Why assembly uses `xor reg, reg` instead of `mov reg, 0`

One of the most common assembly idioms is zeroing a register by XORing it with itself. Since any value XOR itself is always 0, the register is cleared without needing an immediate constant.

xor ax, ax

This works because every bit cancels itself out:

10101010 XOR 10101010 = 00000000

Although this is functionally equivalent to:

mov ax, 0

xor reg, reg is typically preferred because it encodes into fewer bytes, is recognized by modern CPUs as a register-zeroing instruction, and can avoid unnecessary dependencies on the register's previous value. It's a small optimization that has become standard practice in low-level programming.