Am 29.11.2011 05:24, schrieb Ronald S. Bultje:
> Hi,
>
> On Mon, Nov 28, 2011 at 12:58 PM, Aneesh Dogra <[email protected]> wrote:
>> Fixes: http://bugzilla.libav.org/show_bug.cgi?id=102
>> ---
>>  libavcodec/indeo3.c |    2 ++
>>  1 files changed, 2 insertions(+), 0 deletions(-)
>>
>> diff --git a/libavcodec/indeo3.c b/libavcodec/indeo3.c
>> index 1d8f694..596cd27 100644
>> --- a/libavcodec/indeo3.c
>> +++ b/libavcodec/indeo3.c
>> @@ -804,6 +804,8 @@ static int decode_plane(Indeo3DecodeContext *ctx, 
>> AVCodecContext *avctx,
>>     num_vectors = bytestream_get_le32(&data);
>>     ctx->mc_vectors  = num_vectors ? data : 0;
>>
>> +    if (num_vectors * 2 >= data_size)
>> +        return AVERROR_INVALIDDATA;
>>     /* init the bitreader */
>>     init_get_bits(&ctx->gb, &data[num_vectors * 2], data_size << 3);
>>     ctx->skip_bits   = 0;
> This leads to the question whether data_size << 3 is really the proper
> size for this buffer. In reality, it's (data_size - num_vectors * 2)
> << 3.
>
> Fixing this will allow doing things like get_bits_left() further down
> in the stack to prevent more subtle buffer overreads.

Yes. That's absolutely correct. A proper fix would be very appreciated!

Best regards
Maxim
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to