avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11) #15

softworkz · 2022-01-02T03:10:33Z

Fix a recently introduced regression when using QSV VPP.

v2: Fixed commit message wrapping

v3: Use different way to acquire runtime version

v4: Use mfxStatus as return type for get_mfx_version()

cc: "Xiang, Haihao" haihao.xiang-at-intel.com@ffmpeg.org
cc: Soft Works softworkz@hotmail.com

softworkz · 2022-01-02T03:11:50Z

/submit

ffmpeg-codebot · 2022-01-02T03:12:18Z

Submitted as pull.15.ffstaging.FFmpeg.1641093136327.ffmpegagent@gmail.com

To fetch this version into FETCH_HEAD:

git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-15/softworkz/qsv_vpp_regression-v1

To fetch this version to local tag pr-ffstaging-15/softworkz/qsv_vpp_regression-v1:

git fetch --no-tags https://github.com/ffstaging/FFmpeg tag pr-ffstaging-15/softworkz/qsv_vpp_regression-v1

softworkz · 2022-01-02T03:41:31Z

/submit

ffmpeg-codebot · 2022-01-02T03:41:53Z

Submitted as pull.15.v2.ffstaging.FFmpeg.1641094910660.ffmpegagent@gmail.com

To fetch this version into FETCH_HEAD:

git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-15/softworkz/qsv_vpp_regression-v2

To fetch this version to local tag pr-ffstaging-15/softworkz/qsv_vpp_regression-v2:

git fetch --no-tags https://github.com/ffstaging/FFmpeg tag pr-ffstaging-15/softworkz/qsv_vpp_regression-v2

ffmpeg-codebot · 2022-01-04T02:18:43Z

On the FFmpeg mailing list, "Xiang, Haihao" wrote (reply to this):

On Sun, 2022-01-02 at 03:41 +0000, ffmpegagent wrote:
> From: softworkz <softworkz@hotmail.com>
> 
> Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> regression in a way that scaling via vpp_qsv doesn't work any longer
> for devices with an MSDK runtime version lower than 1.19. This is true
> for older CPUs which are stuck at 1.11.
> The commit added checks for the compile-sdk version but it didn't test
> for the runtime version.
> 
> Signed-off-by: softworkz <softworkz@hotmail.com>
> ---
>     avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
>     
>     Fix a recently introduced regression when using QSV VPP.
>     
>     v2: Fixed commit message wrapping
> 
> Published-As: 
> https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-15%2Fsoftworkz%2Fqsv_vpp_regression-v2
> Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-
> 15/softworkz/qsv_vpp_regression-v2
> Pull-Request: https://github.com/ffstaging/FFmpeg/pull/15
> 
> Range-diff vs v1:
> 
>  1:  eea966f702 ! 1:  ccede9d840 avfilter/vpp_qsv: fix regression on older api
> versions (e.g. 1.11)
>      @@ Metadata
>        ## Commit message ##
>           avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
>       
>      -    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> regression in a way
>      -    that scaling via vpp_qsv didn't work any longer for devices with an
> MSDK runtime
>      -    version lower than 1.19. This is true for older CPUs which are stuck
> at 1.11.
>      -    The commit added checks for the compile-sdk version but it didn't
> test for the
>      -    runtime version.
>      +    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
>      +    regression in a way that scaling via vpp_qsv doesn't work any longer
>      +    for devices with an MSDK runtime version lower than 1.19. This is
> true
>      +    for older CPUs which are stuck at 1.11.
>      +    The commit added checks for the compile-sdk version but it didn't
> test
>      +    for the runtime version.
>       
>           Signed-off-by: softworkz <softworkz@hotmail.com>
>       
> 
> 
>  libavfilter/vf_vpp_qsv.c | 32 ++++++++++++++++++++------------
>  1 file changed, 20 insertions(+), 12 deletions(-)
> 
> diff --git a/libavfilter/vf_vpp_qsv.c b/libavfilter/vf_vpp_qsv.c
> index ab58a5777e..09590157e3 100644
> --- a/libavfilter/vf_vpp_qsv.c
> +++ b/libavfilter/vf_vpp_qsv.c
> @@ -139,8 +139,9 @@ static const AVOption options[] = {
>      { "height", "Output video height", OFFSET(oh), AV_OPT_TYPE_STRING, {
> .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
>      { "format", "Output pixel format", OFFSET(output_format_str),
> AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
>      { "async_depth", "Internal parallelization depth, the higher the value
> the higher the latency.", OFFSET(async_depth), AV_OPT_TYPE_INT, { .i64 = 0 },
> 0, INT_MAX, .flags = FLAGS },
> +#ifdef QSV_HAVE_SCALING_CONFIG
>      { "scale_mode", "scale mode: 0=auto, 1=low power, 2=high quality",
> OFFSET(scale_mode), AV_OPT_TYPE_INT, { .i64 = MFX_SCALING_MODE_DEFAULT },
> MFX_SCALING_MODE_DEFAULT, MFX_SCALING_MODE_QUALITY, .flags = FLAGS, "scale
> mode" },
> -
> +#endif
>      { NULL }
>  };
>  
> @@ -304,6 +305,7 @@ static int config_output(AVFilterLink *outlink)
>      QSVVPPParam     param = { NULL };
>      QSVVPPCrop      crop  = { 0 };
>      mfxExtBuffer    *ext_buf[ENH_FILTERS_COUNT];
> +    mfxVersion      mfx_version;
>      AVFilterLink    *inlink = ctx->inputs[0];
>      enum AVPixelFormat in_format;
>  
> @@ -317,6 +319,11 @@ static int config_output(AVFilterLink *outlink)
>      param.ext_buf       = ext_buf;
>      param.async_depth   = vpp->async_depth;
>  
> +    if (MFXQueryVersion(vpp->qsv->session, &mfx_version) < 0) {
> +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
> +        return AVERROR(EINVAL);
> +    }
> +
>      if (inlink->format == AV_PIX_FMT_QSV) {
>           if (!inlink->hw_frames_ctx || !inlink->hw_frames_ctx->data)
>               return AVERROR(EINVAL);
> @@ -467,19 +474,20 @@ static int config_output(AVFilterLink *outlink)
>  #endif
>      }
>  
> -    if (inlink->w != outlink->w || inlink->h != outlink->h) {
>  #ifdef QSV_HAVE_SCALING_CONFIG
> -        memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> -        vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> -        vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> -        vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> -
> -        param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp->scale_conf;
> -#else
> -        av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> -            "not supported with this MSDK version.\n");
> -#endif
> +    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> +        if (QSV_RUNTIME_VERSION_ATLEAST(mfx_version, 1, 19)) {
> +            memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> +            vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> +            vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> +            vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> +
> +            param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> >scale_conf;
> +        } else
> +            av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> +                "not supported with this MSDK version.\n");
>      }
> +#endif
>  
>      if (vpp->use_frc || vpp->use_crop || vpp->deinterlace || vpp->denoise ||
>          vpp->detail || vpp->procamp || vpp->rotate || vpp->hflip ||

LGTM. I wonder whether you installed libmfx and msdk runtime from 
different packages, so the mfx version is higher than the runtime version? 

Thanks
Haihao


> 
> base-commit: d6b2357eddca392ee137cb2a92ff178a0a7d0cce
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

ffmpeg-codebot · 2022-01-04T02:18:45Z

User "Xiang, Haihao" <haihao.xiang-at-intel.com@ffmpeg.org> has been added to the cc: list.

ffmpeg-codebot · 2022-01-04T02:26:45Z

On the FFmpeg mailing list, Soft Works wrote (reply to this):



