Simon Josefsson reports:

_gnutls_handshake_log ("PKCS #1 padding error"); ret = GNUTLS_E_PKCS1_WRONG_PAD; Werner Koch points out that this error message may result in a vulnerability similar to Bleichenbacher's Crypto 98 attack. It is not exactly the same situation -- Bleichenbacher talks about PKCS#1 encryption (block type 1, uses random padding) where this deals with PKCS#1 verification (block type 2, uses 0xFF padding) -- but at a glance, it appears to have similar consequences, but differ in the number of messages required to mount the attack.

Secunia reports:

Mailman can be exploited by malicious people to conduct cross-site scripting and phishing attacks, and cause a DoS (Denial of Service).

1) An error in the logging functionality can be exploited to inject a spoofed log message into the error log via a specially crafted URL.

Successful exploitation may trick an administrator into visiting a malicious web site.

2) An error in the processing of malformed headers which does not follow the RFC 2231 standard can be exploited to cause a DoS (Denial of Service).

3) Some unspecified input isn't properly sanitised before being returned to the user. This can be exploited to execute arbitrary HTML and script code in a user's browser session in context of an affected site.

Kefka reports multiple cross site scripting vulnerabilities within hlstats. The vulnerabilities are caused due to improper checking of variables, allowing an attacker to perform cross site scripting.

The Debian Security Team reports:

Michael Gehring discovered several potential out-of-bounds index accesses in gtetrinet, a multiplayer Tetris-like game, which may allow a remote server to execute arbitrary code

The Joomla development team reports multiple vulnerabilities within the joomla application. Joomla is vulnerable to the following vulnerabilities:

• Improper validation of the mosMail function
• Improper validation of the JosIsValidEmail function.
• Remote code execution in PEAR.php
• Zend Hash del key or index vulnerability

Problem Description

While processing Link Control Protocol (LCP) configuration options received from the remote host, sppp(4) fails to correctly validate option lengths. This may result in data being read or written beyond the allocated kernel memory buffer.

Impact

An attacker able to send LCP packets, including the remote end of a sppp(4) connection, can cause the FreeBSD kernel to panic. Such an attacker may also be able to obtain sensitive information or gain elevated privileges.

Workaround

No workaround is available, but systems which do not use sppp(4) are not vulnerable.

Secunia reports:

Some vulnerabilities have been reported in Horde, which can be exploited by malicious people to conduct phishing and cross-site scripting attacks.

1. Input passed to the "url" parameter in index.php isn't properly verified before it is being used to include an arbitrary web site in a frameset. This can e.g. be exploited to trick a user into believing certain malicious content is served from a trusted web site.
2. Some unspecified input passed in index.php isn't properly sanitised before being returned to the user. This can be exploited to execute arbitrary HTML and script code in a user's browser session in context of an affected site.

The Globus Alliance reports:

The proxy generation tool (grid-proxy-init) creates the file, secures the file to provide access only to owner and writes proxy to the file. A race condition exists between the opening of the proxy credentials file, and making sure it is safe file to write to. The checks to ensure this file is accessible only to the owner take place using the filename after the file is opened for writing, but before any data is written.

Various components of the toolkit use files in shared directories to store information, some being sensitive information. For example, the tool to create proxy certificates, stores the generated proxy certificate by default in /tmp. Specific vulnerabilities in handling such files were reported in myproxy-admin-adduser, grid-ca-sign and grid-security-config.

Ludwig Nussel reports that x11vnc is vulnerable to an authentication bypass vulnerability. The vulnerability is caused by an error in auth.c. This could allow a remote attacker to gain unauthorized and unauthenticated access to the system.

Luigi Auriemma reports three vulnerabilities within alsaplayer:

• The function which handles the HTTP connections is vulnerable to a buffer-overflow that happens when it uses sscanf for copying the URL in the Location's field received from the server into the redirect buffer of only 1024 bytes declared in http_open.
• A buffer-overflow exists in the functions which add items to the playlist when the GTK interface is used (so the other interfaces are not affected by this problem): new_list_item and CbUpdated in interface/gtk/PlaylistWindow.cpp.
• AlsaPlayer automatically queries the CDDB server specified in its configuration (by default freedb.freedb.org) when the user choices the CDDA function for playing audio CDs. The function which queries the server uses a buffer of 20 bytes and one of 9 for storing the category and ID strings received from the server while the buffer which contains this server's response is 32768 bytes long. Naturally for exploiting this bug the attacker must have control of the freedb server specified in the AlsaPlayer's configuration.

These vulnerabilities could allow a remote attacker to execute arbitrary code, possibly gaining access to the system.

The PostgreSQL development team reports:

An attacker able to submit crafted strings to an application that will embed those strings in SQL commands can use invalidly-encoded multibyte characters to bypass standard string-escaping methods, resulting in possible injection of hostile SQL commands into the database. The attacks covered here work in any multibyte encoding.

The widely-used practice of escaping ASCII single quote "'" by turning it into "\'" is unsafe when operating in multibyte encodings that allow 0x5c (ASCII code for backslash) as the trailing byte of a multibyte character; this includes at least SJIS, BIG5, GBK, GB18030, and UHC. An application that uses this conversion while embedding untrusted strings in SQL commands is vulnerable to SQL-injection attacks if it communicates with the server in one of these encodings. While the standard client libraries used with PostgreSQL have escaped "'" in the safe, SQL-standard way of "''" for some time, the older practice remains common.

Multiple vulnerabilities had been reported in various versions of PostgreSQL:

• The EXECUTE restrictions can be bypassed by using the AGGREGATE function, which is missing a permissions check.
• A buffer overflow exists in gram.y which could allow an attacker to execute arbitrary code by sending a large number of arguments to a refcursor function, found in gram.y
• The intagg contributed module allows an attacker to crash the server (Denial of Service) by constructing a malicious crafted array.

Jean-David Maillefer reports a Denial of Service vulnerability within MySQL. The vulnerability is caused by improper checking of the data_format routine, which cause the MySQL server to crash. The crash is triggered by the following code:
"SELECT date_format('%d%s', 1);

The SquirrelMail developers report:

A logged in user could overwrite random variables in compose.php, which might make it possible to read/write other users' preferences or attachments.

The Ruby on Rails blog reports:

With Rails 1.1.0 through 1.1.5 (minus the short-lived 1.1.3), you can trigger the evaluation of Ruby code through the URL because of a bug in the routing code of Rails. This means that you can essentially take down a Rails process by starting something like /script/profiler, as the code will run for a long time and that process will be hung while it happens. Other URLs can even cause data loss.

Clamav team reports:

A heap overflow vulnerability was discovered in libclamav which could cause a denial of service or allow the execution of arbitrary code.

The problem is specifically located in the PE file rebuild function used by the UPX unpacker.

Relevant code from libclamav/upx.c:

memcpy(dst, newbuf, foffset); *dsize = foffset; free(newbuf); cli_dbgmsg("UPX: PE structure rebuilt from compressed file\n"); return 1;

Due to improper validation it is possible to overflow the above memcpy() beyond the allocated memory block.

The Drupal project reports:

A malicious user can execute a cross site scripting attack by enticing someone to visit a Drupal site via a specially crafted link.

Author reports:

Fixed 2 more possible memory allocation attacks. They are similar to the problem we fixed with 1.4.4. This bug can easily be be exploted for a DoS; remote code execution is not entirely impossible.

Secunia reports:

Two vulnerabilities have been reported in Ruby, which can be exploited by malicious people to bypass certain security restrictions.

1. An error in the handling of the "alias" functionality can be exploited to bypass the safe level protection and replace methods called in the trusted level.
2. An error caused due to directory operations not being properly checked can be exploited to bypass the safe level protection and close untainted directory streams.

The Apache Software Foundation and The Apache HTTP Server Project reports:

An off-by-one flaw exists in the Rewrite module, mod_rewrite, as shipped with Apache 1.3 since 1.3.28, 2.0 since 2.0.46, and 2.2 since 2.2.0.

Depending on the manner in which Apache HTTP Server was compiled, this software defect may result in a vulnerability which, in combination with certain types of Rewrite rules in the web server configuration files, could be triggered remotely. For vulnerable builds, the nature of the vulnerability can be denial of service (crashing of web server processes) or potentially allow arbitrary code execution. This issue has been rated as having important security impact by the Apache HTTP Server Security Team.

This flaw does not affect a default installation of Apache HTTP Server. Users who do not use, or have not enabled, the Rewrite module mod_rewrite are not affected by this issue. This issue only affects installations using a Rewrite rule with the following characteristics:

• The RewriteRule allows the attacker to control the initial part of the rewritten URL (for example if the substitution URL starts with $1)
• The RewriteRule flags do NOT include any of the following flags: Forbidden (F), Gone (G), or NoEscape (NE).

Please note that ability to exploit this issue is dependent on the stack layout for a particular compiled version of mod_rewrite. If the compiler used to compile Apache HTTP Server has added padding to the stack immediately after the buffer being overwritten, it will not be possible to exploit this issue, and Apache HTTP Server will continue operating normally.

The Apache HTTP Server project thanks Mark Dowd of McAfee Avert Labs for the responsible reporting of this vulnerability.

A Mozilla Foundation Security Advisory reports of multiple issues. Several of which can be used to run arbitrary code with the privilege of the user running the program.

• MFSA 2006-56 chrome: scheme loading remote content
• MFSA 2006-55 Crashes with evidence of memory corruption (rv:1.8.0.5)
• MFSA 2006-54 XSS with XPCNativeWrapper(window).Function(...)
• MFSA 2006-53 UniversalBrowserRead privilege escalation
• MFSA 2006-52 PAC privilege escalation using Function.prototype.call
• MFSA 2006-51 Privilege escalation using named-functions and redefined "new Object()"
• MFSA 2006-50 JavaScript engine vulnerabilities
• MFSA 2006-49 Heap buffer overwrite on malformed VCard
• MFSA 2006-48 JavaScript new Function race condition
• MFSA 2006-47 Native DOM methods can be hijacked across domains
• MFSA 2006-46 Memory corruption with simultaneous events
• MFSA 2006-45 Javascript navigator Object Vulnerability
• MFSA 2006-44 Code execution through deleted frame reference

Zope team reports:

Unspecified vulnerability in (Zope2) allows local users to obtain sensitive information via unknown attack vectors related to the docutils module and "restructured text".

The Drupal team reports:

Vulnerability: XSS Vulnerability in taxonomy module

It is possible for a malicious user to insert and execute XSS into terms, due to lack of validation on output of the page title. The fix wraps the display of terms in check_plain().

Goober's advisory reports reports that shoutcast is vulnerable to an arbitrary file reading vulnerability:

Impact of the vulnerability depends on the way the product was installed. In general, the vulnerability allows the attacker to read any file which can be read by the Shoutcast server process.

The Samba Team reports:

The smbd daemon maintains internal data structures used track active connections to file and printer shares. In certain circumstances an attacker may be able to continually increase the memory usage of an smbd process by issuing a large number of share connection requests. This defect affects all Samba configurations.

A TWiki Security Alert reports:

The TWiki upload filter already prevents executable scripts such as .php, .php1, .phps, .pl from potentially getting executed by appending a .txt suffix to the uploaded filename. However, PHP and some other types allows additional file suffixes, such as .php.en, .php.1, and .php.2. TWiki does not check for these suffixes, e.g. it is possible to upload php scripts with such suffixes without the .txt filename padding.

This issue can also be worked around with a restrictive web server configuration. See the TWiki Security Alert for more information about how to do this.

The Trac 0.9.6 Release Notes reports:

Fixed reStructuredText breach of privacy and denial of service vulnerability found by Felix Wiemann.

The discovered vulnerability requires docutils to be installed and enabled. Systems that do not have docutils installed or enabled are not vulnerable. As of this version version 0.3.9 or greater of docutils is required for using reStructuredText markup in Trac.

Horde 3.1.2 release announcement:

Security Fixes:

• Closed XSS problems in dereferrer (IE only), help viewer and problem reporting screen.
• Removed unused image proxy code from dereferrer.

The Team Mambo reports that two SQL injection vulnerabilities have been found in Mambo. The vulnerabilities exists due to missing sanitation of the title and catid parameters in the weblinks.php page and can lead to execution of arbitrary SQL code.

phpmyadmin Site reports:

It was possible to craft a request that contains XSS by attacking the "table" parameter.

The webmin development team reports:

An attacker without a login to Webmin can read the contents of any file on the server using a specially crafted URL. All users should upgrade to version 1.290 as soon as possible, or setup IP access control in Webmin.

SecurityFocus reports:

Mutt is prone to a remote buffer-overflow vulnerability. This issue is due to the application's failure to properly bounds-check user-supplied input before copying it to an insufficiently sized memory buffer. This issue may allow remote attackers to execute arbitrary machine code in the context of the affected application. Failed exploit attempts will likely crash the application, denying further service to legitimate users.

Joomla Site reports:

• Secured "Remember Me" functionality against SQL injection attacks
• Secured "Related Items" module against SQL injection attacks
• Secured "Weblinks" submission against SQL injection attacks
• Secured SEF from XSS vulnerability
• Hardened frontend submission forms against spoofing
• Secured mosmsg from misuse
• Hardened mosgetparam by setting variable type to integer if default value is detected as numeric
• Secured com_messages from XSS vulnerability
• Secured getUserStateFromRequest() from XSS vulnerability

Andreas Seltenreich reports that hashcash is prone to a heap overflow vulnerability. This vulnerability is caused by improper checking of memory allocations within the "array_push()" function. An attacker could trigger this vulnerability by passing a lot of "-r" or "-j" flags from the command line, this only applies when the application is configured to allow command line options, or by passing a lot of resource names when the application was started with the "-m" flag set. This could lead to a Denial or Service or could allow remote access to the targeted system.

If GnuPG processes a userid with a very long packet length, GnuPG can crash due to insufficient bounds check. This can result in a denial-of-service condition or potentially execution of arbitrary code with the privileges of the user running GnuPG.

FrSIRT advisory ADV-2006-2356 reports:

Multiple vulnerabilities have been identified in Horde Application Framework, which may be exploited by attackers to execute arbitrary scripting code. These flaws are due to input validation errors in the "test.php" and "templates/problem/problem.inc" scripts that do not validate the "url", "name", "email", "subject" and "message" parameters, which could be exploited by attackers to cause arbitrary scripting code to be executed by the user's browser in the security context of an affected Web site.

Secunia reports:

socsam has discovered a vulnerability in WebCalendar, which can be exploited by malicious people to bypass certain security restrictions and disclose sensitive information.

Input passed to the "includedir" parameter isn't properly verified, before it is used in an "fopen()" call. This can be exploited to load an arbitrary setting file from an external web site.

This can further be exploited to disclose the content of arbitrary files by defining the "user_inc" variable in a malicious setting file.

Successful exploitation requires that "register_globals" is enabled.

Problem Description

A suitably malformed multipart MIME message can cause sendmail to exceed predefined limits on its stack usage.

Impact

An attacker able to send mail to, or via, a server can cause queued messages on the system to not be delivered, by causing the sendmail process which handles queued messages to crash. Note that this will not stop new messages from entering the queue (either from local processes, or incoming via SMTP).

Workaround

No workaround is available, but systems which do not receive email from untrusted sources are not vulnerable.

Multiple vulnerabilities have been reported within dokuwiki. dokuwiki is proven vulnerable to:

• arbitrary PHP code insertion via spellcheck module,
• XSS attack via "Update your account profile,"
• bypassing of ACL controls when enabled.

A Secunia Advisory reports:

Federico L. Bossi Bonin has discovered a weakness in xine-lib, which can be exploited by malicious people to crash certain applications on a user's system.

The weakness is cause due to a heap corruption within the "xineplug_inp_http.so" plugin when handling an overly large reply from the HTTP server. This can be exploited to crash an application that uses the plugin (e.g. gxine).

Problem Description

smbfs does not properly sanitize paths containing a backslash character; in particular the directory name '..\' is interpreted as the parent directory by the SMB/CIFS server, but smbfs handles it in the same manner as any other directory.

Impact

When inside a chroot environment which resides on a smbfs mounted file-system it is possible for an attacker to escape out of this chroot to any other directory on the smbfs mounted file-system.

Workaround

Mount the smbfs file-systems which need to be used with chroot on top, in a way so the chroot directory is exactly on the mount point and not a sub directory

Problem Description

There are two documented methods of restricting access to NIS maps through ypserv(8): through the use of the /var/yp/securenets file, and through the /etc/hosts.allow file. While both mechanisms are implemented in the server, a change in the build process caused the "securenets" access restrictions to be inadvertantly disabled.

