Re: [SR-Users] Core Dump from select stmt

Hello, On 4/17/12 4:52 AM, Akan wrote: you can download now all updates if you install latest git branch 3.2: * http://www.kamailio.org/wiki/install/3.2.x/git For backporting patches, here are some guidelines: * http://www.kamailio.org/wiki/devel/backporting-to-3.2.x Cheers, Daniel > > Thanks > > Nathaniel > > On 4/16/2012 8:31 AM, Daniel-Constantin Mierla wrote: >> Hello, >> >> the issue was produced inside the debugger module, while printing >> config trace, according to the worker process back trace. >> >> Can you try with the patch from following commit: >> >> http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=40ddcf8… >> >> >> If all goes fine with it, I will backport to stable branch. >> >> Cheers, >> Daniel >> >> >> On 4/16/12 2:58 AM, Akan wrote: >>> On Solaris they seem to manage core dumps a little differently. >>> Only 3 processes produced core dumps. I have included the text >>> files of the back trace. >>> >>> Thanks >>> >>> Nathaniel >>> >>> On 4/15/2012 5:34 AM, Daniel-Constantin Mierla wrote: >>>> Hello, >>>> >>>> this one seems from the main process during the shut down >>>> cleanup. Perhaps the core from main process overwritten the one >>>> from worker process. Can you enable one core file per process in >>>> Solaris? In linux is possible via: >>>> >>>> echo "1"> /proc/sys/kernel/core_uses_pid >>>> Then you have to reproduce the case and get the backtraces from >>>> each corefile. >>>> >>>> Cheers, >>>> Daniel >>>> >>>> On 4/14/12 6:58 PM, Akan wrote: >>>>> Here is the output from gdb that you requested: >>>>> >>>>> Core was generated by `./kamailio'. >>>>> Program terminated with signal 10, Bus error. >>>>> #0 0x00000001001c9ff0 in cfg_destory_groups () at >>>>> cfg/cfg_struct.c:306 >>>>> 306 >>>>> old_string = *(char **)(block + group->var_offset + >>>>> mapping[i].offset); >>>>> (gdb) bt full >>>>> #0 0x00000001001c9ff0 in cfg_destory_groups () at >>>>> cfg/cfg_struct.c:306 >>>>> group = 0x1004e2528 >>>>> def = <value optimized out> >>>>> i = <value optimized out> >>>>> group2 = <value optimized out> >>>>> mapping = 0x1004e23d8 >>>>> old_string = <value optimized out> >>>>> #1 cfg_destroy () at cfg/cfg_struct.c:398 >>>>> __FUNCTION__ = "cfg_destroy" >>>>> #2 0x0000000100060d44 in cleanup (show_status=1) at main.c:553 >>>>> memlog = <value optimized out> >>>>> __FUNCTION__ = "cleanup" >>>>> #3 0x0000000100061ae0 in shutdown_children (show_status=<value >>>>> optimized out>, sig=15) at main.c:678 >>>>> No locals. >>>>> #4 0x0000000100063470 in handle_sigs () at main.c:769 >>>>> chld = 0 >>>>> chld_status = 0 >>>>> memlog = <value optimized out> >>>>> #5 0x0000000100066310 in main_loop () at main.c:1713 >>>>> i = <value optimized out> >>>>> pid = <value optimized out> >>>>> si = <value optimized out> >>>>> si_desc = "udp receiver child=3 >>>>> sock=209.252.110.38:5060\000óÐ\000\000\000\001\000BE\000\000\000\000\000\000\000\000\002ÿÿÿÿ\177ÿóÁ\000\000\000\001\000\037z\200", >>>>> '\000' <repeats 