[webkit-reviews] review denied: [Bug 88888] RenderLayer subtrees without any self-painting layer shouldn't be walked during painting : [Attachment 147117] Proposed change v1.

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Jun 12 11:40:13 PDT 2012


Simon Fraser (smfr) <simon.fraser at apple.com> has denied Julien Chaffraix
<jchaffraix at webkit.org>'s request for review:
Bug 88888: RenderLayer subtrees without any self-painting layer shouldn't be
walked during painting
https://bugs.webkit.org/show_bug.cgi?id=88888

Attachment 147117: Proposed change v1.
https://bugs.webkit.org/attachment.cgi?id=147117&action=review

------- Additional Comments from Simon Fraser (smfr) <simon.fraser at apple.com>
View in context: https://bugs.webkit.org/attachment.cgi?id=147117&action=review


> Source/WebCore/rendering/RenderLayer.cpp:446
> +void RenderLayer::setHasSelfPaintingLayerDescendant()

This method name is confusing. It should be
markAncestorsHaveSelfPaintingDescendant() or something.

> Source/WebCore/rendering/RenderLayer.h:878
> +    // If have no self-painting descendants, we don't have to walk our
children during painting.
> +    // This can lead to significant savings, especially if the tree has lots
of non-self-painting layers grouped together
> +    // (e.g. table cells). The current logic is simple and conservative
which will lead to over-marking subtrees (FIXME).

It would be nice to toggle this back to false when possible. Isn't there an
existing tree walk you can piggy-back on to update this flag?


More information about the webkit-reviews mailing list