fix(artery): distribute ActorSelection messages across lanes by target path

[He-Pin]

Motivation

The "must be able to send messages with actorSelection concurrently preserving order" test flakes on CI: 4 sender actors each drive 1000 round-trips via ActorSelection (4004 messages total). With multi-lane artery config (outbound-lanes > 1), all ActorSelection messages are routed to the same outbound queue because selectQueue uses the anchor's UID as the distribution key:

OrdinaryQueueIndex + (math.abs(r.path.uid % outboundLanes))

The anchor for ActorSelection is the root guardian (RootActorPath), whose UID is always 0 (ActorCell.undefinedUid). So math.abs(0 % N) = 0 for any N — all ActorSelection traffic concentrates on lane 0 while other lanes sit idle.

Similarly, inbound lane partitioning uses the wire recipient's UID (root guardian = 0), concentrating all inbound ActorSelection processing on a single inbound lane.

This is not a recent regression — the selectQueue logic has been unchanged since the Pekko fork from Akka.

Modification

Outbound (Association.send): Add a dedicated case sel: ActorSelectionMessage that computes the queue index from the selection's target path elements hash instead of the anchor's UID:

case sel: ActorSelectionMessage =>
  sel.msg match {
    case _: PriorityMessage =>
      // cluster heartbeats stay on control queue
      controlQueue.offer(outboundEnvelope)
    case _ =>
      val queueIndex =
        if (outboundLanes == 1) OrdinaryQueueIndex
        else OrdinaryQueueIndex + ((sel.elements.hashCode() & Int.MaxValue) % outboundLanes)
      queues(queueIndex).offer(outboundEnvelope)
  }

Inbound (ArteryTransport.inboundLanePartitioner): Parse the ActorSelectionMessage's target path from the envelope byte buffer and use it as the destination hash key, distributing inbound ActorSelection processing across all inbound lanes.

PriorityMessage ActorSelection (used by cluster heartbeats) continues to go through the control queue unchanged.

Result

ActorSelection messages are distributed across all outbound and inbound lanes based on target path, eliminating the single-lane throughput bottleneck while preserving per-path message ordering (same target path → same hash → same lane).

Tests

sbt -Dpekko.test.timefactor=2 "remote / Test / testOnly *SendConsistency*"
→ 24/24 passed (all 6 spec variants: Upd/Tcp/TlsTcp × 1-lane/3-lanes)

sbt "remote / mimaReportBinaryIssues" → clean

References

Fixes #3089

[He-Pin]

Superseded by #3092 which has a more optimized CodedInputStream-based protobuf parsing and includes test coverage.