- sr-dev - kamailio.org

usefull git options
by Andrei Pelinescu-Onciul 20 Nov '08

20 Nov '08

Include one line summaries of the merged commits in commit merge messages (a MUST): git config --global merge.log true Use auto-tracking (behave as if -track is added to every git-pull): git config --global branch.autosetupmerge always Note: --global changes the global configuration ( ~/.gitconfig). Without --global the local per repository configuration will be changed (.git/config). You can either use the commands above or edit by hand the config files. Example ~/.gitconfig: [user] name = Andrei Pelinescu-Onciul email = andrei(a)iptel.org signinkey = 2CF56A2D [color] branch = auto status = auto [merge] log = true [branch] autosetupmerge = always [remote "ser"] url= ssh://git.sip-router.org/ser push=cvshead:andrei/cvshead push=sctp:andrei/sctp [url "ssh://git.sip-router.org/"] insteadOf = sip-router: insteadOf = sip_router: [alias] intercommit = !sh -c 'git show "$0" > /tmp/commit1 && git show "$1" > /tmp/commit2 && interdiff /tmp/commit[12] | less -FRS' funcdiff = !sh -c \"git show \\\"\\$0:\\$2\\\" | sed -n \\\"/^[^ \\t].*\\$3(/,/^}/p\\\" > /tmp/.tmp1 &&\n git show \\\"\\$1:\\$2\\\" | sed -n \\\"/^[^ \\t].*\\$3(/,/^}/p\\\" > /tmp/.tmp2 &&\ngit diff /tmp/.tmp1 /tmp/.tmp2\" The insteadOf lines create aliases for the url (so that I can use git pull sip-router:ser_core cvs-head instead of git pull ssh://git.sip-router.org/ser_core cvs-head The aliases create a new git "commands", for more details see http://git.or.cz/gitwiki/Aliases. The funcdiff alias tries to display the differences in _one_ function between 2 different commits, e.g.: git funcdiff <oldrev> <newrev> <path> <function> The intecommit alias is for seeing differences between a submited patch (local commit into your repository) and the possibly modified version applied: git intercommit <orig_commit> <applied_commit>. Andrei

1 0

Karlsruhe meeting minutes
by Andrei Pelinescu-Onciul 20 Nov '08

20 Nov '08

Hi, I've attached a very brief version of the Karlsruhe meeting minutes. It contains input mostly from me and Daniel (so that you know who to blame :-)). Sorry for the long delay. Hope we haven't missed anything and if we did feel free to reply to this thread. Thanks a lot to 1&1 for hosting the meeting and to everyone who was able to come on such a short notice. Andrei

1 0

Re: [SR-Dev] [Serdev] [jan@iptel.org: Re: git repository online]
by Andrei Pelinescu-Onciul 19 Nov '08

19 Nov '08

> > Here is some more information on the git repository: > > > Synchronization with SER CVS > ============================ > The repository git.sip-router.org/ser is pulls commits from the SER CVS > repository automatically. The synchronization works one-way only, changes > commited to the SER CVS are automatically imported into the Git repository, > but not vice versa. In other words if you want to commit something in the > SER CVS repository then you have to do it manually, git won't help you here. To push changes I've done in the git ser repo to ser cvs I use: git log --pretty=oneline # to get the commit id I want to "export" git cvsexportcommit -c -p -w ~/ser.cvs/ -v <commitid> e.g.: git cvsexportcommit -c -p -w ~/ser.cvs/ -v <commitid> a9d929f558a6a620034fac8aa06c503b72c30fd8 in ~/ser.cvs I have an up-to-date checkout of ser from berlios cvs with the following small change: cd ~/ser.cvs; rm ser; ln -s . ser I need this because in the git repo ser is under the ser subdir, while in cvs it's in the "root" dir (or at least this is the way I checkout it). Another way to do it is to generate patches in git (e.g. git format-patch <comitid>) and the apply it manually on cvs, but this requires also a copy-paste for the commit message. > > Clone the repository: > $ git clone git://git.sip-router.org/ser > > and display the list of branches in the cloned repository: > $ git branch -a > > origin/HEAD > origin/MAIN > origin/Maintainer > origin/andrei_cvs-head > origin/cancel_fix > origin/cvs-head > origin/janakj > origin/jj/cvshead > origin/jj/doxygen > origin/master > origin/rel_0_8_11 > origin/rel_0_8_12 > origin/rel_0_8_14 > origin/rel_0_9_0 > origin/rel_2_0_0 > origin/rel_janakj > origin/start > origin/testing_0_8_12 > origin/vendor > > Several of the branches above are automatically updated from CVS, namely > origin/cvs-head receives cvs commits to HEAD and origin/rel_2_0_0 > receives commits to the branch rel_2_0_0 in CVS, and so on. > > These branches will be updated with latest commits from the CVS repository > every time you do: > > $ git pull > > Note that after the steps described above you have the whole CVS commit history > of SER on your local machine, starting with day 1 until today, and including > all branches from the CVS repository. This is convenient because you can > work with the history offline and pretty much all common operations are > blazingly fast because they only work with local files. > Note also that you can push changes only to branches starting with tmp/ , with your <your_username>/ or master (which is unused). The cvs sync branches are read-only, Andrei [...]

