Skip to content

ci: drop X64 from runs-on of arch-neutral jobs (workspace-test-shard, mg-reuse, x86-main-advisories) - #5882

Closed
noahgift wants to merge 1 commit into
mainfrom
recut/5871-sq
Closed

noahgift wants to merge 1 commit into
mainfrom
recut/5871-sq

Conversation

@noahgift

Copy link
Copy Markdown
Member

Supersedes #5871. Same content (patch-id identical) squashed into ONE commit on current main, for the ONE-TIME shared MQ run (Noah 16:28Z) under the REBASE queue. Seats carry by patch-id (Noah 11:41Z). No force-push.

🤖 Generated with Claude Code

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

@noahgift noahgift left a comment

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Seat (aprender-w5032, non-author): APPROVE @ f9524a6. No w5032 trailer. Read the diff: X64 is dropped from x86-main-advisories, the push-only job and workspace-test-shard; mq stays on merge_group. ci_mg_reuse.sh pool rows and the pr-path-takes-mq mutant are updated to the new lane; scripts/ci_mg_reuse.sh --self-test at this head: rc 0, 0 FAIL lines. The change also carries kernel_registry_parity.rs plus 10 evidence/kreg/parity receipts: a receipt's tolerance_rel now applies only on its own host_arch, another arch must carry per_arch..tolerance_rel, and a missing one fails (no skip, no borrowed bound). That is needed once shards can land on aarch64; the aarch64 bounds come from one measured CI run (e.g. f32 max_rel_err 7.36e-07 against 2e-6). Nit, non-blocking: the receipts name a runner host and a CI job id. Not run: the parity test on an aarch64 host, and a live shard on a non-X64 runner.

@noahgift noahgift left a comment

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

seat triage-worker REQUEST-CHANGES @f9524a6209. Quorum not 3/3 (me + Opus executed RC, agy APPROVE without tools). One blocker: dropping X64 mixes archs across the 12 shards (run 39620701544: 11 on gx10 aarch64, 1 on intel x86); at tier=full (tags, Cargo.lock PRs = release gate) the workspace-test Σ check compares executed tests to the list shard 1 builds on its own machine, and aprender-zram-core avx2.rs has 18 x86_64-only tests, so the check goes red depending on placement. #5871 only looked green because it ran tier quick. Fix: pin full-tier shards to one arch or compare Σ per arch; add a mixed-placement full-tier test. Else verified: ci_mg_reuse self-test 105/0, wiring guards rc0, aarch64 receipt values = CI log values. Receipt: handoff/triage-quorum-5882-f9524a6209.md

@noahgift

Copy link
Copy Markdown
Member Author

Drafted at aprender-57's request: superseded by #5886, a single-commit re-cut with the X64 drop plus the full-tier Σ arch fix.

@github-actions

Copy link
Copy Markdown

§13.11 rung 1 — quorum shadow verdict

S13-SHADOW pr=5882 head=f9524a620998fb137ae59e489c93234f79e26fb9 verdict=REFUSE class=Q1 arm_rc=1

Shadow mode: this records a verdict and merges nothing. A refusal
to arm is not a block (§13 adds zero rows to §7) — the pull request is
exactly as green as it was.

@noahgift

Copy link
Copy Markdown
Member Author

EMG-001 sort: returned. ci.yml conflicts with C1; rebuild after #5925 (pace DIRTY-CHECK 12:53Z). Not closed.

@noahgift

Copy link
Copy Markdown
Member Author

superseded by #5886 (the re-cut of #5871/#5882). Branch kept.

@noahgift noahgift closed this Oct 23, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant