Commit Graph
77 Commits
Author SHA1 Message Date
evgeny 7eb6d6ee68 logging: move socket mechanics and periodic probe decisions to trace 2026-10-07 21:14:38 +02:00
evgeny bdcf8b5634 Сохрани накопленные исправления аудио, async и диагностики 2026-10-02 01:49:32 +03:00
evgeny 00e890e0c3 Reserve worker completions before starting background tasks 2026-10-01 17:36:12 +03:00
evgeny fb4343fca6 transport: reality-релей idle-timeout (конфиг, дефолт 5с) + warn при >64 релеях; убрать лимит FD_SETSIZE; проверка socket_id в stcp_server 2026-09-24 12:08:45 +03:00
evgeny 9793733cd5 chatgui: диагностика воспроизведения голосовых сообщений (Windows)
- sound_manager: ma_engine_init/playRawPcm логируют через DEBUG_* (qWarning/qDebug терялись в GUI-приложении без консоли); детальный результат старта PCM (rc/is_playing/cursor)
- main: ранняя инициализация лога в GUI-потоке до SoundManager::init, чтобы диагностика инициализации попала в файл
- messagedelegate: лог клика по Play/timeline (filePath/dlState/существует ли файл)
- voiceplayback: лог заголовка opus (sr/ch/frameSamples)

Попутные правки:
- u_async: фикс двойного close wakeup_pipe (eventfd/Windows)
- chat_core: NULL-guard в get_db/is_initialized
- android: sqlite открывается в utun_instance_init, cleanup stale-потока без утечки fd
- android SettingsScreen: компактная раскладка имени интерфейса
2026-09-21 18:59:04 +03:00
evgeny a7a3666d85 radio: AGC release-rate (без кликов), jitter dead-zone, TX/RX-статистика, FGS-гонка и кэш настроек 2026-09-20 20:07:48 +03:00
evgeny 27e66f1092 log: move socket diagnostics SYS->SOCKET and conn lifecycle ETCP->CONNECTION (avoid poll() flood) 2026-09-20 01:51:40 +03:00
evgeny fae0a7c069 uasync: cast function pointers in socket_add log to void* 2026-09-20 01:18:33 +03:00
evgeny c36cebe357 uasync: add socket name diagnostics and HUP/ERR auto-unregister safety net 2026-09-20 01:11:30 +03:00
evgeny 73fe99b33f dm: прямой обмен сообщениями (DM) + soundtouch + архитектура звонков
- src/dm/: dm_core/dm_crypto/dm_mailbox — прямой обмен сообщениями (DM)
- lib/soundtouch-master/: vendored SoundTouch (изменение темпа/высоты аудио)
- doc/dm_arch.md, doc/chat_call_arch.md: архитектура DM и звонков
- tests: test_dm.c, test_dm_e2e.c
- chat: рефакторинг chat_* модулей, member_sync, invite_build
- chatgui/chatgui-android: синхронизация с новым API чата
- lib/u_async, debug_config, utun_instance: сопутствующие изменения
2026-09-15 21:09:17 +03:00
evgeny 187bcd80b9 uasync: ре-валидация node перед последующими коллбэками (счётчик cb_called) вместо снапшота 2026-09-14 12:56:52 +03:00
evgeny 836c715299 uasync: снапшот полей socket_node в poll()/select() — защита от realloc во время коллбэков (UAF) 2026-09-14 12:47:35 +03:00
evgeny 5c2501b4f3 chat: BGP-driven address sync, member identity verify, join_sig fix
- Адреса узлов вынесены из wire-протокола member_sync/merkle/join/welcome
  в BGP/topo_group: node_addresses теперь пишет только topo_node_sqlite_addrs_put
  (источник истины ni->v4/v6_addrs), классификация config_type/nat_type в sockmeta
