Operating System & Memory Hierarchy i206 Fall 2010 John Chuang Some slides adapted from Glenn Brookshear, Brian Hayes, or Marti Hearst

John Chuang2 Summary of Key Concepts  Data Storage: -Bit storage using transistors -Memory hierarchy -Locality of reference and caching  Operating System: -Process management -Context switch -Threads vs. processes -Synchronization and locks

John Chuang3 Memory Hierarchy Bits & Bytes Binary Numbers Number Systems Gates Boolean Logic Circuits CPU Machine Instructions Assembly Instructions Program Algorithms Application Memory Data compression Compiler/ Interpreter Operating System Data Structures Analysis I/O Memory hierarchy Design Methodologies/ Tools Process Truth table Venn Diagram DeMorgan’s Law Numbers, text, audio, video, image, … Decimal, Hexadecimal, Binary AND, OR, NOT, XOR, NAND, NOR, etc. Register, Cache Main Memory, Secondary Storage Context switch Process vs. Thread Locks and deadlocks Op-code, operands Instruction set arch Lossless v. lossy Info entropy & Huffman code Adders, decoders, Memory latches, ALUs, etc. Data Representation Data storage Principles ALUs, Registers, Program Counter, Instruction Register Network Distributed Systems Security Cryptography Standards & Protocols Inter-process Communication Searching, sorting, Encryption, etc. Stacks, queues, maps, trees, graphs, … Big-O UML, CRC TCP/IP, RSA, … Confidentiality Integrity Authentication … C/S, P2P Caching sockets Formal models Finite automata regex

John Chuang4 Storing 1 Bit of Data  Flip-flop (also called a latch) is a circuit built from gates that can store one bit of data -Setting input A to 1 sets output to 1 -Setting input B to 1 sets output to 0 -While both input lines are 0, the most recently stored value is preserved Adapted from Brookshear Fig 1.5 output Input A Input B Example: This 8-bit memory cell can store a byte of data.

John Chuang5 Data Storage  Where and how do we store data? Brookshear Fig 2.13

John Chuang6 Data Storage  In general, we want data to be located in close proximity of their use (the CPU in this case)  Why do we want to store data at so many different locations within a computer?  Considerations of -Storage capacity -Access latency -Cost -… -Patterns of data access

John Chuang7 Memory Hierarchy Image from

John Chuang8 Storage Technologies  Characteristics -Solid State vs. Mechanical -Volatile vs. Nonvolatile -Read only vs. Read-write (RW) -Expense -Speed -Storage capacity (a function of size & expense)  Types: -Registers: Volatile, small, expensive, fast, solid state, RW -DRAM (dynamic random access memory): Volatile, now large, expensive, fast, solid state, RW -ROM (read-only memory): Nonvolatile, Read-only, cheap, fast -Hard Disk: Nonvolatile, mechanical, RW, cheap, slow -Flash Memory (aka SSD): Nonvolatile, solid state, RW, expensive, fast (Versus hard disk: silent, lighter, faster … but more expensive) -Other external storage: CD-ROM, floppy disks, magnetic tape, …

John Chuang9 Levels of a Memory Hierarchy (circa 2000)

John Chuang10 Data Storage  In general, we want data to be located in close proximity of their use (the CPU in this case)  Why do we want to store data at so many different locations within a computer?  Considerations of -Storage capacity -Access latency -Cost -… -Patterns of data access

John Chuang11 Caching  Data access patterns exhibit locality of reference -Recently referenced item more likely to be referenced again  Caches store items that have been used recently for faster access  Caching can be done at various levels of the memory hierarchy Image from ALU Registers R W R W R W L1/L2 Cache Main Memory R W Disk Storage

John Chuang12 Operating System Bits & Bytes Binary Numbers Number Systems Gates Boolean Logic Circuits CPU Machine Instructions Assembly Instructions Program Algorithms Application Memory Data compression Compiler/ Interpreter Operating System Data Structures Analysis I/O Memory hierarchy Design Methodologies/ Tools Process Truth table Venn Diagram DeMorgan’s Law Numbers, text, audio, video, image, … Decimal, Hexadecimal, Binary AND, OR, NOT, XOR, NAND, NOR, etc. Register, Cache Main Memory, Secondary Storage Context switch Process vs. Thread Locks and deadlocks Op-code, operands Instruction set arch Lossless v. lossy Info entropy & Huffman code Adders, decoders, Memory latches, ALUs, etc. Data Representation Data storage Principles ALUs, Registers, Program Counter, Instruction Register Network Distributed Systems Security Cryptography Standards & Protocols Inter-process Communication Searching, sorting, Encryption, etc. Stacks, queues, maps, trees, graphs, … Big-O UML, CRC TCP/IP, RSA, … Confidentiality Integrity Authentication … C/S, P2P Caching sockets Formal models Finite automata regex

