Exemplar ov_SC06_018 @0x80184c74 (matched), 3 members. Per-sibling whole-binary byte-gate (G3/P9) is the
arbiter: jtbl_carve -> make extract -> remap_hseq -> make build, keep iff byte-identical else
revert. Tally: === func_80184C74: {'BANKED': 3} ===
Committed per family because jtbl_family_bank's per-sibling revert restores from HEAD — an
uncommitted prior family would be lost. Campaign-end R22 verifies the fleet.
Exemplar ov_SC06_020 @0x80180b04 (matched), 5 members. Per-sibling whole-binary byte-gate (G3/P9) is the
arbiter: jtbl_carve -> make extract -> remap_hseq -> make build, keep iff byte-identical else
revert. Tally: === func_80180B04: {'gate-fail': 3, 'BANKED': 2} ===
Committed per family because jtbl_family_bank's per-sibling revert restores from HEAD — an
uncommitted prior family would be lost. Campaign-end R22 verifies the fleet.
Exemplar ov_SC02_028 @0x801884d8 (matched), 15 members. Per-sibling whole-binary byte-gate (G3/P9) is the
arbiter: jtbl_carve -> make extract -> remap_hseq -> make build, keep iff byte-identical else
revert. Tally: === func_801884D8: {'BANKED': 15} ===
Committed per family because jtbl_family_bank's per-sibling revert restores from HEAD — an
uncommitted prior family would be lost. Campaign-end R22 verifies the fleet.
Exemplar ov_SC01_077 @0x80178d40 (matched-ov077), 3 members. Per-sibling whole-binary byte-gate (G3/P9) is the
arbiter: jtbl_carve -> make extract -> remap_hseq -> make build, keep iff byte-identical else
revert. Tally: === func_80178D40: {'BANKED': 2, 'isolate-fail': 1} ===
Committed per family because jtbl_family_bank's per-sibling revert restores from HEAD — an
uncommitted prior family would be lost. Campaign-end R22 verifies the fleet.
Exemplar ov_SC02_027 @0x8018A564 (matched), 22 members, 125 ins each. Per-sibling whole-binary
byte-gate (G3/P9) is the arbiter: jtbl_carve -> make extract -> remap_hseq -> make build, keep iff
byte-identical else revert. Tally: {'BANKED': 21, 'isolate-fail': 1}.
Committed per family because jtbl_family_bank's per-sibling revert restores from HEAD — an
uncommitted prior family would be silently destroyed mid-sweep (the tool refuses a dirty tree for
this reason). Campaign-end R22 verifies the fleet.
NOTE on counting: the jtbl carve creates NEW split files, so a git-diff INCLUDE_ASM tally
over-reports (removals visible, re-additions inside untracked files not). True count settles
against the stub oracle at the campaign-end R22.
scope_data_externs §8d drops the draft's decl of any symbol the TU already declares at file scope.
It keys on the SYMBOL, but a §37 asm-label ALIAS binds a DIFFERENT C identifier to that symbol:
the TU declares `D_801851BC`, it does NOT declare `tbl_D_80187044`. Dropping the alias left the
body referencing an undeclared name, which cc1 reports with no `error:` prefix — so the sweep
classified all 132 siblings as CC1-FAIL(no-diagnostic), i.e. as a codegen wall.
The bitter part: the alias exists PRECISELY BECAUSE the TU declares that symbol with a conflicting
type (a `void (*[])(void)` dispatch table vs this function's 20-byte-stride view). The drop rule
fired on exactly the declarations written to survive it. Why 1 of 2 died was fully determined:
tbl_D_80187048's symbol is not in the TU, so it demoted normally.
Fix: is_asm_alias() — an alias is demoted into the body, never dropped (the identifiers differ, so
it cannot collide with the TU's decl). Control-tested 6 ways incl. self-labels and plain externs.
Measured: func_80132018 3/135 -> 135/135; full re-sweep +16 more. Total +148 members.
R22 clean-fleet 213 passed / 0 failed of 213. tools-health OK, dedup-check 1949/0.
Fleet 96.11 -> 96.15% fn-count, 87.8 -> 87.9% distinct; stubs 14,120 -> 13,972 = -148 (2nd oracle).
CORRECTION TO MY OWN CLAIM (R14): after the probe I said the 58% aggregate was concealing a broad
problem. The re-sweep refuted it — only 16 more banks fleet-wide. The alias class really was one
family; the first read ("outlier") was right and the correction was wrong.
875 sweep failures classified: 231 PLUMBING-other, 141 DIFF (real divergence, only 16%),
136 `conflicting types for cdFileLocTable` (ONE symbol — biggest single class left),
77 CC1-FAIL(no-diagnostic), 26 memcpy, 12 D_80114F24, 11 D_800AE620, 9 D_800183E0.
STILL UNFIXED, and the most dangerous instrument left: the sweep's failure classifier greps for
`error:`, which gcc-2.7.2 never emits on hard errors. Every hard error therefore reads
CC1-FAIL(no-diagnostic). That is how a missing declaration looked like a codegen wall across 132
functions. rtu_match was fixed for this at T0(b); this classifier was not.
family_sweep --hseq --band all -j 8 over every matched-exemplar family: 553 families /
203 overlays / 1,419 banked / 1,023 failed (58%). R22 clean-fleet 213 passed / 0 failed of 213.
tools-health OK, dedup-check 1949 validated / 0 failed.
Fleet: 93.9 -> 94.2% instr / 87.2 -> 87.8% distinct / 95.72 -> 96.11% fn-count.
Second oracle (R34): INCLUDE_ASM stubs 15,542 -> 14,120 = -1,422, equal to the diff-derived net
(1,451 removed - 29 re-added = 1,422 = 1,419 sweep + 3 probe). Three independent counts agree.
B -> C -> P IS ONE CHAIN, NOT THREE WINS. 1,102 of the 1,422 landed in ov_SC02_037 (409),
ov_SC03_107 (364), ov_MAIN_012 (329) — the three newly-onboarded binaries from C, which had never
been wired into the shared-body ecosystem, so every matched exemplar was unreachable from them.
B fixed the declarations, C wired the include, P poured through the opening. A repeat sweep will
NOT pay like this; the opening was one-time.
S47 total: 1,481 functions banked with zero agent drafting, all from removing plumbing.
Two findings recorded, neither fixed (deliberate, costed):
- --band defaults to `substantial`: the first probe returned a confident {"families": 0,
"banked": 0} on a real 135-member `mid` family. Always pass --band all.
- The alias-gather defect: probe on 0x80132018 banked 3/135, all 132 failures classified
CC1-FAIL(no-diagnostic) because gcc-2.7.2 emits no `error:` prefix. Real error is
`tbl_D_80187044' undeclared` — the exemplar declares TWO §37 asm-label aliases and uses both,
family_remap carried one. T7-S1's "gather" class. Measured as an OUTLIER (aggregate 58%),
which is why the sweep ran before the fix.
Refused by design, all named: 50 jr families / 183 member-slots (§53 interlock — it printed its
own coverage and reason), 264 STRUCT, 112 unresolved immediates, 3 not-stub.
C, unblocked by B's declaration conform. 23/23/16 banked across ov_SC03_107, ov_MAIN_012,
ov_SC02_037 — the first non-zero result on this population (S46 got 0/142, then 0/129).
R22 clean-fleet: check-all 213 passed / 0 failed of 213. tools-health OK.
dedup-check 1949 validated / 0 failed, C1 coverage 249295 (= 249233 + 62, independent
confirmation of the count). Fleet 93.9% instr / 87.2% distinct / 95.72% fn-count.
