[webkit-reviews] review canceled: [Bug 109591] RenderLayer::hasOutOfFlowPositionedDescendant should not be affected by invisible children : [Attachment 188002] Patch

[bugzilla-daemon at webkit.org]

vollick at chromium.org has canceled vollick at chromium.org's request for review:
Bug 109591: RenderLayer::hasOutOfFlowPositionedDescendant should not be
affected by invisible children
https://bugs.webkit.org/show_bug.cgi?id=109591

Attachment 188002: Patch
https://bugs.webkit.org/attachment.cgi?id=188002&action=review

------- Additional Comments from vollick at chromium.org
Need to ditch this approach. Glenn and I discussed the following approach which
seems like the right way forward.
 - get rid of m_hasOutOfFlowEtcEtc
 - instead, on the fly in updateNeedsCompositedEtc, we'll
   - check if we have a descendant
   - if so, check if its containing block is us
   - if so, we form a containing block and don't need to worry about
out-of-flow positioned descendants
   - if not, go through the list of positioned descendants maintained in
RenderBlock and if any of them have containing block that is an ancestor of us,
we can't opt in.