_LVL 1 #define CONFIG_ACCOUNTING 0 #define CONFIG_ANALYZER 0 #define CONFIG_AV1_DECODER 1 #define CONFIG_AV1_ENCODER 0 +#define CONFIG_AV1_HIGHBITDEPTH 1 #define CONFIG_BIG_ENDIAN 0 #define CONFIG_BITSTREAM_DEBUG 0 #define CONFIG_COEFFICIENT_RANGE_CHECKING 0 -#define CONFIG_COLLECT_INTER_MODE_RD_STATS 0 +#define CONFIG_COLLECT_COMPONENT_TIMING 0 +#define CONFIG_COLLECT_PARTITION_STATS 0 #define CONFIG_COLLECT_RD_STATS 0 #define CONFIG_DEBUG 0 #define CONFIG_DENOISE 1 +#define CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 1 #define CONFIG_DIST_8X8 0 #define CONFIG_ENTROPY_STATS 0 -#define CONFIG_FILEOPTIONS 1 -#define CONFIG_FIX_GF_LENGTH 1 -#define CONFIG_FP_MB_STATS 0 #define CONFIG_GCC 1 #define CONFIG_GCOV 0 -#define CONFIG_GLOBAL_MOTION_SEARCH 1 #define CONFIG_GPROF 0 +#define CONFIG_HTB_TRELLIS 0 #define CONFIG_INSPECTION 0 #define CONFIG_INTERNAL_STATS 0 #define CONFIG_INTER_STATS_ONLY 0 #define CONFIG_LIBYUV 0 -#define CONFIG_LOWBITDEPTH 1 +#define CONFIG_LPF_MASK 0 #define CONFIG_MAX_DECODE_PROFILE 2 #define CONFIG_MISMATCH_DEBUG 0 #define CONFIG_MULTITHREAD 1 +#define CONFIG_NN_V2 0 #define CONFIG_NORMAL_TILE_MODE 0 #define CONFIG_OS_SUPPORT 1 #define CONFIG_PIC 0 #define CONFIG_RD_DEBUG 0 -#define CONFIG_REDUCED_ENCODER_BORDER 0 +#define CONFIG_REALTIME_ONLY 0 #define CONFIG_RUNTIME_CPU_DETECT 1 #define CONFIG_SHARED 0 #define CONFIG_SHARP_SETTINGS 0 #define CONFIG_SIZE_LIMIT 0 #define CONFIG_SPATIAL_RESAMPLING 1 -#define CONFIG_STATIC 1 +#define CONFIG_SPEED_STATS 0 +#define CONFIG_SUPERRES_IN_RECODE 1 +#define CONFIG_TUNE_VMAF 0 #define CONFIG_WEBM_IO 0 #define DECODE_HEIGHT_LIMIT 0 #define DECODE_WIDTH_LIMIT 0 +#define FORCE_HIGHBITDEPTH_DECODING 0 #define HAVE_AVX 1 #define HAVE_AVX2 1 #define HAVE_DSPR2 0 @@ -76,7 +79,5 @@ #define HAVE_SSSE3 1 #define HAVE_VSX 0 #define HAVE_WXWIDGETS 0 -#define INCLUDE_INSTALL_DIR INSTALLDIR/include #define INLINE inline -#define LIB_INSTALL_DIR INSTALLDIR/lib #endif /* AOM_CONFIG_H_ */ diff --git a/media/libaom/config/linux/x64/config/aom_dsp_rtcd.h b/media/libaom/config/linux/x64/config/aom_dsp_rtcd.h index 2856d4ede7..36c170a05d 100644 --- a/media/libaom/config/linux/x64/config/aom_dsp_rtcd.h +++ b/media/libaom/config/linux/x64/config/aom_dsp_rtcd.h @@ -26,10 +26,10 @@ void aom_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src void aom_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); RTCD_EXTERN void (*aom_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); -void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); +void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); void aom_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); void aom_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); @@ -47,8 +47,8 @@ void aom_convolve8_vert_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t void aom_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*aom_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); +void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); #define aom_convolve_copy aom_convolve_copy_sse2 void aom_dc_128_predictor_16x16_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); @@ -471,35 +471,37 @@ void aom_h_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_h_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_h_predictor_8x8 aom_h_predictor_8x8_sse2 -void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params, const int bd); -#define aom_highbd_blend_a64_d16_mask aom_highbd_blend_a64_d16_mask_c +void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); void aom_highbd_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); +void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); void aom_highbd_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_vmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); void aom_highbd_dc_128_predictor_16x16_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); void aom_highbd_dc_128_predictor_16x16_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); @@ -855,10 +857,10 @@ void aom_highbd_lpf_horizontal_14_c(uint16_t *s, int pitch, const uint8_t *blimi void aom_highbd_lpf_horizontal_14_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); #define aom_highbd_lpf_horizontal_14 aom_highbd_lpf_horizontal_14_sse2 -void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); void aom_highbd_lpf_horizontal_4_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); void aom_highbd_lpf_horizontal_4_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); @@ -1231,10 +1233,10 @@ void aom_highbd_v_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_v_predictor_8x8_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); #define aom_highbd_v_predictor_8x8 aom_highbd_v_predictor_8x8_sse2 -void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); void aom_lpf_horizontal_14_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void aom_lpf_horizontal_14_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -1711,10 +1713,6 @@ void aom_v_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_v_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_v_predictor_8x8 aom_v_predictor_8x8_sse2 -void av1_round_shift_array_c(int32_t *arr, int size, int bit); -void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); -RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); - void aom_dsp_rtcd(void); #ifdef RTCD_C @@ -1788,6 +1786,9 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) aom_dc_top_predictor_64x64 = aom_dc_top_predictor_64x64_avx2; aom_h_predictor_32x32 = aom_h_predictor_32x32_sse2; if (flags & HAS_AVX2) aom_h_predictor_32x32 = aom_h_predictor_32x32_avx2; + aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_c; + if (flags & HAS_SSE4_1) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_sse4_1; + if (flags & HAS_AVX2) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_avx2; aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_sse4_1; aom_highbd_blend_a64_mask = aom_highbd_blend_a64_mask_c; @@ -1989,8 +1990,6 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) aom_v_predictor_64x32 = aom_v_predictor_64x32_avx2; aom_v_predictor_64x64 = aom_v_predictor_64x64_sse2; if (flags & HAS_AVX2) aom_v_predictor_64x64 = aom_v_predictor_64x64_avx2; - av1_round_shift_array = av1_round_shift_array_c; - if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; } #endif diff --git a/media/libaom/config/linux/x64/config/aom_scale_rtcd.h b/media/libaom/config/linux/x64/config/aom_scale_rtcd.h index 6b56795cd4..65c184b83d 100644 --- a/media/libaom/config/linux/x64/config/aom_scale_rtcd.h +++ b/media/libaom/config/linux/x64/config/aom_scale_rtcd.h @@ -59,15 +59,27 @@ void aom_yv12_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buf void aom_yv12_extend_frame_borders_c(struct yv12_buffer_config *ybf, const int num_planes); #define aom_yv12_extend_frame_borders aom_yv12_extend_frame_borders_c -void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_coloc_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_u aom_yv12_partial_coloc_copy_u_c + +void aom_yv12_partial_coloc_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_v aom_yv12_partial_coloc_copy_v_c + +void aom_yv12_partial_coloc_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_y aom_yv12_partial_coloc_copy_y_c + +void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_u aom_yv12_partial_copy_u_c -void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_v aom_yv12_partial_copy_v_c -void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_ybc, int hstart2, int vstart2); #define aom_yv12_partial_copy_y aom_yv12_partial_copy_y_c +int aom_yv12_realloc_with_new_border_c(struct yv12_buffer_config *ybf, int new_border, int byte_alignment, int num_planes); +#define aom_yv12_realloc_with_new_border aom_yv12_realloc_with_new_border_c + void aom_scale_rtcd(void); #ifdef RTCD_C diff --git a/media/libaom/config/linux/x64/config/av1_rtcd.h b/media/libaom/config/linux/x64/config/av1_rtcd.h index d273182086..8f20181b42 100644 --- a/media/libaom/config/linux/x64/config/av1_rtcd.h +++ b/media/libaom/config/linux/x64/config/av1_rtcd.h @@ -31,30 +31,55 @@ struct txfm_param; struct aom_variance_vtable; struct search_site_config; struct yv12_buffer_config; +struct NN_CONFIG; +typedef struct NN_CONFIG NN_CONFIG; + +enum { NONE, RELU, SOFTSIGN, SIGMOID } UENUM1BYTE(ACTIVATION); +#if CONFIG_NN_V2 +enum { SOFTMAX_CROSS_ENTROPY } UENUM1BYTE(LOSS); +struct NN_CONFIG_V2; +typedef struct NN_CONFIG_V2 NN_CONFIG_V2; +struct FC_LAYER; +typedef struct FC_LAYER FC_LAYER; +#endif // CONFIG_NN_V2 + +struct CNN_CONFIG; +typedef struct CNN_CONFIG CNN_CONFIG; +struct CNN_LAYER_CONFIG; +typedef struct CNN_LAYER_CONFIG CNN_LAYER_CONFIG; +struct CNN_THREAD_DATA; +typedef struct CNN_THREAD_DATA CNN_THREAD_DATA; +struct CNN_BRANCH_CONFIG; +typedef struct CNN_BRANCH_CONFIG CNN_BRANCH_CONFIG; +struct CNN_MULTI_OUT; +typedef struct CNN_MULTI_OUT CNN_MULTI_OUT; /* Function pointers return by CfL functions */ typedef void (*cfl_subsample_lbd_fn)(const uint8_t *input, int input_stride, uint16_t *output_q3); +#if CONFIG_AV1_HIGHBITDEPTH typedef void (*cfl_subsample_hbd_fn)(const uint16_t *input, int input_stride, uint16_t *output_q3); +typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, + int dst_stride, int alpha_q3, int bd); +#endif + typedef void (*cfl_subtract_average_fn)(const uint16_t *src, int16_t *dst); typedef void (*cfl_predict_lbd_fn)(const int16_t *src, uint8_t *dst, int dst_stride, int alpha_q3); -typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, - int dst_stride, int alpha_q3, int bd); #ifdef __cplusplus extern "C" { #endif -void apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -RTCD_EXTERN void (*apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +RTCD_EXTERN void (*av1_apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); void av1_build_compound_diffwtd_mask_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); void av1_build_compound_diffwtd_mask_sse4_1(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); @@ -71,42 +96,89 @@ void av1_build_compound_diffwtd_mask_highbd_ssse3(uint8_t *mask, DIFFWTD_MASK_TY void av1_build_compound_diffwtd_mask_highbd_avx2(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); RTCD_EXTERN void (*av1_build_compound_diffwtd_mask_highbd)(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); -void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +int64_t av1_calc_frame_error_c(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_sse2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_avx2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +RTCD_EXTERN int64_t (*av1_calc_frame_error)(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); + +void av1_cnn_activate_c( float **input, int channels, int width, int height, int stride, ACTIVATION layer_activation); +#define av1_cnn_activate av1_cnn_activate_c -void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); +void av1_cnn_add_c( float **input, int channels, int width, int height, int stride, const float **add); +#define av1_cnn_add av1_cnn_add_c -void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_cnn_batchnorm_c(float **image, int channels, int width, int height, int stride, const float *gamma, const float *beta, const float *mean, const float *std); +#define av1_cnn_batchnorm av1_cnn_batchnorm_c + +void av1_cnn_convolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride, int start_idx, int step); +#define av1_cnn_convolve av1_cnn_convolve_c + +void av1_cnn_deconvolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride); +#define av1_cnn_deconvolve av1_cnn_deconvolve_c + +void av1_cnn_predict_c( const float **input, int in_width, int in_height, int in_stride, const CNN_CONFIG *cnn_config, const CNN_THREAD_DATA *thread_data, CNN_MULTI_OUT *output_struct); +#define av1_cnn_predict av1_cnn_predict_c + +void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_convolve_horiz_rs_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); void av1_convolve_horiz_rs_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); RTCD_EXTERN void (*av1_convolve_horiz_rs)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); -void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_dr_prediction_z1_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); -#define av1_dr_prediction_z1 av1_dr_prediction_z1_c +void av1_dr_prediction_z1_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z1)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); void av1_dr_prediction_z2_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z2 av1_dr_prediction_z2_c +void av1_dr_prediction_z2_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z2)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); void av1_dr_prediction_z3_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z3 av1_dr_prediction_z3_c +void av1_dr_prediction_z3_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z3)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); void av1_filter_intra_edge_c(uint8_t *p, int sz, int strength); void av1_filter_intra_edge_sse4_1(uint8_t *p, int sz, int strength); @@ -132,19 +204,19 @@ void av1_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8 void av1_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve8_vert av1_highbd_convolve8_vert_sse2 -void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve_avg av1_highbd_convolve_avg_c @@ -156,54 +228,76 @@ void av1_highbd_convolve_horiz_rs_c(const uint16_t *src, int src_stride, uint16_ void av1_highbd_convolve_horiz_rs_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); RTCD_EXTERN void (*av1_highbd_convolve_horiz_rs)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); -void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_dr_prediction_z1_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z1 av1_highbd_dr_prediction_z1_c +void av1_highbd_dr_prediction_z1_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z1)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); void av1_highbd_dr_prediction_z2_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z2 av1_highbd_dr_prediction_z2_c +void av1_highbd_dr_prediction_z2_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z2)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); void av1_highbd_dr_prediction_z3_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z3 av1_highbd_dr_prediction_z3_c +void av1_highbd_dr_prediction_z3_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z3)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -void av1_highbd_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_avx2(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x16)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_32x32)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_16_add av1_highbd_iwht4x4_16_add_c @@ -211,34 +305,14 @@ void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int des void av1_highbd_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_1_add av1_highbd_iwht4x4_1_add_c -void av1_highbd_jnt_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - void av1_highbd_warp_affine_c(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_highbd_warp_affine_sse4_1(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_highbd_warp_affine)(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); -void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); +void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); void av1_inv_txfm2d_add_16x16_c(const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type, int bd); #define av1_inv_txfm2d_add_16x16 av1_inv_txfm2d_add_16x16_c @@ -304,25 +378,9 @@ void av1_inv_txfm_add_ssse3(const tran_low_t *dqcoeff, uint8_t *dst, int stride, void av1_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); RTCD_EXTERN void (*av1_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_jnt_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_round_shift_array_c(int32_t *arr, int size, int bit); +void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); +RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); int av1_selfguided_restoration_c(const uint8_t *dgd8, int width, int height, int dgd_stride, int32_t *flt0, int32_t *flt1, int flt_stride, @@ -347,6 +405,7 @@ RTCD_EXTERN void (*av1_upsample_intra_edge_high)(uint16_t *p, int sz, int bd); void av1_warp_affine_c(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_warp_affine_sse4_1(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); +void av1_warp_affine_avx2(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_warp_affine)(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); @@ -354,12 +413,26 @@ void av1_wiener_convolve_add_src_sse2(const uint8_t *src, ptrdiff_t src_stride, void av1_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); RTCD_EXTERN void (*av1_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); -void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); +void cdef_copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); + +void cdef_copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); + +void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); int cdef_find_dir_c(const uint16_t *img, int stride, int32_t *var, int coeff_shift); int cdef_find_dir_sse2(const uint16_t *img, int stride, int32_t *var, int coeff_shift); @@ -398,34 +471,20 @@ cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_ssse3(TX_SIZE tx_size); cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_avx2(TX_SIZE tx_size); RTCD_EXTERN cfl_subsample_lbd_fn (*cfl_get_luma_subsampling_444_lbd)(TX_SIZE tx_size); -void copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); - -void copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); - -cfl_predict_hbd_fn get_predict_hbd_fn_c(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_hbd_fn (*get_predict_hbd_fn)(TX_SIZE tx_size); - -cfl_predict_lbd_fn get_predict_lbd_fn_c(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_lbd_fn (*get_predict_lbd_fn)(TX_SIZE tx_size); - -cfl_subtract_average_fn get_subtract_average_fn_c(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_sse2(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_subtract_average_fn (*get_subtract_average_fn)(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_c(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_hbd_fn (*cfl_get_predict_hbd_fn)(TX_SIZE tx_size); + +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_c(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_lbd_fn (*cfl_get_predict_lbd_fn)(TX_SIZE tx_size); + +cfl_subtract_average_fn cfl_get_subtract_average_fn_c(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_sse2(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_subtract_average_fn (*cfl_get_subtract_average_fn)(TX_SIZE tx_size); void av1_rtcd(void); @@ -437,9 +496,9 @@ static void setup_rtcd_internal(void) (void)flags; - apply_selfguided_restoration = apply_selfguided_restoration_c; - if (flags & HAS_SSE4_1) apply_selfguided_restoration = apply_selfguided_restoration_sse4_1; - if (flags & HAS_AVX2) apply_selfguided_restoration = apply_selfguided_restoration_avx2; + av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_c; + if (flags & HAS_SSE4_1) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_sse4_1; + if (flags & HAS_AVX2) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_avx2; av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_c; if (flags & HAS_SSE4_1) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_sse4_1; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_avx2; @@ -449,6 +508,8 @@ static void setup_rtcd_internal(void) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_c; if (flags & HAS_SSSE3) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_ssse3; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_avx2; + av1_calc_frame_error = av1_calc_frame_error_sse2; + if (flags & HAS_AVX2) av1_calc_frame_error = av1_calc_frame_error_avx2; av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_sse2; if (flags & HAS_AVX2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_avx2; av1_convolve_2d_scale = av1_convolve_2d_scale_c; @@ -461,6 +522,21 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) av1_convolve_x_sr = av1_convolve_x_sr_avx2; av1_convolve_y_sr = av1_convolve_y_sr_sse2; if (flags & HAS_AVX2) av1_convolve_y_sr = av1_convolve_y_sr_avx2; + av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_sse2; + if (flags & HAS_SSSE3) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_ssse3; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_avx2; + av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_avx2; + av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_avx2; + av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_avx2; + av1_dr_prediction_z1 = av1_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_dr_prediction_z1 = av1_dr_prediction_z1_avx2; + av1_dr_prediction_z2 = av1_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_dr_prediction_z2 = av1_dr_prediction_z2_avx2; + av1_dr_prediction_z3 = av1_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_dr_prediction_z3 = av1_dr_prediction_z3_avx2; av1_filter_intra_edge = av1_filter_intra_edge_c; if (flags & HAS_SSE4_1) av1_filter_intra_edge = av1_filter_intra_edge_sse4_1; av1_filter_intra_edge_high = av1_filter_intra_edge_high_c; @@ -482,34 +558,39 @@ static void setup_rtcd_internal(void) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_c; if (flags & HAS_SSSE3) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_ssse3; if (flags & HAS_AVX2) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_avx2; + av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_avx2; + av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_avx2; + av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_avx2; + av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_avx2; + av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_avx2; + av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_avx2; + av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_avx2; av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_sse4_1; if (flags & HAS_AVX2) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_avx2; - av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_sse4_1; - av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_sse4_1; - av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_sse4_1; - if (flags & HAS_AVX2) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_avx2; + av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_sse4_1; + av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_sse4_1; av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_sse4_1; - av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_sse4_1; + av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_sse4_1; + av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_sse4_1; av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_sse4_1; - av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_avx2; - av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_avx2; - av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_avx2; - av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_avx2; av1_highbd_warp_affine = av1_highbd_warp_affine_c; if (flags & HAS_SSE4_1) av1_highbd_warp_affine = av1_highbd_warp_affine_sse4_1; av1_highbd_wiener_convolve_add_src = av1_highbd_wiener_convolve_add_src_c; @@ -522,15 +603,8 @@ static void setup_rtcd_internal(void) av1_inv_txfm_add = av1_inv_txfm_add_c; if (flags & HAS_SSSE3) av1_inv_txfm_add = av1_inv_txfm_add_ssse3; if (flags & HAS_AVX2) av1_inv_txfm_add = av1_inv_txfm_add_avx2; - av1_jnt_convolve_2d = av1_jnt_convolve_2d_c; - if (flags & HAS_SSSE3) av1_jnt_convolve_2d = av1_jnt_convolve_2d_ssse3; - if (flags & HAS_AVX2) av1_jnt_convolve_2d = av1_jnt_convolve_2d_avx2; - av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_avx2; - av1_jnt_convolve_x = av1_jnt_convolve_x_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_x = av1_jnt_convolve_x_avx2; - av1_jnt_convolve_y = av1_jnt_convolve_y_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_y = av1_jnt_convolve_y_avx2; + av1_round_shift_array = av1_round_shift_array_c; + if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; av1_selfguided_restoration = av1_selfguided_restoration_c; if (flags & HAS_SSE4_1) av1_selfguided_restoration = av1_selfguided_restoration_sse4_1; if (flags & HAS_AVX2) av1_selfguided_restoration = av1_selfguided_restoration_avx2; @@ -540,8 +614,17 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE4_1) av1_upsample_intra_edge_high = av1_upsample_intra_edge_high_sse4_1; av1_warp_affine = av1_warp_affine_c; if (flags & HAS_SSE4_1) av1_warp_affine = av1_warp_affine_sse4_1; + if (flags & HAS_AVX2) av1_warp_affine = av1_warp_affine_avx2; av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_sse2; if (flags & HAS_AVX2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_avx2; + cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_avx2; + cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_avx2; cdef_filter_block = cdef_filter_block_sse2; if (flags & HAS_SSSE3) cdef_filter_block = cdef_filter_block_ssse3; if (flags & HAS_SSE4_1) cdef_filter_block = cdef_filter_block_sse4_1; @@ -568,22 +651,14 @@ static void setup_rtcd_internal(void) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_c; if (flags & HAS_SSSE3) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_ssse3; if (flags & HAS_AVX2) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_avx2; - copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_avx2; - copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_avx2; - get_predict_hbd_fn = get_predict_hbd_fn_c; - if (flags & HAS_SSSE3) get_predict_hbd_fn = get_predict_hbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_hbd_fn = get_predict_hbd_fn_avx2; - get_predict_lbd_fn = get_predict_lbd_fn_c; - if (flags & HAS_SSSE3) get_predict_lbd_fn = get_predict_lbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_lbd_fn = get_predict_lbd_fn_avx2; - get_subtract_average_fn = get_subtract_average_fn_sse2; - if (flags & HAS_AVX2) get_subtract_average_fn = get_subtract_average_fn_avx2; + cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_avx2; + cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_avx2; + cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_sse2; + if (flags & HAS_AVX2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_avx2; } #endif diff --git a/media/libaom/config/mac/x64/config/aom_config.asm b/media/libaom/config/mac/x64/config/aom_config.asm index 4ab2dedb42..9ce9b2090e 100644 --- a/media/libaom/config/mac/x64/config/aom_config.asm +++ b/media/libaom/config/mac/x64/config/aom_config.asm @@ -1,5 +1,5 @@ ; -; Copyright (c) 2018, Alliance for Open Media. All rights reserved +; Copyright (c) 2021, 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 @@ -14,49 +14,52 @@ ARCH_MIPS equ 0 ARCH_PPC equ 0 ARCH_X86 equ 0 ARCH_X86_64 equ 1 -CONFIG_2PASS_PARTITION_SEARCH_LVL equ 1 CONFIG_ACCOUNTING equ 0 CONFIG_ANALYZER equ 0 CONFIG_AV1_DECODER equ 1 CONFIG_AV1_ENCODER equ 0 +CONFIG_AV1_HIGHBITDEPTH equ 1 CONFIG_BIG_ENDIAN equ 0 CONFIG_BITSTREAM_DEBUG equ 0 CONFIG_COEFFICIENT_RANGE_CHECKING equ 0 -CONFIG_COLLECT_INTER_MODE_RD_STATS equ 0 +CONFIG_COLLECT_COMPONENT_TIMING equ 0 +CONFIG_COLLECT_PARTITION_STATS equ 0 CONFIG_COLLECT_RD_STATS equ 0 CONFIG_DEBUG equ 0 CONFIG_DENOISE equ 1 +CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 equ 1 CONFIG_DIST_8X8 equ 0 CONFIG_ENTROPY_STATS equ 0 -CONFIG_FILEOPTIONS equ 1 -CONFIG_FIX_GF_LENGTH equ 1 -CONFIG_FP_MB_STATS equ 0 CONFIG_GCC equ 1 CONFIG_GCOV equ 0 -CONFIG_GLOBAL_MOTION_SEARCH equ 1 CONFIG_GPROF equ 0 +CONFIG_HTB_TRELLIS equ 0 CONFIG_INSPECTION equ 0 CONFIG_INTERNAL_STATS equ 0 CONFIG_INTER_STATS_ONLY equ 0 CONFIG_LIBYUV equ 0 -CONFIG_LOWBITDEPTH equ 1 +CONFIG_LPF_MASK equ 0 CONFIG_MAX_DECODE_PROFILE equ 2 CONFIG_MISMATCH_DEBUG equ 0 CONFIG_MULTITHREAD equ 1 +CONFIG_NN_V2 equ 0 CONFIG_NORMAL_TILE_MODE equ 0 CONFIG_OS_SUPPORT equ 1 CONFIG_PIC equ 0 CONFIG_RD_DEBUG equ 0 -CONFIG_REDUCED_ENCODER_BORDER equ 0 +CONFIG_REALTIME_ONLY equ 0 CONFIG_RUNTIME_CPU_DETECT equ 1 CONFIG_SHARED equ 0 CONFIG_SHARP_SETTINGS equ 0 CONFIG_SIZE_LIMIT equ 0 CONFIG_SPATIAL_RESAMPLING equ 1 -CONFIG_STATIC equ 1 +CONFIG_SPEED_STATS equ 0 +CONFIG_SUPERRES_IN_RECODE equ 1 +CONFIG_TUNE_VMAF equ 0 CONFIG_WEBM_IO equ 0 DECODE_HEIGHT_LIMIT equ 0 DECODE_WIDTH_LIMIT equ 0 +FORCE_HIGHBITDEPTH_DECODING equ 0 HAVE_AVX equ 1 HAVE_AVX2 equ 1 HAVE_DSPR2 equ 0 diff --git a/media/libaom/config/mac/x64/config/aom_config.h b/media/libaom/config/mac/x64/config/aom_config.h index 9a0be7cd11..12997ea4dd 100644 --- a/media/libaom/config/mac/x64/config/aom_config.h +++ b/media/libaom/config/mac/x64/config/aom_config.