Ulrich Weigand writes:
>
> Ok, the problem is that it goes up the exception chain and finds frames
> that were posted while still on the old stack :-/ I don't think that
> unwinding exceptions across stacks is a good idea; what about the following
> patch that simply catches all exceptions before they leave the large stack?
>
Hallo Ulrich,
thanks, this second patch gets wine to enter the debugger again. As
the crash happens in xlib, --synchronous is needed to get a backtrace.
My other wish is that __TRY __EXCEPT was handled on the big stack too.
In my source tree the failing function (XCopyArea) is surrounded
by a __TRY clause and the corresponding __EXCEPT is not yet
entered. The debugger in entered via XSetErrorHandler -> error_handler
-> DebugBreak.
Is this possible to dispatch the exception? I tried to find
documentation on the error-code that xlib delivers, but didn't find
anything yet.
Bye
Uwe Bonnes [EMAIL PROTECTED]
Free Software: If you contribute nothing, expect nothing
--
