From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757397AbbIUXIp (ORCPT ); Mon, 21 Sep 2015 19:08:45 -0400 Received: from mail-pa0-f52.google.com ([209.85.220.52]:36775 "EHLO mail-pa0-f52.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757184AbbIUXIo (ORCPT ); Mon, 21 Sep 2015 19:08:44 -0400 Date: Mon, 21 Sep 2015 16:08:42 -0700 (PDT) From: David Rientjes X-X-Sender: rientjes@chino.kir.corp.google.com To: Michal Hocko cc: Kyle Walker , akpm@linux-foundation.org, hannes@cmpxchg.org, vdavydov@parallels.com, oleg@redhat.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] mm/oom_kill.c: don't kill TASK_UNINTERRUPTIBLE tasks In-Reply-To: <20150919082237.GB28815@dhcp22.suse.cz> Message-ID: References: <1442512783-14719-1-git-send-email-kwalker@redhat.com> <20150919082237.GB28815@dhcp22.suse.cz> User-Agent: Alpine 2.10 (DEB 1266 2009-07-14) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sat, 19 Sep 2015, Michal Hocko wrote: > Nack to this. TASK_UNINTERRUPTIBLE should be time constrained/bounded > state. Using it as an oom victim criteria makes the victim selection > less deterministic which is undesirable. As much as I am aware of > potential issues with the current implementation, making the behavior > more random doesn't really help. > Agreed, we can't avoid killing a process simply because it is in D state, this isn't an indication that the process will not be able to exit and in the worst case could panic the system if all other processes cannot be oom killed.