Forward-merges PR #91 (off-task encrypt/decrypt worker pools, GSO,
connected UDP), the macOS #102 package-integrity fix, and the AUR
#98 fips-dns fix from master.
Six textual conflicts resolved: wire/noise imports, the K-bit-flip
and rekey-cutover handlers, ActivePeer fields, and an advert test
fixture. One semantic conflict: PR #91's worker-pool code on master
referenced FLAG_SP and MmpPeerState.spin_bit, which next removed in
4aded9a2 (spin bit superseded by the MMP receiver-report timestamp
echo; FMP flags bit 2 reclaimed). The worker acceleration and its
fmp_flags round-trip test are kept; the spin-bit references are
dropped to honor next's wire-format change.
Moves both AEAD layers (ChaCha20-Poly1305, one round per layer per
packet) plus the sendmsg syscall off the rx_loop task onto a per-shard
worker pool, adds per-peer connect(2)-ed UDP with SO_REUSEPORT, and
uses Linux UDP GSO (sendmsg+UDP_SEGMENT — kernel splits one super-skb
into N on-the-wire datagrams in a single TX-stack walk) when packets
in a batch are uniform-size. Same kernel primitive WireGuard's
in-kernel module and BoringTun use to hit 2.5–3.2 Gbps single-stream.
Single TCP stream on a 5-node docker-bridge mesh, 5 x 15 s x P=1:
A→D: 1379 → 2708 Mbps (1.96x, RTT +0.12 ms)
A→E: 1394 → 2663 Mbps (1.91x, RTT +0.11 ms)
E→A: 1406 → 2624 Mbps (1.87x, RTT +0.19 ms)
Static-peer pairs only — every CoV under 3%, 0 outliers, 0% ICMP
loss. The ~+100 µs RTT is the worker queue handoff cost; AEAD +
sendmmsg now run on a separate core in exchange.
What lands:
- src/node/encrypt_worker.rs: std::thread + crossbeam_channel
workers; hash-by-destination dispatch pins a TCP flow to one
worker so wire ordering is preserved; per-worker sendmmsg(2)
batching up to 32; Linux uses sendmsg(2)+UDP_SEGMENT when
packets in a group are uniform-size.
- src/node/decrypt_worker.rs: receive-side mirror. Each shard owns
its session's recv cipher + replay window in a thread-local
HashMap (no shared RwLock/Mutex). Sessions are handed off at
promote_connection and re-registered on K-bit flip / rekey
cutover.
- src/node/handlers/session.rs try_send_session_data_pipelined:
FSP+FMP both seal in-place in the worker on one wire-buffer
alloc; no intermediate inner_plaintext / fsp_payload Vecs.
- src/transport/udp/connected_peer.rs + peer_drain.rs: per-peer
connect(2)-ed UDP socket with SO_REUSEPORT (set on the listen
socket too — without that, EADDRINUSE on activation and every
packet falls back to the wildcard path); the worker sends with
msg_name=NULL and the kernel uses its cached 5-tuple. Tick-
driven activation in handlers/connected_udp.rs, idempotent.
- src/transport/udp/mod.rs: mem::replace the recvmmsg backing buffer
instead of buf.to_vec() per packet — single pointer swap, no
MTU-sized memcpy.
- src/protocol/link.rs SessionDatagramRef: zero-copy borrowed view
used by handle_session_datagram for the bulk local-delivery
path; handle_session_payload takes the borrowed payload
directly (no payload[35..].to_vec()).
- src/transport/mod.rs TransportAddr::from_socket_addr: collapses
the two-alloc from_string(addr.to_string()) pattern to one.
- src/node/handlers/rx_loop.rs: decrypt-fallback drain promoted
ahead of packet_rx in the select! (TCP ACK starvation fix);
interleaved fallback drain every 32 packets inside the rx burst
loop.
- noise::Session: send_cipher_clone / recv_cipher_clone /
recv_replay_snapshot_owned / take_send_counter / accept_replay
so off-task workers can hold a cloned cipher + reserved counter
while the dispatcher keeps replay/counter sequencing serial.
CipherState::cipher_clone returns a refcount-bumped LessSafeKey.
AsyncUdpSocket: AsRawFd so workers issue raw sendmmsg / sendmsg
without going through the tokio reactor.
- Worker pool sizing: both default to num_cpus, overridable via
FIPS_ENCRYPT_WORKERS=N / FIPS_DECRYPT_WORKERS=N. Per-peer
connected UDP can be disabled via FIPS_CONNECTED_UDP=0.