Impact

ypserv(8) will not load or process any of the networks or hosts specified in the /var/yp/securenets file, rendering those access controls ineffective.

Workaround

One possible workaround is to use /etc/hosts.allow for access control, as shown by examples in that file.

Another workaround is to use a firewall (e.g., ipfw(4), ipf(4), or pf(4)) to limit access to RPC functions from untrusted systems or networks, but due to the complexities of RPC, it might be difficult to create a set of firewall rules which accomplish this without blocking all access to the machine in question.

The freeradious development team reports:

Multiple issues exist with version 1.0.4, and all prior versions of the server. Externally exploitable vulnerabilities exist only for sites that use the rlm_sqlcounter module. Those sites may be vulnerable to SQL injection attacks, similar to the issues noted below. All sites that have not deployed the rlm_sqlcounter module are not vulnerable to external exploits.

The issues are:
SQL Injection attack in the rlm_sqlcounter module.
Buffer overflow in the rlm_sqlcounter module, that may cause a server crash.
Buffer overflow while expanding %t, that may cause a server crash.

The freeradius development team reports:

A validation issue exists with the EAP-MSCHAPv2 module in all versions from 1.0.0 (where the module first appeared) to 1.1.0. Insufficient input validation was being done in the EAP-MSCHAPv2 state machine. A malicious attacker could manipulate their EAP-MSCHAPv2 client state machine to potentially convince the server to bypass authentication checks. This bypassing could also result in the server crashing

The SquirrelMail Project Team reports:

A security issue has been uncovered in functions/plugin.php that could allow a remote user to access local files on the server without requiring login. This issue manifests itself if register_globals is enabled, and magic_quotes_gpc is disabled.

Stefan Esser reports:

During the evaluation of DokuWiki for a german/korean wiki of mine a flaw in DokuWiki's spellchecker was discovered, that allows injecting arbitrary PHP commands, by requesting a spellcheck on PHP commands in 'complex curly syntax'.

Because the spellchecker is written as part of the AJAX functionality of DokuWiki, it can be directly called by any website visitor, without the need for a wiki account.

The Drupal team reports:

Vulnerability: SQL injection

A security vulnerability in the database layer allowed certain queries to be submitted to the database without going through Drupal's query sanitizer.

Vulnerability: Execution of arbitrary files

Certain -- alas, typical -- configurations of Apache allows execution of carefully named arbitrary scripts in the files directory. Drupal now will attempt to automatically create a .htaccess file in your "files" directory to protect you.

MySQL reports:

An SQL-injection security hole has been found in multibyte encoding processing. An SQL-injection security hole can include a situation whereby when inserting user supplied data into a database, the user might inject his own SQL statements that the server will execute. With regards to this vulnerability discovered, when character set unaware escaping is used (e.g., addslashes() in PHP), it is possible to bypass it in some multibyte character sets (e.g., SJIS, BIG5 and GBK). As a result, a function like addslashes() is not able to prevent SQL injection attacks. It is impossible to fix this on the server side. The best solution is for applications to use character set aware escaping offered in a function like mysql_real_escape().

Workarounds:

One can use NO_BACKSLASH_ESCAPES mode as a workaround for a bug in mysql_real_escape_string(), if you cannot upgrade your server for some reason. It will enable SQL standard compatibility mode, where backslash is not considered a special character.

Secunia reports:

MySQL have some vulnerabilities, which can be exploited by malicious users to disclose potentially sensitive information and compromise a vulnerable system.

1) An error within the code that generates an error response to an invalid COM_TABLE_DUMP packet can be exploited by an authenticated client to disclosure certain memory content of the server process.

2) A boundary error within the handling of specially crafted invalid COM_TABLE_DUMP packets can be exploited by an authenticated client to cause a buffer overflow and allows arbitrary code execution.

3) An error within the handling of malformed login packets can be exploited to disclosure certain memory content of the server process in the error messages.

Esteban Martinez Fayo reports:

The FrontPage Server Extensions 2002 (included in Windows Sever 2003 IIS 6.0 and available as a separate download for Windows 2000 and XP) has a web page /_vti_bin/_vti_adm/fpadmdll.dll that is used for administrative purposes. This web page is vulnerable to cross site scripting attacks allowing an attacker to run client-side script on behalf of an FPSE user. If the victim is an administrator, the attacker could take complete control of a Front Page Server Extensions 2002 server.

To exploit the vulnerability an attacker can send a specially crafted e-mail message to a FPSE user and then persuade the user to click a link in the e-mail message.

In addition, this vulnerability can be exploited if an attacker hosts a malicious website and persuade the user to visit it.

Jason Duell reports:

Cscope contains an alarming number of buffer overflow vulnerabilities. By a rough count, there are at least 48 places where we blindly sprintf() a file name into a fixed-length buffer of size PATHLEN without checking to see if the file's name is <= PATHLEN. We do similar things with environment variable values.

Secunia reports:

Coppermine Photo Gallery have a vulnerability, which can be exploited by malicious users to compromise a vulnerable system.

The vulnerability is caused due to an error in the handling of file uploads where a filename has multiple file extensions. This can be exploited to upload malicious script files inside the web root (e.g. a PHP script).

Successful exploitation may allow execution of script code depending on the HTTP server configuration (it requires e.g. an Apache server with the "mod_mime" module installed).

Secunia reports:

Coppermine Photo Gallery have a vulnerability, which can be exploited by malicious people to disclose sensitive information.

Input passed to the "file" parameter in "index.php" isn't properly verified, before it is used to include files. This can be exploited to include arbitrary files from local resources.

Example: http://[host]/index.php?file=.//././/././/././/./[file]%00

Successful exploitation requires that "magic_quotes_gpc" is disabled.

Secunia reports:

Coppermine Photo Gallery have a vulnerability, which can be exploited by malicious people and by malicious users to compromise a vulnerable system.

1) Input passed to the "lang" parameter in include/init.inc.php isn't properly verified, before it is used to include files. This can be exploited to include arbitrary files from local resources. The vulnerability can be further exploited by users who are allowed to upload image files to execute arbitrary PHP code.

2) Input passed to the "f" parameter in docs/showdoc.php isn't properly verified, before it is used to include files. This can be exploited to include arbitrary files from local resources on the Windows platform, and remote files from Windows shared folders.

phpMyAdmin security team reports:

It was possible to inject arbitrary SQL commands by forcing an authenticated user to follow a crafted link.

Such issue is quite common in many PHP applications and users should take care what links they follow. We consider these vulnerabilities to be quite dangerous.

RealVNC is susceptible to an authentication-bypass vulnerability. A malicious VNC client can cause a VNC server to allow it to connect without any authentication regardless of the authentication settings configured in the server. Exploiting this issue allows attackers to gain unauthenticated, remote access to the VNC servers.

Secunia reports:

phpLDAPadmin have some vulnerabilities, which can be exploited by malicious users to conduct script insertion attacks and by malicious people to conduct cross-site scripting attacks.

1) Some input isn't properly sanitised before being returned to the user. This can be exploited to execute arbitrary HTML and script code in a user's browser session in context of an affected site.

2) Input passed to the "Container DN", "Machine Name", and "UID Number" parameters in "template_engine.php" isn't properly sanitised before being used. This can be exploited to inject arbitrary HTML and script code, which will be executed in a user's browser session in context of an affected site when the malicious user data is viewed.

JVN reports:

FreeStyleWiki has XSS vulnerability.

Stefano Di Paola reports:

An authenticated user could remotely execute arbitrary commands by taking advantage of a stack overflow.

To take advantage of these flaws an attacker should have direct access to MySQL server communication layer (port 3306 or unix socket). But if used in conjuction with some web application flaws (i.e. php code injection) an attacker could use socket programming (i.e. php sockets) to gain access to that layer.

OS Reviews reports:

If the update of the stats via web front-end is allowed, a remote attacker can execute arbitrary code on the server using a specially crafted request involving the migrate parameter. Input starting with a pipe character ("|") leads to an insecure call to Perl's open function and the rest of the input being executed in a shell. The code is run in the context of the process running the AWStats CGI.

Arbitrary code can be executed by uploading a specially crafted configuration file if an attacker can put a file on the server with chosen file name and content (e.g. by using an FTP account on a shared hosting server). In this configuration file, the LogFile directive can be used to execute shell code following a pipe character. As above, an open call on unsanitized input is the source of this vulnerability.

Secunia reports:

phpWebFTP have a vulnerability, which can be exploited by malicious people to disclose sensitive information.

Input passed to to the "language" parameter in index.php isn't properly verified, before it is used to include files. This can be exploited to include arbitrary files from local resources.

Successful exploitation requires that "magic_quotes_gpc" is disabled.

A Mozilla Foundation Security Advisory reports for deleted object reference when designMode="on"

Martijn Wargers and Nick Mott each described crashes that were discovered to ultimately stem from the same root cause: attempting to use a deleted controller context when designMode was turned on. This generally results in crashing the browser, but in theory references to deleted objects can be abused to run malicious code.

"splices" reported the same crash at the fan site MozillaZine and on Bugtraq, incorrectly describing it as a buffer overflow.

Secunia reports:

A vulnerability has been reported, which can be exploited by malicious people to conduct script insertion attacks.

Input passed using the wiki macro isn't properly sanitised before being used. This can be exploited to inject arbitrary HTML and script code, which will be executed in a user's browser session in context of an affected site when the malicious user data is viewed.

Secunia reports:

A vulnerability has been reported in ClamAV, which can be exploited by malicious people to cause a DoS (Denial of Service) and potentially to compromise a vulnerable system.

The vulnerability is caused due to a boundary error within the HTTP client in the Freshclam command line utility. This can be exploited to cause a stack-based buffer overflow when the HTTP headers received from a web server exceeds 8KB.

Successful exploitation requires that Freshclam is used to download virus signature updates from a malicious mirror web server e.g. via DNS poisoning.

Secunia reports:

A vulnerability has been reported in jabberd, which can be exploited by malicious people to cause a DoS (Denial of Service).

The vulnerability is caused due to an error within the handling of SASL negotiation. This can be exploited to cause a crash by sending a "response" stanza before an "auth" stanza.

Secunia reports:

Cacti have a security issue, which can be exploited by malicious people to execute arbitrary SQL code and potentially compromise a vulnerable system.

The problem is caused due to the presence of the insecure "server.php" test script.

Secunia reports:

Amaya have two vulnerabilities, which can be exploited by malicious people to compromise a user's system.

The vulnerabilities are caused due to boundary errors within the parsing of various attribute values. This can be exploited to cause stack-based buffer overflows when a user opens a specially crafted HTML document containing certain tags with overly long attribute values.

Successful exploitation allows execution of arbitrary code.

Secunia reports:

A security issue has been discovered in LifeType, which can be exploited by malicious people to execute arbitrary SQL code and potentially compromise a vulnerable system.

The problem is caused due to the presence of the insecure "server.php" test script.

Secunia reports:

Multiple vulnerabilities have been reported in Ethereal, which can be exploited by malicious people to cause a DoS (Denial of Service) or compromise a vulnerable system.

The vulnerabilities are caused due to various types of errors including boundary errors, an off-by-one error, an infinite loop error, and several unspecified errors in a multitude of protocol dissectors.

Successful exploitation causes Ethereal to stop responding, consume a large amount of system resources, crash, or execute arbitrary code.

Emmanouel Kellenis reports a denial of service vulnerability within asterisk. The vulnerability is caused by a buffer overflow in "format_jpeg.c". A large JPEG image could trigger this bug, potentially allowing a local attacker to execute arbitrary code.

Gentoo reports:

Andrea Barisani of Gentoo Linux discovered xzgv and zgv allocate insufficient memory when rendering images with more than 3 output components, such as images using the YCCK or CMYK colour space. When xzgv or zgv attempt to render the image, data from the image overruns a heap allocated buffer.

An attacker may be able to construct a malicious image that executes arbitrary code with the permissions of the xzgv or zgv user when attempting to render the image.

FRSIRT reports:

A vulnerability has been identified in CrossFire, which could be exploited by remote attackers to execute arbitrary commands or cause a denial of service. This flaw is due to a buffer overflow error in the "oldsocketmode" module that fails to properly handle overly large requests, which could be exploited by a malicious client to crash or compromise a vulnerable system.

Javier Fernández-Sanguino Peña reports:

The DBI library, the Perl5 database interface, creates a temporary PID file in an insecure manner. This can be exploited by a malicious user to overwrite arbitrary files owned by the person executing the parts of the library.

Dedi Dwianto reports:

A remote user can access the file directly to cause the system to display an error message that indicates the installation path. The resulting error message will disclose potentially sensitive installation path information to the remote attacker.

c0ntexb reports:

There are 2 format string bugs in the latest version of Xine that could be exploited by a malicious person to execute code on the system of a remote user running the media player against a malicious playlist file. By passing a format specifier in the path of a file that is embedded in a remote playlist, it is possible to trigger this bug.

Unspecified vulnerability in the CMU Cyrus Simple Authentication and Security Layer (SASL) library, has unknown impact and remote unauthenticated attack vectors, related to DIGEST-MD5 negotiation.

Problem Description

On "7th generation" and "8th generation" processors manufactured by AMD, including the AMD Athlon, Duron, Athlon MP, Athlon XP, Athlon64, Athlon64 FX, Opteron, Turion, and Sempron, the fxsave and fxrstor instructions do not save and restore the FOP, FIP, and FDP registers unless the exception summary bit (ES) in the x87 status word is set to 1, indicating that an unmasked x87 exception has occurred.

This behaviour is consistent with documentation provided by AMD, but is different from processors from other vendors, which save and restore the FOP, FIP, and FDP registers regardless of the value of the ES bit. As a result of this discrepancy remaining unnoticed until now, the FreeBSD kernel does not restore the contents of the FOP, FIP, and FDP registers between context switches.

Impact

On affected processors, a local attacker can monitor the execution path of a process which uses floating-point operations. This may allow an attacker to steal cryptographic keys or other sensitive information.

Workaround

No workaround is available, but systems which do not use AMD Athlon, Duron, Athlon MP, Athlon XP, Athlon64, Athlon64 FX, Opteron, Turion, or Sempron processors are not vulnerable.

Secunia reports:

The vulnerability is caused due to missing security declarations in "changeMemberPortrait" and "deletePersonalPortrait". This can be exploited to manipulate or delete another user's portrait via the "member_id" parameter.

A Mozilla Foundation Security Advisory reports of multiple issues. Several of which can be used to run arbitrary code with the privilege of the user running the program.

• MFSA 2006-29 Spoofing with translucent windows
• MFSA 2006-28 Security check of js_ValueToFunctionObject() can be circumvented
• MFSA 2006-26 Mail Multiple Information Disclosure
• MFSA 2006-25 Privilege escalation through Print Preview
• MFSA 2006-24 Privilege escalation using crypto.generateCRMFRequest
• MFSA 2006-23 File stealing by changing input type
• MFSA 2006-22 CSS Letter-Spacing Heap Overflow Vulnerability
• MFSA 2006-20 Crashes with evidence of memory corruption (rv:1.8.0.2)
• MFSA 2006-19 Cross-site scripting using .valueOf.call()
• MFSA 2006-18 Mozilla Firefox Tag Order Vulnerability
• MFSA 2006-17 cross-site scripting through window.controllers
• MFSA 2006-16 Accessing XBL compilation scope via valueOf.call()
• MFSA 2006-15 Privilege escalation using a JavaScript function's cloned parent
• MFSA 2006-14 Privilege escalation via XBL.method.eval
• MFSA 2006-13 Downloading executables with "Save Image As..."
• MFSA 2006-12 Secure-site spoof (requires security warning dialog)
• MFSA 2006-11 Crashes with evidence of memory corruption (rv:1.8)
• MFSA 2006-10 JavaScript garbage-collection hazard audit
• MFSA 2006-09 Cross-site JavaScript injection using event handlers

Secunia reports:

A vulnerability has been reported in Mailman, which can be exploited by malicious people to conduct cross-site scripting attacks.

Unspecified input passed to the private archive script is not properly sanitised before being returned to users. This can be exploited to execute arbitrary HTML and script code in a user's browser session in context of a vulnerable site.

Javier Fernandez-Sanguino Pena reports two temporary file vulnerability within f2c. The vulnerabilities are caused due to weak temporary file handling. An attacker could create an symbolic link, causing a local user running f2c to overwrite the symlinked file. This could give the attacker elevated privileges.