THE TOOL REPORTED "BANKED 0 / 129" AND WAS WRONG. gate_stage's ladder hands the same
--verified-out path to harvest_verify on every rung, and each rung opens it for write: stage 0
banked 23 and wrote them, then a later rung that banked nothing truncated the file to 1 byte.
The in-memory list uses += and stayed correct, which is why the JSON verdict listed all 23 names
while the file said nothing. dedup_extend read the file, printed BANKED 0, and took its
`if not banked:` branch — skipping add_members_surgical, so the registry was missing 62
memberships for functions already spliced in and byte-verified.
- Registry repaired by deriving the banked set from git diff (+DEFINE_func_*), not from the
broken file. Post-check: 0 missing.
- ensure_include_revert did NOT fire (added_include False, include already present) — the
P29-S19 defect that once stripped a load-bearing include from 135 binaries stayed closed.
- gate_stage now writes verified_out once at the end from the accumulated truth.
Caught only because bank truth is derived from source (§55b), never from the gate report.
Residue (67) is consistent with the symbols B deliberately left: memcpy 17, ApplyMatrixSV 12,
gte_SetRotMatrix 4, plus 21 CC1-FAIL and 3 DIFF. Not separated: how much of the 62 is B's
conform vs the ladder's own recovery rungs.
Task B, re-scoped from evidence. The 129 dedup_extend failures are 106 conflicting-types /
21 CC1-FAIL / 4 undefined-ref / 3 DIFF — real byte divergence is 2%, and memcpy is 17 of 106,
not the story. Direction reversed too: the byte-true DEF of func_80128ED8 is what the target
.c files already declare; engine_core.h's macro-local extern was the stub-era guess.
Conformed 8 axes to byte-truth (func_8012F14C 2843, func_8012E5CC 2052, func_8012F038 2214,
func_8014C568 1816, func_80128ED8 1524, func_8012C750 406, func_8012C0EC 50, func_80144A04 25).
R22 clean-fleet: check-all 213 passed / 0 failed of 213. Zero functions banked by design.
Tooling (R33/R35) — three guards that asserted completeness over a narrowed population:
- NEW tools/macro_draft.py: a deduped fn has no definition in any .c (body lives in a DEFINE_
macro), so conform_decls had been refusing the largest class it was built for.
- conform_decls skipped engine_core.h wholesale as "a defining TU": 10 stale externs survived
while 1,514 fleet sites moved, and it still printed "axis complete". Skip now scoped to the
defining macro's span.
- Return-axis compare was literal: typedef int/s32 and a missing `extern` faked a return change.
Now compares normalized types.
- §85 consumer scan under-reported (the dangerous direction): a cast between `=` and the call
hid `s0 = (s32 *)func_80144A04(...)`. Now classified by position, validated both ways.
Corrections to my own predictions (R14): the documented scalar-narrowing hazard was benign
across 2,052 sites; the breaks were arity (6 call sites, fixed with §17a-1 fn-ptr casts) and
the consumer-guard gap. A header-only first probe broke ov_SC01_000 — §85 is literal.
Not done, named: memcpy (builtin codegen), ApplyMatrixSV (no DEF), gte_SetRotMatrix (link bug),
func_80147364 (unparseable macro), D_800AE620/D_80126CC4 (data axis). Cookbook §159.
- BANKED: 11 functions at 400-952 ins from the cascade (func_8017D898 952, func_8017CE58 733,
func_801902EC 673, func_8018C2D8 673, func_8018A8D4, func_8017C6F4, func_800CBB38,
func_800CF3A4, +3). check-all 213/213 from a clean tree. 6 near = jr/switch (§53 separate
banking step), 1 failed. The cascade agents wrote 6 new cookbook sections incl. §158.
⚠️ tools-health UNVERIFIED at commit (stale cookbook index fixed, confirming re-run
interrupted) — run it first next session. check-all is the byte oracle and it is green.
- WASTE PREVENTION (Drew: "prevent this from ever happening again, however you need to"):
* tools/validate_targets.py (NEW) — names 5 defect classes (NO-ASM / MID-BODY /
OUT-OF-RANGE / ALREADY-DONE / NO-BOUNDARY), exits non-zero.
* WIRED INTO wave_snapshot so it fails closed — every wave passes through there for its .s
files, so no path from target list to spawned agents bypasses validation. Negative-control:
a 3-target bad list is refused with the exact mid-body offset (+72 bytes of 100).
* The cascade `done()` predicate now short-circuits on SKIPPED as well as MATCH. It tested
only MATCH, so a non-existent target fell Sonnet -> Opus -> Fable and three agents each
proved the same phantom absent: ~29 invalid targets x 3 tiers = 87 of 119 agents, ~9.7M
tokens. A tier that cannot act must END the pipeline, not escalate emptiness.
* docs/accelerators.md A9, including that wave_snapshot's own R32 assertion REFUSED that list
(24 of 57 found) and was routed around — the one instrument warning that was right and ignored.
- B RE-SCOPED (S46-10) and deliberately NOT done: the extend blocker is INTRA-HEADER, not
target-side. engine_core.h declares memcpy FOUR incompatible ways across its DEFINE_ macros;
two in one TU collide. NOT a safe cleanup — the in-tree note at ov_MAIN_012.c:14333 records
that `extern memcpy` disables gcc's builtin and turns an inlined block-move into a CALL, so the
declaration CHANGES CODEGEN. Probe one macro in one binary and byte-gate before any sweep.
- C (dedup_extend over the 129) stays blocked on B. Full context for both in the checkpoint.
- RECOVERY PASS A (wave residue): gate_stage over the 3 big-3 draft dirs recovered
6 sites the bare gate rejected — func_80168664 x3, func_80168F40 x2, func_8012B77C
x1. That is 6 of 19 PLUMBING = ~32%, matching the 16-39% range P29 measured. Batch 1
goes 27 -> 33 of 60. R22 213/213 + tools-health green.
- HONEST SIZING (correcting my own claim): ~32% is NOT "a one-time fix for a ~50% draft
loss". It moves the batch loss from 55% to 45%. Real and free; not transformative.
- WIRED (task 9): dedup_extend now gates through gate_stage (the ladder:
canon_resident_calls -> cast_call_sites -> sig_unify -> harvest_verify) instead of
harvest_verify verbatim. Its 142 candidates failed 0/142 with reasons 118 PLUMBING /
21 CC1-FAIL / 3 DIFF — ~1 in 50 a real byte divergence, the rest declaration conflicts
in the TARGET TU, which is exactly what the ladder reconciles.
GATE_NO_ARITY=1 is forced for the child: gate_stage's arity pre-pass writes the
fleet-shared engine_core.h BEFORE the gate and a failing draft can leave that edit
behind — the F1 defect that broke 141 of 213 binaries in S45. The ladder's other rungs
are draft-local. --ladder can be disabled to restore the old path.
- DOCTRINE (step 3): gate every wave with gate_stage, not bare harvest_verify. Batch 1
needed a second manual pass only because I used the bare gate first.
- LEVERAGE METRIC CORRECTED (R14/R35): the behemoth ranking must use LIVE reach
(unmatched sharers), not total sharers. func_80144B9C reads 770 ins x 141 = 108,570
by total, but 138 of those are already banked — its true weight is 770 x 3 = 2,310.
Same x134 over-count the cookbook records in §25; build_wave_args --rank live exists
precisely for this and I used the wrong ranking. Remaining >=400 ins: 57 distinct
functions / 77 live instances / 38,968 ins, reach ~1.35 => ~0.3% instr-weighted.
