On Tue, Dec 1, 2015 at 7:39 PM, Lou Logan <l...@lrcd.com> wrote: > On Tue, 1 Dec 2015 19:27:49 -0500, Ganesh Ajjanagadde wrote: > >> The slowness comes from the generate code, e.g llrint compiles down to >> a single asm instruction cvttsd2si, while floor(x + 0.5) needs to do >> multiple things. > > How much slower is it?
I don't know and honestly don't care myself as it is not really the selling point of the patch since AFAIK these are not performance critical. The point is really moot - I merely claimed that it is slower, which is obvious from the asm. The more serious reasons for avoiding it are also given in the message. > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-devel _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
