http://bugs.winehq.org/show_bug.cgi?id=24415
--- Comment #5 from Alan J Richardson ajrich@earthlink.net 2010-09-15 23:20:43 CDT --- Created an attachment (id=30781) --> (http://bugs.winehq.org/attachment.cgi?id=30781) Additional backtraces
I backed up my local git to the 1.3.2 release and applied the patch from bug 24332 to generate these three backtraces.
Some notes - the first two were run consecutively from the same terminal. The third was run from a new terminal, and also includes a bit more of the output illustrating a pattern I've noticed: the unhandled exception appears to be a repeat of a handled exception. (Or maybe that's just the first pass feature of the debugger?)
I let the launcher download up to "acceptable" before starting to play, taking advantage of the new "streaming" feature. Allowing the patcher to complete may be a workaround, if the crash is caused by the streaming code. Could those who don't see this crash check in with whether they are streaming or allowed the launcher to completely install the client before attempting play. Testing myself would involve many hours of downloading, especially if it is indeed a workaround and I'd have to start the install over again to reproduce the bug.
I'll try +heap next.