On Aug 27, 2010, at 3:07 PM, John Ralls wrote: > > On Aug 27, 2010, at 11:58 AM, Phillip Heller wrote: > >> >> On Aug 25, 2010, at 9:05 PM, John Ralls wrote: >> >>> >>> I don't see any instances of libigemacintegration.dylib in that stack trace... >>> >>> "Bus Error" on OSX is the same as "SegFault" in Linux. You're dereferencing a pointer that has a bogus value (0x20). You'll have to fire up GDB to find out what pointer and how it's gotten the bad value. >> >> John, >> >> Thanks for the suggestion. Alas, I'm not terribly skilled with gdb, and really am not even sure where to start. >> >> However, I figured I'd try a different approach to debugging. Instead, I've written a very simple demo to identify a couple issues: >> >> Please find the demo code here: >> >> 1) When in the NSApplicationBlockTermination handler, I notice that "another" Command-Q will cause the application to immediately end. So, I have tried attaching a new handler to the signal within the original handler (this new handler is just a True value, such that the application doesn't end). However this still doesn't solve the problem. Any suggestions? >> >> 2) As to the crash, this demo does reveal something. When I quit by destroying the window with the red pill, cmd-q and ok, or appmenu -> quit, there is no problem. When I quit by clicking the dock icon-> quit, it does exit, but not without returning the following console message: >> >> 2010-08-27 14:57:48.454 python[2502:903] *** attempt to pop an unknown autorelease pool (0x200ea00) >> >> I appreciate all of your help and hard work on gtk-osx! > > Phil, > > Um, find the demo code *where*? Oops. http://pastebin.com/d1EFAzyk Sorry about that. --phil
Attachment:
smime.p7s
Description: S/MIME cryptographic signature