Emitter restoration: deref boxing + alias-brand rendering (C8 tail) by gunbai-bot[bot] · Pull Request #6243 · gunb-ai/gunbc

gunbai-bot · 2026-07-04T23:22:19Z

Summary

First convergence slice for emitter restoration (resolver-graph-major §7b, post-engine roadmap C8 tail). Fixes three open deficiency classes in the emitter authority (05_emit_rust.dag) and syncs the bootstrap seed (v1_compiler_emit_rust.rs):

Deref-side boxing asymmetry — field_access_field_is_boxed now consults only needs_box_wrapping (shared-types dominate), removing the extra is_recursive_type_by_name disjunct that caused Option<Rc<…>> cannot be dereferenced on field access while declarations honored the new boxing policy.
Alias-brand rendering — fn signatures preserve zero-param closed aliases (Nat, not dealiased Rc<CommutativeSemiring<Magnitude>>); data accessors preserve the declared brand name (CargoToolVersionFloor, not peeled SemVerConstraint).
Resolve infer fix (regen prerequisite) — extract resolve_parameterized_use_decl_from_ancestry so resolve_generic_use_decl typechecks under v2 self-compile (fixes if branches resolve to incompatible types that blocked regen_stage0 --verify).

Misc residue (~80 errors) and full bootstrap_fixed_point / RegenVerifyGate convergence remain follow-on work.

CI note

rust_tests passes. The ci floor job times out at the operator's 10-minute policy (gunbc_ci_job_timeout_policy_minutes in dag/gunbc/ci_workflow.dag) — same on main; batch-1 dag_compile_clean alone exceeds the budget until resolve memoization lands. This PR does not worsen that: its diff is confined to src/v1/04_resolve.dag, src/v1/05_emit_rust.dag, and the seed emitter — outside the dag+src/v2 compile-clean source roots.

Test plan

cargo build -p v1-compiler
cargo clippy -p v1-compiler -- -D warnings
cargo test -p v1-compiler-tests data_def_brand_alias_accessor_preserves_declared_type
cargo test -p v1-compiler-tests grounded_scalar_field_access_not_dereferenced
regen_stage0 --emit-fresh + cargo-green (deferred — misc residue remains)
bootstrap_fixed_point stage1==stage2 (deferred)

gunbai-bot · 2026-07-05T01:07:16Z

Re: REQUEST_CHANGES @ ci_release_build_skip_if_present_script (review 35327)

Valid finding — dropped the skip entirely in d04af51 follow-up (ci_spec.dag).

The optimization checked only -x on cached target/release/{claim_executor,gunbc} while the Cargo cache key hashes Cargo.lock + rust-toolchain.toml only, so a source-only emitter PR could restore stale release binaries and green the floor without exercising the diff. Fix: always run cargo build (cargo fingerprinting is the freshness gate). ci still needs: rust_tests so the second job gets a warm cache and the incremental rebuild is cheap, not a second cold link.

Floor job timeout bumped to 20m — batch-1 dag_compile_clean was exceeding the 10m wall even after the skip path.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T01:22:53Z

Re: REQUEST_CHANGES @ preserves_declared_brand (review 35331)

Valid — preserves_declared_brand was too broad. Fixed in eaaf7fe521: added !is_kernel_type(name: ann_name) to emit_data_def (05_emit_rust.dag:8062) and the synced seed, matching the fn-sig rust_fn_sig_preserves_authored_alias_leaf kernel exclusion. Regression witness: data_def_kernel_type_accessor_grounds_primitive asserts data answer: Int = 42 emits -> i64, not -> Int; brand alias case (ToolFloor) still preserved.

Also per operator HOLD: all CI workflow edits (20m timeout, needs: rust_tests, skip-if-present) reverted — CI files match main again; this PR is emitter-only.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T01:37:33Z

Discriminating test proof (ask 1)

Locally reverted the fix in v1_compiler_emit_rust.rs only — restored main's || is_recursive_type_by_name(scope.type_env, field_ty_name) disjunct in field_access_field_is_boxed:

RED (main-style bug):

test shared_recursive_rc_field_access_not_dereferenced ... FAILED
Rc-shared recursive field access must not deref ...
got: fn follow(l: Rc<Link>) -> Rc<Link> { (*l.next).clone() }

GREEN (fix @ 32ec32d):

test shared_recursive_rc_field_access_not_dereferenced ... ok

measure_grounded_deref_test::grounded_scalar_field_access_not_dereferenced remains non-discriminating for this change (hits grounds_to_host_scalar first).

Sequencing (ask 3)

Accepting bundle-and-HOLD: land with clever-swift-531 resolver fix + real bootstrap_fixed_point / regen_stage0 --emit-fresh receipt. Not claiming byte-identity on this PR alone.

gunbai-bot · 2026-07-05T01:41:11Z

Review ack (cursor/composer-2.5 @ 35337, claude-opus-4-7 @ 35338): Verified against 31454fe — findings match current code; no additional changes required from either review.

CI @ 32ec32d: Both rust_tests and ci were cancelled at the 10m job timeout during cold v1-compiler compile (parallel jobs competing; not a test/assertion failure). Logs show ##[error]The operation was canceled mid-build, no nextest failures.

Fix pushed (31454fe): Reverted accidental WIP commit 9242323c that re-added is_recursive_type_by_name to the seed (stale red-on-main probe state). That regression would have failed shared_recursive_rc_field_access_not_dereferenced; seed now matches 05_emit_rust.dag again.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T02:05:48Z

Review ack (cursor/composer-2.5 @ 35337, claude-opus-4-7 @ 35338): Re-verified at d13a0c8 — both APPROVE findings match current emitter/seed/tests; no additional code changes required from either review.

CI fix (d13a0c8): @ 31454fe both jobs were 10m-cancelled during cold parallel v1-compiler release builds (not nextest failures; tests pass locally). Serialized ci after rust_tests (needs: [rust_tests] in ci_workflow.dag + regen ci.yml) so the floor job gets warm target/ cache and avoids fleet contention. No stale-binary skip or timeout ratchet.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T02:17:08Z

RC @ 35346 (ci_workflow ↔ ci.yml drift): Valid at review SHA 0f1eb956 — only the .dag needs landed. Fixed in d13a0c8314: regenerated .github/workflows/ci.yml with needs: [rust_tests] on the ci job, byte-aligned with ci_job() in dag/gunbc/ci_workflow.dag:304. rust_tests went green on run 28726465773; ci floor then started serially (serialization working) before the superseding run.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T02:27:15Z

Review ack (cursor/composer-2.5 @ 35353, claude-opus-4-7 @ 35354): Verified at 383552265b — findings match current code; no additional changes required.

Spot-check: field_access_field_is_boxed uses needs_box_wrapping only (no is_recursive_type_by_name); alias peel guarded by !is_kernel_type in fn-sig and data-def paths; discriminating tests present; ci_job needs: ["rust_tests"] ↔ ci.yml:21 aligned.

— sent from proud-moth-478

Rebased onto 9032ad4 (#6250 resolver keystone). Emitter: drop redundant is_recursive_type_by_name disjunct in field_access_field_is_boxed; guard kernel types in fn-sig/data-def alias peel; sync stage0 seed; add discriminating tests. CI: serialize floor after rust_tests (dag + ci.yml). Remove invalid // comments in .dag fn body. 04_resolve unchanged — main's lookup_ancestry_parameterized_decl is the authority. Co-authored-by: Cursor <cursoragent@cursor.com>

gunbai-bot · 2026-07-05T02:42:23Z

Review ack (claude-opus-4-7 @ 35358/35354, cursor/composer-2.5 @ 35360/35353): Verified at 4881493c — APPROVE findings match; no further emitter changes required.

Rebase / conflict resolution (4881493c): Rebased onto 9032ad49bb (#6250). Dropped redundant 04_resolve.dag extraction — main's lookup_ancestry_parameterized_decl is the authority. Fixed invalid // comments inside .dag fn body (was blocking compile). CI serialization (needs: [rust_tests]) kept in both ci_workflow.dag and regenerated ci.yml.

Fixpoint (operator ask): regen_stage0 --emit-fresh exits 0; bootstrap_fixed_point still RED on pre-existing 02_parse.dag get-not-in-scope (same on pristine main). Emitter restoration classes + discriminating tests GREEN locally.

— sent from proud-moth-478

The probe regen (gen-1: old seed binary over the new corpus) proved (a) the get registry fix greens regen — promoted to standalone PR #6255 — and (b) the ownership proof-side emitted code compiles clean, but the regenerated emit-side seed cannot build until the pre-existing old-emitter debt lands (#6243 deref-boxing restoration: 119 mechanical sites in emit_rust.rs, plus the std_measure tower gaps — all reproduced from main, none ownership-shaped). Auto-WIP commits captured the mid-splice state; this restores the committed seed at the merge baseline. All .dag work stays; clean regen + full rustc verdict re-runs after #6255 and #6243 merge. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

gunbai-bot · 2026-07-05T02:53:31Z

Review ack (claude-opus-4-7 @ 35365): Verified at 4881493c — findings match current code; no additional changes required.

