Chromium: Update resource tracking when moving a frame between documents

[Jenn Braithwaite]

Chromium fix for bug reported in 44713.  That patch provides the framework for the fix and fixes the Mac platform.  Separate patches will be submitted for other platforms, thus this separate bug.

[Jenn Braithwaite]

Created attachment 72121 [details]
patch

[Dmitry Titov]

Comment on attachment 72121 [details]
patch

View in context: https://bugs.webkit.org/attachment.cgi?id=72121&action=review

> WebKit/chromium/src/FrameLoaderClientImpl.cpp:1394
> +        oldWebFrame->client()->removeIdentifierForRequest(oldWebFrame, identifier);

It looks like we are passing here wrong webFrame... Do we really need this parameter?

[Jenn Braithwaite]

Created attachment 72261 [details]
updated patch to remove unnecessary frame parameter

[Adam Barth]

Needs fishd for WebKit API review.

[Jenn Braithwaite]

Created attachment 72526 [details]
updated patch - use map.contains()

[Darin Fisher (:fishd, Google)]

Comment on attachment 72526 [details]
updated patch - use map.contains()

API change looks good to me.  Is it possible to write a layout test for this?

[Jenn Braithwaite]

(In reply to comment #6)
> (From update of attachment 72526 [details])
> API change looks good to me.  Is it possible to write a layout test for this?

The layout test is in the patch submitted with bug 44713.  Both that test (fast/frames/iframe-reparenting-fail-load.html) and fast/frames/iframe-reparenting-adopt-node.html cover this change.

[Dmitry Titov]

Comment on attachment 72526 [details]
updated patch - use map.contains()

r=me

[WebKit Commit Bot]

Comment on attachment 72526 [details]
updated patch - use map.contains()

Clearing flags on attachment: 72526

Committed r71297: <http://trac.webkit.org/changeset/71297>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.