The wave-J/K/L draft recovery (S52) showed §176b was under-scoped: it made batched drafts agree
with each other and forgot the file they land in. Banks (a) TU-seeded conflict detection and its
iterative behaviour, (b) per-destination-file keying, and (c) the new recovery variant for a callee
the TU prototypes as (void) while your call must pass an argument. All 11 recovered drafts
re-verified MATCH after repair; none needed a codegen change.
resolve_conflicts() had two defects, both found by the wave-J/K/L draft recovery:
(a) THE SYMBOL TABLE STARTED EMPTY -- only draft-vs-draft was compared, so a draft contradicting
a declaration ALREADY IN the .c reached the rebuild and surfaced only as a compile error and
a bisect. src/800.c carries 'extern void func_8001C9D0(void);' (from banked func_8001C2C4)
while three wave-J drafts declared it (s32)/(void *). The TU now seeds the table, and the
drop report names whether the clash is with the TU itself or an earlier draft.
(b) ONE NAMESPACE FOR ALL FILES -- 'seen' was global across the slate, so two drafts landing in
DIFFERENT .c files could not legally disagree about a symbol. Separate TUs are separate
namespaces; the table is now keyed per destination file (R39: over-refusal discards good work).
On the 11 recovered drafts the new check named 7 real TU conflicts that the old one missed
entirely. All 7 were repaired by adopting the TU's declaration verbatim and casting at the use
site -- including a NEW variant: when the TU's prototype takes no argument and the call must pass
one, cast through a function pointer, ((void (*)(s32))func_8001C9D0)(a0). All 11 re-verified
MATCH afterwards, so the cast is byte-identical in every case.
R39 NC: a synthetic draft re-contradicting the TU is still dropped; the repaired slate is 11/11.
build_wave_atlas.py:
- taken-set was a hardcoded 'abcdefghijkl' wave-letter literal -> glob('.run/wave_*_cards.json')
(R33 derive, don't re-derive). NC: 634 -> 726 taken, strict superset, +92 cards from waves m/n
that the literal would have silently re-issued.
- --exclude-bins defaulted to 'main' on the REFUTED link-defect rationale; default now empty and
the help states the real reason (main cannot be gated incrementally; use tools/gate_main.py).
- new --only-bins allow-list: main waves need it, since gate_main rebuilds once per SLATE.
gate_lane.py: an unhandled gate_stage exception produced no JSON line -> v={} -> '0 banked',
indistinguishable from an honest empty gate (cost 2 cycles in P31). Now a non-zero rc or a missing
JSON line is labelled CRASH with the stderr tail, recorded in the results JSON, listed in the
summary as NEVER GATED, and exits non-zero. R39 NC both directions.
Harvested by a 15-agent workflow over every wave's journal.jsonl (where the long analyses live —
notifications truncate them), each cross-checked against the existing cookbook + gcc-2.7.2-map
before being called novel, then synthesized by residual CLASS rather than by function.
- §176-A STATEMENT ORDER AROUND A CALL is the first thing to check for any schedule/delay-slot/
±1-length residual. fill_simple_delay_slots backward-scans and never hoists an instruction
emitted AFTER a call into that call's slot. The arg-register tell: a delay-slot store whose
address goes through that call's own arg reg PROVES the statement sits before the call.
func_80186764 7/92-off + pins -> MATCH 92/92 by moving one statement (no pin needed);
func_80188528 41/102-off, filed 'irreducible tie-break, permuter fuel' -> MATCH 102/102.
- §176-B a small REGALLOC-PERM is usually NOT allocation. B1: a narrow global that is really
wide_sym+2 must be expressed through the wide object (asm-label alias) or the disambiguator
sees no dependency; func_80185548 MATCH 77, verified with a full HI16/LO16+R_MIPS_26 resolve
oracle. B2: pin the short-lived INTERLOPER out of the way, not the contested value;
func_8018CA74 MATCH 63/63 with a register-choice sweep confirming the mechanism.
- §176-C 🔴 WALL REFUTATION, and I verified the source myself: gcc-2.7.2 sched.c:1704 tests
call_used_regs[i] where every neighbouring line uses regno+i, so for any 1-word register the
test is always call_used_regs[0] (, call-used on MIPS). EVERY hard-reg SET in a block
therefore gets a REG_DEP_ANTI on the last call, while the pseudo arm (sched.c:1732) is guarded
by reg_n_calls_crossed. A PIN CANNOT SCHEDULE AROUND A CALL — sometimes the correct move is to
UNPIN. Refutes the universality of sched.md S11 step 1 and the 'always try pins' reflex.
