When a security vulnerability is found, the right thing to do is to prepare
a patch and circulate it as quickly as possible. At least, that would
appear to be the prevailing wisdom. This
ComputerWorld article, however, takes a different approach: in many
cases, patch circulation should be slowed down, not sped up.
The author is talking, in particular, about vulnerabilities which are found
by "white hat" hackers, as opposed to those which are already being
actively exploited. These vulnerabilities are, presumably, unknown to the
cracker community at the time the patch is prepared. But a security patch
provides an instant road map for anybody looking for vulnerabilities.
Rather than put in some honest work digging through and understanding a
large program, a cracker need only look at the piece of code which is
fixed. The release of a security patch allows administrators to close a
hole, but it also tells the world about the existence and location of that hole. At that
point, the race begins: administrators try to get the patch deployed before
the crackers get their exploits working.
What's needed is a way to give the defenders a larger window of time to
obtain patches before information about the vulnerability they fix is
distributed. Various approaches have been tried to accomplish that goal.
The "vendor-sec" mailing list, for example, helps Linux distributors and
other operating system vendors to all have their updates ready by the time
a vulnerability is announced. Vendor-sec helps, but it does not solve the
problem of actually distributing an update to millions of users.
The OpenSSH project once took a different approach and pushed a major update on users in an attempt to
deploy a security fix without saying what it was; this move was received
poorly, however.
What the ComputerWorld article suggests is that patches should be
distributed in encrypted form. For some period of time, the encrypted
patch is just a useless pile of bits sitting on the disk. This time would
be the window which allows the patch to be distributed without disclosing
the problem which is being fixed. After a given period of time, a key is
distributed which enables the decryption of the patch; at that time, clear
versions of the patch could also be made available. In theory, this
approach would enable the security-conscious users on the net to update
their systems nearly simultaneously as soon as the nature of the problem is
disclosed.
This is a solution which could perhaps work, though steps would have to be
taken to fend off denial-of-service attacks aimed at preventing the
distribution of the decryption key. The provision of encrypted patches
does go somewhat against the spirit of the free software community, and it
could, by some readings, be taken as a violation of the GPL. For almost
all of the security vulnerabilities which are reported, the encrypted patch
mechanism would be far more trouble than it would be worth. The next time
an easily-exploitable vulnerability turns up in a utility like bind or ssh,
however, it might be a nice option to have.
A memory leak has been discovered in mod_ssl that may be triggered by
sending normal HTTP requests to the Apache HTTPS port. An attacker can
exploit this vulnerability to consume all memory available in the server,
thus causing a denial of service condition. This problem has been fixed in
Apache 2.0.49.
Automake may be vulnerable to a symbolic link attack which may allow an
attacker to modify data or escalate their privileges. This is due to
the insecure way Automake creates directories during compilation. An
attacker may be able to create symbolic links in the place of files
contained in the affected directories, which may potentially lead to
elevated privileges due to modification of data.
The cvs client is vulnerable to a pathname vulnerability which can allow a hostile server to overwrite files on the local system. The cvs server is subject to a similar vulnerability which allows the checkout of RCS archives anywhere on the server system. Versions 1.11.15 and 1.12.7 fix the problem.
The 2.4 and 2.6 kernels contain a
vulnerability in the iso9660 (CDROM) filesystem which can be used by a
local attacker to obtain root privileges. The exploit requires creating a
specially-crafted filesystem and getting the kernel to mount it. Many
systems are configured to automatically mount CDs on insertion, however, so
the possibility of this vulnerability being exploited by users with
physical access to the system is real. The 2.4.26 kernel contains the fix,
which will also be merged into the upcoming 2.6.6 release.
The neon WebDAV library contains format string vulnerabilities which may be exploited by a hostile DAV server. This vulnerability exists in utilities which use neon, including cadaver and OpenOffice.org.
The server from the game Scorched 3D is vulnerable to a
format string attack that can lead to a denial of service and
possibly to the execution of arbitrary code.
"fam" (file alteration monitor) watches files and directories for changes and lets interested applications know when something happens. This package has a flaw in its group handling that blocks some legitimate operations while, at the same time, exposing the names of files that should otherwise be invisible.
Steve Kemp and Jaguar discovered a number of buffer overflow
vulnerabilities in vfte, a version of the fte editor which runs on the
Linux console, found in the package fte-console. This program is
setuid root in order to perform certain types of low-level operations
on the console.
GtkHTML is the HTML rendering widget used by the Evolution mail reader.
