mirror of
https://github.com/luxfi/crypto.git
synced 2026-07-27 01:54:50 +00:00
Hanzo AI
7034d335f2
Replace the gpu.go one-liner stub with a real, full accelerator that
exposes the FIPS 204 NTT primitive plus batched sign/verify across all
three parameter sets via one shared NTT kernel.
The package decomposes into:
params.go FIPS 204 parameter table (modes 2/3/5 → k, ℓ, η, τ,
β, γ₁, γ₂, ω, sizes). Single table; the NTT kernel
is mode-independent because every set shares the same
ring Z_q[X]/(X^256+1) with q=8380417.
ntt_cpu.go Pure-Go negacyclic NTT/INTT over the FIPS 204 ring.
Canonical Cooley-Tukey forward + Gentleman-Sande
inverse, FIPS 204 Appendix B zeta table. Constant-
time helpers (reduce32, addModQ, subModQ, mulModQ).
parallel.go Worker-pool helpers (capWorkers, parallelDo) shared
by both engines.
gpu.go Public dispatcher: Accelerator with threshold-routed
NTTForward/Inverse/Mul Batch, SignBatch, VerifyBatch,
Stats. Defaults: ≥16 polys for NTT GPU, ≥8 for sign,
≥16 for verify. Stats counters expose GPU vs CPU
dispatch.
gpu_cgo.go (build tag cgo) — luxfi/accel-backed engine. NTT/Mul
use accel.Lattice().PolynomialNTT/INTT/Mul; batched
sign/verify use MLDSASignBatch / MLDSAVerifyBatch with
mode ∈ {2, 3, 5}. Per-op GPU failures fall through to
a parallel CPU path so the dispatcher contract
("every batch completes") holds.
gpu_nocgo.go (build tag !cgo) — pure-Go engine. NTT goes to
ntt_cpu.go; sign/verify go to circl-backed
mldsa{44,65,87} wrappers. parallelDo fan-out matches
GOMAXPROCS so CPU-only hosts still get multi-core
throughput. No stubs.
Tests cover: FIPS 204 parameter cross-check vs circl, NTT∘INTT
roundtrip, linearity, known vector, schoolbook PolyMul reference,
constant-time helper bounds, threshold routing, batch sign+verify
across all three modes, cross-wrapper equivalence (accelerator
signatures verify under plain mldsa{44,65,87}.Verify).
Both CGO_ENABLED={0,1} build and test green. KAT vectors in
pq/mldsa/mldsa{44,65,87} unchanged.
Apple M1 Max numbers (single-poly NTT, batched sign/verify):
CPU NTTForward ~2.6-3.3µs/op
CPU PolyMul ~7-9µs/op
SignBatch 44 batch64 ~6.5ms (CPU) / ~8.1ms (CGO+accel)
SignBatch 65 batch64 ~6.8ms (CPU) / ~12.3ms
SignBatch 87 batch64 ~23ms (CPU) / ~45ms
VerifyBatch 44 batch64 ~5ms (CPU) / ~6.1ms
VerifyBatch 65 batch64 ~4.7ms (CPU) / ~7.6ms
VerifyBatch 87 batch64 ~7.5ms (CPU) / ~10.1ms
(CGO numbers reflect accel session overhead; actual Metal kernels
for MLDSASignBatch/VerifyBatch land in luxfi/accel and are routed
to CPU until lux-gpu publishes the FIPS 204-aware kernel.)
Mode-44 and mode-87 sign/verify now share the same accelerator
surface as mode-65 — adding new parameter sets in the future is one
row in the params table, not a kernel rewrite.
Lux Crypto
Cryptographic primitives for the Lux Network -- post-quantum signatures, key encapsulation, BLS aggregation, threshold signing, ring signatures, and EVM-compatible secp256k1.
go get github.com/luxfi/crypto
Architecture
luxfi/crypto is the cryptographic foundation for all Lux software. It provides both classical and post-quantum primitives, with automatic CGO acceleration where available (blst for BLS, circl for lattice schemes). The pure-Go fallback path requires no C compiler.
Post-Quantum (NIST FIPS 203/204/205)
| Package | Algorithm | Standard | Security | Key Sizes |
|---|---|---|---|---|
mldsa/ |
ML-DSA | FIPS 204 | 128/192/256-bit (Levels 2/3/5) | 44: 1312/2560 B, 65: 1952/4032 B, 87: 2592/4896 B |
mlkem/ |
ML-KEM | FIPS 203 | 128/192/256-bit | 512: 800/1632 B, 768: 1184/2400 B, 1024: 1568/3168 B |
slhdsa/ |
SLH-DSA (FIPS 205, formerly SPHINCS+) | FIPS 205 | 128/192/256-bit (12 variants) | SHA2/SHAKE, fast/small tradeoff |
pq/ |
Unified PQ interface | -- | Wraps mldsa, mlkem, slhdsa | Mode selection at runtime |
ML-DSA and ML-KEM wrap Cloudflare's circl with ergonomic key serialization. SLH-DSA provides hash-based signatures as a conservative fallback (no lattice assumptions).
