[FFmpeg-devel] [PATCH 2/2] lavc/vaapi_encode_h264: correct bit_rate_scale setting.

[Mark Thompson]

On 31/10/17 02:37, Jun Zhao wrote:
> From d1e105057e93e7c2788d6d684292db9008fbf3ac Mon Sep 17 00:00:00 2001
> From: Jun Zhao <jun.zhao at intel.com>
> Date: Tue, 31 Oct 2017 10:19:08 +0800
> Subject: [PATCH 2/2] lavc/vaapi_encode_h264: correct bit_rate_scale setting.
> 
> As H264 Spec 2012 E.2.2, bit_rate_scale means the max input bit rate.
> 
> Signed-off-by: Jun Zhao <jun.zhao at intel.com>
> Signed-off-by: Wang, Yi A <yi.a.wang at intel.com>
> ---
>  libavcodec/vaapi_encode_h264.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/libavcodec/vaapi_encode_h264.c b/libavcodec/vaapi_encode_h264.c
> index 1d43e934ef..27a810c64e 100644
> --- a/libavcodec/vaapi_encode_h264.c
> +++ b/libavcodec/vaapi_encode_h264.c
> @@ -406,7 +406,7 @@ static int vaapi_encode_h264_init_sequence_params(AVCodecContext *avctx)
>          // Try to scale these to a sensible range so that the
>          // golomb encode of the value is not overlong.
>          hrd->bit_rate_scale =
> -            av_clip_uintp2(av_log2(avctx->bit_rate) - 15 - 6, 4);
> +            av_clip_uintp2(av_log2(FFMAX(avctx->bit_rate, avctx->rc_max_rate)) - 15 - 6, 4);
>          hrd->bit_rate_value_minus1[0] =
>              (avctx->bit_rate >> hrd->bit_rate_scale + 6) - 1;
>  
> -- 
> 2.14.1
> 

Can you offer some more justification for that?  I think the change is wrong: consider under what circumstances overflow/underflow occurs.

- Mark