Add base types for path ops

src/pathops/SkPathOpsPoint.cpp

Index: src/pathops/SkPathOpsPoint.cpp
===================================================================
--- src/pathops/SkPathOpsPoint.cpp	(revision 0)
+++ src/pathops/SkPathOpsPoint.cpp	(revision 0)
@@ -0,0 +1,17 @@
+/*
+ * Copyright 2012 Google Inc.
+ *
+ * Use of this source code is governed by a BSD-style license that can be
+ * found in the LICENSE file.
+ */
+#include "SkPathOpsPoint.h"
+
+SkDVector operator-(const SkDPoint& a, const SkDPoint& b) {

[whunt, 2013/03/22 18:16:06]

This function should be inlined.  If the rational for not inlining it is code
size you'll find that pushing the operators and making the function call are
actually more code bytes than the subtracts.

[caryclark, 2013/03/22 19:38:51]

Again, I think measurement is the key. The compiler (or, more precisely, the
linker) can inline this code by itself.

+    SkDVector v = {a.fX - b.fX, a.fY - b.fY};

[whunt, 2013/03/22 18:16:06]

you'll bet better code with a constructor

[caryclark, 2013/03/22 19:38:51]

As Mike has pointed out, Skia intentionally does not use constructors for
structs.

+    return v;
+}
+
+SkDPoint operator+(const SkDPoint& a, const SkDVector& b) {

[whunt, 2013/03/22 18:16:06]

This function should be inlined.  If the rational for not inlining it is code
size you'll find that pushing the operators and making the function call are
actually more code bytes than the adds.

[caryclark, 2013/03/22 20:05:00]

Noted.

+    SkDPoint v = {a.fX + b.fX, a.fY + b.fY};

[whunt, 2013/03/22 18:16:06]

you'll bet better code with a constructor

[caryclark, 2013/03/22 20:05:00]

Noted.

+    return v;
+}