Commit Graph

17 Commits

Author SHA1 Message Date
Siavash Sameni
31d2306915 feat: per-step encode profiling in send task stats
Some checks failed
Build Release Binaries / build-amd64 (push) Failing after 3m48s
Adds average microsecond timings for each encode step:
- avg_agc_us: AGC processing
- avg_opus_us: Opus encoding
- avg_fec_us: FEC encode + repair generation
- avg_send_us: QUIC send_media
- avg_total_us: sum of above

Logged every 5 seconds in send stats. Resets each interval.
Use to identify which step is bottlenecking the encode loop
on devices where fps drops below 50.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-06 14:18:33 +04:00
Siavash Sameni
4af7c5f94c fix: AudioRing cursor desync + capture thread use-after-free
Some checks failed
Build Release Binaries / build-amd64 (push) Failing after 3m56s
AudioRing (reader-detects-lap architecture):
- Writer NEVER touches read_pos — fixes SPSC invariant violation
- Reader self-corrects when lapped (snaps read_pos forward)
- Power-of-2 capacity (16384 = 341ms) with bitmask indexing
- Added overflow_count and underrun_count diagnostics
- Wired ring health into engine stats and periodic logging

Capture thread use-after-free (drain latch):
- Added CountDownLatch(2) to AudioPipeline
- Audio threads count down after exiting their loops
- teardown() awaits latch (200ms timeout) before destroy()
- Guarantees no in-flight JNI calls when native handle is freed
- stopAudio() no longer nulls pipeline (teardown handles it)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-06 13:28:34 +04:00
Claude
20922455bd fix: send task crash on QUIC congestion + AEC toggle + debug reporter
Root cause: send_media() returns Err(Blocked) when QUIC congestion
window is full. The send task treated ANY send error as fatal (break),
killing the entire call. Now send errors drop the packet and continue.

Also hardened recv task to survive transient errors and added health
logging (recv gap tracking, periodic stats) to both send and recv.

Relay: added comprehensive debug logging — recv gaps, lock contention,
forward latency, send errors — all per-participant with 5s stats.

Other changes:
- AEC toggle in Settings (persisted, applied on next call)
- Debug report: records call audio (WAV), RMS histogram (CSV), logcat,
  stats. Emailed as zip via Android share intent after call ends.
- Replaced LinearProgressIndicator with Box (compose version compat)
- FileProvider for sharing debug zip attachments

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-06 07:38:56 +00:00
Claude
e6564bab57 fix: mic mute crackling + add AEC/NoiseSuppressor + dedup room participants
Mic mute: the send loop now zeros the capture buffer when muted instead
of relying on write_audio() to skip writes. Previously stale ring data
and AGC amplification of near-silence caused crackling artifacts.

AEC: attach Android's hardware AcousticEchoCanceler to the AudioRecord
session. Also attach NoiseSuppressor when available. Both are released
on capture stop.

Room UI: deduplicate participants by fingerprint so ghost entries from
stale relay state don't show duplicate names.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-06 06:06:35 +00:00
Claude
aebf9156c0 fix: dedup participants in UI, wait for QUIC close ack before exiting
UI: deduplicate room participants by fingerprint so ghost entries from
stale relay state don't show duplicates.