1 0

git:sip-router: script: variable number of params for functions
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: andrei/mod_f_params Commit: a2da0c5890f7c02538f0acb3197e6753dc9708bb URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=a2da0c5… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Wed Nov 19 00:10:46 2008 +0100 script: variable number of params for functions - support for true variable number of parameters module functions ( f(sip_msg, param_no, param_array[]) ) - support for functions with 3-6 parameters using kamailio extended interface. The implementation is however different: while in kamailio all module functions where called with 6 parameters (even if they declared only 2), this version will generate different function calls (small performance benefit by avoiding unneeded stack pushes and register saving for the most common 2 parameter functions and works also with other calling conventions). For performance reasons, a separate script engine command is now generated for each type of function: module functions calls with 0-2 parameters (MODULE_T), 3 (MODULE3_T), 4 (MODULE4_T), 5 (MODULE5_T) and 6 (MODULE6_T) parameters. In case of name conflicts with variable param. number function, it's undefined which actual function will get executed (it will be the "first" one, but the order depends on the module loading order a.s.o.). --- action.c | 90 ++++++++++++++++++++++++++++++++++++++++++++++++++++--- cfg.y | 37 +++++++++++++++++++++-- route.c | 5 +++ route_struct.c | 5 +++ route_struct.h | 10 ++++-- sr_module.c | 3 +- 6 files changed, 138 insertions(+), 12 deletions(-) diff --git a/action.c b/action.c index 4dc7aba..ea69e3a 100644 --- a/action.c +++ b/action.c @@ -46,6 +46,7 @@ * (andrei) * 2007-06-14 run_actions & do_action need a ctx or handle now, no more * static vars (andrei) + * 2008-11-18 support for variable parameter module functions (andrei) */ @@ -101,6 +102,7 @@ int do_action(struct run_act_ctx* h, struct action* a, struct sip_msg* msg) struct dest_info dst; char* tmp; char *new_uri, *end, *crt; + void* f; int len; int user; struct sip_uri uri, next_hop; @@ -717,11 +719,89 @@ int do_action(struct run_act_ctx* h, struct action* a, struct sip_msg* msg) break; case MODULE_T: if ( a->val[0].type==MODEXP_ST && a->val[0].u.data && - ((union cmd_export_u*)a->val[0].u.data)->c.function){ - ret=((union cmd_export_u*)a->val[0].u.data)->c.function(msg, - (char*)a->val[2].u.data, - (char*)a->val[3].u.data - ); + (f=((union cmd_export_u*)a->val[0].u.data)->c.function)){ + ret=((cmd_function)f)(msg, + (char*)a->val[2].u.data, + (char*)a->val[3].u.data + ); + if (ret==0) h->run_flags|=EXIT_R_F; + h->last_retcode=ret; + } else { + LOG(L_CRIT,"BUG: do_action: bad module call\n"); + } + break; + /* instead of using the parameter number, we use different names + * for calls to functions with 3, 4, 5, 6 or variable number of + * parameters due to performance reasons */ + case MODULE3_T: + if ( a->val[0].type==MODEXP_ST && a->val[0].u.data && + (f=((union cmd_export_u*)a->val[0].u.data)->c.function)){ + ret=((cmd_function3)f)(msg, + (char*)a->val[2].u.data, + (char*)a->val[3].u.data, + (char*)a->val[4].u.data + ); + if (ret==0) h->run_flags|=EXIT_R_F; + h->last_retcode=ret; + } else { + LOG(L_CRIT,"BUG: do_action: bad module call\n"); + } + break; + case MODULE4_T: + if ( a->val[0].type==MODEXP_ST && a->val[0].u.data && + (f=((union cmd_export_u*)a->val[0].u.data)->c.function)){ + ret=((cmd_function4)f)(msg, + (char*)a->val[2].u.data, + (char*)a->val[3].u.data, + (char*)a->val[4].u.data, + (char*)a->val[5].u.data + ); + if (ret==0) h->run_flags|=EXIT_R_F; + h->last_retcode=ret; + } else { + LOG(L_CRIT,"BUG: do_action: bad module call\n"); + } + break; + case MODULE5_T: + if ( a->val[0].type==MODEXP_ST && a->val[0].u.data && + (f=((union cmd_export_u*)a->val[0].u.data)->c.function)){ + ret=((cmd_function5)f)(msg, + (char*)a->val[2].u.data, + (char*)a->val[3].u.data, + (char*)a->val[4].u.data, + (char*)a->val[5].u.data, + (char*)a->val[6].u.data + ); + if (ret==0) h->run_flags|=EXIT_R_F; + h->last_retcode=ret; + } else { + LOG(L_CRIT,"BUG: do_action: bad module call\n"); + } + break; + case MODULE6_T: + if ( a->val[0].type==MODEXP_ST && a->val[0].u.data && + (f=((union cmd_export_u*)a->val[0].u.data)->c.function)){ + ret=((cmd_function6)f)(msg, + (char*)a->val[2].u.data, + (char*)a->val[3].u.data, + (char*)a->val[4].u.data, + (char*)a->val[5].u.data, + (char*)a->val[6].u.data, + (char*)a->val[7].u.data + ); + if (ret==0) h->run_flags|=EXIT_R_F; + h->last_retcode=ret; + } else { + LOG(L_CRIT,"BUG: do_action: bad module call\n"); + } + break; + case MODULEX_T: + if ( a->val[0].type==MODEXP_ST && a->val[0].u.data && + (f=((union cmd_export_u*)a->val[0].u.data)->c.function)){ + ret=((cmd_function_var)f)(msg, + a->val[1].u.number, + &a->val[2] + ); if (ret==0) h->run_flags|=EXIT_R_F; h->last_retcode=ret; } else { diff --git a/cfg.y b/cfg.y index ab485e6..5adc111 100644 --- a/cfg.y +++ b/cfg.y @@ -88,6 +88,7 @@ * 2007-11-28 added TCP_OPT_{FD_CACHE, DEFER_ACCEPT, DELAYED_ACK, SYNCNT, * LINGER2, KEEPALIVE, KEEPIDLE, KEEPINTVL, KEEPCNT} (andrei) * 2008-01-24 added cfg_var definition (Miklos) + * 2008-11-18 support for variable parameter module functions (andrei) */ %{ @@ -2359,7 +2360,8 @@ cmd: $$=0; yyerror("bad argument, [proto:]host[:port] expected"); } | FORCE_SEND_SOCKET error {$$=0; yyerror("missing '(' or ')' ?"); } - | ID {mod_func_action = mk_action(MODULE_T, 2, MODEXP_ST, NULL, NUMBER_ST, 0); } LPAREN func_params RPAREN { + | ID {mod_func_action = mk_action(MODULE_T, 2, MODEXP_ST, NULL, NUMBER_ST, + 0); } LPAREN func_params RPAREN { mod_func_action->val[0].u.data = find_export_record($1, mod_func_action->val[1].u.number, rt, &u_tmp); @@ -2372,6 +2374,33 @@ cmd: } pkg_free(mod_func_action); mod_func_action=0; + }else{ + switch( ((union cmd_export_u*) + mod_func_action->val[0].u.data)->c.param_no){ + case 0: + case 1: + case 2: + /* MODULE_T used for 0-2 params */ + break; + case 3: + mod_func_action->type=MODULE3_T; + break; + case 4: + mod_func_action->type=MODULE4_T; + break; + case 5: + mod_func_action->type=MODULE5_T; + break; + case 6: + mod_func_action->type=MODULE6_T; + break; + case VAR_PARAM_NO: + mod_func_action->type=MODULEX_T; + break; + default: + yyerror("too many parameters for function\n"); + break; + } } $$ = mod_func_action; } @@ -2385,8 +2414,10 @@ func_params: func_param: NUMBER { if (mod_func_action->val[1].u.number < MAX_ACTIONS-2) { - mod_func_action->val[mod_func_action->val[1].u.number+2].type = NUMBER_ST; - mod_func_action->val[mod_func_action->val[1].u.number+2].u.number = $1; + mod_func_action->val[mod_func_action->val[1].u.number+2].type = + NUMBER_ST; + mod_func_action->val[mod_func_action->val[1].u.number+2].u.number = + $1; mod_func_action->val[1].u.number++; } else { yyerror("Too many arguments\n"); diff --git a/route.c b/route.c index b222bfa..5c39764 100644 --- a/route.c +++ b/route.c @@ -503,6 +503,11 @@ static int fix_actions(struct action* a) break; case MODULE_T: + case MODULE3_T: + case MODULE4_T: + case MODULE5_T: + case MODULE6_T: + case MODULEX_T: cmd = t->val[0].u.data; if (cmd && cmd->c.fixup) { int i; diff --git a/route_struct.c b/route_struct.c index 0c00c61..c5fb76b 100644 --- a/route_struct.c +++ b/route_struct.c @@ -350,6 +350,11 @@ void print_action(struct action* t) DBG("if ("); break; case MODULE_T: + case MODULE3_T: + case MODULE4_T: + case MODULE5_T: + case MODULE6_T: + case MODULEX_T: DBG(" external_module_call("); break; case FORCE_RPORT_T: diff --git a/route_struct.h b/route_struct.h index 814517b..22b92d9 100644 --- a/route_struct.h +++ b/route_struct.h @@ -70,7 +70,7 @@ enum { METHOD_O=1, URI_O, FROM_URI_O, TO_URI_O, SRCIP_O, SRCPORT_O, enum { FORWARD_T=1, SEND_T, DROP_T, LOG_T, ERROR_T, ROUTE_T, EXEC_T, SET_HOST_T, SET_HOSTPORT_T, SET_USER_T, SET_USERPASS_T, SET_PORT_T, SET_URI_T, SET_HOSTPORTTRANS_T, - IF_T, MODULE_T, + IF_T, MODULE_T, MODULE3_T, MODULE4_T, MODULE5_T, MODULE6_T, MODULEX_T, SETFLAG_T, RESETFLAG_T, ISFLAGSET_T , AVPFLAG_OPER_T, LEN_GT_T, PREFIX_T, STRIP_T,STRIP_TAIL_T, @@ -133,13 +133,17 @@ typedef struct { } u; } action_u_t; -#define MAX_ACTIONS 4 +/* maximum internal array/params + * for module function calls val[0] and val[1] store a pointer to the + * function and the number of params, the rest are the function params + */ +#define MAX_ACTIONS (2+6) struct action{ int type; /* forward, drop, log, send ...*/ int count; - action_u_t val[MAX_ACTIONS]; struct action* next; + action_u_t val[MAX_ACTIONS]; }; struct expr* mk_exp(int op, struct expr* left, struct expr* right); diff --git a/sr_module.c b/sr_module.c index d59dc6d..21d89ac 100644 --- a/sr_module.c +++ b/sr_module.c @@ -353,7 +353,8 @@ union cmd_export_u* find_mod_export_record(char* mod, char* name, for(i=0, cmd=(void*)&t->exports->VER.cmds[0]; cmd->VER.name; \ i++, cmd=(void*)&t->exports->VER.cmds[i]){\ if((strcmp(name, cmd->VER.name)==0)&& \ - (cmd->VER.param_no==param_no) && \ + ((cmd->VER.param_no==param_no) || \ + (cmd->VER.param_no==VAR_PARAM_NO)) && \ ((cmd->VER.flags & flags) == flags) \ ){ \ DBG("find_export_record: found <%s> in module %s [%s]\n", \

