[FFmpeg-devel] [PATCH] lavc/qsvenc: add the missing vp9 file

Dennis Mungai dmngaie at gmail.com
Fri Nov 8 00:41:42 EET 2019


On Wed, 6 Nov 2019 at 16:31, Fu, Linjie <linjie.fu at intel.com> wrote:
>
> > -----Original Message-----
> > From: ffmpeg-devel <ffmpeg-devel-bounces at ffmpeg.org> On Behalf Of
> > Dennis Mungai
> > Sent: Wednesday, November 6, 2019 20:36
> > To: FFmpeg development discussions and patches <ffmpeg-
> > devel at ffmpeg.org>
> > Cc: Zhong Li <zhongli_dev at 126.com>
> > Subject: Re: [FFmpeg-devel] [PATCH] lavc/qsvenc: add the missing vp9 file
> >
> > On Tue, 5 Nov 2019 at 12:22, Fu, Linjie <linjie.fu at intel.com> wrote:
> > >
> > > > -----Original Message-----
> > > > From: ffmpeg-devel <ffmpeg-devel-bounces at ffmpeg.org> On Behalf Of
> > > > Dennis Mungai
> > > > Sent: Tuesday, November 5, 2019 13:04
> > > > To: FFmpeg development discussions and patches <ffmpeg-
> > > > devel at ffmpeg.org>
> > > > Cc: Zhong Li <zhongli_dev at 126.com>
> > > > Subject: Re: [FFmpeg-devel] [PATCH] lavc/qsvenc: add the missing vp9
> > file
> > > >
> > > > On Mon, 4 Nov 2019 at 10:12, Zhong Li <zhongli_dev at 126.com> wrote:
> > > > >
> > > > > It is missed in commit 3358380
> > > > >
> > > > > Signed-off-by: Zhong Li <zhongli_dev at 126.com>
> > > > > ---
> > > > >  libavcodec/qsvenc_vp9.c | 113
> > > > ++++++++++++++++++++++++++++++++++++++++++++++++
> > > > >  1 file changed, 113 insertions(+)
> > > > >  create mode 100644 libavcodec/qsvenc_vp9.c
> > > > >
> > > > > diff --git a/libavcodec/qsvenc_vp9.c b/libavcodec/qsvenc_vp9.c
> > > > > new file mode 100644
> > > > > index 0000000..9402f80
> > > > > --- /dev/null
> > > > > +++ b/libavcodec/qsvenc_vp9.c
> > > > > @@ -0,0 +1,113 @@
> > > > > +/*
> > > > > + * Intel MediaSDK QSV based VP9 encoder
> > > > > + *
> > > > > + * 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 <stdint.h>
> > > > > +#include <sys/types.h>
> > > > > +
> > > > > +#include <mfx/mfxvideo.h>
> > > > > +
> > > > > +#include "libavutil/common.h"
> > > > > +#include "libavutil/opt.h"
> > > > > +
> > > > > +#include "avcodec.h"
> > > > > +#include "internal.h"
> > > > > +#include "qsv.h"
> > > > > +#include "qsv_internal.h"
> > > > > +#include "qsvenc.h"
> > > > > +
> > > > > +typedef struct QSVVP9EncContext {
> > > > > +    AVClass *class;
> > > > > +    QSVEncContext qsv;
> > > > > +} QSVVP9EncContext;
> > > > > +
> > > > > +static av_cold int qsv_enc_init(AVCodecContext *avctx)
> > > > > +{
> > > > > +    QSVVP9EncContext *q = avctx->priv_data;
> > > > > +    q->qsv.low_power = 1;
> > > > > +
> > > > > +    return ff_qsv_enc_init(avctx, &q->qsv);
> > > > > +}
> > > > > +
> > > > > +static int qsv_enc_frame(AVCodecContext *avctx, AVPacket *pkt,
> > > > > +                         const AVFrame *frame, int *got_packet)
> > > > > +{
> > > > > +    QSVVP9EncContext *q = avctx->priv_data;
> > > > > +
> > > > > +    return ff_qsv_encode(avctx, &q->qsv, pkt, frame, got_packet);
> > > > > +}
> > > > > +
> > > > > +static av_cold int qsv_enc_close(AVCodecContext *avctx)
> > > > > +{
> > > > > +    QSVVP9EncContext *q = avctx->priv_data;
> > > > > +
