Removing intermediate IntPoint objects when FloatPoint is needed

Sometimes (probably for historical reasons), IntPoint objects
are created where FloatPoint objects are needed. The IntPoint will
(still) be converted to a FloatPoint by the compiler so all that
happens is that the code takes an unnecessary detour.

Need to be careful so that we don't depend on some truncation
performed by the detour via integers.

Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_slimming_paint_v2;master.tryserver.blink:linux_trusty_blink_rel
Change-Id: Ia3b90f2a80cb7a65cfdf78c1370bea0acb79491d
Reviewed-on: https://chromium-review.googlesource.com/1126397
Reviewed-by: Philip Rogers <pdr@chromium.org>
Commit-Queue: Daniel Bratell <bratell@opera.com>
Cr-Commit-Position: refs/heads/master@{#572961}
10 files changed