mirror of
https://github.com/FFmpeg/FFmpeg.git
synced 2024-11-26 19:01:44 +02:00
avcodec: Implement vc1 nvdec hwaccel
This hwaccel is interesting because it also works for wmv3/9 content, which is not supported by the nvidia parser used by cuviddec.
This commit is contained in:
parent
bb4c9d0a8e
commit
912ceba61b
@ -13,7 +13,7 @@ version <next>:
|
||||
- PCE support for extended channel layouts in the AAC encoder
|
||||
- native aptX encoder and decoder
|
||||
- Raw aptX muxer and demuxer
|
||||
- NVIDIA NVDEC-accelerated H.264, HEVC and VP9 hwaccel decoding
|
||||
- NVIDIA NVDEC-accelerated H.264, HEVC, VC1 and VP9 hwaccel decoding
|
||||
- Intel QSV-accelerated overlay filter
|
||||
|
||||
|
||||
|
3
configure
vendored
3
configure
vendored
@ -2740,6 +2740,8 @@ vc1_d3d11va2_hwaccel_select="vc1_decoder"
|
||||
vc1_dxva2_hwaccel_deps="dxva2"
|
||||
vc1_dxva2_hwaccel_select="vc1_decoder"
|
||||
vc1_mmal_hwaccel_deps="mmal"
|
||||
vc1_nvdec_hwaccel_deps="nvdec"
|
||||
vc1_nvdec_hwaccel_select="vc1_decoder"
|
||||
vc1_qsv_hwaccel_deps="libmfx"
|
||||
vc1_vaapi_hwaccel_deps="vaapi"
|
||||
vc1_vaapi_hwaccel_select="vc1_decoder"
|
||||
@ -2763,6 +2765,7 @@ vp9_vaapi_hwaccel_select="vp9_decoder"
|
||||
wmv3_d3d11va_hwaccel_select="vc1_d3d11va_hwaccel"
|
||||
wmv3_d3d11va2_hwaccel_select="vc1_d3d11va2_hwaccel"
|
||||
wmv3_dxva2_hwaccel_select="vc1_dxva2_hwaccel"
|
||||
wmv3_nvdec_hwaccel_select="vc1_nvdec_hwaccel"
|
||||
wmv3_vaapi_hwaccel_select="vc1_vaapi_hwaccel"
|
||||
wmv3_vdpau_hwaccel_select="vc1_vdpau_hwaccel"
|
||||
|
||||
|
@ -864,6 +864,7 @@ OBJS-$(CONFIG_MPEG4_VDPAU_HWACCEL) += vdpau_mpeg4.o
|
||||
OBJS-$(CONFIG_MPEG4_VIDEOTOOLBOX_HWACCEL) += videotoolbox.o
|
||||
OBJS-$(CONFIG_VC1_D3D11VA_HWACCEL) += dxva2_vc1.o
|
||||
OBJS-$(CONFIG_VC1_DXVA2_HWACCEL) += dxva2_vc1.o
|
||||
OBJS-$(CONFIG_VC1_NVDEC_HWACCEL) += nvdec_vc1.o
|
||||
OBJS-$(CONFIG_VC1_QSV_HWACCEL) += qsvdec_other.o
|
||||
OBJS-$(CONFIG_VC1_VAAPI_HWACCEL) += vaapi_vc1.o
|
||||
OBJS-$(CONFIG_VC1_VDPAU_HWACCEL) += vdpau_vc1.o
|
||||
|
@ -111,6 +111,7 @@ static void register_all(void)
|
||||
REGISTER_HWACCEL(VC1_D3D11VA, vc1_d3d11va);
|
||||
REGISTER_HWACCEL(VC1_D3D11VA2, vc1_d3d11va2);
|
||||
REGISTER_HWACCEL(VC1_DXVA2, vc1_dxva2);
|
||||
REGISTER_HWACCEL(VC1_NVDEC, vc1_nvdec);
|
||||
REGISTER_HWACCEL(VC1_VAAPI, vc1_vaapi);
|
||||
REGISTER_HWACCEL(VC1_VDPAU, vc1_vdpau);
|
||||
REGISTER_HWACCEL(VC1_MMAL, vc1_mmal);
|
||||
@ -128,6 +129,7 @@ static void register_all(void)
|
||||
REGISTER_HWACCEL(WMV3_D3D11VA, wmv3_d3d11va);
|
||||
REGISTER_HWACCEL(WMV3_D3D11VA2, wmv3_d3d11va2);
|
||||
REGISTER_HWACCEL(WMV3_DXVA2, wmv3_dxva2);
|
||||
REGISTER_HWACCEL(WMV3_NVDEC, wmv3_nvdec);
|
||||
REGISTER_HWACCEL(WMV3_VAAPI, wmv3_vaapi);
|
||||
REGISTER_HWACCEL(WMV3_VDPAU, wmv3_vdpau);
|
||||
|
||||
|
@ -54,7 +54,9 @@ static int map_avcodec_id(enum AVCodecID id)
|
||||
switch (id) {
|
||||
case AV_CODEC_ID_H264: return cudaVideoCodec_H264;
|
||||
case AV_CODEC_ID_HEVC: return cudaVideoCodec_HEVC;
|
||||
case AV_CODEC_ID_VC1: return cudaVideoCodec_VC1;
|
||||
case AV_CODEC_ID_VP9: return cudaVideoCodec_VP9;
|
||||
case AV_CODEC_ID_WMV3: return cudaVideoCodec_VC1;
|
||||
}
|
||||
return -1;
|
||||
}
|
||||
|
184
libavcodec/nvdec_vc1.c
Normal file
184
libavcodec/nvdec_vc1.c
Normal file
@ -0,0 +1,184 @@
|
||||
/*
|
||||
* VC1 HW decode acceleration through NVDEC
|
||||
*
|
||||
* Copyright (c) 2017 Philip Langdale
|
||||
*
|
||||
* 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
|
||||
*/
|
||||
|
||||
#include "avcodec.h"
|
||||
#include "nvdec.h"
|
||||
#include "decode.h"
|
||||
#include "vc1.h"
|
||||
|
||||
static unsigned char get_ref_idx(AVFrame *frame)
|
||||
{
|
||||
FrameDecodeData *fdd;
|
||||
NVDECFrame *cf;
|
||||
|
||||
if (!frame || !frame->private_ref)
|
||||
return 255;
|
||||
|
||||
fdd = (FrameDecodeData*)frame->private_ref->data;
|
||||
cf = (NVDECFrame*)fdd->hwaccel_priv;
|
||||
|
||||
return cf->idx;
|
||||
}
|
||||
|
||||
static int nvdec_vc1_start_frame(AVCodecContext *avctx, const uint8_t *buffer, uint32_t size)
|
||||
{
|
||||
VC1Context *v = avctx->priv_data;
|
||||
MpegEncContext *s = &v->s;
|
||||
|
||||
NVDECContext *ctx = avctx->internal->hwaccel_priv_data;
|
||||
CUVIDPICPARAMS *pp = &ctx->pic_params;
|
||||
FrameDecodeData *fdd;
|
||||
NVDECFrame *cf;
|
||||
AVFrame *cur_frame = s->current_picture.f;
|
||||
|
||||
int ret;
|
||||
|
||||
ret = ff_nvdec_start_frame(avctx, cur_frame);
|
||||
if (ret < 0)
|
||||
return ret;
|
||||
|
||||
fdd = (FrameDecodeData*)cur_frame->private_ref->data;
|
||||
cf = (NVDECFrame*)fdd->hwaccel_priv;
|
||||
|
||||
*pp = (CUVIDPICPARAMS) {
|
||||
.PicWidthInMbs = (cur_frame->width + 15) / 16,
|
||||
.FrameHeightInMbs = (cur_frame->height + 15) / 16,
|
||||
.CurrPicIdx = cf->idx,
|
||||
.field_pic_flag = v->field_mode,
|
||||
.bottom_field_flag = v->cur_field_type,
|
||||
.second_field = v->second_field,
|
||||
|
||||
.intra_pic_flag = s->pict_type == AV_PICTURE_TYPE_I ||
|
||||
s->pict_type == AV_PICTURE_TYPE_BI,
|
||||
.ref_pic_flag = s->pict_type == AV_PICTURE_TYPE_I ||
|
||||
s->pict_type == AV_PICTURE_TYPE_P,
|
||||
|
||||
.CodecSpecific.vc1 = {
|
||||
.ForwardRefIdx = get_ref_idx(s->last_picture.f),
|
||||
.BackwardRefIdx = get_ref_idx(s->next_picture.f),
|
||||
.FrameWidth = cur_frame->width,
|
||||
.FrameHeight = cur_frame->height,
|
||||
|
||||
.intra_pic_flag = s->pict_type == AV_PICTURE_TYPE_I ||
|
||||
s->pict_type == AV_PICTURE_TYPE_BI,
|
||||
.ref_pic_flag = s->pict_type == AV_PICTURE_TYPE_I ||
|
||||
s->pict_type == AV_PICTURE_TYPE_P,
|
||||
.progressive_fcm = v->fcm == 0,
|
||||
|
||||
.profile = v->profile,
|
||||
.postprocflag = v->postprocflag,
|
||||
.pulldown = v->broadcast,
|
||||
.interlace = v->interlace,
|
||||
.tfcntrflag = v->tfcntrflag,
|
||||
.finterpflag = v->finterpflag,
|
||||
.psf = v->psf,
|
||||
.multires = v->multires,
|
||||
.syncmarker = v->resync_marker,
|
||||
.rangered = v->rangered,
|
||||
.maxbframes = s->max_b_frames,
|
||||
|
||||
.panscan_flag = v->panscanflag,
|
||||
.refdist_flag = v->refdist_flag,
|
||||
.extended_mv = v->extended_mv,
|
||||
.dquant = v->dquant,
|
||||
.vstransform = v->vstransform,
|
||||
.loopfilter = v->s.loop_filter,
|
||||
.fastuvmc = v->fastuvmc,
|
||||
.overlap = v->overlap,
|
||||
.quantizer = v->quantizer_mode,
|
||||
.extended_dmv = v->extended_dmv,
