| 00:15.51 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 00:32.05 | gtjoseph | coreyfarrell: Is "undefined symbol _ast_mem_backtrace_buffer" in the DPMA related to something you're doing? |
| 00:32.43 | coreyfarrell | ugh |
| 00:33.31 | coreyfarrell | yes, it's my fault :( |
| 00:37.18 | gtjoseph | was there a fix? |
| 00:37.38 | gtjoseph | I thought I saw a review by sgriepentrog that said something now worked with the DPMA |
| 00:38.19 | coreyfarrell | yeah that was a fix for the other direction, so DPMA built against 13.5 will work on asterisk 13.0 |
| 00:38.37 | mjordan | coreyfarrell: Foo Bar is my test account. I'll disable it :-) |
| 00:38.55 | coreyfarrell | the problem here is DPMA built against 13.0 uses _ast_mem_backtrace_buffer which is currently missing on branch head |
| 00:39.00 | coreyfarrell | mjordan: I just found it funny |
| 00:39.18 | gtjoseph | gotcha |
| 00:39.25 | gtjoseph | not a critical issue for me now |
| 00:40.12 | gtjoseph | i was just testing a new DPMA a few days ago (which worked) and happened to see the messages fly by today while working on something else |
| 00:40.13 | coreyfarrell | gtjoseph: yeah well I need to get a fix in place soon (before any release that includes r4502). pretty sure r4502 isn't in the current RC's. |
| 00:40.35 | mjordan | it isn't |
| 00:40.42 | coreyfarrell | thought so thanks mjordan |
| 00:42.20 | coreyfarrell | gtjoseph: so I think I'll have to fix this in a separate review, as I need to add the _ast_mem_backtrace_buffer symbol to /branches/13 only - it never existed in 11.x, no need for compatibility in trunk. |
| 00:42.54 | coreyfarrell | it'll just take me a bit to figure out how to create the symbol only on systems with the execinfo.h, since part of the point was to allow Asterisk to build on platforms where that is missing |
| 01:04.33 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 01:08.24 | *** join/#asterisk-dev snuff-work (~blatus@203.87.61.22) |
| 01:08.25 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 02:51.13 | coreyfarrell | mjordan: thanks for backporting all those clang warning fixes! |
| 02:51.22 | mjordan | no worries |
| 02:51.30 | mjordan | we might miss a few, but most aren't too hard to backport |
| 02:52.00 | coreyfarrell | usually more tedious than hard |
| 05:00.18 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-28-58.cablep.bezeqint.net) |
| 05:00.18 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 05:27.46 | *** join/#asterisk-dev Demon_VoIP (~demon@109.60.135.186) |
| 06:24.47 | *** join/#asterisk-dev jagruti_ (~root@182.70.116.47) |
| 06:25.22 | *** part/#asterisk-dev jagruti_ (~root@182.70.116.47) |
| 06:29.11 | *** join/#asterisk-dev jagruti_ (~root@182.70.116.47) |
| 06:29.36 | *** part/#asterisk-dev jagruti_ (~root@182.70.116.47) |
| 06:51.04 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 06:51.04 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 06:58.28 | *** join/#asterisk-dev hehol (~hehol@2001:1438:1009:200:54b3:f7d9:7b04:78b7) |
| 07:17.30 | *** join/#asterisk-dev pchero_work (~pchero@109.70.54.56) |
| 10:19.45 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 10:19.45 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:07.12 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@216-191-106-163.dedicated.allstream.net) |
| 12:37.02 | *** join/#asterisk-dev jameswf (uid27319@gateway/web/irccloud.com/x-zpsdocaqfezuyiyr) |
| 13:00.56 | *** join/#asterisk-dev anonymouz666 (~anonymouz@189-25-66-28.user.veloxzone.com.br) |
| 14:01.05 | *** join/#asterisk-dev sgriepentrog (sgriepentr@nat/digium/x-uunwlxiqklddvujl) |
| 14:05.30 | anonymouz666 | oej: very funny stuff today, the guy here executed "asterisk -rv" and teapot version appears... "what the hell is teapot...?" hehe |
| 14:06.39 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-zaziaaulageebosg) |
| 14:07.01 | oej | he he |
| 14:12.35 | *** join/#asterisk-dev mjordan (mjordan@nat/digium/x-eqaebnqcuzswxnvo) |
| 14:12.36 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 14:16.19 | stefan27 | Looks like I have another seg-fault to track down, yet again triggered by using webrtc. (webrtc seems attract bugs). It is a segmentation fault in libpjnath.so related to a whole forest of unresolved bugs (e.g. https://issues.asterisk.org/jira/browse/ASTERISK-22938) |
| 14:18.35 | mjordan | stefan27: what does your backtrace look like? |
| 14:20.45 | stefan27 | I cannot post the whole trace yet, because I dont know if it needs masking of IPs/passwords but this is the essential part of it: http://pastebin.com/hJxSWssu |
