RE: lockdep oddity

["Hua Zhong" <hzhong@gmail.com>]

Maybe we should define raw __likely/__unlikely which behave the same way as the vanilla and use them in places like spinlocks to
avoid these weird problems.

> * Heiko Carstens <heiko.carstens@de.ibm.com> wrote:
> 
> > The lock validator gives me this (latest -mm and 2.6.18-rc6):
> > 
> > =====================================
> > [ BUG: bad unlock balance detected! ]
> > -------------------------------------
> > swapper/0 is trying to release lock (resource_lock) at:
> > [<0000000000042842>] request_resource+0x52/0x88 but there 
> are no more 
> > locks to release!
> > 
> > The reason is that the BUILD_LOCK_OPS macros in 
> kernel/lockdep.c don't 
> > contain any of the *_acquire calls, while all of the 
> _unlock functions 
> > contain a *_release call. Hence I get immediately unbalanced locks.
> 
> hmmm ... that sounds like a bug. Weird - i recently ran 
> PREEMPT+SMP+LOCKDEP kernels and didnt notice this.
> 
> > Found this will debugging some random memory corruptions 
> that happen 
> > when CONFIG_PROVE_LOCKING and CONFIG_PROFILE_LIKELY are both on.
> > Switching both off or having only one of them on seems to work.
> 
> previously i had some weirdnesses with PROFILE_LIKELY too, 
> they were caused by it generating cross-calls from within 
> lockdep. Do the corruptions go away if you remove all 
> likely() and unlikely() markings from kernel/lockdep.c?
> 
> 	Ingo