Open
Bug 933730
Opened 11 years ago
Updated 2 years ago
"ASSERTION: Appshell already destroyed?" due to making OfflineAudioContext in a dying Window
Categories
(Core :: Web Audio, defect, P3)
Tracking
()
NEW
People
(Reporter: jruderman, Assigned: karlt)
References
Details
(Keywords: assertion, memory-leak, testcase)
Attachments
(3 files)
1. Load the testcase 2. Click the button 3. Quit Firefox Result: assertion during shutdown, and leak ###!!! ASSERTION: Appshell already destroyed?: 'Error', file content/media/MediaStreamGraph.cpp, line 1538
Reporter | ||
Comment 1•11 years ago
|
||
Reporter | ||
Comment 2•11 years ago
|
||
Comment 3•11 years ago
|
||
Karl -- Can I give this bug to you? I know I've been giving a lot to you in the last week; so if you need me to find someone else for this one, just let me know. Thanks!
Assignee: nobody → karlt
Assignee | ||
Comment 4•11 years ago
|
||
Jesse, let me know if this or other bugs are blocking you, and I'll change priorities.
Priority: -- → P2
Reporter | ||
Comment 5•11 years ago
|
||
For this bug, I'm having my fuzzer ignore the assertion, and ignore leaks when the assertion occurs. So I don't think it's blocking much.
Assignee | ||
Comment 6•10 years ago
|
||
When nsCycleCollector_shutdown() is called, nsThreadManager has already shutdown all nsThreads. StartNonRealtimeProcessing() won't get the new nsThread that it expects. I wonder whether we can get DestroyNonRealtimeInstance to destroy the graph without creating the thread, by starting the graph at LIFECYCLE_WAITING_FOR_MAIN_THREAD_CLEANUP. But it also sounds like AudioContext shutdown is happening too early. I wonder whether that could be delayed until after we know there is no script running.
Updated•9 years ago
|
Rank: 29
Comment 7•7 years ago
|
||
Mass change P2->P3 to align with new Mozilla triage process.
Priority: P2 → P3
Updated•2 years ago
|
Severity: normal → S3
You need to log in
before you can comment on or make changes to this bug.
Description
•