PR #23311 opened by Jamaika1 URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/23311 Patch URL: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/23311.patch
# Summary of changes Briefly describe what this PR does and why. <!-- If this PR requires new FATE test samples, attach them to the PR and list their target paths below (relative to the fate-suite root). Attached filenames must match the sample's filename: ```fate-samples # e.g. vorbis/new-sample.ogg ``` --> >From 95ae8c1bf6d1d85b7de01eee80381ce811decd6c Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:27:32 +0000 Subject: [PATCH 01/11] Added image codec AV2 --- libavcodec/libavm.c | 49 ++ libavcodec/libavmdec.c | 315 +++++++++ libavcodec/libavmenc.c | 1465 ++++++++++++++++++++++++++++++++++++++++ 3 files changed, 1829 insertions(+) create mode 100644 libavcodec/libavm.c create mode 100644 libavcodec/libavmdec.c create mode 100644 libavcodec/libavmenc.c diff --git a/libavcodec/libavm.c b/libavcodec/libavm.c new file mode 100644 index 0000000000..4073fa9684 --- /dev/null +++ b/libavcodec/libavm.c @@ -0,0 +1,49 @@ +/* + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +/** + * @file + * AOM common functions + */ + +#include "libavutil/pixdesc.h" +#include "libavm.h" + +void ff_avm_image_copy_16_to_8(AVFrame *pic, struct avm_image *img) +{ + const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(pic->format); + int i; + + for (i = 0; i < desc->nb_components; i++) { + int w = img->d_w; + int h = img->d_h; + int x, y; + + if (i) { + w = (w + img->x_chroma_shift) >> img->x_chroma_shift; + h = (h + img->y_chroma_shift) >> img->y_chroma_shift; + } + + for (y = 0; y < h; y++) { + uint16_t *src = (uint16_t *)(img->planes[i] + y * img->stride[i]); + uint8_t *dst = pic->data[i] + y * pic->linesize[i]; + for (x = 0; x < w; x++) + *dst++ = *src++; + } + } +} diff --git a/libavcodec/libavmdec.c b/libavcodec/libavmdec.c new file mode 100644 index 0000000000..f1606b90f0 --- /dev/null +++ b/libavcodec/libavmdec.c @@ -0,0 +1,315 @@ +/* + * Copyright (c) 2010, Google, Inc. + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +/** + * @file + * AV2 decoder support via libavm + */ + +#include <avm/avm_decoder.h> +#include <avm/avmdx.h> + +#include "libavutil/common.h" +#include "libavutil/cpu.h" +#include "libavutil/hdr_dynamic_metadata.h" +#include "libavutil/imgutils.h" + +#include "avcodec.h" +#include "bytestream.h" +#include "codec_internal.h" +#include "decode.h" +#include "itut35.h" +#include "libavm.h" +#include "profiles.h" + +typedef struct AV2DecodeContext { + struct avm_codec_ctx decoder; +} AV2DecodeContext; + +static av_cold int avm_init(AVCodecContext *avctx, + const struct avm_codec_iface *iface) +{ + AV2DecodeContext *ctx = avctx->priv_data; + struct avm_codec_dec_cfg deccfg = { + .threads = FFMIN(avctx->thread_count ? avctx->thread_count : av_cpu_count(), 16) + }; + + av_log(avctx, AV_LOG_VERBOSE, "%s\n", avm_codec_version_str()); + av_log(avctx, AV_LOG_VERBOSE, "%s\n", avm_codec_build_config()); + + if (avm_codec_dec_init(&ctx->decoder, iface, &deccfg, 0) != AVM_CODEC_OK) { + const char *error = avm_codec_error(&ctx->decoder); + av_log(avctx, AV_LOG_ERROR, "Failed to initialize decoder: %s\n", + error); + return AVERROR(EINVAL); + } + + return 0; +} + +// returns 0 on success, AVERROR_INVALIDDATA otherwise +static int set_pix_fmt(AVCodecContext *avctx, struct avm_image *img) +{ + static const enum AVColorRange color_ranges[] = { + AVCOL_RANGE_MPEG, AVCOL_RANGE_JPEG + }; + avctx->color_range = color_ranges[img->range]; + if (img->cp != AVM_CICP_CP_UNSPECIFIED) + avctx->color_primaries = (enum AVColorPrimaries)img->cp; + if (img->mc != AVM_CICP_MC_UNSPECIFIED) + avctx->colorspace = (enum AVColorSpace)img->mc; + if (img->tc != AVM_CICP_TC_UNSPECIFIED) + avctx->color_trc = (enum AVColorTransferCharacteristic)img->tc; + + switch (img->fmt) { + case AVM_IMG_FMT_I420: + case AVM_IMG_FMT_I42016: + if (img->bit_depth == 8) { + avctx->pix_fmt = img->monochrome ? + AV_PIX_FMT_GRAY8 : AV_PIX_FMT_YUV420P; + avctx->profile = AV_PROFILE_AV1_MAIN; + return 0; + } else if (img->bit_depth == 10) { + avctx->pix_fmt = img->monochrome ? + AV_PIX_FMT_GRAY10 : AV_PIX_FMT_YUV420P10; + avctx->profile = AV_PROFILE_AV1_MAIN; + return 0; + } else if (img->bit_depth == 12) { + avctx->pix_fmt = img->monochrome ? + AV_PIX_FMT_GRAY12 : AV_PIX_FMT_YUV420P12; + avctx->profile = AV_PROFILE_AV1_PROFESSIONAL; + return 0; + } else { + return AVERROR_INVALIDDATA; + } + case AVM_IMG_FMT_I422: + case AVM_IMG_FMT_I42216: + if (img->bit_depth == 8) { + avctx->pix_fmt = AV_PIX_FMT_YUV422P; + avctx->profile = AV_PROFILE_AV1_PROFESSIONAL; + return 0; + } else if (img->bit_depth == 10) { + avctx->pix_fmt = AV_PIX_FMT_YUV422P10; + avctx->profile = AV_PROFILE_AV1_PROFESSIONAL; + return 0; + } else if (img->bit_depth == 12) { + avctx->pix_fmt = AV_PIX_FMT_YUV422P12; + avctx->profile = AV_PROFILE_AV1_PROFESSIONAL; + return 0; + } else { + return AVERROR_INVALIDDATA; + } + case AVM_IMG_FMT_I444: + case AVM_IMG_FMT_I44416: + if (img->bit_depth == 8) { + avctx->pix_fmt = avctx->colorspace == AVCOL_SPC_RGB ? + AV_PIX_FMT_GBRP : AV_PIX_FMT_YUV444P; + avctx->profile = AV_PROFILE_AV1_HIGH; + return 0; + } else if (img->bit_depth == 10) { + avctx->pix_fmt = AV_PIX_FMT_YUV444P10; + avctx->pix_fmt = avctx->colorspace == AVCOL_SPC_RGB ? + AV_PIX_FMT_GBRP10 : AV_PIX_FMT_YUV444P10; + avctx->profile = AV_PROFILE_AV1_HIGH; + return 0; + } else if (img->bit_depth == 12) { + avctx->pix_fmt = avctx->colorspace == AVCOL_SPC_RGB ? + AV_PIX_FMT_GBRP12 : AV_PIX_FMT_YUV444P12; + avctx->profile = AV_PROFILE_AV1_PROFESSIONAL; + return 0; + } else { + return AVERROR_INVALIDDATA; + } + + default: + return AVERROR_INVALIDDATA; + } +} + +static int decode_metadata_itu_t_t35(AVFrame *frame, + const uint8_t *buffer, size_t buffer_size) +{ + if (buffer_size < 6) + return AVERROR(EINVAL); + + GetByteContext bc; + bytestream2_init(&bc, buffer, buffer_size); + + const int country_code = bytestream2_get_byteu(&bc); + const int provider_code = bytestream2_get_be16u(&bc); + const int provider_oriented_code = bytestream2_get_be16u(&bc); + const int application_identifier = bytestream2_get_byteu(&bc); + + // See "HDR10+ AV2 Metadata Handling Specification" v1.0.1, Section 2.1. + if (country_code == ITU_T_T35_COUNTRY_CODE_US + && provider_code == ITU_T_T35_PROVIDER_CODE_SAMSUNG + && provider_oriented_code == 0x0001 + && application_identifier == 0x04) { + // HDR10+ + AVDynamicHDRPlus *hdr_plus = av_dynamic_hdr_plus_create_side_data(frame); + if (!hdr_plus) + return AVERROR(ENOMEM); + + int res = av_dynamic_hdr_plus_from_t35(hdr_plus, bc.buffer, + bytestream2_get_bytes_left(&bc)); + if (res < 0) + return res; + } + + return 0; +} + +static int decode_metadata(AVFrame *frame, const struct avm_image *img) +{ + const size_t num_metadata = avm_img_num_metadata(img); + for (size_t i = 0; i < num_metadata; ++i) { + const avm_metadata_t *metadata = avm_img_get_metadata(img, i); + if (!metadata) + continue; + + switch (metadata->type) { + case OBU_METADATA_TYPE_ITUT_T35: { + int res = decode_metadata_itu_t_t35(frame, metadata->payload, metadata->sz); + if (res < 0) + return res; + break; + } + default: + break; + } + } + return 0; +} + +static int avm_decode(AVCodecContext *avctx, AVFrame *picture, + int *got_frame, AVPacket *avpkt) +{ + AV2DecodeContext *ctx = avctx->priv_data; + const void *iter = NULL; + struct avm_image *img; + int ret; + + if (avm_codec_decode(&ctx->decoder, avpkt->data, avpkt->size, NULL) != + AVM_CODEC_OK) { + const char *error = avm_codec_error(&ctx->decoder); + const char *detail = avm_codec_error_detail(&ctx->decoder); + + av_log(avctx, AV_LOG_ERROR, "Failed to decode frame: %s\n", error); + if (detail) + av_log(avctx, AV_LOG_ERROR, " Additional information: %s\n", + detail); + return AVERROR_INVALIDDATA; + } + + if ((img = avm_codec_get_frame(&ctx->decoder, &iter))) { + if (img->d_w > img->w || img->d_h > img->h) { + av_log(avctx, AV_LOG_ERROR, "Display dimensions %dx%d exceed storage %dx%d\n", + img->d_w, img->d_h, img->w, img->h); + return AVERROR_EXTERNAL; + } + + if ((ret = set_pix_fmt(avctx, img)) < 0) { + av_log(avctx, AV_LOG_ERROR, "Unsupported output colorspace (%d) / bit_depth (%d)\n", + img->fmt, img->bit_depth); + return ret; + } + + if ((int)img->d_w != avctx->width || (int)img->d_h != avctx->height) { + av_log(avctx, AV_LOG_INFO, "dimension change! %dx%d -> %dx%d\n", + avctx->width, avctx->height, img->d_w, img->d_h); + ret = ff_set_dimensions(avctx, img->d_w, img->d_h); + if (ret < 0) + return ret; + } + if ((ret = ff_get_buffer(avctx, picture, 0)) < 0) + return ret; + +#ifdef AVM_CTRL_AVMD_GET_FRAME_FLAGS + { + avm_codec_frame_flags_t flags; + ret = avm_codec_control(&ctx->decoder, AVMD_GET_FRAME_FLAGS, &flags); + if (ret == AVM_CODEC_OK) { + if (flags & AVM_FRAME_IS_KEY) + picture->flags |= AV_FRAME_FLAG_KEY; + else + picture->flags &= ~AV_FRAME_FLAG_KEY; + if (flags & (AVM_FRAME_IS_KEY | AVM_FRAME_IS_INTRAONLY)) + picture->pict_type = AV_PICTURE_TYPE_I; + else