4th Edition: Chapter 1 - John Jay College of Criminal Justice

Download Report

Transcript 4th Edition: Chapter 1 - John Jay College of Criminal Justice

What we will cover…
 Processes





Process Concept
Process Scheduling
Operations on Processes
Interprocess Communication
Communication in Client-Server Systems (Reading Materials)
 Threads



Overview
Multithreading Models
Threading Issues
Lecture 3
1-1
What is a process
 An operating system executes a variety of
programs:



Batch system – jobs
Time-shared systems – user programs or tasks
Single-user Microsoft Windows or Macintosh OS
• User runs many programs
• Word processor, web browser, email
 Informally, a Process is just one such program in
execution; progress in sequential fashion

Similar to any high level language programs code
(C/C++/Java code etc.) written by users
 However, formally, a process is something more
than just the program code (text section)!
Lecture 3
1-2
Process in Memory
 In addition to the text section
 A process includes:



program counter
contents of the processor’s
registers
stack






Contains temporary data
Method parameters
Return addresses
Local variables
data section
While a program is a passive entity, a process is known as an active entity
Lecture 3
1-3
Process State
 As a process executes, goes from creation to
termination, goes through various “states”





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 processor
terminated: The process has finished execution
Lecture 3
1-4
Diagram of Process State
Lecture 3
1-5
Process Control Block (PCB)
 A process contains numerous information
 A system has many processes
 How to manage all the process information
Each process is represented by a Process Control Block
 a table full of information for each process






Process state
Program counter
CPU registers
CPU scheduling information
Memory-management information
I/O status information
Lecture 3
1-6
Process Control Block (PCB)
Lecture 3
1-7
CPU Switch From Process to Process
Lecture 3
1-8
Process Scheduling
 In a multiprogramming environment, there will be many processes


many of them ready to run
Many of them waiting for some other events to occur
 How to manage the architecture?
 Queuing



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
 Processes migrate among these various queues
Lecture 3
1-9
A Representation of Process Scheduling
Lecture 3
1-10
OS Queue structure (implemented with link list)
Lecture 3
1-11
Schedulers
 A process migrates among various queues
 Often more processes are there than can be executed
immediately


Stored in mass-storage devices (typically, disk)
Must be brought into main memory for execution
 OS selects processes in some fashion
 Selection process carried out by a scheduler
 Two schedulers in effect…


Long-term scheduler (or job scheduler) – selects which
processes should be brought into the memory
Short-term scheduler (or CPU scheduler) – selects which
process should be executed next and allocates CPU
Lecture 3
1-12
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
 Long-term scheduler has another big responsibility
 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
 Balance between two types of processes
Lecture 3
1-13
Addition of Medium Term Scheduling
Lecture 3
1-14
Context Switch
 All the earlier mentioned process scheduling has a
trade-off
 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 via a context switch
 Time dependent on hardware support
 Context-switch time is pure overhead; the system does
no useful work while switching
Lecture 3
1-15
Interprocess Communication
 Concurrent processes within a system may be independent or
cooperating
 Cooperating process can affect or be affected by other
processes, including sharing data
 Reasons for cooperating processes:



Information sharing – several users may be interested in a shared file
Computation speedup – break a task into subtasks and work in parallel
Convenience
 Need InterProcess Communication (IPC)
 Two models of IPC


Shared memory
Message passing
Lecture 3
1-16
Communications Models
Shared-memory
Message-passing
Lecture 3
1-17
Shared memory: Producer-Consumer Problem
 Paradigm for cooperating processes
 producer process produces information that is
consumed by a consumer process
 IPC implemented by a shared buffer
 unbounded-buffer places no practical limit on
the size of the buffer
 bounded-buffer assumes that there is a fixed
buffer size
• More practical
• Let’s design!
Lecture 3
1-18
Bounded-Buffer – Shared-Memory Solution design
 Three steps in the design problem
1.
Design the buffer
2.
Design the producer process
3.
Design the consumer process
1.
Shared buffer
(implemented as circular array with two logical pointers: in and out)
#define BUFFER_SIZE 10
typedef struct {
...
} item;
item buffer[BUFFER_SIZE];
int in = 0;
int out = 0;
Lecture 3
1-19
Bounded-Buffer – Producer & Consumer process
design
2. Producer design
while (true) {
/* Produce an item */
while (((in + 1) % BUFFER SIZE) == out)
; /* do nothing -- no free buffers */
buffer[in] = nextProduced;
in = (in + 1) % BUFFER SIZE;
}
3. Consumer design
while (true) {
while (in == out)
; // do nothing -- nothing to consume
// remove an item from the buffer
nextConsumed = buffer[out];
out = (out + 1) % BUFFER SIZE;
}
Lecture 3
1-20
Shared Memory design
 Previous design is correct, but can only
use BUFFER_SIZE-1 elements!!!
 Exercise for you to design a solution
where BUFFER_SIZE items can be in the
buffer