Classical
| Package | Algorithm | Use |
|---|---|---|
bls/ |
BLS12-381 (G1 keys, G2 signatures) | Consensus signatures, aggregation, proof-of-possession |
secp256k1/ |
secp256k1 ECDSA | EVM transaction signing, Ethereum compatibility |
secp256r1/ |
P-256 ECDSA | TLS, WebAuthn, FIDO2 |
ecies/ |
ECIES (secp256k1) | Asymmetric encryption for Ethereum-compatible keys |
Threshold and Multi-Party
| Package | Protocol | Use |
|---|---|---|
threshold/ |
Threshold signature framework | Interface + registry for pluggable threshold schemes |
threshold/bls/ |
BLS threshold signatures | t-of-n BLS signing for consensus |
cggmp21/ |
CGGMP21 (ECDSA threshold) | MPC key generation and signing, Paillier commitments |
Advanced Constructions
| Package | Construction | Use |
|---|---|---|
ring/ |
Ring signatures (LSAG + lattice-based) | Unlinkable signer anonymity |
lamport/ |
Lamport one-time signatures | Hash-based PQ signatures (stateful) |
hpke/ |
Hybrid Public Key Encryption | ML-KEM + X25519 hybrid, KEM factory |
kem/ |
KEM abstraction | ML-KEM, X25519, hybrid combiner |
aead/ |
AEAD ciphers | AES-256-GCM, ChaCha20-Poly1305 |
kdf/ |
Key derivation | HKDF, SLIP-10 HD derivation |
verkle/ |
Verkle tree commitments | State proof compression |
kzg4844/ |
KZG commitments (EIP-4844) | Blob transaction proofs |
ipa/ |
Inner product arguments | Verkle proof backend |
Infrastructure
| Package | Purpose |
|---|---|
common/ |
Address, Hash types (20-byte, 32-byte) |
hash/, hashing/ |
Keccak256, SHA256, RIPEMD160, Blake2b |
rlp/ |
RLP encoding (Ethereum wire format) |
cb58/ |
CB58 encoding (Lux address format) |
cert/ |
TLS certificate management for node identity |
signer/ |
Transaction signing abstraction |
sign/ |
Signature scheme registry |
secret/ |
Secret-safe memory operations (zeroing, constant-time) |
address/ |
Multi-chain address derivation |
gpu/ |
GPU-accelerated modular arithmetic bindings |
bindings/ |
C/Rust FFI exports |
BLS Signatures
import "github.com/luxfi/crypto/bls"
sk, _ := bls.NewSecretKey()
pk := bls.PublicFromSecretKey(sk)
sig := bls.Sign(sk, []byte("block hash"))
valid := bls.Verify(pk, sig, []byte("block hash"))
// Aggregation
aggSig, _ := bls.AggregateSignatures(sig1, sig2, sig3)
aggPK, _ := bls.AggregatePublicKeys(pk1, pk2, pk3)
valid = bls.Verify(aggPK, aggSig, msg)
Post-Quantum Signatures (ML-DSA)
import "github.com/luxfi/crypto/mldsa"
sk, pk, _ := mldsa.GenerateKey(mldsa.MLDSA87) // NIST Level 5
sig, _ := sk.Sign(rand.Reader, data, nil)
valid := pk.VerifySignature(data, sig)
Post-Quantum Key Encapsulation (ML-KEM)
import "github.com/luxfi/crypto/mlkem"
sk, pk, _ := mlkem.GenerateKey(mlkem.MLKEM1024) // NIST Level 5
ciphertext, sharedSecret, _ := pk.Encapsulate()
recovered, _ := sk.Decapsulate(ciphertext)
// sharedSecret == recovered (32 bytes, use as AES key)
Hybrid HPKE
import "github.com/luxfi/crypto/hpke"
// ML-KEM-1024 + X25519 hybrid
suite := hpke.NewHybridSuite()
enc, ct, _ := suite.Seal(recipientPK, plaintext, aad)
pt, _ := suite.Open(recipientSK, enc, ct, aad)
Testing
go test ./... # 382 test functions
go test -bench=. ./... # benchmarks
Compiled test binaries are provided for quick verification:
bls.test-- BLS signature testsmldsa.test-- ML-DSA testsmlkem.test-- ML-KEM testsslhdsa.test-- SLH-DSA testscrypto.test-- Core crypto tests
Papers
- Lux PQ Crypto Suite -- parameter selection and security analysis for ML-DSA, ML-KEM, SLH-DSA
- Lux Hybrid PQ Architecture -- hybrid classical/PQ transition strategy
- Lux Crypto Agility -- algorithm negotiation and migration framework
- Lux Pulsar PQ -- post-quantum ring signatures
- Lux Universal Threshold Signatures -- multi-curve threshold framework
References
- NIST FIPS 203: ML-KEM (Module-Lattice-Based Key-Encapsulation Mechanism)
- NIST FIPS 204: ML-DSA (Module-Lattice-Based Digital Signature Algorithm)
- NIST FIPS 205: SLH-DSA (Stateless Hash-Based Digital Signature Algorithm)
- Cloudflare circl -- underlying lattice implementations
- BLS12-381 -- pairing-friendly curve specification
License
Lux Ecosystem License v1.2. See LICENSE.