Fix handling of negative radius in HTMLAreaElement's coords when in circle state: - https://html.spec.whatwg.org/#attr-area-shape-circle The radius must be non-negative as per the specification. However, our code fails to check. Firefox and Chrome both reject negative negative radius.
Created attachment 288142 [details] Patch
Comment on attachment 288142 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=288142&action=review > Source/WebCore/html/HTMLAreaElement.cpp:163 > + if (radius > 0) This should be >=
Comment on attachment 288142 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=288142&action=review > Source/WebCore/ChangeLog:15 > + Firefox and Chrome both reject negative negative radius. "negative negative" => "negative"
Comment on attachment 288142 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=288142&action=review >> Source/WebCore/html/HTMLAreaElement.cpp:163 >> + if (radius > 0) > > This should be >= Why? If radius is 0 then there is no circle and it would never hit test an element anyway. FYI, I have also checked that Blink is using > 0.
Comment on attachment 288142 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=288142&action=review >> Source/WebCore/html/HTMLAreaElement.cpp:163 >> + if (radius > 0) > > This should be >= Using '>' is OK as-is.. I mean, it does not make sense to describe an ellipse with a 0 radius.
Created attachment 288175 [details] Patch
Comment on attachment 288175 [details] Patch Clearing flags on attachment: 288175 Committed r205565: <http://trac.webkit.org/changeset/205565>
All reviewed patches have been landed. Closing bug.