Linux EFI development
 help / color / mirror / Atom feed
From: James Bottomley <James.Bottomley@HansenPartnership.com>
To: Jarkko Sakkinen <jarkko@kernel.org>,
	Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Matthew Garrett <matthewg@nvidia.com>,
	keyrings@vger.kernel.org,  linux-integrity@vger.kernel.org,
	rafael@kernel.org, linux-pm@vger.kernel.org,
		linux-efi@vger.kernel.org
Subject: Re: [PATCH 01/17] tpm: Define a kernel-owned TPM NV index that can't be modified by userland
Date: Wed, 18 Nov 2026 11:06:55 -0500	[thread overview]
Message-ID: <ba43d150febc8a85247d6c54f57db0835cb8f5de.camel@HansenPartnership.com> (raw)
In-Reply-To: <avnMl7fD156F1fx4@kernel.org>

[Matthew seems preoccupied, so let me try]
On Sun, 2026-11-15 at 17:15 +0200, Jarkko Sakkinen wrote:
> On Sun, Oct 11, 2026 at 04:28:33AM -0700, Matthew Garrett wrote:
> > On Sun, Oct 11, 2026 at 12:27:58PM +0200, James Bottomley wrote:
> > > On Sun, 2026-10-11 at 03:23 -0700, Matthew Garrett wrote:
> > > > Hmm. Well, that's irritating. Which year's XPS is this?
> > > 
> > > It's an XPS-13 9380 Core i7 ... circa 2019 I think.
> > 
> > Yes, that's definitely later than I hoped. Bother. I think this
> > patchset is probably still worthwhile because it does unblock us on
> > at least some machines, and the implementation is intended to be
> > opaque (other than the change in PCR 5 measurements, but we should
> > still get log replay) and can be replaced with something else if we
> > want. The other approach that springs to mind is tying this to
> > having distros switch to a new kernel signing key before enabling
> > this and so having PCR 7 be associated with the feature, but that
> > sounds like an incredibly tedious thing to get right in a reliable
> > way.
> 
> What prevents blocking PCR 5 access from /dev/tpm*? If firmware
> already, does it, this can only add defense-in-depth.

Nothing does, but what's the concern? A random user writing to PCR5 to
disrupt the validity of the next suspend image? This is already an
existing problem for binding secrets release to any secure boot PCRs.
However, if we start blocking access to secure boot PCRs, there's a
high likelihood of unforeseen consequences (like interfering with
unlock once secrets that do capping writes to PCRs to prevent
subsequent releases).

The better mechanisms to address this are probably:

   1. Detect that PCR5 has been disrupted (get the kernel to store the
      on-boot value and check they match) and if so refuse signed
      suspend to disk with a message (this is easy).
   2. Capture the additional writes to PCR5 and store them as part of
      the attestation so the resuming system can prove the current boot
      value of PCR5 matches the attested one when the additional
      extends are replayed (this is a bit more complex).

Regards,

James

  reply	other threads:[~2026-11-18 16:06 UTC|newest]

