[webkit-reviews] review granted: [Bug 30304] struct RGBA32Buffer is twice as big as the old ImageData of Qt : [Attachment 41472] Reduce RGBA32Buffer from 52bytes to 20 bytes per instance (take two)

[bugzilla-daemon at webkit.org]

Eric Seidel <eric at webkit.org> has granted Holger Freyther <zecke at selfish.org>'s
request for review:
Bug 30304: struct RGBA32Buffer is twice as big as the old ImageData of Qt
https://bugs.webkit.org/show_bug.cgi?id=30304

Attachment 41472: Reduce RGBA32Buffer from 52bytes to 20 bytes per instance
(take two)
https://bugs.webkit.org/attachment.cgi?id=41472&action=review

------- Additional Comments from Eric Seidel <eric at webkit.org>
I think  170	     uint8_t m_disposalMethod;
can just be a FrameDisposalMethod
the need to use unsigned, etc, comes from bitfields, or from the fact that some
compilers default enums to being signed.  But here it doesn't seem to be an
issue.	BUt if you're rathe rleave it as uint for consistence, I guess that's
fine too.

Not to jump back into the previous discussion, but I'm confused why this isn't
const:
 401	 IntRect& rect = m_rects[frameIndex];

rect.contains() is a const method, no?

I think you meant 8 for both of these:
36 COMPILE_ASSERT(RGBA32Buffer::FrameStatusLast < 256, FrameStatusFitsIn8Byte);

 37 COMPILE_ASSERT(RGBA32Buffer::DisposeLast < 256, DisposeLastFitsIn9Byte);

This seems fine though.