| 14:21.07 | stefan27 | It's upon receiving RTP (stun) while srtp is still 0x0 and its not RTCP |
| 14:22.09 | mjordan | that's not the same as the issue you linked. |
| 14:22.27 | stefan27 | it's on pj_stun_session_on_rx_pkt () |
| 14:22.33 | mjordan | so? |
| 14:22.35 | mjordan | :-) |
| 14:22.40 | mjordan | that's not the same as the backtrace on that issue. |
| 14:22.47 | mjordan | it's not an abort due to an assertion |
| 14:22.51 | mjordan | it's a crash due to ... something else. |
| 14:22.59 | *** join/#asterisk-dev putnopvut (putnopvut@asterisk/master-of-queues/mmichelson) |
| 14:22.59 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:22.59 | stefan27 | Aha, I never mean't same though, just related |
| 14:23.05 | mjordan | I wouldn't even say that. |
| 14:23.08 | stefan27 | but it may not be very related if you say so |
| 14:23.12 | file | I definitely would not |
| 14:23.18 | *** join/#asterisk-dev pabelanger (~pabelange@asterisk/contributor-and-bug-marshal/pabelanger) |
| 14:23.19 | *** mode/#asterisk-dev [+o pabelanger] by ChanServ |
| 14:23.26 | stefan27 | Can you identify any relating jiras? |
| 14:23.33 | stefan27 | related* |
| 14:23.40 | mjordan | no |
| 14:23.45 | mjordan | I'd file a new issue. |
| 14:23.49 | mjordan | SRTP being NULL doesn't matter. |
| 14:24.00 | mjordan | you aren't using SDES-SRTP, so I would expect that to be NULL. |
| 14:24.14 | stefan27 | I can try upgrading from pjprojects 2.2 to 2.3 |
| 14:24.21 | *** part/#asterisk-dev pabelanger (~pabelange@asterisk/contributor-and-bug-marshal/pabelanger) |
| 14:24.25 | stefan27 | but this happens like once a week at most! |
| 14:24.37 | stefan27 | i need to find a way to provoke this bug |
| 14:24.48 | mjordan | I'd install pjproject with debug symbols, since the crash is happening there |
| 14:39.06 | stefan27 | I tried to install that after crash. For pjprojects i tried creating my own rpm with spec file as in http://pastebin.com/AeMeBz3v it created a debug-info-rpm along my real rpm, the debuginfo rpm contained /usr/lib/debug/usr/lib64/libpjnath.so.2.debug and /usr/lib/debug/usr/lib64/libpjnath.so.debug but installing my debuginfo-rpm gave no additional info when re-doing the gdb-command |
| 14:40.11 | stefan27 | Maybe I made a mistake in the rpmbuild? |
| 14:51.08 | stefan27 | How does debug-symbols work? Is the existence of /usr/lib/debug/usr/lib64/libpjnath.so.2.debug on my machine not enough to make gdb -se "asterisk" -ex "bt full" -ex "thread apply all bt" --batch -c /tmp/core write additional debug-stuff whenever it detects a call like "0x00007ff88b7d1b87 in pj_stun_session_on_rx_pkt () from /lib64/libpjnath.so.2" |
| 15:14.34 | *** join/#asterisk-dev superscrat (asanders@nat/digium/x-cvslhcqtrmwbbqnb) |
| 15:22.51 | *** join/#asterisk-dev gtjoseph (~gtj0@unaffiliated/gtj) |
| 15:23.25 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-bddfaseedzzwxoif) |
| 15:35.57 | stefan27 | so we call pj_ice_sess_on_rx_pkt(rtp->ice, rtcp ? AST_RTP_ICE_COMPONENT_RTCP : AST_RTP_ICE_COMPONENT_RTP, rtcp ? TRANSPORT_SOCKET_RTCP : TRANSPORT_SOCKET_RTP, buf, len, &address, pj_sockaddr_get_len(&address)); can any of those parameters get changed by other threads before this call completes? |
| 15:40.15 | file | no. |
| 15:41.25 | stefan27 | I really need to get debug symbols in pjprojects |
| 16:00.39 | *** join/#asterisk-dev mjordan (mjordan@nat/digium/x-ykufuwsdztfoaxsm) |
| 16:00.39 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 16:01.28 | stefan27 | http://pastebin.com/12TfFNF0 Why is len sent as parameter and not sizeof(buf) by the way? |
| 16:05.15 | stefan27 | we call len = ast_recvfrom(rtcp ? rtp->rtcp->s : rtp->s, buf, size, flags, sa)) earlier: if len>0 is it still possible that len is not equal to sizeof(buf) ? |
| 16:05.55 | stefan27 | I'm just trying to guess possible causes, since it'll take ages to debug by just waiting for it to happen again and hopefully have more debuginfo by then. |
| 16:21.28 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 16:21.28 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 16:34.50 | *** join/#asterisk-dev w9sh (~sph@50-79-224-193-static.hfc.comcastbusiness.net) |
| 16:36.45 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 16:36.45 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 16:54.41 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@216-191-106-163.dedicated.allstream.net) |
| 16:59.38 | *** join/#asterisk-dev leedm777 (leedm777@nat/digium/x-nfvfrvivtlflqqze) |
