Overlap Linux E2E with full Windows API qualification - #2878
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
Independent feature review complete at 8b3f443: SHIP, no direct correctness/coverage blocker. Both full unfiltered API suites remain required by the ci-required workflow merge contract, with OS-specific artifacts and separate hosted workspaces. Expected check labels are preserved in the caller/inner names and canonical policy; actual hosted emitted labels and timing still need confirmation. Reviewer ran 9 focused API/E2E/topology tests, 683-file link check and diff whitespace validation. Coordinator ran 511/511 control tests, docs governance, link check and 161/161 Action pin check. No tests were dropped, no new proof-reuse/selection authority introduced. Round 1 complete. Keep parent-first order; exact-head/current-base hosted qualification and three-minute age remain required. Project Priority II/Review synchronized from Refs #2332/#2336. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 8b3f443288
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".
|
Hosted scheduling observation at head8b3f4432880559807ad3a909207b38aafb81a19f: actual checks are still API Integration / API Integration (ubuntu-latest) and API Integration / API Integration (windows-latest). Both started03:00:22UTC; Linux completed successfully03:08:10, E2E started03:08:23 while Windows API remained in_progress. This directly verifies the intended overlap and retained check labels. The run is still pending; no completed-run savings or merge-readiness claim yet. |
|
Completed hosted measurement for #2878 Required CI run34431587587 at head8b3f4432880559807ad3a909207b38aafb81a19f succeeded. All18 measured jobs completed; aggregate runner time4,779seconds; first job start to final completion1,314seconds (21m54s). Exact retained check results: Linux API468s, Windows API1,275s, E2E548s. Linux finished03:08:10UTC; E2E started03:08:23 and finished03:17:31; Windows finished03:21:37. E2E therefore completed before Windows, confirming the removed serialization while all suites passed. Earlier staged original-head run34423790319 used4,651runner-seconds and1,772seconds elapsed (29m32s). This run is7m38s shorter elapsed but128runner-seconds higher. These are different hosted samples, not a controlled causal estimate or a compute-savings claim. The directly proven result is retained coverage/check names and actual platform overlap. Both runs have one successful attempt; source/metadata/configured Node identity is retained in the read-only observation reports. The current PR is fully green on its stacked base. Parent-first retarget/current-base qualification remains before merge. No production evidence reuse or selection activation. |
…32-linux-e2e-prerequisites
|
Final refresh: #2871 merged as2d50256be7be980f7cb8e8230d12ff44093987da after corrected required run34462829280 passed. This PR is retargeted to main and pushed as96e07710f1835bc226fb441f4e9a312c28d3d05a. The scheduling workflow/transformer/regression files are byte-identical to independently reviewed8b3f44328; all additional changes are the reviewed evidence/state documentation. The corrected verifier is already in the base. Current integrated control512 and docs governance/686links pass; complete full required CI is now requalifying this final head. No linked closing issues were introduced by the rewritten body. Current policy remains shadow, all original full suites remain required, and no production reuse is enabled. |
|
Completed the continuation train: #2863, #2864, #2865, #2867, #2868, #2869, #2871 and #2878 are merged. Final merge f667ae1 contains final head96e07710f1835bc226fb441f4e9a312c28d3d05a. Each PR received current-base full required qualification; all eight project items are Done with populated priorities, and the post-merge checkpoint has zero unresolved review threads. Final required run34465851361 passed18jobs in one attempt:5,500runner-seconds and1,872seconds first-job-to-last-completion (31m12s). Linux API completed10:32:55UTC, E2E started10:32:58 and completed10:43:42, Windows API completed10:55:15. Both original full API labels/suites remain, and overlap is directly proven. This sample is slower than the earlier21m54 sample; neither is a causal speedup or compute-savings estimate. Windows duration variability remains a measurement/diagnostic priority, not a presumed flaky test. No cancellation or retry was needed for the final run. Local cumulative control512passed; the corrected portability head's control510 and seven export tests passed; the regenerated standalone kit separately passed306 (overlapping scopes). The owner-authorized linked-entry verifier fix has independent SHIP review. The r2 kit source iscdc6d8e9fa801d0b90d152935e14c540dba6c835;27payloadfiles verified before ZIP creation and after extraction. The old verifier/archive are superseded. Production selection/reuse remains disabled/shadow. Before activation: complete the tracked producer-contract/inventory/exposure gaps, prove protected execution for one whole-task family, extend the existing canonical gate receipt, connect authenticated landed/full-audit events, and qualify failure recall. The current recall report is not ready. The merged operations guide records the sequence and a recommendation to avoid repeatedly qualifying tightly coupled CI-control stacks. SC-4, SC-6 and SC-7 remain in OUTSTANDING_TASKS; no settings/runner/issuer activation is implied. |
Fresh-context review of this PR found I had searched the wrong issue. The continuation train's tracker is CI-12 #2336, not #2337 or #2324, and #2336 records the session as maintainer-requested and, at 02:40:20Z on 2026-09-10, "the maintainer's 2026-09-10 request to finish implementation, merging, testing and CI improvements". So "no delegation is recorded" was wrong. The row now asks whether that request supplied the per-PR review the amendment wants, and infers nothing either way. Three further corrections the same review prompted: The divergence did not start 2026-09-09. J.1 (#2772) and J.2 (#2787) are both 2026-09-08, and 2026-09-09 is the day the rule was honoured, with the four-PR delegation. The pointer file said the opposite. The row omitted the train's quality record, which changes how question (a) reads. Every PR carried a Codex and an independent review with current-base full required qualification. One confirmed HIGH, a trusted verifier bypassed through a Windows directory junction, was refused at 09:44Z, fixed in cdc6d8e at 09:49:14Z and is an ancestor of #2871's 10:23:22Z merge; export.test.mjs passes 7/7 on current main. The gap is procedural, not a shipped known defect. #2878 is added: it is the eighth merge in the same train. #2868's ci-continuation-observe.yml and #2791's nightly observer are different PRs under different trackers, and #2791 was delegated on 2026-09-09. An unqualified ruling about "the observer" would hit the wrong one, so the row says which is which. Also: name J.2 as where a parked control-plane PR is recorded now that SC-10 is closed; stop hard-coding live connector status in a file whose own preamble says live GitHub outranks it, and restore the standing "a usage-limit notice is informational, not a finding" triage line; correct the CF subset sentence, which spans v0.4 as well; bump Last reviewed. Refs #2337, #2336.
Linux E2E previously waited for the complete API operating-system matrix. Split the required API caller into independent Linux and Windows calls so E2E starts after Linux API while full Windows qualification continues. Both complete test commands, artifacts and policy check labels remain; Windows stays unconditional and required before merge. Unknown or unset reusable-workflow platform input runs both operating systems.
The staging transformer recognizes the legacy and split layouts and rejects disabled or misrouted Windows callers. Existing E2E prerequisites remain. The continuation operations, validation, STATUS and master-plan documents now distinguish deployed staging/observation from reference admission and production reuse, record the verified results, and give the next integration sequence. Known admission-contract, inventory-completeness, replay and recall gaps remain tracked on #2336; policy stays shadow and admission has no required-check authority.
Validation:
Refs #2332, #2336. Exact-head hosted qualification against the final main base remains required before merge. No selection, proof reuse, trigger, permission, branch-protection or runner changes. Rollback removes the Windows caller and Linux platform input together, restoring the reusable workflow's complete default matrix.