sr-dev March 2025

sr-dev@lists.kamailio.org

10 participants
180 discussions

[kamailio/kamailio] rtpengine: fix race condn (same ip:port) in parallel forks due to gencookie (PR #4139)
by Souradeep Bera 18 May '25

18 May '25

- avoid parallel calls to gencookie from generating the same cookie for rtpengine #### Pre-Submission Checklist    - [ ] Commit message has the format required by CONTRIBUTING guide - [ ] Commits are split per component (core, individual modules, libs, utils, ...) - [ ] Each component has a single commit (if not, squash them into one commit) - [ ] No commits to README files for modules (changes must be done to docbook files in `doc/` subfolder, the README file is autogenerated) #### Type Of Change - [ ] Small bug fix (non-breaking change which fixes an issue) - [ ] New feature (non-breaking change which adds new functionality) - [ ] Breaking change (fix or feature that would change existing functionality) #### Checklist:  - [ ] PR should be backported to stable branches - [ ] Tested changes locally - [ ] Related to issue #XXXX (replace XXXX with an open issue number) #### Description  You can view, comment on, or merge this pull request online at: https://github.com/kamailio/kamailio/pull/4139 -- Commit Summary -- * rtpengine: fix race condn assigning same ip:port due to gencookie in parallel forks -- File Changes -- M src/modules/rtpengine/rtpengine.c (3) -- Patch Links -- https://github.com/kamailio/kamailio/pull/4139.patch https://github.com/kamailio/kamailio/pull/4139.diff -- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/pull/4139 You are receiving this because you are subscribed to this thread. Message ID: <kamailio/kamailio/pull/4139(a)github.com>

4 5

[kamailio/kamailio] generate rpm packages using github actions (PR #4194)
by sergey-safarov 17 May '25

17 May '25

#### Pre-Submission Checklist - [x] Commit message has the format required by CONTRIBUTING guide - [x] Commits are split per component (core, individual modules, libs, utils, ...) - [x] Each component has a single commit (if not, squash them into one commit) - [x] No commits to README files for modules (changes must be done to docbook files in `doc/` subfolder, the README file is autogenerated) #### Type Of Change - [x] Small bug fix (non-breaking change which fixes an issue) - [ ] New feature (non-breaking change which adds new functionality) - [ ] Breaking change (fix or feature that would change existing functionality) #### Checklist: - [x] PR should be backported to stable branches - [x] Tested changes locally - [ ] Related to issue #XXXX (replace XXXX with an open issue number) #### Description I have refactored the rpm packaging and am now using the GitHub actions. Now `tar.gz` archive with rpm packages, source rpm and deps rpm files attached to the release message like. You can view, comment on, or merge this pull request online at: https://github.com/kamailio/kamailio/pull/4194 -- Commit Summary -- * generate rpm packages using github actions -- File Changes -- A .github/workflows/rpm.yml (57) -- Patch Links -- https://github.com/kamailio/kamailio/pull/4194.patch https://github.com/kamailio/kamailio/pull/4194.diff -- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/pull/4194 You are receiving this because you are subscribed to this thread. Message ID: <kamailio/kamailio/pull/4194(a)github.com>

2 5

[kamailio/kamailio] rtpengine: select node with weight=0 if no other nodes online (Issue #4180)
by sergey-safarov 13 May '25

13 May '25

sergey-safarov created an issue (kamailio/kamailio#4180) ### Description I have installed two `rtpengine` instances and want to use them as an `active backup`. I want to use `backup` only when `active` is unavailable. Closest settings for this rtpengine node `weight` and want to use settings like ``` MariaDB [kamailio]> select * from rtpengine; +----+-------+---------------------------------+--------+----------+---------------------+ | id | setid | url | weight | disabled | stamp | +----+-------+---------------------------------+--------+----------+---------------------+ | 1 | 0 | udp6:[2005:84c0:bf:11::20]:2223 | 1 | 0 | 1900-01-01 00:00:01 | | 2 | 0 | udp6:[2005:84c0:bf:11::21]:2223 | 0 | 0 | 1900-01-01 00:00:01 | +----+-------+---------------------------------+--------+----------+---------------------+ ``` But in case `weight=0` node is not selected and handled as disabled. It will be fine do not use rtpengine nodes with `weight=0` when other rtpengine with `weight!=0` are available. But if no other nodes are available, use `weight=0` nodes as last resort. ### Expected behavior Do not use rtpengine nodes with `weight=0` when other rtpengine with `weight!=0` are available. But if no other nodes are available, use `weight=0` nodes as last resort. #### Actual observed behavior The node is not selected and handled as disabled when `weight=0`. #### Log Messages ``` rtpengine [rtpengine.c:3430]: rtpp_test(): rtpengine instance <udp6:[2005:84c0:bf:11::21]:2223> found, support for it enabled rtpengine [rtpengine.c:3883]: select_rtpp_node(): rtpengine failed to select new for calllen=27 callid=1-10824@2005:84c0:bf:11::22 rtpengine [rtpengine.c:3183]: rtpp_function_call(): no available proxies ``` -- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/4180 You are receiving this because you are subscribed to this thread. Message ID: <kamailio/kamailio/issues/4180(a)github.com>

3 3

[kamailio/kamailio] WSS-connections show up as <null> in $tls_version and $tls_cypher_info (Issue #4167)
by Christian Berger 12 May '25

12 May '25

ChristianBergerSipgate created an issue (kamailio/kamailio#4167) Hi, we have found a potential bug. The pseudo variables return null for WSS connections. Example config: `xlog("L_INFO", "<mainLogic> REGISTER: TLS or WSS detected: fU=$fU tls_version=$tls_version tls_cipher_info=$tls_cipher_info ua=$ua\n");` Log output: `Mar 7 08:37:28 sip-tcploadbalancer01 /usr/sbin/kamailio[1268940]: { "level": "ERROR", "module": "tls", "file": "tls_select.c", "line": 131, "function": "get_cur_connection", "message": "Transport protocol is not TLS (bug in config)\n" } Mar 7 08:37:28 sip-tcploadbalancer01 /usr/sbin/kamailio[1268940]: { "level": "INFO", "module": "tls", "file": "tls_select.c", "line": 310, "function": "get_version", "message": "TLS connection not found in select_version\n" } Mar 7 08:37:28 sip-tcploadbalancer01 /usr/sbin/kamailio[1268940]: { "level": "ERROR", "module": "tls", "file": "tls_select.c", "line": 131, "function": "get_cur_connection", "message": "Transport protocol is not TLS (bug in config)\n" } Mar 7 08:37:28 sip-tcploadbalancer01 /usr/sbin/kamailio[1268940]: { "level": "INFO", "module": "tls", "file": "tls_select.c", "line": 201, "function": "get_cipher", "message": "TLS connection not found in select_cipher\n" } Mar 7 08:37:28 sip-tcploadbalancer01 /usr/sbin/kamailio[1268940]: { "level": "INFO", "module": "xlog", "file": "xlog.c", "line": 278, "function": "", "message": "<mainLogic> REGISTER: TLS or WSS detected: fU=1125411e0 tls_version=<null> tls_cipher_info=<null> ua=webphone\n" }` As you can see both tls_version and tls_cipher_info return <null> even though the underlying connection is via WSS. We have looked into it, and it seems like `get_cur_connection` in `modules/tls/tls_select.c` only checks for TLS, but not WSS. ``` struct tcp_connection *get_cur_connection(struct sip_msg *msg) { struct tcp_connection *c; if(_tls_pv_con != 0) return _tls_pv_con; if(msg->rcv.proto != PROTO_TLS) { ERR("Transport protocol is not TLS (bug in config)\n"); return 0; } c = tcpconn_get(msg->rcv.proto_reserved1, 0, 0, 0, cfg_get(tls, tls_cfg, con_lifetime)); if(c && c->type != PROTO_TLS) { ERR("Connection found but is not TLS\n"); tcpconn_put(c); return 0; } return c; } ``` We think that checking for `PROTO_WSS` might solve the issue. -- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/4167 You are receiving this because you are subscribed to this thread. Message ID: <kamailio/kamailio/issues/4167(a)github.com>