1 0

photos from Karlsruhe meeting
by Daniel-Constantin Mierla 18 Nov '08

18 Nov '08

Hello, a photo album that includes pictures (thanks to Jerome and Henning) from last week's Karlsruhe meeting of SIP Router project is available at: http://sip-router.org/pub/photos/album-sip-router-karlsruhe/ Cheers, Daniel -- Daniel-Constantin Mierla http://www.asipto.com

1 0

git:sip-router: added .gitignore
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: andrei/mod_f_params Commit: e23943edfbc76ac53a1b5ffac74ab4d1ccd5f5a0 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=e23943e… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Tue Nov 18 00:20:30 2008 +0100 added .gitignore - .gitignore containing patterns for all the files that git should ignore (from *.o, *.d to config.mak, modules.lst a.s.o). --- .gitignore | 28 ++++++++++++++++++++++++++++ 1 files changed, 28 insertions(+), 0 deletions(-) diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..1a32c7e --- /dev/null +++ b/.gitignore @@ -0,0 +1,28 @@ +# ignore autogenerated files +cfg.tab.c +cfg.tab.h +lex.yy.c +librpath.lst +makecfg.lst +config.mak +modules.lst +# ignore dependency files +*.d +# ignore binary files and objects +ser +*.so +*.so.* +*.o +# archives +*.tar.gz +*.deb +*.rpm +# tags +tags +TAGS +# vi swaps +.*.swp +# various other binaries +test/udp_flood +utils/gen_ha1/gen_ha1 +utils/sercmd/sercmd

1 0

git:sip-router: module interface: minor changes from kamailio
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: andrei/mod_f_params Commit: 1dd086b90c043db12f70974d7de46691feb0aa0b URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=1dd086b… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Mon Nov 17 18:21:53 2008 +0100 module interface: minor changes from kamailio - changed callback with no params prototypes from f() into f(void), original commit message: - sync function definitions with declarations for functions w/o parameters - this functions were declared as 'f()', that means they can have a arbitrary number of arguments (because of compatibility to the old style definitions in the C standard), even if they are defined as f(void) - there is no warning reported if somebody call a function like this with some parameters - another potential problem is, when somebody change the definition to accept some parameter, but forgot to change also the declaration. Then the compiler will not report wrong usage of this functions too. - sometimes ago the '-Wold-style-definition' warning was introduced, the definitions get then fixed subsequently, but not all declarations - added free_fixup_function type definition - added some doxygen comments Credits go to: Henning Westerholt <henning.westerholt(a)1und1.de> --- sr_module.h | 19 ++++++++++++++----- 1 files changed, 14 insertions(+), 5 deletions(-) diff --git a/sr_module.h b/sr_module.h index 15aed59..f099f37 100644 --- a/sr_module.h +++ b/sr_module.h @@ -41,6 +41,13 @@ * 2007-06-07 added PROC_INIT, called in the main process context * (same as PROC_MAIN), buf guaranteed to be called before * any other process is forked (andrei) + * 2008-11-17 sip-router version: includes some of the openser/kamailio + * changes: f(void) instead of f(), free_fixup_function() + */ + +/*! + * \file + * \brief modules/plug-in structures declarations */ @@ -53,12 +60,14 @@ #include "route_struct.h" #include "str.h" -typedef struct module_exports* (*module_register)(); +typedef struct module_exports* (*module_register)(void); typedef int (*cmd_function)(struct sip_msg*, char*, char*); typedef int (*fixup_function)(void** param, int param_no); +typedef int (*free_fixup_function)(void** param, int param_no); typedef int (*response_function)(struct sip_msg*); typedef void (*onbreak_function)(struct sip_msg*); -typedef void (*destroy_function)(); +typedef void (*destroy_function)(void); + typedef int (*init_function)(void); typedef int (*child_init_function)(int rank); @@ -193,19 +202,19 @@ extern struct sr_module* modules; /* global module list*/ extern response_function* mod_response_cbks;/* response callback array */ extern int mod_response_cbk_no; /* size of reponse callbacks array */ -int register_builtin_modules(); +int register_builtin_modules(void); int register_module(struct module_exports*, char*, void*); int load_module(char* path); cmd_export_t* find_export_record(char* name, int param_no, int flags); cmd_function find_export(char* name, int param_no, int flags); cmd_function find_mod_export(char* mod, char* name, int param_no, int flags); rpc_export_t* find_rpc_export(char* name, int flags); -void destroy_modules(); +void destroy_modules(void); int init_child(int rank); int init_modules(void); struct sr_module* find_module_by_name(char* mod); -/* +/*! \brief * Find a parameter with given type and return it's * address in memory * If there is no such parameter, NULL is returned

1 0

git:sip-router: kamailio compatibility header files
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: andrei/mod_f_params Commit: 3c72005540cbe07ccc1f6f7c30407e65b933e9f5 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=3c72005… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Tue Nov 18 00:21:51 2008 +0100 kamailio compatibility header files - mi, pvar & statistics basic include files. For now they contain only basic type declarations and defines to allow compilation (so you cannot use them, but the code will compile). - pvar needs more work or replacement --- mi/mi.h | 99 +++++++++++++++++++++++++++++++++++++++ pvar.h | 145 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ statistics.h | 72 +++++++++++++++++++++++++++++ 3 files changed, 316 insertions(+), 0 deletions(-) Diff: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commitdiff;h=3c7…

