Re: Debugging crash in _class_shouldFinalizeOnMainThread

2010-09-01 Thread Barry Wark
On Wed, Sep 1, 2010 at 3:52 PM, Greg Parker wrote: > On Sep 1, 2010, at 11:15 AM, Barry Wark wrote: >> I'm getting intermittent crashes in my test suite for a GC application >> on OS X 10.6.4. The KERN_INVALID_ADDRESS at 0x00090028 occurs >> in >> >> 0   libobjc.A.dylib                   0

Re: Debugging crash in _class_shouldFinalizeOnMainThread

2010-09-01 Thread Sean McBride
On Wed, 1 Sep 2010 12:52:57 -0700, Greg Parker said: >It's likely that a memory smasher or use-after-free of some other object >has clobbered this object's isa. Try running with guardmalloc and >AUTO_USE_GUARDS=YES to see if those uncover a smasher somewhere. And do I recall correctly that one sh

Re: Debugging crash in _class_shouldFinalizeOnMainThread

2010-09-01 Thread Greg Parker
On Sep 1, 2010, at 11:15 AM, Barry Wark wrote: > I'm getting intermittent crashes in my test suite for a GC application > on OS X 10.6.4. The KERN_INVALID_ADDRESS at 0x00090028 occurs > in > > 0 libobjc.A.dylib 0x7fff82ac847b > _class_shouldFinalizeOnMainThread + 4

Debugging crash in _class_shouldFinalizeOnMainThread

2010-09-01 Thread Barry Wark
I'm getting intermittent crashes in my test suite for a GC application on OS X 10.6.4. The KERN_INVALID_ADDRESS at 0x00090028 occurs in 0 libobjc.A.dylib0x7fff82ac847b _class_shouldFinalizeOnMainThread + 4 within the Dispatch queue: Garbage Collection Work Queue