60-agent wave over the big-3 (ov_SC03_107, ov_SC02_037, ov_MAIN_012), size-routed per §157.
- BANKED: 14 distinct functions, 27 sites. 11 of the 14 landed in 2-3 binaries
independently => shared engine code, so each is a propagation candidate.
- Gate: ov_SC03_107 11/24, ov_SC02_037 6/17, ov_MAIN_012 10/19 = 27/60.
check-all 213/213 from clean; tools-health OK; dedup 1949/0.
- THE CONVERSION GAP, MEASURED AGAIN: match_one claimed 53/60, the whole-binary
gate banked 27. The losses are 19 PLUMBING + 3 CC1-FAIL + 11 DIFF — i.e. ~2/3 of
the loss is declaration plumbing, not wrong code. Same ratio P29 measured
(~92% byte-correct drafts, ~27% banking) and the same class that blocked all
142 dedup_extend candidates tonight. Three independent populations, one wall.
Concentrated: D_800AF634 x6, D_800AE620 x3, RotMatrixX x2 (data-decl class ->
reconcile_decls) and func_80146A6C x3, func_801376E8 x3 (DEF-side signature
class -> canon_sig_reconcile v3.2).
- EFFORT A/B: INCONCLUSIVE, recorded as such. Yield 16/20 (default) vs 17/20
(effort:low) is noise, matching P13-T7 at the other end of the ladder — BUT the
positive control failed to materialise: the workflow journal carries only
agentId/key/type, no per-agent token usage, so "low effort is free" and "the
effort parameter silently inherited" remain observationally identical. Not
written into doctrine. A future test needs an EXTERNAL measure (per-agent
wall-clock or tool-call counts), not the agents' self-reported iteration counts.
- Sonnet's 50-59 ins arm claimed 20/20, contributing the larger bodies
(func_80142BB4, func_8012B77C) — §157 size-routing held at the top of its range.
Drew: "make it more multi-threaded... I still see my cpu idle for far too long."
Measured, fixed, and regression-tested against the S46-3 bank as a KNOWN ANSWER.
- THE MEASUREMENT: 31s saturated (33 makes/48 cc1/load 27) then ~25s with ONE build alive
while 31 cores idled, repeating. Causes: ex.map starts in list order so the giants land
last, and apply/restore is single-threaded.
- gate_all -> gate_failures: return EVERY failure the sweep already computed (~138 rounds -> 1).
- Longest-first gate scheduling; results re-sorted into `changed` order so the verdict stays
bit-identical to the serial loop's.
- PER-OVERLAY INDEPENDENT SEARCH, IN PROCESSES. My first cut used threads and the box refuted
it: 0-4 builds alive at load 3, because the work is regex over 15k-line files and 138
"parallel" searches all queued on the GIL. Same logic in a ProcessPoolExecutor: 14-29 builds,
load 34.75, search phase ~100s. Safe because the shared header is written ONCE by the parent
and each overlay owns its own .c files + build/<bin>/. Seeded with one in-process search
first — a pool submitted at once gives every worker an empty suspect list and makes all 138
pay a full bisection. place_in_overlay extracted to module level so the worker and the
in-process apply cannot drift (R33); compiles_standalone's fixed t.c is per-call now.
- THE REGRESSION (the point, not the stopwatch): revert src/+config to pre-bank, re-run the
identical command -> 29 functions (same), 141 overlays byte-identical, 682s vs ~1440s, and
285 exclusions vs ~350 => +62 MORE member instances (249,161). The old prefix-based
necessity probe was OVER-EXCLUDING (charging 4 fns to 9 overlays that did not all need
them); the per-overlay shrink minimises per overlay. The faster path is also more correct —
a timing comparison would never have shown it. R22 213/213 + tools-health green.
- STILL SERIAL, now the actual wall-clock (neither is a build): ~3min setup before the first
gate (registered_addrs() yaml-parsing a 1949-group/249k-instance registry + 213 sig loads)
and ~2.5min of sequential reconcile_caller_extern after the search.
- Captured as defaults: docs/accelerators.md A8 + memory fleet-tool-parallelism-defaults.
cookbook index regenerated (my §155c append left it stale — the gate caught it, exit 1).
The S45p9 blocker is closed, and the recovery loop that kept it from finishing is rewritten.
- BANKED: dedup_propagate --auto-from ov_SC02_037 --recover -> 29 functions propagated,
141 overlays byte-identical, dedup 1920 -> 1949 groups, member instances 246,284 ->
249,099 (+2,815). make clean && extract-all && check-all -> 213 passed / 0 failed (R22).
- WHY IT FINISHED THIS TIME: gate_all -> gate_failures returns EVERY failure from the sweep
that already computed them, and the recovery loop resolves them all per round. Converged in
3 rounds; the old one-overlay-per-sweep design needed ~138. That reframes the S45 run — it
was not nearly done when it died, it had barely started.
- Batching did NOT cost capability: per-overlay necessity probes excluded four of the nine
culprits from only the 9 overlays that needed it (not all 138), and ov_SC07_006 was
RECOVERED by the Part-B caller-extern reconcile instead of excluded.
- Plan phase parallelised: 5 min -> 26 s, plan + skip classification byte-identical. Its
compiles_standalone temp file is per-call now — the fixed `t.c` was the same fake-isolation
class as match_one's shared --work dir (P28 T5), latent until something ran it in parallel.
- docs/accelerators.md (NEW, Drew 2026-08-07): the reusable-workflow ledger — what we learned
late that a future decomp should know on day one, each entry with when we found it, when it
WAS findable, what it cost, and the honest prerequisite where one exists.
STAGE 1 of docs/concurrency-design.md, landed and negative-control proven.
tools/shared_lock.py (NEW) — one reader/writer flock over the FLEET-SHARED state
(src/shared/*, config/overlays.mk, config/dedup.us.yaml, the overlay .c files
propagation rewrites). Per-binary resources keep gate_stage's existing per-binary flock.
- gate_stage takes it SHARED when the gate writes no shared state, EXCLUSIVE when it
does (propagate, or the arity pre-pass enabled) -- so distinct-binary gates still run
concurrently but can never overlap a writer.
- dedup_propagate and fix_arity_callers --apply take it EXCLUSIVE.
- NESTING-AWARE: gate_stage SPAWNS both writers, so a naive child lock would deadlock
against the parent. The holder exports BFM_SHARED_LOCK_HELD and children inherit.
NEGATIVE CONTROLS (all pass):
NC1 a held SHARED lock refuses a non-blocking exclusive writer, loudly, naming the lock
NC2 parent-holds/child-inherits does NOT deadlock (the real risk in this design)
NC3 two readers acquire concurrently (0.00s) -- phase-B parallelism preserved
bulk_harvest docstring CORRECTED: its "propagation is the ONLY writer of the shared
engine_core.h" claim was FALSE as written and had been asserted for phases (F1 -- the
arity pre-pass writes it from inside every worker). Now states what is actually true,
under which two conditions, plus the one-line assertion that detects a violation.
BANKS: wave-3's drafts re-gated on a CLEAN tree -> 15 of 20 banked. The same drafts
previously reported 0 banked / 20 near -- that verdict was 100% an artifact of the
broken tree, which is why they were held as UNJUDGED rather than accepted as failures.
check-all 213 passed / 0 failed. F1 bracketing assertion CLEAN.
Session total banked: 44 functions + func_8015C030 propagated x7.
