Re-bank the two daObjAbuku_c paths #1996 absorbed, and check the ratchet on main

[andrewboudreau]

Main has been out of sync on the CONVERTED ratchet since #1996 merged. Every PR touching src/** inherits the red. #1978 is currently failing for this and it has nothing to do with ov002.

What happened

config/converted-baseline.json banks src/_ZN12daObjAbuku_cD0Ev.cpp and src/_ZN12daObjAbuku_cD1Ev.cpp. #1996 deleted both when it promoted the ov002/daObjAbuku_c translation unit.

No one's checks were wrong — they never overlapped. #1996's green ran against a base predating #2007, and #2007's re-bank (1957 → 2567) is what first banked those two files. By the time #1996 merged, the paths it deletes were sitting in a baseline its CI had never read. The ratchet workflow only ran on: pull_request, so main itself was never checked, and the breakage surfaced on the next unrelated PR instead.

The fix

The removal is legitimate, so it is banked with a reason rather than reverted. tiers_ratchet.py diagnosed it itself:

MOVED -- absorbed into src/actors/daObjAbuku_c.cpp by TU promotion
(ov002/daObjAbuku_c), which fails: No raw offset arithmetic
(*(u32*)(c + 0x74)); No unk_<off> fields; Calls things by real names,
not mangled _Z

A reconstructed TU must spell vague-linkage symbols directly (_ZN7fBase_cnwEj, _ZN8dActor_cC2Ev, _ZN8dActor_cD2Ev) or its range will not link, so no_mangled_refs structurally cannot pass for an absorbing file. Byte-match outranks readability. Both removals are logged to config/converted-backslide-exceptions.jsonl with that reason. No source file changed and no byte moved. Baseline 2567 → 2565.

The push: [main] trigger

The workflow's design note asserted that staleness here "only ever runs one way" — permissive, never falsely red, so no PR ever needs to re-bank in order to pass. That is true for conversions, which only add.

TU promotion is the exception: it removes banked paths. The note now says so, and cites this incident, so the next person reading it does not re-derive it from a red build.

The trigger gates nothing — there is no branch protection and a red main is advisory. What it buys is attribution: the merge that desynchronises the baseline goes red itself, instead of ambushing whoever opens the next PR.

This matters more as the queue drains. Every remaining TU-promotion PR (#1993 with 26 TUs, #1995, #2000, #2002, #2005) will hit the same removal, and each one should bank its own reason rather than leaving it for the next branch to discover.

Verification

tiers_ratchet --check        PASS   baseline 2565   current 2565
pytest tools/test_tiers.py   29 passed
check_dead_references        no new dead references
port_refcheck                405 checked, all resolve
check_src_tu_compiles        88/88

🤖 Generated with Claude Code

https://claude.ai/code/session_01WdCK1xgrJdiJzPCh3bAJfQ

[tangos-validator]

✅ PR validation — Passed

Committed merge introduces no reconstruction or attribution regression.

Full merge validation

Check	Result
Committed test merge	yes
Byte-verified functions	11,042 / 11,347 (97.31%, +0)
Byte-verified code bytes	2,049,508 / 2,211,124 (92.69%, +0)
Claimed, not byte-verified	174 functions, 55,796 bytes (+0)
Perfect source moves	0 R100
Enrolled ranges (delinks complete)	11,058 functions, 2,052,772 bytes (92.84%, +0) -- differs from byte-verified by +16
Contributor credit	0 added, 0 changed, 0 lost
Relocation check	0 checked; no affected slots
Module fidelity	106/106 exact; 100.000000% compared bytes
Code linked from verified source	11,087 functions, 2,066,772 bytes (93.47%)
Module bytes from source	2,066,772 / 3,049,600 (67.8%); 811,492 (26.6%) are data no delink entry reaches
ROM data reproduced from source	447 symbol(s) exact, 242 partial, 15 differ

Byte-verified means the range carries complete in a delinks.txt, so the ROM build compiled it and compared it to the cartridge. The 174 claimed functions have a src/ file named after the symbol with no NONMATCHING banner, and nothing compiles them -- dsd fills their addresses with the ROM's own bytes. Both together are the 11,216 this project calls matched.

_{The private worker commits a test merge, builds the stock ROM profile, compares every executable module, measures matched and source-built code, checks contributor lineage, and verifies affected relocations. The mod profile is opt-in and is not part of this merge gate.}