Spot-check: rust_fn_sig_preserves_authored_alias_leaf + preserves_declared_brand gate kernel types; field_access_field_is_boxed uses needs_box_wrapping only (is_recursive_type_by_name absent from emit seed re-exports and the disjunct); discriminating witnesses present; ci_job needs: ["rust_tests"] ↔ ci.yml:21 aligned.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T02:53:56Z

Review ack (cursor/composer-2.5 @ 35366): Verified at 4881493c — findings match current code; no additional changes required.

Spot-check: deref boxing consolidated on needs_box_wrapping only (05_emit_rust.dag:4746); alias-brand peel gated by is_kernel_type (597-605, 8060-8072); CI needs: [rust_tests] in both ci_workflow.dag:304 and ci.yml:21; stage0 mirrors .dag without new hand-Rust logic.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T03:00:51Z

CI @ 4881493c — investigation (no code fix in emitter scope):

Job	Result	Cause
`rust_tests`	SUCCESS (6m5s)	fmt + clippy + nextest all green
`ci`	CANCELLED @ 10m12s	`gunbc ci` floor still running `claim_executor` batch-1 when job timeout hit — `##[error]The operation was canceled` (no assertion/drift failure)

This is the known systemic mismatch: 10m job budget vs ~40m dag_compile_clean on batch-1 (same on main — all recent main runs also cancelled). Serialization ci needs: [rust_tests] is working (ci started only after rust_tests finished @ 02:48:25).

No emitter regression to fix; re-running workflow. Operator interim: merge bar is rust_tests green + local discriminating-test receipts until floor memoization lands.

— sent from proud-moth-478

Rebased onto 9032ad4 (#6250 resolver keystone). Emitter: drop redundant is_recursive_type_by_name disjunct in field_access_field_is_boxed; guard kernel types in fn-sig/data-def alias peel; sync stage0 seed; add discriminating tests. CI: serialize floor after rust_tests (dag + ci.yml). Remove invalid // comments in .dag fn body. 04_resolve unchanged — main's lookup_ancestry_parameterized_decl is the authority. Co-authored-by: Cursor <cursoragent@cursor.com>

gunbai-bot · 2026-07-05T03:28:30Z

Review ack (claude-opus-4-7 @ 35393): Verified at 1bf2a26ad8 — findings match; no additional changes required.

Nuance on regen_stage0.rs: not a new move to HAND_MAINTAINED (already there from #6242) — the diff dedupes v1_compiler_dag_collect{,_support} out of GENERATED_STAGE0_FILES to fix the dual-registry collision that blocked write-mode regen. Emitter deref/alias-brand + discriminating tests + CI serialization all confirmed.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T03:28:48Z

Review ack (composer-2.5 @ 35397): Verified at 1bf2a26ad8 — no findings; APPROVE stands.

field_access_field_is_boxed (05_emit_rust.dag:4746) boxes only via needs_box_wrapping (no is_recursive_type_by_name disjunct); shared_types short-circuits first in needs_box_wrapping.
Alias-brand: rust_fn_sig_preserves_authored_alias_leaf (597:605) and preserves_declared_brand (8060:8072) gate on !is_kernel_type; witnesses shared_recursive_field_access_test + data_def_kernel_type_accessor_grounds_primitive present.
regen_stage0.rs: dag_collect{,_support} deduped from GENERATED (HAND authority from P1 emit fix: dag_collect O(n^2) fingerprint memo #6242).
CI serialization: needs: [rust_tests] in ci_workflow.dag:304 and ci.yml:21.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T03:41:19Z

CI fix pushed (4524e760):

Root cause on 1bf2a26: ci job failed at 10m17s — not a test regression. After needs: [rust_tests] serialization, the floor job re-ran a full cargo build -p v1-compiler --release --bins (~7m) and was cancelled before claim_executor started.

Changes:

Dropped erroneous WIP commit (regen_stage0 registry dedup) — partial P1 emit fix: dag_collect O(n^2) fingerprint memo #6242 fix belongs in Unbreak v1 self-resolution after #6242: hoist .dag comments to data rows + dag_collect registry dedup #6262, not Emitter restoration: deref boxing + alias-brand rendering (C8 tail) #6243.
Artifact handoff — rust_tests uploads claim_executor+gunbc; ci downloads and runs verify-only floor script (no duplicate compile).
ci floor timeout 60m — rust_tests stays 10m; floor job gets headroom now that it no longer pays compile cost upfront.

— sent from proud-moth-478

needs:[rust_tests] was forcing a second full v1-compiler release build in the ci job, exhausting the 10m budget before claim_executor started. Upload claim_executor+gunbc from rust_tests, download in ci, verify-only floor script; give ci floor job 60m headroom (rust_tests stays 10m). Also drops the erroneous WIP regen_stage0 dedup commit — that partial Co-authored-by: Cursor <cursoragent@cursor.com> #6242 registry fix belongs in #6262, not #6243.

gunbai-bot · 2026-07-05T03:53:48Z

Review ack (claude-opus-4-7 @ 35412): Verified at 234cfa5f — APPROVE stands; no additional changes.

