linux-kernel - [PATCH v7 0/3] Optimize tpm2_read

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-next>] [day] [month] [year] [list]

Message-Id: <20251208050620.339408-1-jarkko@kernel.org>
Date: Mon,  8 Dec 2025 07:06:16 +0200
From: Jarkko Sakkinen <jarkko@...nel.org>
To: linux-integrity@...r.kernel.org
Cc: tpm2@...ts.linux.dev,
	Jarkko Sakkinen <jarkko@...nel.org>,
	James Bottomley <James.Bottomley@...senPartnership.com>,
	Mimi Zohar <zohar@...ux.ibm.com>,
	David Howells <dhowells@...hat.com>,
	Paul Moore <paul@...l-moore.com>,
	James Morris <jmorris@...ei.org>,
	"Serge E. Hallyn" <serge@...lyn.com>,
	keyrings@...r.kernel.org (open list:KEYS-TRUSTED),
	linux-security-module@...r.kernel.org (open list:SECURITY SUBSYSTEM),
	linux-kernel@...r.kernel.org (open list)
Subject: [PATCH v7 0/3] Optimize tpm2_read_public() calls

The main goal is fairly straight-forwrd here.

The aim of these patches is optimize the number of tpm2_read_public() calls
to the bare minimum. 

## About dropping 'parentName' attribute for ASN.1 keys from the patch set

I wrote this section as a remainder as I have facts fresh in my mind so 
that I can return them as soon as there is working group for the ASN.1 
specification. We really need to have this in the spec.

I dropped [1] given that [2] is landing shortly to IETF draft process,
according to James Bottomley [3]. We will return to [1] as soon as draft
process is open for comments. Still, that attribute is super important,
and here is why.

This will cause a overhead as tpm2_unseal_trusted needs to do an
unnecessary (from pure technical perspective) TPM2_ReadPublic command to
acquire TPM name of the parent. This is obviously known at the time of
creation of a key but the information is not stored anywhere by the
key format.

It also aligns badly with TCG specifications as Table 6 of architecture
spec explicitly defines a reference (or name) for transient keys,
persistent keys and NV indexes to be TPM_ALG_ID concatenated together
with the hash of TPMT_PUBLIC. I.e. the file format is using exactly
the opposite what should be use as reference for keys than what it 
should use.

Other benefits are of course auto-discovery of parent for a key file,
which is nasty to do without the name pre-stored.

[1] https://lore.kernel.org/linux-integrity/20251207173210.93765-3-jarkko@kernel.org/
[2] https://www.hansenpartnership.com/draft-bottomley-tpm2-keys.txt
[3] https://lore.kernel.org/linux-integrity/89d90617ba9b7a5eff1d5fad6bb9773033d3c18c.camel@HansenPartnership.com/

Jarkko Sakkinen (3):
  tpm2-sessions: Define TPM2_NAME_MAX_SIZE
  KEYS: trusted: Re-orchestrate tpm2_read_public() calls
  tpm2-sessions: Remove AUTH_MAX_NAMES

 drivers/char/tpm/tpm2-cmd.c               |   3 +-
 drivers/char/tpm/tpm2-sessions.c          | 116 ++++++--------------
 include/linux/tpm.h                       |  19 +++-
 security/keys/trusted-keys/trusted_tpm2.c | 124 ++++++++++++++--------
 4 files changed, 133 insertions(+), 129 deletions(-)

-- 
2.39.5