Quoting Janne Grunau (2016-06-03 19:20:39) > On 2016-05-16 10:26:00 +0200, Anton Khirnov wrote: > > This function does not do any bitstream parsing and it depends on the > > current frame being allocated, so this will allow the frame_start() to > > be moved out eventually. > > It hardly depends on the frame being allocated. Setting the poc directly > in the frame with a pointer as argument is itself *censored*. I'm > undecided if moving ff_h264_init_poc() out is the right thing. Having > the poc as soon as possible might help debugging otoh this type of frame > level debugability is probably not priority for the h264 decoder > anymore. > > So I don't care much either way but the dependence on cur_pic being > allocated is a weak argument for moving this call.
Well, you need to store the frame POC somewhere, preferably in the frame itself. And for that you need to have the frame. In any case, the idea is that this function will only do bitstream parsing, all the other calculations should be somewhere else. -- Anton Khirnov _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
