Re: [PATCH 4/5] oom: the points calculation of child processesmust use find_lock_task_mm() too

[Oleg Nesterov]

On 05/31, KOSAKI Motohiro wrote:
>
> --- a/mm/oom_kill.c
> +++ b/mm/oom_kill.c
> @@ -87,6 +87,7 @@ static struct task_struct *find_lock_task_mm(struct task_struct *p)
>  unsigned long badness(struct task_struct *p, unsigned long uptime)
>  {
>  	unsigned long points, cpu_time, run_time;
> +	struct task_struct *c;
>  	struct task_struct *child;
>  	int oom_adj = p->signal->oom_adj;
>  	struct task_cputime task_time;
> @@ -124,11 +125,13 @@ unsigned long badness(struct task_struct *p, unsigned long uptime)
>  	 * child is eating the vast majority of memory, adding only half
>  	 * to the parents will make the child our kill candidate of choice.
>  	 */
> -	list_for_each_entry(child, &p->children, sibling) {
> -		task_lock(child);
> -		if (child->mm != p->mm && child->mm)
> -			points += child->mm->total_vm/2 + 1;
> -		task_unlock(child);
> +	list_for_each_entry(c, &p->children, sibling) {
> +		child = find_lock_task_mm(c);
> +		if (child) {
> +			if (child->mm != p->mm)
> +				points += child->mm->total_vm/2 + 1;
> +			task_unlock(child);
> +		}

Acked-by: Oleg Nesterov <oleg@xxxxxxxxxx>




And, I think we need another patch on top of this one. Note that
this list_for_each_entry(p->children) can only see the tasks forked
by p, it can't see other children forked by its sub-threads.

IOW, we need

	do {
		list_for_each_entry(c, &t->children, sibling) {
			...
		}
	} while_each_thread(p, t);

Probably the same for oom_kill_process().

What do you think?

Oleg.

--
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/