Re: [sr-dev] [kamailio/kamailio] Makefile.defs: clang as default and lib for 64 bit on FreeBSD (#845)

2016-11-03 Thread Daniel-Constantin Mierla
Is this for all FreeBSD versions or only from a specific version on? Many server are running old versions of distros. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub:

[sr-dev] git:master:76e2a0dd: rtpengine: init all struct to get rid of compile warnings

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: 76e2a0dd05aa1a6fb3148c231d57cd2278ced9d2 URL: https://github.com/kamailio/kamailio/commit/76e2a0dd05aa1a6fb3148c231d57cd2278ced9d2 Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

[sr-dev] git:master:66bb1a55: mi_fifo: close fifo stream in case of errors

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: 66bb1a55f4e1b6e8dbfd8fbc3c6625e356daa734 URL: https://github.com/kamailio/kamailio/commit/66bb1a55f4e1b6e8dbfd8fbc3c6625e356daa734 Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

[sr-dev] git:master:f609a83e: mi_rpc: init structures to avoid access to invalid content in case of errors

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: f609a83e58eccae409a5966a002edba389bbc372 URL: https://github.com/kamailio/kamailio/commit/f609a83e58eccae409a5966a002edba389bbc372 Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

[sr-dev] git:master:0cc0b643: lib/binrpc: memset addr variable to avoind uninitialized fields

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: 0cc0b643ee6442a13f257e3d72b826f257501204 URL: https://github.com/kamailio/kamailio/commit/0cc0b643ee6442a13f257e3d72b826f257501204 Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

[sr-dev] git:master:aeb9b5af: ctl: proper trim for long error messages to avoid out of bounds access

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: aeb9b5af7099ef3f8d3df02dcde1617fe22e4e46 URL: https://github.com/kamailio/kamailio/commit/aeb9b5af7099ef3f8d3df02dcde1617fe22e4e46 Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

Re: [sr-dev] [kamailio/kamailio] Makefile.defs: clang as default and lib for 64 bit on FreeBSD (#845)

2016-11-03 Thread Rick
You are right. I forgot that `clang` is the default only since FreeBSD 10. For the lib dir I am not sure how long this is done. I will do some research. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub:

Re: [sr-dev] [kamailio/kamailio] fix TLS client config selection based on dst ip and port (#842)

2016-11-03 Thread sergey-vb
It is interesting but not easy solution. Because tls_lookup_cfg accepting only one ip and port... Another solution can be in [MANAGE_BRANCH] route add next check condition: ``` if ($rP == "TLS") { $xavp(tls=>server_name) = $rd + ":" + $rp; $xavp(tls=>server_id) = $rd + ":" + $rp; }

[sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio seg (#846)

2016-11-03 Thread Rick
When doing a dispatcher reload with kamcmd kamailio crashes. ``` # kamailio -v version: kamailio 5.0.0-dev6 (x86_64/freebsd) 276ccc flags: STATS: Off, USE_TCP, USE_TLS, USE_SCTP, TLS_HOOKS, USE_RAW_SOCKS, DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MEM, SHM_MMAP, PKG_MALLOC, Q_MALLOC, F_MALLOC,

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio seg (#846)

2016-11-03 Thread Rick
This happens every time I do reload. -- 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/846#issuecomment-258093753___ sr-dev mailing list

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread ng-voice GmbH
Hi Rick, there should be a second core-dump for the crash? The BT shows the shutdown, not where the actual crash happened. As I cannot reproduce the issue (works on my systems with the same version you're running), can you shed a bit more light about your configuration and your data? Thanks,

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Rick
Sorry there is no other core-dump created. Sorry I forgot to tell about the system, it is kamailio build from git-source on a FreeBSD 10.3-RELEASE-p7 machine. I attached gdb to the ctl process and got the following: ``` (gdb) s Single stepping until exit from function _kevent, which has no

Re: [sr-dev] [kamailio/kamailio] fix TLS client config selection based on dst ip and port (#842)

2016-11-03 Thread Daniel-Constantin Mierla
Closed #842. -- 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/pull/842#event-846118189___ sr-dev mailing list sr-dev@lists.sip-router.org

Re: [sr-dev] [kamailio/kamailio] fix TLS client config selection based on dst ip and port (#842)

2016-11-03 Thread Daniel-Constantin Mierla
The prototype of tls_lookup_cfg() can be changed if needed, that's not a problem. The solution with xavp is ok and available now, the one with new match attribute may be an alternative the config simpler. Anyhow, I am closing this one, given there is a solution. If anyone considers to

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Daniel-Constantin Mierla
Output of 'bt full' is more useful. Also get: 'frame 0' + 'info locals'. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub:

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Rick
Ok this is what I did, hope that is correct. ``` gdb /usr/local/sbin/kamailio /var/run/kamailio/kamailio.core ``` ``` (gdb) bt full #0 0x0008098d32c3 in ds_avl_destroy (node_ptr=0x809ed2b40) at dispatch.c:3046 node = (ds_set_t *) 0x809ee72d8 i = 2 dest = (ds_dest_t

[sr-dev] git:master:e415c3a0: dialog: safety check of return pointer to dmq vars field

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: e415c3a0493b15c19b8a7984ddaff0bfd170a5ed URL: https://github.com/kamailio/kamailio/commit/e415c3a0493b15c19b8a7984ddaff0bfd170a5ed Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Daniel-Constantin Mierla
It is needed from the ctl process core file, not from the shutdown core file. You have to enable core file per pid, otherwise it may get overwritten. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub:

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Rick
Ok now I got two core files this one is the second one: ``` (gdb) bt full #0 0x0008098ce9c8 in add_dest2list (id=2, uri={s = 0x80299c05a "sip:5.159.56.120", len = 16}, flags=0, priority=0, attrs=0x7fff4ed0, list_idx=1, setn=0x7fff4f04) at dispatch.c:434 dp = (ds_dest_t *)