- Жёсткая анти-подделка: node_id=derive(x25519), ed25519 сверяется с доверенной
  таблицей nodes, NODEINFO node_id/pubkey mismatch отклоняется как forgery
- Единые member_sync_build_join_msg (112Б) и member_sync_build_update_msg:
  устранён баг _verify_join_sig (join_ts по смещению 80 вместо 104, длина 104→112),
  который рушил приём мемберов; все 7 мест подписи/проверки join_sig через хелпер
- Единое обновление своего мембера chat_core_update_my_member (вместо sync_my_addresses)
- Android standby: пауза таймеров (broadcast/db_sync/merkle_sync/chat_sync) через standby_wait
- name[16] → name[MAX_CONN_NAME_LEN] с truncation-warning в парсере
- conn_mgr: etcp_connect_cancel_by_node (use-after-free при отмене invite),
  промоушен conn в ncd после успешного invite
- Android GUI: foreground-gating опросов, сохранение портов интерфейсов, фикс udp-log flush
- chatgui: персистентный db_path; мелкие DEBUG-логи epoll/pool
2026-08-15 12:36:05 +03:00
evgeny 5d02a6d419 android: debug categories from GUI, fix poll log format, event loop sleeps until traffic
- android_jni_bridge: utun_bridge_set_debug_level actually calls debug_set_level / debug_set_category_level_by_name (was stub)
- SettingsScreen: CATEGORIES updated to match current g_debug_categories from debug_config.c
- ConfigProvider: persist category levels in [debug] section of generated INI config
- u_async: poll() log uses computed timeout_ms instead of confusing %d.%dms format
- u_async: timeout_node char name[16] → const char* name (no truncation, smaller struct)
- memory_pool: pool alloc/free debug logs moved from DEBUG to TRACE
- instance_lite: poll timeout changed from 10ms (100tb) to infinite (-1) — sleeps until traffic
- instance_lite: removed useless heartbeat timer
2026-08-12 09:02:19 +03:00
evgeny d8921f0e3d merkle_sync: fix infinite WAL growth from bg_check
- _recompute_bucket: skip INSERT OR REPLACE when hash unchanged,
  also skip DELETE when row doesn't exist (was always writing)
- bg_check interval: 100ms → 5000ms (was writing 10x/sec × 4KB/page
  into WAL = ~1.4GB over hours)
- bg_timer: round-robin across channels instead of only first
- bg_timer: periodic WAL checkpoint (every ~60s) to prevent
  unbounded WAL growth from any source
- test: sync ms_session struct copy with real definition

- stcp_link: handle race where conn already CLOSED before close
- stcp_client: remove double-free c->free_on_close = cli
- etcp_connections: stcp_link_close on TCP link down
- u_async: round sub-ms timeouts to 1ms to avoid busy-loop
2026-08-09 19:25:36 +03:00
Evgeny 2cba839de4 Убрал DEBUG_CATEGORY_DEBUG: сообщения разбросаны по модульным категориям (+CHAT=25, MEDIA=23) 2026-08-09 17:24:15 +03:00
evgeny 7f901c01e1 etcp: переупорядочить close + защита state==2 во всех API-точках; topo: правильная очистка group nodes и registry; chat/member: новый модуль member_sync; chatgui: обновления транспорта и UI 2026-08-09 12:39:53 +03:00
evgeny 38c9f229bc stcp: ed25519 pubkey exchange in handshake, fix instance destroy segfault, fix chat_tcp_test
- u_async.c: process immediate_queue BEFORE epoll_wait in uasync_poll
