Re: [PATCH 2/5] oom: select_bad_process: PF_EXITING check should take ->mm into account

[Oleg Nesterov <oleg@redhat.com>]

Thanks a lot Kosaki for doing this!

I still can't find the time to play with this code :/

On 05/31, KOSAKI Motohiro wrote:
>
> select_bad_process() checks PF_EXITING to detect the task which is going
> to release its memory, but the logic is very wrong.
>
> 	- a single process P with the dead group leader disables
> 	  select_bad_process() completely, it will always return
> 	  ERR_PTR() while P can live forever
>
> 	- if the PF_EXITING task has already released its ->mm
> 	  it doesn't make sense to expect it is goiing to free
> 	  more memory (except task_struct/etc)
>
> Change the code to ignore the PF_EXITING tasks without ->mm.
>
> --- a/mm/oom_kill.c
> +++ b/mm/oom_kill.c
> @@ -287,7 +287,7 @@ static struct task_struct *select_bad_process(unsigned long *ppoints,
>  		 * the process of exiting and releasing its resources.
>  		 * Otherwise we could get an easy OOM deadlock.
>  		 */
> -		if (p->flags & PF_EXITING) {
> +		if ((p->flags & PF_EXITING) && p->mm) {

(strictly speaking, this change is needed after 3/5 which removes the
 top-level "if (!p->mm)" check in select_bad_process).


I'd like to add a note... with or without this, we have problems
with the coredump. A thread participating in the coredumping
(group-leader in this case) can have PF_EXITING && mm, but this doesn't
mean it is going to exit soon, and the dumper can use a lot more memory.

Otoh, if select_bad_process() chooses the thread which dumps the core,
SIGKILL can't stop it. This should be fixed in do_coredump() paths, this
is the long-standing problem.

And, as it was already discussed, we only check the group-leader here.
But I can't suggest something better.

Oleg.