|
||||
.range_mapy_flag = v->range_mapy_flag,
|
||||
.range_mapy = v->range_mapy,
|
||||
.range_mapuv_flag = v->range_mapuv_flag,
|
||||
.range_mapuv = v->range_mapuv,
|
||||
.rangeredfrm = v->rangeredfrm,
|
||||
}
|
||||
};
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int nvdec_vc1_end_frame(AVCodecContext *avctx)
|
||||
{
|
||||
NVDECContext *ctx = avctx->internal->hwaccel_priv_data;
|
||||
int ret = ff_nvdec_end_frame(avctx);
|
||||
ctx->bitstream = NULL;
|
||||
return ret;
|
||||
}
|
||||
|
||||
static int nvdec_vc1_decode_slice(AVCodecContext *avctx, const uint8_t *buffer, uint32_t size)
|
||||
{
|
||||
NVDECContext *ctx = avctx->internal->hwaccel_priv_data;
|
||||
void *tmp;
|
||||
|
||||
tmp = av_fast_realloc(ctx->slice_offsets, &ctx->slice_offsets_allocated,
|
||||
(ctx->nb_slices + 1) * sizeof(*ctx->slice_offsets));
|
||||
if (!tmp)
|
||||
return AVERROR(ENOMEM);
|
||||
ctx->slice_offsets = tmp;
|
||||
|
||||
if (!ctx->bitstream)
|
||||
ctx->bitstream = (uint8_t*)buffer;
|
||||
|
||||
ctx->slice_offsets[ctx->nb_slices] = buffer - ctx->bitstream;
|
||||
ctx->bitstream_len += size;
|
||||
ctx->nb_slices++;
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int nvdec_vc1_frame_params(AVCodecContext *avctx,
|
||||
AVBufferRef *hw_frames_ctx)
|
||||
{
|
||||
// Each frame can at most have one P and one B reference
|
||||
return ff_nvdec_frame_params(avctx, hw_frames_ctx, 2);
|
||||
}
|
||||
|
||||
AVHWAccel ff_vc1_nvdec_hwaccel = {
|
||||
.name = "vc1_nvdec",
|
||||
.type = AVMEDIA_TYPE_VIDEO,
|
||||
.id = AV_CODEC_ID_VC1,
|
||||
.pix_fmt = AV_PIX_FMT_CUDA,
|
||||
.start_frame = nvdec_vc1_start_frame,
|
||||
.end_frame = nvdec_vc1_end_frame,
|
||||
.decode_slice = nvdec_vc1_decode_slice,
|
||||
.frame_params = nvdec_vc1_frame_params,
|
||||
.init = ff_nvdec_decode_init,
|
||||
.uninit = ff_nvdec_decode_uninit,
|
||||
.priv_data_size = sizeof(NVDECContext),
|
||||
};
|
||||
|
||||
#if CONFIG_WMV3_NVDEC_HWACCEL
|
||||
AVHWAccel ff_wmv3_nvdec_hwaccel = {
|
||||
.name = "wmv3_nvdec",
|
||||
.type = AVMEDIA_TYPE_VIDEO,
|
||||
.id = AV_CODEC_ID_WMV3,
|
||||
.pix_fmt = AV_PIX_FMT_CUDA,
|
||||
.start_frame = nvdec_vc1_start_frame,
|
||||
.end_frame = nvdec_vc1_end_frame,
|
||||
.decode_slice = nvdec_vc1_decode_slice,
|
||||
.frame_params = nvdec_vc1_frame_params,
|
||||
.init = ff_nvdec_decode_init,
|
||||
.uninit = ff_nvdec_decode_uninit,
|
||||
.priv_data_size = sizeof(NVDECContext),
|
||||
};
|
||||
#endif
|
@ -1119,6 +1119,9 @@ static const enum AVPixelFormat vc1_hwaccel_pixfmt_list_420[] = {
|
||||
AV_PIX_FMT_D3D11VA_VLD,
|
||||
AV_PIX_FMT_D3D11,
|
||||
#endif
|
||||
#if CONFIG_VC1_NVDEC_HWACCEL
|
||||
AV_PIX_FMT_CUDA,
|
||||
#endif
|
||||
#if CONFIG_VC1_VAAPI_HWACCEL
|
||||
AV_PIX_FMT_VAAPI,
|
||||
#endif
|
||||
|
@ -29,7 +29,7 @@
|
||||
|
||||
#define LIBAVCODEC_VERSION_MAJOR 58
|
||||
#define LIBAVCODEC_VERSION_MINOR 3
|
||||
#define LIBAVCODEC_VERSION_MICRO 101
|
||||
#define LIBAVCODEC_VERSION_MICRO 102
|
||||
|
||||
#define LIBAVCODEC_VERSION_INT AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \
|
||||
LIBAVCODEC_VERSION_MINOR, \
|
||||
|
Loading…
Reference in New Issue
Block a user