Hello,
finally had a chance to look over it. I guess it is some race,
resulting in a memory space being overwritten. The 200ok for publish
with ringing state is processed more or less at the time when 200ok
for invite comes and dialog changes state.
Can you re-compile with MEMDBG=1 (edit the Makefile.defs or give it
in the command line parameter) and try again? Maybe the memory
manager can catch the fault operation and provide more details.
Cheers,
Daniel
On 5/17/12 6:58 PM, Charles Chance wrote:
Hi Daniel,
I have sent the messages over to you.
I look forward to hear what you can find.
Best,
Charles
------------------------------------------------------------------------
*From:*Daniel-Constantin Mierla [mailto:miconda@gmail.com]
*Sent:* 17 May 2012 15:06
*To:* Charles Chance
*Cc:* 'SIP Router - Kamailio (OpenSER) and SIP Express Router (SER)
-Users Mailing List'
*Subject:* Re: FW: (Devel) Segmentation fault using pua_dialoginfo
Hello,
the content of ps->param is pretty much all invalid. So looks like a
callback on a freed/invalid parameter, could be a double-callback
execution.
Set debug=3 and send all the messages for such case -- it might be
quite a lot of them, so send them to my email address, the mailing
list has a limit of message size.
Cheers,
Daniel
On 5/17/12 2:13 PM, Charles Chance wrote:
Hi,
Requested output as follows:
(gdb) frame 1
#1 publ_cback_func (t=0xb3dc8e38, type=1024, ps=0xbfcdd5d8)
at send_publish.c:246
246 hash_code= core_hash(hentity->pres_uri, NULL, HASH_SIZE);
(gdb) p *ps
$2 = {req = 0x0, rpl = 0xb7c30340, param = 0xb3dc3ff4, code = 412,
flags = 0,
branch = 0, t_rbuf = 0x0, dst = 0x0, send_buf = {s = 0x0, len = 0}}
(gdb) p *(ua_pres_t*)(*ps->param)
$3 = {id = {s = 0x20455942 <Address 0x20455942 out of bounds>,
len = 980445555}, pres_uri = 0x40323531, event = 875444279,
expires = 875703856, desired_expires = 858861105, flag = 808794676,
db_flag = 1394618421, cb_param = 0x322f5049, next = 0xa0d302e,
ua_flag = 979462486, etag = {
s = 0x50495320 <Address 0x50495320 out of bounds>, len =
808333871},
tuple_id = {s = 0x5044552f <Address 0x5044552f out of bounds>,
len = 775436064}, body = 0x322e3034, content_type = {
s = 0x312e3134 <Address 0x312e3134 out of bounds>, len =
1648047155},
watcher_uri = 0x636e6172, call_id = {
s = 0x397a3d68 <Address 0x397a3d68 out of bounds>, len =
1647593320},
to_tag = {s = 0x3435634b <Address 0x3435634b out of bounds>,
len = 842149473}, from_tag = {
s = 0x63306461 <Address 0x63306461 out of bounds>, len =
808334648},
cseq = 1767246349, version = 1394621025, outbound_proxy = 0x322f5049,
extra_headers = 0x552f302e, record_route = {
s = 0x31205044 <Address 0x31205044 out of bounds>, len =
825111097},
remote_contact = {s = 0x312e3836 <Address 0x312e3836 out of bounds>,
len = 976498224}, contact = {
s = 0x36373135 <Address 0x36373135 out of bounds>, len =
1701985073}}
Yes, it is a test server so very happy to arrange for remote access
if required. In the meantime, I will do a little more digging to
try to find out why entity is null.
Cheers,
Charles
--
Daniel-Constantin Mierla -http://www.asipto.com
http://twitter.com/#!/miconda <http://twitter.com/#%21/miconda>
-http://www.linkedin.com/in/miconda
------------------------------------------------------------------------
No virus found in this message.
Checked by AVG -
www.avg.com <http://www.avg.com>
Version: 2012.0.2176 / Virus Database: 2425/5004 - Release Date:
05/16/12
-http://www.linkedin.com/in/miconda
Kamailio Advanced Training, Seattle, USA, Sep 23-26, 2012 -http://asipto.com/u/katu