Re: [FFmpeg-devel] [PATCH v3 2/2] avcodec/mediacodecdec: check ff_Build_SDK_INT return value

["Tomas Härdin" <git@haerdin.se>]

tis 2023-01-10 klockan 19:35 +0800 skrev Zhao Zhili:
> From: Zhao Zhili <zhilizhao@tencent.com>
> 
> ---
> v3: when target API <= 24 ==> when target API < 24
> v2: add comments
> 
>  libavcodec/mediacodecdec.c | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/libavcodec/mediacodecdec.c b/libavcodec/mediacodecdec.c
> index 11f655a9aa..21464900d1 100644
> --- a/libavcodec/mediacodecdec.c
> +++ b/libavcodec/mediacodecdec.c
> @@ -415,7 +415,13 @@ static av_cold int
> mediacodec_decode_init(AVCodecContext *avctx)
>             s->ctx->codec_name, ret);
>  
>      sdk_int = ff_Build_SDK_INT(avctx);
> -    if (sdk_int <= 23 &&
> +    /* ff_Build_SDK_INT can fail when target API < 24 and JVM isn't
> available.
> +     * If we don't check sdk_int > 0, the workaround might be
> enabled by
> +     * mistake.
> +     * JVM is required to make the workaround works reliably. On the
> other hand,
> +     * missing a workaround should not be a serious issue, we do as
> best we can.
> +     */
> +    if (sdk_int > 0 && sdk_int <= 23 &&
>          strcmp(s->ctx->codec_name,
> "OMX.amlogic.mpeg2.decoder.awesome") == 0) {
>          av_log(avctx, AV_LOG_INFO, "Enabling workaround for %s on
> API=%d\n",
>                 s->ctx->codec_name, sdk_int);

Looks OK

/Tomas

_______________________________________________
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".