[SR-Users] Crash kamailio 4.3.1

Federico Cabiddu federico.cabiddu at gmail.com
Fri Sep 11 13:16:56 CEST 2015


Hi Daniel,
thank you for having a look. Regard your remarks:

1) when the cloned transaction structure is passed to ts_onreply, we take
the pointer to its urecord and from this last one we take the transaction
list. So, the transaction list should still be coherent. We could have a
problem if the urecord does not exist anymore (even if I'm wondering how
this could happen since a urecord is removed only when receiving a
TMCB_DESTROY event for the last transaction of a urecord). I could change
this to calculate the hash_id from the ruri in the cloned transaction,
search for it in the hash slot and only if found walk through the
transactions.

2) I checked and it looks to me that we always lock the entry (either with
lock_entry_by_ruri or with lock_entry) before performing an operation on
the table (records or transactions). Did I miss a place where this is not
happening (can happen looking the same code over and over again :))?

I'll keep looking at it.

Regards,
Federico

On Fri, Sep 11, 2015 at 9:10 AM, Daniel-Constantin Mierla <miconda at gmail.com
> wrote:

> Hi Federico,
>
> didn't get the time to write a more detailed report, but my quick look at
> the tsilo code revealed two potential issues:
>
> - the ts transaction structure is cloned in shared memory to be passed as
> parameter to a tm callback. The clone is stil linked to the list with prev
> and next. In the callback, if I got it properly, it starts walking through
> the list, but the list chould have been updated and prev/next can point to
> invalid data structure now.
>
> - there is a hash table that has locks for each slot, but those locks are
> not use and parallel operations (add/remove) can be done in the hash table
> by different kamailio processes
>
> When I get more time I will try to check again and see if those
> suppositions are valid. Meanwhile, maybe you can check as well.
>
> Cheers,
> Daniel
>
>
> On 10/09/15 19:33, Federico Cabiddu wrote:
>
> Hi Thibault,
> have you tried last tsilo code from 4.3.x branch?
> Recently there has been a fix (
> https://github.com/kamailio/kamailio/commit/6ce6803d57dabe287d7d6fa859e93c1df402d821)
> for an issue that may be related to yours.
> I'll keep investigating to see if I can spot something else. In the
> meanwhile could you describe your scenario? Are you storing multiple
> transactions per ruri? Did any of them got a final reply before the crash?
>
> Regards,
>
> Federico
>
> On Thu, Sep 10, 2015 at 3:00 PM, Daniel-Constantin Mierla <
> <miconda at gmail.com>miconda at gmail.com> wrote:
>
>>
>>
>> On 10/09/15 14:36, Thibault Gueslin wrote:
>>
>>
>>
>> 2015-09-10 14:25 GMT+02:00 Daniel-Constantin Mierla <miconda at gmail.com>:
>>
>>> Do you have msrp enabled in configuration file
>>>
>>
>> I don't think so
>>
>>
>> The last frames of backtrace indicates code related to msrp, but might be
>> just some code lines mismatching.
>>
>> The issue seems to be in tsilo. I looke over the code and I spotted some
>> "unclear" mechanisms that can lead to race conditions, which may result in
>> invalid access to memory, as it happens in this case, ptr becomes
>> 0x8b08578b49642454 -- from my short investigation, that is likely to be due
