Messages in this thread | | | Date | Fri, 20 Jul 2007 14:09:09 -0700 | From | "Paul Menage" <> | Subject | Re: [RFC][-mm PATCH 4/8] Memory controller memory accounting (v3) |
| |
On 7/20/07, Paul Menage <menage@google.com> wrote: > > + mem = rcu_dereference(mm->mem_container); > > + /* > > + * For every charge from the container, increment reference > > + * count > > + */ > > + css_get(&mem->css); > > + rcu_read_unlock(); > > It's not clear to me that this is safe. >
Sorry, didn't finish this thought.
Right after the rcu_dereference() the mm could be moved to another container by some other process.
Since there's an rcu_synchronize() after the movement, and the rcu_synchronize() caller holds container_mutex then I guess it's not possible for a third process to delete the container during the RCU section, since it won't be able to acquire the mutex until after the RCU read section completes. So OK, it is safe, at least based on the guarantees made by the current implementation.
Paul - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |