Re: [RFC][mm] [PATCH 4/4] Memory cgroup hierarchy feature selector (v3)

[Li Zefan <lizf@cn.fujitsu.com>]

> +	/*
> +	 * If parent's use_hiearchy is set, we can't make any modifications
> +	 * in the child subtrees. If it is unset, then the change can
> +	 * occur, provided the current cgroup has no children.
> +	 *
> +	 * For the root cgroup, parent_mem is NULL, we allow value to be
> +	 * set if there are no children.
> +	 */
> +	if (!parent_mem || (!parent_mem->use_hierarchy &&
> +				(val == 1 || val == 0))) {
> +		if (list_empty(&cont->children))
> +			mem->use_hierarchy = val;
> +		else
> +			retval = -EBUSY;
> +	} else
> +		retval = -EINVAL;
> +
> +	return retval;
> +}

As I mentioned there is a race here. :(

echo 1 > /memcg/memory.use_hierarchy
 =>if (list_empty(&cont->children))
                                      mkdir /memcg/0
                                       => mem->use_hierarchy = 0
       mem->use_hierarchy = 1;

Now it ends up with parent's use_hierarchy is set but its child's
use_hierarchy is not set.

--
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>