← all exploratory attempts · paired baseline · show in source

breadth-06: Temporal execution queues

Agent report: partial Review pending

Session: completed

docs/spec-fabric-pe-temporal.md, lines 473–480

The CGRA execution-plan importer consumes the same projected logical queue, its complete consumer set, allocation-unit ordinal, and entry capacity. Its runtime reserves every distinct unit in one ingress match group before requesting the selected enqueue actions, commits all matching queue tails together, and retires one physical queue head only on the common consumer handshake. Software channel storage is only the dense per-consumer token view of those physical queues; it cannot multiply physical occupancy, impose a one-token default, or bypass shared-unit occupancy and Fabric grant policy.

Execution provenance

The original pilot recorded the compiler hash and checkout revision separately; its binary build revision was not verified. Coverage follow-ups use separately identified rebuilt binaries. Original simulator binary hashes were not retained.

Original standalone-test agent report

This is exploratory evidence, not an accepted production-format PBT.

No author explanation recorded.

Reported numeric verdicts

not measured

Tested obligations

• P1 declared per-allocation-unit entry capacity is carried unchanged
• P2 per-consumer token view does not multiply physical occupancy (capacity independent of FU-local consumer count)
• P3 no one-token default imposed on a larger declared capacity
• P4 non-positive entry capacity is not accepted (secondary; grounded in the :482-508 resource contract)

Untested obligations

• reserves every distinct unit in one ingress match group before requesting the selected enqueue actions
• commits all matching queue tails together
• retires one physical queue head only on the common consumer handshake
• cannot bypass shared-unit occupancy and Fabric grant policy
• complete consumer set and allocation-unit ordinal as importer inputs

Format

Reproduction

Measured subject code coverage

measured · Measured handwritten Loom subject code only; external LLVM/MLIR and generated code excluded. Includes main executions, including rejected inputs; deliberate controls excluded. No matching native-suite baseline for this executable has been imported, so no coverage gain is claimed. Semantic fidelity remains a separate audit.

Execution: {'subject_exit_codes': {'0': 100, '1': 8}, 'subject_invocations': 108}
Baseline: No matching baseline measured
Minimized cases: not measured

Published evidence