diff --git a/src/nouveau/headers/nvidia/video/nvdec_drv.h b/src/nouveau/headers/nvidia/video/nvdec_drv.h new file mode 100644 index 00000000000..bbfbeda0ffd --- /dev/null +++ b/src/nouveau/headers/nvidia/video/nvdec_drv.h @@ -0,0 +1,1846 @@ +/* + * SPDX-FileCopyrightText: Copyright (c) 1993-2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved. + * SPDX-License-Identifier: MIT + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef __NVDEC_DRV_H_ +#define __NVDEC_DRV_H_ + +// TODO: Many fields can be converted to bitfields to save memory BW +// TODO: Revisit reserved fields for proper alignment and memory savings + +/////////////////////////////////////////////////////////////////////////////// +// NVDEC(MSDEC 5) is a single engine solution, and seperates into VLD, MV, IQT, +// MCFETCH, MC, MCC, REC, DBF, DFBFDMA, HIST etc unit. +// The class(driver to HW) can mainly seperate into VLD parser +// and Decoder part to be consistent with original design. And +// the sequence level info usally set in VLD part. Later codec like +// VP8 won't name in this way. +// MSVLD: Multi-Standard VLD parser. +// +#define ALIGN_UP(v, n) (((v) + ((n)-1)) &~ ((n)-1)) +#define NVDEC_ALIGN(value) ALIGN_UP(value,256) // Align to 256 bytes +#define NVDEC_MAX_MPEG2_SLICE 65536 // at 4096*4096, macroblock count = 65536, 1 macroblock per slice + +#define NVDEC_CODEC_MPEG1 0 +#define NVDEC_CODEC_MPEG2 1 +#define NVDEC_CODEC_VC1 2 +#define NVDEC_CODEC_H264 3 +#define NVDEC_CODEC_MPEG4 4 +#define NVDEC_CODEC_DIVX NVDEC_CODEC_MPEG4 +#define NVDEC_CODEC_VP8 5 +#define NVDEC_CODEC_HEVC 7 +#define NVDEC_CODEC_VP9 9 +#define NVDEC_CODEC_HEVC_PARSER 12 +#define NVDEC_CODEC_AV1 10 + +// AES encryption +enum +{ + AES128_NONE = 0x0, + AES128_CTR = 0x1, + AES128_CBC, + AES128_ECB, + AES128_OFB, + AES128_CTR_LSB16B, + AES128_CLR_AS_ENCRYPT, + AES128_RESERVED = 0x7 +}; + +enum +{ + AES128_CTS_DISABLE = 0x0, + AES128_CTS_ENABLE = 0x1 +}; + +enum +{ + AES128_PADDING_NONE = 0x0, + AES128_PADDING_CARRY_OVER, + AES128_PADDING_RFC2630, + AES128_PADDING_RESERVED = 0x7 +}; + +typedef enum +{ + ENCR_MODE_CTR64 = 0, + ENCR_MODE_CBC = 1, + ENCR_MODE_ECB = 2, + ENCR_MODE_ECB_PARTIAL = 3, + ENCR_MODE_CBC_PARTIAL = 4, + ENCR_MODE_CLEAR_INTO_VPR = 5, // used for clear stream decoding into VPR. + ENCR_MODE_FORCE_INTO_VPR = 6, // used to force decode output into VPR. +} ENCR_MODE; + +// drm_mode configuration +// +// Bit 0:2 AES encryption mode +// Bit 3 CTS (CipherTextStealing) enable/disable +// Bit 4:6 Padding type +// Bit 7:7 Unwrap key enable/disable + +#define AES_MODE_MASK 0x7 +#define AES_CTS_MASK 0x1 +#define AES_PADDING_TYPE_MASK 0x7 +#define AES_UNWRAP_KEY_MASK 0x1 + +#define AES_MODE_SHIFT 0 +#define AES_CTS_SHIFT 3 +#define AES_PADDING_TYPE_SHIFT 4 +#define AES_UNWRAP_KEY_SHIFT 7 + +#define AES_SET_FLAG(M, C, P) ((M & AES_MODE_MASK) << AES_MODE_SHIFT) | \ + ((C & AES_CTS_MASK) << AES_CTS_SHIFT) | \ + ((P & AES_PADDING_TYPE_MASK) << AES_PADDING_TYPE_SHIFT) + +#define AES_GET_FLAG(V, F) ((V & ((AES_##F##_MASK) <<(AES_##F##_SHIFT))) >> (AES_##F##_SHIFT)) + +#define DRM_MODE_MASK 0x7f // Bits 0:6 (0:2 -> AES_MODE, 3 -> AES_CTS, 4:6 -> AES_PADDING_TYPE) +#define AES_GET_DRM_MODE(V) (V & DRM_MODE_MASK) + +enum { DRM_MS_PIFF_CTR = AES_SET_FLAG(AES128_CTR, AES128_CTS_DISABLE, AES128_PADDING_CARRY_OVER) }; +enum { DRM_MS_PIFF_CBC = AES_SET_FLAG(AES128_CBC, AES128_CTS_DISABLE, AES128_PADDING_NONE) }; +enum { DRM_MARLIN_CTR = AES_SET_FLAG(AES128_CTR, AES128_CTS_DISABLE, AES128_PADDING_NONE) }; +enum { DRM_MARLIN_CBC = AES_SET_FLAG(AES128_CBC, AES128_CTS_DISABLE, AES128_PADDING_RFC2630) }; +enum { DRM_WIDEVINE = AES_SET_FLAG(AES128_CBC, AES128_CTS_ENABLE, AES128_PADDING_NONE) }; +enum { DRM_WIDEVINE_CTR = AES_SET_FLAG(AES128_CTR, AES128_CTS_DISABLE, AES128_PADDING_CARRY_OVER) }; +enum { DRM_ULTRA_VIOLET = AES_SET_FLAG(AES128_CTR_LSB16B, AES128_CTS_DISABLE, AES128_PADDING_NONE) }; +enum { DRM_NONE = AES_SET_FLAG(AES128_NONE, AES128_CTS_DISABLE, AES128_PADDING_NONE) }; +enum { DRM_CLR_AS_ENCRYPT = AES_SET_FLAG(AES128_CLR_AS_ENCRYPT, AES128_CTS_DISABLE, AES128_PADDING_NONE)}; + +// SSM entry structure +typedef struct _nvdec_ssm_s { + unsigned int bytes_of_protected_data;//bytes of protected data, follows bytes_of_clear_data. Note: When padding is enabled, it does not include the padding_bytes (1~15), which can be derived by "(16-(bytes_of_protected_data&0xF))&0xF" + unsigned int bytes_of_clear_data:16; //bytes of clear data, located before bytes_of_protected_data + unsigned int skip_byte_blk : 4; //valid when (entry_type==0 && mode = 1) + unsigned int crypt_byte_blk : 4; //valid when (entry_type==0 && mode = 1) + unsigned int skip : 1; //whether this SSM entry should be skipped or not + unsigned int last : 1; //whether this SSM entry is the last one for the whole decoding frame + unsigned int pad : 1; //valid when (entry_type==0 && mode==0 && AES_PADDING_TYPE==AES128_PADDING_RFC2630), 0 for pad_end, 1 for pad_begin + unsigned int mode : 1; //0 for normal mode, 1 for pattern mode + unsigned int entry_type : 1; //0 for DATA, 1 for IV + unsigned int reserved : 3; +} nvdec_ssm_s; /* SubSampleMap, 8bytes */ + +// PASS2 OTF extension structure for SSM support, not exist in nvdec_mpeg4_pic_s (as MPEG4 OTF SW-DRM is not supported yet) +typedef struct _nvdec_pass2_otf_ext_s { + unsigned int ssm_entry_num :16; //specifies how many SSM entries (each in unit of 8 bytes) existed in SET_SUB_SAMPLE_MAP_OFFSET surface + unsigned int ssm_iv_num :16; //specifies how many SSM IV (each in unit of 16 bytes) existed in SET_SUB_SAMPLE_MAP_IV_OFFSET surface + unsigned int real_stream_length; //the real stream length, which is the bitstream length EMD/VLD will get after whole frame SSM processing, sum up of "clear+protected" bytes in SSM entries and removing "non_slice_data/skip". + unsigned int non_slice_data :16; //specifies the first many bytes needed to skip, includes only those of "clear+protected" bytes ("padding" bytes excluded) + unsigned int drm_mode : 7; + unsigned int reserved : 9; +} nvdec_pass2_otf_ext_s; /* 12bytes */ + + +//NVDEC5.0 low latency decoding (partial stream kickoff without context switch), method will reuse HevcSetSliceInfoBufferOffset. +typedef struct _nvdec_substream_entry_s { + unsigned int substream_start_offset; //substream byte start offset to bitstream base address + unsigned int substream_length; //subsream length in byte + unsigned int substream_first_tile_idx : 8; //the first tile index(raster scan in frame) of this substream,max is 255 + unsigned int substream_last_tile_idx : 8; //the last tile index(raster scan in frame) of this substream, max is 255 + unsigned int last_substream_entry_in_frame : 1; //this entry is the last substream entry of this frame + unsigned int reserved : 15; +} nvdec_substream_entry_s;/*low latency without context switch substream entry map,12bytes*/ + + +// GIP + +/* tile border coefficients of filter */ +#define GIP_ASIC_VERT_FILTER_RAM_SIZE 16 /* bytes per pixel */ + +/* BSD control data of current picture at tile border + * 11 * 128 bits per 4x4 tile = 128/(8*4) bytes per row */ +#define GIP_ASIC_BSD_CTRL_RAM_SIZE 4 /* bytes per row */ + +/* 8 dc + 8 to boundary + 6*16 + 2*6*64 + 2*64 -> 63 * 16 bytes */ +#define GIP_ASIC_SCALING_LIST_SIZE (16*64) + +/* tile border coefficients of filter */ +#define GIP_ASIC_VERT_SAO_RAM_SIZE 16 /* bytes per pixel */ + +/* max number of tiles times width and height (2 bytes each), + * rounding up to next 16 bytes boundary + one extra 16 byte + * chunk (HW guys wanted to have this) */ +#define GIP_ASIC_TILE_SIZE ((20*22*2*2+16+15) & ~0xF) + +/* Segment map uses 32 bytes / CTB */ +#define GIP_ASIC_VP9_CTB_SEG_SIZE 32 + +// HEVC Filter FG buffer +#define HEVC_DBLK_TOP_SIZE_IN_SB16 ALIGN_UP(632, 128) // ctb16 + 444 +#define HEVC_DBLK_TOP_BUF_SIZE(w) NVDEC_ALIGN( (ALIGN_UP(w,16)/16 + 2) * HEVC_DBLK_TOP_SIZE_IN_SB16) // 8K: 1285*256 + +#define HEVC_DBLK_LEFT_SIZE_IN_SB16 ALIGN_UP(506, 128) // ctb16 + 444 +#define HEVC_DBLK_LEFT_BUF_SIZE(h) NVDEC_ALIGN( (ALIGN_UP(h,16)/16 + 2) * HEVC_DBLK_LEFT_SIZE_IN_SB16) // 8K: 1028*256 + +#define HEVC_SAO_LEFT_SIZE_IN_SB16 ALIGN_UP(713, 128) // ctb16 + 444 +#define HEVC_SAO_LEFT_BUF_SIZE(h) NVDEC_ALIGN( (ALIGN_UP(h,16)/16 + 2) * HEVC_SAO_LEFT_SIZE_IN_SB16) // 8K: 1542*256 + +// VP9 Filter FG buffer +#define VP9_DBLK_TOP_SIZE_IN_SB64 ALIGN_UP(2000, 128) // 420 +#define VP9_DBLK_TOP_BUF_SIZE(w) NVDEC_ALIGN( (ALIGN_UP(w,64)/64 + 2) * VP9_DBLK_TOP_SIZE_IN_SB64) // 8K: 1040*256 + +#define VP9_DBLK_LEFT_SIZE_IN_SB64 ALIGN_UP(1600, 128) // 420 +#define VP9_DBLK_LEFT_BUF_SIZE(h) NVDEC_ALIGN( (ALIGN_UP(h,64)/64 + 2) * VP9_DBLK_LEFT_SIZE_IN_SB64) // 8K: 845*256 + +// VP9 Hint Dump Buffer +#define VP9_HINT_DUMP_SIZE_IN_SB64 ((64*64)/(4*4)*8) // 8 bytes per CU, 256 CUs(2048 bytes) per SB64 +#define VP9_HINT_DUMP_SIZE(w, h) NVDEC_ALIGN(VP9_HINT_DUMP_SIZE_IN_SB64*((w+63)/64)*((h+63)/64)) + +// used for ecdma debug +typedef struct _nvdec_ecdma_config_s +{ + unsigned int ecdma_enable; // enable/disable ecdma + unsigned short ecdma_blk_x_src; // src start position x , it's 64x aligned + unsigned short ecdma_blk_y_src; // src start position y , it's 8x aligned + unsigned short ecdma_blk_x_dst; // dst start position x , it's 64x aligned + unsigned short ecdma_blk_y_dst; // dst start position y , it's 8x aligned + unsigned short ref_pic_idx; // ref(src) picture index , used to derived source picture base address + unsigned short boundary0_top; // src insided tile/partition region top boundary + unsigned short boundary0_bottom; // src insided tile/partition region bottom boundary + unsigned short boundary1_left; // src insided tile/partition region left boundary + unsigned short boundary1_right; // src insided tile/partition region right boundary + unsigned char blk_copy_flag; // blk_copy enable flag. + // if it's 1 ,ctb_size ==3,ecdma_blk_x_src == boundary1_left and ecdma_blk_y_src == boundary0_top ; + // if it's 0 ,ecdma_blk_x_src == ecdma_blk_x_dst and ecdma_blk_y_src == ecdma_blk_y_dst; + unsigned char ctb_size; // ctb_size .0:64x64,1:32x32,2:16x16,3:8x8 +} nvdec_ecdma_config_s; + +typedef struct _nvdec_status_hevc_s +{ + unsigned int frame_status_intra_cnt; //Intra block counter, in unit of 8x8 block, IPCM block included + unsigned int frame_status_inter_cnt; //Inter block counter, in unit of 8x8 block, SKIP block included + unsigned int frame_status_skip_cnt; //Skip block counter, in unit of 4x4 block, blocks having NO/ZERO texture/coeff data + unsigned int frame_status_fwd_mvx_cnt; //ABS sum of forward MVx, one 14bit MVx(integer) per 4x4 block + unsigned int frame_status_fwd_mvy_cnt; //ABS sum of forward MVy, one 14bit MVy(integer) per 4x4 block + unsigned int frame_status_bwd_mvx_cnt; //ABS sum of backward MVx, one 14bit MVx(integer) per 4x4 block + unsigned int frame_status_bwd_mvy_cnt; //ABS sum of backward MVy, one 14bit MVy(integer) per 4x4 block + unsigned int error_ctb_pos; //[15:0] error ctb position in Y direction, [31:16] error ctb position in X direction + unsigned int error_slice_pos; //[15:0] error slice position in Y direction, [31:16] error slice position in X direction +} nvdec_status_hevc_s; + +typedef struct _nvdec_status_vp9_s +{ + unsigned int frame_status_intra_cnt; //Intra block counter, in unit of 8x8 block, IPCM block included + unsigned int frame_status_inter_cnt; //Inter block counter, in unit of 8x8 block, SKIP block included + unsigned int frame_status_skip_cnt; //Skip block counter, in unit of 4x4 block, blocks having NO/ZERO texture/coeff data + unsigned int frame_status_fwd_mvx_cnt; //ABS sum of forward MVx, one 14bit MVx(integer) per 4x4 block + unsigned int frame_status_fwd_mvy_cnt; //ABS sum of forward MVy, one 14bit MVy(integer) per 4x4 block + unsigned int frame_status_bwd_mvx_cnt; //ABS sum of backward MVx, one 14bit MVx(integer) per 4x4 block + unsigned int frame_status_bwd_mvy_cnt; //ABS sum of backward MVy, one 14bit MVy(integer) per 4x4 block + unsigned int error_ctb_pos; //[15:0] error ctb position in Y direction, [31:16] error ctb position in X direction + unsigned int error_slice_pos; //[15:0] error slice position in Y direction, [31:16] error slice position in X direction +} nvdec_status_vp9_s; + +typedef struct _nvdec_status_s +{ + unsigned int mbs_correctly_decoded; // total numers of correctly decoded macroblocks + unsigned int mbs_in_error; // number of error macroblocks. + unsigned int cycle_count; // total cycles taken for execute. read from PERF_DECODE_FRAME_V register + unsigned int error_status; // report error if any + union + { + nvdec_status_hevc_s hevc; + nvdec_status_vp9_s vp9; + }; + unsigned int slice_header_error_code; // report error in slice header + +} nvdec_status_s; + +// per 16x16 block, used in hevc/vp9 surface of SetExternalMVBufferOffset when error_external_mv_en = 1 +typedef struct _external_mv_s +{ + int mvx : 14; //integrate pixel precision + int mvy : 14; //integrate pixel precision + unsigned int refidx : 4; +} external_mv_s; + +// HEVC +typedef struct _nvdec_hevc_main10_444_ext_s +{ + unsigned int transformSkipRotationEnableFlag : 1; //sps extension for transform_skip_rotation_enabled_flag + unsigned int transformSkipContextEnableFlag : 1; //sps extension for transform_skip_context_enabled_flag + unsigned int intraBlockCopyEnableFlag :1; //sps intraBlockCopyEnableFlag, always 0 before spec define it + unsigned int implicitRdpcmEnableFlag : 1; //sps implicit_rdpcm_enabled_flag + unsigned int explicitRdpcmEnableFlag : 1; //sps explicit_rdpcm_enabled_flag + unsigned int extendedPrecisionProcessingFlag : 1; //sps extended_precision_processing_flag,always 0 in current profile + unsigned int intraSmoothingDisabledFlag : 1; //sps intra_smoothing_disabled_flag + unsigned int highPrecisionOffsetsEnableFlag :1; //sps high_precision_offsets_enabled_flag + unsigned int fastRiceAdaptationEnableFlag: 1; //sps fast_rice_adaptation_enabled_flag + unsigned int cabacBypassAlignmentEnableFlag : 1; //sps cabac_bypass_alignment_enabled_flag, always 0 in current profile + unsigned int sps_444_extension_reserved : 22; //sps reserve for future extension + + unsigned int log2MaxTransformSkipSize : 4 ; //pps extension log2_max_transform_skip_block_size_minus2, 0...5 + unsigned int crossComponentPredictionEnableFlag: 1; //pps cross_component_prediction_enabled_flag + unsigned int chromaQpAdjustmentEnableFlag:1; //pps chroma_qp_adjustment_enabled_flag + unsigned int diffCuChromaQpAdjustmentDepth:2; //pps diff_cu_chroma_qp_adjustment_depth, 0...3 + unsigned int chromaQpAdjustmentTableSize:3; //pps chroma_qp_adjustment_table_size_minus1+1, 1...6 + unsigned int log2SaoOffsetScaleLuma:3; //pps log2_sao_offset_scale_luma, max(0,bitdepth-10),maxBitdepth 16 for future. + unsigned int log2SaoOffsetScaleChroma: 3; //pps log2_sao_offset_scale_chroma + unsigned int pps_444_extension_reserved : 15; //pps reserved + char cb_qp_adjustment[6]; //-[12,+12] + char cr_qp_adjustment[6]; //-[12,+12] + unsigned int HevcFltAboveOffset; // filter above offset respect to filter buffer, 256 bytes unit + unsigned int HevcSaoAboveOffset; // sao above offset respect to filter buffer, 256 bytes unit +} nvdec_hevc_main10_444_ext_s; + +typedef struct _nvdec_hevc_pic_v1_s +{ + // New fields + //hevc main10 444 extensions + nvdec_hevc_main10_444_ext_s hevc_main10_444_ext; + + //HEVC skip bytes from beginning setting for secure + //it is different to the sw_hdr_skip_length who skips the middle of stream of + //the slice header which is parsed by driver + unsigned int sw_skip_start_length : 14; + unsigned int external_ref_mem_dis : 1; + unsigned int error_recovery_start_pos : 2; //0: from start of frame, 1: from start of slice segment, 2: from error detected ctb, 3: reserved + unsigned int error_external_mv_en : 1; + unsigned int reserved0 : 14; + // Reserved bits padding +} nvdec_hevc_pic_v1_s; + +//No versioning in structure: NVDEC2 (T210 and GM206) +//version v1 : NVDEC3 (T186 and GP100) +//version v2 : NVDEC3.1 (GP10x) + +typedef struct _nvdec_hevc_pic_v2_s +{ + // mv-hevc field + unsigned int mv_hevc_enable :1; + unsigned int nuh_layer_id :6; + unsigned int default_ref_layers_active_flag :1; + unsigned int NumDirectRefLayers :6; + unsigned int max_one_active_ref_layer_flag :1; + unsigned int NumActiveRefLayerPics :6; + unsigned int poc_lsb_not_present_flag :1; + unsigned int reserved0 :10; +} nvdec_hevc_pic_v2_s; + +typedef struct _nvdec_hevc_pic_v3_s +{ + // slice level decoding + unsigned int slice_decoding_enable:1;//1: enable slice level decoding + unsigned int slice_ec_enable:1; //1: enable slice error concealment. When slice_ec_enable=1,slice_decoding_enable must be 1; + unsigned int slice_ec_mv_type:2; //0: zero mv; 1: co-located mv; 2: external mv; + unsigned int err_detected_sw:1; //1: indicate sw/driver has detected error already in frame kick mode + unsigned int slice_ec_slice_type:2; //0: B slice; 1: P slice ; others: reserved + unsigned int slice_strm_recfg_en:1; //enable slice bitstream re-configure or not ; + unsigned int reserved:24; + unsigned int HevcSliceEdgeOffset;// slice edge buffer offset which repsect to filter buffer ,256 bytes as one unit +}nvdec_hevc_pic_v3_s; + +typedef struct _nvdec_hevc_pic_s +{ + //The key/IV addr must be 128bit alignment + unsigned int wrapped_session_key[4]; //session keys + unsigned int wrapped_content_key[4]; //content keys + unsigned int initialization_vector[4]; //Ctrl64 initial vector + // hevc_bitstream_data_info + unsigned int stream_len; // stream length in one frame + unsigned int enable_encryption; // flag to enable/disable encryption + unsigned int key_increment : 6; // added to content key after unwrapping + unsigned int encryption_mode : 4; + unsigned int key_slot_index : 4; + unsigned int ssm_en : 1; + unsigned int enable_histogram : 1; // histogram stats output enable + unsigned int enable_substream_decoding: 1; //frame substream kickoff without context switch + unsigned int reserved0 :15; + + // Driver may or may not use based upon need. + // If 0 then default value of 1<<27 = 298ms @ 450MHz will be used in ucode. + // Driver can send this value based upon resolution using the formula: + // gptimer_timeout_value = 3 * (cycles required for one frame) + unsigned int gptimer_timeout_value; + + // general + unsigned char tileformat : 2 ; // 0: TBL; 1: KBL; 2: Tile16x16 + unsigned char gob_height : 3 ; // Set GOB height, 0: GOB_2, 1: GOB_4, 2: GOB_8, 3: GOB_16, 4: GOB_32 (NVDEC3 onwards) + unsigned char reserverd_surface_format : 3 ; + unsigned char sw_start_code_e; // 0: stream doesn't contain start codes,1: stream contains start codes + unsigned char disp_output_mode; // 0: Rec.709 8 bit, 1: Rec.709 10 bit, 2: Rec.709 10 bits -> 8 bit, 3: Rec.2020 10 bit -> 8 bit + unsigned char reserved1; + unsigned int framestride[2]; // frame buffer stride for luma and chroma + unsigned int colMvBuffersize; // collocated MV buffer size of one picture ,256 bytes unit + unsigned int HevcSaoBufferOffset; // sao buffer offset respect to filter buffer ,256 bytes unit . + unsigned int HevcBsdCtrlOffset; // bsd buffer offset respect to filter buffer ,256 bytes unit . + // sps + unsigned short pic_width_in_luma_samples; // :15, 48(?)..16384, multiple of 8 (48 is smallest width supported by NVDEC for CTU size 16x16) + unsigned short pic_height_in_luma_samples; // :15, 8..16384, multiple of 8 + unsigned int chroma_format_idc : 4; // always 1 (=4:2:0) + unsigned int bit_depth_luma : 4; // 8..12 + unsigned int bit_depth_chroma : 4; + unsigned int log2_min_luma_coding_block_size : 4; // 3..6 + unsigned int log2_max_luma_coding_block_size : 4; // 3..6 + unsigned int log2_min_transform_block_size : 4; // 2..5 + unsigned int log2_max_transform_block_size : 4; // 2..5 + unsigned int reserved2 : 4; + + unsigned int max_transform_hierarchy_depth_inter : 3; // 0..4 + unsigned int max_transform_hierarchy_depth_intra : 3; // 0..4 + unsigned int scalingListEnable : 1; // + unsigned int amp_enable_flag : 1; // + unsigned int sample_adaptive_offset_enabled_flag : 1; // + unsigned int pcm_enabled_flag : 1; // + unsigned int pcm_sample_bit_depth_luma : 4; // + unsigned int pcm_sample_bit_depth_chroma : 4; + unsigned int log2_min_pcm_luma_coding_block_size : 4; // + unsigned int log2_max_pcm_luma_coding_block_size : 4; // + unsigned int pcm_loop_filter_disabled_flag : 1; // + unsigned int sps_temporal_mvp_enabled_flag : 1; // + unsigned int strong_intra_smoothing_enabled_flag : 1; // + unsigned int reserved3 : 3; + // pps + unsigned int dependent_slice_segments_enabled_flag : 1; // + unsigned int output_flag_present_flag : 1; // + unsigned int num_extra_slice_header_bits : 3; // 0..7 (normally 0) + unsigned int sign_data_hiding_enabled_flag : 1; // + unsigned int cabac_init_present_flag : 1; // + unsigned int num_ref_idx_l0_default_active : 4; // 1..15 + unsigned int num_ref_idx_l1_default_active : 4; // 1..15 + unsigned int init_qp : 7; // 0..127, support higher bitdepth + unsigned int constrained_intra_pred_flag : 1; // + unsigned int transform_skip_enabled_flag : 1; // + unsigned int cu_qp_delta_enabled_flag : 1; // + unsigned int diff_cu_qp_delta_depth : 2; // 0..3 + unsigned int reserved4 : 5; // + + char pps_cb_qp_offset ; // -12..12 + char pps_cr_qp_offset ; // -12..12 + char pps_beta_offset ; // -12..12 + char pps_tc_offset ; // -12..12 + unsigned int pps_slice_chroma_qp_offsets_present_flag : 1; // + unsigned int weighted_pred_flag : 1; // + unsigned int weighted_bipred_flag : 1; // + unsigned int transquant_bypass_enabled_flag : 1; // + unsigned int tiles_enabled_flag : 1; // (redundant: = num_tile_columns_minus1!=0 || num_tile_rows_minus1!=0) + unsigned int entropy_coding_sync_enabled_flag : 1; // + unsigned int num_tile_columns : 5; // 0..20 + unsigned int num_tile_rows : 5; // 0..22 + unsigned int loop_filter_across_tiles_enabled_flag : 1; // + unsigned int loop_filter_across_slices_enabled_flag : 1; // + unsigned int deblocking_filter_control_present_flag : 1; // + unsigned int deblocking_filter_override_enabled_flag : 1; // + unsigned int pps_deblocking_filter_disabled_flag : 1; // + unsigned int lists_modification_present_flag : 1; // + unsigned int log2_parallel_merge_level : 3; // 2..4 + unsigned int slice_segment_header_extension_present_flag : 1; // (normally 0) + unsigned int reserved5 : 6; + + // reference picture related + unsigned char num_ref_frames; + unsigned char reserved6; + unsigned short longtermflag; // long term flag for refpiclist.bit 15 for picidx 0, bit 14 for picidx 1,... + unsigned char initreflistidxl0[16]; // :5, [refPicidx] 0..15 + unsigned char initreflistidxl1[16]; // :5, [refPicidx] 0..15 + short RefDiffPicOrderCnts[16]; // poc diff between current and reference pictures .[-128,127] + // misc + unsigned char IDR_picture_flag; // idr flag for current picture + unsigned char RAP_picture_flag; // rap flag for current picture + unsigned char curr_pic_idx; // current picture store buffer index,used to derive the store addess of frame buffer and MV + unsigned char pattern_id; // used for dithering to select between 2 tables + unsigned short sw_hdr_skip_length; // reference picture inititial related syntax elements(SE) bits in slice header. + // those SE only decoding once in driver,related bits will flush in HW + unsigned short reserved7; + + // used for ecdma debug + nvdec_ecdma_config_s ecdma_cfg; + + //DXVA on windows + unsigned int separate_colour_plane_flag : 1; + unsigned int log2_max_pic_order_cnt_lsb_minus4 : 4; //0~12 + unsigned int num_short_term_ref_pic_sets : 7 ; //0~64 + unsigned int num_long_term_ref_pics_sps : 6; //0~32 + unsigned int bBitParsingDisable : 1 ; //disable parsing + unsigned int num_delta_pocs_of_rps_idx : 8; + unsigned int long_term_ref_pics_present_flag : 1; + unsigned int reserved_dxva : 4; + //the number of bits for short_term_ref_pic_set()in slice header,dxva API + unsigned int num_bits_short_term_ref_pics_in_slice; + + // New additions + nvdec_hevc_pic_v1_s v1; + nvdec_hevc_pic_v2_s v2; + nvdec_hevc_pic_v3_s v3; + nvdec_pass2_otf_ext_s ssm; + +} nvdec_hevc_pic_s; + +//hevc slice info class +typedef struct _hevc_slice_info_s { + unsigned int first_flag :1;//first slice(s) of frame,must valid for slice EC + unsigned int err_flag :1;//error slice(s) .optional info for EC + unsigned int last_flag :1;//last slice segment(s) of frame,this bit is must be valid when slice_strm_recfg_en==1 or slice_ec==1 + unsigned int conceal_partial_slice :1; // indicate do partial slice error conealment for packet loss case + unsigned int available :1; // indicate the slice bitstream is available. + unsigned int reserved0 :7; + unsigned int ctb_count :20;// ctbs counter inside slice(s) .must valid for slice EC + unsigned int bs_offset; //slice(s) bitstream offset in bitstream buffer (in byte unit) + unsigned int bs_length; //slice(s) bitstream length. It is sum of aligned size and skip size and valid slice bitstream size. + unsigned short start_ctbx; //slice start ctbx ,it's optional,HW can output it in previous slice decoding. + //but this is one check points for error + unsigned short start_ctby; //slice start ctby + } hevc_slice_info_s; + + +//hevc slice ctx class +//slice pos and next slice address +typedef struct _slice_edge_ctb_pos_ctx_s { + unsigned int next_slice_pos_ctbxy; //2d address in raster scan + unsigned int next_slice_segment_addr; //1d address in tile scan +}slice_edge_ctb_pos_ctx_s; + +// next slice's first ctb located tile related information +typedef struct _slice_edge_tile_ctx_s { + unsigned int tileInfo1;// Misc tile info includes tile width and tile height and tile col and tile row + unsigned int tileInfo2;// Misc tile info includes tile start ctbx and start ctby and tile index + unsigned int tileInfo3;// Misc tile info includes ctb pos inside tile +} slice_edge_tile_ctx_s; + +//frame level stats +typedef struct _slice_edge_stats_ctx_s { + unsigned int frame_status_intra_cnt;// frame stats for intra block count + unsigned int frame_status_inter_cnt;// frame stats for inter block count + unsigned int frame_status_skip_cnt;// frame stats for skip block count + unsigned int frame_status_fwd_mvx_cnt;// frame stats for sum of abs fwd mvx + unsigned int frame_status_fwd_mvy_cnt;// frame stats for sum of abs fwd mvy + unsigned int frame_status_bwd_mvx_cnt;// frame stats for sum of abs bwd mvx + unsigned int frame_status_bwd_mvy_cnt;// frame stats for sum of abs bwd mvy + unsigned int frame_status_mv_cnt_ext;// extension bits of sum of abs mv to keep full precision. +}slice_edge_stats_ctx_s; + +//ctx of vpc_edge unit for tile left +typedef struct _slice_vpc_edge_ctx_s { + unsigned int reserved; +}slice_vpc_edge_ctx_s; + +//ctx of vpc_main unit +typedef struct _slice_vpc_main_ctx_s { + unsigned int reserved; +} slice_vpc_main_ctx_s; + +//hevc slice edge ctx class +typedef struct _slice_edge_ctx_s { + //ctb pos + slice_edge_ctb_pos_ctx_s slice_ctb_pos_ctx; + // stats + slice_edge_stats_ctx_s slice_stats_ctx; + // tile info + slice_edge_tile_ctx_s slice_tile_ctx; + //vpc_edge + slice_vpc_edge_ctx_s slice_vpc_edge_ctx; + //vpc_main + slice_vpc_main_ctx_s slice_vpc_main_ctx; +} slice_edge_ctx_s; + +//vp9 + +typedef struct _nvdec_vp9_pic_v1_s +{ + // New fields + // new_var : xx; // for variables with expanded bitlength, comment on why the new bit legth is required + // Reserved bits for padding and/or non-HW specific functionality + unsigned int Vp9FltAboveOffset; // filter above offset respect to filter buffer, 256 bytes unit + unsigned int external_ref_mem_dis : 1; + unsigned int bit_depth : 4; + unsigned int error_recovery_start_pos : 2; //0: from start of frame, 1: from start of slice segment, 2: from error detected ctb, 3: reserved + unsigned int error_external_mv_en : 1; + unsigned int Reserved0 : 24; +} nvdec_vp9_pic_v1_s; + +enum VP9_FRAME_SFC_ID +{ + VP9_LAST_FRAME_SFC = 0, + VP9_GOLDEN_FRAME_SFC, + VP9_ALTREF_FRAME_SFC, + VP9_CURR_FRAME_SFC +}; + +typedef struct _nvdec_vp9_pic_s +{ + // vp9_bitstream_data_info + //Key and IV address must 128bit alignment + unsigned int wrapped_session_key[4]; //session keys + unsigned int wrapped_content_key[4]; //content keys + unsigned int initialization_vector[4]; //Ctrl64 initial vector + unsigned int stream_len; // stream length in one frame + unsigned int enable_encryption; // flag to enable/disable encryption + unsigned int key_increment : 6; // added to content key after unwrapping + unsigned int encryption_mode : 4; + unsigned int sw_hdr_skip_length :14; //vp9 skip bytes setting for secure + unsigned int key_slot_index : 4; + unsigned int ssm_en : 1; + unsigned int enable_histogram : 1; // histogram stats output enable + unsigned int reserved0 : 2; + + // Driver may or may not use based upon need. + // If 0 then default value of 1<<27 = 298ms @ 450MHz will be used in ucode. + // Driver can send this value based upon resolution using the formula: + // gptimer_timeout_value = 3 * (cycles required for one frame) + unsigned int gptimer_timeout_value; + + //general + unsigned char tileformat : 2 ; // 0: TBL; 1: KBL; 2: Tile16x16 + unsigned char gob_height : 3 ; // Set GOB height, 0: GOB_2, 1: GOB_4, 2: GOB_8, 3: GOB_16, 4: GOB_32 (NVDEC3 onwards) + unsigned char reserverd_surface_format : 3 ; + unsigned char reserved1[3]; + unsigned int Vp9BsdCtrlOffset; // bsd buffer offset respect to filter buffer ,256 bytes unit . + + + //ref_last dimensions + unsigned short ref0_width; //ref_last coded width + unsigned short ref0_height; //ref_last coded height + unsigned short ref0_stride[2]; //ref_last stride + + //ref_golden dimensions + unsigned short ref1_width; //ref_golden coded width + unsigned short ref1_height; //ref_golden coded height + unsigned short ref1_stride[2]; //ref_golden stride + + //ref_alt dimensions + unsigned short ref2_width; //ref_alt coded width + unsigned short ref2_height; //ref_alt coded height + unsigned short ref2_stride[2]; //ref_alt stride + + + /* Current frame dimensions */ + unsigned short width; //pic width + unsigned short height; //pic height + unsigned short framestride[2]; // frame buffer stride for luma and chroma + + unsigned char keyFrame :1; + unsigned char prevIsKeyFrame:1; + unsigned char resolutionChange:1; + unsigned char errorResilient:1; + unsigned char prevShowFrame:1; + unsigned char intraOnly:1; + unsigned char reserved2 : 2; + + /* DCT coefficient partitions */ + //unsigned int offsetToDctParts; + + unsigned char reserved3[3]; + //unsigned char activeRefIdx[3];//3 bits + //unsigned char refreshFrameFlags; + //unsigned char refreshEntropyProbs; + //unsigned char frameParallelDecoding; + //unsigned char resetFrameContext; + + unsigned char refFrameSignBias[4]; + char loopFilterLevel;//6 bits + char loopFilterSharpness;//3 bits + + /* Quantization parameters */ + unsigned char qpYAc; + char qpYDc; + char qpChAc; + char qpChDc; + + /* From here down, frame-to-frame persisting stuff */ + + char lossless; + char transform_mode; + char allow_high_precision_mv; + char mcomp_filter_type; + char comp_pred_mode; + char comp_fixed_ref; + char comp_var_ref[2]; + char log2_tile_columns; + char log2_tile_rows; + + /* Segment and macroblock specific values */ + unsigned char segmentEnabled; + unsigned char segmentMapUpdate; + unsigned char segmentMapTemporalUpdate; + unsigned char segmentFeatureMode; /* ABS data or delta data */ + unsigned char segmentFeatureEnable[8][4]; + short segmentFeatureData[8][4]; + char modeRefLfEnabled; + char mbRefLfDelta[4]; + char mbModeLfDelta[2]; + char reserved5; // for alignment + + // New additions + nvdec_vp9_pic_v1_s v1; + nvdec_pass2_otf_ext_s ssm; + +} nvdec_vp9_pic_s; + +#define NVDEC_VP9HWPAD(x, y) unsigned char x[y] + +typedef struct { + /* last bytes of address 41 */ + unsigned char joints[3]; + unsigned char sign[2]; + /* address 42 */ + unsigned char class0[2][1]; + unsigned char fp[2][3]; + unsigned char class0_hp[2]; + unsigned char hp[2]; + unsigned char classes[2][10]; + /* address 43 */ + unsigned char class0_fp[2][2][3]; + unsigned char bits[2][10]; + +} nvdec_nmv_context; + +typedef struct { + unsigned int joints[4]; + unsigned int sign[2][2]; + unsigned int classes[2][11]; + unsigned int class0[2][2]; + unsigned int bits[2][10][2]; + unsigned int class0_fp[2][2][4]; + unsigned int fp[2][4]; + unsigned int class0_hp[2][2]; + unsigned int hp[2][2]; + +} nvdec_nmv_context_counts; + +/* Adaptive entropy contexts, padding elements are added to have + * 256 bit aligned tables for HW access. + * Compile with TRACE_PROB_TABLES to print bases for each table. */ +typedef struct nvdec_vp9AdaptiveEntropyProbs_s +{ + /* address 32 */ + unsigned char inter_mode_prob[7][4]; + unsigned char intra_inter_prob[4]; + + /* address 33 */ + unsigned char uv_mode_prob[10][8]; + unsigned char tx8x8_prob[2][1]; + unsigned char tx16x16_prob[2][2]; + unsigned char tx32x32_prob[2][3]; + unsigned char sb_ymode_probB[4][1]; + unsigned char sb_ymode_prob[4][8]; + + /* address 37 */ + unsigned char partition_prob[2][16][4]; + + /* address 41 */ + unsigned char uv_mode_probB[10][1]; + unsigned char switchable_interp_prob[4][2]; + unsigned char comp_inter_prob[5]; + unsigned char mbskip_probs[3]; + NVDEC_VP9HWPAD(pad1, 1); + + nvdec_nmv_context nmvc; + + /* address 44 */ + unsigned char single_ref_prob[5][2]; + unsigned char comp_ref_prob[5]; + NVDEC_VP9HWPAD(pad2, 17); + + /* address 45 */ + unsigned char probCoeffs[2][2][6][6][4]; + unsigned char probCoeffs8x8[2][2][6][6][4]; + unsigned char probCoeffs16x16[2][2][6][6][4]; + unsigned char probCoeffs32x32[2][2][6][6][4]; + +} nvdec_vp9AdaptiveEntropyProbs_t; + +/* Entropy contexts */ +typedef struct nvdec_vp9EntropyProbs_s +{ + /* Default keyframe probs */ + /* Table formatted for 256b memory, probs 0to7 for all tables followed by + * probs 8toN for all tables. + * Compile with TRACE_PROB_TABLES to print bases for each table. */ + + unsigned char kf_bmode_prob[10][10][8]; + + /* Address 25 */ + unsigned char kf_bmode_probB[10][10][1]; + unsigned char ref_pred_probs[3]; + unsigned char mb_segment_tree_probs[7]; + unsigned char segment_pred_probs[3]; + unsigned char ref_scores[4]; + unsigned char prob_comppred[2]; + NVDEC_VP9HWPAD(pad1, 9); + + /* Address 29 */ + unsigned char kf_uv_mode_prob[10][8]; + unsigned char kf_uv_mode_probB[10][1]; + NVDEC_VP9HWPAD(pad2, 6); + + nvdec_vp9AdaptiveEntropyProbs_t a; /* Probs with backward adaptation */ + +} nvdec_vp9EntropyProbs_t; + +/* Counters for adaptive entropy contexts */ +typedef struct nvdec_vp9EntropyCounts_s +{ + unsigned int inter_mode_counts[7][3][2]; + unsigned int sb_ymode_counts[4][10]; + unsigned int uv_mode_counts[10][10]; + unsigned int partition_counts[16][4]; + unsigned int switchable_interp_counts[4][3]; + unsigned int intra_inter_count[4][2]; + unsigned int comp_inter_count[5][2]; + unsigned int single_ref_count[5][2][2]; + unsigned int comp_ref_count[5][2]; + unsigned int tx32x32_count[2][4]; + unsigned int tx16x16_count[2][3]; + unsigned int tx8x8_count[2][2]; + unsigned int mbskip_count[3][2]; + + nvdec_nmv_context_counts nmvcount; + + unsigned int countCoeffs[2][2][6][6][4]; + unsigned int countCoeffs8x8[2][2][6][6][4]; + unsigned int countCoeffs16x16[2][2][6][6][4]; + unsigned int countCoeffs32x32[2][2][6][6][4]; + + unsigned int countEobs[4][2][2][6][6]; + +} nvdec_vp9EntropyCounts_t; + +// Legacy codecs encryption parameters +typedef struct _nvdec_pass2_otf_s { + unsigned int wrapped_session_key[4]; // session keys + unsigned int wrapped_content_key[4]; // content keys + unsigned int initialization_vector[4];// Ctrl64 initial vector + unsigned int enable_encryption : 1; // flag to enable/disable encryption + unsigned int key_increment : 6; // added to content key after unwrapping + unsigned int encryption_mode : 4; + unsigned int key_slot_index : 4; + unsigned int ssm_en : 1; + unsigned int reserved1 :16; // reserved +} nvdec_pass2_otf_s; // 0x10 bytes + +typedef struct _nvdec_display_param_s +{ + unsigned int enableTFOutput : 1; //=1, enable dbfdma to output the display surface; if disable, then the following configure on tf is useless. + //remap for VC1 + unsigned int VC1MapYFlag : 1; + unsigned int MapYValue : 3; + unsigned int VC1MapUVFlag : 1; + unsigned int MapUVValue : 3; + //tf + unsigned int OutStride : 8; + unsigned int TilingFormat : 3; + unsigned int OutputStructure : 1; //(0=frame, 1=field) + unsigned int reserved0 :11; + int OutputTop[2]; // in units of 256 + int OutputBottom[2]; // in units of 256 + //histogram + unsigned int enableHistogram : 1; // enable histogram info collection. + unsigned int HistogramStartX :12; // start X of Histogram window + unsigned int HistogramStartY :12; // start Y of Histogram window + unsigned int reserved1 : 7; + unsigned int HistogramEndX :12; // end X of Histogram window + unsigned int HistogramEndY :12; // end y of Histogram window + unsigned int reserved2 : 8; +} nvdec_display_param_s; // size 0x1c bytes + +// H.264 +typedef struct _nvdec_dpb_entry_s // 16 bytes +{ + unsigned int index : 7; // uncompressed frame buffer index + unsigned int col_idx : 5; // index of associated co-located motion data buffer + unsigned int state : 2; // bit1(state)=1: top field used for reference, bit1(state)=1: bottom field used for reference + unsigned int is_long_term : 1; // 0=short-term, 1=long-term + unsigned int not_existing : 1; // 1=marked as non-existing + unsigned int is_field : 1; // set if unpaired field or complementary field pair + unsigned int top_field_marking : 4; + unsigned int bottom_field_marking : 4; + unsigned int output_memory_layout : 1; // Set according to picture level output NV12/NV24 setting. + unsigned int reserved : 6; + unsigned int FieldOrderCnt[2]; // : 2*32 [top/bottom] + int FrameIdx; // : 16 short-term: FrameNum (16 bits), long-term: LongTermFrameIdx (4 bits) +} nvdec_dpb_entry_s; + +typedef struct _nvdec_h264_pic_s +{ + nvdec_pass2_otf_s encryption_params; + unsigned char eos[16]; + unsigned char explicitEOSPresentFlag; + unsigned char hint_dump_en; //enable COLOMV surface dump for all frames, which includes hints of "MV/REFIDX/QP/CBP/MBPART/MBTYPE", nvbug: 200212874 + unsigned char reserved0[2]; + unsigned int stream_len; + unsigned int slice_count; + unsigned int mbhist_buffer_size; // to pass buffer size of MBHIST_BUFFER + + // Driver may or may not use based upon need. + // If 0 then default value of 1<<27 = 298ms @ 450MHz will be used in ucode. + // Driver can send this value based upon resolution using the formula: + // gptimer_timeout_value = 3 * (cycles required for one frame) + unsigned int gptimer_timeout_value; + + // Fields from msvld_h264_seq_s + int log2_max_pic_order_cnt_lsb_minus4; + int delta_pic_order_always_zero_flag; + int frame_mbs_only_flag; + int PicWidthInMbs; + int FrameHeightInMbs; + + unsigned int tileFormat : 2 ; // 0: TBL; 1: KBL; 2: Tile16x16 + unsigned int gob_height : 3 ; // Set GOB height, 0: GOB_2, 1: GOB_4, 2: GOB_8, 3: GOB_16, 4: GOB_32 (NVDEC3 onwards) + unsigned int reserverd_surface_format : 27; + + // Fields from msvld_h264_pic_s + int entropy_coding_mode_flag; + int pic_order_present_flag; + int num_ref_idx_l0_active_minus1; + int num_ref_idx_l1_active_minus1; + int deblocking_filter_control_present_flag; + int redundant_pic_cnt_present_flag; + int transform_8x8_mode_flag; + + // Fields from mspdec_h264_picture_setup_s + unsigned int pitch_luma; // Luma pitch + unsigned int pitch_chroma; // chroma pitch + + unsigned int luma_top_offset; // offset of luma top field in units of 256 + unsigned int luma_bot_offset; // offset of luma bottom field in units of 256 + unsigned int luma_frame_offset; // offset of luma frame in units of 256 + unsigned int chroma_top_offset; // offset of chroma top field in units of 256 + unsigned int chroma_bot_offset; // offset of chroma bottom field in units of 256 + unsigned int chroma_frame_offset; // offset of chroma frame in units of 256 + unsigned int HistBufferSize; // in units of 256 + + unsigned int MbaffFrameFlag : 1; // + unsigned int direct_8x8_inference_flag: 1; // + unsigned int weighted_pred_flag : 1; // + unsigned int constrained_intra_pred_flag:1; // + unsigned int ref_pic_flag : 1; // reference picture (nal_ref_idc != 0) + unsigned int field_pic_flag : 1; // + unsigned int bottom_field_flag : 1; // + unsigned int second_field : 1; // second field of complementary reference field + unsigned int log2_max_frame_num_minus4: 4; // (0..12) + unsigned int chroma_format_idc : 2; // + unsigned int pic_order_cnt_type : 2; // (0..2) + int pic_init_qp_minus26 : 6; // : 6 (-26..