Re: [PATCH 1/3] memcg: oom wakeup filter

[KAMEZAWA Hiroyuki]

On Mon, 15 Mar 2010 14:50:13 -0700
Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:

> On Fri, 12 Mar 2010 14:31:37 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> wrote:
> 
> > +static int memcg_oom_wake_function(wait_queue_t *wait,
> > +	unsigned mode, int sync, void *arg)
> > +{
> > +	struct mem_cgroup *wake_mem = (struct mem_cgroup *)arg;
> > +	struct oom_wait_info *oom_wait_info;
> > +
> > +	/* both of oom_wait_info->mem and wake_mem are stable under us */
> > +	oom_wait_info = container_of(wait, struct oom_wait_info, wait);
> > +
> > +	if (oom_wait_info->mem == wake_mem)
> > +		goto wakeup;
> > +	/* if no hierarchy, no match */
> > +	if (!oom_wait_info->mem->use_hierarchy || !wake_mem->use_hierarchy)
> > +		return 0;
> > +	/* check hierarchy */
> > +	if (!css_is_ancestor(&oom_wait_info->mem->css, &wake_mem->css) &&
> > +	    !css_is_ancestor(&wake_mem->css, &oom_wait_info->mem->css))
> > +		return 0;
> > +
> > +wakeup:
> > +	return autoremove_wake_function(wait, mode, sync, arg);
> > +}
> 
> What are the locking rules for calling css_is_ancestor()?

css_is_ancestor is checking css->id (and hierarchy stack). What we need here is
to guarantee css is valid object.

Here, we have reference count of both of oom_wait_info->mem and wake_mem.
Then, ->css is always vaild and it's "id" is stable under us.

Hmm, maybe this comment is too short.
/* both of oom_wait_info->mem and wake_mem are stable under us */
I'll prepare some update.


Thanks,
-Kame



--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/