Flaky TestGetPRHeadDrift (stamp-without-stream), pre-existing on main #617

New issue

Closed

opened 2026-09-15 23:45:13 +00:00 by crueber · 1 comment

crueber commented

2026-09-15 23:45:13 +00:00

Owner

Pre-existing flake, surfaced during #613 review and reproduced on clean origin/main there (3/30 -race) vs 1/30 with the PR applied — same order, code path untouched by #613 (per-env isolation). Fails intermittently in-suite, passes solo. Needs a stress run (-count/-race) and a fix or quarantine decision; not a #613 blocker.

crueber added the

bug

issues

labels

2026-09-15 23:45:13 +00:00

crueber referenced this issue from a commit

2026-09-16 00:01:42 +00:00

Fix flaky TestGetPRHeadDrift stamp-without-stream race (Forgejo #617)

crueber referenced this issue from a pull request that will close it,

2026-09-16 00:01:46 +00:00

Fix flaky TestGetPRHeadDrift stamp-without-stream race (fixes #617) #618

crueber closed this issue

2026-09-16 00:04:18 +00:00

crueber commented

2026-09-16 00:04:18 +00:00

Author

Owner

Fixed by #618 (merged): root-caused as a logical-ordering flake (background mergeable pass wins the race and stamps without streaming); test now drains the pass via waitMergeableDrained (bounded 5s poll) before mutating refs. Test-only, no prod change, no weakened assertions. Verified: 300/300 targeted + full package -race green, cover 96.1%, independent review APPROVE (root cause + fix soundness re-verified in code).

crueber referenced this issue from a commit

2026-09-16 00:04:19 +00:00

Merge pull request 'Fix flaky TestGetPRHeadDrift stamp-without-stream race (fixes #617)' (#618) from fix/issue-617 into main