GtkHTML supplied with versions of Evolution prior to 1.2.4 contain a bug
when handling HTML messages. Alan Cox discovered that certain malformed
messages could cause the Evolution mail component to crash.
According to a security
advisory from the heimdal project: All releases prior to 0.6.1 and
0.5.3 have a cross-realm vulnerability allowing someone with control over a
realm to impersonate anyone in the cross-realm trust path.
A vulnerability was discovered recently in Interchange, an e-commerce
and general HTTP database display system. This vulnerability can be
exploited by an attacker to expose the content of arbitrary variables.
An attacker may learn SQL access information for your Interchange
application and use this information to read and manipulate sensitive
data.
Versions of ipsec-tools prior to 0.2.5 contain a vulnerability wherein the racoon utility fails to verify digital signatures on some packets. This hole can lead to unauthorized connections or man-in-the-middle attacks. See this advisory for details.
kdelibs (and, thus, Konqueror) has a vulnerability where a hostile server can force the disclosure of cookies that should not be presented to it. KDE versions 3.1.3 and later contain a fix.
KDE has issued a security advisory for all
versions of kdepim as distributed with KDE versions 3.1.0 through 3.1.4
inclusive. A carefully crafted .VCF file potentially enables local
attackers to compromise the privacy of a victim's data or execute arbitrary
commands with the victim's privileges. The Common Vulnerabilities and
Exposures project (cve.mitre.org) has assigned the name CAN-2003-0988 to
this issue.
The kernel-utils package contains several utilities that can be used to
control the kernel or machine hardware. In Red Hat Linux 8.0 this package
contains user mode linux (UML) utilities.
The uml_net utility in kernel-utils packages with Red Hat Linux 8.0 was
incorrectly shipped setuid root. This could allow local users to control
certain network interfaces, add and remove arp entries and routes, and put
interfaces in and out of promiscuous mode.
All users of the kernel-utils package should update to these packages that
contain a version of uml_net that is not setuid root.
Alternatively, as a work-around to this vulnerability issue the following
command as root:
Glenn Randers-Pehrson discovered a problem in connection with 16-bit
samples from libpng, an interface for reading and writing PNG
(Portable Network Graphics) format files. The starting offsets for
the loops are calculated incorrectly which causes a buffer overrun
beyond the beginning of the row buffer.
Yuuichi Teranishi discovered a flaw in libxml2 versions prior to 2.6.6.
When fetching a remote resource via FTP or HTTP, libxml2 uses special
parsing routines. These routines can overflow a buffer if passed a very
long URL. If an attacker is able to find an application using libxml2 that
parses remote resources and allows them to influence the URL, then this
flaw could be used to execute arbitrary code.
Matthew Galgoci of Red Hat discovered a Denial of Service (DoS)
vulnerability in versions of Mailman prior to 2.1. An attacker could send
a carefully-crafted message causing mailman to crash. The Common
Vulnerabilities and Exposures project (cve.mitre.org) has assigned the name
CAN-2003-0991 to this issue.
Apache's mod_python module could crash the httpd process if a specific,
malformed query string was sent.
The Apache Foundation has reported that mod_python may be prone to
Denial of Service attacks when handling a malformed query. Mod_python
2.7.9 was released to fix the vulnerability, however, because the
vulnerability has not been fully fixed, version 2.7.10 has been released.
Users of mod_python 3.0.4 are not affected by this vulnerability.
Mozilla 1.4 contains a few vulnerabilities, including disclosure of cookies to the wrong server, a scripting vulnerability which can allow an attacker to run arbitrary code, and an S/MIME vulnerability which can lead to remote denial of service or code execution attacks.
A vulnerability was discovered in mpg321, a command-line mp3 player,
whereby user-supplied strings were passed to printf(3) unsafely. This
vulnerability could be exploited by a remote attacker to overwrite
memory, and possibly execute arbitrary code. In order for this
vulnerability to be exploited, mpg321 would need to play a malicious
mp3 file (including via HTTP streaming).
Some some vulnerabilities exsist in the Nessus NASL scripting engine. To
exploit these flaws, an attacker would need to have a valid Nessus account
as well as the ability to upload arbitrary Nessus plugins in the Nessus
server (this option is disabled by default) or he/she would need to trick a
user somehow into running a specially crafted nasl script. Read the full
advisory for additional information.
netpbm is graphics conversion toolkit made up of a large number of
single-purpose programs. Many of these programs were found to create
temporary files in an insecure manner, which could allow a local
attacker to overwrite files with the privileges of the user invoking a
vulnerable netpbm tool.