> > > > > +    return ff_qsv_enc_close(avctx, &q->qsv);
> > > > > +}
> > > > > +
> > > > > +#define OFFSET(x) offsetof(QSVVP9EncContext, x)
> > > > > +#define VE AV_OPT_FLAG_VIDEO_PARAM |
> > > > AV_OPT_FLAG_ENCODING_PARAM
> > > > > +static const AVOption options[] = {
> > > > > +    QSV_COMMON_OPTS
> > > > > +
> > > > > +    { "profile",   NULL, OFFSET(qsv.profile), AV_OPT_TYPE_INT,   { .i64 =
> > > > MFX_PROFILE_UNKNOWN },   0,       INT_MAX,  VE,  "profile" },
> > > > > +    { "unknown",   NULL, 0,                   AV_OPT_TYPE_CONST, { .i64 =
> > > > MFX_PROFILE_UNKNOWN},   INT_MIN,  INT_MAX,  VE,  "profile" },
> > > > > +    { "profile0",  NULL, 0,                   AV_OPT_TYPE_CONST, { .i64 =
> > > > MFX_PROFILE_VP9_0   },  INT_MIN,  INT_MAX,  VE,  "profile" },
> > > > > +    { "profile1",  NULL, 0,                   AV_OPT_TYPE_CONST, { .i64 =
> > > > MFX_PROFILE_VP9_1   },  INT_MIN,  INT_MAX,  VE,  "profile" },
> > > > > +    { "profile2",  NULL, 0,                   AV_OPT_TYPE_CONST, { .i64 =
> > > > MFX_PROFILE_VP9_2   },  INT_MIN,  INT_MAX,  VE,  "profile" },
> > > > > +    { "profile3",  NULL, 0,                   AV_OPT_TYPE_CONST, { .i64 =
> > > > MFX_PROFILE_VP9_3   },  INT_MIN,  INT_MAX,  VE,  "profile" },
> > > > > +
> > > > > +    { NULL },
> > > > > +};
> > > > > +
> > > > > +static const AVClass class = {
> > > > > +    .class_name = "vp9_qsv encoder",
> > > > > +    .item_name  = av_default_item_name,
> > > > > +    .option     = options,
> > > > > +    .version    = LIBAVUTIL_VERSION_INT,
> > > > > +};
> > > > > +
> > > > > +static const AVCodecDefault qsv_enc_defaults[] = {
> > > > > +    { "b",         "1M"    },
> > > > > +    { "refs",      "0"     },
> > > > > +    { "g",         "250"   },
> > > > > +    { "trellis",   "-1"    },
> > > > > +    { "flags",     "+cgop" },
> > > > > +    { NULL },
> > > > > +};
> > > > > +
> > > > > +AVCodec ff_vp9_qsv_encoder = {
> > > > > +    .name           = "vp9_qsv",
> > > > > +    .long_name      = NULL_IF_CONFIG_SMALL("VP9 video (Intel Quick
> > Sync
> > > > Video acceleration)"),
> > > > > +    .priv_data_size = sizeof(QSVVP9EncContext),
> > > > > +    .type           = AVMEDIA_TYPE_VIDEO,
> > > > > +    .id             = AV_CODEC_ID_VP9,
> > > > > +    .init           = qsv_enc_init,
> > > > > +    .encode2        = qsv_enc_frame,
> > > > > +    .close          = qsv_enc_close,
> > > > > +    .capabilities   = AV_CODEC_CAP_DELAY | AV_CODEC_CAP_HYBRID,
> > > > > +    .pix_fmts       = (const enum AVPixelFormat[]){ AV_PIX_FMT_NV12,
> > > > > +                                                    AV_PIX_FMT_P010,
> > > > > +                                                    AV_PIX_FMT_QSV,
> > > > > +                                                    AV_PIX_FMT_NONE },
> > > > > +    .priv_class     = &class,
> > > > > +    .defaults       = qsv_enc_defaults,
> > > > > +    .caps_internal  = FF_CODEC_CAP_INIT_CLEANUP,
> > > > > +    .wrapper_name   = "qsv",
> > > > > +};
> > > > > --
> > > > > 1.8.3.1
> > > > >
> > > >
> > > >
> > > > Hello Mr Zhong,
> > > >
> > > > What hardware is supported by this encoder?
