[webkit-reviews] review granted: [Bug 3439] mouseover effects can get stuck sometimes due to missing events : [Attachment 7179] proposed fix

[bugzilla-request-daemon at opendarwin.org]

Darin Adler <darin at apple.com> has granted Darin Adler <darin at apple.com>'s
request for review:
Bug 3439: mouseover effects can get stuck sometimes due to missing events
http://bugzilla.opendarwin.org/show_bug.cgi?id=3439

Attachment 7179: proposed fix
http://bugzilla.opendarwin.org/attachment.cgi?id=7179&action=edit

------- Additional Comments from Darin Adler <darin at apple.com>
It's not great to pass the subframe to the passSubframeEventToSubframe
function, but on the other hand, it's my own fault for insisting the subframe
function be removed from MouseEventWithHitTestResults. Eventually we should be
able to clean that up a bit -- I'm not sure it needs to go back through the
Frame and through Macintosh-specific code.

+    RenderObject *renderer = mev.innerNode()->renderer();

New code should put the * next to the class name, not the variable.

+    RefPtr<FrameView> protector(this);

No explanation of why this was added.

I don't understand the positioning of the call to setCursor in the new code.
Wouldn't we need a new call to setCursor after passing the event to the new
subframe?

Ideally I think that oldUnder should be per-page, not per-frame. If we did
that, I think we might be able to simply get rid of oldSubframe.

Despite those doubts and comments, looks fine to land.

r=me