1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-11-26 19:01:44 +02:00
FFmpeg/libavcodec/libsvtav1.c
Limin Wang c5f3143090 avcodec/libsvtav1: properly enforce CQP mode when set in wrapper
SVT-AV1 seems to have switched their default from CQP to CRF in February,
so enforce the controlling option accordingly.

Signed-off-by: Limin Wang <lance.lmwang@gmail.com>
2021-09-29 18:01:16 +08:00

575 lines
19 KiB
C

/*
* Scalable Video Technology for AV1 encoder library plugin
*
* Copyright (c) 2018 Intel Corporation
*
* 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 this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include <stdint.h>
#include <EbSvtAv1ErrorCodes.h>
#include <EbSvtAv1Enc.h>
#include "libavutil/common.h"
#include "libavutil/frame.h"
#include "libavutil/imgutils.h"
#include "libavutil/opt.h"
#include "libavutil/pixdesc.h"
#include "libavutil/avassert.h"
#include "internal.h"
#include "encode.h"
#include "packet_internal.h"
#include "avcodec.h"
#include "profiles.h"
typedef enum eos_status {
EOS_NOT_REACHED = 0,
EOS_SENT,
EOS_RECEIVED
}EOS_STATUS;
typedef struct SvtContext {
const AVClass *class;
EbSvtAv1EncConfiguration enc_params;
EbComponentType *svt_handle;
EbBufferHeaderType *in_buf;
int raw_size;
int max_tu_size;
AVFrame *frame;
AVBufferPool *pool;
EOS_STATUS eos_flag;
// User options.
int hierarchical_level;
int la_depth;
int enc_mode;
int rc_mode;
int scd;
int qp;
int tier;
int tile_columns;
int tile_rows;
} SvtContext;
static const struct {
EbErrorType eb_err;
int av_err;
const char *desc;
} svt_errors[] = {
{ EB_ErrorNone, 0, "success" },
{ EB_ErrorInsufficientResources, AVERROR(ENOMEM), "insufficient resources" },
{ EB_ErrorUndefined, AVERROR(EINVAL), "undefined error" },
{ EB_ErrorInvalidComponent, AVERROR(EINVAL), "invalid component" },
{ EB_ErrorBadParameter, AVERROR(EINVAL), "bad parameter" },
{ EB_ErrorDestroyThreadFailed, AVERROR_EXTERNAL, "failed to destroy thread" },
{ EB_ErrorSemaphoreUnresponsive, AVERROR_EXTERNAL, "semaphore unresponsive" },
{ EB_ErrorDestroySemaphoreFailed, AVERROR_EXTERNAL, "failed to destroy semaphore"},
{ EB_ErrorCreateMutexFailed, AVERROR_EXTERNAL, "failed to create mutex" },
{ EB_ErrorMutexUnresponsive, AVERROR_EXTERNAL, "mutex unresponsive" },
{ EB_ErrorDestroyMutexFailed, AVERROR_EXTERNAL, "failed to destroy mutex" },
{ EB_NoErrorEmptyQueue, AVERROR(EAGAIN), "empty queue" },
};
static int svt_map_error(EbErrorType eb_err, const char **desc)
{
int i;
av_assert0(desc);
for (i = 0; i < FF_ARRAY_ELEMS(svt_errors); i++) {
if (svt_errors[i].eb_err == eb_err) {
*desc = svt_errors[i].desc;
return svt_errors[i].av_err;
}
}
*desc = "unknown error";
return AVERROR_UNKNOWN;
}
static int svt_print_error(void *log_ctx, EbErrorType err,
const char *error_string)
{
const char *desc;
int ret = svt_map_error(err, &desc);
av_log(log_ctx, AV_LOG_ERROR, "%s: %s (0x%x)\n", error_string, desc, err);
return ret;
}
static int alloc_buffer(EbSvtAv1EncConfiguration *config, SvtContext *svt_enc)
{
const int pack_mode_10bit =
(config->encoder_bit_depth > 8) && (config->compressed_ten_bit_format == 0) ? 1 : 0;
const size_t luma_size_8bit =
config->source_width * config->source_height * (1 << pack_mode_10bit);
