Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752875Ab1BVUyx (ORCPT ); Tue, 22 Feb 2011 15:54:53 -0500 Received: from smtp.outflux.net ([198.145.64.163]:33559 "EHLO smtp.outflux.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751182Ab1BVUyv (ORCPT ); Tue, 22 Feb 2011 15:54:51 -0500 Date: Tue, 22 Feb 2011 12:54:13 -0800 From: Kees Cook To: Greg KH Cc: Alan Cox , David Daney , linux-kernel@vger.kernel.org, Eugene Teo , Ralph Campbell , Roland Dreier , Sean Hefty , Hal Rosenstock , Jeremy Fitzhardinge , Konrad Rzeszutek Wilk , Alexander Viro , Miklos Szeredi , "J. Bruce Fields" , Neil Brown , Matthew Wilcox , James Morris , Stephen Smalley , Eric Paris , Nick Piggin , Arnd Bergmann , Ian Campbell , Jarkko Sakkinen , Tejun Heo , Casey Schaufler Subject: Re: [PATCH 2/2] debugfs: only allow root access to debugging interfaces Message-ID: <20110222205413.GH4000@outflux.net> References: <20110222181613.GU4000@outflux.net> <4D640133.9020901@caviumnetworks.com> <20110222184726.GV4000@outflux.net> <20110222191454.GB9991@suse.de> <20110222192532.GY4000@outflux.net> <20110222193418.773ccd4b@lxorguk.ukuu.org.uk> <20110222195018.GA4000@outflux.net> <20110222201610.GA29787@suse.de> <20110222202856.GE4000@outflux.net> <20110222203704.GA7224@suse.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20110222203704.GA7224@suse.de> Organization: Canonical X-HELO: www.outflux.net Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3891 Lines: 81 On Tue, Feb 22, 2011 at 12:37:04PM -0800, Greg KH wrote: > On Tue, Feb 22, 2011 at 12:28:56PM -0800, Kees Cook wrote: > > On Tue, Feb 22, 2011 at 12:16:10PM -0800, Greg KH wrote: > > > On Tue, Feb 22, 2011 at 11:50:18AM -0800, Kees Cook wrote: > > > > On Tue, Feb 22, 2011 at 07:34:18PM +0000, Alan Cox wrote: > > > > > > What system do you proposed to keep these "stupid mistakes" from > > > > > > continuing to happen? If debugfs had already been mode 0700, we could have > > > > > > avoided all of these CVEs, including the full-blown local root escalation. > > > > > > > > > > And all sorts of features would have put themselves in sysfs instead and > > > > > broken no doubt. > > > > > > > > > > > The "no rules" approach to debugfs is not a good idea, IMO. > > > > > > > > > > It's a debugging fs, it needs to be "no rules" other than the obvious > > > > > "don't mount it on production systems" > > > > > > > > Okay, so the debugfs is not supposed to be mounted on a production system. > > > > > > No, not true at all, the "enterprise" distros all mount debugfs for good > > > reason on their systems. > > > > What reasons are those? Or better yet, why do you and Alan Cox disagree on > > this point? > > These distros have made the decision to support the perf interface, > which lives in debugfs, for their customers. I'm not saying that I > disagree with Alan about this, just pointing out the reality of the > situation here. A tool used only by the root user, so the proposed mount mode of 0700 wouldn't break anything. > > > > This seems to be news to a lot of developers trying to use the interfaces > > > > exposed there. It would be nice to say this more loudly. Basically, > > > > a normal system should not depend on anything in the debugfs. I can get > > > > behind that. > > > > > > Again, not true. Mostly all due to the perf interface, fix that to move > > > out of debugfs (patches have been proposed) and this problem will go > > > away. > > > > You can't have "no rules" and "all distros mount debugfs for good reason". > > This is asking for (even more) trouble. If there is something universally > > useful in debugfs (I do not count perf as universally useful -- my parents > > do not use perf), then why is it living in a filesystem with no rules > > (where "no rules" seems to also include "don't break interfaces"). > > Again, "don't break interfaces" is just me saying "don't break the > interfaces I have created in debugfs as they are to be used by all > users." Don't take that as a set-in-stone rule of debugfs at all, it > isn't. > > Again, you are trying to exclude a whole range of useful and valid files > from being used, when there was only a very very very small percentage > created incorrectly. They have now been fixed, and we have the > infrastructure to prevent future ones from being created as well, so I > don't see the issue here anymore. I'm trying to minimize exposure. So far, debugfs has proven itself to be repeatedly dangerous/flawed. I would like to take preventative measures to contain it. Everyone seems to agree that debugfs is useful for debugging, and I don't doubt that. It may also be riddled with potential holes, so why expose an entire tree of debugging interfaces to non-root users? We have %pK to keep kernel addresses out of the hands of non-root users when reading the debugging interfaces, and we have either my patchset or Dan Carpenter's to keep non-root users from writing to these debugging interfaces. There needs to be a way for system owners to be able to protect themselves proactively from debugfs. -Kees -- Kees Cook Ubuntu Security Team -- 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/