ROOT CAUSE of the 141/213 breakage earlier this session (correctly derived this time;
my first attribution to F1 was WRONG -- no arity journal ever touched func_80146A6C and
the arity undo reported success):
dedup_propagate --recover's Part B reconciles a conflicting caller extern and
DELIBERATELY leaves the edit on disk when it buys the byte-match ("keep the reconcile
on disk"). Correct while the fn survives -- but a fn can still be dropped by a LATER
iteration against a different overlay, and when the plan finally emptied, the
"all candidates dropped" sys.exit fired with NO restore. Reconciles kept for
ov_SC07_001..009 were orphaned: no-proto'd caller externs for functions that were
never propagated -> ov_SC07_010 "passing arg 2 of func_80146A6C makes pointer from
integer" -> 141 of 213 binaries failed check-all.
The byte-gate never mis-banked (it fails closed). The real cost was VERDICT VOIDING:
every subsequent gate reported "near" against the broken tree, so two whole batches
(4/4 and 20/20) were mis-read as draft failures when they measured the tree (R35).
FIX: a reconcile LEDGER. Every kept reconcile is recorded against its fn, undone the
moment that fn leaves the plan, and ALL outstanding reconciles are restored before the
failure exit -- so a failed propagation leaves the tree exactly as it found it.
HONESTY: the fix is IMPLEMENTED AND REVIEWED BUT NOT YET PROVEN. The negative control
aimed at the exact failing propagation SUCCEEDED instead (different tree state), so the
guarded path never executed. A targeted test of the ledger is still owed.
Also lands the propagation that control performed: func_8015C030 x7 overlays
(func_80168B70 excluded from 4 SC07 overlays, survived elsewhere). check-all 213/213.
29 byte-gated matches into ov_SC02_037 (626 -> 597 live stubs), from two Haiku/Sonnet
crack waves on the CORRECTED frontier (live INCLUDE_ASM stubs with cached seeds, not
the already-banked reach-141 shared core).
Gated with BOTH safety guards after a live F1 incident (see below):
GATE_NO_ARITY=1 — no fleet-shared engine_core.h writes
--no-propagate — propagation deferred to its own controlled step
F1 bracketing assertion CLEAN (git status --porcelain src/shared config empty).
R22 clean-fleet: make clean && extract-all && check-all -> 213 passed, 0 failed.
F1 CONFIRMED IN PRODUCTION (docs/concurrency-design.md, found by the Fable5 design
pass hours earlier): gate_stage's arity pre-pass (fix_arity_callers --apply) writes
the fleet-shared header + caller externs; when a draft then FAILS to bank the edit can
survive. func_80146A6C failed its gate yet left a caller signature behind, breaking
141 of 213 binaries (ov_SC07_010: 'passing arg 2 makes pointer from integer'). The
byte-gate held throughout — nothing wrong was banked, it failed closed and loud.
Recovered by revert-to-HEAD + deterministic replay from the on-disk drafts.
Cost of the guard, measured: 2 banks (24 -> 22 on the wave-2 batch).
MODEL-LADDER CALIBRATION (independently re-verified, not agent claims):
Haiku 4-27 ins: 43/50 = 86% (~44k tokens/match)
Haiku 30-49 ins: 14/25 = 56%
Haiku >=50 ins: 5/25 = 20% (~177k tokens/match, 4x worse)
=> the documented 'Haiku <=50' band is optimistic; the cliff starts ~30, collapses at 50.
Agent honesty across 100 drafters: 63 MATCH claims, 63 real, 0 false (one apparent
false claim was MY verification missing --o0 on an -O0-cluster function).
CARRIED: the 29 banks are x1 (propagation off); wave-3's 19 verified drafts are
UNJUDGED — their gate ran against the F1-broken tree, so those verdicts were void.
- the gate DECODED from matched C (func_8012832C case 0x300E -> func_80128998 -> streaming
API with &cdFileLocTable[144]) -> scene arithmetic named the 1ST-BOSS arena -> ONE targeted
load captured it at 0x801E4C60
- RETRO-VERIFIED: Phase-3's dumps/ram_castle.bin (2026-06-14) holds it at the SAME address,
same 6,764-B exact prefix — R10 two independent datapoints two months apart;
bossHp_SteamKnight (0x801E4398) lives inside this module's image
- onboarded md_SC02_009 (id 0x3E, TLO 0x4): BYTE-IDENTICAL first build; fleet 213;
R22 213/213; tools-health OK; audit-disc UNCLAIMED 6 -> 5, residue 0
- the last 5 (MAIN/7, MAIN/9, SC03/53/54/56) reclassified emulator->STATIC-RE targets with
decoded leads (memory-map §S45 p3); loc-id map appended to docs/debug-menu-list.txt
- negatives banked: pause menu, memory-box prompt, new-game intro, high/low game, Minku
spawn (slot-A actor 0x15 = md_MAIN_015 candidate naming)
- THE TOUR (Drew driving the retail debug menu; mode-7 hammer over the Redux web API):
all 28 script modules captured live at four byte-verified per-chapter slots
(SC03/73-79 @0x801EF468 ch2-period, SC03/132-138 @0x801E25E8 ch3, SC04/24-30
@0x801E7B28, SC05/23-29 @0x801ED988); the routing law: debug-menu AREA selects the
chapter, each CITY interior streams its own module (member k <-> interior k).
md_MAIN_011/DISELECT byte-proven 24,236/24,240 in RAM; slots A/B/boot R34-verified live.
- MAIN/3 DISCOVERED: the main-menu module (id 0x39, 121,884 B), mis-bucketed as data by
BOTH audit oracles; live byte-proven @0x800CEDF8 (42,632-B exact prefix); onboarded.
- 29 onboardings BYTE-IDENTICAL on first build -> fleet 212; R22 212/212 after three
md_MAIN_003 catches: the A4 DsMix leak; an extract-order-sensitive splat boundary
(bytes: a 1-word data sentinel in .text + fn at +4 -> pinned in symbols file);
corpus.stubs now treats D_*/jtbl_* INCLUDE_ASM as blob includes (mirrors progress.py)
- module-id census (offline, disc-wide): 77 id-law code payloads, 0 further misses;
SC03/55 = confirmed DATA. audit-disc: UNCLAIMED 34 -> 6, residue 0 — the 6 carry
byte-checked negative evidence; next tier = the CD-read tracer
- docs: memory-map §S45 (slots + routing + debug-menu ops), disc-completeness S45
addendum, decision-log R31 entry, docs/debug-menu-list.txt (Drew's transcription)
- .run/s45 evidence allowlisted (tour logs/scripts/rosters); 104 ram dumps LOCAL-ONLY
- new baseline: 93.8% instr / 95.68% fn / 87.2% distinct over 212
- slot A 29/29 (md_MAIN_013..041 @ 0x800CAE08), slot B 6/6 (md_MAIN_042..047 @ 0x800CCB1C),
boot trio 3/3 (md_MAIN_001 [=MAIN/0 twin], md_MAIN_008, md_MAIN_011 @ 0x800CEDF8) — every one
BYTE-IDENTICAL on its FIRST build (byte-corroborating the §S44 loader table for slots A/B/boot)
- TLO roster derived from the §154 id-word law (.run/s45/derive_tlo.py): 0x4 default;
011=0x7C, 025=0xC, 034=0x80, 039=0xC (first-prologue scan)
- new_binary.sh: module hdr carve is now a dot-typed .rodata PAIRED with the c segment —
a header can hold a function's jump table (md_MAIN_034), and standalone rodata emits
.L locals that don't cross objects; bin links in the data block (both refuted by bytes)
- A4 law: symbols.resident.txt dropped from the boot trio's stacks (windows inside the
resident region; DsMix @0x800D1BD8 had minted a phantom fn boundary in md_MAIN_011) —
re-extracted clean, all three byte-identical, phantom gone
- family_sweep --hseq scoped by --only to the 637 families with a matched exemplar AND a member in
the new 3 (2,232 stageable; avoids re-gating the swept-dry fleet). BANKED 290 / 81 failed /
6 skipped (unresolved immediates), every one whole-binary byte-gated; all three SHAs green.
