https://bugs.winehq.org/show_bug.cgi?id=42463
--- Comment #4 from Andrew Eikum aeikum@codeweavers.com --- (In reply to Sebastian Lackner from comment #2)
Created attachment 57310 [details] Proposed patch.
@Jactry: Could you please test that this patch fixes the issue?
I'm not opposed to this, but I don't understand why calling the init function twice would cause the assertion failure. It's protected by a mutex, so they should run in series, and we completely clean up when pulse_test_connect exits.
(See also Bug 40147 and Bug 41102.)