fix(client-shared): resolve High code-review findings (Client.Shared-005, Client.Shared-006)

Client.Shared-005: _activeDataSubscriptions (a plain Dictionary) and the
_activeAlarmSubscription tuple were mutated from the caller thread, the
keep-alive failover path, and DisconnectAsync with no synchronization,
risking bucket corrosion / InvalidOperationException / lost entries.
Added a dedicated _subscriptionLock and wrapped every read/write of that
bookkeeping state inside it (Subscribe/Unsubscribe[Alarms]Async,
Disconnect, Dispose, and the snapshot/clear/re-record steps of
ReplaySubscriptionsAsync). Awaited adapter calls stay outside the lock so
it is never held across I/O.

Client.Shared-006: HandleKeepAliveFailureAsync had only a non-atomic
state check guarding re-entry, so two bad keep-alives could each start a
failover loop, racing to dispose/replace _session and double-replaying
subscriptions. It now claims an atomic _failoverInProgress slot via
Interlocked.CompareExchange; a re-entrant call returns immediately. The
loop body moved to RunFailoverAsync, wrapped in try/finally that resets
the flag.

Tests: added KeepAliveFailure_ReentrantWhileFailoverInFlight_RunsFailoverOnce
and SubscribeAndUnsubscribe_ConcurrentCalls_DoNotCorruptState regression
tests; made the FakeSubscriptionAdapter / FakeSessionAdapter /
FakeSessionFactory test doubles thread-safe (and added a CreateGate hook)
so the concurrency tests exercise production locking rather than fake
state. All 138 Client.Shared tests pass.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Joseph Doherty
2026-05-22 06:20:40 -04:00
parent 3de688f8d6
commit e221371a0c
6 changed files with 248 additions and 61 deletions

View File

@@ -12,15 +12,24 @@ internal sealed class FakeSessionFactory : ISessionFactory
public bool ThrowOnCreate { get; set; }
public string? LastEndpointUrl { get; private set; }
/// <summary>
/// Optional gate that, when set, blocks <see cref="CreateSessionAsync" /> until completed.
/// Lets tests hold a failover loop in-flight to exercise re-entrancy.
/// </summary>
public TaskCompletionSource? CreateGate { get; set; }
public IReadOnlyList<FakeSessionAdapter> CreatedSessions => _createdSessions;
public Task<ISessionAdapter> CreateSessionAsync(
public async Task<ISessionAdapter> CreateSessionAsync(
ApplicationConfiguration config, EndpointDescription endpoint, string sessionName,
uint sessionTimeoutMs, UserIdentity identity, CancellationToken ct)
{
CreateCallCount++;
LastEndpointUrl = endpoint.EndpointUrl;
if (CreateGate != null)
await CreateGate.Task;
if (ThrowOnCreate)
throw new InvalidOperationException("FakeSessionFactory configured to fail.");
@@ -39,7 +48,7 @@ internal sealed class FakeSessionFactory : ISessionFactory
// Ensure endpoint URL matches
session.EndpointUrl = endpoint.EndpointUrl;
_createdSessions.Add(session);
return Task.FromResult<ISessionAdapter>(session);
return session;
}
/// <summary>