- src/perf_profile.rs: optional per-stage timing reporter under
FIPS_PERF=1 (or FIPS_PIPELINE_TRACE=1). Off by default; zero
overhead when disabled.
- All cfg(unix)-gated. Windows continues on the existing tokio-
based send/recv.
Decrypt worker session lifecycle:
- Node::unregister_decrypt_worker_session mirrors the existing
register helper. Wired at the two natural sites that already
iterate peers_by_index: the rekey drain-completion block in
handlers/rekey.rs (drops the worker entry for the old our_index
once the drain window has expired and the cache_key is
unreachable to any in-flight OLD-K packet), and remove_active_peer
in handlers/dispatch.rs (drops the worker entry for each of the
four index slots: current, rekey, pending, previous). Only
our_index is normally registered; unregister_session is fire-
and-forget for missing entries, so calling unconditionally on
all four slots is correct and bounds the cleanup without per-
slot accounting. Without these callers the per-worker sessions
HashMap and the Node's decrypt_registered_sessions set would
grow monotonically per rekey on long-lived peers.
Testing:
- testing/static/scripts/bench-multirun.sh: multi-run iperf3 +
ping bench. N reruns (default 5), median / min / max / CoV % /
per-run outlier flag, avg ping RTT, ICMP loss %, TCP retransmit
total. Plain client→dest labels + topology header. Pre-bench
peer-convergence check (FIPS_BENCH_CONVERGE_SECS, default 15);
per-path route verification via stats.bytes_sent deltas — fails
fast if traffic exits via a non-static-peer link.
- testing/static/docker-compose.yml: passes FIPS_ENCRYPT_WORKERS /
FIPS_DECRYPT_WORKERS / FIPS_PERF through to containers for A/B
benchmarking without rebuilds.
- testing/static/scripts/iperf-test.sh: same plain client→dest
labels + topology header (was multihop/direct/N hop, which
conflated topology distance with on-wire path).
- .config/nextest.toml: synthetic UDP node tests serialized
through a max-threads=1 test group. Localhost handshakes drop
on shared CI runners under parallel load; one-at-a-time keeps
assertions reliable.
- src/node/tests/spanning_tree.rs: repair_missing_edge_handshakes
— retries up to 5 times for synthetic edges whose msg1 was
dropped, with a drain after each edge retry instead of after
each attempt's full burst.
- src/node/decrypt_worker.rs::tests: two unit tests asserting
WorkerMsg::UnregisterSession removes the worker-thread session
HashMap entry (handle_msg_unregister_session_removes_entry) and
is a no-op for never-seen cache_keys
(handle_msg_unregister_session_idempotent_on_unknown_key), which
is the safety invariant the unconditional unregister calls at
the four index slots in remove_active_peer rely on.
- src/node/encrypt_worker.rs::unix_tests
pipelined_send_wire_layout_roundtrips_canonical_decoders: mirrors
the encoder geometry of try_send_session_data_pipelined (no
coords, the common established-session path), runs the worker's
real seal + send via flush_direct_batch_sync, and decodes the
resulting wire packet using only canonical receive-side decoders
(EncryptedHeader::parse, SessionDatagramRef::decode, FSP header
parse, noise::open). Any divergence between the hand-rolled
encoder offsets (fsp_aad_offset, fsp_plaintext_offset) and the
decoders fails at one of the parse / open / decode steps before
the inner-plaintext assertion fires. Complements the existing
fsp_preseal_runs_before_outer_fmp_seal test which covers the
seal-ordering invariant with synthetic headers but does not
exercise the wire-layout invariant.
CHANGELOG.md [Unreleased] # Changed entry added describing the
worker-pool threading model, hash-by-destination dispatch,
sendmmsg/UDP_GSO, per-peer connected UDP, the operator-facing env
vars, and the bench numbers above.
Cherry-picks from mmalmi/master (paths translated from
crates/fips-core/src/ to src/): 9b7c723, 0deb5cb, 13f7339, e036c0e,
3740a68, 3792f83, 8510193, 4910b07, e53f545, e4e2896, 5fe4af5,
1d01ada, 8c37008, e12469e, 6eb2860.
Co-authored-by: Johnathan Corgan <johnathan@corganlabs.com>
The chacha20 crate (RustCrypto) ships SSE2 + soft backends only — on
aarch64 (Apple Silicon, ARM Linux servers, Docker on M-series Macs) it
falls through to a portable software impl at ~600–800 MB/s/core. ring
0.17 wraps BoringSSL's hand-tuned ChaCha20-Poly1305, which dispatches
to NEON on aarch64 and AVX2/AVX-512 on x86_64 — typically 3-5 GB/s/core
on the same hardware.