- plus §176-D (CSE levers in reverse), §176-E (two cc1-probed spellings), §176-F (four residual
verdicts that were lying).
- the section ends with an explicit 'What is NOT banked here' listing 7 mined items judged too
thin — including two whose functions are still INCLUDE_ASM (lever unverifiable) and one whose
narrative CONTRADICTS the banked C. Recorded so they are not silently lost.
Banking the PROCESS idioms this session produced, which were sitting only in commit messages,
tool docstrings and the wave prompt — none of which a future session reads.
- §176a VERIFICATION-LAYER LAWS: match_one verifies SHAPE not SYMBOL IDENTITY (masks
jal/HI16/LO16 — a wrong callee or wrong global reports MATCH; func_8002A234 cost 5 gate
attempts); a detector is ADVISORY and the gate is the ARBITER (3 wave-G drafts withheld on
symfix flags all banked unchanged); a verifier that can pass WITHOUT BUILDING is worse than
none (stale-binary false pass); an all-zeros gate result is a NULL not a finding; and the
general rule — before believing a measurement, run the control that would make it FAIL.
- §176b BATCH-GATING MECHANICS: gate cost scales with (binary,TU) GROUPS not drafts; batched
drafts must agree with EACH OTHER (type conflicts, duplicate typedefs); compatibility compares
TYPE SIGNATURES ONLY but the declarator suffix matters (too-strict and too-coarse both bit me);
conflict-dropped drafts are recoverable via cast-at-use; a COMPILE error names its own culprit
so only a BYTE mismatch needs bisection.
- §176c MAIN CANNOT BE GATED INCREMENTALLY — psyq_integrate/ld_interleave rewrite the .ld;
byte-proven both ways including with NO draft substituted. Use tools/gate_main.py.
Agent-discovered matching idioms are being mined from all 14 wave journals in parallel and land
next as §176.
- 48 atlas mass cards on ov_SC02_000 with --min-ins 40: 45/48 shape-verified, 42 banked,
ONE gate group. stubs 11,549 -> 11,477. distinct-code 90.4%.
- the band question is answered: 94% draft at avg 65 ins (up to 119), after 98% at avg 51 in
wave M. The mass lane is NOT size-limited in the band the instruction-weighted metric tracks.
- 2 NEARs enqueued with unusually deep analyses, both reusable beyond their own functions:
* func_80189C6C (close=2): magic 0x66666667 + mfhi-shift 5 DECODES to a plain /80 — write
'(x<<12)/80' and let gcc synthesize its own magic multiply. Residual root-caused against
REAL cc1 -da RTL dumps to loop.c move_movables desirability (threshold 29 vs measured
insn_count 24-26). Matches the func_80015A74 hard-tail class.
* func_80185840 (close=3): 'register u32 zr __asm__("$0"); c = val + zr;' reproduces the
non-coalescing addu-zero copy that cookbook §52a had classified as a WALL; plus 'r = K;'
before a SINGLE-armed if is what lets reorg's backward scan steal the li into the delay
slot (any two-arm spelling needs the eager target-thread steal, which never fires).
- waves C-N: ~448 banked, stubs 12,059 -> 11,549, fleet 95.4% instr, 213/213 after every
batch, main 0 -> 175 matched.
- records the three campaign-changing findings (main is open and was never hard; gate-group
packing is the throughput lever; match_one verifies SHAPE not SYMBOL IDENTITY), the tools
built/fixed, the prioritized open work, and the methodological lesson.
- cron be8fb48c cancelled; nothing in flight at handoff.
- 44 atlas mass cards on ov_SC04_011, avg 51 ins (up to 112) -> 43/44 shape-verified,
40 banked, ONE gate group. stubs 11,589 -> 11,549.
- data point that matters for the endgame: the mass lane holds at ~98% draft on the LARGER
band (51 avg vs the 12-42 cousins the night started with). Since the public metric is
instruction-weighted, that is the band that moves it — and it is in reach of haiku/sonnet,
not only the frontier tier.
- 1 near enqueued (func_80188A30, close=9: gcc reorders a beqz + delay-slot nop; core logic
verified correct) as grinder fuel.