Engine: after select! ends, call close_now() + connection.closed() with
500ms timeout to wait for the relay to acknowledge the CONNECTION_CLOSE.
Previously the close frame was queued but the runtime died before quinn
could retransmit if the first packet was lost.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-06 05:40:06 +00:00
Claude
9bbaec6b35 fix: use shutdown_timeout so QUIC CONNECTION_CLOSE actually gets sent
shutdown_background() killed the tokio runtime before quinn could send the
CONNECTION_CLOSE frame on the wire, so the relay never knew the client left.
Now use shutdown_timeout(500ms) to give quinn time to flush the close frame,
matching the desktop client pattern (which uses 2s timeout).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-06 05:20:20 +00:00
Claude
a9c4260b4e fix: close QUIC connection on hangup so relay removes participant immediately
stop_call() now calls close_now() on the stored transport handle before
killing the tokio runtime. This sends a QUIC CONNECTION_CLOSE frame so
the relay's recv loop breaks immediately, triggering leave() + RoomUpdate
broadcast. Previously the runtime was killed first, so transport.close()
never ran and the relay kept stale participants until idle timeout.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-06 04:58:24 +00:00
Claude
0835c36d0f feat: settings page with persistence, client alias in handshake, fix null fingerprints
Some checks failed
Build Release Binaries / build-amd64 (push) Failing after 3m34s
- Add SettingsScreen with identity (alias, key backup/restore), audio defaults,
  server management, network prefs, and default room
- SettingsRepository persists all settings via SharedPreferences
- Auto-generate random display names on first launch (e.g. "Swift Wolf")
- Thread alias through CallOffer → relay handshake → RoomUpdate broadcast
- Derive caller fingerprint from identity key in relay handshake (fixes null
  fingerprints when --auth-url is not set)
- Persist identity seed for stable fingerprints across reconnects
- Add alias field to SignalMessage::CallOffer (serde default for backward compat)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-06 03:56:33 +00:00
Claude
2d4b8eebd5 feat: RoomUpdate protocol — broadcast participant list on join/leave
- Add RoomUpdate signal message to wzp-proto with participant count + list
- Add RoomParticipant struct (fingerprint + optional alias)
- Store fingerprint/alias in relay Participant struct
- Broadcast RoomUpdate to all room members on join and leave
- Add signal recv task in Android engine to handle RoomUpdate
- Surface room_participant_count + room_participants in CallStats JSON
- Show "X in room" with participant names in Android in-call UI

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-05 18:12:24 +00:00
Claude
a23d9f5e41 feat: foreground service, dB gain sliders, speaker routing, live network stats
- Wire CallService foreground service for background calls (microphone type)
- Add Voice Volume + Mic Gain sliders (-20 to +20 dB) applied in Kotlin
- Connect AudioRouteManager for real speaker toggle via AudioManager
- Feed quinn QUIC RTT into PathMonitor, display Loss/RTT/Jitter from live data
- Nuclear teardown between calls — recreate engine + audio pipeline each call
- Fix re-entrant teardown loop from CallService notification callback
- Park audio threads as daemons to avoid libcrypto TLS destructor crash on exit
- Remove duplicate wakelocks from Activity (service owns them now)
- Strip AEC + denoise from capture path, keep AGC only (incremental approach)
- Fix .so copy target: libwzp_android.so not libwzp.so

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-05 17:45:00 +00:00
Claude
b3e56ecbd8 feat: add AGC to capture + playout paths, add server UI, DNS resolve
- Wire AutoGainControl on both capture (mic → encode) and playout
  (decode → speaker) paths to normalize volume levels
- Add server list with add/remove custom server dialog
- Add IPv4/IPv6 preference toggle for DNS resolution
- Resolve DNS hostnames to IP in Kotlin before passing to Rust engine
- Revert to IP addresses for default servers (DNS still broken on QUIC)

AGC confirmed working — voice levels noticeably improved in testing.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-05 14:02:33 +00:00
Claude
bf91cf25bd feat: add real audio pipeline with Opus + RaptorQ FEC
- AudioPipeline: Kotlin AudioRecord/AudioTrack on JVM threads, PCM
  shuttled to Rust via lock-free ring buffers + JNI
- FEC: RaptorQ fountain codes on encode (5 frames/block, 20% repair
  ratio for GOOD profile), decoder feeds repair symbols for recovery
- Real audio level meter from mic RMS (replaces fake animation)
- Room name editable in UI (default: "android")
- Relay changed to pangolin.manko.yoga:4433
- Stats overlay shows FEC recovered count
- CallState now synced from polled stats (fixes "Connecting" stuck bug)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-05 12:33:59 +00:00
Claude
af85a49e86 fix: eliminate all native thread creation — run everything single-threaded
pthread_create crashes on Android due to static bionic __init_tcb stubs
in the Rust std prebuilt rlibs. This is unfixable without rebuilding std.

