On 2013/7/12 0:22, Michal Hocko wrote:

On Thu 11-07-13 08:44:08, Tejun Heo wrote:

quoted

Hello, Michal.

On Thu, Jul 11, 2013 at 11:33:00AM +0200, Michal Hocko wrote:

quoted

+static inline
+struct mem_cgroup *vmpressure_to_mem_cgroup(struct vmpressure *vmpr)
+{
+	return container_of(vmpr, struct mem_cgroup, vmpressure);
+}
+
+void vmpressure_pin_memcg(struct vmpressure *vmpr)
+{
+	struct mem_cgroup *memcg = vmpressure_to_mem_cgroup(vmpr);
+
+	css_get(&memcg->css);
+}
+
+void vmpressure_unpin_memcg(struct vmpressure *vmpr)
+{
+	struct mem_cgroup *memcg = vmpressure_to_mem_cgroup(vmpr);
+
+	css_put(&memcg->css);
+}

So, while this *should* work, can't we just cancel/flush the work item
from offline? 

I would rather not put vmpressure clean up code into memcg offlining.
We have reference counting for exactly this purposes so it feels strange
to overcome it like that.

I'd agree with Tejun here. Asynchrously should be avoided if not necessary,
and the change would be simpler. There's already a vmpressure_init() in
mem_cgroup_css_alloc(), so it doesn't seem bad to do vmpressure cleanup
in memcg.