Re: [PATCH 0/4] keys: Introduce a keys frontend for attestation reports

[Dionna Amalie Glaze]

> For an ephemeral TPM, the EK should be guaranteed to be random and
> therefore non repeating, so there's not much need for the nonce to add
> non-repeatability.  So, in theory, the vTPM/EK binding can be published
> once and relied on even for multiple different tenant endpoints, sort
> of like the EK cert for a physical TPM.
>

Okay that sounds reasonable.

Regarding my other comment about daemons, we might already be in that
state for containers even without the sysfs proposal, given that the
sev-guest device requires root.
We'd need a daemon to provide protected access to the attestation
report (e.g., https://github.com/confidential-containers/attestation-agent)
so that's a bit of a sad situation.

-- 
-Dionna Glaze, PhD (she/her)