John Chuang13 Abstract View of System Components Source: Silberschatz, Galvin, Gagne

John Chuang14 What is an Operating System?  A software program that acts as an intermediary between a user of a computer and the computer hardware  Operating system functions: -Oversee operation of computer -Store and retrieve files -Schedule programs for execution -Execute programs

John Chuang15 OS Components  Shell (command interpreter)  Kernel -Process manager -Memory manager -File manager -I/O systems manager -Networking -Scheduler/dispatcher -Etc. shell user kernel

John Chuang16 Process Management  A process is a program in execution -A process needs certain resources, including CPU time, memory, files, and I/O devices, to accomplish its task  The operating system is responsible for the following activities in connection with process management -Process creation and deletion -Process suspension and resumption -Provision of mechanisms for: -process synchronization -process communication

John Chuang17 Process State  As a process executes, it changes state -new: The process is being created. -running: Instructions are being executed. -waiting: The process is waiting for some event to occur. -ready: The process is waiting to be assigned to a process. -terminated: The process has finished execution. Source: Silberschatz, Galvin, Gagne

John Chuang18 Resource Sharing  Many processes share hardware resources -CPU, I/O (mouse, keyboard, monitor, disk, …)  Processes must take turns -Context switch (aka process switch)  Operating system serves as coordinator/ scheduler Brookshear Fig 3.6

John Chuang19 Context Switch  When CPU switches to another process, the system must save the state of the old process and load the saved state for the new process  Process control block (PCB, on right) contains information associated with each process  Context-switch time is overhead; the system does no useful work while switching -A system is said to be “thrashing” when it spends more time performing context switches than useful work Source: Silberschatz, Galvin, Gagne

John Chuang20 Threads  A thread (or lightweight process) is a basic unit of CPU utilization  Each thread has its own -program counter -stack -register set  The threads share: -text section (program) -data section (global variables) Source: Silberschatz, Galvin, Gagne

John Chuang21 Threads versus Multiple Processes  Threads are cheaper to create than processes (~10-20x)  Switching to a different thread in same process is much cheaper than switching to a different process (~5-50x) since no context switching required  Threads within same process can share data and other resources more conveniently and efficiently (than Inter-Process Communication)  Threads within a process are not protected from one another

John Chuang22 Synchronization  Scenario: two threads want to update count (current value is 4) -Thread 1: count = count + 1 -Thread 2: count = count - 1  What is final value of count ? -(a) 3 -(b) 4 -(c) 5 -(d) any of the above -(e) none of the above

John Chuang23 Synchronization  LOAD R1, XY  LOAD R2, #1  ADD R3, R1, R2  STORE R3, XY  LOAD R4, XY  LOAD R5, #-1  ADD R6, R4, R5  STORE R6, XY

John Chuang24 Synchronization  LOAD R1, XY  LOAD R2, #1  ADD R3, R1, R2  STORE R3, XY  LOAD R4, XY  LOAD R5, #-1  ADD R6, R4, R5  STORE R6, XY  LOAD R1, XY  LOAD R4, XY  LOAD R2, #1  LOAD R5, #-1  ADD R3, R1, R2  ADD R6, R4, R5  STORE R3, XY  STORE R6, XY

John Chuang25 Lesson  Even a seemingly simple operation like count++ is not atomic, and can be subject to race conditions

John Chuang26 Synchronization  Use locks to coordinate actions of competing processes  Also known as mutual exclusion (mutex)  Simplest form of lock is a semaphore -Process must acquire semaphore before entering critical section; release semaphore upon exit initialize_lock()... acquire_lock() try: change_shared_data() finally: release_lock()

John Chuang27 Summary of Key Concepts  Data Storage: -Bit storage using transistors -Memory hierarchy -Locality of reference and caching  Operating System: -Process management -Context switch -Threads vs. processes -Synchronization and locks