>> to following a ->next field in a freed structure.
>>
>> Not being the author of tsilo module, I can't do much more right now. I
>> will open an issue on bug tracker explaining what I found, assigning
>> Federico (cc-ed, author of the module) to analyze and see if anything is
>> wrong there.
>>
>> Cheers,
>> Daniel
>>
>>
>>
>>
>>> Send the output from gdb for next commands:
>>>
>>> frame 0
>>>
>>
>> #0  ts_append (msg=0x7f883bfd5490, ruri=0x7fff65fbfe30,
>> table=0x7f883bf7e390 "location") at ts_append.c:60
>> 60 in ts_append.c
>>
>>
>>> info locals
>>>
>> _r = 0x7f88339bcfd0
>>
>> ptr = 0x8b08578b49642454
>>
>> res = 2
>> __FUNCTION__ = "ts_append"
>>
>>
>>> p *msg
>>>
>>
>>
>>
>>> p *ruri
>>> p *ptr
>>> p *_r
>>>
>>>
>> | p *msg
>>
>> $21 = {id = 2, pid = 31171, tval = {tv_sec = 1441885042, tv_usec =
>> 162339}, fwd_send_flags = {f = 0 '\000', blst_imask = 0 '\000'},
>> rpl_send_flags = {f = 0 '\000', blst_imask = 0 '\000'}, first_line = {type
>> = 1,
>>     flags = 1, len = 46, u = {request = {method = {
>>           s = 0x7f88339b8910 "REGISTER sip:sip-staging.serveur.com
>> SIP/2.0\r\nVia: SIP/2.0/TCP 172.16.224.222:52242;branch=z9hG4bK20965be26f6a6324;rport\r\nContact:
>> <sip:toto4.toto.com.Ipod_tgu at 172.16.224.222:52242;transport=tcp>;expi"...,
>> len = 8}, uri = {
>>           s = 0x7f88339b8919 "sip:sip-staging.serveur.com
>> SIP/2.0\r\nVia: SIP/2.0/TCP 172.16.224.222:52242;branch=z9hG4bK20965be26f6a6324;rport\r\nContact:
>> <sip:toto4.toto.com.Ipod_tgu at 172.16.224.222:52242;transport=tcp>;expires=900\r\n"...,
>> len = 27}, version = {
>>           s = 0x7f88339b8935 "SIP/2.0\r\nVia: SIP/2.0/TCP
>> 172.16.224.222:52242;branch=z9hG4bK20965be26f6a6324;rport\r\nContact:
>> <sip:toto4.toto.com.Ipod_tgu at 172.16.224.222:52242;transport=tcp>;expires=900\r\nMax-Forwards:
>> 69\r\nTo: <sip:t"..., len = 7}, method_value = 32}, reply = {version = {
>>           s = 0x7f88339b8910 "REGISTER sip:sip-staging.serveur.com
>> SIP/2.0\r\nVia: SIP/2.0/TCP 172.16.224.222:52242;branch=z9hG4bK20965be26f6a6324;rport\r\nContact:
>> <sip:toto4.toto.com.Ipod_tgu at 172.16.224.222:52242;transport=tcp>;expi"...,
>> len = 8}, status = {
>>           s = 0x7f88339b8919 "sip:sip-staging.serveur.com
>> SIP/2.0\r\nVia: SIP/2.0/TCP 172.16.224.222:52242;branch=z9hG4bK20965be26f6a6324;rport\r\nContact:
>> <sip:toto4.toto.com.Ipod_tgu at 172.16.224.222:52242;transport=tcp>;expires=900\r\n"...,
>> len = 27}, reason = {
>>           s = 0x7f88339b8935 "SIP/2.0\r\nVia: SIP/2.0/TCP
>> 172.16.224.222:52242;branch=z9hG4bK20965be26f6a6324;rport\r\nContact:
>> <sip:toto4.toto.com.Ipod_tgu at 172.16.224.222:52242;transport=tcp>;expires=900\r\nMax-Forwards:
>> 69\r\nTo: <sip:t"..., len = 7}, statuscode = 32}}}, via1 =
>> 0x7f883bfa7020, via2 = 0x0, headers = 0x7f883bfa6f30, last_header =
>> 0x7f883bfac220, parsed_flag = 18446744073709551615, h_via1 =
>> 0x7f883bfa6f30,
>>   h_via2 = 0x0, callid = 0x7f883bfac400, to = 0x7f883bfac688, cseq =
>> 0x7f883bfac388, from = 0x7f883bfa6fa8, contact = 0x7f883bfac778,
>> maxforwards = 0x7f883bfac700, route = 0x0, record_route = 0x0,
>>   content_type = 0x0, content_length = 0x7f883bfac220, authorization =
>> 0x0, expires = 0x0, proxy_auth = 0x0, supported = 0x7f883bfb6300, require =
>> 0x0, proxy_require = 0x0, unsupported = 0x0,
>>   allow = 0x7f883bfac298, event = 0x0, accept = 0x0, accept_language =
>> 0x0, organization = 0x0, priority = 0x0, subject = 0x0, user_agent =
>> 0x7f883bfac310, server = 0x0, content_disposition = 0x0, diversion = 0x0,
>>   rpid = 0x0, refer_to = 0x0, session_expires = 0x0, min_se = 0x0,
>> sipifmatch = 0x0, subscription_state = 0x0, date = 0x0, identity = 0x0,
>> identity_info = 0x0, pai = 0x0, ppi = 0x0, path = 0x0, privacy = 0x0,
>>   body = 0x0, eoh = 0x7f88339b8b2b "\r\n", unparsed = 0x7f88339b8b2b
>> "\r\n", rcv = {src_ip = {af = 2, len = 4, u = {addrl = {3334267998, 0},
>> addr32 = {3334267998, 0, 0, 0}, addr16 = {58462, 50876, 0, 0, 0, 0, 0,
>>           0}, addr = "^\344\274\306", '\000' <repeats 11 times>}}, dst_ip
>> = {af = 2, len = 4, u = {addrl = {2667915013, 0}, addr32 = {2667915013, 0,
>> 0, 0}, addr16 = {9989, 40709, 0, 0, 0, 0, 0, 0},
>>         addr = "\005'\005\237", '\000' <repeats 11 times>}}, src_port =
>> 52243, dst_port = 5060, proto_reserved1 = 3, proto_reserved2 = 0, src_su =
>> {s = {sa_family = 2,
>>         sa_data = "\314\023^\344\274\306\000\000\000\000\000\000\000"},
>> sin = {sin_family = 2, sin_port = 5068, sin_addr = {s_addr = 3334267998},
>> sin_zero = "\000\000\000\000\000\000\000"}, sin6 = {
>>         sin6_family = 2, sin6_port = 5068, sin6_flowinfo = 3334267998,
>> sin6_addr = {__in6_u = {__u6_addr8 = "\000\000\000\000\000\000\000\000
>> \206\233\063\210\177\000", __u6_addr16 = {0, 0, 0, 0, 34336, 13211,
>>               32648, 0}, __u6_addr32 = {0, 0, 865830432, 32648}}},
>> sin6_scope_id = 865796800}}, bind_address = 0x7f883bfb6bd0, proto = 2
>> '\002'},
>>   buf = 0x7f88339b8910 "REGISTER sip:sip-staging.serveur.com
>> SIP/2.0\r\nVia: SIP/2.0/TCP 172.16.224.222:52242;branch=z9hG4bK20965be26f6a6324;rport\r\nContact:
>> <sip:toto4.toto.com.Ipod_tgu at 172.16.224.222:52242;transport=tcp>;expi"...,
>> len = 541, new_uri = {s = 0x0, len = 0}, dst_uri = {s = 0x0, len = 0},
>> parsed_uri_ok = 1, parsed_uri = {user = {s = 0x0, len = 0}, passwd = {s =
>> 0x0, len = 0}, host = {
>>       s = 0x7f88339b891d "sip-staging.serveur.com SIP/2.0\r\nVia:
>> SIP/2.0/TCP 172.16.224.222:52242;branch=z9hG4bK20965be26f6a6324;rport\r\nContact:
>> <sip:toto4.toto.com.Ipod_tgu at 172.16.224.222:52242;transport=tcp>;expires=900\r\nMax-"...,
>> len = 23}, port = {s = 0x0, len = 0}, params = {s = 0x0, len = 0},
>> sip_params = {s = 0x0, len = 0}, headers = {s = 0x0, len = 0}, port_no = 0,
>> proto = 0, type = SIP_URI_T,
>>     flags = (unknown: 0), transport = {s = 0x0, len = 0}, ttl = {s = 0x0,
>> len = 0}, user_param = {s = 0x0, len = 0}, maddr = {s = 0x0, len = 0},
>> method = {s = 0x0, len = 0}, lr = {s = 0x0, len = 0}, r2 = {s = 0x0,
>>       len = 0}, gr = {s = 0x0, len = 0}, transport_val = {s = 0x0, len =
>> 0}, ttl_val = {s = 0x0, len = 0}, user_param_val = {s = 0x0, len = 0},
>> maddr_val = {s = 0x0, len = 0}, method_val = {s = 0x0, len = 0},
>>     lr_val = {s = 0x0, len = 0}, r2_val = {s = 0x0, len = 0}, gr_val = {s
>> = 0x0, len = 0}}, parsed_orig_ruri_ok = 0, parsed_orig_ruri = {user = {s =
>> 0x0, len = 0}, passwd = {s = 0x0, len = 0}, host = {s = 0x0,
>>       len = 0}, port = {s = 0x0, len = 0}, params = {s = 0x0, len = 0},
>> sip_params = {s = 0x0, len = 0}, headers = {s = 0x0, len = 0}, port_no = 0,
