The ud2 instruction is an architecturally undefined opcode on x86 that reliably raises an "invalid opcode" exception. Compilers use it to mark "unreachable" code, ensuring a crash if execution ever reaches it. Originally, no official undefined instruction existed, so programmers relied on byte sequences like 0F FF or 0F B9 to trigger the exception. These sequences were internally decoded as instructions with two unused parameters. When Intel introduced changes that broke 0F FF, software relying on it failed, illustrating Hyrum's Law. To provide a stable solution, Intel officially created ud2, retroactively naming the old variants ud0 and ud1. Unlike its predecessors, ud2 is a clean two-byte instruction with no parameters, avoiding decoding issues that could cause access violations instead of invalid opcode exceptions. This makes ud2 the consistent, recommended choice for triggering an intentional crash.

The naming reflects the history: 0F FF became ud0, 0F B9 became ud1, and the new official instruction was named ud2. The key advantage of ud2 is its architectural guarantee, ensuring predictable behavior regardless of memory page presence or processor implementation details.