[IRCServices] 'recursive' crashing (Abort trap 6)

[Jille Timmermans]

Hello,

I am running ircservices-5.0.63 on FreeBSD 6; and (for the second time)
it crashed on an Abort trap, and kept on crashing, and flooded the log
up to the available disk-space.
IRC services is linked to UnrealIRCd 3.2.7; which in turn is linked to a
self-written server (bot).

Log:
Nov 11 16:01:03 2008] PANIC! signal 6, buffer = :USERX, :Ping timeout
[Nov 11 16:01:03 2008] Services terminating: Abort trap: 6
[Nov 11 16:01:03 2008] FATAL: Caught signal 6 (Abort trap: 6) while
shutting down
[Nov 11 16:01:03 2008] PANIC! signal 6, buffer = :USERX, :Ping timeout
[Nov 11 16:01:03 2008] Services terminating: Abort trap: 6
[Nov 11 16:01:03 2008] PANIC! signal 6, buffer = :USERX, :Ping timeout
[Nov 11 16:01:03 2008] Services terminating: Abort trap: 6
[Nov 11 16:01:03 2008] modules: BUG: trying to unuse module
`chanserv/main' from module `misc/xml-import' but caller not found in
user list!
[Nov 11 16:01:03 2008] PANIC! signal 6, buffer = :USERX, :Ping timeout
[Nov 11 16:01:03 2008] Services terminating: Abort trap: 6
[Nov 11 16:01:03 2008] PANIC! signal 6, buffer = :USERX, :Ping timeout
[Nov 11 16:01:03 2008] Services terminating: Abort trap: 6


And the last 2 lines repeat themselves untill the disk is full.

Earlier that day, a few of these messages occur:
[Nov 11 01:15:57 2008] user: do_part: no channel record for USERY on
#CHAN1 (bug?)
[Nov 11 02:45:19 2008] user: do_part: no channel record for USERY on
#CHAN2 (bug?)
[Nov 11 03:04:22 2008] user: do_part: no channel record for USERZ on
#CHAN1 (bug?)

I unfortunately don't have a coredump; neither I build with debugging;
neither I can reproduce it.
The time between this and the previous crash is quite long.

Anyone got a clue what this is ?
I am willing to help hunting this bug down and providing extra information.

-- Jille