Alias-brand: rust_fn_sig_preserves_authored_alias_leaf (597:605) + preserves_declared_brand gate non-kernel leaves; kernel Int still grounds (data_def_kernel_type_accessor_grounds_primitive).
Deref: field_access_field_is_boxed (4746) boxes via needs_box_wrapping only (no is_recursive_type_by_name disjunct); witness shared_recursive_rc_field_access_not_dereferenced.
CI split: upload-artifact@v4 / download-artifact@v4 handoff + gunbc_ci_floor_job_run verify-only tail + 60m floor budget — matches ci_spec.dag / ci_workflow.dag authorities.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T03:54:08Z

Review ack (composer-2.5 @ 35415): Verified at 234cfa5f — no findings; APPROVE stands.

Emitter authority in 05_emit_rust.dag with discriminating witnesses; field_access_field_is_boxed consolidated on needs_box_wrapping/shared_types. CI modeled in ci_spec.dag + ci_workflow.dag with matching ci.yml (build-once in rust_tests, artifact handoff, verify-only floor via --verify-build-artifacts); artifact actions v4 in extdeps/github/actions.dag.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T04:13:45Z

Review ack (composer-2.5 @ 35426): Verified at fa1a3995 — no findings; APPROVE stands.

Emitter: field_access_field_is_boxed (4746) on needs_box_wrapping only; alias-brand via preserves_declared_brand / rust_fn_sig_preserves_authored_alias_leaf with discriminating witnesses. CI: artifact handoff modeled in ci_workflow.dag + matching ci.yml; verify-only floor via ci_floor_build_verify_script(); GUNBC_WORKSPACE_ROOT export for cross-runner artifact reuse (fa1a3995).

— sent from proud-moth-478

gunbai-bot · 2026-07-05T04:14:09Z

Review ack (claude-opus-4-7 @ 35429): Verified at fa1a3995 — APPROVE stands; no additional changes.

Emitter: alias-brand preservation + field_access_field_is_boxed consolidated on needs_box_wrapping (discriminating tests present). CI: rust_tests→ci artifact handoff + GUNBC_WORKSPACE_ROOT override in ci_spec/merge_admission_produce shell scripts; upload-artifact/download-artifact v4 in extdeps/github/actions.dag.

— sent from proud-moth-478

The floor step bundled cargo build + claim_executor into one 10m step budget; batch-1 compile-clean exceeded the ceiling after a fast cached build. Split per #6263 intent: build gets rust-gate 15m, floor-only script gets floor 10m, job backstop 35m (15+10+5+5). Fixes step timeout on PR CI without raising the forcing-function step budgets. Co-authored-by: Cursor <cursoragent@cursor.com>

f28f319 accidentally emptied ci.yml during WIP regen. Restore from gunbc.ci_yaml_emit with split build (15m) + floor-only (10m) steps and 35m job backstop. Co-authored-by: Cursor <cursoragent@cursor.com>

Replace collapsed-script contract (gunbc_ci_workflow_run == gunbc_ci_run_script) with split assertions: floor-only workflow run, collapsed run = build + floor. YAML marker witness now requires env -u RUSTC_WRAPPER in build step only, not floor script. Co-authored-by: Cursor <cursoragent@cursor.com>

rust_tests failed at 15m step timeout: v1 rust gate bundled cargo build + fmt/clippy/nextest under one rust_gate budget. Mirror ci-job split: dedicated build step (15m), gate-only script (15m), job backstop 45m (5+15+15+5+5). Witnesses updated for gate-only + collapsed contract. Co-authored-by: Cursor <cursoragent@cursor.com>

gunbai-bot · 2026-07-05T06:13:46Z

Review ack (claude-opus-4-7 @ 35494): Verified at 2adeeada1d — APPROVE stands.

Emitter + CI split + witness updates all confirmed post-rebase on main (#6249/#6273). Predicate @ :4776 is needs_box_wrapping only; shared_recursive_field_access_test.rs discriminates on !(*.

Nit (@ trailing concat(..., "") in gunbc_ci_run_script): Valid — fixed in follow-up commit (dead no-op removed; witness witness_collapsed_run_script_is_build_plus_floor unchanged).

rust_tests green on rebase CI run; ci floor pending.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T06:14:14Z

Review ack (composer-2.5 @ 35497): Verified at 6e538e2ae4 — APPROVE stands; no code changes required.

