Messages in this thread | | | Date | Wed, 22 Jun 2011 16:01:35 -0700 (PDT) | From | David Rientjes <> | Subject | Re: [PATCH 3/6] oom: kill younger process first |
| |
On Wed, 22 Jun 2011, KOSAKI Motohiro wrote:
> diff --git a/include/linux/sched.h b/include/linux/sched.h > index e4e6d7b..392ff30 100644 > --- a/include/linux/sched.h > +++ b/include/linux/sched.h > @@ -2257,6 +2257,9 @@ static inline unsigned long wait_task_inactive(struct task_struct *p, > #define next_task(p) \ > list_entry_rcu((p)->tasks.next, struct task_struct, tasks) > > +#define prev_task(p) \ > + list_entry((p)->tasks.prev, struct task_struct, tasks) > + > #define for_each_process(p) \ > for (p = &init_task ; (p = next_task(p)) != &init_task ; ) > > @@ -2269,6 +2272,14 @@ extern bool current_is_single_threaded(void); > #define do_each_thread(g, t) \ > for (g = t = &init_task ; (g = t = next_task(g)) != &init_task ; ) do > > +/* > + * Similar to do_each_thread(). but two difference are there. > + * - traverse tasks reverse order (i.e. younger to older) > + * - caller must hold tasklist_lock. rcu_read_lock isn't enough > +*/ > +#define do_each_thread_reverse(g, t) \ > + for (g = t = &init_task ; (g = t = prev_task(g)) != &init_task ; ) do > + > #define while_each_thread(g, t) \ > while ((t = next_thread(t)) != g) >
I've already ack'd the patch, but if there is another version posted as a result of our discussion of using euid in the heuristic, I think it would be helpful to reiterate in this comment that, like do_each_thread(), do_each_thread_reverse() is not break-safe either. It might end up preventing a bug down the road.
| |