Secunia reports:

The vulnerabilities are caused due to integer overflow errors in "libmpdemux/asfheader.c" within the handling of an ASF file, and in "libmpdemux/aviheader.c" when parsing the "indx" chunk in an AVI file. This can be exploited to cause heap-based buffer overflows via a malicious ASF file, or via a AVI file with specially-crafted "wLongsPerEntry" and "nEntriesInUse" values in the "indx" chunk.

The KDE team reports:

Kaffeine can produce a buffer overflow in http_peek() while creating HTTP request headers for fetching remote playlists, which under certain circumstances could be used to crash the application and/or execute arbitrary code.

Renaud Lifchitz reports a vulnerability within thunderbird. The vulnerability is caused by improper checking of javascript scripts. This could lead to javascript code execution which can lead to information disclosure or a denial of service (application crash). This vulnerability is present even if javascript had been disabled in the preferences.

phpMyAdmin security announcement:

It was possible to conduct an XSS attack with a direct call to some scripts under the themes directory.

Secunia reports:

A vulnerability has been reported in phpMyAdmin, which can be exploited by malicious people to conduct cross-site scripting attacks.

Input passed to the "set_theme" parameter isn't properly sanitised before being returned to the user. This can be exploited to execute arbitrary HTML and script code in a user's browser session in context of an affected site.

Secunia reports:

Some vulnerabilities have been reported in ClamAV, which potentially can be exploited by malicious people to cause a DoS (Denial of Service) and compromise a vulnerable system.

An unspecified integer overflow error exists in the PE header parser in "libclamav/pe.c". Successful exploitation requires that the ArchiveMaxFileSize option is disabled.

Some format string errors in the logging handling in "shared/output.c" may be exploited to execute arbitrary code.

An out-of-bounds memory access error in the "cli_bitset_test()" function in "ibclamav/others.c" may be exploited to cause a crash.

The mediawiki development team reports a vulnerability within the mediawiki application. The vulnerability is caused by improper checking of inline style attributes. This could result in the execution of arbitrary javascript code in Microsoft Internet Explorer. It appears that other browsers are not affected by this vulnerability.

Ubuntu reports:

A buffer overflow was found in the "pnmtopng" conversion program. By tricking an user (or automated system) to process a specially crafted PNM image with pnmtopng, this could be exploited to execute arbitrary code with the privileges of the user running pnmtopng.

Jean-Sébastien Guay-Leroux report a vulnerability within the zoo archiver. The vulnerability which is present in the fullpath() function (from the misc.c file) is caused by improper checking of user supplied data. The data returned to the buffer can be up to 512 bytes, while the buffer is created to hold 256 bytes. This could result in a buffer overflow which could allow remote code execution.

The mediawiki development team reports that there is an site scripting vulnerability within mediawiki. The vulnerability is caused by improper checking of encoded links which could allow the injection of html in the output generated by mediawiki. This could lead to cross site scripting attacks against mediawiki installations.

Secunia reports:

Some vulnerabilities have been reported in Dia, which potentially can be exploited by malicious people to compromise a user's system.

The vulnerabilities are caused due to boundary errors within the XFig import plugin. This can be exploited to cause buffer overflows and may allow arbitrary code execution when a specially-crafted FIG file is imported.

Hendrik Weimer reports:

OpenVPN clients are a bit too generous when accepting configuration options from a server. It is possible to transmit environment variables to client-side shell scripts. There are some filters in place to prevent obvious nonsense, however they don't catch the good old LD_PRELOAD trick. All we need is to put a file onto the client under a known location (e.g. by returning a specially crafted document upon web access) and we have a remote root exploit. But since the attack may only come from authenticated servers, this threat is greatly reduced.

Samba Security Advisory:

The machine trust account password is the secret shared between a domain controller and a specific member server. Access to the member server machine credentials allows an attacker to impersonate the server in the domain and gain access to additional information regarding domain users and groups.

The winbindd daemon writes the clear text of server's machine credentials to its log file at level 5. The winbindd log files are world readable by default and often log files are requested on open mailing lists as tools used to debug server misconfigurations.

This affects servers configured to use domain or ads security and possibly Samba domain controllers as well (if configured to use winbindd).

Nathan Dors of the Pubcookie Project reports:

Non-persistent XSS vulnerabilities were found in the Pubcookie Apache module (mod_pubcookie) and ISAPI filter. These components mishandle untrusted data when printing responses to the browser. This makes them vulnerable to carefully crafted requests containing script or HTML. If an attacker can lure an unsuspecting user to visit carefully staged content, the attacker can use it to redirect the user to a vulnerable Pubcookie application server and attempt to exploit the XSS vulnerabilities.

These vulnerabilities are classified as *high* due to the nature and purpose of Pubcookie application servers for user authentication and Web Single Sign-on (SSO). An attacker who injects malicious script through the vulnerabilities might steal private Pubcookie data including a user's authentication assertion ("granting") cookies and application session cookies.

Nathan Dors of the Pubcookie Project reports:

Multiple non-persistent XSS vulnerabilities were found in the Pubcookie login server's compiled binary "index.cgi" CGI program. The CGI program mishandles untrusted data when printing responses to the browser. This makes the program vulnerable to carefully crafted requests containing script or HTML. If an attacker can lure an unsuspecting user to visit carefully staged content, the attacker can use it to redirect the user to his or her local Pubcookie login page and attempt to exploit the XSS vulnerabilities.

These vulnerabilities are classified as *critical* due to the nature and purpose of the Pubcookie login server for user authentication and Web Single Sign-on (SSO). Specific threats include:

• An attacker who injects malicious script through the vulnerabilities might steal senstive user data including a user's authentication credentials (usernames and passwords);
• An attacker who injects malicious script through the vulnerabilities might steal private Pubcookie data including a user's authentication assertion ("granting") cookies and SSO ("login") session cookies;
• An attacker who injects HTML tags through the vulnerabilities might deface a site's Pubcookie login page for a single visit by a single user (i.e. a non-persistent defacement).

At the heart of these threats lies a violation of the user's trust in the Pubcookie login server.

Freeradius Security Contact reports:

Insufficient input validation was being done in the EAP-MSCHAPv2 state machine. A malicious attacker could manipulate their EAP-MSCHAPv2 client state machine to potentially convince the server to bypass authentication checks. This bypassing could also result in the server crashing.

Horde 3.1.1 release announcement:

Major changes compared to Horde 3.1 are:

• Fix for remote code execution vulnerability in the help viewer, discovered by Jan Schneider from the Horde team.

Secunia Advisories Reports:

A boundary error when processing SWF files can be exploited to cause a buffer overflow. This may allow execution of arbitrary code on the user's system.

iDefense Reports:

Remote exploitation of a heap-based buffer overflow in RealNetwork Inc's RealPlayer could allow the execution of arbitrary code in the context of the currently logged in user.

In order to exploit this vulnerability, an attacker would need to entice a user to follow a link to a malicious server. Once the user visits a website under the control of an attacker, it is possible in a default install of RealPlayer to force a web-browser to use RealPlayer to connect to an arbitrary server, even when it is not the default application for handling those types, by the use of embedded object tags in a webpage. This may allow automated exploitation when the page is viewed.

Problem Description

A race condition has been reported to exist in the handling by sendmail of asynchronous signals.

Impact

A remote attacker may be able to execute arbitrary code with the privileges of the user running sendmail, typically root.

Workaround

There is no known workaround other than disabling sendmail.

Problem Description

The opiepasswd(1) program uses getlogin(2) to identify the user calling opiepasswd(1). In some circumstances getlogin(2) will return "root" even when running as an unprivileged user. This causes opiepasswd(1) to allow an unpriviled user to configure OPIE authentication for the root user.

Impact

In certain cases an attacker able to run commands as a non privileged users which have not explicitly logged in, for example CGI scripts run by a web server, is able to configure OPIE access for the root user. If the attacker is able to authenticate as root using OPIE authentication, for example if "PermitRootLogin" is set to "yes" in sshd_config or the attacker has access to a local user in the "wheel" group, the attacker can gain root privileges.

Workaround

Disable OPIE authentication in PAM:

# sed -i "" -e /opie/s/^/#/ /etc/pam.d/*

or

Remove the setuid bit from opiepasswd:

# chflags noschg /usr/bin/opiepasswd

# chmod 555 /usr/bin/opiepasswd

# chflags schg /usr/bin/opiepasswd

Problem Description

IPsec provides an anti-replay service which when enabled prevents an attacker from successfully executing a replay attack. This is done through the verification of sequence numbers. A programming error in the fast_ipsec(4) implementation results in the sequence number associated with a Security Association not being updated, allowing packets to unconditionally pass sequence number verification checks.

Impact

An attacker able to to intercept IPSec packets can replay them. If higher level protocols which do not provide any protection against packet replays (e.g., UDP) are used, this may have a variety of effects.

Workaround

No workaround is available.

Daniel Stone of X.Org reports:

During the analysis of results from the Coverity code review of X.Org, we discovered a flaw in the server that allows local users to execute arbitrary code with root privileges, or cause a denial of service by overwriting files on the system, again with root privileges.

A Project heimdal Security Advisory reports:

The telnet client program in Heimdal has buffer overflows in the functions slc_add_reply() and env_opt_add(), which may lead to remote code execution.

The telnetd server program in Heimdal has buffer overflows in the function getterminaltype, which may lead to remote code execution.

The rshd server in Heimdal has a privilege escalation bug when storing forwarded credentials. The code allowes a user to overwrite a file with its credential cache, and get ownership of the file.

A Project cURL Security Advisory reports:

libcurl uses the given file part of a TFTP URL in a manner that allows a malicious user to overflow a heap-based memory buffer due to the lack of boundary check.

This overflow happens if you pass in a URL with a TFTP protocol prefix ("tftp://"), using a valid host and a path part that is longer than 512 bytes.

The affected flaw can be triggered by a redirect, if curl/libcurl is told to follow redirects and an HTTP server points the client to a tftp URL with the characteristics described above.

Drupal reports:

Mail header injection vulnerability.

Linefeeds and carriage returns were not being stripped from email headers, raising the possibility of bogus headers being inserted into outgoing email.

This could lead to Drupal sites being used to send unwanted email.

Session fixation vulnerability.

If someone creates a clever enough URL and convinces you to click on it, and you later log in but you do not log off then the attacker may be able to impersonate you.

XSS vulnerabilities.

Some user input sanity checking was missing. This could lead to possible cross-site scripting (XSS) attacks.

XSS can lead to user tracking and theft of accounts and services.

Security bypass in menu.module.

If you use menu.module to create a menu item, the page you point to will be accessible to all, even if it is an admin page.

Secunia advisory SA19246:

Paul Craig has discovered a vulnerability in Horde, which can be exploited by malicious people to disclose sensitive information. Input passed to the "url" parameter in "services/go.php" isn't properly verified, before it is used in a "readfile()" call. This can be exploited to disclose the content of arbitrary files via e.g. the "php://" protocol wrapper.

The vulnerability has been confirmed in version 3.0.9 and has also been reported in prior versions.

Provided and/or discovered by: Paul Craig, Security-Assessment.com.

Adobe reports:

Critical vulnerabilities have been identified in Flash Player that could allow an attacker who successfully exploits these vulnerabilities to take control of the affected system. A malicious SWF must be loaded in Flash Player by the user for an attacker to exploit these vulnerabilities.

Flash Player 8 update (8.0.24.0), and Flash Player 7 update (7.0.63.0) address security vulnerabilities in previous versions of Flash Player, which could lead to the potential execution of arbitrary code. These vulnerabilities could be accessed through content delivered from a remote location via the users web browser, email client, or other applications that include or reference the Flash Player.

Problem description:

A part of the NFS server code charged with handling incoming RPC messages via TCP had an error which, when the server received a message with a zero-length payload, would cause a NULL pointer dereference which results in a kernel panic. The kernel will only process the RPC messages if a userland nfsd daemon is running.

Impact:

The NULL pointer deference allows a remote attacker capable of sending RPC messages to an affected FreeBSD system to crash the FreeBSD system.

Workaround:

1. Disable the NFS server: set the nfs_server_enable variable to "NO" in /etc/rc.conf, and reboot.

   Alternatively, if there are no active NFS clients (as listed by the showmount(8) utility), simply killing the mountd and nfsd processes should suffice.

2. Add firewall rules to block RPC traffic to the NFS server from untrusted hosts.

Problem description:

Because OpenSSH and OpenPAM have conflicting designs (one is event- driven while the other is callback-driven), it is necessary for OpenSSH to fork a child process to handle calls to the PAM framework. However, if the unprivileged child terminates while PAM authentication is under way, the parent process incorrectly believes that the PAM child also terminated. The parent process then terminates, and the PAM child is left behind.

Due to the way OpenSSH performs internal accounting, these orphaned PAM children are counted as pending connections by the master OpenSSH server process. Once a certain number of orphans has accumulated, the master decides that it is overloaded and stops accepting client connections.

Impact:

By repeatedly connecting to a vulnerable server, waiting for a password prompt, and closing the connection, an attacker can cause OpenSSH to stop accepting client connections until the system restarts or an administrator manually kills the orphaned PAM processes.

Workaround:

The following command will show a list of orphaned PAM processes:

# pgrep -lf 'sshd.*\[pam\]'

The following command will kill orphaned PAM processes:

# pkill -f 'sshd.*\[pam\]'

To prevent OpenSSH from leaving orphaned PAM processes behind, perform one of the following:

1. Disable PAM authentication in OpenSSH. Users will still be able to log in using their Unix password, OPIE or SSH keys.

   To do this, execute the following commands as root:

   # echo 'UsePAM no' >>/etc/ssh/sshd_config

   # echo 'PasswordAuthentication yes' >>/etc/ssh/sshd_config

   # /etc/rc.d/sshd restart

2. If disabling PAM is not an option - if, for instance, you use RADIUS authentication, or store user passwords in an SQL database - you may instead disable privilege separation. However, this may leave OpenSSH vulnerable to hitherto unknown bugs, and should be considered a last resort.

   To do this, execute the following commands as root:

   # echo 'UsePrivilegeSeparation no' >>/etc/ssh/sshd_config

   # /etc/rc.d/sshd restart

Werner Koch reports:

In the aftermath of the false positive signature verfication bug (announced 2006-02-15) more thorough testing of the fix has been done and another vulnerability has been detected. This new problem affects the use of *gpg* for verification of signatures which are _not_ detached signatures. The problem also affects verification of signatures embedded in encrypted messages; i.e. standard use of gpg for mails.

The Mplayer team reports:

A potential buffer overflow was found in the ASF demuxer. Arbitrary remote code execution is possible (under the user ID running the player) when streaming an ASF file from a malicious server or local code execution (under the user ID running the player) if a malicious ASF file is played locally.

SSH Communications Security Corp reports a format string vulnerability in their SFTP server. This vulnerability could cause a user with SCP/SFTP access only to get permission to execute also other commands. It could also allow user A to create a special file that when accessed by user B allows user A to execute commands as user B.

GNU tar is vulnerable to a buffer overflow, caused by improper bounds checking of the PAX extended headers. By tricking an user into processing a specially crafted tar archive, this could be exploited to execute arbitrary code with the privileges of the user.

Some vulnerabilities have been reported in Bugzilla, which can be exploited by malicious users to conduct SQL injection attacks, and by malicious people to disclose sensitive information and conduct script insertion attacks.

Multiple vulnerabilities has been discovered since 1.4.5, including IMAP injection as well as some XSS issues.

Yan Feng reports a format string vulnerability in gedit. This vulnerability could cause a denial of service with a binary file that contains format string characters within the filename. It had been reported that web browsers and email clients can be configured to provide a filename as an argument to gedit.:

SecurityFocus reports that WebCalendar is affected by an unauthorized access vulnerability. The vulnerability is caused by improper checking of the authentication mechanism before access is being permitted to the "assistant_edit.php" file.

Chris Evans reports that AbiWord is vulnerable to multiple stack-based buffer overflow vulnerabilities. This is caused by improper checking of the user-supplied data before it is being copied to an too small buffer. The vulnerability is triggered when someone is importing RTF files.

The PostgreSQL team reports:

Due to inadequate validity checking, a user could exploit the special case that SET ROLE normally uses to restore the previous role setting after an error. This allowed ordinary users to acquire superuser status, for example.

Werner Koch reports:

The Gentoo project identified a security related bug in GnuPG. When using any current version of GnuPG for unattended signature verification (e.g. by scripts and mail programs), false positive signature verification of detached signatures may occur.

