[FFmpeg-devel] [PATCH 1/4] avformat/iamf_parse: Check extradata size

[James Almer]

On 6/23/2025 9:57 AM, Andreas Rheinhardt wrote:
> James Almer:
>> On 6/23/2025 9:44 AM, Michael Niedermayer wrote:
>>> On Fri, Jun 20, 2025 at 12:28:13AM +0200, Andreas Rheinhardt wrote:
>>>> Michael Niedermayer:
>>>>> Fixes: Assertion n>=0 && n<=32 failed at ./libavcodec/get_bits.h:406
>>>>> Fixes: 398527871/clusterfuzz-testcase-minimized-
>>>>> ffmpeg_dem_IAMF_fuzzer-6602025714647040
>>>>>
>>>>> Found-by: continuous fuzzing process https://github.com/google/oss-
>>>>> fuzz/tree/master/projects/ffmpeg
>>>>> Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
>>>>> ---
>>>>>    libavformat/iamf_parse.c | 2 ++
>>>>>    1 file changed, 2 insertions(+)
>>>>>
>>>>> diff --git a/libavformat/iamf_parse.c b/libavformat/iamf_parse.c
>>>>> index 71497876ac3..330e01733dd 100644
>>>>> --- a/libavformat/iamf_parse.c
>>>>> +++ b/libavformat/iamf_parse.c
>>>>> @@ -305,6 +305,8 @@ static int update_extradata(AVCodecParameters
>>>>> *codecpar)
>>>>>            skip_bits(&gb, 4);
>>>>>            put_bits(&pb, 4, codecpar->ch_layout.nb_channels); // set
>>>>> channel config
>>>>>            ret = put_bits_left(&pb);
>>>>> +        if (ret < 0)
>>>>> +            return AVERROR_INVALIDDATA;
>>>>>            while (ret >= 32) {
>>>>>               put_bits32(&pb, get_bits_long(&gb, 32));
>>>>>               ret -= 32;
>>>>
>>>> There is only one way for put_bits_left() to return a negative value: If
>>>> there is more data in the internal buffer than can be written out. And
>>>> this scenario is already a violation of the PutBit API. Given that the
>>>> size of the internal buffer depends upon the arch, it could be that one
>>>> would have already hit an assert in case one is not using x64. In other
>>>> words, your check is too late.
>>>
>>> the patches puprose was mainly to show that
>>> 3f9420132441345b7ccd57001f230bb98f655696
>>> was insufficient to fix 398527871
>>>
>>> I do not expect my patch would be the correct solution even if the
>>> check is done earlier. IAMF is cursed
>>
>> Does increasing buf from 6 bytes to 8 or more fix it? I see putbits may
>> do an AV_W*64(), so six bytes sounds like it was never safe.
>>
> That is only executed when the internal bit buffer is full; you will
> never reach it on x64. The problem is that you initialize the put bits
> buffer with FFMIN(codecpar->extradata_size, sizeof(buf)) instead of
> sizeof(buf). If this were not so, there would always be bits left.
> But this only fixes the API violations, it does not guarantee that the
> written data is actually correct. What is actually in the data that gets
> written in the loop?
The remaining bits in the GetBitContext buffer in order to fill the 
PutBitContext buffer.

The following probably fixes it, based on what you said.

> diff --git a/libavformat/iamf_parse.c b/libavformat/iamf_parse.c
> index 71497876ac..19c549d4ac 100644
> --- a/libavformat/iamf_parse.c
> +++ b/libavformat/iamf_parse.c
> @@ -288,7 +288,7 @@ static int update_extradata(AVCodecParameters *codecpar)
>          uint8_t buf[6];
>          int size = FFMIN(codecpar->extradata_size, sizeof(buf));
> 
> -        init_put_bits(&pb, buf, size);
> +        init_put_bits(&pb, buf, sizeof(buf));
>          ret = init_get_bits8(&gb, codecpar->extradata, size);
>          if (ret < 0)
>              return ret;
> @@ -312,6 +312,9 @@ static int update_extradata(AVCodecParameters *codecpar)
>          put_bits(&pb, ret, get_bits_long(&gb, ret));
>          flush_put_bits(&pb);
> 
> +        if (get_bits_left(&gb) < 0)
> +           return AVERROR_INVALIDDATA;
> +
>          memcpy(codecpar->extradata, buf, put_bytes_output(&pb));
>          break;
>      }


-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature.asc
Type: application/pgp-signature
Size: 495 bytes
Desc: OpenPGP digital signature
URL: <https://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20250623/29ded4ee/attachment.sig>