if (flags & AVM_FRAME_IS_SWITCH) + picture->pict_type = AV_PICTURE_TYPE_SP; + else + picture->pict_type = AV_PICTURE_TYPE_P; + } + } +#endif + + av_reduce(&picture->sample_aspect_ratio.num, + &picture->sample_aspect_ratio.den, + picture->height * img->r_w, + picture->width * img->r_h, + INT_MAX); + ff_set_sar(avctx, picture->sample_aspect_ratio); + + if ((img->fmt & AVM_IMG_FMT_HIGHBITDEPTH) && img->bit_depth == 8) + ff_avm_image_copy_16_to_8(picture, img); + else { + const uint8_t *planes[4] = { img->planes[0], img->planes[1], img->planes[2] }; + const int stride[4] = { img->stride[0], img->stride[1], img->stride[2] }; + + av_image_copy(picture->data, picture->linesize, planes, + stride, avctx->pix_fmt, img->d_w, img->d_h); + } + ret = decode_metadata(picture, img); + if (ret < 0) { + av_log(avctx, AV_LOG_ERROR, "Failed to decode metadata\n"); + return ret; + } + *got_frame = 1; + } + return avpkt->size; +} + +static av_cold int avm_free(AVCodecContext *avctx) +{ + AV2DecodeContext *ctx = avctx->priv_data; + avm_codec_destroy(&ctx->decoder); + return 0; +} + +static av_cold int av2_init(AVCodecContext *avctx) +{ + return avm_init(avctx, avm_codec_av2_dx()); +} + +const FFCodec ff_libavm_av2_decoder = { + .p.name = "libavm-av2", + CODEC_LONG_NAME("libavm AV2"), + .p.type = AVMEDIA_TYPE_VIDEO, + .p.id = AV_CODEC_ID_AV2, + .priv_data_size = sizeof(AV2DecodeContext), + .init = av2_init, + .close = avm_free, + FF_CODEC_DECODE_CB(avm_decode), + .p.capabilities = AV_CODEC_CAP_OTHER_THREADS | AV_CODEC_CAP_DR1, + .caps_internal = FF_CODEC_CAP_NOT_INIT_THREADSAFE | + FF_CODEC_CAP_AUTO_THREADS, + //.p.profiles = NULL_IF_CONFIG_SMALL(ff_av1_profiles), + .p.wrapper_name = "libavm", +}; diff --git a/libavcodec/libavmenc.c b/libavcodec/libavmenc.c new file mode 100644 index 0000000000..bd5a22c213 --- /dev/null +++ b/libavcodec/libavmenc.c @@ -0,0 +1,1465 @@ +/* + * Copyright (c) 2010, Google, Inc. + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +/** + * @file + * AV2 encoder support via libavm + */ + +#include <limits.h> + +#define AVM_DISABLE_CTRL_TYPECHECKS 1 +#include <avm/avm_encoder.h> +#include <avm/avmcx.h> + +#define MAX_TILE_WIDTH (4096) // Max Tile width in pixels +#define MAX_TILE_AREA (4096 * 2304) // Maximum tile area in pixels +#define MAX_TILE_ROWS 64 +#define MAX_TILE_COLS 64 + +#include "libavutil/avassert.h" +#include "libavutil/base64.h" +#include "libavutil/cpu.h" +#include "libavutil/hdr_dynamic_metadata.h" +#include "libavutil/imgutils.h" +#include "libavutil/mathematics.h" +#include "libavutil/mem.h" +#include "libavutil/opt.h" +#include "libavutil/pixdesc.h" + +#include "avcodec.h" +#include "bsf.h" +#include "bytestream.h" +#include "codec_internal.h" +#include "dovi_rpu.h" +#include "encode.h" +#include "internal.h" +#include "itut35.h" +#include "libavm.h" +#include "profiles.h" + +/* + * Portion of struct avm_codec_cx_pkt from avm_encoder.h. + * One encoded frame returned from the library. + */ +struct FrameListData { + void *buf; /**< compressed data buffer */ + size_t sz; /**< length of compressed data */ + int64_t pts; /**< time stamp to show frame + (in timebase units) */ + unsigned long duration; /**< duration to show frame + (in timebase units) */ + uint32_t flags; /**< flags for this frame */ + uint64_t sse[4]; + int have_sse; /**< true if we have pending sse[] */ + uint64_t frame_number; + struct FrameListData *next; +}; + +typedef struct AVMEncoderContext { + AVClass *class; + AVBSFContext *bsf; + DOVIContext dovi; + struct avm_codec_ctx encoder; + struct avm_image rawimg; + struct avm_fixed_buf twopass_stats; + unsigned twopass_stats_size; + struct FrameListData *coded_frame_list; + int cpu_used; + int auto_alt_ref; + int arnr_strength; + int aq_mode; + int lag_in_frames; + int qp; + int static_thresh; + int denoise_noise_level; + int denoise_block_size; + uint64_t sse[4]; + int have_sse; /**< true if we have pending sse[] */ + uint64_t frame_number; + int rc_undershoot_pct; + int rc_overshoot_pct; + int minsection_pct; + int maxsection_pct; + int frame_parallel; + int tile_cols, tile_rows; + int tile_cols_log2, tile_rows_log2; + avm_superblock_size_t superblock_size; + int uniform_tiles; + int row_mt; + int enable_cdef; + int enable_global_motion; + int enable_intrabc; + int enable_restoration; + int usage; + int tune; + int still_picture; + int enable_rect_partitions; + int enable_angle_delta; + int enable_cfl_intra; + int enable_paeth_intra; + int enable_smooth_intra; + int enable_intra_edge_filter; + int enable_palette; + int enable_flip_idtx; + int reduced_tx_type_set; + int use_intra_dct_only; + int use_inter_dct_only; + int use_intra_default_tx_only; + int enable_ref_frame_mvs; + int enable_interinter_wedge; + int enable_interintra_wedge; + int enable_interintra_comp; + int enable_masked_comp; + int enable_onesided_comp; + int enable_smooth_interintra; + int enable_diff_wtd_comp; + AVDictionary *avm_params; +} AVMContext; + +#define OFFSET(x) offsetof(AVMContext, x) + +static const char *const ctlidstr[] = { + [AVME_SET_CPUUSED] = "AVME_SET_CPUUSED", + [AVME_SET_QP] = "AVME_SET_QP", + [AVME_SET_ENABLEAUTOALTREF] = "AVME_SET_ENABLEAUTOALTREF", + [AVME_SET_ARNR_MAXFRAMES] = "AVME_SET_ARNR_MAXFRAMES", + [AVME_SET_ARNR_STRENGTH] = "AVME_SET_ARNR_STRENGTH", + [AVME_SET_STATIC_THRESHOLD] = "AVME_SET_STATIC_THRESHOLD", + [AV2E_SET_COLOR_RANGE] = "AV2E_SET_COLOR_RANGE", + [AV2E_SET_COLOR_PRIMARIES] = "AV2E_SET_COLOR_PRIMARIES", + [AV2E_SET_MATRIX_COEFFICIENTS] = "AV2E_SET_MATRIX_COEFFICIENTS", + [AV2E_SET_TRANSFER_CHARACTERISTICS] = "AV2E_SET_TRANSFER_CHARACTERISTICS", + [AV2E_SET_AQ_MODE] = "AV2E_SET_AQ_MODE", + [AV2E_SET_FRAME_PARALLEL_DECODING] = "AV2E_SET_FRAME_PARALLEL_DECODING", + [AV2E_SET_SUPERBLOCK_SIZE] = "AV2E_SET_SUPERBLOCK_SIZE", + [AV2E_SET_TILE_COLUMNS] = "AV2E_SET_TILE_COLUMNS", + [AV2E_SET_TILE_ROWS] = "AV2E_SET_TILE_ROWS", + [AV2E_SET_ENABLE_RESTORATION] = "AV2E_SET_ENABLE_RESTORATION", + [AV2E_SET_ROW_MT] = "AV2E_SET_ROW_MT", + [AV2E_SET_DENOISE_NOISE_LEVEL] = "AV2E_SET_DENOISE_NOISE_LEVEL", + [AV2E_SET_DENOISE_BLOCK_SIZE] = "AV2E_SET_DENOISE_BLOCK_SIZE", + [AV2E_SET_MAX_REFERENCE_FRAMES] = "AV2E_SET_MAX_REFERENCE_FRAMES", + [AV2E_SET_ENABLE_GLOBAL_MOTION] = "AV2E_SET_ENABLE_GLOBAL_MOTION", + [AV2E_SET_ENABLE_INTRABC] = "AV2E_SET_ENABLE_INTRABC", + [AV2E_SET_ENABLE_CDEF] = "AV2E_SET_ENABLE_CDEF", + [AVME_SET_TUNING] = "AVME_SET_TUNING", + [AV2E_SET_ENABLE_1TO4_PARTITIONS] = "AV2E_SET_ENABLE_1TO4_PARTITIONS", + [AV2E_SET_ENABLE_RECT_PARTITIONS] = "AV2E_SET_ENABLE_RECT_PARTITIONS", + [AV2E_SET_ENABLE_ANGLE_DELTA] = "AV2E_SET_ENABLE_ANGLE_DELTA", + [AV2E_SET_ENABLE_CFL_INTRA] = "AV2E_SET_ENABLE_CFL_INTRA", + [AV2E_SET_ENABLE_INTRA_EDGE_FILTER] = "AV2E_SET_ENABLE_INTRA_EDGE_FILTER", + [AV2E_SET_ENABLE_PAETH_INTRA] = "AV2E_SET_ENABLE_PAETH_INTRA", + [AV2E_SET_ENABLE_SMOOTH_INTRA] = "AV2E_SET_ENABLE_SMOOTH_INTRA", + [AV2E_SET_ENABLE_PALETTE] = "AV2E_SET_ENABLE_PALETTE", + [AV2E_SET_ENABLE_FLIP_IDTX] = "AV2E_SET_ENABLE_FLIP_IDTX", + [AV2E_SET_INTRA_DCT_ONLY] = "AV2E_SET_INTRA_DCT_ONLY", + [AV2E_SET_INTER_DCT_ONLY] = "AV2E_SET_INTER_DCT_ONLY", + [AV2E_SET_INTRA_DEFAULT_TX_ONLY] = "AV2E_SET_INTRA_DEFAULT_TX_ONLY", + [AV2E_SET_REDUCED_TX_TYPE_SET] = "AV2E_SET_REDUCED_TX_TYPE_SET", + [AV2E_SET_ENABLE_DIFF_WTD_COMP] = "AV2E_SET_ENABLE_DIFF_WTD_COMP", + [AV2E_SET_ENABLE_INTERINTER_WEDGE] = "AV2E_SET_ENABLE_INTERINTER_WEDGE", + [AV2E_SET_ENABLE_INTERINTRA_WEDGE] = "AV2E_SET_ENABLE_INTERINTRA_WEDGE", + [AV2E_SET_ENABLE_MASKED_COMP] = "AV2E_SET_ENABLE_MASKED_COMP", + [AV2E_SET_ENABLE_INTERINTRA_COMP] = "AV2E_SET_ENABLE_INTERINTRA_COMP", + [AV2E_SET_ENABLE_ONESIDED_COMP] = "AV2E_SET_ENABLE_ONESIDED_COMP", + [AV2E_SET_REDUCED_REFERENCE_SET] = "AV2E_SET_REDUCED_REFERENCE_SET", + [AV2E_SET_ENABLE_SMOOTH_INTERINTRA] = "AV2E_SET_ENABLE_SMOOTH_INTERINTRA", + [AV2E_SET_ENABLE_REF_FRAME_MVS] = "AV2E_SET_ENABLE_REF_FRAME_MVS", + [AV2E_GET_SEQ_LEVEL_IDX] = "AV2E_GET_SEQ_LEVEL_IDX", + [AV2_GET_NEW_FRAME_IMAGE] = "AV2_GET_NEW_FRAME_IMAGE", +}; + +static av_cold void log_encoder_error(AVCodecContext *avctx, const char *desc) +{ + AVMContext *ctx = avctx->priv_data; + const char *error = avm_codec_error(&ctx->encoder); + const char *detail = avm_codec_error_detail(&ctx->encoder); + + av_log(avctx, AV_LOG_ERROR, "%s: %s\n", desc, error); + if (detail) + av_log(avctx, AV_LOG_ERROR, " Additional information: %s\n", detail); +} + +static av_cold void dump_enc_cfg(AVCodecContext *avctx, + const struct avm_codec_enc_cfg *cfg, + int level) +{ + int width = -30; + + av_log(avctx, level, "avm_codec_enc_cfg\n"); + av_log(avctx, level, "generic settings\n" + " %*s%u\n %*s%u\n %*s%u\n %*s%u\n %*s%u\n" + " %*s%u\n %*s%u\n" + " %*s{%u/%u}\n %*s%d\n %*s%u\n", + width, "g_usage:", cfg->g_usage, + width, "g_threads:", cfg->g_threads, + width, "g_profile:", cfg->g_profile, + width, "g_w:", cfg->g_w, + width, "g_h:", cfg->g_h, + width, "g_bit_depth:", cfg->g_bit_depth, + width, "g_input_bit_depth:", cfg->g_input_bit_depth, + width, "g_timebase:", cfg->g_timebase.num, cfg->g_timebase.den, + width, "g_pass:", cfg->g_pass, + width, "g_lag_in_frames:", cfg->g_lag_in_frames); + av_log(avctx, level, "rate