Skip to content

Rebase trustee v0.17#160

Merged
lmilleri merged 247 commits intoopenshift:mainfrom
lmilleri:try-merge
Feb 9, 2026
Merged

Rebase trustee v0.17#160
lmilleri merged 247 commits intoopenshift:mainfrom
lmilleri:try-merge

Conversation

@lmilleri
Copy link

@lmilleri lmilleri commented Jan 21, 2026

List of changes:

  • rebase to trustee v0.17
  • revert rust version hack (ubi9 builder did not have the same upstream rust version so we had to downgrade, now with ubi10 we're able to get rid of the workaround)
  • revert downstream amd caching certificate changes (adopt upstream solution)
  • UBI 10 container image
  • fix hermetic build
  • use trusted konflux digests

dependabot bot and others added 30 commits September 22, 2025 08:28
Bumps [clap](https://github.com/clap-rs/clap) from 4.5.42 to 4.5.48.
- [Release notes](https://github.com/clap-rs/clap/releases)
- [Changelog](https://github.com/clap-rs/clap/blob/master/CHANGELOG.md)
- [Commits](clap-rs/clap@clap_complete-v4.5.42...clap_complete-v4.5.48)

---
updated-dependencies:
- dependency-name: clap
  dependency-version: 4.5.48
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Tyler Fanelli <tfanelli@redhat.com>
With the move to ITA v2 appraisal API that supports composite
evidence, the token claim paths got changed. The claims we
used previously are now under 'tdx'.

Signed-off-by: Mikko Ylinen <mikko.ylinen@intel.com>
Update the rust-ear dependency to latest version.

Signed-off-by: Tyler Fanelli <tfanelli@redhat.com>
CompositeEvidence, RuntimeData, and InitData were all being deserialized
with each KBS attestation request. Therefore, they were further
integrated into the KBS protocol and added to kbs-types. Update
kbs-types and use its definitions of these structs.

Signed-off-by: Tyler Fanelli <tfanelli@redhat.com>
kbs-types 0.14.0 included changes in which the Azure vTPM strings were
hyphenated.

Signed-off-by: Tyler Fanelli <tfanelli@redhat.com>
kbs-types 0.14.0 included changes in which the Azure vTPM strings were
hyphenated.

Signed-off-by: Tyler Fanelli <tfanelli@redhat.com>
updated guest-components to a revision that is using the same
kbs-types version.

Signed-off-by: Magnus Kulke <magnuskulke@microsoft.com>
Signed-off-by: Tyler Fanelli <tfanelli@redhat.com>
Signed-off-by: Tyler Fanelli <tfanelli@redhat.com>
Signed-off-by: Tyler Fanelli <tfanelli@redhat.com>
Signed-off-by: Tyler Fanelli <tfanelli@redhat.com>
Currently the docker compose deployment uses non-secure token keys. As
more people use the docker compose deployment with real use cases
(rather than development, as it was first intended), let's enable secure
tokens.

We may want to rework the setup container and move this code into a
script, but for now let's just add the logic here.

Signed-off-by: Tobin Feldman-Fitzthum <tfeldmanfitz@nvidia.com>
The IBM Z docs seem to have be re-organised so update the links
to try and point to the new versions

Signed-off-by: stevenhorsman <steven@uk.ibm.com>
Remove the `set_default` calls when building KbsConfig from a file.

Add the `#[serde(default)]` to KbsConfig to leverage the defaults already
implemented for the children structs.

Signed-off-by: Camilla Conte <cconte@redhat.com>
Implements two commands:
- Run, to launch the API Server in the foreground.
- Keygen, to generate a new admin key pair.

Signed-off-by: Camilla Conte <cconte@redhat.com>
Co-authored-by: Claude <noreply@anthropic.com>
Signed-off-by: Camilla Conte <cconte@redhat.com>
Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Similar to our extractors, the verifiers can now optionally take a
configuration file.

This config is not used to enable/disable the verifiers. This is done by
the rust features. Instead, this is to specify any configuration
information that might be needed by a verifier.

Currently only the nvidia verifier has any such configuration (which
doesn't do anything yet). We probably should use this for the CCA
verifier, which has its own separate config file. This will be left to
another PR.

Signed-off-by: Tobin Feldman-Fitzthum <tfeldmanfitz@nvidia.com>
Use remote NRAS service to validate NVIDIA devices.

By default we will continue to use the local verifier because users are
expected to enter a licensing agreement with NVIDIA to use NRAS. Once
they've done that, they can switch to the remote verifier, which uses
NRAS.

Signed-off-by: Tobin Feldman-Fitzthum <tfeldmanfitz@nvidia.com>
The remote NVIDIA verifier handles the RIMs for the user and does not
expose the raw measurements as TCB Claims.

Instead, there are several claims relating to the attestation process,
such as whether the cert chain is valid, whether a rim could be found
for the driver version, etc.

NRAS does provide an overall boolean attestation result, but rather than
using this, let's use the other claims to calculate the AR4SI vector.

No reference values are required to use this policy, but you can specify
allowed driver and vbios versions,which will be checked.

