----- Forwarded message from Eugeny Korekin <[EMAIL PROTECTED]> -----
Date: Sat, 4 Jan 2003 16:12:49 +0600
From: Eugeny Korekin <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Subject: [jabberd] jabberd crashes
Delivered-To: [EMAIL PROTECTED]
Delivered-To: [EMAIL PROTECTED]
X-OriginalArrivalTime: 04 Jan 2003 10:12:49.0602 (UTC) FILETIME=[D5CC8220:01C2B3D9]
X-MailScanner: Found to be clean
X-MailScanner-SpamCheck:
Reply-To: [EMAIL PROTECTED]
I have jabberd 1.4.2 on solaris 8(sparc), and it crashes often.
Here is one of backtraces:
...
This GDB was configured as "sparc-sun-solaris2.8"...
Core was generated by `jabberd/jabberd -D -B -h im.ftc.ru'.
Program terminated with signal 6, Abort.
...
(gdb) bt
#0 0xff01bed4 in _libc_kill () from /usr/lib/libc.so.1
#1 0xfefb57c4 in abort () from /usr/lib/libc.so.1
#2 0x25f78 in pmalloc (p=0x0, size=16) at pool.c:168
#3 0x2a6d4 in spool_new (p=0x0) at str.c:119
#4 0x239cc in jid_full (id=0x2e85b8) at jid.c:167
#5 0x20e60 in xdb_act (xc=0xb9580, owner=0x2e85b8,
ns=0xfef60c00 "jabber:iq:auth", act=0xfef60c10 "check", match=0x0,
data=0x563fd8) at xdb.c:277
#6 0xfef4e42c in mod_auth_plain_jane (m=0x1d0518, arg=0x0)
at mod_auth_plain.c:70
#7 0xfef47b14 in js_mapi_call (si=0xba080, e=5, packet=0x564eb0,
user=0x5ed998, s=0x0) at modules.c:160
#8 0xfef48184 in js_authreg (arg=0x564eb0) at authreg.c:70
#9 0x2130c in mtq_main (arg=0xe7f20) at mtq.c:151
#10 0xff3069c0 in pth_spawn_trampoline () from /usr/local/lib/libpth.so.14
in frame #8:
(gdb) p user->id->user
$2 = 0x0
Is that wrong?
In the end of debug log:
Sat Jan 4 10:23:40 2003 mio.c:620 mio while loop, working
Sat Jan 4 10:23:40 2003 mio.c:380 _mio_accept calling accept on fd #9
Sat Jan 4 10:23:40 2003 mio_ssl.c:238 SSL accepting socket with new session 3f5bc8
Sat Jan 4 10:23:40 2003 mio_ssl.c:251 Accept blocked, returning
Sat Jan 4 10:23:40 2003 mio.c:407 new socket accepted (fd: 53, ip: 172.16.254.163,
port: 1321)
Sat Jan 4 10:23:40 2003 mio.c:607 mio while loop top
Sat Jan 4 10:23:40 2003 mio.c:620 mio while loop, working
Sat Jan 4 10:23:40 2003 mio.c:607 mio while loop top
Sat Jan 4 10:23:40 2003 mio.c:620 mio while loop, working
Sat Jan 4 10:23:40 2003 mio.c:760 MIO read from socket 24: <xdb id='274'
type='error' to='sessions' from='xdb_auth_cpile' ns='jabber:iq:auth'/>
Sat Jan 4 10:23:40 2003 base_accept.c:120 process XML: m:24F878 state:3, arg:B1868,
x:976758
Sat Jan 4 10:23:40 2003 deliver.c:474 DELIVER 1:sessions <xdb id='274' type='error'
to='sessions' from='xdb_auth_cpile' ns='jabber:iq:auth'/>
Sat Jan 4 10:23:40 2003 deliver.c:678 delivering to instance 'sessions'
Sat Jan 4 10:23:40 2003 xdb.c:52 xdb_results checking xdb packet <xdb id='274'
type='error' to='sessions' from='xdb_auth_cpile' ns='jabber:iq:auth'/>
Memory Leak! [pmalloc received NULL pool, unable to track allocation, exiting]
Debug output is enabled, can not background.
New session for [EMAIL PROTECTED]/Рабо�а
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
New session for [EMAIL PROTECTED]/Just another jabber client
Sat Jan 4 10:23:41 2003 dnsrv.c:155 dnsrv: Read error on coprocess(1): 9 Bad file
number
Sat Jan 4 10:23:41 2003 dnsrv.c:169 DNSRV CHILD: out of loop.. exiting normal
Debug output is enabled, can not background.
We are dying from abort() in pmalloc, which receives 0 as arg.
I am using module xdb_auth_cpile module to autentificate users trough ldap, does that
matters?
xdb_auth_cpile is separate process, which sends and receives messages to/from jabberd
at the end of xdb_auth_cpile debug log:
SEND: <xdb id='274' type='error' to='sessions' from='xdb_auth_cpile'
ns='jabber:iq:auth'/>
startTag: creating new node for xdb
dispatching xdb
SEND: <xdb id='275' type='error' to='sessions' from='xdb_auth_cpile'
ns='jabber:iq:auth'/>
RECV: <xdb type='get' to='[EMAIL PROTECTED]' from='sessions' ns='jabber:iq:auth'
id='276'/><xdb type='get' to='[EMAIL PROTECTED]' from='sessions' ns='jabber:iq:auth'
id='277'/><xdb type='get' to='[EMAIL PROTECTED]' from='sessions' ns='jabber:iq:auth'
id='278'/><xdb type='get' to='[EMAIL PROTECTED]' from='sessions' ns='jabber:iq:auth'
id='279'/><xdb type='get' to='[EMAIL PROTECTED]' from='sessions' ns='jabber:iq:auth'
id='280'/>
startTag: creating new node for xdb
dispatching xdb
SEND: <xdb id='276' type='result' to='sessions' from='xdb_auth_cpile'
ns='jabber:iq:auth'><query xmlns='jabber:iq:auth'/></xdb>
send: Cannot determine peer address at
/usr/local/lib/perl5/site_perl/5.6.1/Jabber/Connection.pm line 402
i think, xdb_auth_cpile sees, that jabberd are dead and dies too.
WBW, Eugeny <az at ftc ru>
_______________________________________________
Jabberd mailing list
[EMAIL PROTECTED]
http://jabberstudio.org/cgi-bin/mailman/listinfo/jabberd
----- End forwarded message -----
WBW, Eugeny <az at ftc ru>
_______________________________________________
jdev mailing list
[EMAIL PROTECTED]
http://mailman.jabber.org/listinfo/jdev