This problem affects the tool *gpgv*, as well as using "gpg --verify" to imitate gpgv, if only the exit code of the process is used to decide whether a detached signature is valid. This is a plausible mode of operation for gpgv.

If, as suggested, the --status-fd generated output is used to decide whether a signature is valid, no problem exists. In particular applications making use of the GPGME library[2] are not affected.

Pizzashack reports:

Max Vozeler has reported a problem whereby rssh can allow users who have shell access to systems where rssh is installed (and rssh_chroot_helper is installed SUID) to gain root access to the system, due to the ability to chroot to arbitrary locations. There are a lot of potentially mitigating factors, but to be safe you should upgrade immediately.

Roger Dingledine reports:

If you offer a Tor hidden service, an adversary who can run a fast Tor server and who knows some basic statistics can find the location of your hidden service in a matter of minutes to hours.

Tavis Ormandy reports:

The bash shell uses the value of the PS4 environment variable (after expansion) as a prefix for commands run in execution trace mode. Execution trace mode (xtrace) is normally set via bash's -x command line option or interactively by running "set -o xtrace". However, it may also be enabled by placing the string "xtrace" in the SHELLOPTS environment variable before bash is started.

A malicious user with sudo access to a shell script that uses bash can use this feature to run arbitrary commands for each line of the script.

The Secure Science Corporation reports that libtomcrypt is vulnerable to a weak signature scheme. This allows an attacker to create a valid random signature and use that to sign arbitrary messages without requiring the private key.

r0t reports:

Mantis contains a flaw that allows a remote cross site scripting attack. This flaw exists because input passed to "target_field" parameter in "view_filters_page.php" is not properly sanitised before being returned to the user. This could allow a user to create a specially crafted URL that would execute arbitrary code in a user's browser within the trust relationship between the browser and the server, leading to a loss of integrity.

Multiple vulnerabilities have been reported within phpbb. phpbb is proven vulnerable to:

• script insertion,
• bypassing of protetion mechanisms,
• multiple cross site scripting vulnerabilities,
• SQL injection,
• arbitrary code execution

The postgresql development team reports:

The more severe of the two errors is that the functions that support client-to-server character set conversion can be called from SQL commands by unprivileged users, but these functions are not designed to be safe against malicious choices of argument values. This problem exists in PostgreSQL 7.3.* through 8.0.*. The recommended fix is to disable public EXECUTE access for these functions. This does not affect normal usage of the functions for character set conversion, but it will prevent misuse.

The other error is that the contrib/tsearch2 module misdeclares several functions as returning type "internal" when they do not have any "internal" argument. This breaks the type safety of "internal" by allowing users to construct SQL commands that invoke other functions accepting "internal" arguments. The consequences of this have not been investigated in detail, but it is certainly at least possible to crash the backend.

Eric Romang reports a temporary file creation vulnerability within heartbeat. The vulnerability is caused by hardcoded temporary file usage. This can cause an attacker to create an arbitrary symlink causing the application to overwrite the symlinked file with the permissions of the user executing the application.

The KDE team reports:

kpdf, the KDE pdf viewer, shares code with xpdf. xpdf contains a heap based buffer overflow in the splash rasterizer engine that can crash kpdf or even execute arbitrary code.

The Perl Development page reports:

Dyad Security recently released a security advisory explaining how in certain cases, a carefully crafted format string passed to sprintf can cause a buffer overflow. This buffer overflow can then be used by an attacker to execute code on the machine. This was discovered in the context of a design problem with the Webmin administration package that allowed a malicious user to pass unchecked data into sprintf.

Francesco Ongaro reports that phpicalendar is vulnerable for a cross site scripting attack. The vulnerability is caused by improper validation of the index.php file allowing attackers to include an arbitrary file with the .php extension

The phpicalendar team reports that there is an unspecified vulnerability within phpicalendar. This seems to be a file disclosure vulnerability caused by improper checking of the template parsing function. This would allow an attacker to disclose any file readable by the user under which the webserver runs.

Problem description:

When insufficient memory is available to handle an incoming selective acknowledgement, the TCP/IP stack may enter an infinite loop.

Impact:

By opening a TCP connection and sending a carefully crafted series of packets, an attacker may be able to cause a denial of service.

Workaround:

On FreeBSD 5.4, the net.inet.tcp.sack.enable sysctl can be used to disable the use of SACK:

# sysctl net.inet.tcp.sack.enable=0

No workaround is available for FreeBSD 5.3.

Problem description:

A logic bug in pf's IP fragment cache may result in a packet fragment being inserted twice, violating a kernel invariant.

Impact:

By sending carefully crafted sequence of IP packet fragments, a remote attacker can cause a system running pf with a ruleset containing a 'scrub fragment crop' or 'scrub fragment drop-ovl' rule to crash.

Workaround:

Do not use 'scrub fragment crop' or 'scrub fragment drop-ovl' rules on systems running pf. In most cases, such rules can be replaced by 'scrub fragment reassemble' rules; see the pf.conf(5) manual page for more details.

Systems which do not use pf, or use pf but do not use the aforementioned rules, are not affected by this issue.

Problem description:

A buffer allocated from the kernel stack may not be completely initialized before being copied to userland. [CVE-2006-0379]

A logic error in computing a buffer length may allow too much data to be copied into userland. [CVE-2006-0380]

Impact:

Portions of kernel memory may be disclosed to local users. Such memory might contain sensitive information, such as portions of the file cache or terminal buffers. This information might be directly useful, or it might be leveraged to obtain elevated privileges in some way. For example, a terminal buffer might include a user-entered password.

Workaround:

No workaround is available.

Problem description:

An integer overflow in the handling of corrupt IEEE 802.11 beacon or probe response frames when scanning for existing wireless networks can result in the frame overflowing a buffer.

Impact:

An attacker able broadcast a carefully crafted beacon or probe response frame may be able to execute arbitrary code within the context of the FreeBSD kernel on any system scanning for wireless networks.

Workaround:

No workaround is available, but systems without IEEE 802.11 hardware or drivers loaded are not vulnerable.

Problem description:

The firewall maintains a pointer to layer 4 header information in the event that it needs to send a TCP reset or ICMP error message to discard packets. Due to incorrect handling of IP fragments, this pointer fails to get initialized.

Impact:

An attacker can cause the firewall to crash by sending ICMP IP fragments to or through firewalls which match any reset, reject or unreach actions.

Workaround:

Change any reset, reject or unreach actions to deny. It should be noted that this will result in packets being silently discarded.

Mitre CVE reports:

Format string vulnerability in main.cpp in kpopup 0.9.1-0.9.5pre2 allows local users to cause a denial of service (segmentation fault) and possibly execute arbitrary code via format string specifiers in command line arguments.

misc.cpp in KPopup 0.9.1 trusts the PATH variable when executing killall, which allows local users to elevate their privileges by modifying the PATH variable to reference a malicious killall program.

SecurityFocus credits "b0f" b0fnet@yahoo.com

Problem description:

A number of issues has been discovered in cpio:

When creating a new file, cpio closes the file before setting its permissions. (CVE-2005-1111)

When extracting files cpio does not properly sanitize file names to filter out ".." components, even if the --no-absolute-filenames option is used. (CVE-2005-1229)

When adding large files (larger than 4 GB) to a cpio archive on 64-bit platforms an internal buffer might overflow. (CVE-2005-4268)

Impact

The first problem can allow a local attacker to change the permissions of files owned by the user executing cpio providing that they have write access to the directory in which the file is being extracted. (CVE-2005-1111)

The lack of proper file name sanitation can allow an attacker to overwrite arbitrary local files when extracting files from a cpio archive. (CVE-2005-1229)

The buffer-overflow on 64-bit platforms could lead cpio to a Denial-of-Service situation (crash) or possibly execute arbitrary code with the permissions of the user running cpio. (CVE-2005-4268)

Workaround

Use a different utility to create and extract cpio archives, for example pax(1) or (on FreeBSD 5.3 or later) tar(1). If this is not possible, do not extract untrusted archives and when running on 64-bit platforms do not add untrusted files to cpio archives.

Problem description

The ispell_op function used by ee(1) while executing spell check operations employs an insecure method of temporary file generation. This method produces predictable file names based on the process ID and fails to confirm which path will be over written with the user.
It should be noted that ispell does not have to be installed in order for this to be exploited. The option simply needs to be selected.

Impact

These predictable temporary file names are problematic because they allow an attacker to take advantage of a race condition in order to execute a symlink attack, which could allow them to overwrite files on the system in the context of the user running the ee(1) editor.

Workaround

Instead of invoking ispell through ee(1), invoke it directly.

Problem description

The "sort_offline" function used by texindex(1) employs the "maketempname" function, which produces predictable file names and fails to validate that the paths do not exist.

Impact

These predictable temporary file names are problematic because they allow an attacker to take advantage of a race condition in order to execute a symlink attack, which could enable them to overwrite files on the system in the context of the user running the texindex(1) utility.

Workaround

No workaround is available, but the problematic code is only executed if the input file being processed is 500kB or more in length; as a result, users working with documents of less than several hundred pages are very unlikely to be affected.

Problem description

A temporary file is created, used, deleted, and then re-created with the same name. This creates a window during which an attacker could replace the file with a link to another file. While cvsbug(1) is based on the send-pr(1) utility, this problem does not exist in the version of send-pr(1) distributed with FreeBSD.
In FreeBSD 4.10 and 5.3, some additional problems exist concerning temporary file usage in both cvsbug(1) and send-pr(1).

Impact

A local attacker could cause data to be written to any file to which the user running cvsbug(1) (or send-pr(1) in FreeBSD 4.10 and 5.3) has write access. This may cause damage in itself (e.g., by destroying important system files or documents) or may be used to obtain elevated privileges.

Workaround

Do not use the cvsbug(1) utility on any system with untrusted users.
Do not use the send-pr(1) utility on a FreeBSD 4.10 or 5.3 system with untrusted users.

Sun Microsystems reports:

The SGE 6.0u7_1 release fixes a security bug which can allow malicious users to gain root access.

Matthias Andree reports:

Fetchmail contains a bug that causes itself to crash when bouncing a message to the originator or to the local postmaster. The crash happens after the bounce message has been sent, when fetchmail tries to free the dynamic array of failed addresses, and calls the free() function with an invalid pointer.

The Zero Day Initiative reports:

This vulnerability allows remote attackers to execute arbitrary code on vulnerable Clam AntiVirus installations. Authentication is not required to exploit this vulnerability.

This specific flaw exists within libclamav/upx.c during the unpacking of executable files compressed with UPX. Due to an invalid size calculation during a data copy from the user-controlled file to heap allocated memory, an exploitable memory corruption condition is created.

Juan J. Marítnez reports:

The milter crashes while processing a headerless message

Impact: bogom crashes and sendmail moves it to error state

Matthias Andree reports:

Bogofilter's/bogolexer's input handling in version 0.96.2 was not keeping track of its output buffers properly and could overrun a heap buffer if the input contained words whose length exceeded 16,384 bytes, the size of flex's input buffer. A "word" here refers to a contiguous run of input octets that was not '_' and did not match at least one of ispunct(), iscntrl() or isspace().

Matthias Andree reports:

When using Unicode databases (default in more recent bogofilter installations), upon encountering invalid input sequences, bogofilter or bogolexer could overrun a malloc()'d buffer, corrupting the heap, while converting character sets. Bogofilter would usually be processing untrusted data received from the network at that time.

This problem was aggravated by an unrelated bug that made bogofilter process binary attachments as though they were text, and attempt charset conversion on them. Given the MIME default character set, US-ASCII, all input octets in the range 0x80...0xff were considered invalid input sequences and could trigger the heap corruption.

A rxvt-unicode changelog reports:

SECURITY FIX: on systems using openpty, permissions were not correctly updated on the tty device and were left as world-readable and world-writable (likely in original rxvt, too), and were not restored properly. Affected are only systems where non-unix ptys were used (such as most BSDs). Found, patched and debugged by Ryan Beasley.

The Apache HTTP Server Project reports:

A flaw in mod_imap when using the Referer directive with image maps. In certain site configurations a remote attacker could perform a cross-site scripting attack if a victim can be forced to visit a malicious URL using certain web browsers.

Kurt Fitzner reports a buffer overflow vulnerability within nbd. This could potentially allow the execution of arbitrary code on the nbd server.

Max Vozeler reports:

If ALL the following conditions are true, administrators using scponly-4.1 or older may be at risk of a local privilege escalation exploit:

• the chrooted setuid scponlyc binary is installed
• regular non-scponly users have interactive shell access to the box
• a user executable dynamically linked setuid binary (such as ping) exists on the same file system mount as the user's home directory
• the operating system supports an LD_PRELOAD style mechanism to overload dynamic library loading

Pekka Pessi also reports:

If ANY the following conditions are true, administrators using scponly-4.1 or older may be at risk of a local privilege escalation exploit:

• scp compatibility is enabled
• rsync compatibility is enabled

The fetchmail team reports:

Fetchmail contains a bug that causes an application crash when fetchmail is configured for multidrop mode and the upstream mail server sends a message without headers. As fetchmail does not record this message as "previously fetched", it will crash with the same message if it is re-executed, so it cannot make progress. A malicious or broken-into upstream server could thus cause a denial of service in fetchmail clients.

Secunia Research reports:

Input passed to the "t_core_path" parameter in "bug_sponsorship_list_view_inc.php" isn't properly verified, before it used to include files. This can be exploited to include arbitrary files from external and local resources.

r0t reports:

Mantis contains a flaw that allows a remote cross site scripting attack. This flaw exists because input passed to "target_field" parameter in "view_filters_page.php" isn't properly sanitised before being returned to the user. This could allow a user to create a specially crafted URL that would execute arbitrary code in a user's browser within the trust relationship between the browser and the server, leading to a loss of integrity.

Announce of Mnemo H3 (2.0.3) (final):

This [2.0.3] is a security release that fixes cross site scripting vulnerabilities in several of the notepad name and note data fields. None of the vulnerabilities can be exploited by unauthenticated users; however, we strongly recommend that all users of Mnemo 2.0.2 upgrade to 2.0.3 as soon as possible.

Announce of Nag H3 (2.0.4) (final):

This [2.0.4] is a security release that fixes cross site scripting vulnerabilities in several of the tasklist name and task data fields. None of the vulnerabilities can be exploited by unauthenticated users; however, we strongly recommend that all users of Nag 2.0.3 upgrade to 2.0.4 as soon as possible.

Announce of Turba H3 (2.0.5) (final):

This [2.0.5] is a security release that fixes cross site scripting vulnerabilities in several of the address book name and contact data fields. None of the vulnerabilities can be exploited by unauthenticated users; however, we strongly recommend that all users of Turba 2.0.4 upgrade to 2.0.5 as soon as possible.

Announce of Kronolith H3 (2.0.6) (final):

This [2.0.6] is a security release that fixes cross site scripting vulnerabilities in several of the calendar name and event data fields. None of the vulnerabilities can be exploited by unauthenticated users; however, we strongly recommend that all users of Kronolith 2.0.5 upgrade to 2.0.6 as soon as possible.

Announce of Horde H3 3.0.8 (final):

This [3.0.8] is a security release that fixes cross site scripting vulnerabilities in several of Horde's templates. None of the vulnerabilities can be exploited by unauthenticated users; however, we strongly recommend that all users of Horde 3.0.7 upgrade to 3.0.8 as soon as possible.

A Project cURL Security Advisory reports:

libcurl's URL parser function can overflow a malloced buffer in two ways, if given a too long URL.

1 - pass in a URL with no protocol (like "http://") prefix, using no slash and the string is 256 bytes or longer. This leads to a single zero byte overflow of the malloced buffer.

2 - pass in a URL with only a question mark as separator (no slash) between the host and the query part of the URL. This leads to a single zero byte overflow of the malloced buffer.

Both overflows can be made with the same input string, leading to two single zero byte overwrites.

The affected flaw cannot be triggered by a redirect, but the long URL must be passed in "directly" to libcurl. It makes this a "local" problem. Of course, lots of programs may still pass in user-provided URLs to libcurl without doing much syntax checking of their own, allowing a user to exploit this vulnerability.

Secunia reports:

Stefan Esser has reported a vulnerability in phpMyAdmin, which can be exploited by malicious people to conduct cross-site scripting attacks, disclose sensitive information, and compromise a vulnerable system.

The vulnerability is caused due to an error in the register_globals emulation layer in "grab_globals.php" where the "import_blacklist" variable is not properly protected from being overwritten. This can be exploited to execute arbitrary HTML and script code in a user's browser session in context of an affected site, and include arbitrary files from external and local resources.