control settings\n" + " %*s%d\n %*s%u\n", + width, "rc_end_usage:", cfg->rc_end_usage, + width, "rc_target_bitrate:", cfg->rc_target_bitrate); + av_log(avctx, level, "quantizer settings\n" + " %*s%u\n %*s%u\n", + width, "rc_min_quantizer:", cfg->rc_min_quantizer, + width, "rc_max_quantizer:", cfg->rc_max_quantizer); + av_log(avctx, level, "bitrate tolerance\n" + " %*s%u\n %*s%u\n", + width, "rc_undershoot_pct:", cfg->rc_undershoot_pct, + width, "rc_overshoot_pct:", cfg->rc_overshoot_pct); + av_log(avctx, level, "decoder buffer model\n" + " %*s%u\n %*s%u\n %*s%u\n", + width, "rc_buf_sz:", cfg->rc_buf_sz, + width, "rc_buf_initial_sz:", cfg->rc_buf_initial_sz, + width, "rc_buf_optimal_sz:", cfg->rc_buf_optimal_sz); + av_log(avctx, level, "2 pass rate control settings\n" + " %*s%u\n %*s%u\n", + width, "rc_2pass_vbr_minsection_pct:", cfg->rc_2pass_vbr_minsection_pct, + width, "rc_2pass_vbr_maxsection_pct:", cfg->rc_2pass_vbr_maxsection_pct); + av_log(avctx, level, "keyframing settings\n" + " %*s%d\n %*s%u\n %*s%u\n", + width, "kf_mode:", cfg->kf_mode, + width, "kf_min_dist:", cfg->kf_min_dist, + width, "kf_max_dist:", cfg->kf_max_dist); + av_log(avctx, level, "tile settings\n" + " %*s%d\n %*s%d\n", + width, "tile_width_count:", cfg->tile_width_count, + width, "tile_height_count:", cfg->tile_height_count); + av_log(avctx, level, "\n"); +} + +static void coded_frame_add(void *list, struct FrameListData *cx_frame) +{ + struct FrameListData **p = list; + + while (*p) + p = &(*p)->next; + *p = cx_frame; + cx_frame->next = NULL; +} + +static av_cold void free_coded_frame(struct FrameListData *cx_frame) +{ + av_freep(&cx_frame->buf); + av_freep(&cx_frame); +} + +static av_cold void free_frame_list(struct FrameListData *list) +{ + struct FrameListData *p = list; + + while (p) { + list = list->next; + free_coded_frame(p); + p = list; + } +} + +static av_cold int codecctl_int(AVCodecContext *avctx, + int id, + int val) +{ + AVMContext *ctx = avctx->priv_data; + char buf[80]; + int width = -30; + int res; + + snprintf(buf, sizeof(buf), "%s:", ctlidstr[id]); + av_log(avctx, AV_LOG_DEBUG, " %*s%d\n", width, buf, val); + + res = avm_codec_control(&ctx->encoder, id, val); + if (res != AVM_CODEC_OK) { + snprintf(buf, sizeof(buf), "Failed to set %s codec control", + ctlidstr[id]); + log_encoder_error(avctx, buf); + return AVERROR(EINVAL); + } + + return 0; +} + +static int add_hdr_plus(AVCodecContext *avctx, struct avm_image *img, const AVFrame *frame) +{ + // Check for HDR10+ + AVFrameSideData *side_data = + av_frame_get_side_data(frame, AV_FRAME_DATA_DYNAMIC_HDR_PLUS); + if (!side_data) + return 0; + + size_t payload_size; + AVDynamicHDRPlus *hdr_plus = (AVDynamicHDRPlus *)side_data->buf->data; + int res = av_dynamic_hdr_plus_to_t35(hdr_plus, NULL, &payload_size); + if (res < 0) { + log_encoder_error(avctx, "Error finding the size of HDR10+"); + return res; + } + + uint8_t *hdr_plus_buf; + // Extra bytes for the country code, provider code, provider oriented code and app id. + const size_t hdr_plus_buf_size = payload_size + 6; + hdr_plus_buf = av_malloc(hdr_plus_buf_size); + if (!hdr_plus_buf) + return AVERROR(ENOMEM); + + uint8_t *payload = hdr_plus_buf; + // See "HDR10+ AV2 Metadata Handling Specification" v1.0.1, Section 2.1. + bytestream_put_byte(&payload, ITU_T_T35_COUNTRY_CODE_US); + bytestream_put_be16(&payload, ITU_T_T35_PROVIDER_CODE_SAMSUNG); + bytestream_put_be16(&payload, 0x0001); // provider_oriented_code + bytestream_put_byte(&payload, 0x04); // application_identifier + + res = av_dynamic_hdr_plus_to_t35(hdr_plus, &payload, &payload_size); + if (res < 0) { + av_free(hdr_plus_buf); + log_encoder_error(avctx, "Error encoding HDR10+ from side data"); + return res; + } + + res = avm_img_add_metadata(img, OBU_METADATA_TYPE_ITUT_T35, + hdr_plus_buf, hdr_plus_buf_size, AVM_MIF_ANY_FRAME); + av_free(hdr_plus_buf); + if (res < 0) { + log_encoder_error(avctx, "Error adding HDR10+ to avm_img"); + return res; + } + return 0; +} + +static int add_hdr_smpte2094_app5(AVCodecContext *avctx, struct avm_image *img, + const AVFrame *frame) +{ + AVFrameSideData *side_data = + av_frame_get_side_data(frame, AV_FRAME_DATA_DYNAMIC_HDR_SMPTE_2094_APP5); + if (!side_data) + return 0; + + size_t payload_size; + AVDynamicHDRSmpte2094App5 *hdr = (AVDynamicHDRSmpte2094App5 *)side_data->buf->data; + int res = av_dynamic_hdr_smpte2094_app5_to_t35(hdr, NULL, &payload_size); + if (res < 0) { + log_encoder_error(avctx, "Error finding the size of HDR SMPTE-2094-50"); + return res; + } + + uint8_t *hdr_buf; + // Extra bytes for the country code, provider code, provider oriented code. + const size_t hdr_buf_size = payload_size + 5; + hdr_buf = av_malloc(hdr_buf_size); + if (!hdr_buf) + return AVERROR(ENOMEM); + + uint8_t *payload = hdr_buf; + bytestream_put_byte(&payload, ITU_T_T35_COUNTRY_CODE_US); + bytestream_put_be16(&payload, ITU_T_T35_PROVIDER_CODE_SMPTE); + bytestream_put_be16(&payload, 0x0001); // provider_oriented_code + + res = av_dynamic_hdr_smpte2094_app5_to_t35(hdr, &payload, &payload_size); + if (res < 0) { + av_free(hdr_buf); + log_encoder_error(avctx, "Error encoding HDR SMPTE-2094-50 from side data"); + return res; + } + + res = avm_img_add_metadata(img, OBU_METADATA_TYPE_ITUT_T35, + hdr_buf, hdr_buf_size, AVM_MIF_ANY_FRAME); + av_free(hdr_buf); + if (res < 0) { + log_encoder_error(avctx, "Error adding HDR SMPTE-2094-50 to avm_img"); + return res; + } + return 0; +} + +static av_cold int codecctl_imgp(AVCodecContext *avctx, + int id, + struct avm_image *img) +{ + AVMContext *ctx = avctx->priv_data; + char buf[80]; + int res; + + snprintf(buf, sizeof(buf), "%s:", ctlidstr[id]); + + res = avm_codec_control(&ctx->encoder, id, img); + if (res != AVM_CODEC_OK) { + snprintf(buf, sizeof(buf), "Failed to get %s codec control", + ctlidstr[id]); + log_encoder_error(avctx, buf); + return AVERROR(EINVAL); + } + + return 0; +} + +static av_cold int avm_free(AVCodecContext *avctx) +{ + AVMContext *ctx = avctx->priv_data; + avm_codec_destroy(&ctx->encoder); + avm_img_remove_metadata(&ctx->rawimg); + av_freep(&ctx->twopass_stats.buf); + av_freep(&avctx->stats_out); + free_frame_list(ctx->coded_frame_list); + av_bsf_free(&ctx->bsf); + ff_dovi_ctx_unref(&ctx->dovi); + return 0; +} + +static int set_pix_fmt(AVCodecContext *avctx, avm_codec_caps_t codec_caps, + struct avm_codec_enc_cfg *enccfg, avm_codec_flags_t *flags, + avm_img_fmt_t *img_fmt) +{ + const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(avctx->pix_fmt); + enccfg->g_bit_depth = enccfg->g_input_bit_depth = desc->comp[0].depth; + switch (avctx->pix_fmt) { + case AV_PIX_FMT_GRAY8: + enccfg->monochrome = 1; + av_fallthrough; + case AV_PIX_FMT_YUV420P: + enccfg->g_profile = 1; + *img_fmt = AVM_IMG_FMT_I420; + return 0; + case AV_PIX_FMT_YUV422P: + enccfg->g_profile = 3; + *img_fmt = AVM_IMG_FMT_I422; + return 0; + case AV_PIX_FMT_YUV444P: + case AV_PIX_FMT_GBRP: + enccfg->g_profile = 4; + *img_fmt = AVM_IMG_FMT_I444; + return 0; + case AV_PIX_FMT_GRAY10: + case AV_PIX_FMT_GRAY12: + enccfg->monochrome = 1; + av_fallthrough; + case AV_PIX_FMT_YUV420P10: + case AV_PIX_FMT_YUV420P12: + enccfg->g_profile = + enccfg->g_bit_depth > 10 ? 5 : 1; + *img_fmt = AVM_IMG_FMT_I42016; + return 0; + break; + case AV_PIX_FMT_YUV422P10: + case AV_PIX_FMT_YUV422P12: + enccfg->g_profile = + enccfg->g_bit_depth > 10 ? 5 : 3; + *img_fmt = AVM_IMG_FMT_I42216; + return 0; + break; + case AV_PIX_FMT_YUV444P10: + case AV_PIX_FMT_YUV444P12: + case AV_PIX_FMT_GBRP10: + case AV_PIX_FMT_GBRP12: + enccfg->g_profile = + enccfg->g_bit_depth > 10 ? 5 : 4; + *img_fmt = AVM_IMG_FMT_I44416; + return 0; + break; + default: + break; + } + av_log(avctx, AV_LOG_ERROR, "Unsupported pixel format.\n"); + return AVERROR_INVALIDDATA; +} + +static void set_color_range(AVCodecContext *avctx) +{ + avm_color_range_t avm_cr; + switch (avctx->color_range) { + case AVCOL_RANGE_UNSPECIFIED: + case AVCOL_RANGE_MPEG: avm_cr = AVM_CR_STUDIO_RANGE; break; + case AVCOL_RANGE_JPEG: avm_cr = AVM_CR_FULL_RANGE; break; + default: + av_log(avctx, AV_LOG_WARNING, "Unsupported color range (%d)\n", + avctx->color_range); + return; + } + + codecctl_int(avctx, AV2E_SET_COLOR_RANGE, avm_cr); +} + +static int count_uniform_tiling(int dim, int sb_size, int tiles_log2) +{ + int sb_dim = (dim + sb_size - 1) / sb_size; + int tile_dim = (sb_dim + (1 << tiles_log2) - 1) >> tiles_log2; + av_assert0(tile_dim > 0); + return (sb_dim + tile_dim - 1) / tile_dim; +} + +static int choose_tiling(AVCodecContext *avctx, + struct avm_codec_enc_cfg *enccfg) +{ + AVMContext *ctx = avctx->priv_data; + int sb_256x256_possible, sb_128x128_possible, sb_size, sb_width, sb_height; + int uniform_rows, uniform_cols; + int uniform_64x64_possible, uniform_128x128_possible, uniform_256x256_possible; + int tile_size, rounding, i; + + if (ctx->tile_cols_log2 >= 0) + ctx->tile_cols = 1 << ctx->tile_cols_log2; + if (ctx->tile_rows_log2 >= 0) + ctx->tile_rows = 1 << ctx->tile_rows_log2; + + if (ctx->tile_cols == 0) { + ctx->tile_cols = (avctx->width + MAX_TILE_WIDTH - 1) / + MAX_TILE_WIDTH; + if (ctx->tile_cols > 1) { + av_log(avctx, AV_LOG_DEBUG, "Automatically using %d tile " + "columns to fill width.