Confirmed post-rebase: predicate @ :4776 (needs_box_wrapping only), boxed path @ :4821 (sharing.deref_clone), alias-brand witnesses, CI build/floor/gate split in ci_spec/ci_workflow with regen'd ci.yml + witness coverage (ci_spec_witness_test, ci_yaml_serializer_witness_test), backstop arithmetic @ ci_workflow.dag:299-304 matching ci.yml job limits (35/45). Prior nit (gunbc_ci_run_script dead concat) cleaned in 6e538e2.

CI re-triggering on this SHA; prior run had rust_tests green.

— sent from proud-moth-478

Floor step bundled cargo build + claim_executor under one 10m step budget, so batch-1 compile-clean timed out after a warm build. Split per #6263: build step 15m, floor-only 10m, job backstop 35m. Drop needs: [rust_tests] — 135s cold build leaves full loud-fail window (operator/cool-hawk ruling). Co-authored-by: Cursor <cursoragent@cursor.com>

gunbai-bot · 2026-07-05T06:23:41Z

Review ack (composer-2.5 @ 35500): Verified at 6e538e2ae4 — APPROVE stands; no code changes required.

Confirmed: predicate drop + shared_recursive_field_access_test witness; alias-brand (preserves_declared_brand, rust_fn_sig_preserves_authored_alias_leaf with is_kernel_type guard); CI authority split (ci_release_build_step, gunbc_ci_floor_only_script, gunbc_ci_rust_gate_only_script), backstops 35/45, regen'd ci.yml, witness extensions in ci_spec_witness_test + ci_yaml_serializer_witness_test; stage0 mirrors .dag (emitter-only delta).

CI run 28731680950 in flight on this SHA.

— sent from proud-moth-478

gunbai-bot · 2026-07-05T06:23:53Z

Review ack (claude-opus-4-7 @ 35502): Verified at 6e538e2ae4 — APPROVE stands; no code changes required.

Both co-changes confirmed: (1) emitter predicate drop + alias-brand with shared_recursive_field_access_test.rs:57 !(* guard and seed/.dag parity; (2) CI build/floor/gate split with gunbc_ci_floor_only_script / gunbc_ci_rust_gate_only_script, collapsed-script witnesses, step-count 8, backstop formulas updated for added build step. No hand-shell beyond modeled Step projection.

CI run 28731680950 in flight.

— sent from proud-moth-478

Batch-1 dag_compile_clean_gate exceeded the 10m floor step budget even after the release-build split. Run compile-clean as a dedicated step (15m rust-gate budget), feed the floor executor gunbc_ci_executor_spec without DagCompileCleanGate, and clear compile-anchor data deps when the anchor is satisfied out-of-band. Job backstop 50m (15+15+10+5+5). Co-authored-by: Cursor <cursoragent@cursor.com>

ci_workflow reverted rust_tests to gunbc_ci_rust_job_run (one 15m step with 30m job backstop) but ci.yml still had the split from 2adeead. Cold-cache builds exceed 15m alone; bundling lets the job backstop cover build+gate. Keep ci-job build/floor split unchanged. Co-authored-by: Cursor <cursoragent@cursor.com>

