sr-dev August 2020

sr-dev@lists.kamailio.org

17 participants
244 discussions

git:master:a929de4e: xhttp_prom: add support for histogram metrics.

by Vicente Hernando Ara

Module: kamailio Branch: master Commit: a929de4ec810a210a8840e997b8357f33a7a6fd1 URL: https://github.com/kamailio/kamailio/commit/a929de4ec810a210a8840e997b8357f… Author: Vicente Hernando Ara <vhernando(a)sonoc.io> Committer: Vicente Hernando Ara <vhernando(a)sonoc.io> Date: 2020-08-10T19:52:59+02:00 xhttp_prom: add support for histogram metrics. --- Modified: src/modules/xhttp_prom/doc/xhttp_prom_admin.xml Modified: src/modules/xhttp_prom/prom.c Modified: src/modules/xhttp_prom/prom.h Modified: src/modules/xhttp_prom/prom_metric.c Modified: src/modules/xhttp_prom/prom_metric.h Modified: src/modules/xhttp_prom/xhttp_prom.c Modified: src/modules/xhttp_prom/xhttp_prom.h --- Diff: https://github.com/kamailio/kamailio/commit/a929de4ec810a210a8840e997b8357f… Patch: https://github.com/kamailio/kamailio/commit/a929de4ec810a210a8840e997b8357f…

3 years, 8 months

[kamailio/kamailio] t_forward_nonack not decoding/unescaping URI (#2255)

by Julien Chavanton

### Description SIP OPTION relayed using `t_forward_nonack()` not being URI decoded  ### Troubleshooting #### Reproduction register with PATH and use SIP OPTION keep alive or simply send relay a SIP OPTIONS with a route. #### Debugging Data  ```#1 0x00007fa3067a998f in t_forward_nonack (t=0x7fa3246fce88, p_msg=0x7fa3885aa1c0, proxy=0x0, proto=0) at t_fwd.c:1729 1729 raise(SIGSEGV); (gdb) l 1724 ser_error=MIN_int(lowest_ret, E_CFG); 1725 return -1; 1726 } 1727 if(lowest_ret!=E_CFG) { 1728 LM_ERR("failure to add branches\n"); 1729 raise(SIGSEGV); << raising exception to get the coredump 1730 } 1731 ser_error=lowest_ret; 1732 return lowest_ret; 1733 } ``` #### Log Messages ``` 4(34) ERROR: tm [ut.h:245]: uri2dst2(): bad_uri: sip:247.25.169.51:52307%3Btransport%3Dtls 4(34) ERROR: tm [t_fwd.c:1728]: t_forward_nonack(): failure to add branches ``` ### Possible Solutions This patch can fix the problem from the routing script : ``` if (is_method("OPTIONS") && uri!=myself) { $var(path_uri) = $(hdr(Route){re.subst,/<(.*)>/\1/}); if (defined $(var(path_uri){uri.params}) && $(var(path_uri){uri.params}{param.value,received}) != "") { $var(received) = $(var(path_uri){uri.params}{param.value,received}); if ($var(received) =~ "sip:.*:.*") { $du = $var(received); $avp(received) = $(var(path_uri){uri.params}{param.value,received}); avp_subst("$avp(received)", "/%3Btransport%3Dtcp/;transport=tcp/ig"); if ($avp(received) =~ "sip:.*:.*") { $du = $avp(received); } } t_relay(); } ``` ### Additional Information tested on 5.3.2, something changed during 5.3.0.dev, I will try to pin point the commit -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/2255

3 years, 8 months

git:master:6d928b59: ctl: skip handling rpc commands if kamilio is in shutdown phase

