Implicit coversion operators from integer geometry types to floating point.

ui/gfx/size_unittest.cc

Index: ui/gfx/size_unittest.cc
diff --git a/ui/gfx/size_unittest.cc b/ui/gfx/size_unittest.cc
new file mode 100644
index 0000000000000000000000000000000000000000..65d58cafde380b84ac81e31cc3ae8d77097619f5
--- /dev/null
+++ b/ui/gfx/size_unittest.cc
@@ -0,0 +1,31 @@
+// Copyright (c) 2012 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#include "ui/gfx/size_base.h"
+
+#include "testing/gtest/include/gtest/gtest.h"
+#include "ui/gfx/size.h"
+#include "ui/gfx/size_f.h"
+
+namespace ui {
+
+static int test_sizef(const gfx::SizeF& s) {
+  return s.width();
+}
+
+TEST(SizeTest, ToSizeF) {
+  // Check that implicit conversion from integer to float compiles.
+  gfx::Size a(10, 20);
+  float width = test_sizef(a);
+  EXPECT_EQ(width, a.width());
+
+  gfx::SizeF b(10, 20);
+  bool equals = a == b;
+  EXPECT_EQ(true, equals);
+
+  equals = b == a;
+  EXPECT_EQ(true, equals);
+}
+
+}  // namespace ui