[Webkit-unassigned] [Bug 192586] [WebAudio] Call AudioContext::uninitialize() immediately when the AudioContext is stopped

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=192586

--- Comment #13 from Yacine Bandou <bandou.yacine at gmail.com> ---
(In reply to Darin Adler from comment #11)
> (In reply to Yacine Bandou from comment #0)
> > This modification can now be reverted without regression in WebAudio tests.
> 
> I’m sorry; I don’t understand why this is the bar for knowing whether the
> code is needed. Do we have coverage for this teardown scenario in the
> WebAudio tests? The original patch when this change was made did not add any
> new tests.
> 
> It’s absolutely clear that we want to stop immediately.
> 
> What’s not so clear is if the problems with reentrancy and active DOM object
> teardown that originally motivated this change are covered with tests and
> have been solved. I am concerned about the test coverage.

Sorry for this late response.

The original patch has added a new test see https://trac.webkit.org/changeset/94608/webkit

This fix is somehow a revert of this original patch, and for me the non regression is guaranteed by this test.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20190122/162335f8/attachment.html>