by Daniel-Constantin Mierla

Module: kamailio Branch: master Commit: 6d928b59a3dbd6eb291da3d3936be6e8f18588af URL: https://github.com/kamailio/kamailio/commit/6d928b59a3dbd6eb291da3d3936be6e… Author: Daniel-Constantin Mierla <miconda(a)gmail.com> Committer: Daniel-Constantin Mierla <miconda(a)gmail.com> Date: 2020-08-10T16:23:11+02:00 ctl: skip handling rpc commands if kamilio is in shutdown phase - related to #2433 --- Modified: src/modules/ctl/binrpc_run.c --- Diff: https://github.com/kamailio/kamailio/commit/6d928b59a3dbd6eb291da3d3936be6e… Patch: https://github.com/kamailio/kamailio/commit/6d928b59a3dbd6eb291da3d3936be6e… --- diff --git a/src/modules/ctl/binrpc_run.c b/src/modules/ctl/binrpc_run.c index 064b32b487..6323370b95 100644 --- a/src/modules/ctl/binrpc_run.c +++ b/src/modules/ctl/binrpc_run.c @@ -627,10 +627,15 @@ int process_rpc_req(unsigned char* buf, int size, int* bytes_needed, rpc_export_t* rpc_e; struct binrpc_ctx f_ctx; struct binrpc_parse_ctx* ctx; - + + if(ksr_shutdown_phase()) { + /* during shutdown - no more RPC command handling */ + return 0; + } + if (size<BINRPC_MIN_PKT_SIZE){ *bytes_needed=BINRPC_MIN_PKT_SIZE-size; - return 0; /* more data , nothing processed */ + return 0; /* more data, nothing processed */ } err=init_binrpc_ctx(&f_ctx, buf, size, sh); ctx=&f_ctx.in.ctx;

3 years, 8 months

git:master:6da946e4: core: keep a global flag in shm for shut down phase

by Daniel-Constantin Mierla

Module: kamailio Branch: master Commit: 6da946e4a3d6d6949dd6babe9cf3919edc0b4519 URL: https://github.com/kamailio/kamailio/commit/6da946e4a3d6d6949dd6babe9cf3919… Author: Daniel-Constantin Mierla <miconda(a)gmail.com> Committer: Daniel-Constantin Mierla <miconda(a)gmail.com> Date: 2020-08-10T16:15:04+02:00 core: keep a global flag in shm for shut down phase - set when destroy modules is starting - useful to check from other processes if shut down started --- Modified: src/core/sr_module.c Modified: src/core/sr_module.h Modified: src/main.c --- Diff: https://github.com/kamailio/kamailio/commit/6da946e4a3d6d6949dd6babe9cf3919… Patch: https://github.com/kamailio/kamailio/commit/6da946e4a3d6d6949dd6babe9cf3919… --- diff --git a/src/core/sr_module.c b/src/core/sr_module.c index 2c76ee2a1b..545845fdc5 100644 --- a/src/core/sr_module.c +++ b/src/core/sr_module.c @@ -109,6 +109,24 @@ unsigned int set_modinit_delay(unsigned int v) return r; } +/* shut down phase for instance - kept in shared memory */ +static int *_ksr_shutdown_phase = 0; + +int ksr_shutdown_phase_init(void) +{ + if(_ksr_shutdown_phase) { + _ksr_shutdown_phase = (int*)shm_malloc(sizeof(int)); + } + return 0; +} +/** + * return destroy modules phase state + */ +int ksr_shutdown_phase(void) +{ + return (_ksr_shutdown_phase)?(*_ksr_shutdown_phase):0; +} + /* keep state if server is in destroy modules phase */ static int _sr_destroy_modules_phase = 0; @@ -738,6 +756,10 @@ void destroy_modules() struct sr_module* t, *foo; _sr_destroy_modules_phase = 1; + if(_ksr_shutdown_phase!=NULL) { + *_ksr_shutdown_phase = 1; + } + /* call first destroy function from each module */ t=modules; while(t) { diff --git a/src/core/sr_module.h b/src/core/sr_module.h index 1f833a5e95..3014a88dad 100644 --- a/src/core/sr_module.h +++ b/src/core/sr_module.h @@ -544,5 +544,7 @@ int is_rpc_worker(int rank); unsigned int set_modinit_delay(unsigned int v); int destroy_modules_phase(void); +int ksr_shutdown_phase_init(void); +int ksr_shutdown_phase(void); #endif /* sr_module_h */ diff --git a/src/main.c b/src/main.c index b2a2040631..9294bc3d94 100644 --- a/src/main.c +++ b/src/main.c @@ -2542,6 +2542,8 @@ int main(int argc, char** argv) if (ksr_route_locks_set_init()<0) goto error; + ksr_shutdown_phase_init(); + /* init lookup for core event routes */ sr_core_ert_init();

