Re: [FFmpeg-devel] [PATCH 1/6] avcodec/osq: Check that pkt_offset does not exceed pkt size

From: Paul B Mahol <onemda@gmail.com>
To: FFmpeg development discussions and patches <ffmpeg-devel@ffmpeg.org>
Subject: Re: [FFmpeg-devel] [PATCH 1/6] avcodec/osq: Check that pkt_offset does not exceed pkt size
Date: Fri, 22 Sep 2023 19:36:08 +0200
Message-ID: <CAPYw7P7w3Kze22LyS3tsb0Lu__kohXMy5G-TDD+cu1gHXt5YSg@mail.gmail.com> (raw)
In-Reply-To: <20230922164748.GY8640@pb2>

On 9/22/23, Michael Niedermayer <michael@niedermayer.cc> wrote:
> On Thu, Sep 21, 2023 at 08:14:31PM +0200, Paul B Mahol wrote:
>> On Thu, Sep 21, 2023 at 8:09 PM Michael Niedermayer
>> <michael@niedermayer.cc>
>> wrote:
>>
>> > Fixes: out of array access
>> > Fixes:
>> > 62164/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_OSQ_fuzzer-6227491892887552
>> > Fixes:
>> > 62164/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_OSQ_fuzzer-6268561729126400
>> > Fixes:
>> > 62164/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_OSQ_fuzzer-6414805046788096
>> > Fixes:
>> > 62164/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_OSQ_fuzzer-6538151088488448
>> > Fixes:
>> > 62164/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_OSQ_fuzzer-6608131540779008
>> >
>> > 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@niedermayer.cc>
>> > ---
>> >  libavcodec/osq.c | 3 +++
>> >  1 file changed, 3 insertions(+)
>> >
>> > diff --git a/libavcodec/osq.c b/libavcodec/osq.c
>> > index e7f11691d2e..bcc75fef6fc 100644
>> > --- a/libavcodec/osq.c
>> > +++ b/libavcodec/osq.c
>> > @@ -408,6 +408,9 @@ static int osq_receive_frame(AVCodecContext *avctx,
>> > AVFrame *frame)
>> >      GetBitContext *gb = &s->gb;
>> >      int ret, n;
>> >
>> > +    if (s->pkt_offset > s->pkt->size)
>> > +        s->pkt_offset = 0;
>> >
>>
>> This is more hack than real fix.
>
> why ?
>
> pkt->size is reset outside the codec, so either it needs to be
> checked on codec entry or the codec should not use
> internal->in_pkt and expect its size to be conserved
> or implement flush() or something
>
> ff_decode_flush_buffers() for example will clear teh packet
>
> if you prefer i can implement flush() and reset pkt_offset in it
> that probably would achieve teh same
> just say if you prefer that ?

Yup, that is much cleaner, go ahead with that solution with flush().
I forgot about flush() case completely.

>
> thx
>
> [...]
> --
> Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
>
> When the tyrant has disposed of foreign enemies by conquest or treaty, and
> there is nothing more to fear from them, then he is always stirring up
> some war or other, in order that the people may require a leader. -- Plato
>
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".