1 0

git:sip-router: dual module interface support: ser and kamailio
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: andrei/mod_f_params Commit: 2873d3842c3bf2585200f381473a5b3f171ab8b0 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=2873d38… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Tue Nov 18 00:48:59 2008 +0100 dual module interface support: ser and kamailio Added support for both ser and kamailio module interfaces: a module just needs to define its module interface prior to including sr_module.h (e.g. by adding EXTRA_DEFS+=-DSER_MOD_INTERFACE for a ser module or EXTRA_DEFS+=-DKAMAILIO_MOD_INTERFACE to the module Makefile). This way ser and kamailio modules can be mixed at will with only one Makefile change. Under the hood, now each module declares its interface version (by exporting a global symbol named module_interface_ver) and the internal module loader and module export finder were updated to take the interface version into account. Internally the core works now with a generic module_export_u union. --- action.c | 13 ++- cfg.y | 8 +- core_cmd.c | 6 +- modparam.c | 12 +- route.c | 8 +- sr_module.c | 393 ++++++++++++++++++++++++++++++++++++++++++++--------------- sr_module.h | 172 ++++++++++++++++++++++++-- 7 files changed, 482 insertions(+), 130 deletions(-) Diff: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commitdiff;h=287…

1 0

git:sip-router: kamailio LM_* compatibility macros
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: andrei/mod_f_params Commit: 4f69388d8139bed903b383cc1673930eb8dac1ef URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=4f69388… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Tue Nov 18 00:58:18 2008 +0100 kamailio LM_* compatibility macros - added LM_* macros (LM_ERR, LM_BUG a.s.o) for compatibility with kamailio (for now they just use ser ERR, BUG ... macros) --- dprint.h | 21 +++++++++++++++++++++ 1 files changed, 21 insertions(+), 0 deletions(-) diff --git a/dprint.h b/dprint.h index 9535adc..41e9a45 100644 --- a/dprint.h +++ b/dprint.h @@ -244,13 +244,34 @@ int log_facility_fixup(void *handle, str *name, void **val); #define WARN(...) LOG(L_WARN, "WARNING: " LOC_INFO __VA_ARGS__) #define INFO(...) LOG(L_INFO, "INFO: " LOC_INFO __VA_ARGS__) #define BUG(...) LOG(L_CRIT, "BUG: " LOC_INFO __VA_ARGS__) + #define NOTICE(...) LOG(L_NOTICE, "NOTICE: " LOC_INFO __VA_ARGS__) + #define ALERT(...) LOG(L_ALERT, "ALERT: " LOC_INFO __VA_ARGS__) + #define CRIT(...) LOG(L_CRIT, "CRITICAL: " LOC_INFO __VA_ARGS__) #else #define DEBUG(fmt, args...) DBG("DEBUG: " LOC_INFO fmt, ## args) #define ERR(fmt, args...) LOG(L_ERR, "ERROR: " LOC_INFO fmt, ## args) #define WARN(fmt, args...) LOG(L_WARN, "WARN: " LOC_INFO fmt, ## args) #define INFO(fmt, args...) LOG(L_INFO, "INFO: " LOC_INFO fmt, ## args) #define BUG(fmt, args...) LOG(L_CRIT, "BUG: " LOC_INFO fmt, ## args) + #define NOTICE(fmt, args...) \ + LOG(L_NOTICE, "NOTICE: " LOC_INFO fmt, ## args) + #define ALERT(fmt, args...) \ + LOG(L_ALERT, "ALERT: " LOC_INFO fmt, ## args) + #define CRIT(fmt, args...) \ + LOG(L_CRIT, "CRITICAL: " LOC_INFO fmt, ## args) #endif +/* kamailio/openser compatibility */ + +#define LM_GEN1 LOG + +#define LM_ALERT ALERT +#define LM_CRIT CRIT +#define LM_ERR ERR +#define LM_WARN WARN +#define LM_NOTICE +#define LM_INFO INFO +#define LM_DBG DEBUG + #endif /* ifndef dprint_h */

1 0

git:sip-router: tm makefile: declare SER_MOD_INTERFACE
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: andrei/mod_f_params Commit: f265669f32e0b6fc880db16d1fdd8da0b69c6426 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=f265669… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Tue Nov 18 01:01:18 2008 +0100 tm makefile: declare SER_MOD_INTERFACE --- modules/tm/Makefile | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/modules/tm/Makefile b/modules/tm/Makefile index 10c4f42..0c60918 100644 --- a/modules/tm/Makefile +++ b/modules/tm/Makefile @@ -9,5 +9,6 @@ include ../../Makefile.defs auto_gen= NAME=tm.so LIBS= +DEFS+= -DSER_MOD_INTERFACE include ../../Makefile.modules

1 0

git:sip-router: Merge branch 'module_interface' into mod_f_params
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: andrei/mod_f_params Commit: 3ab704801e5065c575cb57e23560458c93864926 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=3ab7048… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Tue Nov 18 01:13:21 2008 +0100 Merge branch 'module_interface' into mod_f_params * module_interface: tm makefile: declare SER_MOD_INTERFACE kamailio LM_* compatibility macros dual module interface support: ser and kamailio kamailio compatibility header files added .gitignore module interface: minor changes from kamailio ---

1 0

git:sip-router: added .gitignore
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: andrei/mod_f_params Commit: 94476acf20b8f084ef95d461989b3132a07c4281 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=94476ac… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Tue Nov 18 00:20:30 2008 +0100 added .gitignore - .gitignore containing patterns for all the files that git should ignore (from *.o, *.d to config.mak, modules.lst a.s.o). --- .gitignore | 28 ++++++++++++++++++++++++++++ 1 files changed, 28 insertions(+), 0 deletions(-) diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..1a32c7e --- /dev/null +++ b/.gitignore @@ -0,0 +1,28 @@ +# ignore autogenerated files +cfg.tab.c +cfg.tab.h +lex.yy.c +librpath.lst +makecfg.lst +config.mak +modules.lst +# ignore dependency files +*.d +# ignore binary files and objects +ser +*.so +*.so.* +*.o +# archives +*.tar.gz +*.deb +*.rpm +# tags +tags +TAGS +# vi swaps +.*.swp +# various other binaries +test/udp_flood +utils/gen_ha1/gen_ha1 +utils/sercmd/sercmd

