[webkit-reviews] review granted: [Bug 71244] CSS 2.1 failure: border-conflict-element-* : [Attachment 115980] Patch

[bugzilla-daemon at webkit.org]

Julien Chaffraix <jchaffraix at webkit.org> has granted Robert Hogan
<robert at webkit.org>'s request for review:
Bug 71244: CSS 2.1 failure: border-conflict-element-*
https://bugs.webkit.org/show_bug.cgi?id=71244

Attachment 115980: Patch
https://bugs.webkit.org/attachment.cgi?id=115980&action=review

------- Additional Comments from Julien Chaffraix <jchaffraix at webkit.org>
> If I had correctly implemented the second patch and converted
CollapsedBorderValue::rowOffset() to return an unsigned int,
border-conflict-element-021.htm would have started failing. My understanding is
that casting from signed->unsigned->signed ensured that in the end the
underflowed value was *not* used when deciding where in the table the
rowgroup's border was.

My bad, I missed the double signed <-> unsigned cast which would cancel any
visible change in the resolution.


> So border-conflict-element-021.htm tests for the potential regression you
spotted and it didn't fail on patch '2' because the logic that checks the
position of the border's cell was using a signed value rather than an unsigned.


Thanks for double checking, appreciated!