const size_t luma_size_10bit =
(config->encoder_bit_depth > 8 && pack_mode_10bit == 0) ? luma_size_8bit : 0;
EbSvtIOFormat *in_data;
svt_enc->raw_size = (luma_size_8bit + luma_size_10bit) * 3 / 2;
// allocate buffer for in and out
svt_enc->in_buf = av_mallocz(sizeof(*svt_enc->in_buf));
if (!svt_enc->in_buf)
return AVERROR(ENOMEM);
svt_enc->in_buf->p_buffer = av_mallocz(sizeof(*in_data));
if (!svt_enc->in_buf->p_buffer)
return AVERROR(ENOMEM);
svt_enc->in_buf->size = sizeof(*svt_enc->in_buf);
return 0;
}
static int config_enc_params(EbSvtAv1EncConfiguration *param,
AVCodecContext *avctx)
{
SvtContext *svt_enc = avctx->priv_data;
const AVPixFmtDescriptor *desc;
param->source_width = avctx->width;
param->source_height = avctx->height;
desc = av_pix_fmt_desc_get(avctx->pix_fmt);
param->encoder_bit_depth = desc->comp[0].depth;
if (desc->log2_chroma_w == 1 && desc->log2_chroma_h == 1)
param->encoder_color_format = EB_YUV420;
else if (desc->log2_chroma_w == 1 && desc->log2_chroma_h == 0)
param->encoder_color_format = EB_YUV422;
else if (!desc->log2_chroma_w && !desc->log2_chroma_h)
param->encoder_color_format = EB_YUV444;
else {
av_log(avctx, AV_LOG_ERROR , "Unsupported pixel format\n");
return AVERROR(EINVAL);
}
if (avctx->profile != FF_PROFILE_UNKNOWN)
param->profile = avctx->profile;
if (avctx->level != FF_LEVEL_UNKNOWN)
param->level = avctx->level;
if ((param->encoder_color_format == EB_YUV422 || param->encoder_bit_depth > 10)
&& param->profile != FF_PROFILE_AV1_PROFESSIONAL ) {
av_log(avctx, AV_LOG_WARNING, "Forcing Professional profile\n");
param->profile = FF_PROFILE_AV1_PROFESSIONAL;
} else if (param->encoder_color_format == EB_YUV444 && param->profile != FF_PROFILE_AV1_HIGH) {
av_log(avctx, AV_LOG_WARNING, "Forcing High profile\n");
param->profile = FF_PROFILE_AV1_HIGH;
}
// Update param from options
param->hierarchical_levels = svt_enc->hierarchical_level;
param->enc_mode = svt_enc->enc_mode;
param->tier = svt_enc->tier;
param->rate_control_mode = svt_enc->rc_mode;
param->scene_change_detection = svt_enc->scd;
param->qp = svt_enc->qp;
param->target_bit_rate = avctx->bit_rate;
if (avctx->gop_size > 0)
param->intra_period_length = avctx->gop_size - 1;
if (avctx->framerate.num > 0 && avctx->framerate.den > 0) {
param->frame_rate_numerator = avctx->framerate.num;
param->frame_rate_denominator = avctx->framerate.den;
} else {
param->frame_rate_numerator = avctx->time_base.den;
param->frame_rate_denominator = avctx->time_base.num * avctx->ticks_per_frame;
}
param->enable_tpl_la = !!param->rate_control_mode;
if (param->rate_control_mode) {
param->max_qp_allowed = avctx->qmax;
param->min_qp_allowed = avctx->qmin;
}
/* 2 = IDR, closed GOP, 1 = CRA, open GOP */
param->intra_refresh_type = avctx->flags & AV_CODEC_FLAG_CLOSED_GOP ? 2 : 1;
if (svt_enc->la_depth >= 0)
param->look_ahead_distance = svt_enc->la_depth;
param->tile_columns = svt_enc->tile_columns;
param->tile_rows = svt_enc->tile_rows;
return 0;
}
static int read_in_data(EbSvtAv1EncConfiguration *param, const AVFrame *frame,
EbBufferHeaderType *header_ptr)
{
EbSvtIOFormat *in_data = (EbSvtIOFormat *)header_ptr->p_buffer;
ptrdiff_t linesizes[4];
size_t sizes[4];
int bytes_shift = param->encoder_bit_depth > 8 ? 1 : 0;
