fix(triage): diff the flake-gate file list against the pinned base OID - #9464
Merged
Conversation
The "Record changed test files for the flakiness gate" step runs its `git diff 'HEAD^1' HEAD` inside the env -i scrubbed child. On the persistent pool, resolving the `^1` parent there intermittently fails with "Could not access 'HEAD^1'" — the shallow merge-ref object store is left unreadable by prior `--depth=2` fetches — which takes down the whole verify lane. The "Pin agent inputs" step already captures the base OID to $RUNNER_TEMP/verify-base-oid while .git is still root-owned. Diff against that content-addressed OID instead of re-resolving the parent: it needs no parent walk and is the same value the workflow already trusts for its post-build re-pin. Add a pin asserting the record step reads the recorded OID rather than re-deriving HEAD^1.
Collaborator
ytahdn
approved these changes
Aug 19, 2026
yiliang114
enabled auto-merge
August 19, 2026 07:08
chiga0
approved these changes
Aug 19, 2026
qwen-code-ci-bot
requested changes
Aug 19, 2026
qwen-code-ci-bot
left a comment
qwen-code-ci-bot
left a comment
Collaborator
There was a problem hiding this comment.
Thanks @yiliang114 — the underlying problem is real and verified: both linked Qwen Triage runs on #9432 (32222732873 and 32222952508) failed in their verify job with error: Could not access 'HEAD^1' (confirmed in the job log — the record step dies right after the env -i scrub re-exec, exactly as described). Diffing against the already-pinned base OID instead of re-resolving the parent is a sensible direction. Before this moves to code review, one housekeeping item: the PR body doesn't use the repository's pull request template.
Merged PRs here follow that template consistently, and reviewers rely on the fixed sections to find what they need. ## What this PR does is there, but the rest of the required headings are missing:
## Why it's needed— the body carries## WhyHEAD^1is fragile hereinstead## Reviewer Test Plan(with### How to verify,### Evidence (Before & After), and the### Tested onOS table) — the body carries## Verificationinstead## Risk & Scope## Linked Issues— the body carries## Referencesinstead- the Chinese translation inside
<details>— the template asks for the full body to be mirrored in Chinese
Nearly everything is already written in the current body — the what/why prose, the failure links, the node --test verification — so this should be a reformat rather than a rewrite. Please restructure it into the template sections: https://github.com/QwenLM/qwen-code/blob/main/.github/pull_request_template.md
One note while you're at it: the Tested on table wants ✅ /
Once the body follows the template, a re-run (@qwen-code /triage) will pick it up and continue the review. 🙏
中文说明
感谢 @yiliang114 —— 底层问题真实存在且已核实：#9432 上链接的两次 Qwen Triage 运行（32222732873 和 32222952508）均在 verify job 中以 error: Could not access 'HEAD^1' 失败（已在 job 日志中确认——record 步骤恰好在 env -i scrub 重新 exec 之后挂掉，与 PR 描述完全一致）。用已经固定下来的 base OID 做 diff、而不是重新解析父提交，方向是合理的。在进入代码审查之前，有一个流程问题：这个 PR 的正文没有使用本仓库的 pull request 模板。
本仓库已合并的 PR 都遵循该模板，reviewer 依赖这些固定章节来快速定位所需信息。## What this PR does 已有，但其余必需章节缺失：
## Why it's needed—— 正文目前用的是## WhyHEAD^1is fragile here## Reviewer Test Plan（含### How to verify、### Evidence (Before & After)以及### Tested on操作系统表格）—— 正文目前用的是## Verification## Risk & Scope## Linked Issues—— 正文目前用的是## References<details>中的中文翻译 —— 模板要求将正文完整翻译为中文
其实所需内容在现有正文里都已经写到了——改动与动机描述、失败链接、node --test 验证——所以只需重新排版，不必重写。请按模板章节重新组织：https://github.com/QwenLM/qwen-code/blob/main/.github/pull_request_template.md
另外提醒一点：Tested on 表格需要按操作系统标注 ✅ /
正文符合模板后，重新触发（@qwen-code /triage）即可继续审查。🙏
