[JJenkx]

What

The standalone write-watch CMake project registers write_watch_execute-fault and write_watch_read-fault twice on current main. Configuration stops before either macOS write-watch job can build. Remove the redundant final loop, preserving one registration for each fault mode and all other tests. No C++ test, production code or timeout changes.

Tested

Pinned main 51c31ea17e9f66cb36fe9655ec31c81610bfcfc9, Linux host with CMake 4.4.4/Ninja. Actual unmodified standalone configure fails with both duplicate-name diagnostics; the fixed configure/generate passes and CTest JSON contains 18 unique tests, including each fault mode exactly once. The same before/after result holds in a real Darwin CMake cross-configuration on Linux, with an injected assertion verifying APPLE=1, CMAKE_SYSTEM_NAME=Darwin and CMAKE_HOST_SYSTEM_NAME=Linux. This is registration/configuration proof, not an Apple SDK build or macOS execution.

A Linux compilation attempt of the macOS-only fixtures failed at their unconditional mach/mach.h includes; it is retained as excluded wrong-host evidence. No runtime pass is claimed. Conventions and whitespace pass. The original #12468 macOS CI failures reproduce this main-source configuration error; its published scanner branch already has only one fault loop and is unchanged.

Configure logs, registration manifests, patch and source proofs, SHA-256 8cece16d735c276255db5e3c6617a50957dc86e743d12ccf894c579fb52dacd8 (download verified).

Checklist

• Based on current main; no other open PR implements the same functions
• One topic per PR; follow-ups go in a new PR
• No comments in code except technical debt
• Unimplemented paths throw (NotImplemented_nid_no_patch); silent stubs are listed under Silent stubs in their library's technical debt file
• No notes, investigation .md files or images added to the repository (attach them to this PR)
• New third-party code is a submodule built from source
• Generic behaviour, not specific to one title
• Depends on: none
• Fixes: none
• Unblocks: none
• GPUs tested: n/a
• AI-assisted: yes (Codex)

[MarcZX10]

Reviewed — looks good

The exact-head diff removes only the duplicate execute-fault/read-fault registration loop; the remaining registrations and their 30-second timeout stay intact.

What I ran

Head 75b97c359a77bc7781a44d3e768765abcfd214eb; PR base de2ba4cecb30ffb1c04d7e7dd693b03eace96b1b. A merge preview against current main 42d009f5f37c8f388b01fbfe38d2e704c792c46f was clean (tree c8dffe07254a8804fa5dca9a1a350ef91ee0d13e). Environment: Linux x86_64, Python 3.11.2. I did not configure, build, or run CTest locally.

python3 tools/check_conventions.py --base de2ba4cecb30ffb1c04d7e7dd693b03eace96b1b
git diff --check de2ba4cecb30ffb1c04d7e7dd693b03eace96b1b...75b97c359a77bc7781a44d3e768765abcfd214eb
git merge-tree --write-tree 42d009f5f37c8f388b01fbfe38d2e704c792c46f 75b97c359a77bc7781a44d3e768765abcfd214eb
sha256sum /tmp/anyps5-13293-evidence.tar.gz

Results

Check	Result
Conventions	Passed.
Diff check	Passed; no whitespace errors.
Current-main merge preview	Clean.
Evidence artifact	Downloaded and SHA-256 matched the PR's stated digest. Its included baseline configure log shows both duplicate-name errors; the fixed configure log succeeds, and its generated CTest manifest contains 18 unique tests with each fault mode once.
Test-quality heuristic	Not executable from the isolated PR checkout because the local-only helper script is absent. Manual CMake inspection confirms Recorder.cpp is registered in agc_driver_recorder_tests; this PR changes only the standalone write-watch CMake file.
Hosted CI	Intake, eligibility, portability and report passed. Windows, conventions and labels were pending at the latest snapshot; Linux and specialized jobs were skipped.

Unedited local check output

All conventions checks passed.
c8dffe07254a8804fa5dca9a1a350ef91ee0d13e
8cece16d735c276255db5e3c6617a50957dc86e743d12ccf894c579fb52dacd8  /tmp/anyps5-13293-evidence.tar.gz

Negative control

The change adds no assertion or test logic, so I did not run a code-mutation control. The author-provided evidence artifact records the baseline configuration failing on duplicate test names and the changed configuration generating all 18 unique registrations.

Findings

Blocking: None.

Non-blocking: The artifact establishes configuration and test registration, not test execution. The PR accurately states that macOS runtime and Apple SDK compilation were not verified; the current macOS CI jobs were still pending.

Sources

• Changed CMake registrations at the reviewed head
• Verified configure logs and generated registration manifests (SHA-256 8cece16d735c276255db5e3c6617a50957dc86e743d12ccf894c579fb52dacd8)

Not verified

No local configure, build, CTest execution, Apple SDK build, or macOS runtime. The cross-configuration and registration evidence is from the verified author-provided artifact; pending hosted checks may add platform evidence.

[adriadam10]

Closing as a duplicate of #13220, which was opened earlier and has the identical diff (removes the same redundant foreach(mode execute-fault read-fault) block in core/libs/tests/write_watch/CMakeLists.txt, same blob a912f98cc1). The change itself is correct; it will be reviewed there. Feel free to reopen if #13220 is dropped.