A phpMyAdmin security advisory reports:

It was possible to conduct an XSS attack via the HTTP_HOST variable; also, some scripts in the libraries directory that handle header generation were vulnerable to XSS.

Secunia reports:

Simon Kilvington has reported a vulnerability in FFmpeg libavcodec, which can be exploited by malicious people to cause a DoS (Denial of Service) and potentially to compromise a user's system.

The vulnerability is caused due to a boundary error in the "avcodec_default_get_buffer()" function of "utils.c" in libavcodec. This can be exploited to cause a heap-based buffer overflow when a specially-crafted 1x1 ".png" file containing a palette is read.

Secunia reports:

A vulnerability has been reported in Trac, which can be exploited by malicious people to conduct SQL injection attacks.

Some unspecified input passed in the search module isn't properly sanitised before being used in a SQL query. This can be exploited to manipulate SQL queries by injecting arbitrary SQL code.

Secunia reports:

Some vulnerabilities have been reported in Drupal, which can be exploited by malicious people to bypass certain security restrictions, and conduct script insertion and HTTP response splitting attacks.

1) An input validation error in the filtering of HTML code can be exploited to inject arbitrary JavaScript code in submitted content, which will be executed in a user's browser session in context of an affected site when the malicious user data is viewed. Successful exploitation requires that the user has access to the full HTML input format. Ref: sa-2005-007

2) An input validation error in the attachment handling can be exploited to upload a malicious image with embedded HTML and script content, which will be executed in a user's browser session in context of an affected site when viewed directly with the Microsoft Internet Explorer browser. This can also be exploited to inject arbitrary HTTP headers, which will be included in the response sent to the user. Ref: sa-2005-008

3) The problem is that it is possible to bypass the "access user profile" permission. However, this cannot be exploited to modify data. Successful exploitation requires that the server runs PHP 5. Ref: sa-2005-009

Opera reports:

It is possible to make a form input that looks like an image link. If the form input has a "title" attribute, the status bar will show the "title". A "title" which looks like a URL can mislead the user, since the title can say http://nice.familiar.com/, while the form action can be something else.

Opera's tooltip says "Title:" before the title text, making a spoof URL less convincing. A user who has enabled the status bar and disabled tooltips can be affected by this. Neither of these settings are Opera's defaults.

This exploit is mostly of interest to users who disable JavaScript. If JavaScript is enabled, any link target or form action can be overridden by the script. The tooltip and the statusbar can only be trusted to show the true location if JavaScript is disabled.

Java code using LiveConnect methods to remove a property of a JavaScript object may in some cases use null pointers that can make Opera crash. This crash is not exploitable and such code is rare on the web.

An Opera Advisory reports:

Opera for UNIX uses a wrapper shell script to start up Opera. This shell script reads the input arguments, like the file names or URLs that Opera is to open. It also performs some environment checks, for example whether Java is available and if so, where it is located.

This wrapper script can also run commands embedded in the URL, so that a specially crafted URL can make arbitrary commands run on the recipient's machine. Users who have other programs set up to use Opera to open Web links are vulnerable to this flaw. For these users, clicking a Web link in for example OpenOffice.org or Evolution can run a command that was put into the link.

A Secunia Advisory reports:

peter MC tachatte has discovered a vulnerability in Mambo, which can be exploited by malicious people to manipulate certain information and compromise a vulnerable system.

The vulnerability is caused due to an error in the "register_globals" emulation layer in "globals.php" where certain arrays used by the system can be overwritten. This can be exploited to include arbitrary files from external and local resources via the "mosConfig_absolute_path" parameter.

Successful exploitation requires that "register_globals" is disabled.

Ghostscript is affected by an insecure temporary file creation vulnerability. This issue is likely due to a design error that causes the application to fail to verify the existence of a file before writing to it.

An attacker may leverage this issue to overwrite arbitrary files with the privileges of an unsuspecting user that activates the vulnerable application. Reportedly this issue is unlikely to facilitate privilege escalation.

Announce of Horde 3.0.7 (final):

This [3.0.7] is a security release that fixes cross site scripting vulnerabilities in two of Horde's MIME viewers. These holes could for example be exploited by an attacker sending specially crafted emails to Horde's webmail client IMP. The attack could be used to steal users' identity information, taking over users' sessions, or changing users' settings.

As a hotfix the css and tgz MIME drivers can be disabled by removing their entries from the $mime_drivers_map['horde']['registered'] list in horde/config/mime_drivers.php.

A phpMyAdmin security advisory reports:

Some scripts in phpMyAdmin are vulnerable to an HTTP Response Splitting attack.

Severity:

We consider these vulnerabilities to be serious. However, they can only be triggered on systems running with register_globals = on.

A Secunia Advisory reports:

Christopher Kunz has reported a vulnerability in phpSysInfo, which can be exploited by malicious people to manipulate certain information.

The vulnerability is caused due to an error in the "register_globals" emulation layer where certain arrays used by the system can be overwritten. This can be exploited to execute arbitrary HTML and script code in a user's browser session and include arbitrary files from local resources.

A Secunia Advisory reports:

A vulnerability has been reported in Macromedia Flash Player, which can be exploited by malicious people to compromise a user's system.

The vulnerability is caused due to missing validation of the frame type identifier that is read from a SWF file. This value is used as an index in Flash.ocx to reference an array of function pointers. This can be exploited via a specially crafted SWF file to cause the index to reference memory that is under the attacker's control, which causes Flash Player to use attacker supplied values as function pointers.

Successful exploitation allows execution of arbitrary code.

A Secunia Advisory reports:

Lostmon has reported some vulnerabilities in Flyspray, which can be exploited by malicious people to conduct cross-site scripting attacks.

Some input isn't properly sanitised before being returned to the user. This can be exploited to execute arbitrary HTML and script code in a user's browser session in context of an affected site.

A Secunia Advisory reports:

A vulnerability has been reported in SpamAssassin, which can be exploited by malicious people to cause a DoS (Denial of Service).

The vulnerability is caused due to the use of an inefficient regular expression in "/SpamAssassin/Message.pm" to parse email headers. This can cause perl to crash when it runs out of stack space and can be exploited via a malicious email that contains a large number of recipients.

Jens Steube reports that qpopper is vulnerable to a privilege escalation vulnerability. qpopper does not properly drop root privileges so that user supplied configuration and trace files can be processed with root privileges. This could allow a local attacker to create or modify arbitrary files.

qpopper is also affected by improper umask settings which could allow users to create group or world-writeable files, possibly allowing an attacker to overwrite arbitrary files.

Gregory Beaver reports:

A standard feature of the PEAR installer implemented in all versions of PEAR can lead to the execution of arbitrary PHP code upon running the "pear" command or loading the Web/Gtk frontend.

To be vulnerable, a user must explicitly install a publicly released malicious package using the PEAR installer, or explicitly install a package that depends on a malicious package.

James Yonan reports:

If the TCP server accept() call returns an error status, the resulting exception handler may attempt to indirect through a NULL pointer, causing a segfault. Affects all OpenVPN 2.0 versions.

James Yonan reports:

A format string vulnerability in the foreign_option function in options.c could potentially allow a malicious or compromised server to execute arbitrary code on the client. Only non-Windows clients are affected. The vulnerability only exists if (a) the client's TLS negotiation with the server succeeds, (b) the server is malicious or has been compromised such that it is configured to push a maliciously crafted options string to the client, and (c) the client indicates its willingness to accept pushed options from the server by having "pull" or "client" in its configuration file (Credit: Vade79).

A Secunia Advisory reports:

Some vulnerabilities have been reported in PHP, which can be exploited by malicious people to conduct cross-site scripting attacks, bypass certain security restrictions, and potentially compromise a vulnerable system.

A Secunia Advisory reports:

Some vulnerabilities have been reported in Skype, which can be exploited by malicious people to cause a DoS or to compromise a user's system.

A Secunia Advisory reports:

M.A.Young has reported a vulnerability in Squid, which can be exploited by malicious people to cause a DoS (Denial of Service).

The vulnerability is caused due to an error in handling certain FTP server responses. This can be exploited to crash Squid by visiting a malicious FTP server via the proxy.

A Secunia Advisory reports:

Remco Verhoef has discovered a vulnerability in Basic Analysis and Security Engine (BASE), which can be exploited by malicious users to conduct SQL injection attacks.

The fetchmail team reports:

The fetchmailconf program before and excluding version 1.49 opened the run control file, wrote the configuration to it, and only then changed the mode to 0600 (rw-------). Writing the file, which usually contains passwords, before making it unreadable to other users, can expose sensitive password information.

Ulf Härnhammar reports:

When Lynx connects to an NNTP server to fetch information about the available articles in a newsgroup, it will call a function called HTrjis() with the information from certain article headers. The function adds missing ESC characters to certain data, to support Asian character sets. However, it does not check if it writes outside of the char array buf, and that causes a remote stack-based buffer overflow.

Ruby home page reports:

The Object Oriented Scripting Language Ruby supports safely executing an untrusted code with two mechanisms: safe level and taint flag on objects.

A vulnerability has been found that allows bypassing these mechanisms.

By using the vulnerability, arbitrary code can be executed beyond the restrictions specified in each safe level. Therefore, Ruby has to be updated on all systems that use safe level to execute untrusted code.

Ariel Berkman reports:

Unlike most of the supported image formats in xloadimage, the NIFF image format can store a title name of arbitrary length as part of the image file.

When xloadimage is processing a loaded image, it is creating a new Image object and then writing the processed image to it. At that point, it will also copy the title from the old image to the newly created image.

The 'zoom', 'reduce', and 'rotate' functions are using a fixed length buffer to construct the new title name when an image processing is done. Since the title name in a NIFF format is of varying length, and there are insufficient buffer size validations, the buffer can be overflowed.

Jennifer Steffens reports:

The Back Orifice preprocessor contains a stack-based buffer overflow. This vulnerability could be leveraged by an attacker to execute code remotely on a Snort sensor where the Back Orifice preprocessor is enabled. However, there are a number of factors that make remote code execution difficult to achieve across different builds of Snort on different platforms, even on the same platform with different compiler versions, and it is more likely that an attacker could use the vulnerability as a denial of service attack.

The Back Orifice preprocessor can be disabled by commenting out the line "preprocessor bo" in snort.conf. This can be done in any text editor using the following procedure:

1. Locate the line "preprocessor bo"
2. Comment out this line by preceding it with a hash (#). The new line will look like "#preprocessor bo"
3. Save the file
4. Restart snort

WebCalendar is proven vulnerable to a remote file inclusion vulnerability. The send_reminders.php does not properly verify the "includedir" parameter, giving remote attackers the possibility to include local and remote files. These files can be used by the attacker to gain access to the system.

Michael Dipper wrote:

A vulnerability has been discovered in gallery, which allows remote users unauthorized access to files on the webserver.

A remote user accessing gallery over the web may use specially crafted HTTP parameters to access arbitrary files located on the webserver. All files readable by the webserver process are subject to disclosure. The vulnerability is *not* restricted to the webserver's document root but extends to the whole server file space.

The vulnerabilty may be used by any anonymous user, there is no login to the application required.

Vulnerability:

Such applications are affected if they use the option SSL_OP_MSIE_SSLV2_RSA_PADDING. This option is implied by use of SSL_OP_ALL, which is intended to work around various bugs in third-party software that might prevent interoperability. The SSL_OP_MSIE_SSLV2_RSA_PADDING option disables a verification step in the SSL 2.0 server supposed to prevent active protocol-version rollback attacks. With this verification step disabled, an attacker acting as a "man in the middle" can force a client and a server to negotiate the SSL 2.0 protocol even if these parties both support SSL 3.0 or TLS 1.0. The SSL 2.0 protocol is known to have severe cryptographic weaknesses and is supported as a fallback only.

Applications using neither SSL_OP_MSIE_SSLV2_RSA_PADDING nor SSL_OP_ALL are not affected. Also, applications that disable use of SSL 2.0 are not affected.

A phpMyAdmin security announcement reports:

In libraries/grab_globals.lib.php, the $__redirect parameter was not correctly validated, opening the door to a local file inclusion attack.

We consider this vulnerability to be serious.

A Zope Hotfix Alert reports:

This hotfix resolves a security issue with docutils.

Affected are possibly all Zope instances that expose RestructuredText functionalies to untrusted users through the web.

Gentoo Linux Security Advisory reports:

Ulf Harnhammar discovered a format string bug in the routines handling CDDB server response contents.

An attacker could submit malicious information about an audio CD to a public CDDB server (or impersonate a public CDDB server). When the victim plays this CD on a multimedia frontend relying on xine-lib, it could end up executing arbitrary code.

FrSIRT reports:

A vulnerability has been identified in UW-IMAP, which could be exploited by remote attackers to execute arbitrary commands. This flaw is due to a stack overflow error in the "mail_valid_net_parse_work()" [src/c-client/mail.c] function that does not properly handle specially crafted mailbox names containing a quote (") character, which could be exploited by authenticated remote attackers to execute arbitrary commands with the privileges of the IMAP server.

Emanuel Haupt reports:

Someone who controls an FTP server that weex will log in to can set up malicious data in the account that weex will use, and that will cause a format string bug that will allow remote code execution. It will only happen when weex is first run or when its cache files are rebuilt with the -r option, though. The vulnerability was found by Ulf Harnhammar.

Shaun Colley reports:

When generating error and warning messages, picasm copies strings into fixed length buffers without bounds checking.

If an attacker could trick a user into assembling a source file with a malformed 'error' directive, arbitrary code could be executed with the privileges of the user. This could result in full system compromise.

The uim developers reports:

Masanari Yamamoto discovered that incorrect use of environment variables in uim. This bug causes privilege escalation if setuid/setgid applications was linked to libuim.

This bug appears in 'immodule for Qt' enabled Qt. (Normal Qt is also safe.) In some distribution, mlterm is also an setuid/setgid application.

A Debian Security Advisory reports:

Javier Fernández-Sanguino Peña discovered several insecure temporary file uses in cfengine, a tool for configuring and maintaining networked machines, that can be exploited by a symlink attack to overwrite arbitrary files owned by the user executing cfengine, which is probably root.

Gentoo Linux Security Advisory reports:

Clam AntiVirus is vulnerable to a buffer overflow in "libclamav/upx.c" when processing malformed UPX-packed executables. It can also be sent into an infinite loop in "libclamav/fsg.c" when processing specially-crafted FSG-packed executables.

By sending a specially-crafted file an attacker could execute arbitrary code with the permissions of the user running Clam AntiVirus, or cause a Denial of Service.

A Mozilla Foundation Security Advisory reports of multiple issues:

Heap overrun in XBM image processing

jackerror reports that an improperly terminated XBM image ending with space characters instead of the expected end tag can lead to a heap buffer overrun. This appears to be exploitable to install or run malicious code on the user's machine.

Thunderbird does not support the XBM format and is not affected by this flaw.

Crash on "zero-width non-joiner" sequence

Mats Palmgren discovered that a reported crash on Unicode sequences with "zero-width non-joiner" characters was due to stack corruption that may be exploitable.

XMLHttpRequest header spoofing

It was possible to add illegal and malformed headers to an XMLHttpRequest. This could have been used to exploit server or proxy flaws from the user's machine, or to fool a server or proxy into thinking a single request was a stream of separate requests. The severity of this vulnerability depends on the value of servers which might be vulnerable to HTTP request smuggling and similar attacks, or which share an IP address (virtual hosting) with the attacker's page.

For users connecting to the web through a proxy this flaw could be used to bypass the same-origin restriction on XMLHttpRequests by fooling the proxy into handling a single request as multiple pipe-lined requests directed at arbitrary hosts. This could be used, for example, to read files on intranet servers behind a firewall.

Object spoofing using XBL <implements>

moz_bug_r_a4 demonstrated a DOM object spoofing bug similar to MFSA 2005-55 using an XBL control that <implements> an internal interface. The severity depends on the version of Firefox: investigation so far indicates Firefox 1.0.x releases don't expose any vulnerable functionality to interfaces spoofed in this way, but that early Deer Park Alpha 1 versions did.

XBL was changed to no longer allow unprivileged controls from web content to implement XPCOM interfaces.

JavaScript integer overflow

Georgi Guninski reported an integer overflow in the JavaScript engine. We presume this could be exploited to run arbitrary code under favorable conditions.