Thread overview: 63+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-08 13:20 [RFC] Make hibernation work with lockdown Matthew Garrett
2026-10-08 13:20 ` [PATCH 01/17] tpm: Define a kernel-owned TPM NV index that can't be modified by userland Matthew Garrett
2026-10-08 13:41   ` Matthew Garrett
2026-10-08 16:24     ` Jarkko Sakkinen
2026-10-08 16:23   ` Jarkko Sakkinen
2026-10-09  8:33     ` Matthew Garrett
2026-10-10 21:36       ` Jarkko Sakkinen
2026-10-08 17:06   ` Ilias Apalodimas
2026-10-10  9:22   ` James Bottomley
2026-10-10 21:26     ` Matthew Garrett
2026-10-11  6:04       ` James Bottomley
2026-10-11 10:23         ` Matthew Garrett
2026-10-11 10:27           ` James Bottomley
2026-10-11 11:28             ` Matthew Garrett
2026-11-15 15:15               ` Jarkko Sakkinen
2026-11-18 16:06                 ` James Bottomley [this message]
2026-11-18 16:30                   ` Jarkko Sakkinen
2026-11-18 16:44                     ` Jarkko Sakkinen
2026-11-18 16:45                       ` Jarkko Sakkinen
2026-10-11  8:21       ` James Bottomley
2026-11-13  0:40   ` Jarkko Sakkinen
2026-11-13  0:45     ` Jarkko Sakkinen
2026-11-13  0:50       ` Jarkko Sakkinen
2026-10-08 13:20 ` [PATCH 02/17] efi: Add a mechanism to modify TPM state depending on kernel security features Matthew Garrett
2026-10-08 16:38   ` Jarkko Sakkinen
2026-10-08 13:20 ` [PATCH 03/17] tpm: Allow tpm2_start_auth_session() to start an audit session Matthew Garrett
2026-11-13  2:17   ` Jarkko Sakkinen
2026-10-08 13:20 ` [PATCH 04/17] tpm: Log commands executed in " Matthew Garrett
2026-11-15 19:35   ` Jarkko Sakkinen
2026-10-08 13:20 ` [PATCH 05/17] tpm: Add a kernel attestation key and signed audit digest retrieval Matthew Garrett
2026-10-08 16:45   ` James Bottomley
2026-10-09  8:29     ` Matthew Garrett
2026-10-08 13:20 ` [PATCH 06/17] tpm: Use TPM2_NV_ReadPublic to read NV index names Matthew Garrett
2026-10-08 13:20 ` [PATCH 07/17] tpm: Add in-kernel support for reading NV indices Matthew Garrett
2026-10-08 13:20 ` [PATCH 08/17] tpm: Add NV define, undefine and write helpers Matthew Garrett
2026-10-08 13:20 ` [PATCH 09/17] tpm: Provision the kernel NV index at registration Matthew Garrett
2026-10-08 13:20 ` [PATCH 10/17] tpm: Move the bounds-checked response reader to a header Matthew Garrett
2026-10-08 13:20 ` [PATCH 11/17] tpm: Add kernel signing key creation with audited provenance Matthew Garrett
2026-10-08 17:00   ` James Bottomley
2026-10-09  8:31     ` Matthew Garrett
2026-10-08 13:20 ` [PATCH 12/17] tpm: Add signing with the kernel signing key Matthew Garrett
2026-10-08 13:20 ` [PATCH 13/17] tpm: Add verification of kernel signing key provenance Matthew Garrett
2026-10-10  9:34   ` James Bottomley
2026-10-10 21:18     ` Matthew Garrett
2026-10-10  9:42   ` James Bottomley
2026-10-10 21:20     ` Matthew Garrett
2026-11-13  2:25   ` Jarkko Sakkinen
2026-10-08 13:20 ` [PATCH 14/17] PM: hibernate: Add image digest and signature page infrastructure Matthew Garrett
2026-10-08 13:20 ` [PATCH 15/17] PM: hibernate: Sign and verify images with a kernel-generated TPM key Matthew Garrett
2026-10-08 13:20 ` [PATCH 16/17] PM: hibernate: Refuse to verify images with a virtual TPM Matthew Garrett
2026-10-08 13:20 ` [PATCH 17/17] PM: hibernate: Allow hibernation under lockdown with signed images Matthew Garrett
2026-10-08 16:53   ` Jarkko Sakkinen
2026-10-09  8:31     ` Matthew Garrett
2026-10-10 21:29       ` Jarkko Sakkinen
2026-10-08 15:56 ` [RFC] Make hibernation work with lockdown Jarkko Sakkinen
2026-10-23 15:28 ` Jarkko Sakkinen
2026-10-23 19:13   ` Jarkko Sakkinen
2026-10-31  3:01     ` Jarkko Sakkinen
2026-11-07 22:34 ` Neal Gompa
2026-11-10  1:36 ` Jarkko Sakkinen
2026-11-13  0:06   ` Jarkko Sakkinen
2026-11-15 15:02 ` Jarkko Sakkinen
2026-11-15 15:08   ` Jarkko Sakkinen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ba43d150febc8a85247d6c54f57db0835cb8f5de.camel@HansenPartnership.com \
    --to=james.bottomley@hansenpartnership.com \
    --cc=jarkko@kernel.org \
    --cc=keyrings@vger.kernel.org \
    --cc=linux-efi@vger.kernel.org \
    --cc=linux-integrity@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=matthewg@nvidia.com \
    --cc=mjg59@srcf.ucam.org \
    --cc=rafael@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox