Messages in this thread | | | Date | Wed, 03 Oct 2012 09:05:31 +0530 | From | "Srivatsa S. Bhat" <> | Subject | Re: Lockdep complains about commit 1331e7a1bb ("rcu: Remove _rcu_barrier() dependency on __stop_machine()") |
| |
On 10/03/2012 03:47 AM, Jiri Kosina wrote: > On Wed, 3 Oct 2012, Srivatsa S. Bhat wrote: > >> I don't see how this circular locking dependency can occur.. If you are using SLUB, >> kmem_cache_destroy() releases slab_mutex before it calls rcu_barrier(). If you are >> using SLAB, kmem_cache_destroy() wraps its whole operation inside get/put_online_cpus(), >> which means, it cannot run concurrently with a hotplug operation such as cpu_up(). So, I'm >> rather puzzled at this lockdep splat.. > > I am using SLAB here. > > The scenario I think is very well possible: > > > CPU 0 CPU 1 > kmem_cache_destroy()
What about the get_online_cpus() right here at CPU0 before calling mutex_lock(slab_mutex)? How can the cpu_up() proceed on CPU1?? I still don't get it... :(
(kmem_cache_destroy() uses get/put_online_cpus() around acquiring and releasing slab_mutex).
Regards, Srivatsa S. Bhat
> mutex_lock(slab_mutex) > _cpu_up() > cpu_hotplug_begin() > mutex_lock(cpu_hotplug.lock) > rcu_barrier() > _rcu_barrier() > get_online_cpus() > mutex_lock(cpu_hotplug.lock) > (blocks, CPU 1 has the mutex) > __cpu_notify() > mutex_lock(slab_mutex) > > Deadlock. > > Right? >
| |