Fix instances of clang -Wbitfield-enum-conversion in Blink

third_party/WebKit/Source/core/svg/properties/SVGAnimatedProperty.cpp

Index: third_party/WebKit/Source/core/svg/properties/SVGAnimatedProperty.cpp
diff --git a/third_party/WebKit/Source/core/svg/properties/SVGAnimatedProperty.cpp b/third_party/WebKit/Source/core/svg/properties/SVGAnimatedProperty.cpp
index a6a9a698d8d03dcd90e0e31e0df5405243469220..c188b1e89431fff8632fb56f32980e15e99fb41d 100644
--- a/third_party/WebKit/Source/core/svg/properties/SVGAnimatedProperty.cpp
+++ b/third_party/WebKit/Source/core/svg/properties/SVGAnimatedProperty.cpp
@@ -40,7 +40,9 @@ SVGAnimatedPropertyBase::SVGAnimatedPropertyBase(
     const QualifiedName& attributeName,
     CSSPropertyID cssPropertyId)
     : m_type(type),
-      m_cssPropertyId(cssPropertyId),
+      // Explicitly cast cssPropertyId to silence warnings about bitfield
+      // truncation.

[Nico, 2017/03/14 18:50:00]

Maybe also say why this is safe?

[Reid Kleckner, 2017/03/14 19:37:25]

The DCHECK_EQ is already checking that this is safe.

I'm also not really a fan of these sorts of comments. I'd rather let someone who
wants to know why there's a cast try to remove it and find out what the compiler
says. If someone makes this not a bitfield, for example, this comment will
become stale, and the cast superfluous.

[Nico, 2017/03/14 19:52:24]

Which dcheck? The `DCHECK_EQ(this->cssPropertyId(), cssPropertyId);`?

Imho `// Safe because values that don't fit in bitfield are aliases which aren't
passed to this function, see static_assert in header` or something like that
makes it obvious that the author has thought about this. When i blame code, I
often wonder if something's intentional the way it is or it just happens to be
that way, and "just happens to be that way" is true fairly often. So I find
comments like this helpful.

+      m_cssPropertyId(static_cast<unsigned>(cssPropertyId)),
       m_contextElement(contextElement),
       m_attributeName(attributeName) {
   DCHECK(m_contextElement);