[FFmpeg-devel] [PATCH 6/8] avcodec/vvc/ctu: Check palette_escape_val

Andreas Rheinhardt andreas.rheinhardt at outlook.com
Sun Jun 22 09:41:16 EEST 2025


Nuo Mi:
> On Fri, Jun 20, 2025 at 8:40 AM Michael Niedermayer <michael at niedermayer.cc>
> wrote:
> 
>> Fixes: integer overflow
>> Fixes:
>> 418314174/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_VVC_fuzzer-4871731867353088
>>
>> Found-by: continuous fuzzing process
>> https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
>> Signed-off-by
>> <https://github.com/google/oss-fuzz/tree/master/projects/ffmpegSigned-off-by>:
>> Michael Niedermayer <michael at niedermayer.cc>
>> ---
>>  libavcodec/vvc/ctu.c | 2 ++
>>  1 file changed, 2 insertions(+)
>>
>> diff --git a/libavcodec/vvc/ctu.c b/libavcodec/vvc/ctu.c
>> index ba4c89b1d1b..7fa2b496389 100644
>> --- a/libavcodec/vvc/ctu.c
>> +++ b/libavcodec/vvc/ctu.c
>> @@ -2053,6 +2053,8 @@ static int palette_subblock_data(VVCLocalContext *lc,
>>                  const int v = PALETTE_INDEX(xc, yc);
>>                  if (v == esc) {
>>                      const int coeff = ff_vvc_palette_escape_val(lc);
>>
> A check for < 0 is needed.

The check below is performed as unsigned comparison. If coeff were < 0,
then (unsigned)coeff is > INT_MAX and the check below triggers.

> 
>> +                    if (coeff >= (1U << sps->bit_depth))
>> +                        return AVERROR_INVALIDDATA;
> 
>                      const int pixel = av_clip_intp2(RSHIFT(coeff * scale,
>> 6), sps->bit_depth);
>>                      PALETTE_SET_PIXEL(xc, yc, pixel);
>>                  } else {
>> --
>> 2.49.0


More information about the ffmpeg-devel mailing list