From mboxrd@z Thu Jan 1 00:00:00 1970 From: ebiederm@xmission.com (Eric W. Biederman) Subject: Re: [PATCH] sysfs: Don't allow the creation of symlinks we can't remove Date: Thu, 22 Jul 2010 03:35:29 -0700 Message-ID: References: <20100621150826.762ac9f2.akpm@linux-foundation.org> <20100621222938.GA20583@suse.de> <20100621231058.GA1066@suse.de> <20100622035631.GA3755@suse.de> <1279792459.12439.0.camel@jlt3.sipsolutions.net> <1279793435.12439.3.camel@jlt3.sipsolutions.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Greg KH , Andrew Morton , "Rafael J. Wysocki" , "Maciej W. Rozycki" , Kay Sievers , Greg KH , netdev To: Johannes Berg Return-path: Received: from out01.mta.xmission.com ([166.70.13.231]:60329 "EHLO out01.mta.xmission.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754944Ab0GVKfh (ORCPT ); Thu, 22 Jul 2010 06:35:37 -0400 In-Reply-To: <1279793435.12439.3.camel@jlt3.sipsolutions.net> (Johannes Berg's message of "Thu\, 22 Jul 2010 12\:10\:34 +0200") Sender: netdev-owner@vger.kernel.org List-ID: Johannes Berg writes: > On Thu, 2010-07-22 at 03:05 -0700, Eric W. Biederman wrote: > >> >> Detect this problem up front and simply don't create symlinks we won't >> >> be able to remove later. This prevents symlink leakage and fails in >> >> a much clearer and more understandable way. >> > >> > Eric, I was looking into sysfs netns support for wireless, and with this >> > patch applied I just get the warning and no network interfaces. >> >> The warning patch just makes things fail faster. Although I get some of the >> wireless interfaces for hwsim when I use this one. > > Hmm, I didn't. To be clear I just get hwsim0. Not wlan0 or wlan1. >> > Was there any patch that was supposed to fix hwsim? >> >> - If you have my patches that fix CONFIG_SYSFS_DEPRECATED, >> you should find everything works there. > > But then I was carrying those two patches too. > >> As for a proper fix I have just resent my one liner to >> drives/base/core.c I can't think of a better option right now. >> >> For hwsim it is arguable, but the behaviour of sysfs for the >> bluetooth bnep driver is very clearly a 3 year old regression, >> and the cause is exactly the same. > > Right, it actually starts working again with that patch you sent. > However, netns support is really broken: > > Do we have a convenient command line tool to do this? I remember there being a different netlink message from normal network devices. > root@kvm:~# ip link > 3: wlan0: mtu 1500 qdisc noop state DOWN qlen 1000 > link/ether 02:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff > 7: lo: mtu 16436 qdisc noop state DOWN > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > root@kvm:~# ls /sys/class/net/ > eth0 hwsim0 lo wlan1 wlan2 I think this is actually the output of something working. I expect after you created a new netns you didn't mount a new instance of /sys. /sys remembers which netns you had when you mounted it. So you have to mount /sys again so you can see the /sys/class/net for the network namespace you are in. Eric