linux-kernel - Re: [PATCH] tracefs: Do not allocate and free proxy

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20191011182748.23d6de31@gandalf.local.home>
Date:   Fri, 11 Oct 2019 18:27:48 -0400
From:   Steven Rostedt <rostedt@...dmis.org>
To:     Florian Weimer <fw@...eb.enyo.de>
Cc:     Linus Torvalds <torvalds@...ux-foundation.org>,
        LKML <linux-kernel@...r.kernel.org>,
        Matthew Garrett <matthewgarrett@...gle.com>,
        James Morris James Morris <jmorris@...ei.org>,
        LSM List <linux-security-module@...r.kernel.org>,
        Linux API <linux-api@...r.kernel.org>,
        Ben Hutchings <ben@...adent.org.uk>,
        Al Viro <viro@...iv.linux.org.uk>
Subject: Re: [PATCH] tracefs: Do not allocate and free proxy_ops for
 lockdown

On Fri, 11 Oct 2019 23:46:20 +0200
Florian Weimer <fw@...eb.enyo.de> wrote:

> * Steven Rostedt:
> 
> > Once locked down is set, can it ever be undone without rebooting?  
> 
> I think this is the original intent with such patches, yes.  But then
> reality interferes and people add some escape hatch, so that it's
> possible again to load arbitrary kernel modules.  And for servers, you
> can't have a meaningful physical presence check, so you end up with a
> lot of complexity for something that offers absolutely zero gains in
> security.
> 
> The other practical issue is that general-purpose Linux distributions
> cannot prevent kernel downgrades, so even if there's a
> cryptographically signed chain from the firmware to the kernel, you
> can boot last year's kernel, use a root-to-ring-0 exploit to disable
> its particular implementation of lockdown, and then kexec the real
> kernel with lockdown disabled.
> 
> I'm sure that kernel lockdown has applications somewhere, but for
> general-purpose distributions (who usually want to support third-party
> kernel modules), it's an endless source of problems that wouldn't
> exist without it.

I just decided to keep the two separate. The tracing_disable is
permanent (unless you actually do something that writes into kernel
memory to change the variable). When set, there's nothing to clear it.

Thus, I decided not to couple that with lockdown, and let the lockdown
folks do whatever they damn well please ;-)

-- Steve