avcodec/nvdec: Implement vp8 hwaccel

This commit is contained in:
Philip Langdale 2017-11-19 08:42:39 -08:00
parent f69e9365f6
commit 1da9851e34
8 changed files with 110 additions and 2 deletions

View File

@ -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, MPEG-1/2/4, VC1 and VP9 hwaccel decoding
- NVIDIA NVDEC-accelerated H.264, HEVC, MPEG-1/2/4, VC1, VP8/9 hwaccel decoding
- Intel QSV-accelerated overlay filter
- mcompand audio filter
- acontrast audio filter

2
configure vendored
View File

@ -2746,6 +2746,8 @@ vc1_vaapi_hwaccel_deps="vaapi"
vc1_vaapi_hwaccel_select="vc1_decoder"
vc1_vdpau_hwaccel_deps="vdpau"
vc1_vdpau_hwaccel_select="vc1_decoder"
vp8_nvdec_hwaccel_deps="nvdec"
vp8_nvdec_hwaccel_select="vp8_decoder"
vp8_vaapi_hwaccel_deps="vaapi VAPictureParameterBufferVP8"
vp8_vaapi_hwaccel_select="vp8_decoder"
vp9_d3d11va_hwaccel_deps="d3d11va DXVA_PicParams_VP9"

View File

@ -871,6 +871,7 @@ 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
OBJS-$(CONFIG_VP8_NVDEC_HWACCEL) += nvdec_vp8.o
OBJS-$(CONFIG_VP8_VAAPI_HWACCEL) += vaapi_vp8.o
OBJS-$(CONFIG_VP9_D3D11VA_HWACCEL) += dxva2_vp9.o
OBJS-$(CONFIG_VP9_DXVA2_HWACCEL) += dxva2_vp9.o

View File

@ -59,6 +59,7 @@ extern const AVHWAccel ff_vc1_dxva2_hwaccel;
extern const AVHWAccel ff_vc1_nvdec_hwaccel;
extern const AVHWAccel ff_vc1_vaapi_hwaccel;
extern const AVHWAccel ff_vc1_vdpau_hwaccel;
extern const AVHWAccel ff_vp8_nvdec_hwaccel;
extern const AVHWAccel ff_vp8_vaapi_hwaccel;
extern const AVHWAccel ff_vp9_d3d11va_hwaccel;
extern const AVHWAccel ff_vp9_d3d11va2_hwaccel;

View File

@ -58,6 +58,7 @@ static int map_avcodec_id(enum AVCodecID id)
case AV_CODEC_ID_MPEG2VIDEO: return cudaVideoCodec_MPEG2;
case AV_CODEC_ID_MPEG4: return cudaVideoCodec_MPEG4;
case AV_CODEC_ID_VC1: return cudaVideoCodec_VC1;
case AV_CODEC_ID_VP8: return cudaVideoCodec_VP8;
case AV_CODEC_ID_VP9: return cudaVideoCodec_VP9;
case AV_CODEC_ID_WMV3: return cudaVideoCodec_VC1;
}

97
libavcodec/nvdec_vp8.c Normal file
View File

@ -0,0 +1,97 @@
/*
* VP8 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 "internal.h"
#include "vp8.h"
static unsigned char safe_get_ref_idx(VP8Frame *frame)
{
return frame ? ff_nvdec_get_ref_idx(frame->tf.f) : 255;
}
static int nvdec_vp8_start_frame(AVCodecContext *avctx, const uint8_t *buffer, uint32_t size)
{
VP8Context *h = avctx->priv_data;
NVDECContext *ctx = avctx->internal->hwaccel_priv_data;
CUVIDPICPARAMS *pp = &ctx->pic_params;
FrameDecodeData *fdd;
NVDECFrame *cf;
AVFrame *cur_frame = h->framep[VP56_FRAME_CURRENT]->tf.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,
.CodecSpecific.vp8 = {
.width = cur_frame->width,
.height = cur_frame->height,
.first_partition_size = h->header_partition_size,
.LastRefIdx = safe_get_ref_idx(h->framep[VP56_FRAME_PREVIOUS]),
.GoldenRefIdx = safe_get_ref_idx(h->framep[VP56_FRAME_GOLDEN]),
.AltRefIdx = safe_get_ref_idx(h->framep[VP56_FRAME_GOLDEN2]),
.frame_type = !h->keyframe,
.version = h->profile,
.show_frame = !h->invisible,
.update_mb_segmentation_data = h->segmentation.enabled ? h->segmentation.update_feature_data : 0,
}
};
return 0;
}
static int nvdec_vp8_frame_params(AVCodecContext *avctx,
AVBufferRef *hw_frames_ctx)
{
// VP8 uses a fixed size pool of 3 possible reference frames
return ff_nvdec_frame_params(avctx, hw_frames_ctx, 3);
}
AVHWAccel ff_vp8_nvdec_hwaccel = {
.name = "vp8_nvdec",
.type = AVMEDIA_TYPE_VIDEO,
.id = AV_CODEC_ID_VP8,
.pix_fmt = AV_PIX_FMT_CUDA,
.start_frame = nvdec_vp8_start_frame,
.end_frame = ff_nvdec_simple_end_frame,
.decode_slice = ff_nvdec_simple_decode_slice,
.frame_params = nvdec_vp8_frame_params,
.init = ff_nvdec_decode_init,
.uninit = ff_nvdec_decode_uninit,
.priv_data_size = sizeof(NVDECContext),
};

View File

@ -29,7 +29,7 @@
#define LIBAVCODEC_VERSION_MAJOR 58
#define LIBAVCODEC_VERSION_MINOR 6
#define LIBAVCODEC_VERSION_MICRO 100
#define LIBAVCODEC_VERSION_MICRO 101
#define LIBAVCODEC_VERSION_INT AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \
LIBAVCODEC_VERSION_MINOR, \

View File

@ -2601,6 +2601,9 @@ int vp78_decode_frame(AVCodecContext *avctx, void *data, int *got_frame,
enum AVPixelFormat pix_fmts[] = {
#if CONFIG_VP8_VAAPI_HWACCEL
AV_PIX_FMT_VAAPI,
#endif
#if CONFIG_VP8_NVDEC_HWACCEL
AV_PIX_FMT_CUDA,
#endif
AV_PIX_FMT_YUV420P,
AV_PIX_FMT_NONE,
@ -2949,6 +2952,9 @@ AVCodec ff_vp8_decoder = {
.hw_configs = (const AVCodecHWConfigInternal*[]) {
#if CONFIG_VP8_VAAPI_HWACCEL
HWACCEL_VAAPI(vp8),
#endif
#if CONFIG_VP8_NVDEC_HWACCEL
HWACCEL_NVDEC(vp8),
#endif
NULL
},