Privilege escalation using about: scheme

heatsync and shutdown report two different ways to bypass the restriction on loading high privileged "chrome" pages from an unprivileged "about:" page. By itself this is harmless--once the "about" page's privilege is raised the original page no longer has access--but should this be combined with a same-origin violation this could lead to arbitrary code execution.

Chrome window spoofing

moz_bug_r_a4 demonstrates a way to get a blank "chrome" canvas by opening a window from a reference to a closed window. The resulting window is not privileged, but the normal browser UI is missing and can be used to construct a spoof page without any of the safety features of the browser chrome designed to alert users to phishing sites, such as the address bar and the status bar.

A Secunia Advisory reports:

Peter Zelezny has discovered a vulnerability in Firefox, which can be exploited by malicious people to compromise a user's system.

The vulnerability is caused due to the shell script used to launch Firefox parsing shell commands that are enclosed within backticks in the URL provided via the command line. This can e.g. be exploited to execute arbitrary shell commands by tricking a user into following a malicious link in an external application which uses Firefox as the default browser.

Marc Stern reports an off-by-one vulnerability in within mod_ssl. The vulnerability lies in mod_ssl's Certificate Revocation List (CRL). If Apache is configured to use a CRL this could allow an attacker to crash a child process causing a Denial of Service.

A Squirrelmail Advisory reports:

An extract($_POST) was done in options_identities.php which allowed for an attacker to set random variables in that file. This could lead to the reading (and possible writing) of other people's preferences, cross site scripting or writing files in webserver-writable locations.

Allocating large pixmaps by a client can trigger an integer overflow in the X server, potentially leading to execution of arbitrary code with elevated (root) privileges.

Imran Ghory reports a vulnerability within unzip. The vulnerability is caused by a race condition between extracting an archive and changing the permissions of the extracted files. This would give an attacker enough time to remove a file and hardlink it to another file owned by the user running unzip. When unzip changes the permissions of the file it could give the attacker access to files that normally would not have been accessible for others.

Tom Ferris reports:

A buffer overflow vulnerability exists within Firefox version 1.0.6 and all other prior versions which allows for an attacker to remotely execute arbitrary code on an affected host.

The problem seems to be when a hostname which has all dashes causes the NormalizeIDN call in nsStandardURL::BuildNormalizedSpec to return true, but is sets encHost to an empty string. Meaning, Firefox appends 0 to approxLen and then appends the long string of dashes to the buffer instead.

Note: It is possible to disable IDN support as a workaround to protect against this buffer overflow. How to do this is described on the What Firefox and Mozilla users should know about the IDN buffer overflow security issue web page.

Michael Krax reports a vulnerability within htdig. The vulnerability lies within an unsanitized config parameter, allowing a malicious attacker to execute arbitrary scripting code on the target's browser. This might allow the attacker to obtain the user's cookies which are associated with the site, including cookies used for authentication.

The squid patches page notes:

After certain slightly odd requests Squid crashes with a segmentation fault in sslConnectTimeout.

The squid patches page notes:

Squid crashes with the above assertion failure [assertion failed: store.c:523: "e->store_status == STORE_PENDING"] in certain conditions involving aborted requests.

Problem description

A DNSSEC-related validator function in BIND 9.3.0 contains an inappropriate internal consistency test. When this test is triggered, named(8) will exit.

Impact

On systems with DNSSEC enabled, a remote attacker may be able to inject a specially crafted packet that will cause the internal consistency test to trigger, and named(8) to terminate. As a result, the name server will no longer be available to service requests.

Workaround

DNSSEC is not enabled by default, and the "dnssec-enable" directive is not normally present. If DNSSEC has been enabled, disable it by changing the "dnssec-enable" directive to "dnssec-enable no;" in the named.conf(5) configuration file.

An ISC advisory reports a buffer overrun vulnerability within bind. The vulnerability could result in a Denial of Service. A workaround is available by disabling recursion and glue fetching.

Several filename-related stack overflow bugs allow a local attacker to elevate its privileges to the games group, since urban is installed setgid games.

Issue discovered and fixed by <shaun@rsc.cx>.

There is a command injection vulnerability in admin page of fswiki.

A SITIC Vulnerability Advisory reports:

Evolution suffers from several format string bugs when handling data from remote sources. These bugs lead to crashes or the execution of arbitrary assembly language code.

1. The first format string bug occurs when viewing the full vCard data attached to an e-mail message.
2. The second format string bug occurs when displaying contact data from remote LDAP servers.
3. The third format string bug occurs when displaying task list data from remote servers.
4. The fourth, and least serious, format string bug occurs when the user goes to the Calendars tab to save task list data that is vulnerable to problem 3 above. Other calendar entries that do not come from task lists are also affected.

Luke Howard reports:

If a pam_ldap client authenticates against an LDAP server that returns a passwordPolicyResponse control, but omits the optional "error" field of the PasswordPolicyResponseValue, then the LDAP authentication result will be ignored and the authentication step will always succeed.

The pcre library is vulnerable to a buffer overflow vulnerability due to insufficient validation of quantifier values. This could lead execution of arbitrary code with the permissions of the program using pcre by way of a specially crated regular expression.

Ulf Harnhammar has discovered a remotely exploitable buffer overflow in Elm e-mail client when parsing the Expires header of an e-mail message:

The attacker only needs to send the victim an e-mail message. When the victim with that message in his or her inbox starts Elm or simply views the inbox in an already started copy of Elm, the buffer overflow will happen immediately. The overflow is stack-based, and it gives full control over EIP, EBP and EBX. It is caused by a bad sscanf(3) call, using a format string containing "%s" to copy from a long char array to a shorter array.

James Yonan reports:

If two or more client machines try to connect to the server at the same time via TCP, using the same client certificate, and when --duplicate-cn is not enabled on the server, a race condition can crash the server with "Assertion failed at mtcp.c:411"

James Yonan reports:

A malicious [authenticated] client in "dev tap" ethernet bridging mode could theoretically flood the server with packets appearing to come from hundreds of thousands of different MAC addresses, causing the OpenVPN process to deplete system virtual memory as it expands its internal routing table.

James Yonan reports:

If the client sends a packet which fails to decrypt on the server, the OpenSSL error queue is not properly flushed, which can result in another unrelated client instance on the server seeing the error and responding to it, resulting in disconnection of the unrelated client.

James Yonan reports:

DoS attack against server when run with "verb 0" and without "tls-auth". If a client connection to the server fails certificate verification, the OpenSSL error queue is not properly flushed, which can result in another unrelated client instance on the server seeing the error and responding to it, resulting in disconnection of the unrelated client.

A tor advisory reports

Tor clients can completely loose anonymity, confidentiality, and data integrity if the first Tor server in their path is malicious. Specifically, if the Tor client chooses a malicious Tor server for her first hop in the circuit, that server can learn all the keys she negotiates for the rest of the circuit (or just spoof the whole circuit), and then read and/or modify all her traffic over that circuit.

A Adobe Security Advisory reports:

The identified vulnerability is a buffer overflow within a core application plug-in, which is part of Adobe Acrobat and Adobe Reader. If a malicious file were opened it could trigger a buffer overflow as the file is being loaded into Adobe Acrobat and Adobe Reader. A buffer overflow can cause the application to crash and increase the risk of malicious code execution.

A Hardened-PHP Project Security Advisory reports:

When the library parses XMLRPC requests/responses, it constructs a string of PHP code, that is later evaluated. This means any failure to properly handle the construction of this string can result in arbitrary execution of PHP code.

This new injection vulnerability is cause by not properly handling the situation, when certain XML tags are nested in the parsed document, that were never meant to be nested at all. This can be easily exploited in a way, that user-input is placed outside of string delimiters within the evaluation string, which obviously results in arbitrary code execution.

Note that several applications contains an embedded version on XML_RPC, therefor making them the vulnerable to the same code injection vulnerability.

An iDEFENSE Security Advisory reports:

Remote exploitation of an input validation vulnerability in AWStats allows remote attackers to execute arbitrary commands.

The problem specifically exists because of insufficient input filtering before passing user-supplied data to an eval() function. As part of the statistics reporting function, AWStats displays information about the most common referrer values that caused users to visit the website. The referrer data is used without proper sanitation in an eval() statement, resulting in the execution of arbitrary perl code.

Successful exploitation results in the execution of arbitrary commands with permissions of the web service. Exploitation will not occur until the stats page has been regenerated with the tainted referrer values from the http access log. Note that AWStats is only vulnerable in situations where at least one URLPlugin is enabled.

Wojtek Kaniewski reports:

Multiple vulnerabilities have been found in libgadu, a library for handling Gadu-Gadu instant messaging protocol. It is a part of ekg, a Gadu-Gadu client, but is widely used in other clients. Also some of the user contributed scripts were found to behave in an insecure manner.

• integer overflow in libgadu (CVE-2005-1852) that could be triggered by an incomming message and lead to application crash and/or remote code execution
• insecure file creation (CVE-2005-1850) and shell command injection (CVE-2005-1851) in other user contributed scripts (discovered by Marcin Owsiany and Wojtek Kaniewski)
• several signedness errors in libgadu that could be triggered by an incomming network data or an application passing invalid user input to the library
• memory alignment errors in libgadu that could be triggered by an incomming message and lead to bus errors on architectures like SPARC
• endianness errors in libgadu that could cause invalid behaviour of applications on big-endian architectures

The GAIM team reports:

A remote user could cause Gaim to crash on some systems by sending the Gaim user a file whose filename contains certain invalid characters. It is unknown what combination of systems are affected, but it is suspected that Windows users and systems with older versions of GTK+ are especially susceptible.

The GAIM team reports:

A remote AIM or ICQ user can cause a buffer overflow in Gaim by setting an away message containing many AIM substitution strings (such as %t or %n).

xpdf is vulnerable to a denial of service vulnerability which can cause xpdf to create an infinitely large file, thereby filling up the /tmp partition, when opening a specially crafted PDF file.

Note that several applications contains an embedded version of xpdf, therefor making them the vulnerable to the same DoS. In CUPS this vulnerability would cause the pdftops filter to crash.

Jose Antonio Coret reports that GForge contains multiple Cross Site Scripting vulnerabilities and an e-mail flood vulnerability:

The login form is also vulnerable to XSS (Cross Site Scripting) attacks. This may be used to launch phising attacks by sending HTML e-mails (i.e.: saying that you need to upgrade to the latest GForge version due to a security problem) and putting in the e-mail an HTML link that points to an specially crafted url that inserts an html form in the GForge login page and when the user press the login button, he/she send the credentials to the attackers website.

The 'forgot your password?' feature allows a remote user to load a certain URL to cause the service to send a validation e-mail to the specified user's e-mail address. There is no limit to the number of messages sent over a period of time, so a remote user can flood the target user's secondary e-mail address. E-Mail Flood, E-Mail bomber.

Postnuke Security Announcementss reports of the following vulnerabilities:

• missing input validation within /modules/Messages/readpmsg.php
• possible path disclosure within /user.php
• possible path disclosure within /modules/News/article.php
• possible remote code injection within /includes/pnMod.php
• possible cross-site-scripting in /index.php

• remote code injection via xml rpc library

A Secunia Advisory reports:

Some vulnerabilities have been reported in Mambo, where some have unknown impacts and others can be exploited by malicious people to conduct spoofing and SQL injection attacks.

1. Input passed to the "user_rating" parameter when voting isn't properly sanitised before being used in a SQL query. This can be exploited to manipulate SQL queries by injecting arbitrary SQL code.
2. Some unspecified vulnerabilities in the "mosDBTable" class and the "DOMIT" library have an unknown impact.
3. An unspecified error in the "administrator/index3.php" script can be exploited to spoof session IDs.

Problem description

A programming error in the implementation of the AES-XCBC-MAC algorithm for authentication resulted in a constant key being used instead of the key specified by the system administrator.

Impact

If the AES-XCBC-MAC algorithm is used for authentication in the absence of any encryption, then an attacker may be able to forge packets which appear to originate from a different system and thereby succeed in establishing an IPsec session. If access to sensitive information or systems is controlled based on the identity of the source system, this may result in information disclosure or privilege escalation.

Problem description

A fixed-size buffer is used in the decompression of data streams. Due to erronous analysis performed when zlib was written, this buffer, which was belived to be sufficiently large to handle any possible input stream, is in fact too small.

Impact

A carefully constructed compressed data stream can result in zlib overwriting some data structures. This may cause applications to halt, resulting in a denial of service; or it may result in an attacker gaining elevated privileges.

Problem description

Due to insufficient parameter checking of the node type during device creation, any user can expose hidden device nodes on devfs mounted file systems within their jail. Device nodes will be created in the jail with their normal default access permissions.

Impact

Jailed processes can get access to restricted resources on the host system. For jailed processes running with superuser privileges this implies access to all devices on the system. This level of access can lead to information leakage and privilege escalation.

The ProFTPD release notes states:

sean <infamous42md at hotpop.com> found two format string vulnerabilities, one in mod_sql's SQLShowInfo directive, and one involving the 'ftpshut' utility. Both can be considered low risk, as they require active involvement on the part of the site administrator in order to be exploited.

These vulnerabilities could potentially lead to information disclosure, a denial-of-server situation, or execution of arbitrary code with the permissions of the user running ProFTPD.

When nbsmtp is executed in debug mode, server messages will be printed to stdout and logged via syslog. Syslog is used insecurely and user-supplied format characters are directly fed to the syslog function, which results in a format string vulnerability.

Under some circumstances, an SMTP server may be able to abuse this vulnerability in order to alter the nbsmtp process and execute malicious code.

Sylpheed is vulnerable to a buffer overflow when displaying emails with attachments that have MIME-encoded file names. This could be used by a remote attacker to crash sylpheed potentially allowing execution of arbitrary code with the permissions of the user running sylpheed.

A phpMyAdmin security announcement reports:

The convcharset parameter was not correctly validated, opening the door to a XSS attack.

Serge Mister and Robert Zuccherato reports that the OpenPGP protocol is vulnerable to a cryptographic attack when using symmetric encryption in an automated way.

David Shaw reports about the impact:

This attack, while very significant from a cryptographic point of view, is not generally effective in the real world. To be specific, unless you have your OpenPGP program set up as part of an automated system to accept encrypted messages, decrypt them, and then provide a response to the submitter, then this does not affect you at all.

Note that the fix in GnuPG does note completely eliminate the potential problem:

These patches disable a portion of the OpenPGP protocol that the attack is exploiting. This change should not be user visible. With the patch in place, this attack will not work using a public-key encrypted message. It will still work using a passphrase-encrypted message.

Georgi Guninski discovered a way to construct Vim modelines that execute arbitrary shell commands. The vulnerability can be exploited by including shell commands in modelines that call the glob() or expand() functions. An attacker could trick an user to read or edit a trojaned file with modelines enabled, after which the attacker is able to execute arbitrary commands with the privileges of the user.

Note: It is generally recommended that VIM users use set nomodeline in ~/.vimrc to avoid the possibility of trojaned text files.

A Gentoo Linux Security Advisory reports:

Tavis Ormandy of the Gentoo Linux Security Audit Team discovered a stack based buffer overflow in the libTIFF library when reading a TIFF image with a malformed BitsPerSample tag.

Successful exploitation would require the victim to open a specially crafted TIFF image, resulting in the execution of arbitrary code.

A Secunia Advisory reports:

Secunia Research has discovered a vulnerability in Opera, which can be exploited by malicious people to conduct cross-site scripting attacks and retrieve a user's files.

The vulnerability is caused due to Opera allowing a user to drag e.g. an image, which is actually a "javascript:" URI, resulting in cross-site scripting if dropped over another site. This may also be used to populate a file upload form, resulting in uploading of arbitrary files to a malicious web site.

Successful exploitation requires that the user is tricked into dragging and dropping e.g. an image or a link.

A Secunia Advisory reports:

Secunia Research has discovered a vulnerability in Opera, which can be exploited by malicious people to trick users into executing malicious files.

The vulnerability is caused due to an error in the handling of extended ASCII codes in the download dialog. This can be exploited to spoof the file extension in the file download dialog via a specially crafted "Content-Disposition" HTTP header.

Successful exploitation may result in users being tricked into executing a malicious file via the download dialog, but requires that the "Arial Unicode MS" font (ARIALUNI.TTF) has been installed on the system.

