On 02.01.2016 23:36, Luca Barbato wrote: > On 02/01/16 23:26, Andreas Cadhalpun wrote: >> On 02.01.2016 23:03, Luca Barbato wrote: >>> On 02/01/16 20:01, Andreas Cadhalpun wrote: >>>> It's documented to only support reading 1-25 bits. >>>> >>>> Signed-off-by: Andreas Cadhalpun <andreas.cadhal...@googlemail.com> >>>> --- >>>> libavcodec/vorbisdec.c | 4 ++-- >>>> 1 file changed, 2 insertions(+), 2 deletions(-) >>>> >>> >>> When do we get in this situation? >> >> I'm not sure what you mean. > > Is it something exceptional (it is a corner case that might > be rejected before we get to that place) or is just something overlooked?
I think it's just something overlooked. >> The first one happens if bits is 0, e.g. because vc->audio_channels is 1. >> The second one happens if rangebits is 0. > > Looks like we are getting at least some of places in which bits 0 is valid > maybe a get_bits0 would simplify at least the looks of some of those cases ^^ Maybe, but on the other hand there are only about a dozen or so places in the complete code base that would benefit from that. Do you think that's enough to justify adding another helper function? Best regards, Andreas _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
