Last updated April 10, 2022

Status: Tags: Links: ) CMPT 295 - Introduction to Computer Systems

CMPT 295 Lectures

Principles

4 - Data Representation, Integer Arithmetic

• Data overflow from adding, truncation
• Unsigned int
  • True sum is without overflow, actual is when considering overflow
• Signed int
  • Becomes negative since the last digit is stored as negative
  • Discarding carry out is same as modular arithmetic
  • If you know true sum and want to find actual, subtract $2^w$
  • If negative + positive, then true sum = actual sum
• Carry out is the extra number not fit into memoryz
• Subtraction is addition but with a negativve number

Multiplication

• needs 2x the capacity of the valuesz

• unsigned and signed addition formulas for true and actual sums

5 - Upsizing

• When you upsize, you add 0’s to left
• If int and short have same value and you compare, short will be upgraded to an int
  • ui > us is true since ui = xbeef0000 and us = 0000beef
  • if signed though, then ui < us since ui msb turns ui negative
    • us gets padded with 1’sdddz
• turning unsigned to signed pads left with 1’s

6

• Scientific vs normalization

Normalization ?

• Move decimal until left-most 1 is on the left of it
• Exponent is number moved

Why add 1 to fracz

7

• Binary rounding rules

19

• Memory Stack
• Caller vs Callee
• Example to try

20

• Recursion
• Example to try

22

• Buffer Overflow
• Canary Value
• Code Injection Attack

23

• Floating Point Data
• Microprocessor
• Instruction Set Architecture (ISA)

24

MIPS

26

• Creating an ISA (x295M)

29

• Microprocessor Sequential Execution

30

• Clock register
• Microprocessor Staged Execution

31

• Microprocessor Pipeline Execution

33

• Code Optimization

35

• Microprocessor Superscalar Execution

36

• Locality

References:

Created:: 2022-03-03 20:43