- stcp_link.c: defensive NULL etcp_conn/etcp_link in stcp_link_close
- utun_instance.c: move stcp_server_list_destroy_all to Phase J (before ETCP),
  add deferred drain in both Phase J and L
- stcp.h/c: STCP_HS_ENC_CLIENT 6→38, STCP_HS_ENC_SERVER 7→39,
  exchange ed25519_pubkey in both handshake directions
- etcp_connections.c: copy peer_ed25519 from STCP to ETCP_CONN in
  etcp_link_enter_ready_tcp; guard link->etcp in tcp_link_close_cb
- chat_sync.c: fix lk->conn NULL dereference for TCP links
- test_stcp.c: update stcp_server_create/stcp_client_connect callsites
- tools/chat_tcp_test: change transport udp→tcp
2026-08-07 00:48:27 +03:00
evgeny d773f1724a refactor: STCP connections as ETCP links (is_tcp)
- Add is_tcp, tcp_link, tcp_reconnect_* fields to ETCP_LINK
- Remove transport_link from ETCP_CONN, add tcp_link_count
- stcp_link: replace embedded etcp_conn with pointer; add etcp_link backpointer
- stcp_link: link_rx_cb routes ETCP data to etcp_conn_input
- stcp_link: call on_close_cb in close
- New TCP functions: etcp_link_enter_ready_tcp, etcp_tcp_link_start_connect,
  etcp_tcp_link_start_reconnect, tcp_link_close_cb, etcp_tcp_send
- etcp_link_new: handle NULL conn for TCP links; skip enter_init for TCP
- etcp_link_close: close stcp_link for TCP
- etcp_encrypt_send: redirect TCP to etcp_tcp_send (skips AES-CCM)
- tcp_server_on_link: create ETCP_LINK(is_tcp=1) + enter_ready_tcp
- etcp_connect: TCP addrs -> etcp_tcp_link_start_connect
- Remove ctx->tcp_link/tcp_ready
- test_stcp_traffic: find stcp_link via links->is_tcp instead of transport_link
- 66/68 tests pass; 2 STCP tests expected to fail (integration not complete)
2026-08-06 22:14:32 +03:00
evgeny be0e76c9a6 debug: реорганизация категорий — SYS(1), ETCP(4) расширен, CONNECTION(3) очищен; fix stcp_client.c sock_err до connect()
Категории:
- DEBUG_CATEGORY_SYS=1: u_async + ll_queue + memory + timing (было 4)
- DEBUG_CATEGORY_ETCP=4: ETCP + normalizer + BBR (было 3)
- DEBUG_CATEGORY_CONNECTION=3: только события статуса линка/подключения (INFO)
- DEBUG_CATEGORY_KEEPALIVE=21: только периодический мониторинг (DEBUG)
- Удалены: NORMALIZER, BBR, CONNECTIVITY
- Устаревшие DB_SYNC → CHAT_SYNC + MEMBER_SYNC (предыдущий коммит)

CONNECTION+INFO — только: Connection UP/DOWN, Link UP/down, Link initialized/added,
  NAT address init/changed, init server/client.
CONNECTION+DEBUG — retry, collision, reinit, socket register.
KEEPALIVE — init_timer SUPPRESSED, timer started, keepalive skipped.
REINIT — ETCP+INFO.
Connection status changed — GENERAL (пользовательский callback).
Dedup link_status: if(old_link_status) перед Link down.
Убран дублирующий Conn:%s из Link down, добавлен link_id в client added link.
Молчаливые ошибки: +DEBUG_ERROR в 15 местах (merkle_sync, member_sync, db_sync).

Fix: stcp_client.c — socket_get_error() вызывался ДО connect(), проверяя
застарелое errno вместо ошибки connect(). Неблокирующий connect() возвращает
-1+EINPROGRESS (норма), но сравнивался с мусором → ложный FAIL.
Все 4 STCP-теста исправлены (68/68 pass).