Solution: run the entire call (QUIC connect, handshake, media send/recv)
on a single tokio current_thread runtime. The JNI startCall() now blocks,
so Kotlin dispatches it to Dispatchers.IO (JVM thread, not pthread).

Audio pipeline temporarily simplified to silence frames — will restore
once threading is solved (either via Java Thread or rebuilding std).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-05 09:52:28 +00:00
Claude
9d9ce4706d fix: use current_thread tokio runtime — avoid pthread_create SEGV on Android
Multi-thread tokio runtime crashes with SEGV_ACCERR in __init_tcb
during pthread_create on Android (static bionic stubs from CRT).
Switch to current_thread runtime which runs network I/O on the
calling thread without spawning additional OS threads.

Also: clean up build.rs — use only libc++_shared.so (dynamic),
remove getauxval_fix.c hack, remove static c++/c++abi linking.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-05 09:27:46 +00:00
Claude
8d5f6fe044 feat: wire QUIC transport, JNI bridge, connect UI + add docs
- Replace raw FFI with proper `jni` crate for string marshalling
- Wire QUIC transport in engine: connect to relay, crypto handshake
  (CallOffer/CallAnswer, X25519+Ed25519), send/recv MediaPackets
- Feed received packets into jitter buffer (was previously ignored)
- Add connect screen UI with CALL button (idle state) and in-call
  controls (mute, speaker, hang up, live stats)
- Hardcode relay 172.16.81.125:4433, room "android"
- Add comprehensive docs in docs/android/:
  architecture.md (8 mermaid diagrams), build-guide.md,
  debugging.md, maintenance.md, roadmap.md

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-05 04:43:49 +00:00
Claude
e7b1c3372a feat: Android VoIP client — Phase 2 (JNI bridge, Compose UI, AEC pipeline wiring)
- JNI bridge with 8 extern functions (init, startCall, stopCall, setMute,
  setSpeaker, getStats, forceProfile, destroy) with panic catching
- Kotlin engine layer: WzpEngine JNI wrapper, WzpCallback interface,
  CallStats data class with JSON deserialization
- Jetpack Compose UI: InCallScreen with quality indicator (green/yellow/red),
  mute/speaker/hangup buttons, stats overlay, duration timer
- CallActivity with RECORD_AUDIO permission handling, Material3 theme
- CallService foreground service with WakeLock, WiFi lock, notification
- AudioRouteManager for speaker/earpiece/Bluetooth SCO switching
- AEC wired into CallEncoder pipeline: AEC → AGC → denoise → silence → encode
- AEC farend reference fed from decode path to encode path in pipeline
- Engine exposes set_aec_enabled/set_agc_enabled via AtomicBool flags

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-04 18:16:38 +00:00
Claude
26e9c55f1f feat: Android VoIP client — Phase 1 (audio quality, network adaptation, crate skeleton)
- New wzp-android crate with Oboe C++ backend, lock-free SPSC ring buffers,
  engine orchestrator, codec pipeline, and Android Gradle project structure
- AEC (NLMS adaptive filter), AGC (two-stage with fast attack/slow release),
  windowed-sinc FIR resampler replacing linear interpolation (wzp-codec)
- Opus encoder tuning: complexity 7 default, set_expected_loss support
- Mobile jitter buffer: asymmetric EMA (fast up/slow down), handoff spike
  detection with 2s cooldown, configurable safety margin
- Network-aware quality control: cellular-specific thresholds, faster
  downgrade on cellular, proactive tier drop on WiFi→cellular handoff,
  FEC ratio boost during network transitions
- Handoff detection in PathMonitor via RTT jitter spike analysis

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-04 18:07:55 +00:00