+25) + int chroma_qp_index_offset : 5; // : 5 (-12..+12) + int second_chroma_qp_index_offset : 5; // : 5 (-12..+12) + + unsigned int weighted_bipred_idc : 2; // : 2 (0..2) + unsigned int CurrPicIdx : 7; // : 7 uncompressed frame buffer index + unsigned int CurrColIdx : 5; // : 5 index of associated co-located motion data buffer + unsigned int frame_num : 16; // + unsigned int frame_surfaces : 1; // frame surfaces flag + unsigned int output_memory_layout : 1; // 0: NV12; 1:NV24. Field pair must use the same setting. + + int CurrFieldOrderCnt[2]; // : 32 [Top_Bottom], [0]=TopFieldOrderCnt, [1]=BottomFieldOrderCnt + nvdec_dpb_entry_s dpb[16]; + unsigned char WeightScale[6][4][4]; // : 6*4*4*8 in raster scan order (not zig-zag order) + unsigned char WeightScale8x8[2][8][8]; // : 2*8*8*8 in raster scan order (not zig-zag order) + + // mvc setup info, must be zero if not mvc + unsigned char num_inter_view_refs_lX[2]; // number of inter-view references + char reserved1[14]; // reserved for alignment + signed char inter_view_refidx_lX[2][16]; // DPB indices (must also be marked as long-term) + + // lossless decode (At the time of writing this manual, x264 and JM encoders, differ in Intra_8x8 reference sample filtering) + unsigned int lossless_ipred8x8_filter_enable : 1; // = 0, skips Intra_8x8 reference sample filtering, for vertical and horizontal predictions (x264 encoded streams); = 1, filter Intra_8x8 reference samples (JM encoded streams) + unsigned int qpprime_y_zero_transform_bypass_flag : 1; // determines the transform bypass mode + unsigned int reserved2 : 30; // kept for alignment; may be used for other parameters + + nvdec_display_param_s displayPara; + nvdec_pass2_otf_ext_s ssm; + +} nvdec_h264_pic_s; + +// VC-1 Scratch buffer +typedef enum _vc1_fcm_e +{ + FCM_PROGRESSIVE = 0, + FCM_FRAME_INTERLACE = 2, + FCM_FIELD_INTERLACE = 3 +} vc1_fcm_e; + +typedef enum _syntax_vc1_ptype_e +{ + PTYPE_I = 0, + PTYPE_P = 1, + PTYPE_B = 2, + PTYPE_BI = 3, //PTYPE_BI is not used to config register NV_CNVDEC_VLD_PIC_INFO_COMMON. field NV_CNVDEC_VLD_PIC_INFO_COMMON_PIC_CODING_VC1 is only 2 bits. I and BI pictures are configured with same value. Please refer to manual. + PTYPE_SKIPPED = 4 +} syntax_vc1_ptype_e; + +// 7.1.1.32, Table 46 etc. +enum vc1_mvmode_e +{ + MVMODE_MIXEDMV = 0, + MVMODE_1MV = 1, + MVMODE_1MV_HALFPEL = 2, + MVMODE_1MV_HALFPEL_BILINEAR = 3, + MVMODE_INTENSITY_COMPENSATION = 4 +}; + +// 9.1.1.42, Table 105 +typedef enum _vc1_fptype_e +{ + FPTYPE_I_I = 0, + FPTYPE_I_P, + FPTYPE_P_I, + FPTYPE_P_P, + FPTYPE_B_B, + FPTYPE_B_BI, + FPTYPE_BI_B, + FPTYPE_BI_BI +} vc1_fptype_e; + +// Table 43 (7.1.1.31.2) +typedef enum _vc1_dqprofile_e +{ + DQPROFILE_ALL_FOUR_EDGES = 0, + DQPROFILE_DOUBLE_EDGE = 1, + DQPROFILE_SINGLE_EDGE = 2, + DQPROFILE_ALL_MACROBLOCKS = 3 +} vc1_dqprofile_e; + +typedef struct _nvdec_vc1_pic_s +{ + nvdec_pass2_otf_s encryption_params; + unsigned char eos[16]; // to pass end of stream data separately if not present in bitstream surface + unsigned char prefixStartCode[4]; // used for dxva to pass prefix start code. + unsigned int bitstream_offset; // offset in words from start of bitstream surface if there is gap. + unsigned char explicitEOSPresentFlag; // to indicate that eos[] is used for passing end of stream data. + unsigned char reserved0[3]; + unsigned int stream_len; + unsigned int slice_count; + unsigned int scratch_pic_buffer_size; + + // Driver may or may not use based upon need. + // If 0 then default value of 1<<27 = 298ms @ 450MHz will be used in ucode. + // Driver can send this value based upon resolution using the formula: + // gptimer_timeout_value = 3 * (cycles required for one frame) + unsigned int gptimer_timeout_value; + + // Fields from vc1_seq_s + unsigned short FrameWidth; // actual frame width + unsigned short FrameHeight; // actual frame height + + unsigned char profile; // 1 = SIMPLE or MAIN, 2 = ADVANCED + unsigned char postprocflag; + unsigned char pulldown; + unsigned char interlace; + + unsigned char tfcntrflag; + unsigned char finterpflag; + unsigned char psf; + unsigned char tileFormat : 2 ; // 0: TBL; 1: KBL; 2: Tile16x16 + unsigned char gob_height : 3 ; // Set GOB height, 0: GOB_2, 1: GOB_4, 2: GOB_8, 3: GOB_16, 4: GOB_32 (NVDEC3 onwards) + unsigned char reserverd_surface_format : 3 ; + + // simple,main + unsigned char multires; + unsigned char syncmarker; + unsigned char rangered; + unsigned char maxbframes; + + // Fields from vc1_entrypoint_s + unsigned char dquant; + unsigned char panscan_flag; + unsigned char refdist_flag; + unsigned char quantizer; + + unsigned char extended_mv; + unsigned char extended_dmv; + unsigned char overlap; + unsigned char vstransform; + + // Fields from vc1_scratch_s + char refdist; + char reserved1[3]; // for alignment + + // Fields from vld_vc1_pic_s + vc1_fcm_e fcm; + syntax_vc1_ptype_e ptype; + int tfcntr; + int rptfrm; + int tff; + int rndctrl; + int pqindex; + int halfqp; + int pquantizer; + int postproc; + int condover; + int transacfrm; + int transacfrm2; + int transdctab; + int pqdiff; + int abspq; + int dquantfrm; + vc1_dqprofile_e dqprofile; + int dqsbedge; + int dqdbedge; + int dqbilevel; + int mvrange; + enum vc1_mvmode_e mvmode; + enum vc1_mvmode_e mvmode2; + int lumscale; + int lumshift; + int mvtab; + int cbptab; + int ttmbf; + int ttfrm; + int bfraction; + vc1_fptype_e fptype; + int numref; + int reffield; + int dmvrange; + int intcompfield; + int lumscale1; // type was char in ucode + int lumshift1; // type was char in ucode + int lumscale2; // type was char in ucode + int lumshift2; // type was char in ucode + int mbmodetab; + int imvtab; + int icbptab; + int fourmvbptab; + int fourmvswitch; + int intcomp; + int twomvbptab; + // simple,main + int rangeredfrm; + + // Fields from pdec_vc1_pic_s + unsigned int HistBufferSize; // in units of 256 + // frame buffers + unsigned int FrameStride[2]; // [y_c] + unsigned int luma_top_offset; // offset of luma top field in units of 256 + unsigned int luma_bot_offset; // offset of luma bottom field in units of 256 + unsigned int luma_frame_offset; // offset of luma frame in units of 256 + unsigned int chroma_top_offset; // offset of chroma top field in units of 256 + unsigned int chroma_bot_offset; // offset of chroma bottom field in units of 256 + unsigned int chroma_frame_offset; // offset of chroma frame in units of 256 + + unsigned short CodedWidth; // entrypoint specific + unsigned short CodedHeight; // entrypoint specific + + unsigned char loopfilter; // entrypoint specific + unsigned char fastuvmc; // entrypoint specific + unsigned char output_memory_layout; // picture specific + unsigned char ref_memory_layout[2]; // picture specific 0: fwd, 1: bwd + unsigned char reserved3[3]; // for alignment + + nvdec_display_param_s displayPara; + nvdec_pass2_otf_ext_s ssm; + +} nvdec_vc1_pic_s; + +// MPEG-2 +typedef struct _nvdec_mpeg2_pic_s +{ + nvdec_pass2_otf_s encryption_params; + unsigned char eos[16]; + unsigned char explicitEOSPresentFlag; + unsigned char reserved0[3]; + unsigned int stream_len; + unsigned int slice_count; + + // Driver may or may not use based upon need. + // If 0 then default value of 1<<27 = 298ms @ 450MHz will be used in ucode. + // Driver can send this value based upon resolution using the formula: + // gptimer_timeout_value = 3 * (cycles required for one frame) + unsigned int gptimer_timeout_value; + + // Fields from vld_mpeg2_seq_pic_info_s + short FrameWidth; // actual frame width + short FrameHeight; // actual frame height + unsigned char picture_structure; // 0 => Reserved, 1 => Top field, 2 => Bottom field, 3 => Frame picture. Table 6-14. + unsigned char picture_coding_type; // 0 => Forbidden, 1 => I, 2 => P, 3 => B, 4 => D (for MPEG-2). Table 6-12. + unsigned char intra_dc_precision; // 0 => 8 bits, 1=> 9 bits, 2 => 10 bits, 3 => 11 bits. Table 6-13. + char frame_pred_frame_dct; // as in section 6.3.10 + char concealment_motion_vectors; // as in section 6.3.10 + char intra_vlc_format; // as in section 6.3.10 + unsigned char tileFormat : 2 ; // 0: TBL; 1: KBL; 2: Tile16x16 + unsigned char gob_height : 3 ; // Set GOB height, 0: GOB_2, 1: GOB_4, 2: GOB_8, 3: GOB_16, 4: GOB_32 (NVDEC3 onwards) + unsigned char reserverd_surface_format : 3 ; + + char reserved1; // always 0 + char f_code[4]; // as in section 6.3.10 + + // Fields from pdec_mpeg2_picture_setup_s + unsigned short PicWidthInMbs; + unsigned short FrameHeightInMbs; + unsigned int pitch_luma; + unsigned int pitch_chroma; + unsigned int luma_top_offset; + unsigned int luma_bot_offset; + unsigned int luma_frame_offset; + unsigned int chroma_top_offset; + unsigned int chroma_bot_offset; + unsigned int chroma_frame_offset; + unsigned int HistBufferSize; + unsigned short output_memory_layout; + unsigned short alternate_scan; + unsigned short secondfield; + /******************************/ + // Got rid of the union kept for compatibility with NVDEC1. + // Removed field mpeg2, and kept rounding type. + // NVDEC1 ucode is not using the mpeg2 field, instead using codec type from the methods. + // Rounding type should only be set for Divx3.11. + unsigned short rounding_type; + /******************************/ + unsigned int MbInfoSizeInBytes; + unsigned int q_scale_type; + unsigned int top_field_first; + unsigned int full_pel_fwd_vector; + unsigned int full_pel_bwd_vector; + unsigned char quant_mat_8x8intra[64]; + unsigned char quant_mat_8x8nonintra[64]; + unsigned int ref_memory_layout[2]; //0:for fwd; 1:for bwd + + nvdec_display_param_s displayPara; + nvdec_pass2_otf_ext_s ssm; + +} nvdec_mpeg2_pic_s; + +// MPEG-4 +typedef struct _nvdec_mpeg4_pic_s +{ + nvdec_pass2_otf_s encryption_params; + unsigned char eos[16]; + unsigned char explicitEOSPresentFlag; + unsigned char reserved2[3]; // for alignment + unsigned int stream_len; + unsigned int slice_count; + unsigned int scratch_pic_buffer_size; + + // Driver may or may not use based upon need. + // If 0 then default value of 1<<27 = 298ms @ 450MHz will be used in ucode. + // Driver can send this value based upon resolution using the formula: + // gptimer_timeout_value = 3 * (cycles required for one frame) + unsigned int gptimer_timeout_value; + + // Fields from vld_mpeg4_seq_s + short FrameWidth; // :13 video_object_layer_width + short FrameHeight; // :13 video_object_layer_height + char vop_time_increment_bitcount; // : 5 1..16 + char resync_marker_disable; // : 1 + unsigned char tileFormat : 2 ; // 0: TBL; 1: KBL; 2: Tile16x16 + unsigned char gob_height : 3 ; // Set GOB height, 0: GOB_2, 1: GOB_4, 2: GOB_8, 3: GOB_16, 4: GOB_32 (NVDEC3 onwards) + unsigned char reserverd_surface_format : 3 ; + char reserved3; // for alignment + + // Fields from pdec_mpeg4_picture_setup_s + int width; // : 13 + int height; // : 13 + + unsigned int FrameStride[2]; // [y_c] + unsigned int luma_top_offset; // offset of luma top field in units of 256 + unsigned int luma_bot_offset; // offset of luma bottom field in units of 256 + unsigned int luma_frame_offset; // offset of luma frame in units of 256 + unsigned int chroma_top_offset; // offset of chroma top field in units of 256 + unsigned int chroma_bot_offset; // offset of chroma bottom field in units of 256 + unsigned int chroma_frame_offset; // offset of chroma frame in units of 256 + + unsigned int HistBufferSize; // in units of 256, History buffer size + + int trd[2]; // : 16, temporal reference frame distance (only needed for B-VOPs) + int trb[2]; // : 16, temporal reference B-VOP distance from fwd reference frame (only needed for B-VOPs) + + int divx_flags; // : 16 (bit 0: DivX interlaced chroma rounding, bit 1: Divx 4 boundary padding, bit 2: Divx IDCT) + + short vop_fcode_forward; // : 1...7 + short vop_fcode_backward; // : 1...7 + + unsigned char interlaced; // : 1 + unsigned char quant_type; // : 1 + unsigned char quarter_sample; // : 1 + unsigned char short_video_header; // : 1 + + unsigned char curr_output_memory_layout; // : 1 0:NV12; 1:NV24 + unsigned char ptype; // picture type: 0 for PTYPE_I, 1 for PTYPE_P, 2 for PTYPE_B, 3 for PTYPE_BI, 4 for PTYPE_SKIPPED + unsigned char rnd; // : 1, rounding mode + unsigned char alternate_vertical_scan_flag; // : 1 + + unsigned char top_field_flag; // : 1 + unsigned char reserved0[3]; // alignment purpose + + unsigned char intra_quant_mat[64]; // : 64*8 + unsigned char nonintra_quant_mat[64]; // : 64*8 + unsigned char ref_memory_layout[2]; //0:for fwd; 1:for bwd + unsigned char reserved1[34]; // 256 byte alignemnt till now + + nvdec_display_param_s displayPara; + +} nvdec_mpeg4_pic_s; + +// VP8 +enum VP8_FRAME_TYPE +{ + VP8_KEYFRAME = 0, + VP8_INTERFRAME = 1 +}; + +enum VP8_FRAME_SFC_ID +{ + VP8_GOLDEN_FRAME_SFC = 0, + VP8_ALTREF_FRAME_SFC, + VP8_LAST_FRAME_SFC, + VP8_CURR_FRAME_SFC +}; + +typedef struct _nvdec_vp8_pic_s +{ + nvdec_pass2_otf_s encryption_params; + + // Driver may or may not use based upon need. + // If 0 then default value of 1<<27 = 298ms @ 450MHz will be used in ucode. + // Driver can send this value based upon resolution using the formula: + // gptimer_timeout_value = 3 * (cycles required for one frame) + unsigned int gptimer_timeout_value; + + unsigned short FrameWidth; // actual frame width + unsigned short FrameHeight; // actual frame height + + unsigned char keyFrame; // 1: key frame; 0: not + unsigned char version; + unsigned char tileFormat : 2 ; // 0: TBL; 1: KBL; 2: Tile16x16 + unsigned char gob_height : 3 ; // Set GOB height, 0: GOB_2, 1: GOB_4, 2: GOB_8, 3: GOB_16, 4: GOB_32 (NVDEC3 onwards) + unsigned char reserverd_surface_format : 3 ; + unsigned char errorConcealOn; // 1: error conceal on; 0: off + + unsigned int firstPartSize; // the size of first partition(frame header and mb header partition) + + // ctx + unsigned int HistBufferSize; // in units of 256 + unsigned int VLDBufferSize; // in units of 1 + // current frame buffers + unsigned int FrameStride[2]; // [y_c] + unsigned int luma_top_offset; // offset of luma top field in units of 256 + unsigned int luma_bot_offset; // offset of luma bottom field in units of 256 + unsigned int luma_frame_offset; // offset of luma frame in units of 256 + unsigned int chroma_top_offset; // offset of chroma top field in units of 256 + unsigned int chroma_bot_offset; // offset of chroma bottom field in units of 256 + unsigned int chroma_frame_offset; // offset of chroma frame in units of 256 + + nvdec_display_param_s displayPara; + + // decode picture buffere related + char current_output_memory_layout; + char output_memory_layout[3]; // output NV12/NV24 setting. item 0:golden; 1: altref; 2: last + + unsigned char segmentation_feature_data_update; + unsigned char reserved1[3]; + + // ucode return result + unsigned int resultValue; // ucode return the picture header info; includes copy_buffer_to_golden etc. + unsigned int partition_offset[8]; // byte offset to each token partition (used for encrypted streams only) + + nvdec_pass2_otf_ext_s ssm; + +} nvdec_vp8_pic_s; // size is 0xc0 + +// PASS1 + +//Sample means the entire frame is encrypted with a single IV, and subsample means a given frame may be encrypted in multiple chunks with different IVs. +#define NUM_SUBSAMPLES 32 + +typedef struct _bytes_of_data_s +{ + unsigned int clear_bytes; // clear bytes per subsample + unsigned int encypted_bytes; // encrypted bytes per subsample + +} bytes_of_data_s; + +typedef struct _nvdec_pass1_input_data_s +{ + bytes_of_data_s sample_size[NUM_SUBSAMPLES]; // clear/encrypted bytes per subsample + unsigned int initialization_vector[NUM_SUBSAMPLES][4]; // Ctrl64 initial vector per subsample + unsigned char IvValid[NUM_SUBSAMPLES]; // each element will tell whether IV is valid for that subsample or not. + unsigned int stream_len; // encrypted bitstream size. + unsigned int clearBufferSize; // allocated size of clear buffer size + unsigned int reencryptBufferSize; // allocated size of reencrypted buffer size + unsigned int vp8coeffPartitonBufferSize; // allocated buffer for vp8 coeff partition buffer + unsigned int PrevWidth; // required for VP9 + unsigned int num_nals :16; // number of subsamples in a frame + unsigned int drm_mode : 8; // DRM mode + unsigned int key_sel : 4; // key select from keyslot + unsigned int codec : 4; // codecs selection + unsigned int TotalSizeOfClearData; // Used with Pattern based encryption + unsigned int SliceHdrOffset; // This is used with pattern mode encryption where data before slice hdr comes in clear. + unsigned int EncryptBlkCnt :16; + unsigned int SkipBlkCnt :16; +} nvdec_pass1_input_data_s; + +#define VP8_MAX_TOKEN_PARTITIONS 8 +#define VP9_MAX_FRAMES_IN_SUPERFRAME 8 + +typedef struct _nvdec_pass1_output_data_s +{ + unsigned int clear_header_size; // h264/vc1/mpeg2/vp8, decrypted pps/sps/part of slice header info, 128 bits aligned + unsigned int reencrypt_data_size; // h264/vc1/mpeg2, slice level data, vp8 mb header info, 128 bits aligned + unsigned int clear_token_data_size; // vp8, clear token data saved in VPR, 128 bits aligned + unsigned int key_increment : 6; // added to content key after unwrapping + unsigned int encryption_mode : 4; // encryption mode + unsigned int bReEncrypted : 1; // set to 0 if no re-encryption is done. + unsigned int bvp9SuperFrame : 1; // set to 1 for vp9 superframe + unsigned int vp9NumFramesMinus1 : 3; // set equal to numFrames-1 for vp9superframe. Max 8 frames are possible in vp9 superframe. + unsigned int reserved1 :17; // reserved, 32 bit alignment + unsigned int wrapped_session_key[4]; // session keys + unsigned int wrapped_content_key[4]; // content keys + unsigned int initialization_vector[4]; // Ctrl64 initial vector + union { + unsigned int partition_size[VP8_MAX_TOKEN_PARTITIONS]; // size of each token partition (used for encrypted streams of VP8) + unsigned int vp9_frame_sizes[VP9_MAX_FRAMES_IN_SUPERFRAME]; // frame size information for all frames in vp9 superframe. + }; + unsigned int vp9_clear_hdr_size[VP9_MAX_FRAMES_IN_SUPERFRAME]; // clear header size for each frame in vp9 superframe. +} nvdec_pass1_output_data_s; + + +/***************************************************** + AV1 +*****************************************************/ +typedef struct _scale_factors_reference_s{ + short x_scale_fp; // horizontal fixed point scale factor + short y_scale_fp; // vertical fixed point scale factor +}scale_factors_reference_s; + +typedef struct _frame_info_t{ + unsigned short width; // in pixel, av1 support arbitray resolution + unsigned short height; + unsigned short stride[2]; // luma and chroma stride in 16Bytes + unsigned int frame_buffer_idx; // TBD :clean associate the reference frame and frame buffer id to lookup base_addr +} frame_info_t; + +typedef struct _ref_frame_struct_s{ + frame_info_t info; + scale_factors_reference_s sf; // scalefactor for reference frame and current frame size, driver can calculate it + unsigned char sign_bias : 1; // calcuate based on frame_offset and current frame offset + unsigned char wmtype : 2; // global motion parameters : identity,translation,rotzoom,affine + unsigned char reserved_rf : 5; + short frame_off; // relative offset to current frame + short roffset; // relative offset from current frame +} ref_frame_struct_s; + +typedef struct _av1_fgs_cfg_t{ + //from AV1 spec 5.9.30 Film Grain Params syntax + unsigned short apply_grain : 1; + unsigned short overlap_flag : 1; + unsigned short clip_to_restricted_range : 1; + unsigned short chroma_scaling_from_luma : 1; + unsigned short num_y_points_b : 1; // flag indicates num_y_points>0 + unsigned short num_cb_points_b : 1; // flag indicates num_cb_points>0 + unsigned short num_cr_points_b : 1; // flag indicates num_cr_points>0 + unsigned short scaling_shift : 4; + unsigned short reserved_fgs : 5; + unsigned short sw_random_seed; + short cb_offset; + short cr_offset; + char cb_mult; + char cb_luma_mult; + char cr_mult; + char cr_luma_mult; +} av1_fgs_cfg_t; + + +typedef struct _nvdec_av1_pic_s +{ + nvdec_pass2_otf_s encryption_params; + + nvdec_pass2_otf_ext_s ssm; + + av1_fgs_cfg_t fgs_cfg; + + // Driver may or may not use based upon need. + // If 0 then default value of 1<<27 = 298ms @ 450MHz will be used in ucode. + // Driver can send this value based upon resolution using the formula: + // gptimer_timeout_value = 3 * (cycles required for one frame) + unsigned int gptimer_timeout_value; + + unsigned int stream_len; // stream length. + unsigned int reserved12; // skip bytes length to real frame data . + + //sequence header + unsigned int use_128x128_superblock : 1; // superblock 128x128 or 64x64, 0:64x64, 1: 128x128 + unsigned int chroma_format : 2; // 1:420, others:reserved for future + unsigned int bit_depth : 4; // bitdepth + unsigned int enable_filter_intra : 1; // tool enable in seq level, 0 : disable 1: frame header control + unsigned int enable_intra_edge_filter : 1; + unsigned int enable_interintra_compound : 1; + unsigned int enable_masked_compound : 1; + unsigned int enable_dual_filter : 1; // enable or disable vertical and horiz filter selection + unsigned int reserved10 : 1; // 0 - disable order hint, and related tools + unsigned int reserved0 : 3; + unsigned int enable_jnt_comp : 1; // 0 - disable joint compound modes + unsigned int reserved1 : 1; + unsigned int enable_cdef : 1; + unsigned int reserved11 : 1; + unsigned int enable_fgs : 1; + unsigned int enable_substream_decoding : 1; //enable frame substream kickoff mode without context switch + unsigned int reserved2 : 10; // reserved bits + + //frame header + unsigned int frame_type : 2; // 0:Key frame, 1:Inter frame, 2:intra only, 3:s-frame + unsigned int show_frame : 1; // show frame flag + unsigned int reserved13 : 1; + unsigned int disable_cdf_update : 1; // disable CDF update during symbol decoding + unsigned int allow_screen_content_tools : 1; // screen content tool enable + unsigned int cur_frame_force_integer_mv : 1; // AMVR enable + unsigned int scale_denom_minus9 : 3; // The denominator minus9 of the superres scale + unsigned int allow_intrabc : 1; // IBC enable + unsigned int allow_high_precision_mv : 1; // 1/8 precision mv enable + unsigned int interp_filter : 3; // interpolation filter : EIGHTTAP_REGULAR,.... + unsigned int switchable_motion_mode : 1; // 0: simple motion mode, 1: SIMPLE, OBMC, LOCAL WARP + unsigned int use_ref_frame_mvs : 1; // 1: current frame can use the previous frame mv information, MFMV + unsigned int refresh_frame_context : 1; // backward update flag + unsigned int delta_q_present_flag : 1; // quantizer index delta values are present in the block level + unsigned int delta_q_res : 2; // left shift will apply to decoded quantizer index delta values + unsigned int delta_lf_present_flag : 1; // specified whether loop filter delta values are present in the block level + unsigned int delta_lf_res : 2; // specifies the left shift will apply to decoded loop filter values + unsigned int delta_lf_multi : 1; // seperate loop filter deltas for Hy,Vy,U,V edges + unsigned int reserved3 : 1; + unsigned int coded_lossless : 1; // 1 means all segments use lossless coding. Frame is fully lossless, CDEF/DBF will disable + unsigned int tile_enabled : 1; // tile enable + unsigned int reserved4 : 2; + unsigned int superres_is_scaled : 1; // frame level frame for using_superres + unsigned int reserved_fh : 1; + + unsigned int tile_cols : 8; // horizontal tile numbers in frame, max is 64 + unsigned int tile_rows : 8; // vertical tile numbers in frame, max is 64 + unsigned int context_update_tile_id : 16; // which tile cdf will be seleted as the backward update CDF, MAXTILEROW=64, MAXTILECOL=64, 12bits + + unsigned int cdef_damping_minus_3 : 2; // controls the amount of damping in the deringing filter + unsigned int cdef_bits : 2; // the number of bits needed to specify which CDEF filter to apply + unsigned int frame_tx_mode : 3; // 0:ONLY4x4,3:LARGEST,4:SELECT + unsigned int frame_reference_mode : 2; // single,compound,select + unsigned int skip_mode_flag : 1; // skip mode + unsigned int skip_ref0 : 4; + unsigned int skip_ref1 : 4; + unsigned int allow_warp : 1; // sequence level & frame level warp enable + unsigned int reduced_tx_set_used : 1; // whether the frame is restricted to oa reduced subset of the full set of transform types + unsigned int ref_scaling_enable : 1; + unsigned int reserved5 : 1; + unsigned int reserved6 : 10; // reserved bits + unsigned short superres_upscaled_width; // upscale width, frame_size_with_refs() syntax,restoration will use it + unsigned short superres_luma_step; + unsigned short superres_chroma_step; + unsigned short superres_init_luma_subpel_x; + unsigned short superres_init_chroma_subpel_x; + + /*frame header qp information*/ + unsigned char base_qindex; // the maximum qp is 255 + char y_dc_delta_q; + char u_dc_delta_q; + char v_dc_delta_q; + char u_ac_delta_q; + char v_ac_delta_q; + unsigned char qm_y; // 4bit: 0-15 + unsigned char qm_u; + unsigned char qm_v; + + /*cdef, need to update in the new spec*/ + unsigned int cdef_y_pri_strength; // 4bit for one, max is 8 + unsigned int cdef_uv_pri_strength; // 4bit for one, max is 8 + unsigned int cdef_y_sec_strength : 16; // 2bit for one, max is 8 + unsigned int cdef_uv_sec_strength : 16; // 2bit for one, max is 8 + + /*segmentation*/ + unsigned char segment_enabled; + unsigned char segment_update_map; + unsigned char reserved7; + unsigned char segment_temporal_update; + short segment_feature_data[8][8]; + unsigned char last_active_segid; // The highest numbered segment id that has some enabled feature. + unsigned char segid_preskip; // Whether the segment id will be read before the skip syntax element. + // 1: the segment id will be read first. + // 0: the skip syntax element will be read first. + unsigned char prevsegid_flag; // 1 : previous segment id is available + unsigned char segment_quant_sign : 8; // sign bit for segment alternative QP + + /*loopfilter*/ + unsigned char filter_level[2]; + unsigned char filter_level_u; + unsigned char filter_level_v; + unsigned char lf_sharpness_level; + char lf_ref_deltas[8]; // 0 = Intra, Last, Last2+Last3, GF, BRF, ARF2, ARF + char lf_mode_deltas[2]; // 0 = ZERO_MV, MV + + /*restoration*/ + unsigned char lr_type ; // restoration type. Y:bit[1:0];U:bit[3:2],V:bit[5:4] + unsigned char lr_unit_size; // restoration unit size 0:32x32, 1:64x64, 2:128x128,3:256x256; Y:bit[1:0];U:bit[3:2],V:bit[5:4] + + //general + frame_info_t current_frame; + ref_frame_struct_s ref_frame[7]; // Last, Last2, Last3, Golden, BWDREF, ALTREF2, ALTREF + + unsigned int use_temporal0_mvs : 1; + unsigned int use_temporal1_mvs : 1; + unsigned int use_temporal2_mvs : 1; + unsigned int mf1_type : 3; + unsigned int mf2_type : 3; + unsigned int mf3_type : 3; + unsigned int reserved_mfmv : 20; + + short mfmv_offset[3][7]; // 3: mf0~2, 7: Last, Last2, Last3, Golden, BWDREF, ALTREF2, ALTREF + char mfmv_side[3][7]; // flag for reverse offset great than 0 + // MFMV relative offset from the ref frame(reference to reference relative offset) + + unsigned char tileformat : 2; // 0: TBL; 1: KBL; + unsigned char gob_height : 3; // Set GOB height, 0: GOB_2, 1: GOB_4, 2: GOB_8, 3: GOB_16, 4: GOB_32 (NVDEC3 onwards) + unsigned char errorConcealOn : 1; // this field is not used, use ctrl_param.error_conceal_on to enable error concealment in ucode, + // always set NV_CNVDEC_GIP_ERR_CONCEAL_CTRL_ON = 1 to enable error detect in hw + unsigned char reserver8 : 2; // reserve + + unsigned char stream_error_detection : 1; + unsigned char mv_error_detection : 1; + unsigned char coeff_error_detection : 1; + unsigned char reserved_eh : 5; + + // Filt neighbor buffer offset + unsigned int Av1FltTopOffset; // filter top buffer offset respect to filter buffer, 256 bytes unit + unsigned int Av1FltVertOffset; // filter vertical buffer offset respect to filter buffer, 256 bytes unit + unsigned int Av1CdefVertOffset; // cdef vertical buffer offset respect to filter buffer, 256 bytes unit + unsigned int Av1LrVertOffset; // lr vertical buffer offset respect to filter buffer, 256 bytes unit + unsigned int Av1HusVertOffset; // hus vertical buffer offset respect to filter buffer, 256 bytes unit + unsigned int Av1FgsVertOffset; // fgs vertical buffer offset respect to filter buffer, 256 bytes unit + + unsigned int enable_histogram : 1; + unsigned int sw_skip_start_length : 14; //skip start length + unsigned int reserved_stat : 17; + +} nvdec_av1_pic_s; + +////////////////////////////////////////////////////////////////////// +// AV1 Buffer structure +////////////////////////////////////////////////////////////////////// +typedef struct _AV1FilmGrainMemory + { + unsigned char scaling_lut_y[256]; + unsigned char scaling_lut_cb[256]; + unsigned char scaling_lut_cr[256]; + short cropped_luma_grain_block[4096]; + short cropped_cb_grain_block[1024]; + short cropped_cr_grain_block[1024]; +} AV1FilmGrainMemory; + +typedef struct _AV1TileInfo_OLD +{ + unsigned char width_in_sb; + unsigned char height_in_sb; + unsigned char tile_start_b0; + unsigned char tile_start_b1; + unsigned char tile_start_b2; + unsigned char tile_start_b3; + unsigned char tile_end_b0; + unsigned char tile_end_b1; + unsigned char tile_end_b2; + unsigned char tile_end_b3; + unsigned char padding[6]; +} AV1TileInfo_OLD; + +typedef struct _AV1TileInfo +{ + unsigned char width_in_sb; + unsigned char padding_w; + unsigned char height_in_sb; + unsigned char padding_h; +} AV1TileInfo; + +typedef struct _AV1TileStreamInfo +{ + unsigned int tile_start; + unsigned int tile_end; + unsigned char padding[8]; +} AV1TileStreamInfo; + + +// AV1 TileSize buffer +#define AV1_MAX_TILES 256 +#define AV1_TILEINFO_BUF_SIZE_OLD NVDEC_ALIGN(AV1_MAX_TILES * sizeof(AV1TileInfo_OLD)) +#define AV1_TILEINFO_BUF_SIZE NVDEC_ALIGN(AV1_MAX_TILES * sizeof(AV1TileInfo)) + +// AV1 TileStreamInfo buffer +#define AV1_TILESTREAMINFO_BUF_SIZE NVDEC_ALIGN(AV1_MAX_TILES * sizeof(AV1TileStreamInfo)) + +// AV1 SubStreamEntry buffer +#define MAX_SUBSTREAM_ENTRY_SIZE 32 +#define AV1_SUBSTREAM_ENTRY_BUF_SIZE NVDEC_ALIGN(MAX_SUBSTREAM_ENTRY_SIZE * sizeof(nvdec_substream_entry_s)) + +// AV1 FilmGrain Parameter buffer +#define AV1_FGS_BUF_SIZE NVDEC_ALIGN(sizeof(AV1FilmGrainMemory)) + +// AV1 Temporal MV buffer +#define AV1_TEMPORAL_MV_SIZE_IN_64x64 256 // 4Bytes for 8x8 +#define AV1_TEMPORAL_MV_BUF_SIZE(w, h) ALIGN_UP( ALIGN_UP(w,128) * ALIGN_UP(h,128) / (64*64) * AV1_TEMPORAL_MV_SIZE_IN_64x64, 4096) + +// AV1 SegmentID buffer +#define AV1_SEGMENT_ID_SIZE_IN_64x64 128 // (3bits + 1 pad_bits) for 4x4 +#define AV1_SEGMENT_ID_BUF_SIZE(w, h) ALIGN_UP( ALIGN_UP(w,128) * ALIGN_UP(h,128) / (64*64) * AV1_SEGMENT_ID_SIZE_IN_64x64, 4096) + +// AV1 Global Motion buffer +#define AV1_GLOBAL_MOTION_BUF_SIZE NVDEC_ALIGN(7*32) + +// AV1 Intra Top buffer +#define AV1_INTRA_TOP_BUF_SIZE NVDEC_ALIGN(8*8192) + +// AV1 Histogram buffer +#define AV1_HISTOGRAM_BUF_SIZE NVDEC_ALIGN(1024) + +// AV1 Filter FG buffer +#define AV1_DBLK_TOP_SIZE_IN_SB64 ALIGN_UP(1920, 128) +#define AV1_DBLK_TOP_BUF_SIZE(w) NVDEC_ALIGN( (ALIGN_UP(w,64)/64 + 2) * AV1_DBLK_TOP_SIZE_IN_SB64) + +#define AV1_DBLK_LEFT_SIZE_IN_SB64 ALIGN_UP(1536, 128) +#define AV1_DBLK_LEFT_BUF_SIZE(h) NVDEC_ALIGN( (ALIGN_UP(h,64)/64 + 2) * AV1_DBLK_LEFT_SIZE_IN_SB64) + +#define AV1_CDEF_LEFT_SIZE_IN_SB64 ALIGN_UP(1792, 128) +#define AV1_CDEF_LEFT_BUF_SIZE(h) NVDEC_ALIGN( (ALIGN_UP(h,64)/64 + 2) * AV1_CDEF_LEFT_SIZE_IN_SB64) + +#define AV1_HUS_LEFT_SIZE_IN_SB64 ALIGN_UP(12544, 128) +#define AV1_ASIC_HUS_LEFT_BUFFER_SIZE(h) NVDEC_ALIGN( (ALIGN_UP(h,64)/64 + 2) * AV1_HUS_LEFT_SIZE_IN_SB64) +#define AV1_HUS_LEFT_BUF_SIZE(h) 2*AV1_ASIC_HUS_LEFT_BUFFER_SIZE(h) // Ping-Pong buffers + +#define AV1_LR_LEFT_SIZE_IN_SB64 ALIGN_UP(1920, 128) +#define AV1_LR_LEFT_BUF_SIZE(h) NVDEC_ALIGN( (ALIGN_UP(h,64)/64 + 2) * AV1_LR_LEFT_SIZE_IN_SB64) + +#define AV1_FGS_LEFT_SIZE_IN_SB64 ALIGN_UP(320, 128) +#define AV1_FGS_LEFT_BUF_SIZE(h) NVDEC_ALIGN( (ALIGN_UP(h,64)/64 + 2) * AV1_FGS_LEFT_SIZE_IN_SB64) + +// AV1 Hint Dump Buffer +#define AV1_HINT_DUMP_SIZE_IN_SB64 ((64*64)/(4*4)*8) // 8 bytes per CU, 256 CUs(2048 bytes) per SB64 +#define AV1_HINT_DUMP_SIZE_IN_SB128 ((128*128)/(4*4)*8) // 8 bytes per CU,1024 CUs(8192 bytes) per SB128 +#define AV1_HINT_DUMP_SIZE(w, h) NVDEC_ALIGN(AV1_HINT_DUMP_SIZE_IN_SB128*((w+127)/128)*((h+127)/128)) // always use SB128 for allocation + + +/******************************************************************* + New H264 +********************************************************************/ +typedef struct _nvdec_new_h264_pic_s +{ + nvdec_pass2_otf_s encryption_params; + unsigned char eos[16]; + unsigned char explicitEOSPresentFlag; + unsigned char hint_dump_en; //enable COLOMV surface dump for all frames, which includes hints of "MV/REFIDX/QP/CBP/MBPART/MBTYPE", nvbug: 200212874 + unsigned char reserved0[2]; + unsigned int stream_len; + unsigned int slice_count; + unsigned int mbhist_buffer_size; // to pass buffer size of MBHIST_BUFFER + + // Driver may or may not use based upon need. + // If 0 then default value of 1<<27 = 298ms @ 450MHz will be used in ucode. + // Driver can send this value based upon resolution using the formula: + // gptimer_timeout_value = 3 * (cycles required for one frame) + unsigned int gptimer_timeout_value; + + // Fields from msvld_h264_seq_s + int log2_max_pic_order_cnt_lsb_minus4; + int delta_pic_order_always_zero_flag; + int frame_mbs_only_flag; + int PicWidthInMbs; + int FrameHeightInMbs; + + unsigned int tileFormat : 2 ; // 0: TBL; 1: KBL; 2: Tile16x16 + unsigned int gob_height : 3 ; // Set GOB height, 0: GOB_2, 1: GOB_4, 2: GOB_8, 3: GOB_16, 4: GOB_32 (NVDEC3 onwards) + unsigned int reserverd_surface_format : 27; + + // Fields from msvld_h264_pic_s + int entropy_coding_mode_flag; + int pic_order_present_flag; + int num_ref_idx_l0_active_minus1; + int num_ref_idx_l1_active_minus1; + int deblocking_filter_control_present_flag; + int redundant_pic_cnt_present_flag; + int transform_8x8_mode_flag; + + // Fields from mspdec_h264_picture_setup_s + unsigned int pitch_luma; // Luma pitch + unsigned int pitch_chroma; // chroma pitch + + unsigned int luma_top_offset; // offset of luma top field in units of 256 + unsigned int luma_bot_offset; // offset of luma bottom field in units of 256 + unsigned int luma_frame_offset; // offset of luma frame in units of 256 + unsigned int chroma_top_offset; // offset of chroma top field in units of 256 + unsigned int chroma_bot_offset; // offset of chroma bottom field in units of 256 + unsigned int chroma_frame_offset; // offset of chroma frame in units of 256 + unsigned int HistBufferSize; // in units of 256 + + unsigned int MbaffFrameFlag : 1; // + unsigned int direct_8x8_inference_flag: 1; // + unsigned int weighted_pred_flag : 1; // + unsigned int constrained_intra_pred_flag:1; // + unsigned int ref_pic_flag : 1; // reference picture (nal_ref_idc != 0) + unsigned int field_pic_flag : 1; // + unsigned int bottom_field_flag : 1; // + unsigned int second_field : 1; // second field of complementary reference field + unsigned int log2_max_frame_num_minus4: 4; // (0..12) + unsigned int chroma_format_idc : 2; // + unsigned int pic_order_cnt_type : 2; // (0..2) + int pic_init_qp_minus26 : 6; // : 6 (-26..