From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753082AbaHDWQO (ORCPT ); Mon, 4 Aug 2014 18:16:14 -0400 Received: from a.ns.miles-group.at ([95.130.255.143]:65275 "EHLO radon.swed.at" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752710AbaHDWQM (ORCPT ); Mon, 4 Aug 2014 18:16:12 -0400 Message-ID: <53E0062A.8030204@nod.at> Date: Tue, 05 Aug 2014 00:16:10 +0200 From: Richard Weinberger User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.6.0 MIME-Version: 1.0 To: Ram Pai CC: "Eric W. Biederman" , linux-fsdevel , Al Viro , Christoph Hellwig , Paul McKenney , Jeff Mahoney , sahne@0x90.at, "linux-kernel@vger.kernel.org" Subject: Re: MNT_DETACH and mount namespace issue References: <1406728756-32443-1-git-send-email-richard@sigma-star.at> <53D959A7.5070702@nod.at> <53DAC069.7070606@nod.at> <20140801154448.GE24719@ram.oc3035372033.ibm.com> <53DBE894.3020900@nod.at> <87ha1v6ewb.fsf@x220.int.ebiederm.org> <874mxs2of1.fsf@x220.int.ebiederm.org> <53DFF8E7.2060906@nod.at> <20140804221051.GH24719@ram.oc3035372033.ibm.com> In-Reply-To: <20140804221051.GH24719@ram.oc3035372033.ibm.com> X-Enigmail-Version: 1.6 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Am 05.08.2014 00:10, schrieb Ram Pai: > On Mon, Aug 04, 2014 at 11:19:35PM +0200, Richard Weinberger wrote: >> Am 04.08.2014 18:46, schrieb Eric W. Biederman: >>> Richard Weinberger writes: >> >> /proc is propagating into another mount namespaces that does not care. >> This happens because systemd creates for several services a mount namespace and sets >> the root tree to MS_SHARED. > > if propagations are not needed, than set the root of the new mount > namespace to MS_PRIVATE first and then set it to MS_SHARED. > > MS_PRIVATE will delink the propagations, and MS_SHARED later will enable > the new mounts to propagate to whoever wants them. AFAICT this would break systemd's PrivateTmp feature. :( They want propagation. Such that a systemd service has a private /tmp but sees freshly mounted filesystems after setting up the namespace. Thanks, //richard