Make ordinary Noise reconnects atomic and race-safe (#1463)

Replaces destroy-then-rebuild Noise re-handshakes with an atomic reconnect protocol: prepared XX message-1 handoff tokens (claim-once, invalidated by crossed inbound initiations), receive-only quarantine of the established transport while an inbound replacement proves identity (promote on success, rollback+cooldown on failure/timeout), deterministic lower-peerID-wins crossed-initiator resolution, and a per-link-epoch BLE revalidation policy that re-proves cached sessions inside the same bleQueue critical section as the link rebind. On main, a fresh msg1 simply destroys an established session and rekey spans two non-atomic barriers.

Includes the review fix: timeout-restores defer outbound queue draining until the convergence retry completes (restore reason plumbed end-to-end), so DMs are never drained under keys a restarted peer already discarded — with a deterministic interleaving test. Identity-mismatch restores drain immediately. Full local suite 1768 tests green.
This commit is contained in:
jack
2026-07-26 13:04:56 +02:00
committed by GitHub
parent e9275cb3d8
commit 78a81e5b57
17 changed files with 3248 additions and 179 deletions
@@ -12,6 +12,7 @@ import Testing
@testable import BitFoundation // to avoid unnecessary public's
@testable import bitchat
@Suite("Integration Tests", .serialized)
struct IntegrationTests {
private var helper = TestNetworkHelper()
@@ -272,8 +273,18 @@ struct IntegrationTests {
// Re-establish Noise handshake explicitly via managers
do {
let m1 = try helper.noiseManagers["Bob"]!.initiateHandshake(with: helper.nodes["Alice"]!.peerID)
let m2 = try helper.noiseManagers["Alice"]!.handleIncomingHandshake(from: helper.nodes["Bob"]!.peerID, message: m1)!
let m3 = try helper.noiseManagers["Bob"]!.handleIncomingHandshake(from: helper.nodes["Alice"]!.peerID, message: m2)!
let m2 = try #require(
try helper.noiseManagers["Alice"]!.handleIncomingHandshake(
from: helper.nodes["Bob"]!.peerID,
message: m1
)
)
let m3 = try #require(
try helper.noiseManagers["Bob"]!.handleIncomingHandshake(
from: helper.nodes["Alice"]!.peerID,
message: m2
)
)
_ = try helper.noiseManagers["Alice"]!.handleIncomingHandshake(from: helper.nodes["Bob"]!.peerID, message: m3)
} catch {
Issue.record("Failed to re-establish Noise session after restart: \(error)")
@@ -8,6 +8,7 @@
import Foundation
import CryptoKit
import Testing
@testable import BitFoundation // to avoid unnecessary public's
@testable import bitchat
@@ -27,9 +28,14 @@ final class TestNetworkHelper {
node.mockNickname = name
nodes[name] = node
// Create/replace Noise manager for this node
// This synchronous helper directly drives all three XX messages and
// has no transport callback loop for delayed collision recovery.
let key = Curve25519.KeyAgreement.PrivateKey()
noiseManagers[name] = NoiseSessionManager(localStaticKey: key, keychain: mockKeychain)
noiseManagers[name] = NoiseSessionManager(
localStaticKey: key,
keychain: mockKeychain,
recentInitiatorCompletionGracePeriod: 0
)
return node
}
@@ -108,8 +114,18 @@ final class TestNetworkHelper {
let peer2ID = nodes[node2]?.peerID else { return }
let msg1 = try manager1.initiateHandshake(with: peer2ID)
let msg2 = try manager2.handleIncomingHandshake(from: peer1ID, message: msg1)!
let msg3 = try manager1.handleIncomingHandshake(from: peer2ID, message: msg2)!
let msg2 = try #require(
try manager2.handleIncomingHandshake(
from: peer1ID,
message: msg1
)
)
let msg3 = try #require(
try manager1.handleIncomingHandshake(
from: peer2ID,
message: msg2
)
)
_ = try manager2.handleIncomingHandshake(from: peer1ID, message: msg3)
}
}