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018, Alliance for Open Media. All rights reserved + * Copyright (c) 2021, 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 @@ -16,49 +16,52 @@ #define ARCH_PPC 0 #define ARCH_X86 0 #define ARCH_X86_64 1 -#define CONFIG_2PASS_PARTITION_SEARCH_LVL 1 #define CONFIG_ACCOUNTING 0 #define CONFIG_ANALYZER 0 #define CONFIG_AV1_DECODER 1 #define CONFIG_AV1_ENCODER 0 +#define CONFIG_AV1_HIGHBITDEPTH 1 #define CONFIG_BIG_ENDIAN 0 #define CONFIG_BITSTREAM_DEBUG 0 #define CONFIG_COEFFICIENT_RANGE_CHECKING 0 -#define CONFIG_COLLECT_INTER_MODE_RD_STATS 0 +#define CONFIG_COLLECT_COMPONENT_TIMING 0 +#define CONFIG_COLLECT_PARTITION_STATS 0 #define CONFIG_COLLECT_RD_STATS 0 #define CONFIG_DEBUG 0 #define CONFIG_DENOISE 1 +#define CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 1 #define CONFIG_DIST_8X8 0 #define CONFIG_ENTROPY_STATS 0 -#define CONFIG_FILEOPTIONS 1 -#define CONFIG_FIX_GF_LENGTH 1 -#define CONFIG_FP_MB_STATS 0 #define CONFIG_GCC 1 #define CONFIG_GCOV 0 -#define CONFIG_GLOBAL_MOTION_SEARCH 1 #define CONFIG_GPROF 0 +#define CONFIG_HTB_TRELLIS 0 #define CONFIG_INSPECTION 0 #define CONFIG_INTERNAL_STATS 0 #define CONFIG_INTER_STATS_ONLY 0 #define CONFIG_LIBYUV 0 -#define CONFIG_LOWBITDEPTH 1 +#define CONFIG_LPF_MASK 0 #define CONFIG_MAX_DECODE_PROFILE 2 #define CONFIG_MISMATCH_DEBUG 0 #define CONFIG_MULTITHREAD 1 +#define CONFIG_NN_V2 0 #define CONFIG_NORMAL_TILE_MODE 0 #define CONFIG_OS_SUPPORT 1 #define CONFIG_PIC 0 #define CONFIG_RD_DEBUG 0 -#define CONFIG_REDUCED_ENCODER_BORDER 0 +#define CONFIG_REALTIME_ONLY 0 #define CONFIG_RUNTIME_CPU_DETECT 1 #define CONFIG_SHARED 0 #define CONFIG_SHARP_SETTINGS 0 #define CONFIG_SIZE_LIMIT 0 #define CONFIG_SPATIAL_RESAMPLING 1 -#define CONFIG_STATIC 1 +#define CONFIG_SPEED_STATS 0 +#define CONFIG_SUPERRES_IN_RECODE 1 +#define CONFIG_TUNE_VMAF 0 #define CONFIG_WEBM_IO 0 #define DECODE_HEIGHT_LIMIT 0 #define DECODE_WIDTH_LIMIT 0 +#define FORCE_HIGHBITDEPTH_DECODING 0 #define HAVE_AVX 1 #define HAVE_AVX2 1 #define HAVE_DSPR2 0 @@ -76,7 +79,5 @@ #define HAVE_SSSE3 1 #define HAVE_VSX 0 #define HAVE_WXWIDGETS 0 -#define INCLUDE_INSTALL_DIR INSTALLDIR/include #define INLINE inline -#define LIB_INSTALL_DIR INSTALLDIR/lib #endif /* AOM_CONFIG_H_ */ diff --git a/media/libaom/config/mac/x64/config/aom_dsp_rtcd.h b/media/libaom/config/mac/x64/config/aom_dsp_rtcd.h index 2856d4ede7..36c170a05d 100644 --- a/media/libaom/config/mac/x64/config/aom_dsp_rtcd.h +++ b/media/libaom/config/mac/x64/config/aom_dsp_rtcd.h @@ -26,10 +26,10 @@ void aom_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src void aom_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); RTCD_EXTERN void (*aom_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); -void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); +void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); void aom_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); void aom_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); @@ -47,8 +47,8 @@ void aom_convolve8_vert_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t void aom_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*aom_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); +void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); #define aom_convolve_copy aom_convolve_copy_sse2 void aom_dc_128_predictor_16x16_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); @@ -471,35 +471,37 @@ void aom_h_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_h_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_h_predictor_8x8 aom_h_predictor_8x8_sse2 -void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params, const int bd); -#define aom_highbd_blend_a64_d16_mask aom_highbd_blend_a64_d16_mask_c +void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); void aom_highbd_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); +void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); void aom_highbd_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_vmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); void aom_highbd_dc_128_predictor_16x16_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); void aom_highbd_dc_128_predictor_16x16_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); @@ -855,10 +857,10 @@ void aom_highbd_lpf_horizontal_14_c(uint16_t *s, int pitch, const uint8_t *blimi void aom_highbd_lpf_horizontal_14_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); #define aom_highbd_lpf_horizontal_14 aom_highbd_lpf_horizontal_14_sse2 -void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); void aom_highbd_lpf_horizontal_4_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); void aom_highbd_lpf_horizontal_4_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); @@ -1231,10 +1233,10 @@ void aom_highbd_v_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_v_predictor_8x8_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); #define aom_highbd_v_predictor_8x8 aom_highbd_v_predictor_8x8_sse2 -void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); void aom_lpf_horizontal_14_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void aom_lpf_horizontal_14_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -1711,10 +1713,6 @@ void aom_v_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_v_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_v_predictor_8x8 aom_v_predictor_8x8_sse2 -void av1_round_shift_array_c(int32_t *arr, int size, int bit); -void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); -RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); - void aom_dsp_rtcd(void); #ifdef RTCD_C @@ -1788,6 +1786,9 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) aom_dc_top_predictor_64x64 = aom_dc_top_predictor_64x64_avx2; aom_h_predictor_32x32 = aom_h_predictor_32x32_sse2; if (flags & HAS_AVX2) aom_h_predictor_32x32 = aom_h_predictor_32x32_avx2; + aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_c; + if (flags & HAS_SSE4_1) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_sse4_1; + if (flags & HAS_AVX2) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_avx2; aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_sse4_1; aom_highbd_blend_a64_mask = aom_highbd_blend_a64_mask_c; @@ -1989,8 +1990,6 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) aom_v_predictor_64x32 = aom_v_predictor_64x32_avx2; aom_v_predictor_64x64 = aom_v_predictor_64x64_sse2; if (flags & HAS_AVX2) aom_v_predictor_64x64 = aom_v_predictor_64x64_avx2; - av1_round_shift_array = av1_round_shift_array_c; - if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; } #endif diff --git a/media/libaom/config/mac/x64/config/aom_scale_rtcd.h b/media/libaom/config/mac/x64/config/aom_scale_rtcd.h index 6b56795cd4..65c184b83d 100644 --- a/media/libaom/config/mac/x64/config/aom_scale_rtcd.h +++ b/media/libaom/config/mac/x64/config/aom_scale_rtcd.h @@ -59,15 +59,27 @@ void aom_yv12_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buf void aom_yv12_extend_frame_borders_c(struct yv12_buffer_config *ybf, const int num_planes); #define aom_yv12_extend_frame_borders aom_yv12_extend_frame_borders_c -void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_coloc_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_u aom_yv12_partial_coloc_copy_u_c + +void aom_yv12_partial_coloc_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_v aom_yv12_partial_coloc_copy_v_c + +void aom_yv12_partial_coloc_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_y aom_yv12_partial_coloc_copy_y_c + +void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_u aom_yv12_partial_copy_u_c -void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_v aom_yv12_partial_copy_v_c -void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_ybc, int hstart2, int vstart2); #define aom_yv12_partial_copy_y aom_yv12_partial_copy_y_c +int aom_yv12_realloc_with_new_border_c(struct yv12_buffer_config *ybf, int new_border, int byte_alignment, int num_planes); +#define aom_yv12_realloc_with_new_border aom_yv12_realloc_with_new_border_c + void aom_scale_rtcd(void); #ifdef RTCD_C diff --git a/media/libaom/config/mac/x64/config/av1_rtcd.h b/media/libaom/config/mac/x64/config/av1_rtcd.h index d273182086..8f20181b42 100644 --- a/media/libaom/config/mac/x64/config/av1_rtcd.h +++ b/media/libaom/config/mac/x64/config/av1_rtcd.h @@ -31,30 +31,55 @@ struct txfm_param; struct aom_variance_vtable; struct search_site_config; struct yv12_buffer_config; +struct NN_CONFIG; +typedef struct NN_CONFIG NN_CONFIG; + +enum { NONE, RELU, SOFTSIGN, SIGMOID } UENUM1BYTE(ACTIVATION); +#if CONFIG_NN_V2 +enum { SOFTMAX_CROSS_ENTROPY } UENUM1BYTE(LOSS); +struct NN_CONFIG_V2; +typedef struct NN_CONFIG_V2 NN_CONFIG_V2; +struct FC_LAYER; +typedef struct FC_LAYER FC_LAYER; +#endif // CONFIG_NN_V2 + +struct CNN_CONFIG; +typedef struct CNN_CONFIG CNN_CONFIG; +struct CNN_LAYER_CONFIG; +typedef struct CNN_LAYER_CONFIG CNN_LAYER_CONFIG; +struct CNN_THREAD_DATA; +typedef struct CNN_THREAD_DATA CNN_THREAD_DATA; +struct CNN_BRANCH_CONFIG; +typedef struct CNN_BRANCH_CONFIG CNN_BRANCH_CONFIG; +struct CNN_MULTI_OUT; +typedef struct CNN_MULTI_OUT CNN_MULTI_OUT; /* Function pointers return by CfL functions */ typedef void (*cfl_subsample_lbd_fn)(const uint8_t *input, int input_stride, uint16_t *output_q3); +#if CONFIG_AV1_HIGHBITDEPTH typedef void (*cfl_subsample_hbd_fn)(const uint16_t *input, int input_stride, uint16_t *output_q3); +typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, + int dst_stride, int alpha_q3, int bd); +#endif + typedef void (*cfl_subtract_average_fn)(const uint16_t *src, int16_t *dst); typedef void (*cfl_predict_lbd_fn)(const int16_t *src, uint8_t *dst, int dst_stride, int alpha_q3); -typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, - int dst_stride, int alpha_q3, int bd); #ifdef __cplusplus extern "C" { #endif -void apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -RTCD_EXTERN void (*apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +RTCD_EXTERN void (*av1_apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); void av1_build_compound_diffwtd_mask_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); void av1_build_compound_diffwtd_mask_sse4_1(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); @@ -71,42 +96,89 @@ void av1_build_compound_diffwtd_mask_highbd_ssse3(uint8_t *mask, DIFFWTD_MASK_TY void av1_build_compound_diffwtd_mask_highbd_avx2(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); RTCD_EXTERN void (*av1_build_compound_diffwtd_mask_highbd)(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); -void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +int64_t av1_calc_frame_error_c(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_sse2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_avx2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +RTCD_EXTERN int64_t (*av1_calc_frame_error)(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); + +void av1_cnn_activate_c( float **input, int channels, int width, int height, int stride, ACTIVATION layer_activation); +#define av1_cnn_activate av1_cnn_activate_c -void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); +void av1_cnn_add_c( float **input, int channels, int width, int height, int stride, const float **add); +#define av1_cnn_add av1_cnn_add_c -void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_cnn_batchnorm_c(float **image, int channels, int width, int height, int stride, const float *gamma, const float *beta, const float *mean, const float *std); +#define av1_cnn_batchnorm av1_cnn_batchnorm_c + +void av1_cnn_convolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride, int start_idx, int step); +#define av1_cnn_convolve av1_cnn_convolve_c + +void av1_cnn_deconvolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride); +#define av1_cnn_deconvolve av1_cnn_deconvolve_c + +void av1_cnn_predict_c( const float **input, int in_width, int in_height, int in_stride, const CNN_CONFIG *cnn_config, const CNN_THREAD_DATA *thread_data, CNN_MULTI_OUT *output_struct); +#define av1_cnn_predict av1_cnn_predict_c + +void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_convolve_horiz_rs_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); void av1_convolve_horiz_rs_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); RTCD_EXTERN void (*av1_convolve_horiz_rs)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); -void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_dr_prediction_z1_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); -#define av1_dr_prediction_z1 av1_dr_prediction_z1_c +void av1_dr_prediction_z1_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z1)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); void av1_dr_prediction_z2_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z2 av1_dr_prediction_z2_c +void av1_dr_prediction_z2_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z2)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); void av1_dr_prediction_z3_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z3 av1_dr_prediction_z3_c +void av1_dr_prediction_z3_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z3)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); void av1_filter_intra_edge_c(uint8_t *p, int sz, int strength); void av1_filter_intra_edge_sse4_1(uint8_t *p, int sz, int strength); @@ -132,19 +204,19 @@ void av1_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8 void av1_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve8_vert av1_highbd_convolve8_vert_sse2 -void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve_avg av1_highbd_convolve_avg_c @@ -156,54 +228,76 @@ void av1_highbd_convolve_horiz_rs_c(const uint16_t *src, int src_stride, uint16_ void av1_highbd_convolve_horiz_rs_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); RTCD_EXTERN void (*av1_highbd_convolve_horiz_rs)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); -void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_dr_prediction_z1_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z1 av1_highbd_dr_prediction_z1_c +void av1_highbd_dr_prediction_z1_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z1)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); void av1_highbd_dr_prediction_z2_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z2 av1_highbd_dr_prediction_z2_c +void av1_highbd_dr_prediction_z2_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z2)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); void av1_highbd_dr_prediction_z3_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z3 av1_highbd_dr_prediction_z3_c +void av1_highbd_dr_prediction_z3_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z3)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -void av1_highbd_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_avx2(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x16)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_32x32)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_16_add av1_highbd_iwht4x4_16_add_c @@ -211,34 +305,14 @@ void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int des void av1_highbd_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_1_add av1_highbd_iwht4x4_1_add_c -void av1_highbd_jnt_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - void av1_highbd_warp_affine_c(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_highbd_warp_affine_sse4_1(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_highbd_warp_affine)(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); -void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); +void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); void av1_inv_txfm2d_add_16x16_c(const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type, int bd); #define av1_inv_txfm2d_add_16x16 av1_inv_txfm2d_add_16x16_c @@ -304,25 +378,9 @@ void av1_inv_txfm_add_ssse3(const tran_low_t *dqcoeff, uint8_t *dst, int stride, void av1_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); RTCD_EXTERN void (*av1_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_jnt_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_round_shift_array_c(int32_t *arr, int size, int bit); +void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); +RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); int av1_selfguided_restoration_c(const uint8_t *dgd8, int width, int height, int dgd_stride, int32_t *flt0, int32_t *flt1, int flt_stride, @@ -347,6 +405,7 @@ RTCD_EXTERN void (*av1_upsample_intra_edge_high)(uint16_t *p, int sz, int bd); void av1_warp_affine_c(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_warp_affine_sse4_1(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); +void av1_warp_affine_avx2(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_warp_affine)(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); @@ -354,12 +413,26 @@ void av1_wiener_convolve_add_src_sse2(const uint8_t *src, ptrdiff_t src_stride, void av1_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); RTCD_EXTERN void (*av1_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); -void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); +void cdef_copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); + +void cdef_copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); + +void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); int cdef_find_dir_c(const uint16_t *img, int stride, int32_t *var, int coeff_shift); int cdef_find_dir_sse2(const uint16_t *img, int stride, int32_t *var, int coeff_shift); @@ -398,34 +471,20 @@ cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_ssse3(TX_SIZE tx_size); cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_avx2(TX_SIZE tx_size); RTCD_EXTERN cfl_subsample_lbd_fn (*cfl_get_luma_subsampling_444_lbd)(TX_SIZE tx_size); -void copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); - -void copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); - -cfl_predict_hbd_fn get_predict_hbd_fn_c(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_hbd_fn (*get_predict_hbd_fn)(TX_SIZE tx_size); - -cfl_predict_lbd_fn get_predict_lbd_fn_c(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_lbd_fn (*get_predict_lbd_fn)(TX_SIZE tx_size); - -cfl_subtract_average_fn get_subtract_average_fn_c(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_sse2(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_subtract_average_fn (*get_subtract_average_fn)(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_c(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_hbd_fn (*cfl_get_predict_hbd_fn)(TX_SIZE tx_size); + +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_c(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_lbd_fn (*cfl_get_predict_lbd_fn)(TX_SIZE tx_size); + +cfl_subtract_average_fn cfl_get_subtract_average_fn_c(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_sse2(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_subtract_average_fn (*cfl_get_subtract_average_fn)(TX_SIZE tx_size); void av1_rtcd(void); @@ -437,9 +496,9 @@ static void setup_rtcd_internal(void) (void)flags; - apply_selfguided_restoration = apply_selfguided_restoration_c; - if (flags & HAS_SSE4_1) apply_selfguided_restoration = apply_selfguided_restoration_sse4_1; - if (flags & HAS_AVX2) apply_selfguided_restoration = apply_selfguided_restoration_avx2; + av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_c; + if (flags & HAS_SSE4_1) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_sse4_1; + if (flags & HAS_AVX2) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_avx2; av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_c; if (flags & HAS_SSE4_1) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_sse4_1; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_avx2; @@ -449,6 +508,8 @@ static void setup_rtcd_internal(void) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_c; if (flags & HAS_SSSE3) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_ssse3; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_avx2; + av1_calc_frame_error = av1_calc_frame_error_sse2; + if (flags & HAS_AVX2) av1_calc_frame_error = av1_calc_frame_error_avx2; av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_sse2; if (flags & HAS_AVX2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_avx2; av1_convolve_2d_scale = av1_convolve_2d_scale_c; @@ -461,6 +522,21 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) av1_convolve_x_sr = av1_convolve_x_sr_avx2; av1_convolve_y_sr = av1_convolve_y_sr_sse2; if (flags & HAS_AVX2) av1_convolve_y_sr = av1_convolve_y_sr_avx2; + av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_sse2; + if (flags & HAS_SSSE3) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_ssse3; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_avx2; + av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_avx2; + av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_avx2; + av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_avx2; + av1_dr_prediction_z1 = av1_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_dr_prediction_z1 = av1_dr_prediction_z1_avx2; + av1_dr_prediction_z2 = av1_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_dr_prediction_z2 = av1_dr_prediction_z2_avx2; + av1_dr_prediction_z3 = av1_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_dr_prediction_z3 = av1_dr_prediction_z3_avx2; av1_filter_intra_edge = av1_filter_intra_edge_c; if (flags & HAS_SSE4_1) av1_filter_intra_edge = av1_filter_intra_edge_sse4_1; av1_filter_intra_edge_high = av1_filter_intra_edge_high_c; @@ -482,34 +558,39 @@ static void setup_rtcd_internal(void) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_c; if (flags & HAS_SSSE3) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_ssse3; if (flags & HAS_AVX2) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_avx2; + av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_avx2; + av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_avx2; + av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_avx2; + av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_avx2; + av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_avx2; + av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_avx2; + av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_avx2; av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_sse4_1; if (flags & HAS_AVX2) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_avx2; - av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_sse4_1; - av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_sse4_1; - av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_sse4_1; - if (flags & HAS_AVX2) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_avx2; + av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_sse4_1; + av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_sse4_1; av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_sse4_1; - av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_sse4_1; + av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_sse4_1; + av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_sse4_1; av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_sse4_1; - av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_avx2; - av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_avx2; - av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_avx2; - av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_avx2; av1_highbd_warp_affine = av1_highbd_warp_affine_c; if (flags & HAS_SSE4_1) av1_highbd_warp_affine = av1_highbd_warp_affine_sse4_1; av1_highbd_wiener_convolve_add_src = av1_highbd_wiener_convolve_add_src_c; @@ -522,15 +603,8 @@ static void setup_rtcd_internal(void) av1_inv_txfm_add = av1_inv_txfm_add_c; if (flags & HAS_SSSE3) av1_inv_txfm_add = av1_inv_txfm_add_ssse3; if (flags & HAS_AVX2) av1_inv_txfm_add = av1_inv_txfm_add_avx2; - av1_jnt_convolve_2d = av1_jnt_convolve_2d_c; - if (flags & HAS_SSSE3) av1_jnt_convolve_2d = av1_jnt_convolve_2d_ssse3; - if (flags & HAS_AVX2) av1_jnt_convolve_2d = av1_jnt_convolve_2d_avx2; - av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_avx2; - av1_jnt_convolve_x = av1_jnt_convolve_x_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_x = av1_jnt_convolve_x_avx2; - av1_jnt_convolve_y = av1_jnt_convolve_y_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_y = av1_jnt_convolve_y_avx2; + av1_round_shift_array = av1_round_shift_array_c; + if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; av1_selfguided_restoration = av1_selfguided_restoration_c; if (flags & HAS_SSE4_1) av1_selfguided_restoration = av1_selfguided_restoration_sse4_1; if (flags & HAS_AVX2) av1_selfguided_restoration = av1_selfguided_restoration_avx2; @@ -540,8 +614,17 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE4_1) av1_upsample_intra_edge_high = av1_upsample_intra_edge_high_sse4_1; av1_warp_affine = av1_warp_affine_c; if (flags & HAS_SSE4_1) av1_warp_affine = av1_warp_affine_sse4_1; + if (flags & HAS_AVX2) av1_warp_affine = av1_warp_affine_avx2; av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_sse2; if (flags & HAS_AVX2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_avx2; + cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_avx2; + cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_avx2; cdef_filter_block = cdef_filter_block_sse2; if (flags & HAS_SSSE3) cdef_filter_block = cdef_filter_block_ssse3; if (flags & HAS_SSE4_1) cdef_filter_block = cdef_filter_block_sse4_1; @@ -568,22 +651,14 @@ static void setup_rtcd_internal(void) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_c; if (flags & HAS_SSSE3) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_ssse3; if (flags & HAS_AVX2) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_avx2; - copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_avx2; - copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_avx2; - get_predict_hbd_fn = get_predict_hbd_fn_c; - if (flags & HAS_SSSE3) get_predict_hbd_fn = get_predict_hbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_hbd_fn = get_predict_hbd_fn_avx2; - get_predict_lbd_fn = get_predict_lbd_fn_c; - if (flags & HAS_SSSE3) get_predict_lbd_fn = get_predict_lbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_lbd_fn = get_predict_lbd_fn_avx2; - get_subtract_average_fn = get_subtract_average_fn_sse2; - if (flags & HAS_AVX2) get_subtract_average_fn = get_subtract_average_fn_avx2; + cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_avx2; + cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_avx2; + cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_sse2; + if (flags & HAS_AVX2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_avx2; } #endif diff --git a/media/libaom/config/win/ia32/config/aom_config.asm b/media/libaom/config/win/ia32/config/aom_config.asm index 6ae776c7c4..baf05cb910 100644 --- a/media/libaom/config/win/ia32/config/aom_config.asm +++ b/media/libaom/config/win/ia32/config/aom_config.asm @@ -1,5 +1,5 @@ ; -; Copyright (c) 2018, Alliance for Open Media. All rights reserved +; Copyright (c) 2021, 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 @@ -14,49 +14,52 @@ ARCH_MIPS equ 0 ARCH_PPC equ 0 ARCH_X86 equ 1 ARCH_X86_64 equ 0 -CONFIG_2PASS_PARTITION_SEARCH_LVL equ 1 CONFIG_ACCOUNTING equ 0 CONFIG_ANALYZER equ 0 CONFIG_AV1_DECODER equ 1 CONFIG_AV1_ENCODER equ 0 +CONFIG_AV1_HIGHBITDEPTH equ 1 CONFIG_BIG_ENDIAN equ 0 CONFIG_BITSTREAM_DEBUG equ 0 CONFIG_COEFFICIENT_RANGE_CHECKING equ 0 -CONFIG_COLLECT_INTER_MODE_RD_STATS equ 0 +CONFIG_COLLECT_COMPONENT_TIMING equ 0 +CONFIG_COLLECT_PARTITION_STATS equ 0 CONFIG_COLLECT_RD_STATS equ 0 CONFIG_DEBUG equ 0 CONFIG_DENOISE equ 1 +CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 equ 1 CONFIG_DIST_8X8 equ 0 CONFIG_ENTROPY_STATS equ 0 -CONFIG_FILEOPTIONS equ 1 -CONFIG_FIX_GF_LENGTH equ 1 -CONFIG_FP_MB_STATS equ 0 CONFIG_GCC equ 1 CONFIG_GCOV equ 0 -CONFIG_GLOBAL_MOTION_SEARCH equ 1 CONFIG_GPROF equ 0 +CONFIG_HTB_TRELLIS equ 0 CONFIG_INSPECTION equ 0 CONFIG_INTERNAL_STATS equ 0 CONFIG_INTER_STATS_ONLY equ 0 CONFIG_LIBYUV equ 0 -CONFIG_LOWBITDEPTH equ 1 +CONFIG_LPF_MASK equ 0 CONFIG_MAX_DECODE_PROFILE equ 2 CONFIG_MISMATCH_DEBUG equ 0 CONFIG_MULTITHREAD equ 1 +CONFIG_NN_V2 equ 0 CONFIG_NORMAL_TILE_MODE equ 0 CONFIG_OS_SUPPORT equ 1 CONFIG_PIC equ 0 CONFIG_RD_DEBUG equ 0 -CONFIG_REDUCED_ENCODER_BORDER equ 0 +CONFIG_REALTIME_ONLY equ 0 CONFIG_RUNTIME_CPU_DETECT equ 1 CONFIG_SHARED equ 0 CONFIG_SHARP_SETTINGS equ 0 CONFIG_SIZE_LIMIT equ 0 CONFIG_SPATIAL_RESAMPLING equ 1 -CONFIG_STATIC equ 1 +CONFIG_SPEED_STATS equ 0 +CONFIG_SUPERRES_IN_RECODE equ 1 +CONFIG_TUNE_VMAF equ 0 CONFIG_WEBM_IO equ 0 DECODE_HEIGHT_LIMIT equ 0 DECODE_WIDTH_LIMIT equ 0 +FORCE_HIGHBITDEPTH_DECODING equ 0 HAVE_AVX equ 1 HAVE_AVX2 equ 1 HAVE_DSPR2 equ 0 diff --git a/media/libaom/config/win/ia32/config/aom_config.h b/media/libaom/config/win/ia32/config/aom_config.h index 090a455778..fc1271c743 100644 --- a/media/libaom/config/win/ia32/config/aom_config.h +++ b/media/libaom/config/win/ia32/config/aom_config.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018, Alliance for Open Media. All rights reserved + * Copyright (c) 2021, 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 @@ -16,49 +16,52 @@ #define ARCH_PPC 0 #define ARCH_X86 1 #define ARCH_X86_64 0 -#define CONFIG_2PASS_PARTITION_SEARCH_LVL 1 #define CONFIG_ACCOUNTING 0 #define CONFIG_ANALYZER 0 #define CONFIG_AV1_DECODER 1 #define CONFIG_AV1_ENCODER 0 +#define CONFIG_AV1_HIGHBITDEPTH 1 #define CONFIG_BIG_ENDIAN 0 #define CONFIG_BITSTREAM_DEBUG 0 #define CONFIG_COEFFICIENT_RANGE_CHECKING 0 -#define CONFIG_COLLECT_INTER_MODE_RD_STATS 0 +#define CONFIG_COLLECT_COMPONENT_TIMING 0 +#define CONFIG_COLLECT_PARTITION_STATS 0 #define CONFIG_COLLECT_RD_STATS 0 #define CONFIG_DEBUG 0 #define CONFIG_DENOISE 1 +#define CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 1 #define CONFIG_DIST_8X8 0 #define CONFIG_ENTROPY_STATS 0 -#define CONFIG_FILEOPTIONS 1 -#define CONFIG_FIX_GF_LENGTH 1 -#define CONFIG_FP_MB_STATS 0 #define CONFIG_GCC 1 #define CONFIG_GCOV 0 -#define CONFIG_GLOBAL_MOTION_SEARCH 1 #define CONFIG_GPROF 0 +#define CONFIG_HTB_TRELLIS 0 #define CONFIG_INSPECTION 0 #define CONFIG_INTERNAL_STATS 0 #define CONFIG_INTER_STATS_ONLY 0 #define CONFIG_LIBYUV 0 -#define CONFIG_LOWBITDEPTH 1 +#define CONFIG_LPF_MASK 0 #define CONFIG_MAX_DECODE_PROFILE 2 #define CONFIG_MISMATCH_DEBUG 0 #define CONFIG_MULTITHREAD 1 +#define CONFIG_NN_V2 0 #define CONFIG_NORMAL_TILE_MODE 0 #define CONFIG_OS_SUPPORT 1 #define CONFIG_PIC 0 #define CONFIG_RD_DEBUG 0 -#define CONFIG_REDUCED_ENCODER_BORDER 0 +#define CONFIG_REALTIME_ONLY 0 #define CONFIG_RUNTIME_CPU_DETECT 1 #define CONFIG_SHARED 0 #define CONFIG_SHARP_SETTINGS 0 #define CONFIG_SIZE_LIMIT 0 #define CONFIG_SPATIAL_RESAMPLING 1 -#define CONFIG_STATIC 1 +#define CONFIG_SPEED_STATS 0 +#define CONFIG_SUPERRES_IN_RECODE 1 +#define CONFIG_TUNE_VMAF 0 #define CONFIG_WEBM_IO 0 #define DECODE_HEIGHT_LIMIT 0 #define DECODE_WIDTH_LIMIT 0 +#define FORCE_HIGHBITDEPTH_DECODING 0 #define HAVE_AVX 1 #define HAVE_AVX2 1 #define HAVE_DSPR2 0 @@ -76,7 +79,5 @@ #define HAVE_SSSE3 1 #define HAVE_VSX 0 #define HAVE_WXWIDGETS 0 -#define INCLUDE_INSTALL_DIR INSTALLDIR/include #define INLINE inline -#define LIB_INSTALL_DIR INSTALLDIR/lib #endif /* AOM_CONFIG_H_ */ diff --git a/media/libaom/config/win/ia32/config/aom_dsp_rtcd.h b/media/libaom/config/win/ia32/config/aom_dsp_rtcd.h index 5b3762f109..9018213bd6 100644 --- a/media/libaom/config/win/ia32/config/aom_dsp_rtcd.h +++ b/media/libaom/config/win/ia32/config/aom_dsp_rtcd.h @@ -26,10 +26,10 @@ void aom_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src void aom_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); RTCD_EXTERN void (*aom_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); -void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); +void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); void aom_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); void aom_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); @@ -47,9 +47,9 @@ void aom_convolve8_vert_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t void aom_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*aom_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -RTCD_EXTERN void (*aom_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); +void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +RTCD_EXTERN void (*aom_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); void aom_dc_128_predictor_16x16_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); void aom_dc_128_predictor_16x16_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); @@ -471,33 +471,37 @@ void aom_h_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_h_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); RTCD_EXTERN void (*aom_h_predictor_8x8)(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); -void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params, const int bd); -#define aom_highbd_blend_a64_d16_mask aom_highbd_blend_a64_d16_mask_c +void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); void aom_highbd_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); +void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); void aom_highbd_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_vmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); void aom_highbd_dc_128_predictor_16x16_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); void aom_highbd_dc_128_predictor_16x16_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); @@ -853,10 +857,10 @@ void aom_highbd_lpf_horizontal_14_c(uint16_t *s, int pitch, const uint8_t *blimi void aom_highbd_lpf_horizontal_14_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14)(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); -void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); void aom_highbd_lpf_horizontal_4_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); void aom_highbd_lpf_horizontal_4_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); @@ -1229,10 +1233,10 @@ void aom_highbd_v_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_v_predictor_8x8_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); RTCD_EXTERN void (*aom_highbd_v_predictor_8x8)(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); -void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); void aom_lpf_horizontal_14_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void aom_lpf_horizontal_14_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -1709,10 +1713,6 @@ void aom_v_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_v_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); RTCD_EXTERN void (*aom_v_predictor_8x8)(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); -void av1_round_shift_array_c(int32_t *arr, int size, int bit); -void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); -RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); - void aom_dsp_rtcd(void); #ifdef RTCD_C @@ -1955,6 +1955,9 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE2) aom_h_predictor_8x4 = aom_h_predictor_8x4_sse2; aom_h_predictor_8x8 = aom_h_predictor_8x8_c; if (flags & HAS_SSE2) aom_h_predictor_8x8 = aom_h_predictor_8x8_sse2; + aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_c; + if (flags & HAS_SSE4_1) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_sse4_1; + if (flags & HAS_AVX2) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_avx2; aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_sse4_1; aom_highbd_blend_a64_mask = aom_highbd_blend_a64_mask_c; @@ -1962,8 +1965,10 @@ static void setup_rtcd_internal(void) aom_highbd_blend_a64_vmask = aom_highbd_blend_a64_vmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_vmask = aom_highbd_blend_a64_vmask_sse4_1; aom_highbd_convolve8_horiz = aom_highbd_convolve8_horiz_c; + if (flags & HAS_SSE2) aom_highbd_convolve8_horiz = aom_highbd_convolve8_horiz_sse2; if (flags & HAS_AVX2) aom_highbd_convolve8_horiz = aom_highbd_convolve8_horiz_avx2; aom_highbd_convolve8_vert = aom_highbd_convolve8_vert_c; + if (flags & HAS_SSE2) aom_highbd_convolve8_vert = aom_highbd_convolve8_vert_sse2; if (flags & HAS_AVX2) aom_highbd_convolve8_vert = aom_highbd_convolve8_vert_avx2; aom_highbd_convolve_copy = aom_highbd_convolve_copy_c; if (flags & HAS_SSE2) aom_highbd_convolve_copy = aom_highbd_convolve_copy_sse2; @@ -2367,8 +2372,6 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE2) aom_v_predictor_8x4 = aom_v_predictor_8x4_sse2; aom_v_predictor_8x8 = aom_v_predictor_8x8_c; if (flags & HAS_SSE2) aom_v_predictor_8x8 = aom_v_predictor_8x8_sse2; - av1_round_shift_array = av1_round_shift_array_c; - if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; } #endif diff --git a/media/libaom/config/win/ia32/config/aom_scale_rtcd.h b/media/libaom/config/win/ia32/config/aom_scale_rtcd.h index 6b56795cd4..65c184b83d 100644 --- a/media/libaom/config/win/ia32/config/aom_scale_rtcd.h +++ b/media/libaom/config/win/ia32/config/aom_scale_rtcd.h @@ -59,15 +59,27 @@ void aom_yv12_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buf void aom_yv12_extend_frame_borders_c(struct yv12_buffer_config *ybf, const int num_planes); #define aom_yv12_extend_frame_borders aom_yv12_extend_frame_borders_c -void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_coloc_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_u aom_yv12_partial_coloc_copy_u_c + +void aom_yv12_partial_coloc_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_v aom_yv12_partial_coloc_copy_v_c + +void aom_yv12_partial_coloc_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_y aom_yv12_partial_coloc_copy_y_c + +void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_u aom_yv12_partial_copy_u_c -void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_v aom_yv12_partial_copy_v_c -void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_ybc, int hstart2, int vstart2); #define aom_yv12_partial_copy_y aom_yv12_partial_copy_y_c +int aom_yv12_realloc_with_new_border_c(struct yv12_buffer_config *ybf, int new_border, int byte_alignment, int num_planes); +#define aom_yv12_realloc_with_new_border aom_yv12_realloc_with_new_border_c + void aom_scale_rtcd(void); #ifdef RTCD_C diff --git a/media/libaom/config/win/ia32/config/av1_rtcd.h b/media/libaom/config/win/ia32/config/av1_rtcd.h index a2a6f14ff5..7ecd05f380 100644 --- a/media/libaom/config/win/ia32/config/av1_rtcd.h +++ b/media/libaom/config/win/ia32/config/av1_rtcd.h @@ -31,30 +31,55 @@ struct txfm_param; struct aom_variance_vtable; struct search_site_config; struct yv12_buffer_config; +struct NN_CONFIG; +typedef struct NN_CONFIG NN_CONFIG; + +enum { NONE, RELU, SOFTSIGN, SIGMOID } UENUM1BYTE(ACTIVATION); +#if CONFIG_NN_V2 +enum { SOFTMAX_CROSS_ENTROPY } UENUM1BYTE(LOSS); +struct NN_CONFIG_V2; +typedef struct NN_CONFIG_V2 NN_CONFIG_V2; +struct FC_LAYER; +typedef struct FC_LAYER FC_LAYER; +#endif // CONFIG_NN_V2 + +struct CNN_CONFIG; +typedef struct CNN_CONFIG CNN_CONFIG; +struct CNN_LAYER_CONFIG; +typedef struct CNN_LAYER_CONFIG CNN_LAYER_CONFIG; +struct CNN_THREAD_DATA; +typedef struct CNN_THREAD_DATA CNN_THREAD_DATA; +struct CNN_BRANCH_CONFIG; +typedef struct CNN_BRANCH_CONFIG CNN_BRANCH_CONFIG; +struct CNN_MULTI_OUT; +typedef struct CNN_MULTI_OUT CNN_MULTI_OUT; /* Function pointers return by CfL functions */ typedef void (*cfl_subsample_lbd_fn)(const uint8_t *input, int input_stride, uint16_t *output_q3); +#if CONFIG_AV1_HIGHBITDEPTH typedef void (*cfl_subsample_hbd_fn)(const uint16_t *input, int input_stride, uint16_t *output_q3); +typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, + int dst_stride, int alpha_q3, int bd); +#endif + typedef void (*cfl_subtract_average_fn)(const uint16_t *src, int16_t *dst); typedef void (*cfl_predict_lbd_fn)(const int16_t *src, uint8_t *dst, int dst_stride, int alpha_q3); -typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, - int dst_stride, int alpha_q3, int bd); #ifdef __cplusplus extern "C" { #endif -void apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -RTCD_EXTERN void (*apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +RTCD_EXTERN void (*av1_apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); void av1_build_compound_diffwtd_mask_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); void av1_build_compound_diffwtd_mask_sse4_1(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); @@ -71,42 +96,89 @@ void av1_build_compound_diffwtd_mask_highbd_ssse3(uint8_t *mask, DIFFWTD_MASK_TY void av1_build_compound_diffwtd_mask_highbd_avx2(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); RTCD_EXTERN void (*av1_build_compound_diffwtd_mask_highbd)(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); -void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +int64_t av1_calc_frame_error_c(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_sse2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_avx2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +RTCD_EXTERN int64_t (*av1_calc_frame_error)(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); + +void av1_cnn_activate_c( float **input, int channels, int width, int height, int stride, ACTIVATION layer_activation); +#define av1_cnn_activate av1_cnn_activate_c -void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); +void av1_cnn_add_c( float **input, int channels, int width, int height, int stride, const float **add); +#define av1_cnn_add av1_cnn_add_c -void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_cnn_batchnorm_c(float **image, int channels, int width, int height, int stride, const float *gamma, const float *beta, const float *mean, const float *std); +#define av1_cnn_batchnorm av1_cnn_batchnorm_c + +void av1_cnn_convolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride, int start_idx, int step); +#define av1_cnn_convolve av1_cnn_convolve_c + +void av1_cnn_deconvolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride); +#define av1_cnn_deconvolve av1_cnn_deconvolve_c + +void av1_cnn_predict_c( const float **input, int in_width, int in_height, int in_stride, const CNN_CONFIG *cnn_config, const CNN_THREAD_DATA *thread_data, CNN_MULTI_OUT *output_struct); +#define av1_cnn_predict av1_cnn_predict_c + +void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_convolve_horiz_rs_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); void av1_convolve_horiz_rs_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); RTCD_EXTERN void (*av1_convolve_horiz_rs)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); -void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_dr_prediction_z1_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); -#define av1_dr_prediction_z1 av1_dr_prediction_z1_c +void av1_dr_prediction_z1_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z1)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); void av1_dr_prediction_z2_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z2 av1_dr_prediction_z2_c +void av1_dr_prediction_z2_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z2)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); void av1_dr_prediction_z3_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z3 av1_dr_prediction_z3_c +void av1_dr_prediction_z3_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z3)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); void av1_filter_intra_edge_c(uint8_t *p, int sz, int strength); void av1_filter_intra_edge_sse4_1(uint8_t *p, int sz, int strength); @@ -129,19 +201,19 @@ void av1_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint void av1_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve8_vert av1_highbd_convolve8_vert_c -void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve_avg av1_highbd_convolve_avg_c @@ -153,54 +225,76 @@ void av1_highbd_convolve_horiz_rs_c(const uint16_t *src, int src_stride, uint16_ void av1_highbd_convolve_horiz_rs_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); RTCD_EXTERN void (*av1_highbd_convolve_horiz_rs)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); -void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_dr_prediction_z1_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z1 av1_highbd_dr_prediction_z1_c +void av1_highbd_dr_prediction_z1_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z1)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); void av1_highbd_dr_prediction_z2_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z2 av1_highbd_dr_prediction_z2_c +void av1_highbd_dr_prediction_z2_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z2)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); void av1_highbd_dr_prediction_z3_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z3 av1_highbd_dr_prediction_z3_c +void av1_highbd_dr_prediction_z3_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z3)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -void av1_highbd_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_avx2(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x16)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_32x32)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_16_add av1_highbd_iwht4x4_16_add_c @@ -208,34 +302,14 @@ void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int des void av1_highbd_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_1_add av1_highbd_iwht4x4_1_add_c -void av1_highbd_jnt_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - void av1_highbd_warp_affine_c(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_highbd_warp_affine_sse4_1(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_highbd_warp_affine)(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); -void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); +void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); void av1_inv_txfm2d_add_16x16_c(const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type, int bd); #define av1_inv_txfm2d_add_16x16 av1_inv_txfm2d_add_16x16_c @@ -301,25 +375,9 @@ void av1_inv_txfm_add_ssse3(const tran_low_t *dqcoeff, uint8_t *dst, int stride, void av1_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); RTCD_EXTERN void (*av1_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_jnt_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_round_shift_array_c(int32_t *arr, int size, int