[PATCH 2/8] cgroup: kill CSS_REMOVED
tj at kernel.org
Wed Oct 31 19:33:29 UTC 2012
On Wed, Oct 31, 2012 at 08:16:02PM +0100, Michal Hocko wrote:
> Sure, it cannot detect races but this wasn't the intention of the check.
> The more important part is that memcg can outlive css and we want to
> catch bugs when we try to charge to an already dead memcg.
cgroup will provide mechanisms to perform such checks properly
synchronized, so let's not leak internals like CSS rercnt going
negative outside cgroup.c.
More information about the Containers