4 5

[kamailio/kamailio] cmake: switching deb rules (Issue #4053)
by Victor Seva 10 May '25

10 May '25

Opening this generic issue to track issues when trying to switch deb package generation to cmake: -- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/4053 You are receiving this because you are subscribed to this thread. Message ID: <kamailio/kamailio/issues/4053(a)github.com>

5 56

[kamailio/kamailio] Segfault when using kamailio 5.8.4 as a presence proxy (Issue #4102)
by Joe Mordica 08 May '25

08 May '25

### Description We are now using kamailio on the edge as a presence proxy to scale presence. We have 3 internal kamailio instances handling the actual subscriptions and publishing. This works well, however, once per 24-36hours we experience a crash on the border proxy instance. We did not experience this when using the border proxy as a presence server (there was no proxying SUBSCRIBE). ### Troubleshooting #### Reproduction I believe this is happening when a customers network randomly experiences issues where it doesn't receive our replies to subscribes (and continues to send a subscribe). Logs are below. #### Debugging Data  Here's the output while replacing any sensitive info in the payload. ``` GNU gdb (Debian 13.1-3) 13.1 Copyright (C) 2023 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-linux-gnu". Type "show configuration" for configuration details. For bug reporting instructions, please see: <https://www.gnu.org/software/gdb/bugs/>. Find the GDB manual and other documentation resources online at: <http://www.gnu.org/software/gdb/documentation/>. For help, type "help". Type "apropos word" to search for commands related to "word"... Reading symbols from /usr/sbin/kamailio... Reading symbols from /usr/lib/debug/.build-id/ce/c19d0ed1a928e4e25a84ae20d77e9c904a4892.debug... warning: Can't open file /dev/zero (deleted) during file-backed mapping note processing [New LWP 26] [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". Core was generated by `kamailio -DD -E -m 8000 -M 512 -f /etc/kamailio/kamailio.cfg -P /var/run/kamail'. Program terminated with signal SIGSEGV, Segmentation fault. #0 via_builder (len=len@entry=0x7ffed7e3ecb8, msg=msg@entry=0x7ff9529548b0, send_info=send_info@entry=0x7ff764d20bb0, branch=branch@entry=0x7ffed7e3ecc0, extra_params=0x0, hp=hp@entry=0x7ffed7e3ebb0) at core/msg_translator.c:2910 2910 core/msg_translator.c: No such file or directory. (gdb) bt full #0 via_builder (len=len@entry=0x7ffed7e3ecb8, msg=msg@entry=0x7ff9529548b0, send_info=send_info@entry=0x7ff764d20bb0, branch=branch@entry=0x7ffed7e3ecc0, extra_params=0x0, hp=hp@entry=0x7ffed7e3ebb0) at core/msg_translator.c:2910 via_len = <optimized out> extra_len = <optimized out> line_buf = <optimized out> max_len = <optimized out> via_prefix_len = <optimized out> address_str = <optimized out> port_str = 0x0 send_sock = 0x0 comp_len = <optimized out> comp_name_len = <optimized out> port = <optimized out> proto = <optimized out> ip = {af = 536870912, len = 0, u = {addrl = {534820896, 1586192}, addr32 = {534820896, 0, 1586192, 0}, addr16 = {47136, 8160, 0, 0, 13328, 24, 0, 0}, addr = " \270\340\037\000\000\000\000\0204\030\000\000\000\000"}} from = 0x0 local_addr = {s = {sa_family = 16, sa_data = "#\000\000\000\000\000\020\000\000\000\000\000\000"}, sin = {sin_family = 16, sin_port = 35, sin_addr = {s_addr = 0}, sin_zero = "\020\000\000\000\000\000\000"}, sin6 = {sin6_family = 16, sin6_port = 35, sin6_flowinfo = 0, sin6_addr = {__in6_u = { __u6_addr8 = "\020\000\000\000\000\000\000\000\003\000\000\000\000\000\000", __u6_addr16 = {16, 0, 0, 0, 3, 0, 0, 0}, __u6_addr32 = {16, 0, 3, 0}}}, sin6_scope_id = 3000082688}, sas = {ss_family = 16, __ss_padding = "#\000\000\000\000\000\020\000\000\000\000\000\000\000\003\000\000\000\000\000\000\000\000\241\321\262cg\357`@\000\000\000\000\000\000\000X\f-\177!V\000\000\020\016\217R\371\177\000\000\001\000\000\000!V\000\000*\000\000\000\000\000\000\000\001\000\000\000\376\177\000\000\004\360;\177!V\000\000\250,A\177!V\000\000\000\355\343\327\376\177\000\000\000\241\321\262cg\357`h\256\006R\371\177\000", __ss_align = 94701866020241}} con = 0x0 rxavp = 0x0 xname = {s = 0x0, len = -2141140176} __func__ = "via_builder" #1 0x000056217f155030 in create_via_hf (len=len@entry=0x7ffed7e3ecb8, msg=msg@entry=0x7ff9529548b0, send_info=send_info@entry=0x7ff764d20bb0, branch=branch@entry=0x7ffed7e3ecc0) at core/msg_translator.c:3232 via = <optimized out> extra_params = {s = 0x0, len = 0} hp = {host = 0x56217f4fb380 <default_global_address>, port = 0x56217f4fb370 <default_global_port>} sbuf = "\331\023\000\000\001", '\000' <repeats 18 times> --Type <RET> for more, q to quit, c to continue without paging--info locals slen = <optimized out> xparams = <optimized out> id_buf = <optimized out> id_len = 0 __func__ = "create_via_hf" __llevel = <optimized out> __kld = <optimized out> #2 0x000056217f1592f7 in build_req_buf_from_sip_req (msg=msg@entry=0x7ff9529548b0, returned_len=returned_len@entry=0x7ffed7e3eeac, send_info=send_info@entry=0x7ff764d20bb0, mode=mode@entry=128) at core/msg_translator.c:2096 len = 581 new_len = <optimized out> received_len = 0 rport_len = 0 uri_len = 0 via_len = 0 body_delta = 0 line_buf = 0x0 received_buf = 0x0 rport_buf = 0x0 new_buf = 0x0 buf = 0x56217f5866c0 <buf> "SUBSCRIBE sip:74@sip.domain.co SIP/2.0\r\nVia: SIP/2.0/UDP 12.171.207.82:25744;branch=z9hG4bK1805670803;rport\r\nFrom: \"Heather Myers\" <sip:1086448@sip.domain.co>;tag=1805518213\r\nTo: <sip:74@sip.domain.co>\r\nCa"... path_buf = {s = 0x0, len = 0} offset = 0 s_offset = 0 size = <optimized out> via_anchor = 0x7ff952951300 via_lump = <optimized out> via_rm = <optimized out> via_insert_param = 0x0 path_anchor = <optimized out> path_lump = <optimized out> branch = {s = 0x7ff952954ed8 "z9hG4bKc7ed.5819b4295fe83f981a2fc19b7f739d7e.0", len = 46} flags = 262273 --Type <RET> for more, q to quit, c to continue without paging--list udp_mtu = <optimized out> di = {send_sock = 0x0, to = {s = {sa_family = 18608, sa_data = "\225R\371\177\000\000\002\000\023\331\n4\001\r"}, sin = {sin_family = 18608, sin_port = 21141, sin_addr = { s_addr = 32761}, sin_zero = "\002\000\023\331\n4\001\r"}, sin6 = {sin6_family = 18608, sin6_port = 21141, sin6_flowinfo = 32761, sin6_addr = {__in6_u = { __u6_addr8 = "\002\000\023\331\n4\001\r\000\000\000\000\000\000\000", __u6_addr16 = {2, 55571, 13322, 3329, 0, 0, 0, 0}, __u6_addr32 = {3641901058, 218182666, 0, 0}}}, sin6_scope_id = 0}, sas = {ss_family = 18608, __ss_padding = "\225R\371\177\000\000\002\000\023\331\n4\001\r", '\000' <repeats 103 times>, __ss_align = 0}}, id = 0, send_flags = {f = 0, blst_imask = 0}, proto = 0 '\000', proto_pad0 = 0 '\000', proto_pad1 = 0} ret = <optimized out> __func__ = "build_req_buf_from_sip_req" error00 = <optimized out> #3 0x00007ff951fbe90d in prepare_new_uac (t=t@entry=0x7ff764d20880, i_req=i_req@entry=0x7ff9529548b0, branch=branch@entry=0, uri=<optimized out>, uri@entry=0x7ff9529548e8, path=<optimized out>, path@entry=0x7ff952954f60, next_hop=<optimized out>, fsocket=0x7ff9528f0e10, snd_flags=..., fproto=<optimized out>, flags=<optimized out>, instance=<optimized out>, ruid=<optimized out>, location_ua=<optimized out>) at ./src/modules/tm/t_fwd.c:482 shbuf = 0x0 add_rm_backup = <optimized out> body_lumps_backup = <optimized out> parsed_uri_bak = {user = { s = 0x56217f5866ce <buf+14> "74(a)sip.domain.co SIP/2.0\r\nVia: SIP/2.0/UDP 12.171.207.82:25744;branch=z9hG4bK1805670803;rport\r\nFrom: \"Heather Myers\" <sip:1086448@sip.domain.co>;tag=1805518213\r\nTo: <sip:74@sip.domain.co>\r\nCall-ID: 0_18056"..., len = 2}, passwd = {s = 0x0, len = 0}, host = { s = 0x56217f5866d1 <buf+17> "sip.domain.co SIP/2.0\r\nVia: SIP/2.0/UDP 12.171.207.82:25744;branch=z9hG4bK1805670803;rport\r\nFrom: \"Heather Myers\" <sip:1086448@sip.domain.co>;tag=1805518213\r\nTo: <sip:74@sip.domain.co>\r\nCall-ID: 0_18056078"..., len = 11}, port = {s = 0x0, len = 0}, params = {s = 0x0, len = 0}, sip_params = {s = 0x0, len = 0}, headers = { s = 0x0, len = 0}, port_no = 0, proto = 0, type = SIP_URI_T, flags = 0, transport = {s = 0x0, len = 0}, ttl = {s = 0x0, len = 0}, user_param = {s = 0x0, len = 0}, maddr = { s = 0x0, len = 0}, method = {s = 0x0, len = 0}, lr = {s = 0x0, len = 0}, r2 = {s = 0x0, len = 0}, gr = {s = 0x0, len = 0}, transport_val = {s = 0x0, len = 0}, ttl_val = { s = 0x0, len = 0}, user_param_val = {s = 0x0, len = 0}, maddr_val = {s = 0x0, len = 0}, method_val = {s = 0x0, len = 0}, lr_val = {s = 0x0, len = 0}, r2_val = {s = 0x0, len = 0}, gr_val = {s = 0x0, len = 0}} ret = -1 len = 32761 parsed_uri_ok_bak = <optimized out> free_new_uri = 1 msg_uri_bak = {s = 0x0, len = 0} dst_uri_bak = {s = 0x7ff9529525f0 "sip:10.52.1.13:5081", len = 19} dst_uri_backed_up = 1 path_bak = {s = 0x0, len = 0} free_path = 1 instance_bak = {s = 0x0, len = 0} ``` #### Log Messages  ``` ERROR 2025-01-10T06:00:31.847348563Z [resource.labels.containerName: kamailio] 8(28) CRITICAL: {1 1 SUBSCRIBE 0_1907143137(a)192.168.100.28} tm [timer.h:188]: _set_fr_retr(): already added: 0x7ff763e62250 , tl=0x7ff763e62270!!! ERROR 2025-01-10T06:00:31.847388483Z [resource.labels.containerName: kamailio] 8(28) CRITICAL: {1 1 SUBSCRIBE 0_1907143137(a)192.168.100.28} tm [t_fwd.c:1613]: t_send_branch(): BUG: retransmission already started for: 0x7ff763e62250 ERROR 2025-01-10T06:00:31.847394809Z [resource.labels.containerName: kamailio] 8(28) ERROR: {1 1 SUBSCRIBE 0_1907143137(a)192.168.100.28} sl [sl_funcs.c:428]: sl_reply_error(): stateless error reply used: No error (0/SL) ERROR 2025-01-10T06:00:31.847400738Z [resource.labels.containerName: kamailio] 8(28) BUG: {1 1 SUBSCRIBE 0_1907143137(a)192.168.100.28} tm [t_lookup.c:2052]: t_unref(): REQ_ERR DELAYED should have been caught much earlier for 0x7ff763e61f70: 24 (hex 18) ERROR 2025-01-10T06:00:47.346744773Z [resource.labels.containerName: kamailio] 45(65) ERROR: <core> [core/tcp_main.c:4818]: tcpconn_main_timeout(): connect 64.63.142.90:12190 failed (timeout) ERROR 2025-01-10T06:00:58.274915089Z [resource.labels.containerName: kamailio] 45(65) CRITICAL: <core> [core/pass_fd.c:281]: receive_fd(): EOF on 20 ERROR 2025-01-10T06:01:02.715789076Z [resource.labels.containerName: kamailio] 0(1) ALERT: <core> [main.c:805]: handle_sigs(): child process 26 exited by a signal 11 ERROR 2025-01-10T06:01:02.715827333Z [resource.labels.containerName: kamailio] 0(1) ALERT: <core> [main.c:809]: handle_sigs(): core was generated ERROR 2025-01-10T06:01:02.718815287Z [resource.labels.containerName: kamailio] 45(65) CRITICAL: <core> [core/pass_fd.c:281]: receive_fd(): EOF on 25 ERROR 2025-01-10T06:01:02.720985493Z [resource.labels.containerName: kamailio] 45(65) CRITICAL: <core> [core/pass_fd.c:281]: receive_fd(): EOF on 23 ERROR 2025-01-10T06:01:02.725450552Z [resource.labels.containerName: kamailio] 45(65) CRITICAL: <core> [core/pass_fd.c:281]: receive_fd(): EOF on 30 ``` #### SIP Traffic  ``` (paste your sip traffic here) ``` ### Possible Solutions I can atleast say that when this segfault happens that these logs are always present right before the crash. However, I also see instances of these logs without a crash immediately following. On the proxy we are doing this is withindlg route to send subscribes to the presence route: ``` if KSR.is_SUBSCRIBE() and KSR.is_myself_ruri() then -- in-dialog subscribe requests ksr_route_presence(); KSR.x.exit(); end ``` Any then this is our ksr_route_presence(): ``` function ksr_route_presence() if not KSR.is_SUBSCRIBE() then return 1; end local fuser = KSR.kx.get_fuser(); if fuser=='' then KSR.xlog.xinfo('Subscription attempt without username from IP ' .. KSR.kx.get_srcip() .. ' - Rejecting\n'); KSR.sl.sl_send_reply(404, 'Not Found'); KSR.x.exit(); end KSR.pv.sets('$avp(s:tenant)', string.match(fuser, '-(.*)')); local setId = '6' .. KSR.kx.get_def('DEFAULT_DISPATCH_SET'); local modeId = '7'; if KSR.dispatcher.ds_select_dst(setId, modeId)<0 then KSR.sl.send_reply(404, 'No destination'); KSR.x.exit(); end KSR.corex.set_send_socket(KSR.kx.get_def('PRIVATE_LISTEN_IP') .. ':' .. KSR.kx.get_def('INTERNAL_PORT')); KSR.tm.t_on_failure('ksr_route_rtf_dispatch'); ksr_route_relay(); end ``` ### Additional Information * **Kamailio Version** - output of `kamailio -v` ``` version: kamailio 5.8.4 (x86_64/linux) flags: USE_TCP, USE_TLS, USE_SCTP, TLS_HOOKS, USE_RAW_SOCKS, DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MMAP, PKG_MALLOC, MEM_JOIN_FREE, Q_MALLOC, F_MALLOC, TLSF_MALLOC, DBG_SR_MEMORY, USE_FUTEX, FAST_LOCK-ADAPTIVE_WAIT, USE_DNS_CACHE, USE_DNS_FAILOVER, USE_NAPTR, USE_DST_BLOCKLIST, HAVE_RESOLV_RES, TLS_PTHREAD_MUTEX_SHARED ADAPTIVE_WAIT_LOOPS 1024, MAX_RECV_BUFFER_SIZE 262144, MAX_SEND_BUFFER_SIZE 262144, MAX_URI_SIZE 1024, BUF_SIZE 65535, DEFAULT PKG_SIZE 8MB poll method support: poll, epoll_lt, epoll_et, sigio_rt, select. id: unknown compiled with gcc 12.2.0 ``` * **Operating System**:  ``` PRETTY_NAME="Debian GNU/Linux 12 (bookworm)" NAME="Debian GNU/Linux" VERSION_ID="12" VERSION="12 (bookworm)" VERSION_CODENAME=bookworm ID=debian HOME_URL="https://www.debian.org/" SUPPORT_URL="https://www.debian.org/support" BUG_REPORT_URL="https://bugs.debian.org/" ``` ``` Linux gke-us-south1-external-sip-800ca69e-qyb3 5.15.0-1048-gke #53-Ubuntu SMP Tue Nov 28 00:39:01 UTC 2023 x86_64 GNU/Linux ``` -- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/4102 You are receiving this because you are subscribed to this thread. Message ID: <kamailio/kamailio/issues/4102(a)github.com>

