Le 9 avr. 2010 à 08:08, Greg Parker a écrit : > In this circumstance, mach_msg() will block if the queue is full, and either > succeed later or time out. Neither of those paths should end up at HALT. > (mach_msg() may do other things in other circumstances.) > > Presumably mach_msg() returned something other than MACH_MSG_SUCCESS or > MACH_SEND_TIMED_OUT in your example. What's in register rax in your crash log?
rax: 0x0000000010000003 So I suppose it's what somebody mentioned before: the runLoop has already died. When the runLoop is exited, my code resets the CFRunLoopRef to NULL: under this condition, no further call to CFRunLoopStop is possible. But I assume the extraneous call happened exactly between the exit of the run loop and the reset of the pointer (at first sight, I see no reason why CFRunLoopStop would be called twice, but…). I now protect the call to CFRunLoopStop() by a flag that I set to true when the function is called for the first time, so it might solve this issue. Thanks for helping, Vincent_______________________________________________ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
