CHAPTER 4 - PROCESSES and CHAPTER 5 - THREADS

1. CHAPTER 4 - PROCESSESandCHAPTER 5 - THREADS CGS 3763 - Operating System Concepts UCF, Spring 2004

2. Process Concept • An operating system executes a variety of programs: • Batch system – jobs • Time-shared systems – user programs or tasks • Textbook uses the terms job and process almost interchangeably. • For this class assume a job is a program in executable form waiting to be brought into the computer system • A process is a program in execution and includes: • Process Control Block • Program Counter • A process is created when it is assigned memory and a PCB is created (by the OS) to hold its status

3. Process States • As a process executes, it moves through “states”: • new: The job is waiting to be turned into a process • during process creation, memory is allocated for the job’s instruction and data segments and a PCB is populated. • ready: The process is waiting to be assigned the CPU • running: Instructions are being executed by the CPU • the # of processes in the running state can be no greater than the # of processors (CPUs) in the system • waiting: The process is waiting for some event to occur • often associated with explicit requests for I/O operations • terminated: The process has finished execution • resources assigned to the process are reclaimed

4. Diagram of Process State

5. Queue or State? • Some states in the previous diagram are actually queues • New or Job queue – set of all processes waiting to enter the system. • Ready queue – set of all processes residing in main memory, ready and waiting to execute. • Waiting - In this class, we have somewhat abstracted away the idea of a queue for this state. In reality, processes may be placed in a device queue to wait for access to a particular I/O device. • Processes are selected from queued states by schedulers

6. Process Schedulers • Short-term scheduler (or CPU scheduler) • selects which process in the ready queue should be executed next and allocates CPU. • invoked very frequently (milliseconds)  (must be fast). • Long-term scheduler (or job scheduler) • selects which processes should be created and brought into the ready queue from the job queue. • invoked infrequently (seconds, minutes)  (may be slow). • controls the degree of multiprogramming. • Medium-term scheduler • Helps manage process mix by swapping in/out processes

7. Addition of Medium Term Scheduling

8. Process Mix • Processes can be described as either: • I/O-bound process • spends more time doing I/O than computations • many short CPU bursts. • CPU-bound process • spends more time doing computations • few very long CPU bursts. • Need to strike a balance between the two. • Otherwise either the CPU or I/O devices underutilized.

9. Moving Between States • Events as well as schedulers can cause a process to move from one state to another • Trap - during execution, process encounters and error. OS traps error and may abnormally end (abend) the process moving it from running to terminate. • SVC(end) - process ends voluntarily and moves from running to terminate • SVC(I/O) - process requests OS perform some I/O operation. If synchronous I/O, process moves from running to waiting state. • I/O Hardware Interrupt - signals the completion of some I/O operation for which a process was waiting. Process can be moved from waiting to ready. • Timer Interrupt - signals that a process has used up its current time slice (timesharing systems). Process returned from running to ready state to await its next turn.

10. Resource Allocation • A process requires certain resources in order to execute • Memory, CPU Time, I/O Devices, etc. • Resources can be allocated in one of two ways: • Static Allocation - resources assigned at start of process, released during termination • Can cause reduction in throughput • Dynamic Allocation - resources assigned as needed while process running • Can cause deadlock • Resources can be “shared” to reduce conflicts • Example: Print spooling

11. Process Control Block (PCB) • Saves the status of a process • Process state • Program counter • CPU registers • Scheduling information • e.g., priority • Memory-management information • e.g., Base and Limit Registers • Accounting information • I/O status information • Pointer to next PCB • PCB generated during process creation

12. Process Control Block (PCB)

13. PCBs Stored by OS as Linked Lists

14. CPU Switches From Process to Process

15. Context Switching • 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. • Context switch may involve more than one change in the program counter • Process 1 executing • OS manages the switch • Process 2 starts executing • Context-switch time is overhead; the system does no useful work while switching. • Time dependent on hardware support.

16. Threads (Lightweight Process) • Used to reduce context switching overhead • Allows sharing of instructions, data, files and other resources among several related tasks • Threads also share a common PCB • Each thread has its own “thread descriptor” • Program Counter • Register Set • Stack • Control of CPU can be shared among threads associated with the same process without a full-blown context switch. • Only change of PC and registers required.