+25) + int chroma_qp_index_offset : 5; // : 5 (-12..+12) + int second_chroma_qp_index_offset : 5; // : 5 (-12..+12) + + unsigned int weighted_bipred_idc : 2; // : 2 (0..2) + unsigned int CurrPicIdx : 7; // : 7 uncompressed frame buffer index + unsigned int CurrColIdx : 5; // : 5 index of associated co-located motion data buffer + unsigned int frame_num : 16; // + unsigned int frame_surfaces : 1; // frame surfaces flag + unsigned int output_memory_layout : 1; // 0: NV12; 1:NV24. Field pair must use the same setting. + + int CurrFieldOrderCnt[2]; // : 32 [Top_Bottom], [0]=TopFieldOrderCnt, [1]=BottomFieldOrderCnt + nvdec_dpb_entry_s dpb[16]; + unsigned char WeightScale[6][4][4]; // : 6*4*4*8 in raster scan order (not zig-zag order) + unsigned char WeightScale8x8[2][8][8]; // : 2*8*8*8 in raster scan order (not zig-zag order) + + // mvc setup info, must be zero if not mvc + unsigned char num_inter_view_refs_lX[2]; // number of inter-view references + char reserved1[14]; // reserved for alignment + signed char inter_view_refidx_lX[2][16]; // DPB indices (must also be marked as long-term) + + // lossless decode (At the time of writing this manual, x264 and JM encoders, differ in Intra_8x8 reference sample filtering) + unsigned int lossless_ipred8x8_filter_enable : 1; // = 0, skips Intra_8x8 reference sample filtering, for vertical and horizontal predictions (x264 encoded streams); = 1, filter Intra_8x8 reference samples (JM encoded streams) + unsigned int qpprime_y_zero_transform_bypass_flag : 1; // determines the transform bypass mode + unsigned int reserved2 : 30; // kept for alignment; may be used for other parameters + + nvdec_display_param_s displayPara; + nvdec_pass2_otf_ext_s ssm; + +} nvdec_new_h264_pic_s; + +// golden crc struct dumped into surface +// for each part, if golden crc compare is enabled, one interface is selected to do crc calculation in vmod. +// vmod's crc is compared with cmod's golden crc (4*32 bits), and compare reuslt is written into surface. +typedef struct +{ + // input + unsigned int dbg_crc_enable_partb : 1; // Eable flag for enable/disable interface crc calculation in NVDEC HW's part b + unsigned int dbg_crc_enable_partc : 1; // Eable flag for enable/disable interface crc calculation in NVDEC HW's part c + unsigned int dbg_crc_enable_partd : 1; // Eable flag for enable/disable interface crc calculation in NVDEC HW's part d + unsigned int dbg_crc_enable_parte : 1; // Eable flag for enable/disable interface crc calculation in NVDEC HW's part e + unsigned int dbg_crc_intf_partb : 6; // For partb to select which interface to compare crc. see DBG_CRC_PARTE_INTF_SEL for detailed control value for each interface + unsigned int dbg_crc_intf_partc : 6; // For partc to select which interface to compare crc. see DBG_CRC_PARTE_INTF_SEL for detailed control value for each interface + unsigned int dbg_crc_intf_partd : 6; // For partd to select which interface to compare crc. see DBG_CRC_PARTE_INTF_SEL for detailed control value for each interface + unsigned int dbg_crc_intf_parte : 6; // For parte to select which interface to compare crc. see DBG_CRC_PARTE_INTF_SEL for detailed control value for each interface + unsigned int reserved0 : 4; + + unsigned int dbg_crc_partb_golden[4]; // Golden crc values for part b + unsigned int dbg_crc_partc_golden[4]; // Golden crc values for part c + unsigned int dbg_crc_partd_golden[4]; // Golden crc values for part d + unsigned int dbg_crc_parte_golden[4]; // Golden crc values for part e + + // output + unsigned int dbg_crc_comp_partb : 4; // Compare result for part b + unsigned int dbg_crc_comp_partc : 4; // Compare result for part c + unsigned int dbg_crc_comp_partd : 4; // Compare result for part d + unsigned int dbg_crc_comp_parte : 4; // Compare result for part e + unsigned int reserved1 : 16; + + unsigned char reserved2[56]; +}nvdec_crc_s; // 128 Bytes + +#endif // __DRV_NVDEC_H_ diff --git a/src/nouveau/headers/nvidia/video/nvenc_drv.h b/src/nouveau/headers/nvidia/video/nvenc_drv.h new file mode 100644 index 00000000000..d21c23beb35 --- /dev/null +++ b/src/nouveau/headers/nvidia/video/nvenc_drv.h @@ -0,0 +1,3830 @@ +/* + * SPDX-FileCopyrightText: Copyright (c) 1993-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved. + * SPDX-License-Identifier: MIT + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#ifndef NVENC_DRV_H +#define NVENC_DRV_H +// +// CLASS NV_C9B7_NVENC +// =================== +// +// +// 1 - INTRODUCTION +// +// The NVENC is a video encoder engine. Currently only H.264 baseline, main +// and high profile are supported. For baseline, only 1 slicegroup is allowed. +// NVENC has built-in motion estimation that enables it to do encode on its own +// but also has the ability to work with external hints that could be produced +// by another motion search implementation such as CEA.
+// NVENC motion search is divided into a simple integer search module and a high +// perf subpel refinement module. The integer search capability of NVENC is +// limited (perf wise) so for higher perf the CEA approach should be used. +// Motion search/estimation (ME) results in a set of vectors with scores +// which it will evaluate for encoding cost and eligibility. It also searches +// intra encoding modes for the lowest cost encoding, picks a final winner +// from all inter and intra modes that have been evaluated and encodes the +// macroblock according to that mode. Skip and direct modes are also checked. +// Search and encoding involve forward dct and quantization, inverse dct and +// quantization, reconstruction, and optionally deblocking. Entropy encode can +// do either CAVLC or CABAC. +// +// +// 1.4. CLASS OPERATION +// ------------------------ +// +// *** TBD *** +// +// If MVC is enabled, HW need to write the prefix NAL unit(14) when current view is a base view. +// As SVC is not supported in nvenc, so svc_extension_flag should be always false. Then, in the prefix NAL unit, +// only the syntax elements(H.7.3.1.1 NAL unit header MVC extension syntax) in the prefix NAL unit should be written: +// nal_unit_header_mvc_extension( ) { +// non_idr_flag +// priority_id +// view_id +// temporal_id +// anchor_pic_flag +// inter_view_flag +// reserved_one_bit +// } +// +// 1.5. RESTRICTIONS +// ------------------------ +// +// (1) Reference pictures to be accessed by NVENC must be in tiled 16x16 format +// (2) Input pictures are either all in BL format or all in tiled 16x16 format +// (3) Input pictures and reference pictures must be of the kind specified by +// 'pic_struct' (either fields or frames). NVENC cannot encode an interlaced +// input picture (2 fields) as a frame or vice versa, and cannot use frame +// reference pictures when encoding a field picture, or field reference +// pictures when encoding a frame. +// +// +// +// 2 - APPLICATION MEMORY STRUCTURES +// +// 2.1. DRV PICTURE SETUP BUFFER +// ----------------------------- +// +// DrvPictureSetupBuffer contains sequence, picture, and slice level control +// data required for encoding. +// +// Input pictures (to be encoded) can be any size (HxV) but the encoded output +// is always an integer number of macroblocks high and wide, which means that +// H and V are rounded up to the next multiple of 16 (H'xV') and all pixels +// within this range are used to encode the macroblocks. NVENC does not clear, +// pad, duplicate or otherwise modify any pixels values within the H'xV' sized +// input picture contained in the input surface. +// The values for frame/pic_width_minus1 and frame/pic_height_minus1 contained +// in the config structures are the original width and height (display size) +// of the pictures, and from this the nvenc app will derive the cropping +// information. Cropping will only be done on the right and bottom side of the +// picture. + +typedef struct +{ + unsigned int luma_log2_weight_denom : 3; // the base 2 logarithm of the denominator for all luma weighting factors. The value of luma_log2_weight_denom shall be in the range of 0 to 7, inclusive + unsigned int chroma_log2_weight_denom : 3; // the base 2 logarithm of the denominator for all chroma weighting factors. The value of chroma_log2_weight_denom shall be in the range of 0 to 7, inclusive. + unsigned int reversed : 26; + + unsigned char luma_weight_l0_flag[16]; // equal to 1 specifies that weighting factors for the luma component of list 0 prediction are present, luma_weight_l0_flag[8~15] not used now + unsigned char chroma_weight_l0_flag[16]; // equal to 1 specifies that weighting factors for the chroma prediction values of list 0 prediction are present, chroma_weight_l0_flag[8~15] not used now + + unsigned char luma_weight_l1_flag[16]; // have the same semantics as luma_weight_l0_flag, chroma_weight_l0_flag, not used now + unsigned char chroma_weight_l1_flag[16]; // not used now + + short luma_weight_l0[16]; // is the weighting factor applied to the luma prediction value for list 0 prediction using RefPicList0[ i ]. When luma_weight_l0_flag is equal to 1, the value of luma_weight_l0[ i ] shall be in the range of -128 to 127, inclusive. -128 to 255 for H265 + short luma_offset_l0[16]; // is the additive offset applied to the luma prediction value for list 0 prediction using RefPicList0[ i ]. The value of luma_offset_l0[ i ] shall be in the range of -128 to 127, inclusive. -2^9 to 2^9-1 for h265 + short luma_weight_l1[16]; // is the weighting factor applied to the luma prediction value for list 0 prediction using RefPicList0[ i ]. When luma_weight_l0_flag is equal to 1, the value of luma_weight_l0[ i ] shall be in the range of -128 to 127, inclusive. -128 to 255 for H265 + short luma_offset_l1[16]; // is the additive offset applied to the luma prediction value for list 0 prediction using RefPicList0[ i ]. The value of luma_offset_l0[ i ] shall be in the range of -128 to 127, inclusive. -2^9 to 2^9-1 for h265 + short chroma_weight_l0[2][16]; // has the similar meaning with luma_weight/offset + short chroma_offset_l0[2][16]; // has the similar meaning with luma_weight/offset + short chroma_weight_l1[2][16]; // has the similar meaning with luma_weight/offset + short chroma_offset_l1[2][16]; // has the similar meaning with luma_weight/offset + +} nvenc_pred_weight_table_s; // 452 bytes + +typedef struct +{ + unsigned short frame_width_minus1; // frame width in pixels minus 1, range 0-4095 + unsigned short frame_height_minus1; // frame height in pixels minus 1, range 0-4095 + unsigned short sfc_pitch; // pitch of luma plane + unsigned short sfc_pitch_chroma; // pitch of chroma plane + unsigned short sfc_trans_mode; // least 3 significient bits are used to stand for 8 modes, normal, xflip, yflip etc. + unsigned short reserved2; // pad to int + + // do not use offset from cmod trace as golden, cmod behavior is different. + unsigned int luma_top_frm_offset; // offset of luma top field or frame in units of 256 + unsigned int luma_bot_offset; // offset of luma bottom field in units of 256. Not used if frame format. + unsigned int chroma_top_frm_offset; // offset of chroma top field or frame, or offset of chroma U(Cb) plane in planar mode, both in units of 256 + unsigned int chroma_bot_offset; // offset of chroma bottom field, not used if frame format. Or offset of chroma V(Cr) plane in planar mode. Both in units of 256. + unsigned int block_height : 7; // BL mapping block height setting + unsigned int tiled_16x16 : 1; // Surface is 16x16 tiled instead of BL mapped (must be 1 for refpics) + unsigned int memory_mode : 2; // 0: semi-planar, 1: planar + unsigned int nv21_enable : 1; // the surface format is yuv or yvu + unsigned int input_bl_mode : 2; // the input block linear mode: 0~gpu bl; 1~tegra bl; 2~naive bl this is only for cmod and cmod/plugin + unsigned int reserved : 19; // pad to NvU32 +} nvenc_h264_surface_cfg_s; // 32 bytes + +typedef struct +{ + unsigned int profile_idc : 8; // seq_parameters(_ext) regs + unsigned int level_idc : 8; + unsigned int chroma_format_idc : 2; // 0=monochrome, 1=yuv4:2:0, 2-3=not supported + unsigned int pic_order_cnt_type : 2; // only support values 0 and 2 + unsigned int log2_max_frame_num_minus4 : 4; + unsigned int log2_max_pic_order_cnt_lsb_minus4 : 4; + unsigned int frame_mbs_only : 1; + //mvc + unsigned int stereo_mvc_enable : 1; //if this bit is enabled, the profile_idc should be 128 + unsigned int separate_colour_plane_flag : 1; //if this bit is enabled, the chroma_format_idc should be 3 + unsigned int lossless_qpprime_flag : 1; //if this bit is enabled and QP=0, bypass trans/q and encoded as lossless MB (hori or vert) +} nvenc_h264_sps_data_s; // 4 bytes + +typedef struct +{ + unsigned int pic_param_set_id : 8; // picture parameter set identification + unsigned int entropy_coding_mode_flag : 1; // select entropy copding mode: cabac, cavlc + unsigned int num_ref_idx_l0_active_minus1 : 5; // number of currently active reference pictures in list 0 minus 1, range 0..30 + unsigned int num_ref_idx_l1_active_minus1 : 5; // number of currently active reference pictures in list 1 minus 1, range 0..30 + unsigned int weighted_bipred_idc : 2; // weighted prediction mode for B: only 0 (default) and 2 (implicit) are supported + int pic_init_qp_minus26 : 6; // initial QP value, range -26..+25 + int chroma_qp_index_offset : 5; // offset to add to chroma QPy for QPc table indexing for Cb, range -12..+12 + + int second_chroma_qp_index_offset : 5; // offset to add to chroma QPy for QPc table indexing for Cr, range -12..+12 + unsigned int constrained_intra_pred_flag : 1; // if set, intra prediction can only use pixels from macroblocks that are also intra coded + unsigned int deblocking_filter_control_present_flag : 1;// if set, deblock filter controls syntax elements will be present/encoded in the stream + unsigned int transform_8x8_mode_flag : 1; // if set, enables the use of transform size 8x8 (flags will be encoded in stream per macroblock) + unsigned int pic_order_present_flag : 1; // if set, pic_order control info is encoded in the stream + unsigned int weighted_pred_flag : 1; + unsigned int reserved0 : 22; // pad to full int +} nvenc_h264_pps_data_s; // 8 bytes + +typedef struct +{ + unsigned int self_temporal_stamp_l0 : 3; // stamp to use for L0 integer search in stamp based mode + unsigned int self_temporal_stamp_l1 : 3; // stamp to use for L1 integer search in stamp based mode + unsigned int self_temporal_explicit : 1; // explicitly evaluate these vectors + unsigned int self_temporal_search : 1; // integer search enable for this hint + unsigned int self_temporal_refine : 1; // subpel search enable for this hint + unsigned int self_temporal_enable : 1; // hint enable (enables processing + fetching of data) (only changeable at first slice in picture) + + unsigned int coloc_stamp_l0 : 3; // stamp to use for L0 integer search in stamp based mode + unsigned int coloc_stamp_l1 : 3; // stamp to use for L1 integer search in stamp based mode + unsigned int coloc_explicit : 1; // explicitly evaluate these vectors + unsigned int coloc_search : 1; // integer search enable for this hint + unsigned int coloc_refine : 1; // subpel search enable for this hint + unsigned int coloc_enable : 1; // hint enable (enables processing + fetching of data) (only changeable at first slice in picture) + + unsigned int self_spatial_stamp_l0 : 3; // stamp to use for L0 integer search in stamp based mode + unsigned int self_spatial_stamp_l1 : 3; // stamp to use for L1 integer search in stamp based mode + unsigned int self_spatial_explicit : 1; // explicitly evaluate these vectors + unsigned int self_spatial_search : 1; // integer search enable for this hint + unsigned int self_spatial_refine : 1; // subpel search enable for this hint + unsigned int self_spatial_enable : 1; // hint enable (enables processing + fetching of data) (only changeable at first slice in picture) + unsigned int reserved0 : 2; // pad to full int + + unsigned int external_stamp_l0_refidx0_stamp : 3; // Stamp to use for external L0 hint with refidx = 0 + unsigned int external_stamp_l0_refidx1_stamp : 3; // Stamp to use for external L0 hint with refidx = 1 + unsigned int external_stamp_l0_refidx2_stamp : 3; // Stamp to use for external L0 hint with refidx = 2 + unsigned int external_stamp_l0_refidx3_stamp : 3; // Stamp to use for external L0 hint with refidx = 3 + unsigned int external_stamp_l0_refidx4_stamp : 3; // Stamp to use for external L0 hint with refidx = 4 + unsigned int external_stamp_l0_refidx5_stamp : 3; // Stamp to use for external L0 hint with refidx = 5 + unsigned int external_stamp_l0_refidx6_stamp : 3; // Stamp to use for external L0 hint with refidx = 6 + unsigned int external_stamp_l0_refidx7_stamp : 3; // Stamp to use for external L0 hint with refidx = 7 + unsigned int external_stamp_l1_refidx0_stamp : 3; // Stamp to use for external L1 hint with refidx = 0 + unsigned int external_stamp_l1_refidx1_stamp : 3; // Stamp to use for external L1 hint with refidx = 1 + unsigned int reserved1 : 2; // pad to full int + + + unsigned int external_stamp_l1_refidx2_stamp : 3; // Stamp to use for external L1 hint with refidx = 2 + unsigned int external_stamp_l1_refidx3_stamp : 3; // Stamp to use for external L1 hint with refidx = 3 + unsigned int external_stamp_l1_refidx4_stamp : 3; // Stamp to use for external L1 hint with refidx = 4 + unsigned int external_stamp_l1_refidx5_stamp : 3; // Stamp to use for external L1 hint with refidx = 5 + unsigned int external_stamp_l1_refidx6_stamp : 3; // Stamp to use for external L1 hint with refidx = 6 + unsigned int external_stamp_l1_refidx7_stamp : 3; // Stamp to use for external L1 hint with refidx = 7 + + + unsigned int external_explicit : 1; // explicitly evaluate these vectors + unsigned int external_search : 1; // integer search enable for this hint + unsigned int external_refine : 1; // subpel search enable for this hint + unsigned int external_enable : 1; // hint enable (enables processing + fetching of data) (only changeable at first slice in picture) + + unsigned int const_mv_stamp_l0 : 3; // stamp to use for L0 integer search in stamp based mode + unsigned int const_mv_stamp_l1 : 3; // stamp to use for L1 integer search in stamp based mode + unsigned int const_mv_explicit : 1; // explicitly evaluate these vectors + unsigned int const_mv_search : 1; // integer search enable for this hint + unsigned int const_mv_refine : 1; // subpel search enable for this hint + unsigned int const_mv_enable : 1; // hint enable (enables processing + fetching of data) + + unsigned int stamp_refidx1_stamp : 3; // Stamp to use for multi-ref hint with refidx = 1 + unsigned int stamp_refidx2_stamp : 3; // Stamp to use for multi-ref hint with refidx = 2 + unsigned int stamp_refidx3_stamp : 3; // Stamp to use for multi-ref hint with refidx = 3 + unsigned int stamp_refidx4_stamp : 3; // Stamp to use for multi-ref hint with refidx = 4 + unsigned int stamp_refidx5_stamp : 3; // Stamp to use for multi-ref hint with refidx = 5 + unsigned int stamp_refidx6_stamp : 3; // Stamp to use for multi-ref hint with refidx = 6 + unsigned int stamp_refidx7_stamp : 3; // Stamp to use for multi-ref hint with refidx = 7 + unsigned int reserved2 : 11; // pad to full int + +} nvenc_h264_me_hint_cfg_s; //16 bytes + +typedef struct +{ + unsigned int mvx_frac : 2; // vector X component fraction + int mvx_int : 12; // vector X component integer portion + unsigned int reserved_x : 2; // padding X to 16 bit + unsigned int mvy_frac : 2; // vector Y component fraction + int mvy_int : 10; // vector Y component integer portion + unsigned int reserved_y : 4; // padding Y to 16 bit +} nvenc_h264_me_const_vec_s; // 4 bytes + +typedef struct +{ + unsigned int bitmask[2] ; // 8x8 search point pattern bitmask + unsigned int hor_adjust : 1; // shift hor based on lsb of pred + unsigned int ver_adjust : 1; // shift ver based on lsb of pred + unsigned int reserved : 30; +} nvenc_h264_me_stamp_s; // 12 bytes + +typedef struct +{ + unsigned int mv_limit_enable : 1; // 0=disable, 1=enable set the absolute MV range (Cisco flip camera request) + unsigned int left_mvx_frac : 2; // vector X component fraction + int left_mvx_int : 12; // vector X component integer portion + unsigned int reserved1 : 1; // padding X to 16 bit + unsigned int top_mvy_frac : 2; // vector Y component fraction + int top_mvy_int : 10; // vector Y component integer portion + unsigned int reserved2 : 4; // padding Y to 16 bit + + unsigned int right_mvx_frac : 2; // vector X component fraction + int right_mvx_int : 12; // vector X component integer portion + unsigned int reserved3 : 2; // padding X to 16 bit + unsigned int bottom_mvy_frac : 2; // vector Y component fraction + int bottom_mvy_int : 10; // vector Y component integer portion + unsigned int reserved4 : 4; // padding Y to 16 bit +} nvenc_h264_me_limit_mv_s; // 8 bytes + +typedef struct +{ + unsigned char teb_K ; // TEB K + unsigned char teb_N ; // TEB N + unsigned char teb_P ; // TEB P + unsigned char teb_S : 4; // TEB S + unsigned char teb_mode : 4; // TEB mode +} nvenc_me_tebif_s; + +typedef struct +{ + unsigned int me_predictor_mode : 1; // 0=MDP, 1=const + unsigned int refinement_mode : 1; // 0=hpel, 1=qpel + unsigned int lambda_mode : 1; // 0=MDP, 1=const + unsigned int const_lambda : 11; // U8.3 format lambda, used if lambda_mode=1 + unsigned int refine_on_search_enable : 1; // 0=disable, 1=enable subpel refine for search candidates + unsigned int me_only_mode : 1; + unsigned int fps_mvcost : 1; + unsigned int sps_mvcost : 1; + unsigned int sps_cost_func : 1; + unsigned int me_icc_mode_mad : 1; + + unsigned int sps_filter : 3; + unsigned int mc_filter : 3; + unsigned int vc1_fastuv_mc : 1; + unsigned int vc1_mc_rnd : 1; + unsigned int mbc_ctrl_arbitor : 2; + unsigned int mv_only_enable : 1; + unsigned int average_mvhint_enable : 1; + + nvenc_h264_me_limit_mv_s limit_mv; + nvenc_h264_me_hint_cfg_s predsrc; // predictor sources + nvenc_h264_me_const_vec_s l0_hint; // constant MV used as L0 hint + nvenc_h264_me_const_vec_s l1_hint; // constant MV used as L1 hint + nvenc_h264_me_const_vec_s l0_pred; // constant MV used as L0 predictor + nvenc_h264_me_const_vec_s l1_pred; // constant MV used as L1 predictor + nvenc_h264_me_stamp_s shape0; // priority 0 stamping pattern + nvenc_h264_me_stamp_s shape1; // priority 1 stamping pattern + nvenc_h264_me_stamp_s shape2; // priority 2 stamping pattern + nvenc_h264_me_stamp_s shape3; // priority 3 stamping pattern + nvenc_h264_me_stamp_s shape4; // priority 4 stamping pattern + nvenc_h264_me_stamp_s shape5; // priority 5 stamping pattern + nvenc_h264_me_stamp_s shape6; // priority 6 stamping pattern + nvenc_h264_me_stamp_s shape7; // priority 7 stamping pattern + nvenc_me_tebif_s teb_para; // TEB parameter + unsigned int mbc_mb_size : 9; // MBC size in MB + unsigned int rmvp_source_half_scaled : 1; // only support picture size half scale down in horz&vert + unsigned int partDecisionMadeByFPP : 1; // enable CU32 FPP; + unsigned int penalty_factor_FPP : 6; // MV with cost no larger than min_cost + ((min_cost * factor) >> 8) will go to SPS, not used from nvenc7 + unsigned int spatial_hint_pattern : 4; + unsigned int temporal_hint_pattern : 6; + unsigned int Cu16partDecisionMadeByFPP : 1; // enable CU16 making part decision + unsigned int sps_mvp_from_fbm : 1; //1:sps use fbm'MVP and lambda;0:based on me_predictor_mode and lambda_mode to set the mvp and lambda for sps + unsigned int fbm_select_best_cu16_parttype_num : 3; // #winner partitions selected for cu16x16. valid range is 0~7; + unsigned int fbm_op_winner_num_p_frame : 3; // FBM output number of integer winners for each PU. when fbm_op_winner_num_p_frame_cu8 has non-zero value, this field means the PU winners for CU64~CU16, i.e. w/o CU8 + unsigned int fbm_op_winner_num_b_frame_l0 : 3; // FBM output number of integer winners for each PU + unsigned int fbm_op_winner_num_b_frame_l1 : 3; // FBM output number of integer winners for each PU + unsigned int fbm_select_best_cu32_parttype_num : 3; // cu32 select best parttype num + unsigned int sps_evaluate_merge_cand_num : 3; // number of merg candidate for sps to explicit evaluate + unsigned int fps_quad_thresh_hold : 10; + unsigned int external_hint_order : 1; //0:Insert external hint at first ref; 1:insert external hint at last ref; + unsigned int coloc_hint_order : 1; //0:Insert coloc hint at first ref; 1: insert coloc hint at last ref. + unsigned int reserved1 : 5; + short ct_threshold ; // for OFS only, threshold in Census Transform calculation + //(1<> 8) + + unsigned int l0_part_16x16_enable : 1; // enable L0 prediction for 16x16 + unsigned int l0_part_16x8_enable : 1; // enable L0 prediction for 16x8 + unsigned int l0_part_8x16_enable : 1; // enable L0 prediction for 8x16 + unsigned int l0_part_8x8_enable : 1; // enable L0 prediction for 8x8 + unsigned int l0_part_8x4_enable : 1; // enable L0 prediction for 8x4 + unsigned int l0_part_4x8_enable : 1; // enable L0 prediction for 4x8 + unsigned int l0_part_4x4_enable : 1; // enable L0 prediction for 4x4 + unsigned int l0_part_reserved : 1; // reserved for future L0 prediction extensions + unsigned int l1_part_16x16_enable : 1; // enable L1 prediction for 16x16 + unsigned int l1_part_16x8_enable : 1; // enable L1 prediction for 16x8 + unsigned int l1_part_8x16_enable : 1; // enable L1 prediction for 8x16 + unsigned int l1_part_8x8_enable : 1; // enable L1 prediction for 8x8 + unsigned int l1_part_reserved : 4; // reserved for future L1 prediction extensions + unsigned int bi_part_16x16_enable : 1; // enable Bi prediction for 16x16 + unsigned int bi_part_16x8_enable : 1; // enable Bi prediction for 16x8 + unsigned int bi_part_8x16_enable : 1; // enable Bi prediction for 8x16 + unsigned int bi_part_8x8_enable : 1; // enable Bi prediction for 8x8 + unsigned int bi_part_reserved : 4; // reserved for future Bi prediction extensions + unsigned int bdirect_mode : 2; // 0: disable, 1: spatial, 2: temporal + unsigned int bskip_enable : 1; // enable b_skip encoding (requires b_direct encoding) + unsigned int pskip_enable : 1; // enable p_skip encoding + unsigned int special_reserved : 4; // reserved for future special prediction extensions + + short bias_inter_16x16; // bias for inter 16x16 (subtracts from inter 16x16 cost) + short bias_inter_16x8; // bias for inter 16x8 (subtracts from inter 16x8 cost) + + short bias_inter_8x16; // bias for inter 8x16 (subtracts from inter 8x16 cost) + short bias_inter_8x8; // bias for inter 8x8 (subtracts from inter 8x8 cost) + + short bias_pskip; // bias for skip (subtracts from skip cost) + short bias_bdir; // bias for bdirect (subtracts from bdirect cost) + + short bias_intra_over_inter; // bias for any intra (subtracts from intra cost) + short bias_intra_16x16; // bias for intra 16x16 (subtracts from intra 16x16 cost) + + short bias_intra_8x8; // bias for intra 8x8 (subtracts from intra 8x8 cost) + short bias_intra_4x4; // bias for intra 4x4 (subtracts from intra 4x4 cost) + + short bias_intra_most_prob; // bias for intra most probable mode + unsigned short mv_cost_bias; // bias applied to internal motion vector cost calculation + + unsigned short intra_nxn_bias_multiplier; // 0..255 range; used for internal bias calculation + unsigned short intra_most_prob_bias_multiplier; // 0..31 range; used for internal bias calculation + + short pskip_bias_multiplier; // S16 range; used for internal bias calculation + short bdirect_bias_multiplier; // S16 range; used for internal bias calculation + + unsigned short pskip_esc_threshold; // if pskip cost is below this, disable full search & attempt pskip encode + unsigned short bdirect_esc_threshold; // if bdirect cost is below this, disable full search & attempt bdirect encode + + unsigned short early_intra_disable_mpeb_threshold; // if early intra cost is below this, disable full search & select intra + unsigned short tempmv_wt_spread_threshold; // The threshold for the sum of differences between 8x8s and 16x16 mvs. Used for temporal mv weight + unsigned int tempmv_wt_distort_threshold : 16; // The threshold for the cost of the 16x16 mv. Used for temporal mv weight + unsigned int mv_cost_predictor_control : 1; // 0=use 16x16 predictor only; 1=use accurate predictor (slower but better) + unsigned int mv_cost_enable : 1; // enable mv cost calculations in mode decision + unsigned int intra_most_prob_force_on : 1; // always evaluate most probably 4x4 or 8x8 intra mode during full search + unsigned int early_intra_mode_control : 2; // 0=inter; 1=eval_intra; 2=ext_hint; 3=none (low perf,best qual) + unsigned int early_intra_mode_type_16x16dc : 1; // evaluate 16x16dc if early_intra_mode_control==1 + unsigned int early_intra_mode_type_16x16h : 1; // evaluate 16x16h if early_intra_mode_control==1 + unsigned int early_intra_mode_type_16x16v : 1; // evaluate 16x16v if early_intra_mode_control==1 + unsigned int early_ip_is_final : 1; // if set, result of early intra is final (disable full intra search) + unsigned int chroma_eval_mode : 1; // for intra chroma search 0 = U and V; 1 = U only + unsigned int ip_search_mode : 3; // which intra sizes to search before making IP decision b0=4x4,b1=8x8,b2=16x16 + unsigned int multiply_bias_with_lambda : 1; // setting this to 1 will result in bias being multiplied by lambda + unsigned int force_ipcm : 1; // 0=normal encode; 1=force mpeb/mpec encode as IPCM + unsigned int early_termination_ip1 : 1; // enable IP1 early termination + short bias_favor_intra_16x16_early; // S16 range; used for internal bias calculation + unsigned short priority_ipred_type_ip1 : 3; // which intra type search first for early termination, must be a subset of ip_search_mode + unsigned short intra_ssd_cnt_4x4 : 4; // SSD cost evaluation between N top intra modes selected by satd + unsigned short intra_ssd_cnt_8x8 : 4; // SSD cost evaluation between N top intra modes selected by satd + unsigned short intra_ssd_cnt_16x16 : 4; // SSD cost evaluation between N top intra modes selected by satd + unsigned short skip_evaluate_enable : 1; // 0: not check skip ssd cost; 1, evaluate skip ssd cost + unsigned int rdo_level : 3; // 0:mode decision by satd 1:mode decision by ssd + unsigned int tu_search_num : 3; // tu search part num per MB + unsigned int luma_residual_zero_eval : 2; // luma residual zero cost evaluate, bit0:intra; bit1:inter + unsigned int num_1div8_lambda_ssd : 7; // number of 1/8 lambda ssd for lambda_coef delta. lambda_coef = (1+num_1div8_lambda_ssd*1/8)*lambda_ssd + unsigned int ofs_mvx_range : 2; // control mvx precision/range, 0: ouptput format: S10.5, range[-1024, 1023]; 1: output format: S11.4, range[-2048, 2047]; 2: output format: S12.3, range[-4096, 4095]; 3: output format: S12.2, range[-4096, 4095] + unsigned int reserved4 : 15; // pad to 32 bytes + short bias_tu_4x4; // bias for TU 4x4 + short bias_tu_8x8; // bias for TU 8x8 + unsigned int ssim_rdo : 2; // SSIM-RDO: 0 disable, 1: enable SW SSIM-RDO, 2: enable HW SSIM-RDO, + unsigned int calc_ssim : 1; // Calculate SSIM distortion: 0 disable, 1: enable. ssim_rdo = 1 => calc_ssim = 1 + unsigned int reserved5 : 29; // reserved + unsigned int ssd_over_ssim_factor; // distortion ssd/ssim ratio, in U24.8 format + unsigned int reserved6[13]; // reserved +} nvenc_h264_md_control_s; // 128 bytes + +typedef struct +{ + unsigned short qpp_run_vector_4x4; // cost values for 4x4 transform (16 bit total vector) + unsigned short qpp_run_vector_8x8[3]; // cost values for 8x8 transform (48 bit total vector) + // 2 bits each for first 12 coefs, then 1 bit each for next 24 + unsigned char qpp_luma8x8_cost; // luma 8x8 cost threshold, 0 = throw out all coefs, range 0-15 + unsigned char qpp_luma16x16_cost; // luma 16x16 cost threshold, 0 = throw out all coefs, range 0-15 + unsigned char qpp_chroma_cost; // chroma cost threshold, 0 = throw out all coefs, range 0-15 + unsigned char qpp_mode : 2; // 0 = OFF, 1 = 8x8, 2 = 16x16_8x8 + unsigned char reserved1 : 