I've rechecked documentation for the "vpx_codec_control" - can't find
any guarantee about the last parameter when the function fails, thus
prefer to be on the safe side.

On Tue, Apr 5, 2022 at 3:59 PM James Almer <jamrial@gmail.com> wrote:
>
> On 4/5/2022 9:15 AM, Danil Chapovalov wrote:
> > ---
> >   libavcodec/libvpxenc.c | 9 ++++++++-
> >   1 file changed, 8 insertions(+), 1 deletion(-)
> >
> > diff --git a/libavcodec/libvpxenc.c b/libavcodec/libvpxenc.c
> > index dff1d06b0e..0705863450 100644
> > --- a/libavcodec/libvpxenc.c
> > +++ b/libavcodec/libvpxenc.c
> > @@ -1250,6 +1250,8 @@ static int storeframe(AVCodecContext *avctx, struct FrameListData *cx_frame,
> >       int ret = ff_get_encode_buffer(avctx, pkt, cx_frame->sz, 0);
> >       uint8_t *side_data;
> >       int pict_type;
> > +    int quality;
> > +    VPxContext *ctx = avctx->priv_data;
> >
> >       if (ret < 0)
> >           return ret;
> > @@ -1264,7 +1266,12 @@ static int storeframe(AVCodecContext *avctx, struct FrameListData *cx_frame,
> >           pict_type = AV_PICTURE_TYPE_P;
> >       }
> >
> > -    ff_side_data_set_encoder_stats(pkt, 0, cx_frame->sse + 1,
> > +    ret = vpx_codec_control(&ctx->encoder, VP8E_GET_LAST_QUANTIZER_64, &quality);
> > +    if (ret != VPX_CODEC_OK) {
>
> Is the last argument guaranteed to be untouched when the return value of
> vpx_codec_control() is not VPX_CODEC_OK? If so, you can initialize
> quality to 0 above, and remove this part.
>
> > +        quality = 0;
> > +    }
> > +
> > +    ff_side_data_set_encoder_stats(pkt, quality * FF_QP2LAMBDA, cx_frame->sse + 1,
> >                                      cx_frame->have_sse ? 3 : 0, pict_type);
> >
> >       if (cx_frame->have_sse) {
> _______________________________________________
> 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".