Q: cgroup: Questions about possible issues in cgroup locking

Frederic Weisbecker fweisbec at gmail.com
Wed Feb 1 16:28:55 UTC 2012


On Wed, Dec 21, 2011 at 02:08:48PM +0100, Oleg Nesterov wrote:
> On 12/21, Frederic Weisbecker wrote:
> > - Is the check for use_task_css_set_links in cgroup_post_fork() safe? given
> > it is checked outside css_set_lock?
> >
> > Imagine this:
> >
> > CPU 0                                                        CPU 1
> > ----                                                         -----
> >
> > cgroup_enable_task_cg() {
> > 	uset_tasks_css_set_links = 1
> > 	for_each_thread() {
> > 		add tasks in the list
> > 	}
> > }
> >                                                            do_fork() {
> >                                                                cgroup_post_fork() {
> >                                                                      use_tasks_css_set_links appears
> >                                                                      to be equal to 0 due to write/read
> >                                                                       not flushed. New task won't
> >                                                                       appear to the list.
> 
> Yes, I was thinking about this too.
> 
> Or (I think) they can race "contrariwise". CPU_1 creates the new child,
> then CPU_0 sets uset_tasks_css_set_links = 1. But afaics there is no any
> guarantee that CPU_0 sees the result of list_add_tail_rcu().

Exactly! In fact even if RCU was safe with while_each_thread() it wouldn't
be enough for us because of that.

I fear we need the read_lock(tasklist_lock) here, with a pair of smp
barriers to ensure use_task_css_set_links update is visible as
expected.

I'll try to cook something.


More information about the Containers mailing list