- each draft compiles STANDALONE so it carries its own 'typedef struct {...} SVECTOR;'. Once
one such function banks, that typedef lives in src/800.c forever and every later draft
defining its own collides — a C89 duplicate-typedef error, not a byte miss. harvest_verify
already handles this; gate_main did not, and wave L lost a verified-correct draft to it.
- strip_dup_typedefs() drops typedefs the destination file (or an earlier body in the same
batch) already defines, and keeps novel ones.
- R39 NC: drops the duplicate, keeps the novel one, leaves the function body untouched, and is
a strict no-op when there are no duplicates.
- records the night's totals and, more importantly, HOW to run the main lane: gate_main.py
batch clean-rebuild, never gate_lane/gate_stage (incremental + .ld rewrite = false diff).
- notes the one mechanical improvement left (strip duplicate typedefs on substitution) and
that the ~10 conflict-dropped main drafts are verified-correct and recoverable.
- 44 atlas mass cards -> 44/44 shape-verified, 42 banked (2 dropped: a duplicate SVECTOR
typedef and a u8[] vs char[] decl conflict). main stubs 955 -> 913. fleet stubs -> 11,589.
- first wave carrying law 1c (match_one masks relocations => verifies SHAPE not SYMBOL
IDENTITY; re-check every symbol against the target .s relocation lines after MATCH).
- the compile-error shortcut paid for itself: named 'SVECTOR at src/800.c:94' and the exact
draft in seconds, where the old bisect path burned 28 minutes producing nothing.
- KNOWN NEXT IMPROVEMENT: gate_main should strip duplicate typedefs on substitution the way
harvest_verify already does — src/800.c carries a local SVECTOR from a previously banked
function, so any later draft defining its own collides. Mechanical, recurring, cheap.
- 44 atlas mass cards on main -> 44/44 standalone, 41 banked after 3 in-TU decl-conflict drops.
gate_main clean rebuild -> 143dbb89 BYTE-IDENTICAL, then full-fleet R22 213/213.
main stubs 996 -> 955; main REAL 92 -> 133 (incl. propagated/dedup credit).
- took 5 attempts; each exposed a real defect, the last one substantive:
* conflict checker too coarse (u8 D_x == u8 D_x[]) -> fixed + NC'd
* my pkill pattern matched its own shell -> stop shell process-matching
* FALSE PASS: sha() read a stale binary when the build failed -> rm output + check returncode
* bisect burned clean rebuilds; the compiler names the culprit -> read the error instead
* THE REAL BUG: func_8002A234 stored to the WRONG GLOBALS (v1->D_80078EE8/0->D_80078EE4,
target is the reverse). 2 bytes, both at %lo relocation offsets.
- STANDING CAVEAT (write this into the wave prompt): match_one MASKS jal/HI16/LO16, so it
verifies INSTRUCTION SHAPE, NOT SYMBOL IDENTITY. A draft that calls the wrong function or
stores to the wrong global passes standalone every time. Only the whole-binary gate sees it
— same class as the PsyQ symbol-name errors in waves F/G.
- clean_build() ran 'make build' and then sha()'d build/us/SLUS_007.26 off disk. If the build
FAILED (compile error), the PREVIOUS successful binary was still there, so sha() returned the
good hash and the tool reported BYTE-IDENTICAL for a build that never ran.
- that is exactly how it claimed '43 banked' for wave K on a TU that did not compile; the
clean-fleet R22 caught it ([FAIL] main). A verifier that can pass without building is worse
than no verifier.
- fix: rm the output before building, and treat a non-zero make return as no-hash/never-pass.
- also: unbuffered print (a 16-min run looked hung with an empty log) and read the compile
error to name the culprit instead of bisecting at a full clean rebuild per step.
- gate_main reported wave K BYTE-IDENTICAL; the clean fleet R22 then failed [FAIL] main.
Three drafts declared D_80078D98 inconsistently (1 scalar, 2 array) and my conflict checker
could not see the difference: typesig() split on the symbol and kept only the prefix.
- fixed to retain the declarator suffix ('' vs '[]'); NC'd both directions — the wave-K
conflict is now caught, wave J's known answer (34/5) is unchanged.
- BOTH failure modes now documented in the tool: v1 too STRICT (compared parameter names,
discarded 2 good drafts), v2 too COARSE (ignored [], passed a real conflict). R22 caught
what the tool missed, which is exactly why the clean-rebuild rule exists.
- phase log: recorded the night's methodological lesson — every serious stall was an
instrument trusted without a control, never the compiler.
Drew T