[webkit-reviews] review granted: [Bug 104517] [CSSRegions] Assertion failure in Node::detach (!renderer || renderer->inRenderFlowThread()) : [Attachment 186616] incorporated Julien Chaffraix's feedback (with updated ChangeLogs)

[bugzilla-daemon at webkit.org]

Julien Chaffraix <jchaffraix at webkit.org> has granted Mihai Maerean
<mmaerean at adobe.com>'s request for review:
Bug 104517: [CSSRegions] Assertion failure in Node::detach (!renderer ||
renderer->inRenderFlowThread())
https://bugs.webkit.org/show_bug.cgi?id=104517

Attachment 186616: incorporated Julien Chaffraix's feedback (with updated
ChangeLogs)
https://bugs.webkit.org/attachment.cgi?id=186616&action=review

------- Additional Comments from Julien Chaffraix <jchaffraix at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=186616&action=review


r=me

> LayoutTests/ChangeLog:16
> +	   The RenderObject::inRenderFlowThread bit could have become
disconnected from the fact that the RenderObject
> +	   has (or not) an enclosing RenderFlowThread.
> +	   The cause of this was that, when setting or removing the parent of a
RenderObject, the inRenderFlowThread flags
> +	   wasn't being set/reset for the children too.
> +	   This is now fixed by calling the new
setInRenderFlowThreadIncludingDescendants.
> +
> +	   The ASSERT was hit for anonymous blocks when detaching the document.

> +
> +	   The test starts from such an anonymous block and the detaches the
body of document.

Nit: We usually don't copy the WebCore ChangeLog here. (Most of the time I keep
the LayoutTests ChangeLog empty unless there is something important to
mention).