> -----Original Message-----
> From: Xiang, Haihao <haihao.xiang@intel.com>
> Sent: Tuesday, January 4, 2022 3:16 AM
> To: ffmpeg-devel@ffmpeg.org
> Cc: softworkz@hotmail.com
> Subject: Re: [FFmpeg-devel] [PATCH v2] avfilter/vpp_qsv: fix regression on
> older api versions (e.g. 1.11)
> 
> On Sun, 2022-01-02 at 03:41 +0000, ffmpegagent wrote:
> > From: softworkz <softworkz@hotmail.com>
> >
> > Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > regression in a way that scaling via vpp_qsv doesn't work any longer
> > for devices with an MSDK runtime version lower than 1.19. This is true
> > for older CPUs which are stuck at 1.11.
> > The commit added checks for the compile-sdk version but it didn't test
> > for the runtime version.
> >
> > Signed-off-by: softworkz <softworkz@hotmail.com>
> > ---
> >     avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
> >
> >     Fix a recently introduced regression when using QSV VPP.
> >
> >     v2: Fixed commit message wrapping
> >
> > Published-As:
> > https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-
> 15%2Fsoftworkz%2Fqsv_vpp_regression-v2
> > Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-
> > 15/softworkz/qsv_vpp_regression-v2
> > Pull-Request: https://github.com/ffstaging/FFmpeg/pull/15
> >
> > Range-diff vs v1:
> >
> >  1:  eea966f702 ! 1:  ccede9d840 avfilter/vpp_qsv: fix regression on older
> api
> > versions (e.g. 1.11)
> >      @@ Metadata
> >        ## Commit message ##
> >           avfilter/vpp_qsv: fix regression on older api versions (e.g.
> 1.11)
> >
> >      -    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > regression in a way
> >      -    that scaling via vpp_qsv didn't work any longer for devices with
> an
> > MSDK runtime
> >      -    version lower than 1.19. This is true for older CPUs which are
> stuck
> > at 1.11.
> >      -    The commit added checks for the compile-sdk version but it didn't
> > test for the
> >      -    runtime version.
> >      +    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> >      +    regression in a way that scaling via vpp_qsv doesn't work any
> longer
> >      +    for devices with an MSDK runtime version lower than 1.19. This is
> > true
> >      +    for older CPUs which are stuck at 1.11.
> >      +    The commit added checks for the compile-sdk version but it didn't
> > test
> >      +    for the runtime version.
> >
> >           Signed-off-by: softworkz <softworkz@hotmail.com>
> >
> >
> >
> >  libavfilter/vf_vpp_qsv.c | 32 ++++++++++++++++++++------------
> >  1 file changed, 20 insertions(+), 12 deletions(-)
> >
> > diff --git a/libavfilter/vf_vpp_qsv.c b/libavfilter/vf_vpp_qsv.c
> > index ab58a5777e..09590157e3 100644
> > --- a/libavfilter/vf_vpp_qsv.c
> > +++ b/libavfilter/vf_vpp_qsv.c
> > @@ -139,8 +139,9 @@ static const AVOption options[] = {
> >      { "height", "Output video height", OFFSET(oh), AV_OPT_TYPE_STRING, {
> > .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
> >      { "format", "Output pixel format", OFFSET(output_format_str),
> > AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
> >      { "async_depth", "Internal parallelization depth, the higher the value
> > the higher the latency.", OFFSET(async_depth), AV_OPT_TYPE_INT, { .i64 = 0
> },
> > 0, INT_MAX, .flags = FLAGS },
> > +#ifdef QSV_HAVE_SCALING_CONFIG
> >      { "scale_mode", "scale mode: 0=auto, 1=low power, 2=high quality",
> > OFFSET(scale_mode), AV_OPT_TYPE_INT, { .i64 = MFX_SCALING_MODE_DEFAULT },
> > MFX_SCALING_MODE_DEFAULT, MFX_SCALING_MODE_QUALITY, .flags = FLAGS, "scale
> > mode" },
> > -
> > +#endif
> >      { NULL }
> >  };
> >
> > @@ -304,6 +305,7 @@ static int config_output(AVFilterLink *outlink)
> >      QSVVPPParam     param = { NULL };
> >      QSVVPPCrop      crop  = { 0 };
> >      mfxExtBuffer    *ext_buf[ENH_FILTERS_COUNT];
> > +    mfxVersion      mfx_version;
> >      AVFilterLink    *inlink = ctx->inputs[0];
> >      enum AVPixelFormat in_format;
> >
> > @@ -317,6 +319,11 @@ static int config_output(AVFilterLink *outlink)
> >      param.ext_buf       = ext_buf;
> >      param.async_depth   = vpp->async_depth;
> >
> > +    if (MFXQueryVersion(vpp->qsv->session, &mfx_version) < 0) {
> > +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
> > +        return AVERROR(EINVAL);
> > +    }
> > +
> >      if (inlink->format == AV_PIX_FMT_QSV) {
> >           if (!inlink->hw_frames_ctx || !inlink->hw_frames_ctx->data)
> >               return AVERROR(EINVAL);
> > @@ -467,19 +474,20 @@ static int config_output(AVFilterLink *outlink)
> >  #endif
> >      }
> >
> > -    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> >  #ifdef QSV_HAVE_SCALING_CONFIG
> > -        memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> > -        vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> > -        vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> > -        vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> > -
> > -        param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> >scale_conf;
> > -#else
> > -        av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> > -            "not supported with this MSDK version.\n");
> > -#endif
> > +    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> > +        if (QSV_RUNTIME_VERSION_ATLEAST(mfx_version, 1, 19)) {
> > +            memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> > +            vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> > +            vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> > +            vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> > +
> > +            param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> > >scale_conf;
> > +        } else
> > +            av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> > +                "not supported with this MSDK version.\n");
> >      }
> > +#endif
> >
> >      if (vpp->use_frc || vpp->use_crop || vpp->deinterlace || vpp->denoise
> ||
> >          vpp->detail || vpp->procamp || vpp->rotate || vpp->hflip ||
> 
> LGTM. I wonder whether you installed libmfx and msdk runtime from
> different packages, so the mfx version is higher than the runtime version?

It's about a user with a Broadwell CPU on Windows. The latest drivers' MSDK 
version is 1.11 and our ffmpeg is compiled against MSDK 1.35.
(a setup that is basically working fine, just the runtime checks are missing 
at several places)

Thanks,
softworkz

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

ffmpeg-codebot · 2022-01-04T02:26:46Z

User Soft Works <softworkz@hotmail.com> has been added to the cc: list.

ffmpeg-codebot · 2022-01-04T02:38:51Z

On the FFmpeg mailing list, "Xiang, Haihao" wrote (reply to this):

On Tue, 2022-01-04 at 02:25 +0000, Soft Works wrote:
> > -----Original Message-----
> > From: Xiang, Haihao <haihao.xiang@intel.com>
> > Sent: Tuesday, January 4, 2022 3:16 AM
> > To: ffmpeg-devel@ffmpeg.org
> > Cc: softworkz@hotmail.com
> > Subject: Re: [FFmpeg-devel] [PATCH v2] avfilter/vpp_qsv: fix regression on
> > older api versions (e.g. 1.11)
> > 
> > On Sun, 2022-01-02 at 03:41 +0000, ffmpegagent wrote:
> > > From: softworkz <softworkz@hotmail.com>
> > > 
> > > Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > > regression in a way that scaling via vpp_qsv doesn't work any longer
> > > for devices with an MSDK runtime version lower than 1.19. This is true
> > > for older CPUs which are stuck at 1.11.
> > > The commit added checks for the compile-sdk version but it didn't test
> > > for the runtime version.
> > > 
> > > Signed-off-by: softworkz <softworkz@hotmail.com>
> > > ---
> > >     avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
> > > 
> > >     Fix a recently introduced regression when using QSV VPP.
> > > 
> > >     v2: Fixed commit message wrapping
> > > 
> > > Published-As:
> > > https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-
> > 
> > 15%2Fsoftworkz%2Fqsv_vpp_regression-v2
> > > Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-
> > > 15/softworkz/qsv_vpp_regression-v2
> > > Pull-Request: https://github.com/ffstaging/FFmpeg/pull/15
> > > 
> > > Range-diff vs v1:
> > > 
> > >  1:  eea966f702 ! 1:  ccede9d840 avfilter/vpp_qsv: fix regression on older
> > 
> > api
> > > versions (e.g. 1.11)
> > >      @@ Metadata
> > >        ## Commit message ##
> > >           avfilter/vpp_qsv: fix regression on older api versions (e.g.
> > 
> > 1.11)
> > > 
> > >      -    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > > regression in a way
> > >      -    that scaling via vpp_qsv didn't work any longer for devices with
> > 
> > an
> > > MSDK runtime
> > >      -    version lower than 1.19. This is true for older CPUs which are
> > 
> > stuck
> > > at 1.11.
> > >      -    The commit added checks for the compile-sdk version but it
> > > didn't
> > > test for the
> > >      -    runtime version.
> > >      +    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > >      +    regression in a way that scaling via vpp_qsv doesn't work any
> > 
> > longer
> > >      +    for devices with an MSDK runtime version lower than 1.19. This
> > > is
> > > true
> > >      +    for older CPUs which are stuck at 1.11.
> > >      +    The commit added checks for the compile-sdk version but it
> > > didn't
> > > test
> > >      +    for the runtime version.
> > > 
> > >           Signed-off-by: softworkz <softworkz@hotmail.com>
> > > 
> > > 
> > > 
> > >  libavfilter/vf_vpp_qsv.c | 32 ++++++++++++++++++++------------
> > >  1 file changed, 20 insertions(+), 12 deletions(-)
> > > 
> > > diff --git a/libavfilter/vf_vpp_qsv.c b/libavfilter/vf_vpp_qsv.c
> > > index ab58a5777e..09590157e3 100644
> > > --- a/libavfilter/vf_vpp_qsv.c
> > > +++ b/libavfilter/vf_vpp_qsv.c
> > > @@ -139,8 +139,9 @@ static const AVOption options[] = {
> > >      { "height", "Output video height", OFFSET(oh), AV_OPT_TYPE_STRING, {
> > > .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
> > >      { "format", "Output pixel format", OFFSET(output_format_str),
> > > AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
> > >      { "async_depth", "Internal parallelization depth, the higher the
> > > value
> > > the higher the latency.", OFFSET(async_depth), AV_OPT_TYPE_INT, { .i64 = 0
> > 
> > },
> > > 0, INT_MAX, .flags = FLAGS },
> > > +#ifdef QSV_HAVE_SCALING_CONFIG
> > >      { "scale_mode", "scale mode: 0=auto, 1=low power, 2=high quality",
> > > OFFSET(scale_mode), AV_OPT_TYPE_INT, { .i64 = MFX_SCALING_MODE_DEFAULT },
> > > MFX_SCALING_MODE_DEFAULT, MFX_SCALING_MODE_QUALITY, .flags = FLAGS, "scale
> > > mode" },
> > > -
> > > +#endif
> > >      { NULL }
> > >  };
> > > 
> > > @@ -304,6 +305,7 @@ static int config_output(AVFilterLink *outlink)
> > >      QSVVPPParam     param = { NULL };
> > >      QSVVPPCrop      crop  = { 0 };
> > >      mfxExtBuffer    *ext_buf[ENH_FILTERS_COUNT];
> > > +    mfxVersion      mfx_version;
> > >      AVFilterLink    *inlink = ctx->inputs[0];
> > >      enum AVPixelFormat in_format;
> > > 
> > > @@ -317,6 +319,11 @@ static int config_output(AVFilterLink *outlink)
> > >      param.ext_buf       = ext_buf;
> > >      param.async_depth   = vpp->async_depth;
> > > 
> > > +    if (MFXQueryVersion(vpp->qsv->session, &mfx_version) < 0) {
> > > +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
> > > +        return AVERROR(EINVAL);
> > > +    }
> > > +
> > >      if (inlink->format == AV_PIX_FMT_QSV) {
> > >           if (!inlink->hw_frames_ctx || !inlink->hw_frames_ctx->data)
> > >               return AVERROR(EINVAL);
> > > @@ -467,19 +474,20 @@ static int config_output(AVFilterLink *outlink)
> > >  #endif
> > >      }
> > > 
> > > -    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> > >  #ifdef QSV_HAVE_SCALING_CONFIG
> > > -        memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> > > -        vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> > > -        vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> > > -        vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> > > -
> > > -        param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> > > scale_conf;
> > > -#else
> > > -        av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> > > -            "not supported with this MSDK version.\n");
> > > -#endif
> > > +    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> > > +        if (QSV_RUNTIME_VERSION_ATLEAST(mfx_version, 1, 19)) {
> > > +            memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> > > +            vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> > > +            vpp->scale_conf.Header.BufferSz    =
> > > sizeof(mfxExtVPPScaling);
> > > +            vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> > > +
> > > +            param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> > > > scale_conf;
> > > 
> > > +        } else
> > > +            av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> > > +                "not supported with this MSDK version.\n");
> > >      }
> > > +#endif
> > > 
> > >      if (vpp->use_frc || vpp->use_crop || vpp->deinterlace || vpp->denoise
> > > > 
> > > 
> > >          vpp->detail || vpp->procamp || vpp->rotate || vpp->hflip ||
> > 
> > LGTM. I wonder whether you installed libmfx and msdk runtime from
> > different packages, so the mfx version is higher than the runtime version?
> 
> It's about a user with a Broadwell CPU on Windows. The latest drivers' MSDK 
> version is 1.11 and our ffmpeg is compiled against MSDK 1.35.
> (a setup that is basically working fine, just the runtime checks are missing 
> at several places)

