-subpixel accurate for scaling down by passing in x and dx.
-blend for bilinear use single multiply instead of 2 multiplies.
-filter col do 2 pixels at a time
-assembly do address munging
-assembly avoid wait on store for core2
BUG=none
TEST=none
Review URL: https://webrtc-codereview.appspot.com/397012

git-svn-id: http://libyuv.googlecode.com/svn/trunk@187 16f28f9a-4ce2-e073-06de-1de4eb20be90