4; // padding + //add by cl + unsigned char quant_intra_sat_flag : 1; // 0: no saturation is applied; 1: saturation is applied + unsigned char quant_inter_sat_flag : 1; // 0: no saturation is applied; 1: saturation is applied + unsigned int quant_intra_sat_limit : 16; // quantization saturation limit for intra MB + unsigned int quant_inter_sat_limit : 16; // quantization saturation limit for inter MB + //~ + unsigned short dz_4x4_YI[16]; // deadzone for 4x4 transform of Luma Intra + unsigned short dz_4x4_YP[16]; // deadzone for 4x4 transform of Luma Inter + unsigned short dz_4x4_CI; // deadzone for 4x4 transform of Chroma Intra + unsigned short dz_4x4_CP; // deadzone for 4x4 transform of Chroma Inter + unsigned short dz_8x8_YI[16]; // deadzone for 8x8 transform of Luma Intra, 16 values are mapped to 64 coefs + unsigned short dz_8x8_YP[16]; // deadzone for 8x8 transform of Luma Inter, 16 values are mapped to 64 coefs + unsigned int reserved2[11]; // pad to 192 bytes to allow 3 reads of 64 bytes in ucode + // this is optimal size as 160=5 reads of 32 will exceed dma queue +} nvenc_h264_quant_control_s; // 192 bytes + +typedef struct +{ + unsigned int reorder_l0_cmd_count : 4; + unsigned int reorder_l1_cmd_count : 4; + unsigned int mmco_cmd_count : 4; + unsigned int no_output_of_prior_pic_flag : 1; + unsigned int long_term_ref_flag : 1; + unsigned int reserved :18; +} nvenc_h264_refpic_cmd_s; // 4 bytes + +typedef struct +{ + // ref_pic_list_reorder_cmd includes both opcode and op arguments + unsigned int reordering_of_pic_nums_idc : 3; // Bit 31-29 (0-5 (Table 7-7) + unsigned int abs_diff_pic_num_minus1 :17; // (opcode=0/1) + unsigned int long_term_pic_num : 4; // (opcode=2): + unsigned int abs_diff_view_idx_minus1 : 4; // (opcode=4/5): + unsigned int reserved : 4; +} nvenc_h264_ref_pic_reorder_s; // 4 bytes + +typedef struct +{ + unsigned int mmco_cmd_id : 3; // totally 6 commands + + unsigned int difference_of_pic_nums_minus1 :17; // 1) mark a short_term ref as long (mmco=3); 2) mark a short_term as un-used (mmco=1) + unsigned int long_term_pic_num : 3; // [0 to MaxLongTermFrameIdx ]; mark a long as "unused for ref" (mmco=2); + unsigned int long_term_frame_idx : 3; // [0 to MaxLongTermFrameIdx] 1) mark a picture to long (mmco=3, mmco=6) + unsigned int max_long_term_frame_idx_plus1 : 4; // [0, num_ref_frames] max long term frm indexes allowed for long term pictures. + // max_long_term_frame_idx_plus1=0 -> MaxLongTermFrameIdx="no-long-term" + // max_long_term_frame_idx_plus1>0 -> MaxLongTermFrameIdx= max_long_term_frame_idx_plus1-1 + unsigned int reserved : 2; +} nvenc_h264_mmco_s; // 4 bytes + +typedef struct +{ + // int0 + unsigned int num_mb : 19; // number of macroblocks in this slice, support upto 8kx8k + unsigned int qp_avr : 8; // 6bit value used by hw. Other bits can be used for rounding. + unsigned int reserved : 5; + // int1 + unsigned int slice_tgt_rate; // target bit rate (size) (if RC mode == slice) + + // int2 + char slice_alpha_c0_offset_div2 : 4; + char slice_beta_offset_div2 : 4; + unsigned char qp_slice_min; // min slice qp value + unsigned short qp_slice_max : 8; // max slice qp value + unsigned short force_intra : 1; // force entire slice to be intra + unsigned short disable_deblocking_filter_idc : 2; + unsigned short cabac_init_idc : 2; + unsigned short disable_slice_header : 1; // if set to 1, disables HW slice header output + unsigned short reserved0 : 2; + + // int3 + unsigned int num_ref_idx_active_override_flag : 1; + unsigned int num_ref_idx_l0_active_minus1 : 5; // [0.31] + unsigned int num_ref_idx_l1_active_minus1 : 5; // [0,31] + unsigned int reserved1 : 9; // padding + unsigned int me_control_idx : 4; // index in array of nvenc_h264_me_control_s to use for this slice + unsigned int md_control_idx : 4; // index in array of nvenc_h264_md_control_s to use for this slice + unsigned int q_control_idx : 4; // index in array of nvenc_h264_quant_control_s to use for this slice + + // int 4 + unsigned int limit_slice_top_boundary : 1; // limit ME to slice boundaries + unsigned int limit_slice_bot_boundary : 1; // limit ME to slice boundaries + unsigned int limit_slice_left_boundary : 1; // limit ME to slice boundaries + unsigned int limit_slice_right_boundary : 1; // limit ME to slice boundaries + unsigned int ROI_enable : 1; // Enable region of interest processing + unsigned int ROI_me_control_idx : 4; + unsigned int ROI_md_control_idx : 4; + unsigned int ROI_q_control_idx : 4; + unsigned int ROI_qp_delta : 8; // QP delta in ROI relative to picture outside ROI + unsigned int wp_control_idx : 4; // index in array of nvenc_pred_weight_table_s to use for this slice + unsigned int reserved2 : 3; // reserved for alignment + + // int 5 + unsigned int ROI_top_mbx : 8; // top left mbx of the ROI + unsigned int ROI_top_mby : 8; + unsigned int ROI_bot_mbx : 8; // bottom right mbx of the ROI + unsigned int ROI_bot_mby : 8; + + // int 6: ref_pic_cmd_cfg + nvenc_h264_refpic_cmd_s ref_pic_cmd_cfg; + + // int 7-22: RefPicListReorder/Modification: Defined in Section 7.4.3.1 + nvenc_h264_ref_pic_reorder_s ref_pic_list_reorder_cmd[2][8]; // size shall not exceed num_ref_idx_lx_active_minus1 + 1 (8+8=16) + + //int 23-30: MMCO: Defined in Section: 7.4.3.3 + nvenc_h264_mmco_s mmco_cmd[8] ; + + // int 31 + unsigned int reserved3; + +} nvenc_h264_slice_control_s; // 128 bytes + + +// Quarter-resolution modifier flags for first_pass_source_half_scaled +#define NVENC_QRES_FLAG_ENABLE 0x1 // Set if quarter-resolution is enabled +#define NVENC_QRES_FLAG_DECIMATED 0x10 // Set to indicate that subsampling was performed with decimation rather than low-pass +#define NVENC_QRES_FLAG_PASS2REFS 0x20 // Set to indicate that 1st-pass uses subsampled 2nd pass references instead of self-generated references + +// Lookahead macros for ab_beta(INFO1) and prev_act(INFO2) +#define NVENC_LOOKAHEAD_INFO1(Wp2Wi) ((unsigned char)(Wp2Wi)) // clamp((int)(Wp*256.0/Wi), 0, 255) -> ratio of inter over intra complexity, 0=no lookahead +// Lookahead INFO2: (depth<<8)|(nextI) +// nextI: 1..255=distance to next scene change, 0=unknown or greater than lookahead depth; +// depth: 1..255=max lookahead depth (0=no lookahead) +#define NVENC_LOOKAHEAD_INFO2(nextI,depth) (((unsigned short)(nextI)) | (((unsigned short)(depth))<<8)) + +// HRD conformance and misc RC flags +#define NVENC_RC_HRD_VCL 0x1 // vcl_cpb_size and vcl_bitrate +#define NVENC_RC_HRD_NAL 0x2 // nal_cpb_size and nal_bitrate +#define NVENC_RC_HRD_STRICTGOPTARGET 0x4 // Strict GOP target (minimizes GOP-to-GOP rate fluctuations) +#define NVENC_RC_FIXED_PBRATIO 0x8U // Disable dynamic B-frame rhopbi adjustments + + // needed only if picture level RC is enabled; as part of picture level RC, it will also perform HRD verification + // The contents of this file is produced by the driver during sequence level operations and should not be modified. +typedef struct +{ + unsigned char hrd_type; // nal (=2) and vcl (=1) type + unsigned char QP[3]; // QP for 0:P picture, ,1:B picture 2; I picture + unsigned char minQP[3]; // min QP for 0:P picture, ,1:B picture 2; I picture + unsigned char maxQP[3]; // max QP for 0:P picture, ,1:B picture 2; I picture + unsigned char maxQPD; // max QP delta between two consecutive QP updates + unsigned char baseQPD; // initial QP delta between two consecutive QP updates + int rhopbi[3]; // 23.8 signed fixed point quant ratios P/I, B/I, I/I(=1) + + int framerate; // fps + unsigned int buffersize; // total buffer size + + int nal_cpb_size; // size in bits + int nal_bitrate; // rate in bps + int vcl_cpb_size; // size in bits + int vcl_bitrate; // rate in bps + + unsigned int gop_length; // I period, gop_length == 0xffffffff is used for infinite gop length + int Np; // 27.4 signed fixed (gop_length + num_b_frames -1)/(num_b_frames + 1) + int Bmin; // 23.8 signed fixed point min buffer level, updated by driver at sequence level only. + int Ravg; // 23.8 signed fixed point average rate, updated by driver at sequence level only. + int R; // 23.8 signed fixed point rate + + unsigned char ab_alpha; // VBR/CBR: min/target quality level (1..51); rcmode4: weight of prev frame activity compared to part of the current picture + unsigned char ab_beta; // lookahead info1: ratio of inter over intra complexity (fix8); rcmode4: weight of current activity compared to neighbors + unsigned short prev_act; // lookahead info2: nextI, depth and valid flag; rcmode4: average activity level of the first reference (L0) + + unsigned char aqMode : 3; // 0: disable adaptive quantization, 1: AQ mode fast, 2: AQ mode full + unsigned char dump_aq_stats : 1; // 1: reads aq stats + unsigned char is_emphasis_level : 1; // 1: enables emphasis level map, treats QP map as level map + unsigned char reserve0 : 3; // Reserved for alignment + unsigned char single_frame_VBV; // 1: VBV buffer size is set to average frame size; 0: otherwise + unsigned char two_pass_rc; // 0: single pass rc , 1: first pass of 2 pass rc 2: second pass of 2 pass rc + unsigned char rc_class; // reserved + + unsigned char first_pass_source_half_scaled; // 0 : first pass on half resolution 1: first pass on full resolution (along with NVENC_QRES_FLAG_XXX) + unsigned char iSizeRatioX; // ratio between I picture target size over average picture size numerator + unsigned char iSizeRatioY; // ratio between I picture target size over average picture size denominator + unsigned char ext_scene_change_flag; // Scene change flag set as hint by external preprocessing unit. 0: No scene change, 1: Current picture is first in scene + unsigned int ext_intra_satd; // If non zero this represents the intra SATD for current picture computed by external preprocessing unit + unsigned char ext_picture_rc_hints; // picture rc hints are set by external source. 0: no hints, 1: hints are avaialble for current picture + unsigned char session_max_qp; // Current frame qp will never exceed this value when ext_picture_rc_hints is set + unsigned char reserve[2]; +} nvenc_h264_rc_s; // 88 bytes + +typedef struct +{ + unsigned int var_min; + unsigned int var_max; + unsigned int var_avg; + unsigned char reserved[4]; +} nvenc_aq_stat_s; // 16 bytes + +typedef struct +{ + char l0[8]; // reference picture list 0 + char l1[8]; // reference picture list 1 + char temp_dist_l0[8]; // temporal distance of each ref pic in list 0 DiffPicOrderCnt( currPicOrField, pic0) ) + char temp_dist_l1[8]; // temporal distance of each ref pic in list 1 DiffPicOrderCnt( currPicOrField, pic1) ) + short dist_scale_factor[8][8]; // (h264 spec eq 8-203) array [refidx1][refidx0] signed 11 bit values + unsigned int diff_pic_order_cnt_zero[2]; // This is a 2-dimensional array of booleans (1 bit flags) indicating that DiffPicOrderCnt(refidx1, refidx0) == 0 + unsigned int max_slice_size; // for use in slice_mode 1 only: dynamic insertion at/before this boundary (#bytes) + unsigned int max_byte_count_before_resid_zero; // Maximum byte count before micrcode forces the residuals to zero. + int delta_pic_order_cnt_bottom; // direct value of syntax element to be put in every slice of the picture + unsigned short frame_num; // value for slice header syntax element + unsigned short pic_order_cnt_lsb; // value of the corresponding syntax element + unsigned short idr_pic_id; + unsigned short colour_plane_id; + unsigned int longTermFlag; // Bits 7:0 for pictures in RefPicList0 and bits 23:16 for pictures in RefPicList1 + unsigned char num_forced_slices_minus1; // number of forced slice boundary locations minus1(number of entries in h264_slice_control_s) + unsigned char num_me_controls_minus1; // number of nvenc_h264_me_control_s entries in array minus1 + unsigned char num_md_controls_minus1; // number of nvenc_h264_md_control_s entries in array minus1 + unsigned char num_q_controls_minus1; // number of nvenc_h264_quant_control_s entries in array minus1 + unsigned int slice_control_offset; // offset from start of top level "nvenc_h264_drv_pic_setup_s" structure + // to start of array of "nvenc_h264_slice_control_s" structures + unsigned int me_control_offset; // offset from start of top level "nvenc_h264_drv_pic_setup_s" structure + // to start of array of "nvenc_h264_me_control_s" structures + unsigned int md_control_offset; // offset from start of top level "nvenc_h264_drv_pic_setup_s" structure + // to start of array of "nvenc_h264_md_control_s" structures + unsigned int q_control_offset; // offset from start of top level "nvenc_h264_drv_pic_setup_s" structure + // to start of array of "nvenc_h264_quant_control_s" structures + unsigned int hist_buf_size; // size in bytes of the buffer allocated for history + unsigned int bitstream_buf_size; // size in bytes of the buffer allocated for bitstream slice/mb data + unsigned int bitstream_start_pos; // start position in bitstream buffer where data should be written (byte offset) + unsigned int pic_struct : 2; // 0 = frame, 1 = top/first field, 2 = bot/second field + unsigned int pic_type : 2; // 0 = P, 1 = B, 2 = I, 3 = IDR + unsigned int ref_pic_flag : 1; // reference picture (0 = no, 1 = yes) + unsigned int slice_mode : 1; // 0 = dynamic slice insertion based on slice size + // 1 = static slice insertion based on slice control array + // (insert slices as defined in array, if all array elements used + // before end of picture is reached, start over with entry 0) + unsigned int ipcm_rewind_enable : 1; + unsigned int cabac_zero_word_enable : 1; + + //mvc + unsigned int base_view : 1; // 1: current view component is a base view; 0: non-base view + unsigned int priority_id : 6; // priority identifier for the NAL unit + unsigned int view_id : 10; // view identifier for the NAL unit + unsigned int temporal_id : 3; // temporal identifier for the NAL unit + unsigned int anchor_pic_flag : 1; // equal to 1 specifies that the current access unit is an anchor access unit + unsigned int inter_view_flag : 1; // equal to 1 specifies that the current view component may be used for inter-view prediction by other view components in the current access unit + unsigned int opm_enable : 1; // check with OPM if encryption needed + unsigned int temporal_hint8x8_enable : 1; // 8x8 hint enable + char cur_interview_ref_pic; //current interview ref pic + char prev_interview_ref_pic; //previous interview ref pic to replace current interview ref pic when write the regs ME_MBC_REFLIST0_DPBLUT and ME_MBC_REFLIST1_DPBLUT + unsigned char codec : 3; + unsigned char e4byteStartCode : 1; // if enable to 1, all the slices in a picture will use 4 byte start code 00000001(used in rtp mode), original is 000001 + unsigned char qpfifo : 1; // 0 = Polling mode, 1 = Interrupt mode. + unsigned int me_candidates_output_struct : 1; // 0 = me-only mode, 1 = MV & cost for all blocks (hybrid format) + unsigned int me_candidates_output_enable : 1; // hybrid mode results configuration (see me_candidates_output_struct). 0 = results disabled. 1 = results enabled (for fullscale image). + unsigned char medma_output_cost_mode : 1; // 0 = residual cost only. 1 = residual cost + mv cost + unsigned char intraRefreshCount; // Set to number of pictures over which intra refresh will happen. Set to 0 for non intra refresh picture + unsigned int mpec_threshold : 30; // collect mpec stats after threshold mbs + // when mpec_stat_on is enabled and mpec_threshold is not equal to picWidthInMbs, + // rc mode 0 will be used to collect stats + unsigned int me_lambda_fifo_enable : 1; // lambda map fifo enable. 1: enable. 0: disable + unsigned int me_candidates_rdo_mode : 1; // me candidates rdo mode enable. 1: enable. 0: disable. + unsigned int slice_stat_offset; // offset from start of top level "nvenc_stat_data_s" structure + // to start of array of "nvenc_slice_stat_s" structure + unsigned int mpec_stat_offset; // offset from start of top level "nvenc_stat_data_s" structure + // to start of array of "nvenc_mpec_stat_s" structure + unsigned int wp_control_offset; // offset from start of top level "nvenc_h264_drv_pic_setup_s" structure + // to start of array of "nvenc_pred_weight_table_s" structure + // to start of array of "nvenc_mpec_stat_s" structure + + unsigned char num_wp_controls_minus1; // number of nvenc_pred_weight_table_s entries in array minus 1 + unsigned char b_as_ref : 1; // use b frame as reference + unsigned char ofs_mode : 2; // ofs mode. 0: OFF; 1: Optical Flow; 2: Stereo (output mvx only), 3: Stereo (output mvx/mvy) + unsigned char new_subframe : 1; + unsigned char bit_depth_minus_8 : 4; + unsigned char stripID; + unsigned char strips_in_frame; + unsigned short slice_encoding_row_start; // start mb row for slice encoding + unsigned short slice_encoding_row_num; // num of mb row for slice encoding + + unsigned char segment_enable : 1; // used by ofs mode only for segment map enable flag, 0: (default) disable/ 1: enable + unsigned char background_seg : 5; // used by ofs mode only when segment map is enabled, to specify the segment id of background + unsigned char input_sub_sampling : 1; // used by PDMA 2x2 sub-sampling + unsigned char output_sub_sampling : 1; // used by MPEB 2x2 sub-sampling output + unsigned char segment_P2 ; + unsigned char segment_thr ; + unsigned char accurate_skip_mv : 1; + unsigned char intra_sel_4x4_enable : 1; // seletive intra4x4 , 0:disable(always do intra4x4 serach), 1:enable(enable or disable intra4x4 search based on pixel variance) + unsigned int enable_source_image_padding : 1; // enable/disable HW source image padding + unsigned int luma_padding_mode : 2; // HW souce padding mode for luma: 0: copy frame boundary, 1: zero padding, 2: half value, i.e 1<<(bitdepth-1) + unsigned int chroma_padding_mode : 2; // HW souce padding mode for chroma: 0: copy frame boundary, 1: zero padding, 2: half value, i.e 1<<(bitdepth-1) + unsigned int stats_fifo_enable : 1; // enable per-mb short stats fifo + unsigned int chroma_skip_threshold_4x4 : 16; // low 3 bits is the factional, to check if chroma residual is big in skip mb, can't simply set chroma residual to 0 even though luma has been decided to skip + unsigned int intra_sel_4x4_threshold : 16; // disable intra4x4 search when intra_sel_4x4_enable=true and max variance of 8x8 blocks in current MB is less than intra_sel_4x4_threshold + unsigned int aq_stat_offset; // start of array of "nvenc_aq_stat_s" structure + unsigned int act_stat_offset; // offset from start of "nvenc_rc_pic_stat_s" structure, + // in the RC stats buffer. This field points to the start + // of data for external (delta-)QP map. + unsigned int stats_fifo_offset; // offset from start of top level "nvenc_stat_data_s" structure +} nvenc_h264_pic_control_s; // 276 bytes + + +// This is the top level config structure passed by driver to nvenc. It must be aligned +// to a 256B boundary. The first integer in this struct is a 'magic' number which driver +// must set to a specific value, and which the main Falcon app must check to see if it +// matches the expected value. This prevents running wrong combinations of traces/drvr +// code and falcon code. Encoding is as follows: +// Bit 31..16 = 0xc9b7 +// Bit 15..8 = Version number +// Bit 7..0 = Revision number +// This number changes whenever there is a change to the class. If the change is small and +// backward compatible, only the revision number is incremented. If the change is major or +// not backward compatible, the revision number is reset to zero and the version number +// incremented. Falcon app should check if the upper 24 bits match the expected value, and +// terminate with error code "NvC9B7EncErrorH264BadMagic" if there is a mismatch. The value +// for the current class is defined as: + +#ifdef NV_NVENC_8_2 +#define NV_NVENC_DRV_MAGIC_VALUE 0xC9B70006 +#elif NV_NVENC_8_0 +#define NV_NVENC_DRV_MAGIC_VALUE 0xC8B70006 +#elif NV_NVENC_7_3 +#define NV_NVENC_DRV_MAGIC_VALUE 0xC7B70006 +#elif NV_NVENC_7_2 +#define NV_NVENC_DRV_MAGIC_VALUE 0xC4B70006 +#elif NV_NVENC_7_0 +#define NV_NVENC_DRV_MAGIC_VALUE 0xC5B70006 +#elif NV_NVENC_6_8 +#define NV_NVENC_DRV_MAGIC_VALUE 0xB6B70006 +#elif NV_NVENC_6_6 +#define NV_NVENC_DRV_MAGIC_VALUE 0xB4B70006 +#elif NV_NVENC_6_4 +#define NV_NVENC_DRV_MAGIC_VALUE 0xC3B70006 +#elif NV_NVENC_6_2 +#define NV_NVENC_DRV_MAGIC_VALUE 0xC2B70006 +#elif NV_NVENC_6_0 +#define NV_NVENC_DRV_MAGIC_VALUE 0xc1b70006 +#elif NV_NVENC_5_0 +#define NV_NVENC_DRV_MAGIC_VALUE 0xd0b70006 +#elif NV_NVENC_1_0 +#define NV_NVENC_DRV_MAGIC_VALUE 0xc0b70006 +#elif NV_MSENC_2_0 +#define NV_MSENC_DRV_MAGIC_VALUE 0xa0b70006 +#else +#define NV_MSENC_DRV_MAGIC_VALUE 0x90b70006 +#endif + +#define HIST_BLOCK_SIZE 192 +// Move all local define of these values to this file +//align to current AV1HIST MAS +const static int AV1_HIST_PIXEL_FRAME_SIZE = (8 * 8192); +const static int AV1_HIST_MV_SB_SIZE = (256 * 4); + +typedef struct +{ + // 256 aligned + unsigned int magic; // magic number, see text above 4 bytes + nvenc_h264_surface_cfg_s refpic_cfg; // defines layout of reference pictures 32 bytes + nvenc_h264_surface_cfg_s input_cfg; // defines layout of input pictures 32 bytes + nvenc_h264_surface_cfg_s outputpic_cfg; // defines layout of reconstructed pictures 32 bytes + nvenc_h264_sps_data_s sps_data; // 4 bytes + nvenc_h264_pps_data_s pps_data; // 8 bytes + nvenc_h264_rc_s rate_control; // Rate Control information 88 bytes + nvenc_h264_pic_control_s pic_control; // 276 bytes + nvenc_h264_surface_cfg_s half_scaled_outputpic_cfg; // defines layout of 2x2 subsampled reconstructed picture 32 bytes + unsigned int gpTimer_timeout_val; // GPTimer cycle count set from driver 4 bytes + +} nvenc_h264_drv_pic_setup_s; // 512 bytes + + +// **************************************************************************** +// **************************************************************************** +// H265 +// **************************************************************************** +// **************************************************************************** +typedef struct +{ + unsigned short frame_width_minus1; // frame width in pixels minus 1, range 0-4095 + unsigned short frame_height_minus1; // frame height in pixels minus 1, range 0-4095 + unsigned short sfc_pitch; // pitch of luma plane + unsigned short sfc_pitch_chroma; // pitch of chroma plane + unsigned short sfc_trans_mode; // least 3 significient bits are used to stand for 8 modes, normal, xflip, yflip etc. + unsigned short reserved2; // pad to int + + unsigned int luma_top_frm_offset; // offset of luma top field or frame in units of 256 + unsigned int luma_bot_offset; // offset of luma bottom field in units of 256. Not used if frame format. + unsigned int chroma_top_frm_offset; // offset of chroma top field or frame, or offset of chroma U(Cb) plane in planar mode, both in units of 256 + unsigned int chroma_bot_offset; // offset of chroma bottom field, not used if frame format. Or offset of chroma V(Cr) plane in planar mode. Both in units of 256. + unsigned int block_height : 7; // BL mapping block height setting + unsigned int tiled_16x16 : 1; // Surface is 16x16 tiled instead of BL mapped (must be 1 for refpics) + unsigned int memory_mode : 2; // 0: semi-planar, 1: planar + unsigned int nv21_enable : 1; // the surface format is yuv or yvu + unsigned int input_bl_mode : 2; // the input block linear mode: 0~gpu bl; 1~tegra bl; 2~naive bl this is only for cmod and cmod/plugin + unsigned int reserved : 19; // pad to NvU32 +} nvenc_h265_surface_cfg_s; // 32 bytes + +typedef struct +{ + unsigned int chroma_format_idc : 2; // 0=monochrome, 1=yuv4:2:0, 3=yuv4:4:4, 2=not supported + unsigned int separate_colour_plane_flag : 1; //if this bit is enabled, the chroma_format_idc should be 3 + unsigned int log2_max_pic_order_cnt_lsb_minus4 : 4; + unsigned int min_cu_size : 2; //0-8x8, 1-16x16, 2-32x32, 3-64x64 + unsigned int max_cu_size : 2; //0-8x8, 1-16x16, 2-32x32, 3-64x64 + unsigned int min_tu_size : 2; //0-4x4, 1-8x8, 2-16x16, 3-32x32 + unsigned int max_tu_size : 2; //0-4x4, 1-8x8, 2-16x16, 3-32x32 + unsigned int reserved0 : 1; // pad to full short + + unsigned int max_transform_depth_inter : 2; // max_transform_hierarchy_depth_inter + unsigned int max_transform_depth_intra : 2; // max_transform_hierarchy_depth_intra + unsigned int pcm_enabled_flag : 1; // if this bit is 1 , PCM data is present is the sequence + unsigned int pcm_loop_filter_disable_flag : 1; // Disable loop filter and SAO for CU with pcm_flag to 1 + unsigned int long_term_ref_pics_present_flag : 1; // if set to 1 , it indicates long term pictures are used for inter prediction. + unsigned int amp_enabled_flag : 1; // if set asymmetric motion partitions can be used for inter predictions. + unsigned int strong_intra_smoothing_enabled_flag : 1; // if set , bilinear interpolation is conditionally used filtering process for intra prediction. + unsigned int num_short_term_ref_pic_sets : 7; // max value is 64 + unsigned int num_long_term_ref_pics_sps : 5; + unsigned int bit_depth_minus_8 : 4; // luma/chroma should have the same value + unsigned int sample_adaptive_offset_enabled_flag : 1; // set 1 specifies that the sample adaptive offset process is applied to the reconstructed picture after the deblocking filter process + unsigned int sps_temporal_mvp_enabled_flag : 1; // if set, slice_temporal_mvp_enabled_flag is present in slice header + unsigned int reserved1 : 21; +} nvenc_h265_sps_data_s; // 8 bytes + +typedef struct +{ + unsigned int pic_param_set_id : 8; // picture parameter set identification + unsigned int num_ref_idx_l0_active_minus1 : 5; // number of currently active reference pictures in list 0 minus 1, range 0..30 + unsigned int num_ref_idx_l1_active_minus1 : 5; // number of currently active reference pictures in list 1 minus 1, range 0..30 + unsigned int constrained_intra_pred_flag : 1; // if set, intra prediction can only use pixels from macroblocks that are also intra coded + unsigned int transquant_bypass_enabled_flag : 1; // if set, transform ,scaling and inloop filtering is bypassed. + unsigned int transform_skip_enabled_flag : 1; // if set , it enables transform bypass during residual coding. + unsigned int deblocking_filter_override_enabled_flag : 1; // it set, the deblocking_filter_override_flag is encoded in slice segment header + unsigned int pps_loop_filter_across_slices_enabled_flag : 1; // if set, it enables loop filter across slice boundary. + unsigned int cu_qp_delta_enabled_flag : 1; // if set , it enables cu_qp_delta_abs is encoded at coding unit layer. + unsigned int lists_modification_present_flag : 1; // if set , it indicates reference picture list modification command might be present is slice segment header. + unsigned int output_flag_present_flag : 1; // if set, pic_output syntax element is added to slice syntax header + unsigned int cabac_init_present_flag : 1; // if set, cabac_init_flag is present in slice segment header. + unsigned int pps_slice_chroma_qp_offsets_present : 1; // if set to 1, cb_qp_offset and cr_qp_offset is signalled in the slice segment header. + unsigned int weighted_pred_flag : 1; // weighted_pred_flag equal to 1 specifies that weighted prediction is applied to P slices + unsigned int weighted_bipred_flag : 1; // weighted_bipred_flag equal to 1 specifies that weighted prediction is applied to B slices + unsigned int reserved : 2; + + int pic_init_qp_minus26 : 6; // initial QP value, range -26..+25. -38...+25 for 10bit, -74...+25 for 16bits. TODO: eable this syntex for pps + int pps_cb_qp_offset : 5; + int pps_cr_qp_offset : 5; + unsigned int dependent_slice_segments_enabled_flag : 1; + unsigned int sign_data_hiding_enabled_flag : 1; + unsigned int diff_cu_qp_delta_depth : 2; + unsigned int log2_parallel_merge_level_minus2 : 3; + unsigned int num_extra_slice_header_bits : 3; + + int reserved1 : 6; +} nvenc_h265_pps_data_s; // 8 bytes + + // needed only if picture level RC is enabled; as part of picture level RC, it will also perform HRD verification + // The contents of this file is produced by the driver during sequence level operations and should not be modified. +typedef struct +{ + unsigned char hrd_type; // nal (=2) and vcl (=1) type + unsigned char QP[3]; // QP for 0:P picture, ,1:B picture 2; I picture + unsigned char minQP[3]; // min QP for 0:P picture, ,1:B picture 2; I picture + unsigned char maxQP[3]; // max QP for 0:P picture, ,1:B picture 2; I picture + unsigned char maxQPD; // max QP delta between two consecutive QP updates + unsigned char baseQPD; // initial QP delta between two consecutive QP updates + int rhopbi[3]; // 23.8 signed fixed point quant ratios P/I, B/I, I/I(=1) + + int framerate; // fps + unsigned int buffersize; // total buffer size + + int nal_cpb_size; // size in bits + int nal_bitrate; // rate in bps + int vcl_cpb_size; // size in bits + int vcl_bitrate; // rate in bps + + unsigned int gop_length; // I period, gop_length == 0xffffffff is used for infinite gop length + int Np; // 27.4 signed fixed (gop_length + num_b_frames -1)/(num_b_frames + 1) + int Bmin; // 23.8 signed fixed point min buffer level, updated by driver at sequence level only. + int Ravg; // 23.8 signed fixed point average rate, updated by driver at sequence level only. + int R; // 23.8 signed fixed point rate + + unsigned char ab_alpha; // VBR/CBR: min/target quality level (1..51); rcmode4: weight of prev frame activity compared to part of the current picture + unsigned char ab_beta; // lookahead info1: ratio of inter over intra complexity (fix8); rcmode4: weight of current activity compared to neighbors + unsigned short prev_act; // lookahead info2: NVENC_LOOKAHEAD_INFO2; rcmode4: average activity level of the first reference (L0) + + unsigned char reserved0; + unsigned char single_frame_vbv; // 1: VBV buffer size is set to average frame size; 0: otherwise + unsigned char two_pass_rc; // 0: single pass rc , 1: first pass of 2 pass rc 2: second pass of 2 pass rc + unsigned char rc_class; // reserved + unsigned char first_pass_source_half_scaled; // 0 : first pass on half resolution 1: first pass on full resolution (along with NVENC_QRES_FLAG_XXX) + unsigned char i_size_ratio_x; // ratio between I picture target size over average picture size numerator + unsigned char i_size_ratio_y; // ratio between I picture target size over average picture size denominator + unsigned char aqMode; // 0: disable adaptive quantization 1: AQ mode fast (unimplemented) 2: AQ mode full (unimplemented) 3: External qp delta map + unsigned char dump_aq_stats; // 1: dump aq stats + unsigned char ext_scene_change_flag; // Scene change flag set as hint by external preprocessing unit. 0: No scene change, 1: Current picture is first in scene + unsigned char ext_picture_rc_hints; // picture rc hints are set by external source. 