[FFmpeg-devel] [PATCH] avcodec/hq_hqa: Check info size

[Hendrik Leppkes]

On Sat, May 30, 2020 at 2:01 AM Michael Niedermayer
<michael at niedermayer.cc> wrote:
>
> Fixes: assertion failure
> Fixes: 21079/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_HQ_HQA_fuzzer-5737046523248640
>
> Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
> Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
> ---
>  libavcodec/hq_hqa.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/libavcodec/hq_hqa.c b/libavcodec/hq_hqa.c
> index eec2e980b3..8404e80ec8 100644
> --- a/libavcodec/hq_hqa.c
> +++ b/libavcodec/hq_hqa.c
> @@ -321,7 +321,7 @@ static int hq_hqa_decode_frame(AVCodecContext *avctx, void *data,
>          int info_size;
>          bytestream2_skip(&ctx->gbc, 4);
>          info_size = bytestream2_get_le32(&ctx->gbc);
> -        if (bytestream2_get_bytes_left(&ctx->gbc) < info_size) {
> +        if (info_size < 0 || bytestream2_get_bytes_left(&ctx->gbc) < info_size) {
>              av_log(avctx, AV_LOG_ERROR, "Invalid INFO size (%d).\n", info_size);
>              return AVERROR_INVALIDDATA;
>          }
> --

bytestream2_get_le32 returns an unsigned type, wouldn't it be better
to make info_size unsigned and avoid the type cast and signed overflow
that results in this check failing?

- Hendrik