[FFmpeg-devel] [PATCH] libavutil/hwcontext_qsv: fix a bug for mapping qsv frame to vaapi

[James Almer]

On 9/9/2021 6:13 AM, Wenbin Chen wrote:
> Command below failed.
> ffmpeg -v verbose -init_hw_device vaapi=va:/dev/dri/renderD128
> -init_hw_device qsv=qs at va -hwaccel qsv -hwaccel_device qs
> -filter_hw_device va -c:v h264_qsv
> -i 1080P.264 -vf "hwmap,format=vaapi" -c:v h264_vaapi output.264
> 
> Cause: Assign pair->first directly to data[3] in vaapi frame.
> pair->first is *VASurfaceID while data[3] in vaapi frame is
> VASurfaceID. I fix this line of code. Now the command above works.
> 
> Signed-off-by: Wenbin Chen <wenbin.chen at intel.com>
> ---
>   libavutil/hwcontext_qsv.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/libavutil/hwcontext_qsv.c b/libavutil/hwcontext_qsv.c
> index d431e71eab..6539cae619 100644
> --- a/libavutil/hwcontext_qsv.c
> +++ b/libavutil/hwcontext_qsv.c
> @@ -781,7 +781,7 @@ static int qsv_map_from(AVHWFramesContext *ctx,
>       case AV_HWDEVICE_TYPE_VAAPI:
>       {
>           mfxHDLPair *pair = (mfxHDLPair*)surf->Data.MemId;
> -        child_data = pair->first;
> +        child_data = (uint8_t*)(intptr_t)*(VASurfaceID*)pair->first;

You can probably remove the intptr_t casting.

Also, shouldn't this same fix be done for all three child device types 
used in this function? Which for that matter, child_data seems to be set 
for a d3d11va child device, but then never used.

>           break;
>       }
>   #endif
>