[Webkit-unassigned] [Bug 49468] [Qt] Assertion failure in DocumentLoader::commitData when loading a media document

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu May 5 15:54:30 PDT 2011


https://bugs.webkit.org/show_bug.cgi?id=49468


Joe Wild <joseph.wild at nokia.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #91036|0                           |1
        is obsolete|                            |
  Attachment #92484|                            |review?, commit-queue?
               Flag|                            |




--- Comment #5 from Joe Wild <joseph.wild at nokia.com>  2011-05-05 15:54:30 PST ---
Created an attachment (id=92484)
 --> (https://bugs.webkit.org/attachment.cgi?id=92484&action=review)
Moved canceling to FrameLoaderClientQt::committedLoad

I had some doubts about the test for this, so submitted as a separate
patch to review.

Cancel loads in FrameLoaderClientQt::committedLoad that will
be handled by the MediaDocument handler. If they are allowed
to continue they will assert/reset.

Canceling here can cause the history stack to end up in the wrong state,
so there is some code here to clean that up.

Also, I included an test case for this in a separate patch.  This test
should pass on all platforms, but is only really tested with the
right on Qt and when the right plugins are loaded.  Without these
the test won't get to the place where the error occurred.

It was starting to get messy to get this test to always work so I just
added a simple test.  Some of the things that stop this from working.
  * run-webkit-tests doesn't find the right plugins since it defines
    its own QTWEBKIT_PLUGIN_PATH value.  I had to copy the plugin to
    my local test directory.
  * Only Qt DumpRenderTree supports layoutTestController.handleErrorPages()
    needed to execute the code for this test.
  * And finally even under Qt this case is not handled since it only
    supports this error handling mechanism for the main frame.  I tried
    not using a "meta refresh" instead of the iframe for the test, but
    took the test to a generated page for the plugin where I had no control.

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



More information about the webkit-unassigned mailing list