linux-kernel - Re: [RFC PATCH v2 0/4] tsm: Runtime measurement registers ABI

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]

Message-Id: <6DEAEC08-420C-46A9-8877-EBF60331A931@google.com>
Date: Mon, 19 Aug 2024 14:25:15 -0700
From: Qinkun Bao <qinkun@...gle.com>
To: Samuel Ortiz <sameo@...osinc.com>
Cc: Dan Williams <dan.j.williams@...el.com>,
 Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@...ux.intel.com>,
 "Yao, Jiewen" <jiewen.yao@...el.com>,
 "Xing, Cedric" <cedric.xing@...el.com>,
 biao.lu@...el.com,
 James Bottomley <James.Bottomley@...senPartnership.com>,
 linux-coco@...ts.linux.dev,
 linux-integrity@...r.kernel.org,
 linux-kernel@...r.kernel.org,
 Vishal Annapurve <vannapurve@...gle.com>,
 Chong Cai <chongc@...gle.com>,
 Dan Middleton <dan.middleton@...el.com>,
 Alex Crisan <alex.crisan@...el.com>
Subject: Re: [RFC PATCH v2 0/4] tsm: Runtime measurement registers ABI

A gentle ping on this email thread. We have tested the patch series [1] and will release a product based on the patch series. 

If the patch series can not get upstreamed, the whole confidential computing community can not have a way to measure the workload with RTMRs. Without the patch, RTMR3 is completely unused. The patch works perfectly for our usage case (Like the existing TPM ABI, the raw measurement is taken instead of recording the log entry.). Assuming RTMR serves as an alternative to TPM-based measurement, migrating existing software to the RTMR could be greatly simplified by developing an ABI that resembles the TPM.

I don’t object to having an ABI to take the log entry. For our usage case, we use the Canonical event log [2] to measure the workload. I do think that we should NOT block the patch series for several months to solve an issue that TPM can not solve. 

Link:
[1] https://lore.kernel.org/lkml/20240128212532.2754325-1-sameo@rivosinc.com/
[2] https://trustedcomputinggroup.org/resource/canonical-event-log-format/