RESEARCH RECORD

P/N521 Cross-Limb Borrow Regression

Cryptography · Integrity

Mobile view is optimised for record-by-record inspection. Use the section menu below; desktop keeps the full route context visible.

Record sections

SUMMARY

Record summary

CLAIM

A permanent regression fixture preserves the subtraction overflow boundary for P/N521 limb arithmetic.

NON-CLAIM

This does not prove production certification or all arithmetic invariants.

Public page exposes only sanitized, source-linked evidence. Private endpoints, secrets, and raw witness material remain out of scope.

THREAT / INVARIANT

Threat: An optimization could reintroduce a cross-limb borrow defect.

Invariant: The regression fixture must remain bound by manifest and aggregate report hashes.

Trigger: regression-order-sub-cross-limb-borrow

Root cause: Cross-limb borrow handling required permanent regression coverage.

TOOLING

Reference implementation: fenrua-kernel regression fixture and generated regression report.

Tools used: Node.js validation, native differential tests, sanitizer differential tests, Git provenance.

Tool versions: Node.js v24.18.0 plus pinned fenrua-kernel evidence links.

COMMANDS

npm run validate
node scripts/test-kernel-telemetry.mjs

EVIDENCE

Results: regression_001_p521_sub_overflow.bin and pinned GitHub regression report.

Evidence confirmation: Review confidence is represented only where linked public evidence exists.

Evidence hash: See linked evidence artifact.

Fix revision: 390f7aeef778ce93db12e16028bc3a788b643c2d

REGRESSION

regression_001_p521_sub_overflow.bin

Kernel impact: Integrity primitive boundary clarified.

Utility impact: Public interface remains maturity-labelled and limitation-adjacent.

LIMITATIONS

Browser does not replay binary fixture bytes; public page exposes hashes and pinned links.

Supersession: Active; not superseded.

REPRODUCTION

npm run validate
node scripts/test-kernel-telemetry.mjs