\n", ctx->tile_cols); + } + } + av_assert0(ctx->tile_cols > 0); + if (ctx->tile_rows == 0) { + int max_tile_width = + FFALIGN((FFALIGN(avctx->width, 128) + + ctx->tile_cols - 1) / ctx->tile_cols, 128); + ctx->tile_rows = + (max_tile_width * FFALIGN(avctx->height, 128) + + MAX_TILE_AREA - 1) / MAX_TILE_AREA; + if (ctx->tile_rows > 1) { + av_log(avctx, AV_LOG_DEBUG, "Automatically using %d tile " + "rows to fill area.\n", ctx->tile_rows); + } + } + av_assert0(ctx->tile_rows > 0); + + if ((avctx->width + 63) / 64 < ctx->tile_cols || + (avctx->height + 63) / 64 < ctx->tile_rows) { + av_log(avctx, AV_LOG_ERROR, "Invalid tile sizing: frame not " + "large enough to fit specified tile arrangement.\n"); + return AVERROR(EINVAL); + } + if (ctx->tile_cols > MAX_TILE_COLS || + ctx->tile_rows > MAX_TILE_ROWS) { + av_log(avctx, AV_LOG_ERROR, "Invalid tile sizing: AV2 does " + "not allow more than %dx%d tiles.\n", + MAX_TILE_COLS, MAX_TILE_ROWS); + return AVERROR(EINVAL); + } + if (avctx->width / ctx->tile_cols > MAX_TILE_WIDTH) { + av_log(avctx, AV_LOG_ERROR, "Invalid tile sizing: AV2 does " + "not allow tiles of width greater than %d.\n", + MAX_TILE_WIDTH); + return AVERROR(EINVAL); + } + + ctx->superblock_size = AVM_SUPERBLOCK_SIZE_DYNAMIC; + + if (ctx->tile_cols == 1 && ctx->tile_rows == 1) { + av_log(avctx, AV_LOG_DEBUG, "Using a single tile.\n"); + return 0; + } + + sb_256x256_possible = + (avctx->width + 255) / 256 >= ctx->tile_cols && + (avctx->height + 255) / 256 >= ctx->tile_rows; + + sb_128x128_possible = + (avctx->width + 127) / 128 >= ctx->tile_cols && + (avctx->height + 127) / 128 >= ctx->tile_rows; + + ctx->tile_cols_log2 = ctx->tile_cols == 1 ? 0 : + av_log2(ctx->tile_cols - 1) + 1; + ctx->tile_rows_log2 = ctx->tile_rows == 1 ? 0 : + av_log2(ctx->tile_rows - 1) + 1; + + uniform_cols = count_uniform_tiling(avctx->width, + 64, ctx->tile_cols_log2); + uniform_rows = count_uniform_tiling(avctx->height, + 64, ctx->tile_rows_log2); + av_log(avctx, AV_LOG_DEBUG, "Uniform with 64x64 superblocks " + "-> %dx%d tiles.\n", uniform_cols, uniform_rows); + uniform_64x64_possible = uniform_cols == ctx->tile_cols && + uniform_rows == ctx->tile_rows; + + if (sb_256x256_possible) { + uniform_cols = count_uniform_tiling(avctx->width, + 256, ctx->tile_cols_log2); + uniform_rows = count_uniform_tiling(avctx->height, + 256, ctx->tile_rows_log2); + av_log(avctx, AV_LOG_DEBUG, "Uniform with 256x256 superblocks " + "-> %dx%d tiles.\n", uniform_cols, uniform_rows); + uniform_256x256_possible = uniform_cols == ctx->tile_cols && + uniform_rows == ctx->tile_rows; + } else { + av_log(avctx, AV_LOG_DEBUG, "256x256 superblocks not possible.\n"); + uniform_256x256_possible = 0; + } + + if (sb_128x128_possible) { + uniform_cols = count_uniform_tiling(avctx->width, + 128, ctx->tile_cols_log2); + uniform_rows = count_uniform_tiling(avctx->height, + 128, ctx->tile_rows_log2); + av_log(avctx, AV_LOG_DEBUG, "Uniform with 128x128 superblocks " + "-> %dx%d tiles.\n", uniform_cols, uniform_rows); + uniform_128x128_possible = uniform_cols == ctx->tile_cols && + uniform_rows == ctx->tile_rows; + } else { + av_log(avctx, AV_LOG_DEBUG, "128x128 superblocks not possible.\n"); + uniform_128x128_possible = 0; + } + + ctx->uniform_tiles = 1; + if (uniform_64x64_possible && uniform_128x128_possible) { + av_log(avctx, AV_LOG_DEBUG, "Using uniform tiling with dynamic " + "superblocks (tile_cols_log2 = %d, tile_rows_log2 = %d).\n", + ctx->tile_cols_log2, ctx->tile_rows_log2); + return 0; + } + if (uniform_64x64_possible && !sb_128x128_possible) { + av_log(avctx, AV_LOG_DEBUG, "Using uniform tiling with 64x64 " + "superblocks (tile_cols_log2 = %d, tile_rows_log2 = %d).\n", + ctx->tile_cols_log2, ctx->tile_rows_log2); + ctx->superblock_size = AVM_SUPERBLOCK_SIZE_64X64; + return 0; + } + if (uniform_128x128_possible && !sb_256x256_possible) { + av_log(avctx, AV_LOG_DEBUG, "Using uniform tiling with 128x128 " + "superblocks (tile_cols_log2 = %d, tile_rows_log2 = %d).\n", + ctx->tile_cols_log2, ctx->tile_rows_log2); + ctx->superblock_size = AVM_SUPERBLOCK_SIZE_128X128; + return 0; + } + if (uniform_256x256_possible) { + av_log(avctx, AV_LOG_DEBUG, "Using uniform tiling with 256x256 " + "superblocks (tile_cols_log2 = %d, tile_rows_log2 = %d).\n", + ctx->tile_cols_log2, ctx->tile_rows_log2); + ctx->superblock_size = AVM_SUPERBLOCK_SIZE_256X256; + return 0; + } + ctx->uniform_tiles = 0; + + if (sb_256x256_possible) { + sb_size = 256; + ctx->superblock_size = AVM_SUPERBLOCK_SIZE_256X256; + } else if (sb_128x128_possible) { + sb_size = 128; + ctx->superblock_size = AVM_SUPERBLOCK_SIZE_128X128; + } else { + sb_size = 64; + ctx->superblock_size = AVM_SUPERBLOCK_SIZE_64X64; + } + av_log(avctx, AV_LOG_DEBUG, "Using fixed tiling with %dx%d " + "superblocks (tile_cols = %d, tile_rows = %d).\n", + sb_size, sb_size, ctx->tile_cols, ctx->tile_rows); + + enccfg->tile_width_count = ctx->tile_cols; + enccfg->tile_height_count = ctx->tile_rows; + + sb_width = (avctx->width + sb_size - 1) / sb_size; + sb_height = (avctx->height + sb_size - 1) / sb_size; + + tile_size = sb_width / ctx->tile_cols; + rounding = sb_width % ctx->tile_cols; + for (i = 0; i < ctx->tile_cols; i++) { + enccfg->tile_widths[i] = tile_size + + (i < rounding / 2 || + i > ctx->tile_cols - 1 - (rounding + 1) / 2); + } + + tile_size = sb_height / ctx->tile_rows; + rounding = sb_height % ctx->tile_rows; + for (i = 0; i < ctx->tile_rows; i++) { + enccfg->tile_heights[i] = tile_size + + (i < rounding / 2 || + i > ctx->tile_rows - 1 - (rounding + 1) / 2); + } + + return 0; +} + + +static const struct { + int avm_enum; + unsigned offset; +} option_map[] = { + { AVME_SET_ENABLEAUTOALTREF, OFFSET(auto_alt_ref) }, + { AVME_SET_ARNR_STRENGTH, OFFSET(arnr_strength) }, + { AV2E_SET_ENABLE_CDEF, OFFSET(enable_cdef) }, + { AV2E_SET_ENABLE_RESTORATION, OFFSET(enable_restoration) }, + { AV2E_SET_ENABLE_RECT_PARTITIONS, OFFSET(enable_rect_partitions) }, + { AV2E_SET_ENABLE_ANGLE_DELTA, OFFSET(enable_angle_delta) }, + { AV2E_SET_ENABLE_CFL_INTRA, OFFSET(enable_cfl_intra) }, + { AV2E_SET_ENABLE_INTRA_EDGE_FILTER, OFFSET(enable_intra_edge_filter) }, + { AV2E_SET_ENABLE_PAETH_INTRA, OFFSET(enable_paeth_intra) }, + { AV2E_SET_ENABLE_SMOOTH_INTRA, OFFSET(enable_smooth_intra) }, + { AV2E_SET_ENABLE_PALETTE, OFFSET(enable_palette) }, + { AV2E_SET_ENABLE_FLIP_IDTX, OFFSET(enable_flip_idtx) }, + { AV2E_SET_INTRA_DCT_ONLY, OFFSET(use_intra_dct_only) }, + { AV2E_SET_INTER_DCT_ONLY, OFFSET(use_inter_dct_only) }, + { AV2E_SET_INTRA_DEFAULT_TX_ONLY, OFFSET(use_intra_default_tx_only) }, + { AV2E_SET_REDUCED_TX_TYPE_SET, OFFSET(reduced_tx_type_set) }, + { AV2E_SET_ENABLE_REF_FRAME_MVS, OFFSET(enable_ref_frame_mvs) }, + { AV2E_SET_ENABLE_DIFF_WTD_COMP, OFFSET(enable_diff_wtd_comp) }, + { AV2E_SET_ENABLE_INTERINTER_WEDGE, OFFSET(enable_interinter_wedge) }, + { AV2E_SET_ENABLE_MASKED_COMP, OFFSET(enable_masked_comp) }, + { AV2E_SET_ENABLE_INTERINTRA_COMP, OFFSET(enable_interintra_comp) }, + { AV2E_SET_ENABLE_INTERINTRA_WEDGE, OFFSET(enable_interintra_wedge) }, + { AV2E_SET_ENABLE_ONESIDED_COMP, OFFSET(enable_onesided_comp) }, + { AV2E_SET_ENABLE_SMOOTH_INTERINTRA, OFFSET(enable_smooth_interintra) }, +}; + +static av_cold int avm_init(AVCodecContext *avctx, + const struct avm_codec_iface *iface) +{ + AVMContext *ctx = avctx->priv_data; + const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(avctx->pix_fmt); + struct avm_codec_enc_cfg enccfg = { 0 }; + avm_codec_flags_t flags = + (avctx->flags & AV_CODEC_FLAG_PSNR) ? AVM_CODEC_USE_PSNR : 0; + int res; + avm_img_fmt_t img_fmt; + avm_codec_caps_t codec_caps = avm_codec_get_caps(iface); + + av_log(avctx, AV_LOG_INFO, "%s\n", avm_codec_version_str()); + av_log(avctx, AV_LOG_VERBOSE, "%s\n", avm_codec_build_config()); + + if ((res = avm_codec_enc_config_default(iface, &enccfg, ctx->usage)) != AVM_CODEC_OK) { + av_log(avctx, AV_LOG_ERROR, "Failed to get config: %s\n", + avm_codec_err_to_string(res)); + return AVERROR(EINVAL); + } + + if (set_pix_fmt(avctx, codec_caps, &enccfg, &flags, &img_fmt)) + return AVERROR(EINVAL); + + if(!avctx->bit_rate) + if(avctx->rc_max_rate || avctx->rc_buffer_size || avctx->rc_initial_buffer_occupancy) { + av_log( avctx, AV_LOG_ERROR, "Rate control parameters set without a bitrate\n"); + return AVERROR(EINVAL); + } + + dump_enc_cfg(avctx, &enccfg, AV_LOG_DEBUG); + + enccfg.g_w = avctx->width; + enccfg.g_h = avctx->height; + enccfg.g_timebase.num = avctx->time_base.num; + enccfg.g_timebase.den = avctx->time_base.den; + enccfg.g_threads = + FFMIN(avctx->thread_count ? avctx->thread_count : av_cpu_count(), 64); + + if (ctx->lag_in_frames >= 0) + enccfg.g_lag_in_frames = ctx->lag_in_frames; + + if (avctx->flags & AV_CODEC_FLAG_PASS1) + enccfg.g_pass = AVM_RC_FIRST_PASS; + else if (avctx->flags & AV_CODEC_FLAG_PASS2) + enccfg.g_pass = AVM_RC_LAST_PASS; + else + enccfg.g_pass = AVM_RC_ONE_PASS; + + if (avctx->rc_min_rate == avctx->rc_max_rate && + avctx->rc_min_rate == avctx->bit_rate && avctx->bit_rate) { + enccfg.rc_end_usage = AVM_CBR; + } else if (ctx->qp >= 0) { + enccfg.rc_end_usage = AVM_CQ; + if (!avctx->bit_rate) + enccfg.rc_end_usage = AVM_Q; + } + + if (avctx->bit_rate) { + enccfg.rc_target_bitrate = av_rescale_rnd(avctx->bit_rate, 1, 1000, + AV_ROUND_NEAR_INF); + } else if (enccfg.rc_end_usage != AVM_Q) { + enccfg.rc_end_usage = AVM_Q; + ctx->qp = 40; + av_log(avctx, AV_LOG_WARNING, + "Neither bitrate nor constrained quality specified, using default QP of %d\n", + ctx->qp); + } + + if (avctx->qmin >= 0) + enccfg.rc_min_quantizer = avctx->qmin; + if (avctx->qmax >= 0) { + enccfg.rc_max_quantizer = avctx->qmax; + } else if (!ctx->qp) { + enccfg.rc_max_quantizer = 0; + } + + if (enccfg.rc_end_usage == AVM_CQ || enccfg.rc_end_usage == AVM_Q) { + if (ctx->qp < enccfg.rc_min_quantizer || ctx->qp > enccfg.rc_max_quantizer) { + av_log(avctx, AV_LOG_ERROR, + "CQ level %d must be between minimum and maximum quantizer value (%d-%d)\n", + ctx->qp, enccfg.rc_min_quantizer, enccfg.rc_max_quantizer); + return AVERROR(EINVAL); + } + } + + //enccfg.rc_dropframe_thresh = 0; + + // 0-100 (0 => CBR, 100 => VBR) + if (ctx->minsection_pct >= 0) + enccfg.rc_2pass_vbr_minsection_pct = ctx->minsection_pct; + else if (avctx->bit_rate) + enccfg.rc_2pass_vbr_minsection_pct = + avctx->rc_min_rate * 100LL / avctx->bit_rate; + if (ctx->maxsection_pct >= 0) + enccfg.rc_2pass_vbr_maxsection_pct = ctx->maxsection_pct; + else if (avctx->rc_max_rate) + enccfg.rc_2pass_vbr_maxsection_pct = + avctx->rc_max_rate * 100LL / avctx->bit_rate; + + if (avctx->rc_buffer_size) + enccfg.rc_buf_sz = + avctx->rc_buffer_size * 1000LL / avctx->bit_rate; + if (avctx->rc_initial_buffer_occupancy) + enccfg.rc_buf_initial_sz = + avctx->rc_initial_buffer_occupancy * 1000LL / avctx->bit_rate; + enccfg.rc_buf_optimal_sz = enccfg.rc_buf_sz * 5 / 6; + + if (ctx->rc_undershoot_pct >= 0) + enccfg.rc_undershoot_pct = ctx->rc_undershoot_pct; + if (ctx->rc_overshoot_pct >= 0) + enccfg.rc_overshoot_pct = ctx->rc_overshoot_pct; + + // _enc_init() will balk if kf_min_dist differs from max w/AVM_KF_AUTO + if (avctx->keyint_min >= 0 && avctx->keyint_min == avctx->gop_size) + enccfg.kf_min_dist = avctx->keyint_min; + if (avctx->gop_size >= 0) + enccfg.kf_max_dist = avctx->gop_size; + + if (enccfg.g_pass == AVM_RC_FIRST_PASS) + enccfg.g_lag_in_frames = 0; + else if (enccfg.g_pass == AVM_RC_LAST_PASS) { + int decode_size, ret; + + if (!avctx->stats_in) { + av_log(avctx, AV_LOG_ERROR, "No stats file for second pass\n"); + return AVERROR_INVALIDDATA; + } + + ctx->twopass_stats.sz = strlen(avctx->stats_in) * 3 / 4; + ret = av_reallocp(&ctx->twopass_stats.buf, ctx->twopass_stats.sz); + if (ret < 0) { + av_log(avctx, AV_LOG_ERROR, + "Stat buffer alloc (%zu bytes) failed\n", + ctx->twopass_stats.sz); + ctx->twopass_stats.sz = 0; + return ret; + } + decode_size = av_base64_decode(ctx->twopass_stats.buf, avctx->stats_in, + ctx->twopass_stats.sz); + if (decode_size < 0) { + av_log(avctx, AV_LOG_ERROR, "Stat buffer decode failed\n"); + return AVERROR_INVALIDDATA; + } + + ctx->twopass_stats.sz = decode_size; + } + + /* 0-3: For non-zero values the encoder increasingly optimizes for reduced + * complexity playback on low powered devices at the expense of encode + * quality. */ + if (avctx->profile != AV_PROFILE_UNKNOWN) + enccfg.g_profile = avctx->profile; + + enccfg.g_error_resilient = 0; + + res = choose_tiling(avctx, &enccfg); + if (res < 0) + return res; + + if (ctx->still_picture) { + // Set the maximum number of frames to 1. This will let libavm set + // still_picture and reduced_still_picture_header to 1 in the Sequence + // Header as required by AVIF still images. + enccfg.g_limit = 1; + // Reduce memory usage for still images. + enccfg.g_lag_in_frames = 0; + // All frames will be key frames. + enccfg.kf_max_dist = AVM_KF_DISABLED; + enccfg.kf_mode = AVM_KF_DISABLED; + } + + /* Construct Encoder Context */ + res = avm_codec_enc_init(&ctx->encoder, iface, &enccfg, flags); + if (res != AVM_CODEC_OK) { + dump_enc_cfg(avctx, &enccfg, AV_LOG_WARNING); + log_encoder_error(avctx, "Failed to initialize encoder"); + return AVERROR(EINVAL); + } + dump_enc_cfg(avctx, &enccfg, AV_LOG_DEBUG); + + // codec control failures are currently treated only as warnings + av_log(avctx, AV_LOG_DEBUG, "avm_codec_control\n"); + codecctl_int(avctx, AVME_SET_CPUUSED, ctx->cpu_used); + + for (size_t i = 0; i < FF_ARRAY_ELEMS(option_map); ++i) { + int val = *(int*)((char*)ctx + option_map[i].offset); + if (val >= 0) + codecctl_int(avctx, option_map[i].avm_enum, val); + } + codecctl_int(avctx, AVME_SET_STATIC_THRESHOLD, ctx->static_thresh); + if ((enccfg.rc_end_usage == AVM_CQ) || (enccfg.rc_end_usage == AVM_Q)) + codecctl_int(avctx, AVME_SET_QP, ctx->qp); + if (ctx->tune >= 0) + codecctl_int(avctx, AVME_SET_TUNING, ctx->tune); + + if (desc->flags & AV_PIX_FMT_FLAG_RGB) { + codecctl_int(avctx, AV2E_SET_COLOR_PRIMARIES, AVCOL_PRI_BT709); + codecctl_int(avctx, AV2E_SET_MATRIX_COEFFICIENTS, AVCOL_SPC_RGB); + codecctl_int(avctx, AV2E_SET_TRANSFER_CHARACTERISTICS, AVCOL_TRC_IEC61966_2_1); + } else { + codecctl_int(avctx, AV2E_SET_COLOR_PRIMARIES, avctx->color_primaries); + codecctl_int(avctx, AV2E_SET_MATRIX_COEFFICIENTS, avctx->colorspace); + codecctl_int(avctx, AV2E_SET_TRANSFER_CHARACTERISTICS, avctx->color_trc); + } + if (ctx->aq_mode >= 0) + codecctl_int(avctx, AV2E_SET_AQ_MODE, ctx->aq_mode); + if (ctx->frame_parallel >= 0) + codecctl_int(avctx, AV2E_SET_FRAME_PARALLEL_DECODING, ctx->frame_parallel); + set_color_range(avctx); + + codecctl_int(avctx, AV2E_SET_SUPERBLOCK_SIZE, ctx->superblock_size); + if (ctx->uniform_tiles) { + codecctl_int(avctx, AV2E_SET_TILE_COLUMNS, ctx->tile_cols_log2); + codecctl_int(avctx, AV2E_SET_TILE_ROWS, ctx->tile_rows_log2); + } + + if (ctx->denoise_noise_level >= 0) + codecctl_int(avctx, AV2E_SET_DENOISE_NOISE_LEVEL, ctx->denoise_noise_level); + if (ctx->denoise_block_size >= 0) + codecctl_int(avctx, AV2E_SET_DENOISE_BLOCK_SIZE, ctx->denoise_block_size); + if (ctx->enable_global_motion >= 0) + codecctl_int(avctx, AV2E_SET_ENABLE_GLOBAL_MOTION, ctx->enable_global_motion); + if (avctx->refs >= 3) { + codecctl_int(avctx, AV2E_SET_MAX_REFERENCE_FRAMES, avctx->refs); + } + if (ctx->row_mt >= 0) + codecctl_int(avctx, AV2E_SET_ROW_MT, ctx->row_mt); + if (ctx->enable_intrabc >= 0) + codecctl_int(avctx, AV2E_SET_ENABLE_INTRABC, ctx->enable_intrabc); + + { + const AVDictionaryEntry *en = NULL; + + while ((en = av_dict_iterate(ctx->avm_params, en))) { + int ret = avm_codec_set_option(&ctx->encoder, en->key, en->value); + if (ret != AVM_CODEC_OK) { + log_encoder_error(avctx, en->key); + return AVERROR_EXTERNAL; + } + } + } + + // provide dummy value to initialize wrapper, values will be updated each _encode() + avm_img_wrap(&ctx->rawimg, img_fmt, avctx->width, avctx->height, 1, + (unsigned char*)1); + + if (enccfg.g_bit_depth >= 10) + ctx->rawimg.bit_depth = enccfg.g_bit_depth; + + ctx->dovi.logctx = avctx; + if ((res = ff_dovi_configure(&ctx->dovi, avctx)) < 0) + return res; + + AVCPBProperties *cpb_props = ff_encode_add_cpb_side_data(avctx); + if (!cpb_props) + return AVERROR(ENOMEM); + + if (enccfg.rc_end_usage == AVM_CBR || + enccfg.g_pass != AVM_RC_ONE_PASS) { + cpb_props->max_bitrate = avctx->rc_max_rate; + cpb_props->min_bitrate = avctx->rc_min_rate; + cpb_props->avg_bitrate = avctx->bit_rate; + } + cpb_props->buffer_size = avctx->rc_buffer_size; + + return 0; +} + +static inline void cx_pktcpy(AVMContext *ctx, + struct FrameListData *dst, + const struct avm_codec_cx_pkt *src) +{ + dst->pts = src->data.frame.pts; + dst->duration = src->data.frame.duration; + dst->flags = src->data.frame.flags; + dst->sz = src->data.frame.sz; + dst->buf = src->data.frame.buf; + dst->frame_number = ++ctx->frame_number; + dst->have_sse = ctx->have_sse; + if (ctx->have_sse) { + /* associate last-seen SSE to the frame. */ + /* Transfers ownership from ctx to dst. */ + memcpy(dst->sse, ctx->sse, sizeof(dst->sse)); + ctx->have_sse = 0; + } +} + +/** + * Store coded frame information in format suitable for return from encode2(). + * + * Write information from @a cx_frame to @a pkt + * @return packet data size on success + * @return a negative AVERROR on error + */ +static int storeframe(AVCodecContext *avctx, struct FrameListData *cx_frame, + AVPacket *pkt) +{ + AVMContext *ctx = avctx->priv_data; + enum AVPictureType pict_type; + int ret = ff_get_encode_buffer(avctx, pkt, cx_frame->sz, 0); + if (ret < 0) { + av_log(avctx, AV_LOG_ERROR, + "Error getting output packet of size %zu.\n", cx_frame->sz); + return ret; + } + memcpy(pkt->data, cx_frame->buf, pkt->size); + pkt->pts = pkt->dts = cx_frame->pts; + pkt->duration = cx_frame->duration; + + if (!!