getifaddrs() ifa_flags contains interface flags (IFF_UP etc.), NOT address
flags (IFA_F_TEMPORARY). Since IFA_F_TEMPORARY == IFA_F_SECONDARY == 0x01,
and 0x01 in interface flags is IFF_UP, ALL addresses appeared as 'temporary',
making the permanent>temporary priority non-functional.
New netlink-based get_interface_ipv6_addr_nl() reads real IFA_FLAGS from
RTM_GETADDR response, correctly identifying:
- Temporary: IFA_F_TEMPORARY=1, IFA_F_MANAGETEMPADDR=0
- Stable/mngtmp: IFA_F_TEMPORARY=1, IFA_F_MANAGETEMPADDR=1
- Legacy static: IFA_F_TEMPORARY=0
Verified: wlo1 permanent (3a7a:eff:fef5:9637) now correctly selected
for prefer_stable=1, temporary (1fb5:b7c:7538:dcf4) for prefer_stable=0.
- CM_V6_ANY for :: socket (matches all v6 addr classes)
- sin6_scope_id from s->netif_index (not from local_addr where :: has scope=0)
- getnameinfo in stcp_link_connect to preserve scope_id in addr string
- scope_id in cm_invite_tcp_connect, ncd_create_links, connect_create_links_v6, cm_bg_ping_to_node
- ncd_down_cb: defer callback removal+close via uasync_call_soon to avoid
UAF in etcp_cbk_fire iteration (removing ncd_up_cb corrupts saved next ptr)
- ncd_recv_control_handler: use deferred close for CLOSE handlers instead of
direct etcp_connection_close to avoid nested etcp_cbk_fire UAF
- node_conn_direct_open/open_node: use ni->node_name if available for
conn name, fallback to 'n_<node_id>' instead of NULL
- init_connections: build ni_tmp in link loop with only current link's addr,
preventing NCD from creating links for ALL addrs on first socket
- utun_instance_reload: same per-link ni_tmp fix
- init_connections/reload: set ni_tmp.node_name = client->name for
test find_client_link by name
- chat_sync.c: move pending_invite clear from INFO_RESP to WELCOME,
post GUI_EVT_CONNECT_RESULT on invite success in cs_handle_welcome
- etcp_connections.c: NORM_DECRYPT_TRY/CRYPTO_DECRYPT_OK DEBUG_DEBUG->TRACE
- secure_channel.c: ENCRYPT/DECRYPT DEBUG_DEBUG->TRACE
- conn_mgr_init(TOPO_GROUP* group) — работает внутри заданной группы
- conn_mgr_connect_node: загрузка узла из SQLite если нет в BGP-таблице
- topo_node_sqlite_node_load(): загрузка TOPO_NODE (ref=0) из БД
- cm_start_phase_direct: db_loaded-ветка — перебор всех адресов без NAT-фильтра
- cm_cleanup_db_node: удаление TOPO_NODEQ из группы при неудаче, только если нет путей
- cm_start_phase_indirect: исправлено зависание при best_candidates==0 (cm_deliver_result)
- conn_mgr_set_direct_timeout_ms(): конфигурируемый таймаут фазы DIRECT
- conn_mgr_destroy до etcp_router_destroy (исправлен порядок)
- test_conn_mgr: +2 сценария (DB-успех, DB-недоступен), проверка topo_group
ready_cbks/etcp_conn_set_ready_cbk/add_ready/remove_ready → init_cbks/..._init_cbk
ca_ready_cb → ca_init_cb
cm_direct_ready_cb → cm_direct_init_cb
cm_reverse_ready_cb → cm_reverse_init_cb
ntp_node_on_conn_ready → ntp_node_on_conn_init
connect_ready_cb → connect_init_cb
init accurately describes when conn->initialized=1, before keepalives establish UP.
Also fire etcp_fire_conn_status INIT at this point instead of create time.
- Fix wrong categories: BGP/DEBUG -> ETCP for connection lifecycle events
- Fix level: WARN -> DEBUG for normal out-of-order pkts before init
- Fix level: INFO -> DEBUG for per-packet TX DATA spam
- Remove duplicate socket init logs (GENERAL duplicates of ETCP)
- Add missing INFO: INIT_RESPONSE received (client), INIT_RESPONSE sent (server)
- Unify all connection establishment messages under ETCP category
- Remove redundant DEBUG address prints in INIT send (info already in INFO msg)
- PING/PONG operations: BGP -> CONNECTION
- DIRECT IP detection: BGP -> NAT
- find_third_node: use nd->inst->connections instead of group->senders_list
- on_conn_up callback auto-calls trigger_checks (no BGP handshake needed)
- Remove group param from link_ready, request_check_all — nat_detection self-sufficient
- Remove trigger_checks + route_ping_cancel_for_conn from public API (internal only)
- topo_group.c: removed last nat_detection_trigger_checks call — completely decoupled
- Header: 4-line summary, no duplicate function list, 7 public functions
etcp_conn_set_peer_node_id at state=0 updates conn->peer_node_id but
does not reindex the queue entry (key stays 0). Check the queue entry
directly to detect unindexed outbound connections.