>> proto = 0, type = ERROR_URI_T, flags = (unknown: 0), transport = {
>>       s = 0x0, len = 0}, ttl = {s = 0x0, len = 0}, user_param = {s = 0x0,
>> len = 0}, maddr = {s = 0x0, len = 0}, method = {s = 0x0, len = 0}, lr = {s
>> = 0x0, len = 0}, r2 = {s = 0x0, len = 0}, gr = {s = 0x0,
>>       len = 0}, transport_val = {s = 0x0, len = 0}, ttl_val = {s = 0x0,
>> len = 0}, user_param_val = {s = 0x0, len = 0}, maddr_val = {s = 0x0, len =
>> 0}, method_val = {s = 0x0, len = 0}, lr_val = {s = 0x0, len = 0},
>>     r2_val = {s = 0x0, len = 0}, gr_val = {s = 0x0, len = 0}}, add_rm =
>> 0x0, body_lumps = 0x0, reply_lump = 0x7f883bfa3878, add_to_branch_s =
>> '\000' <repeats 57 times>, add_to_branch_len = 0, hash_index = 39024,
>>   msg_flags = 129, flags = 32, set_global_address = {s = 0x0, len = 0},
>> set_global_port = {s = 0x0, len = 0}, force_send_socket = 0x0, path_vec =
>> {s = 0x0, len = 0}, instance = {s = 0x0, len = 0}, reg_id = 0,
>>   ruid = {s = 0x0, len = 0}, location_ua = {s = 0x0, len = 0}, ldv =
>> {flow = {decoded = 0, rcv = {src_ip = {af = 0, len = 0, u = {addrl = {0,
>> 0}, addr32 = {0, 0, 0, 0}, addr16 = {0, 0, 0, 0, 0, 0, 0, 0},
>>             addr = '\000' <repeats 15 times>}}, dst_ip = {af = 0, len =
>> 0, u = {addrl = {0, 0}, addr32 = {0, 0, 0, 0}, addr16 = {0, 0, 0, 0, 0, 0,
>> 0, 0}, addr = '\000' <repeats 15 times>}}, src_port = 0,
>>         dst_port = 0, proto_reserved1 = 0, proto_reserved2 = 0, src_su =
>> {s = {sa_family = 0, sa_data = '\000' <repeats 13 times>}, sin =
>> {sin_family = 0, sin_port = 0, sin_addr = {s_addr = 0},
>>             sin_zero = "\000\000\000\000\000\000\000"}, sin6 =
>> {sin6_family = 0, sin6_port = 0, sin6_flowinfo = 0, sin6_addr = {__in6_u =
>> {__u6_addr8 = '\000' <repeats 15 times>, __u6_addr16 = {0, 0, 0, 0, 0, 0,
>>                   0, 0}, __u6_addr32 = {0, 0, 0, 0}}}, sin6_scope_id =
>> 0}}, bind_address = 0x0, proto = 0 '\000'}}}}
>>
>>
>>
>> (gdb) p *ruri
>> $22 = {s = 0x7f883bf317e8 "toto4.toto.com.Ipod_tgu", len = 23}
>>
>> (gdb) p *ptr
>> Cannot access memory at address 0x8b08578b49642454
>>
>> (gdb) p *_r
>> $23 = {ruri = {s = 0x7f88339bd040 "toto4.toto.com.Ipod_tgu", len = 23},
>> rurihash = 164669906, entry = 0x7f883398af08, transactions =
>> 0x7f88339bd090, next = 0x0, prev = 0x0}
>>
>>
>>>
>>> Cheers,
>>> Daniel
>>>
>>>
>>> On 10/09/15 14:07, Thibault Gueslin wrote:
>>>
>>> [Thread debugging using libthread_db enabled]
>>>
>>> Using host libthread_db library
>>> "/lib/x86_64-linux-gnu/libthread_db.so.1".
>>>
>>> Core was generated by `/usr/sbin/kamailio -f /etc/kamailio/kamailio.cfg
>>> -P /var/run/kamailio/kamailio.'.
>>>
>>> Program terminated with signal SIGSEGV, Segmentation fault.
>>>
>>> #0  ts_append (msg=0x7f883bfd5490, ruri=0x7fff65fbfe30,
>>> table=0x7f883bf7e390 "location") at ts_append.c:60
>>>
>>> 60 ts_append.c: No such file or directory.
>>>
>>> (gdb) bt
>>>
>>> #0  ts_append (msg=0x7f883bfd5490, ruri=0x7fff65fbfe30,
>>> table=0x7f883bf7e390 "location") at ts_append.c:60
>>>
>>> #1  0x00007f8837f638fc in w_ts_append (_msg=0x7f883bfd5490,
>>> _table=0x7f883bf7e390 "location", _ruri=0x7f88339aa634 "\002") at
>>> tsilo.c:225
>>>
>>> #2  0x0000000000534db0 in do_action (h=0x7fff65fc1a40, a=0x7f883bf7d278,
>>> msg=0x7f883bfd5490) at action.c:1059
>>>
>>> #3  0x00000000005339e8 in run_actions (h=0x7f8837f65540, a=0x5,
>>> msg=0x7f88339aa634) at action.c:1548
>>>
>>> #4  0x00000000005352e6 in do_action (h=0x7fff65fc1a40, a=0x7f883bf7dc50,
>>> msg=0x7f883bfd5490) at action.c:1048
>>>
>>> #5  0x00000000005339e8 in run_actions (h=0x7f8837f65540, a=0x5,
>>> msg=0x7f88339aa634) at action.c:1548
>>>
>>> #6  0x00000000005352e6 in do_action (h=0x7fff65fc1a40, a=0x7f883bf7df48,
>>> msg=0x7f883bfd5490) at action.c:1048
>>>
>>> #7  0x00000000005339e8 in run_actions (h=0x7f8837f65540, a=0x5,
>>> msg=0x7f88339aa634) at action.c:1548
>>>
>>> #8  0x00000000005354a3 in do_action (h=0x7fff65fc1a40, a=0x7f883bfa8ab8,
>>> msg=0x7f883bfd5490) at action.c:677
>>>
>>> #9  0x00000000005339e8 in run_actions (h=0x7f8837f65540, a=0x5,
>>> msg=0x7f88339aa634) at action.c:1548
>>>
>>> #10 0x00000000005354a3 in do_action (h=0x7fff65fc1a40, a=0x7f883bfa8ab8,
>>> msg=0x7f883bfd5490) at action.c:677
>>>
>>> #11 0x00000000005339e8 in run_actions (h=0x7f8837f65540, a=0x5,
>>> msg=0x7f88339aa634) at action.c:1548
>>>
>>> #12 0x00000000005352e6 in do_action (h=0x7fff65fc1a40, a=0x7f883bf56068,
>>> msg=0x7f883bfd5490) at action.c:1048
>>>
>>> #13 0x00000000005339e8 in run_actions (h=0x7f8837f65540, h at entry=0x7fff65fc1a40,
>>> a=0x5, a at entry=0x7f883bf4f958, msg=0x7f88339aa634, msg at entry=0x7f883bfd5490)
>>> at action.c:1548
>>>
>>> #14 0x000000000053f885 in run_top_route (a=0x7f883bf4f958,
>>> msg=0x7f883bfd5490, c=<optimized out>) at action.c:1634
>>>
>>> #15 0x00000000005407e6 in receive_msg (buf=0x0, len=5,
>>> rcv_info=0x7f88339b8638) at receive.c:196
>>>
>>> #16 0x00000000005b806b in tcp_read_req (con=0x7f88339b8620,
>>> bytes_read=0x7fff65fc1d50, read_flags=0x7fff65fc1d54) at tcp_read.c:1382
>>>
>>> #17 0x00000000005bb4e1 in handle_io (fm=0xb, events=5, idx=865773108) at
>>> tcp_read.c:1568
>>>
>>> #18 0x00000000005c191a in io_wait_loop_epoll (h=<optimized out>,
>>> t=<optimized out>, repeat=<optimized out>) at io_wait.h:1061
>>>
>>> #19 tcp_receive_loop (unix_sock=938890560) at tcp_read.c:1733
>>>
>>> #20 0x00000000004d3447 in tcp_init_children () at tcp_main.c:4787
>>>
>>> #21 0x0000000000506863 in main_loop () at main.c:1658
>>>
>>> #22 0x000000000041b944 in main (argc=0, argv=0x0) at main.c:2533
>>>
>>>
>>> It seems to be on reception of REGISTER from client.
>>>
>>>
>>>
>>>
>>> --
>>> Daniel-Constantin Mierlahttp://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda
>>> Book: SIP Routing With Kamailio - http://www.asipto.com
>>> Kamailio Advanced Training, Sep 28-30, 2015, in Berlin - http://asipto.com/u/kat
>>>
>>>
>>
>> --
>> Daniel-Constantin Mierlahttp://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda
>> Book: SIP Routing With Kamailio - http://www.asipto.com
>> Kamailio Advanced Training, Sep 28-30, 2015, in Berlin - http://asipto.com/u/kat
>>
>>
>
> --
> Daniel-Constantin Mierlahttp://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda
> Book: SIP Routing With Kamailio - http://www.asipto.com
> Kamailio Advanced Training, Sep 28-30, 2015, in Berlin - http://asipto.com/u/kat
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sip-router.org/pipermail/sr-users/attachments/20150911/bbe46d18/attachment.html>


More information about the sr-users mailing list