— Qwen Code · qwen3.8-max
Contributor
Code Coverage Summary
CLI Package - Full Text ReportCore Package - Full Text ReportFor detailed HTML reports, please see the 'coverage-reports-22.x-ubuntu-latest' artifact from the main CI run. |
Collaborator
|
Released in v0.21.15. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What this PR does
The "Record changed test files for the flakiness gate" step runs
git diff 'HEAD^1' HEADinside theenv -iscrubbed child. On the persistent pool, resolving the^1parent there intermittently fails witherror: Could not access 'HEAD^1', which takes down the whole verify lane (observed on #9432: runs 32222732873 and 32222952508).The "Pin agent inputs" step already captures the base OID to
$RUNNER_TEMP/verify-base-oidwhile.gitis still root-owned. This PR diffs against that content-addressed OID instead of re-resolvingHEAD^1— no parent walk, and it's the exact value the workflow already trusts for its post-build re-pin. The OID is guarded by a hex-shape check (mirroring the existing post-build re-pin), and the test pins are updated to match.Why it's needed
The step runs after the one-shot
env -ire-exec (the startup-channel scrub). ResolvingHEAD^1requires re-reading the merge commit's parents off the shallow object store; prior--depth=2fetches on the persistent pool leave that store intermittently unreadable, so the parent walk fails. The OID needs no walk.Reviewer Test Plan
How to verify
Run
node --test .github/scripts/qwen-triage-workflow.test.mjs. The structural suiteqwen-triage: flakiness gate (#9125)should pass, includingrecords the changed-test list BEFORE the workspace is handed to the build user(the test containing both pins).Evidence (Before & After)
Non-user-visible CI/workflow change.
git diff 'HEAD^1' HEADand intermittently failed witherror: Could not access 'HEAD^1'(see fix(core): clarify list_agents excludes Agent Team teammates #9432 runs 32222732873 and 32222952508)."$BASE_OID" HEAD; the structural suite passes.Tested on
Environment (optional)
N/A — unit tests only.
Risk & Scope
HEAD^1unreadable. This PR only stops re-resolving the parent; it does not fix the pool.verify-base-oidvalue it already captured for the post-build re-pin.Linked Issues
No dedicated tracking issue. Fixes the flakiness-gate failure observed on #9432 (runs 32222732873, 32222952508).
中文说明
这个 PR 做了什么
"Record changed test files for the flakiness gate" 步骤在
env -i清理过的子进程里执行git diff 'HEAD^1' HEAD。在持久化 pool 上，在那里解析^1父提交会间歇性失败并报error: Could not access 'HEAD^1'，进而拖垮整个 verify lane（在 #9432 上观察到：运行 32222732873 和 32222952508）。"Pin agent inputs" 步骤在
.git仍由 root 拥有时就已经把 base OID 捕获到$RUNNER_TEMP/verify-base-oid。本 PR 改为对这个内容寻址的 OID 做 diff，而不是重新解析HEAD^1—— 不需要父提交遍历，而且用的正是 workflow 在 post-build re-pin 时已经信任的那个值。OID 增加了一个十六进制形状的校验（对齐现有的 post-build re-pin），测试的精确匹配也同步更新。为什么需要
该步骤在一次性
env -i重新 exec（startup-channel scrub）之后运行。解析HEAD^1需要从 shallow object store 重新读取 merge commit 的父提交；持久化 pool 上先前的--depth=2fetch 会让那个 store 间歇性地不可读，于是父提交遍历失败。而 OID 不需要遍历。Reviewer 测试计划
如何验证
运行
node --test .github/scripts/qwen-triage-workflow.test.mjs。结构化套件qwen-triage: flakiness gate (#9125)应通过，包括records the changed-test list BEFORE the workspace is handed to the build user（包含两个 pin 的那个测试）。证据（前后对比）
非用户可见的 CI/workflow 改动。
git diff 'HEAD^1' HEAD，间歇性失败并报error: Could not access 'HEAD^1'（见 fix(core): clarify list_agents excludes Agent Team teammates #9432 的运行 32222732873 和 32222952508）。"$BASE_OID" HEAD做 diff；结构化套件通过。测试环境
环境（可选）
N/A —— 仅单元测试。
风险与范围
HEAD^1不可读的持久化 pool 底层 shallow object store 损坏问题。本 PR 只是不再重新解析父提交，并没有修复 pool。verify-base-oid值。关联 Issue
没有专门的追踪 issue。修复了在 #9432 上观察到的 flakiness-gate 失败（运行 32222732873、32222952508）。