F-Secure Internet Gatekeeper for Linux Kernel 6.31

Multiple stack-based buffer overflows in the get_header function in header.c for LHA 1.14, as used in products such as Barracuda Spam Firewall, allow remote attackers or local users to execute arbitrary code via long directory or file names in an LHA archive, which triggers the overflow when testing or extracting the archive.

Buffer overflow in multiple F-Secure Anti-Virus products and versions for Windows and Linux, including Anti-Virus for Windows Servers 5.52 and earlier, Internet Security 2004, 2005 and 2006, and Anti-Virus for Linux Servers 4.64 and earlier, allows remote attackers to execute arbitrary code via crafted ZIP archives.

Multiple content security gateway and antivirus products allow remote attackers to bypass content restrictions via MIME messages that use multiple MIME fields with the same name, which may be interpreted differently by mail clients.

Multiple content security gateway and antivirus products allow remote attackers to bypass content restrictions via MIME messages that use whitespace in an unusual fashion, which may be interpreted differently by mail clients.

Multiple content security gateway and antivirus products allow remote attackers to bypass content restrictions via MIME messages that use malformed quoting in MIME headers, parameters, and values, including (1) fields that should not be quoted, (2) duplicate quotes, or (3) missing leading or trailing quote characters, which may be interpreted differently by mail clients.

Multiple content security gateway and antivirus products allow remote attackers to bypass content restrictions via MIME messages that use non-standard but frequently supported Content-Transfer-Encoding values such as (1) uuencode, (2) mac-binhex40, and (3) yenc, which may be interpreted differently by mail clients.

Multiple content security gateway and antivirus products allow remote attackers to bypass content restrictions via MIME messages that use non-standard separator characters, or use standard separators incorrectly, within MIME headers, fields, parameters, or values, which may be interpreted differently by mail clients.

Multiple content security gateway and antivirus products allow remote attackers to bypass content restrictions via MIME messages that use fields that use RFC2047 encoding, which may be interpreted differently by mail clients.

Multiple content security gateway and antivirus products allow remote attackers to bypass content restrictions via MIME messages that use RFC2231 encoding, which may be interpreted differently by mail clients.

Multiple content security gateway and antivirus products allow remote attackers to bypass content restrictions via MIME encapsulation that uses RFC822 comment fields, which may be interpreted as other fields by mail clients.

Multiple directory traversal vulnerabilities in LHA 1.14 allow remote attackers or local users to create arbitrary files via an LHA archive containing filenames with (1) .. sequences or (2) absolute pathnames with double leading slashes ("//absolute/path").

Multiple interpretation error in various F-Secure Anti-Virus products, including Workstation 5.43 and earlier, Windows Servers 5.50 and earlier, MIMEsweeper 5.50 and earlier, Anti-Virus for Linux Servers and Gateways 4.61 and earlier, and other products, allows remote attackers to bypass antivirus protection via a compressed file with both local and global headers set to zero, which does not prevent the compressed file from being opened on the target system.

The Content Scanner Server in F-Secure Anti-Virus for Microsoft Exchange 6.21 and earlier, F-Secure Anti-Virus for Microsoft Exchange 6.01 and earlier, and F-Secure Internet Gatekeeper 6.32 and earlier allow remote attackers to cause a denial of service (service crash due to unhandled exception) via a certain malformed packet.