Thanks for answer, I will apply your patch soon.

BRs
Haihao


_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

ffmpeg-codebot · 2022-01-04T02:42:48Z

On the FFmpeg mailing list, Soft Works wrote (reply to this):



> -----Original Message-----
> From: Xiang, Haihao <haihao.xiang@intel.com>
> Sent: Tuesday, January 4, 2022 3:37 AM
> To: ffmpeg-devel@ffmpeg.org; softworkz@hotmail.com
> Subject: Re: [FFmpeg-devel] [PATCH v2] avfilter/vpp_qsv: fix regression on
> older api versions (e.g. 1.11)
> 
> On Tue, 2022-01-04 at 02:25 +0000, Soft Works wrote:
> > > -----Original Message-----
> > > From: Xiang, Haihao <haihao.xiang@intel.com>
> > > Sent: Tuesday, January 4, 2022 3:16 AM
> > > To: ffmpeg-devel@ffmpeg.org
> > > Cc: softworkz@hotmail.com
> > > Subject: Re: [FFmpeg-devel] [PATCH v2] avfilter/vpp_qsv: fix regression
> on
> > > older api versions (e.g. 1.11)
> > >
> > > On Sun, 2022-01-02 at 03:41 +0000, ffmpegagent wrote:
> > > > From: softworkz <softworkz@hotmail.com>
> > > >
> > > > Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > > > regression in a way that scaling via vpp_qsv doesn't work any longer
> > > > for devices with an MSDK runtime version lower than 1.19. This is true
> > > > for older CPUs which are stuck at 1.11.
> > > > The commit added checks for the compile-sdk version but it didn't test
> > > > for the runtime version.
> > > >
> > > > Signed-off-by: softworkz <softworkz@hotmail.com>
> > > > ---
> > > >     avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
> > > >
> > > >     Fix a recently introduced regression when using QSV VPP.
> > > >
> > > >     v2: Fixed commit message wrapping
> > > >
> > > > Published-As:
> > > > https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-
> > >
> > > 15%2Fsoftworkz%2Fqsv_vpp_regression-v2
> > > > Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-
> ffstaging-
> > > > 15/softworkz/qsv_vpp_regression-v2
> > > > Pull-Request: https://github.com/ffstaging/FFmpeg/pull/15
> > > >
> > > > Range-diff vs v1:
> > > >
> > > >  1:  eea966f702 ! 1:  ccede9d840 avfilter/vpp_qsv: fix regression on
> older
> > >
> > > api
> > > > versions (e.g. 1.11)
> > > >      @@ Metadata
> > > >        ## Commit message ##
> > > >           avfilter/vpp_qsv: fix regression on older api versions (e.g.
> > >
> > > 1.11)
> > > >
> > > >      -    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > > > regression in a way
> > > >      -    that scaling via vpp_qsv didn't work any longer for devices
> with
> > >
> > > an
> > > > MSDK runtime
> > > >      -    version lower than 1.19. This is true for older CPUs which
> are
> > >
> > > stuck
> > > > at 1.11.
> > > >      -    The commit added checks for the compile-sdk version but it
> > > > didn't
> > > > test for the
> > > >      -    runtime version.
> > > >      +    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > > >      +    regression in a way that scaling via vpp_qsv doesn't work any
> > >
> > > longer
> > > >      +    for devices with an MSDK runtime version lower than 1.19.
> This
> > > > is
> > > > true
> > > >      +    for older CPUs which are stuck at 1.11.
> > > >      +    The commit added checks for the compile-sdk version but it
> > > > didn't
> > > > test
> > > >      +    for the runtime version.
> > > >
> > > >           Signed-off-by: softworkz <softworkz@hotmail.com>
> > > >
> > > >
> > > >
> > > >  libavfilter/vf_vpp_qsv.c | 32 ++++++++++++++++++++------------
> > > >  1 file changed, 20 insertions(+), 12 deletions(-)
> > > >
> > > > diff --git a/libavfilter/vf_vpp_qsv.c b/libavfilter/vf_vpp_qsv.c
> > > > index ab58a5777e..09590157e3 100644
> > > > --- a/libavfilter/vf_vpp_qsv.c
> > > > +++ b/libavfilter/vf_vpp_qsv.c
> > > > @@ -139,8 +139,9 @@ static const AVOption options[] = {
> > > >      { "height", "Output video height", OFFSET(oh), AV_OPT_TYPE_STRING,
> {
> > > > .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
> > > >      { "format", "Output pixel format", OFFSET(output_format_str),
> > > > AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
> > > >      { "async_depth", "Internal parallelization depth, the higher the
> > > > value
> > > > the higher the latency.", OFFSET(async_depth), AV_OPT_TYPE_INT, { .i64
> = 0
> > >
> > > },
> > > > 0, INT_MAX, .flags = FLAGS },
> > > > +#ifdef QSV_HAVE_SCALING_CONFIG
> > > >      { "scale_mode", "scale mode: 0=auto, 1=low power, 2=high quality",
> > > > OFFSET(scale_mode), AV_OPT_TYPE_INT, { .i64 = MFX_SCALING_MODE_DEFAULT
> },
> > > > MFX_SCALING_MODE_DEFAULT, MFX_SCALING_MODE_QUALITY, .flags = FLAGS,
> "scale
> > > > mode" },
> > > > -
> > > > +#endif
> > > >      { NULL }
> > > >  };
> > > >
> > > > @@ -304,6 +305,7 @@ static int config_output(AVFilterLink *outlink)
> > > >      QSVVPPParam     param = { NULL };
> > > >      QSVVPPCrop      crop  = { 0 };
> > > >      mfxExtBuffer    *ext_buf[ENH_FILTERS_COUNT];
> > > > +    mfxVersion      mfx_version;
> > > >      AVFilterLink    *inlink = ctx->inputs[0];
> > > >      enum AVPixelFormat in_format;
> > > >
> > > > @@ -317,6 +319,11 @@ static int config_output(AVFilterLink *outlink)
> > > >      param.ext_buf       = ext_buf;
> > > >      param.async_depth   = vpp->async_depth;
> > > >
> > > > +    if (MFXQueryVersion(vpp->qsv->session, &mfx_version) < 0) {
> > > > +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
> > > > +        return AVERROR(EINVAL);
> > > > +    }
> > > > +
> > > >      if (inlink->format == AV_PIX_FMT_QSV) {
> > > >           if (!inlink->hw_frames_ctx || !inlink->hw_frames_ctx->data)
> > > >               return AVERROR(EINVAL);
> > > > @@ -467,19 +474,20 @@ static int config_output(AVFilterLink *outlink)
> > > >  #endif
> > > >      }
> > > >
> > > > -    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> > > >  #ifdef QSV_HAVE_SCALING_CONFIG
> > > > -        memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> > > > -        vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> > > > -        vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> > > > -        vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> > > > -
> > > > -        param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> > > > scale_conf;
> > > > -#else
> > > > -        av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> > > > -            "not supported with this MSDK version.\n");
> > > > -#endif
> > > > +    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> > > > +        if (QSV_RUNTIME_VERSION_ATLEAST(mfx_version, 1, 19)) {
> > > > +            memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> > > > +            vpp->scale_conf.Header.BufferId    =
> MFX_EXTBUFF_VPP_SCALING;
> > > > +            vpp->scale_conf.Header.BufferSz    =
> > > > sizeof(mfxExtVPPScaling);
> > > > +            vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> > > > +
> > > > +            param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> > > > > scale_conf;
> > > >
> > > > +        } else
> > > > +            av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> > > > +                "not supported with this MSDK version.\n");
> > > >      }
> > > > +#endif
> > > >
> > > >      if (vpp->use_frc || vpp->use_crop || vpp->deinterlace || vpp-
> >denoise
> > > > >
> > > >
> > > >          vpp->detail || vpp->procamp || vpp->rotate || vpp->hflip ||
> > >
> > > LGTM. I wonder whether you installed libmfx and msdk runtime from
> > > different packages, so the mfx version is higher than the runtime
> version?
> >
> > It's about a user with a Broadwell CPU on Windows. The latest drivers' MSDK
> > version is 1.11 and our ffmpeg is compiled against MSDK 1.35.
> > (a setup that is basically working fine, just the runtime checks are
> missing
> > at several places)
> 
> Thanks for answer, I will apply your patch soon.