0: no hints, 1: hints are avaialble for current picture + unsigned char session_max_qp; // Current frame qp will never exceed this value when ext_picture_rc_hints is set + unsigned int ext_intra_satd; // If non zero this represents the intra SATD for current picture computed by external preprocessing unit + unsigned char reserved2[8]; +} nvenc_h265_rc_s; // 96 bytes + +typedef struct +{ + int inter_ref_pic_set_prediction_flag : 1; + int delta_idx_minus1 : 6; + int delta_rps_sign : 1; + int abs_delta_rps_minus1 : 16; + int reserved : 8; + unsigned short used_by_curr_pic_flag; + unsigned short use_delta_flag; + short num_negative_pics; + short num_positive_pics; + short delta_poc_s0_minus1[16]; + unsigned short used_by_curr_pic_s0_flag; + short delta_poc_s1_minus1[16]; + unsigned short used_by_curr_pic_s1_flag; +}nvenc_h265_strps_s; // 80bytes multiple of 4 + +typedef struct +{ + int num_long_term_sps : 5; + int num_long_term_pics : 5; + int reserved0 : 22; + unsigned char lt_idx_sps[16]; + unsigned short poc_lsb_lt[16]; + unsigned short used_by_curr_pic_lt_flag; + unsigned short delta_poc_msb_present_flag; + int delta_poc_msb_cycle_lt[16]; +}nvenc_h265_ltrps_s; // 120bytes multiple of 4 + +typedef struct +{ + int NumPocTotalCurr : 5; + int ref_pic_list_mod_flag_l0 : 1; + int ref_pic_list_mod_flag_l1 : 1; + int reserved0 : 25; + unsigned char list_entry_l0[16]; + unsigned char list_entry_l1[16]; +}nvenc_h265_ref_pic_list_modification_s; //36bytes multiple of 4 + +typedef struct +{ + unsigned int vps_poc_lsb_not_present_flag : 1; + unsigned int vps_default_ref_layers_active_flag : 1; + unsigned int vps_max_one_active_ref_layer_flag : 1; + unsigned int vps_num_direct_ref_layers : 6; + unsigned int slh_discardable_flag : 1; // mv_hevc, equal to 1 specifies that the coded picture is not used as a reference picture + unsigned int slh_cross_layer_bla_flag : 1; // mv_hevc + unsigned int slh_num_active_ref_layer_pics : 6; // active ref layer pics of current picture + unsigned int slh_inter_layer_pred_enabled_flag : 1; + unsigned int slh_num_inter_layer_ref_pics_minus1 : 6; + unsigned int reserved0 : 8; + unsigned char slh_inter_layer_pred_layer_idc[8]; //NumPicTotalCurr <= 8 +}nvenc_mv_hevc_cfg_s; //12bytes multiple of 4 + +typedef struct +{ + char l0[8]; // reference picture list 0 + char l1[8]; // reference picture list 1 + char temp_dist_l0[8]; // temporal distance of each ref pic in list 0 DiffPicOrderCnt( currPicOrField, pic0) ) + char temp_dist_l1[8]; // temporal distance of each ref pic in list 1 DiffPicOrderCnt( currPicOrField, pic1) ) + short dist_scale_factor[8][8]; // (h264 spec eq 8-203) array [refidx1][refidx0] signed 11 bit values + unsigned int diff_pic_order_cnt_zero[2]; // This is a 2-dimensional array of booleans (1 bit flags) indicating that DiffPicOrderCnt(refidx1, refidx0) == 0 + unsigned int max_slice_size; // for use in slice_mode 1 only: dynamic insertion at/before this boundary (#bytes) + unsigned int max_byte_count_before_resid_zero; // Maximum byte count before micrcode forces the residuals to zero. + + unsigned short no_output_of_prior_pics_flag : 1; // used to modify the DPB after decoding a IDR or BLA picture + unsigned short pic_output_flag : 1; // affects the decoded picture output and removal processes as specified in Annex C + unsigned short mode_decision_only_flag : 1; // only output h265 mode and mocomp result, no recon and bitstream output + unsigned short cu_early_terminatiom_enable : 1; // 0: disable CU early termination, 1: enable CU early termination + unsigned short input_sub_sampling : 1; // used by PDMA 2x2 sub-sampling + unsigned short output_sub_sampling : 1; // used by MPEB 2x2 sub-sampling output + unsigned short reserved0 : 10; + unsigned short pic_order_cnt_lsb; // value of the corresponding syntax element + + unsigned char num_forced_slices_minus1; // number of forced slice boundary locations minus1(number of entries in h264_slice_control_s) + unsigned char num_me_controls_minus1; // number of nvenc_h264_me_control_s entries in array minus1 + unsigned char num_md_controls_minus1; // number of nvenc_h264_md_control_s entries in array minus1 + unsigned char num_q_controls_minus1; // number of nvenc_h264_quant_control_s entries in array minus1 + unsigned int slice_control_offset; // offset from start of top level "nvenc_h264_drv_pic_setup_s" structure + // to start of array of "nvenc_h264_slice_control_s" structures + unsigned int me_control_offset; // offset from start of top level "nvenc_h264_drv_pic_setup_s" structure + // to start of array of "nvenc_h264_me_control_s" structures + unsigned int md_control_offset; // offset from start of top level "nvenc_h264_drv_pic_setup_s" structure + // to start of array of "nvenc_h264_md_control_s" structures + unsigned int q_control_offset; // offset from start of top level "nvenc_h264_drv_pic_setup_s" structure + // to start of array of "nvenc_h264_quant_control_s" structures + unsigned int hist_buf_size; // size in bytes of the buffer allocated for history + unsigned int bitstream_buf_size; // size in bytes of the buffer allocated for bitstream slice/mb data + unsigned int bitstream_start_pos; // start position in bitstream buffer where data should be written (byte offset) + unsigned int pic_type : 2; // 0 = P, 1 = B, 2 = I, 3 = IDR + unsigned int ref_pic_flag : 1; // reference picture (0 = no, 1 = yes) + unsigned int slice_mode : 1; // 0 = dynamic slice insertion based on slice size + // 1 = static slice insertion based on slice control array + // (insert slices as defined in array, if all array elements used + // before end of picture is reached, start over with entry 0) + unsigned int cabac_zero_word_enable : 1; + unsigned int codec : 3; + unsigned int ceil_log2_pic_size_in_ctbs : 6; // ceillog2(picSizeInCtbs) for HW needs too much area/time for 16kx16k case + unsigned int sao_luma_mode : 5; + unsigned int sao_chroma_mode : 5; + unsigned int reserved1 : 8; + + unsigned int mpec_threshold; // collect mpec stats after threshold mbs + // when mpec_stat_on is enabled and mpec_threshold is not equal to picWidthInMbs, + // rc mode 0 will be used to collect stats + unsigned int slice_stat_offset; // offset from start of top level "nvenc_h265_stat_data_s" structure + // to start of array of "nvenc_h265_slice_stat_s" structure + unsigned int mpec_stat_offset; // offset from start of top level "nvenc_h265_stat_data_s" structure + // to start of array of "nvenc_h265_mpec_stat_s" structure + unsigned int aq_stat_offset; // to start of array of "nvenc_h265_aq_stat_s" structure + unsigned int short_term_ref_pic_set_sps_flag : 1; // if set to 1 ,the short term RPS is signalled using SPS, if set to 0 the short term RPS is encoded in the current slice header + unsigned int short_term_ref_pic_set_idx : 6; // used to signal index into the list of the short term RPS in the SPS. + unsigned int e4byteStartCode : 1; // if enable to 1, all the slices in a picture will use 4 byte start code 00000001(used in rtp mode), original is 000001 + unsigned int nal_unit_type : 6; // specifies the type of RBSP data structure contained in the NAL unit + unsigned int nuh_layer_id : 6; // reserved and shall be equal to 0. + unsigned int nuh_temporal_id_plus1 : 3; // specifies the temporal_ip_plus1 of RBSP data structure contained in the NAL unit + unsigned int qpfifo : 1; // 0 = Polling mode, 1 = Interrupt mode. + unsigned int enable_source_image_padding : 1; // enable/disable HW source image padding + unsigned int luma_padding_mode : 2; // HW souce padding mode for luma: 0: copy frame boundary, 1: zero padding, 2: half value, i.e 1<<(bitdepth-1) + unsigned int chroma_padding_mode : 2; // HW souce padding mode for chroma: 0: copy frame boundary, 1: zero padding, 2: half value, i.e 1<<(bitdepth-1) + unsigned int stats_fifo_enable : 1; // enable per-ctb short stats fifo + unsigned int reserved2 : 2; + + nvenc_h265_strps_s short_term_rps; // if short_term_ref_pic_set_sps_flag is set to 0, the nvenc_h265_strps_s structure is signalled in the slice segment segment header. + nvenc_h265_ltrps_s long_term_rps; // if long_term_ref_pics_present_flag is set to 1 in SPS , nvenc_h265_ltrps_s structure is signalled in the slice segment segment header + nvenc_h265_ref_pic_list_modification_s ref_pic_list_modification; // if lists_modification_present_flag is set to 1, then nvenc_h265_ref_pic_list_modification_s structure is encoded in slice segment header. + unsigned int rc_pass2_offset; + + unsigned int wp_control_offset; // offset from start of top level "nvenc_h265_drv_pic_setup_s" structure + // to start of array of "nvenc_pred_weight_table_s" structure + + unsigned char num_wp_controls_minus1; // number of nvenc_pred_weight_table_s entries in array minus 1 + unsigned char stripID; + unsigned char strips_in_frame; + unsigned char slice_idx_start; + unsigned short slice_encoding_row_start; // start ctb row for slice encoding + unsigned short slice_encoding_row_num; // num of ctb row for slice encoding + + nvenc_mv_hevc_cfg_s mv_hevc_cfg; + unsigned int chroma_skip_threshold_4x4 : 16; // low 3 bits is the factional, to check if chroma residual is big in skip mb, can't simply set chroma residual to 0 even though luma has been decided to skip + unsigned int longTermFlag : 16; // Bits 7:0 for pictures in RefPicList0 and bits 15:8 for pictures in RefPicList1 + + char poc_entry_l0[8]; // ref pic list to ref pic entry map + char poc_entry_l1[8]; // ref pic list to ref pic entry map + unsigned int same_poc[2]; // set to 1 when the poc of cur ref pic and neighbor ref pic are same, otherwise set it to 0. + unsigned int act_stat_offset; // offset from start of "nvenc_rc_pic_stat_s" structure, points to data for external QP map + unsigned int stats_fifo_offset; // offset from start of top level "nvenc_stat_data_s" structure +} nvenc_h265_pic_control_s; // 536 bytes + +typedef struct +{ + // 256 aligned + unsigned int magic; // magic number, see text above // 4 bytes + nvenc_h265_surface_cfg_s refpic_cfg; // defines layout of reference pictures // 32 bytes + nvenc_h265_surface_cfg_s input_cfg; // defines layout of input pictures // 32 bytes + nvenc_h265_surface_cfg_s outputpic_cfg; // defines layout of reconstructed pictures // 32 bytes + nvenc_h265_sps_data_s sps_data; // 8 bytes + nvenc_h265_pps_data_s pps_data; // 8 bytes + nvenc_h265_rc_s rate_control; // Rate Control information // 96 bytes + nvenc_h265_pic_control_s pic_control; //536 bytes + nvenc_h265_surface_cfg_s half_scaled_outputpic_cfg; // defines layout of 2x2 subsampled reconstructed picture // 32 bytes + unsigned int gpTimer_timeout_val; // GPTimer cycle count set from driver 4 bytes + unsigned char reserved[240]; // Pad to 1024 bytes + +} nvenc_h265_drv_pic_setup_s; + +typedef struct +{ + unsigned short intra_refresh_cnt; // distance between forced-intra MBs in slice; 0 means OFF + unsigned short intra_refresh_offs; // first forced-intra MB in slice + + unsigned int h265_intra_luma4x4_mode_enable; //bitmask indicate which intra luma 4x4 modes to enable in ip1 intra search + unsigned int h265_intra_luma8x8_mode_enable; //bitmask indicate which intra luma 8x8 modes to enable in ip1 intra search + unsigned int h265_intra_luma16x16_mode_enable; //bitmask indicate which intra luma 16x16 modes to enable in ip1 intra search + unsigned int h265_intra_luma32x32_mode_enable; //bitmask indicate which intra luma 32x32 modes to enable in ip1 intra search + unsigned int h265_intra_luma64x64_mode_enable; //bitmask indicate which intra luma 64x64 modes to enable in ip1 intra search + unsigned int h265_intra_luma_mode_enable_leftbits : 30; //bitmask indicate intra luma left modes(32~34) from ip1 luma4x4 to ip2 luma64x64. + unsigned int reserved0 : 2; + + + unsigned char cu16x16_part_2Nx2N_enable : 1; // enable prediction for 2Nx2N + unsigned char cu16x16_part_2NxN_enable : 1; // enable prediction for 2NxN + unsigned char cu16x16_part_Nx2N_enable : 1; // enable prediction for Nx2N + unsigned char cu16x16_part_NxN_enable : 1; // enable prediction for NxN + unsigned char cu16x16_part_2NxnU_enable : 1; // enable prediction for 2NxnU + unsigned char cu16x16_part_2NxnD_enable : 1; // enable prediction for 2NxnD + unsigned char cu16x16_part_nLx2N_enable : 1; // enable prediction for nLx2N + unsigned char cu16x16_part_nRx2N_enable : 1; // enable prediction for nRx2N + + unsigned char cu32x32_part_2Nx2N_enable : 1; // enable prediction for 2Nx2N + unsigned char cu32x32_part_2NxN_enable : 1; // enable prediction for 2NxN + unsigned char cu32x32_part_Nx2N_enable : 1; // enable prediction for Nx2N + unsigned char cu32x32_part_NxN_enable : 1; // enable prediction for NxN + unsigned char cu32x32_part_2NxnU_enable : 1; // enable prediction for 2NxnU + unsigned char cu32x32_part_2NxnD_enable : 1; // enable prediction for 2NxnD + unsigned char cu32x32_part_nLx2N_enable : 1; // enable prediction for nLx2N + unsigned char cu32x32_part_nRx2N_enable : 1; // enable prediction for nRx2N + + unsigned char cu64x64_part_2Nx2N_enable : 1; // enable prediction for 2Nx2N + unsigned char cu64x64_part_2NxN_enable : 1; // enable prediction for 2NxN + unsigned char cu64x64_part_Nx2N_enable : 1; // enable prediction for Nx2N + unsigned char cu64x64_part_NxN_enable : 1; // enable prediction for NxN + unsigned char cu64x64_part_2NxnU_enable : 1; // enable prediction for 2NxnU + unsigned char cu64x64_part_2NxnD_enable : 1; // enable prediction for 2NxnD + unsigned char cu64x64_part_nLx2N_enable : 1; // enable prediction for nLx2N + unsigned char cu64x64_part_nRx2N_enable : 1; // enable prediction for nRx2N + + unsigned char intra_mode_4x4_enable : 1; // enable intra prediction in mdp for 4x4 + unsigned char intra_mode_8x8_enable : 1; // enable intra prediction in mdp for 8x8 + unsigned char intra_mode_16x16_enable : 1; // enable intra prediction in mdp for 16x16 + unsigned char intra_mode_32x32_enable : 1; // enable intra prediction in mdp for 32x32 + unsigned char intra_mode_64x64_enable : 1; // enable intra prediction in mdp for 32x32 + unsigned char priority_ipred_type_ip1 : 3; // which intra type search first for early termination, must be a subset of ip_search_mode. reserved, not implemented in HW + + unsigned int h265_intra_chroma_mode_enable : 5; //bitmask indicate which intra_chroma_pred_mode to enable + unsigned int inter_penalty_factor_for_ip1 : 6; // early teminate ip1 intra search if intra cost is larger than inter_cost + ((inter_cost * factor) >> 8) + unsigned int tusize_4x4_enable : 1; //enable 4x4 transform + unsigned int tusize_8x8_enable : 1; //enable 8x8 transform + unsigned int tusize_16x16_enable : 1; //enable 16x16 transform + unsigned int tusize_32x32_enable : 1; //enable 32x32 transform + unsigned int intra_tusize_decision : 1; //enable intra tu size decision, reserved, not implemented in HW + unsigned int pskip_enable : 1; //enable p_skip encoding + unsigned int tu_search_basedon_pusize : 1; //enable tu search based on pu size + unsigned int reserved1 : 14; // reserved for future special prediction extensions + + short bias_part_2Nx2N; //bias for 2Nx2N PU + short bias_part_2NxN; //bias for 2NxN PU + short bias_part_Nx2N; //bias for Nx2N PU + short bias_part_NxN; //bias for NxN PU + short bias_part_2NxnU; //bias for 2NxnU PU + short bias_part_2NxnD; //bias for 2NxnD PU + short bias_part_nLx2N; //bias for nLx2N PU + short bias_part_nRxN; //bias for nRxN PU + + short bias_cusize_8x8; //bias for 8x8 CU + short bias_cusize_16x16; //bias for 16x16 CU + short bias_cusize_32x32; //bias for 32x32 CU + short bias_cusize_64x64; //bias for 64x64 CU + short bias_pskip; // bias for pskip (subtracts from pskip cost) + + short bias_intra_over_inter; // bias for any intra (subtracts from intra cost) + short bias_intra_most_prob; // bias for intra most probable mode + short bias_qpp; + short bias_transform_skip; // bias for transform skip mode + + unsigned short mv_cost_bias; // bias applied to internal motion vector cost calculation + unsigned short pskip_esc_threshold; // if pskip cost is below this, disable full search & attempt pskip encode + + unsigned short tempmv_wt_spread_threshold; // The threshold for the sum of differences between 8x8s and 16x16 mvs. Used for temporal mv weight + unsigned int tempmv_wt_distort_threshold : 16; // The threshold for the cost of the 16x16 mv. Used for temporal mv weight + unsigned int mv_cost_predictor_control : 1; // 0=use 16x16 predictor only; 1=use accurate predictor (slower but better) + unsigned int mv_cost_enable : 1; // enable mv cost calculations in mode decision + unsigned int chroma_eval_mode : 1; // for intra chroma search 0 = U and V; 1 = U only + unsigned int ip_search_mode : 4; // which intra sizes to search before making IP decision b0=4x4,b1=8x8,b2=16x16,b3=32x32 + unsigned int multiply_bias_with_lambda : 1; // setting this to 1 will result in bias being multiplied by lambda + unsigned int force_ipcm : 1; // 0=normal encode; 1=force mpeb/mpec encode as IPCM + unsigned int early_termination_ip1 : 1; // enable IP1 early termination + unsigned int tu_search_num_cu16 : 3; // tu search part num in cu16 + unsigned int tu_search_num_cu32 : 3; // tu search part num in cu32 + unsigned int early_termination_mode_ip1; //enable intra mode search for early termination, 32 bits for the 35 modes other than H, V and DC, these 3 modes enabled as default + short bias_tu_4x4; // bias for TU 4x4 + short bias_tu_8x8; // bias for TU 8x8 + short bias_tu_16x16; // bias for TU 16x16 + short bias_tu_32x32; // bias for TU 32x32 + short num_1div8_lambda_ssd : 7; // number of 1/8 lambda ssd for lambda_coef delta. lambda_coef = (1+num_1div8_lambda_ssd*1/8)*lambda_ssd + short skip_evaluate_enable_cu8 : 1; // 0: not check skip ssd cost; 1, evaluate skip ssd cost + short skip_evaluate_enable_cu16 : 1; // 0: not check skip ssd cost; 1, evaluate skip ssd cost + short skip_evaluate_enable_cu32 : 1; // 0: not check skip ssd cost; 1, evaluate skip ssd cost + short skip_evaluate_enable_cu64 : 1; // 0: not check skip ssd cost; 1, evaluate skip ssd cost + short reserved2 : 1; + short intra_ssd_cnt_4x4 : 4; // SSD cost evaluation between N top intra modes selected by satd + short intra_ssd_cnt_8x8 : 4; // SSD cost evaluation between N top intra modes selected by satd + short intra_ssd_cnt_16x16 : 4; // SSD cost evaluation between N top intra modes selected by satd + short intra_ssd_cnt_32x32 : 4; // SSD cost evaluation between N top intra modes selected by satd + short rdo_level : 3; // 0:cu size decided by satd in P frame 1:cu size decided by ssd + short reserved3 : 1; // pad to short + + short bias_skip; // bias for skip mv's ssd (subtracts from skip mv's ssd cost) + short bias_intra_32x32; // bias for intra 32x32 (subtracts from intra 32x32 cost) + short bias_intra_16x16; // bias for intra 16x16 (subtracts from intra 16x16 cost) + short bias_intra_8x8; // bias for intra 8x8 (subtracts from intra 8x8 cost) + short bias_intra_4x4; // bias for intra 4x4 (subtracts from intra 4x4 cost) + unsigned char reserved[22]; +} nvenc_h265_md_control_s; // 128 bytes + +typedef struct +{ + unsigned int num_ctus : 19; // number of ctus in this slice, support upto 8kx8k + unsigned int qp_avr : 8; // 6bit value used by hw. Other bits can be used for rounding. + unsigned int reserved0 : 5; + + unsigned int dependent_slice_segment_flag : 1; // slice segment header syntax element present or not + unsigned int slice_type : 2; // 0 = B, 1 = P, 2 = I + unsigned int five_minus_max_num_merge_cand : 3; // specifies the maximum number of merging MVP candidates supported + unsigned int cabac_init_flag : 1; + unsigned int deblocking_filter_override_flag : 1; // specifies the deblocking control params is preset in slice header. + unsigned int slice_deblocking_filter_disabled_flag : 1; // if set to 1 , it disables deblcoking for the current slice. + // If deblocking_filter_override_flag is 0, this is set to pps_deblocking_filter_disabled_flag + unsigned int slice_loop_filter_across_slices_enabled_flag : 1; // if set to 1, it enables loop filter accross slice boundary. + unsigned int qp_slice_min : 8; // min slice qp value + unsigned int qp_slice_max : 8; // max slice qp value + unsigned int force_intra : 1; + unsigned int slice_sao_luma_flag : 1; // equal to 1 specifies that SAO is enabled for the luma component in the current slice + unsigned int slice_sao_chroma_flag : 1; // equal to 1 specifies that SAO is enabled for the chroma component in the current slice + unsigned int reserved1 : 3; + + int slice_beta_offset_div2 : 4; // if deblocking_filter_override_flag = 1 , then slice_beta_offset_div2 is signalled in slice segment header + int slice_tc_offset_div2 : 4; // if deblocking_filter_override_flag = 1 , then slice_tc_offset_div2 is signalled in slice segment header. + int slice_cb_qp_offset : 5; // offset added to pps_cb_qp_offset. It shall be in the range -12 to +12. + int slice_cr_qp_offset : 5; // offset added to pps_cr_qp_offset. It shall be in the range -12 to +12. + unsigned int slice_temporal_mvp_enabled_flag : 1; // set 1, temporal motion vector predictors may be used + unsigned int collocated_from_l0_flag : 1; // set 1, the collocated picture for tmvp is derived from ref. pic. list 0 + unsigned int collocated_ref_idx : 5; // specifies ref. index of the collocated picture for tmvp + int reserved2 : 7; + + unsigned int num_ref_idx_active_override_flag : 1; + unsigned int num_ref_idx_l0_active_minus1 : 5; // [0.31] + unsigned int num_ref_idx_l1_active_minus1 : 5; // [0,31] + unsigned int me_control_idx : 4; // index in array of nvenc_h264_me_control_s to use for this slice + unsigned int md_control_idx : 4; // index in array of nvenc_h264_md_control_s to use for this slice + unsigned int q_control_idx : 4; // index in array of nvenc_h264_quant_control_s to use for this slice + unsigned int wp_control_idx : 4; // index in array of nvenc_pred_weight_table_s to use for this slice + unsigned int reserved3 : 5; + + unsigned int limit_slice_top_boundary : 1; // limit ME to slice boundaries + unsigned int limit_slice_bot_boundary : 1; // limit ME to slice boundaries + unsigned int limit_slice_left_boundary : 1; // limit ME to slice boundaries + unsigned int limit_slice_right_boundary : 1; // limit ME to slice boundaries + unsigned int ROI_enable : 1; // Enable region of interest processing + unsigned int ROI_me_control_idx : 4; + unsigned int ROI_md_control_idx : 4; + unsigned int ROI_q_control_idx : 4; + unsigned int ROI_qp_delta : 8; // QP delta in ROI relative to picture outside ROI + unsigned int reserved4 : 7; // reserved for alignment + + unsigned int ROI_top_mbx : 16; // top left mbx of the ROI + unsigned int ROI_top_mby : 16; // top left mby of the ROI + unsigned int ROI_bot_mbx : 16; // bottom right mbx of the ROI + unsigned int ROI_bot_mby : 16; // bottom right mby of the ROI + + unsigned char reserved5[36]; // padding to 64 bytes +} nvenc_h265_slice_control_s; // 64 bytes + +// h265 not support QPP, reserved it but not used +typedef struct +{ + // QPP has not been defined for H265 + unsigned short qpp_run_vector_4x4; // cost values for 4x4 transform (16 bit total vector) + unsigned short qpp_run_vector_8x8[3]; // cost values for 8x8 transform (48 bit total vector) + // 2 bits each for first 12 coefs, then 1 bit each for next 24 + unsigned char qpp_luma8x8_cost; // luma 8x8 cost threshold, 0 = throw out all coefs, range 0-15 + unsigned char qpp_luma16x16_cost; // luma 16x16 cost threshold, 0 = throw out all coefs, range 0-15 + unsigned char qpp_chroma_cost; // chroma cost threshold, 0 = throw out all coefs, range 0-15 + unsigned char qpp_mode : 2; // 0 = OFF, 1 = 8x8, 2 = 16x16_8x8 + unsigned char reserved0 : 4; // padding + + unsigned int reserved1; // padding to 16 bytes +} nvenc_h265_quant_control_s; // 16 bytes + +//***************************************************************************** +// VP9 +//***************************************************************************** + +//vp9 filter config +typedef struct +{ + unsigned char filter_level[8]; // loop filter strength for each segment. + char ref_delta[4]; // Loop filter strength adjustments based on frame type (intra, inter) + char mode_delta[2]; // Loop filter strength adjustments based on mode (zero, new mv) + unsigned char disable : 1; // disable loop-filter + unsigned char sharpness : 3; // sharpness level + unsigned char baselevel_gt32 : 1; // filter base level greater than 32 flag + unsigned char reserved : 3; + unsigned char reserved2; +}nvenc_vp9_filter_cfg_s; //16 bytes + +typedef struct +{ + unsigned int round_chroma_ac : 3; // round for chroma AC + unsigned int round_chroma_dc : 3; // round for chroma DC + unsigned int round_luma_ac : 3; // round for luma AC + unsigned int round_luma_dc : 3; // round for luma DC + unsigned int round_delta_inter : 3; // round delta for inter + unsigned int reserved : 17; +}vp9_round_cfg_s; + +//vp9 quant config +typedef struct +{ + unsigned char qindex_seg[8]; // Base frame qindex (index in quant lookup table) for each segment 0...255 + int delta_y_dc : 5; // explicit qindex adjustment for y dccoefficient, -15...15 + int delta_uv_dc : 5; // qindex adjustment for uv accoefficient, -15...15 + int delta_uv_ac : 5; // qindex adjustment for uv dccoefficient, -15...15 + int reserved : 17; + //round parameters + vp9_round_cfg_s round_seg[8]; // round config for each segment +}nvenc_vp9_quant_cfg_s; //44 bytes + +// vp9 dma config for each tile +typedef struct +{ + unsigned int tile_width : 6; // tile width in super blocks(64x64) + unsigned int tile_height : 6; // tile height in super blocks(64x64) + unsigned int tile_pos_x : 6; // tile x-position in super blocks unit + unsigned int reserved : 14; +}nvenc_vp9_modedma_cfg_s; + +//vp9 smalle config +typedef struct +{ + unsigned int high_prec_mv_enable : 1; // high precision mv enable + unsigned int temporal_mv_enable : 1; // temporal mv enable + unsigned int transform_mode : 3; // transform mode: 0: only4x4, 1: allow_8x8, 2: allow_16x16, 3: allow_32x32, 4:transform_mode_select + unsigned int lossless_enable : 1; // lossless encoding enable + unsigned int mocomp_filter_type : 3; // mocomp filter type:0 eight-tap smooth, 1: eight-tap, 2:eight-tap sharp, 3: bilinear, 4:switchable + unsigned int tile_enable : 1; // enable tile encoding + unsigned int segment_enable : 1; // enable segment + unsigned int segment_upd_enable : 1; // enable explicit map update for current frame. If true, then segment_temp_enable says how to update + unsigned int segment_temp_enable : 1; // Type of map update + unsigned int reserved : 19; +}nvenc_vp9_smalle_config_s; + +//vp9 frame-level parameters +typedef struct +{ + unsigned int pic_width : 13; // picture width in pixels + unsigned int pic_height : 13; // picture height in pixels + unsigned int rec_disable : 1; // disable recon output + unsigned int type : 1; // picture type: 0 Intra/1 Inter + unsigned int refframe : 2; // reference frame type used 1 = last 2= golden 3= alt-ref + unsigned int reserved : 2; + + //ref-frame param + unsigned char ref_index[8]; // ref-frame index for each segment + //skip param + unsigned char skip_mode[8]; // skip-mode enable for each segment: skip mode implies both the use of a (0,0) motion vector and that no residual will be coded + //filter param + nvenc_vp9_filter_cfg_s filter_cfg; // filter config + //quant param + nvenc_vp9_quant_cfg_s quant_cfg; // quant & round config + + nvenc_vp9_smalle_config_s smalle_cfg; // small-e config + unsigned int num_tile_controls_minus1; // number of nvenc_vp9_tile_control_s entries in array minus1 + unsigned int tile_controls_offset; // offset from start of top level "nvenc_vp9_pic_control_s" structure + // to start of array of "nvenc_vp9_tile_control_s" structures + + unsigned int bitstream_buf_start_pos; // bitstream buffer offset start position byte offset + unsigned int bitstream_buf_size; // bitstream buffer size in bytes + + unsigned int combined_linebuf_size; // combined line buffer size in bytes + + unsigned int filterline_linebuf_size; // filter line line-buffer size in bytes + + unsigned int filtercol_linebuf_size; // filter column line-buffer size in bytes +}nvenc_vp9_pic_control_s; //112 bytes + +typedef struct +{ + unsigned int col_pos : 6; // tile position in Super Block of Y axis + unsigned int bs_size : 24; // tile size in Super Block + unsigned int reserved : 2; + nvenc_vp9_modedma_cfg_s mode_dma_cfg; // mode-dma config for this tile + unsigned char tile_idx; // tile index in the picture + unsigned char reserved2[55]; +}nvenc_vp9_tile_control_s; //64bytes + +typedef struct +{ + unsigned char QP[3]; // QP for 0:P picture, ,1:B picture 2; I picture + unsigned char minQP[3]; // min QP for 0:P picture, ,1:B picture 2; I picture + unsigned char maxQP[3]; // max QP for 0:P picture, ,1:B picture 2; I picture + unsigned char reserved[3]; + int rhopbi[3]; // 23.8 signed fixed point quant ratios P/I, B/I, I/I(=1) + + int framerate; // fps + unsigned int buffersize; // total buffer size + + int nal_cpb_size; // size in bits + int nal_bitrate; // rate in bps + + int gop_length; // I period + int Bmin; // 23.8 signed fixed point min buffer level, updated by driver at sequence level only. + int Ravg; // 23.8 signed fixed point average rate, updated by driver at sequence level only. + int R; // 23.8 signed fixed point rate + + +} nvenc_vp9_rc_s; // 56 bytes, currently keep the same as vp8 + + +typedef struct +{ + unsigned int picture_index; // value received from SetPictureIndex method + unsigned int error_status : 2; // report error if any + unsigned int ucode_error_status : 30; // report error status from ucode to driver + unsigned int total_byte_count; // total bitstream size in bytes + unsigned int bitstream_start_pos; // byte_offset where VP9 BS DMA start writing bitstream + unsigned short pic_type; // inter/key frame type + unsigned short avgQP; // average QP + unsigned int cycle_count; // total cycles taked for execute. Written out only when DumpCycleCount in SetControlParams is set + unsigned int intra_cu32x32_count; // number of intra 32x32 CUs. + unsigned int inter_cu32x32_count; // number of inter 32x32 CUs. + unsigned int intra_cu16x16_count; // number of intra 16x16 CUs. + unsigned int inter_cu16x16_count; // number of inter 16x16 CUs. + unsigned int intra_cu8x8_count; // number of intra 8x8 CUs. + unsigned int inter_cu8x8_count; // number of inter 8x8 CUs. + unsigned int cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int total_intra_cost; // sum of all intra best mode's SATDs, each mb has one intra SATD. + unsigned int total_inter_cost; // sum of all inter best mode's SATDs, each mb has one inter SATD. + +} nvenc_vp9_pic_stat_s; // 64 bytes + +// **************************************************************************** +// **************************************************************************** +// VP8 +// **************************************************************************** +// **************************************************************************** + +// TO DO: Find out if we should define # maxSegments as 4 +#define MAXSEGMENTS 4 + +typedef struct +{ + unsigned short frame_width_minus1; // frame width in pixels minus 1, range 0-4095 + unsigned short frame_height_minus1; // frame height in pixels minus 1, range 0-4095 + + unsigned int sfc_pitch : 10; // pitch of luma plane in macorblock units( i.e stride >>4) + unsigned int sfc_pitch_chroma : 10; // pitch of chroma plane in macroblock units + unsigned int reserved2 : 12; + + unsigned int sfc_trans_mode : 4; // least 3 significient bits are used to stand for 8 modes, normal, xflip, yflip etc. + unsigned int block_height : 7; // BL mapping block height setting + unsigned int tiled_16x16 : 1; // Surface is 16x16 tiled instead of BL mapped (must be 1 for refpics) + unsigned int memory_mode : 2; // 0: semi-planar, 1: planar + unsigned int input_bl_mode : 2; // the input block linear mode: 0~gpu bl; 1~tegra bl; 2~naive bl this is only for cmod and cmod/plugin + unsigned int reserved : 16; // pad to int + + unsigned int luma_top_frm_offset; // offset of luma top field or frame in units of 256 + unsigned int luma_bot_offset; // offset of luma bottom field in units of 256. Not used if frame format. + unsigned int chroma_top_frm_offset; // offset of chroma top field or frame, or offset of chroma U(Cb) plane in planar mode, both in units of 256 + unsigned int chroma_bot_offset; // offset of chroma bottom field, not used if frame format. Or offset of chroma V(Cr) plane in planar mode. Both in units of 256. + +} nvenc_vp8_surface_cfg_s; // 28 bytes + +typedef struct +{ + //unsigned int coeff_partitioning_enabled : 1; // 0: coefficient partitioning disabled 1: coeffient partitioning enabled (SE: redundant, same as multi_token_partition!=0) + //unsigned int num_segments_minus_one : 2; // number of segments + unsigned int horiz_scale : 2; // 0=1, 1=5/4, 2=5/3, 3=2 + unsigned int vert_scale : 2; // 0=1, 1=5/4, 2=5/3, 3=2 + //unsigned int golden_frame_enable : 1; // 0: golden frame disabled, 1: golden frame enabled + //unsigned int alt_ref_frame_enable : 1; // 0: alt_ref_frame_disabled, 1: alt_ref_frame_enabled + //unsigned int adaptive_golden_frame_selection : 1; // 0: fixed distance between golden frames 1: adaptive golden frame selection + //unsigned int adaptive_altref_frame_selection : 1; // 0: fixed distance between alt ref frames 1: adaptive alt ref frame selection + unsigned int reserved : 28; // Align to word boundary + + //unsigned int golden_frame_interval : 16; // golden frame interval if adaptive_golden_frame_selection is fixed + //unsigned int altref_frame_interval : 16; // altref frame interval if adaptive_altref_frame_selection is fixed + +} nvenc_vp8_sequence_data_s; // 4 bytes + +typedef struct +{ + unsigned int frame_type : 1; // 0=key frame, 1=interframe + unsigned int version : 3; // VP8 version, also affects luma/chroma mocomp: 0=bicubic/bicubic, 1,2=bilinear/bilinear, 3..7=bilinear/fullpel (SE: perhaps separate flags for bilinear/fullpel?) + unsigned int show_frame : 1; // 0=don't display, 1=display + unsigned int clr_type : 1; // color space: 0=ITU-R BT.601, 1=reserved (multiplier free) + unsigned int clamp_type : 1; // 0=clamping of reconstructed samples required, 1=no clamping necessary (SE: not supported by NVENC?) + // + unsigned int filter_type : 2; // 0=normal, 1=simple 2: disable (SE: filter_type==2 is redundant, same as filter_level==0) + unsigned int filter_level : 6; // loop filter strength (1..63), 0=disabled + unsigned int sharpness_level : 3; // loop filter sharpness (0..7) + unsigned int mode_ref_lf_delta_enabled : 1; // apply mb param dependent loop filter level deltas: 0=disabled, 1=enabled; if disabled the following parameters are not used + unsigned int mode_ref_lf_delta_update : 1; // update ref_lf_deltas and mode_lf_deltas: 0=keep previous, 1=update + unsigned int ref_lf_deltas_update_0 : 1; // update ref_lf_deltas[0]: 0=keep previous, 1=update + unsigned int ref_lf_deltas_update_1 : 1; // update ref_lf_deltas[1]: 0=keep previous, 1=update + unsigned int ref_lf_deltas_update_2 : 1; // update ref_lf_deltas[2]: 0=keep previous, 1=update + unsigned int ref_lf_deltas_update_3 : 1; // update ref_lf_deltas[3]: 0=keep previous, 1=update + unsigned int mode_lf_deltas_update_0 : 1; // update mode_lf_deltas[0]: 0=keep previous, 1=update + unsigned int mode_lf_deltas_update_1 : 1; // update mode_lf_deltas[1]: 0=keep previous, 1=update + unsigned int mode_lf_deltas_update_2 : 1; // update mode_lf_deltas[2]: 0=keep previous, 1=update + unsigned int mode_lf_deltas_update_3 : 1; // update mode_lf_deltas[3]: 0=keep previous, 1=update + unsigned int reserved0 : 4; // Align to word boundary + signed char ref_lf_deltas[4]; // reference frame dependent loop filter level delta (-63..+63) + // Index 0: macroblock is INTRA + // 1: macroblock is INTER and reference frame is LAST + // 2: macroblock is INTER and reference frame is GOLDEN + // 3: macroblock is INTER and reference frame is ALTREF + // lower 7 bits of each entry used as a signed value -63..+63, MS Bit is ignored by HW (SE: invalid values should be rejected?) + signed char mode_lf_deltas[4]; // macroblock mode dependent loop filter level delta (-63..+63) + // Index 0: B_PRED (macroblock is INTRA 4x4) + // 1: ZEROMV (macroblock is INTER and mode is ZEROMV) + // 2: MV (macroblock is INTER and mode is neither ZEROMV nor SPLITMV) + // 3: SPLITMV (macroblock is INTER and mode is SPLITMV (not 16x16)) + // lower 7 bits of each entry used as a signed value -63..+63, MS Bit is ignored by HW (SE: invalid values should be rejected?) + // Other frame header parameters + unsigned int multi_token_partition : 2; // number of tcoeff token partitions (0,1,2,3 -> 1,2,4,8): num_part = 1 << multi_token_partition + unsigned int refresh_entropy_probs : 1; // 0=restore previous probs at end of frame, 1=keep updated probs for next frame + unsigned int mb_no_coeff_skip : 1; // per-mb skip flag, 0=disable 1:enable (misnomer?) + unsigned int update_ymode_prob : 1; // update ymode probs, 0=no, 1=yes + unsigned int update_uvmode_prob : 1; // update uvmode probs, 0=no, 1=yes + //unsigned int coef_update_probs : 1; // update tcoeff token probs, 0=no, 1=yes (SE: isnt't this redundant with coef_probs_update array?) + + unsigned int base_qindex : 7; // quantizer index (0..127) + + unsigned int refresh_last_frame : 1; // use decoded frame as last reference frame + unsigned int refresh_golden_frame : 1; // use decoded frame as golden reference frame + unsigned int refresh_alt_ref_frame : 1; // use decoded frame as alt_ref reference frame + unsigned int copy_buffer_to_gf : 2; // update golden frame from: 0=none (keep), 1=last_frame, 2=alt_ref_frame + unsigned int copy_buffer_to_arf : 2; // update alt_ref frame from: 0=none (keep), 1=last_frame, 2=golden_frame + unsigned int ref_frame_sign_bias_gf : 1; // golden frame is future frame (invert mv sign) + unsigned int ref_frame_sign_bias_arf : 1; // alt_ref frame is future frame (invert mv sign) + unsigned int fixed_ref_frame_probs : 1; //use fixed ref frame probability + unsigned int reserved1 : 9; // Align to word boundary + + + signed int y1dc_delta_q : 5; // quantizer delta for luma DC (-15..+15) + signed int y2dc_delta_q : 5; // quantizer delta for WHT DC (-15..+15) + signed int y2ac_delta_q : 5; // quantizer delta for WHT AC (-15..+15) + signed int uvdc_delta_q : 5; // quantizer delta for chroma DC (-15..+15) + signed int uvac_delta_q : 5; // quantizer delta for chroma AC (-15..+15) + + // + unsigned int segmentation_enabled : 1; // 0=disabled, 1=enabled; if disabled the following parameters are not used + unsigned int update_mb_segmentation_map : 1; // per-mb segment ID map update: 0=keep previous, 1=update (SE: have to provide the actual map) + unsigned int update_mb_segmentation_data : 1; // mb_segment_abs_delta, seg_filter_level, and seg_qindex update: 0=keep previous, 1=update + unsigned int mb_segment_abs_delta : 1; // seg_filter_level and seg_qindex are: 0=delta values, 1=absolute values + unsigned int reserved2 : 3; // Align to word boundary + signed char seg_filter_level[MAXSEGMENTS]; // per-segment loop filter levels + // if mb_segment_abs_delta==0: relative to filter_level (-63..+63) + // if mb_segment_abs_delta==1: absolute value (0..63) + signed char seg_qindex[MAXSEGMENTS]; // per-segment qindex values + // if mb_segment_abs_delta==0: relative to base_qindex (-127..