An Ethreal Security Advisories reports:

Our testing program has turned up several more security issues:

• The LDAP dissector could free static memory and crash.
• The AgentX dissector could crash.
• The 802.3 dissector could go into an infinite loop.
• The PER dissector could abort.
• The DHCP dissector could go into an infinite loop.
• The BER dissector could abort or loop infinitely.
• The MEGACO dissector could go into an infinite loop.
• The GIOP dissector could dereference a null pointer.
• The SMB dissector was susceptible to a buffer overflow.
• The WBXML could dereference a null pointer.
• The H1 dissector could go into an infinite loop.
• The DOCSIS dissector could cause a crash.
• The SMPP dissector could go into an infinite loop.
• SCTP graphs could crash.
• The HTTP dissector could crash.
• The SMB dissector could go into a large loop.
• The DCERPC dissector could crash.
• Several dissectors could crash while reassembling packets.

Steve Grubb at Red Hat found the following issues:

• The CAMEL dissector could dereference a null pointer.
• The DHCP dissector could crash.
• The CAMEL dissector could crash.
• The PER dissector could crash.
• The RADIUS dissector could crash.
• The Telnet dissector could crash.
• The IS-IS LSP dissector could crash.
• The NCP dissector could crash.

iDEFENSE found the following issues:

• Several dissectors were susceptible to a format string overflow.

Impact:

It may be possible to make Ethereal crash, use up available memory, or run arbitrary code by injecting a purposefully malformed packet onto the wire or by convincing someone to read a malformed packet trace file.

A Watchfire whitepaper reports an vulnerability in the Apache webserver. The vulnerability can be exploited by malicious people causing cross site scripting, web cache poisoining, session hijacking and most importantly the ability to bypass web application firewall protection. Exploiting this vulnerability requires multiple carefully crafted HTTP requests, taking advantage of an caching server, proxy server, web application firewall etc. This only affects installations where Apache is used as HTTP proxy in combination with the following web servers:

• IIS/6.0 and 5.0
• Apache 2.0.45 (as web server)
• apache 1.3.29
• WebSphere 5.1 and 5.0
• WebLogic 8.1 SP1
• Oracle9iAS web server 9.0.2
• SunONE web server 6.1 SP4

An Secunia Advisory reports:

Neel Mehta and Alex Wheeler have reported some vulnerabilities in Clam AntiVirus, which can be exploited by malicious people to cause a DoS (Denial of Service) or compromise a vulnerable system.

1. Two integer overflow errors in "libclamav/tnef.c" when processing TNEF files can be exploited to cause a heap-based buffer overflow via a specially crafted TNEF file with a length value of -1 in the header.
2. An integer overflow error in "libclamav/chmunpack.c" can be exploited to cause a heap-based buffer overflow via a specially crafted CHM file with a chunk entry that has a filename length of -1.
3. A boundary error in "libclamav/fsg.c" when processing a FSG compressed file can cause a heap-based buffer overflow.

The ISC DHCP programs are vulnerable to several format string vulnerabilities which may allow a remote attacker to execute arbitrary code with the permissions of the DHCP programs, typically root for the DHCP server.

Multiple cross-site scripting (XSS) vulnerabilities in eGroupware before 1.0.0.007 allow remote attackers to inject arbitrary web script or HTML via the (1) ab_id, (2) page, (3) type, or (4) lang parameter to index.php or (5) category_id parameter.

Multiple SQL injection vulnerabilities in index.php in eGroupware before 1.0.0.007 allow remote attackers to execute arbitrary SQL commands via the (1) filter or (2) cats_app parameter.

In fetchmail 6.2.5.1, the the remote code injection via POP3 UIDL was fixed, but a denial of service attack was introduced:

Two possible NULL-pointer dereferences allow a malicous POP3 server to crash fetchmail by respondig with UID lines containing only the article number but no UID (in violation of RFC-1939), or a message without Message-ID when no UIDL support is available.

Natanael Copa reports that dnrd is vulnerable to a remote buffer overflow and a remote stack overflow. These vulnerabilities can be triggered by sending invalid DNS packets to dnrd.

The buffer overflow could potentially be used to execute arbitrary code with the permissions of the dnrd daemon. Note that dnrd runs in an chroot environment and runs as non-root.

The stack overflow vulnerability can cause dnrd to crash.

The LDAP backend in PowerDNS has issues with escaping queries which could cause connection errors. This would make it possible for a malicious user to temporarily blank domains.

This is known to affect all releases prior to 2.9.18.

fetchmail's POP3/UIDL code does not truncate received UIDs properly. A malicious or compromised POP3 server can thus corrupt fetchmail's stack and inject code when fetchmail is using UIDL, either through configuration, or as a result of certain server capabilities. Note that fetchmail is run as root on some sites, so an attack might compromise the root account and thus the whole machine.

A KDE Security Advisory explains:

Kate / Kwrite create 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.

Depending on the system security settings, backup files might be readable by other users. Kate / Kwrite are network transparent applications and therefore this vulnerability might not be restricted to local users.

The Mozilla Foundation reports of multiple security vulnerabilities in Firefox and Mozilla:

• MFSA 2005-56 Code execution through shared function objects
• MFSA 2005-55 XHTML node spoofing
• MFSA 2005-54 Javascript prompt origin spoofing
• MFSA 2005-53 Standalone applications can run arbitrary code through the browser
• MFSA 2005-52 Same origin violation: frame calling top.focus()
• MFSA 2005-51 The return of frame-injection spoofing
• MFSA 2005-50 Possibly exploitable crash in InstallVersion.compareTo()
• MFSA 2005-49 Script injection from Firefox sidebar panel using data:
• MFSA 2005-48 Same-origin violation with InstallTrigger callback
• MFSA 2005-47 Code execution via "Set as Wallpaper"
• MFSA 2005-46 XBL scripts ran even when Javascript disabled
• MFSA 2005-45 Content-generated event vulnerabilities

Kuba Zygmunt discovered a flaw in the input validation routines of Drupal's filter mechanism. An attacker could execute arbitrary PHP code on a target site when public comments or postings are allowed.

A Securityreason.com advisory reports that various cross site scripting vulnerabilities have been found in phpSysInfo. Input is not properly sanitised before it is returned to the user. A malicious person could exploit this to execute arbitrary HTML and script code in a users browser session. Also it is possible to view the full path of certain scripts by accessing them directly.

A Zataz advisory reports that MySQL contains a security flaw which could allow a malicious local user to inject arbitrary SQL commands during the initial database creation process.

The problem lies in the mysql_install_db script which creates temporary files based on the PID used by the script.

A Gentoo advisory reports:

Net-SNMP creates temporary files in an insecure manner, possibly allowing the execution of arbitrary code.

A malicious local attacker could exploit a race condition to change the content of the temporary files before they are executed by fixproc, possibly leading to the execution of arbitrary code. A local attacker could also create symbolic links in the temporary files directory, pointing to a valid file somewhere on the filesystem. When fixproc is executed, this would result in the file being overwritten.

phpBB is vulnerable to remote exploitation of an input validation vulnerability allows attackers to read the contents of arbitrary system files under the privileges of the webserver. This also allows remote attackers to unlink arbitrary system files under the privileges of the webserver.

A Zataz advisory reports that shtool contains a security flaw which could allow a malicious local user to create or overwrite the contents of arbitrary files. The attacker could fool a user into executing the arbitrary file possibly executing arbitrary code.

A Secunia Advisory reports:

A vulnerability has been reported in phpPgAdmin, which can be exploited by malicious people to disclose sensitive information.

Input passed to the "formLanguage" parameter in "index.php" isn't properly verified, before it is used to include files. This can be exploited to include arbitrary files from local resources.

Successful exploitation requires that "magic_quotes_gpc" is disabled.

The pear-XML_RPC release notes reports that the following issues has been fixed:

Eliminate path disclosure vulnerabilities by suppressing error messages when eval()'ing.

Eliminate path disclosure vulnerability by catching bogus parameters submitted to XML_RPC_Value::serializeval().

Eric Romang reports that ekg creates temporary files in an insecure manner. This can be exploited by an attacker using a symlink attack to overwrite arbitrary files and possibly execute arbitrary commands with the permissions of the user running ekg.

A Bugzilla Security Advisory reports:

Any user can change any flag on any bug, even if they don't have access to that bug, or even if they can't normally make bug changes. This also allows them to expose the summary of a bug.

Bugs are inserted into the database before they are marked as private, in Bugzilla code. Thus, MySQL replication can lag in between the time that the bug is inserted and when it is marked as private (usually less than a second). If replication lags at this point, the bug summary will be accessible to all users until replication catches up. Also, on a very slow machine, there may be a pause longer than a second that allows users to see the title of the newly-filed bug.

Insecure file permissions, network access control and DNS usage put systems that use Legato NetWorker at risk.

When the software is running, several files that contain sensitive information are created with insecure permissions. The information exposed include passwords and can therefore be used for privilege elevation.

An empty "servers" file, which should normally contain hostnames of authorized backup servers, may allow unauthorized backups to be made. Sensitive information can be extracted from these backups.

When reverse DNS fails for the Legato client IP a weak authorization scheme, containing a flaw that allows unauthorized access, is used. This may allow unauthorized access.

Secunia Research reports:

Secunia has discovered a security issue in Adobe Reader for Linux, which can be exploited by malicious, local users to gain knowledge of sensitive information.

The problem is caused due to temporary files being created with permissions based on a user's umask in the "/tmp" folder under certain circumstances when documents are opened.

Successful exploitation allows an unprivileged user to read arbitrary users' documents.

An iDEFENSE Security Advisory reports:

Remote exploitation of an input validation error in Clam AntiVirus ClamAV allows attackers to cause a denial of service condition.

The vulnerability specifically exists due to insufficient validation on cabinet file header data. The ENSURE_BITS() macro fails to check for zero length reads, allowing a carefully constructed cabinet file to cause an infinite loop.

ClamAV is used in a number of mail gateway products. Successful exploitation requires an attacker to send a specially constructed CAB file through a mail gateway or personal anti-virus client utilizing the ClamAV scanning engine. The infinate loop will cause the ClamAV software to use all available processor resources, resulting in a denial of service or severe degradation to system performance. Remote exploitation can be achieved by sending a malicious file in an e-mail message or during an HTTP session.

An iDEFENSE Security Advisory reports:

Remote exploitation of an input validation error in Clam AntiVirus ClamAV allows attackers to cause a denial of service condition.

The vulnerability specifically exists due to improper behavior during exceptional conditions.

Successful exploitation allows attackers to exhaust file descriptors pool and memory. Anti-virus detection functionality will fail if there is no file descriptors available with which to open files. Remote exploitation can be achieved by sending a malicious file in an e-mail message or during an HTTP session.

Problem Description

An error in the handling of corrupt compressed data streams can result in a buffer being overflowed.

Impact

By carefully crafting a corrupt compressed data stream, an attacker can overwrite data structures in a zlib-using application. This may cause the application to halt, causing a denial of service; or it may result in the attacker gaining elevated privileges.

An Adobe Security Advisory reports:

A vulnerability within Adobe Reader has been identified. Under certain circumstances, remote exploitation of a buffer overflow in Adobe Reader could allow an attacker to execute arbitrary code.

If exploited, it could allow the execution of arbitrary code under the privileges of the local user. Remote exploitation is possible if the malicious PDF document is sent as an email attachment or if the PDF document is accessed via a web link.

A Net-SNMP release announcement reports:

A security vulnerability has been found in Net-SNMP releases that could allow a denial of service attack against Net-SNMP agent's which have opened a stream based protocol (EG, TCP but not UDP; it should be noted that Net-SNMP does not by default open a TCP port).

Stefan Esser reports:

Wrongly implemented user input filters lead to multiple SQL Injection vulnerabilities which can lead f.e. to disclosure of the admin password hash.

Wrongly implemented user input filters allows injection of user input into executed commandline.

Alberto Trivero posted his Remote Command Execution Exploit for Cacti <= 0.8.6d to Bugtraq on the 22th June. Having analysed his bug we come to the conclusion, that the malfunctioning input filters, which were already mentioned in the previous advisory are also responsible for this bug still being exploitable.

A HTTP headers bypass switch can also be used to completely bypass the authentification system of Cacti. As admin it is possible to execute shell commands with the permission of the webserver.

While looking at the source of Cacti a HTTP headers bypass switch was discovered, that also switches off a call to session_start() and the manual application of addslashes() in case of magic_quotes_gpc=Off.

When register_globals is turned on* an attacker can use this switch to disables Cacti's use of PHP's session support and therefore supply the session variables on his own through f.e. the URL. Additionally using the switch renders several SQL statements vulnerable to SQL Injections attacks, when magic_quotes_gpc is turned off, which is the recommended setting.

Logged in as an admin it is possible to issue shell commands.

(*) register_globals is turned off by default since PHP 4.2 but is activated on most servers because of older scripts requiring it.

GulfTech Security Research reports:

There are a number of vulnerabilities in WordPress that may allow an attacker to ultimately run arbitrary code on the vulnerable system. These vulnerabilities include SQL Injection, Cross Site Scripting, and also issues that may aid an attacker in social engineering.

A Gentoo Linux Security Advisory reports:

Due to a lack of input validation, WordPress is vulnerable to SQL injection and XSS attacks.

An attacker could use the SQL injection vulnerabilites to gain information from the database. Furthermore the cross-site scripting issues give an attacker the ability to inject and execute malicious script code or to steal cookie-based authentication credentials, potentially compromising the victim's browser.

FrSIRT Advisory reports:

A vulnerability was identified in phpBB, which may be exploited by attackers to compromise a vulnerable web server. This flaw is due to an input validation error in the "viewtopic.php" script that does not properly filter the "highlight" parameter before calling the "preg_replace()" function, which may be exploited by remote attackers to execute arbitrary PHP commands with the privileges of the web server.

GulfTech Security Research Team reports:

PEAR XML_RPC is vulnerable to a very high risk php code injection vulnerability due to unsanatized data being passed into an eval() call.

Problem Description

The ipfw tables lookup code caches the result of the last query. The kernel may process multiple packets concurrently, performing several concurrent table lookups. Due to an insufficient locking, a cached result can become corrupted that could cause some addresses to be incorrectly matched against a lookup table.

Impact

When lookup tables are used with ipfw, packets may on very rare occasions incorrectly match a lookup table. This could result in a packet being treated contrary to the defined packet filtering ruleset. For example, a packet may be allowed to pass through when it should have been discarded.

The problem can only occur on Symmetric Multi-Processor (SMP) systems, or on Uni Processor (UP) systems with the PREEMPTION kernel option enabled (not the default).

Workaround

a) Do not use lookup tables.

OR

b) Disable concurrent processing of packets in the network stack by setting the "debug.mpsafenet=0" tunable:

# echo "debug.mpsafenet=0" << /boot/loader.conf

Problem Description

Two problems have been discovered relating to the extraction of bzip2-compressed files. First, a carefully constructed invalid bzip2 archive can cause bzip2 to enter an infinite loop. Second, when creating a new file, bzip2 closes the file before setting its permissions.

Impact

The first problem can cause bzip2 to extract a bzip2 archive to an infinitely large file. If bzip2 is used in automated processing of untrusted files this could be exploited by an attacker to create an denial-of-service situation by exhausting disk space or by consuming all available cpu time.

The second problem can allow a local attacker to change the permissions of local files owned by the user executing bzip2 providing that they have write access to the directory in which the file is being extracted.

Workaround

Do not uncompress bzip2 archives from untrusted sources and do not uncompress files in directories where untrusted users have write access.

Problem Description

Two problems have been discovered in the FreeBSD TCP stack.

First, when a TCP packets containing a timestamp is received, inadequate checking of sequence numbers is performed, allowing an attacker to artificially increase the internal "recent" timestamp for a connection.

Second, a TCP packet with the SYN flag set is accepted for established connections, allowing an attacker to overwrite certain TCP options.

Impact

Using either of the two problems an attacker with knowledge of the local and remote IP and port numbers associated with a connection can cause a denial of service situation by stalling the TCP connection. The stalled TCP connection my be closed after some time by the other host.

Workaround

In some cases it may be possible to defend against these attacks by blocking the attack packets using a firewall. Packets used to effect either of these attacks would have spoofed source IP addresses.