[sr-dev] git:master:0369e5ab: dispatcher: coherent indentation

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: 0369e5ab3811a09cf1b1615950401967217bfce4 URL: https://github.com/kamailio/kamailio/commit/0369e5ab3811a09cf1b1615950401967217bfce4 Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

[sr-dev] git:master:a471e472: auth_ephemeral: re-ordered code so access to sip msg is done after NULL check

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: a471e4723206cdce7e9e319af344145b92444698 URL: https://github.com/kamailio/kamailio/commit/a471e4723206cdce7e9e319af344145b92444698 Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

[sr-dev] git:master:98f9cace: dispatcher: replaced some fields init with memset to 0

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: 98f9cace74adc33f42a6a6aca4d0adc1bcc81248 URL: https://github.com/kamailio/kamailio/commit/98f9cace74adc33f42a6a6aca4d0adc1bcc81248 Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Daniel-Constantin Mierla
The dp0 pointer ends up in an invalid value: ``` dp0 = (ds_dest_t *) 0xc0c0c0c0 ``` Likely a buffer overflow or a mis-shifting of pointer values. Can you grab the latest master branch and try again? I pushed some memset inits in dispatcher, likely not to solve the issue, but it's better to

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Rick
While I was try to test I am getting this with latest master: ``` # kamailio -c loading modules under config path: /usr/local/lib64/kamailio/modules/ 0(67414) ERROR: [sr_module.c:582]: load_module(): could not open module : /usr/local/lib64/kamailio/modules/alias_db.so: Undefined symbol

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Rick
the `inline int set_alias_to_pvar(struct sip_msg* _msg, str *alias, int no, void *p)` doesn't work for me with clang, when changed to `static inline int set_alias_to_pvar(struct sip_msg* _msg, str *alias, int no, void *p)` it works -- You are receiving this because you are subscribed to this

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Rick
With latest right now I didn't get a crash. -- 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/846#issuecomment-258176437___ sr-dev mailing

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Daniel-Constantin Mierla
If you get other issues not related to dispatcher, open a different item here, otherwise is not easy to track what's wrong on a topic. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub:

[sr-dev] git:master:45e94dca: alias_db: remove inline to function provided as parameter to another one

2016-11-03 Thread Daniel-Constantin Mierla
Module: kamailio Branch: master Commit: 45e94dca8e1fc2b7bae6cece0d3ac19fbc5e430d URL: https://github.com/kamailio/kamailio/commit/45e94dca8e1fc2b7bae6cece0d3ac19fbc5e430d Author: Daniel-Constantin Mierla Committer: Daniel-Constantin Mierla Date:

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Daniel-Constantin Mierla
Closed #846. -- 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/846#event-846591847___ sr-dev mailing list sr-dev@lists.sip-router.org

Re: [sr-dev] [kamailio/kamailio] dispatcher: kamcmd dispatcher.reload kamailio SIGCHLD (#846)

2016-11-03 Thread Daniel-Constantin Mierla
I removed the inline from the alias_db function, pushed the change to the master. Given that the dispatcher is not exposing the issue, I am closing this one. If something new appears, then maybe is better to open a new item in the bug tracker. -- You are receiving this because you are

Re: [sr-dev] [kamailio/kamailio] Presence: Fix startup inconsistency in presentity hash table (#809)

2016-11-03 Thread Emmanuel Schmidbauer
this commit broke presence for us. we use NSQ + presence, after a phone reboots and we send a NOTIFY with confirmed state, presence module sends a NOTIFY with terminated state. please revert this commit. see the log and SIP packet below: ``` Nov 3 09:23:52 /usr/local/sbin/kamailio[26288]:

Re: [sr-dev] [kamailio/kamailio] Presence: Fix startup inconsistency in presentity hash table (#809)

2016-11-03 Thread Daniel-Constantin Mierla
@eschmidbauer - is it affecting the runtime behaviour, because it should have been targeting only the startup procedure? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub:

Re: [sr-dev] [kamailio/kamailio] Presence: Fix startup inconsistency in presentity hash table (#809)

2016-11-03 Thread Emmanuel Schmidbauer
@miconda yes, we are seeing this issue during runtime. not on startup we reverted the commit and problem went away -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub:

Re: [sr-dev] [kamailio/kamailio] Presence: Fix startup inconsistency in presentity hash table (#809)

2016-11-03 Thread lazedo
the problem with this commit is for expired column = 0 which is usually set on "early" state -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub:

Re: [sr-dev] [kamailio/kamailio] Presence: Fix startup inconsistency in presentity hash table (#809)

2016-11-03 Thread Emmanuel Schmidbauer
@lazedo you can see that we sending expires=1 in our NSQ message and in nsq_pua.c ```C struct json_object *ExpiresObj = nsq_json_get_object(json_obj, BLF_JSON_EXPIRES); if (ExpiresObj != NULL) { expires = json_object_get_int(ExpiresObj); if

Re: [sr-dev] [kamailio/kamailio] Presence: Fix startup inconsistency in presentity hash table (#809)

2016-11-03 Thread lazedo
@eschmidbauer yep, but that's wrong by definition, we had the same problem. try sending `expires = 3600` , it will work. i'm ok with this commit, it could be improved if we were able to ignore the `0` which we interpret as `do not expire` -- You are receiving this because you are subscribed

Re: [sr-dev] [kamailio/kamailio] Presence: Fix startup inconsistency in presentity hash table (#809)

2016-11-03 Thread Emmanuel Schmidbauer
it's not entirely clear to me what this PR fixes. What about adding a module parameter (default = enabled) to enable this code? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: