On 08/10/2012 09:27 PM, Kamezawa Hiroyuki wrote:

quoted

+bool __memcg_kmem_new_page(gfp_t gfp, void *_handle, int order)

quoted

+{
+	struct mem_cgroup *memcg;
+	struct mem_cgroup **handle = (struct mem_cgroup **)_handle;
+	bool ret = true;
+	size_t size;
+	struct task_struct *p;
+
+	*handle = NULL;
+	rcu_read_lock();
+	p = rcu_dereference(current->mm->owner);
+	memcg = mem_cgroup_from_task(p);
+	if (!memcg_kmem_enabled(memcg))
+		goto out;
+
+	mem_cgroup_get(memcg);
+

This mem_cgroup_get() will be a potentioal performance problem.
Don't you have good idea to avoid accessing atomic counter here ?
I think some kind of percpu counter or a feature to disable "move task"
will be a help.

I have just sent out a proposal to deal with this. I tried the trick of
marking only the first charge and last uncharge, and it works quite
alright at the cost of a bit test on most calls to memcg_kmem_charge.

Please let me know what you think.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>