Signed-off-by: Tobin Feldman-Fitzthum <tfeldmanfitz@nvidia.com>
Add a readme desribing some of the differences between the local and
remote verifier and linking to the EULA.

Signed-off-by: Tobin Feldman-Fitzthum <tfeldmanfitz@nvidia.com>
It expects trusted AK public keys to be pre-configured to verify
the quote sent from the client

Example configuration:
```json
{
  "verifier_config": {
    "tpm_verifier": {
      "trusted_ak_keys_dir": "/etc/tpm/trusted_ak_keys",
      "max_trusted_ak_keys": 100
    }
  }
}
```

Assisted by: AI in coming up with a lazy iterator for optimally reading
the public keys from a directory

Signed-off-by: Pradipta Banerjee <pradipta.banerjee@gmail.com>
Signed-off-by: Alice Frosi <afrosi@redhat.com>
Sets up TPM using swtpm emulator to run the
tests

You'll need root privileges to run this test.
This is because setup-swtpm needs access to statedir /var/lib/swtpm-localca
that is owned by swtpm:root.

Likewise the tpm device created is owned by tss:root.

make install-swtpm-dependencies setup-swtpm
export TEE=tpm
sudo -E TEE=$TEE make test-bgcheck

Signed-off-by: Pradipta Banerjee <pradipta.banerjee@gmail.com>
Ignore the error for now. Add a FIXME note

Signed-off-by: Pradipta Banerjee <pradipta.banerjee@gmail.com>
Builds fail with the error
"Can't locate Time/Piece.pm in @inc"

Signed-off-by: Pradipta Banerjee <pradipta.banerjee@gmail.com>
The tdx v6 instances are available in the CoCo subscription now.

Signed-off-by: Magnus Kulke <magnuskulke@microsoft.com>
Bumps [env_logger](https://github.com/rust-cli/env_logger) from 0.10.2 to 0.11.8.
- [Release notes](https://github.com/rust-cli/env_logger/releases)
- [Changelog](https://github.com/rust-cli/env_logger/blob/main/CHANGELOG.md)
- [Commits](rust-cli/env_logger@v0.10.2...v0.11.8)

---
updated-dependencies:
- dependency-name: env_logger
  dependency-version: 0.11.8
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Make sure that the admin endpoints cannot be accessed when using the
wrong admin key.

Signed-off-by: Tobin Feldman-Fitzthum <tfeldmanfitz@nvidia.com>
AR4SI trust vector names are using `-`. However, rego policy does not
support variable names like `file-system`, thus we can only use
`file_system` as variable names.

This patch uses a map "result" to use the `file-system` like member
names to align with the ar4si standard.

Note that regorus does not support eval rule like
```
data.policy.result["file-system"]
```

thus we only eval rule `data.policy.result` and process it in the code
logic.

Fixes confidential-containers#972

Signed-off-by: Xynnn007 <xynnn@linux.alibaba.com>
@lmilleri lmilleri force-pushed the try-merge branch 3 times, most recently from 0cfd174 to 74d89e8 Compare January 30, 2026 16:48
@lmilleri
Copy link
Author

lmilleri commented Feb 2, 2026

/retest

@lmilleri lmilleri force-pushed the try-merge branch 3 times, most recently from 09ff9b3 to 20d03ce Compare February 2, 2026 14:19
@red-hat-konflux
Copy link

Caution

There are some errors in your PipelineRun template.

PipelineRun Error
trustee-pull-request.yaml yaml validation error: line 437: mapping values are not allowed in this context

@lmilleri lmilleri force-pushed the try-merge branch 4 times, most recently from da1b239 to 4872249 Compare February 2, 2026 15:29
@lmilleri lmilleri marked this pull request as ready for review February 2, 2026 15:57
@openshift-ci openshift-ci bot removed the do-not-merge/work-in-progress Indicates that a PR should not merge because it is a work in progress. label Feb 2, 2026
@lmilleri lmilleri requested a review from dbkreling February 2, 2026 15:57
@lmilleri lmilleri changed the title Try merge Rebase trustee v0.17 Feb 2, 2026
}

fn get_sa_list(sa_list: &[c_char; 450]) -> Value {
fn get_sa_list(sa_list: &[c_char; 320]) -> Value {
Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@mythi please check if this change is correct

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@lmilleri yes that is ok for the old DCAP version

Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Workaround to fix the build due to openssl incompatibility

Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Signed-off-by: Leonardo Milleri <lmilleri@redhat.com>
Copy link

@dbkreling dbkreling left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I also added a comment about commenting on three revert commits that are in this PR, for the case there's a need to clarify their reasons for the future.
But that's nothing I see that prevents my approval.
I will trust the dependabot changes are working OK, supported by the test results on this PR. Approving.


# Package UBI image.
FROM registry.access.redhat.com/ubi9
FROM registry.access.redhat.com/ubi10:10.1-1767602397

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

These versions although providing more control, require more attention in case the image is affected by a CVE, an unavoidable trade-off. Just pointing out to highlight our approach.

@lmilleri lmilleri merged commit 4485761 into openshift:main Feb 9, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Comments