Quoting Andreas Cadhalpun (2015-03-05 22:45:37) > On 05.03.2015 16:04, Anton Khirnov wrote: > > Quoting Andreas Cadhalpun (2015-02-26 01:27:34) > >> However it might still be a good idea to apply above patch, because the > >> decoder/parser could set avctx->channels to 0, even if the demuxer has set > >> it to > >> something positive. > > > > Yes, that's a pervasive problem I'm going to address in a rather huge > > set I'm working on. In the meantime, I think it'd be best to just store > > the value in the private data to make sure no one messes with it. > > You mean as an additional field in the ADXDemuxerContext? > > I'm not sure that would be a good thing, because then it might be possible > for the demuxer and the decoder to disagree about the number of channels. >
That would mean there's a bug in one of them. The current API allows a caller to mess with the stream codec context, so in theory the caller can write anything in there. A demuxer should not depend on this being sane. -- Anton Khirnov _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
