Add support for compute runtime quote policy by martintomazic · Pull Request #6474 · oasisprotocol/oasis-core

martintomazic · 2026-03-16T20:24:52Z

Replaces #6471 closes #6387.

netlify · 2026-03-16T20:24:58Z

✅ Deploy Preview for oasisprotocol-oasis-core canceled.

Name	Link
🔨 Latest commit	`c4fb335`
🔍 Latest deploy log	https://app.netlify.com/projects/oasisprotocol-oasis-core/deploys/69ca35c64786c8000890123b

martintomazic · 2026-03-17T14:49:52Z

.buildkite/code.pipeline.yml

+      - .buildkite/scripts/test_e2e.sh --timeout 20m
+        --scenario e2e/runtime/runtime-encryption
+        --scenario e2e/runtime/compute-policy
+


e2e test with mocked TEE was instrumental to realizing that without compute_policy on the rust side of the protocol, RHP would fail due to the missing field.

To test that compute policy is applied e2e you can disable pcs in the quote policy (fixture) and run:

export OASIS_UNSAFE_ALLOW_DEBUG_ENCLAVES=1 export OASIS_UNSAFE_MOCK_TEE=1 export OASIS_TEE_HARDWARE=intel-sgx make .buildkite/scripts/test_e2e.sh --scenario e2e/runtime/compute-policy

The test would gets stuck (as expected).

Full SGX test would be desired here (easy). It would be also nice to add another one, where we e.g. set MinTCBEvaluationDataNumber and verify via the control status that the registration failed... Still not optimal as technically we would catch node local attestation to halt early and not consensus rejecting byzantine compute.

Testing that rofl nodes pass on the lower TCB but compute node is rejected would be even harder as we would need two different SGX environments to make it work.

What do you think?

Yes, it would be great if we can test it in real SGX.

For the two different environments, let's open a separate issue to test that E2E.

martintomazic · 2026-03-18T09:03:17Z

go/oasis-node/cmd/node/node.go

 		n.dataDir,
 		n.chainContext,
 		n.Identity,
+		willRegisterComputeRuntime,


Previously I was passing runtime roles (see). Now we only need bool flag.

No matter what name I choose here, passing the variable through workers, feels like abstraction leak.

I could avoid all this and use the config.GlobalConfig directly in the common.GetQuotePolicy. I hate global state though, but here might be acceptable with a TODO issue given this shows abstractions are not best in the first place?

Yes, this seems awkward. For example the common worker constructor is already doing config lookups so it seems fine to have them there.

Removing the use of global config and explicitly passing it instead is a different issue.

martintomazic · 2026-03-18T09:15:04Z

runtime/src/consensus/registry.rs

+        /// The compute runtime quote policy.
+        #[cbor(optional)]
+        compute_policy: sgx::QuotePolicy,


This field is ignored for now and only here to pass e2e test.

In practice we probably need new RHP method (similar to node identity), to signal whether compute policy should be respected or not. Technically not breaking?

The runtime should sign with its RAK whether the intention of this enclave is to be used as the one registering on the consensus (having access to keymanager) or is the intention to be used in the less constrained environments (default policy only). Similar solution to what is done for the node_id, i.e. make it part of the hash structure and sign with RAK and validate signature on the consensus side.
Do we really need this / what are possible attack vectors without it?

I would avoid this extra verification for now as I don't see what attack this would prevent. It is other runtimes and the consensus layer that are enforcing this policy.

codecov · 2026-03-18T09:38:05Z

Codecov Report

❌ Patch coverage is 88.98305% with 13 lines in your changes missing coverage. Please review.
✅ Project coverage is 64.43%. Comparing base (76002fa) to head (b795a2e).
⚠️ Report is 11 commits behind head on master.

Files with missing lines	Patch %	Lines
go/common/node/node.go	78.57%	0 Missing and 3 partials ⚠️
go/common/node/sgx.go	91.66%	1 Missing and 1 partial ⚠️
go/common/sgx/pcs/pcs.go	0.00%	2 Missing ⚠️
go/oasis-node/cmd/node/node.go	75.00%	1 Missing and 1 partial ⚠️
go/worker/common/committee/node.go	92.30%	0 Missing and 2 partials ⚠️
go/consensus/cometbft/apps/scheduler/scheduler.go	88.88%	0 Missing and 1 partial ⚠️
go/worker/keymanager/worker.go	50.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #6474      +/-   ##
==========================================
- Coverage   64.79%   64.43%   -0.36%     
==========================================
  Files         699      699              
  Lines       68206    68264      +58     
==========================================
- Hits        44193    43988     -205     
- Misses      18999    19205     +206     
- Partials     5014     5071      +57

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

go/common/node/node.go

go/common/node/sgx.go

kostko · 2026-03-24T05:01:51Z

go/oasis-node/cmd/node/node.go

 		n.dataDir,
 		n.chainContext,
 		n.Identity,
+		willRegisterComputeRuntime,


Yes, this seems awkward. For example the common worker constructor is already doing config lookups so it seems fine to have them there.