Same wire format. ChaCha20-Poly1305 is byte-deterministic for a given
(key, nonce, plaintext, aad), so any correct AEAD implementation
produces identical ciphertext. The full noise test suite covers this
implicitly: IK and XK roundtrip handshakes, replay window correctness,
multi-message nonce sequencing, and 100-message stress all pass at
1129/1129 (the lib's full `cargo test` count) — these only succeed if
ring's output matches what the receiver's existing replay-window
decrypt path expects.
Implementation notes:
* `LessSafeKey` (and `UnboundKey`) deliberately do not implement
Clone for safety. `CipherState`'s manual Clone impl rebuilds it
from the retained 32-byte key — cheap for ChaCha20-Poly1305 since
construction is essentially a key copy + a constant-time check.
* The keyed AEAD is now cached in `CipherState.cipher` instead of
being re-derived per packet. This was already a perf win for the
chacha20poly1305 backend (`new_from_slice` per packet was hot in
profiles); for ring it's a bigger win because `LessSafeKey`
construction also derives the Poly1305 key.
* Public `Vec<u8>`-returning API preserved. New module-private
`seal`/`open` helpers wrap ring's `seal_in_place_append_tag` /
`open_in_place` so the per-packet allocation pattern is local to
one place.
* `EndToEndState::Established` triggers `clippy::large_enum_variant`
after the swap (`NoiseSession` grew from ~600 to ~1.5 KB because
ring precomputes the Poly1305 key state at construction). That
precomputation is the win — boxing the variant would re-add an
indirection per packet and work against it. `#[allow]`'d at the
enum decl with a justifying comment.
ring is widely deployed (rustls, hyper-rustls, AWS SDK, …) and a
pure-Rust crate (uses BoringSSL's asm via a vendored build). It
introduces no new C toolchain requirements that aren't already there
for any rustls user.
Bench data from a downstream consumer of this crate (Docker e2e,
DURATION=10, identical hardware before/after, aarch64 Linux on
Apple Silicon):
2-node direct (A↔B):
TCP 1-stream 437 → 1097 Mbps (2.51×)
TCP 4-stream 439 → 1109 Mbps (2.53×)
TCP 8-stream 445 → 1069 Mbps (2.40×)
UDP @1000 Mbit 599/40% loss → 1000 Mbps lossless
ping under load ~0.6 ms (unchanged)
3-node forced transit (A → C → B):
TCP 1-stream 438 → 1019 Mbps (2.33×)
TCP 4-stream 421 → 982 Mbps (2.33×)
TCP 8-stream 443 → 1031 Mbps (2.33×)
UDP @1000 Mbit 475/52% loss → 1000 Mbps lossless
ping under load 7.68 ms / 215 ms max → 0.72 ms / 3.6 ms max
The relay-path lift is the cleanest tell on the bottleneck: the
transit node was crypto-bound (single-threaded soft chacha couldn't
keep up with offered rate), so the queue accumulated under load. With
NEON the relay isn't crypto-bound and the queue stops accumulating —
the 215ms ping-tail collapses to 3.6ms.
- Delete PROTOCOL_NAME_IK, PROTOCOL_NAME_XK and all IK/XK methods
- Remove Ik/Xk variants from NoisePattern enum
- Rename XX methods to drop xx_ prefix (sole pattern)
- Rename HandshakeMessageType variants from NoiseIKMsg1/NoiseIKMsg2
to Msg1/Msg2, add Msg3 variant (0x03) for XX 3-message flow
- Fix doc comments with correct XX message sizes and descriptions
- Fix stale SessionSetup/SessionAck handshake payload size comments
Replace the 2-message IK handshake with a 3-message XX handshake for
FMP link establishment. XX requires no prior knowledge of the peer's
static key — both identities are revealed during the handshake
(responder in msg2, initiator in msg3). This is the foundation for
the forklift upgrade that enables rolling protocol upgrades.
Changes:
- Noise XX state machine alongside IK/XK (8 unit tests)
- Protocol negotiation payload codec: format byte, packed version
min/max, 64-bit feature bitfield, TLV extensions (11 unit tests)
- FMP wire format version 0→1, msg3 header/builder, TCP stream framing
- FMP handshake switched to XX: PeerConnection 3-message flow,
handle_msg1 simplified (no identity), handle_msg2 sends msg3 and
promotes initiator, new handle_msg3 promotes responder with
restart/rekey/cross-connection detection
- Rekey handshake switched to XX with negotiation payload hash chain
fix (decrypt-and-discard in complete_rekey_msg2/msg3)
- Negotiation payload in msg2/msg3 (FMP version [1,1], features=0)
- Debug logging for handshake promotion paths
- Integration test convergence timeouts adjusted for extra round-trip
Squashed commits:
- Add Noise XX state machine alongside IK/XK
- Add protocol negotiation payload codec
- FMP wire format prep: version 1, msg3 header support
- Switch FMP handshake from Noise IK to XX
- Increase convergence timeouts for XX 3-message handshake
- Fix negotiation hash chain desync in rekey handshake
Bump rand (0.8→0.10), rtnetlink (0.14→0.20), tun (0.7→0.8),
simple-dns (0.9→0.11), socket2 (0.5→0.6), and criterion (0.5→0.8).