Thanks - just wait what the others will say, whether they want to have it in 5.0
or not..

sw
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

ffmpeg-codebot · 2022-01-04T08:11:22Z

On the FFmpeg mailing list, "Xiang, Haihao" wrote (reply to this):

On Sun, 2022-01-02 at 03:41 +0000, ffmpegagent wrote:
> From: softworkz <softworkz@hotmail.com>
> 
> Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> regression in a way that scaling via vpp_qsv doesn't work any longer
> for devices with an MSDK runtime version lower than 1.19. This is true
> for older CPUs which are stuck at 1.11.
> The commit added checks for the compile-sdk version but it didn't test
> for the runtime version.
> 
> Signed-off-by: softworkz <softworkz@hotmail.com>
> ---
>     avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
>     
>     Fix a recently introduced regression when using QSV VPP.
>     
>     v2: Fixed commit message wrapping
> 
> Published-As: 
> https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-15%2Fsoftworkz%2Fqsv_vpp_regression-v2
> Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-
> 15/softworkz/qsv_vpp_regression-v2
> Pull-Request: https://github.com/ffstaging/FFmpeg/pull/15
> 
> Range-diff vs v1:
> 
>  1:  eea966f702 ! 1:  ccede9d840 avfilter/vpp_qsv: fix regression on older api
> versions (e.g. 1.11)
>      @@ Metadata
>        ## Commit message ##
>           avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
>       
>      -    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> regression in a way
>      -    that scaling via vpp_qsv didn't work any longer for devices with an
> MSDK runtime
>      -    version lower than 1.19. This is true for older CPUs which are stuck
> at 1.11.
>      -    The commit added checks for the compile-sdk version but it didn't
> test for the
>      -    runtime version.
>      +    Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
>      +    regression in a way that scaling via vpp_qsv doesn't work any longer
>      +    for devices with an MSDK runtime version lower than 1.19. This is
> true
>      +    for older CPUs which are stuck at 1.11.
>      +    The commit added checks for the compile-sdk version but it didn't
> test
>      +    for the runtime version.
>       
>           Signed-off-by: softworkz <softworkz@hotmail.com>
>       
> 
> 
>  libavfilter/vf_vpp_qsv.c | 32 ++++++++++++++++++++------------
>  1 file changed, 20 insertions(+), 12 deletions(-)
> 
> diff --git a/libavfilter/vf_vpp_qsv.c b/libavfilter/vf_vpp_qsv.c
> index ab58a5777e..09590157e3 100644
> --- a/libavfilter/vf_vpp_qsv.c
> +++ b/libavfilter/vf_vpp_qsv.c
> @@ -139,8 +139,9 @@ static const AVOption options[] = {
>      { "height", "Output video height", OFFSET(oh), AV_OPT_TYPE_STRING, {
> .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
>      { "format", "Output pixel format", OFFSET(output_format_str),
> AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
>      { "async_depth", "Internal parallelization depth, the higher the value
> the higher the latency.", OFFSET(async_depth), AV_OPT_TYPE_INT, { .i64 = 0 },
> 0, INT_MAX, .flags = FLAGS },
> +#ifdef QSV_HAVE_SCALING_CONFIG
>      { "scale_mode", "scale mode: 0=auto, 1=low power, 2=high quality",
> OFFSET(scale_mode), AV_OPT_TYPE_INT, { .i64 = MFX_SCALING_MODE_DEFAULT },
> MFX_SCALING_MODE_DEFAULT, MFX_SCALING_MODE_QUALITY, .flags = FLAGS, "scale
> mode" },
> -
> +#endif
>      { NULL }
>  };
>  
> @@ -304,6 +305,7 @@ static int config_output(AVFilterLink *outlink)
>      QSVVPPParam     param = { NULL };
>      QSVVPPCrop      crop  = { 0 };
>      mfxExtBuffer    *ext_buf[ENH_FILTERS_COUNT];
> +    mfxVersion      mfx_version;
>      AVFilterLink    *inlink = ctx->inputs[0];
>      enum AVPixelFormat in_format;
>  
> @@ -317,6 +319,11 @@ static int config_output(AVFilterLink *outlink)
>      param.ext_buf       = ext_buf;
>      param.async_depth   = vpp->async_depth;
>  
> +    if (MFXQueryVersion(vpp->qsv->session, &mfx_version) < 0) {

It resulted in segmentation fault in testing because vpp->qsv is NULL here. Note
vpp->qsv is set in ff_qsvvpp_create.

Thanks
Haihao


> +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
> +        return AVERROR(EINVAL);
> +    }
> +
>      if (inlink->format == AV_PIX_FMT_QSV) {
>           if (!inlink->hw_frames_ctx || !inlink->hw_frames_ctx->data)
>               return AVERROR(EINVAL);
> @@ -467,19 +474,20 @@ static int config_output(AVFilterLink *outlink)
>  #endif
>      }
>  
> -    if (inlink->w != outlink->w || inlink->h != outlink->h) {
>  #ifdef QSV_HAVE_SCALING_CONFIG
> -        memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> -        vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> -        vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> -        vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> -
> -        param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp->scale_conf;
> -#else
> -        av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> -            "not supported with this MSDK version.\n");
> -#endif
> +    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> +        if (QSV_RUNTIME_VERSION_ATLEAST(mfx_version, 1, 19)) {
> +            memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> +            vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> +            vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> +            vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> +
> +            param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> >scale_conf;
> +        } else
> +            av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> +                "not supported with this MSDK version.\n");
>      }
> +#endif
>  
>      if (vpp->use_frc || vpp->use_crop || vpp->deinterlace || vpp->denoise ||
>          vpp->detail || vpp->procamp || vpp->rotate || vpp->hflip ||
> 
> base-commit: d6b2357eddca392ee137cb2a92ff178a0a7d0cce
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

softworkz · 2022-01-07T03:58:11Z

/submit

ffmpeg-codebot · 2022-01-07T03:58:41Z

Submitted as pull.15.v3.ffstaging.FFmpeg.1641527917915.ffmpegagent@gmail.com

To fetch this version into FETCH_HEAD:

git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-15/softworkz/qsv_vpp_regression-v3