2026-08-05 10:43:32 +03:00
Evgeny 2b10f53466 diagnostics: ncd connect timer SET/FIRE logs + WSAEWOULDBLOCK fix + cm_invite_fail details
- u_async.c: log ncd_connect timer expiration calculation (tb, now_tb, exp_tb, exp_ms, delta_tb)
- node_conn_direct.c: SET logs at 4 timer creation points with value_tb/now_tb/path; TIMEOUT log with now_tb
- conn_mgr_core.c: cm_invite_fail entry log + CLOSE/SKIP branch detail
- stcp_client.c: fix WSAEWOULDBLOCK (10035) treated as error on Windows (add ERR_WOULDBLOCK check)
2026-08-04 18:16:53 +03:00
Evgeny 6a35fc4b45 u_async: add pipe() fallback for non-Linux POSIX wakeup (FreeBSD) 2026-08-04 15:01:36 +03:00
evgeny e1e54df4f4 fix: invite не затирает имя узла, имя читается из member_sync peers.userinfo
- conn_mgr_core.c: убрана запись node_name из cm_handle_invite_info_resp
- topo_node_sqlite.c: name в nodes не перезаписывается пустой строкой
- android_jni_bridge.c: bridge_get_node_name читает peers_<ch>.userinfo если nodes.name пуст
2026-08-03 18:50:11 +03:00
evgeny ce5a84da24 fix: uasync epoll event ordering + test_uasync_socket_race fix 2026-07-29 11:10:55 +03:00
evgeny 064bed06f2 android: fix key regeneration, Clear DB, uasync eventfd, udp log from config
- jni_on_event: fix JNI_EDETACHED — AttachCurrentThread/DetachCurrentThread
- u_async.c: pipe→eventfd for reliable cross-thread wakeup
  add process_posted_tasks in poll() path, fix EINTR (saved_errno)
- bridge_debug_hook: keep udp_log_write, remove mutex from udp_log
- memory_pool.c: alloc/free logs INFO→DEBUG (reduce spam)
- instance_lite.c: chat_event_set_handler before ensure_keys,
  KEYS_GENERATED events, fchmod 0600 on keys.conf,
  UDP log from config [log_udp] section, regenerate_trampoline
- config_parser: parse [log_udp] ip/port
- NativeLib.kt: remove callbacksSet, simplify init/stop/destroy
- ChatViewModel.kt: unified event callback in init, pubKey StateFlow,
  requestRegenerateKeys, clearState fix
- SettingsScreen.kt: pubKey.collectAsState(), Clear DB via restart()
- HeadlessService.kt: remove setUdpLogTarget (now from config)
- android_udp_log.c: simplified — no mutex, buffer+send+timer in one thread
- utun_instance.c: NULL-check tcp_connections before destroy
- chat_event.h: +CHAT_EVT_KEYS_GENERATED 11, -NODE_CHANGED
- logreceiver: UDP log receiver tool
2026-07-28 16:29:33 +03:00
Evgeny a47ef86810 uasync: fix stall detection — record timestamp after poll wait, not before 2026-07-27 19:42:35 +03:00
Evgeny 48bfe644c4 uasync: add event loop stall detection (warn if >10ms between poll entries) 2026-07-27 19:35:13 +03:00
Evgeny 27906e9f91 db_sync: односторонний sync, sparse-таблица, vp в SEND_DATA, PUSH-релей 2026-07-25 15:00:21 +03:00
Evgeny de7841f33a chatgui: create group, unified logging, fix uasync_post epoll bug
- Add CreateGroupDialog with group name input
- Generate X25519+Ed25519 keypairs, random uint64 channel_id,
  Ed25519 signature, write to DB via uasync thread
- chat_core_create_channel(): create msg table, topo_node_sqlite_channel_put()
- GUI_EVT_CHANNEL_UPDATED notification to reload channel list
- Per-category debug config: debug_level= + debug_categories= in [gui]
- Unified logging via debug_config: all threads write to same file/console
- debug_enable_console(1) to mirror file output to stderr
- DEBUG_LEVEL_DISABLED=-1: truly disable category, NONE=0 fallback to global
- Fix uasync_post under epoll: process_posted_tasks was never called
- Fix MainWindow constructor parameter ordering (dbPath vs debugFile)
- debug_ui.h: append mode ('a' instead of 'w')
2026-07-12 17:39:20 +03:00
Evgeny 8962f58ea4 uasync: replace raw socket_node pointer with index-based handle — fix UAF after realloc 2026-07-02 17:08:24 +03:00
Evgeny de8cce521a all 51 tests pass: fd=0 protection + clean socks_http_proxy test
- socket_compat: never close fd=0
- socks_http_proxy: rewritten cleanly, 2 workers, 25 reqs each, up to 100KB
- uasync_socket_race: dup2-based fd=0 reservation
- u_async: removed DIAG spam
- tcp_io: removed debug fprintf

Result: 51/51 tests pass.