1 0

git:sip-router: added .gitignore
by Andrei Pelinescu-Onciul 18 Nov '08

18 Nov '08

Module: sip-router Branch: master Commit: 94476acf20b8f084ef95d461989b3132a07c4281 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=94476ac… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Tue Nov 18 00:20:30 2008 +0100 added .gitignore - .gitignore containing patterns for all the files that git should ignore (from *.o, *.d to config.mak, modules.lst a.s.o). --- .gitignore | 28 ++++++++++++++++++++++++++++ 1 files changed, 28 insertions(+), 0 deletions(-) diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..1a32c7e --- /dev/null +++ b/.gitignore @@ -0,0 +1,28 @@ +# ignore autogenerated files +cfg.tab.c +cfg.tab.h +lex.yy.c +librpath.lst +makecfg.lst +config.mak +modules.lst +# ignore dependency files +*.d +# ignore binary files and objects +ser +*.so +*.so.* +*.o +# archives +*.tar.gz +*.deb +*.rpm +# tags +tags +TAGS +# vi swaps +.*.swp +# various other binaries +test/udp_flood +utils/gen_ha1/gen_ha1 +utils/sercmd/sercmd

1 0

git:sip-router: tm makefile: declare SER_MOD_INTERFACE
by Andrei Pelinescu-Onciul 17 Nov '08

17 Nov '08

Module: sip-router Branch: andrei/module_interface Commit: f265669f32e0b6fc880db16d1fdd8da0b69c6426 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=f265669… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Tue Nov 18 01:01:18 2008 +0100 tm makefile: declare SER_MOD_INTERFACE --- modules/tm/Makefile | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/modules/tm/Makefile b/modules/tm/Makefile index 10c4f42..0c60918 100644 --- a/modules/tm/Makefile +++ b/modules/tm/Makefile @@ -9,5 +9,6 @@ include ../../Makefile.defs auto_gen= NAME=tm.so LIBS= +DEFS+= -DSER_MOD_INTERFACE include ../../Makefile.modules

1 0

git:sip-router: doc/NEWS: minor update
by Andrei Pelinescu-Onciul 16 Nov '08

16 Nov '08

Module: sip-router Branch: master Commit: b5358fc9f8ae48b04c26c1f2e55988b782b48f33 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=b5358fc… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Sun Nov 16 18:29:06 2008 +0000 doc/NEWS: minor update - added missing sigio_rt poll method to tcp_poll_method --- NEWS | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/NEWS b/NEWS index 0b0d031..f9c0867 100644 --- a/NEWS +++ b/NEWS @@ -577,8 +577,8 @@ new config variables: tcp_connection_lifetime = value (s) - how long the lifetime of a tcp connection will be exteneded after an IO event (accept, connect, read, write). Default: 120 s. - tcp_poll_method = poll|select|epoll_et|epoll-lt|kqueue|devpoll - poll - method used (by default the best one for the current OS is selected) + tcp_poll_method = poll|select|sigio_rt|epoll_et|epoll-lt|kqueue|devpoll - + poll method used (by default the best one for the current OS is selected) tcp_max_connections = no. - maximum number of tcp connections (if the number is exceeded no new tcp connections will be accepted). Default: 2048.

1 0

git:sip-router: doc/NEWS: minor update
by Andrei Pelinescu-Onciul 16 Nov '08

16 Nov '08

Module: sip-router Branch: ser_core_cvs Commit: b5358fc9f8ae48b04c26c1f2e55988b782b48f33 URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=b5358fc… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Sun Nov 16 18:29:06 2008 +0000 doc/NEWS: minor update - added missing sigio_rt poll method to tcp_poll_method --- NEWS | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/NEWS b/NEWS index 0b0d031..f9c0867 100644 --- a/NEWS +++ b/NEWS @@ -577,8 +577,8 @@ new config variables: tcp_connection_lifetime = value (s) - how long the lifetime of a tcp connection will be exteneded after an IO event (accept, connect, read, write). Default: 120 s. - tcp_poll_method = poll|select|epoll_et|epoll-lt|kqueue|devpoll - poll - method used (by default the best one for the current OS is selected) + tcp_poll_method = poll|select|sigio_rt|epoll_et|epoll-lt|kqueue|devpoll - + poll method used (by default the best one for the current OS is selected) tcp_max_connections = no. - maximum number of tcp connections (if the number is exceeded no new tcp connections will be accepted). Default: 2048.

1 0

git:test: added .gitignore
by Andrei Pelinescu-Onciul 16 Nov '08

16 Nov '08

Module: test Branch: master Commit: 078dc10293224d69cf567c42d7471ce09f8989aa URL: http://git.sip-router.org/cgi-bin/gitweb.cgi/test/?a=commit;h=078dc10293224… Author: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Committer: Andrei Pelinescu-Onciul <andrei(a)iptel.org> Date: Sun Nov 16 18:57:14 2008 +0100 added .gitignore - patterns for all the files that git should ignore (from *.o, *.d to config.mak, modules.lst a.s.o) --- .gitignore | 28 ++++++++++++++++++++++++++++ 1 files changed, 28 insertions(+), 0 deletions(-) diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..1a32c7e --- /dev/null +++ b/.gitignore @@ -0,0 +1,28 @@ +# ignore autogenerated files +cfg.tab.c +cfg.tab.h +lex.yy.c +librpath.lst +makecfg.lst +config.mak +modules.lst +# ignore dependency files +*.d +# ignore binary files and objects +ser +*.so +*.so.* +*.o +# archives +*.tar.gz +*.deb +*.rpm +# tags +tags +TAGS +# vi swaps +.*.swp +# various other binaries +test/udp_flood +utils/gen_ha1/gen_ha1 +utils/sercmd/sercmd