From the advisory:
"During a pen-test we stumbled across a nasty bug in OpenSSH-portable
with PAM support enabled (via the --with-pam configure script switch). This
bug allows a remote attacker to identify valid users on vulnerable systems,
through a simple timing attack. The vulnerability is easy to exploit and
may have high severity, if combined with poor password policies and other
security problems that allow local privilege escalation."
Paul Szabo discovered a number of bugs in suidperl, a helper
program to run perl scripts with setuid privileges. By exploiting
these bugs, an attacker could abuse suidperl to discover information
about files (such as testing for their existence and some of their
permissions) that should not be accessible to unprivileged users.
The postfix MTA, versions through 1.1.12 (but not 2.0) is subject to two remotely exploitable denial of service vulnerabilities; see this advisory from Michal Zalewski for details.
PWLib is a cross-platform class library designed to support the OpenH323
project. OpenH323 provides an implementation of the ITU H.323
teleconferencing protocol, used by packages such as Gnome Meeting.
A test suite for the H.225 protocol (part of the H.323 family) provided by
the NISCC uncovered bugs in PWLib prior to version 1.6.0. An attacker
could trigger these bugs by sending carefully crafted messages to an
application. The effects of such an attack can vary depending on the
application, but would usually result in a Denial of Service. The Common
Vulnerabilities and Exposures project (cve.mitre.org) has assigned the name
CAN-2004-0097 to this issue.
Samba, a LanManager-like file and printer server for Unix, was found
to contain a vulnerability whereby a local user could use the "smbmnt"
utility, which is setuid root, to mount a file share from a remote
server which contained setuid programs under the control of the user.
These programs could then be executed to gain privileges on the local
system.
The shar utility (as found in the sharutils package through version 4.2.1) suffers from a stack-based buffer overflow vulnerability; see this advisory for details.
A bug was found in the processing of %-encoded characters in a URL in
versions of Squid 2.5.STABLE4 and earlier. If a Squid configuration uses
Access Control Lists (ACLs), a remote attacker could create URLs that would
not be correctly tested against Squid's ACLs, potentially allowing clients
to access prohibited URLs.
The tar utility does not properly filter file names containing
"../", meaning that a hostile archive can, if unpacked by an
unsuspecting user, overwrite any file that is writable by that user. GNU
tar versions 1.13.19 and earlier are vulnerable; unzip through version 5.42
has the same vulnerability.
George Bakos discovered flaws in the ISAKMP decoding routines of tcpdump
versions prior to 3.8.1. The Common Vulnerabilities and Exposures project
(cve.mitre.org) has assigned the name CAN-2003-0989 to this issue.
Jonathan Heusser discovered two additional flaws in the ISAKMP decoding
routines of tcpdump versions up to and including 3.8.1. The Common
Vulnerabilities and Exposures project (cve.mitre.org) has assigned the name
CAN-2004-0057 to this issue.
Jonathan Heusser discovered a flaw in the print_attr_string function in the
RADIUS decoding routines for tcpdump 3.8.1 and earlier. The Common
Vulnerabilities and Exposures project (cve.mitre.org) has assigned the name
CAN-2004-0055 to this issue.
Remote attackers could potentially exploit these issues by sending
carefully-crafted packets to a victim. If the victim uses tcpdump, these
packets could result in a denial of service, or possibly execute arbitrary
code as the 'pcap' user.
TCPDUMP v3.8.1 and earlier versions contain multiple flaws in the packet
display functions for the ISAKMP protocol. Upon receiving specially
crafted ISAKMP packets, TCPDUMP will try to read beyond the end of the
packet capture buffer and crash. More information is available in this Rapid7 advisory.
This vulnerability,
originally thought to be confined to BSD-derived systems, was first covered
in the July 26th Security
Summary. It is now known that Linux telnet daemons are vulnerable as
well.
Shaun Colley discovered a problem in xine-ui, the xine video player
user interface. A script contained in the package to possibly remedy
a problem or report a bug does not create temporary files in a secure
fashion. This could allow a local attacker to overwrite files with
the privileges of the user invoking xine.
James Grimmelmann reports
from the Digital Cops in a
Virtual Environment conference held last month. "One estimate of
MyDoom's effects puts its total damage at $38 billion. Oh, really? Well,
Hurricane Isabel did 'only' $4 billion. These oft-quoted estimates of virus
damage are, shall we say, perhaps overstated?"