5 31

[kamailio/kamailio] Kamailio blocks in SSL_accept (Issue #4185)
by Mathias Schneuwly 08 May '25

08 May '25

schneuwlym created an issue (kamailio/kamailio#4185) ### Description This issue is a follow up of the discussion in the [mailing list](https://lists.kamailio.org/mailman3/hyperkitty/list/sr-users@lists.ka… We are using Kamailio 5.8.4 in a yocto scarthgap based image. We noticed on multiple nodes, that suddenly phones lost their registrations and the log showed messages like `send_fd(): sendmsg failed sending 55 on 52: Too many references: cannot splice (109)`. After debugging we guess that the `send_fd` messages are just a symptom, not the cause. Generating a backtrace using `kamctl trap` shows that all TCP receiver are waiting for a `pthread_rwlock_...` lock. Our deployments, where we randomly see the issue, are very small; 10-30 registrations. I've tested kamailio version 5.8.4, 5.8.5, 5.8.6 and 6.0.1 and I was able to reproduce the issue with all versions. I've also tested version 5.5.4, but with an older yocto (dunfell) which used openssl 1.1.1w, **which did not block!** Yocto Scarthgap uses OpenSSL 3.2.3 and we also updated to version 3.4.1, which didn't show any difference. ### Troubleshooting #### Reproduction Meanwhile we found a way to reproduce the issue within seconds. It requires a bit of brute force, but it locks kamailio reliable. We are generating 3 parallel SIP REGISTER streams with sipexer to 3 different accounts. Following an example of one stream: ``` watch -n 0.1 "./sipexer -register -vl 3 -co -com -ex 3600 -fu "2" -au "2" -ap "test" -tk <path to key>.key -tc <path to certificate>.cer -ti -sd -ru sip:<domain> -cu sip:2@<domain> tls:<fqdn>:5061;sleep 0.1;./sipexer -register -vl 3 -co -com -ex 0 -fu "2" -au "2" -ap "test" -tk <path to key>.key -tc <path to cert>.cer -ti -sd -ru sip:<domain> -cu sip:2@<domain> tls:<fqdn>:5061" ``` Our dialplan is quite complex, but I was also able to reproduce the issue with the default [dialplan](https://raw.githubusercontent.com/kamailio/kamailio/refs/heads/5.… with the following changes: ``` 169a170,171 > #!define WITH_TLS > 205a208,219 > alias=udp:<domain>:5060 > alias=udp:127.0.0.1:5060 > alias=udp:10.0.6.1:5060 > alias=tls:<domain> > alias=tls:10.0.6.1:5061 > alias=tls:10.165.42.76:5061 > alias=tls:<domain>:5061 > alias=udp:<domain> > listen=udp:127.0.0.1:5060 > listen=tls:10.0.6.1:5061 > listen=udp:10.0.6.1:5060 > listen=tls:10.165.42.76:5061 498c512 < modparam("tls", "config", "/usr/local/etc/kamailio/tls.cfg") --- > modparam("tls", "config", "/etc/kamailio/kamailio-tls.cfg") ``` ``` # cat /etc/kamailio/kamailio-tls.cfg [server:default] method = TLSv1+ verify_certificate = yes require_certificate = yes private_key = "<path to key>.key" certificate = "<path to cert>.crt" ca_list = "<path to ca>.pem" [client:default] method = TLSv1+ verify_certificate = yes require_certificate = yes private_key = "<path to key>.key" certificate = "<path to cert>.crt" ca_list = "<path to ca>.pem" ``` #### Debugging Data Following the backtrace of one tcp listener process. The full backtrace can be found [here](https://github.com/user-attachments/files/19430313/gdb_kamailio_20250… ``` ---start 4608 ----------------------------------------------------- [Thread debugging using libthread_db enabled] Using host libthread_db library "/usr/lib/libthread_db.so.1". __futex_abstimed_wait_common (futex_word=futex_word@entry=0x7fc855758a0c, expected=expected@entry=3, clockid=clockid@entry=0, abstime=abstime@entry=0x0, private=<optimized out>, cancel=cancel@entry=false) at futex-internal.c:103 warning: 103 futex-internal.c: No such file or directory #0 __futex_abstimed_wait_common (futex_word=futex_word@entry=0x7fc855758a0c, expected=expected@entry=3, clockid=clockid@entry=0, abstime=abstime@entry=0x0, private=<optimized out>, cancel=cancel@entry=false) at futex-internal.c:103 err = -512 clockbit = 256 op = <optimized out> #1 0x00007fc85a59429c in __GI___futex_abstimed_wait64 (futex_word=futex_word@entry=0x7fc855758a0c, expected=expected@entry=3, clockid=clockid@entry=0, abstime=abstime@entry=0x0, private=<optimized out>) at futex-internal.c:128 No locals. #2 0x00007fc85a59d830 in __pthread_rwlock_wrlock_full64 (abstime=0x0, clockid=0, rwlock=0x7fc855758a00) at /usr/src/debug/glibc/2.39+git/nptl/pthread_rwlock_common.c:730 private = <optimized out> wf = <optimized out> err = <optimized out> prefer_writer = false may_share_futex_used_flag = true wpf = <optimized out> ready = <optimized out> r = <optimized out> may_share_futex_used_flag = <optimized out> r = <optimized out> done = <optimized out> wpf = <optimized out> ready = <optimized out> __value = <optimized out> prefer_writer = <optimized out> private = <optimized out> wf = <optimized out> err = <optimized out> w = <optimized out> w = <optimized out> private = <optimized out> err = <optimized out> w = <optimized out> wf = <optimized out> wf = <optimized out> __value = <optimized out> #3 ___pthread_rwlock_wrlock (rwlock=0x7fc855758a00) at pthread_rwlock_wrlock.c:26 result = <optimized out> #4 0x00007fc859820b49 in CRYPTO_THREAD_write_lock (lock=<optimized out>) at /usr/src/debug/openssl/3.2.3/crypto/threads_pthread.c:108 No locals. #5 0x00007fc8599bd5da in ossl_prov_drbg_generate (drbg=0x7fc855710040, out=0x7fc856471980 "", outlen=48, strength=256, prediction_resistance=0, adin=0x7ffeec8b67e0 "@", adinlen=8) at /usr/src/debug/openssl/3.2.3/providers/implementations/rands/drbg.c:645 fork_id = <optimized out> reseed_required = 0 ret = 0 __func__ = "ossl_prov_drbg_generate" err = <optimized out> now = <optimized out> #6 0x00007fc8599bd9ca in ossl_drbg_get_seed (vdrbg=<optimized out>, pout=0x7ffeec8b6880, entropy=<optimized out>, min_len=<optimized out>, max_len=<optimized out>, prediction_resistance=0, adin=0x7ffeec8b6838 "\220'\037V\310\177", adin_len=8) at /usr/src/debug/openssl/3.2.3/providers/implementations/rands/drbg.c:175 drbg = 0x7fc855710040 bytes_needed = 48 buffer = 0x7fc856471980 "" __func__ = "ossl_drbg_get_seed" #7 0x00007fc8599bc907 in get_entropy (drbg=<optimized out>, drbg@entry=0x7fc8561f2790, pout=pout@entry=0x7ffeec8b6880, entropy=<optimized out>, min_len=48, max_len=48, prediction_resistance=prediction_resistance@entry=0) at /usr/src/debug/openssl/3.2.3/providers/implementations/rands/drbg.c:245 bytes = <optimized out> p_str = 256 __func__ = "get_entropy" #8 0x00007fc8599bd26c in ossl_prov_drbg_reseed_unlocked (drbg=drbg@entry=0x7fc8561f2790, prediction_resistance=prediction_resistance@entry=0, ent=ent@entry=0x0, ent_len=ent_len@entry=0, adin=adin@entry=0x0, adinlen=adinlen@entry=0) at /usr/src/debug/openssl/3.2.3/providers/implementations/rands/drbg.c:573 entropy = 0x0 entropylen = 0 __func__ = "ossl_prov_drbg_reseed_unlocked" #9 