It can be costly.
Created attachment 392254 [details] wip
Created attachment 392258 [details] wip
Created attachment 392270 [details] wip
Created attachment 392295 [details] wip
Created attachment 392312 [details] patch
Created attachment 392313 [details] patch
Comment on attachment 392313 [details] patch I've got a bit of a compat/regression concern here, but let's see what happens...also YAY.
Comment on attachment 392313 [details] patch View in context: https://bugs.webkit.org/attachment.cgi?id=392313&action=review > Source/WebCore/ChangeLog:14 > + This appears to be 3-4% Speedometer progression. Nice!
Created attachment 392385 [details] patch
Created attachment 392393 [details] patch
Created attachment 392399 [details] patch
Comment on attachment 392399 [details] patch Clearing flags on attachment: 392399 Committed r257839: <https://trac.webkit.org/changeset/257839>
All reviewed patches have been landed. Closing bug.
<rdar://problem/60035288>
Comment on attachment 392399 [details] patch View in context: https://bugs.webkit.org/attachment.cgi?id=392399&action=review > Source/WebCore/dom/Element.cpp:3352 > + for (auto& element : ancestorsOfType<Element>(*this)) { This should probably be the flat / composed tree ancestors, right? I suspect this would crash otherwise in some shadow DOM situations involving slotting.
Good point!
Created attachment 392406 [details] followup
Reopening for a followup
Created attachment 392410 [details] patch
Created attachment 392411 [details] followup
Comment on attachment 392411 [details] followup Clearing flags on attachment: 392411 Committed r257846: <https://trac.webkit.org/changeset/257846>