int ret, frame_size;
for (int i = 0; i < 4; i++)
linesizes[i] = frame->linesize[i];
ret = av_image_fill_plane_sizes(sizes, frame->format, frame->height,
linesizes);
if (ret < 0)
return ret;
frame_size = 0;
for (int i = 0; i < 4; i++) {
if (sizes[i] > INT_MAX - frame_size)
return AVERROR(EINVAL);
frame_size += sizes[i];
}
in_data->luma = frame->data[0];
in_data->cb = frame->data[1];
in_data->cr = frame->data[2];
in_data->y_stride = AV_CEIL_RSHIFT(frame->linesize[0], bytes_shift);
in_data->cb_stride = AV_CEIL_RSHIFT(frame->linesize[1], bytes_shift);
in_data->cr_stride = AV_CEIL_RSHIFT(frame->linesize[2], bytes_shift);
header_ptr->n_filled_len = frame_size;
return 0;
}
static av_cold int eb_enc_init(AVCodecContext *avctx)
{
SvtContext *svt_enc = avctx->priv_data;
EbErrorType svt_ret;
int ret;
svt_enc->eos_flag = EOS_NOT_REACHED;
svt_ret = svt_av1_enc_init_handle(&svt_enc->svt_handle, svt_enc, &svt_enc->enc_params);
if (svt_ret != EB_ErrorNone) {
return svt_print_error(avctx, svt_ret, "Error initializing encoder handle");
}
ret = config_enc_params(&svt_enc->enc_params, avctx);
if (ret < 0) {
av_log(avctx, AV_LOG_ERROR, "Error configuring encoder parameters\n");
return ret;
}
svt_ret = svt_av1_enc_set_parameter(svt_enc->svt_handle, &svt_enc->enc_params);
if (svt_ret != EB_ErrorNone) {
return svt_print_error(avctx, svt_ret, "Error setting encoder parameters");
}
svt_ret = svt_av1_enc_init(svt_enc->svt_handle);
if (svt_ret != EB_ErrorNone) {
return svt_print_error(avctx, svt_ret, "Error initializing encoder");
}
if (avctx->flags & AV_CODEC_FLAG_GLOBAL_HEADER) {
EbBufferHeaderType *headerPtr = NULL;
svt_ret = svt_av1_enc_stream_header(svt_enc->svt_handle, &headerPtr);
if (svt_ret != EB_ErrorNone) {
return svt_print_error(avctx, svt_ret, "Error building stream header");
}
avctx->extradata_size = headerPtr->n_filled_len;
avctx->extradata = av_mallocz(avctx->extradata_size + AV_INPUT_BUFFER_PADDING_SIZE);
if (!avctx->extradata) {
av_log(avctx, AV_LOG_ERROR,
"Cannot allocate AV1 header of size %d.\n", avctx->extradata_size);
return AVERROR(ENOMEM);
}
memcpy(avctx->extradata, headerPtr->p_buffer, avctx->extradata_size);
svt_ret = svt_av1_enc_stream_header_release(headerPtr);
if (svt_ret != EB_ErrorNone) {
return svt_print_error(avctx, svt_ret, "Error freeing stream header");
}
}
svt_enc->frame = av_frame_alloc();
if (!svt_enc->frame)
return AVERROR(ENOMEM);
return alloc_buffer(&svt_enc->enc_params, svt_enc);
}
static int eb_send_frame(AVCodecContext *avctx, const AVFrame *frame)
{
SvtContext *svt_enc = avctx->priv_data;
EbBufferHeaderType *headerPtr = svt_enc->in_buf;
int ret;
if (!frame) {
EbBufferHeaderType headerPtrLast;
if (svt_enc->eos_flag == EOS_SENT)
return 0;
headerPtrLast.n_alloc_len = 0;
headerPtrLast.n_filled_len = 0;
headerPtrLast.n_tick_count = 0;
headerPtrLast.p_app_private = NULL;
headerPtrLast.p_buffer = NULL;
headerPtrLast.flags = EB_BUFFERFLAG_EOS;
svt_av1_enc_send_picture(svt_enc->svt_handle, &headerPtrLast);
svt_enc->eos_flag = EOS_SENT;
return 0;
}
ret = read_in_data(&svt_enc->enc_params, frame, headerPtr);
if (ret < 0)
return ret;
headerPtr->flags = 0;
headerPtr->p_app_private = NULL;
headerPtr->pts = frame->pts;
