Re: [PATCH 5/5] cgroup: fix a race between cgroup_mount() and cgroup_kill_sb()

[Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>]

On Fri, Jun 27, 2014 at 02:32:33PM +0800, Li Zefan wrote:
> > cgroup_mount()
> > {
> > 	mutex_lock();
> > 	lookup_cgroup_root();
> > 	if (root isn't killed yet)
> > 		root->this_better_stay_alive++;
> > 	mutex_unlock();
> > 	kernfs_mount();
> > }
> > 
> > cgroup_kill_sb()
> > {
> > 	mutex_lock();
> > 	if (check whether root can be killed)
> > 		percpu_ref_kill();
> > 	mutex_unlock();
> > 	if (the above condition was true)
> > 		kernfs_kill_sb();
> > }
> > 
> 
> This looks nasty, and I don't think it's correct. If we skip the call
> to kernfs_kill_sb(), kernfs_super_info won't be freed but super_block
> will, so we will end up either leaking memory or accessing invalid
> memory. There are other problems like returning with sb->s_umount still
> held.

Yeah, right, the conditional shouldn't be on kernfs_kill_sb().  It
should only be on percpu_ref_kill().  kernfs mount code will wait out
the dead sb and create a new one; however, this is still not feasible
because we don't have a place to dec ->this_better_stay_alive as
there's no umount callback.  :(

Thanks.

-- 
tejun