1 1

sip router meeting in Karlsruhe
by Daniel-Constantin Mierla 12 Nov '08

12 Nov '08

Hello, because of taking several days off, a complete report might show up a bit later, apologizes for that. Now a (short) summary... First, many thanks thanks to 1&1 for hosting the event and everyone that participated in such short notice. There were representatives from 10 companies: - 1&1 - FhG Fokus - Telio - Asipto - iptelorg/Tekelec - Voztelecom - Iptego - Itsyscom - Longphone - Basis AudioNet Some pictures should be published in the near future. After going in short introduction and presentation of the goals from the point of view of each project, we focused on: - identification of potential points of conflicts and how to get to a resolution in such case - code integration for common layer of the first phase - future development and proposals of new features - management of the larger eco-system that includes related projects and business entities I will send several emails detailing what was discussed and proposed there in few days. meanwhile, the outlines: - it is hard to avoid conflicts just by some clear and strict rules, so the common sense should lead the collaboration and discussions - GIT repository should be up in several days so the work can start, with a time line of 2-3 months from now to get core and tm in a very good shape of integration - another meeting shall be set in about 3 months time, to allow enough time for people to be able to attend, adjust the development and look more deep at the future. While the a lot of focus in the next months will be on integration, development of new features won't stop -- for examples, steps to a partial asynchronous processing are undertaking, couple of new modules are planned for release, several other modules to bring new functionalities - we should encourage and promote the development of related applications, like web interfaces, management tools, applications servers -- the add value for community and business Cheers, Daniel -- Daniel-Constantin Mierla http://www.asipto.com

1 0

[OpenSIPS-Users] [Fwd: [Serdev] the sip router project]
by Adrian Georgescu 05 Nov '08

05 Nov '08

As Bogdan already revealed, our intentions is to produce a new design for the core that will address the scalability and security requirement in a completely different manner than the original project because with the passing of time the requirements of the customers of the original product have also changed. I believe that both projects will learn from each other and grow better by going their own separate ways. As an eco-system we have matured enough in the last five years to be able to learn from and admit about our own mistakes in the past and look now at a better future. Regards, Adrian 5 nov 2008 kl. 11.41 skrev Bogdan-Andrei Iancu: > After 7 year of SER/OpenSER I (and many other) got to simplest > conclusion that the current design is not able to sustain the progress > of SER / OpenSER (like scripting, async calls, integration, scaling, > etc) ? Mainly because SER was design 7 years ago when there was only > stateless processing, no TCP, etc.... > > And I personally do not see any future in keep trying to patch the > existing design as it has no future Bogdan, Thanks for the good insight in why you don't want to join the effort, but go your own way. If you have different ideas for the core than what you understand from the other teams, then there's no reason for you to join the effort. I did not realize that was the case, now I do understand you better. > nature invented forking for seeking new alternatives! A good way of expressing it :-) At the same time, there's something to be said for the old, solid and good non-forked original product... Abba is still more enjoyable than all the forks... Ha ha! Cheers, /O

1 0

Re: [SR-Dev] [OpenSIPS-Users] [Fwd: [Serdev] the sip router project]
by Johansson Olle E 05 Nov '08

05 Nov '08

5 nov 2008 kl. 11.41 skrev Bogdan-Andrei Iancu: > After 7 year of SER/OpenSER I (and many other) got to simplest > conclusion that the current design is not able to sustain the progress > of SER / OpenSER (like scripting, async calls, integration, scaling, > etc) ? Mainly because SER was design 7 years ago when there was only > stateless processing, no TCP, etc.... > > And I personally do not see any future in keep trying to patch the > existing design as it has no future Bogdan, Thanks for the good insight in why you don't want to join the effort, but go your own way. If you have different ideas for the core than what you understand from the other teams, then there's no reason for you to join the effort. I did not realize that was the case, now I do understand you better. > nature invented forking for seeking new alternatives! A good way of expressing it :-) At the same time, there's something to be said for the old, solid and good non-forked original product... Abba is still more enjoyable than all the forks... Ha ha! Cheers, /O

1 0

Re: [SR-Dev] [OpenSIPS-Users] [Fwd: [Serdev] the sip router project]
by jh＠tutpro.com 05 Nov '08

05 Nov '08

Bogdan-Andrei Iancu writes: > Have you ever consider a fork as an option driven by technical needs?? > Like to do something totally different than you have and than the other > people want to do. > > After 7 year of SER/OpenSER I (and many other) got to simplest > conclusion that the current design is not able to sustain the progress > of SER / OpenSER (like scripting, async calls, integration, scaling, > etc) ? Mainly because SER was design 7 years ago when there was only > stateless processing, no TCP, etc.... bogdan, the above makes sense, but i have never been able to figure out why you could not implement the new designs as part of openser project that you yourself started. if it was due to people, that may happen any time again (unless you work alone) and has nothing to do with technical needs. -- juha

