http://bugs.winehq.org/show_bug.cgi?id=18799
--- Comment #13 from Stefan Dösinger stefandoesinger@gmx.at 2009-12-13 06:13:25 --- Fwiw, new backtraces of the same crash are giving little to no hints towards what is going wrong here.
What I don't understand about this bug is that the game is crashing far away from a buffer call, so I don't see why the hack fixes it. It obviously does help, but I don't see why.
Suspicion 1 is that the there is a random memory corruption that got triggered by the buffer changes. If that's the case this will be very hard to debug.
The other suspicion is that the game accesses the buffer memory after unlock. If that's the case there would be a correlation between the bad read/write address and a pointer returned by buffer->map. That would be a game bug, but maybe one that never happens to trigger a crash on Windows.