2026-07-02 00:58:47 +03:00
Evgeny f06ebd2ab4 fix: protect fd 0 from close in socket_close_wrapper + dup2 reserve in tests
- socket_close_wrapper: never close sock==0 (prevents fd 0 reuse cycle)
- test_uasync_socket_race: dup2-based fd 0 reservation, pass 800/800
- tcp_io: SOCKET_INVALID before on_error (previous commit)
- Plus earlier fixes: uasync gen counter, ll_queue call_soon_id, pkt_normalizer

Result: test_uasync_socket_race PASSES, 50/51 tests pass
2026-07-02 00:50:36 +03:00
Evgeny a97a38d867 fix: set tc->sock=SOCKET_INVALID before on_error in handle_error
В одном epoll_wait batch три события для fd: EPOLLERR (освобождает tc),
EPOLLIN (read_cb), EPOLLOUT (write_cb). read_cb/write_cb guard проверяет
tc->sock==SOCKET_INVALID, но sock обнулялся ПОСЛЕ on_error.
Перенос sock=SOCKET_INVALID до on_error защищает от use-after-free.

Доказано: без фикса segfault на tcp_io.c:374 (write_cb → tc->connected).
С фиксом тест работает без крашей.
2026-07-02 00:43:37 +03:00
Evgeny ccdbab95a8 fix: use-after-free in pkt_normalizer + ll_queue waiter cancel + uasync gen counter for epoll fd-reuse
- pkt_normalizer: save recvpart->len before queue_data_put (may free synchronously)
- ll_queue: save uasync_call_soon handle in queue_waiter_handle, cancel on free
- uasync: generation counter in epoll data.u64 to skip stale events after fd reuse
- uasync: keep fd_to_index mapping after socket_array_remove for inactive nodes
- test: test_uasync_socket_race — 4 children × 200 iterations, currently flaky
2026-07-01 23:51:31 +03:00
Evgeny 075c4f4583 memory_pool: name field — идентификация пула в диагностике коррапшна 2026-06-09 22:20:25 +03:00
Evgeny 5982b011b4 etcp: restore etcp_link_update_inflight_lim + state machine fixes + diagnostics
- etcp_link_update_inflight_lim: self-contained method with clamping,
  send_blocked_inflight check, loadbalancer_link_ready on cwnd increase
- etcp_conn_on_inflight_lim_changed: recalc optimal_inflight + resume input
- etcp_ack_recv: use etcp_link_update_inflight_lim instead of manual set
- etcp_request_pkt: fix line 928 check old link (inf_pkt->last_link) not new
- etcp_link_close: recalc optimal_inflight on link removal (both branches)
- etcp_socket_remove: fix dangling pointer loop (nullify after close)
- uasync_print_resources: show timer names, remain time, deleted entries
- diagnostics in utun_instance_destroy and test_etcp_reconnect
2026-06-09 16:40:37 +03:00
Evgeny c56ec0b0ed uasync: добавить отладку poll→fd/rev, timer→имя при разрешении событий 2026-06-04 01:57:10 +03:00
Evgeny 63f019530b feat: tcp_io library + uasync socket toggle + ll_queue empty_callback
lib/u_async: uasync_set_socket_read/write — O(1) EPOLL_CTL_MOD toggle per socket
lib/ll_queue: queue_set_empty_callback — one-shot deferred callback on queue drain
lib/tcp_io: TCP connection management via uasync + ll_queue with pool alloc
  - read: direct recv into pool buffer, high/low water flow control
  - write: autofetch via ll_queue callback, partial send via write_buf
  - on_fin/on_error/on_flushed callbacks, FIN deferred until read_queue drained
