[webkit-reviews] review denied: [Bug 62116] REGRESSION(r87980): Null pointer dereference in RenderTextControl::setInnerText() : [Attachment 96072] Patch

[bugzilla-daemon at webkit.org]

Dimitri Glazkov (Google) <dglazkov at chromium.org> has denied Kent Tamura
<tkent at chromium.org>'s request for review:
Bug 62116: REGRESSION(r87980): Null pointer dereference in
RenderTextControl::setInnerText()
https://bugs.webkit.org/show_bug.cgi?id=62116

Attachment 96072: Patch
https://bugs.webkit.org/attachment.cgi?id=96072&action=review

------- Additional Comments from Dimitri Glazkov (Google)
<dglazkov at chromium.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=96072&action=review

I think this approach is wrong. We should not be special-casing id handling
inside shadow DOM. Or maybe I just don't understand why we would want to do
that?

> Source/WebCore/dom/Element.cpp:961
> +    if (hasID() && m_attributeMap && treeScope() == document()) {

Why do we have to have a special way of handling shadow DOM nodes here? Other
than working around the bug you're fixing? I don't think this is right.