On Sat, 28 Jan 2017 18:35:01 +0100 Nicolas George <geo...@nsup.org> wrote:
> Le nonidi 9 pluviôse, an CCXXV, Muhammad Faiz a écrit : > > So, if ff_inlink_make_frame_writable() takes AVFrame**, it will be > > incompatible with framequeue framework, because it stores AVFrame*, so > > it will contain dangling AVFrame*. > > There is nothing "incompatible" about it, what would it even mean? It > just can not be used as is on this specific call site; it was not meant > to. It is a smaller, more versatile brick. You just need to add the > little bit of mortar to have it fit here. > > It is not possible to "make" a frame writable, the original name is a > misnomer; the only thing possible is to create a new copy that is Well, that's what the existing API calls it. In several places. > uniquely referenced and therefore writable. If that is needed, an extra > copy can be made to have the new frame take the place of the old one. > But that is only rarely needed; making that copy always like you propose It doesn't always copy. > following wm4's bad advice Please cease your childish attacks. > will just make the code less efficient and > more awkward. Amazing. We're talking about copying a few bytes additional in a case that copies an entire frame of media data. And it's only "rarely needed", as you said just above. (To remind other readers, ff_inlink_make_frame_writable() exits immediately if the frame is writable.) _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
