Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751863AbYAOOkA (ORCPT ); Tue, 15 Jan 2008 09:40:00 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1750840AbYAOOjv (ORCPT ); Tue, 15 Jan 2008 09:39:51 -0500 Received: from fxip-0047f.externet.hu ([88.209.222.127]:51482 "EHLO pomaz-ex.szeredi.hu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750752AbYAOOjt (ORCPT ); Tue, 15 Jan 2008 09:39:49 -0500 To: serue@us.ibm.com CC: miklos@szeredi.hu, serue@us.ibm.com, akpm@linux-foundation.org, hch@infradead.org, viro@ftp.linux.org.uk, ebiederm@xmission.com, kzak@redhat.com, linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, containers@lists.osdl.org, util-linux-ng@vger.kernel.org In-reply-to: <20080115142115.GC4453@sergelap.austin.ibm.com> (serue@us.ibm.com) Subject: Re: [patch 8/9] unprivileged mounts: propagation: inherit owner from parent References: <20080108113502.184459371@szeredi.hu> <20080108113632.895453887@szeredi.hu> <20080114231340.GG6704@sergelap.austin.ibm.com> <20080115142115.GC4453@sergelap.austin.ibm.com> Message-Id: From: Miklos Szeredi Date: Tue, 15 Jan 2008 15:37:37 +0100 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2028 Lines: 49 > > > > On mount propagation, let the owner of the clone be inherited from the > > > > parent into which it has been propagated. Also if the parent has the > > > > "nosuid" flag, set this flag for the child as well. > > > > > > What about nodev? > > > > Hmm, I think the nosuid thing is meant to prevent suid mounts being > > introduced into a "suidless" namespace. This doesn't apply to dev > > mounts, which are quite safe in a suidless environment, as long as the > > user is not able to create devices. But that should be taken care of > > by capability tests. > > > > I'll update the description. > > Hmm, > > Part of me wants to say the safest thing for now would be to refuse > mounts propagation from non-user mounts to user mounts. > > I assume you're thinking about a fully user-mounted chroot, where > the user woudl still want to be able to stick in a cdrom and have > it automounted under /mnt/cdrom, propagated from the root mounts ns? Right. > But then are there no devices which the user could create on a floppy > while inserted into his own laptop, owned by his own uid, then insert > into this machine, and use the device under the auto-mounted /dev/floppy > to gain inappropriate access? I assume, that the floppy and cdrom are already mounted with nosuid,nodev. The problem case is I think is if a sysadmin does some mounting in the initial namespace, and this is propagated into the fully user-mounted namespace (or chroot), so that a mount with suid binaries slips in. Which is bad, because the user may be able rearange the namespace, to trick the suid program to something it should not do. OTOH, a mount with devices can't be abused this way, since it is not possible to gain privileges to files/devices just by rearanging the mounts. Miklos -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/