bit); +void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); +RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); int av1_selfguided_restoration_c(const uint8_t *dgd8, int width, int height, int dgd_stride, int32_t *flt0, int32_t *flt1, int flt_stride, @@ -344,6 +402,7 @@ RTCD_EXTERN void (*av1_upsample_intra_edge_high)(uint16_t *p, int sz, int bd); void av1_warp_affine_c(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_warp_affine_sse4_1(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); +void av1_warp_affine_avx2(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_warp_affine)(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); @@ -351,12 +410,26 @@ void av1_wiener_convolve_add_src_sse2(const uint8_t *src, ptrdiff_t src_stride, void av1_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); RTCD_EXTERN void (*av1_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); -void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); +void cdef_copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); + +void cdef_copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); + +void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); int cdef_find_dir_c(const uint16_t *img, int stride, int32_t *var, int coeff_shift); int cdef_find_dir_sse2(const uint16_t *img, int stride, int32_t *var, int coeff_shift); @@ -395,34 +468,20 @@ cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_ssse3(TX_SIZE tx_size); cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_avx2(TX_SIZE tx_size); RTCD_EXTERN cfl_subsample_lbd_fn (*cfl_get_luma_subsampling_444_lbd)(TX_SIZE tx_size); -void copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); - -void copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); - -cfl_predict_hbd_fn get_predict_hbd_fn_c(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_hbd_fn (*get_predict_hbd_fn)(TX_SIZE tx_size); - -cfl_predict_lbd_fn get_predict_lbd_fn_c(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_lbd_fn (*get_predict_lbd_fn)(TX_SIZE tx_size); - -cfl_subtract_average_fn get_subtract_average_fn_c(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_sse2(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_subtract_average_fn (*get_subtract_average_fn)(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_c(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_hbd_fn (*cfl_get_predict_hbd_fn)(TX_SIZE tx_size); + +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_c(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_lbd_fn (*cfl_get_predict_lbd_fn)(TX_SIZE tx_size); + +cfl_subtract_average_fn cfl_get_subtract_average_fn_c(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_sse2(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_subtract_average_fn (*cfl_get_subtract_average_fn)(TX_SIZE tx_size); void av1_rtcd(void); @@ -434,9 +493,9 @@ static void setup_rtcd_internal(void) (void)flags; - apply_selfguided_restoration = apply_selfguided_restoration_c; - if (flags & HAS_SSE4_1) apply_selfguided_restoration = apply_selfguided_restoration_sse4_1; - if (flags & HAS_AVX2) apply_selfguided_restoration = apply_selfguided_restoration_avx2; + av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_c; + if (flags & HAS_SSE4_1) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_sse4_1; + if (flags & HAS_AVX2) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_avx2; av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_c; if (flags & HAS_SSE4_1) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_sse4_1; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_avx2; @@ -446,6 +505,9 @@ static void setup_rtcd_internal(void) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_c; if (flags & HAS_SSSE3) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_ssse3; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_avx2; + av1_calc_frame_error = av1_calc_frame_error_c; + if (flags & HAS_SSE2) av1_calc_frame_error = av1_calc_frame_error_sse2; + if (flags & HAS_AVX2) av1_calc_frame_error = av1_calc_frame_error_avx2; av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_c; if (flags & HAS_SSE2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_sse2; if (flags & HAS_AVX2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_avx2; @@ -462,6 +524,25 @@ static void setup_rtcd_internal(void) av1_convolve_y_sr = av1_convolve_y_sr_c; if (flags & HAS_SSE2) av1_convolve_y_sr = av1_convolve_y_sr_sse2; if (flags & HAS_AVX2) av1_convolve_y_sr = av1_convolve_y_sr_avx2; + av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_sse2; + if (flags & HAS_SSSE3) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_ssse3; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_avx2; + av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_avx2; + av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_avx2; + av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_avx2; + av1_dr_prediction_z1 = av1_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_dr_prediction_z1 = av1_dr_prediction_z1_avx2; + av1_dr_prediction_z2 = av1_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_dr_prediction_z2 = av1_dr_prediction_z2_avx2; + av1_dr_prediction_z3 = av1_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_dr_prediction_z3 = av1_dr_prediction_z3_avx2; av1_filter_intra_edge = av1_filter_intra_edge_c; if (flags & HAS_SSE4_1) av1_filter_intra_edge = av1_filter_intra_edge_sse4_1; av1_filter_intra_edge_high = av1_filter_intra_edge_high_c; @@ -484,34 +565,39 @@ static void setup_rtcd_internal(void) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_c; if (flags & HAS_SSSE3) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_ssse3; if (flags & HAS_AVX2) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_avx2; + av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_avx2; + av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_avx2; + av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_avx2; + av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_avx2; + av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_avx2; + av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_avx2; + av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_avx2; av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_sse4_1; if (flags & HAS_AVX2) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_avx2; - av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_sse4_1; - av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_sse4_1; - av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_sse4_1; - if (flags & HAS_AVX2) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_avx2; + av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_sse4_1; + av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_sse4_1; av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_sse4_1; - av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_sse4_1; + av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_sse4_1; + av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_sse4_1; av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_sse4_1; - av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_avx2; - av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_avx2; - av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_avx2; - av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_avx2; av1_highbd_warp_affine = av1_highbd_warp_affine_c; if (flags & HAS_SSE4_1) av1_highbd_warp_affine = av1_highbd_warp_affine_sse4_1; av1_highbd_wiener_convolve_add_src = av1_highbd_wiener_convolve_add_src_c; @@ -524,18 +610,8 @@ static void setup_rtcd_internal(void) av1_inv_txfm_add = av1_inv_txfm_add_c; if (flags & HAS_SSSE3) av1_inv_txfm_add = av1_inv_txfm_add_ssse3; if (flags & HAS_AVX2) av1_inv_txfm_add = av1_inv_txfm_add_avx2; - av1_jnt_convolve_2d = av1_jnt_convolve_2d_c; - if (flags & HAS_SSSE3) av1_jnt_convolve_2d = av1_jnt_convolve_2d_ssse3; - if (flags & HAS_AVX2) av1_jnt_convolve_2d = av1_jnt_convolve_2d_avx2; - av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_avx2; - av1_jnt_convolve_x = av1_jnt_convolve_x_c; - if (flags & HAS_SSE2) av1_jnt_convolve_x = av1_jnt_convolve_x_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_x = av1_jnt_convolve_x_avx2; - av1_jnt_convolve_y = av1_jnt_convolve_y_c; - if (flags & HAS_SSE2) av1_jnt_convolve_y = av1_jnt_convolve_y_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_y = av1_jnt_convolve_y_avx2; + av1_round_shift_array = av1_round_shift_array_c; + if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; av1_selfguided_restoration = av1_selfguided_restoration_c; if (flags & HAS_SSE4_1) av1_selfguided_restoration = av1_selfguided_restoration_sse4_1; if (flags & HAS_AVX2) av1_selfguided_restoration = av1_selfguided_restoration_avx2; @@ -545,9 +621,20 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE4_1) av1_upsample_intra_edge_high = av1_upsample_intra_edge_high_sse4_1; av1_warp_affine = av1_warp_affine_c; if (flags & HAS_SSE4_1) av1_warp_affine = av1_warp_affine_sse4_1; + if (flags & HAS_AVX2) av1_warp_affine = av1_warp_affine_avx2; av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_c; if (flags & HAS_SSE2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_sse2; if (flags & HAS_AVX2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_avx2; + cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_c; + if (flags & HAS_SSE2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_avx2; + cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_c; + if (flags & HAS_SSE2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_avx2; cdef_filter_block = cdef_filter_block_c; if (flags & HAS_SSE2) cdef_filter_block = cdef_filter_block_sse2; if (flags & HAS_SSSE3) cdef_filter_block = cdef_filter_block_ssse3; @@ -576,25 +663,15 @@ static void setup_rtcd_internal(void) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_c; if (flags & HAS_SSSE3) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_ssse3; if (flags & HAS_AVX2) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_avx2; - copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_c; - if (flags & HAS_SSE2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_avx2; - copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_c; - if (flags & HAS_SSE2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_avx2; - get_predict_hbd_fn = get_predict_hbd_fn_c; - if (flags & HAS_SSSE3) get_predict_hbd_fn = get_predict_hbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_hbd_fn = get_predict_hbd_fn_avx2; - get_predict_lbd_fn = get_predict_lbd_fn_c; - if (flags & HAS_SSSE3) get_predict_lbd_fn = get_predict_lbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_lbd_fn = get_predict_lbd_fn_avx2; - get_subtract_average_fn = get_subtract_average_fn_c; - if (flags & HAS_SSE2) get_subtract_average_fn = get_subtract_average_fn_sse2; - if (flags & HAS_AVX2) get_subtract_average_fn = get_subtract_average_fn_avx2; + cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_avx2; + cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_avx2; + cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_c; + if (flags & HAS_SSE2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_sse2; + if (flags & HAS_AVX2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_avx2; } #endif diff --git a/media/libaom/config/win/mingw32/config/aom_config.asm b/media/libaom/config/win/mingw32/config/aom_config.asm index 6ae776c7c4..baf05cb910 100644 --- a/media/libaom/config/win/mingw32/config/aom_config.asm +++ b/media/libaom/config/win/mingw32/config/aom_config.asm @@ -1,5 +1,5 @@ ; -; Copyright (c) 2018, Alliance for Open Media. All rights reserved +; Copyright (c) 2021, 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 @@ -14,49 +14,52 @@ ARCH_MIPS equ 0 ARCH_PPC equ 0 ARCH_X86 equ 1 ARCH_X86_64 equ 0 -CONFIG_2PASS_PARTITION_SEARCH_LVL equ 1 CONFIG_ACCOUNTING equ 0 CONFIG_ANALYZER equ 0 CONFIG_AV1_DECODER equ 1 CONFIG_AV1_ENCODER equ 0 +CONFIG_AV1_HIGHBITDEPTH equ 1 CONFIG_BIG_ENDIAN equ 0 CONFIG_BITSTREAM_DEBUG equ 0 CONFIG_COEFFICIENT_RANGE_CHECKING equ 0 -CONFIG_COLLECT_INTER_MODE_RD_STATS equ 0 +CONFIG_COLLECT_COMPONENT_TIMING equ 0 +CONFIG_COLLECT_PARTITION_STATS equ 0 CONFIG_COLLECT_RD_STATS equ 0 CONFIG_DEBUG equ 0 CONFIG_DENOISE equ 1 +CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 equ 1 CONFIG_DIST_8X8 equ 0 CONFIG_ENTROPY_STATS equ 0 -CONFIG_FILEOPTIONS equ 1 -CONFIG_FIX_GF_LENGTH equ 1 -CONFIG_FP_MB_STATS equ 0 CONFIG_GCC equ 1 CONFIG_GCOV equ 0 -CONFIG_GLOBAL_MOTION_SEARCH equ 1 CONFIG_GPROF equ 0 +CONFIG_HTB_TRELLIS equ 0 CONFIG_INSPECTION equ 0 CONFIG_INTERNAL_STATS equ 0 CONFIG_INTER_STATS_ONLY equ 0 CONFIG_LIBYUV equ 0 -CONFIG_LOWBITDEPTH equ 1 +CONFIG_LPF_MASK equ 0 CONFIG_MAX_DECODE_PROFILE equ 2 CONFIG_MISMATCH_DEBUG equ 0 CONFIG_MULTITHREAD equ 1 +CONFIG_NN_V2 equ 0 CONFIG_NORMAL_TILE_MODE equ 0 CONFIG_OS_SUPPORT equ 1 CONFIG_PIC equ 0 CONFIG_RD_DEBUG equ 0 -CONFIG_REDUCED_ENCODER_BORDER equ 0 +CONFIG_REALTIME_ONLY equ 0 CONFIG_RUNTIME_CPU_DETECT equ 1 CONFIG_SHARED equ 0 CONFIG_SHARP_SETTINGS equ 0 CONFIG_SIZE_LIMIT equ 0 CONFIG_SPATIAL_RESAMPLING equ 1 -CONFIG_STATIC equ 1 +CONFIG_SPEED_STATS equ 0 +CONFIG_SUPERRES_IN_RECODE equ 1 +CONFIG_TUNE_VMAF equ 0 CONFIG_WEBM_IO equ 0 DECODE_HEIGHT_LIMIT equ 0 DECODE_WIDTH_LIMIT equ 0 +FORCE_HIGHBITDEPTH_DECODING equ 0 HAVE_AVX equ 1 HAVE_AVX2 equ 1 HAVE_DSPR2 equ 0 diff --git a/media/libaom/config/win/mingw32/config/aom_config.h b/media/libaom/config/win/mingw32/config/aom_config.h index 090a455778..fc1271c743 100644 --- a/media/libaom/config/win/mingw32/config/aom_config.h +++ b/media/libaom/config/win/mingw32/config/aom_config.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018, Alliance for Open Media. All rights reserved + * Copyright (c) 2021, 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 @@ -16,49 +16,52 @@ #define ARCH_PPC 0 #define ARCH_X86 1 #define ARCH_X86_64 0 -#define CONFIG_2PASS_PARTITION_SEARCH_LVL 1 #define CONFIG_ACCOUNTING 0 #define CONFIG_ANALYZER 0 #define CONFIG_AV1_DECODER 1 #define CONFIG_AV1_ENCODER 0 +#define CONFIG_AV1_HIGHBITDEPTH 1 #define CONFIG_BIG_ENDIAN 0 #define CONFIG_BITSTREAM_DEBUG 0 #define CONFIG_COEFFICIENT_RANGE_CHECKING 0 -#define CONFIG_COLLECT_INTER_MODE_RD_STATS 0 +#define CONFIG_COLLECT_COMPONENT_TIMING 0 +#define CONFIG_COLLECT_PARTITION_STATS 0 #define CONFIG_COLLECT_RD_STATS 0 #define CONFIG_DEBUG 0 #define CONFIG_DENOISE 1 +#define CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 1 #define CONFIG_DIST_8X8 0 #define CONFIG_ENTROPY_STATS 0 -#define CONFIG_FILEOPTIONS 1 -#define CONFIG_FIX_GF_LENGTH 1 -#define CONFIG_FP_MB_STATS 0 #define CONFIG_GCC 1 #define CONFIG_GCOV 0 -#define CONFIG_GLOBAL_MOTION_SEARCH 1 #define CONFIG_GPROF 0 +#define CONFIG_HTB_TRELLIS 0 #define CONFIG_INSPECTION 0 #define CONFIG_INTERNAL_STATS 0 #define CONFIG_INTER_STATS_ONLY 0 #define CONFIG_LIBYUV 0 -#define CONFIG_LOWBITDEPTH 1 +#define CONFIG_LPF_MASK 0 #define CONFIG_MAX_DECODE_PROFILE 2 #define CONFIG_MISMATCH_DEBUG 0 #define CONFIG_MULTITHREAD 1 +#define CONFIG_NN_V2 0 #define CONFIG_NORMAL_TILE_MODE 0 #define CONFIG_OS_SUPPORT 1 #define CONFIG_PIC 0 #define CONFIG_RD_DEBUG 0 -#define CONFIG_REDUCED_ENCODER_BORDER 0 +#define CONFIG_REALTIME_ONLY 0 #define CONFIG_RUNTIME_CPU_DETECT 1 #define CONFIG_SHARED 0 #define CONFIG_SHARP_SETTINGS 0 #define CONFIG_SIZE_LIMIT 0 #define CONFIG_SPATIAL_RESAMPLING 1 -#define CONFIG_STATIC 1 +#define CONFIG_SPEED_STATS 0 +#define CONFIG_SUPERRES_IN_RECODE 1 +#define CONFIG_TUNE_VMAF 0 #define CONFIG_WEBM_IO 0 #define DECODE_HEIGHT_LIMIT 0 #define DECODE_WIDTH_LIMIT 0 +#define FORCE_HIGHBITDEPTH_DECODING 0 #define HAVE_AVX 1 #define HAVE_AVX2 1 #define HAVE_DSPR2 0 @@ -76,7 +79,5 @@ #define HAVE_SSSE3 1 #define HAVE_VSX 0 #define HAVE_WXWIDGETS 0 -#define INCLUDE_INSTALL_DIR INSTALLDIR/include #define INLINE inline -#define LIB_INSTALL_DIR INSTALLDIR/lib #endif /* AOM_CONFIG_H_ */ diff --git a/media/libaom/config/win/mingw32/config/aom_dsp_rtcd.h b/media/libaom/config/win/mingw32/config/aom_dsp_rtcd.h index 5b3762f109..9018213bd6 100644 --- a/media/libaom/config/win/mingw32/config/aom_dsp_rtcd.h +++ b/media/libaom/config/win/mingw32/config/aom_dsp_rtcd.h @@ -26,10 +26,10 @@ void aom_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src void aom_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); RTCD_EXTERN void (*aom_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); -void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); +void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); void aom_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); void aom_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); @@ -47,9 +47,9 @@ void aom_convolve8_vert_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t void aom_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*aom_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -RTCD_EXTERN void (*aom_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); +void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +RTCD_EXTERN void (*aom_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); void aom_dc_128_predictor_16x16_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); void aom_dc_128_predictor_16x16_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); @@ -471,33 +471,37 @@ void aom_h_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_h_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); RTCD_EXTERN void (*aom_h_predictor_8x8)(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); -void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params, const int bd); -#define aom_highbd_blend_a64_d16_mask aom_highbd_blend_a64_d16_mask_c +void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); void aom_highbd_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); +void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); void aom_highbd_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_vmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); void aom_highbd_dc_128_predictor_16x16_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); void aom_highbd_dc_128_predictor_16x16_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); @@ -853,10 +857,10 @@ void aom_highbd_lpf_horizontal_14_c(uint16_t *s, int pitch, const uint8_t *blimi void aom_highbd_lpf_horizontal_14_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14)(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); -void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); void aom_highbd_lpf_horizontal_4_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); void aom_highbd_lpf_horizontal_4_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); @@ -1229,10 +1233,10 @@ void aom_highbd_v_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_v_predictor_8x8_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); RTCD_EXTERN void (*aom_highbd_v_predictor_8x8)(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); -void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); void aom_lpf_horizontal_14_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void aom_lpf_horizontal_14_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -1709,10 +1713,6 @@ void aom_v_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_v_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); RTCD_EXTERN void (*aom_v_predictor_8x8)(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); -void av1_round_shift_array_c(int32_t *arr, int size, int bit); -void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); -RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); - void aom_dsp_rtcd(void); #ifdef RTCD_C @@ -1955,6 +1955,9 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE2) aom_h_predictor_8x4 = aom_h_predictor_8x4_sse2; aom_h_predictor_8x8 = aom_h_predictor_8x8_c; if (flags & HAS_SSE2) aom_h_predictor_8x8 = aom_h_predictor_8x8_sse2; + aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_c; + if (flags & HAS_SSE4_1) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_sse4_1; + if (flags & HAS_AVX2) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_avx2; aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_sse4_1; aom_highbd_blend_a64_mask = aom_highbd_blend_a64_mask_c; @@ -1962,8 +1965,10 @@ static void setup_rtcd_internal(void) aom_highbd_blend_a64_vmask = aom_highbd_blend_a64_vmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_vmask = aom_highbd_blend_a64_vmask_sse4_1; aom_highbd_convolve8_horiz = aom_highbd_convolve8_horiz_c; + if (flags & HAS_SSE2) aom_highbd_convolve8_horiz = aom_highbd_convolve8_horiz_sse2; if (flags & HAS_AVX2) aom_highbd_convolve8_horiz = aom_highbd_convolve8_horiz_avx2; aom_highbd_convolve8_vert = aom_highbd_convolve8_vert_c; + if (flags & HAS_SSE2) aom_highbd_convolve8_vert = aom_highbd_convolve8_vert_sse2; if (flags & HAS_AVX2) aom_highbd_convolve8_vert = aom_highbd_convolve8_vert_avx2; aom_highbd_convolve_copy = aom_highbd_convolve_copy_c; if (flags & HAS_SSE2) aom_highbd_convolve_copy = aom_highbd_convolve_copy_sse2; @@ -2367,8 +2372,6 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE2) aom_v_predictor_8x4 = aom_v_predictor_8x4_sse2; aom_v_predictor_8x8 = aom_v_predictor_8x8_c; if (flags & HAS_SSE2) aom_v_predictor_8x8 = aom_v_predictor_8x8_sse2; - av1_round_shift_array = av1_round_shift_array_c; - if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; } #endif diff --git a/media/libaom/config/win/mingw32/config/aom_scale_rtcd.h b/media/libaom/config/win/mingw32/config/aom_scale_rtcd.h index 6b56795cd4..65c184b83d 100644 --- a/media/libaom/config/win/mingw32/config/aom_scale_rtcd.h +++ b/media/libaom/config/win/mingw32/config/aom_scale_rtcd.h @@ -59,15 +59,27 @@ void aom_yv12_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buf void aom_yv12_extend_frame_borders_c(struct yv12_buffer_config *ybf, const int num_planes); #define aom_yv12_extend_frame_borders aom_yv12_extend_frame_borders_c -void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_coloc_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_u aom_yv12_partial_coloc_copy_u_c + +void aom_yv12_partial_coloc_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_v aom_yv12_partial_coloc_copy_v_c + +void aom_yv12_partial_coloc_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_y aom_yv12_partial_coloc_copy_y_c + +void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_u aom_yv12_partial_copy_u_c -void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_v aom_yv12_partial_copy_v_c -void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_ybc, int hstart2, int vstart2); #define aom_yv12_partial_copy_y aom_yv12_partial_copy_y_c +int aom_yv12_realloc_with_new_border_c(struct yv12_buffer_config *ybf, int new_border, int byte_alignment, int num_planes); +#define aom_yv12_realloc_with_new_border aom_yv12_realloc_with_new_border_c + void aom_scale_rtcd(void); #ifdef RTCD_C diff --git a/media/libaom/config/win/mingw32/config/av1_rtcd.h b/media/libaom/config/win/mingw32/config/av1_rtcd.h index a2a6f14ff5..7ecd05f380 100644 --- a/media/libaom/config/win/mingw32/config/av1_rtcd.h +++ b/media/libaom/config/win/mingw32/config/av1_rtcd.h @@ -31,30 +31,55 @@ struct txfm_param; struct aom_variance_vtable; struct search_site_config; struct yv12_buffer_config; +struct NN_CONFIG; +typedef struct NN_CONFIG NN_CONFIG; + +enum { NONE, RELU, SOFTSIGN, SIGMOID } UENUM1BYTE(ACTIVATION); +#if CONFIG_NN_V2 +enum { SOFTMAX_CROSS_ENTROPY } UENUM1BYTE(LOSS); +struct NN_CONFIG_V2; +typedef struct NN_CONFIG_V2 NN_CONFIG_V2; +struct FC_LAYER; +typedef struct FC_LAYER FC_LAYER; +#endif // CONFIG_NN_V2 + +struct CNN_CONFIG; +typedef struct CNN_CONFIG CNN_CONFIG; +struct CNN_LAYER_CONFIG; +typedef struct CNN_LAYER_CONFIG CNN_LAYER_CONFIG; +struct CNN_THREAD_DATA; +typedef struct CNN_THREAD_DATA CNN_THREAD_DATA; +struct CNN_BRANCH_CONFIG; +typedef struct CNN_BRANCH_CONFIG CNN_BRANCH_CONFIG; +struct CNN_MULTI_OUT; +typedef struct CNN_MULTI_OUT CNN_MULTI_OUT; /* Function pointers return by CfL functions */ typedef void (*cfl_subsample_lbd_fn)(const uint8_t *input, int input_stride, uint16_t *output_q3); +#if CONFIG_AV1_HIGHBITDEPTH typedef void (*cfl_subsample_hbd_fn)(const uint16_t *input, int input_stride, uint16_t *output_q3); +typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, + int dst_stride, int alpha_q3, int bd); +#endif + typedef void (*cfl_subtract_average_fn)(const uint16_t *src, int16_t *dst); typedef void (*cfl_predict_lbd_fn)(const int16_t *src, uint8_t *dst, int dst_stride, int alpha_q3); -typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, - int dst_stride, int alpha_q3, int bd); #ifdef __cplusplus extern "C" { #endif -void apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -RTCD_EXTERN void (*apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +RTCD_EXTERN void (*av1_apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); void av1_build_compound_diffwtd_mask_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); void av1_build_compound_diffwtd_mask_sse4_1(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); @@ -71,42 +96,89 @@ void av1_build_compound_diffwtd_mask_highbd_ssse3(uint8_t *mask, DIFFWTD_MASK_TY void av1_build_compound_diffwtd_mask_highbd_avx2(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); RTCD_EXTERN void (*av1_build_compound_diffwtd_mask_highbd)(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); -void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +int64_t av1_calc_frame_error_c(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_sse2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_avx2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +RTCD_EXTERN int64_t (*av1_calc_frame_error)(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); + +void av1_cnn_activate_c( float **input, int channels, int width, int height, int stride, ACTIVATION layer_activation); +#define av1_cnn_activate av1_cnn_activate_c -void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); +void av1_cnn_add_c( float **input, int channels, int width, int height, int stride, const float **add); +#define av1_cnn_add av1_cnn_add_c -void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_cnn_batchnorm_c(float **image, int channels, int width, int height, int stride, const float *gamma, const float *beta, const float *mean, const float *std); +#define av1_cnn_batchnorm av1_cnn_batchnorm_c + +void av1_cnn_convolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride, int start_idx, int step); +#define av1_cnn_convolve av1_cnn_convolve_c + +void av1_cnn_deconvolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride); +#define av1_cnn_deconvolve av1_cnn_deconvolve_c + +void av1_cnn_predict_c( const float **input, int in_width, int in_height, int in_stride, const CNN_CONFIG *cnn_config, const CNN_THREAD_DATA *thread_data, CNN_MULTI_OUT *output_struct); +#define av1_cnn_predict av1_cnn_predict_c + +void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_convolve_horiz_rs_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); void av1_convolve_horiz_rs_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); RTCD_EXTERN void (*av1_convolve_horiz_rs)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); -void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_dr_prediction_z1_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); -#define av1_dr_prediction_z1 av1_dr_prediction_z1_c +void av1_dr_prediction_z1_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z1)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); void av1_dr_prediction_z2_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z2 av1_dr_prediction_z2_c +void av1_dr_prediction_z2_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z2)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); void av1_dr_prediction_z3_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z3 av1_dr_prediction_z3_c +void av1_dr_prediction_z3_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z3)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); void av1_filter_intra_edge_c(uint8_t *p, int sz, int strength); void av1_filter_intra_edge_sse4_1(uint8_t *p, int sz, int strength); @@ -129,19 +201,19 @@ void av1_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint void av1_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve8_vert av1_highbd_convolve8_vert_c -void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve_avg av1_highbd_convolve_avg_c @@ -153,54 +225,76 @@ void av1_highbd_convolve_horiz_rs_c(const uint16_t *src, int src_stride, uint16_ void av1_highbd_convolve_horiz_rs_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); RTCD_EXTERN void (*av1_highbd_convolve_horiz_rs)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); -void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_dr_prediction_z1_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z1 av1_highbd_dr_prediction_z1_c +void av1_highbd_dr_prediction_z1_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z1)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); void av1_highbd_dr_prediction_z2_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z2 av1_highbd_dr_prediction_z2_c +void av1_highbd_dr_prediction_z2_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z2)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); void av1_highbd_dr_prediction_z3_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z3 av1_highbd_dr_prediction_z3_c +void av1_highbd_dr_prediction_z3_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z3)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -void av1_highbd_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_avx2(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x16)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_32x32)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_16_add av1_highbd_iwht4x4_16_add_c @@ -208,34 +302,14 @@ void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int des void av1_highbd_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_1_add av1_highbd_iwht4x4_1_add_c -void av1_highbd_jnt_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - void av1_highbd_warp_affine_c(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_highbd_warp_affine_sse4_1(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_highbd_warp_affine)(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); -void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); +void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); void av1_inv_txfm2d_add_16x16_c(const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type, int bd); #define av1_inv_txfm2d_add_16x16 av1_inv_txfm2d_add_16x16_c @@ -301,25 +375,9 @@ void av1_inv_txfm_add_ssse3(const tran_low_t *dqcoeff, uint8_t *dst, int stride, void av1_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); RTCD_EXTERN void (*av1_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_jnt_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_round_shift_array_c(int32_t *arr, int size, int bit); +void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); +RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); int av1_selfguided_restoration_c(const uint8_t *dgd8, int width, int height, int dgd_stride, int32_t *flt0, int32_t *flt1, int flt_stride, @@ -344,6 +402,7 @@ RTCD_EXTERN void (*av1_upsample_intra_edge_high)(uint16_t *p, int sz, int bd); void av1_warp_affine_c(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_warp_affine_sse4_1(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); +void av1_warp_affine_avx2(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_warp_affine)(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); @@ -351,12 +410,26 @@ void av1_wiener_convolve_add_src_sse2(const uint8_t *src, ptrdiff_t src_stride, void av1_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); RTCD_EXTERN void (*av1_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); -void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); +void cdef_copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); + +void cdef_copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); + +void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); int cdef_find_dir_c(const uint16_t *img, int stride, int32_t *var, int coeff_shift); int cdef_find_dir_sse2(const uint16_t *img, int stride, int32_t *var, int coeff_shift); @@ -395,34 +468,20 @@ cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_ssse3(TX_SIZE tx_size); cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_avx2(TX_SIZE tx_size); RTCD_EXTERN cfl_subsample_lbd_fn (*cfl_get_luma_subsampling_444_lbd)(TX_SIZE tx_size); -void copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); - -void copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); - -cfl_predict_hbd_fn get_predict_hbd_fn_c(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_hbd_fn (*get_predict_hbd_fn)(TX_SIZE tx_size); - -cfl_predict_lbd_fn get_predict_lbd_fn_c(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_lbd_fn (*get_predict_lbd_fn)(TX_SIZE tx_size); - -cfl_subtract_average_fn get_subtract_average_fn_c(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_sse2(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_subtract_average_fn (*get_subtract_average_fn)(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_c(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_hbd_fn (*cfl_get_predict_hbd_fn)(TX_SIZE tx_size); + +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_c(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_lbd_fn (*cfl_get_predict_lbd_fn)(TX_SIZE tx_size); + +cfl_subtract_average_fn cfl_get_subtract_average_fn_c(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_sse2(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_subtract_average_fn (*cfl_get_subtract_average_fn)(TX_SIZE tx_size); void av1_rtcd(void); @@ -434,9 +493,9 @@ static void setup_rtcd_internal(void) (void)flags; - apply_selfguided_restoration = apply_selfguided_restoration_c; - if (flags & HAS_SSE4_1) apply_selfguided_restoration = apply_selfguided_restoration_sse4_1; - if (flags & HAS_AVX2) apply_selfguided_restoration = apply_selfguided_restoration_avx2; + av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_c; + if (flags & HAS_SSE4_1) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_sse4_1; + if (flags & HAS_AVX2) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_avx2; av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_c; if (flags & HAS_SSE4_1) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_sse4_1; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_avx2; @@ -446,6 +505,9 @@ static void setup_rtcd_internal(void) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_c; if (flags & HAS_SSSE3) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_ssse3; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_avx2; + av1_calc_frame_error = av1_calc_frame_error_c; + if (flags & HAS_SSE2) av1_calc_frame_error = av1_calc_frame_error_sse2; + if (flags & HAS_AVX2) av1_calc_frame_error = av1_calc_frame_error_avx2; av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_c; if (flags & HAS_SSE2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_sse2; if (flags & HAS_AVX2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_avx2; @@ -462,6 +524,25 @@ static void setup_rtcd_internal(void) av1_convolve_y_sr = av1_convolve_y_sr_c; if (flags & HAS_SSE2) av1_convolve_y_sr = av1_convolve_y_sr_sse2; if (flags & HAS_AVX2) av1_convolve_y_sr = av1_convolve_y_sr_avx2; + av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_sse2; + if (flags & HAS_SSSE3) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_ssse3; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_avx2; + av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_avx2; + av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_avx2; + av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_c; + if (flags & HAS_SSE2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_avx2; + av1_dr_prediction_z1 = av1_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_dr_prediction_z1 = av1_dr_prediction_z1_avx2; + av1_dr_prediction_z2 = av1_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_dr_prediction_z2 = av1_dr_prediction_z2_avx2; + av1_dr_prediction_z3 = av1_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_dr_prediction_z3 = av1_dr_prediction_z3_avx2; av1_filter_intra_edge = av1_filter_intra_edge_c; if (flags & HAS_SSE4_1) av1_filter_intra_edge = av1_filter_intra_edge_sse4_1; av1_filter_intra_edge_high = av1_filter_intra_edge_high_c; @@ -484,34 +565,39 @@ static void setup_rtcd_internal(void) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_c; if (flags & HAS_SSSE3) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_ssse3; if (flags & HAS_AVX2) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_avx2; + av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_avx2; + av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_avx2; + av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_avx2; + av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_avx2; + av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_avx2; + av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_avx2; + av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_avx2; av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_sse4_1; if (flags & HAS_AVX2) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_avx2; - av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_sse4_1; - av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_sse4_1; - av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_sse4_1; - if (flags & HAS_AVX2) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_avx2; + av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_sse4_1; + av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_sse4_1; av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_sse4_1; - av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_sse4_1; + av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_sse4_1; + av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_sse4_1; av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_sse4_1; - av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_avx2; - av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_avx2; - av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_avx2; - av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_avx2; av1_highbd_warp_affine = av1_highbd_warp_affine_c; if (flags & HAS_SSE4_1) av1_highbd_warp_affine = av1_highbd_warp_affine_sse4_1; av1_highbd_wiener_convolve_add_src = av1_highbd_wiener_convolve_add_src_c; @@ -524,18 +610,8 @@ static void setup_rtcd_internal(void) av1_inv_txfm_add = av1_inv_txfm_add_c; if (flags & HAS_SSSE3) av1_inv_txfm_add = av1_inv_txfm_add_ssse3; if (flags & HAS_AVX2) av1_inv_txfm_add = av1_inv_txfm_add_avx2; - av1_jnt_convolve_2d = av1_jnt_convolve_2d_c; - if (flags & HAS_SSSE3) av1_jnt_convolve_2d = av1_jnt_convolve_2d_ssse3; - if (flags & HAS_AVX2) av1_jnt_convolve_2d = av1_jnt_convolve_2d_avx2; - av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_avx2; - av1_jnt_convolve_x = av1_jnt_convolve_x_c; - if (flags & HAS_SSE2) av1_jnt_convolve_x = av1_jnt_convolve_x_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_x = av1_jnt_convolve_x_avx2; - av1_jnt_convolve_y = av1_jnt_convolve_y_c; - if (flags & HAS_SSE2) av1_jnt_convolve_y = av1_jnt_convolve_y_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_y = av1_jnt_convolve_y_avx2; + av1_round_shift_array = av1_round_shift_array_c; + if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; av1_selfguided_restoration = av1_selfguided_restoration_c; if (flags & HAS_SSE4_1) av1_selfguided_restoration = av1_selfguided_restoration_sse4_1; if (flags & HAS_AVX2) av1_selfguided_restoration = av1_selfguided_restoration_avx2; @@ -545,9 +621,20 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE4_1) av1_upsample_intra_edge_high = av1_upsample_intra_edge_high_sse4_1; av1_warp_affine = av1_warp_affine_c; if (flags & HAS_SSE4_1) av1_warp_affine = av1_warp_affine_sse4_1; + if (flags & HAS_AVX2) av1_warp_affine = av1_warp_affine_avx2; av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_c; if (flags & HAS_SSE2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_sse2; if (flags & HAS_AVX2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_avx2; + cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_c; + if (flags & HAS_SSE2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_avx2; + cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_c; + if (flags & HAS_SSE2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_avx2; cdef_filter_block = cdef_filter_block_c; if (flags & HAS_SSE2) cdef_filter_block = cdef_filter_block_sse2; if (flags & HAS_SSSE3) cdef_filter_block = cdef_filter_block_ssse3; @@ -576,25 +663,15 @@ static void setup_rtcd_internal(void) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_c; if (flags & HAS_SSSE3) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_ssse3; if (flags & HAS_AVX2) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_avx2; - copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_c; - if (flags & HAS_SSE2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_avx2; - copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_c; - if (flags & HAS_SSE2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_avx2; - get_predict_hbd_fn = get_predict_hbd_fn_c; - if (flags & HAS_SSSE3) get_predict_hbd_fn = get_predict_hbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_hbd_fn = get_predict_hbd_fn_avx2; - get_predict_lbd_fn = get_predict_lbd_fn_c; - if (flags & HAS_SSSE3) get_predict_lbd_fn = get_predict_lbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_lbd_fn = get_predict_lbd_fn_avx2; - get_subtract_average_fn = get_subtract_average_fn_c; - if (flags & HAS_SSE2) get_subtract_average_fn = get_subtract_average_fn_sse2; - if (flags & HAS_AVX2) get_subtract_average_fn = get_subtract_average_fn_avx2; + cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_avx2; + cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_avx2; + cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_c; + if (flags & HAS_SSE2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_sse2; + if (flags & HAS_AVX2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_avx2; } #endif diff --git a/media/libaom/config/win/mingw64/config/aom_config.asm b/media/libaom/config/win/mingw64/config/aom_config.asm index 4ab2dedb42..9ce9b2090e 100644 --- a/media/libaom/config/win/mingw64/config/aom_config.asm +++ b/media/libaom/config/win/mingw64/config/aom_config.asm @@ -1,5 +1,5 @@ ; -; Copyright (c) 2018, Alliance for Open Media. All rights reserved +; Copyright (c) 2021, 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 @@ -14,49 +14,52 @@ ARCH_MIPS equ 0 ARCH_PPC equ 0 ARCH_X86 equ 0 ARCH_X86_64 equ 1 -CONFIG_2PASS_PARTITION_SEARCH_LVL equ 1 CONFIG_ACCOUNTING equ 0 CONFIG_ANALYZER equ 0 CONFIG_AV1_DECODER equ 1 CONFIG_AV1_ENCODER equ 0 +CONFIG_AV1_HIGHBITDEPTH equ 1 CONFIG_BIG_ENDIAN equ 0 CONFIG_BITSTREAM_DEBUG equ 0 CONFIG_COEFFICIENT_RANGE_CHECKING equ 0 -CONFIG_COLLECT_INTER_MODE_RD_STATS equ 0 +CONFIG_COLLECT_COMPONENT_TIMING equ 0 +CONFIG_COLLECT_PARTITION_STATS equ 0 CONFIG_COLLECT_RD_STATS equ 0 CONFIG_DEBUG equ 0 CONFIG_DENOISE equ 1 +CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 equ 1 CONFIG_DIST_8X8 equ 0 CONFIG_ENTROPY_STATS equ 0 -CONFIG_FILEOPTIONS equ 1 -CONFIG_FIX_GF_LENGTH equ 1 -CONFIG_FP_MB_STATS equ 0 CONFIG_GCC equ 1 CONFIG_GCOV equ 0 -CONFIG_GLOBAL_MOTION_SEARCH equ 1 CONFIG_GPROF equ 0 +CONFIG_HTB_TRELLIS equ 0 CONFIG_INSPECTION equ 0 CONFIG_INTERNAL_STATS equ 0 CONFIG_INTER_STATS_ONLY equ 0 CONFIG_LIBYUV equ 0 -CONFIG_LOWBITDEPTH equ 1 +CONFIG_LPF_MASK equ 0 CONFIG_MAX_DECODE_PROFILE equ 2 CONFIG_MISMATCH_DEBUG equ 0 CONFIG_MULTITHREAD equ 1 +CONFIG_NN_V2 equ 0 CONFIG_NORMAL_TILE_MODE equ 0 CONFIG_OS_SUPPORT equ 1 CONFIG_PIC equ 0 CONFIG_RD_DEBUG equ 0 -CONFIG_REDUCED_ENCODER_BORDER equ 0 +CONFIG_REALTIME_ONLY equ 0 CONFIG_RUNTIME_CPU_DETECT equ 1 CONFIG_SHARED equ 0 CONFIG_SHARP_SETTINGS equ 0 CONFIG_SIZE_LIMIT equ 0 CONFIG_SPATIAL_RESAMPLING equ 1 -CONFIG_STATIC equ 1 +CONFIG_SPEED_STATS equ 0 +CONFIG_SUPERRES_IN_RECODE equ 1 +CONFIG_TUNE_VMAF equ 0 CONFIG_WEBM_IO equ 0 DECODE_HEIGHT_LIMIT equ 0 DECODE_WIDTH_LIMIT equ 0 +FORCE_HIGHBITDEPTH_DECODING equ 0 HAVE_AVX equ 1 HAVE_AVX2 equ 1 HAVE_DSPR2 equ 0 diff --git a/media/libaom/config/win/mingw64/config/aom_config.h b/media/libaom/config/win/mingw64/config/aom_config.h index 9a0be7cd11..12997ea4dd 100644 --- a/media/libaom/config/win/mingw64/config/aom_config.h +++ b/media/libaom/config/win/mingw64/config/aom_config.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018, Alliance for Open Media. All rights reserved + * Copyright (c) 2021, 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 @@ -16,49 +16,52 @@ #define ARCH_PPC 0 #define ARCH_X86 0 #define ARCH_X86_64 1 -#define CONFIG_2PASS_PARTITION_SEARCH_LVL 1 #define CONFIG_ACCOUNTING 0 #define CONFIG_ANALYZER 0 #define CONFIG_AV1_DECODER 1 #define CONFIG_AV1_ENCODER 0 +#define CONFIG_AV1_HIGHBITDEPTH 1 #define CONFIG_BIG_ENDIAN 0 #define CONFIG_BITSTREAM_DEBUG 0 #define CONFIG_COEFFICIENT_RANGE_CHECKING 0 -#define CONFIG_COLLECT_INTER_MODE_RD_STATS 0 +#define CONFIG_COLLECT_COMPONENT_TIMING 0 +#define CONFIG_COLLECT_PARTITION_STATS 0 #define CONFIG_COLLECT_RD_STATS 0 #define CONFIG_DEBUG 0 #define CONFIG_DENOISE 1 +#define CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 1 #define CONFIG_DIST_8X8 0 #define CONFIG_ENTROPY_STATS 0 -#define CONFIG_FILEOPTIONS 1 -#define CONFIG_FIX_GF_LENGTH 1 -#define CONFIG_FP_MB_STATS 0 #define CONFIG_GCC 1 #define CONFIG_GCOV 0 -#define CONFIG_GLOBAL_MOTION_SEARCH 1 #define CONFIG_GPROF 0 +#define CONFIG_HTB_TRELLIS 0 #define CONFIG_INSPECTION 0 #define CONFIG_INTERNAL_STATS 0 #define CONFIG_INTER_STATS_ONLY 0 #define CONFIG_LIBYUV 0 -#define CONFIG_LOWBITDEPTH 1 +#define CONFIG_LPF_MASK 0 #define CONFIG_MAX_DECODE_PROFILE 2 #define CONFIG_MISMATCH_DEBUG 0 #define CONFIG_MULTITHREAD 1 +#define CONFIG_NN_V2 0 #define CONFIG_NORMAL_TILE_MODE 0 #define CONFIG_OS_SUPPORT 1 #define CONFIG_PIC 0 #define CONFIG_RD_DEBUG 0 -#define CONFIG_REDUCED_ENCODER_BORDER 0 +#define CONFIG_REALTIME_ONLY 0 #define CONFIG_RUNTIME_CPU_DETECT 1 #define CONFIG_SHARED 0 #define CONFIG_SHARP_SETTINGS 0 #define CONFIG_SIZE_LIMIT 0 #define CONFIG_SPATIAL_RESAMPLING 1 -#define CONFIG_STATIC 1 +#define CONFIG_SPEED_STATS 0 +#define CONFIG_SUPERRES_IN_RECODE 1 +#define CONFIG_TUNE_VMAF 0 #define CONFIG_WEBM_IO 0 #define DECODE_HEIGHT_LIMIT 0 #define DECODE_WIDTH_LIMIT 0 +#define FORCE_HIGHBITDEPTH_DECODING 0 #define HAVE_AVX 1 #define HAVE_AVX2 1 #define HAVE_DSPR2 0 @@ -76,7 +79,5 @@ #define HAVE_SSSE3 1 #define HAVE_VSX 0 #define HAVE_WXWIDGETS 0 -#define INCLUDE_INSTALL_DIR INSTALLDIR/include #define INLINE inline -#define LIB_INSTALL_DIR INSTALLDIR/lib #endif /* AOM_CONFIG_H_ */ diff --git a/media/libaom/config/win/mingw64/config/aom_dsp_rtcd.h b/media/libaom/config/win/mingw64/config/aom_dsp_rtcd.h index 2856d4ede7..36c170a05d 100644 --- a/media/libaom/config/win/mingw64/config/aom_dsp_rtcd.h +++ b/media/libaom/config/win/mingw64/config/aom_dsp_rtcd.h @@ -26,10 +26,10 @@ void aom_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src void aom_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); RTCD_EXTERN void (*aom_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); -void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); +void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); void aom_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); void aom_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); @@ -47,8 +47,8 @@ void aom_convolve8_vert_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t void aom_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*aom_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); +void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); #define aom_convolve_copy aom_convolve_copy_sse2 void aom_dc_128_predictor_16x16_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); @@ -471,35 +471,37 @@ void aom_h_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_h_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_h_predictor_8x8 aom_h_predictor_8x8_sse2 -void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params, const int bd); -#define aom_highbd_blend_a64_d16_mask aom_highbd_blend_a64_d16_mask_c +void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); void aom_highbd_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); +void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); void aom_highbd_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_vmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); void aom_highbd_dc_128_predictor_16x16_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); void aom_highbd_dc_128_predictor_16x16_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); @@ -855,10 +857,10 @@ void aom_highbd_lpf_horizontal_14_c(uint16_t *s, int pitch, const uint8_t *blimi void aom_highbd_lpf_horizontal_14_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); #define aom_highbd_lpf_horizontal_14 aom_highbd_lpf_horizontal_14_sse2 -void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); void aom_highbd_lpf_horizontal_4_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); void aom_highbd_lpf_horizontal_4_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); @@ -1231,10 +1233,10 @@ void aom_highbd_v_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_v_predictor_8x8_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); #define aom_highbd_v_predictor_8x8 aom_highbd_v_predictor_8x8_sse2 -void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); void aom_lpf_horizontal_14_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void aom_lpf_horizontal_14_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -1711,10 +1713,6 @@ void aom_v_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_v_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_v_predictor_8x8 