http://bugs.winehq.org/show_bug.cgi?id=16051
--- Comment #17 from Jeremy White jwhite@codeweavers.com 2009-07-13 15:25:56 --- Blech. My tests with those patches were invalid. You do get somewhat inconsistent behavior at startup. Mostly you'll get a c0000005 exception that will start a chain reaction leading to death. But other times you'll get a hang with some other thread dying off unexpectedly.
I have a hack patch that I started using to speed up my testing; it's not a fix (it's an anti-fix :-/); it just gets those c0000005 exceptions to crash out more quickly to make them easier to study. I'll attach it for future reference.