chromium-rust-v0.53 WASM to jvm-v1.2 WebSocket/mplex timeout

Nobody has claimed this yet.

**Repos:**
- [libp2p/rust-libp2p](https://github.com/libp2p/rust-libp2p) WASM build (commit `b7914e407da34c99fb76dcc300b3d44b9af97fac`)
- [libp2p/jvm-libp2p](https://github.com/libp2p/jvm-libp2p) (commit `737456de5d96db030aa04c8527585af1a510898f`)

**Transport:** ws, **Secure:** noise, **Muxer:** mplex

## Summary

chromium-rust-v0.53 (WASM build running in Chromium) times out when dialing jvm-v1.2 as listener via ws/noise/mplex. The JVM listener starts successfully and publishes its multiaddr, but the WASM dialer never connects. The test hits the 180-second timeout with no output from the dialer container.

The yamux variant of the same test (`chromium-rust-v0.53 x jvm-v1.2 (ws, noise, yamux)`) passes, confirming the issue is specific to mplex muxer negotiation between these two implementations.

## Failing tests (1)

- `chromium-rust-v0.53 x jvm-v1.2 (ws, noise, mplex)`

## Passing variant

- `chromium-rust-v0.53 x jvm-v1.2 (ws, noise, yamux)` — passes in 13s

## Error output

```
Test timed out after 180s
```

No output from the dialer container. The listener log shows successful Redis connection and multiaddr advertisement:

```
Connection established to Redis (Jedis{Connection{DefaultJedisSocketFactory{proxy-b7ff7ce3:6379}}})
Starting listener with advertisedAddress: /ip4/172.18.0.5/tcp/12120/ws/p2p/12D3KooWEZT8uKLqgX2tg6vzf9EaSKET9c1sREZiY5PE9zG9hnfv
```

## Root cause analysis

The WASM rust-libp2p v0.53 dialer and JVM jvm-libp2p v1.2 listener fail to negotiate the mplex muxer over a ws/noise connection. Since the yamux variant works, the noise handshake and WebSocket transport are functioning correctly between these implementations. The failure occurs during or after muxer negotiation when mplex is selected.

Possible causes:
- Incompatible mplex protocol version or framing between rust-libp2p WASM v0.53 and jvm-libp2p v1.2
- The WASM environment may handle mplex stream initialization differently, causing a deadlock during muxer setup
- jvm-libp2p's mplex implementation may expect a different multistream-select sequence than what the WASM client sends

## Notes

This is distinct from the `chromium-rust-v0.53 x python-v0.x (ws, noise, mplex)` timeout (see `chromium-rust-v053-python-ws-timeout.md`), which involves a different listener implementation. The chromium-rust-v0.53 WASM dialer connects successfully to rust and go listeners via ws/noise/mplex, so the issue is specific to the JVM listener's mplex handling.

Start with the failing `chromium-rust-v0.53 x jvm-v1.2 (ws, noise, mplex)` interoperability test and compare it with the passing yamux variant. Inspect the pinned rust-libp2p and jvm-libp2p commits and the related `chromium-rust-v053-python-ws-timeout.md`; done means the mplex test no longer times out while the yamux and other listener cases remain passing.

Written by the indexing model from the issue text.