- Session total into the big 3: 4,836 h_exact + 290 template = 5,126 member-functions.
- Remaining stubs 624+717+710 = 2,051 = the ~802 novel functions x instances + the genuinely
failed/unstageable tier (the new frontier).
- tools/dedup_extend.py over the clean tree (the prior run correctly REFUSED my uncommitted tree,
H4 — that refusal was the tail I misread as a result; and my earlier '0 stubs left' was a
single-file grep -c display artifact, caught before being reported).
- BANKED 4,836 / 5,016 planned (1,612 DEFINE_func per binary; 180 skipped incl. 8/binary
verbose-form). Each splice byte-gated; all three binaries remain BYTE-IDENTICAL (SHA re-checked
per binary post-run). engine_core.h include added -> audit-binaries green again (R36).
- Remaining stubs: ov_MAIN_012 712 · ov_SC02_037 822 · ov_SC03_107 802 = the h_norm-only tier +
the ~802 novel functions (the new frontier).
- Three uncompressed (PAC type-1) overlays at the standard 0x80128158 slot, onboarded via the new
tools/new_binary.sh, each byte-identical at 100% INCLUDE_ASM on the FIRST build:
ov_MAIN_012 d6b3e8b9 (383,783 B, 2,324 fns)
ov_SC02_037 b0c5394a (661,903 B, 2,434 fns)
ov_SC03_107 87d02b57 (474,087 B, 2,414 fns)
This also BYTE-PROVES the statically derived base (the §S44 loader table + the 500:1 h_exact
vote): a wrong vram could not have produced byte-identical images once symbols resolve.
- Fleet: 140 -> 143 binaries. audit-binaries currently FAILS on all three by design (no
engine_core.h include yet — the SC07-blindness check working as built); dedup_extend is the fix
and the next commit.
- Registered by the script: overlays.mk blocks, check.sha, symbols seeds, the 3 BINARIES dicts.
family map regenerated (3,577 target families / 279 with a matched sib — the new binaries'
members now visible).
- R22 CLEAN-FLEET: 140 passed, 0 failed of 140. Fleet 12,502,519/13,160,961 = 94.997% instr
(+18,146 instructions this session, 23 functions). 393 instructions from the 95.000% bar.
- REDO of the S43-9 retraction, done correctly through harvest_verify (splice/build/keep-iff-
byte-identical/revert) instead of hand-building. 5/5 banked, each re-verified three ways:
image SHA == locked SHA, stub gone, real definition present.
ov_SC03_101/func_801814F8 · ov_SC03_104/func_80184934 · ov_SC04_003/func_8017E4F4 ·
ov_SC04_005/func_80181054 · ov_SC04_007/func_8017FF08
- THE DEFECT THIS PROVES: family_sweep --hseq reported this family 0/5 with
"PLUMBING: parse error before 'unsigned'" — but the remapped drafts are byte-CORRECT. The only
`unsigned` in the draft is INSIDE A COMMENT, so a gate-pipeline transform is eating a `/*` opener
and turning comment text into code. Per-transform runs on the draft alone all preserve it, so it
needs the gate's real invocation (--src-file) to reproduce. NOT YET PINNED — and it is silently
costing banks in every sweep it touches. Next: run the three transforms with --src-file and diff.
- Workaround that banked them: carry the exemplar's typedefs by hand (the family_remap _carry_macros
gap, §146/§152) and gate directly, bypassing the sweep's recovery ladder.
- Also killed a self-inflicted infinite poll: an `until ! pgrep -f "permuter_ils.py <fn>"` loop whose
pattern matched its OWN bash command line, so the condition could never go false (spun 2h30m).
Same family as the day's other defects: a check that cannot return the answer that ends it.
- func_8018D98C (ov_SC06_033, 710 ins): MATCH, gated, carved into its own split
(src/ov_SC06_033/ov_SC06_033_jr_8018D98C.c); image matches its locked SHA; stub gone.
NOT a family — `find asm -name func_8018D98C.s` returns exactly one file, so this banks 1x710.
The prompt's "renderer sibling" premise was wrong: it is a 12-state entity state machine over
jtbl_801CF234; func_8017C6F4's C shares nothing with it. Structurally exact on the first draft.
- §153 THE ADDRESS-REMATERIALISATION LAUNDER (third zero-emission asm lever, after §148-C's allocno
numerator and §151's blocked scheduler tick): an `&SYM` used as an argument >=2x in ONE cse basic
block gets its pseudos unified (4 refs), so local-alloc.c:1080's remat path (needs reg_n_refs==2)
never fires and global.c:388 hands it a CALLEE-SAVED register, cascading a rename. 14 probes prove
no respelling reaches it (do/while splits cse1; cse2 puts it back). Cure, zero bytes, one per site
in its own block: `{ s32 _m = (s32)&SYM; __asm__ __volatile__("" : "=r"(_m) : "0"(_m)); f(x,_m,y); }`
— the volatile asm is never entered in cse's table AND sets _m, emptying the equivalence class.
Placement is load-bearing (#APP is a scheduling barrier); with two address args, launder BOTH.
- INTEGRATION CAUTION: the agent's TU-CONFORMED variant gated DIFF while the PLAIN one banked.
rtu_match MATCHing does not promise a decl-rewritten variant survives the real build — gate the
plain variant first.
- R22 clean-fleet still owed (one agent remains on asm/); this and the 12 family banks are
incremental-gated (§130) until it runs.
⚠️ R22 CLEAN-FLEET OWED (two agents still reading asm/, so `make clean` is unsafe). Each of the 12
was gated whole-binary AND independently re-checked against its own config/check.<bin>.sha (12/12),
stubs confirmed replaced — but incremental (§130). Treat as UNCONFIRMED until the clean run.
- THREE isolated cheap-Opus agents, briefed with §150/§151 + the mandatory all-drafts scan,
CONVERGED INDEPENDENTLY: func_8017C6F4's 947-ins body exists in 12 OVERLAYS under 5 DIFFERENT
NAMES at 6 DIFFERENT ADDRESSES, each differing by exactly TWO per-overlay symbols (screen-rect
helper + 64x64 cell table). Gated 12/12, 0 failed. 11,364 ins from this morning's single crack.
- WHY IT HID ~30 PHASES (cookbook §152): name-keyed grouping scattered it across 5 names,
address-keyed across 6 addresses (and the address collides with an unrelated 15-ins body in 3
other overlays), and h_seq-keyed scattered it too — which is why the Phase-26 sweeps missed it.
THE KEY IS BYTE SIZE: `grep -rl 'nonmatching .*, 0xECC' asm/*/nonmatchings/*/` returns exactly
the 12, reads the asm (cannot go stale like family_hseq.json), no false positives. Refines the
Phase-26 "h_seq is spent" finding: h_seq is worth exactly ONE size-keyed sweep behind each FRESH
core crack — here it paid 11:1.
- TWO CAUTIONS THAT TRAVEL WITH IT: (1) a MASKED tool cannot validate a remap — match_one and
rtu_match both mask jal/%hi/%lo, exactly the fields a remap edits, so a wrong symbol map still
reports MATCH; gate remaps by the whole-binary SHA only. (2) a stale residual is NOT evidence two
functions differ — I briefed "func_8017C59C scores 340, different body"; refuted in one command
(that 340 came from a pre-§150-fix draft, which scores nonzero against its own target too).