3 years, 8 months

[kamailio/kamailio] Redirect to freeswitch after answer kamailio UA. (#2434)

by amitpaltech

Hello, I am newbie to Kamailio, after a lot searching kamailio doc , I am unable to find change in configuration file (kamailio.cfg). Which redirect the call from kamailio to freeswitch if the call is answered in kamailio. Thanks. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/2434

3 years, 8 months

[kamailio/kamailio] Reports -DCOMPILER='"gcc 0.2.1"' when compiling with gcc 10.2.1 on Fedora (#2425)

by Anthony Messina

Likely non-critical, but during the build, reports `-DCOMPILER='"gcc 0.2.1"'` when compiling with gcc 10.2.1 on Fedora 32 x86_64. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/2425

3 years, 8 months

git:master:79e77380: Makefile.defs: map gcc v10.x to 9.0+ options

by Daniel-Constantin Mierla

Module: kamailio Branch: master Commit: 79e7738027a40affa4c7f606bf55a3ad0fe31bc4 URL: https://github.com/kamailio/kamailio/commit/79e7738027a40affa4c7f606bf55a3a… Author: Daniel-Constantin Mierla <miconda(a)gmail.com> Committer: Daniel-Constantin Mierla <miconda(a)gmail.com> Date: 2020-08-05T16:38:01+02:00 Makefile.defs: map gcc v10.x to 9.0+ options - detect properly the 10.x gcc version number, GH #2425 --- Modified: src/Makefile.defs --- Diff: https://github.com/kamailio/kamailio/commit/79e7738027a40affa4c7f606bf55a3a… Patch: https://github.com/kamailio/kamailio/commit/79e7738027a40affa4c7f606bf55a3a… --- diff --git a/src/Makefile.defs b/src/Makefile.defs index d65a1e2f6b..44fb9d665f 100644 --- a/src/Makefile.defs +++ b/src/Makefile.defs @@ -275,7 +275,7 @@ ifneq (,$(findstring gcc, $(CC_LONGVER))) RPAREN=) CC_VER:=$(word 1,$(CC)) $(shell $(CC) - --version 2>/dev/null|head -n 1|\ sed -e 's/([^$(RPAREN)]*)//g' \ - -e 's/^.*$[0-9][0-9]*\.[0-9][0-9]*\.[0-9][0-9]*$.*/\1/'\ + -e 's/^.*[^0-9]$[0-9][0-9]*\.[0-9][0-9]*\.[0-9][0-9]*$.*/\1/'\ -e 's/^[^0-9].*$[0-9][0-9]*\.[0-9][0-9]*$.*/\1/') # CC_VER:=$(word 1,$(CC)) $(shell $(CC) - --version|head -n 1|cut -d" " -f 3\ # |sed -e 's/^.*$[0-9][0-9]*\.[0-9][0-9]*\.[0-9][0-9]*$.*/\1/'\ @@ -306,7 +306,9 @@ ifneq (,$(findstring gcc, $(CC_LONGVER))) -e 's/8\.[0-9]\..*/8.0+/' \ -e 's/8\.[0-9]$$/8.0+/' \ -e 's/9\.[0-9]\..*/9.0+/' \ - -e 's/9\.[0-9]$$/9.0+/') + -e 's/9\.[0-9]$$/9.0+/' \ + -e 's/10\.[0-9]\..*/9.0+/' \ + -e 's/10\.[0-9]$$/9.0+/') ifeq (,$(strip $(filter-out 3.0 3.4 4.x 4.2+ 4.5+ 5.0+ 6.0+ 7.0+ 8.0+ 9.0+,$(CC_SHORTVER)))) # dependencies can be generated on-the-fly while compiling *.c CC_MKDEP_OPTS=-MMD -MP

3 years, 9 months

registrar lookup() and xavp_rcd

by Alex Balashov

Hi, The documentation for `registrar` says that record properties can be stored in the XAVP set in the `xavp_rcd` modparam when `registered()` is called. But it does not say that this XAVP will be populated when a `lookup()` is done. And yet, it is populated; I have recently benefited from this. I think this is a really bit of functionality to mention, otherwise one is left guessing how many registrar operations are required to obtain usrloc binding data. I will be happy to push a documentation fix to correct this, but I wanted to clarify whether this is just an oversight or whether this is unexpected behaviour. Thanks, -- Alex -- Alex Balashov | Principal | Evariste Systems LLC Tel: +1-706-510-6800 / +1-800-250-5920 (toll-free) Web: http://www.evaristesys.com/, http://www.csrpswitch.com/