> > > > On  CFL, even with low power encoding enabled, I get:
> > > >
> > > > [vp9_qsv @ 0x55dff61e65c0] Current profile is unsupported
> > > > [vp9_qsv @ 0x55dff61e65c0] Selected ratecontrol mode is unsupported
> > > > [vp9_qsv @ 0x55dff61e65c0] Low power mode is unsupported
> > > > [vp9_qsv @ 0x55dff61e65c0] Current frame rate is unsupported
> > > > [vp9_qsv @ 0x55dff61e65c0] Current picture structure is unsupported
> > > > [vp9_qsv @ 0x55dff61e65c0] Current resolution is unsupported
> > > > [vp9_qsv @ 0x55dff61e65c0] Current pixel format is unsupported
> > > > [vp9_qsv @ 0x55dff61e65c0] some encoding parameters are not
> > supported
> > > > by the QSV runtime. Please double check the input parameters.
> > > > Error initializing output stream 0:0 -- Error while opening encoder
> > > > for output stream #0:0 - maybe incorrect parameters such as bit_rate,
> > > > rate, width or height
> > > > [aac @ 0x55dff6184300] Qavg: 166.914
> > > > [aac @ 0x55dff6184300] 2 frames left in the queue on closing
> > > > Conversion failed!
> > >
> > > If media-driver is used, vp9 is supported since ICL:
> > > https://github.com/intel/media-
> > driver/blob/master/docs/media_features.md#supported-encoding-input-
> > format-and-max-resolution
> > >
> > > Also you could check with "vainfo" in the terminal to see the supported
> > VAEntrypoint.
> > >
> > > $ vainfo
> > > ...
> > >      VAProfileVP9Profile0            : VAEntrypointVLD
> > >       VAProfileVP9Profile0            : VAEntrypointEncSliceLP
> > >       VAProfileVP9Profile1            : VAEntrypointVLD
> > >       VAProfileVP9Profile1            : VAEntrypointEncSliceLP
> > >       VAProfileVP9Profile2            : VAEntrypointVLD
> > >       VAProfileVP9Profile2            : VAEntrypointEncSliceLP
> > >       VAProfileVP9Profile3            : VAEntrypointVLD
> > >       VAProfileVP9Profile3            : VAEntrypointEncSliceLP
> > > ...
> > >
> > > - linjie
> >
> > I see, thanks.
> >
> > Will this ever be back-ported to Kabylake+?
> > These systems *can* handle VP9 encoding, at least in VAAPI.
>
> VP9 encoding on Kaby lake is supported by intel-vaapi driver(i965) for VAAPI.
> IIRC, MSDK seems to only support iHD:
> https://github.com/Intel-Media-SDK/MediaSDK/wiki/Intel-media-stack-on-Ubuntu#available-drivers
>
> - linjie

I see, thanks for the clarification.
In the future, can customers expect this feature (in iHD) to be
back-ported to prior Intel GPUs with the VP9 encoder built in?
Is there any technical merit as to why iHD does not expose VP9
encoding in KBL+ whereas the i965 driver does?


More information about the ffmpeg-devel mailing list