(cx_frame->flags & AVM_FRAME_IS_KEY)) { + pkt->flags |= AV_PKT_FLAG_KEY; + pict_type = AV_PICTURE_TYPE_I; + } else if (cx_frame->flags & AVM_FRAME_IS_INTRAONLY) { + pict_type = AV_PICTURE_TYPE_I; + } else { + pict_type = AV_PICTURE_TYPE_P; + } + + ff_encode_add_stats_side_data(pkt, 0, cx_frame->sse + 1, + cx_frame->have_sse ? 3 : 0, pict_type); + + if (cx_frame->have_sse) { + int i; + for (i = 0; i < 3; ++i) { + avctx->error[i] += cx_frame->sse[i + 1]; + } + cx_frame->have_sse = 0; + } + + if (avctx->flags & AV_CODEC_FLAG_GLOBAL_HEADER) { + ret = av_bsf_send_packet(ctx->bsf, pkt); + if (ret < 0) { + av_log(avctx, AV_LOG_ERROR, "extract_extradata filter " + "failed to send input packet\n"); + return ret; + } + ret = av_bsf_receive_packet(ctx->bsf, pkt); + + if (ret < 0) { + av_log(avctx, AV_LOG_ERROR, "extract_extradata filter " + "failed to receive output packet\n"); + return ret; + } + } + return pkt->size; +} + +/** + * Queue multiple output frames from the encoder, returning the front-most. + * In cases where avm_codec_get_cx_data() returns more than 1 frame append + * the frame queue. Return the head frame if available. + * @return Stored frame size + * @return AVERROR(EINVAL) on output size error + * @return AVERROR(ENOMEM) on coded frame queue data allocation error + */ +static int queue_frames(AVCodecContext *avctx, AVPacket *pkt_out) +{ + AVMContext *ctx = avctx->priv_data; + const struct avm_codec_cx_pkt *pkt; + const void *iter = NULL; + int size = 0; + + if (ctx->coded_frame_list) { + struct FrameListData *cx_frame = ctx->coded_frame_list; + /* return the leading frame if we've already begun queueing */ + size = storeframe(avctx, cx_frame, pkt_out); + if (size < 0) + return size; + ctx->coded_frame_list = cx_frame->next; + free_coded_frame(cx_frame); + } + + /* consume all available output from the encoder before returning. buffers + * are only good through the next avm_codec call */ + while ((pkt = avm_codec_get_cx_data(&ctx->encoder, &iter))) { + switch (pkt->kind) { + case AVM_CODEC_CX_FRAME_PKT: + if (!size) { + struct FrameListData cx_frame; + + /* avoid storing the frame when the list is empty and we haven't yet + * provided a frame for output */ + av_assert0(!ctx->coded_frame_list); + cx_pktcpy(ctx, &cx_frame, pkt); + size = storeframe(avctx, &cx_frame, pkt_out); + if (size < 0) + return size; + } else { + struct FrameListData *cx_frame = + av_malloc(sizeof(struct FrameListData)); + + if (!cx_frame) { + av_log(avctx, AV_LOG_ERROR, + "Frame queue element alloc failed\n"); + return AVERROR(ENOMEM); + } + cx_pktcpy(ctx, cx_frame, pkt); + cx_frame->buf = av_malloc(cx_frame->sz); + + if (!cx_frame->buf) { + av_log(avctx, AV_LOG_ERROR, + "Data buffer alloc (%zu bytes) failed\n", + cx_frame->sz); + av_freep(&cx_frame); + return AVERROR(ENOMEM); + } + memcpy(cx_frame->buf, pkt->data.frame.buf, pkt->data.frame.sz); + coded_frame_add(&ctx->coded_frame_list, cx_frame); + } + break; + case AVM_CODEC_STATS_PKT: + { + struct avm_fixed_buf *stats = &ctx->twopass_stats; + uint8_t *tmp = av_fast_realloc(stats->buf, + &ctx->twopass_stats_size, + stats->sz + + pkt->data.twopass_stats.sz); + if (!tmp) { + av_freep(&stats->buf); + stats->sz = 0; + av_log(avctx, AV_LOG_ERROR, "Stat buffer realloc failed\n"); + return AVERROR(ENOMEM); + } + stats->buf = tmp; + memcpy((uint8_t *)stats->buf + stats->sz, + pkt->data.twopass_stats.buf, pkt->data.twopass_stats.sz); + stats->sz += pkt->data.twopass_stats.sz; + break; + } + case AVM_CODEC_PSNR_PKT: + { + av_assert0(!ctx->have_sse); + ctx->sse[0] = pkt->data.psnr.sse[0]; + ctx->sse[1] = pkt->data.psnr.sse[1]; + ctx->sse[2] = pkt->data.psnr.sse[2]; + ctx->sse[3] = pkt->data.psnr.sse[3]; + ctx->have_sse = 1; + break; + } + case AVM_CODEC_CUSTOM_PKT: + // ignore unsupported/unrecognized packet types + break; + } + } + + return size; +} + +static enum AVPixelFormat avmfmt_to_pixfmt(struct avm_image *img) +{ + switch (img->fmt) { + case AVM_IMG_FMT_I420: + case AVM_IMG_FMT_I42016: + if (img->bit_depth == 8) + return img->monochrome ? AV_PIX_FMT_GRAY8 : AV_PIX_FMT_YUV420P; + else if (img->bit_depth == 10) + return img->monochrome ? AV_PIX_FMT_GRAY10 : AV_PIX_FMT_YUV420P10; + else + return img->monochrome ? AV_PIX_FMT_GRAY12 : AV_PIX_FMT_YUV420P12; + case AVM_IMG_FMT_I422: + case AVM_IMG_FMT_I42216: + if (img->bit_depth == 8) + return AV_PIX_FMT_YUV422P; + else if (img->bit_depth == 10) + return AV_PIX_FMT_YUV422P10; + else + return AV_PIX_FMT_YUV422P12; + case AVM_IMG_FMT_I444: + case AVM_IMG_FMT_I44416: + if (img->bit_depth == 8) + return AV_PIX_FMT_YUV444P; + else if (img->bit_depth == 10) + return AV_PIX_FMT_YUV444P10; + else + return AV_PIX_FMT_YUV444P12; + }; + return AV_PIX_FMT_NONE; +} + +static int avm_encode(AVCodecContext *avctx, AVPacket *pkt, + const AVFrame *frame, int *got_packet) +{ + AVMContext *ctx = avctx->priv_data; + struct avm_image *rawimg = NULL; + int64_t timestamp = 0; + unsigned long duration = 0; + int res, coded_size; + avm_enc_frame_flags_t flags = 0; + AVFrameSideData *sd; + + if (frame) { + rawimg = &ctx->rawimg; + avm_img_remove_metadata(rawimg); + rawimg->planes[AVM_PLANE_Y] = frame->data[0]; + rawimg->planes[AVM_PLANE_U] = frame->data[1]; + rawimg->planes[AVM_PLANE_V] = frame->data[2]; + rawimg->stride[AVM_PLANE_Y] = frame->linesize[0]; + rawimg->stride[AVM_PLANE_U] = frame->linesize[1]; + rawimg->stride[AVM_PLANE_V] = frame->linesize[2]; + timestamp = frame->pts; + + if (frame->duration > ULONG_MAX) { + av_log(avctx, AV_LOG_WARNING, + "Frame duration too large: %"PRId64"\n", frame->duration); + } else if (frame->duration) + duration = frame->duration; + else if (avctx->framerate.num > 0 && avctx->framerate.den > 0) + duration = av_rescale_q(1, av_inv_q(avctx->framerate), avctx->time_base); + else { + duration = 1; + } + + switch (frame->color_range) { + case AVCOL_RANGE_MPEG: + rawimg->range = AVM_CR_STUDIO_RANGE; + break; + case AVCOL_RANGE_JPEG: + rawimg->range = AVM_CR_FULL_RANGE; + break; + } + + avm_img_remove_metadata(rawimg); + sd = av_frame_get_side_data(frame, AV_FRAME_DATA_DOVI_METADATA); + if (ctx->dovi.cfg.dv_profile && sd) { + const AVDOVIMetadata *metadata = (const AVDOVIMetadata *)sd->data; + uint8_t *t35; + int size; + if ((res = ff_dovi_rpu_generate(&ctx->dovi, metadata, FF_DOVI_WRAP_T35, + &t35, &size)) < 0) + return res; + res = avm_img_add_metadata(rawimg, OBU_METADATA_TYPE_ITUT_T35, + t35, size, AVM_MIF_ANY_FRAME); + av_free(t35); + if (res != AVM_CODEC_OK) + return AVERROR(ENOMEM); + } else if (ctx->dovi.cfg.dv_profile) { + av_log(avctx, AV_LOG_ERROR, "Dolby Vision enabled, but received frame " + "without AV_FRAME_DATA_DOVI_METADATA\n"); + return AVERROR_INVALIDDATA; + } + + if (frame->pict_type == AV_PICTURE_TYPE_I) + flags |= AVM_EFLAG_FORCE_KF; + + res = add_hdr_plus(avctx, rawimg, frame); + if (res < 0) + return res; + + res = add_hdr_smpte2094_app5(avctx, rawimg, frame); + if (res < 0) + return res; + } + + res = avm_codec_encode(&ctx->encoder, rawimg, timestamp, duration, flags); + if (res != AVM_CODEC_OK) { + log_encoder_error(avctx, "Error encoding frame"); + return AVERROR_INVALIDDATA; + } + coded_size = queue_frames(avctx, pkt); + if (coded_size < 0) + return coded_size; + + if (!frame && avctx->flags & AV_CODEC_FLAG_PASS1) { + size_t b64_size = AV_BASE64_SIZE(ctx->twopass_stats.sz); + + avctx->stats_out = av_malloc(b64_size); + if (!avctx->stats_out) { + av_log(avctx, AV_LOG_ERROR, "Stat buffer alloc (%zu bytes) failed\n", + b64_size); + return AVERROR(ENOMEM); + } + av_base64_encode(avctx->stats_out, b64_size, ctx->twopass_stats.buf, + ctx->twopass_stats.sz); + } + + *got_packet = !!coded_size; + + if (*got_packet && avctx->flags & AV_CODEC_FLAG_RECON_FRAME) { + AVCodecInternal *avci = avctx->internal; + struct avm_image img; + + av_frame_unref(avci->recon_frame); + + res = codecctl_imgp(avctx, AV2_GET_NEW_FRAME_IMAGE, &img); + if (res < 0) + return res; + + avci->recon_frame->format = avmfmt_to_pixfmt(&img); + if (avci->recon_frame->format == AV_PIX_FMT_NONE) { + av_log(ctx, AV_LOG_ERROR, + "Unhandled reconstructed frame colorspace: %d\n", + img.fmt); + return AVERROR(ENOSYS); + } + + avci->recon_frame->width = img.d_w; + avci->recon_frame->height = img.d_h; + + res = av_frame_get_buffer(avci->recon_frame, 0); + if (res < 0) + return res; + + if ((img.fmt & AVM_IMG_FMT_HIGHBITDEPTH) && img.bit_depth == 8) + ff_avm_image_copy_16_to_8(avci->recon_frame, &img); + else { + const uint8_t *planes[4] = { img.planes[0], img.planes[1], img.planes[2] }; + const int stride[4] = { img.stride[0], img.stride[1], img.stride[2] }; + + av_image_copy(avci->recon_frame->data, avci->recon_frame->linesize, planes, + stride, avci->recon_frame->format, img.d_w, img.d_h); + } + } + + return 0; +} + +static const enum AVPixelFormat av2_pix_fmts_highbd_with_gray[] = { + AV_PIX_FMT_YUV420P, + AV_PIX_FMT_YUV422P, + AV_PIX_FMT_YUV444P, + AV_PIX_FMT_GBRP, + AV_PIX_FMT_YUV420P10, + AV_PIX_FMT_YUV422P10, + AV_PIX_FMT_YUV444P10, + AV_PIX_FMT_YUV420P12, + AV_PIX_FMT_YUV422P12, + AV_PIX_FMT_YUV444P12, + AV_PIX_FMT_GBRP10, + AV_PIX_FMT_GBRP12, + AV_PIX_FMT_GRAY8, + AV_PIX_FMT_GRAY10, + AV_PIX_FMT_GRAY12, + AV_PIX_FMT_NONE +}; + +static av_cold int av2_init(AVCodecContext *avctx) +{ + return avm_init(avctx, avm_codec_av2_cx()); +} + +#define VE AV_OPT_FLAG_VIDEO_PARAM | AV_OPT_FLAG_ENCODING_PARAM +static const AVOption options[] = { + { "cpu-used", "Quality/Speed ratio modifier", OFFSET(cpu_used), AV_OPT_TYPE_INT, {.i64 = 1}, 0, 8, VE}, + { "auto-alt-ref", "Enable use of alternate reference " + "frames (2-pass only)", OFFSET(auto_alt_ref), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 2, VE}, + { "lag-in-frames", "Number of frames to look ahead at for " + "alternate reference frame selection", OFFSET(lag_in_frames), AV_OPT_TYPE_INT, {.i64 = -1}, -1, INT_MAX, VE}, + { "arnr-strength", "altref noise reduction filter strength", OFFSET(arnr_strength), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 6, VE}, + { "aq-mode", "adaptive quantization mode", OFFSET(aq_mode), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 4, VE, .unit = "aq_mode"}, + { "none", "Aq not used", 0, AV_OPT_TYPE_CONST, {.i64 = 0}, 0, 0, VE, .unit = "aq_mode"}, + { "variance", "Variance based Aq", 0, AV_OPT_TYPE_CONST, {.i64 = 1}, 0, 0, VE, .unit = "aq_mode"}, + { "complexity", "Complexity based Aq", 0, AV_OPT_TYPE_CONST, {.i64 = 2}, 0, 0, VE, .unit = "aq_mode"}, + { "cyclic", "Cyclic Refresh Aq", 0, AV_OPT_TYPE_CONST, {.i64 = 3}, 0, 0, VE, .unit = "aq_mode"}, + { "qp", "Select the quality for constant quality mode", offsetof(AVMContext, qp), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 255, VE }, + { "static-thresh", "A change threshold on blocks below which they will be skipped by the encoder", OFFSET(static_thresh), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, INT_MAX, VE }, + { "denoise-noise-level", "Amount of noise to be removed", OFFSET(denoise_noise_level), AV_OPT_TYPE_INT, {.i64 = -1}, -1, INT_MAX, VE}, + { "denoise-block-size", "Denoise block size ", OFFSET(denoise_block_size), AV_OPT_TYPE_INT, {.i64 = -1}, -1, INT_MAX, VE}, + { "undershoot-pct", "Datarate undershoot (min) target (%)", OFFSET(rc_undershoot_pct), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 100, VE}, + { "overshoot-pct", "Datarate overshoot (max) target (%)", OFFSET(rc_overshoot_pct), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 1000, VE}, + { "minsection-pct", "GOP min bitrate (% of target)", OFFSET(minsection_pct), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 100, VE}, + { "maxsection-pct", "GOP max bitrate (% of target)", OFFSET(maxsection_pct), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 5000, VE}, + { "frame-parallel", "Enable frame parallel decodability features", OFFSET(frame_parallel), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "tiles", "Tile columns x rows", OFFSET(tile_cols), AV_OPT_TYPE_IMAGE_SIZE, { .str = NULL }, 0, 0, VE }, + { "tile-columns", "Log2 of number of tile columns to use", OFFSET(tile_cols_log2), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 6, VE}, + { "tile-rows", "Log2 of number of tile rows to use", OFFSET(tile_rows_log2), AV_OPT_TYPE_INT, {.i64 = -1}, -1, 6, VE}, + { "row-mt", "Enable row based multi-threading", OFFSET(row_mt), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-cdef", "Enable CDEF filtering", OFFSET(enable_cdef), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-global-motion", "Enable global motion", OFFSET(enable_global_motion), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-intrabc", "Enable intra block copy prediction mode", OFFSET(enable_intrabc), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-restoration", "Enable Loop Restoration filtering", OFFSET(enable_restoration), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "usage", "Quality and compression efficiency vs speed trade-off", OFFSET(usage), AV_OPT_TYPE_INT, {.i64 = 0}, 0, INT_MAX, VE, .unit = "usage"}, + { "good", "Good quality", 0, AV_OPT_TYPE_CONST, {.i64 = 0 /* AVM_USAGE_GOOD_QUALITY */}, 0, 0, VE, .unit = "usage"}, + { "tune", "The metric that the encoder tunes for. Automatically chosen by the encoder by default", OFFSET(tune), AV_OPT_TYPE_INT, {.i64 = -1}, -1, AVM_TUNE_SSIM, VE, .unit = "tune"}, + { "psnr", NULL, 0, AV_OPT_TYPE_CONST, {.i64 = AVM_TUNE_PSNR}, 0, 0, VE, .unit = "tune"}, + { "ssim", NULL, 0, AV_OPT_TYPE_CONST, {.i64 = AVM_TUNE_SSIM}, 0, 0, VE, .unit = "tune"}, + { "still-picture", "Encode in single frame mode (typically used for still AVIF images).", OFFSET(still_picture), AV_OPT_TYPE_BOOL, {.i64 = 0}, -1, 1, VE }, + { "dolbyvision", "Enable Dolby Vision RPU coding", OFFSET(dovi.enable), AV_OPT_TYPE_BOOL, {.i64 = FF_DOVI_AUTOMATIC }, -1, 1, VE, .unit = "dovi" }, + { "auto", NULL, 0, AV_OPT_TYPE_CONST, {.i64 = FF_DOVI_AUTOMATIC}, .flags = VE, .unit = "dovi" }, + { "enable-rect-partitions", "Enable rectangular partitions", OFFSET(enable_rect_partitions), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-angle-delta", "Enable angle delta intra prediction", OFFSET(enable_angle_delta), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-cfl-intra", "Enable chroma predicted from luma intra prediction", OFFSET(enable_cfl_intra), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-intra-edge-filter", "Enable intra edge filter", OFFSET(enable_intra_edge_filter), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-smooth-intra", "Enable smooth intra prediction mode", OFFSET(enable_smooth_intra), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-paeth-intra", "Enable paeth predictor in intra prediction", OFFSET(enable_paeth_intra), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-palette", "Enable palette prediction mode", OFFSET(enable_palette), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-flip-idtx", "Enable extended transform type", OFFSET(enable_flip_idtx), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "reduced-tx-type-set", "Use reduced set of transform types", OFFSET(reduced_tx_type_set), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "use-intra-dct-only", "Use DCT only for INTRA modes", OFFSET(use_intra_dct_only), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "use-inter-dct-only", "Use DCT only for INTER modes", OFFSET(use_inter_dct_only), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "use-intra-default-tx-only", "Use default-transform only for INTRA modes", OFFSET(use_intra_default_tx_only), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-ref-frame-mvs", "Enable temporal mv prediction", OFFSET(enable_ref_frame_mvs), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-diff-wtd-comp", "Enable difference-weighted compound", OFFSET(enable_diff_wtd_comp), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-onesided-comp", "Enable one sided compound", OFFSET(enable_onesided_comp), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-interinter-wedge", "Enable interinter wedge compound", OFFSET(enable_interinter_wedge), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-interintra-wedge", "Enable interintra wedge compound", OFFSET(enable_interintra_wedge), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-masked-comp", "Enable masked compound", OFFSET(enable_masked_comp), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-interintra-comp", "Enable interintra compound", OFFSET(enable_interintra_comp), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "enable-smooth-interintra", "Enable smooth interintra mode", OFFSET(enable_smooth_interintra), AV_OPT_TYPE_BOOL, {.i64 = -1}, -1, 1, VE}, + { "avm-params", "Set libavm options using a :-separated list of key=value pairs", OFFSET(avm_params), AV_OPT_TYPE_DICT, { 0 }, 0, 0, VE }, + { NULL }, +}; + +static const FFCodecDefault defaults[] = { + { "b", "0" }, + { "qmin", "-1" }, + { "qmax", "-1" }, + { "g", "-1" }, + { "keyint_min", "-1" }, + { NULL }, +}; + +static const AVClass class_avm = { + .class_name = "libavm-av2 encoder", + .item_name = av_default_item_name, + .option = options, + .version = LIBAVUTIL_VERSION_INT, +}; + +FFCodec ff_libavm_av2_encoder = { + .p.name = "libavm-av2", + CODEC_LONG_NAME("libavm AV2"), + .p.type = AVMEDIA_TYPE_VIDEO, + .p.id = AV_CODEC_ID_AV2, + .p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_DELAY | + AV_CODEC_CAP_ENCODER_RECON_FRAME | + AV_CODEC_CAP_OTHER_THREADS, + .color_ranges = AVCOL_RANGE_MPEG | AVCOL_RANGE_JPEG, + .p.priv_class = &class_avm, + .p.wrapper_name = "libavm", + .priv_data_size = sizeof(AVMContext), + .init = av2_init, + FF_CODEC_ENCODE_CB(avm_encode), + .close = avm_free, + .caps_internal = FF_CODEC_CAP_NOT_INIT_THREADSAFE | + FF_CODEC_CAP_INIT_CLEANUP | + FF_CODEC_CAP_AUTO_THREADS, + CODEC_PIXFMTS_ARRAY(av2_pix_fmts_highbd_with_gray), + .defaults = defaults, +}; -- 2.52.0 >From 631a14d93da56aa504d924ad0cdb667c1eacb0c4 Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:30:51 +0000 Subject: [PATCH 02/11] Added image codec AV2 --- libavformat/img2enc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavformat/img2enc.c b/libavformat/img2enc.c index b11f62d85d..176cee69a9 100644 --- a/libavformat/img2enc.c +++ b/libavformat/img2enc.c @@ -297,7 +297,7 @@ const FFOutputFormat