[FFmpeg-devel] [PATCH 2/3] vorbisdec: replace get_bits with get_bitsz where n can be 0

[Andreas Cadhalpun]

On 03.01.2016 22:50, Michael Niedermayer wrote:
> On Sun, Jan 03, 2016 at 07:50:39PM +0100, Andreas Cadhalpun wrote:
>>  vorbisdec.c |    5 +++++
>>  1 file changed, 5 insertions(+)
>> ba151dadb72b6c74e1139decf4b32c8676ddc58e  0001-vorbisdec-reject-rangebits-0.patch
>> From d740a59b6e099c90504d55c65923def1ad6de234 Mon Sep 17 00:00:00 2001
>> From: Andreas Cadhalpun <Andreas.Cadhalpun at googlemail.com>
>> Date: Sun, 3 Jan 2016 19:11:24 +0100
>> Subject: [PATCH 1/2] vorbisdec: reject rangebits 0
>>
>> This causes non-unique elements in floor_setup->data.t1.list, which
>> makes the stream undecodable according to the specification.
>>
>> Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun at googlemail.com>
>> ---
>>  libavcodec/vorbisdec.c | 5 +++++
>>  1 file changed, 5 insertions(+)
>>
>> diff --git a/libavcodec/vorbisdec.c b/libavcodec/vorbisdec.c
>> index f773afa..2792af1 100644
>> --- a/libavcodec/vorbisdec.c
>> +++ b/libavcodec/vorbisdec.c
>> @@ -573,6 +573,11 @@ static int vorbis_parse_setup_hdr_floors(vorbis_context *vc)
>>                  return AVERROR(ENOMEM);
>>  
>>              rangebits = get_bits(gb, 4);
>> +            if (!rangebits) {
>> +                av_log(vc->avctx, AV_LOG_ERROR,
>> +                       "A rangebits value of 0 is not compliant with the Vorbis I specification.\n");
>> +                return AVERROR_INVALIDDATA;
>> +            }
> 
> this assumes partitions > 0
> iam not sure if this is required or not, the spec does not seem to
> explicitly state it

OK, so let's only error out if partitions is > 0.
Updated patch attached.

Best regards,
Andreas

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-vorbisdec-reject-rangebits-0-with-non-0-partitions.patch
Type: text/x-diff
Size: 1256 bytes
Desc: not available
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20160103/29572068/attachment.patch>