1. 1 Chapter 3: Processes  Process Concept  Process Scheduling  Operations on Processes  Cooperating Processes  Interprocess Communication  Communication in Client-Server Systems

2. 2 Process Management  A program in a computer system is know as:  Batch system – job  Time-shared systems – user program or task or process  The terms job and process and task is almost interchangeable.  Process – “a program in execution”; “unit of work in a time sharing system”. A program must reside in memory in order to run  Program (passive) VS. Process (active). A process needs certain resources, including CPU time, memory, files, and I/O devices, to accomplish its task.  Several processes can run the same program at the same time.  In what condition this can happen?

3. 3 Process Concept  The operating system is responsible for the following activities in connection with process management.  Process creation and deletion.  Providing identity for the process  process suspension and resumption.  To balance the work load of the system  Provision of mechanisms for:  process synchronization  process communication

4. 4 Processes The Process Model  Multiprogramming of four programs  Conceptual model of 4 independent, sequential processes  Only one program active at any instant

5. 5 Process State  As a process executes, it changes state  new: The process is being created.  Resources are allocated by OS  running: Instructions are being executed.  Only one process is in this state (in uni-processor system)  waiting: The process is waiting for some event to occur.  I/O operation or OS events  ready: The process is waiting to be assigned to a CPU.  Different queue scheduling exist  terminated: The process has finished execution.  Resources are reclaimed by OS

6. 6 Diagram of Process State Queue of jobs Queue of procs Queues of procs

7. 7 Process Control Block (PCB) Process Identity Registers: ACC, Stack, Index General purpose, Flags (PSW)

8. 8 Process Control Block (PCB) Information associated with each process. A PCB is the identity of a process.  Process state  Process id  Program counter  CPU registers  CPU scheduling information  Memory-management information  Accounting information  I/O status information

9. 9 Details of PCB used for Implementation of Processes Fields of a process table entry

10. 10 CPU Switch From Process to Process Cpu scheduler For interrupt

11. 11 Process Scheduling Queues  Job queue – set of all processes in the system.  Ready queue – set of all processes residing in main memory, ready and waiting to execute.  Device queues – set of processes waiting for an I/O device.  Process migration between the various queues.

12. 12 Ready Queue And Various I/O Device Queues Ready queue Device Queues: processes are waiting to be signaled

13. 13 Representation of Process Scheduling Cause of releasing CPU by process Signal is generated by

14. 14 Schedulers  Long-term scheduler (or job scheduler) – selects which processes should be brought into the ready queue, to balance between I/O-bound and CPU-bound processes  Short-term scheduler (or CPU scheduler) – selects which process should be executed next and allocates CPU  Fairness; priority  Medium-term scheduler – swaps the processes In and out between memory and swap-device in response to the change of characteristics of the processes in using memory or I/O devices.

15. 15 Addition of Medium Term Scheduling

16. 16 Schedulers (Cont.)  Short-term scheduler is invoked very frequently (milliseconds)  (must be fast).  Long-term scheduler is invoked very infrequently (seconds, minutes)  (may be slow).  The long-term scheduler controls the degree of multiprogramming.  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.  What happens if all processes are I/O-bound? Or CPU-bound?

17. Process Suspension Auxiliary Slides from the Stallings Book 17

18. 18 A tree of processes on a typical Solaris

19. 19 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.  Context-switch time is overhead; the system does no useful work while switching.  Hardware can help a lot to reduce the context switch time  Time dependent on hardware support.  Multiple sets of registers  Saving the address space  Threads as solution to context switch  overhead

20. 20 Process Creation  Parent process create children processes, which, in turn create other processes, forming a tree of processes.  Resource sharing  Parent and children share all resources.  Children share subset of parent’s resources.  Parent and child share no resources.  Execution  Parent and children execute concurrently.  Parent waits until children terminate.

21. 21 Process Creation (Cont.)  Address space  Child duplicate of parent.  Child has a program loaded into it.  UNIX examples  fork system call creates new process  exec system call used after a fork to replace the process’ memory space with a new program.