ff_image2_muxer = { .p.long_name = NULL_IF_CONFIG_SMALL("image2 sequence"), .p.extensions = "bmp,dpx,exr,jls,jpeg,jpg,jxs,jxl,ljpg,pam,pbm,pcx,pfm,pgm,pgmyuv,phm," "png,ppm,sgi,tga,tif,tiff,jp2,j2c,j2k,xwd,sun,ras,rs,im1,im8," - "im24,sunras,vbn,xbm,xface,pix,y,avif,qoi,hdr,wbmp", + "im24,sunras,vbn,xbm,xface,pix,y,av2,avif,qoi,hdr,wbmp", .priv_data_size = sizeof(VideoMuxData), .p.video_codec = AV_CODEC_ID_MJPEG, .write_header = write_header, -- 2.52.0 >From 16a1694cf642d4fb883b2582aca03239a4288c38 Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:32:36 +0000 Subject: [PATCH 03/11] Added image codec AV2 --- libavformat/demux.c | 1 + 1 file changed, 1 insertion(+) diff --git a/libavformat/demux.c b/libavformat/demux.c index 55085210de..336d0c7dd8 100644 --- a/libavformat/demux.c +++ b/libavformat/demux.c @@ -112,6 +112,7 @@ static int set_codec_from_probe_data(AVFormatContext *s, AVStream *st, { "ac3", AV_CODEC_ID_AC3, AVMEDIA_TYPE_AUDIO }, { "aptx", AV_CODEC_ID_APTX, AVMEDIA_TYPE_AUDIO }, { "av1", AV_CODEC_ID_AV1, AVMEDIA_TYPE_VIDEO }, + { "av2", AV_CODEC_ID_AV2, AVMEDIA_TYPE_VIDEO }, { "dts", AV_CODEC_ID_DTS, AVMEDIA_TYPE_AUDIO }, { "dvbsub", AV_CODEC_ID_DVB_SUBTITLE, AVMEDIA_TYPE_SUBTITLE }, { "dvbtxt", AV_CODEC_ID_DVB_TELETEXT, AVMEDIA_TYPE_SUBTITLE }, -- 2.52.0 >From 0a8296d674f8ab4a794cb0cf2a7216133a31a427 Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:33:45 +0000 Subject: [PATCH 04/11] Added image codec AV2 --- libavformat/img2.c | 1 + 1 file changed, 1 insertion(+) diff --git a/libavformat/img2.c b/libavformat/img2.c index 2c69a932da..78b4018b08 100644 --- a/libavformat/img2.c +++ b/libavformat/img2.c @@ -27,6 +27,7 @@ #include "img2.h" #define IMG_TAGS(TAG) \ + TAG(AV2, av2 ) \ TAG(MJPEG, jpeg ) \ TAG(MJPEG, jpg ) \ TAG(MJPEG, jps ) \ -- 2.52.0 >From e9835babcfa1fc33e15136b8ba842659d364527f Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:36:39 +0000 Subject: [PATCH 05/11] Added image codec AV2 --- libavformat/isom_tags.c | 1 + 1 file changed, 1 insertion(+) diff --git a/libavformat/isom_tags.c b/libavformat/isom_tags.c index 1cd655b06c..5e63f9f7b5 100644 --- a/libavformat/isom_tags.c +++ b/libavformat/isom_tags.c @@ -160,6 +160,7 @@ const AVCodecTag ff_codec_movvideo_tags[] = { { AV_CODEC_ID_VP8, MKTAG('v', 'p', '0', '8') }, /* VP8 */ { AV_CODEC_ID_VP9, MKTAG('v', 'p', '0', '9') }, /* VP9 */ { AV_CODEC_ID_AV1, MKTAG('a', 'v', '0', '1') }, /* AV1 */ + { AV_CODEC_ID_AV2, MKTAG('a', 'v', '0', '2') }, /* AV2 */ { AV_CODEC_ID_MPEG1VIDEO, MKTAG('m', '1', 'v', ' ') }, { AV_CODEC_ID_MPEG1VIDEO, MKTAG('m', '1', 'v', '1') }, /* Apple MPEG-1 Camcorder */ -- 2.52.0 >From 47e5a35e388fca7672e24eaa28f2a6026074ce79 Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:38:08 +0000 Subject: [PATCH 06/11] Added image codec AV2 --- libavformat/ivfenc.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/libavformat/ivfenc.c b/libavformat/ivfenc.c index 9feaea3516..50a8a6c039 100644 --- a/libavformat/ivfenc.c +++ b/libavformat/ivfenc.c @@ -31,10 +31,11 @@ static int ivf_init(AVFormatContext *s) { AVCodecParameters *par = s->streams[0]->codecpar; - if (!(par->codec_id == AV_CODEC_ID_AV1 || + if (!(par->codec_id == AV_CODEC_ID_AV2 || + par->codec_id == AV_CODEC_ID_AV1 || par->codec_id == AV_CODEC_ID_VP8 || par->codec_id == AV_CODEC_ID_VP9)) { - av_log(s, AV_LOG_ERROR, "Currently only VP8, VP9 and AV1 are supported!\n"); + av_log(s, AV_LOG_ERROR, "Currently only VP8, VP9, AV1 and AV2 are supported!\n"); return AVERROR(EINVAL); } @@ -61,7 +62,8 @@ static int ivf_write_header(AVFormatContext *s) avio_wl16(pb, 32); // header length avio_wl32(pb, par->codec_id == AV_CODEC_ID_VP9 ? AV_RL32("VP90") : - par->codec_id == AV_CODEC_ID_VP8 ? AV_RL32("VP80") : AV_RL32("AV01")); + par->codec_id == AV_CODEC_ID_VP8 ? AV_RL32("VP80") : + par->codec_id == AV_CODEC_ID_AV1 ? AV_RL32("AV01") : AV_RL32("AV02")); avio_wl16(pb, par->width); avio_wl16(pb, par->height); avio_wl32(pb, s->streams[0]->time_base.den); @@ -108,6 +110,7 @@ static const AVCodecTag codec_ivf_tags[] = { { AV_CODEC_ID_VP8, MKTAG('V', 'P', '8', '0') }, { AV_CODEC_ID_VP9, MKTAG('V', 'P', '9', '0') }, { AV_CODEC_ID_AV1, MKTAG('A', 'V', '0', '1') }, + { AV_CODEC_ID_AV2, MKTAG('A', 'V', '0', '2') }, { AV_CODEC_ID_NONE, 0 } }; -- 2.52.0 >From 518aa6b3771aeef6537dc3518b4b0bb4ecd4ea2b Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:38:59 +0000 Subject: [PATCH 07/11] Added image codec AV2 --- libavcodec/libavm.h | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) create mode 100644 libavcodec/libavm.h diff --git a/libavcodec/libavm.h b/libavcodec/libavm.h new file mode 100644 index 0000000000..c2c66eb8bf --- /dev/null +++ b/libavcodec/libavm.h @@ -0,0 +1,33 @@ +/* + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +/** + * @file + * AOM common functions + */ + +#ifndef AVCODEC_LIBAOM_H +#define AVCODEC_LIBAOM_H + +#include <avm/avm_image.h> + +#include "libavutil/frame.h" + +void ff_avm_image_copy_16_to_8(AVFrame *pic, struct avm_image *img); + +#endif /* AVCODEC_LIBAOM_H */ -- 2.52.0 >From 040e215269eb310522c0c1806edc2e05a811a0dd Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:40:33 +0000 Subject: [PATCH 08/11] Added image codec AV2 --- libavcodec/codec_desc.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/libavcodec/codec_desc.c b/libavcodec/codec_desc.c index 99e7c0de3d..b56726d7e0 100644 --- a/libavcodec/codec_desc.c +++ b/libavcodec/codec_desc.c @@ -2017,6 +2017,13 @@ static const AVCodecDescriptor codec_descriptors[] = { .props = AV_CODEC_PROP_LOSSY | AV_CODEC_PROP_LOSSLESS, .mime_types= MT("image/webp"), }, + { + .id = AV_CODEC_ID_AV2, + .type = AVMEDIA_TYPE_VIDEO, + .name = "av2", + .long_name = NULL_IF_CONFIG_SMALL("Alliance for Open Media AV2"), + .props = AV_CODEC_PROP_LOSSY | AV_CODEC_PROP_LOSSLESS, + }, /* various PCM "codecs" */ { -- 2.52.0 >From b071c5939eb38283887c34d67c66071433215f3f Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:41:39 +0000 Subject: [PATCH 09/11] Added image codec AV2 --- libavcodec/codec_id.h | 1 + 1 file changed, 1 insertion(+) diff --git a/libavcodec/codec_id.h b/libavcodec/codec_id.h index bb2c16a958..d7b8b909f0 100644 --- a/libavcodec/codec_id.h +++ b/libavcodec/codec_id.h @@ -333,6 +333,7 @@ enum AVCodecID { AV_CODEC_ID_PRORES_RAW, AV_CODEC_ID_JPEGXS, AV_CODEC_ID_WEBP_ANIM, + AV_CODEC_ID_AV2, /* various PCM "codecs" */ AV_CODEC_ID_FIRST_AUDIO = 0x10000, ///< A dummy id pointing at the start of audio codecs -- 2.52.0 >From c11054bcae132ecf04aeb18a66cafa695af23e31 Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:48:37 +0000 Subject: [PATCH 10/11] Added image codec AV2 --- libavcodec/Makefile | 3 +++ 1 file changed, 3 insertions(+) diff --git a/libavcodec/Makefile b/libavcodec/Makefile index ac3c4d1978..50244fe67f 100644 --- a/libavcodec/Makefile +++ b/libavcodec/Makefile @@ -1172,6 +1172,8 @@ OBJS-$(CONFIG_PCM_ALAW_AT_ENCODER) += audiotoolboxenc.o OBJS-$(CONFIG_PCM_MULAW_AT_ENCODER) += audiotoolboxenc.o OBJS-$(CONFIG_LIBAOM_AV1_DECODER) += libaomdec.o libaom.o OBJS-$(CONFIG_LIBAOM_AV1_ENCODER) += libaomenc.o libaom.o +OBJS-$(CONFIG_LIBAVM_AV2_DECODER) += libavmdec.o libavm.o +OBJS-$(CONFIG_LIBAVM_AV2_ENCODER) += libavmenc.o libavm.o OBJS-$(CONFIG_LIBARIBB24_DECODER) += libaribb24.o ass.o OBJS-$(CONFIG_LIBARIBCAPTION_DECODER) += libaribcaption.o ass.o OBJS-$(CONFIG_LIBCODEC2_DECODER) += libcodec2.o @@ -1346,6 +1348,7 @@ SKIPHEADERS-$(CONFIG_DXVA2) += dxva2.h dxva2_internal.h SKIPHEADERS-$(CONFIG_JNI) += ffjni.h SKIPHEADERS-$(CONFIG_LCMS2) += fflcms2.h SKIPHEADERS-$(CONFIG_LIBAOM) += libaom.h +SKIPHEADERS-$(CONFIG_LIBAVM) += libavm.h SKIPHEADERS-$(CONFIG_LIBJXL) += libjxl.h SKIPHEADERS-$(CONFIG_LIBVPX) += libvpx.h SKIPHEADERS-$(CONFIG_LIBWEBP_ENCODER) += libwebpenc_common.h -- 2.52.0 >From 63320c65a49953f97af93ae68388b8fb2171a248 Mon Sep 17 00:00:00 2001 From: Jamaika1 <[email protected]> Date: Tue, 2 Jun 2026 04:50:25 +0000 Subject: [PATCH 11/11] Added image codec AV2 --- libavcodec/allcodecs.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/libavcodec/allcodecs.c b/libavcodec/allcodecs.c index 0815d46f79..2fd9880b5d 100644 --- a/libavcodec/allcodecs.c +++ b/libavcodec/allcodecs.c @@ -783,6 +783,7 @@ extern const FFCodec ff_pcm_mulaw_at_decoder; extern const FFCodec ff_qdmc_at_decoder; extern const FFCodec ff_qdm2_at_decoder; extern FFCodec ff_libaom_av1_encoder; +extern FFCodec ff_libavm_av2_encoder; /* preferred over libaribb24 */ extern const FFCodec ff_libaribcaption_decoder; extern const FFCodec ff_libaribb24_decoder; @@ -860,6 +861,7 @@ extern const FFCodec ff_amrnb_mediacodec_decoder; extern const FFCodec ff_amrwb_mediacodec_decoder; extern const FFCodec ff_h263_v4l2m2m_encoder; extern const FFCodec ff_libaom_av1_decoder; +extern const FFCodec ff_libavm_av2_decoder; /* hwaccel hooks only, so prefer external decoders */ extern const FFCodec ff_av1_decoder; extern const FFCodec ff_av1_cuvid_decoder; -- 2.52.0 _______________________________________________ ffmpeg-devel mailing list -- [email protected] To unsubscribe send an email to [email protected]