svt_av1_enc_send_picture(svt_enc->svt_handle, headerPtr);
return 0;
}
static AVBufferRef *get_output_ref(AVCodecContext *avctx, SvtContext *svt_enc, int filled_len)
{
if (filled_len > svt_enc->max_tu_size) {
const int max_frames = 8;
int max_tu_size;
if (filled_len > svt_enc->raw_size * max_frames) {
av_log(avctx, AV_LOG_ERROR, "TU size > %d raw frame size.\n", max_frames);
return NULL;
}
max_tu_size = 1 << av_ceil_log2(filled_len);
av_buffer_pool_uninit(&svt_enc->pool);
svt_enc->pool = av_buffer_pool_init(max_tu_size + AV_INPUT_BUFFER_PADDING_SIZE, NULL);
if (!svt_enc->pool)
return NULL;
svt_enc->max_tu_size = max_tu_size;
}
av_assert0(svt_enc->pool);
return av_buffer_pool_get(svt_enc->pool);
}
static int eb_receive_packet(AVCodecContext *avctx, AVPacket *pkt)
{
SvtContext *svt_enc = avctx->priv_data;
EbBufferHeaderType *headerPtr;
AVFrame *frame = svt_enc->frame;
EbErrorType svt_ret;
AVBufferRef *ref;
int ret = 0, pict_type;
if (svt_enc->eos_flag == EOS_RECEIVED)
return AVERROR_EOF;
ret = ff_encode_get_frame(avctx, frame);
if (ret < 0 && ret != AVERROR_EOF)
return ret;
if (ret == AVERROR_EOF)
frame = NULL;
ret = eb_send_frame(avctx, frame);
if (ret < 0)
return ret;
av_frame_unref(svt_enc->frame);
svt_ret = svt_av1_enc_get_packet(svt_enc->svt_handle, &headerPtr, svt_enc->eos_flag);
if (svt_ret == EB_NoErrorEmptyQueue)
return AVERROR(EAGAIN);
ref = get_output_ref(avctx, svt_enc, headerPtr->n_filled_len);
if (!ref) {
av_log(avctx, AV_LOG_ERROR, "Failed to allocate output packet.\n");
svt_av1_enc_release_out_buffer(&headerPtr);
return AVERROR(ENOMEM);
}
pkt->buf = ref;
pkt->data = ref->data;
memcpy(pkt->data, headerPtr->p_buffer, headerPtr->n_filled_len);
memset(pkt->data + headerPtr->n_filled_len, 0, AV_INPUT_BUFFER_PADDING_SIZE);
pkt->size = headerPtr->n_filled_len;
pkt->pts = headerPtr->pts;
pkt->dts = headerPtr->dts;
switch (headerPtr->pic_type) {
case EB_AV1_KEY_PICTURE:
pkt->flags |= AV_PKT_FLAG_KEY;
// fall-through
case EB_AV1_INTRA_ONLY_PICTURE:
pict_type = AV_PICTURE_TYPE_I;
break;
case EB_AV1_INVALID_PICTURE:
pict_type = AV_PICTURE_TYPE_NONE;
break;
default:
pict_type = AV_PICTURE_TYPE_P;
break;
}
if (headerPtr->pic_type == EB_AV1_NON_REF_PICTURE)
pkt->flags |= AV_PKT_FLAG_DISPOSABLE;
if (headerPtr->flags & EB_BUFFERFLAG_EOS)
svt_enc->eos_flag = EOS_RECEIVED;
ff_side_data_set_encoder_stats(pkt, headerPtr->qp * FF_QP2LAMBDA, NULL, 0, pict_type);
svt_av1_enc_release_out_buffer(&headerPtr);
return 0;
}
static av_cold int eb_enc_close(AVCodecContext *avctx)
{
SvtContext *svt_enc = avctx->priv_data;
if (svt_enc->svt_handle) {
svt_av1_enc_deinit(svt_enc->svt_handle);
svt_av1_enc_deinit_handle(svt_enc->svt_handle);
}
if (svt_enc->in_buf) {
av_free(svt_enc->in_buf->p_buffer);
av_freep(&svt_enc->in_buf);
}
av_buffer_pool_uninit(&svt_enc->pool);
av_frame_free(&svt_enc->frame);
return 0;
}
#define OFFSET(x) offsetof(SvtContext, x)
#define VE AV_OPT_FLAG_VIDEO_PARAM | AV_OPT_FLAG_ENCODING_PARAM
static const AVOption options[] = {
{ "hielevel", "Hierarchical prediction levels setting", OFFSET(hierarchical_level),
AV_OPT_TYPE_INT, { .i64 = 4 }, 3, 4, VE , "hielevel"},
{ "3level", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = 3 }, INT_MIN, INT_MAX, VE, "hielevel" },