aom_v_predictor_8x8_sse2 -void av1_round_shift_array_c(int32_t *arr, int size, int bit); -void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); -RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); - void aom_dsp_rtcd(void); #ifdef RTCD_C @@ -1788,6 +1786,9 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) aom_dc_top_predictor_64x64 = aom_dc_top_predictor_64x64_avx2; aom_h_predictor_32x32 = aom_h_predictor_32x32_sse2; if (flags & HAS_AVX2) aom_h_predictor_32x32 = aom_h_predictor_32x32_avx2; + aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_c; + if (flags & HAS_SSE4_1) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_sse4_1; + if (flags & HAS_AVX2) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_avx2; aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_sse4_1; aom_highbd_blend_a64_mask = aom_highbd_blend_a64_mask_c; @@ -1989,8 +1990,6 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) aom_v_predictor_64x32 = aom_v_predictor_64x32_avx2; aom_v_predictor_64x64 = aom_v_predictor_64x64_sse2; if (flags & HAS_AVX2) aom_v_predictor_64x64 = aom_v_predictor_64x64_avx2; - av1_round_shift_array = av1_round_shift_array_c; - if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; } #endif diff --git a/media/libaom/config/win/mingw64/config/aom_scale_rtcd.h b/media/libaom/config/win/mingw64/config/aom_scale_rtcd.h index 6b56795cd4..65c184b83d 100644 --- a/media/libaom/config/win/mingw64/config/aom_scale_rtcd.h +++ b/media/libaom/config/win/mingw64/config/aom_scale_rtcd.h @@ -59,15 +59,27 @@ void aom_yv12_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buf void aom_yv12_extend_frame_borders_c(struct yv12_buffer_config *ybf, const int num_planes); #define aom_yv12_extend_frame_borders aom_yv12_extend_frame_borders_c -void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_coloc_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_u aom_yv12_partial_coloc_copy_u_c + +void aom_yv12_partial_coloc_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_v aom_yv12_partial_coloc_copy_v_c + +void aom_yv12_partial_coloc_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_y aom_yv12_partial_coloc_copy_y_c + +void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_u aom_yv12_partial_copy_u_c -void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_v aom_yv12_partial_copy_v_c -void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_ybc, int hstart2, int vstart2); #define aom_yv12_partial_copy_y aom_yv12_partial_copy_y_c +int aom_yv12_realloc_with_new_border_c(struct yv12_buffer_config *ybf, int new_border, int byte_alignment, int num_planes); +#define aom_yv12_realloc_with_new_border aom_yv12_realloc_with_new_border_c + void aom_scale_rtcd(void); #ifdef RTCD_C diff --git a/media/libaom/config/win/mingw64/config/av1_rtcd.h b/media/libaom/config/win/mingw64/config/av1_rtcd.h index d273182086..8f20181b42 100644 --- a/media/libaom/config/win/mingw64/config/av1_rtcd.h +++ b/media/libaom/config/win/mingw64/config/av1_rtcd.h @@ -31,30 +31,55 @@ struct txfm_param; struct aom_variance_vtable; struct search_site_config; struct yv12_buffer_config; +struct NN_CONFIG; +typedef struct NN_CONFIG NN_CONFIG; + +enum { NONE, RELU, SOFTSIGN, SIGMOID } UENUM1BYTE(ACTIVATION); +#if CONFIG_NN_V2 +enum { SOFTMAX_CROSS_ENTROPY } UENUM1BYTE(LOSS); +struct NN_CONFIG_V2; +typedef struct NN_CONFIG_V2 NN_CONFIG_V2; +struct FC_LAYER; +typedef struct FC_LAYER FC_LAYER; +#endif // CONFIG_NN_V2 + +struct CNN_CONFIG; +typedef struct CNN_CONFIG CNN_CONFIG; +struct CNN_LAYER_CONFIG; +typedef struct CNN_LAYER_CONFIG CNN_LAYER_CONFIG; +struct CNN_THREAD_DATA; +typedef struct CNN_THREAD_DATA CNN_THREAD_DATA; +struct CNN_BRANCH_CONFIG; +typedef struct CNN_BRANCH_CONFIG CNN_BRANCH_CONFIG; +struct CNN_MULTI_OUT; +typedef struct CNN_MULTI_OUT CNN_MULTI_OUT; /* Function pointers return by CfL functions */ typedef void (*cfl_subsample_lbd_fn)(const uint8_t *input, int input_stride, uint16_t *output_q3); +#if CONFIG_AV1_HIGHBITDEPTH typedef void (*cfl_subsample_hbd_fn)(const uint16_t *input, int input_stride, uint16_t *output_q3); +typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, + int dst_stride, int alpha_q3, int bd); +#endif + typedef void (*cfl_subtract_average_fn)(const uint16_t *src, int16_t *dst); typedef void (*cfl_predict_lbd_fn)(const int16_t *src, uint8_t *dst, int dst_stride, int alpha_q3); -typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, - int dst_stride, int alpha_q3, int bd); #ifdef __cplusplus extern "C" { #endif -void apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -RTCD_EXTERN void (*apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +RTCD_EXTERN void (*av1_apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); void av1_build_compound_diffwtd_mask_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); void av1_build_compound_diffwtd_mask_sse4_1(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); @@ -71,42 +96,89 @@ void av1_build_compound_diffwtd_mask_highbd_ssse3(uint8_t *mask, DIFFWTD_MASK_TY void av1_build_compound_diffwtd_mask_highbd_avx2(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); RTCD_EXTERN void (*av1_build_compound_diffwtd_mask_highbd)(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); -void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +int64_t av1_calc_frame_error_c(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_sse2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_avx2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +RTCD_EXTERN int64_t (*av1_calc_frame_error)(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); + +void av1_cnn_activate_c( float **input, int channels, int width, int height, int stride, ACTIVATION layer_activation); +#define av1_cnn_activate av1_cnn_activate_c -void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); +void av1_cnn_add_c( float **input, int channels, int width, int height, int stride, const float **add); +#define av1_cnn_add av1_cnn_add_c -void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_cnn_batchnorm_c(float **image, int channels, int width, int height, int stride, const float *gamma, const float *beta, const float *mean, const float *std); +#define av1_cnn_batchnorm av1_cnn_batchnorm_c + +void av1_cnn_convolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride, int start_idx, int step); +#define av1_cnn_convolve av1_cnn_convolve_c + +void av1_cnn_deconvolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride); +#define av1_cnn_deconvolve av1_cnn_deconvolve_c + +void av1_cnn_predict_c( const float **input, int in_width, int in_height, int in_stride, const CNN_CONFIG *cnn_config, const CNN_THREAD_DATA *thread_data, CNN_MULTI_OUT *output_struct); +#define av1_cnn_predict av1_cnn_predict_c + +void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_convolve_horiz_rs_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); void av1_convolve_horiz_rs_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); RTCD_EXTERN void (*av1_convolve_horiz_rs)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); -void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_dr_prediction_z1_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); -#define av1_dr_prediction_z1 av1_dr_prediction_z1_c +void av1_dr_prediction_z1_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z1)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); void av1_dr_prediction_z2_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z2 av1_dr_prediction_z2_c +void av1_dr_prediction_z2_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z2)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); void av1_dr_prediction_z3_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z3 av1_dr_prediction_z3_c +void av1_dr_prediction_z3_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z3)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); void av1_filter_intra_edge_c(uint8_t *p, int sz, int strength); void av1_filter_intra_edge_sse4_1(uint8_t *p, int sz, int strength); @@ -132,19 +204,19 @@ void av1_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8 void av1_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve8_vert av1_highbd_convolve8_vert_sse2 -void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve_avg av1_highbd_convolve_avg_c @@ -156,54 +228,76 @@ void av1_highbd_convolve_horiz_rs_c(const uint16_t *src, int src_stride, uint16_ void av1_highbd_convolve_horiz_rs_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); RTCD_EXTERN void (*av1_highbd_convolve_horiz_rs)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); -void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_dr_prediction_z1_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z1 av1_highbd_dr_prediction_z1_c +void av1_highbd_dr_prediction_z1_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z1)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); void av1_highbd_dr_prediction_z2_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z2 av1_highbd_dr_prediction_z2_c +void av1_highbd_dr_prediction_z2_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z2)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); void av1_highbd_dr_prediction_z3_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z3 av1_highbd_dr_prediction_z3_c +void av1_highbd_dr_prediction_z3_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z3)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -void av1_highbd_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_avx2(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x16)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_32x32)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_16_add av1_highbd_iwht4x4_16_add_c @@ -211,34 +305,14 @@ void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int des void av1_highbd_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_1_add av1_highbd_iwht4x4_1_add_c -void av1_highbd_jnt_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - void av1_highbd_warp_affine_c(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_highbd_warp_affine_sse4_1(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_highbd_warp_affine)(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); -void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); +void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); void av1_inv_txfm2d_add_16x16_c(const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type, int bd); #define av1_inv_txfm2d_add_16x16 av1_inv_txfm2d_add_16x16_c @@ -304,25 +378,9 @@ void av1_inv_txfm_add_ssse3(const tran_low_t *dqcoeff, uint8_t *dst, int stride, void av1_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); RTCD_EXTERN void (*av1_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_jnt_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_round_shift_array_c(int32_t *arr, int size, int bit); +void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); +RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); int av1_selfguided_restoration_c(const uint8_t *dgd8, int width, int height, int dgd_stride, int32_t *flt0, int32_t *flt1, int flt_stride, @@ -347,6 +405,7 @@ RTCD_EXTERN void (*av1_upsample_intra_edge_high)(uint16_t *p, int sz, int bd); void av1_warp_affine_c(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_warp_affine_sse4_1(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); +void av1_warp_affine_avx2(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_warp_affine)(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); @@ -354,12 +413,26 @@ void av1_wiener_convolve_add_src_sse2(const uint8_t *src, ptrdiff_t src_stride, void av1_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); RTCD_EXTERN void (*av1_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); -void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); +void cdef_copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); + +void cdef_copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); + +void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); int cdef_find_dir_c(const uint16_t *img, int stride, int32_t *var, int coeff_shift); int cdef_find_dir_sse2(const uint16_t *img, int stride, int32_t *var, int coeff_shift); @@ -398,34 +471,20 @@ cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_ssse3(TX_SIZE tx_size); cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_avx2(TX_SIZE tx_size); RTCD_EXTERN cfl_subsample_lbd_fn (*cfl_get_luma_subsampling_444_lbd)(TX_SIZE tx_size); -void copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); - -void copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); - -cfl_predict_hbd_fn get_predict_hbd_fn_c(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_hbd_fn (*get_predict_hbd_fn)(TX_SIZE tx_size); - -cfl_predict_lbd_fn get_predict_lbd_fn_c(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_lbd_fn (*get_predict_lbd_fn)(TX_SIZE tx_size); - -cfl_subtract_average_fn get_subtract_average_fn_c(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_sse2(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_subtract_average_fn (*get_subtract_average_fn)(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_c(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_hbd_fn (*cfl_get_predict_hbd_fn)(TX_SIZE tx_size); + +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_c(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_lbd_fn (*cfl_get_predict_lbd_fn)(TX_SIZE tx_size); + +cfl_subtract_average_fn cfl_get_subtract_average_fn_c(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_sse2(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_subtract_average_fn (*cfl_get_subtract_average_fn)(TX_SIZE tx_size); void av1_rtcd(void); @@ -437,9 +496,9 @@ static void setup_rtcd_internal(void) (void)flags; - apply_selfguided_restoration = apply_selfguided_restoration_c; - if (flags & HAS_SSE4_1) apply_selfguided_restoration = apply_selfguided_restoration_sse4_1; - if (flags & HAS_AVX2) apply_selfguided_restoration = apply_selfguided_restoration_avx2; + av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_c; + if (flags & HAS_SSE4_1) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_sse4_1; + if (flags & HAS_AVX2) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_avx2; av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_c; if (flags & HAS_SSE4_1) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_sse4_1; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_avx2; @@ -449,6 +508,8 @@ static void setup_rtcd_internal(void) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_c; if (flags & HAS_SSSE3) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_ssse3; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_avx2; + av1_calc_frame_error = av1_calc_frame_error_sse2; + if (flags & HAS_AVX2) av1_calc_frame_error = av1_calc_frame_error_avx2; av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_sse2; if (flags & HAS_AVX2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_avx2; av1_convolve_2d_scale = av1_convolve_2d_scale_c; @@ -461,6 +522,21 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) av1_convolve_x_sr = av1_convolve_x_sr_avx2; av1_convolve_y_sr = av1_convolve_y_sr_sse2; if (flags & HAS_AVX2) av1_convolve_y_sr = av1_convolve_y_sr_avx2; + av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_sse2; + if (flags & HAS_SSSE3) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_ssse3; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_avx2; + av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_avx2; + av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_avx2; + av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_avx2; + av1_dr_prediction_z1 = av1_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_dr_prediction_z1 = av1_dr_prediction_z1_avx2; + av1_dr_prediction_z2 = av1_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_dr_prediction_z2 = av1_dr_prediction_z2_avx2; + av1_dr_prediction_z3 = av1_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_dr_prediction_z3 = av1_dr_prediction_z3_avx2; av1_filter_intra_edge = av1_filter_intra_edge_c; if (flags & HAS_SSE4_1) av1_filter_intra_edge = av1_filter_intra_edge_sse4_1; av1_filter_intra_edge_high = av1_filter_intra_edge_high_c; @@ -482,34 +558,39 @@ static void setup_rtcd_internal(void) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_c; if (flags & HAS_SSSE3) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_ssse3; if (flags & HAS_AVX2) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_avx2; + av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_avx2; + av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_avx2; + av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_avx2; + av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_avx2; + av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_avx2; + av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_avx2; + av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_avx2; av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_sse4_1; if (flags & HAS_AVX2) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_avx2; - av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_sse4_1; - av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_sse4_1; - av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_sse4_1; - if (flags & HAS_AVX2) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_avx2; + av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_sse4_1; + av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_sse4_1; av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_sse4_1; - av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_sse4_1; + av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_sse4_1; + av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_sse4_1; av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_sse4_1; - av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_avx2; - av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_avx2; - av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_avx2; - av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_avx2; av1_highbd_warp_affine = av1_highbd_warp_affine_c; if (flags & HAS_SSE4_1) av1_highbd_warp_affine = av1_highbd_warp_affine_sse4_1; av1_highbd_wiener_convolve_add_src = av1_highbd_wiener_convolve_add_src_c; @@ -522,15 +603,8 @@ static void setup_rtcd_internal(void) av1_inv_txfm_add = av1_inv_txfm_add_c; if (flags & HAS_SSSE3) av1_inv_txfm_add = av1_inv_txfm_add_ssse3; if (flags & HAS_AVX2) av1_inv_txfm_add = av1_inv_txfm_add_avx2; - av1_jnt_convolve_2d = av1_jnt_convolve_2d_c; - if (flags & HAS_SSSE3) av1_jnt_convolve_2d = av1_jnt_convolve_2d_ssse3; - if (flags & HAS_AVX2) av1_jnt_convolve_2d = av1_jnt_convolve_2d_avx2; - av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_avx2; - av1_jnt_convolve_x = av1_jnt_convolve_x_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_x = av1_jnt_convolve_x_avx2; - av1_jnt_convolve_y = av1_jnt_convolve_y_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_y = av1_jnt_convolve_y_avx2; + av1_round_shift_array = av1_round_shift_array_c; + if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; av1_selfguided_restoration = av1_selfguided_restoration_c; if (flags & HAS_SSE4_1) av1_selfguided_restoration = av1_selfguided_restoration_sse4_1; if (flags & HAS_AVX2) av1_selfguided_restoration = av1_selfguided_restoration_avx2; @@ -540,8 +614,17 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE4_1) av1_upsample_intra_edge_high = av1_upsample_intra_edge_high_sse4_1; av1_warp_affine = av1_warp_affine_c; if (flags & HAS_SSE4_1) av1_warp_affine = av1_warp_affine_sse4_1; + if (flags & HAS_AVX2) av1_warp_affine = av1_warp_affine_avx2; av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_sse2; if (flags & HAS_AVX2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_avx2; + cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_avx2; + cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_avx2; cdef_filter_block = cdef_filter_block_sse2; if (flags & HAS_SSSE3) cdef_filter_block = cdef_filter_block_ssse3; if (flags & HAS_SSE4_1) cdef_filter_block = cdef_filter_block_sse4_1; @@ -568,22 +651,14 @@ static void setup_rtcd_internal(void) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_c; if (flags & HAS_SSSE3) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_ssse3; if (flags & HAS_AVX2) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_avx2; - copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_avx2; - copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_avx2; - get_predict_hbd_fn = get_predict_hbd_fn_c; - if (flags & HAS_SSSE3) get_predict_hbd_fn = get_predict_hbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_hbd_fn = get_predict_hbd_fn_avx2; - get_predict_lbd_fn = get_predict_lbd_fn_c; - if (flags & HAS_SSSE3) get_predict_lbd_fn = get_predict_lbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_lbd_fn = get_predict_lbd_fn_avx2; - get_subtract_average_fn = get_subtract_average_fn_sse2; - if (flags & HAS_AVX2) get_subtract_average_fn = get_subtract_average_fn_avx2; + cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_avx2; + cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_avx2; + cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_sse2; + if (flags & HAS_AVX2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_avx2; } #endif diff --git a/media/libaom/config/win/x64/config/aom_config.asm b/media/libaom/config/win/x64/config/aom_config.asm index 4ab2dedb42..9ce9b2090e 100644 --- a/media/libaom/config/win/x64/config/aom_config.asm +++ b/media/libaom/config/win/x64/config/aom_config.asm @@ -1,5 +1,5 @@ ; -; Copyright (c) 2018, Alliance for Open Media. All rights reserved +; Copyright (c) 2021, 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 @@ -14,49 +14,52 @@ ARCH_MIPS equ 0 ARCH_PPC equ 0 ARCH_X86 equ 0 ARCH_X86_64 equ 1 -CONFIG_2PASS_PARTITION_SEARCH_LVL equ 1 CONFIG_ACCOUNTING equ 0 CONFIG_ANALYZER equ 0 CONFIG_AV1_DECODER equ 1 CONFIG_AV1_ENCODER equ 0 +CONFIG_AV1_HIGHBITDEPTH equ 1 CONFIG_BIG_ENDIAN equ 0 CONFIG_BITSTREAM_DEBUG equ 0 CONFIG_COEFFICIENT_RANGE_CHECKING equ 0 -CONFIG_COLLECT_INTER_MODE_RD_STATS equ 0 +CONFIG_COLLECT_COMPONENT_TIMING equ 0 +CONFIG_COLLECT_PARTITION_STATS equ 0 CONFIG_COLLECT_RD_STATS equ 0 CONFIG_DEBUG equ 0 CONFIG_DENOISE equ 1 +CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 equ 1 CONFIG_DIST_8X8 equ 0 CONFIG_ENTROPY_STATS equ 0 -CONFIG_FILEOPTIONS equ 1 -CONFIG_FIX_GF_LENGTH equ 1 -CONFIG_FP_MB_STATS equ 0 CONFIG_GCC equ 1 CONFIG_GCOV equ 0 -CONFIG_GLOBAL_MOTION_SEARCH equ 1 CONFIG_GPROF equ 0 +CONFIG_HTB_TRELLIS equ 0 CONFIG_INSPECTION equ 0 CONFIG_INTERNAL_STATS equ 0 CONFIG_INTER_STATS_ONLY equ 0 CONFIG_LIBYUV equ 0 -CONFIG_LOWBITDEPTH equ 1 +CONFIG_LPF_MASK equ 0 CONFIG_MAX_DECODE_PROFILE equ 2 CONFIG_MISMATCH_DEBUG equ 0 CONFIG_MULTITHREAD equ 1 +CONFIG_NN_V2 equ 0 CONFIG_NORMAL_TILE_MODE equ 0 CONFIG_OS_SUPPORT equ 1 CONFIG_PIC equ 0 CONFIG_RD_DEBUG equ 0 -CONFIG_REDUCED_ENCODER_BORDER equ 0 +CONFIG_REALTIME_ONLY equ 0 CONFIG_RUNTIME_CPU_DETECT equ 1 CONFIG_SHARED equ 0 CONFIG_SHARP_SETTINGS equ 0 CONFIG_SIZE_LIMIT equ 0 CONFIG_SPATIAL_RESAMPLING equ 1 -CONFIG_STATIC equ 1 +CONFIG_SPEED_STATS equ 0 +CONFIG_SUPERRES_IN_RECODE equ 1 +CONFIG_TUNE_VMAF equ 0 CONFIG_WEBM_IO equ 0 DECODE_HEIGHT_LIMIT equ 0 DECODE_WIDTH_LIMIT equ 0 +FORCE_HIGHBITDEPTH_DECODING equ 0 HAVE_AVX equ 1 HAVE_AVX2 equ 1 HAVE_DSPR2 equ 0 diff --git a/media/libaom/config/win/x64/config/aom_config.h b/media/libaom/config/win/x64/config/aom_config.h index 9a0be7cd11..12997ea4dd 100644 --- a/media/libaom/config/win/x64/config/aom_config.h +++ b/media/libaom/config/win/x64/config/aom_config.