22. 22 Process Creation

23. 23 Process Creation in POSIX #include Int main ( ) { int pid; /* fork another process */ pid = fork (); if (pid < 0) { /*error occurred */ fprintf (stderr, “Fork Failed”); exit (-1); } else if (pid == 0) { /* child process */ execlp (“/bin/ls”, “ls”, NULL); } else { /* parent process */ /* parent will wait for the child to complete */ wait (NULL); printf (“Child Complete”); exit (0); } C program forking a separate process

24. 24 Process Creation in Win32 NOT IN EXAM

25. 25 Process Creation in Java NOT IN EXAM

26. 26 Processes Tree on a UNIX System

27. 27 “Process Status” Command ps aux  a: display information about other users' processes as well as yours.  u: display information associated with the following keywords: user, pid, %cpu, %mem, vsz, rss, tt, state, start, time and command.  x: display information about processes without controlling terminals

28. 28 Process status data RSS: the real memory (resident set) size of the process (in 1024 byte units). PID: process ID %CPU: the cpu utilization of the process; %MEM: the percentage of real memory used by this process. VSZ: virtual size in Kbytes (alias vsize) TT: an abbreviation for the pathname of the controlling terminal, if any. STAT: symbolic process state STARTED: the time the command started TIME: accumulated cpu time, user + system

29. 29 Releasing a crashed Unix terminal From FAQ: UWaterloo Question: my terminal has crashed and I can not log out from my session. How can I terminate my session?  In some occasions the terminal crashes and you do not have any control on it. Some reasons are: 1) lack of memory when you are running a memory demanding program such as ”web browser" or "matlab" on a low memory terminal, 2) bugs in an application program you run, and 3) non-functionality of the mouse, i.e., you can not activate the console window to log-off.  In such cases, resetting the terminal (or machine) is not always a good idea, since rebooting may take a long time, or you may not be allowed to reset it anyway. In order to release a crashed terminal, you should login into your account from another terminal and kill the shell process of your crashed session. Upon login, you should see the status of your processes using the process status command " ps " with appropriate switches.

30. 30 Releasing a crashed Unix terminal …  ps -ef | grep yourUserId  prints the status of all your processes on a machine. The following is an example:  % ps -ef | grep ksartipi  UID PID PPID C STIME TTY TIME CMD  ksartipi 9002 9000 0 11:10:43 pts/13 0:01 -tcsh  ksartipi 11670 11666 0 Mar 22 pts/9 0:00 -tcsh  ksartipi 7051 1 1 09:45:30 ? 0:34 emacs-20.2es  ksartipi 9092 9002 0 11:11:56 pts/13 0:00 grep ksartipi  You will see several lines, each relating to a different process. You should locate the shell process of the crashed session. This is done by looking at the process ids (PID) of different processes and recognizing your current shell process and your crashed shell process. Your current shell process has the PID number close to the other programs that you run in your new session. In the above example, the first line belongs to the currently created shell process since its PID is close to the PID of the "grep ksartipi" command (two PIDs 9002 and 9092), as opposed to the PID of the shell process in the second line (11670).  Now, you can kill the crashed shell process by the command: "kill -9 11670" at the Unix prompt and release the crashed terminal.

31. 31 Process Termination Reasons for Termination:  Process executes last statement and asks the operating system to terminate it (exit n).  Output data from child to parent (via wait PID).  Process’ resources are deallocated by operating system.  Parent may terminate execution of children processes (abort).  Child has exceeded allocated resources.  Task assigned to child is no longer required.  Parent is exiting.  Operating system does not allow child to continue if its parent terminates.  Cascading termination.