Part of Assignment 1
Lecture 3
1-21
Interprocess Communication – Message Passing
 Processes communicate with each other without
resorting to shared memory
 IPC facility provides two operations:


 If


send(message) – message size fixed or variable
receive(message)
P and Q wish to communicate, they need to:
establish a communication link between them
exchange messages via send/receive
Lecture 3
1-22
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
 A link is associated with exactly one pair of
communicating processes
 Between each pair there exists exactly one link
 Symmetric (both sender & receiver must name the
other to communicate)
 Asymmetric (receiver not required to name the
sender)
Lecture 3
1-23
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
Lecture 3
1-24
Communications in Client-Server Systems
 Socket connection
Lecture 3
1-25
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
Lecture 3
1-26
Socket Communication
Lecture 3
1-27
Threads
 Process model discussed so far assumed that a process
was sequentially executed program with a single thread
 Increased scale of computing
 putting pressure on programmers, challenges include
•
•
•
•
•
Dividing activities
Balance
Data splitting
Data dependency
Testing and debugging
 Think of a busy web server!
Lecture 3
1-28
Single and Multithreaded Processes
Lecture 3
1-29
Benefits
 Responsiveness
 Resource Sharing
 Economy
 Scalability
Lecture 3
1-30
Multithreaded Server Architecture
Lecture 3
1-31
Concurrent Execution on a Single-core System
Lecture 3
1-32
Parallel Execution on a Multicore System
Lecture 3
1-33
User and Kernel Threads
 User threads: Thread management done by
user-level threads library
 Kernel threads: Supported by the Kernel
Windows XP
 Solaris
 Linux
 Mac OS X

Lecture 3
1-34
Multithreading Models
 Many-to-One
 One-to-One
 Many-to-Many
Lecture 3
1-35
Many-to-One
 Many user-level threads mapped to single
kernel thread
 Examples:
Solaris Green Threads
 GNU Portable Threads

Lecture 3
1-36
One-to-One
 Each user-level thread maps to kernel thread
 Examples
 Windows NT/XP/2000
 Linux
Lecture 3
1-37
Many-to-Many Model
 Allows many user level threads to be
mapped to many kernel threads
 Allows the operating system to
create a sufficient number of kernel
threads
 Solaris prior to version 9
Lecture 3
1-38
Many-to-Many Model
Lecture 3
1-39
Threading Issues
 Thread cancellation of target thread
 Dynamic unbound usage of threads
Lecture 3
1-40
Thread Cancellation
 Terminating a thread before it has
finished
 General approaches:

Asynchronous cancellation terminates the
target thread immediately
• Problems?

Deferred cancellation allows the target
thread to periodically check if it should be
cancelled
Lecture 3
1-41
Dynamic usage of threads
 Create thread as and when needed
 Disadvantages:
Amount of time to create a thread
 Nevertheless, this thread will be discarded
once it has completed work; no reusage
 No bound on the total number of threads
created in the system

• may result in severe resource scarcity
Lecture 3
1-42
Solution: Thread Pools
 Create a number of threads in a pool where
they await work
 Advantages:
Usually faster to service a request with an existing
thread than create a new thread
 Allows the number of threads in the application(s)
to be bound to the size of the pool

 Almost all modern OS provide kernel support
for threads: Windows XP, MAC, Linux…
Lecture 3
1-43