Removing the use of global config and explicitly passing it instead is a different issue.

kostko · 2026-03-24T08:55:58Z

runtime/src/consensus/registry.rs

+        /// The compute runtime quote policy.
+        #[cbor(optional)]
+        compute_policy: sgx::QuotePolicy,


I would avoid this extra verification for now as I don't see what attack this would prevent. It is other runtimes and the consensus layer that are enforcing this policy.

See if we can trigger real quote verification. This will still get rejected at local verification instead of consensus.

martintomazic force-pushed the martin/feature/compute-policy branch from d0c991e to 859e7a8 Compare March 17, 2026 14:35

martintomazic commented Mar 17, 2026

View reviewed changes

martintomazic force-pushed the martin/feature/compute-policy branch from 859e7a8 to b795a2e Compare March 18, 2026 08:55

martintomazic commented Mar 18, 2026

View reviewed changes

martintomazic marked this pull request as ready for review March 18, 2026 09:40

martintomazic requested review from kostko, peternose and ptrus as code owners March 18, 2026 09:40

martintomazic mentioned this pull request Mar 18, 2026

Per role quote policies #6471

Closed

kostko reviewed Mar 24, 2026

View reviewed changes

This was referenced Mar 24, 2026

CI: Add additional runner with lower SGX TCB for e2e tests #6478

Open

go/worker/registration: Clean dead registration signer code #6481

Open

martintomazic added 6 commits March 30, 2026 10:09

go: Use CapabilityTEEVerifyParams instead of long param list

670f837

go/common/node: Add compute runtime policy to SgxConstraints

a31543e

go: Validate attestation against compute policy before registration

7c8bd59

go/common: Add compute runtime policy sanity test

800fd92

go/oasis-test-runner: Add e2e test for compute runtime quote policy

15d297e

e2e: Enxure compute policy rejects registation (POC)

c4fb335

See if we can trigger real quote verification. This will still get rejected at local verification instead of consensus.

martintomazic force-pushed the martin/feature/compute-policy branch from b795a2e to c4fb335 Compare March 30, 2026 08:35

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add support for compute runtime quote policy#6474

Add support for compute runtime quote policy#6474
martintomazic wants to merge 6 commits intomasterfrom
martin/feature/compute-policy

martintomazic commented Mar 16, 2026 •

edited

Loading

Uh oh!

netlify bot commented Mar 16, 2026 •

edited

Loading

Uh oh!

martintomazic Mar 17, 2026 •

edited

Loading

Uh oh!

kostko Mar 24, 2026 •

edited

Loading

Uh oh!

martintomazic Mar 18, 2026 •

edited

Loading

Uh oh!

kostko Mar 24, 2026

Uh oh!

martintomazic Mar 18, 2026

Uh oh!

kostko Mar 24, 2026

Uh oh!

codecov bot commented Mar 18, 2026 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

kostko Mar 24, 2026

Uh oh!

kostko Mar 24, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

martintomazic commented Mar 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

netlify bot commented Mar 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ Deploy Preview for oasisprotocol-oasis-core canceled.

Uh oh!

martintomazic Mar 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kostko Mar 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

martintomazic Mar 18, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kostko Mar 24, 2026

Choose a reason for hiding this comment

Uh oh!

martintomazic Mar 18, 2026

Choose a reason for hiding this comment

Uh oh!

kostko Mar 24, 2026

Choose a reason for hiding this comment

Uh oh!

codecov bot commented Mar 18, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

Uh oh!

kostko Mar 24, 2026

Choose a reason for hiding this comment

Uh oh!

kostko Mar 24, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

martintomazic commented Mar 16, 2026 •

edited

Loading

netlify bot commented Mar 16, 2026 •

edited

Loading

martintomazic Mar 17, 2026 •

edited

Loading

kostko Mar 24, 2026 •

edited

Loading

martintomazic Mar 18, 2026 •

edited

Loading

codecov bot commented Mar 18, 2026 •

edited

Loading