  - getpeername for pre-connected socket detection
tcp_proxy_server: migrated to tcp_io (22→13 fields, 437→270 lines, -7 functions)
tests: test_tcp_io (5 tests), test_tcp_proxy_server rewritten for tcp_conn
2026-06-03 16:24:46 +03:00
Evgeny 8994d7a834 update timeout_heap and u_async: O(1) cancel, heap cancel bubble-up fix, epoll callback safety, uasync_stop, socket_array u_realloc 2026-05-29 01:11:37 +03:00
Evgeny 4d5afe5ae9 fix uasync: posted_tasks delivery on Linux, create cleanup leaks, atomic array expand 2026-05-14 03:32:23 +03:00
Evgeny 1a4b0a9de6 fix stale pointer after realloc in uasync socket/event handlers 2026-05-14 01:24:06 +03:00
Evgeny d0af315ff6 fix uasync set_timeout(0) ordering, dummynet shaper initial delay, add tests
- uasync_set_timeout(0): use FIFO immediate_queue instead of timeout_heap
  to guarantee execution AFTER all currently-expired heap timers
- uasync_cancel_timeout: search immediate_queue for timeout=0 timers
- dummynet: start initial shaper with bandwidth-based delay instead of 0
  (prevents queue from never filling up when packets arrive sequentially)
- test_dummynet: rename get_time_us -> now_us (conflict with u_async.h)
- test_u_async_comprehensive: fix immediate_timeouts, add timeout_ordering test
2026-05-12 13:42:16 +03:00
Evgeny 61fee27478 congestion: burst-based BW measurement + adaptive inflight (slow start/cong avoidance/loss response)
New features:
- get_time_us() in lib/u_async.c/h for µs-precision timing
- Burst measurement (ETCP_SECTION_MEAS_TS 0x07, MEAS_RESP 0x08, FILLER 0x09)
  - 12 MTU packets sent at line rate, receiver measures inter-packet gaps
  - Burst triggered when link saturated + min 500ms interval
  - Result: measured BW -> bandwidth for shaper, burst_target_bdp for window cap
- 3-phase adaptive inflight in link_stats_timer_cb:
  - Slow start: windows doubles each RTT, exits on loss/RTT spike/threshold
  - Congestion avoidance: additive increase, proportional RTT shrink, multiplicative loss shrink
  - Burst target cap: inflight_lim <= burst_target_bdp * 1.1
- Dynamic optimal_inflight = sum(link->inflight_lim_bytes), recalculated on change
- Adaptive win_timebase = rtt/2 (was hardcoded 10ms)

Fixes:
- dummynet.c: shaper formula fixed (80000 -> 80)
- dummynet.c: queue_data_put call fixed (3 args -> 2)
- etcp.c: SIZE ERROR format string (missing log_name argument)
- etcp_request_pkt: FILLER leaves room for PAYLOAD overhead

Tests:
- test_etcp_congestion.c: single-threaded, 5-phase BW/delay/loss test
  - sender + receiver + dummynet in one uasync
  - CSV metrics output every 100ms
- All 29 existing tests pass
2026-04-26 00:17:32 +03:00
Evgeny d74ab1c1c6 Add name parameter to uasync_set_timeout for timer identification in dumps 2026-04-21 19:06:45 +03:00
jeka c1d2d26cae 1 2026-04-15 09:00:12 +03:00
Evgeny 189a93c71d fix cpu consume at idle 2026-04-03 03:29:51 +03:00
Evgeny 8030161ea6 -warn 2026-03-27 15:14:07 +03:00
jeka 65b6b6c150 fix ack timestamp 2026-03-13 22:35:26 +03:00
jeka 60d97a90b1 1 2026-03-08 17:35:00 +03:00
jeka 2beb945bb0 1 2026-03-08 17:31:56 +03:00