Changed ComputedStyle's EVisibility to be an enum class and fixed naming

third_party/WebKit/Source/core/style/ComputedStyle.h

Index: third_party/WebKit/Source/core/style/ComputedStyle.h
diff --git a/third_party/WebKit/Source/core/style/ComputedStyle.h b/third_party/WebKit/Source/core/style/ComputedStyle.h
index fee328e23171e67405ef60f6aeeb23f708d4012b..a66a3f692703a6fc7afc027125741241d2e39597 100644
--- a/third_party/WebKit/Source/core/style/ComputedStyle.h
+++ b/third_party/WebKit/Source/core/style/ComputedStyle.h
@@ -298,7 +298,7 @@ protected:
         m_inheritedData.m_captionSide = initialCaptionSide();
         m_inheritedData.m_listStyleType = initialListStyleType();
         m_inheritedData.m_listStylePosition = initialListStylePosition();
-        m_inheritedData.m_visibility = initialVisibility();
+        m_inheritedData.m_visibility = static_cast<unsigned>(initialVisibility());

[Bugs Nash, 2016/07/28 01:25:12]

why not make m_inheritedData.m_visibility store as a EVisibility instead? Seems
like the code would be cleaner and more type safe

[esprehn, 2016/07/29 01:53:09]

I'm not sure if type safe bit field enums work in all our compilers (MSVC was
bad years ago), you should do some static_assert's about the size of
ComputedStyle's various objects and then try messing with this stuff to see if
MSVC can deal with enum class bitfields. It used to realign the types which
wasted lots of space.

[sashab, 2016/07/29 06:25:50]

Yup; thought I replied to this, thanks Elliott. I'm happy to do that as a
follow-up. For now, the design doc contains a bunch of research I've done to
make sure these static casts don't cost us any cycles (they don't).

The packing example from MSDN seems to align unsigned int bitfields to 8-bit
boundaries:
https://msdn.microsoft.com/en-us/library/ewwyfdbe.aspx

Also, from https://msdn.microsoft.com/en-us/library/1d48zaa8.aspx, "Adjacent bit
fields are packed into the same 1-, 2-, or 4-byte allocation unit if the
integral types are the same size and if the next bit field fits into the current
allocation unit without crossing the boundary imposed by the common alignment
requirements of the bit fields.".

But then, on https://msdn.microsoft.com/en-us/library/9f53hsws.aspx, it suggests
that it doesn't do that alignment for unsigned.

This will need more investigation and I'm happy to do that in parallel, but for
now we can safely leave things as they are.

         m_inheritedData.m_textAlign = initialTextAlign();
         m_inheritedData.m_textTransform = initialTextTransform();
         m_inheritedData.m_textUnderline = false;
@@ -1468,9 +1468,9 @@ public:
     void setTextTransform(ETextTransform v) { m_inheritedData.m_textTransform = v; }
 
     // visibility
-    static EVisibility initialVisibility() { return VISIBLE; }
+    static EVisibility initialVisibility() { return EVisibility::Visible; }
     EVisibility visibility() const { return static_cast<EVisibility>(m_inheritedData.m_visibility); }
-    void setVisibility(EVisibility v) { m_inheritedData.m_visibility = v; }
+    void setVisibility(EVisibility v) { m_inheritedData.m_visibility = static_cast<unsigned>(v); }
 
     // white-space inherited
     static EWhiteSpace initialWhiteSpace() { return NORMAL; }