morluto/rea

Fork 0

mirror of https://github.com/morluto/rea.git synced 2026-11-05 07:47:29 +00:00

[PR #7048] [MERGED] test(firmware): keep producer fixtures within supported Node limits #7603

New issue

Closed

opened 2026-11-03 13:39:03 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-11-03 13:39:03 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/morluto/rea/pull/7048
Author: @Plastic-time
Created: 10/24/2026
Status: ✅ Merged
Merged: 10/28/2026
Merged by: @morluto

Base: main ← Head: codex/fix-firmware-fixture-node-compat-20261024

📝 Commits (2)

da9e3cd test(firmware): bound fixture hashing and avoid Node startup warnings
15c9c97 test(firmware): apply the Wasm flag only on Node 22

📊 Changes

2 files changed (+22 additions, -3 deletions)

View changed files

📝 tests/fixtures/firmware/producer.mjs (+19 -3)
📝 tests/fixtures/firmware/provider.ts (+3 -0)

📄 Description

Summary

Keep the synthetic firmware producers reliable on the supported Node 22/24/26 lines without changing production resource limits or weakening test assertions.

Related to #5672: these failures were found while rerunning its cancellation fix after @CassieFox23 requested fresh verification. The recorded findings reproduced on unchanged upstream. This PR is independently based on 238914974; #5672 is the discovery context, not a dependency.

Problem and expected behavior

Node 22.23.4 prepends a V8 flag-conflict warning when the fixture starts with --jitless, breaking the two exact missing-report stderr assertions.
Node 26.12.1 intermittently aborts the oversized-child producer with Worklist::Segment::Create out-of-memory under the existing 1 GiB address-space limit, before the intended publication/continuation assertions run.

The synthetic producer should finish within the real limiter and emit its intended diagnostics.

Change and scope

Pass --no-expose-wasm only on Node 22, where it prevents the startup warning. Newer tested versions do not accept this option.
Hash the actual oversized file with one reusable 64 KiB buffer and a file handle closed in finally, reducing chunk-allocation pressure.
Preserve the 128 MiB + 1 byte child, actual digest, existing assertions, and production limits. Only two fixture files change.

Contribution scope: not applicable — ordinary test-fixture maintenance within the existing boundary.

Contract and boundary impact

Owner: synthetic firmware producer launch and report generation.
CLI/MCP, provider contracts, evidence semantics, generated metadata and package behavior: unchanged.
Test effects: launch the existing synthetic Node processes; read and write disposable fixture files. No real firmware engine or target execution, new network access, or containment relaxation.

Evidence and regression coverage

Linux matrix, exact candidate 15c9c975:

Node	Six existing boundary/MCP files
22.23.4	66 passed
24.22.1	66 passed
26.12.1	66 passed

Unchanged 238914974 reproduces the two Node 22 failures. The Node 26 failure is intermittent: its suite passed in this matrix, but two of three standalone producer repeats aborted; the earlier baseline suite also failed the oversized-child case.

A same-runner Node 26 comparison then ran each revision 20 times under the unchanged 1 GiB limit: baseline 14 aborts/20, candidate 0/20. Every successful output was independently checked with Python for size 134,217,729 and SHA-256 matching its report. This is a bounded sample, not a failure-rate estimate.

For completed standalone Node 26 runs on that runner, peak RSS was 64,824–65,036 KiB on baseline versus 49,852–50,272 KiB on the candidate; elapsed time was 0.12–0.13 s versus 0.14–0.15 s, and user+system CPU was 0.15–0.16 s versus 0.13–0.14 s. These are /usr/bin/time -v producer-process measurements, not full extraction benchmarks or cold/warm cache comparisons. Agent-token costs were not measured.

Validation performed

npm run test:focused -- the five firmware boundary files plus tests/boundary/mcp/firmwareMcp.test.ts — 66 passed on each Node version.
npm run check on Node 24/Linux — passed, including typecheck, built lint, formatting, knip, metadata and architecture guards; lint reported 0 errors and 2744 warnings.
git diff --check — passed.

Compatibility, safety, and release

No production or release changes. The temporary verification workflows stay on separate fork branches. Real Unblob/Binwalk and other operating systems were not exercised by this change; the MCP test uses the synthetic producer. Full-suite coverage remains with upstream CI.

Author readiness

Human owner: @Plastic-time

I reviewed the final diff and relevant existing code, understand the change and its edge cases, and take responsibility for the complete PR, including AI-assisted work.
I checked the evidence above, confirmed it supports the current change, and recorded meaningful validation gaps.
I declare this PR ready for initial maintainer review.

Before merge

This PR addresses a concrete problem or an agreed enhancement.
Material scope and authority questions are resolved, and any required maintainer scope decision covers the final diff.
The PR has one focused outcome and the title follows type(scope): outcome.
Related issue is linked, or the reason for not linking one is stated above.
Tests cover changed observable behavior and meaningful failure paths.
Owning docs, contracts, and generated metadata are updated where needed.
User-visible CLI/MCP changes include representative output.
I checked the final diff for secrets, unrelated cleanup, and unsupported claims.

