[webkit-reviews] review canceled: [Bug 86787] Update CachedResourceLoader lifetime to allow for main resources : [Attachment 142588] RefCounted<CachedResourceLoader>

[bugzilla-daemon at webkit.org]

Nate Chapin <japhet at chromium.org> has canceled Nate Chapin
<japhet at chromium.org>'s request for review:
Bug 86787: Update CachedResourceLoader lifetime to allow for main resources
https://bugs.webkit.org/show_bug.cgi?id=86787

Attachment 142588: RefCounted<CachedResourceLoader>
https://bugs.webkit.org/attachment.cgi?id=142588&action=review

------- Additional Comments from Nate Chapin <japhet at chromium.org>
I wasn't clearing CachedResourceLoader::m_document at the end of the Document's
life, which could theoretically lead to a stale pointer (I wasn't able to
convince myself that the Document will always outlive the DocumentLoader,
correct me if there's something I don't know). 

Turns out severing the Document<->CachedResourceLoader connection is more
complicated than I thought, because there are actually cases where multiple
Documents will end up with the same CachedResourceLoader.  I think I've made
this code sufficiently defensive (and commented to death), but I may be missing
something.