32. 32 Interprocess Communication Message PassingShared Memory

33. 33 Cooperating Processes  Independent process cannot affect or be affected by the execution of another process.  Cooperating process can affect or be affected by the execution of another process  Advantages of process cooperation  Information sharing:  Via memory  Computation speed-up  Multi-CPU system  Convenience:  Running several applications: editor; printer; compiler

34. 34 Producer-Consumer Problem  Paradigm for cooperating processes, producer process produces information that is consumed by a consumer process.  unbounded-buffer places no practical limit on the size of the buffer.  bounded-buffer assumes that there is a fixed buffer size.

35. 35 Bounded-Buffer – Shared-Memory Solution  Shared data #define BUFFER_SIZE 12 (I.e., 0 …11) Typedef struct {... /* different fields for each item */ } item; item buffer[BUFFER_SIZE]; int in = 0; int out = 0;  The proposed solution (next slide) can only use BUFFER_SIZE-1 elements

36. 36 Bounded-Buffer – Producer Process item nextProduced; while (1) { while (((in + 1) % BUFFER_SIZE) == out) ; /* do nothing; buffer is full */ buffer[in] = nextProduced; in = (in + 1) % BUFFER_SIZE; } 0 11 0 producer consumer 0 11 Ring-buffer 12 slots out in out in out Full Half-full Empty

37. 37 Bounded-Buffer – Consumer Process item nextConsumed; while (1) { while (out == in) ; /* do nothing; buffer is empty */ nextConsumed = buffer[out]; out = (out + 1) % BUFFER_SIZE; } 0 11 0 producer consumer 0 11 Ring-buffer 12 slots out in Empty Half-fullFull

38. 38 Interprocess Communication (IPC) Mechanism for processes to communicate and to synchronize their actions.  Message system – processes communicate with each other without resorting to shared variables.  IPC facility provides two operations:  Send (message) – message size fixed or variable  Receive (message)  If P and Q wish to communicate, they need to:  establish a logical communication link between them  exchange messages via send/receive  Implementation of communication link  physical (e.g., shared memory, hardware bus)  logical (e.g., logical properties)

39. 39 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?  What is the capacity of a link?  Is the size of a message that the link can accommodate fixed or variable?  Is a link unidirectional or bi-directional?

40. 40 Direct Communication  Processes must name each other explicitly:  send (P, message) – send a message to process P  receive(Q, message) – receive a 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.

41. 41 Indirect Communication  Messages are directed 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.  Link may be unidirectional or bi-directional.

42. 42 Indirect Communication  Operations  create a new mailbox  send and receive messages through mailbox  destroy a mailbox  Primitives are defined as: send (A, message) – send a message to mailbox A receive (A, message) – receive a message from mailbox A

43. 43 Indirect Communication  Mailbox sharing  P 1, P 2, and P 3 share mailbox A.  P 1, sends; P 2 and P 3 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.

44. 44 Synchronization  Message passing may be either blocking or non-blocking  Blocking is considered synchronous  Blocking send has the sender block until the message is received  Blocking receive has the receiver block until a message is available  Non-blocking is considered asynchronous  Non-blocking send has the sender send the message and continue  Non-blocking receive has the receiver receive a valid message or null

45. 45 Buffering  Queue of messages attached to the link; implemented in one of three ways. 1. Zero capacity – 0 messages Sender must wait for receiver (rendezvous). 2. Bounded capacity – finite length of n messages Sender must wait if link full. 3.Unbounded capacity – infinite length Sender never waits.

46. Examples of IPC Systems - POSIX  POSIX Shared Memory  Process first creates shared memory segment segment id = shmget(IPC PRIVATE, size, S_IRUSR | S_IWUSR);  Process wanting access to that shared memory must attach to it shared memory = (char *) shmat(id, NULL, 0);  Now the process could write to the shared memory sprintf(shared memory, "Writing to shared memory");  When done a process can detach the shared memory from its address space shmdt(shared memory);

47. Examples of IPC Systems - Mach  Mach communication is message based  Even system calls are messages  Each task gets two mailboxes at creation- Kernel and Notify  Only three system calls needed for message transfer msg_send(), msg_receive(), msg_rpc()  Mailboxes needed for commuication, created via port_allocate()