- OPEN TOOL DEFECT (R32): family_remap's unit backscan halts at the first #define, so it carried
16/16 gte macros and 0/10 typedefs, silently — the §146 gap from the other side.
- MY ERROR, RETRACTED IN THE LOG (S43-9): I reported the 263x5 cluster as "5 byte-identical, 1,315
ins". FALSE — the drafts had been reverted, so I measured the INCLUDE_ASM STUB BASELINE, which is
byte-identical by construction. R34's trap, self-inflicted by hand-building instead of using
harvest_verify. Nothing was banked there; the cluster is UNRESOLVED. ("41 behemoth drafts" was
likewise a file count — 79 files, 20 distinct functions.)
- func_8017EF68 MATCH 969/969, re-verified by me, gated: ov_SC06_000 byte-identical at da4a26ff.
- MECHANISM (from cc1's own -dR trace, not inferred): the r3000 machine description gives the
memory unit load-ready-cost 2 / store 1, so blockage(load,store)=2 — a LOAD CAN NEVER BE PICKED
IN THE TICK IMMEDIATELY AFTER A STORE PICK. sched2 therefore always wedges one ready ALU insn
between the lw and the sh, and the target's zero-wedge order is UNREACHABLE BY ANY STATEMENT
ORDER. That is why ~20 documented hand variants AND the repaired permuter both floored at 2.
The draft's own §49 sched1-LUID story was incomplete — real but secondary.
- THE LEVER (cookbook §151, "the ghost wedge"): a zero-emission tied in/out asm
`__asm__("" : "=r"(v) : "0"(v), "r"(rival));` — 0 bytes, but a schedulable insn that absorbs the
blocked tick, and it sets reg_n_sets(v)=2 which also kills sched1's birthing boost (one
instrument, both passes). Two measured fallouts: rival-read in the same asm (22->12), then a
second re-tie on a HIGH-REF host to restore allocno live-length parity (each in-loop insn is +1
live length for every loop-spanning allocno; a trio of invariant addresses sat exactly on
allocno_compare's integer-floor boundary). Host choice empirical: pkt=MATCH, ot=705, double=10.
- ✅ R22 CLEAN-FLEET: make clean && extract-all && check-all -> 140 passed, 0 failed of 140.
This DISCHARGES the [R22 PENDING] caveat on commit:1484 — all five banks are confirmed, not
incremental-build artifacts (§130).
- FLEET: 96.63% fn-count / 94.9% instr-weighted (12,489,130/13,160,961) / 89.2% distinct-code;
0 NON_MATCHING (G4); dedup 1919 groups. Session +4,757 ins from 2 cracks x 5 binaries.
Distance to P30's 95% instr bar: 13,782 ins (was 18,539 at session start).
⚠️ R22 CLEAN-FLEET VERIFY IS OWED, NOT DONE. All four gates below were INCREMENTAL builds
(§130: an incremental build can report BYTE-IDENTICAL for a change a clean build cannot link).
Committed now only to protect the work — a second Fable5 agent is reading asm/, so `make clean`
would destroy its inputs mid-run. The clean-fleet run follows the moment that agent finishes;
treat these four banks as UNCONFIRMED until then.
- THE CRACK (Drew approved the Fable5 escalation, R27): byte-exact, PIN-FREE, 947 ins. My §147-E
"qty_compare tie, unreachable from source" diagnosis was WRONG. The residual was VARIABLE
IDENTITY: (1) the X-pass and Y-pass min/max intermediates are DIFFERENT variables (8, not 4
reused); (2) mnc/mxc do not exist — the cell clamps reuse the prim-loop mn/mx (X) and mny/my (Y).
Ablations: split-only 63, reuse-only 624, conjunction MATCH. That is also why S42's "separate
X vs Y variables" probe was filed as a failure (it was half the fix), and why every allocator
lever was inert — pins, §148-C sliders, declaration order and 14 permuter restarts cannot reach
a draft with the wrong NUMBER OF PSEUDOS.
- VERIFIED INDEPENDENTLY BEFORE BELIEVING IT (R14): I re-ran match_one -> MATCH (947 ins), then
the whole-binary gate per binary.
- BANKED ×4 (every 948-ins sibling of this body), each byte-identical:
ov_SC03_126 c48a8bb8 · ov_SC03_003 898bf52a · ov_SC04_021 33614234 · ov_SC05_019 3f5b4f13.
family_remap produced all three siblings cleanly.
- §146 SEEN AGAIN: all three siblings first failed with `PLUMBING: parse error before 'MTX_C6F4'`
— _carry_macros carries #defines but NOT typedefs; prepending the 9 typedef lines fixed all
three. That label is legible ONLY because of this session's classifier fix; before it, it read
"CC1-FAIL: make: *** Error N" and cost a manual splice-and-rebuild each.
- cookbook §150 (decode register ownership from the MATCHING diff regions before touching the
allocator; per-instance register asymmetry ⇒ per-instance variables; the deleted-self-move tell
and the global.c:719-vs-:729 death-before-store exemption behind it). §147-E corrected: it named
the wrong allocator — these are global.c allocnos, not local qty_compare quantities.
Answering "did we do S4?" honestly: NO, not properly. The earlier pass re-gated only the NEWEST
stored draft per draft-exemplar head (8 banked of 35). S6 then proved that is sampling, not scanning
-- its giant's match was the 9th of 31 drafts, and my first pass had reported "closeness 40".
Redone with EVERY stored draft run through match_one, over the 39 draft-exemplar heads + Drew's
named large-function list (38 targets, 33 with drafts on disk):
14 of 33 targets MATCH from a stored draft (some had 51-57 drafts each)
-> 6 banked first pass, +1 after recover_giant = 7 banked
-> including func_8018057C (897 ins), which was on the "needs an agent" list
The 14 came overwhelmingly from ov_SC01_077 -- exactly the heads where only the newest draft had
been tried. The winning drafts sit in .run/_a10_sample-cn-cast-rc/, .run/drafts-wave-cn-cast/,
.run/drafts-wave-cn/, .run/ab-exp/opus-cn/, .run/backlog_drafts/ -- i.e. spread across many
historical pipelines, which is precisely why "newest" is the wrong selector.
7 still open after recovery (5 near, 2 failed) -- integration classes, drafts kept in .run/s41/rec/.
VERIFIED: make clean && make extract-all && make check-all -> 140 passed, 0 failed of 140.
Fleet 12483035 -> 12484373 instr; distinct +1,338 / +7 uniq; fn-count +7. instr-weighted 94.9%.
audit-digest OK. 0 NON_MATCHING (G4).
STILL OPEN from S4: the 263x5 cluster (0x80182fd4 exemplar) sweeps 0/5 with `parse error before
'unsigned'` in the spliced draft -- NOT the missing-type class, undiagnosed, do not assume codegen.
And the 2 resident stubs with gate-rejected match_one-MATCH drafts remain untouched.
THE RULE (cookbook §146, now paid for twice): SCAN every stored draft, never sample. A head with 57
drafts has 57 chances, and the pipelines that produced them differ in ways that matter.
The two functions the roadmap has carried as PERMANENT WALLS since Phase 24 are matched in all 138
overlays. Neither needed a siege. Both matched from drafts ALREADY ON DISK.
func_80178004 165 ins x 138 = 22,770 Phase 26: Fable5, ~477k tokens, "intrinsic 3-integer
regalloc wall". THREE stored drafts report match_one
MATCH today; one banked first try, no new work.
func_801412A8 198 ins x 138 = 27,324 close=29/110 since Phase 24. Matched from 1 of 31 stored
drafts + the §37/§124 alias.