An Ethreal Security Advisories reports:

An aggressive testing program as well as independent discovery has turned up a multitude of security issues:

• The ANSI A dissector was susceptible to format string vulnerabilities. Discovered by Bryan Fulton.
• The GSM MAP dissector could crash.
• The AIM dissector could cause a crash.
• The DISTCC dissector was susceptible to a buffer overflow. Discovered by Ilja van Sprundel
• The FCELS dissector was susceptible to a buffer overflow. Discovered by Neil Kettle
• The SIP dissector was susceptible to a buffer overflow. Discovered by Ejovi Nuwere.
• The KINK dissector was susceptible to a null pointer exception, endless looping, and other problems.
• The LMP dissector was susceptible to an endless loop.
• The Telnet dissector could abort.
• The TZSP dissector could cause a segmentation fault.
• The WSP dissector was susceptible to a null pointer exception and assertions.
• The 802.3 Slow protocols dissector could throw an assertion.
• The BER dissector could throw assertions.
• The SMB Mailslot dissector was susceptible to a null pointer exception and could throw assertions.
• The H.245 dissector was susceptible to a null pointer exception.
• The Bittorrent dissector could cause a segmentation fault.
• The SMB dissector could cause a segmentation fault and throw assertions.
• The Fibre Channel dissector could cause a crash.
• The DICOM dissector could attempt to allocate large amounts of memory.
• The MGCP dissector was susceptible to a null pointer exception, could loop indefinitely, and segfault.
• The RSVP dissector could loop indefinitely.
• The DHCP dissector was susceptible to format string vulnerabilities, and could abort.
• The SRVLOC dissector could crash unexpectedly or go into an infinite loop.
• The EIGRP dissector could loop indefinitely.
• The ISIS dissector could overflow a buffer.
• The CMIP, CMP, CMS, CRMF, ESS, OCSP, PKIX1Explitit, PKIX Qualified, and X.509 dissectors could overflow buffers.
• The NDPS dissector could exhaust system memory or cause an assertion, or crash.
• The Q.931 dissector could try to free a null pointer and overflow a buffer.
• The IAX2 dissector could throw an assertion.
• The ICEP dissector could try to free the same memory twice.
• The MEGACO dissector was susceptible to an infinite loop and a buffer overflow.
• The DLSw dissector was susceptible to an infinite loop.
• The RPC dissector was susceptible to a null pointer exception.
• The NCP dissector could overflow a buffer or loop for a large amount of time.
• The RADIUS dissector could throw an assertion.
• The GSM dissector could access an invalid pointer.
• The SMB PIPE dissector could throw an assertion.
• The L2TP dissector was susceptible to an infinite loop.
• The SMB NETLOGON dissector could dereference a null pointer.
• The MRDISC dissector could throw an assertion.
• The ISUP dissector could overflow a buffer or cause a segmentation fault.
• The LDAP dissector could crash.
• The TCAP dissector could overflow a buffer or throw an assertion.
• The NTLMSSP dissector could crash.
• The Presentation dissector could overflow a buffer.
• Additionally, a number of dissectors could throw an assertion when passing an invalid protocol tree item length.

Roger Dingledine reports:

The Tor 0.1.0.10 release from a few days ago includes a fix for a bug that might allow an attacker to read arbitrary memory (maybe even keys) from an exit server's process space. We haven't heard any reports of exploits yet, but hey.

An iDEFENSE Security Advisory reports:

Remote exploitation of a heap-based buffer overflow vulnerability in the RealText file format parser within various versions of RealNetworks Inc.'s RealPlayer could allow attackers to execute arbitrary code.

Nobuhiro IMAI reports:

the default value modification on Module#public_instance_methods (from false to true) breaks s.add_handler(XMLRPC::iPIMethods("sample"), MyHandler.new) style security protection.

This problem could allow a remote attacker to execute arbitrary commands on XMLRPC server of libruby.

iDEFENSE security group disclosed potential SQL injection attacks from unchecked user input and two security holes regarding potential cross site scripting attacks

A Secunia Advisory reports:

Secunia Research has discovered a vulnerability in Opera, which can be exploited by malicious people to steal content or to perform actions on other web sites with the privileges of the user.

Normally, it should not be possible for the XMLHttpRequest object to access resources from outside the domain of which the object was opened. However, due to insufficient validation of server side redirects, it is possible to circumvent this restriction.

A Secunia Advisory reports:

Secunia Research has discovered a vulnerability in Opera, which can be exploited by malicious people to conduct cross-site scripting attacks and to read local files.

The vulnerability is caused due to Opera not properly restricting the privileges of "javascript:" URLs when opened in e.g. new windows or frames.

A Secunia Advisory reports:

Secunia Research has discovered a vulnerability in Opera, which can be exploited by malicious people to conduct cross-site scripting attacks against users.

The vulnerability is caused due to input not being sanitised, when Opera generates a temporary page for displaying a redirection when "Automatic redirection" is disabled (not default setting).

Todd C. Miller reports:

A race condition in Sudo's command pathname handling prior to Sudo version 1.6.8p9 that could allow a user with Sudo privileges to run arbitrary commands.

Exploitation of the bug requires that the user be allowed to run one or more commands via Sudo and be able to create symbolic links in the filesystem. Furthermore, a sudoers entry giving another user access to the ALL pseudo-command must follow the user's sudoers entry for the race to exist.

Stefan Esser reports:

Trac's wiki and ticket systems allows to add attachments to wiki entries and bug tracker tickets. These attachments are stored within directories that are determined by the id of the corresponding ticket or wiki entry.

Due to a missing validation of the id parameter it is possible for an attacker to supply arbitrary paths to the upload and attachment viewer scripts. This means that a potential attacker can retrieve any file accessible by the webserver user.

Additionally it is possible to upload arbitrary files (up to a configured file length) to any place the webserver has write access too.

For obvious reasons this can lead to the execution of arbitrary code if it possible to upload files to the document root or it's subdirectories. One example of a configuration would be f.e. running Trac and s9y/wordpress with writeable content directories on the same webserver.

Another potential usage of this exploit would be to abuse Trac powered webservers as storage for f.e. torrent files.

A Secunia security advisory reports:

Two vulnerabilities have been reported in Razor-agents, which can be exploited by malicious people to cause a DoS (Denial of Service).

1. An unspecified error in the preprocessing of certain HTML messages can be exploited to crash the application.
2. A bug in the discovery logic causes Razor-agents to go into an infinite loop and consume a large amount of memory when discovery fails.

Apache SpamAssassin Security Team reports:

Apache SpamAssassin 3.0.4 was recently released, and fixes a denial of service vulnerability in versions 3.0.1, 3.0.2, and 3.0.3. The vulnerability allows certain misformatted long message headers to cause spam checking to take a very long time.

While the exploit has yet to be seen in the wild, we are concerned that there may be attempts to abuse the vulnerability in the future. Therefore, we strongly recommend all users of these versions upgrade to Apache SpamAssassin 3.0.4 as soon as possible.

A SquirrelMail Security Advisory reports:

Several cross site scripting (XSS) vulnerabilities have been discovered in SquirrelMail versions 1.4.0 - 1.4.4.

The vulnerabilities are in two categories: the majority can be exploited through URL manipulation, and some by sending a specially crafted email to a victim. When done very carefully, this can cause the session of the user to be hijacked.

Sverre H. Huseby discovered a vulnerability in Adobe Acrobat and Adobe Reader. Under certain circumstances, using XML scripts it is possible to discover the existence of local files.

Problem Description

Two problems related to extraction of files exist in gzip:

The first problem is that gzip does not properly sanitize filenames containing "/" when uncompressing files using the -N command line option.

The second problem is that gzip does not set permissions on newly extracted files until after the file has been created and the file descriptor has been closed.

Impact

The first problem can allow an attacker to overwrite arbitrary local files when uncompressing a file using the -N command line option.

The second problem can allow a local attacker to change the permissions of arbitrary local files, on the same partition as the one the user is uncompressing a file on, by removing the file the user is uncompressing and replacing it with a hardlink before the uncompress operation is finished.

Workaround

Do not use the -N command line option on untrusted files and do not uncompress files in directories where untrusted users have write access.

Problem Description

Several tcpdump protocol decoders contain programming errors which can cause them to go into infinite loops.

Impact

An attacker can inject specially crafted packets into the network which, when processed by tcpdump, could lead to a denial-of-service. After the attack, tcpdump would no longer capture traffic, and would potentially use all available processor time.

Jacopo Ottaviani reports that Gaim can be crashed by being offered files with names containing non-ASCII characters via the Yahoo! protocol.

The GAIM team reports:

Remote attackers can cause a denial of service (crash) via a malformed MSN message that leads to a memory allocation of a large size, possibly due to an integer signedness error.

A web server running Gallery can be exploited for arbitrary PHP code execution through the use of a maliciously crafted URL.

Gallery includes several cross-site scripting vulnerabilities that could allow malicious content to be injected.

A KDE Security Advisory explains:

Overview

KStars includes support for the Instrument Neutral Distributed Interface (INDI). The build system of this extra 3rd party software contained an installation hook to install fliccd (part of INDI) as SUID root application.

Erik Sjölund discovered that the code contains several vulnerabilities that allow stack based buffer overflows.

Impact

If the fliccd binary is installed as suid root, it enables root privilege escalation for local users, or, if the daemon is actually running (which it does not by default) and is running as root, remote root privilege escalation.

3APA3A reports:

If programmer fails to check socket number before using select() or fd_set macros, it's possible to overwrite memory behind fd_set structure. Very few select() based application actually check FD_SETSIZE value. [...]

Depending on vulnerable application it's possible to overwrite portions of memory. Impact is close to off-by-one overflows, code execution doesn't seems exploitable.

Matthias Andree reports:

A vulnerability was found in the fetchnews program (the NNTP client) that may under some circumstances cause a wait for input that never arrives, fetchnews "hangs". [...]

As only one fetchnews program can run at a time, subsequently started fetchnews and texpire programs will terminate. [...]

Upgrade your leafnode package to version 1.11.3.

An STG Security Advisory reports:

GForge CVS module made by Dragos Moinescu and another module made by Ronald Petty have a directory traversal vulnerability. [...] malicious attackers can read arbitrary directory lists.

The CRAM-MD5 authentication support of the University of Washington IMAP and POP3 servers contains a vulnerability that may allow an attacker to bypass authentication and impersonate arbitrary users. Only installations with CRAM-MD5 support configured are affected.

The Squid team reported several denial-of-service vulnerabilities related to the handling of DNS responses and NT Lan Manager messages. These may allow an attacker to crash the Squid cache.

Sebastian Krahmer discovered that the racoon ISAKMP daemon could be crashed with a maliciously crafted UDP packet. No authentication is required in order to perform the attack.

Tavis Ormandy discovered several integer overflows in xli's image size handling. A maliciously crafted image may be able to cause a heap buffer overflow and execute arbitrary code.

Tavis Ormandy discovered that xli and xloadimage attempt to decompress images by piping them through gunzip or similar decompression tools. Unfortunately, the unsanitized file name is included as part of the command. This is dangerous, as in some situations, such as mailcap processing, an attacker may control the input file name. As a result, an attacker may be able to cause arbitrary command execution.

In 2001, zen-parse discovered a buffer overflow in xloadimage's FACES image loader. A maliciously crafted image could cause xloadimage to execute arbitrary code. A published exploit exists for this vulnerability.

In 2005, Rob Holland discovered that the same vulnerability was present in xli.

Stanislav Brabec discovered errors in yamt's path name handling that lead to buffer overflows and directory traversal issues. When processing a file with a maliciously crafted ID3 tag, yamt might overwrite arbitrary files or possibly execute arbitrary code.

The SuSE package ChangeLog contains:

• Several security fixes (#49337):
• directory traversal in rename
• directory traversal in sort
• buffer overflow in sort
• buffer overflow in rename

A Debian Security Advisory reports:

Erik Sjölund discovered that programs linked against xview are vulnerable to a number of buffer overflows in the XView library. When the overflow is triggered in a program which is installed setuid root a malicious user could perhaps execute arbitrary code as privileged user.

The X display locking program xtrlock contains an integer overflow bug. It is possible for an attacker with physical access to the system to bypass the display lock.

Trevor Johnson reported that the Red Hat Linux RPMs used by linux_base contained multiple older vulnerabilities, such as a DNS resolver issue and critical bugs in X font handling and XPM image handling.

A SquirrelMail Security Advisory reports:

SquirrelMail 1.4.4 has been released to resolve a number of security issues disclosed below. It is strongly recommended that all running SquirrelMail prior to 1.4.4 upgrade to the latest release.

Remote File Inclusion

Manoel Zaninetti reported an issue in src/webmail.php which would allow a crafted URL to include a remote web page. This was assigned CAN-2005-0103 by the Common Vulnerabilities and Exposures.

Cross Site Scripting Issues

A possible cross site scripting issue exists in src/webmail.php that is only accessible when the PHP installation is running with register_globals set to On. This issue was uncovered internally by the SquirrelMail Development team. This isssue was assigned CAN-2005-0104 by the Common Vulnerabilities and Exposures.

A second issue which was resolved in the 1.4.4-rc1 release was uncovered and assigned CAN-2004-1036 by the Common Vulnerabilities and Exposures. This issue could allow a remote user to send a specially crafted header and cause execution of script (such as javascript) in the client browser.

Local File Inclusion

A possible local file inclusion issue was uncovered by one of our developers involving custom preference handlers. This issue is only active if the PHP installation is running with register_globals set to On.

Erik Sjölund discovered a vulnerabilitiy in Sympa. The queue application processes messages received via aliases. It contains a buffer overflow in the usage of sprintf. In some configurations, it may allow an attacker to execute arbitrary code as the sympa user.

Florian Weimer wrote:

Mailman 2.1.5 uses weak auto-generated passwords for new subscribers. These passwords are assigned when members subscribe without specifying their own password (either by email or the web frontend). Knowledge of this password allows an attacker to gain access to the list archive even though she's not a member and the archive is restricted to members only. [...]

This means that only about 5 million different passwords are ever generated, a number that is in the range of brute force attacks -- you only have to guess one subscriber address (which is usually not that hard).

Barry Warsaw reports:

Today I am releasing Mailman 2.1.5, a bug fix release [...] This version also contains a fix for an exploit that could allow 3rd parties to retrieve member passwords. It is thus highly recommended that all existing sites upgrade to the latest version.

Oliver Karow discovered cross-site scripting issues in the Apache Jakarta Tomcat manager. The developers refer to the issues as minor.

A Secunia security advisory reports:

A vulnerability has been reported in FreeStyle Wiki and FSWikiLite, which can be exploited by malicious people to conduct script insertion attacks.

Input passed in uploaded attachments is not properly sanitised before being used. This can be exploited to inject arbitrary HTML and script code, which will be executed in a user's browser session in context of an affected site when the malicious attachment is viewed.

A Gentoo Advisory reports:

The FreeRADIUS server is vulnerable to an SQL injection attack and a buffer overflow, possibly resulting in disclosure and modification of data and Denial of Service.

A Debian Advisory reports:

Jens Steube discovered that ppxp, yet another PPP program, does not release root privileges when opening potentially user supplied log files. This can be tricked into opening a root shell.

A RST/GHC Advisory reports that there is an format string vulnerability in oops. The vulnerability can be found in the MySQL/PgSQL authentication module. Succesful exploitation may allow execution of arbitrary code.

The developers of cdrdao report that there is a potential root exploit in the software. In order to be able to succesfully exploit this vulnerability cdrdao must be installed setuid root. When succesfully exploited a local user might get escalated privileges. By default this port is not installed setuid root.

The GAIM team reports:

Potential remote denial of service bug resulting from not checking a pointer for non-NULL before passing it to strncmp, which results in a crash. This can be triggered by a remote client sending an SLP message with an empty body.

The GAIM team reports that GAIM is vulnerable to a denial-of-service vulnerability which can cause GAIM to crash:

It is possible for a remote user to overflow a static buffer by sending an IM containing a very large URL (greater than 8192 bytes) to the Gaim user. This is not possible on all protocols, due to message length restrictions. Jabber are SILC are known to be vulnerable.

Problem description and impact

When running on processors supporting Hyper-Threading Technology, it is possible for a malicious thread to monitor the execution of another thread.

Information may be disclosed to local users, allowing in many cases for privilege escalation. For example, on a multi-user system, it may be possible to steal cryptographic keys used in applications such as OpenSSH or SSL-enabled web servers.

NOTE: Similar problems may exist in other simultaneous multithreading implementations, or even some systems in the absence of simultaneous multithreading. However, current research has only demonstrated this flaw in Hyper-Threading Technology, where shared memory caches are used.

Workaround

Systems not using processors with Hyper-Threading Technology support are not affected by this issue. On systems which are affected, the security flaw can be eliminated by setting the "machdep.hlt_logical_cpus" tunable:

# echo "machdep.hlt_logical_cpus=1" >> /boot/loader.conf

The system must be rebooted in order for tunables to take effect.

Use of this workaround is not recommended on "dual-core" systems, as this workaround will also disable one of the processor cores.