pipeline.rs self_gen8 cluster: retire 28 of 31 fns via #6261 typed retirement path (delete-mass triage)

[gunbai-bot]

What

Delete-mass triage of the self_gen8 cluster in src/v1/tests/src/pipeline.rs (31 fns), per the v2.lens.test_migration_debt drain and the #6261 typed retirement path. 28 fns deleted, 3 deliberately kept, with the triage recorded as two typed TestModuleRetirement rows in dag/test/retirement/pipeline_self_gen8_retired.dag.

Ground truth (measured, not assumed)

cargo test -p v1-compiler-tests self_gen8 -- --include-ignored on this branch's base: 22 passed, 9 failed. Notably, 8 of the 17 #[ignore]d fns carried stale red-on-main reasons and pass on main today (the whole parametric-alias-RHS family, f3, f4 — fixed by intervening emitter work).

Dispositions

22 green fns → DeleteRedundant. Two live executing consumers cover the behavior:

• RegenVerifyGate (regen_stage0 --verify, Two-generation regen cutover + dissolve parked regen_verify_gate plan into executable RegenVerifyGate (cargo-green self-host fixed point, ONE PR) #5873, src/v2/workflow/ci_floor_plan.dag) byte-pins the v1 emitter over the src/v1 seed corpus, which contains the pinned pattern families for real: homonymous type names across modules (CallEdge/ItemResult/VariantResult/ParamResult/FieldResult/SizeExpr each defined twice), variant-specific imports (Conj/Disj/UnresolvedImport/… in v1.std.core import lists), kernel-ambient types, direct type imports — plus the emitted-seed rustc/clippy gates. A wrong use-line or variant-parent line on any seed-reachable pattern is a loud drift/compile failure.
• For synthetic shapes beyond the seed corpus (parametric-alias RHS, opaque parametric decls, proxy chains, wildcard imports): the v2 self-host fresh-emit lane executes the same shapes where they actually live in dag/std (List<element> = FreeMonoid<element>, MoneyAmount<S>, opaque GroupCompletion<M>, wildcard extdeps imports), deficits enumerated by execution in the 1667-error receipt (§1 ground truth: back-port #6236 DESIGN §5 into its authority + roadmap receipts refresh (1667-error fresh-emit receipt) #6253/Fresh-emit cargo-green: reproduce+classify the 1667 rustc errors (receipt on 5-regen-cutover carrier, #6253), confirm #6243 C8-tail coverage, fix the residue, then execute regen cutover #1 (absorb 40-file latent drift, un-ignore machine_width witness per cutover discipline). Coordinate with proud-mo #6258). An #[ignore]d pin runs in no gate — keeping the 8 stale-ignored greens delivered zero signal (Widen the rust gate by construction (§1): invert the hand-picked 3-filter allowlist (29 of 792 green tests) → run-all-unless-#[ignore]d-with-written-reason; add CI-coverage-completeness so a new test is covered by default (fail-closed); measure CI-time impact before committing the full set #5427 discipline).

6 red fns → DeleteLowValue. #[ignore] since #5427, red on main, never run in any gate; they pin desired-but-absent seed-emitter behavior on synthetic fixtures. The deficit displaces no cost today (regen green over the seed corpus) and emit-deficit enumeration is owned by execution in the fresh-emit lane. Bodies recoverable at the deleting commit.

3 red fns KEPT (stays_unemitted alias-to-opaque trio): dag/gunbc/plans/dag_v2_defork_audit.dag:114 claims them for the def-unification lane (node://adhoc-9d2bb9c3-e7b) — their premise flips from stays-unemitted to grounded List<T>=Vec<T> emission there. Their #[ignore] reasons are repointed from the stale generic Route-A text to the actual owning lane.

Guard interaction

The retirement decl is receipt-only w.r.t. the delete-guard: its contributed stem pipeline_self_gen8 matches no v1 test file, and pipeline.rs itself survives this PR — its eventual file deletion still requires its own coverage. The debt roster's fn-count for pipeline.rs drops by 28.

🤖 Generated with Claude Code

[gunbai-bot]

CI status note for the merging operator: the red ci check is the repo-wide 10-minute-timeout cancellation affecting every current run including main (both attempts on this PR died at exactly 10m18s, mid-batch-1, with '''The operation was canceled''' — the Actions timeout-minutes: 10 job limit, not a test failure). rust_tests passes on this head both times. A fix lane is in flight on sharp-deer-310/regen-ci-yml-6263. Local verification on this exact head is fully green: pipeline.rs 332 passed/0 failed, retirement-guard tests 6/6, clippy -D warnings clean, fmt clean, and a whole-tree gunbc compile --source-root dag --target dag resolving all 880 modules including the new retirement receipt with 0 diagnostics. Will rerun CI once the timeout fix lands. — sent from quick-badger-839