Michal Hocko wrote: > On Wed 18-05-16 22:30:14, Tetsuo Handa wrote: > > Even if you call p->signal->oom_score_adj == OOM_SCORE_ADJ_MIN case a bug, > > (p->flags & PF_KTHREAD) || is_global_init(p) case is still possible. > > I couldn't care less about such a case to be honest, and that is not a > reason the cripple the code for such an insanity. There simply doesn't > make any sense to share init's mm with a different task. The global init called vfork(), and the child tried to call execve() with large argv/envp array, and the child got OOM-killed is possible. > OK, this looks correct. Strictly speaking the patch is missing any note > on _why_ this is needed or an improvement. I would add something like > the following: > " > Although the original code was correct it was quite inefficient because > each thread group was scanned num_threads times which can be a lot > especially with processes with many threads. Even though the OOM is > extremely cold path it is always good to be as effective as possible > when we are inside rcu_read_lock() - aka unpreemptible context. > " rcu_read_lock() is not always unpreemptible context. rcu_read_lock() says: In non-preemptible RCU implementations (TREE_RCU and TINY_RCU), it is illegal to block while in an RCU read-side critical section. In preemptible RCU implementations (PREEMPT_RCU) in CONFIG_PREEMPT kernel builds, RCU read-side critical sections may be preempted, but explicit blocking is illegal. Finally, in preemptible RCU implementations in real-time (with -rt patchset) kernel builds, RCU read-side critical sections may be preempted and they may also block, but only when acquiring spinlocks that are subject to priority inheritance. We will need preempt_disable() if we want to make out_of_memory() return as fast as possible. > > > Signed-off-by: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx> > > Cc: David Rientjes <rientjes@xxxxxxxxxx> > > Cc: Oleg Nesterov <oleg@xxxxxxxxxx> > > Acked-by: Michal Hocko <mhocko@xxxxxxxx> Thank you. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>