17. Single and Multithreaded Processes

18. Benefits of Threads • Responsiveness • Faster due to reduced context switching time • Process can continue doing useful work while waiting for some event (isn’t blocked) • Resource Sharing (shared memory/code/data) • Economy • can get more done with same processor • less memory required • Utilization of multiprocessor architectures • different threads can run on different processors

19. Different Types of Threads • User Level Threads • thread management done by user-level library • e.g., POSIX Pthreads, Mach C-threads, Solaris threads • Kernel Level Threads • supported by the Kernel • e.g., Windows 95/98/NT/2000, Solaris, Linux

20. Multithreading Models • Many-to-One • Many user-level threads mapped to single kernel thread. • Used on systems that do not support kernel threads. • One-to-One • Each user-level thread maps to kernel thread. • Many-to-Many Model • Allows many user level threads to be mapped to many kernel threads. • Operating system to create a sufficient number of kernel threads.

21. Many-to-One Model

22. One-to-one Model

23. Many-to-Many Model

24. Cooperating Processes • Independent processes cannot affect or be affected by the execution of another process. • Dependent processes can affect or be affected by the execution of another process • a.k.a., Cooperating Processes • Processes may cooperate for: • Information sharing • Computation speed-up (Requires 2 or more CPUs) • Modularity • Convenience

25. Interprocess Communication (IPC) • Mechanism needed for processes to communicate and to synchronize their actions. • Shared Memory • Tightly coupled systems • Single processor systems allowing overlapping base & limit registers • Mutli-threaded systems (between threads associated with same process) • Message Passing • Processes communicate with each other without resorting to shared variables. • Uses send and receive operations to pass information • Better for loosely coupled / distributed systems • Can use both mechanisms on same system

26. Message Passing • If P and Q wish to communicate, they need to: • establish a communication link between them • exchange messages via send/receive • Implementation of communication link • physical (e.g., hardware bus, highspeed network) • logical (e.g., direct vs. indirect, other logical properties) • Implementation Questions • How are links established? • Can a link be associated with more than two processes? • How many links can there be between every pair of communicating processes? • Does the link support variable or fixed size messages? • Is a link unidirectional or bi-directional?

27. Direct Communication • Processes must name each other explicitly: • send (P, message) – send message to process P • receive(Q, message) – receive message from process Q • Properties of communication link • Links are established automatically. • A link is associated with exactly one pair of communicating processes. • Between each pair there exists exactly one link. • The link may be unidirectional, but is usually bi-directional. • Changing name of process causes problems • All references to old name must be found & modified • Requires re-compilation of affected programs

28. Indirect Communication • Messages are directed to and received from mailboxes (also referred to as ports). • Each mailbox has a unique id. • Processes can communicate only if they share a mailbox. • Properties of communication link • Link established only if processes share a common mailbox. • A link may be associated with many processes. • Each pair of processes may share several communication links (requires multiple mailboxes). • Link may be unidirectional or bi-directional. • No names to change, more modular

29. Indirect Communication Operations • Create a new mailbox • User/Application can create mailboxes through shared memory • Otherwise, mailboxes created by OS at the request of a user/application process • Send and receive messages through mailbox • Destroy a mailbox • User/Application process can destroy any mailbox created in shared memory • OS can destroy mailbox at request of mailbox owner (a user/application process) • OS can destroy unused mailboxes during garbage collection

30. Indirect Communication • Mailbox sharing • P1, P2, and P3 share mailbox A. • P1, sends; P2 and P3 receive. • Who gets the message? • Solutions • Allow a link to be associated with at most two processes. • Allow only one process at a time to execute a receive operation. • Allow the system to select arbitrarily the receiver. Sender is notified who the receiver was.

31. Message Synchronization • Message passing may be either blocking or non-blocking. • Blocking is considered synchronous • Process must wait until send or receive completed • Blocking Send • Blocking Receive • Non-blocking is considered asynchronous • Process can continue executing while waiting for send or receive to complete • Non-blocking Send • Non-blocking Receive

32. Buffering • Message queues attached to the link can be implemented in one of three ways. • Zero capacity – 0 messages • Sender must wait for receiver (rendezvous). • Bounded capacity – finite length of n messages • Sender must wait if link’s message queue full. • Unbounded capacity – infinite number of messages • Sender never waits.