1 0

the sip router project
by Daniel-Constantin Mierla 05 Nov '08

05 Nov '08

Hello Everybody, We are pleased to announce to you the SIP Router project. It aims to build a solid open source SIP routing platform, based on collaboration of the SIP Express Router (SER) and Kamailio (OpenSER) teams. Developers of these two projects believe that an united and non-conflicting environment will bring many benefits to them, community members and companies: * bring together the developers and user communities of both projects * reduce maintenance overhead * avoid duplicated efforts in development * develop a core framework that is flexible, extensible and scalable * promote and build a solid open source SIP server project * ensure business credibility * make future forking undesirable, this harms everybody, affects credibility and business You are welcome to join! Visit the web site at: http://sip-router.org There is a meeting in Karlsruhe, Germany, on Nov 10, 2008, hosted by 1&1, where the developers and community members have the chance to discuss and tune the last aspects of the new project. We are looking to see many of you there: http://sip-router.org/index.php/meeting/ We hope this is a great news for you, thanks to the effort of main developers and management teams of the two projects. We invite you to join the new mailing list for further discussions: http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-dev Daniel, Jiri Kamailio Management Team SER Management Team

10 10

Re: [SR-Dev] [OpenSIPS-Users] [Fwd: [Serdev] the sip router project]
by Adrian Georgescu 04 Nov '08

04 Nov '08

Hello Jiri, For myself and Bogdan things are pretty clear. We both have companies that closely depend on the success of the project and we have contracts with customers to which we deliver. There is no doubt about who we are, what we do, what is our motivation and driving factor. Clarity is something customers need to have for the sake of their current and future investments. It is pretty unclear at this stage how the new project will shape up. Merging and forking can lead to inimaginable places as we have seen before several times and is nothing bad in the end as long as the eco- system keeps going and the customers invest further in it, which happend a lot in the last four years. The real customers are the ones that can decide upon the best solution for their own interests and we shall let them express their choice once they have enough information and measurements. Regards, Adrian On Nov 4, 2008, at 2:46 PM, Jiri Kuthan wrote: > Folks, > > while there is not yet such a remarkable progress to be shared with > those > I have chosen to talk to on the opensips side, the invitation is of > course > very open to anyone with sincere interest in unforking. > > -jiri > > --- > Hello Everybody, > > We are pleased to announce to you the SIP Router project. > > It aims to build a solid open source SIP routing platform, based on > collaboration of the SIP Express Router (SER) and Kamailio (OpenSER) > teams. > > Developers of these two projects believe that an united and > non-conflicting environment will bring many benefits to them, > community > members and companies: > > * bring together the developers and user communities of both > projects > * reduce maintenance overhead > * avoid duplicated efforts in development > * develop a core framework that is flexible, extensible and > scalable > * promote and build a solid open source SIP server project > * ensure business credibility > * make future forking undesirable, this harms everybody, affects > credibility and business > > You are welcome to join! Visit the web site at: > > http://sip-router.org > > There is a meeting in Karlsruhe, Germany, on Nov 10, 2008, hosted by > 1&1, where the developers and community members have the chance to > discuss and tune the last aspects of the new project. We are looking > to > see many of you there: > http://sip-router.org/index.php/meeting/ > > We hope this is a great news for you, thanks to the effort of main > developers and management teams of the two projects. We invite you to > join the new mailing list for further discussions: > > http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-dev > > Daniel, Jiri > Kamailio Management Team > SER Management Team > > > > _______________________________________________ > Users mailing list > Users(a)lists.opensips.org > http://lists.opensips.org/cgi-bin/mailman/listinfo/users

4 3

Re: [SR-Dev] [Kamailio-Users] the sip router project
by Johansson Olle E 04 Nov '08

04 Nov '08

4 nov 2008 kl. 16.24 skrev Bogdan-Andrei Iancu: > Hi Olle, > > Thank you for your thoughts. On a first view, it looks interesting, > but I'm missing some points here (important points): > > 1) as OpenSER was forked from SER because different views (and the > OpenSER view proved to be a very popular and successful one), I > wonder why, Kamilio is getting back to SER? not sharing any more the > OpenSER view as claimed? because such merging will definitely have a > great impact on the dynamical and openness of the projects (like > releases, contributions, driving the project) > > 2) this major change of perspective (at least for kamilio) was a > backstage decision, kept secret from the community - shouldn't be in > the interest of the community to say if going back to the roots > (merging into SER) is something wanted or not? it somehow > contradicts the self existence of OpenSER, right? > > 3) the benefits you mentions are mainly optimization of the internal > project activities and not optimizations of the outcome - what the > project will deliver. And I guess this is the most important. We > already went though the experience of large devel community, > frameworks, etc but with no outcome for more than 2 years... > >> From my personal perspective, the new project looks more like SER >> absorbing Kamilio (considering the sizes, the companies behind each > project, the resources, and the man-power behind each project). > > And at the moment I would like preserve the OpenSER vision and to > have an open source project (a standalone one), far away from the > "control" of any Big Brother ;) Bogdan, Thanks for your feedback, which clearly indicates where you stand today. We will have to see what the community says, but so far, I've only seen cheers and applauds. How this will work out in the future is something no one really can guess at this point. The more people that joins this effort and forms it, the better. The time to make it right is now, by working together and aligning the codebases as described, then working towards the future with new releases. We'll see if the combined forces can produce any deliverys or not. Today, we can only hope and guess. I think it's a great starting point that they have started talking, and agreed to work together on some parts. To me, it seems to be room for two final products stil, sharing the same core. One Kamailio and one SER. After the developer meetings, I think it's time to call for a joint community meeting to have people active in the community - not only developers - in the same room, discussing benefits of working together. I don't want to argue your standpoint, just hope that you will follow the progress and see if there's some benefits for OpenSIPS to join at some stage. Regardless, the GPL license allows you to benefit from the work as always. I might be naive, but I do have a positive feeling about this :-) /O

3 2

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

sr-dev