Daniele Orlandi wrote:
> I'm looking at proto_unregister() in linux-2.6.13:
> 
> Il calls kmem_cache_destroy() while holding proto_list_lock:
> 
> void proto_unregister(struct proto *prot)
> {
>         write_lock(&proto_list_lock);
> 
>         if (prot->slab != NULL) {
>                 kmem_cache_destroy(prot->slab);
> 
> 
> However, kmem_cache_destroy() may sleep:
> 
>         /* Find the cache in the chain of caches. */
>         down(&cache_chain_sem);
>         ^^^^^^^^^^^^^^^^^^^^^^^
> 
> Am I seeing it right?

You're right, good catch. This patch fixes it by moving the lock
down to the list-operation which it is supposed to protect.