[dborgards]

What does this change?

The SDO block-upload server regressions used 300 ms negative windows to guess when an old batch had stopped, while RawCan echo deadlines still ran on host time. Retain the existing ordered-send Task on its session and expose a narrow internal actor-only snapshot, so the tests can await the original batch even after that session ends. Both fixtures now use a separate VirtualClock confirmation actor with real RawCan echo matching.

Session-end cases wait for the actual batch completion before asserting no further segments. The partial-ACK case captures the distinct retransmission task, positively drains the remaining segments, awaits both batches and verifies the complete sequence 1–20. Abort, index, payload/count and ordering assertions remain. The timeout row uses a 500 ms configured server timeout and steps to 750 ms, before RawCan's unchanged one-second deadline, explicitly checking SdoProtocolTimedOut. No public API or protocol behavior changes; the production additions only retain/snapshot an existing Task.

Refs #362. Final issue inventory remains to be checked after merge; this PR does not close the issue automatically.

Type of change

• docs / test / refactor / chore / ci — no release

Validation

Six focused theory cases pass. Three actual source mutation configurations were runtime-red and restored: restore runtime RawCan binding (all six fail), ignore ended-session identity, and ignore superseding batch generation. Runner: /private/tmp/cankit-362-server-batch-mutations.py.

Full local gates passed: 3043 net10 tests with coverage, net48 compilation, CI build, formatting, nine verified packages, 50 engineering tests and traceability. All Cobertura classes/nested types were inspected. Both changed executable entries are covered, including both getter branches (100%); no changed missing/partials. Public API baseline is unchanged.

Checklist

• Build with -p:CI=true, test (net10.0), dotnet format --verify-no-changes, and pack + eng/verify-packages.py all pass locally
• New behaviour is covered by a test (six existing regressions with positive task/timer witnesses)
• The requirement or ADR this relates to is referenced (Tests: wall-clock-gated tests after #92 — the 2026-10-21 inventory #362, FR-CO-004)

[cursor]

PR Summary

Low Risk
Test-only harness plus internal task retention; no change to CANopen protocol handling or public surface.

Overview
Hardens SDO block-upload server regression tests by replacing flaky wall-clock sleeps with deterministic witnesses of ordered-send batch completion.

Production code only retains the existing SendOrderedControlFrames task on SdoBlockServerSession.UploadBatch and adds an internal SdoBlockUploadBatchForTests snapshot so tests can capture and await the in-flight batch (and a distinct retransmission batch after a partial ACK) even after the session slot is cleared. No wire or public API behavior changes.

Both affected theory fixtures now use VirtualClock + ClockedCanBusService for RawCan echo timing. Session-end cases await firstBatch before asserting no segments 2–20; the partial-ACK path drains echoes and asserts a full 1–20 sequence once. The timeout row uses a 500 ms SdoServerTimeout and asserts SdoProtocolTimedOut on the final abort.

^{Reviewed by Cursor Bugbot for commit be41d48. Bugbot is set up for automated code reviews on this repo. Configure here.}

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-27T06:55:53.710474Z	be41d48	Manual request

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[dborgards]

@codex review

Please review the current head be41d48. The PR-opened review summary has remained Running since 06:30 UTC with no submitted review or findings. CI, CodeQL, Codecov and both final CI coverage reports are green; this request is to obtain the required completed head review before the quiet period and merge.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Breezy!

Reviewed commit: be41d48b73

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".