+127) + // if mb_segment_abs_delta==1: absolute value (0..127) + +} nvenc_vp8_picture_data_s; // 32 bytes + +typedef struct +{ + unsigned int early_intra_mode_control : 2; // 0=inter; 1=eval_intra; 2=ext_hint; 3=none (low perf,best qual) + unsigned int early_intra_mode_type_16x16dc : 1; // evaluate 16x16dc if early_intra_mode_control==1 + unsigned int early_intra_mode_type_16x16h : 1; // evaluate 16x16h if early_intra_mode_control==1 + unsigned int early_intra_mode_type_16x16v : 1; // evaluate 16x16v if early_intra_mode_control==1 + unsigned int early_ip_is_final : 1; // if set, result of early intra is final (disable full intra search) + unsigned int chroma_eval_mode : 1; // for intra chroma search 0 = U and V; 1 = U only + unsigned int nearest_mv_enable : 1; // enable nearest mv mode for 16x16 + unsigned int near_mv_enable : 1; // enable near mv mode for 16x16 + unsigned int zero_mv_enable : 1; // enable zero mv mode 16x16 + unsigned int new_mv_enable : 1; // enable new mv mode 16x16 + unsigned int above4x4_mv_enable : 1; // enable above4x4 mv mode for VP8 SPLITMV modes + unsigned int left4x4_mv_enable : 1; // enable left4x4 mv mode for VP8 SPLITMV modes + unsigned int zero4x4_mv_enable : 1; // enable zero4x4 mv mode for VP8 SPLITMV modes + unsigned int new4x4_mv_enable : 1; // enable new4x4 mv mode for VP8 SPLITMV modes + unsigned int l0_part_16x16_enable : 1; // enable L0 prediction for 16x16 + unsigned int l0_part_16x8_enable : 1; // enable L0 prediction for 16x8 + unsigned int l0_part_8x16_enable : 1; // enable L0 prediction for 8x16 + unsigned int l0_part_8x8_enable : 1; // enable L0 prediction for 8x8 + unsigned int pskip_enable : 1; // enable p_skip encoding (still used by ME for special modes, can this be removed?) + //unsigned int use_variance : 1; // to decide between inter and intra (Most likely wont be used) + unsigned int explicit_eval_nearestmv : 1; // enable explicit search vp8 nearest mv + unsigned int explicit_eval_nearmv : 1; // enable explicit search vp8 near mv + unsigned int explicit_eval_left_topmv : 1; // enable explicit search vp8 8x8 left_top mv + unsigned int explicit_eval_left_bottommv : 1; // enable explicit search vp8 8x8 left_bottom mv + unsigned int explicit_eval_top_leftmv : 1; // enable explicit search vp8 8x8 top_left mv + unsigned int explicit_eval_top_rightmv : 1; // enable explicit search vp8 8x8 top_right mv + unsigned int explicit_eval_zeromv : 1; // enable explicit search vp8 zero mv + unsigned int refframe_pic : 2; // reference frame type used 1 = last 2= golden 3= alt-ref + unsigned int reserved0 : 3; // Pad to int + + unsigned int intra_luma4x4_mode_enable : 9; // bitmask indicating which intra luma 4x4 modes to enable + unsigned int intra_luma16x16_mode_enable : 4; // bitmask indicating which intra luma 16x16 modes to enable + unsigned int intra_chroma_mode_enable : 4; // bitmask indicating which intra chroma modes to enable + unsigned int intra_tm_mode_enable : 1; // bitmask indicating intra 4x4 true motion mode enable + unsigned int ip_search_mode : 3; // which intra sizes to search before making IP decision b0=4x4,b2=16x16 // We dont need a bit for 8x8 for VP8 but it might be simpler for reg field to keep it? + unsigned int mv_cost_enable : 1; // enable mv cost calculations in mode decision + unsigned int reserved1 : 10; // pad to full int + + unsigned short early_intra_disable_mpeb_threshold; // if early intra cost is below this, disable full search & select intra + //unsigned short lambda_multiplier_mvcost; // used for mv cost calculation + short bias_favor_intra_16x16_early; // bias for early intra (subtracts from early intra 16x16 cost) + + unsigned short lambda_multiplier_intra[4]; // used for intra mode cost calculation [segment_id] + + unsigned int mdpLambdaCoefMd :16; + unsigned int mdpLambdaCoefMv :16; + + short bias_intra_16x16; // bias for intra 16x16 (subtracts from intra 16x16 cost) + short bias_intra_4x4; // bias for intra 4x4 (subtracts from intra 4x4 cost) + + short bias_inter_16x16; // bias for inter 16x16 (subtracts from inter 16x16 cost) + short bias_inter_16x8; // bias for inter 16x8 (subtracts from inter 16x8 cost) + + short bias_inter_8x16; // bias for inter 8x16 (subtracts from inter 8x16 cost) + short bias_inter_8x8; // bias for inter 8x8 (subtracts from inter 8x8 cost) + + short bias_nearmv; // bias for nearmv (subtracts from nearmv cost) for 16x16 + short bias_nearestmv; // bias for nearestmv (subtracts from nearestmv cost) for 16x16 + + short bias_zeromv; // bias for zeromv (subtracts from zeromv cost) for 16x16 + short bias_newmv; // bias for newmv (subtracts from newmv cost) for 16x16 + + short bias_above4x4mv; // bias for above4x4mv (subtracts from above4x4 cost) for VP8 SPLITMV modes + short bias_left4x4mv; // bias for left4x4mv (subtracts from left4x4 cost) for VP8 SPLITMV modes + + short bias_zero4x4mv; // bias for zero4x4mv (subtracts from zero4x4 cost) for VP8 SPLITMV modes + short bias_new4x4mv; // bias for new4x4mv (subtracts from new4x4 cost) for VP8 SPLITMV modes + + short mv_cost_bias; // bias applied to internal motion vector cost calculation + short bias_intra_over_inter; // bias for any intra (subtracts from intra cost) + unsigned short intra_refresh_cnt; // distance between forced-intra MBs; 0 means OFF + unsigned short intra_refresh_offs; // first forced-intra MB + + unsigned short tempmv_wt_spread_threshold; // The threshold for the sum of difference between 8x8 and 16x16 mvs. Used for temporal mv weight. + unsigned short tempmv_wt_distort_threshold; // The threshold for cost of the 16x16 mv. Used for temporal mv weight. +} nvenc_vp8_md_control_s; // 64 bytes + +typedef struct +{ + unsigned int self_temporal_stamp_l0 : 3; // stamp to use for L0 integer search in stamp based mode + //unsigned int self_temporal_stamp_l1 : 3; // stamp to use for L1 integer search in stamp based mode + unsigned int self_temporal_explicit : 1; // explicitly evaluate these vectors + unsigned int self_temporal_search : 1; // integer search enable for this hint + unsigned int self_temporal_refine : 1; // subpel search enable for this hint + unsigned int self_temporal_enable : 1; // hint enable (enables processing + fetching of data) (only changeable at first slice in picture) + + unsigned int coloc_stamp_l0 : 3; // stamp to use for L0 integer search in stamp based mode + //unsigned int coloc_stamp_l1 : 3; // stamp to use for L1 integer search in stamp based mode + unsigned int coloc_explicit : 1; // explicitly evaluate these vectors + unsigned int coloc_search : 1; // integer search enable for this hint + unsigned int coloc_refine : 1; // subpel search enable for this hint + unsigned int coloc_enable : 1; // hint enable (enables processing + fetching of data) (only changeable at first slice in picture) + + unsigned int self_spatial_stamp_l0 : 3; // stamp to use for L0 integer search in stamp based mode + //unsigned int self_spatial_stamp_l1 : 3; // stamp to use for L1 integer search in stamp based mode + unsigned int self_spatial_explicit : 1; // explicitly evaluate these vectors + unsigned int self_spatial_search : 1; // integer search enable for this hint + unsigned int self_spatial_refine : 1; // subpel search enable for this hint + unsigned int self_spatial_enable : 1; // hint enable (enables processing + fetching of data) (only changeable at first slice in picture) + + unsigned int reserved0 : 11; // Pad to int + + unsigned int external_stamp_lastref_stamp : 3; // Stamp to use for external L0 hint with refidx = 0 + unsigned int external_stamp_goldref_stamp : 3; // Stamp to use for external L0 hint with refidx = 1 + unsigned int external_stamp_altref_stamp : 3; // Stamp to use for external L0 hint with refidx = 2 + unsigned int external_explicit : 1; // explicitly evaluate these vectors + unsigned int external_search : 1; // integer search enable for this hint + unsigned int external_refine : 1; // subpel search enable for this hint + unsigned int external_enable : 1; // hint enable (enables processing + fetching of data) (only changeable at first slice in picture) + + unsigned int const_mv_stamp_l0 : 3; // stamp to use for L0 integer search in stamp based mode + //unsigned int const_mv_stamp_l1 : 3; // stamp to use for L1 integer search in stamp based mode + unsigned int const_mv_explicit : 1; // explicitly evaluate these vectors + unsigned int const_mv_search : 1; // integer search enable for this hint + unsigned int const_mv_refine : 1; // subpel search enable for this hint + unsigned int const_mv_enable : 1; // hint enable (enables processing + fetching of data) + + unsigned int reserved1 : 12; // pad to int + +} nvenc_vp8_me_hint_cfg_s; // 8 bytes + +// Same as H264 me_limit_mv struct +typedef struct +{ + unsigned int mv_limit_enable : 1; // 0=disable, 1=enable set the absolute MV range (Cisco flip camera request) + unsigned int left_mvx_frac : 2; // vector X component fraction + int left_mvx_int : 12; // vector X component integer portion + unsigned int reserved1 : 1; // padding X to 16 bit + unsigned int top_mvy_frac : 2; // vector Y component fraction + int top_mvy_int : 10; // vector Y component integer portion + unsigned int reserved2 : 4; // padding Y to 16 bit + + unsigned int right_mvx_frac : 2; // vector X component fraction + int right_mvx_int : 12; // vector X component integer portion + unsigned int reserved3 : 2; // padding X to 16 bit + unsigned int bottom_mvy_frac : 2; // vector Y component fraction + int bottom_mvy_int : 10; // vector Y component integer portion + unsigned int reserved4 : 4; // padding Y to 16 bit + +} nvenc_vp8_me_limit_mv_s; // 8 bytes + +typedef struct +{ + unsigned int mvx_frac : 2; // vector X component fraction + int mvx_int : 12; // vector X component integer portion + unsigned int reserved_x : 2; // padding X to 16 bit + unsigned int mvy_frac : 2; // vector Y component fraction + int mvy_int : 10; // vector Y component integer portion + unsigned int reserved_y : 4; // padding Y to 16 bit + +} nvenc_vp8_me_const_vec_s; // 4 bytes + +// TO DO: We are wasting ~7 ints here we can combine this to pack hor_adjust and ver_adjust in diff int +// but dont do it if we are going to throw it away in padding anyways +typedef struct +{ + unsigned int bitmask[2] ; // 8x8 search point pattern bitmask + unsigned int hor_adjust : 1; // shift hor based on lsb of pred + unsigned int ver_adjust : 1; // shift ver based on lsb of pred + unsigned int reserved : 30; + +} nvenc_vp8_me_stamp_s; // 12 bytes + +typedef struct +{ + unsigned int me_predictor_mode : 1; // 0=MDP, 1=const + unsigned int refinement_mode : 1; // 0=hpel, 1=qpel + unsigned int lambda_mode : 1; // 0=MDP, 1=const + unsigned int const_lambda : 11; // U8.3 format lambda, used if lambda_mode=1 + unsigned int refine_on_search_enable : 1; // 0=disable, 1=enable subpel refine for search candidates + unsigned int me_only_mode : 1; + unsigned int fps_mvcost : 1; + unsigned int sps_mvcost : 1; + unsigned int sps_cost_func : 1; + unsigned int me_icc_mode_mad : 1; + unsigned int sps_filter : 3; + unsigned int mc_filter : 3; + unsigned int mbc_ctrl_arbitor : 2; + unsigned int average_mvhint_enable : 1; + unsigned int rmvp_source_half_scaled : 1; // only support picture size half scale down in horz&vert + unsigned int reserved : 2; + + nvenc_vp8_me_limit_mv_s limit_mv; + nvenc_vp8_me_hint_cfg_s predsrc; // predictor sources + nvenc_vp8_me_const_vec_s l0_hint; // constant MV used as L0 hint + nvenc_vp8_me_const_vec_s l1_hint; // constant MV used as L1 hint + nvenc_vp8_me_const_vec_s l0_pred; // constant MV used as L0 predictor + // should this be made an array? + nvenc_vp8_me_stamp_s shape0; // priority 0 stamping pattern + nvenc_vp8_me_stamp_s shape1; // priority 1 stamping pattern + nvenc_vp8_me_stamp_s shape2; // priority 2 stamping pattern + nvenc_vp8_me_stamp_s shape3; // priority 3 stamping pattern + nvenc_vp8_me_stamp_s shape4; // priority 4 stamping pattern + nvenc_vp8_me_stamp_s shape5; // priority 5 stamping pattern + nvenc_vp8_me_stamp_s shape6; // priority 6 stamping pattern + nvenc_vp8_me_stamp_s shape7; // priority 7 stamping pattern + + nvenc_me_tebif_s teb_para; // teb parameters + unsigned char mbc_mb_size; // MBC size in MB + unsigned char tempmv_scale; // temporal mv scale + unsigned char reserved1[2]; + +} nvenc_vp8_me_control_s; // 136 bytes + +typedef struct +{ + unsigned int qpp_run_vector_4x4 : 16; // cost values for 4x4 transform (16 bit total vector) + unsigned int qpp_luma8x8_cost : 4; // luma 8x8 cost threshold, 0 = throw out all coefs, range 0-15 + unsigned int qpp_luma16x16_cost : 4; // luma 16x16 cost threshold, 0 = throw out all coefs, range 0-15 + unsigned int qpp_chroma_cost : 4; // chroma cost threshold, 0 = throw out all coefs, range 0-15 + unsigned int qpp_mode : 2; // 0 = OFF, 1 = 8x8, 2 = 16x16_8x8 + //unsigned int quant_intra_sat_flag : 1; // 0: no saturation is applied; 1: saturation is applied + //unsigned int quant_inter_sat_flag : 1; // 0: no saturation is applied; 1: saturation is applied + unsigned int reserved : 2; + + //unsigned int quant_intra_sat_limit : 16; // quantization saturation limit for intra MB + //unsigned int quant_inter_sat_limit : 16; // quantization saturation limit for inter MB + + unsigned short dz_4x4_YI[16]; // deadzone for 4x4 transform of Luma Intra + unsigned short dz_4x4_YP[16]; // deadzone for 4x4 transform of Luma Inter + unsigned short dz_4x4_CI; // deadzone for 4x4 transform of Chroma Intra + unsigned short dz_4x4_CP; // deadzone for 4x4 transform of Chroma Inter + +} nvenc_vp8_quant_control_s; // 72 bytes + +typedef struct +{ + //char temp_dist_l0[3]; // temporal distance of each ref pic (last, gold, altref) + //char reserved; + unsigned int hist_buf_size; // size in bytes of the buffer allocated for history + unsigned int bitstream_buf_size; // size in bytes of the buffer allocated for bitstream slice/mb data + unsigned int bitstream_start_pos; // start position in bitstream buffer where data should be written (byte offset) + unsigned int bitstream_residual_buf_size; // size in bytes of the buffer allocated for residual partition bitstream slice/mb data + unsigned int bitstream_residual_start_pos[8]; // start position in residual partition bitstream buffer where data for residual partitions should be written (byte offset) + +} nvenc_vp8_pic_control_s; // 48 bytes + +typedef struct +{ + //unsigned char hrd_type; // nal (=2) and vcl (=1) type + unsigned char QP[3]; // QP for 0:P picture, ,1:B picture 2; I picture + unsigned char minQP[3]; // min QP for 0:P picture, ,1:B picture 2; I picture + unsigned char maxQP[3]; // max QP for 0:P picture, ,1:B picture 2; I picture + //unsigned char maxQPD; // max QP delta between two consecutive QP updates + //unsigned char baseQPD; // initial QP delta between two consecutive QP updates + unsigned char reserved[3]; + int rhopbi[3]; // 23.8 signed fixed point quant ratios P/I, B/I, I/I(=1) + + int framerate; // fps + unsigned int buffersize; // total buffer size + + int nal_cpb_size; // size in bits + int nal_bitrate; // rate in bps + //int vcl_cpb_size; // size in bits + //int vcl_bitrate; // rate in bps + + int gop_length; // I period + //int Np; // 27.4 signed fixed (gop_length + num_b_frames -1)/(num_b_frames + 1) + int Bmin; // 23.8 signed fixed point min buffer level, updated by driver at sequence level only. + int Ravg; // 23.8 signed fixed point average rate, updated by driver at sequence level only. + int R; // 23.8 signed fixed point rate + + //unsigned char ab_alpha; // weight of prev frame activity compared to part of the current picture + //unsigned char ab_beta; // weight of current activity compared to neighbors + //unsigned short prev_act; // average activity level of the first reference (L0) +} nvenc_vp8_rc_s; // 56 bytes + +typedef struct +{ + // 256 aligned + unsigned int magic; // magic number + nvenc_vp8_surface_cfg_s refpic_cfg; // 28 bytes + nvenc_vp8_surface_cfg_s input_cfg; // 28 bytes + nvenc_vp8_surface_cfg_s outputpic_cfg; // 28 bytes + nvenc_vp8_sequence_data_s sequence_data; // 4 bytes + nvenc_vp8_picture_data_s picture_data; // 32 bytes + nvenc_vp8_pic_control_s pic_control; // 48 bytes + nvenc_vp8_me_control_s me_control; // 136 bytes + nvenc_vp8_md_control_s md_control; // 64 bytes + nvenc_vp8_quant_control_s quant_control; // 72 bytes + nvenc_vp8_rc_s rate_control; // 56 bytes + unsigned char reserved[16]; // pad to 512 bytes + +} nvenc_vp8_drv_pic_setup_s; // 512 bytes + + +// 2.2. Colocated Motion Data Buffer +// --------------------------------- +// +// This buffer will be used to store co-located motion data. This data is produced for +// pictures that are marked as reference pictures. The buffer is not needed for non- +// reference pictures and in such case the corresponding method may be skipped. The +// data in this buffer will be needed for future pictures encoded by both CEA and NVENC. +// There are two of these buffers: one output, written only if the current picture is a +// reference picture, and one input, read if the current picture is being encoded +// as a B picture, and possible also during P picture to serve as hints for motion search. +// Each buffer holds coloc data for all macroblocks in the picture. This +// data is packed into an array of 16 entries of 4 bytes. Each entry holds one motion +// vector and 6 bits of additional data, which is formatted as follows: +// +// coldata array entry N: contents of 'e': +// 0, 4, 8, 12 0 = bottom field flag for ref picture +// 4..1 = reference picture number (index in DPB) +// 5 = zero +// 1, 5, 9, 13 3..0 = ColZero flag for blocks N+2 .. N-1 +// 5..4 = subpartition size for 8x8 block (N-1)/4 +// (0=8x8, 1=8x4, 2=4x8, 3=4x4) +// 15 0 = field mode flag +// 1 = intra coded +// 3..2 = partition size for macroblock +// (0=16x16, 1=16x8, 2=8x16, 3=8x8) +// 5..4 = zero +// all others 5..0 = zero +// +// Note that motion vectors are duplicated for each 4x4 index from the top-left 4x4 block +// of each (sub)partition. Note also that since NVENC does not currently support sub-8x8 +// inter partitioning, the 4 vectors within an 8x8 are replicated values. + +typedef struct +{ + unsigned int mvx : 14; // bit 0..13: x component of MV + unsigned int mvy : 12; // bit 14..25: y component of MV + unsigned int e : 6; // bit 26..31: extra data, different per entry +} nvenc_h264_coloc_entry_s; + +typedef struct // 64 byte aligned block +{ + nvenc_h264_coloc_entry_s coldata[16]; // col data for 1 macroblock = array of 16 entries +} nvenc_h264_coloc_mb_s; + + +// 2.3. History Buffer +// ------------------- +// +// This buffer will be used to keep history data (data produced for one macroblock that +// must be kept until the macroblock in the row directly below is processed). This buffer +// should be at least as large as (((frame_width_minus1+16) / 16) + 16)*128 bytes. +// +// After the execute completes, the data in this buffer is no longer needed and the +// buffer can be discarded/deallocated. +// +// The internal structure of the history record is implementation specific and therefore +// subject to change. The data is produced and consumed by NVENC only and should never be +// interpreted by any other engine or by software. The current structure is defined in +// the following visio document in the HW development tree: +// +// doc/nvenc/arch/hist_layout.vsd +// +// 2.4. Reference and Input Picture Buffer +// --------------------------------------- +// +// Reference Picture Buffer contains reference pictures produced by NVENC during +// encode opeations of previous pictures. The Input Picture Buffer contains the +// current picture that needs to be encoded into a bitstream. +// +// Each picture is actually a separate buffer with its own surface offset method. +// Relative position of luma, chroma fields are defined in the picture setup buffer +// and shown in the figure below. +// +// Driver needs to send SetRefPicOffset#() method for only the used reference pictures +// for the current execute. +// +// Pictures are by default stored in NV24 field or NV12 frame format. NVENC will +// always output in the same format as the current picture that is being encoded +// (frame or field). True picaff (using frame refs to encode a field or using field +// refs to encode a frame) is NOT supported. +// +// The input picture must contain fields if the picture is to be encoded as fields, +// and must contain frames if the picture is to be encoded as a frame. +// +// +// +---------------+ <-- RefPicOffset (start of buffer) +// |XXXXXXXXXXXXXXX| +// +---------------+ <-- luma_top_frm_offset (in picture setup buffer, in units of 256 bytes) +// |TOP/FRM luma | +// | | +// +XXXXXXXXXXXXXXX+ +// +---------------+ <-- luma_bot_offset +// |BOTTOM luma | +// | | +// +XXXXXXXXXXXXXXX+ +// +---------------+ <-- chroma_top_frm_offset +// |TOP/FRM chroma | +// +XXXXXXXXXXXXXXX+ +// +---------------+ <-- chroma_bot_offset +// |BOTTOM chroma | +// +XXXXXXXXXXXXXXX+ +// +---------------+ <-- end of buffer +// +// +// 2.5. CEA Hints Data Buffer +// --------------------------- +// +// CEA (a CUDA based integer motion search kernel) can provide hints to NVENC. +// These hints can either be taken as explicit vectors to evaluate or as guides +// for NVENC's built-in integer search. +// In this mode of operation, CEA and NVENC typically work on the same frame, +// with CEA slightly leading (by a few macroblock rows). NVENC would then read +// in the data already produced by CEA for the current picture and use it. CEA +// has to provide information on how many macroblocks worth of data it has +// already completed and written out into this buffer, so that NVENC can read +// up to that point. The buffer must be big enough to contain all data for the +// entire frame. +// +// This buffer consists of 2 sections: a header and a data section. One such +// buffer must be provided/allocated by driver for each picture that needs to +// be encoded. The header section must be prefilled with zeroes by driver. The +// data section can remain uninitialized. +// +// The data section contains the motion vector candidates for each macroblock. +// It is written by CEA and read by NVENC. +// +// The header section contains flow control and status information produced by +// CEA for this picture, and contains an offset to the start of the data +// section. The header section begins at offset 0 in this buffer. + + +typedef struct +{ + unsigned int cand_cnt_blk_16x16 : 4; // 3:0 Number of candidates per block types + unsigned int cand_cnt_blk_16x8 : 4; // 7:4 + unsigned int cand_cnt_blk_8x16 : 4; // 11:8 + unsigned int cand_cnt_blk_8x8 : 4; // 15:12 + unsigned int cand_cnt_blk_8x4 : 4; // 19:16 + unsigned int cand_cnt_blk_4x8 : 4; // 23:20 + unsigned int cand_cnt_blk_4x4 : 4; // 27:24 + unsigned int cand_bdirect_spatial_zmv : 1; // 28 enable Spatial direct enable ZMV 4x4 scores + unsigned int cand_bdirect_spatial_nzmv : 1; // 29 enable Spatial direct enable non-zero MV 4x4 scores + unsigned int cand_bdirect_temporal : 1; // 30 enable Temporal direct mode. Only valid for Bi-type. + unsigned int cand_pskip : 1; // 31 enable p-skip mode. Only valid for L0 controls. +} cea_cand_count_blktype_s; + +typedef struct +{ + // 64 aligned + unsigned int picture_index; // picture index given by driver (via method) + unsigned int data_start; // start offset in buffer of data section which must be 256B aligned + // the following fields are only there right now to make it easier to implement the cmodel plugin mode + // they might be removed at a later stage + unsigned int cea_lambda; + cea_cand_count_blktype_s blk_counts[3]; + int reserved[10]; //padded by 8 bytes to align the end of the structure to 32 bytes for dmread operations +} cea_hints_header_s; + + +// Please refer to the document named "nvenc_cea_mvinfo_format.doc" for more information. +// This document can be found in the nvenc HW development tree under doc/nvenc/arch + +typedef struct +{ + int mvx : 12; // 11:00 x component of integer pixel MV (relative to current MB) S12.0 + int mvy : 10; // 21:12 y component of integer pixel MV (relative to current MB) S10.0 + int refidx : 5; // 26:22 reference index (31=invalid) + int dir : 1; // 27:27 0=L0 1=L1 + int part_type : 2; // 29:28 0=16x16 1=16x8 2=8x16 3=8x8 (blocks in partition must be consecutive) + int last_of_part: 1; // 30:30 last MV of (sub) partition + int last_of_mb : 1; // 31:31 last MV of macroblock +} cea_hint_s; + +// External hint data structure for Orin AV1 +typedef struct +{ + short refidx : 5; // 4:0 reference index (31=invalid) + short direction : 1; // 5:5 0=L0 1=L1 + short bi : 1; // 6:6 0=single mv, 1=compound mv + short partition_type : 3; // 9:7 parition type: 0: 2NX2N, 1:2NxN, 2:Nx2N. reserved 3bits for modes in later projects + short x8 : 3; // 12:10 x8: current parition's top left x position in 8 pixel unit + short last_of_cu : 1; // 13:13 last mv of current CU + short last_of_sb : 1; // 14:14 last mv of current SB + short reserved0 : 1; + + short mvx : 14; // 13:0 mvx, in S12.2 format + short cu_size : 2; // 15:14 CU size: 0: 8x8, 1: 16x16, 2:32x32, 3:64x64 + + short mvy : 12; // 11:0 mvy, in S10.2 format + short y8 : 3; // 14:12 y8: current parition's top left y position in 8 pixel unit + short reserved1 : 1; +} av1_hint_data; //6 bytes + + +typedef struct +{ + unsigned int segID[4]; //segment ID for one MB. each 4x4 block has a segID which is 8 bits, total is 8bitx16 = 128 bits. maximun segments supported is 32 +} cea_segment_s; + + + +// 2.6. Encoder Status Buffer +// --------------------------- +// +// This buffer provides some additional data about the encoded bit stream that can +// be used for rate control and multi-pass encoding as well as status information +// that can be used for debug. + +typedef struct +{ + unsigned int picture_index; // value received from SetPictureIndex method + unsigned int error_status : 2; // report error if any + unsigned int ucode_error_status : 30; // report error status from ucode to driver + unsigned int total_bit_count; // picture size in bits + unsigned int type1_bit_count; // type1 bit count for the entire picture + unsigned short pic_type; // copied from pic_type in nvenc_h264_pic_control_s + unsigned short num_slices; // number of slices produced + unsigned short ave_activity; // report average activity if activity based RC enabled + unsigned short avgQP; // average QP + unsigned int cycle_count; // total cycles taked for execute. Written out only when DumpCycleCount in SetControlParams is set + int hrdFullness; + unsigned int bitstream_start_pos; // byte_offset where mpec will start writing slice header + unsigned int last_valid_byte_offset; // greatest offset in output buffer which is assured to have valid compressed data + unsigned short intra_mb_count; // number of intra MBs. + unsigned short inter_mb_count; // number of inter MBs. + unsigned int cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int total_intra_cost; // sum of all intra best mode's SATDs, each mb has one intra SATD. + unsigned int total_inter_cost; // sum of all inter best mode's SATDs, each mb has one inter SATD. + unsigned int complexity; // sum of mbRowBitcount * current QS + short average_mvx; // average mvx info, can be used for ME hint + short average_mvy; // average mvy info, can be used for ME hint + unsigned short actual_min_qp_used; // actual minimum qp used in the frame + unsigned short actual_max_qp_used; // actual maximum qp used in the frame + unsigned int hw_perf_me_sps; // me slice per second + unsigned int hw_perf_me_fps; // me frame per second + unsigned int hw_perf_mpec_active_cnt; // mpec active count + unsigned int hw_perf_mpeb_active_cnt; // mpeb active count + unsigned long long int total_dist_ssd; // sum of all ssd distortion of the MB best mode + unsigned int total_dist_ssim; // sum of all ssim distortion of the MB best mode + unsigned int best_satd_l0; // sum of all best satd cost of L0 + unsigned int best_satd_l1; // sum of all best satd cost of L1 + unsigned int reserved[5]; // align to 128 bytes + +} nvenc_pic_stat_s; // 128 bytes + + +typedef struct +{ + unsigned int slice_offset; // slice offset in bytes + unsigned int slice_size; // slice size in bytes + unsigned int slh_bit_count; // slice header size in bits + unsigned char reserved[4]; +} nvenc_slice_stat_s; // 16 bytes per status + + +typedef struct +{ + unsigned int intra_res_bit_count; // #bits spent on intra residuals + unsigned int inter_res_bit_count; // #bits spent on inter residuals + unsigned int intra_mode_bit_count; // #bits spent on intra modes (mbtype, pred mode) + unsigned int inter_mode_bit_count; // #bits spent on inter modes (mbtype, submbtype, refidx, mv) + unsigned int ipcm_bit_count; // # bits spent on MB mode that caused IPCM rewind. + unsigned int type1_bit_count; // count of bits in type I bitstream (excludes leading_zero_8bits, + // start_codeprefix_one_3bytes and trailing_zero_8bits) + unsigned int total_bit_count; // total number of bits produced + + unsigned int total_bin_count; // total number of bins produced by cabac + unsigned int cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int cumulative_ipcm_cost; // sum of all inter SA(T)Ds for the MBs that caused IPCM rewind + unsigned int cumulative_skip_cost; // sum of all P and B skip MB SA(T)Ds + + unsigned short intra_mb_count; // number of intra MBs. + unsigned short inter_mb_count; // number of inter MBs. + unsigned short skip_mb_count; // number of skip MBs. + unsigned short ipcm_mb_count; // number of ipcm MBs. + + unsigned int total_intra_cost; // sum of all intra best mode's SATDs, each ctb has one intra SATD. + unsigned int total_inter_cost; // sum of all inter best mode's SATDs, each ctb has one inter SATD. +} nvenc_mpec_stat_s; // 64 bytes per status + +typedef struct +{ + unsigned int qp_val : 8 ; // qp used for encoding this block + unsigned int total_bit_count : 24 ; // total number of bits produced + unsigned int res_cost : 24 ; // satd cost of residual error + unsigned int flags : 8 ; // flags +} nvenc_stats_fifo_s; // 8 bytes per stat + +typedef struct +{ + nvenc_pic_stat_s pic_stat; // 64Bytes + nvenc_slice_stat_s slice_stat[((4096/16)*(4096/16)) + 3]; // To account for 1mb/slice and 4kx4k, plus 3 for alignment + nvenc_mpec_stat_s mpec_stat[((4096/16)*(4096/16)) + 3]; // To account for stat/1mb and 4kx4k, plus 3 for alignment + nvenc_aq_stat_s aq_stats[((4096/16)*(4096/16)) + 3]; // AQ stats from PDMA + nvenc_stats_fifo_s stats_fifo[((4096/16)*(4096/16)) + 3]; +} nvenc_stat_data_s; + +typedef struct +{ + unsigned int picture_index; // value received from SetPictureIndex method + unsigned int error_status : 2; // report error if any + unsigned int ucode_error_status : 30; // report error status from ucode to driver + unsigned int total_bit_count; // picture size in bits + unsigned int type1_bit_count; // type1 bit count for the entire picture + unsigned short pic_type; // copied from pic_type in nvenc_h264_pic_control_s + unsigned short num_slices; // number of slices produced, maxinum slice number is 65535 + unsigned short ave_activity; // report average activity if activity based RC enabled + unsigned short avgQP; // average QP + unsigned int cycle_count; // total cycles taked for execute. Written out only when DumpCycleCount in SetControlParams is set + int hrdFullness; + unsigned int bitstream_start_pos; // byte_offset where mpec will start writing slice header + unsigned int last_valid_byte_offset; // greatest offset in output buffer which is assured to have valid compressed data + unsigned int intra_cu32x32_count; // number of intra 32x32 CUs. + unsigned int inter_cu32x32_count; // number of inter 32x32 CUs. + unsigned int intra_cu16x16_count; // number of intra 16x16 CUs. + unsigned int inter_cu16x16_count; // number of inter 16x16 CUs. + unsigned int intra_cu8x8_count; // number of intra 8x8 CUs. + unsigned int inter_cu8x8_count; // number of inter 8x8 CUs. + unsigned int cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int total_intra_cost; // sum of all intra best mode's SATDs, each mb has one intra SATD. + unsigned int total_inter_cost; // sum of all inter best mode's SATDs, each mb has one inter SATD. + short average_mvx; // average mvx info, can be used for ME hint + short average_mvy; // average mvy info, can be used for ME hint + unsigned short actual_min_qp_used; // actual minimum qp used in the frame + unsigned short actual_max_qp_used; // actual maximum qp used in the frame + unsigned int hw_perf_me_sps; // me slice per second + unsigned int hw_perf_me_fps; // me frame per second + unsigned int hw_perf_mpec_active_cnt; // mpec active count + unsigned int hw_perf_mpeb_active_cnt; // mpeb active count + unsigned char reserved[24]; +} nvenc_h265_pic_stat_s; // 128 bytes + +typedef struct +{ + unsigned int slice_offset; // slice offset in bytes + unsigned int slice_size; // slice size in bytes + unsigned int slh_bit_count; // slice header size in bits + unsigned int slice_size_ctbs; // slice size in ctbs +}nvenc_h265_slice_stat_s; // 16 bytes per status + +typedef struct +{ + unsigned int intra_res_bit_count; // #bits spent on intra residuals + unsigned int inter_res_bit_count; // #bits spent on inter residuals + unsigned int intra_mode_bit_count; // #bits spent on intra modes (mbtype, pred mode) + unsigned int inter_mode_bit_count; // #bits spent on inter modes (mbtype, submbtype, refidx, mv) + unsigned int ipcm_bit_count; // # bits spent on MB mode that caused IPCM rewind. + unsigned int type1_bit_count; // count of bits in type I bitstream (excludes leading_zero_8bits, + // start_codeprefix_one_3bytes and trailing_zero_8bits) + unsigned int total_bit_count; // total number of bits produced + + unsigned int total_bin_count; // total number of bins produced by cabac + unsigned int cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int cumulative_ipcm_cost; // sum of all inter SA(T)Ds for the MBs that caused IPCM rewind + unsigned int cumulative_skip_cost; // sum of all P and B skip MB SA(T)Ds + unsigned int total_intra_cost; // sum of all intra best mode's SATDs, each ctb has one intra SATD. + unsigned int total_inter_cost; // sum of all inter best mode's SATDs, each ctb has one inter SATD. + + unsigned int intra_cu32x32_count; // number of intra 32x32 CUs. + unsigned int inter_cu32x32_count; // number of inter 32x32 CUs. + unsigned int skip_cu32x32_count; // number of skip 32x32 CUs. + unsigned int ipcm_cu32x32_count; // number of ipcm 32x32 CUs. + unsigned int intra_cu16x16_count; // number of intra 16x16 CUs. + unsigned int inter_cu16x16_count; // number of inter 16x16 CUs. + unsigned int skip_cu16x16_count; // number of skip 16x16 CUs. + unsigned int ipcm_cu16x16_count; // number of ipcm 16x16 CUs. + unsigned int intra_cu8x8_count; // number of intra 8x8 CUs. + unsigned int inter_cu8x8_count; // number of inter 8x8 CUs. + unsigned int skip_cu8x8_count; // number of skip 8x8 CUs. + unsigned int ipcm_cu8x8_count; // number of ipcm 8x8 CUs. + + unsigned int total_emulated_bit_count; // total number of emulated bits in total_bit_count + unsigned char reserved[20]; +} nvenc_h265_mpec_stat_s; // 128 bytes per status + +typedef struct +{ + unsigned int var_min; + unsigned int var_max; + unsigned int var_avg; + unsigned char reserved[4]; +} nvenc_h265_aq_stat_s; // 16 bytes + +typedef struct +{ + unsigned int qp_val : 8 ; // qp used for encoding this block + unsigned int total_bit_count : 24 ; // total number of bits produced + unsigned int res_cost : 24 ; // satd cost of residual error + unsigned int flags : 8 ; // flags +} nvenc_h265_stats_fifo_s; // 8 bytes per stat + +typedef struct +{ + nvenc_h265_pic_stat_s pic_stat; // 64Bytes + nvenc_h265_slice_stat_s slice_stat[((4096/16)*(4096/16)) + 3]; // To account for 1mb/slice and 4kx4k, plus 3 for alignment. support upto 65535 slices + nvenc_h265_mpec_stat_s mpec_stat[((4096/16)*(4096/16)) + 3]; // To account for stat/1mb and 4kx4k, plus 3 for alignment. support upto 65535 slices + nvenc_h265_aq_stat_s aq_stats [((4096/16)*(4096/16)) + 3]; // AQ stats from pdma. support upto 65535 slices + nvenc_h265_stats_fifo_s stats_fifo[((4096/16)*(4096/16)) + 3]; +} nvenc_h265_stat_data_s; + +typedef struct +{ + unsigned int picture_index; // value received from SetPictureIndex method + unsigned int error_status : 2; // report error if any + unsigned int ucode_error_status : 30; // report error status from ucode to driver + unsigned int total_bit_count; // picture size in bits + unsigned int type1_bit_count; // type1 bit count for the entire picture + unsigned short num_tiles; // number of tiles produced + unsigned char av1_start_QP; // AV1 QP value at frame encoding start (to be inserted into frame header by driver) + unsigned char loop_filter_level; // loop filer value computed by ucode (to be inserted into frame header by driver) + unsigned int pic_type : 14; // copied from pic_type in nvenc_h264_pic_control_s + unsigned int cdef_damping_minus_3 : 2; // CDEF damping value computed by ucode (to be inserted into frame header by driver) + int avgQP : 16; // average QP + unsigned int cycle_count; // total cycles taked for execute. Written out only when DumpCycleCount in SetControlParams is set + unsigned int hrdFullness; + unsigned int bitstream_start_pos; // byte_offset where mpec will start writing slice header + unsigned int intra_cu64x64_count; // number of intra 