548258 - ASSERTION: Serial number requested for unrecognized pointer! Are you memmoving a refcounted object? when cycle collecting nsGenericElement after loading tinderbox

Status: RESOLVED INCOMPLETE

(Core :: DOM: Core & HTML, defect, P5)

Description

[Henri Sivonen (:hsivonen) (temporarily away from Bugzilla)]

(Redescribed from bug 514807).

Steps to reproduce:
 1) Enable leak logging for nsGenericElement
 2) Load http://tinderbox.mozilla.org/showbuilds.cgi?tree=Firefox
 3) Quit the app.

Actual results:
Lots of assertions like
###!!! ASSERTION: Serial number requested for unrecognized pointer!  Are you memmoving a refcounted object?: 'serialno != 0', file /Users/Shared/Projects/mozilla-html5/xpcom/base/nsTraceRefcntImpl.cpp, line 996
###!!! ASSERTION: Serial number requested for unrecognized pointer!  Are you memmoving a refcounted object?: 'serialno != 0', file /Users/Shared/Projects/mozilla-html5/xpcom/base/nsTraceRefcntImpl.cpp, line 1050

(Looking at the stacks shows that this is cycle collector releasing stuff.)

Expected results:
No assertions.

Additional info:
Happens both with the old parser and the HTML5 parser.

Comment 1

[Boris Zbarsky [:bzbarsky]]

And is this a recent regression, or has it been around for a while?

Comment 2

[Peter Van der Beken [:peterv]]

Any stacks? I doubt this is a CC problem. I've seen this for documents, because they're added as mutation observers to themselves. We end up refcounting them from LastRelease where we addref the mutation observers, but we've already removed the object from the serial number hash (from the last release).

Comment 3

[Firefox Bug Husbandry Bot]

https://bugzilla.mozilla.org/show_bug.cgi?id=1472046

Move all DOM bugs that haven’t been updated in more than 3 years and has no one currently assigned to P5.

If you have questions, please contact :mdaly.

Comment 4

[u597032]

Page no longer active