3 years, 9 months

git:5.4:a24d6c42: evapi: use port in getaddrinfo() otherwise it does not bind as expected

by Daniel-Constantin Mierla

Module: kamailio Branch: 5.4 Commit: a24d6c421fac3487f45a352955ba1a2d3e3cf2b7 URL: https://github.com/kamailio/kamailio/commit/a24d6c421fac3487f45a352955ba1a2… Author: Daniel-Constantin Mierla <miconda(a)gmail.com> Committer: Daniel-Constantin Mierla <miconda(a)gmail.com> Date: 2020-08-04T15:21:52+02:00 evapi: use port in getaddrinfo() otherwise it does not bind as expected - regression added by cc098ecdfb556bf5efaab72fa767c773000f9fb6 (cherry picked from commit 301302531c630803222e7f14a1c113b1a400f7a1) --- Modified: src/modules/evapi/evapi_dispatch.c --- Diff: https://github.com/kamailio/kamailio/commit/a24d6c421fac3487f45a352955ba1a2… Patch: https://github.com/kamailio/kamailio/commit/a24d6c421fac3487f45a352955ba1a2… --- diff --git a/src/modules/evapi/evapi_dispatch.c b/src/modules/evapi/evapi_dispatch.c index 0f95291bad..1b8bff4402 100644 --- a/src/modules/evapi/evapi_dispatch.c +++ b/src/modules/evapi/evapi_dispatch.c @@ -650,7 +650,7 @@ int evapi_run_dispatcher(char *laddr, int lport) memset(&ai_hints, 0, sizeof(struct addrinfo)); ai_hints.ai_family = AF_UNSPEC; /* allow IPv4 or IPv6 */ ai_hints.ai_socktype = SOCK_STREAM; /* stream socket */ - ai_ret = getaddrinfo(laddr, NULL, &ai_hints, &ai_res); + ai_ret = getaddrinfo(laddr, int2str(lport, NULL), &ai_hints, &ai_res); if (ai_ret != 0) { LM_ERR("getaddrinfo failed: %d %s\n", ai_ret, gai_strerror(ai_ret)); return -1;

3 years, 9 months

git:master:30130253: evapi: use port in getaddrinfo() otherwise it does not bind as expected

by Daniel-Constantin Mierla

Module: kamailio Branch: master Commit: 301302531c630803222e7f14a1c113b1a400f7a1 URL: https://github.com/kamailio/kamailio/commit/301302531c630803222e7f14a1c113b… Author: Daniel-Constantin Mierla <miconda(a)gmail.com> Committer: Daniel-Constantin Mierla <miconda(a)gmail.com> Date: 2020-08-04T15:21:17+02:00 evapi: use port in getaddrinfo() otherwise it does not bind as expected - regression added by cc098ecdfb556bf5efaab72fa767c773000f9fb6 --- Modified: src/modules/evapi/evapi_dispatch.c --- Diff: https://github.com/kamailio/kamailio/commit/301302531c630803222e7f14a1c113b… Patch: https://github.com/kamailio/kamailio/commit/301302531c630803222e7f14a1c113b… --- diff --git a/src/modules/evapi/evapi_dispatch.c b/src/modules/evapi/evapi_dispatch.c index 0f95291bad..1b8bff4402 100644 --- a/src/modules/evapi/evapi_dispatch.c +++ b/src/modules/evapi/evapi_dispatch.c @@ -650,7 +650,7 @@ int evapi_run_dispatcher(char *laddr, int lport) memset(&ai_hints, 0, sizeof(struct addrinfo)); ai_hints.ai_family = AF_UNSPEC; /* allow IPv4 or IPv6 */ ai_hints.ai_socktype = SOCK_STREAM; /* stream socket */ - ai_ret = getaddrinfo(laddr, NULL, &ai_hints, &ai_res); + ai_ret = getaddrinfo(laddr, int2str(lport, NULL), &ai_hints, &ai_res); if (ai_ret != 0) { LM_ERR("getaddrinfo failed: %d %s\n", ai_ret, gai_strerror(ai_ret)); return -1;

3 years, 9 months

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

sr-dev August 2020