[Webkit-unassigned] [Bug 38424] add support for text/html-sandboxed on sandboxed iframes

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=38424


Patrik Persson <patrik.j.persson at ericsson.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #64832|0                           |1
        is obsolete|                            |
  Attachment #65708|                            |review?, commit-queue?
               Flag|                            |




--- Comment #21 from Patrik Persson <patrik.j.persson at ericsson.com>  2010-08-27 06:37:49 PST ---
Created an attachment (id=65708)
 --> (https://bugs.webkit.org/attachment.cgi?id=65708)
Patch with more tests

(In reply to comment #17)

Thanks for the fast reply. We didn't mean to intrude on your night sleep (judging by the time of your message)...

> Can you add a test for loading a text/html-sandboxed document inside a plain iframe in a sandboxed iframe?

Added (LayoutTests/http/tests/security/sandboxed-iframe-inherited-html-sandboxed.html)

> Do we still need to call closeRemainingWindowsWhenDone to clean up properly?  (I might have gotten the name wrong.)

Added call to setCloseRemainingWindowsWhenComplete() in html-sandboxed-mainwindow.html.

> nbr => generally we don't like these sorts of abbreviations.

Changed to numberOfTrustedWindows.

> LayoutTests/http/tests/security/html-sandboxed-mainwindow.html:31
>  +      for (var i = 0; i < 5; i++) {
> Why do we open so many windows?  That seems like of silly.

I added a comment in the test case to explain our reasoning. Essentially, we're trying to determine whether an asynchronous, unbounded event (rendering a loaded document in a new window) was prevented. So, how long do we wait? We came up with this (admittedly silly) solution of a set of very similar, but allowed, windows. If all those allowed windows were loaded, and the disallowed one wasn't, we should be fine.

Yes, silly indeed. We'd love to figure out a more direct way of testing this.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.