WHY func_801412A8 LOOKED INTRINSIC (worth understanding — match_one is structurally blind to it):
the TU declares `extern int func_801412A8(int,int,int,int,int,int)` and its callers USE the return
(`param_1 = func_801412A8(...)`), while the byte-true definition is
`Prim_1412A8 *(Prim_1412A8 *, int, int, int, u16, u16)`. Narrow params cannot agree with an `int`
prototype and the no-prototype escape is illegal once a param promotes, so NEITHER side can move --
and the resulting byte difference is in the CALLERS, which match_one never compiles. The §37/§124
def-side asm-label alias decouples them: the TU decl keeps governing the call sites (codegen
untouched), the definition keeps its byte-true signature.
THEN PROPAGATION RETURNED 0/137 TWICE, both times a missing TYPE, not codegen:
family_remap's `_carry_macros` carries file-scope #defines but (a) NOT typedefs, and (b) is NOT
TRANSITIVE -- it brought addPrim_1412A8 and stopped, though that macro calls setaddr/getaddr and
getaddr casts to PTag_1412A8. Lifted Env_1412A8 / PTag_1412A8 / Prim_1412A8 + OT/getaddr/setaddr
into src/shared/engine_types.h (inside the include guard) -> 137/137, 0 failed.
MY ERROR, CAUGHT BY THE GATE: I lifted the typedefs but did not STRIP them from ov_SC01_077.c, so
they were declared twice and gcc-2.7.2 rejects a repeated typedef even when identical -- the lesson
already recorded at the foot of engine_types.h. R22 came back 139/140 with [FAIL] ov_SC01_077 (the
exemplar's own overlay). Stripped, re-verified, 140/140. A proper lift strips the source;
build_engine_types --strip does both and I did it by hand.
Also a measurement error worth recording: I checked whether the draft defined Prim_1412A8 with a
plain `grep -c` -- which matches inside `addPrim_1412A8` -- and briefly concluded the carry worked.
Substring false positive; the same shape as reading a `return` as a declaration.
VERIFIED: make clean && make extract-all && make check-all -> 140 passed, 0 failed of 140.
Fleet 12432941 -> 12483035 instr (+50,094 -- EXACTLY the two giants x138); fn-count +276;
instr-weighted 94.5% -> 94.8%. audit-digest OK. 0 NON_MATCHING (G4).
THE RULE THIS BUYS: re-measure a wall before respecting it, and SCAN every stored draft rather than
sampling (my first pass checked 8 of 31 and reported "closeness 40" for a function whose MATCH was
in the 9th). Four minutes of re-measurement was worth 50,094 instructions.
Propagation behind every crack, same session (the multiplier the waves exist for). 19 newly-banked
exemplars from waves 1+2, all in the family_sweep lane (0 has_mid_jr):
87 candidate members / 10,212 ins -> 61 BANKED / 26 failed across 39 overlays
The 26 that did not bank are the known plumbing shapes, not codegen: 20 CC1-FAIL + 5 callee
`conflicting types` (func_8017EFA0 x3, func_8012B23C x2) -- the same classes the S40 recovery ladder
already has levers for (§17a-1 no-proto + call-site cast; recover_giant block-scoping). Left open
deliberately rather than force-banked (P9); they are the cheapest fuel on the board next session.
TOOLING GAP RECORDED: the sweep's classifier writes "CC1-FAIL: make: *** Error 33" WITHOUT the actual
cc1 message, so 20 of 26 failures carry no actionable reason. Diagnosing one currently requires
manually splicing the draft into its TU and rebuilding (done twice this session). The classifier
should capture cc1 stderr the way harvest_verify already does -- worth fixing before the next big
sweep, or every CC1-FAIL costs a manual reproduction.
VERIFIED: make clean && make extract-all && make check-all -> 140 passed, 0 failed of 140.
Fleet 12425854 -> 12432941 instr (+7,087); distinct +6,145 / +51 uniq; fn-count +61.
instr-weighted back to 94.5% ON THE HONEST (post-main-regen) denominator of 13,160,961.
audit-digest OK. 0 NON_MATCHING (G4).
Two ultracode waves over the open-only h_norm clusters (the pool nobody had ever aimed a wave at),
pool VERIFIED from the sigs first (R14).
wave 1 8 targets 8/8 match_one 5/8 gate first pass -> 8/8 after recovery
wave 2 16 targets 16/16 match_one 14/16 gate first pass -> 16/16 after recovery
THE HEADLINE IS NOT 24/24 -- IT IS THAT NOT ONE FAILURE WAS CODEGEN. All six first-pass gate
failures were TU-integration plumbing, each with an already-documented lever:
func_801802EC redefinition of morph_lerp strip the §77 PROBE LAYER (the draft carries types +
a static inline so match_one can compile standalone;
the real TU already defines them -- scaffolding is
not part of the bank)
func_8018B238 conflicting types D_80115158 recover_giant: draft declared it file-scope as a
struct array, TU declares u8[] BLOCK-scope inside
other functions -> block-scope the draft's externs
func_8017EF54 conflicting types (SELF) §37/§124 def-side asm-label alias (TU declares
void f(void) for no-arg callers; byte-true def takes
s32 in $a0; no-proto escape illegal once a param
promotes)
func_80183D78 conflicting types (callee) recover_giant
func_8017F278 conflicting types func_80146C3C §17a-1: the fleet canonical is the NO-PROTOTYPE
form + the intended signature applied AT THE CALL
SITE; a concrete prototype collides with it
(wave-2's 14 first-pass banks needed nothing -- the wave-1 lessons were folded into the prompt)
=> the gate number measures INTEGRATION, not matching. Run the recovery ladder before recording a
wave's yield or the metrics under-report the drafters and send the next wave hunting walls that are
not there. docs/wave-metrics.md S40-1.
POOL VERIFICATION (R14, and it cut both ways): the frontier report's cluster pool MEASURED
1,677 clusters / 5,795 fns / 319,755 ins at a 3.68x multiplier vs its claimed 1,689 / 5,956 /
326,261 at 2.7x -- within 2-4%, and the multiplier is BETTER than claimed. The SAME document's whale
claim was 3/4 wrong. Verify each claim separately; do not accept or reject a source wholesale.
ALSO: 24/24 members propagated from wave 1's 5 banked exemplars (0 failed) -- the same machinery
that returned 0/39 before this session's cast_call_sites fix.
NEW IDIOMS, distilled in-session (R16/R30):
§144 the LITERAL'S SPELLING picks the immediate encoding (`cnt + 0xff` vs `cnt - 1`: mod-256
identical, both one addiu, but gcc emits 0x00FF vs 0xFFFF from the source text)
§145a combine_givs ANCHOR RULE -- the address-giv group anchors on the LAST address-giv in SOURCE
order (record_giv prepends, combine_givs takes the head); store order decides the base and a
wrong choice spawns a third induction register
§145b a bare `p = r;` is a COMBINE BARRIER (can_combine_p/use_crosses_set_p) -- it preserves a
pointer-bump addiu that combine would otherwise fold into every MEM offset
§145c chained assignment `a=b=c=0` emits stores RIGHT-TO-LEFT
VERIFIED: make clean && make extract-all && make check-all -> 140 passed, 0 failed of 140.