0x00007fc8599bd68e in ossl_prov_drbg_generate (adinlen=0, adin=0x0, prediction_resistance=0, strength=<optimized out>, outlen=32, out=0x7fc8564717c8 "", drbg=0x7fc8561f2790) at /usr/src/debug/openssl/3.2.3/providers/implementations/rands/drbg.c:697 fork_id = <optimized out> reseed_required = <optimized out> ret = <optimized out> fork_id = <optimized out> reseed_required = <optimized out> ret = <optimized out> __func__ = "ossl_prov_drbg_generate" err = <optimized out> now = <optimized out> #10 ossl_prov_drbg_generate (drbg=0x7fc8561f2790, out=0x7fc8564717c8 "", outlen=32, strength=<optimized out>, prediction_resistance=0, adin=0x0, adinlen=0) at /usr/src/debug/openssl/3.2.3/providers/implementations/rands/drbg.c:634 fork_id = <optimized out> reseed_required = 0 ret = 0 __func__ = "ossl_prov_drbg_generate" err = <optimized out> now = <optimized out> #11 0x00007fc8597e75c6 in evp_rand_generate_locked (ctx=ctx@entry=0x7fc8561f26f0, out=out@entry=0x7fc8564717c8 "", outlen=outlen@entry=32, strength=strength@entry=0, prediction_resistance=<optimized out>, prediction_resistance@entry=0, addin=0x0, addin_len=0) at /usr/src/debug/openssl/3.2.3/crypto/evp/evp_rand.c:570 chunk = 32 max_request = 65536 params = {{key = 0x7fc8599e9521 "max_request", data_type = 2, data = 0x7ffeec8b6968, data_size = 8, return_size = 8}, {key = 0x0, data_type = 0, data = 0x0, data_size = 0, return_size = 0}} __func__ = "evp_rand_generate_locked" #12 0x00007fc8597e840b in EVP_RAND_generate (ctx=0x7fc8561f26f0, out=out@entry=0x7fc8564717c8 "", outlen=outlen@entry=32, strength=strength@entry=0, prediction_resistance=prediction_resistance@entry=0, addin=addin@entry=0x0, addin_len=0) at /usr/src/debug/openssl/3.2.3/crypto/evp/evp_rand.c:592 res = <optimized out> #13 0x00007fc8598f9282 in RAND_bytes_ex (ctx=warning: could not convert 'ossl_lib_ctx_st' from the host encoding (ANSI_X3.4-1968) to UTF-32. This normally should not happen, please file a bug report. 0x0, buf=0x7fc8564717c8 "", num=32, strength=0) at /usr/src/debug/openssl/3.2.3/crypto/rand/rand_lib.c:380 rand = <optimized out> meth = <optimized out> __func__ = "RAND_bytes_ex" #14 0x00007fc859bdb62c in def_generate_session_id (ssl=ssl@entry=0x7fc856436e50, id=id@entry=0x7fc8564717c8 "", id_len=id_len@entry=0x7ffeec8b6ae4) at /usr/src/debug/openssl/3.2.3/ssl/ssl_sess.c:326 retry = 0 #15 0x00007fc859bdc132 in ssl_generate_session_id (s=s@entry=0x7fc856436e50, ss=ss@entry=0x7fc856471570) at /usr/src/debug/openssl/3.2.3/ssl/ssl_sess.c:409 tmp = 32 cb = 0x7fc859bdb600 <def_generate_session_id> ssl = <optimized out> __func__ = "ssl_generate_session_id" #16 0x00007fc859bdc31e in ssl_get_new_session (s=s@entry=0x7fc856436e50, session=session@entry=1) at /usr/src/debug/openssl/3.2.3/ssl/ssl_sess.c:464 ss = 0x7fc856471570 __func__ = "ssl_get_new_session" #17 0x00007fc859c3f169 in tls_early_post_process_client_hello (s=0x7fc856436e50) at /usr/src/debug/openssl/3.2.3/ssl/statem/statem_srvr.c:1898 i = <optimized out> loop = <optimized out> sctx = 0x7fc855d24280 j = <optimized out> dgrd = DOWNGRADE_NONE protverr = <optimized out> c = <optimized out> clienthello = 0x7fc856470a90 al = 80 id = <optimized out> comp = 0x0 ciphers = 0x7fc856471330 scsvs = 0x7fc8564713c0 ssl = 0x7fc856436e50 j = <optimized out> i = <optimized out> al = <optimized out> protverr = <optimized out> loop = <optimized out> id = <optimized out> comp = <optimized out> c = <optimized out> ciphers = <optimized out> scsvs = <optimized out> clienthello = <optimized out> dgrd = <optimized out> sctx = <optimized out> ssl = <optimized out> __func__ = "tls_early_post_process_client_hello" err = <optimized out> cipher = <optimized out> trc_out = <optimized out> pos = <optimized out> pref_cipher = <optimized out> master_key_length = <optimized out> m = <optimized out> comp_id = <optimized out> k = <optimized out> m = <optimized out> nn = <optimized out> v = <optimized out> done = <optimized out> o = <optimized out> #18 tls_post_process_client_hello (s=0x7fc856436e50, wst=<optimized out>) at /usr/src/debug/openssl/3.2.3/ssl/statem/statem_srvr.c:2295 rv = <optimized out> cipher = <optimized out> ssl = 0x7fc856436e50 err = <optimized out> __func__ = "tls_post_process_client_hello" #19 0x00007fc859c2bc1b in read_state_machine (s=0x7fc856436e50) at /usr/src/debug/openssl/3.2.3/ssl/statem/statem.c:712 mt = 1 pkt = {curr = 0x7fc856466125 "", remaining = 0} max_message_size = 0x7fc859c3e3e0 <ossl_statem_server_max_message_size> ret = <optimized out> process_message = 0x7fc859c42400 <ossl_statem_server_process_message> post_process_message = 0x7fc859c414f0 <ossl_statem_server_post_process_message> st = 0x7fc856436ee0 len = 273 transition = 0x7fc859c3d330 <ossl_statem_server_read_transition> cb = 0x7fc859c91d90 <sr_ssl_ctx_info_callback> ssl = 0x7fc856436e50 st = <optimized out> ret = <optimized out> mt = <optimized out> len = <optimized out> transition = <optimized out> pkt = <optimized out> process_message = <optimized out> post_process_message = <optimized out> max_message_size = <optimized out> cb = <optimized out> ssl = <optimized out> __func__ = "read_state_machine" #20 state_machine (s=0x7fc856436e50, server=1) at /usr/src/debug/openssl/3.2.3/ssl/statem/statem.c:478 buf = 0x0 cb = 0x7fc859c91d90 <sr_ssl_ctx_info_callback> st = 0x7fc856436ee0 ret = -1 ssret = <optimized out> ssl = 0x7fc856436e50 __func__ = "state_machine" #21 0x00007fc859cbf036 in tls_accept (c=c@entry=0x7fc856461be0, error=error@entry=0x7ffeec8b6dd8) at tls_server.c:457 ret = <optimized out> ssl = 0x7fc856436e50 cert = <optimized out> tls_c = 0x7fc8564171b0 tls_log = <optimized out> __func__ = "tls_accept" ``` ### Additional Information * **Kamailio Version** - output of `kamailio -v` ``` version: kamailio 5.8.4 (x86_64/linux) f1d83d flags: USE_TCP, USE_TLS, USE_SCTP, TLS_HOOKS, USE_RAW_SOCKS, DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MMAP, PKG_MALLOC, MEM_JOIN_FREE, Q_MALLOC, F_MALLOC, TLSF_MALLOC, DBG_SR_MEMORY, USE_FUTEX, FAST_LOCK-ADAPTIVE_WAIT, USE_DNS_CACHE, USE_DNS_FAILOVER, USE_NAPTR, USE_DST_BLOCKLIST, HAVE_RESOLV_RES ADAPTIVE_WAIT_LOOPS 1024, MAX_RECV_BUFFER_SIZE 262144, MAX_SEND_BUFFER_SIZE 262144, MAX_URI_SIZE 1024, BUF_SIZE 65535, DEFAULT PKG_SIZE 8MB poll method support: poll, epoll_lt, epoll_et, sigio_rt, select. id: f1d83d compiled on 10:25:58 Nov 12 2024 with x86_64-ruag-linux-gcc 13.3.0 ``` * **Operating System**: It is a yocto scarthgap based image (Version 5.0.6 LTS) ``` Linux ttel 6.6.54-yocto-standard #1 SMP PREEMPT_DYNAMIC Tue Oct 8 03:03:40 UTC 2024 x86_64 GNU/Linux ``` -- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/4185 You are receiving this because you are subscribed to this thread. Message ID: <kamailio/kamailio/issues/4185(a)github.com>

