From mboxrd@z Thu Jan 1 00:00:00 1970 From: Michal Hocko Subject: Re: memcg creates an unkillable task in 3.11-rc2 Date: Tue, 30 Jul 2013 14:31:20 +0200 Message-ID: <20130730123120.GA15847@dhcp22.suse.cz> References: <20130723174711.GE21100@mtj.dyndns.org> <8761vui4cr.fsf@xmission.com> <20130729075939.GA4678@dhcp22.suse.cz> <87ehahg312.fsf@xmission.com> <20130729095109.GB4678@dhcp22.suse.cz> <20130729161026.GD22605@mtj.dyndns.org> <87r4eh70yg.fsf@xmission.com> <51F71DE2.4020102@huawei.com> <87ppu0a298.fsf_-_@tw-ebiederman.twitter.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <87ppu0a298.fsf_-_-HxuHnoDHeQZYhcs0q7wBk77fW72O3V7zAL8bYrjMMd8@public.gmane.org> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org Errors-To: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org To: "Eric W. Biederman" Cc: Glauber Costa , containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Johannes Weiner , Tejun Heo , cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Linus Torvalds , kent.overstreet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org On Tue 30-07-13 01:19:31, Eric W. Biederman wrote: [...] > Hmm. Looking farther I see what is going on. And it has nothing to do > with the freezer. (I have commented out that code and reproduced it > without the freezer to be doubly certain). > > > On the exit path exit_robust_list is triggering a page fault to fault a > page back in. Which since we have no memory causes the exit path > to get stuck in mem_cgroup_handle_oom. Hmm, interesting. I assume the exit is caused by the SIGKILL, right? If yes, then why it hasn't coughed early in __mem_cgroup_try_charge > Which means the following change should fix the hang. I will test it in just > a second. > > The problem is that we only handled pending fatal signals and exiting > processes when the OOM logic was enabled. Sigh. If this is about PF_EXITING then we should handle that earlier in the code chain (__mem_cgroup_try_charge sounds like a proper place). Let me think about that. Thanks for debugging this! > Eric > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c > index 00a7a66..5998a57 100644 > --- a/mm/memcontrol.c > +++ b/mm/memcontrol.c > @@ -1792,16 +1792,6 @@ static void mem_cgroup_out_of_memory(struct mem_cgroup *memcg, gfp_t gfp_mask, > unsigned int points = 0; > struct task_struct *chosen = NULL; > > - /* > - * If current has a pending SIGKILL or is exiting, then automatically > - * select it. The goal is to allow it to allocate so that it may > - * quickly exit and free its memory. > - */ > - if (fatal_signal_pending(current) || current->flags & PF_EXITING) { > - set_thread_flag(TIF_MEMDIE); > - return; > - } > - > check_panic_on_oom(CONSTRAINT_MEMCG, gfp_mask, order, NULL); > totalpages = mem_cgroup_get_limit(memcg) >> PAGE_SHIFT ? : 1; > for_each_mem_cgroup_tree(iter, memcg) { > @@ -2220,7 +2210,15 @@ static bool mem_cgroup_handle_oom(struct mem_cgroup *memcg, gfp_t mask, > mem_cgroup_oom_notify(memcg); > spin_unlock(&memcg_oom_lock); > > - if (need_to_kill) { > + /* > + * If current has a pending SIGKILL or is exiting, then automatically > + * select it. The goal is to allow it to allocate so that it may > + * quickly exit and free its memory. > + */ > + if (fatal_signal_pending(current) || current->flags & PF_EXITING) { > + set_thread_flag(TIF_MEMDIE); > + finish_wait(&memcg_oom_waitq, &owait.wait); > + } else if (need_to_kill) { > finish_wait(&memcg_oom_waitq, &owait.wait); > mem_cgroup_out_of_memory(memcg, mask, order); > } else { -- Michal Hocko SUSE Labs