48. Examples of IPC Systems – Windows XP  Message-passing centric via local procedure call (LPC) facility  Only works between processes on the same system  Uses ports (like mailboxes) to establish and maintain communication channels  Communication works as follows:  The client opens a handle to the subsystem’s connection port object  The client sends a connection request  The server creates two private communication ports and returns the handle to one of them to the client  The client and server use the corresponding port handle to send messages or callbacks and to listen for replies

49. 49 Simulating Shared Memory in Java

50. 50 Bounded-Buffer – Message Passing Solution

51. 51 Bounded-Buffer – Message Passing Solution

52. 52 Bounded-Buffer – Message Passing Solution The Producer

53. 53 Bounded-Buffer – Message Passing Solution The Consumer

54. 54 Message Passing in Windows XP

55. 55 Client-Server Communication  Sockets  Remote Procedure Calls  Remote Method Invocation (Java)

56. 56 Sockets  A socket is defined as an endpoint for communication.  Concatenation of IP address and port  The socket 161.25.19.8:1625 refers to port 1625 on host 161.25.19.8  Communication consists between a pair of sockets, one for each process.

57. 57 Socket Communication 1- Client sends request 2- Server waits for request 3- Socket link is established

58. 58 Socket communication details

59. 59 Sockets in Unix Server makes the socket known to the system using bind() so that the clients can use it

60. Run socket example for directory listing Client: [Kamrans-MacBook: Socket-program] K_Sartipi%./client ~K_Sartipi Kamrans-MacBook.local Server: [Kamrans-MacBook:Socket-program] K_Sartipi%./server & 60

61. 61 * Time-of-day server listening to port 6013. * Figure 3.19 import java.net.*; import java.io.*; public class DateServer { public static void main(String[] args) try { ServerSocket sock = new ServerSocket(6013); // now listen for connections while (true) { Socket client = sock.accept(); // we have a connection PrintWriter pout = new PrintWriter(client.getOutputStream(), true); // write the Date to the socket pout.println(new java.util.Date().toString()); // close the socket and resume listening for more connections client.close(); } catch (IOException ioe) { System.err.println(ioe); }

62. 62 * Client program requesting current date from server. * Figure 3.20 import java.net.*; import java.io.*; public class DateClient { public static void main(String[] args) { try { // this could be changed to an IP name or address other than the localhost Socket sock = new Socket("127.0.0.1",6013); InputStream in = sock.getInputStream(); BufferedReader bin = new BufferedReader(new InputStreamReader(in)); // read the date from the socket String line; while( (line = bin.readLine()) != null) System.out.println(line); // close the socket connection sock.close(); } catch (IOException ioe) { System.err.println(ioe); }

63. 63 Network Address of a Computer dig birkhoff.cas.mcmaster.ca Ping birkhoff.cas.mcmaster.ca

64. 64 Remote Procedure Calls (RPC)  Remote procedure call (RPC) abstracts procedure calls between processes on networked systems.  Client & server stubs – client-side proxy for the actual procedure on the server.  The client-side stub locates the server and marshalls the parameters.  The server-side stub receives this message, unpacks the marshalled parameters, and performs the procedure on the server.

65. 65 Execution of RPC  XDR  H andles the Differences of data Representations in Client and Server sides  Binding  Local: compile / load / execute times  Remote: hard coded if port is known; otherwise use of Matchmaker Daemon  Reliability of RPC:  Local never fails  Remote: at-most-once strategy

66. 66 Remote Method Invocation  Remote Method Invocation (RMI) is a Java mechanism similar to RPCs.  RMI allows a Java program on one machine to invoke a method on a remote object.

67. 67 Marshalling Parameters - Local objects passed by copy (object serialization) - Remote objects passed by reference

68. 68 Marshalling Parameters

69. 69 RMI Example NOT IN EXAM

70. 70 RMI Example NOT IN EXAM

71. 71 RMI Example NOT IN EXAM