I think you missed the issue. Consider the following code: jmp foo push ax // 50 foo: int 21h // CD 21 So now we want to set a

I think you missed the issue. Consider the following code: jmp foo push ax // 50 foo: int 21h // CD 21 So now we want to set a breakpoint on the "push ax". If we do it using more than a single byte, it will overwrite the int 21h instruction and the code will mess up and likely crash. That's why 1 byte is used for breakpoints.