Bug 22510 - Crash at WebCore::AccessibilityRenderObject::activeDescendant() on Google Reader with ARIA
Summary: Crash at WebCore::AccessibilityRenderObject::activeDescendant() on Google Rea...
Status: VERIFIED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Accessibility (show other bugs)
Version: 528+ (Nightly build)
Hardware: Mac OS X 10.5
: P1 Critical
Assignee: Nobody
URL: http://www.google.com/reader/axs
Keywords: InRadar, NeedsReduction, Regression
Depends on:
Blocks:
 
Reported: 2008-11-26 04:01 PST by Robert Blaut
Modified: 2008-12-02 10:49 PST (History)
2 users (show)

See Also:


Attachments
crash log (25.06 KB, text/plain)
2008-11-26 04:02 PST, Robert Blaut
no flags Details
patch to fix crash (1.38 KB, patch)
2008-11-30 12:37 PST, chris fleizach
sullivan: review+
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Robert Blaut 2008-11-26 04:01:15 PST
I experience replicable crashes on Google Reader (ARIA enabled) in the latest build: r38760 on Mac OS X 10.5

Steps to reproduce crash:

1) Go to http://www.google.com/reader/axs
2) Press "?" on keyboard. Info page should appear.
3) Press "?" again. Wait a couple of seconds.
4) WebKit crashes.

Crashlog attached.
Comment 1 Robert Blaut 2008-11-26 04:02:08 PST
Created attachment 25518 [details]
crash log
Comment 2 Mark Rowe (bdash) 2008-11-26 15:18:50 PST
<rdar://problem/6404263>
Comment 3 chris fleizach 2008-11-30 12:37:52 PST
Created attachment 25616 [details]
patch to fix crash
Comment 4 Cameron Zwarich (cpst) 2008-11-30 12:50:07 PST
Comment on attachment 25616 [details]
patch to fix crash

Is there some way to test this in a layout test?
Comment 5 chris fleizach 2008-11-30 13:04:19 PST
i couldn't find a good way to test this behavior with the layout test engine. if there are any ideas let me know
Comment 6 chris fleizach 2008-11-30 13:43:53 PST
(In reply to comment #4)
> (From update of attachment 25616 [details] [review])
> Is there some way to test this in a layout test?
> 

i couldn't find a good way to test this behavior with the layout test engine.
if there are any ideas let me know
Comment 7 chris fleizach 2008-11-30 21:36:36 PST
http://trac.webkit.org/changeset/38855