/srcreg-bank-sync-audit — shared backend data-register hazards

Ground-truth precedence: the live ISA doc (tt-isa-docs MCP, fetched each run) outranks every rule, table, and example baked into this skill — treat those as dated illustrations. If the live ISA doc contradicts a baked rule here, do NOT silently proceed: surface the conflict to the user and ask whether the baked rule should be overwritten, discarded, or kept. Default to the ISA doc.

MANDATORY — before any verdict, read the shared grounding policy. The per-architecture source ladder (which docs to consult), the ground-or-abstain rule, and the Source preflight (list the sources you'll consult with their reachability + hierarchy, then PAUSE for the user) are defined once in race-audit-all → .claude/skills/race-audit-all/SKILL.md. Your FIRST action is to Read that file and follow its "Ground-truth source ladder", "Ground-or-abstain", and "Source preflight" sections — they are load-bearing: a verdict produced without them is ungrounded and MUST NOT be reported. If that file genuinely cannot be read, say so and abstain rather than proceed ungrounded. (If you were spawned by a race-audit-all sweep — your prompt already lists the confirmed sources — skip the Source preflight and do not pause; the orchestrator ran it once.)

Coverage — floor, not ceiling. The grep patterns and site lists in this skill are a seed, not an exhaustive enumeration. After running them, widen the search with full reasoning. The techniques here are illustrative examples, not the allowed set — use any approach your reasoning suggests, including ones not listed: e.g. semantic search (by behavior/effect, not just token), resolving macros / wrappers / typedefs / indirection the literal pattern can't match, following the call graph to callers and callees, and diffing the WH/BH/QSR variants to catch a site present in one arch and missing in another. If you can find a hazard, primitive, or site the encoded patterns don't cover — by any means — pursue and report it; do not clamp a stronger analysis to this list or to these techniques. State any residual coverage gaps explicitly (no silent caps).

Execution — parallel by default. When enumeration yields more than a few sites/files, fan out concurrent Agent calls by default (one per file/subsystem, a fresh context each), saturating the available concurrency (~10–16 at once); go inline only for a trivial set. The per-file fan-out described under Thoroughness is the default, not an exhaustive-only option. The cross-referencing/synthesis of results stays sequential (it must follow the per-unit findings). The heavyweight Workflow tool still requires explicit multi-agent opt-in — it is the opt-in exhaustive tier, not the default. Don't over-spawn a tiny diff.

Persisting results — single writer, incremental. Agents only return their findings; they never write a shared file (no concurrent-write clobbering). If findings are persisted to a file, the orchestrator/caller is the sole writer and appends each wave's returns as they arrive — incremental, never only-at-the-end — so an interrupt preserves every completed wave's findings.

The bug class (precise)

The backend data memories are shared and have their own hardware flow control, distinct from config registers, Tensix semaphores, mailboxes, and CBs:

• SrcA / SrcB — each has 2 banks carrying an AllowedClient ∈ {Unpackers, MatrixUnit}, plus four bank-pointer bits (MatrixUnit::SrcABank/SrcBBank, Unpackers[0/1]::SrcBank). The unpacker fills a bank and hands it to the FPU (set data-valid); the FPU consumes and hands it back. The Wait Gate enforces this in hardware: an FPU instruction stalls until the relevant bank's AllowedClient == MatrixUnit; UNPACR can start but stalls mid-execution until AllowedClient is appropriate. Software must keep the two sides' bank pointers in lockstep and place the valid/clear at the right point.
• Dst and LReg exist once (not per-thread). Threads can overwrite each other's data. The math↔pack Dst handoff rides the MATH_PACK semaphore (owned by semaphore-handshake-audit); cross-thread LReg is the (declared-but-unused) mutex::SFPU (also that audit). THIS audit owns the parts they don't: bank-flip / dvalid correctness, and any Dst/LReg sharing not mediated by those primitives.

A desync → the FPU reads a bank the unpacker is still filling, or a thread clobbers a live Dst/LReg → silent data corruption (rarely deadlock).