{ "4level", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = 4 }, INT_MIN, INT_MAX, VE, "hielevel" },
{ "la_depth", "Look ahead distance [0, 120]", OFFSET(la_depth),
AV_OPT_TYPE_INT, { .i64 = -1 }, -1, 120, VE },
{ "preset", "Encoding preset [0, 8]",
OFFSET(enc_mode), AV_OPT_TYPE_INT, { .i64 = MAX_ENC_PRESET }, 0, MAX_ENC_PRESET, VE },
{ "tier", "Set operating point tier", OFFSET(tier),
AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 1, VE, "tier" },
{ "main", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = 0 }, 0, 0, VE, "tier" },
{ "high", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = 1 }, 0, 0, VE, "tier" },
FF_AV1_PROFILE_OPTS
#define LEVEL(name, value) name, NULL, 0, AV_OPT_TYPE_CONST, \
{ .i64 = value }, 0, 0, VE, "avctx.level"
{ LEVEL("2.0", 20) },
{ LEVEL("2.1", 21) },
{ LEVEL("2.2", 22) },
{ LEVEL("2.3", 23) },
{ LEVEL("3.0", 30) },
{ LEVEL("3.1", 31) },
{ LEVEL("3.2", 32) },
{ LEVEL("3.3", 33) },
{ LEVEL("4.0", 40) },
{ LEVEL("4.1", 41) },
{ LEVEL("4.2", 42) },
{ LEVEL("4.3", 43) },
{ LEVEL("5.0", 50) },
{ LEVEL("5.1", 51) },
{ LEVEL("5.2", 52) },
{ LEVEL("5.3", 53) },
{ LEVEL("6.0", 60) },
{ LEVEL("6.1", 61) },
{ LEVEL("6.2", 62) },
{ LEVEL("6.3", 63) },
{ LEVEL("7.0", 70) },
{ LEVEL("7.1", 71) },
{ LEVEL("7.2", 72) },
{ LEVEL("7.3", 73) },
#undef LEVEL
{ "rc", "Bit rate control mode", OFFSET(rc_mode),
AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 2, VE , "rc"},
{ "cqp", "Constant quantizer", 0, AV_OPT_TYPE_CONST, { .i64 = 0 }, INT_MIN, INT_MAX, VE, "rc" },
{ "vbr", "Variable Bit Rate, use a target bitrate for the entire stream", 0, AV_OPT_TYPE_CONST, { .i64 = 1 }, INT_MIN, INT_MAX, VE, "rc" },
{ "cvbr", "Constrained Variable Bit Rate, use a target bitrate for each GOP", 0, AV_OPT_TYPE_CONST,{ .i64 = 2 }, INT_MIN, INT_MAX, VE, "rc" },
{ "qp", "Quantizer to use with cqp rate control mode", OFFSET(qp),
AV_OPT_TYPE_INT, { .i64 = 50 }, 0, 63, VE },
{ "sc_detection", "Scene change detection", OFFSET(scd),
AV_OPT_TYPE_BOOL, { .i64 = 0 }, 0, 1, VE },
{ "tile_columns", "Log2 of number of tile columns to use", OFFSET(tile_columns), AV_OPT_TYPE_INT, {.i64 = 0}, 0, 4, VE},
{ "tile_rows", "Log2 of number of tile rows to use", OFFSET(tile_rows), AV_OPT_TYPE_INT, {.i64 = 0}, 0, 6, VE},
{NULL},
};
static const AVClass class = {
.class_name = "libsvtav1",
.item_name = av_default_item_name,
.option = options,
.version = LIBAVUTIL_VERSION_INT,
};
static const AVCodecDefault eb_enc_defaults[] = {
{ "b", "7M" },
{ "flags", "+cgop" },
{ "g", "-1" },
{ "qmin", "0" },
{ "qmax", "63" },
{ NULL },
};
const AVCodec ff_libsvtav1_encoder = {
.name = "libsvtav1",
.long_name = NULL_IF_CONFIG_SMALL("SVT-AV1(Scalable Video Technology for AV1) encoder"),
.priv_data_size = sizeof(SvtContext),
.type = AVMEDIA_TYPE_VIDEO,
.id = AV_CODEC_ID_AV1,
.init = eb_enc_init,
.receive_packet = eb_receive_packet,
.close = eb_enc_close,
.capabilities = AV_CODEC_CAP_DELAY | AV_CODEC_CAP_OTHER_THREADS,
.caps_internal = FF_CODEC_CAP_AUTO_THREADS | FF_CODEC_CAP_INIT_CLEANUP,
.pix_fmts = (const enum AVPixelFormat[]){ AV_PIX_FMT_YUV420P,
AV_PIX_FMT_YUV420P10,
AV_PIX_FMT_NONE },
.priv_class = &class,
.defaults = eb_enc_defaults,
.wrapper_name = "libsvtav1",
};