Pick multiport lanes by inside flow hash, not routine index

Lanes were collapsing onto lane 0 and staying there. Nebula writes a
peer's inbound packets to the tun queue matching the socket they arrived
on, and tun_flow_update teaches the kernel to steer that flow's outbound
packets to the same queue, preferring what it learned over the hash. So
while a tunnel's lanes were still down -- which is every tunnel, for its
first moments -- all traffic arrived on socket 0, pinning every flow to
queue 0 on both hosts for as long as it stayed busy. With the lane
following the routine, that meant lane 0 forever.

Two changes break the loop:

Choose the lane from the packet's own 5-tuple hash, so lane spread no
longer depends on tun steering at all. The hash is symmetric, and the
high-addressed side rotates its choice by the low side's port offset, so
a flow's two directions land on partner lanes with exact reverse
4-tuples and each arrives through the conntrack entry the other's probe
opened.

Seed every lane as demanded at lane-set creation, so the first traffic
tick probes them all at once rather than waiting for a flow to hash onto
each. Lanes have to be up before the flows are, not after.

A routine may now send on any lane, so it holds a SendBatch per lane,
built lazily and all borrowing one shared arena -- the slab is the
expensive part, and a batch per (routine, lane) would otherwise cost
hundreds of megabytes. full() counts across the batches so the total
outstanding stays bounded as before.

That also means several routines can write to one socket, which was
already true for base traffic under multiport: the linux batchWriter's
sendmmsg scratch had no lock. Take a mutex there, once per flush.

overlay/batch/tx_batch.go

@@ -13,19 +13,36 @@ type batchWriter interface {
 // One SendBatch is owned by each listenIn goroutine; no locking is needed.
 // Slots are backed by an Arena (see its docs)
 type SendBatch struct {
-	out   batchWriter
-	bufs  [][]byte
-	dsts  []netip.AddrPort
-	arena *Arena
+	out  batchWriter
+	bufs [][]byte
+	dsts []netip.AddrPort
+	// arena backs the slots. borrowed means someone else owns it and Flush must
+	// leave it alone.
+	arena    *Arena
+	borrowed bool
 }
 
 // NewSendBatch makes a SendBatch with batchCap slots and an arenaSize byte buffer for slices to back those slots
 func NewSendBatch(out batchWriter, batchCap, arenaSize int) *SendBatch {
+	return newSendBatch(out, batchCap, NewArena(arenaSize), false)
+}
+
+// NewSendBatchSharedArena makes a SendBatch that borrows arena rather than
+// allocating one. Several batches over different sockets can then split a single
+// slab, which is what multiport lanes need: the slab is the expensive part of a
+// SendBatch and one routine may hold a batch per lane. Flush leaves the arena
+// alone, so the owner must Reset it once every batch sharing it is drained.
+func NewSendBatchSharedArena(out batchWriter, batchCap int, arena *Arena) *SendBatch {
+	return newSendBatch(out, batchCap, arena, true)
+}
+
+func newSendBatch(out batchWriter, batchCap int, arena *Arena, borrowed bool) *SendBatch {
 	return &SendBatch{
-		out:   out,
-		bufs:  make([][]byte, 0, batchCap),
-		dsts:  make([]netip.AddrPort, 0, batchCap),
-		arena: NewArena(arenaSize),
+		out:      out,
+		bufs:     make([][]byte, 0, batchCap),
+		dsts:     make([]netip.AddrPort, 0, batchCap),
+		arena:    arena,
+		borrowed: borrowed,
 	}
 }
 
@@ -54,6 +71,8 @@ func (b *SendBatch) Flush() (int, error) {
 	clear(b.bufs)
 	b.bufs = b.bufs[:0]
 	b.dsts = b.dsts[:0]
-	b.arena.Reset()
+	if !b.borrowed {
+		b.arena.Reset()
+	}
 	return written, err
 }