To fetch this version to local tag pr-ffstaging-15/softworkz/qsv_vpp_regression-v3:

git fetch --no-tags https://github.com/ffstaging/FFmpeg tag pr-ffstaging-15/softworkz/qsv_vpp_regression-v3

ffmpeg-codebot · 2022-01-07T06:49:32Z

On the FFmpeg mailing list, "Xiang, Haihao" wrote (reply to this):

On Fri, 2022-01-07 at 03:58 +0000, ffmpegagent wrote:
> From: softworkz <softworkz@hotmail.com>
> 
> Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> regression in a way that scaling via vpp_qsv doesn't work any longer
> for devices with an MSDK runtime version lower than 1.19. This is true
> for older CPUs which are stuck at 1.11.
> The commit added checks for the compile-sdk version but it didn't test
> for the runtime version.
> 
> Signed-off-by: softworkz <softworkz@hotmail.com>
> ---
>     avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
>     
>     Fix a recently introduced regression when using QSV VPP.
>     
>     v2: Fixed commit message wrapping v3: Use different way to acquire
>     runtime version
> 
> Published-As: 
> https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-15%2Fsoftworkz%2Fqsv_vpp_regression-v3
> Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-
> 15/softworkz/qsv_vpp_regression-v3
> Pull-Request: https://github.com/ffstaging/FFmpeg/pull/15
> 
> Range-diff vs v2:
> 
>  1:  ccede9d840 ! 1:  cbf53d8ef4 avfilter/vpp_qsv: fix regression on older api
> versions (e.g. 1.11)
>      @@ Commit message
>           Signed-off-by: softworkz <softworkz@hotmail.com>
>       
>        ## libavfilter/vf_vpp_qsv.c ##
>      +@@
>      + #include "libavutil/opt.h"
>      + #include "libavutil/eval.h"
>      + #include "libavutil/hwcontext.h"
>      ++#include "libavutil/hwcontext_qsv.h"
>      + #include "libavutil/pixdesc.h"
>      + #include "libavutil/mathematics.h"
>      + 
>       @@ libavfilter/vf_vpp_qsv.c: static const AVOption options[] = {
>            { "height", "Output video height", OFFSET(oh), AV_OPT_TYPE_STRING,
> { .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
>            { "format", "Output pixel format", OFFSET(output_format_str),
> AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
>      @@ libavfilter/vf_vpp_qsv.c: static const AVOption options[] = {
>            { NULL }
>        };
>        
>      +@@ libavfilter/vf_vpp_qsv.c: static int config_input(AVFilterLink
> *inlink)
>      +     return 0;
>      + }
>      + 
>      ++static int get_mfx_version(const AVFilterContext *ctx, mfxVersion
> *mfx_version)
>      ++{
>      ++    const AVFilterLink *inlink = ctx->inputs[0];
>      ++    AVBufferRef *device_ref;
>      ++    AVHWDeviceContext *device_ctx;
>      ++    AVQSVDeviceContext *device_hwctx;
>      ++
>      ++    if (inlink->hw_frames_ctx) {
>      ++        AVHWFramesContext *frames_ctx = (AVHWFramesContext *)inlink-
> >hw_frames_ctx->data;
>      ++        device_ref = frames_ctx->device_ref;
>      ++    } else if (ctx->hw_device_ctx) {
>      ++        device_ref = ctx->hw_device_ctx;
>      ++    } else {
>      ++        // Unavailable hw context doesn't matter in pass-through mode,
>      ++        // so don't error here but let runtime version checks fail by
> setting to 0.0
>      ++        mfx_version->Major = 0;
>      ++        mfx_version->Minor = 0;
>      ++        return 0;
>      ++    }
>      ++
>      ++    device_ctx   = (AVHWDeviceContext *)device_ref->data;
>      ++    device_hwctx = device_ctx->hwctx;
>      ++
>      ++    return MFXQueryVersion(device_hwctx->session, mfx_version);

Thanks for the new patch version, it works well for me.

It would be better to change the return type of get_mfx_version from int to
mfxStatus if returning MFXQueryVersion(device_hwctx->session, mfx_version) here.
Otherwise map the returned mfx status to ffmpeg status, then return ffmpeg
status.

BRs
Haihao


>      ++}
>      ++
>      + static int config_output(AVFilterLink *outlink)
>      + {
>      +     AVFilterContext *ctx = outlink->src;
>       @@ libavfilter/vf_vpp_qsv.c: static int config_output(AVFilterLink
> *outlink)
>            QSVVPPParam     param = { NULL };
>            QSVVPPCrop      crop  = { 0 };
>      @@ libavfilter/vf_vpp_qsv.c: static int config_output(AVFilterLink
> *outlink)
>            param.ext_buf       = ext_buf;
>            param.async_depth   = vpp->async_depth;
>        
>      -+    if (MFXQueryVersion(vpp->qsv->session, &mfx_version) < 0) {
>      ++    if (get_mfx_version(ctx, &mfx_version) < 0) {
>       +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
>       +        return AVERROR(EINVAL);
>       +    }
> 
> 
>  libavfilter/vf_vpp_qsv.c | 59 ++++++++++++++++++++++++++++++++--------
>  1 file changed, 47 insertions(+), 12 deletions(-)
> 
> diff --git a/libavfilter/vf_vpp_qsv.c b/libavfilter/vf_vpp_qsv.c
> index ab58a5777e..ca7033cdde 100644
> --- a/libavfilter/vf_vpp_qsv.c
> +++ b/libavfilter/vf_vpp_qsv.c
> @@ -26,6 +26,7 @@
>  #include "libavutil/opt.h"
>  #include "libavutil/eval.h"
>  #include "libavutil/hwcontext.h"
> +#include "libavutil/hwcontext_qsv.h"
>  #include "libavutil/pixdesc.h"
>  #include "libavutil/mathematics.h"
>  
> @@ -139,8 +140,9 @@ static const AVOption options[] = {
>      { "height", "Output video height", OFFSET(oh), AV_OPT_TYPE_STRING, {
> .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
>      { "format", "Output pixel format", OFFSET(output_format_str),
> AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
>      { "async_depth", "Internal parallelization depth, the higher the value
> the higher the latency.", OFFSET(async_depth), AV_OPT_TYPE_INT, { .i64 = 0 },
> 0, INT_MAX, .flags = FLAGS },
> +#ifdef QSV_HAVE_SCALING_CONFIG
>      { "scale_mode", "scale mode: 0=auto, 1=low power, 2=high quality",
> OFFSET(scale_mode), AV_OPT_TYPE_INT, { .i64 = MFX_SCALING_MODE_DEFAULT },
> MFX_SCALING_MODE_DEFAULT, MFX_SCALING_MODE_QUALITY, .flags = FLAGS, "scale
> mode" },
> -
> +#endif
>      { NULL }
>  };
>  
> @@ -297,6 +299,32 @@ static int config_input(AVFilterLink *inlink)
>      return 0;
>  }
>  
> +static int get_mfx_version(const AVFilterContext *ctx, mfxVersion
> *mfx_version)
> +{
> +    const AVFilterLink *inlink = ctx->inputs[0];
> +    AVBufferRef *device_ref;
> +    AVHWDeviceContext *device_ctx;
> +    AVQSVDeviceContext *device_hwctx;
> +
> +    if (inlink->hw_frames_ctx) {
> +        AVHWFramesContext *frames_ctx = (AVHWFramesContext *)inlink-
> >hw_frames_ctx->data;
> +        device_ref = frames_ctx->device_ref;
> +    } else if (ctx->hw_device_ctx) {
> +        device_ref = ctx->hw_device_ctx;
> +    } else {
> +        // Unavailable hw context doesn't matter in pass-through mode,
> +        // so don't error here but let runtime version checks fail by setting
> to 0.0
> +        mfx_version->Major = 0;
> +        mfx_version->Minor = 0;
> +        return 0;
> +    }
> +
> +    device_ctx   = (AVHWDeviceContext *)device_ref->data;
> +    device_hwctx = device_ctx->hwctx;
> +
> +    return MFXQueryVersion(device_hwctx->session, mfx_version);
> +}
> +
>  static int config_output(AVFilterLink *outlink)
>  {
>      AVFilterContext *ctx = outlink->src;
> @@ -304,6 +332,7 @@ static int config_output(AVFilterLink *outlink)
>      QSVVPPParam     param = { NULL };
>      QSVVPPCrop      crop  = { 0 };
>      mfxExtBuffer    *ext_buf[ENH_FILTERS_COUNT];
> +    mfxVersion      mfx_version;
>      AVFilterLink    *inlink = ctx->inputs[0];
>      enum AVPixelFormat in_format;
>  
> @@ -317,6 +346,11 @@ static int config_output(AVFilterLink *outlink)
>      param.ext_buf       = ext_buf;
>      param.async_depth   = vpp->async_depth;
>  
> +    if (get_mfx_version(ctx, &mfx_version) < 0) {
> +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
> +        return AVERROR(EINVAL);
> +    }
> +
>      if (inlink->format == AV_PIX_FMT_QSV) {
>           if (!inlink->hw_frames_ctx || !inlink->hw_frames_ctx->data)
>               return AVERROR(EINVAL);
> @@ -467,19 +501,20 @@ static int config_output(AVFilterLink *outlink)
>  #endif
>      }
>  
> -    if (inlink->w != outlink->w || inlink->h != outlink->h) {
>  #ifdef QSV_HAVE_SCALING_CONFIG
> -        memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> -        vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> -        vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> -        vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> -
> -        param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp->scale_conf;
> -#else
> -        av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> -            "not supported with this MSDK version.\n");
> -#endif
> +    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> +        if (QSV_RUNTIME_VERSION_ATLEAST(mfx_version, 1, 19)) {
> +            memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> +            vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> +            vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> +            vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> +
> +            param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> >scale_conf;
> +        } else
> +            av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> +                "not supported with this MSDK version.\n");
>      }
> +#endif
>  
>      if (vpp->use_frc || vpp->use_crop || vpp->deinterlace || vpp->denoise ||
>          vpp->detail || vpp->procamp || vpp->rotate || vpp->hflip ||
> 
> base-commit: 242ed971cb005157488b9a21942d9fb4be4d0347
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Commit 8b83dad introduced a regression in a way that scaling via vpp_qsv doesn't work any longer for devices with an MSDK runtime version lower than 1.19. This is true for older CPUs which are stuck at 1.11. The commit added checks for the compile-sdk version but it didn't test for the runtime version. Signed-off-by: softworkz <softworkz@hotmail.com>