Migrate all rand call sites: thread_rng()→rng(), gen()→random(),
gen_range()→random_range(), RngCore→Rng trait. Work around secp256k1
0.30 requiring rand 0.8 by generating random bytes directly and
constructing SecretKey from slice.
Migrate rtnetlink to builder-based API: LinkSetRequest replaced with
LinkUnspec builder + change(), RouteAddRequest replaced with
RouteMessageBuilder.
Remove bloom benchmark (criterion 0.8 incompatible with old harness
config).
The session-layer handshake now uses the 3-message XK pattern instead
of the 2-message IK pattern, providing stronger initiator identity
hiding. The initiator static key is deferred to msg3 and encrypted
under the es+ee DH chain, so eavesdroppers cannot identify the
initiator from the handshake.
XK pattern: -> e, es (msg1) / <- e, ee + epoch (msg2) / -> s, se + epoch (msg3)
Key changes:
- Add XK handshake methods alongside existing IK methods in noise module
- Add SessionMsg3 wire format and FSP_PHASE_MSG3 (0x03) prefix
- Replace Responding state with AwaitingMsg3 in session state machine
- Rewrite session handlers: handle_session_setup defers identity to msg3,
handle_session_ack processes msg2 and sends msg3, new handle_session_msg3
completes the responder handshake and registers identity
- Link-layer (FMP) continues to use Noise IK unchanged
- Add comprehensive XK unit tests and update all integration tests
Each node generates a random 8-byte startup epoch, encrypted inside
both Noise IK handshake messages (msg1 and msg2). When a peer's msg1
arrives with a different epoch than the stored value, the node tears
down the stale session and processes the msg1 as a new connection,
enabling near-instant restart detection instead of the 30-second
dead timeout.
Wire format impact:
- msg1: 82 -> 106 bytes (added 24-byte encrypted epoch after ss DH)
- msg2: 33 -> 57 bytes (added 24-byte encrypted epoch after se DH)
- Wire msg1: 90 -> 114 bytes, wire msg2: 45 -> 69 bytes
Rename FIPS Link Protocol (FLP) to FIPS Mesh Protocol (FMP)
The "Link Protocol" name understated the layer's scope — spanning tree
construction, bloom filter routing, greedy forwarding, and mesh-wide
coordination go well beyond link-level concerns. Rename fips-link-layer.md
to fips-mesh-layer.md, update FLP→FMP throughout docs and source code
(FLP_VERSION→FMP_VERSION, wire.rs, rx_loop.rs, spanning_tree.rs).
New SVG illustrations
- Protocol stack: color-coded layer diagram replacing ASCII art
- OSI mapping: side-by-side comparison with traditional networking layers
- Bloom filter propagation: 6-node tree with sender-colored filter boxes
showing split-horizon computation per link
- Routing decision flowchart: 5-step priority chain with candidate ranking
by tree distance and link performance
- Coordinate discovery: sequence diagram showing LookupRequest propagation,
response caching, and SessionSetup cache warming
Redesigned existing SVGs
- Architecture overview: uniform node layout, U-shaped encrypted link
connectors, separate end-to-end session line
- Node architecture: split Router Core into FSP and FMP layers, reorganize
transports into Overlay/Shared Medium/Point-to-Point categories
- Identity derivation: wider boxes, visible encode arrow, dashed npub line
fips-intro.md revisions
- Add inline references to prior work: Yggdrasil/Ironwood for coordinate
routing, Noise Protocol Framework for IK handshakes, WireGuard for
index-based session dispatch, Wikipedia for bloom filters, split-horizon,
and greedy embedding
- Add explanatory paragraphs after bloom filter diagram describing
split-horizon filter computation and candidate selection behavior
- Simplify transport abstraction language, remove I2P/LoRa references
- Fix LookupRequest wording ("propagates" not "floods"), note intermediate
node coordinate caching on lookup responses
- Rewrite architecture overview prose to match redesigned diagrams
Johnathan Corgan
Martti MalmiandJohnathan Corgan