linux-kernel - Re: [PATCH v2 6/7] tpm: expose spaces via a device link /dev/tpms<n>

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: Windows password security audit tool. GUI, reports in PDF.

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

Date:   Fri, 24 Feb 2017 13:52:01 -0700
From:   Jason Gunthorpe <jgunthorpe@...idianresearch.com>
To:     James Bottomley <James.Bottomley@...senPartnership.com>
Cc:     Jarkko Sakkinen <jarkko.sakkinen@...ux.intel.com>,
        tpmdd-devel@...ts.sourceforge.net,
        linux-security-module@...r.kernel.org, dhowells@...hat.com,
        Peter Huewe <peterhuewe@....de>,
        Marcel Selhorst <tpmdd@...horst.net>,
        open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 6/7] tpm: expose spaces via a device link /dev/tpms<n>

On Fri, Feb 24, 2017 at 03:29:15PM -0500, James Bottomley wrote:
> On Fri, 2017-02-24 at 11:11 -0700, Jason Gunthorpe wrote:
> > On Fri, Feb 24, 2017 at 07:39:22PM +0200, Jarkko Sakkinen wrote:
> > 
> > > > I think therefore that tpmns<n> for TPM Namespace would be very
> > > > appropriate.
> > > 
> > > Makes sense. We can go with tpmns.
> > 
> > When we have talked about TPM namespaces in the past it has been
> > around the idea of restricting which TPMs the namespace has access 
> > too and changing the 'kernel tpm' for that namespace.
> 
> Well, you know, nothing in the TPM Space code prevents us from exposing
> the namespace so that it could be shared.  However, I think the
> namespace follows connect (device open) paradigm is pretty much the
> behaviour everyone (including the kernel) wants, mostly because TPM2
> has such a tiny amount of resources that you're always dealing with
> loadable keys meaning you don't really want to see anyone else's
> volatile state.

I'm not arguing with that use model, I am asking what do you want to
call the future feature that restricts which TPMs a process can view
if you want to use the word namespace for the resource manager?

This is something Stephen B has been exploring in conjunction with
vtpm. (eg restrict a container to only use a single vtpm and ban it
from the system tpm)

Jason