OpenBSD 3.4 and NetBSD 1.6 and 1.6.1 allow remote attackers to cause a denial of service (crash) by sending an IPv6 packet with a small MTU to a listening port and then issuing a TCP connect to that port.

The IPv6 URI parsing routines in the apr-util library for Apache 2.0.50 and earlier allow remote attackers to cause a denial of service (child process crash) via a certain URI, as demonstrated using the Codenomicon HTTP Test Tool.

The setsockopt call in the KAME Project IPv6 implementation, as used in FreeBSD 5.2, does not properly handle certain IPv6 socket options, which could allow attackers to read kernel memory and cause a system panic.

Buffer overflow in the getaddrinfo function in Python 2.2 before 2.2.2, when IPv6 support is disabled, allows remote attackers to execute arbitrary code via an IPv6 address that is obtained using DNS.

Apache 2 before 2.0.47, when running on an IPv6 host, allows attackers to cause a denial of service (CPU consumption by infinite loop) when the FTP proxy server fails to create an IPv6 socket.

The IPv6 capability in IRIX 6.5.19 allows remote attackers to cause a denial of service (hang) in inetd via port scanning.

Unknown vulnerability in the IPv6 capability in IRIX 6.5.19 causes snoop to process packets as the root user, with unknown implications.

The OSI dissector in Ethereal 0.9.12 and earlier allows remote attackers to cause a denial of service and possibly execute arbitrary code via invalid IPv4 or IPv6 prefix lengths, possibly triggering a buffer overflow.

Solaris 8 with IPv6 enabled allows remote attackers to cause a denial of service (kernel panic) via a crafted IPv6 packet.

Buffer overflow in ZMailer before 2.99.51_1 allows remote attackers to execute arbitrary code during HELO processing from an IPv6 address, possibly using an address that resolves to a long hostname.

Unknown vulnerability in IPV6 functionality for DCE daemons (1) dced or (2) rpcd on HP-UX 11.11 allows attackers to cause a denial of service (crash) via an attack that modifies internal data.