5 10

[kamailio/kamailio] ims_diameter_server not freeing shared memory (Issue #4116)
by MGrimby 07 May '25

07 May '25

4 5

[kamailio/kamailio] cdp: restructure locking order to prevent rare deadlock (PR #4191)
by Morten Tryfoss 06 May '25

06 May '25

#### Pre-Submission Checklist    - [x] Commit message has the format required by CONTRIBUTING guide - [x] Commits are split per component (core, individual modules, libs, utils, ...) - [x] Each component has a single commit (if not, squash them into one commit) - [x] No commits to README files for modules (changes must be done to docbook files in `doc/` subfolder, the README file is autogenerated) #### Type Of Change - [x] Small bug fix (non-breaking change which fixes an issue) - [ ] New feature (non-breaking change which adds new functionality) - [ ] Breaking change (fix or feature that would change existing functionality) #### Checklist:  - [ ] PR should be backported to stable branches - [x] Tested changes locally - [ ] Related to issue #XXXX (replace XXXX with an open issue number) #### Description We've been running the ims_charging module in production for quite some time, without any issues. A couple of days ago the whole instance freezed, but I was able to get a dump before the stuck processes before restarting. The traffic volume has gradually increased over time, so it's likely related to that. I then did a summary of the different types of stuck processes. Which locks they're holding, and what they're waiting for: ``` holding lock: AAASessionsLock waiting for lock: lock_get(peer_list_lock); #1 0x00007fe70eba8cde in futex_get (lock=0x7fe6e4a5c680) at ../../core/futexlock.h:108 #2 0x00007fe70ebaaca2 in get_peer_by_fqdn (fqdn=0x7fe6e4a5ba30) at peermanager.c:259 #3 0x00007fe70ebb64b7 in get_first_connected_route (cdp_session=0x7fe6e5cc5350, r=0x7fe6e4a5ba30, app_id=4, vendor_id=10415) at routing.c:115 #4 0x00007fe70ebb9a37 in get_routing_peer (cdp_session=0x7fe6e5cc5350, m=0x7fe6e5447ab0) at routing.c:293 #5 0x00007fe70ebcaf8c in AAASendMessage (message=0x7fe6e5447ab0, callback_f=0x7fe707a9e323 <resume_on_initial_ccr>, callback_param=0x7fe6e5297110) at diameter_comm.c:139 #6 0x00007fe707a9d3e0 in Ro_Send_CCR (msg=0x7fe70f011100, dlg=0x7fe6e5c571f0, dir=0, reservation_units=30, incoming_trunk_id=0x7ffe50efb060, outgoing_trunk_id=0x7ffe50efb050, pani=0x7ffe50efaee0, action=0x7fe70efa4730, tindex=10484, tlabel=677172593) at ims_ro.c:1511 #7 0x00007fe707a8285d in ki_ro_ccr (msg=0x7fe70f011100, s_route_name=0x7ffe50efb080, s_direction=0x7ffe50efb070, reservation_units=30, s_incoming_trunk_id=0x7ffe50efb060, s_outgoing_trunk_id=0x7ffe50efb050) at ims_charging_mod.c:742 #8 0x00007fe707a7bf01 in w_ro_ccr (msg=0x7fe70f011100, c_route_name=0x7fe70ef8b8d0 "\220\311\371\016\347\177", c_direction=0x7fe70ef8b980 "p\240\371\016\347\177", reservation_units=30, c_incoming_trunk_id=0x7fe70ef8ba30 "p\241\371\016\347\177", c_outgoing_trunk_id=0x7fe70ef8bae0 "\360\241\371\016\347\177") at ims_charging_mod.c:507 #9 0x00000000004858d8 in do_action (h=0x7ffe50efb970, a=0x7fe70ef99e20, msg=0x7fe70f011100) at core/action.c:1144 #10 0x00000000004928d6 in run_actions (h=0x7ffe50efb970, a=0x7fe70ef99e20, msg=0x7fe70f011100) at core/action.c:1618 #11 0x0000000000492f52 in run_actions_safe (h=0x7ffe50eff1e0, a=0x7fe70ef99e20, msg=0x7fe70f011100) at core/action.c:1681 #12 0x0000000000450156 in rval_get_long (h=0x7ffe50eff1e0, msg=0x7fe70f011100, i=0x7ffe50efbec8, rv=0x7fe70ef9c1d8, cache=0x0) at core/rvalue.c:973 #13 0x0000000000454d24 in rval_expr_eval_long (h=0x7ffe50eff1e0, msg=0x7fe70f011100, res=0x7ffe50efbec8, rve=0x7fe70ef9c1d0) at core/rvalue.c:1854 #14 0x0000000000454d52 in rval_expr_eval_long (h=0x7ffe50eff1e0, msg=0x7fe70f011100, res=0x7ffe50efc448, rve=0x7fe70ef9b920) at core/rvalue.c:1864 #15 0x00000000004850ce in do_action (h=0x7ffe50eff1e0, a=0x7fe70ef9b070, msg=0x7fe70f011100) at core/action.c:1097 -- holding lock: AAASessionsLock waiting for lock: lock_get(peer_list_lock); #1 0x00007fe70eba8d4f in futex_get (lock=0x7fe6e4a5c680) at ../../core/futexlock.h:121 #2 0x00007fe70ebaaca2 in get_peer_by_fqdn (fqdn=0x7ffe50efab90) at peermanager.c:259 #3 0x00007fe70ebb8e89 in get_routing_peer (cdp_session=0x7fe6e5ab6910, m=0x7fe6e5435be0) at routing.c:252 #4 0x00007fe70ebcaf8c in AAASendMessage (message=0x7fe6e5435be0, callback_f=0x7fe707a95edc <resume_on_termination_ccr>, callback_param=0x0) at diameter_comm.c:139 #5 0x00007fe707a95b02 in send_ccr_stop_with_param (ro_session=0x7fe6e5ab65e0, code=0, reason=0x7ffe50efb060) at ims_ro.c:1181 #6 0x00007fe707a72ff7 in dlg_terminated (dlg=0x7fe6e623d7a0, type=64, termcode=0, reason=0x7fe707ab72b3 "normal call clearing", _params=0x7fe707f67280 <params>) at dialog.c:249 #7 0x00007fe707a6a729 in dlg_callback_received (dlg=0x7fe6e623d7a0, type=64, _params=0x7fe707f67280 <params>) at dialog.c:25 #8 0x00007fe707d341b9 in run_dlg_callbacks (type=64, dlg=0x7fe6e623d7a0, req=0x7fe70f011100, rpl=0x0, dir=1, dlg_data=0x0) at dlg_cb.c:271 #9 0x00007fe707cf4db4 in dlg_terminated (req=0x7fe70f011100, dlg=0x7fe6e623d7a0, dir=1) at dlg_handlers.c:413 #10 0x00007fe707cfddeb in dlg_onroute (req=0x7fe70f011100, route_params=0x7ffe50efb6d0, param=0x0) at dlg_handlers.c:1097 #11 0x00007fe70ad285f6 in run_rr_callbacks (req=0x7fe70f011100, rr_param=0x7ffe50efb7c0) at rr_cb.c:96 #12 0x00007fe70ad3ae92 in after_loose (_m=0x7fe70f011100, preloaded=0) at loose.c:1021 #13 0x00007fe70ad3b5ce in loose_route_mode (_m=0x7fe70f011100, _mode=0) at loose.c:1056 #14 0x00007fe70ad3f74f in w_loose_route (msg=0x7fe70f011100, p1=0x0, p2=0x0) at rr_mod.c:273 #15 0x00000000004855ff in do_action (h=0x7ffe50efc390, a=0x7fe70efe0d40, msg=0x7fe70f011100) at core/action.c:1121 -- holding lock: lock_get(peer_list_lock); waiting for lock: lock_get(p->lock); #1 0x00007fe70eba8d4f in futex_get (lock=0x7fe6e4a5cbd0) at ../../core/futexlock.h:121 #2 0x00007fe70ebab0ae in peer_timer (now=1742807320, ptr=0x0) at peermanager.c:286 #3 0x00007fe70ebd0f39 in timer_loop () at timer.c:116 #4 0x00007fe70ebd21b2 in timer_process (returns=0) at timer.c:216 #5 0x00007fe70eb8ccf8 in diameter_peer_start (blocking=0) at diameter_peer.c:350 #6 0x00007fe70eb7cbb2 in cdp_child_init (rank=0) at cdp_mod.c:272 -- holding lock: lock_get(p->lock) waiting for lock: AAASessionsLock #1 0x00007fe70ebf2597 in futex_get (lock=0x7fe6e4a5d490) at ../../core/futexlock.h:108 #2 0x00007fe70ebf26f1 in AAASessionsLock (hash=0) at session.c:79 #3 0x00007fe70ebf5e6e in cdp_get_session (id=...) at session.c:316 #4 0x00007fe70eba6892 in Snd_Message (p=0x7fe6e4a5c880, msg=0x7fe6e63998d0) at peerstatemachine.c:1237 #5 0x00007fe70eba003e in sm_process (p=0x7fe6e4a5c880, event=Send_Message, msg=0x7fe6e63998d0, peer_locked=0, sock=0) at peerstatemachine.c:429 #6 0x00007fe70ebcbdc6 in AAASendMessage (message=0x7fe6e63998d0, callback_f=0x7fe707a8f208 <resume_on_interim_ccr>, callback_param=0x7fe6e6190a90) at diameter_comm.c:166 #7 0x00007fe707a8edb3 in send_ccr_interim (ro_session=0x7fe6e5399160, used=60, reserve=30) at ims_ro.c:847 #8 0x00007fe707a68bd6 in ro_session_ontimeout (tl=0x7fe6e5399200) at ro_timer.c:513 #9 0x00007fe707a63078 in ro_timer_routine (ticks=114862426, attr=0x0) at ro_timer.c:279 #10 0x00000000004fd33e in compat_old_handler (ti=1837798827, tl=0x7fe6e4cf6260, data=0x7fe6e4cf6260) at core/timer.c:980 #11 0x00000000004fde7a in slow_timer_main () at core/timer.c:1103 #12 0x000000000042e4e7 in main_loop () at main.c:1911 #13 0x000000000043876c in main (argc=10, argv=0x7ffe50f001c8) at main.c:3236 ``` In `get_first_connected_route()` in `routing.c` of the cdp module there are two places `get_peer_by_fqdn()` are called. One of them has an unlock/relock of the session list before and after (and a comment about holding two locks at a time), while the other doesn't. I'm currently testing two version of this. The first by just doing the some relock for the other `get_peer_by_fqdn()`, but this PR got an approach for maybe fixing more latent issues. `sm_process()` also got some strange handling by `Rcv_Process()` after the peer lock is released, which seems to be about the same thing. My deadlock is for the `Snd_Message()` equivalent. I've now removed this queueing behaviour, and instead doing a re-lock of the peer to (hopefully) have the same locking order as the other operations. The problem with this thing is how rare it occurs. Just wanted to share my findings for others with more knowledge to the cdp module for comments, suggestions and hopefully some extra testing. You can view, comment on, or merge this pull request online at: https://github.com/kamailio/kamailio/pull/4191 -- Commit Summary -- * cdp: restructure locking order to prevent rare deadlock -- File Changes -- M src/modules/cdp/peerstatemachine.c (25) M src/modules/cdp/routing.c (6) -- Patch Links -- https://github.com/kamailio/kamailio/pull/4191.patch https://github.com/kamailio/kamailio/pull/4191.diff -- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/pull/4191 You are receiving this because you are subscribed to this thread. Message ID: <kamailio/kamailio/pull/4191(a)github.com>

