[Webkit-unassigned] [Bug 98823] http://www.robohornet.org/tests/addcol.html spends 15% of time in HTMLTableRowsCollection::rowAfter

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=98823





--- Comment #1 from Eric Seidel <eric at webkit.org>  2012-10-09 16:58:22 PST ---
I believe why Ryosuke was suggesting is making: invalidateNodeListCachesInAncestors smarter, possibly to pass an (optional) Node* along to Document::invalidateNodeListCaches, and I assume then to check that node pointer in HTMLTableRowsCollection or perhaps some baseclass which is aware of the fact that HTMLTableRowsCollection only cares about changes to its direct children (and in this case, only to addition or removal of HTMLTableRow elements.

I would guess the behavior would be that if Node* == 0, then we always invalidate everything, and if a Node* is given, we only invalidate if the Node* parentNode is the root?  Presumably we may have to pass the root pointer along too?

I suspect rniwa has the design better sorted out in his head.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.