Prevent mis-nested fullscreen in iframes

Prevent mis-nested fullscreen in iframes

http://fullscreen.spec.whatwg.org/#fullscreen-element-ready-check
https://www.w3.org/Bugs/Public/show_bug.cgi?id=26568

BUG=403741
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=180460

[philipj_slow, 2014-08-17 21:42:18]

test

[philipj_slow, 2014-08-17 22:14:45]

PTAL. The test would fail without this change.

elementReady() is recursive and will crash for sufficiently deeply nested
iframes. Unfortunately a test case for it would use too much memory and be too
slow, here's an attempt that I had to leave running over lunch before it crashed
at least several thousand iframes deep, without this CL applied:
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/3120

Given that we can apparently already crash for very deep iframe nesting I'm not
sure what the right thing to do is. Avoiding the recursion is simple but makes
the code slightly uglier IMHO.

[philipj_slow, 2014-08-17 22:20:49]

It looks like for node trees, there are functions like Node::baseURI that are
recursive and crash for deeply nested trees, not sure what to make of that.

[falken, 2014-08-18 03:09:43]

lgtm

> It looks like for node trees, there are functions like Node::baseURI that are
> recursive and crash for deeply nested trees, not sure what to make of that.

Right, I think typically Blink code doesn't try to avoid deep recursion. I guess
it's OK to just let the browser terminate the renderer when that happens.

[philipj_slow, 2014-08-18 12:53:46]

rebase

[philipj_slow, 2014-08-18 12:54:00]

set_commit

[philipj_slow, 2014-08-18 12:54:05]

The CQ bit was checked by philipj@opera.com

[commit-bot: I haz the power, 2014-08-18 12:54:41]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/philipj@opera.com/485443002/60001

[commit-bot: I haz the power, 2014-08-18 13:52:07]

Committed patchset #4 (60001) as 180460