softworkz · 2022-01-07T06:58:51Z

/preview

ffmpeg-codebot · 2022-01-07T06:59:16Z

Preview email sent as pull.15.v4.ffstaging.FFmpeg.1641538754699.ffmpegagent@gmail.com

softworkz · 2022-01-07T07:01:12Z

/submit

ffmpeg-codebot · 2022-01-07T07:01:33Z

Submitted as pull.15.v4.ffstaging.FFmpeg.1641538891098.ffmpegagent@gmail.com

To fetch this version into FETCH_HEAD:

git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-15/softworkz/qsv_vpp_regression-v4

To fetch this version to local tag pr-ffstaging-15/softworkz/qsv_vpp_regression-v4:

git fetch --no-tags https://github.com/ffstaging/FFmpeg tag pr-ffstaging-15/softworkz/qsv_vpp_regression-v4

ffmpeg-codebot · 2022-01-07T07:03:24Z

On the FFmpeg mailing list, Soft Works wrote (reply to this):



> -----Original Message-----
> From: Xiang, Haihao <haihao.xiang@intel.com>
> Sent: Friday, January 7, 2022 7:48 AM
> To: ffmpeg-devel@ffmpeg.org
> Cc: haihao.xiang-at-intel.com@ffmpeg.org; softworkz@hotmail.com
> Subject: Re: [FFmpeg-devel] [PATCH v3] avfilter/vpp_qsv: fix regression on
> older api versions (e.g. 1.11)
> 
> On Fri, 2022-01-07 at 03:58 +0000, ffmpegagent wrote:
> > From: softworkz <softworkz@hotmail.com>
> >
> > Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > regression in a way that scaling via vpp_qsv doesn't work any longer
> > for devices with an MSDK runtime version lower than 1.19. This is true
> > for older CPUs which are stuck at 1.11.
> > The commit added checks for the compile-sdk version but it didn't test
> > for the runtime version.
> >
> > Signed-off-by: softworkz <softworkz@hotmail.com>
> > ---
> >     avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
> >
> >     Fix a recently introduced regression when using QSV VPP.
> >
> >     v2: Fixed commit message wrapping v3: Use different way to acquire
> >     runtime version
> >
> > Published-As:
> > https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-
> 15%2Fsoftworkz%2Fqsv_vpp_regression-v3
> > Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-
> > 15/softworkz/qsv_vpp_regression-v3
> > Pull-Request: https://github.com/ffstaging/FFmpeg/pull/15
> >
> > Range-diff vs v2:
> >
> >  1:  ccede9d840 ! 1:  cbf53d8ef4 avfilter/vpp_qsv: fix regression on older
> api
> > versions (e.g. 1.11)
> >      @@ Commit message
> >           Signed-off-by: softworkz <softworkz@hotmail.com>
> >
> >        ## libavfilter/vf_vpp_qsv.c ##
> >      +@@
> >      + #include "libavutil/opt.h"
> >      + #include "libavutil/eval.h"
> >      + #include "libavutil/hwcontext.h"
> >      ++#include "libavutil/hwcontext_qsv.h"
> >      + #include "libavutil/pixdesc.h"
> >      + #include "libavutil/mathematics.h"
> >      +
> >       @@ libavfilter/vf_vpp_qsv.c: static const AVOption options[] = {
> >            { "height", "Output video height", OFFSET(oh),
> AV_OPT_TYPE_STRING,
> > { .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
> >            { "format", "Output pixel format", OFFSET(output_format_str),
> > AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
> >      @@ libavfilter/vf_vpp_qsv.c: static const AVOption options[] = {
> >            { NULL }
> >        };
> >
> >      +@@ libavfilter/vf_vpp_qsv.c: static int config_input(AVFilterLink
> > *inlink)
> >      +     return 0;
> >      + }
> >      +
> >      ++static int get_mfx_version(const AVFilterContext *ctx, mfxVersion
> > *mfx_version)
> >      ++{
> >      ++    const AVFilterLink *inlink = ctx->inputs[0];
> >      ++    AVBufferRef *device_ref;
> >      ++    AVHWDeviceContext *device_ctx;
> >      ++    AVQSVDeviceContext *device_hwctx;
> >      ++
> >      ++    if (inlink->hw_frames_ctx) {
> >      ++        AVHWFramesContext *frames_ctx = (AVHWFramesContext *)inlink-
> > >hw_frames_ctx->data;
> >      ++        device_ref = frames_ctx->device_ref;
> >      ++    } else if (ctx->hw_device_ctx) {
> >      ++        device_ref = ctx->hw_device_ctx;
> >      ++    } else {
> >      ++        // Unavailable hw context doesn't matter in pass-through
> mode,
> >      ++        // so don't error here but let runtime version checks fail
> by
> > setting to 0.0
> >      ++        mfx_version->Major = 0;
> >      ++        mfx_version->Minor = 0;
> >      ++        return 0;
> >      ++    }
> >      ++
> >      ++    device_ctx   = (AVHWDeviceContext *)device_ref->data;
> >      ++    device_hwctx = device_ctx->hwctx;
> >      ++
> >      ++    return MFXQueryVersion(device_hwctx->session, mfx_version);
> 
> Thanks for the new patch version, it works well for me.
> 
> It would be better to change the return type of get_mfx_version from int to
> mfxStatus if returning MFXQueryVersion(device_hwctx->session, mfx_version)
> here.
> Otherwise map the returned mfx status to ffmpeg status, then return ffmpeg
> status.

Yes, makes sense. Update on the way..

sw

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

ffmpeg-codebot · 2022-01-10T07:05:09Z

On the FFmpeg mailing list, "Xiang, Haihao" wrote (reply to this):

