CS208 Introduction to Computer Systems
Monday, 13 October 2025

+ Today
- Assembly language intro & experiments
- New assignment due Friday
- (Soon: a few little additions to data representation discussions)
- (Exam: I'm hoping to return it Wednesday)

+ Compiler Explorer
- https://godbolt.org/
- C sources --> assembly language
- Settings
- A maximum function

    int maximum(int x, int y) {
        int result = x;
        if (y > x) {
            result = y;
        }
        return result;
    }

  What does the corresponding assembly look like?

- Write a function int f(int a, int b) that does:
    return a * b

- Try some more
    return a + b
    return 5 * a + 9 * b
    try other formulas

- Try an if-statement
    return the smaller of and b

    min (very similar to max)

- Try a for-loop
    return 1 + 2 + 3 + ... + a

+ Between now and Wednesday
- Get the assignment package
- Read the assignment
- Read the readme
- Start looking at the code
- See if you can solve one or two puzzles

=======

+ Layers of abstraction

+ Intel chip/instruction-set history
~1970: 4004
1970s-1980s: 8008, 8080, 8086 : 8-bit machines
    every instruction was some # of bytes
    registers were 8 bits

    registers were named a, b, c, etc.

1982: 80186 16-bit registers
    registers were named ax, bx, cx, etc. (x = "extended")

80286

1985: 80386 32-bit registers
    registers were named eax, ebx, ecx, etc. (e, x = "extended")

much later...
    64-bit registers: rax, rbx, etc. (r = register)

+ Notes on those experiments
...
- mov vs. movl vs. movq
    movl -- l = long = 32 bits (I'm sorry, not the C "long")
    movq -- q = quad = 64 bits (still sorry)
    movb -- b = byte
    etc.

  eax = 32 bits (goes with l)
  rax = 64 bits (goes with q)

- keep your eyes on eax vs. rax vs. ax etc.
- Intel vs. AT&T notation
    - oh no, the documentation is backwards!

+ Look at the assignment

+ What's going on?
- C compiles to machine language
-