Files
AILang/bench/orchestrator-stats/2026-05-18-iter-embedding-abi-m2.1.json
T
Brummel fbeeadeba5 iter embedding-abi-m2.1 (DONE 5-9): swarm-safe ctx ABI complete + sanitiser-verified
Completes Embedding ABI — M2.1 (Tasks 1-4 committed c9a84b3 as the
known-good subset; this commit lands Tasks 5-9 from the re-dispatch
against the Boss-corrected plan).

- T5: swarm.c per-thread-ctx scalar-swarm capability demo (no
  SHARED_CTX) + rc_accounting_shared_ctx_is_flagged_by_tsan added to
  the item-1 harness driver (de-globalisation negative-control teeth,
  where a shared ctx genuinely races on ctx->alloc_count).
- T6: M1 host.c migrated to the ctx ABI; embed_e2e green (s==25
  through the ctx-threaded forwarder).
- T7: DESIGN.md current-state update (provisional-until-M3 narrowed
  to the value/record layout; Decision-10 per-thread-ctx note);
  docs_honesty-pinned sentence + (con Int) snippet preserved verbatim.
- T8: regression gate — workspace green; null-ctx fallback
  byte-behaviour preserved.
- T9: bench-trio — compile_check/cross_lang clean; check.py
  tracked-P2 noise only, causally exonerated (zero codegen/runtime
  diff vs HEAD), NO baseline ratified.

Boss-verified directly (not on report trust): docs_honesty_pin 5/0,
design_schema_drift 8/0, embed_export_hash_stable 1/0,
embed_export_gate 6/0, embed_e2e 1/0, embed_staticlib_alloc_guard
2/0, print_no_leak_pin 1/0 + e2e rc_stats 4/0; the tsan coherent-stop
proven by hand — per-ctx clean (exit 0, 0 warnings), shared-ctx a
real ThreadSanitizer data race at rc.c:157/:208 (exit 66).

Known debt (journal + roadmap): DESIGN.md '## Embedding ABI (M1)'
header still says (M1) while the body mirrors M2 — out of Task 7's
scope-limited regions by design; post-audit doc-honesty tidy.

Milestone-close audit next.
2026-05-18 17:31:50 +02:00

25 lines
2.4 KiB
JSON

{
"iter_id": "embedding-abi-m2.1",
"date": "2026-05-18",
"mode": "standard",
"outcome": "DONE",
"tasks_total": 9,
"tasks_completed": 9,
"reloops_per_task": { "1": 0, "2": 0, "3": 0, "4": 0, "5": 0, "6": 0, "7": 0, "8": 0, "9": 0 },
"review_loops_spec": 0,
"review_loops_quality": 0,
"blocked_reason": null,
"blocked_task": null,
"redispatch": true,
"redispatch_task_range": [5, 9],
"first_dispatch": {
"outcome": "PARTIAL",
"tasks_completed": 4,
"blocked_task": 5,
"blocked_reason": "spec-ambiguous",
"resolution": "Boss spec-defect repair: swarm.c shared-ctx negative control is structurally unable to race for a non-allocating scalar kernel (@__ail_tls_ctx is __thread; the kernel writes no ctx field). Spec amended + plan Task 5 restructured: swarm.c = per-thread-ctx capability demo only; the de-globalisation negative control moved to the rc-accounting harness where ctx->alloc_count genuinely races. Plan Task 3 expected-string transcription error (@ail_backtest_step -> @ail_embed_backtest_step_step) also Boss-corrected.",
"committed_at": "c9a84b33b3e7806491153a5bd3037be1b8ef9d31"
},
"notes": "Re-dispatch Tasks 5-9 all reached approved quality with zero spec/quality re-loops. Task 5: swarm.c per-thread-ctx capability demo (no SHARED_CTX) GREEN + rc_accounting_shared_ctx_is_flagged_by_tsan negative control GREEN (shared-ctx genuinely raced, teeth proven non-vacuous). Task 6: host.c ctx-ABI migration (confirmed expected RED first, then s==25 GREEN); embed_e2e.rs unchanged by design (already links libailang_rt.a by absolute path). Task 7: DESIGN.md M2 current-state + Decision-10 note; pinned bare-scalar sentence + (con Int) snippet preserved verbatim, docs_honesty_pin/design_schema_drift green unmodified. Task 8: full workspace 0 failing suites (DONE_WITH_CONCERNS: embed_export_gate is 6 tests not plan's 8 - pre-existing plan-number artefact, gate unchanged). Task 9: compile_check + cross_lang exit 0; check.py exit 1 on 4 tracked-P2 noise metrics, causally exonerated by source-level invariant (zero codegen/runtime files in working tree => bench binary byte-identical to HEAD), no baseline ratified, plan's git-stash recipe deliberately skipped to protect the uncommitted working tree on re-dispatch. Carried-forward concerns: Task 3 plan-transcription error (Boss-corrected); '## Embedding ABI (M1)' header now describes M2 state (out of task scope, journaled as known debt)."
}