Fleet 12420375 -> 12425854 instr (+5,479); distinct +5,479 / +43 uniq; fn-count +43.
audit-digest OK. 0 NON_MATCHING (G4). Cost: 3.73M subagent tokens across 24 agents, 0 errors.
THE BUG. tools/cast_call_sites.py classifies a declaration line with
^([ \t]*)(extern\s+)?([A-Za-z_][\w \t\*]*?)\b([A-Za-z_]\w*)\s*\(([^;{]*)\)\s*;
Feed it a return statement and `return` is a perfectly good identifier where a type is expected:
return func_8012CB64((s32)out, -0xC0, 0x40, -0x60, 0);
^^^^^^ captured as the return TYPE, func_8012CB64 as the DECLARED NAME
so the "rewrite this decl to canonical" path REPLACED the statement with
`extern s32 func_8012CB64(s32,s32,s32,s32,s32);`, DELETING the return. In C89 a declaration after a
statement is a parse error, so the damage surfaced as a bare syntax error in the DRAFT -- reading as
the draft's fault, not the tool's. 9 of 9 staged members of family 0x801848dc lost their return.
fix: a keyword guard (a declaration's type-specifier can never begin with a statement keyword)
family_sweep --hseq --band all over 5 families: 0/39 -> 18/39 banked (only the guard changed)
⚠️ AND THE TRAP INSIDE THE FIX: the obvious R33 move is "route it through cdecl". CHECKED, and it is
WRONG -- cdecl.parse() is a DECLARATOR-GRAMMAR parser that assumes it was handed a declaration; it
reports `return func_X(...);` as declaring func_X and `if (f(a));` as declaring `if`.
Statement-vs-declaration is a question cdecl does not answer. Routing there would have been a silent
non-fix that looked principled. §134's law still holds for line-SHAPE masking; this is a different
question.
BLAST RADIUS (measured, not assumed -- R14): cast_call_sites is in gate_stage's DEFAULT pipeline
(canon_resident_calls -> cast_call_sites -> sig_unify -> harvest_verify) and has been since Phase 20.
Of 44,833 stored drafts, 318 (0.7%) carry a `return f(...);` line this mis-reads, across 67 callees
(func_8014F468 x41, func_8014F6F4 x37, func_8014F74C x32, ratan2 x25). Every one, every time it
passed the gate pipeline, lost its return and failed as PLUMBING. Part of the historical plumbing
tail is this bug.
ALSO IN THIS COMMIT
- S5 CALIBRATION WAVE (8 agents, ultracode, 1.31M tokens). Pool VERIFIED FIRST (R14 -- Fable's whale
claim was 3/4 wrong): measured 1,677 clusters / 5,795 fns / 319,755 ins at a 3.68x multiplier vs
its claimed 1,689 / 5,956 / 326,261 at 2.7x -- its numbers hold, and the multiplier is BETTER.
Result: 8/8 match_one MATCH (close=0), and 5/8 banked whole-binary -- the §52b/§61 gap is
integration, not codegen. Banked: func_801822E0 func_8017EC98 func_801851A8 func_80189A34
func_80188E10 (693 ins x1 before propagation). Not banked: func_8018B238 (FAILED),
func_8017EF54 + func_801802EC (NEAR) -- drafts kept in .run/wave-s40/ for recovery.
- 18 member-banks from the re-run sweep (the cross-address free-h_exact pool: h_exact-identical at
DIFFERENT addresses, which dedup_propagate correctly refuses since it assumes position-locking --
family_sweep is the right lane).
- cookbook §143 (this bug + the cdecl trap + the blast radius); index regenerated.
VERIFIED: make clean && make extract-all && make check-all -> 140 passed, 0 failed of 140.
Fleet 12419169 -> 12420375 instr; distinct +1,526 / +5 uniq; fn-count +23. audit-digest OK.
0 NON_MATCHING (G4).
NEW IDIOM FROM THE WAVE, not yet folded into §31 (agent was told to write only its draft): a byte
counter must be spelled `cnt + 0xff`, NOT `cnt - 1`. Both are mod-256 identical and both compile to
one addiu, but gcc-2.7.2 picks the immediate encoding from the SOURCE SPELLING (0xFFFF vs 0x00FF).
Also flagged: .run/ghidra_c/func_8017EF54.c is a stale decompile of the WRONG function.
Measured the h_exact free pool from the bytes rather than trusting the frontier report's
numbers (R14 — its whale claim was 3/4 wrong: it said the whale was open in all four SC07
overlays; three were already banked and I closed the fourth earlier this session).
MEASURED: 215 open function-instances / 8,763 instructions are byte-identical (h_exact,
including reloc payloads) to an already-matched function. ONE class is 86% of that pool:
func_801758FC — 55 ins, same address in all 138 overlays, matched in ov_SC01_000 only,
OPEN in the other 137 => 7,535 instructions.
h_exact means identical INCLUDING jal/lui/%lo reloc immediates, so the matched body compiles
byte-identically at every member with NO remap (dedup_extend's correctness argument, §14).
dedup_propagate --addr authored it once as DEFINE_func_801758FC() in engine_core.h and
instantiated it at all 137 open sites in address order.
[ OK ] 138 overlays byte-identical after propagation; 1 new group in config/dedup.us.yaml
VERIFIED: make clean && make extract-all && make check-all -> 140 passed, 0 failed of 140.
Fleet instr 12411467 -> 12419002 = +7,535 EXACTLY; fn-count +137; instr-weighted crosses to
94.5%. distinct-code unchanged BY DESIGN -- the class was already matched in ov_SC01_000, so
the 137 add fleet instructions but no new DISTINCT function. audit-digest OK. 0 NON_MATCHING.
Note this function had been sitting in the stored-draft backlog for ov_SC06_030 and
ov_SC07_010 and re-gated "no" earlier tonight -- because gating a DRAFT is the wrong move for
an h_exact class. The right move is propagating the already-MATCHED body. Same function, two
routes, and only one of them is free.
Remaining free pool after this: 78 instances / 1,228 ins across 32 classes.
The 39 draft-exemplar families all have their exemplar still OPEN in ov_SC01_077 -- a
draft-exemplar family cannot propagate until its head banks. Re-gated the newest stored
ov077 draft per head, in 4 small batches (§61: a wide harvest broke a TU in S38).
Set aside the top 4 heads (65% of the pool's weight, all known-hard): func_801412A8 +
func_80178004 ARE S6's two giant walls (198x138 + 165x138 = 50,094 ins riding on 2 cracks),
func_8017C974 is today's byte-proven close=47, func_8017C294 its 246-ins neighbour.
batch 0 1/9 batch 1 3/9 batch 2 4/9 batch 3 0/8 = 8/35 (23%)
BANKED: func_8017EC7C func_8018281C func_801820DC func_80182988 func_80183BAC
func_80183AF0 func_80183CF4 func_80182E7C
(+474 ins x1 now; ~1,441 ins of templatable weight behind them once their families propagate.)
CALIBRATION REFINEMENT (docs/calibration.md, S39): this population re-gates at 23%, vs 8%
for the general stored pool and 4/6 for fresh post-repair drafts. Three different populations,
three different rates -- which is exactly why the rule is "re-gate what a repair plausibly
touched", not "re-gate the ledger". ov_SC01_077 is the split-heaviest overlay, so the S38
alias-deletion repair plausibly touched all of these.
VERIFIED: make clean && make extract-all && make check-all -> 140 passed, 0 failed of 140.
Fleet 12410275 -> 12410749 instr (+474), distinct +474 / +8 uniq, fn-count +8. audit-digest OK.
0 NON_MATCHING (G4).
Drew T