On Fri, 2022-01-07 at 07:01 +0000, ffmpegagent wrote:
> From: softworkz <softworkz@hotmail.com>
> 
> Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> regression in a way that scaling via vpp_qsv doesn't work any longer
> for devices with an MSDK runtime version lower than 1.19. This is true
> for older CPUs which are stuck at 1.11.
> The commit added checks for the compile-sdk version but it didn't test
> for the runtime version.
> 
> Signed-off-by: softworkz <softworkz@hotmail.com>
> ---
>     avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
>     
>     Fix a recently introduced regression when using QSV VPP.
>     
>     v2: Fixed commit message wrapping
>     
>     v3: Use different way to acquire runtime version
>     
>     v4: Use mfxStatus as return type for get_mfx_version()
> 
> Published-As: 
> https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-15%2Fsoftworkz%2Fqsv_vpp_regression-v4
> Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-
> 15/softworkz/qsv_vpp_regression-v4
> Pull-Request: https://github.com/ffstaging/FFmpeg/pull/15
> 
> Range-diff vs v3:
> 
>  1:  cbf53d8ef4 ! 1:  bf53154420 avfilter/vpp_qsv: fix regression on older api
> versions (e.g. 1.11)
>      @@ libavfilter/vf_vpp_qsv.c: static int config_input(AVFilterLink
> *inlink)
>            return 0;
>        }
>        
>      -+static int get_mfx_version(const AVFilterContext *ctx, mfxVersion
> *mfx_version)
>      ++static mfxStatus get_mfx_version(const AVFilterContext *ctx, mfxVersion
> *mfx_version)
>       +{
>       +    const AVFilterLink *inlink = ctx->inputs[0];
>       +    AVBufferRef *device_ref;
>      @@ libavfilter/vf_vpp_qsv.c: static int config_input(AVFilterLink
> *inlink)
>       +        // so don't error here but let runtime version checks fail by
> setting to 0.0
>       +        mfx_version->Major = 0;
>       +        mfx_version->Minor = 0;
>      -+        return 0;
>      ++        return MFX_ERR_NONE;
>       +    }
>       +
>       +    device_ctx   = (AVHWDeviceContext *)device_ref->data;
>      @@ libavfilter/vf_vpp_qsv.c: static int config_output(AVFilterLink
> *outlink)
>            param.ext_buf       = ext_buf;
>            param.async_depth   = vpp->async_depth;
>        
>      -+    if (get_mfx_version(ctx, &mfx_version) < 0) {
>      ++    if (get_mfx_version(ctx, &mfx_version) != MFX_ERR_NONE) {
>       +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
>       +        return AVERROR(EINVAL);
>       +    }
> 
> 
>  libavfilter/vf_vpp_qsv.c | 59 ++++++++++++++++++++++++++++++++--------
>  1 file changed, 47 insertions(+), 12 deletions(-)
> 
> diff --git a/libavfilter/vf_vpp_qsv.c b/libavfilter/vf_vpp_qsv.c
> index ab58a5777e..5c96703fd3 100644
> --- a/libavfilter/vf_vpp_qsv.c
> +++ b/libavfilter/vf_vpp_qsv.c
> @@ -26,6 +26,7 @@
>  #include "libavutil/opt.h"
>  #include "libavutil/eval.h"
>  #include "libavutil/hwcontext.h"
> +#include "libavutil/hwcontext_qsv.h"
>  #include "libavutil/pixdesc.h"
>  #include "libavutil/mathematics.h"
>  
> @@ -139,8 +140,9 @@ static const AVOption options[] = {
>      { "height", "Output video height", OFFSET(oh), AV_OPT_TYPE_STRING, {
> .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
>      { "format", "Output pixel format", OFFSET(output_format_str),
> AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
>      { "async_depth", "Internal parallelization depth, the higher the value
> the higher the latency.", OFFSET(async_depth), AV_OPT_TYPE_INT, { .i64 = 0 },
> 0, INT_MAX, .flags = FLAGS },
> +#ifdef QSV_HAVE_SCALING_CONFIG
>      { "scale_mode", "scale mode: 0=auto, 1=low power, 2=high quality",
> OFFSET(scale_mode), AV_OPT_TYPE_INT, { .i64 = MFX_SCALING_MODE_DEFAULT },
> MFX_SCALING_MODE_DEFAULT, MFX_SCALING_MODE_QUALITY, .flags = FLAGS, "scale
> mode" },
> -
> +#endif
>      { NULL }
>  };
>  
> @@ -297,6 +299,32 @@ static int config_input(AVFilterLink *inlink)
>      return 0;
>  }
>  
> +static mfxStatus get_mfx_version(const AVFilterContext *ctx, mfxVersion
> *mfx_version)
> +{
> +    const AVFilterLink *inlink = ctx->inputs[0];
> +    AVBufferRef *device_ref;
> +    AVHWDeviceContext *device_ctx;
> +    AVQSVDeviceContext *device_hwctx;
> +
> +    if (inlink->hw_frames_ctx) {
> +        AVHWFramesContext *frames_ctx = (AVHWFramesContext *)inlink-
> >hw_frames_ctx->data;
> +        device_ref = frames_ctx->device_ref;
> +    } else if (ctx->hw_device_ctx) {
> +        device_ref = ctx->hw_device_ctx;
> +    } else {
> +        // Unavailable hw context doesn't matter in pass-through mode,
> +        // so don't error here but let runtime version checks fail by setting
> to 0.0
> +        mfx_version->Major = 0;
> +        mfx_version->Minor = 0;
> +        return MFX_ERR_NONE;
> +    }
> +
> +    device_ctx   = (AVHWDeviceContext *)device_ref->data;
> +    device_hwctx = device_ctx->hwctx;
> +
> +    return MFXQueryVersion(device_hwctx->session, mfx_version);
> +}
> +
>  static int config_output(AVFilterLink *outlink)
>  {
>      AVFilterContext *ctx = outlink->src;
> @@ -304,6 +332,7 @@ static int config_output(AVFilterLink *outlink)
>      QSVVPPParam     param = { NULL };
>      QSVVPPCrop      crop  = { 0 };
>      mfxExtBuffer    *ext_buf[ENH_FILTERS_COUNT];
> +    mfxVersion      mfx_version;
>      AVFilterLink    *inlink = ctx->inputs[0];
>      enum AVPixelFormat in_format;
>  
> @@ -317,6 +346,11 @@ static int config_output(AVFilterLink *outlink)
>      param.ext_buf       = ext_buf;
>      param.async_depth   = vpp->async_depth;
>  
> +    if (get_mfx_version(ctx, &mfx_version) != MFX_ERR_NONE) {
> +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
> +        return AVERROR(EINVAL);
> +    }
> +
>      if (inlink->format == AV_PIX_FMT_QSV) {
>           if (!inlink->hw_frames_ctx || !inlink->hw_frames_ctx->data)
>               return AVERROR(EINVAL);
> @@ -467,19 +501,20 @@ static int config_output(AVFilterLink *outlink)
>  #endif
>      }
>  
> -    if (inlink->w != outlink->w || inlink->h != outlink->h) {
>  #ifdef QSV_HAVE_SCALING_CONFIG
> -        memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> -        vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> -        vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> -        vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> -
> -        param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp->scale_conf;
> -#else
> -        av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> -            "not supported with this MSDK version.\n");
> -#endif
> +    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> +        if (QSV_RUNTIME_VERSION_ATLEAST(mfx_version, 1, 19)) {
> +            memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> +            vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> +            vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> +            vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> +
> +            param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> >scale_conf;
> +        } else
> +            av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> +                "not supported with this MSDK version.\n");
>      }
> +#endif
>  
>      if (vpp->use_frc || vpp->use_crop || vpp->deinterlace || vpp->denoise ||
>          vpp->detail || vpp->procamp || vpp->rotate || vpp->hflip ||
> 
> base-commit: 242ed971cb005157488b9a21942d9fb4be4d0347

LGTM and applied, thx!

-Haihao


_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

ffmpeg-codebot · 2022-01-10T07:19:09Z

On the FFmpeg mailing list, Soft Works wrote (reply to this):