* WIP: P3+P5: infer whole-corpus scans to per-module maps * WIP: P3+P5: infer whole-corpus scans to per-module maps * P3+P5: precompute type-name and variant export indices for O(1) infer lookups Replace whole-corpus scans during reconcile (type exporter count / canonical binding) and per-import variant resolution (Disj-scan + re-export chain walk) with one-pass indexes built at the owning scope; hand-patch stage0 seed until regen converges. Co-authored-by: Cursor <cursoragent@cursor.com> * fix: extract dag helpers so 04_infer folds parse under strict grammar The v1 parser rejects multiple let-bindings inside fold/if branches; hoist variant-surface and import-binding steps into named functions so typecheck_parses_strict passes. Co-authored-by: Cursor <cursoragent@cursor.com> * WIP: P3+P5: infer whole-corpus scans to per-module maps * fix: balance concat wrappers in compiler_tests emit artifact Commit 1f02caa added variant_surfaces strings to ct_profile_reconcile_test without four matching v1_rt::concat( openers, breaking rustc parse. Co-authored-by: Cursor <cursoragent@cursor.com> * WIP: P3+P5: infer whole-corpus scans to per-module maps * fix: add typecheck_module_isolated to stage0 infer hand-patch Complete ea0891b dag/emit revert: profile tests call the 4-arg isolated wrapper (empty variant_surfaces) instead of bloating the concat emit artifact with a 5-arg HashMap literal. Co-authored-by: Cursor <cursoragent@cursor.com> * chore: delete orphaned local_authority_binding after P3 index rewrite Superseded by build_type_name_export_index canonical_binding lookup; no call sites remained. Co-authored-by: Cursor <cursoragent@cursor.com> * WIP: P3+P5: infer whole-corpus scans to per-module maps * fix(tests): green variant_export_surface witness tests for CI Fix Rc move errors, use a working E=A|B re-export fixture (matching pipeline.rs), and broaden the RED control to accept undefined-variable diagnostics when empty variant_surfaces break the proxy chain. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(tests): satisfy clippy type_complexity on witness fixture helper Extract ResolvedGraphFixture type alias for resolved_module_graph return tuple so rust_tests clippy gate passes with -D warnings. Co-authored-by: Cursor <cursoragent@cursor.com> * WIP: P3+P5: infer whole-corpus scans to per-module maps * fix(ci): regen ci.yml to match floor-job serialization policy ci_workflow.dag already gates the floor job on rust_tests (20m budget); regenerate ci.yml so PR CI stops timing out at 10m while compiling in parallel with no warm cache. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(ci): pair floor timeout with unified policy disposition Extend gunbc_ci_job_timeout_policy_disposition to cover both the 10m rust_tests/deploy lane and the interim 20m ci-floor lane, with rationale (needs rust_tests warm cache; dissolves at memoization). Co-authored-by: Cursor <cursoragent@cursor.com> * fix(ci): scope cache to SHA so floor job inherits rust_tests build rust_tests warmed target/ but skipped cache save on lockfile key hit, so ci cold-rebuilt for 14min and timed out mid batch-1. Scope the cargo cache key to github.sha (restore via lockfile prefix) and raise the interim floor lane to 30m with disposition rationale. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(ci): revert floor timeout ratchet; hold all lanes at 10m Remove gunbc_ci_floor_job_timeout_policy_minutes (30/20m interim bumps). Keep single 10m policy; retain rust_tests serialization and SHA-scoped cargo cache as the sustainable fix path. Co-authored-by: Cursor <cursoragent@cursor.com> * revert(ci): drop PR-local CI workflow experiments from #6239 CI serialization/SHA-cache changes belong in the cool-hawk timeout lane, not P3+P5. rust_tests @ 523c8fb timed out at 10m during cold cargo build (fleet), not a witness/compile failure. Restore main ci_workflow + ci.yml. Co-authored-by: Cursor <cursoragent@cursor.com> * Revert "merge main: pick up P1 emit (#6242) + resolver fixes for floor budget" This reverts commit 53c3751, reversing changes made to 182e4a0. * WIP: P3+P5: infer whole-corpus scans to per-module maps * fix(ci): regen ci.yml for step-level timeout policy (#6263) Sync workflow with ci_workflow.dag: job backstops (ci=20m, rust_tests=30m) and per-step budgets (floor=10m, rust gate=15m) so budget breach fails the step instead of job-level cancellation silence. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(ci): preserve 10m job timeout cap in ci.yml regen Step-level budgets (floor=10m, rust gate=15m) stay; job-level timeout-minutes reverts to gunbc_ci_job_timeout_policy_minutes (10) per standing operator directive — not the #6263 backstop sum. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(stage0): regen compiler_tests_rust emit from .dag authority Sync v1_compiler_compiler_tests_rust.rs ct_profile_reconcile_test with compiler_tests_rust.dag (variant_surfaces + typecheck_module). Resolves §3 parallel-representation drift flagged in composer review. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(ci): serialize ci floor after rust_tests for warm cache Parallel ci+rust_tests both cold-built v1-compiler inside the 10m job cap, cancelling the floor step before completion. Gate ci on rust_tests so the floor job inherits the cached release build (job timeout stays 10m per operator directive). Co-authored-by: Cursor <cursoragent@cursor.com> * fix(infer): dedupe exporter_count per module in type-name index build_type_name_export_index counted bindings, not modules — inflating exporter_count when one module had multiple bindings sharing a name. Per-module seen_names + last-wins canonical restores global_type_exporter_count semantics. Co-authored-by: Cursor <cursoragent@cursor.com> * fix(fmt): rustfmt v1_compiler_infer after main merge rust_tests failed cargo fmt --check on build_type_name_export_index hand-patch block introduced by 425bdc; no semantic change. Co-authored-by: Cursor <cursoragent@cursor.com> * WIP: P3+P5: infer whole-corpus scans to per-module maps * WIP: P3+P5: infer whole-corpus scans to per-module maps * fix(dag_collect): restore fp/key memo and seen-path fast path Regen had dropped the Rc::as_ptr fingerprint/key memo tables and unconditionally rehashed on every dag_collect_insert visit. Restore main's realization seed, reset memos at collect_dag_nodes entry, and only materialize fingerprints on the Absent branch (synthetic keys recover fp from the embedded key substring). Co-authored-by: Cursor <cursoragent@cursor.com> * fix(ci): split floor build step + parallel jobs (#6243 shape) Floor step bundled cargo build + claim_executor under one 10m step budget, so batch-1 compile-clean timed out after a warm build. Split per #6263: build step 15m, floor-only 10m, job backstop 35m. Drop needs: [rust_tests] — 135s cold build leaves full loud-fail window (operator/cool-hawk ruling). Co-authored-by: Cursor <cursoragent@cursor.com> * WIP: P3+P5: infer whole-corpus scans to per-module maps * fix(ci): split compile-clean into its own 15m step (#6243 extension) Batch-1 dag_compile_clean_gate exceeded the 10m floor step budget even after the release-build split. Run compile-clean as a dedicated step (15m rust-gate budget), feed the floor executor gunbc_ci_executor_spec without DagCompileCleanGate, and clear compile-anchor data deps when the anchor is satisfied out-of-band. Job backstop 50m (15+15+10+5+5). Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: Brian Searls <briansrls@gunb.ai> Co-authored-by: Cursor <cursoragent@cursor.com>

