Eliminated instructions by using better neon instructions and rearranging the loop. On a Nexus 7, vpxenc (in realtime mode, speed -12) reported a performance improvement of ~1.0%. Change-Id: I6b1700e79318f647ea67ef25e954c308932950ec
On a Nexus 7, vpxenc (in realtime mode, speed -12) reported a performance improvement of ~12.4% Change-Id: Id29d215acf58bb108489e218a259adf74b4768d7