4 13

[kamailio/kamailio] Memory usage increases everytime tls.reload is executed (Issue #3823)
by Sebastian Denz 06 May '25

06 May '25

### Description We are using Kamailio 5.7.4 on Debian 12 (from http://deb.kamailio.org/kamailio57) with rtpengine as an Edgeproxy for our clients. The instance terminates SIP/TLS (with Cliencertificates) and forwards the SIP Traffic to internal systems. After some days we are getting errors like this `tls_complete_init(): tls: ssl bug #1491 workaround: not enough memory for safe operation: shm=7318616 threshold1=8912896` First we thought Kamailio just doesnt have enough memory, so we doubled it.. But after some days the Logmessage (and Userissues) occured again. So we monitored the shmmem statistics and found that used and max_used are constantly growing til it reaches the limit. As i mentioned we are using client-certificates and so we are also using the CRL feature. We do have a systemd-timer which fetches the CRL every hour and runs 'kamcmd tls.reload' when finished. Our tls.cfg looks like this: ``` [server:default] method = TLSv1.2+ private_key = /etc/letsencrypt/live/hostname.de/privkey.pem certificate = /etc/letsencrypt/live/hostname.de/fullchain.pem ca_list = /etc/kamailio/ca_list.pem ca_path = /etc/kamailio/ca_list.pem crl = /etc/kamailio/combined.crl.pem verify_certificate = yes require_certificate = yes [client:default] verify_certificate = yes require_certificate = yes ``` After testing a bit we found that every time tls.reload is executed Kamailio consumes a bit more memory which eventually leads to all the memory being consumed which leads to issues for our users. See following example: ``` [0][root@edgar-dev:~]# while true ; do /usr/sbin/kamcmd tls.reload ; /usr/sbin/kamcmd core.shmmem ; sleep 1 ; done Ok. TLS configuration reloaded. { total: 268435456 free: 223001520 used: 41352552 real_used: 45433936 max_used: 45445968 fragments: 73 } Ok. TLS configuration reloaded. { total: 268435456 free: 222377960 used: 41975592 real_used: 46057496 max_used: 46069232 fragments: 78 } Ok. TLS configuration reloaded. { total: 268435456 free: 221748664 used: 42604992 real_used: 46686792 max_used: 46698080 fragments: 77 } Ok. TLS configuration reloaded. { total: 268435456 free: 221110832 used: 43242408 real_used: 47324624 max_used: 47335608 fragments: 81 } ^C [130][root@edgar-dev:~]# ``` ### Troubleshooting #### Reproduction Everytime tls.reload is called the memory consumptions grows.. #### Debugging Data  ``` If you let me know what would be interesting for tracking this down, i am happy to provide logs/debugging data! ``` #### Log Messages  ``` If you let me know what would be interesting for tracking this down, i am happy to provide logs/debugging data! ``` #### SIP Traffic SIP doesnt seem to be relevant here ### Possible Solutions Calling tls.reload less often or restart kamailio before memory is consumed ;) ### Additional Information ``` version: kamailio 5.7.4 (x86_64/linux) flags: USE_TCP, USE_TLS, USE_SCTP, TLS_HOOKS, USE_RAW_SOCKS, DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MMAP, PKG_MALLOC, MEM_JOIN_FREE, Q_MALLOC, F_MALLOC, TLSF_MALLOC, DBG_SR_MEMORY, USE_FUTEX, FAST_LOCK-ADAPTIVE_WAIT, USE_DNS_CACHE, USE_DNS_FAILOVER, USE_NAPTR, USE_DST_BLOCKLIST, HAVE_RESOLV_RES, TLS_PTHREAD_MUTEX_SHARED ADAPTIVE_WAIT_LOOPS 1024, MAX_RECV_BUFFER_SIZE 262144, MAX_URI_SIZE 1024, BUF_SIZE 65535, DEFAULT PKG_SIZE 8MB poll method support: poll, epoll_lt, epoll_et, sigio_rt, select. id: unknown compiled with gcc 12.2.0 ``` * **Operating System**: ``` * Debian GNU/Linux 12 (bookworm) * Linux edgar-dev 6.1.0-20-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.85-1 (2024-04-11) x86_64 GNU/Linux ``` -- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/3823 You are receiving this because you are subscribed to this thread. Message ID: <kamailio/kamailio/issues/3823(a)github.com>

9 31

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

sr-dev March 2025