🐤 Nightly Canary: upcoming-toolchain regression — e954c33 #613

Activity

1. github-actions

   assigned

   githubrobbi

   on Aug 24, 2026
2. github-actions

   added

   ci-failureAutomated CI/CD failure notificationAutomated CI/CD failure notification

   ci-failure-nightly-canary

   on Aug 24, 2026
3. githubrobbi commented on Aug 24, 2026

   

   githubrobbi

   on Aug 24, 2026

   Collaborator

   More actions

   Root cause: upstream rustc/clippy regression — no code defect on our side

   The floating nightly this run used, 1.100.0-nightly (fb6531d55 2026-08-23), is the nightly that enabled the next-generation trait solver (rust-lang/rust#160895). Under the next solver, clippy's large_futures lint ICEs while computing layout_of on async-fn opaque types — upstream rust-lang/rust#161563 / rust-lang/rust#161495. That lint is in clippy::pedantic, which this workspace denies, and it runs layout on every async fn's future — hence the ICE storm across all three platforms, including inside tokio's own opaques. cargo check on the same nightly is unaffected; only the clippy leg ICEs.

   Fix status (all upstream, already done): rust-lang/rust-clippy#17601 (merged 2026-08-23 17:02 UTC) fixes the lint under the next solver; the Clippy subtree sync rust-lang/rust#161416 carrying it merged 2026-08-24 02:35 UTC — after the 08-24 nightly cutoff, so the first fixed toolchain is nightly-2026-08-25.

   Action for us:

   • No dependency bump applies (toolchain-side, unlike the ethnum episode) and no source change is needed.
   • Lint posture unchanged and non-negotiable: large_futures stays denied — no allow, no lint-set trimming to dodge an upstream ICE.
   • Pin-bump guidance per this canary's purpose: skip nightly-2026-08-23/24 — the next just toolchain-sync lands on ≥ nightly-2026-08-25. That bump also brings the next-gen trait solver onto the pinned gate; if the stricter solver surfaces new lint findings, they get real fixes, not exceptions.
   • Verification: workflow_dispatch the Nightly Canary once nightly-2026-08-25 is published (~2026-08-25 morning UTC); close this issue on green. Do not re-dispatch before then — the current latest nightly still lacks the fix and would file a duplicate tracker.
4. github-actions commented on Aug 25, 2026

   

   github-actionsBot

   on Aug 25, 2026 – with GitHub Actions

   ContributorAuthor

   More actions

   🐤 Nightly Canary failed on main

   Field	Value
   Workflow	🐤 Nightly Canary
   Run	#18
   Attempt	1
   Branch	main
   Commit	910364a
   Triggered by	@githubrobbi

   Filed by ci-failure-notify.yml after the 60-second auto-rerun-transient window: auto-rerun-transient saw no transient signature in the failed-job logs → genuine failure on first attempt.

   ⚠️ This is an early-warning signal from the floating-nightly canary, not a broken pin. main still builds on the pinned toolchain. Investigate before the next pin bump; if it is an upstream/dep regression, the fix is usually a dependency bump (cf. the ethnum 1.5.2→1.5.3 episode), not a toolchain downgrade.

   📧 Notification sent to: cifail@uffs.io
5. githubrobbi commented on Aug 25, 2026

   

   githubrobbi

   on Aug 25, 2026

   Collaborator

   More actions

   Closing as verified fixed: manual canary dispatch on nightly-2026-08-25 (first toolchain with the clippy large_futures next-solver fix, rust-lang/rust#161416) shows zero unexpected rigid alias in layout_of errors — the ICE this issue tracked is gone. The run is still red from a different, unrelated regression in the same nightly (clippy's new empty_enums lint false-positive on zerocopy FromBytes derive expansions), tracked separately with full analysis in #617 so each canary issue names exactly one root cause. Pin-bump guidance: skip nightly-2026-08-23/24 (the ICE) AND hold before ≥ 08-25 until the zerocopy false positive resolves — see the successor issue.
6. githubrobbi

   closed this as completedon Aug 25, 2026