Chop quads exactly on the clip bounds, so we don't spend CPU cycles walking them
when we're above or below the clip.

Still to do:
- chop in X to avoid 16.16. overflow in the edgelist
- apply the same logic for cubics (tho much harder math)



git-svn-id: http://skia.googlecode.com/svn/trunk@88 2bbb7eff-a529-9590-31e7-b0007b416f81
9 files changed