On Mon, Feb 10, 2025 at 05:34:38PM +0100, Andreas Rheinhardt wrote:
> Michael Niedermayer:
> > Fixes: double free
> > Fixes: 393523547/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_AAC_LATM_fuzzer-6740617236905984
> > 
> > Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
> > Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
> > ---
> >  libavcodec/aac/aacdec.c      | 19 +++++++++++++++++--
> >  libavcodec/aac/aacdec_usac.c |  3 ++-
> >  2 files changed, 19 insertions(+), 3 deletions(-)
> > 
> > diff --git a/libavcodec/aac/aacdec.c b/libavcodec/aac/aacdec.c
> > index 8d50ad6d095..16259b5ada9 100644
> > --- a/libavcodec/aac/aacdec.c
> > +++ b/libavcodec/aac/aacdec.c
> > @@ -421,6 +421,21 @@ static uint64_t sniff_channel_order(uint8_t (*layout_map)[3], int tags)
> >      return layout;
> >  }
> >  
> > +static void copy_oc(OutputConfiguration *dst, OutputConfiguration *src)
> > +{
> > +    int err = 0;
> 
> Seems unused.

i saw it after sending the patch


> 
> > +
> > +    for(int i = 0; i < dst->usac.nb_elems; i++)
> > +        av_freep(&dst->usac.elems[i].ext.pl_data);
> > +
> > +    *dst = *src;
> > +
> > +    for(int i = 0; i < dst->usac.nb_elems; i++) {
> > +        AACUsacElemConfig *e = &dst->usac.elems[i];
> > +        e->ext.pl_data = av_memdup(e->ext.pl_data, e->ext.pl_data_offset);
> 
> Unchecked allocation. Furthermore, the *dst = *src makes cleanup on
> error here a PITA.

> Would making pl_data reference-counted (via
> RefStruct) work instead?

likely, yes
do you want to implement that ?

thx

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

In a rich man's house there is no place to spit but his face.
-- Diogenes of Sinope