64x64 CUs. + unsigned int inter_cu64x64_count; // number of inter 64x64 CUs. + unsigned int intra_cu32x32_count; // number of intra 32x32 CUs. + unsigned int inter_cu32x32_count; // number of inter 32x32 CUs. + unsigned int intra_cu16x16_count; // number of intra 16x16 CUs. + unsigned int inter_cu16x16_count; // number of inter 16x16 CUs. + unsigned int intra_cu8x8_count; // number of intra 8x8 CUs. + unsigned int inter_cu8x8_count; // number of inter 8x8 CUs. + unsigned int cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int total_intra_cost; // sum of all intra best mode's SATDs, each mb has one intra SATD. + unsigned int total_inter_cost; // sum of all inter best mode's SATDs, each mb has one inter SATD. + short average_mvx; // average mvx info, can be used for ME hint + short average_mvy; // average mvy info, can be used for ME hint + short actual_min_qp_used; // actual minimum qp used in the frame + short actual_max_qp_used; // actual maximum qp used in the frame + unsigned int hw_perf_me_sps; // me slice per second + unsigned int hw_perf_me_fps; // me frame per second + unsigned int hw_perf_entropy_active_cnt; // Entropy active count + unsigned int hw_perf_rdo_active_cnt; // RDO active count + unsigned short ave_activity; // report average activity if activity based RC enabled + unsigned char reserved[18]; +} nvenc_av1_pic_stat_s; // 128 bytes + +typedef struct +{ + unsigned int tile_offset; // tile offset in bytes + unsigned int tile_size; // tile size in bytes + unsigned short tile_bitstream_idx; // Index of bistream base address the tile offset refers to (for multi entropy cores use - set to 0 otherwise) + unsigned short tile_idx; // Tile index at strip level + unsigned int tile_size_sbs; // tile size in sbs +}nvenc_av1_tile_stat_s; // 16 bytes per status + +typedef struct +{ + unsigned int res_bit_count; // #bits spent on residuals + unsigned int mode_bit_count; // #bits spent on modes (mbtype, pred mode) + unsigned int type1_bit_count; // count of bits in type I bitstream (excludes leading_zero_8bits, + // start_codeprefix_one_3bytes and trailing_zero_8bits) + unsigned int total_bit_count; // total number of bits produced + + unsigned int total_bin_count; // total number of bins produced by arithmetic coder + unsigned int cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int total_intra_cost; // sum of all intra best mode's SATDs, each ctb has one intra SATD. + unsigned int total_inter_cost; // sum of all inter best mode's SATDs, each ctb has one inter SATD. + + unsigned int intra_cu64x64_count; // number of intra 64x64 CUs. + unsigned int inter_cu64x64_count; // number of inter 64x64 CUs. + unsigned int intra_cu32x32_count; // number of intra 32x32 CUs. + unsigned int inter_cu32x32_count; // number of inter 32x32 CUs. + unsigned int intra_cu16x16_count; // number of intra 16x16 CUs. + unsigned int inter_cu16x16_count; // number of inter 16x16 CUs. + unsigned int intra_cu8x8_count; // number of intra 8x8 CUs. + unsigned int inter_cu8x8_count; // number of inter 8x8 CUs. + + + unsigned int total_emulated_bit_count; // total number of emulated bits in total_bit_count + unsigned char reserved[56]; +} nvenc_av1_mpec_stat_s; // 128 bytes per status + +typedef struct +{ + unsigned int var_min; + unsigned int var_max; + unsigned int var_avg; + unsigned char reserved[4]; +} nvenc_av1_aq_stat_s; // 16 bytes + +typedef struct +{ + unsigned int qp_val : 8 ; // qp used for encoding this block + unsigned int total_bit_count : 24 ; // total number of bits produced + unsigned int res_cost : 24 ; // satd cost of residual error + unsigned int flags : 8 ; // flags +} nvenc_av1_stats_fifo_s; // 8 bytes per stat + +typedef struct +{ + nvenc_av1_pic_stat_s pic_stat; // 64Bytes + nvenc_av1_tile_stat_s tile_stat[128]; // maximum tile number is 128 per AV1 spec + nvenc_av1_mpec_stat_s mpec_stat[((8192 / 64)*(8192 / 64)) ]; // To account for stat/1SB and 8kx8k, plus 3 for alignment? + nvenc_av1_aq_stat_s aq_stats[((8192 / 64)*(8192 / 64)) ]; // AQ stats from pdma + nvenc_av1_stats_fifo_s stats_fifo[((8192 / 64)*(8192 / 64)) ]; +} nvenc_av1_stat_data_s; + + +typedef struct +{ + unsigned int intra_res_bit_count; // #bits spent on intra residuals + unsigned int inter_res_bit_count; // #bits spent on inter residuals + unsigned int intra_mode_bit_count; // #bits spent on intra modes + unsigned int inter_mode_bit_count; // #bits spent on inter modes + unsigned int total_bit_count; // total number of bits produced + unsigned int total_bin_count; // total number of bins produced by bac + unsigned int cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int cumulative_skip_cost; // sum of all P skip MB SA(T)Ds + unsigned short intra_mb_count; // number of intra MBs. + unsigned short inter_mb_count; // number of inter MBs. + unsigned short skip_mb_count; // number of skip MBs. + unsigned short reserved0; // align to a word boundary +} nvenc_vp8_segment_stat_s; // 44 Bytes + +typedef struct +{ + unsigned int bitstream_byte_count; // output bytes in main partition bitstream buffer + unsigned int bitstream_res_byte_count[8]; // output bytes in residual partition bitstream buffers + unsigned int cycle_count; // total cycles taked for execute. Written out only when DumpCycleCount in SetControlParams is set + unsigned int picture_index; // value received from SetPictureIndex method + unsigned int error_status; // report error if any + unsigned int total_bit_count; // picture size in bits /*(sum of header's & all the partition's bits. actually redundant)*/ + unsigned short pic_type; // copied from pic_type in nvenc_h264_pic_control_s + unsigned short ave_activity; // report average activity if activity based RC enabled + unsigned short num_segments; // number of segments produced + unsigned short base_qindex; // base_qindex used for the frame + nvenc_vp8_segment_stat_s segment_stat[4]; + unsigned int total_intra_cost; // sum of all intra best mode's SATDs, each mb has one intra SATD. + unsigned int total_inter_cost; // sum of all inter best mode's SATDs, each mb has one inter SATD. + unsigned int reserved1[3]; +} nvenc_vp8_stat_data_s; // 256 Bytes + +// Stats stored in HW format. +typedef struct +{ + unsigned int coef_token_cnt[3][3][4][8][4]; // [value/4][ctx][value%4][band][type] 4608 bytes (4096 + 512) + unsigned int mv_stats[20][4]; +} nvenc_vp8_token_data_s; + + // 8 bytes + + +////////////////////////////////////////////////////////////////// +// **************************************************************************** +// **************************************************************************** +// AV1 +// **************************************************************************** +// **************************************************************************** +//AV1 loop filter config +typedef struct +{ + unsigned char filter_level[4]; // loop filter strength based on edge direction and component - vertical/horizontal edge luma (0,1) and chroma (2,3) + char ref_delta[8]; // Loop filter level adjustments based on reference frame type - 0 = Intra, Last, Last2+Last3, GF, BRF, ARF2, AR + int mode_delta_0 : 8; // Loop filter level adjustments based on mode type - 0 = ZERO_MV, MV + int mode_delta_1 : 8; // Loop filter level adjustments based on mode type - 0 = ZERO_MV, MV + unsigned int update_ref_delta_flag : 8; // update delta for reference i if bit i is set + unsigned int update_mode_delta0 : 1; + unsigned int update_mode_delta1 : 1; + unsigned int delta_enabled : 1; // disable loop-filter + unsigned int sharpness : 3; // sharpness level + unsigned int reserved : 2; +}nvenc_av1_loop_filter_cfg_s; //16 bytes + +//AV1 CDEF config +// NOTE: the following is mandated by the current HW implementation: +// y_strength_0 and uv_strength_0 must always be set to 0 +// Luma/chroma strength combination 0 must always be (y_strength_0, uv_strength_0) i.e. (0,0) +typedef struct +{ + unsigned int cdef_damping_minus_3 : 2; // controls the amount of damping in the deringing filter + unsigned int y_strength_0 : 6; // specify (strength primary filter 0 << 2) | strength secondary filter 0 (luma) + unsigned int y_strength_1 : 6; // specify (strength primary filter 1 << 2) | strength secondary filter 1 (luma) + unsigned int y_strength_2 : 6; // specify (strength primary filter 2 << 2) | strength secondary filter 2 (luma) + unsigned int y_strength_3 : 6; // specify (strength primary filter 3 << 2) | strength secondary filter 3 (luma) + unsigned int y_strength_4 : 6; // specify (strength primary filter 4 << 2) | strength secondary filter 4 (luma). Not supported. + unsigned int y_strength_5 : 6; // specify (strength primary filter 5 << 2) | strength secondary filter 5 (luma). Not supported. + unsigned int y_strength_6 : 6; // specify (strength primary filter 6 << 2) | strength secondary filter 6 (luma). Not supported. + unsigned int y_strength_7 : 6; // specify (strength primary filter 7 << 2) | strength secondary filter 7 (luma). Not supported. + unsigned int cdef_bits : 2; // support up to 3. the number of bits needed to specify w2ich CDEF filter to apply + unsigned int uv_strength_0 : 6; // specify (strength primary filter 0 << 2) | strength secondary filter 0 (chroma) + unsigned int uv_strength_1 : 6; // specify (strength primary filter 1 << 2) | strength secondary filter 1 (chroma) + unsigned int uv_strength_2 : 6; // specify (strength primary filter 2 << 2) | strength secondary filter 2 (chroma) + unsigned int uv_strength_3 : 6; // specify (strength primary filter 3 << 2) | strength secondary filter 3 (chroma) + unsigned int uv_strength_4 : 6; // specify (strength primary filter 4 << 2) | strength secondary filter 4 (chroma). Not supported. + unsigned int uv_strength_5 : 6; // specify (strength primary filter 5 << 2) | strength secondary filter 5 (chroma). Not supported. + unsigned int uv_strength_6 : 6; // specify (strength primary filter 6 << 2) | strength secondary filter 6 (chroma). Not supported. + unsigned int reserved0 : 2; // Reserved bits + unsigned int uv_strength_7 : 6; // specify (strength primary filter 7 << 2) | strength secondary filter 7 (chroma). Not supported. + unsigned int cdef_y_strength_num_minus_1 : 3; // HW will only considers y strength values from index 0 to cdef_y_strength_num_minus_1 (from 0 to 7) + unsigned int cdef_uv_strength_num_minus_1 : 3; // HW will only considers uv strength values from index 0 to cdef_uv_strength_num_minus_1 (from 0 to 7) + unsigned int cdef_combined_idx_minus_1 : 3; // Specify max number of unique (y_strength_x, uv_strength_y) combinations for HW to consider (from 0 to 7) + // bits[2:0] must be set to 0 + unsigned int reserved1 : 17; // Reserved bits + unsigned int cdef_combined_luma_idx : 24; // List of (cdef_combined_idx_minus_1 + 1) y strength indexes, each index using 3 bits and bits[3(n+1)-1:3n] corresponding to index of y_strength in combination n + unsigned int reserved2 : 8; // Reserved bits + unsigned int cdef_combined_chroma_idx : 24; // List of (cdef_combined_idx_minus_1 + 1) uv strength indexes, each index using 3 bits and bits[3(n+1)-1:3n] corresponding to index of uv_strength in combination n + // bits[2:0] must be set to 0 + unsigned int reserved3 : 8; // Reserved bits +}nvenc_av1_cdef_cfg_s; //24 bytes + +//AV1 Loop Restoration Filter config +typedef struct +{ + unsigned int y_lr_type : 2; // luma restoration type + unsigned int uv_lr_type : 2; // chroma restoration type + unsigned int y_lr_unit_size : 2; // luma restoration unit size 0:32x32, 1:64x64, 2:128x128,3:256x256; + unsigned int uv_lr_unit_size : 2; // chroma restoration unit size 0:32x32, 1:64x64, 2:128x128,3:256x256; + unsigned int lr_codebook_enable : 16; // bit enable for up to 16 codebooks + unsigned int lr_pe_on_edge_enable : 1; // enable bit for LR parameter estimation on picture boundary SBs + unsigned int reserved : 7; // Reserved bits +}nvenc_av1_lr_filter_cfg_s; // 4 bytes + +//AV1 quant config +typedef struct +{ + unsigned char base_q_idx; // Base frame qindex (index in quant lookup table) 0...255 + char delta_y_dc; // explicit qindex adjustment for y dccoefficient, -63...63 + char delta_uv_dc; // qindex adjustment for uv accoefficient, -63...63 + char delta_uv_ac; // qindex adjustment for uv dccoefficient, -63...63 +}nvenc_av1_quant_cfg_s; //4 bytes + +typedef struct +{ + unsigned short frame_width_minus1; // frame width in pixels minus 1, range 191-8191 + unsigned short frame_height_minus1; // frame height in pixels minus 1, range 127-8191 + unsigned short sfc_pitch; // pitch of luma plane + unsigned short sfc_pitch_chroma; // pitch of chroma plane + unsigned short sfc_trans_mode; // least 3 significient bits are used to stand for 8 modes, normal, xflip, yflip etc. + unsigned short reserved2; // pad to NvS32 + + unsigned int luma_top_frm_offset; // offset of luma top field or frame in units of 256 + unsigned int luma_bot_offset; // (deprecated) offset of luma bottom field in units of 256. Not used if frame format. + unsigned int chroma_top_frm_offset; // offset of chroma top field or frame, or offset of chroma U(Cb) plane in planar mode, both in units of 256 + unsigned int chroma_bot_offset; // (deprecated) offset of chroma bottom field, not used if frame format. Or offset of chroma V(Cr) plane in planar mode. Both in units of 256. + unsigned int block_height : 7; // BL mapping block height setting + unsigned int tiled_16x16 : 1; // (deprecated) Surface is 16x16 tiled instead of BL mapped (must be 1 for refpics) (deprecated) + unsigned int memory_mode : 2; // (deprecated) 0: semi-planar, 1: planar + unsigned int nv21_enable : 1; // (deprecated) the surface format is yuv or yvu + unsigned int input_bl_mode : 2; // the input block linear mode: 0~gpu bl; 1~tegra bl; 2~naive bl this is only for cmod and cmod/plugin. TBL only for Orin. + unsigned int src_in_lsb : 1; // pixel is LSB aligned in 16bit container, only valid for high-bitdepth + unsigned int reserved : 18; // pad to NvU32 +} nvenc_av1_surface_cfg_s; // 32 bytes + +typedef struct +{ + //sequence header + unsigned int min_cu_size : 2; //0-8x8, 1-16x16, 2-32x32, 3-64x64. Orin AV1 supports 1~3 + unsigned int max_cu_size : 2; //0-8x8, 1-16x16, 2-32x32, 3-64x64 Orin AV1 supports 1~3 + + unsigned int chroma_format_idc : 2; // 0=not supported, 1=yuv4:2:0, 2=yuv4:2:2, 3=yuv4:4:4. Only support YUV420 + unsigned int monochrome_flag : 1; // only luma component if flag is set. Not supported. + unsigned int bit_depth_minus_8 : 4; // luma/chroma should have the same value. support 0 (8bit), 2(10bit), 4(12bit) for Orin AV1 + unsigned int input_bit_depth_minus8 : 4; // Internal high-bitdepth support in HW + // the field is to specify the bit-depth of source yuv. the decoded yuv bitdepth is specified nvenc_av1_seq_hdr_data_s::bit_depth_minus_8; HW will do the bitdepth conversion from source yuv-> recon yuv + // + unsigned int enable_filter_intra : 1; // enables/disables filterintra. Not supported + unsigned int enable_intra_edge_filter : 1; // enables/disables corner/edge/upsampling. Not supported + unsigned int enable_interintra_compound : 1; // enables/disables interintra_compound. Not supported + unsigned int enable_masked_compound : 1; // enables/disables masked compound. Not supported + unsigned int enable_dual_filter : 1; // enable or disable vertical and horiz filter selection. Not supported + unsigned int force_screen_content_tools : 2; // 0 - force off. Not supported. + // 1 - force on + // 2 - adaptive + unsigned int force_integer_mv : 2; // Not supported.0 - Not to force. MV can be in 1/4 or 1/8 + // 1 - force to integer + // 2 - adaptive + unsigned int enable_jnt_comp : 1; // enables/disables joint compound modes. not supported. Not supported. + unsigned int enable_ref_frame_mvs : 1; // enables/disables ref frame mvs. not supported. Not supported. + unsigned int enable_warped_motion : 1; // enables/disables warped motion for sequence. Not supported. + unsigned int enable_superres : 1; // enables/disables superres for sequence. Not supported. + unsigned int enable_cdef : 1; // enables/disables CDEF + unsigned int enable_restoration : 1; // enables/disables restoration filter, only support self-guided filter + unsigned int reserved0 : 2; // reserved bits +} nvenc_av1_seq_hdr_data_s; // 4 bytes + +//frame header +typedef struct +{ + unsigned int frame_type : 2; // 0:Key frame, 1:Inter frame, 2:intra only, 3:s-frame + unsigned int disable_cdf_update : 1; // disable CDF update during symbol decoding + unsigned int allow_screen_content_tools : 1; // screen content tool enable. Not supported. + unsigned int force_integer_mv : 1; // AMVR enable. Not supported. + unsigned int allow_high_precision_mv : 1; // 1/8 precision mv enabl. Not supported. + unsigned int switchable_motion_mode : 1; // 0: simple motion mode, 1: SIMPLE, OBMC, LOCAL WARP. Not supported, fixed to 0. + unsigned int is_filter_switchable : 1; // 1: filter selection is signaled at block level + unsigned int use_ref_frame_mvs : 1; // Not supported. enable/disable use of motion vector info from previous frame + unsigned int frame_end_update_cdf : 1; // enable/disable CDF update at end of frame + unsigned int interp_filter : 3; // interpolation filter : EIGHTTAP,...., SWITCHABLE. Only support EIGHTTAP &EIGHTTAP_SMOOTH. + unsigned int delta_q_present_flag : 1; // quantizer index delta values are present in the block level + unsigned int delta_q_res : 2; // left shift will apply to decoded quantizer index delta values. Not supported. + unsigned int allow_intrabc : 1; // IBC enable. Not supported. + unsigned int delta_lf_present_flag : 1; // specify whether loop filter delta values are present in the block level. Not supported. + unsigned int delta_lf_res : 2; // specify the left shift will apply to decoded loop filter values. Not supported. + unsigned int delta_lf_multi : 1; // seperate loop filter deltas for Hy,Vy,U,V edges. Not supported. + unsigned int using_qmatrix : 1; // enable quant matrix. Not supported. + unsigned int frame_tx_mode : 2; // 0:LARGEST (32x32), 1:SELECT, 2: ONLY4x4 (lossless only) + unsigned int frame_reference_mode : 2; // 0: single, 1: select, 2: compound + unsigned int skip_mode_flag : 1; // skip mode present flag. Not supported. + unsigned int allow_warped_motion : 1; // frame level warped motion enable. Not supported. + unsigned int reduced_tx_set_flag : 1; // whether the frame is restricted to a reduced subset of the full set of transform types. Not supported. + unsigned int coded_lossless : 1; // whether lossless coding is used for frame or not + unsigned int reserved0 : 2; // reserved bits + + unsigned int ssim_rdo_enable : 1; //enable variance based SSIM-RDO + unsigned int ssim_rdo_alt_frame_scale : 8; //frame level scaling factor for altref frame + unsigned int bicompound_enable : 1; //enable bi-compound for B-frames. + unsigned int unicompound_enable : 1; //enable uni-compound for P/B-frames. (only support in Pframes) + unsigned int mc_hfilter : 4; //bitmask for horizontal filters enabled 0: EIGHTTAP, 1: EIGHTTAP_SMOOTH, + unsigned int mc_vfilter : 4; //bitmask for vertical filters enabled 0: EIGHTTAP, 1: EIGHTTAP_SMOOTH, + + unsigned int segmentation_enable : 1; //enable segmentation + unsigned int segmentation_update_map : 1; //enable segmentation update map ,force 1 + unsigned int segment_temporal_update : 1; //enable segmentation temporal update,force 0 + unsigned int segIdPreSkip : 1; //segment id pre skip encoding ,force 1 + unsigned int LastActiveSegId : 3; //highest number of active segment . force 2 + + unsigned int reserved1 : 6; + + + + unsigned char ref_frame_sign_bias[8]; //Used by AV1 MVP. A sign bias equal to 0 indicates that the reference frame is a forwards reference ( + nvenc_av1_loop_filter_cfg_s lf_cfg; // loop filter configuration + nvenc_av1_cdef_cfg_s cdef_cfg; // CDEF configuration + nvenc_av1_quant_cfg_s quant_cfg; // quantizer configuration + nvenc_av1_lr_filter_cfg_s lrf_cfg; // loop restoration filter configuration +} nvenc_av1_frame_hdr_data_s; // 64 bytes + +// needed only if picture level RC is enabled; as part of picture level RC, it will also perform HRD verification +// The contents of this file is produced by the driver during sequence level operations and should not be modified. +typedef struct +{ + unsigned char hrd_type; // nal (=2) and vcl (=1) type + unsigned char QP[3]; // QP for 0:P picture, ,1:B picture 2; I picture + unsigned char minQP[3]; // min QP for 0:P picture, ,1:B picture 2; I picture + unsigned char maxQP[3]; // max QP for 0:P picture, ,1:B picture 2; I picture + char maxQPD; // max QP delta between two consecutive QP updates + char baseQPD; // initial QP delta between two consecutive QP updates + unsigned int rhopbi[3]; // 23.8 unsigned fixed point quant ratios P/I, B/I, I/I(=1) + + unsigned int framerate; // fps + unsigned int buffersize; // total buffer size + + unsigned int nal_cpb_size; // size in bits + unsigned int nal_bitrate; // rate in bps + unsigned int vcl_cpb_size; // size in bits + unsigned int vcl_bitrate; // rate in bps + + unsigned int gop_length; // I period, gop_length == 0xffffffff is used for infinite gop length + unsigned int Np; // 27.4 unsigned fixed (gop_length + num_b_frames -1)/(num_b_frames + 1) + unsigned int Bmin; // 23.8 unsigned fixed point min buffer level, updated by driver at sequence level only. + unsigned int Ravg; // 23.8 unsigned fixed point average rate, updated by driver at sequence level only. + unsigned int R; // 23.8 unsigned fixed point rate + + unsigned char ab_alpha; // VBR/CBR: min/target quality level (1..51); rcmode4: weight of prev frame activity compared to part of the current picture + unsigned char ab_beta; // lookahead info1: ratio of inter over intra complexity (fix8); rcmode4: weight of current activity compared to neighbors + unsigned short prev_act; // lookahead info2: NVENC_LOOKAHEAD_INFO2; rcmode4: average activity level of the first reference (L0) + + unsigned char reserved0; + unsigned char single_frame_vbv; // 1: VBV buffer size is set to average frame size; 0: otherwise + unsigned char two_pass_rc; // 0: single pass rc , 1: first pass of 2 pass rc 2: second pass of 2 pass rc + unsigned char rc_class; // reserved + unsigned char first_pass_source_half_scaled; // 0 : first pass on half resolution 1: first pass on full resolution (along with NVENC_QRES_FLAG_XXX) + unsigned char i_size_ratio_x; // ratio between I picture target size over average picture size numerator + unsigned char i_size_ratio_y; // ratio between I picture target size over average picture size denominator + unsigned char aqMode; // 0: disable adaptive quantization 1: AQ mode fast (unimplemented) 2: AQ mode full (unimplemented) 3: External qp delta map + unsigned char dump_aq_stats; // 1: dump aq stats + unsigned char ext_scene_change_flag; // Scene change flag set as hint by external preprocessing unit. 0: No scene change, 1: Current picture is first in scene + unsigned char ext_picture_rc_hints; // picture rc hints are set by external source. 0: no hints, 1: hints are avaialble for current picture + unsigned char session_max_qp; // Current frame qp will never exceed this value when ext_picture_rc_hints is set + unsigned int ext_intra_satd; // If non zero this represents the intra SATD for current picture computed by external preprocessing unit + unsigned char reserved2[8]; +} nvenc_av1_rc_s; // 96 bytes +////////////////////////////////////////////////////////////////// + + +typedef struct +{ + char l0[4]; // reference picture list 0, maximum 4 forward reference frame + char l1[3]; // reference picture list 1, maximum 3 backward reference frame + char temp_dist_l0[4]; // temporal distance of each ref pic in list 0 DiffPicOrderCnt( currPicOrField, pic0) ) + char temp_dist_l1[3]; // temporal distance of each ref pic in list 1 DiffPicOrderCnt( currPicOrField, pic1) ) + char poc_entry_l0[4]; // ref pic list to ref pic entry map + char poc_entry_l1[3]; // ref pic list to ref pic entry map + char reserved0[3]; + + unsigned int max_byte_count_before_resid_zero; // Maximum byte count before microcode forces the residuals to zero. + + unsigned int mode_decision_only_flag : 1; // only output av1 mode and mocomp result, no recon and bitstream output. Not supported. + unsigned int cu_early_terminatiom_enable : 1; // 0: disable CU early termination, 1: enable CU early termination. Not supported. + unsigned int pic_output_flag : 1; // enable recon picture output + unsigned int input_sub_sampling : 1; // used by PDMA 2x2 sub-sampling. Not supported. + unsigned int output_sub_sampling : 1; // used by MPEB 2x2 sub-sampling output. Not supported. + unsigned int num_ref_idx_l0_active : 3; // [0..4] + unsigned int num_ref_idx_l1_active : 3; // [0..3] + unsigned int tile_idx_start : 8; // index of entry in av1_tile_control_s corresponding to first tile in strip + unsigned int num_strip_tiles_minus1 : 8; // number of tiles in strip (= num_tiles_minus1 if single strip) + unsigned int force_lf_on_flag : 1; // minimum level of loop filter is set to 1 instead of 0 when flag is set + unsigned int force_lf_off_flag : 1; // maximum level of loop filter is set to 0 when flag is set (i.e. filtering off) + unsigned int vp9_skip_chroma : 1; // HW will only process luma and skip chroma. used for vp9 autoprofile TODO: temporary add here, need to move to VP9 dedicated structrue. + unsigned int reserved1 : 2; + + unsigned int max_tu_depth_inter : 2; // max_transform_hierarchy_depth_inter. Orin AV1 supports 0~2 + unsigned int max_tu_depth_intra : 2; // max_transform_hierarchy_depth_intra. Orin AV1 supports 0 + unsigned int num_tiles_minus1 : 8; // number of entries in av1_tile_control_s + unsigned int enable_source_image_padding : 1; // enable/disable HW source image padding + unsigned int luma_padding_mode : 2; // HW souce padding mode for luma: 0: copy frame boundary, 1: zero padding, 2: half value, i.e 1<<(bitdepth-1) + unsigned int chroma_padding_mode : 2; // HW souce padding mode for chroma: 0: copy frame boundary, 1: zero padding, 2: half value, i.e 1<<(bitdepth-1) + unsigned int reserved2 : 15; + unsigned int prob_tables_offset; // When non-zero, ucode will adjust the base address of the probability tables surfaces based on picture QP value as follows + unsigned int tile_control_offset; // offset from start of top level "nvenc_av1_drv_pic_setup_s" structure + unsigned int hist_buf_size; // size in bytes of the buffer allocated for history + unsigned int bitstream_buf_size; // size in bytes of the buffer allocated for bitstream slice/mb data + unsigned int bitstream_start_pos; //start position in bitstream buffer where data should be written (byte offset) + unsigned int pic_type : 2; // 0 = P, 1 = B, 2 = I, 3 = IDR + unsigned int ref_pic_flag : 1; // reference picture (0 = no, 1 = yes) + unsigned int codec : 3; + unsigned int qpfifo : 1; // 0 = Polling mode, 1 = Interrupt mode. + unsigned int num_tile_cols : 8; // Number of tile columns in picture + unsigned int num_tile_rows : 8; // Number of tile rows in picture + unsigned int ctx_update_tile_id : 8; // update cdf for the specified tile id. value 255 is reserved to enable HW auto update CDF from tile with maximun tile data length + unsigned int stats_fifo_enable : 1; // enable per-sb short stats fifo + + // 1 = static slice insertion based on slice control array + // (insert slices as defined in array, if all array elements used + // before end of picture is reached, start over with entry 0) + + unsigned int mpec_threshold; // collect mpec stats after threshold mbs + // when mpec_stat_on is enabled and mpec_threshold is not equal to picWidthInMbs, + unsigned int tile_stat_offset; // offset from start of top level "nvenc_av1_stat_data_s" structure + // rc mode 0 will be used to collect stats + unsigned int mpec_stat_offset; // offset from start of top level "nvenc_stat_data_s" structure + // to start of array of "nvenc_mpec_stat_s" structure + unsigned int aq_stat_offset; // to start of array of "nvenc_av1_aq_stat_s" structure + + unsigned int rc_pass2_offset; + + unsigned int act_stat_offset; // offset from start of "nvenc_rc_pic_stat_s" structure, points to data for external QP map + unsigned int external_hint_number; // total external hint number for current frame, maximun external hints per SB: 128. + + unsigned char stripID; // strip index (from 0 to strips_in_frame-1) + unsigned char strips_in_frame; // total number of strips in frame + unsigned short strip_width_in_sbs; // strip width in SB unit,must be frame_width + unsigned short strip_height_in_sbs; // strip height in SB unit + unsigned short strip_width_offset_in_sbs; // Column index of first SB in strip ,must be 0 + unsigned short strip_height_offset_in_sbs; // Row index of first SB in strip + unsigned short reserved5; + + unsigned int stats_fifo_offset; // offset from start of top level "nvenc_stat_data_s" structure + unsigned char reserved6[8]; +} nvenc_av1_pic_control_s; // 112 bytes + +typedef struct +{ + unsigned int cu8x8_part_2Nx2N_enable : 1; // enable prediction for Inter CU8x8 2Nx2N + unsigned int cu16x16_part_2Nx2N_enable : 1; // enable prediction for Inter CU16x16 2Nx2N + unsigned int cu16x16_part_2NxN_enable : 1; // enable prediction for Inter CU16x16 2NxN + unsigned int cu16x16_part_Nx2N_enable : 1; // enable prediction for Inter CU16x16 Nx2N + unsigned int cu32x32_part_2Nx2N_enable : 1; // enable prediction for Inter CU32x32 2Nx2N + unsigned int cu32x32_part_2NxN_enable : 1; // enable prediction for Inter CU32x32 2NxN + unsigned int cu32x32_part_Nx2N_enable : 1; // enable prediction for Inter CU32x32 Nx2N + unsigned int cu64x64_part_2Nx2N_enable : 1; // enable prediction for Inter CU64x64 2Nx2N + unsigned int cu64x64_part_2NxN_enable : 1; // enable prediction for Inter CU64x64 2NxN + unsigned int cu64x64_part_Nx2N_enable : 1; // enable prediction for Inter CU64x64 Nx2N + unsigned int tusize_4x4_enable : 1; //enable 4x4 transform + unsigned int tusize_8x8_enable : 1; //enable 8x8/8x4/4x8 transform + unsigned int tusize_16x16_enable : 1; //enable 16x16/16x8/8x16 transform + unsigned int tusize_32x32_enable : 1; //enable 32x32/32x16/16x32 transform + unsigned int tusize_64x64_enable : 1; //enable 64x64/64x32/32x16 transform + unsigned int intra_tusize_decision : 1; //enable intra tu size decision, reserved, not implemented in HW + unsigned int pskip_enable : 1; //enable p_skip encoding. Not supported + unsigned int tu_search_basedon_pusize : 1; //enable tu search based on pu size. Not supported + unsigned int multiply_bias_with_lambda : 1; // setting this to 1 will result in bias being multiplied by lambda + unsigned int skip_evaluate_enable_cu8 : 1; // 0: not check skip ssd cost; 1, evaluate skip ssd cost. Not supported + unsigned int skip_evaluate_enable_cu16 : 1; // 0: not check skip ssd cost; 1, evaluate skip ssd cost. Not supported + unsigned int skip_evaluate_enable_cu32 : 1; // 0: not check skip ssd cost; 1, evaluate skip ssd cost. Not supported + unsigned int skip_evaluate_enable_cu64 : 1; // 0: not check skip ssd cost; 1, evaluate skip ssd cost. Not supported + unsigned int inter_rdo_cnt_8x8 : 3; // #RD cost evaluation for inter top N SATD winner partitions of CU8x8. support 0~4 + unsigned int inter_rdo_cnt_16x16 : 3; // #RD cost evaluation for inter top N SATD winner partitions of CU16x16. support 0~4 + unsigned int inter_rdo_cnt_32x32 : 3; // #RD cost evaluation for inter top N SATD winner partitions of CU32x32. support 0~4 + + unsigned int inter_rdo_cnt_64x64 : 3; // #RD cost evaluation for inter top N SATD winner partitions of Cu64x64. support 0~4 + unsigned int intra_rdo_cnt_4x4 : 3; // RD cost evaluation for N top satd winner modes of intra 4x4. Not supported + unsigned int intra_rdo_cnt_8x8 : 3; // RD cost evaluation for N top satd winner modes of intra 8x8. support 0~4 + unsigned int intra_rdo_cnt_8x4 : 3; // RD cost evaluation for N top satd winner modes of intra 8x4. support 0~4 + unsigned int intra_rdo_cnt_4x8 : 3; // RD cost evaluation for N top satd winner modes of intra 4x8. support 0~4 + unsigned int intra_rdo_cnt_16x16 : 3; // RD cost evaluation for N top satd winner modes of intra 16x16. support 0~4 + unsigned int intra_rdo_cnt_16x8 : 3; // RD cost evaluation for N top satd winner modes of intra 16x8. support 0~4 + unsigned int intra_rdo_cnt_8x16 : 3; // RD cost evaluation for N top satd winner modes of intra 8x16. support 0~4 + unsigned int intra_rdo_cnt_32x32 : 3; // RD cost evaluation for N top satd winner modes of intra 32x32. support 0~4 + unsigned int intra_rdo_cnt_32x16 : 3; // RD cost evaluation for N top satd winner modes of intra 32x16. support 0~4 + unsigned int luma_all_zero_check : 1; // enable TU level luma coefficient all zero check + unsigned int chroma_all_zero_check : 1; // enable TU level chroma coefficient all zero check. Not supported + + unsigned int intra_rdo_cnt_16x32 : 3; // RD cost evaluation for N top satd winner modes of intra 16x32. support 0~4 + unsigned int intra_rdo_cnt_64x64 : 3; // RD cost evaluation for N top satd winner modes of intra 64x64. support 0~4 + unsigned int intra_rdo_cnt_64x32 : 3; // RD cost evaluation for N top satd winner modes of intra 64x32. support 0~4 + unsigned int intra_rdo_cnt_32x64 : 3; // RD cost evaluation for N top satd winner modes of intra 32x64. support 0~4 + unsigned int num_1div8_lambda_ssd : 7; // number of 1/8 lambda ssd for lambda_coef delta. lambda_coef = (1+num_1div8_lambda_ssd*1/8)*lambda_ssd + unsigned int mv_cost_enable : 1; // enable mv cost calculations in mode decision + unsigned int reserved0 : 12; + + //Intra Luma angular delta mode controls: begin + unsigned int intra_luma_dir_delta_64x64_enable_0; // bit mask for enabling angular delta modes: {D135, D45, H, V} * {-3,-2,-1,0,1,2,3} + unsigned int intra_luma_dir_delta_64x64_enable_1; // bit mask for enabling angular delta modes: {D67, D203, D157, D113} * {-3,-2,-1,0,1,2,3} + unsigned int intra_luma_dir_delta_64x32_enable_0; + unsigned int intra_luma_dir_delta_64x32_enable_1; // for 64x32 and 32x64 + + unsigned int intra_luma_dir_delta_32x32_enable_0; // bit mask for enabling angular delta modes: {D135, D45, H, V} * {-3,-2,-1,0,1,2,3} + unsigned int intra_luma_dir_delta_32x32_enable_1; // bit mask for enabling angular delta modes: {D67, D203, D157, D113} * {-3,-2,-1,0,1,2,3} + unsigned int intra_luma_dir_delta_32x16_enable_0; // for 32x16 and 16x32 + unsigned int intra_luma_dir_delta_32x16_enable_1; + + unsigned int intra_luma_dir_delta_16x16_enable_0; // bit mask for enabling angular delta modes: {D135, D45, H, V} * {-3,-2,-1,0,1,2,3} + unsigned int intra_luma_dir_delta_16x16_enable_1; // bit mask for enabling angular delta modes: {D67, D203, D157, D113} * {-3,-2,-1,0,1,2,3} + unsigned int intra_luma_dir_delta_16x8_enable_0; // for 8x16 and 16x8 + unsigned int intra_luma_dir_delta_16x8_enable_1; + + unsigned int intra_luma_dir_delta_8x8_enable_0; // bit mask for enabling angular delta modes: {D135, D45, H, V} * {-3,-2,-1,0,1,2,3} + unsigned int intra_luma_dir_delta_8x8_enable_1; // bit mask for enabling angular delta modes: {D67, D203, D157, D113} * {-3,-2,-1,0,1,2,3} + //Intra Luma angular delta mode controls: end + + // 0:DC_PRED, 1: V_PRED, 2: H_PRED, 3:D45_PRED, + // 4:D135_PRED, 5:D113_PRED, 6:D157_PRED, 7: D203_PRED, + // 8: D67_PRED, 9:SMOOTH_PRED, 10:SMOOTH_V_PRED, 11:SMOOTH_H_PRED + // 12: PAETH + unsigned short intra_luma_base_mode_4x4_enable; //bit mask of intra luma base 4x4 mode: + unsigned short intra_luma_base_mode_8x8_enable; //bit mask of intra luma base 8x8/8x4/4x8 mode + unsigned short intra_luma_base_mode_16x16_enable; //bit mask of intra luma base 16x16/16x8/8x16 modes + unsigned short intra_luma_base_mode_32x32_enable; //bit mask of intra luma base 32x32/32x16/16x32 modes + unsigned short intra_luma_base_mode_64x64_enable; //bit mask of intra luma base 64x64/32x64/64x32 modes + + short bias_intra_over_inter; // bias for any intra (subtracts from intra cost) + unsigned short mv_cost_bias; // bias applied to internal motion vector cost calculation + + short bias_tu_4x4; // bias for TU 4x4 + short bias_tu_8x8; // bias for TU 8x8 + short bias_tu_16x16; // bias for TU 16x16 + short bias_tu_32x32; // bias for TU 32x32 + short bias_tu_64x64; // bias for TU 64x64 + short bias_tu_8x4; // bias for TU 8x4 + short bias_tu_4x8; // bias for TU 4x8 + short bias_tu_16x8; // bias for TU 16x8 + short bias_tu_8x16; // bias for TU 8x16 + short bias_tu_32x16; // bias for TU 32x16 + short bias_tu_16x32; // bias for TU 16x32 + + short bias_skip; // bias for skip mv's ssd (subtracts from skip mv's ssd cost). Not supported. + short bias_inter_64x64; // bias for inter 64x64 (subtracts from inter 64x64 cost) + short bias_inter_64x32; // bias for inter 64x32 (subtracts from inter 64x32 cost) + short bias_inter_32x64; // bias for inter 32x64 (subtracts from inter 32x64 cost) + short bias_inter_32x32; // bias for inter 32x32 (subtracts from inter 32x32 cost) + short bias_inter_32x16; // bias for inter 32x16 (subtracts from inter 32x16 cost) + short bias_inter_16x32; // bias for inter 16x32 (subtracts from inter 16x32 cost) + short bias_inter_16x16; // bias for inter 16x16 (subtracts from inter 16x16 cost) + short bias_inter_16x8; // bias for inter 16x8 (subtracts from inter 16x8 cost) + short bias_inter_8x16; // bias for inter 8x16 (subtracts from inter 8x16 cost) + short bias_inter_8x8; // bias for inter 8x8 (subtracts from inter 8x8 cost) + short bias_intra_64x64; // bias for intra 64x64 (subtracts from intra 64x64 cost) + short bias_intra_64x32; // bias for intra 64x32 (subtracts from intra 64x32 cost) + short bias_intra_32x64; // bias for intra 32x64 (subtracts from intra 32x64 cost) + short bias_intra_32x32; // bias for intra 32x32 (subtracts from intra 32x32 cost) + short bias_intra_32x16; // bias for intra 32x16 (subtracts from intra 32x16 cost) + short bias_intra_16x32; // bias for intra 16x32 (subtracts from intra 16x32 cost) + short bias_intra_16x16; // bias for intra 16x16 (subtracts from intra 16x16 cost) + short bias_intra_16x8; // bias for intra 16x8 (subtracts from intra 16x8 cost) + short bias_intra_8x16; // bias for intra 8x16 (subtracts from intra 8x16 cost) + short bias_intra_8x8; // bias for intra 8x8 (subtracts from intra 8x8 cost) + short bias_intra_8x4; // bias for intra 8x4 (subtracts from intra 8x4 cost) + short bias_intra_4x8; // bias for intra 4x8 (subtracts from intra 4x8 cost) + short bias_intra_4x4; // bias for intra 4x4 (subtracts from intra 4x4 cost) + + unsigned short tx_type_search_mask[2]; // bit mask of tx type enabled. bit 0~15, corresponding to: + // tx type DCT_DCT, ADST_DCT, DCT_ADST, ADST_ADST, FLIPADST_DCT, + // DCT_FLIPADST, FLIPADST_FLIPADST, ADST_FLIPADST, FLIPADST_ADST, + // IDTX, V_DCT, H_DCT, V_ADST, H_ADST, V_FLIPADST, H_FLIPADST, + // 0: control for RDO stage. 