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/morluto/rea/pull/7048 **Author:** [@Plastic-time](https://github.com/Plastic-time) **Created:** 10/24/2026 **Status:** ✅ Merged **Merged:** 10/28/2026 **Merged by:** [@morluto](https://github.com/morluto) **Base:** `main` ← **Head:** `codex/fix-firmware-fixture-node-compat-20261024` --- ### 📝 Commits (2) - [`da9e3cd`](https://github.com/morluto/rea/commit/da9e3cda08bedccad0007c77ab572702376cd472) test(firmware): bound fixture hashing and avoid Node startup warnings - [`15c9c97`](https://github.com/morluto/rea/commit/15c9c975f04b7ebe55ea8caf4b8214fd017b21c5) test(firmware): apply the Wasm flag only on Node 22 ### 📊 Changes **2 files changed** (+22 additions, -3 deletions) <details> <summary>View changed files</summary> 📝 `tests/fixtures/firmware/producer.mjs` (+19 -3) 📝 `tests/fixtures/firmware/provider.ts` (+3 -0) </details> ### 📄 Description ## Summary Keep the synthetic firmware producers reliable on the supported Node 22/24/26 lines without changing production resource limits or weakening test assertions. Related to #5672: these failures were found while rerunning its cancellation fix after @CassieFox23 requested fresh verification. The [recorded findings](https://github.com/morluto/rea/pull/5672#issuecomment-6390314319) reproduced on unchanged upstream. This PR is independently based on `238914974`; #5672 is the discovery context, not a dependency. ## Problem and expected behavior - Node 22.23.4 prepends a V8 flag-conflict warning when the fixture starts with `--jitless`, breaking the two exact missing-report stderr assertions. - Node 26.12.1 intermittently aborts the oversized-child producer with `Worklist::Segment::Create` out-of-memory under the existing 1 GiB address-space limit, before the intended publication/continuation assertions run. The synthetic producer should finish within the real limiter and emit its intended diagnostics. ## Change and scope - Pass `--no-expose-wasm` only on Node 22, where it prevents the startup warning. Newer tested versions do not accept this option. - Hash the actual oversized file with one reusable 64 KiB buffer and a file handle closed in `finally`, reducing chunk-allocation pressure. - Preserve the 128 MiB + 1 byte child, actual digest, existing assertions, and production limits. Only two fixture files change. Contribution scope: not applicable — ordinary test-fixture maintenance within the existing boundary. ## Contract and boundary impact - Owner: synthetic firmware producer launch and report generation. - CLI/MCP, provider contracts, evidence semantics, generated metadata and package behavior: unchanged. - Test effects: launch the existing synthetic Node processes; read and write disposable fixture files. No real firmware engine or target execution, new network access, or containment relaxation. ## Evidence and regression coverage [Linux matrix](https://github.com/Plastic-time/rea/actions/runs/40139521896), exact candidate `15c9c975`: | Node | Six existing boundary/MCP files | | --- | --- | | 22.23.4 | 66 passed | | 24.22.1 | 66 passed | | 26.12.1 | 66 passed | Unchanged `238914974` reproduces the two Node 22 failures. The Node 26 failure is intermittent: its suite passed in this matrix, but two of three standalone producer repeats aborted; the [earlier baseline suite](https://github.com/Plastic-time/rea/actions/runs/40139359203/job/121007608422) also failed the oversized-child case. A [same-runner Node 26 comparison](https://github.com/Plastic-time/rea/actions/runs/40139666282) then ran each revision 20 times under the unchanged 1 GiB limit: baseline **14 aborts/20**, candidate **0/20**. Every successful output was independently checked with Python for size 134,217,729 and SHA-256 matching its report. This is a bounded sample, not a failure-rate estimate. For completed standalone Node 26 runs on that runner, peak RSS was 64,824–65,036 KiB on baseline versus 49,852–50,272 KiB on the candidate; elapsed time was 0.12–0.13 s versus 0.14–0.15 s, and user+system CPU was 0.15–0.16 s versus 0.13–0.14 s. These are `/usr/bin/time -v` producer-process measurements, not full extraction benchmarks or cold/warm cache comparisons. Agent-token costs were not measured. ## Validation performed - `npm run test:focused --` the five firmware boundary files plus `tests/boundary/mcp/firmwareMcp.test.ts` — 66 passed on each Node version. - `npm run check` on Node 24/Linux — passed, including typecheck, built lint, formatting, knip, metadata and architecture guards; lint reported 0 errors and 2744 warnings. - `git diff --check` — passed. ## Compatibility, safety, and release No production or release changes. The temporary verification workflows stay on separate fork branches. Real Unblob/Binwalk and other operating systems were not exercised by this change; the MCP test uses the synthetic producer. Full-suite coverage remains with upstream CI. ## Author readiness Human owner: @Plastic-time - [x] I reviewed the final diff and relevant existing code, understand the change and its edge cases, and take responsibility for the complete PR, including AI-assisted work. - [x] I checked the evidence above, confirmed it supports the current change, and recorded meaningful validation gaps. - [x] I declare this PR ready for initial maintainer review. ## Before merge - [x] This PR addresses a concrete problem or an agreed enhancement. - [x] Material scope and authority questions are resolved, and any required maintainer scope decision covers the final diff. - [x] The PR has one focused outcome and the title follows `type(scope): outcome`. - [x] Related issue is linked, or the reason for not linking one is stated above. - [x] Tests cover changed observable behavior and meaningful failure paths. - [x] Owning docs, contracts, and generated metadata are updated where needed. - [x] User-visible CLI/MCP changes include representative output. - [x] I checked the final diff for secrets, unrelated cleanup, and unsupported claims. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi

2026-11-03 13:39:03 +00:00

closed this issue
added the
pull-request
label

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

morluto/rea#7603

No description provided.

Rows
Columns