> -----Original Message-----
> From: Xiang, Haihao <haihao.xiang@intel.com>
> Sent: Monday, January 10, 2022 8:03 AM
> To: ffmpeg-devel@ffmpeg.org
> Cc: haihao.xiang-at-intel.com@ffmpeg.org; softworkz@hotmail.com
> Subject: Re: [FFmpeg-devel] [PATCH v4] avfilter/vpp_qsv: fix regression on
> older api versions (e.g. 1.11)
> 
> On Fri, 2022-01-07 at 07:01 +0000, ffmpegagent wrote:
> > From: softworkz <softworkz@hotmail.com>
> >
> > Commit 8b83dad82512a6948b63408f964463b063ad24c9 introduced a
> > regression in a way that scaling via vpp_qsv doesn't work any longer
> > for devices with an MSDK runtime version lower than 1.19. This is true
> > for older CPUs which are stuck at 1.11.
> > The commit added checks for the compile-sdk version but it didn't test
> > for the runtime version.
> >
> > Signed-off-by: softworkz <softworkz@hotmail.com>
> > ---
> >     avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11)
> >
> >     Fix a recently introduced regression when using QSV VPP.
> >
> >     v2: Fixed commit message wrapping
> >
> >     v3: Use different way to acquire runtime version
> >
> >     v4: Use mfxStatus as return type for get_mfx_version()
> >
> > Published-As:
> > https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-
> 15%2Fsoftworkz%2Fqsv_vpp_regression-v4
> > Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-
> > 15/softworkz/qsv_vpp_regression-v4
> > Pull-Request: https://github.com/ffstaging/FFmpeg/pull/15
> >
> > Range-diff vs v3:
> >
> >  1:  cbf53d8ef4 ! 1:  bf53154420 avfilter/vpp_qsv: fix regression on older
> api
> > versions (e.g. 1.11)
> >      @@ libavfilter/vf_vpp_qsv.c: static int config_input(AVFilterLink
> > *inlink)
> >            return 0;
> >        }
> >
> >      -+static int get_mfx_version(const AVFilterContext *ctx, mfxVersion
> > *mfx_version)
> >      ++static mfxStatus get_mfx_version(const AVFilterContext *ctx,
> mfxVersion
> > *mfx_version)
> >       +{
> >       +    const AVFilterLink *inlink = ctx->inputs[0];
> >       +    AVBufferRef *device_ref;
> >      @@ libavfilter/vf_vpp_qsv.c: static int config_input(AVFilterLink
> > *inlink)
> >       +        // so don't error here but let runtime version checks fail
> by
> > setting to 0.0
> >       +        mfx_version->Major = 0;
> >       +        mfx_version->Minor = 0;
> >      -+        return 0;
> >      ++        return MFX_ERR_NONE;
> >       +    }
> >       +
> >       +    device_ctx   = (AVHWDeviceContext *)device_ref->data;
> >      @@ libavfilter/vf_vpp_qsv.c: static int config_output(AVFilterLink
> > *outlink)
> >            param.ext_buf       = ext_buf;
> >            param.async_depth   = vpp->async_depth;
> >
> >      -+    if (get_mfx_version(ctx, &mfx_version) < 0) {
> >      ++    if (get_mfx_version(ctx, &mfx_version) != MFX_ERR_NONE) {
> >       +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
> >       +        return AVERROR(EINVAL);
> >       +    }
> >
> >
> >  libavfilter/vf_vpp_qsv.c | 59 ++++++++++++++++++++++++++++++++--------
> >  1 file changed, 47 insertions(+), 12 deletions(-)
> >
> > diff --git a/libavfilter/vf_vpp_qsv.c b/libavfilter/vf_vpp_qsv.c
> > index ab58a5777e..5c96703fd3 100644
> > --- a/libavfilter/vf_vpp_qsv.c
> > +++ b/libavfilter/vf_vpp_qsv.c
> > @@ -26,6 +26,7 @@
> >  #include "libavutil/opt.h"
> >  #include "libavutil/eval.h"
> >  #include "libavutil/hwcontext.h"
> > +#include "libavutil/hwcontext_qsv.h"
> >  #include "libavutil/pixdesc.h"
> >  #include "libavutil/mathematics.h"
> >
> > @@ -139,8 +140,9 @@ static const AVOption options[] = {
> >      { "height", "Output video height", OFFSET(oh), AV_OPT_TYPE_STRING, {
> > .str="w*ch/cw" }, 0, 255, .flags = FLAGS },
> >      { "format", "Output pixel format", OFFSET(output_format_str),
> > AV_OPT_TYPE_STRING, { .str = "same" }, .flags = FLAGS },
> >      { "async_depth", "Internal parallelization depth, the higher the value
> > the higher the latency.", OFFSET(async_depth), AV_OPT_TYPE_INT, { .i64 = 0
> },
> > 0, INT_MAX, .flags = FLAGS },
> > +#ifdef QSV_HAVE_SCALING_CONFIG
> >      { "scale_mode", "scale mode: 0=auto, 1=low power, 2=high quality",
> > OFFSET(scale_mode), AV_OPT_TYPE_INT, { .i64 = MFX_SCALING_MODE_DEFAULT },
> > MFX_SCALING_MODE_DEFAULT, MFX_SCALING_MODE_QUALITY, .flags = FLAGS, "scale
> > mode" },
> > -
> > +#endif
> >      { NULL }
> >  };
> >
> > @@ -297,6 +299,32 @@ static int config_input(AVFilterLink *inlink)
> >      return 0;
> >  }
> >
> > +static mfxStatus get_mfx_version(const AVFilterContext *ctx, mfxVersion
> > *mfx_version)
> > +{
> > +    const AVFilterLink *inlink = ctx->inputs[0];
> > +    AVBufferRef *device_ref;
> > +    AVHWDeviceContext *device_ctx;
> > +    AVQSVDeviceContext *device_hwctx;
> > +
> > +    if (inlink->hw_frames_ctx) {
> > +        AVHWFramesContext *frames_ctx = (AVHWFramesContext *)inlink-
> > >hw_frames_ctx->data;
> > +        device_ref = frames_ctx->device_ref;
> > +    } else if (ctx->hw_device_ctx) {
> > +        device_ref = ctx->hw_device_ctx;
> > +    } else {
> > +        // Unavailable hw context doesn't matter in pass-through mode,
> > +        // so don't error here but let runtime version checks fail by
> setting
> > to 0.0
> > +        mfx_version->Major = 0;
> > +        mfx_version->Minor = 0;
> > +        return MFX_ERR_NONE;
> > +    }
> > +
> > +    device_ctx   = (AVHWDeviceContext *)device_ref->data;
> > +    device_hwctx = device_ctx->hwctx;
> > +
> > +    return MFXQueryVersion(device_hwctx->session, mfx_version);
> > +}
> > +
> >  static int config_output(AVFilterLink *outlink)
> >  {
> >      AVFilterContext *ctx = outlink->src;
> > @@ -304,6 +332,7 @@ static int config_output(AVFilterLink *outlink)
> >      QSVVPPParam     param = { NULL };
> >      QSVVPPCrop      crop  = { 0 };
> >      mfxExtBuffer    *ext_buf[ENH_FILTERS_COUNT];
> > +    mfxVersion      mfx_version;
> >      AVFilterLink    *inlink = ctx->inputs[0];
> >      enum AVPixelFormat in_format;
> >
> > @@ -317,6 +346,11 @@ static int config_output(AVFilterLink *outlink)
> >      param.ext_buf       = ext_buf;
> >      param.async_depth   = vpp->async_depth;
> >
> > +    if (get_mfx_version(ctx, &mfx_version) != MFX_ERR_NONE) {
> > +        av_log(ctx, AV_LOG_ERROR, "Failed to query mfx version.\n");
> > +        return AVERROR(EINVAL);
> > +    }
> > +
> >      if (inlink->format == AV_PIX_FMT_QSV) {
> >           if (!inlink->hw_frames_ctx || !inlink->hw_frames_ctx->data)
> >               return AVERROR(EINVAL);
> > @@ -467,19 +501,20 @@ static int config_output(AVFilterLink *outlink)
> >  #endif
> >      }
> >
> > -    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> >  #ifdef QSV_HAVE_SCALING_CONFIG
> > -        memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> > -        vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> > -        vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> > -        vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> > -
> > -        param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> >scale_conf;
> > -#else
> > -        av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> > -            "not supported with this MSDK version.\n");
> > -#endif
> > +    if (inlink->w != outlink->w || inlink->h != outlink->h) {
> > +        if (QSV_RUNTIME_VERSION_ATLEAST(mfx_version, 1, 19)) {
> > +            memset(&vpp->scale_conf, 0, sizeof(mfxExtVPPScaling));
> > +            vpp->scale_conf.Header.BufferId    = MFX_EXTBUFF_VPP_SCALING;
> > +            vpp->scale_conf.Header.BufferSz    = sizeof(mfxExtVPPScaling);
> > +            vpp->scale_conf.ScalingMode        = vpp->scale_mode;
> > +
> > +            param.ext_buf[param.num_ext_buf++] = (mfxExtBuffer*)&vpp-
> > >scale_conf;
> > +        } else
> > +            av_log(ctx, AV_LOG_WARNING, "The QSV VPP Scale option is "
> > +                "not supported with this MSDK version.\n");
> >      }
> > +#endif
> >
> >      if (vpp->use_frc || vpp->use_crop || vpp->deinterlace || vpp->denoise
> ||
> >          vpp->detail || vpp->procamp || vpp->rotate || vpp->hflip ||
> >
> > base-commit: 242ed971cb005157488b9a21942d9fb4be4d0347
> 
> LGTM and applied, thx!
> 
> -Haihao
> 

Thanks
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

softworkz · 2022-05-25T00:16:05Z

Merged upstream

softworkz force-pushed the qsv_vpp_regression branch from eea966f to ccede9d Compare January 2, 2022 03:39

softworkz force-pushed the qsv_vpp_regression branch from ccede9d to cbf53d8 Compare January 7, 2022 03:56

softworkz force-pushed the qsv_vpp_regression branch from cbf53d8 to bf53154 Compare January 7, 2022 06:57

softworkz closed this May 25, 2022

softworkz reopened this May 27, 2022

softworkz closed this May 27, 2022

softworkz deleted the qsv_vpp_regression branch October 11, 2022 12:55

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11) #15

avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11) #15

softworkz commented Jan 2, 2022 •

edited

Loading

softworkz commented Jan 2, 2022

ffmpeg-codebot bot commented Jan 2, 2022

softworkz commented Jan 2, 2022

ffmpeg-codebot bot commented Jan 2, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

softworkz commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

softworkz commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

softworkz commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 10, 2022

ffmpeg-codebot bot commented Jan 10, 2022

softworkz commented May 25, 2022

avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11) #15

avfilter/vpp_qsv: fix regression on older api versions (e.g. 1.11) #15

Conversation

softworkz commented Jan 2, 2022 • edited Loading

softworkz commented Jan 2, 2022

ffmpeg-codebot bot commented Jan 2, 2022

softworkz commented Jan 2, 2022

ffmpeg-codebot bot commented Jan 2, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

ffmpeg-codebot bot commented Jan 4, 2022

softworkz commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

softworkz commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

softworkz commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 7, 2022

ffmpeg-codebot bot commented Jan 10, 2022

ffmpeg-codebot bot commented Jan 10, 2022

softworkz commented May 25, 2022

softworkz commented Jan 2, 2022 •

edited

Loading