| 17:10.16 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 17:10.16 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 18:00.47 | *** part/#asterisk-dev Marquis42 (~mrbmw@d14-69-246-193.try.wideopenwest.com) |
| 18:16.29 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 18:16.29 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 18:18.09 | *** join/#asterisk-dev Stefan26 (Stefan0@h105n2-s-oe-a15.ias.bredband.telia.com) |
| 18:30.37 | *** join/#asterisk-dev mjordan (mjordan@nat/digium/x-xhkghzykctbrnlno) |
| 18:30.37 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 19:21.09 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-28-58.cablep.bezeqint.net) |
| 19:21.09 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 19:21.50 | *** join/#asterisk-dev file (~file@asterisk/developer-and-muffin-lover/file) |
| 19:21.50 | *** mode/#asterisk-dev [+o file] by ChanServ |
| 19:32.07 | *** join/#asterisk-dev justsomedood (~warren@mail.serverplus.com) |
| 19:40.41 | *** join/#asterisk-dev pchero (~pchero@0x555140b5.adsl.cybercity.dk) |
| 19:43.22 | *** join/#asterisk-dev file (~file@asterisk/developer-and-muffin-lover/file) |
| 19:43.22 | *** mode/#asterisk-dev [+o file] by ChanServ |
| 19:43.59 | justsomedood | I'm patching asterisk for the asterisk manager events to escape control characters in the message format, i.e. \r and \n with a "\\r" and "'\\n" so that something like a callerid with a newline doesn't mess up a program parsing asterisk manager output. the function manager_event() uses the printf formatter to create the string to send, so I added a new formatter tag "M" for asterisk manager string output. However, asterisk is using the -Wf |
| 19:43.59 | justsomedood | The only other way I can think to do it is just escape the output before calling manager_event() and the like, but since the string will be longer a free() always has to be done afterwards. Is there a better way to do it that I'm missing?< |
| 19:51.59 | *** join/#asterisk-dev file (~file@asterisk/developer-and-muffin-lover/file) |
| 19:51.59 | *** mode/#asterisk-dev [+o file] by ChanServ |
| 19:57.07 | mjordan | justsomedood: I would escape the output before sending it out the door. |
| 19:58.11 | mjordan | in fact |
| 19:58.16 | mjordan | you can probably do it in astman_append |
| 19:58.40 | mjordan | hm |
| 19:58.43 | mjordan | well, not quite. |
| 19:58.57 | mjordan | It's unfortunate that send_string doesn't take in an ast_str |
| 19:59.43 | mjordan | to avoid the malloc/free, you could use an AST_THREADSTORAGE that wraps an ast_str |
| 20:00.39 | justsomedood | I was hoping I could do it in manager_event, like: "CallerIDNum: %M\r\n" where M is the manager escape. It works fine, it just give tons of compiler warning |
| 20:02.38 | justsomedood | Ok, I'll just do the escaping before calling the manager_event() |
| 20:14.14 | mjordan | compiler warnings => ERROR |
| 20:14.21 | mjordan | we don't allow compiler warnings in :-) |
| 20:14.28 | mjordan | you may want to take a look at the coding guidelines |
| 20:14.47 | mjordan | https://wiki.asterisk.org/wiki/display/AST/Coding+Guidelines |
| 20:26.40 | [TK]D-Fender | packs up to head home |
| 20:29.01 | justsomedood | Understood, I wish there were a way to tell -Wformat of the new printf tags, but alas there isn't. Thanks for linking the guidelines, though, that will save everybody some time when I submit it I'm sure :D |
| 20:42.28 | *** join/#asterisk-dev file (~file@asterisk/developer-and-muffin-lover/file) |
| 20:42.28 | *** mode/#asterisk-dev [+o file] by ChanServ |
| 20:51.11 | *** join/#asterisk-dev [TK]D-Fender (~joe@64.235.216.2) |
| 21:13.30 | *** join/#asterisk-dev file (~file@asterisk/developer-and-muffin-lover/file) |
| 21:13.30 | *** mode/#asterisk-dev [+o file] by ChanServ |
| 21:34.02 | *** join/#asterisk-dev file (~file@asterisk/developer-and-muffin-lover/file) |
| 21:34.02 | *** mode/#asterisk-dev [+o file] by ChanServ |
| 22:05.25 | justsomedood | Oh yeah, second question, I'm currently using asterisk 11 in production and would be testing on that. Is it ok to make a patch for the current 11 branch, and then make another for 13? |
| 22:35.14 | mjordan | sure, but I suspect that your 11 patch will apply okay to 13. If it does, you can just make the 11 patch. |
| 22:36.28 | *** part/#asterisk-dev mjordan (mjordan@nat/digium/x-xhkghzykctbrnlno) |
| 22:38.45 | *** join/#asterisk-dev DivideBy0 (~DivideBy0@unaffiliated/divideby0x0) |
| 22:52.43 | *** join/#asterisk-dev file (~file@asterisk/developer-and-muffin-lover/file) |
| 22:52.43 | *** mode/#asterisk-dev [+o file] by ChanServ |
| 22:54.06 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-zaziaaulageebosg) |