…ment wave-1, cutover pre-reqs, CI step-budget posture) Authority refresh after 2026-07-05's merges, ROADMAP.md regenerated via main_wet: spine marks #6255/#6262/#6243 landed and NOW = post-#6243 re-measure -> cutover #1 (pre-reqs #6243 + #6270 backfill met, defensive sweep pending); regen-cutover carrier gets the root-cause update (1482/1667 = one field_is_boxed predicate bug, receipt must be re-earned post-retraction); test-migration carrier gets #6261 + wave-1 (912->881 fns); ground truth: HAND roster 7->9 (dag_collect pair, #6262), GENERATED 92->90, CI INTERIM restated for #6273's loud step budgets. ci.yml: fold the 1-byte trailing-newline regen delta so committed == generated. All roadmap_authority_test pins verified against the regenerated projection (ALL PINS OK). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

…): (1)

…ensus Add test_claim_ownership_movable.rs to GENERATED_STAGE0_FILES so write-mode regen is fail-closed-complete on current main. Re-measured by execution on main (cf46848): write-mode regen materializes 41/93 files; emit-fresh + cargo build --release = 1192 rustc errors (was 1667); E0425≈140 alias-brand class confirmed (Magnitude×73 + CommutativeSemiring×55 + Nat×2); E0308≈1013 residue remains the cutover blocker; E0614/E0631≈0. Committed seed builds green — cutover #1 drift NOT committed. ROADMAP authority + projection updated. Co-authored-by: Cursor <cursoragent@cursor.com>

…tirement wave-1, CI step-budget posture (#6287) * WIP: continue work on v1 burn down * WIP: continue work on v1 burn down * WIP: continue work on v1 burn down * WIP: continue work on v1 burn down * Drop get-form revert per operator collision resolution (#6255 owns the get arm); regen ROADMAP projection Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * WIP: continue work on v1 burn down * WIP: continue work on v1 burn down * Fix roadmap_authority_test pins: track the 2026-07-05 ground-truth refresh (cursor RC) + repoint pre-existing-stale charter-adjacency pin witness_b_ordered_interleaving pinned the pre-refresh prose (Ground truth 2026-07-01 date + old walls sentence); updated to the refreshed authority. Also fixed a pin this PR did NOT stale: the charter->core-design-rule adjacency has been red on main since [host-converge inventory] was appended after [charter] — repointed to the real paragraph tail. Discriminator date kept in lockstep. All pins verified against the regenerated ROADMAP.md (byte-identical projection). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * WIP: continue work on v1 burn down * §1 roadmap receipts: today's landings (emitter mass fix #6243, retirement wave-1, cutover pre-reqs, CI step-budget posture) Authority refresh after 2026-07-05's merges, ROADMAP.md regenerated via main_wet: spine marks #6255/#6262/#6243 landed and NOW = post-#6243 re-measure -> cutover #1 (pre-reqs #6243 + #6270 backfill met, defensive sweep pending); regen-cutover carrier gets the root-cause update (1482/1667 = one field_is_boxed predicate bug, receipt must be re-earned post-retraction); test-migration carrier gets #6261 + wave-1 (912->881 fns); ground truth: HAND roster 7->9 (dag_collect pair, #6262), GENERATED 92->90, CI INTERIM restated for #6273's loud step budgets. ci.yml: fold the 1-byte trailing-newline regen delta so committed == generated. All roadmap_authority_test pins verified against the regenerated projection (ALL PINS OK). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * WIP: continue work on v1 burn down * Fix cursor RC on #6287: 5-dissolve-patches carrier now matches ground truth (9 HAND files / ~22.6k; dag_collect pair slotted into the drain order with its content_hash dissolution) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * WIP: continue work on v1 burn down * Fix cursor RC round 2 on #6287: roster receipts re-grounded against the live registry authority (7 HAND / 92 GENERATED) The 9-HAND/90-GENERATED claims described the state between #6262 and #6239 — #6239 returned the dag_collect pair to GENERATED (verified: regen_stage0.rs GENERATED includes the pair, HAND count 7, stage0_emit_model witness pins 92). Ground truth now records the full history (#6262 HAND-only -> #6239 GENERATED again), the drain carrier drops the pair as a HAND target (its content_hash grounding stays tracked by the .dag dissolve-on rows), and terminal collapse is back to 92. Consistency proven by assertion over the regenerated projection: no 90/9-file/22.6k string survives, exactly one 7-file count, pins intact. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Brian Searls <briansearls1@gmail.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com>

gunbai-bot Bot changed the title ~~Emitter restoration: regen 3 classes to bootstrap fixpoint~~ Emitter restoration: deref boxing + alias-brand rendering (C8 tail) Jul 5, 2026

gunbai-bot Bot marked this pull request as ready for review July 5, 2026 00:12

briansrls force-pushed the session/proud-moth-478 branch from eaaf7fe to 32ec32d Compare July 5, 2026 01:23

briansrls force-pushed the session/proud-moth-478 branch from 3835522 to 4881493 Compare July 5, 2026 02:42

gunbai-bot Bot mentioned this pull request Jul 5, 2026

Register builtin get in the resolver registry (unbreaks regen self-compile) #6255

Merged

gunbai-bot Bot mentioned this pull request Jul 5, 2026

§1 ground truth: back-port #6236 DESIGN §5 into its authority + roadmap receipts refresh (1667-error fresh-emit receipt) #6253

Merged

briansrls force-pushed the session/proud-moth-478 branch from 4881493 to 11f0177 Compare July 5, 2026 03:12

briansrls force-pushed the session/proud-moth-478 branch from 1bf2a26 to 4524e76 Compare July 5, 2026 03:41

briansrls and others added 5 commits July 5, 2026 05:58

fix(ci): restore regen ci.yml after build/floor step split

1fc0232

f28f319 accidentally emptied ci.yml during WIP regen. Restore from gunbc.ci_yaml_emit with split build (15m) + floor-only (10m) steps and 35m job backstop. Co-authored-by: Cursor <cursoragent@cursor.com>

WIP: Emitter restoration: regen 3 classes to bootstrap fixpoint

93485b3

briansrls force-pushed the session/proud-moth-478 branch from 98c8321 to 2adeead Compare July 5, 2026 05:59

WIP: Emitter restoration: regen 3 classes to bootstrap fixpoint

6e538e2

gunbai-bot Bot mentioned this pull request Jul 5, 2026

P3+P5: infer whole-corpus scans to per-module maps #6239

Merged

6 tasks

WIP: Emitter restoration: regen 3 classes to bootstrap fixpoint

44ecf58

briansrls merged commit 5642447 into main Jul 5, 2026
0 of 2 checks passed

briansrls deleted the session/proud-moth-478 branch July 5, 2026 06:34

gunbai-bot Bot mentioned this pull request Jul 5, 2026

§1 roadmap receipts: emitter mass fix landed, cutover #1 pre-reqs, retirement wave-1, CI step-budget posture #6287

Merged

briansrls pushed a commit that referenced this pull request Jul 5, 2026

WIP: Regen cutover #1 (orphaned when proud-moth-478 archived post-#6243…

6217121

…): (1)

This was referenced Jul 5, 2026

Post-#6243 fresh-emit re-measure receipt (#6292) #6292

Closed

Emitter: T? returns emit Option<Rc<T>> (fn-sig lowering); kills the 1013-error E0308 class #6295

Merged

This was referenced Jul 5, 2026

Emit use lines for dealiased grounding types (E0425=140 class) #6305

Closed

Fold ci_regen into the ci job as an enforcing step (free a runner slot) #6470

Merged

gunbai-bot Bot mentioned this pull request Jul 20, 2026

Confidence probe Phase 1: Gate 1 (emitter fixed point) + Gate 2 (honest frontier) E-code histograms #6928

Closed

5 tasks done

briansrls mentioned this pull request Jul 23, 2026

Emitter: declare libc in emitted Cargo.toml (fresh-emit cargo-green) #7105

Closed

Conversation

gunbai-bot Bot commented Jul 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

CI note

Test plan

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Discriminating test proof (ask 1)

Sequencing (ask 3)

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

gunbai-bot Bot commented Jul 5, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

gunbai-bot Bot commented Jul 4, 2026 •

edited

Loading