- chat_core.c: CREATE TABLE IF NOT EXISTS peers_* in ensure_channel_ready
- utun_node.cpp: GUI_ERROR redirected to debug log file via guiDebugFile()
- etcp_connections.c: reuse unindexed outbound conn (peer_node_id==0)
on cross-connect INIT to avoid 'Can not insert link to socket'
- etcp_connections: keepalive SEND/RECV/DECRYPT_OK: DEBUG→TRACE; DECRYPT_FAIL: WARN→DEBUG
- db_sync_on_conn_up: skip sync if !conn->initialized, wait for next UP event
- db_sync peer_check_timer+instance_add: also check conn->initialized
- chat_sync cs_on_conn_up: ignore conn_up before initialized (cs_send fails otherwise)
- db_sync_on_conn_up log now includes init= and links=
Protocol: added collision field to INIT_REQUEST_PKT (1 byte before ed25519_pubkey).
When two peers connect simultaneously, the one with is_server=0 (API-created
outbound link) is the master. Resolution by smaller node_id:
- We have master link + smaller node_id → send our INIT with collision=1
- We have master link + larger node_id → yield, process as slave
- Remote sends collision=1 → become slave unconditionally
- session_id change always forces reinit (genuine restart)
Removed unconditional etcp_conn_reinit on INIT_RESPONSE(0x03) — guarded by
reset_done flag. Server-side reinit now has collision check BEFORE reset_done
guard; session_id change penetrates reset_done protection.
Closes cross-connect reinit loop causing endless UP/DOWN flapping.
In etcp_conn_ready / etcp_on_up / etcp_on_down / etcp_connection_create /
tcp_server_on_link, callback chains were iterated with:
while (cbe) { cbe->fn(...); cbe = cbe->next; }
If the callback removes itself from the chain (e.g. ca_ready_cb calls
etcp_conn_remove_ready_cbk which u_free's the entry), cbe->next reads
freed memory → SIGSEGV.
Fixed by saving next pointer before invoking the callback:
while (cbe) { n = cbe->next; cbe->fn(...); cbe = n; }
Problem: when two peers connect simultaneously, each incoming
INIT_REQUEST(0x02) or INIT_RESPONSE(0x03) triggers etcp_conn_reinit
unconditionally, causing endless UP/DOWN flapping loop.
Fix: add reset_done flag to ETCP_CONN:
- 0 at creation and after explicit reinit (reinit allowed)
- set to 1 in etcp_conn_ready (connection stable, block reinit)
Three call sites guarded with !conn->reset_done:
- client: handle_init_response_client (INIT_RESPONSE 0x03)
- server: existing link INIT_REQUEST processing
- server: new link INIT_REQUEST processing
send_reset logic preserved unconditionally — only etcp_conn_reinit
itself is blocked when already stable.
- RECV: log every packet (src addr, link found, session_ready, decrypt result)
- RECV: log init decrypt rejection with packet code (catches 0x03 INIT_RESPONSE
silently dropped by init path)
- SEND: log destination addr+fd in send_init_response and etcp_encrypt_send
- REINIT: log full reason in server and client paths (code, session, got_init,
initialized, links_up state before reinit)
When two peers send INIT simultaneously, the first link occupies the socket.
Incoming INIT arriving after finds the existing outbound link by (ip,port)
instead of failing with 'Can not insert link to socket'.
new_conn is only true for the first INIT after server restart.
Subsequent INITs from same peer (recovery loop) see new_conn=0
and revert to old logic, sending 0x05 (no reset) and overwriting
the first 0x03 response.
got_initial_pkt stays 0 until client actually sends seq=1, so
server keeps sending 0x03 across all INIT retries until client
resets and starts from seq=1.
When server restarts, new ETCP_CONN has session_id=0 (calloc).
If client sends INIT_REQUEST_NOINIT (0x04) with session_id=0,
the condition conn->session_id != session_id is false (both 0),
so no reset was sent. Client kept old seq numbers while server
expected seq=1, causing 'Waiting for initial packet' deadlock.
Fix: add || new_conn to the condition — when server created a
fresh connection (was restarted), always send reset to client.
Server now checks both code==ETCP_INIT_REQUEST and session_id mismatch
to decide send_reset. Previously only session_id was checked, so if
session matched (both sides had 00000000), reinit was skipped even
when client explicitly requested reset (code 0x02).
Client side already correct: handle_init_response_client:1473 calls
etcp_conn_reinit if pkt_code==ETCP_INIT_RESPONSE (0x03).
evgeny
Evgeny