Skip to content

seals: six stale after #12695 changed single-element bracket codegen #13284

Closed as not planned

Description

@gHashTag

Master's Spec Guards workflow has been red since 20cabcd (merge of #12695). Its last green run was on 13e9fbd. The failing step is "Does every seal still describe what the compiler emits" (tools/check_seal_currency.py): 3208 seals scanned, 6 report a STALE generated-code hash on all four backends (rust, zig, c, verilog).

#12695 changed bootstrap/src/compiler.rs: a bracketed single element with a declared [1]T annotation is now a list, not a dimension. That changes the code emitted for these six specs. None of the specs changed; their seals now describe the old output. Every PR that runs Spec Guards inherits the red (seen on #13211, whose only change is a blog spec).

Boundary

Seal files under .trinity/seals/, one per spec below. No spec, compiler or tool change.

  • specs/analog/linear.t27 -> .trinity/seals/analog_AnalogLinear.json
  • specs/analog/linear_zoh.t27 -> .trinity/seals/analog_AnalogLinearZoh.json
  • specs/port/trios/crates/trios-chat/rings/CR-CHAT-03/src/external_psk_id_provenance.t27
  • specs/port/trios/crates/trios-chat/rings/CR-CHAT-03/src/pcs_healing.t27
  • specs/port/trios/crates/trios-chat/rings/CR-CHAT-03/src/reinit_freshness.t27
  • specs/port/trios/crates/trios-chat/rings/CR-CHAT-07/src/blind_signature_sender_token.t27

User Scenarios

  • A contributor opens a PR that does not touch these specs. Spec Guards is red on it anyway, and they have to work out that the red is inherited before they can trust the rest of the run.
  • A reviewer reading the master dashboard sees Spec Guards red and cannot tell a real seal regression from this known one.

Requirements

  • Re-save the six seals with a t27c built from master at or after 20cabcd (t27c seal <spec> --save, then tri seals sync-twins). Do not run this from a worktree under a home path; the seal can embed it and fail scan.
  • Do not hand-edit the hashes and do not add the six to tools/seal_baseline.txt as stale: the specs did not move, the compiler did, so a re-save is the honest fix.
  • The run belongs on a T27 OS pane or the lab, not on the Mac.

Success Criteria

  • python3 tools/check_seal_currency.py on master reports STALE generated-code hash : 0.
  • The Spec Guards workflow is green on the first master commit after the re-save.
  • The PR diff is the six seal files only (twins included if sync-twins writes them).

Refs #12695

Activity

  1. added
    shape-refusedClosed by the issue-shape gate (specs/policy/issue_shape.t27); a passing edit reopens it
    on Nov 7, 2026
  2. github-actions commented on Nov 7, 2026

    @github-actions
    Contributor

    Closed by the issue-shape gate (specs/policy/issue_shape.t27, #7498): an issue must have the template's sections. Edit the body; an edit that passes reopens it.

    • missing: ## Requirements -- numbered obligations, FR-001: ... MUST ...
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    shape-refusedClosed by the issue-shape gate (specs/policy/issue_shape.t27); a passing edit reopens it

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions