From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750986AbaLPBrm (ORCPT ); Mon, 15 Dec 2014 20:47:42 -0500 Received: from mga11.intel.com ([192.55.52.93]:34815 "EHLO mga11.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750722AbaLPBrl (ORCPT ); Mon, 15 Dec 2014 20:47:41 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.97,862,1389772800"; d="scan'208";a="429401377" Message-ID: <1418694459.5745.198.camel@intel.com> Subject: Re: [LKP] [userns] BUG: unable to handle kernel NULL pointer dereference at (null) From: Huang Ying To: "Eric W. Biederman" , Wu Fengguang Cc: LKML , LKP ML Date: Tue, 16 Dec 2014 09:47:39 +0800 In-Reply-To: <87iohcx1k1.fsf@x220.int.ebiederm.org> References: <1418613661.5745.164.camel@intel.com> <87iohcx1k1.fsf@x220.int.ebiederm.org> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.12.9-1 Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 2014-12-15 at 11:31 -0600, Eric W. Biederman wrote: > Huang Ying writes: > > > FYI, we noticed the below changes on > > > > git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace.git for-testing > > commit bbea5f5532501fdd67f46442ba7b1122d7ff3123 > > ("userns: Add a knob to disable setgroups on a per user namespace basis") > > Thank you. > > I am quite puzzled by this failure. There was an similar failure when > /proc/[pid]/setgroups was read (if I recall correctly). I don't see how > that change could result at failures during open or failures during > boot. I added a new file to proc which any reasonable system should > leave alone. Are you by chance running trinity during boot? Yes. We are running trinity during boot. > If the reproducer gave me any clue about which file that was opened or > which code path this happened on I would be bery interested. If for no > other reason that to confirm that I have fixed the issue. This is just boot test, with trinity running after kernel boot. Fengguang, do you have more information? Best Regards, Huang, Ying > I have rewritten the implementation of /proc/[pid]/setgroups so it is > simpler and more robust and does not have any errors I can detect. > > Thank you very much for picking up my for-testing branch and beating up > on it. > > Eric