Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756892AbdLOQfj (ORCPT ); Fri, 15 Dec 2017 11:35:39 -0500 Received: from mx2.suse.de ([195.135.220.15]:36376 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756822AbdLOQfh (ORCPT ); Fri, 15 Dec 2017 11:35:37 -0500 Date: Fri, 15 Dec 2017 17:35:34 +0100 From: Michal Hocko To: David Rientjes Cc: Andrew Morton , Andrea Arcangeli , Benjamin Herrenschmidt , Paul Mackerras , Oded Gabbay , Alex Deucher , Christian =?iso-8859-1?Q?K=F6nig?= , David Airlie , Joerg Roedel , Doug Ledford , Jani Nikula , Mike Marciniszyn , Sean Hefty , Dimitri Sivanich , Boris Ostrovsky , =?iso-8859-1?B?Suly9G1l?= Glisse , Paolo Bonzini , Radim =?utf-8?B?S3LEjW3DocWZ?= , linux-kernel@vger.kernel.org, linux-mm@kvack.org Subject: Re: [patch v2 2/2] mm, oom: avoid reaping only for mm's with blockable invalidate callbacks Message-ID: <20171215163534.GB16951@dhcp22.suse.cz> References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.9.1 (2017-09-22) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2449 Lines: 60 On Thu 14-12-17 13:31:00, David Rientjes wrote: > This uses the new annotation to determine if an mm has mmu notifiers with > blockable invalidate range callbacks to avoid oom reaping. Otherwise, the > callbacks are used around unmap_page_range(). Do you have any example where this helped? KVM guest oom killed I guess? > Signed-off-by: David Rientjes Acked-by: Michal Hocko > --- > mm/oom_kill.c | 21 +++++++++++---------- > 1 file changed, 11 insertions(+), 10 deletions(-) > > diff --git a/mm/oom_kill.c b/mm/oom_kill.c > --- a/mm/oom_kill.c > +++ b/mm/oom_kill.c > @@ -514,15 +514,12 @@ static bool __oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm) > } > > /* > - * If the mm has notifiers then we would need to invalidate them around > - * unmap_page_range and that is risky because notifiers can sleep and > - * what they do is basically undeterministic. So let's have a short > + * If the mm has invalidate_{start,end}() notifiers that could block, > * sleep to give the oom victim some more time. > * TODO: we really want to get rid of this ugly hack and make sure that > - * notifiers cannot block for unbounded amount of time and add > - * mmu_notifier_invalidate_range_{start,end} around unmap_page_range > + * notifiers cannot block for unbounded amount of time > */ > - if (mm_has_notifiers(mm)) { > + if (mm_has_blockable_invalidate_notifiers(mm)) { > up_read(&mm->mmap_sem); > schedule_timeout_idle(HZ); > goto unlock_oom; > @@ -565,10 +562,14 @@ static bool __oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm) > * count elevated without a good reason. > */ > if (vma_is_anonymous(vma) || !(vma->vm_flags & VM_SHARED)) { > - tlb_gather_mmu(&tlb, mm, vma->vm_start, vma->vm_end); > - unmap_page_range(&tlb, vma, vma->vm_start, vma->vm_end, > - NULL); > - tlb_finish_mmu(&tlb, vma->vm_start, vma->vm_end); > + const unsigned long start = vma->vm_start; > + const unsigned long end = vma->vm_end; > + > + tlb_gather_mmu(&tlb, mm, start, end); > + mmu_notifier_invalidate_range_start(mm, start, end); > + unmap_page_range(&tlb, vma, start, end, NULL); > + mmu_notifier_invalidate_range_end(mm, start, end); > + tlb_finish_mmu(&tlb, start, end); > } > } > pr_info("oom_reaper: reaped process %d (%s), now anon-rss:%lukB, file-rss:%lukB, shmem-rss:%lukB\n", -- Michal Hocko SUSE Labs