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018, Alliance for Open Media. All rights reserved + * Copyright (c) 2021, 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 @@ -16,49 +16,52 @@ #define ARCH_PPC 0 #define ARCH_X86 0 #define ARCH_X86_64 1 -#define CONFIG_2PASS_PARTITION_SEARCH_LVL 1 #define CONFIG_ACCOUNTING 0 #define CONFIG_ANALYZER 0 #define CONFIG_AV1_DECODER 1 #define CONFIG_AV1_ENCODER 0 +#define CONFIG_AV1_HIGHBITDEPTH 1 #define CONFIG_BIG_ENDIAN 0 #define CONFIG_BITSTREAM_DEBUG 0 #define CONFIG_COEFFICIENT_RANGE_CHECKING 0 -#define CONFIG_COLLECT_INTER_MODE_RD_STATS 0 +#define CONFIG_COLLECT_COMPONENT_TIMING 0 +#define CONFIG_COLLECT_PARTITION_STATS 0 #define CONFIG_COLLECT_RD_STATS 0 #define CONFIG_DEBUG 0 #define CONFIG_DENOISE 1 +#define CONFIG_DISABLE_FULL_PIXEL_SPLIT_8X8 1 #define CONFIG_DIST_8X8 0 #define CONFIG_ENTROPY_STATS 0 -#define CONFIG_FILEOPTIONS 1 -#define CONFIG_FIX_GF_LENGTH 1 -#define CONFIG_FP_MB_STATS 0 #define CONFIG_GCC 1 #define CONFIG_GCOV 0 -#define CONFIG_GLOBAL_MOTION_SEARCH 1 #define CONFIG_GPROF 0 +#define CONFIG_HTB_TRELLIS 0 #define CONFIG_INSPECTION 0 #define CONFIG_INTERNAL_STATS 0 #define CONFIG_INTER_STATS_ONLY 0 #define CONFIG_LIBYUV 0 -#define CONFIG_LOWBITDEPTH 1 +#define CONFIG_LPF_MASK 0 #define CONFIG_MAX_DECODE_PROFILE 2 #define CONFIG_MISMATCH_DEBUG 0 #define CONFIG_MULTITHREAD 1 +#define CONFIG_NN_V2 0 #define CONFIG_NORMAL_TILE_MODE 0 #define CONFIG_OS_SUPPORT 1 #define CONFIG_PIC 0 #define CONFIG_RD_DEBUG 0 -#define CONFIG_REDUCED_ENCODER_BORDER 0 +#define CONFIG_REALTIME_ONLY 0 #define CONFIG_RUNTIME_CPU_DETECT 1 #define CONFIG_SHARED 0 #define CONFIG_SHARP_SETTINGS 0 #define CONFIG_SIZE_LIMIT 0 #define CONFIG_SPATIAL_RESAMPLING 1 -#define CONFIG_STATIC 1 +#define CONFIG_SPEED_STATS 0 +#define CONFIG_SUPERRES_IN_RECODE 1 +#define CONFIG_TUNE_VMAF 0 #define CONFIG_WEBM_IO 0 #define DECODE_HEIGHT_LIMIT 0 #define DECODE_WIDTH_LIMIT 0 +#define FORCE_HIGHBITDEPTH_DECODING 0 #define HAVE_AVX 1 #define HAVE_AVX2 1 #define HAVE_DSPR2 0 @@ -76,7 +79,5 @@ #define HAVE_SSSE3 1 #define HAVE_VSX 0 #define HAVE_WXWIDGETS 0 -#define INCLUDE_INSTALL_DIR INSTALLDIR/include #define INLINE inline -#define LIB_INSTALL_DIR INSTALLDIR/lib #endif /* AOM_CONFIG_H_ */ diff --git a/media/libaom/config/win/x64/config/aom_dsp_rtcd.h b/media/libaom/config/win/x64/config/aom_dsp_rtcd.h index 2856d4ede7..36c170a05d 100644 --- a/media/libaom/config/win/x64/config/aom_dsp_rtcd.h +++ b/media/libaom/config/win/x64/config/aom_dsp_rtcd.h @@ -26,10 +26,10 @@ void aom_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src void aom_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); RTCD_EXTERN void (*aom_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); -void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); -RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby); +void aom_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +void aom_blend_a64_mask_avx2(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); +RTCD_EXTERN void (*aom_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh); void aom_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); void aom_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h); @@ -47,8 +47,8 @@ void aom_convolve8_vert_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t void aom_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*aom_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); +void aom_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); +void aom_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h); #define aom_convolve_copy aom_convolve_copy_sse2 void aom_dc_128_predictor_16x16_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); @@ -471,35 +471,37 @@ void aom_h_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_h_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_h_predictor_8x8 aom_h_predictor_8x8_sse2 -void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params, const int bd); -#define aom_highbd_blend_a64_d16_mask aom_highbd_blend_a64_d16_mask_c +void aom_highbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +void aom_highbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params, const int bd); void aom_highbd_blend_a64_hmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_hmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_hmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); -RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, int bd); +void aom_highbd_blend_a64_mask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +void aom_highbd_blend_a64_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); +RTCD_EXTERN void (*aom_highbd_blend_a64_mask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, int bd); void aom_highbd_blend_a64_vmask_c(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); void aom_highbd_blend_a64_vmask_sse4_1(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); RTCD_EXTERN void (*aom_highbd_blend_a64_vmask)(uint8_t *dst, uint32_t dst_stride, const uint8_t *src0, uint32_t src0_stride, const uint8_t *src1, uint32_t src1_stride, const uint8_t *mask, int w, int h, int bd); -void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +void aom_highbd_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bd); -void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); -RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); +void aom_highbd_convolve_copy_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +void aom_highbd_convolve_copy_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); +RTCD_EXTERN void (*aom_highbd_convolve_copy)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int filter_x_stride, const int16_t *filter_y, int filter_y_stride, int w, int h, int bd); void aom_highbd_dc_128_predictor_16x16_c(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); void aom_highbd_dc_128_predictor_16x16_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); @@ -855,10 +857,10 @@ void aom_highbd_lpf_horizontal_14_c(uint16_t *s, int pitch, const uint8_t *blimi void aom_highbd_lpf_horizontal_14_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); #define aom_highbd_lpf_horizontal_14 aom_highbd_lpf_horizontal_14_sse2 -void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); -RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limt1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_c(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_sse2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +void aom_highbd_lpf_horizontal_14_dual_avx2(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); +RTCD_EXTERN void (*aom_highbd_lpf_horizontal_14_dual)(uint16_t *s, int pitch, const uint8_t *blimit0, const uint8_t *limit0, const uint8_t *thresh0, const uint8_t *blimit1, const uint8_t *limit1, const uint8_t *thresh1,int bd); void aom_highbd_lpf_horizontal_4_c(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); void aom_highbd_lpf_horizontal_4_sse2(uint16_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int bd); @@ -1231,10 +1233,10 @@ void aom_highbd_v_predictor_8x8_c(uint16_t *dst, ptrdiff_t y_stride, const uint1 void aom_highbd_v_predictor_8x8_sse2(uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd); #define aom_highbd_v_predictor_8x8 aom_highbd_v_predictor_8x8_sse2 -void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); -RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subx, int suby, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_c(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_sse4_1(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +void aom_lowbd_blend_a64_d16_mask_avx2(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); +RTCD_EXTERN void (*aom_lowbd_blend_a64_d16_mask)(uint8_t *dst, uint32_t dst_stride, const CONV_BUF_TYPE *src0, uint32_t src0_stride, const CONV_BUF_TYPE *src1, uint32_t src1_stride, const uint8_t *mask, uint32_t mask_stride, int w, int h, int subw, int subh, ConvolveParams *conv_params); void aom_lpf_horizontal_14_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void aom_lpf_horizontal_14_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -1711,10 +1713,6 @@ void aom_v_predictor_8x8_c(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *abov void aom_v_predictor_8x8_sse2(uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left); #define aom_v_predictor_8x8 aom_v_predictor_8x8_sse2 -void av1_round_shift_array_c(int32_t *arr, int size, int bit); -void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); -RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); - void aom_dsp_rtcd(void); #ifdef RTCD_C @@ -1788,6 +1786,9 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) aom_dc_top_predictor_64x64 = aom_dc_top_predictor_64x64_avx2; aom_h_predictor_32x32 = aom_h_predictor_32x32_sse2; if (flags & HAS_AVX2) aom_h_predictor_32x32 = aom_h_predictor_32x32_avx2; + aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_c; + if (flags & HAS_SSE4_1) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_sse4_1; + if (flags & HAS_AVX2) aom_highbd_blend_a64_d16_mask = aom_highbd_blend_a64_d16_mask_avx2; aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_c; if (flags & HAS_SSE4_1) aom_highbd_blend_a64_hmask = aom_highbd_blend_a64_hmask_sse4_1; aom_highbd_blend_a64_mask = aom_highbd_blend_a64_mask_c; @@ -1989,8 +1990,6 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) aom_v_predictor_64x32 = aom_v_predictor_64x32_avx2; aom_v_predictor_64x64 = aom_v_predictor_64x64_sse2; if (flags & HAS_AVX2) aom_v_predictor_64x64 = aom_v_predictor_64x64_avx2; - av1_round_shift_array = av1_round_shift_array_c; - if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; } #endif diff --git a/media/libaom/config/win/x64/config/aom_scale_rtcd.h b/media/libaom/config/win/x64/config/aom_scale_rtcd.h index 6b56795cd4..65c184b83d 100644 --- a/media/libaom/config/win/x64/config/aom_scale_rtcd.h +++ b/media/libaom/config/win/x64/config/aom_scale_rtcd.h @@ -59,15 +59,27 @@ void aom_yv12_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buf void aom_yv12_extend_frame_borders_c(struct yv12_buffer_config *ybf, const int num_planes); #define aom_yv12_extend_frame_borders aom_yv12_extend_frame_borders_c -void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_coloc_copy_u_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_u aom_yv12_partial_coloc_copy_u_c + +void aom_yv12_partial_coloc_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_v aom_yv12_partial_coloc_copy_v_c + +void aom_yv12_partial_coloc_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +#define aom_yv12_partial_coloc_copy_y aom_yv12_partial_coloc_copy_y_c + +void aom_yv12_partial_copy_u_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_u aom_yv12_partial_copy_u_c -void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, struct yv12_buffer_config *dst_bc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_v_c(const struct yv12_buffer_config *src_bc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_bc, int hstart2, int vstart2); #define aom_yv12_partial_copy_v aom_yv12_partial_copy_v_c -void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, struct yv12_buffer_config *dst_ybc, int hstart, int hend, int vstart, int vend); +void aom_yv12_partial_copy_y_c(const struct yv12_buffer_config *src_ybc, int hstart1, int hend1, int vstart1, int vend1, struct yv12_buffer_config *dst_ybc, int hstart2, int vstart2); #define aom_yv12_partial_copy_y aom_yv12_partial_copy_y_c +int aom_yv12_realloc_with_new_border_c(struct yv12_buffer_config *ybf, int new_border, int byte_alignment, int num_planes); +#define aom_yv12_realloc_with_new_border aom_yv12_realloc_with_new_border_c + void aom_scale_rtcd(void); #ifdef RTCD_C diff --git a/media/libaom/config/win/x64/config/av1_rtcd.h b/media/libaom/config/win/x64/config/av1_rtcd.h index d273182086..8f20181b42 100644 --- a/media/libaom/config/win/x64/config/av1_rtcd.h +++ b/media/libaom/config/win/x64/config/av1_rtcd.h @@ -31,30 +31,55 @@ struct txfm_param; struct aom_variance_vtable; struct search_site_config; struct yv12_buffer_config; +struct NN_CONFIG; +typedef struct NN_CONFIG NN_CONFIG; + +enum { NONE, RELU, SOFTSIGN, SIGMOID } UENUM1BYTE(ACTIVATION); +#if CONFIG_NN_V2 +enum { SOFTMAX_CROSS_ENTROPY } UENUM1BYTE(LOSS); +struct NN_CONFIG_V2; +typedef struct NN_CONFIG_V2 NN_CONFIG_V2; +struct FC_LAYER; +typedef struct FC_LAYER FC_LAYER; +#endif // CONFIG_NN_V2 + +struct CNN_CONFIG; +typedef struct CNN_CONFIG CNN_CONFIG; +struct CNN_LAYER_CONFIG; +typedef struct CNN_LAYER_CONFIG CNN_LAYER_CONFIG; +struct CNN_THREAD_DATA; +typedef struct CNN_THREAD_DATA CNN_THREAD_DATA; +struct CNN_BRANCH_CONFIG; +typedef struct CNN_BRANCH_CONFIG CNN_BRANCH_CONFIG; +struct CNN_MULTI_OUT; +typedef struct CNN_MULTI_OUT CNN_MULTI_OUT; /* Function pointers return by CfL functions */ typedef void (*cfl_subsample_lbd_fn)(const uint8_t *input, int input_stride, uint16_t *output_q3); +#if CONFIG_AV1_HIGHBITDEPTH typedef void (*cfl_subsample_hbd_fn)(const uint16_t *input, int input_stride, uint16_t *output_q3); +typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, + int dst_stride, int alpha_q3, int bd); +#endif + typedef void (*cfl_subtract_average_fn)(const uint16_t *src, int16_t *dst); typedef void (*cfl_predict_lbd_fn)(const int16_t *src, uint8_t *dst, int dst_stride, int alpha_q3); -typedef void (*cfl_predict_hbd_fn)(const int16_t *src, uint16_t *dst, - int dst_stride, int alpha_q3, int bd); #ifdef __cplusplus extern "C" { #endif -void apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -void apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); -RTCD_EXTERN void (*apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_c(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_sse4_1(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +void av1_apply_selfguided_restoration_avx2(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); +RTCD_EXTERN void (*av1_apply_selfguided_restoration)(const uint8_t *dat, int width, int height, int stride, int eps, const int *xqd, uint8_t *dst, int dst_stride, int32_t *tmpbuf, int bit_depth, int highbd); void av1_build_compound_diffwtd_mask_c(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); void av1_build_compound_diffwtd_mask_sse4_1(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w); @@ -71,42 +96,89 @@ void av1_build_compound_diffwtd_mask_highbd_ssse3(uint8_t *mask, DIFFWTD_MASK_TY void av1_build_compound_diffwtd_mask_highbd_avx2(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); RTCD_EXTERN void (*av1_build_compound_diffwtd_mask_highbd)(uint8_t *mask, DIFFWTD_MASK_TYPE mask_type, const uint8_t *src0, int src0_stride, const uint8_t *src1, int src1_stride, int h, int w, int bd); -void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +int64_t av1_calc_frame_error_c(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_sse2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +int64_t av1_calc_frame_error_avx2(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); +RTCD_EXTERN int64_t (*av1_calc_frame_error)(const uint8_t *const ref, int stride, const uint8_t *const dst, int p_width, int p_height, int p_stride); + +void av1_cnn_activate_c( float **input, int channels, int width, int height, int stride, ACTIVATION layer_activation); +#define av1_cnn_activate av1_cnn_activate_c -void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params); +void av1_cnn_add_c( float **input, int channels, int width, int height, int stride, const float **add); +#define av1_cnn_add av1_cnn_add_c -void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_cnn_batchnorm_c(float **image, int channels, int width, int height, int stride, const float *gamma, const float *beta, const float *mean, const float *std); +#define av1_cnn_batchnorm av1_cnn_batchnorm_c + +void av1_cnn_convolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride, int start_idx, int step); +#define av1_cnn_convolve av1_cnn_convolve_c + +void av1_cnn_deconvolve_c( const float **input, int in_width, int in_height, int in_stride, const CNN_LAYER_CONFIG *layer_config, float **output, int out_stride); +#define av1_cnn_deconvolve av1_cnn_deconvolve_c + +void av1_cnn_predict_c( const float **input, int in_width, int in_height, int in_stride, const CNN_CONFIG *cnn_config, const CNN_THREAD_DATA *thread_data, CNN_MULTI_OUT *output_struct); +#define av1_cnn_predict av1_cnn_predict_c + +void av1_convolve_2d_copy_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_copy_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_copy_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_scale_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +void av1_convolve_2d_scale_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_scale)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params); + +void av1_convolve_2d_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_2d_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_2d_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_convolve_horiz_rs_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); void av1_convolve_horiz_rs_sse4_1(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); RTCD_EXTERN void (*av1_convolve_horiz_rs)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn); -void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_convolve_x_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_x_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_x_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_convolve_y_sr_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_convolve_y_sr_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_convolve_y_sr)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); + +void av1_dist_wtd_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +void av1_dist_wtd_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); +RTCD_EXTERN void (*av1_dist_wtd_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params); void av1_dr_prediction_z1_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); -#define av1_dr_prediction_z1 av1_dr_prediction_z1_c +void av1_dr_prediction_z1_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z1)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int dx, int dy); void av1_dr_prediction_z2_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z2 av1_dr_prediction_z2_c +void av1_dr_prediction_z2_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z2)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_above, int upsample_left, int dx, int dy); void av1_dr_prediction_z3_c(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); -#define av1_dr_prediction_z3 av1_dr_prediction_z3_c +void av1_dr_prediction_z3_avx2(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); +RTCD_EXTERN void (*av1_dr_prediction_z3)(uint8_t *dst, ptrdiff_t stride, int bw, int bh, const uint8_t *above, const uint8_t *left, int upsample_left, int dx, int dy); void av1_filter_intra_edge_c(uint8_t *p, int sz, int strength); void av1_filter_intra_edge_sse4_1(uint8_t *p, int sz, int strength); @@ -132,19 +204,19 @@ void av1_highbd_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8 void av1_highbd_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve8_vert av1_highbd_convolve8_vert_sse2 -void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_sse2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_copy_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_copy_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int x_step_qn, const int subpel_y_q4, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_scale_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_scale)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int x_step_qn, const int subpel_y_qn, const int y_step_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_2d_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_2d_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps); #define av1_highbd_convolve_avg av1_highbd_convolve_avg_c @@ -156,54 +228,76 @@ void av1_highbd_convolve_horiz_rs_c(const uint16_t *src, int src_stride, uint16_ void av1_highbd_convolve_horiz_rs_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); RTCD_EXTERN void (*av1_highbd_convolve_horiz_rs)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const int16_t *x_filters, int x0_qn, int x_step_qn, int bd); -void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_x_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_x_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_ssse3(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_convolve_y_sr_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_convolve_y_sr)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); + +void av1_highbd_dist_wtd_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +void av1_highbd_dist_wtd_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_dist_wtd_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_qn, const int subpel_y_qn, ConvolveParams *conv_params, int bd); void av1_highbd_dr_prediction_z1_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z1 av1_highbd_dr_prediction_z1_c +void av1_highbd_dr_prediction_z1_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z1)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int dx, int dy, int bd); void av1_highbd_dr_prediction_z2_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z2 av1_highbd_dr_prediction_z2_c +void av1_highbd_dr_prediction_z2_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z2)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_above, int upsample_left, int dx, int dy, int bd); void av1_highbd_dr_prediction_z3_c(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -#define av1_highbd_dr_prediction_z3 av1_highbd_dr_prediction_z3_c +void av1_highbd_dr_prediction_z3_avx2(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); +RTCD_EXTERN void (*av1_highbd_dr_prediction_z3)(uint16_t *dst, ptrdiff_t stride, int bw, int bh, const uint16_t *above, const uint16_t *left, int upsample_left, int dx, int dy, int bd); -void av1_highbd_inv_txfm_add_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_avx2(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_16x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_16x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_16x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x16_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x16)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_32x32_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_32x32)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x4)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_4x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_4x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x16_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x16)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x4_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x4)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_c(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); +RTCD_EXTERN void (*av1_highbd_inv_txfm_add_8x8)(const tran_low_t *input, uint8_t *dest, int stride, const TxfmParam *txfm_param); void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_16_add av1_highbd_iwht4x4_16_add_c @@ -211,34 +305,14 @@ void av1_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int des void av1_highbd_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride, int bd); #define av1_highbd_iwht4x4_1_add av1_highbd_iwht4x4_1_add_c -void av1_highbd_jnt_convolve_2d_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_2d_copy_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_2d_copy_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_2d_copy)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_x_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_x_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_x)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - -void av1_highbd_jnt_convolve_y_c(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_sse4_1(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -void av1_highbd_jnt_convolve_y_avx2(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); -RTCD_EXTERN void (*av1_highbd_jnt_convolve_y)(const uint16_t *src, int src_stride, uint16_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params, int bd); - void av1_highbd_warp_affine_c(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_highbd_warp_affine_sse4_1(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_highbd_warp_affine)(const int32_t *mat, const uint16_t *ref, int width, int height, int stride, uint16_t *pred, int p_col, int p_row, int p_width, int p_height, int p_stride, int subsampling_x, int subsampling_y, int bd, ConvolveParams *conv_params, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); -void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); -RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bps); +void av1_highbd_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +void av1_highbd_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); +RTCD_EXTERN void (*av1_highbd_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params, int bd); void av1_inv_txfm2d_add_16x16_c(const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type, int bd); #define av1_inv_txfm2d_add_16x16 av1_inv_txfm2d_add_16x16_c @@ -304,25 +378,9 @@ void av1_inv_txfm_add_ssse3(const tran_low_t *dqcoeff, uint8_t *dst, int stride, void av1_inv_txfm_add_avx2(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); RTCD_EXTERN void (*av1_inv_txfm_add)(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param); -void av1_jnt_convolve_2d_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_ssse3(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_2d_copy_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_2d_copy_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_2d_copy)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_x_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_x_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_x)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); - -void av1_jnt_convolve_y_c(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_sse2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -void av1_jnt_convolve_y_avx2(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); -RTCD_EXTERN void (*av1_jnt_convolve_y)(const uint8_t *src, int src_stride, uint8_t *dst, int dst_stride, int w, int h, const InterpFilterParams *filter_params_x, const InterpFilterParams *filter_params_y, const int subpel_x_q4, const int subpel_y_q4, ConvolveParams *conv_params); +void av1_round_shift_array_c(int32_t *arr, int size, int bit); +void av1_round_shift_array_sse4_1(int32_t *arr, int size, int bit); +RTCD_EXTERN void (*av1_round_shift_array)(int32_t *arr, int size, int bit); int av1_selfguided_restoration_c(const uint8_t *dgd8, int width, int height, int dgd_stride, int32_t *flt0, int32_t *flt1, int flt_stride, @@ -347,6 +405,7 @@ RTCD_EXTERN void (*av1_upsample_intra_edge_high)(uint16_t *p, int sz, int bd); void av1_warp_affine_c(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_warp_affine_sse4_1(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); +void av1_warp_affine_avx2(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); RTCD_EXTERN void (*av1_warp_affine)(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, int16_t alpha, int16_t beta, int16_t gamma, int16_t delta); void av1_wiener_convolve_add_src_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); @@ -354,12 +413,26 @@ void av1_wiener_convolve_add_src_sse2(const uint8_t *src, ptrdiff_t src_stride, void av1_wiener_convolve_add_src_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); RTCD_EXTERN void (*av1_wiener_convolve_add_src)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, const ConvolveParams *conv_params); -void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); -RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int max, int coeff_shift); +void cdef_copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +void cdef_copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); + +void cdef_copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +void cdef_copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); +RTCD_EXTERN void (*cdef_copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); + +void cdef_filter_block_c(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_ssse3(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_sse4_1(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +void cdef_filter_block_avx2(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); +RTCD_EXTERN void (*cdef_filter_block)(uint8_t *dst8, uint16_t *dst16, int dstride, const uint16_t *in, int pri_strength, int sec_strength, int dir, int pri_damping, int sec_damping, int bsize, int coeff_shift); int cdef_find_dir_c(const uint16_t *img, int stride, int32_t *var, int coeff_shift); int cdef_find_dir_sse2(const uint16_t *img, int stride, int32_t *var, int coeff_shift); @@ -398,34 +471,20 @@ cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_ssse3(TX_SIZE tx_size); cfl_subsample_lbd_fn cfl_get_luma_subsampling_444_lbd_avx2(TX_SIZE tx_size); RTCD_EXTERN cfl_subsample_lbd_fn (*cfl_get_luma_subsampling_444_lbd)(TX_SIZE tx_size); -void copy_rect8_16bit_to_16bit_c(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -void copy_rect8_16bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_16bit_to_16bit)(uint16_t *dst, int dstride, const uint16_t *src, int sstride, int v, int h); - -void copy_rect8_8bit_to_16bit_c(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_ssse3(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_sse4_1(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -void copy_rect8_8bit_to_16bit_avx2(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); -RTCD_EXTERN void (*copy_rect8_8bit_to_16bit)(uint16_t *dst, int dstride, const uint8_t *src, int sstride, int v, int h); - -cfl_predict_hbd_fn get_predict_hbd_fn_c(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_hbd_fn get_predict_hbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_hbd_fn (*get_predict_hbd_fn)(TX_SIZE tx_size); - -cfl_predict_lbd_fn get_predict_lbd_fn_c(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_ssse3(TX_SIZE tx_size); -cfl_predict_lbd_fn get_predict_lbd_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_predict_lbd_fn (*get_predict_lbd_fn)(TX_SIZE tx_size); - -cfl_subtract_average_fn get_subtract_average_fn_c(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_sse2(TX_SIZE tx_size); -cfl_subtract_average_fn get_subtract_average_fn_avx2(TX_SIZE tx_size); -RTCD_EXTERN cfl_subtract_average_fn (*get_subtract_average_fn)(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_c(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_hbd_fn cfl_get_predict_hbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_hbd_fn (*cfl_get_predict_hbd_fn)(TX_SIZE tx_size); + +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_c(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_ssse3(TX_SIZE tx_size); +cfl_predict_lbd_fn cfl_get_predict_lbd_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_predict_lbd_fn (*cfl_get_predict_lbd_fn)(TX_SIZE tx_size); + +cfl_subtract_average_fn cfl_get_subtract_average_fn_c(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_sse2(TX_SIZE tx_size); +cfl_subtract_average_fn cfl_get_subtract_average_fn_avx2(TX_SIZE tx_size); +RTCD_EXTERN cfl_subtract_average_fn (*cfl_get_subtract_average_fn)(TX_SIZE tx_size); void av1_rtcd(void); @@ -437,9 +496,9 @@ static void setup_rtcd_internal(void) (void)flags; - apply_selfguided_restoration = apply_selfguided_restoration_c; - if (flags & HAS_SSE4_1) apply_selfguided_restoration = apply_selfguided_restoration_sse4_1; - if (flags & HAS_AVX2) apply_selfguided_restoration = apply_selfguided_restoration_avx2; + av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_c; + if (flags & HAS_SSE4_1) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_sse4_1; + if (flags & HAS_AVX2) av1_apply_selfguided_restoration = av1_apply_selfguided_restoration_avx2; av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_c; if (flags & HAS_SSE4_1) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_sse4_1; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask = av1_build_compound_diffwtd_mask_avx2; @@ -449,6 +508,8 @@ static void setup_rtcd_internal(void) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_c; if (flags & HAS_SSSE3) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_ssse3; if (flags & HAS_AVX2) av1_build_compound_diffwtd_mask_highbd = av1_build_compound_diffwtd_mask_highbd_avx2; + av1_calc_frame_error = av1_calc_frame_error_sse2; + if (flags & HAS_AVX2) av1_calc_frame_error = av1_calc_frame_error_avx2; av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_sse2; if (flags & HAS_AVX2) av1_convolve_2d_copy_sr = av1_convolve_2d_copy_sr_avx2; av1_convolve_2d_scale = av1_convolve_2d_scale_c; @@ -461,6 +522,21 @@ static void setup_rtcd_internal(void) if (flags & HAS_AVX2) av1_convolve_x_sr = av1_convolve_x_sr_avx2; av1_convolve_y_sr = av1_convolve_y_sr_sse2; if (flags & HAS_AVX2) av1_convolve_y_sr = av1_convolve_y_sr_avx2; + av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_sse2; + if (flags & HAS_SSSE3) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_ssse3; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d = av1_dist_wtd_convolve_2d_avx2; + av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_2d_copy = av1_dist_wtd_convolve_2d_copy_avx2; + av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_x = av1_dist_wtd_convolve_x_avx2; + av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_sse2; + if (flags & HAS_AVX2) av1_dist_wtd_convolve_y = av1_dist_wtd_convolve_y_avx2; + av1_dr_prediction_z1 = av1_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_dr_prediction_z1 = av1_dr_prediction_z1_avx2; + av1_dr_prediction_z2 = av1_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_dr_prediction_z2 = av1_dr_prediction_z2_avx2; + av1_dr_prediction_z3 = av1_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_dr_prediction_z3 = av1_dr_prediction_z3_avx2; av1_filter_intra_edge = av1_filter_intra_edge_c; if (flags & HAS_SSE4_1) av1_filter_intra_edge = av1_filter_intra_edge_sse4_1; av1_filter_intra_edge_high = av1_filter_intra_edge_high_c; @@ -482,34 +558,39 @@ static void setup_rtcd_internal(void) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_c; if (flags & HAS_SSSE3) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_ssse3; if (flags & HAS_AVX2) av1_highbd_convolve_y_sr = av1_highbd_convolve_y_sr_avx2; + av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d = av1_highbd_dist_wtd_convolve_2d_avx2; + av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_2d_copy = av1_highbd_dist_wtd_convolve_2d_copy_avx2; + av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_x = av1_highbd_dist_wtd_convolve_x_avx2; + av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_c; + if (flags & HAS_SSE4_1) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_sse4_1; + if (flags & HAS_AVX2) av1_highbd_dist_wtd_convolve_y = av1_highbd_dist_wtd_convolve_y_avx2; + av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z1 = av1_highbd_dr_prediction_z1_avx2; + av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z2 = av1_highbd_dr_prediction_z2_avx2; + av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_c; + if (flags & HAS_AVX2) av1_highbd_dr_prediction_z3 = av1_highbd_dr_prediction_z3_avx2; av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_sse4_1; if (flags & HAS_AVX2) av1_highbd_inv_txfm_add = av1_highbd_inv_txfm_add_avx2; - av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x16 = av1_highbd_inv_txfm_add_16x16_sse4_1; - av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x8 = av1_highbd_inv_txfm_add_16x8_sse4_1; - av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_sse4_1; - if (flags & HAS_AVX2) av1_highbd_inv_txfm_add_32x32 = av1_highbd_inv_txfm_add_32x32_avx2; + av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_16x4 = av1_highbd_inv_txfm_add_16x4_sse4_1; + av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x16 = av1_highbd_inv_txfm_add_4x16_sse4_1; av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x4 = av1_highbd_inv_txfm_add_4x4_sse4_1; - av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_c; - if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x16 = av1_highbd_inv_txfm_add_8x16_sse4_1; + av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_4x8 = av1_highbd_inv_txfm_add_4x8_sse4_1; + av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_c; + if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x4 = av1_highbd_inv_txfm_add_8x4_sse4_1; av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_c; if (flags & HAS_SSE4_1) av1_highbd_inv_txfm_add_8x8 = av1_highbd_inv_txfm_add_8x8_sse4_1; - av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d = av1_highbd_jnt_convolve_2d_avx2; - av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_2d_copy = av1_highbd_jnt_convolve_2d_copy_avx2; - av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_x = av1_highbd_jnt_convolve_x_avx2; - av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_c; - if (flags & HAS_SSE4_1) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_sse4_1; - if (flags & HAS_AVX2) av1_highbd_jnt_convolve_y = av1_highbd_jnt_convolve_y_avx2; av1_highbd_warp_affine = av1_highbd_warp_affine_c; if (flags & HAS_SSE4_1) av1_highbd_warp_affine = av1_highbd_warp_affine_sse4_1; av1_highbd_wiener_convolve_add_src = av1_highbd_wiener_convolve_add_src_c; @@ -522,15 +603,8 @@ static void setup_rtcd_internal(void) av1_inv_txfm_add = av1_inv_txfm_add_c; if (flags & HAS_SSSE3) av1_inv_txfm_add = av1_inv_txfm_add_ssse3; if (flags & HAS_AVX2) av1_inv_txfm_add = av1_inv_txfm_add_avx2; - av1_jnt_convolve_2d = av1_jnt_convolve_2d_c; - if (flags & HAS_SSSE3) av1_jnt_convolve_2d = av1_jnt_convolve_2d_ssse3; - if (flags & HAS_AVX2) av1_jnt_convolve_2d = av1_jnt_convolve_2d_avx2; - av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_2d_copy = av1_jnt_convolve_2d_copy_avx2; - av1_jnt_convolve_x = av1_jnt_convolve_x_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_x = av1_jnt_convolve_x_avx2; - av1_jnt_convolve_y = av1_jnt_convolve_y_sse2; - if (flags & HAS_AVX2) av1_jnt_convolve_y = av1_jnt_convolve_y_avx2; + av1_round_shift_array = av1_round_shift_array_c; + if (flags & HAS_SSE4_1) av1_round_shift_array = av1_round_shift_array_sse4_1; av1_selfguided_restoration = av1_selfguided_restoration_c; if (flags & HAS_SSE4_1) av1_selfguided_restoration = av1_selfguided_restoration_sse4_1; if (flags & HAS_AVX2) av1_selfguided_restoration = av1_selfguided_restoration_avx2; @@ -540,8 +614,17 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE4_1) av1_upsample_intra_edge_high = av1_upsample_intra_edge_high_sse4_1; av1_warp_affine = av1_warp_affine_c; if (flags & HAS_SSE4_1) av1_warp_affine = av1_warp_affine_sse4_1; + if (flags & HAS_AVX2) av1_warp_affine = av1_warp_affine_avx2; av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_sse2; if (flags & HAS_AVX2) av1_wiener_convolve_add_src = av1_wiener_convolve_add_src_avx2; + cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_16bit_to_16bit = cdef_copy_rect8_16bit_to_16bit_avx2; + cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse2; + if (flags & HAS_SSSE3) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_ssse3; + if (flags & HAS_SSE4_1) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_sse4_1; + if (flags & HAS_AVX2) cdef_copy_rect8_8bit_to_16bit = cdef_copy_rect8_8bit_to_16bit_avx2; cdef_filter_block = cdef_filter_block_sse2; if (flags & HAS_SSSE3) cdef_filter_block = cdef_filter_block_ssse3; if (flags & HAS_SSE4_1) cdef_filter_block = cdef_filter_block_sse4_1; @@ -568,22 +651,14 @@ static void setup_rtcd_internal(void) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_c; if (flags & HAS_SSSE3) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_ssse3; if (flags & HAS_AVX2) cfl_get_luma_subsampling_444_lbd = cfl_get_luma_subsampling_444_lbd_avx2; - copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_16bit_to_16bit = copy_rect8_16bit_to_16bit_avx2; - copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse2; - if (flags & HAS_SSSE3) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_ssse3; - if (flags & HAS_SSE4_1) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_sse4_1; - if (flags & HAS_AVX2) copy_rect8_8bit_to_16bit = copy_rect8_8bit_to_16bit_avx2; - get_predict_hbd_fn = get_predict_hbd_fn_c; - if (flags & HAS_SSSE3) get_predict_hbd_fn = get_predict_hbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_hbd_fn = get_predict_hbd_fn_avx2; - get_predict_lbd_fn = get_predict_lbd_fn_c; - if (flags & HAS_SSSE3) get_predict_lbd_fn = get_predict_lbd_fn_ssse3; - if (flags & HAS_AVX2) get_predict_lbd_fn = get_predict_lbd_fn_avx2; - get_subtract_average_fn = get_subtract_average_fn_sse2; - if (flags & HAS_AVX2) get_subtract_average_fn = get_subtract_average_fn_avx2; + cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_hbd_fn = cfl_get_predict_hbd_fn_avx2; + cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_c; + if (flags & HAS_SSSE3) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_ssse3; + if (flags & HAS_AVX2) cfl_get_predict_lbd_fn = cfl_get_predict_lbd_fn_avx2; + cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_sse2; + if (flags & HAS_AVX2) cfl_get_subtract_average_fn = cfl_get_subtract_average_fn_avx2; } #endif diff --git a/media/libaom/sources.mozbuild b/media/libaom/sources.mozbuild index 9396376375..e6ddf7b79d 100644 --- a/media/libaom/sources.mozbuild +++ b/media/libaom/sources.mozbuild @@ -1,4 +1,4 @@ -# This file is automatically generated. Do not edit. +# This file is generated. Do not edit. files = { 'ARM_EXPORTS': [ @@ -34,13 +34,14 @@ files = { '/media/libaom/src/aom_dsp/arm/intrapred_neon.c', '/media/libaom/src/aom_dsp/arm/loopfilter_neon.c', '/media/libaom/src/aom_dsp/arm/subtract_neon.c', + '/media/libaom/src/aom_dsp/avg.c', '/media/libaom/src/aom_dsp/binary_codes_reader.c', + '/media/libaom/src/aom_dsp/bitreader.c', '/media/libaom/src/aom_dsp/bitreader_buffer.c', '/media/libaom/src/aom_dsp/bitwriter_buffer.c', '/media/libaom/src/aom_dsp/blend_a64_hmask.c', '/media/libaom/src/aom_dsp/blend_a64_mask.c', '/media/libaom/src/aom_dsp/blend_a64_vmask.c', - '/media/libaom/src/aom_dsp/daalaboolreader.c', '/media/libaom/src/aom_dsp/entcode.c', '/media/libaom/src/aom_dsp/entdec.c', '/media/libaom/src/aom_dsp/fft.c', @@ -109,8 +110,8 @@ files = { '/media/libaom/src/av1/decoder/decoder.c', '/media/libaom/src/av1/decoder/decodetxb.c', '/media/libaom/src/av1/decoder/detokenize.c', - '/media/libaom/src/av1/decoder/dthread.c', '/media/libaom/src/av1/decoder/obu.c', + '/media/libaom/src/av1/encoder/arm/neon/av1_error_neon.c', '/media/libaom/src/av1/encoder/arm/neon/quantize_neon.c', ], 'GENERIC_EXPORTS': [ @@ -140,13 +141,14 @@ files = { '/media/libaom/src/aom/src/aom_integer.c', '/media/libaom/src/aom_dsp/aom_convolve.c', '/media/libaom/src/aom_dsp/aom_dsp_rtcd.c', + '/media/libaom/src/aom_dsp/avg.c', '/media/libaom/src/aom_dsp/binary_codes_reader.c', + '/media/libaom/src/aom_dsp/bitreader.c', '/media/libaom/src/aom_dsp/bitreader_buffer.c', '/media/libaom/src/aom_dsp/bitwriter_buffer.c', '/media/libaom/src/aom_dsp/blend_a64_hmask.c', '/media/libaom/src/aom_dsp/blend_a64_mask.c', '/media/libaom/src/aom_dsp/blend_a64_vmask.c', - '/media/libaom/src/aom_dsp/daalaboolreader.c', '/media/libaom/src/aom_dsp/entcode.c', '/media/libaom/src/aom_dsp/entdec.c', '/media/libaom/src/aom_dsp/fft.c', @@ -202,7 +204,6 @@ files = { '/media/libaom/src/av1/decoder/decoder.c', '/media/libaom/src/av1/decoder/decodetxb.c', '/media/libaom/src/av1/decoder/detokenize.c', - '/media/libaom/src/av1/decoder/dthread.c', '/media/libaom/src/av1/decoder/obu.c', ], 'IA32_EXPORTS': [ @@ -232,13 +233,14 @@ files = { '/media/libaom/src/aom/src/aom_integer.c', '/media/libaom/src/aom_dsp/aom_convolve.c', '/media/libaom/src/aom_dsp/aom_dsp_rtcd.c', + '/media/libaom/src/aom_dsp/avg.c', '/media/libaom/src/aom_dsp/binary_codes_reader.c', + '/media/libaom/src/aom_dsp/bitreader.c', '/media/libaom/src/aom_dsp/bitreader_buffer.c', '/media/libaom/src/aom_dsp/bitwriter_buffer.c', '/media/libaom/src/aom_dsp/blend_a64_hmask.c', '/media/libaom/src/aom_dsp/blend_a64_mask.c', '/media/libaom/src/aom_dsp/blend_a64_vmask.c', - '/media/libaom/src/aom_dsp/daalaboolreader.c', '/media/libaom/src/aom_dsp/entcode.c', '/media/libaom/src/aom_dsp/entdec.c', '/media/libaom/src/aom_dsp/fft.c', @@ -251,11 +253,14 @@ files = { '/media/libaom/src/aom_dsp/x86/aom_high_subpixel_8t_sse2.asm', '/media/libaom/src/aom_dsp/x86/aom_high_subpixel_bilinear_sse2.asm', '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_intrin_avx2.c', + '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_intrin_sse2.c', '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_intrin_ssse3.c', '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_sse2.asm', '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_ssse3.asm', '/media/libaom/src/aom_dsp/x86/aom_subpixel_bilinear_sse2.asm', '/media/libaom/src/aom_dsp/x86/aom_subpixel_bilinear_ssse3.asm', + '/media/libaom/src/aom_dsp/x86/avg_intrin_avx2.c', + '/media/libaom/src/aom_dsp/x86/avg_intrin_sse2.c', '/media/libaom/src/aom_dsp/x86/blend_a64_hmask_sse4.c', '/media/libaom/src/aom_dsp/x86/blend_a64_mask_avx2.c', '/media/libaom/src/aom_dsp/x86/blend_a64_mask_sse4.c', @@ -263,14 +268,15 @@ files = { '/media/libaom/src/aom_dsp/x86/fft_avx2.c', '/media/libaom/src/aom_dsp/x86/fft_sse2.c', '/media/libaom/src/aom_dsp/x86/highbd_convolve_avx2.c', + '/media/libaom/src/aom_dsp/x86/highbd_convolve_sse2.c', '/media/libaom/src/aom_dsp/x86/highbd_convolve_ssse3.c', + '/media/libaom/src/aom_dsp/x86/highbd_intrapred_asm_sse2.asm', '/media/libaom/src/aom_dsp/x86/highbd_intrapred_sse2.c', - '/media/libaom/src/aom_dsp/x86/highbd_intrapred_sse2_asm.asm', '/media/libaom/src/aom_dsp/x86/highbd_loopfilter_avx2.c', '/media/libaom/src/aom_dsp/x86/highbd_loopfilter_sse2.c', + '/media/libaom/src/aom_dsp/x86/intrapred_asm_sse2.asm', '/media/libaom/src/aom_dsp/x86/intrapred_avx2.c', '/media/libaom/src/aom_dsp/x86/intrapred_sse2.c', - '/media/libaom/src/aom_dsp/x86/intrapred_sse2_asm.asm', '/media/libaom/src/aom_dsp/x86/intrapred_ssse3.c', '/media/libaom/src/aom_dsp/x86/inv_wht_sse2.asm', '/media/libaom/src/aom_dsp/x86/loopfilter_sse2.c', @@ -325,7 +331,6 @@ files = { '/media/libaom/src/av1/common/warped_motion.c', '/media/libaom/src/av1/common/x86/av1_convolve_horiz_rs_sse4.c', '/media/libaom/src/av1/common/x86/av1_convolve_scale_sse4.c', - '/media/libaom/src/av1/common/x86/av1_highbd_convolve_sse4.c', '/media/libaom/src/av1/common/x86/av1_inv_txfm_avx2.c', '/media/libaom/src/av1/common/x86/av1_inv_txfm_ssse3.c', '/media/libaom/src/av1/common/x86/av1_txfm_sse4.c', @@ -357,6 +362,8 @@ files = { '/media/libaom/src/av1/common/x86/reconinter_ssse3.c', '/media/libaom/src/av1/common/x86/selfguided_avx2.c', '/media/libaom/src/av1/common/x86/selfguided_sse4.c', + '/media/libaom/src/av1/common/x86/warp_plane_avx2.c', + '/media/libaom/src/av1/common/x86/warp_plane_sse2.c', '/media/libaom/src/av1/common/x86/warp_plane_sse4.c', '/media/libaom/src/av1/common/x86/wiener_convolve_avx2.c', '/media/libaom/src/av1/common/x86/wiener_convolve_sse2.c', @@ -365,7 +372,6 @@ files = { '/media/libaom/src/av1/decoder/decoder.c', '/media/libaom/src/av1/decoder/decodetxb.c', '/media/libaom/src/av1/decoder/detokenize.c', - '/media/libaom/src/av1/decoder/dthread.c', '/media/libaom/src/av1/decoder/obu.c', ], 'X64_EXPORTS': [ @@ -395,13 +401,14 @@ files = { '/media/libaom/src/aom/src/aom_integer.c', '/media/libaom/src/aom_dsp/aom_convolve.c', '/media/libaom/src/aom_dsp/aom_dsp_rtcd.c', + '/media/libaom/src/aom_dsp/avg.c', '/media/libaom/src/aom_dsp/binary_codes_reader.c', + '/media/libaom/src/aom_dsp/bitreader.c', '/media/libaom/src/aom_dsp/bitreader_buffer.c', '/media/libaom/src/aom_dsp/bitwriter_buffer.c', '/media/libaom/src/aom_dsp/blend_a64_hmask.c', '/media/libaom/src/aom_dsp/blend_a64_mask.c', '/media/libaom/src/aom_dsp/blend_a64_vmask.c', - '/media/libaom/src/aom_dsp/daalaboolreader.c', '/media/libaom/src/aom_dsp/entcode.c', '/media/libaom/src/aom_dsp/entdec.c', '/media/libaom/src/aom_dsp/fft.c', @@ -414,11 +421,14 @@ files = { '/media/libaom/src/aom_dsp/x86/aom_high_subpixel_8t_sse2.asm', '/media/libaom/src/aom_dsp/x86/aom_high_subpixel_bilinear_sse2.asm', '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_intrin_avx2.c', + '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_intrin_sse2.c', '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_intrin_ssse3.c', '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_sse2.asm', '/media/libaom/src/aom_dsp/x86/aom_subpixel_8t_ssse3.asm', '/media/libaom/src/aom_dsp/x86/aom_subpixel_bilinear_sse2.asm', '/media/libaom/src/aom_dsp/x86/aom_subpixel_bilinear_ssse3.asm', + '/media/libaom/src/aom_dsp/x86/avg_intrin_avx2.c', + '/media/libaom/src/aom_dsp/x86/avg_intrin_sse2.c', '/media/libaom/src/aom_dsp/x86/blend_a64_hmask_sse4.c', '/media/libaom/src/aom_dsp/x86/blend_a64_mask_avx2.c', '/media/libaom/src/aom_dsp/x86/blend_a64_mask_sse4.c', @@ -426,14 +436,15 @@ files = { '/media/libaom/src/aom_dsp/x86/fft_avx2.c', '/media/libaom/src/aom_dsp/x86/fft_sse2.c', '/media/libaom/src/aom_dsp/x86/highbd_convolve_avx2.c', + '/media/libaom/src/aom_dsp/x86/highbd_convolve_sse2.c', '/media/libaom/src/aom_dsp/x86/highbd_convolve_ssse3.c', + '/media/libaom/src/aom_dsp/x86/highbd_intrapred_asm_sse2.asm', '/media/libaom/src/aom_dsp/x86/highbd_intrapred_sse2.c', - '/media/libaom/src/aom_dsp/x86/highbd_intrapred_sse2_asm.asm', '/media/libaom/src/aom_dsp/x86/highbd_loopfilter_avx2.c', '/media/libaom/src/aom_dsp/x86/highbd_loopfilter_sse2.c', + '/media/libaom/src/aom_dsp/x86/intrapred_asm_sse2.asm', '/media/libaom/src/aom_dsp/x86/intrapred_avx2.c', '/media/libaom/src/aom_dsp/x86/intrapred_sse2.c', - '/media/libaom/src/aom_dsp/x86/intrapred_sse2_asm.asm', '/media/libaom/src/aom_dsp/x86/intrapred_ssse3.c', '/media/libaom/src/aom_dsp/x86/inv_wht_sse2.asm', '/media/libaom/src/aom_dsp/x86/loopfilter_sse2.c', @@ -487,7 +498,6 @@ files = { '/media/libaom/src/av1/common/warped_motion.c', '/media/libaom/src/av1/common/x86/av1_convolve_horiz_rs_sse4.c', '/media/libaom/src/av1/common/x86/av1_convolve_scale_sse4.c', - '/media/libaom/src/av1/common/x86/av1_highbd_convolve_sse4.c', '/media/libaom/src/av1/common/x86/av1_inv_txfm_avx2.c', '/media/libaom/src/av1/common/x86/av1_inv_txfm_ssse3.c', '/media/libaom/src/av1/common/x86/av1_txfm_sse4.c', @@ -519,6 +529,8 @@ files = { '/media/libaom/src/av1/common/x86/reconinter_ssse3.c', '/media/libaom/src/av1/common/x86/selfguided_avx2.c', '/media/libaom/src/av1/common/x86/selfguided_sse4.c', + '/media/libaom/src/av1/common/x86/warp_plane_avx2.c', + '/media/libaom/src/av1/common/x86/warp_plane_sse2.c', '/media/libaom/src/av1/common/x86/warp_plane_sse4.c', '/media/libaom/src/av1/common/x86/wiener_convolve_avx2.c', '/media/libaom/src/av1/common/x86/wiener_convolve_sse2.c', @@ -527,7 +539,6 @@ files = { '/media/libaom/src/av1/decoder/decoder.c', '/media/libaom/src/av1/decoder/decodetxb.c', '/media/libaom/src/av1/decoder/detokenize.c', - '/media/libaom/src/av1/decoder/dthread.c', '/media/libaom/src/av1/decoder/obu.c', ], } |