OpenBSD and NetBSD permit usermode code to kill the display server and write to the X.Org /dev/xf86 device, which allows local users with root privileges to reduce securelevel by replacing the System Management Mode (SMM) handler via a write to an SMRAM address within /dev/xf86 (aka the video card memory-mapped I/O range), and then launching the new handler via a System Management Interrupt (SMI), as demonstrated by a write to Programmed I/O port 0xB2.
Integer overflow in banner/banner.c in FreeBSD, NetBSD, and OpenBSD might allow local users to modify memory via a long banner. NOTE: CVE and multiple third parties dispute this issue. Since banner is not setuid, an exploit would not cross privilege boundaries in normal operations. This issue is not a vulnerability
ld.so in FreeBSD, NetBSD, and possibly other BSD distributions does not remove certain harmful environment variables, which allows local users to gain privileges by passing certain environment variables to loading processes. NOTE: this issue has been disputed by a third party, stating that it is the responsibility of the application to properly sanitize the environment
Integer signedness error in the fw_ioctl (FW_IOCTL) function in the FireWire (IEEE-1394) drivers (dev/firewire/fwdev.c) in various BSD kernels, including DragonFlyBSD, FreeBSD 5.5, MidnightBSD 0.1-CURRENT before 20061115, NetBSD-current before 20061116, NetBSD-4 before 20061203, and TrustedBSD, allows local users to read arbitrary memory contents via certain negative values of crom_buf->len in an FW_GCROM command. NOTE: this issue has been labeled as an integer overflow, but it is more like an integer signedness error.
The Xsession script, as used by X Display Manager (xdm) in NetBSD before 20060212, X.Org before 20060317, and Solaris 8 through 10 before 20061006, allows local users to overwrite arbitrary files, or read another user's Xsession errors file, via a symlink attack on a /tmp/xses-$USER file.
The TCP implementation in various BSD operating systems (tcp_input.c) does not properly block connections to broadcast addresses, which could allow remote attackers to bypass intended filters via packets with a unicast link layer address and an IP broadcast address.