1: control for RECON stage + + unsigned int zbin_factor_lt_148 : 8; //Aggresive DZ control for zbin factor when quant value < 148. in 1/128 unit + unsigned int zbin_factor_gt_148 : 8; //Aggresive DZ control for zbin factor when quant value >= 148. in 1/128 unit + unsigned int round_factor_eq_0 : 8; // TBD + unsigned int round_factor_ne_0 : 8; // TBD +} nvenc_av1_md_control_s; // 160 bytes + +typedef nvenc_h264_me_control_s nvenc_me_control_s; + +typedef struct +{ + // 256 aligned + unsigned int magic; // magic number, see text above // 4 bytes + nvenc_av1_surface_cfg_s refpic_cfg; // defines layout of reference pictures // 32 bytes + nvenc_av1_surface_cfg_s input_cfg; // defines layout of input pictures // 32 bytes + nvenc_av1_surface_cfg_s outputpic_cfg; // defines layout of reconstructed pictures // 32 bytes + + nvenc_av1_seq_hdr_data_s seqhdr_data; // 4 bytes + nvenc_av1_frame_hdr_data_s framehdr_data; // 64 bytes + nvenc_av1_rc_s rate_control; // Rate Control information // 96 bytes + nvenc_av1_pic_control_s pic_control; // 112 bytes + nvenc_av1_surface_cfg_s half_scaled_outputpic_cfg; // defines layout of 2x2 subsampled reconstructed picture // 32 bytes + nvenc_av1_md_control_s md_control; // 160 bytes + nvenc_me_control_s me_control; // 192 bytes + unsigned int gpTimer_timeout_val; // GPTimer cycle count set from driver + unsigned char reserved0[4]; +} nvenc_av1_drv_pic_setup_s; //768 bytes + +typedef nvenc_av1_drv_pic_setup_s nvenc_vp9_drv_pic_setup_s; //TODO: Orin VP9 will share the same class as AV1. Define new one for Hopper VP9 + +typedef struct +{ + unsigned int tile_top_sbx : 16; // top left sbx of the tile + unsigned int tile_top_sby : 16; // top left sby of the tile + unsigned int tile_bot_sbx : 16; // bottom right sbx of the tile + unsigned int tile_bot_sby : 16; // bottom right sby of the tile + + unsigned int qp_tile_min : 8; // min slice qp value + unsigned int qp_tile_max : 8; // max slice qp value + unsigned int qp_avr : 8; // 8bit value used by hw + unsigned int tile_type : 2; // 0 = B, 1 = P, 2 = I + unsigned int force_intra : 1; + unsigned int tile_end_update_cdf : 1; // enable/disable CDF update at end of tile + unsigned int limit_tile_top_boundary : 1; // limit ME to tile boundaries + unsigned int limit_tile_bot_boundary : 1; // limit ME to tile boundaries + unsigned int limit_tile_left_boundary : 1; // limit ME to tile boundaries + unsigned int limit_tile_right_boundary : 1; // limit ME to tile boundaries + + unsigned int tile_idx : 8; // tile index at picture level (potentially different from index in array of nvenc_av1_tile_control_s when doing strip encoding) + unsigned int disable_tile_top_filters : 1; //disable filters at tile top boundary + unsigned int disable_tile_bottom_filters : 1; //disable filters at tile bottom boundary + unsigned int disable_next_tile_top_filters : 1; //disable filters at next tile top boundary + unsigned int disable_next_tile_bottom_filters : 1; //disable filters at next tile bottom boundary + unsigned int reserved0 : 20; + + unsigned int next_tile_top_sbx : 16; // top left sbx of next tile + unsigned int next_tile_top_sby : 16; // top left sby of next tile + unsigned int next_tile_bot_sbx : 16; // bottom right sbx of next tile + unsigned int next_tile_bot_sby : 16; // bottom right sby of next tile + + unsigned short intra_refresh_cnt; // distance between forced-intra SBs in slice; 0 means OFF + unsigned short intra_refresh_offs; // first forced-intra SB in slice + + unsigned char reserved2[4]; // padding to 32 bytes +} nvenc_av1_tile_control_s; // 32 bytes + +// AV1 not support QPP, reserved it but not used +typedef struct +{ + // QPP has not been defined for AV1 + unsigned short qpp_run_vector_4x4; // cost values for 4x4 transform (16 bit total vector) + unsigned short qpp_run_vector_8x8[3]; // cost values for 8x8 transform (48 bit total vector) + // 2 bits each for first 12 coefs, then 1 bit each for next 24 + unsigned int qpp_luma8x8_cost : 8; // luma 8x8 cost threshold, 0 = throw out all coefs, range 0-15 + unsigned int qpp_luma16x16_cost : 8; // luma 16x16 cost threshold, 0 = throw out all coefs, range 0-15 + unsigned int qpp_chroma_cost : 8; // chroma cost threshold, 0 = throw out all coefs, range 0-15 + unsigned int qpp_mode : 2; // 0 = OFF, 1 = 8x8, 2 = 16x16_8x8 + unsigned int reserved0 : 6; // padding + + unsigned int reserved1; // padding to 16 bytes +} nvenc_av1_quant_control_s; // 16 bytes + +// Status buffer holds for debug related information +// +// General guideline for using CRC interface debug feature: +// 1. Generate the golden CRC (e.g. using cmodel) for for all interfaces +// 2. program the golden crc to NV_CNVENC_SAFETY_CRC_GOLDEN_PART_[X]_INTERFACE_NAME, and config the SAFETY_CRC_CFG register (with values from dbg_crc_init_enable/comp_enable/addr_comp/mem_wr_dis), set SAFETY_CRC_CTRL_EN = 1 (TZ) +// 3. kick off HW and trigger HW to compare width golden by writing SAFETY_CRC_CFG_COMP_EN when finish processing the frame; compare result could read from NV_CNVENC_SAFETY_CRC_STATUS_PART_[X] +// 4. raw HW CRC values could be read from NV_CNVENC_SAFETY_CRC_HW_PART_[X]_INTERFACE_NAME +typedef struct +{ + unsigned int dbg_crc_enable : 1; //Global enable flag for enable/disable interface crc calculation in NVENC HW + unsigned int dbg_crc_init_enable: 1; //Enable crc initialization + unsigned int dbg_crc_comp_enable: 1; //Enable crc golden comparision + unsigned int dbg_crc_addr_comp : 1; //Enable crc calculation for fbif address + unsigned int dbg_crc_mem_wr_dis : 1; //Disable fb memory write + unsigned int reserved0 : 27; + + //Golden CRC values for each partition + unsigned int dbg_crc_partb_golden[32]; + unsigned int dbg_crc_partf_golden[32]; + unsigned int dbg_crc_partg_golden[32]; + unsigned int dbg_crc_parth_golden[32]; + unsigned int dbg_crc_parti_golden[32]; + unsigned int dbg_crc_partj_golden[32]; + unsigned int dbg_crc_partk_golden[32]; + unsigned int dbg_crc_partl_golden[35]; + + //Compare status register for each partition. + //each bit corresponding to the compare status for one interface (0: pass; 1: fail) + //The number of valid bits for each partition shall equal to the number of golden crcs for that partition + unsigned int dbg_crc_comp_partb; + unsigned int dbg_crc_comp_partf; + unsigned int dbg_crc_comp_partg; + unsigned int dbg_crc_comp_parth; + unsigned int dbg_crc_comp_parti; + unsigned int dbg_crc_comp_partj; + unsigned int dbg_crc_comp_partk; + unsigned int dbg_crc_comp_partl[2]; //split into two compare status registers + + + //HW CRC values generated from each partition. the number shall match with golden CRC number + unsigned int dbg_crc_partb_crc[32]; + unsigned int dbg_crc_partf_crc[32]; + unsigned int dbg_crc_partg_crc[32]; + unsigned int dbg_crc_parth_crc[32]; + unsigned int dbg_crc_parti_crc[32]; + unsigned int dbg_crc_partj_crc[32]; + unsigned int dbg_crc_partk_crc[32]; + unsigned int dbg_crc_partl_crc[35]; + + + //HW verification purpose control fields + unsigned int lambda_ssd; //lambda for SSD cost + unsigned int lambda_satd; //lambda for SATD cost + unsigned int lambda_ld; //lambda for laplacian cost + unsigned char reserved1[180]; +} nvenc_debug_status_s; //2304 bytes + +// 2.7. Bit Stream Output Data Buffer +// ---------------------------------- +// +// This buffer is used to store the encoded bit stream data at slice header +// level and below. +// The driver must ensure that there is enough free space in this buffer to +// contain the entire output produced for the current execute, including any +// headers that will be produced. Output will be written incrementally +// from the starting position relative to the 256 byte aligned start location +// specified for this buffer. +// +// 2.8. Motion Estimation Predictor Data Buffer +// -------------------------------------------- +// +// This buffer is used to store by the motion estimation unit to store vector +// hints that can be used as explicit candidates or as guides for motion +// search in future pictures. +// During encode one such surface can be written and another can be read. +// +// 2.9. Rate Control Process Buffer +// -------------------------------- +// +// This buffer is used to store persistent rate control information. The same +// buffer will be shared in all executes of a particular stream. +// +// This buffer is used only if picture level rate control is enabled ((rc_mode & 3) != 0) + +// Macro to initialize Bavg in VBR mode +#define NVENC_VBR_INIT_BAVG(framerate, ravg) (0) // zero-centered vbr fullness +// Macro to initialize Whist and Wpbi +#define NVENC_RC_INIT_COMPLEXITY(bpp_at_qs1) ((int)((bpp_at_qs1)*16)) // use this macro to decouple from implementation-dependent precision + +typedef struct +{ + + int nal_cpb_fullness; // 28.4 fixed point + int vcl_cpb_fullness; // 28.4 fixed point + + int Wpbi[3]; // 27.4 signed fixed point + int Whist[3][3]; // 27.4 signed fixed point [pbi][t] + int np; // 27.4 signed fixed point + int nb; // 27.4 signed fixed point + int t; // 27.4 signed fixed point + + int Rmean; // 23.8 signed fixed point peak bitrate (bits/pixel) + int Bavg; // 23.8 signed fixed point + + short average_mvx; // average mvx info, can be used for ME hint + short average_mvy; // average mvy info, can be used for ME hint + + int prevQS; // previous second pass Qs derived from rate + short prevQP; // previous 2nd pass qp + short fwdQP; // QP of forward reference + + unsigned char refQPcol[48]; // co-located mbrow qp of the most recently coded reference + unsigned char Rcol[48]; // co-located rate (5.3 precision, relative to avg rate) + unsigned char ref1QP[48]; // first-pass reference qp (when 1st pass is using its own reference frames) + + unsigned int ext_intra_satd_Hist[2]; // external satd History + unsigned char ext_scd_Hist; // scene change flag History + unsigned char scene_change; // last coded frame was a scene change + unsigned char qpp_modifier; // current qpp adjustment (subtracted from qpp thresholds) + unsigned char Ksatd; // pass1 satd to pass2 complexity measure + signed char Werr; // model deviation: Wpass2 / Wpass1 - 1.0, .8 signed fixed-point [0.5x..1.5x] + signed char WerrI; // same as above for I-frames + unsigned char avgBf; // 3.5 fixed point, average number of consecutive B-frames + unsigned char reserved; + unsigned int printf_offset; // Using for ucode print + unsigned char reserved2[4]; + +} nvenc_persistent_state_s; // 256Bytes + +// 2.a. Ucode Probability State +// -------------------------------- +// +// This buffer is used to store probability state computed and used during the last picture. +// This is used to compute the probability updates for current picture +// + +// DO NOT rearrange this structure. The alignment is crucial for dmreads +typedef struct +{ + unsigned char coef_probs[4][8][3][11+5]; // 1536 bytes = 4*(3*128) + // Padding cause this has to be read in parts + unsigned char ymode_prob[4]; // 4 bytes + unsigned char uvmode_prob[3]; // 3 bytes + unsigned char reserved0; // Pad to word boundary + unsigned char mv_context[2][19]; // 38 bytes + unsigned char reserved1[18]; // Align to 32 byte boundary + +}nvenc_vp8_prob_state_s; + + +// 2.b. Rate Control stat buffer used for 2 pass encoding +// ------------------------------------------------------ +// +// This buffer is used to store rate control information for 2 pass encoding. +// The same buffer will be shared in both the passes of the encoding. +// The first pass will write MPEC stats to the buffer and 2nd pass will read +// the stats from the same buffer for QP calculation. The layout of buffer is +// msenc_rc_pic_stat_s followed by msenc_mbrow_stat_s for each mb row. +// + +typedef struct +{ + unsigned int intra_mb_count; // number of intra MBs. + unsigned int inter_mb_count; // number of inter MBs. + unsigned int cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int total_bit_count; + unsigned int avgQP; + unsigned int complexity; + //split encoding + unsigned int strip_complexity; + unsigned int strip_intra_mb_count; + unsigned char reserved[220]; +}nvenc_rc_pic_stat_s; //256 bytes + +typedef struct +{ + unsigned int mbrow_intra_mb_count : 12; // number of intra MBs + unsigned int mbrow_inter_mb_count : 12; // number of inter MBs + unsigned int mbrow_avgQP : 8; // average QP + unsigned int mbrow_cumulative_intra_cost; // sum of all intra SA(T)Ds + unsigned int mbrow_cumulative_inter_cost; // sum of all inter SA(T)Ds + unsigned int mbrow_bits_per_row; +}nvenc_mbrow_stat_s; + +typedef struct +{ + signed char act_qp_buf[4096]; // Buffer to store qp modifier based on avg activity for a MB row +}nvenc_act_stat_s; + +// 2.b. Generic Map Buffer +// -------------------------------- +// +// This buffer is used to store qp map when rcmode==4 & lambda_map when enabled. the structure is in run/level, run==0 means end. +// +typedef struct +{ + unsigned short run; // number of blocks with the same value + unsigned short level; // value +} nvenc_generic_map_s; + +typedef nvenc_generic_map_s nvenc_qp_map_s; +typedef nvenc_generic_map_s nvenc_lambda_map_s; + + +// 2.d. Encrypt Params buffer +// -------------------------------- +// +// This buffer is used by SEC / TSEC engine which will encrypt the encoded stream +// MSENC will output the # of 16 byte blocks present in the encoded bitstream to noOfInputBlocks field. +// MSENC ucode does not use any other field besides noOfInputBlocks. +// These will be populated by driver / TSEC ucode +// Currently used only for MSENC 2.0 for T114 +typedef struct _hdcp_encrypt_param_s +{ + unsigned int sessionId; // <>out + unsigned int crcOffset; // >>out + unsigned long long inputCtr; // >>out + unsigned long long pesPriv[2]; // >>out + unsigned int retCode; // >>out + unsigned char reserved2[4]; +} hdcp_encrypt_param_s; + +// Error codes used across all codecs +typedef enum +{ + E_NONE = 0U, + E_GPTMR_TIMEOUT, + E_MAGIC_NUMBER_MISMATCH, + E_PICSETUP_OFFSET, + E_ENC_STATUS_OUTPUT_OFFSET, + E_OUTPUT_BITSTREAM_OFFSET, + E_INSUFFICIENT_WRDMA_BUF, + E_MDP_ERROR, + E_RDO_ERROR, + E_ME_PROGRAMMING, + E_ME_HW, + E_TFBIF_TIMEOUT, + E_SKIP_ENCODING, + E_INPUT_BITSTREAM_OFFSET, + E_INPUT_HISTORY_OFFSET, + E_MEPRED_OFFSET, + E_EXTHINT_OFFSET, + E_REFPIC_OFFSET, + E_QPMAP_OFFSET, + E_COLLOC_OFFSET, + E_PROB_OFFSET, + E_ENTROPY_OFFSET, + E_FILTER_OFFSET, + E_SSIM_RDO_OFFSET, + E_RC_ERROR, + E_SMALLE_BS_OVERFLOW, + E_SMALLE_METHODSERROR, + E_ENTROPY_ERR, + E_RECON_OVERFLOW, + TOTAL_ERR_CODES +} ERROR_CODES; + +// 3 - APPLICATION SPECIFIC METHOD REGISTERS +// +// The NVENC application method registers are accessed directly by the driver. +// The range assignment is 0x400 - 0x7FC: +// +// +// 3.1 - Common Methods +// -------------------- +// +// SetControlParams() +// ------------------ +// This method sets the control parameters that are used in this execution +// of the encode subclass. +// +// CODEC_TYPE field should be set as +// 3: H.264 // same codec enum as for msdec +// +// FORCE_OUTPIC indicates that the encoded picture (after deblocking if that is enabled) +// must be written out to the buffer specified by SetOutRefPic(), even if the +// current picture is not marked as a reference picture. +// +// FORCE_OUTCOL indicates that the colocated vector data must be written out to the +// buffer specified by SetOutColocData(), even if the current picture is +// not marked as a reference picture. +// +// GPTIMER_ON, this flag is used in activating an app based timer with a timeout value +// less than watchdog timer. When this app level timer expires, various hardware +// submodules in NVENC will be reset, the status will be informed in flowControl +// buffer if DEBUG_MODE is OFF (0) and app will return. +// +// ME_ONLY, indicates whether to run in normal mode of operation or activate only +// motion estimation and write out search vector results +// 0 = encode bitstream +// 1 = only ME, do not run mode decision/recon/entropy encode +// +// SLICE_STAT_ON indicates if slice level statistics dumping is enabled. If enabled, +// following the persistent state information, slice level statistics will be +// output in the Encoder Status buffer. It is assumed that the number of slices +// will be limited to picture height in units of MBs. If there are more slices in the +// picture, driver should turn off slice stat output, by resetting this flag to 0. +// +// MPEC_STAT_ON indicates if nMB level mpec statistics dumping is enabled. If SLICE_STAT_ON +// is enabled, MPEC_STAT_ON will be automatically get disabled and all the slice level +// mpec statistics will be dumped to nvenc_mpec_stat_s. +// +// DEBUG_MODE indicates whether the app will return non-zero error codes to OS in case +// of an error detected by the app. +// +// If the value is set to 1, upon detecting an error the app will NOT do a +// soft reset of hw, the app will return appropriate error code to OS. This +// shall be used during debugging phase. +// +// If the value is set to 0, upon detecting an error the app will do a soft +// reset of hw and return with zero to OS. This shall be used for product. +// +// SUBFRAME_MODE used in low latency applications where client need to read the bitstream +// data at slice boundary before frame done. +// 0: disable +// 1: flush at slice end and report last_valid_byte_offset +// 2: report slice offsets without flushing +// 3: report slice offsets with flush at slice end +// +// RC_MODE contains the rate control mode. +// 0: Constant QP +// 1: picture level VBR +// 2: picture level CBR +// 3: picture level VBR with const QP +// 4: QP Map +// 16: (same as 0) +// 17: mbrow level VBR +// 18: mbrow level CBR (QP is updated every macroblock row) +// 19: mbrow level VBR with const QP +// 32: macroblock level activity based rate control algorithm +// +// +// 31 24 20 16 15 12 8 4 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | |D| |G|S|M|C|P| | CODEC | SetControlParams +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SET_CTRL_PARAM 0x00000700 /* -W-4R */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_CODEC_TYPE 3:0 /* -W-UF */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_CODEC_TYPE_H264 0x00000003 +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_FORCE_OUTPIC 8:8 /* -W-UF */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_FORCE_OUTCOL 9:9 /* -W-UF */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_ME_ONLY 10:10 /* -W-UF */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_SLICE_STAT_ON 11:11 /* -W-UF */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_GPTIMER_ON 12:12 /* -W-UF */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_MPEC_STAT_ON 15:15 /* -W-UF */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_DEBUG_MODE 16:16 /* -W-UF */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_SUBFRAME_MODE 18:17 /* -W-UF */ +// #define NV_C9B7_NVENC_SET_CTRL_PARAM_RC_MODE 31:24 /* -W-UF */ +// +// +// SetPictureIndex() +// ----------------- +// +// This method sets the picture index to NVENC. For the same picture, CEA and NVENC +// should receive the same picture index. Picture index should be incremented by 1 +// in each new execute. CEA will put this in the CEA Hints data buffer. +// +// 31 24 23 16 15 8 7 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | INDEX | SetPictureIndex +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SET_PICTURE_INDEX 0x00000704 /* -W-4R */ +// #define NV_C9B7_NVENC_SET_PICTURE_INDEX_INDEX 31:0 /* -W-UF */ +// +// +// SetInRcData() +// ------------------ +// This method sets the location of the RCDataBuffer. It sets the +// upper bits (32 most significant bits) of the 40-bit offset where the buffer +// is located. This buffer will have RC configuration data from CEA or an app. +// +// The values of OFFSET is defined and initialized to zero during the +// instance's creation. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInDrvPicSetup +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_RC_DATA 0x0000070c /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_RC_DATA_OFFSET 31:0 /* -W-UF */ +// +// SetInDrvPicSetup() +// ------------------ +// This method sets the location of the DrvPictureSetupBuffer. It sets the +// upper bits (32 most significant bits) of the 40-bit offset where the buffer +// is located. +// +// The values of OFFSET is defined and initialized to zero during the +// instance's creation. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInDrvPicSetup +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_DRV_PIC_SETUP 0x00000710 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_DRV_PIC_SETUP_OFFSET 31:0 /* -W-UF */ +// +// +// SetInCEAHintsData() +// ------------------- +// This method sets the CEA hints data buffer offset. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInCEAHintsData +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_CEA_HINTS_DATA 0x00000714 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_CEA_HINTS_DATA_OFFSET 31:0 /* -W-UF */ +// +// +// SetOutEncStatus() +// ----------------- +// This method sets the encoder status buffer offset. It sets the upper bits +// (32 most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetOutEncStatus +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETOUT_ENC_STATUS 0x00000718 /* -W-4R */ +// #define NV_C9B7_NVENC_SETOUT_ENC_STATUS_OFFSET 31:0 /* -W-UF */ +// +// +// SetOutBitstream() +// ----------------- +// This method sets the bitstream output data buffer offset. It sets the upper +// bits (32 most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetOutBitstream +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETOUT_BITSTREAM 0x0000071c /* -W-4R */ +// #define NV_C9B7_NVENC_SETOUT_BITSTREAM_OFFSET 31:0 /* -W-UF */ +// +// +// SetIOHistory() +// -------------- +// This method sets the history buffer offset. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetIOHistory +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIO_HISTORY 0x00000720 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIO_HISTORY_OFFSET 31:0 /* -W-UF */ +// +// +// SetIORCProcess() +// ---------------- +// This method sets the RC Process buffer offset. It sets the upper bits +// (32 most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetIORCProcess +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIO_RC_PROCESS 0x00000724 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIO_PROCESS_OFFSET 31:0 /* -W-UF */ +// +// +// SetInColocData() +// ---------------- +// This method sets the input colocated motion data buffer offset. It sets the upper +// bits (32 most significant bits) of the 40-bit offset where the buffer is located. +// This is used only for encoding of B pictures. If the current picture is an I or P +// picture, this method can be omitted. +// +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInColocData +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_COLOC_DATA 0x00000728 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_COLOC_DATA_OFFSET 31:0 /* -W-UF */ +// +// +// SetOutColocData() +// ----------------- +// This method sets the output colocated motion data buffer offset. It sets the upper +// bits (32 most significant bits) of the 40-bit offset where the buffer is located. +// This is used only for encoding of pictures that are marked as reference pictures. +// If the current picture is not marked as a reference picture, this method can be omitted. +// +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetOutColocData +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETOUT_COLOC_DATA 0x0000072c /* -W-4R */ +// #define NV_C9B7_NVENC_SETOUT_COLOC_DATA_OFFSET 31:0 /* -W-UF */ +// +// +// SetOutRefPic() +// -------------- +// This method sets the offset for the tiled 16x16 output reference picture produced by +// the encode of the current picture. It sets the upper bits (32 most significant bits) +// of the 40-bit offset where the buffer is located. This method must be sent if the +// current picture is marked as a reference picture, or the FORCE_OUTPIC bit is set in +// the SetControlParams(). +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetOutRefPic +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETOUT_REFPIC 0x00000730 /* -W-4R */ +// #define NV_C9B7_NVENC_SETOUT_REFPIC_OFFSET 31:0 /* -W-UF */ +// +// +// SetInCurPic() +// ------------- +// This method sets the offset for current input picture to be encoded. It sets the upper +// bits (32 most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInCurPic +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_CURPIC 0x00000734 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_CURPIC_OFFSET 31:0 /* -W-UF */ +// +// +// SetInMEPredData() +// ----------------- +// This method sets the offset for motion estimation predictor input data buffer. +// This is the buffer containing vector hints produced by an earlier execute of NVENC. +// The method sets the upper bits (32 most significant bits) of the 40-bit offset where +// the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInMEPredData +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_MEPRED_DATA 0x00000738 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_MEPRED_DATA_OFFSET 31:0 /* -W-UF */ +// +// +// SetOutMEPredData() +// ------------------ +// This method sets the offset for motion estimation predictor output data Buffer. +// It sets the upper bits (32 most significant bits) of the 40-bit offset where the +// buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetOutMEPredData +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETOUT_MEPRED_DATA 0x0000073c /* -W-4R */ +// #define NV_C9B7_NVENC_SETOUT_MEPRED_DATA_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic0() +// -------------- +// This method sets the offset for reference picture 0. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic0 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC0 0x00000400 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC0_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic1() +// -------------- +// This method sets the offset for reference picture 1. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic1 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC1 0x00000404 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC1_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic2() +// -------------- +// This method sets the offset for reference picture 2. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic2 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC2 0x00000408 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC2_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic3() +// -------------- +// This method sets the offset for reference picture 3. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic3 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC3 0x0000040c /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC3_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic4() +// -------------- +// This method sets the offset for reference picture 4. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic4 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC4 0x00000410 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC4_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic5() +// -------------- +// This method sets the offset for reference picture 5. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic5 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC5 0x00000414 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC5_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic6() +// -------------- +// This method sets the offset for reference picture 6. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic6 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC6 0x00000418 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC6_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic7() +// -------------- +// This method sets the offset for reference picture 7. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic7 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC7 0x0000041c /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC7_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic8() +// -------------- +// This method sets the offset for reference picture 8. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic8 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC8 0x00000420 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC8_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic9() +// -------------- +// This method sets the offset for reference picture 9. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic9 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC9 0x00000424 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC9_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic10() +// --------------- +// This method sets the offset for reference picture 10. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic10 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC10 0x00000428 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC10_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic11() +// --------------- +// This method sets the offset for reference picture 11. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic11 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC11 0x0000042c /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC11_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic12() +// --------------- +// This method sets the offset for reference picture 12. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic12 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC12 0x00000430 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC12_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic13() +// --------------- +// This method sets the offset for reference picture 13. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic13 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC13 0x00000434 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC13_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic14() +// --------------- +// This method sets the offset for reference picture 14. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic14 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC14 0x00000438 /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC14_OFFSET 31:0 /* -W-UF */ +// +// +// SetInRefPic15() +// --------------- +// This method sets the offset for reference picture 15. It sets the upper bits (32 +// most significant bits) of the 40-bit offset where the buffer is located. +// +// 31 16 0 +// .-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-. +// | OFFSET | SetInRefPic15 +// `-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-' +// +// #define NV_C9B7_NVENC_SETIN_REFPIC15 0x0000043c /* -W-4R */ +// #define NV_C9B7_NVENC_SETIN_REFPIC15_OFFSET 31:0 /* -W-UF */ +// +// +// +// +// 4 - APPLICATION SPECIFIC ERROR CODES +// +// The upper 4 bit indicate the codec type. The lower 28 bit are the error code. +// +// 5 - Some shared defines +#define TEB_BUF_SIZE (256 * 1024) +#define HIST_BLOCK_SIZE 192 + +#define H264_HIST_BLOCK_SIZE 224 +#define H265_HIST_BLOCK_SIZE H264_HIST_BLOCK_SIZE +#define VP9_HIST_BLOCK_SIZE H265_HIST_BLOCK_SIZE + +//HIST BLOCK structre has not been finalized, subject to change. +#define AV1_HIST_BLOCK_SIZE VP9_HIST_BLOCK_SIZE + +// Errors +#define NvC9B7EncErrorH264ApptimerExpired 0x30000001 +#define NvC9B7EncErrorH264InvalidInput 0x30000002 +#define NvC9B7EncErrorH264HwerrInterrupt 0x30000003 +#define NvC9B7EncErrorH264BadMagic 0x30000004 + +// Appendix A - KEY +// +// Read +// ' ' = Other Information +// '-' = Field is part of a write-only register +// 'C' = Value read is always the same, constant value line follows (C) +// 'R' = Value is read +// +// Write +// ' ' = Other Information +// '-' = Must not be written (D), value ignored when written (R,A,F) +// 'W' = Can be written +// +// Internal State +// ' ' = Other Information +// '-' = No internal state +// 'X' = Internal state, initial value is unknown +// 'I' = Internal state, initial value is known and follows (I) +// 'V' = Internal state, initialize at volatile reset +// 'N' = Internal state, initial value and initialize at volatile reset +// 'C' = Internal state, initial value at object creation +// +// Declaration/Size +// ' ' = Other Information +// '-' = Does Not Apply +// 'V' = Type is void +// 'U' = Type is unsigned integer +// 'S' = Type is signed integer +// 'F' = Type is IEEE floating point +// '1' = Byte size (008) +// '2' = Short size (016) +// '3' = Three byte size (024) +// '4' = Word size (032) +// '8' = Double size (064) +// +// Define Indicator +// ' ' = Other Information +// 'D' = Device +// 'M' = Memory +// 'R' = Register +// 'A' = Array of Registers +// 'F' = Field +// 'V' = Value +// 'C' = Clear value +// 'S' = Set value +// 'T' = Call task function + + +#endif