Files
Antje WorringandHanzo Dev 1d97b27e47 pulsar/jasmin: make threshold round1/round2/combine compile (codegen-only)
These 3 threshold-signature sources type-checked but failed in the Jasmin
backend. Fixed with codegen-only restructuring — NO algorithm, value, wire
layout, or constant-time change:
- combine: decomposed an over-blended x86 effective address into a base reg;
  split a shifted-OR source into shift-into-temp then OR.
- round1: relieved register pressure (spill output ptrs, extract the
  expandA+matvec+decompose phase into a helper frame, pack derive_rho_prime
  inputs into one preimage buffer, u32 stores for y_state).
- round2: an export fn had 13 pointer params (Jasmin caps at 6) — marshalled
  them into a caller-supplied table; gave polyvecl_scalar_mont distinct in/out
  ptrs; routed per-peer MAC tables through one buffer; stack-return digest
  helpers; spilled keccak-crossing loop-invariants.
Shared lib tweaks (consumed by these files): mac/transcript/lagrange/seed .jinc
stack-return + u64-index + BE-via-temp. All #[ct] contracts and #declassify
pragmas preserved (CT type-checked at compile). Full Jasmin corpus: 23/23
compile (was 20/23).

Co-authored-by: Hanzo Dev <dev@hanzo.ai>
2026-06-02 12:59:58 -07:00
..

Pulsar — Jasmin high-assurance track

This directory holds the Jasmin sources for the Pulsar high- assurance implementation track. Jasmin (https://github.com/jasmin-lang/jasmin) is a low-level cryptographic implementation language with a verified compiler whose generated assembly is bit-identical to the source-level semantics and admits machine-checked side-channel (constant-time) guarantees through the EasyCrypt companion proof system (https://github.com/EasyCrypt/easycrypt).

The libjade project (https://github.com/formosa-crypto/libjade) ships Jasmin sources for ML-DSA (Dilithium) and ML-KEM (Kyber) with EasyCrypt-machine-checked functional correctness against FIPS 204 / FIPS 203 and constant-time proofs against the leakage model of Barbosa et al. (Eurocrypt 2021). NIST PQC submissions that integrate libjade-verified primitives sit in the high-assurance evaluation track alongside the formosa-crypto reference work.

For Pulsar, the high-assurance plan splits along the Class N1 / Class N4 boundary of the construction itself:

Layer What Jasmin verifies Source of truth
Single-party ML-DSA-65 core Functional equivalence to FIPS 204 + constant-time over secret key and nonce paths libjade (jasmin/ml-dsa-65/, vendored via fetch script)
Threshold layer (Round-1 commit, Round-2 response, Combine) Functional correctness of the round protocol's polynomial-vector arithmetic + constant-time over each party's secret share new (jasmin/threshold/, this submission)

The Class N1 byte-equality claim composes the two: a Pulsar threshold signature is a single-party FIPS 204 signature whose underlying (z, h, c̃) components have been computed by an honest quorum running the threshold protocol. So the Pulsar Jasmin proof chain is

libjade ML-DSA-65 (functional ≡ FIPS 204)
  ∘ Pulsar threshold (functional ≡ single-party computation under
                        honest quorum + share-additivity)
  ⇒ Pulsar output ≡ FIPS 204 output (Class N1)

with constant-time taken as a side condition on every secret-dependent control- and memory-access path.

Status — initial track

The high-assurance scaffolding shipped at submission carries:

  1. The libjade ML-DSA-65 single-party baseline as the verified core (fetched on demand via ml-dsa-65/fetch.sh; not vendored into this repository).
  2. Threshold-layer Jasmin sources at threshold/round1.jazz (~400 lines), threshold/round2.jazz (~600 lines), and threshold/combine.jazz (~400 lines) plus shared lib/ (~1200 lines). The jasmin-ct gate runs blocking on all three (3 / 3 green per scripts/checks/jasmin.sh).
  3. EasyCrypt theories at ../proofs/easycrypt/ — 13 files compile with the admit budget hard-pinned at 0 / 0 by scripts/checks/ec-admits.sh. The Class N1 byte-equality theorem is a proved lemma at Pulsar_N1_Extracted.pulsar_n1_byte_equality_extracted. The residual axioms (22 total: 17 narrow implementation-refinement + 5 Lean-bridged algebraic) are enumerated in ../docs/proof-axiom-inventory.md with a per-axiom closure plan.

How to fetch libjade

cd jasmin/ml-dsa-65
./fetch.sh                  # clones libjade at the pinned commit

The fetch script pins libjade to a specific commit so the submission artifact reproduces deterministically. The libjade tree itself is not committed to this repository — it is fetched on demand. This keeps the submission tarball small and avoids re-licensing complexity (libjade is MIT-licensed; we cite + link rather than redistribute).

How to check

../scripts/check-high-assurance.sh

The script is skip-friendly: if jasminc or easycrypt is not on the system PATH it prints a clear skip message and exits 0. When the tools are present it compiles each .jazz file and runs easycrypt check on each .ec file.

Tool installation

Citations

  • Almeida, Barbosa, Barthe, Blot, Grégoire, Laporte, Oliveira, Pacheco, Schwabe, Strub. The last mile: High-assurance and high-speed cryptographic implementations. IEEE S&P 2020.
  • Bond, Hawblitzel, Maillard, Protzenko et al. EverCrypt: A Fast, Verified, Cross-Platform Cryptographic Provider. IEEE S&P 2020.
  • Barbosa, Barthe, Doczkal, Don, Fehr, Grégoire, Huang, Hülsing, Lee, Wu. Fixing and Mechanizing the Security Proof of FiatShamir with Aborts and Dilithium. CRYPTO 2023.
  • libjade ML-DSA implementation — https://github.com/formosa-crypto/libjade