14 times>, "gÐÿÿÿÿÿÿÿýÿÿÿÿÿÿÿýÿÿÿÿÿÿÿýÿÿÿÿÿÿÿý" >>>>> #6 0x00000001001f7e94 in main (argc=<value optimized out>, >>>>> argv=<value optimized out>) at main.c:2475 >>>>> cfg_stream = <value optimized out> >>>>> c = <value optimized out> >>>>> r = -1 >>>>> tmp = 0x0 >>>>> tmp_len = 0 >>>>> options = 0x100222858 >>>>> ":f:cm:M:dVhEb:l:L:n:vrRDTN:W:w:t:u:g:P:G:SQ:O:a:A:" >>>>> seed = 660384647 >>>>> rfd = <value optimized out> >>>>> debug_save = <value optimized out> >>>>> debug_flag = <value optimized out> >>>>> dont_fork_cnt = <value optimized out> >>>>> n_lst = 0x100000 >>>>> p = <value optimized out> >>>>> (gdb) >>>>> >>>>> >>>>> Thanks >>>>> >>>>> Nathaniel >>>>> >>>>> On 4/14/2012 3:10 AM, Daniel-Constantin Mierla wrote: >>>>>> Hello, >>>>>> >>>>>> from the log messages, you got a core dump file. Locate it >>>>>> (should be in / if you haven't specified -w 'path' or solaris >>>>>> does not have a special directory for corefiles) and get the >>>>>> back trace: >>>>>> >>>>>> gdb /path/to/kamailio /path/to/corefile >>>>>> >>>>>> Then inside gdb execute 'bt full' and send the output here. >>>>>> >>>>>> Cheers, >>>>>> Daniel >>>>>> >>>>>> On 4/14/12 2:56 AM, Akan Technology wrote: >>>>>>> hello, >>>>>>> >>>>>>> I am running kamailio 3.2.2 on Solaris 10 64 bit and I am >>>>>>> getting a core dump on this statement: >>>>>>> >>>>>>> $ru = "sip:" + $rU + "@" + $sel(cfg_get.registrar1.server_ip) >>>>>>> + >>>>>>> $sel(cfg_get.registrar1.server_port); >>>>>>> >>>>>>> >>>>>>> Here is how the variable is defined: >>>>>>> >>>>>>> registrar1.server_ip = "xxx.xxx.xxx.xxx" desc "Registrar >>>>>>> server IP address" >>>>>>> registrar1.server_port = "5080" desc "Registrar server Port" >>>>>>> >>>>>>> >>>>>>> Here is the output from the logs with debug: >>>>>>> >>>>>>> 10(25373) ERROR: *** cfgtrace: >>>>>>> c=[/opt/kamailio-3.2/etc/kamailio/kamailio.cfg] l=1072 a=65 >>>>>>> n=assign >>>>>>> 10(25373) DEBUG: <core> [select.c:424]: Calling SELECT 1004b2bc0 >>>>>>> 10(25373) DEBUG: <core> [select.c:424]: Calling SELECT 1004b4da0 >>>>>>> 10(25373) ERROR: *** cfgtrace:14(25377) : <core> >>>>>>> [pass_fd.c:293]: ERROR: receive_fd: EOF on 22 >>>>>>> 14(25377) DEBUG: <core> [tcp_main.c:3555]: DBG: >>>>>>> handle_ser_child: dead child 10, pid 25373 (shutting down?) >>>>>>> 14(25377) DEBUG: <core> [io_wait.h:617]: DBG: io_watch_del >>>>>>> (1003738b8, 22, 0, 0x0) fd_no=18 called >>>>>>> 14(25377) DEBUG: <core> [tcp_main.c:3316]: DBG: >>>>>>> handle_tcp_child: dead tcp child 0 (pid 25373, no 10) >>>>>>> (shutting down?) >>>>>>> 14(25377) DEBUG: <core> [io_wait.h:617]: DBG: io_watch_del >>>>>>> (1003738b8, 24, 1, 0x0) fd_no=17 called >>>>>>> 0(25363) ALERT: <core> [main.c:751]: child process 25373 >>>>>>> exited by a signal 11 >>>>>>> 0(25363) ALERT: <core> [main.c:754]: core was generated >>>>>>> 0(25363) INFO: <core> [main.c:766]: INFO: terminating due to >>>>>>> SIGCHLD >>>>>>> 14(25377) INFO: <core> [main.c:817]: INFO: signal 15 received