PROCESS CONCEPT

REFERENCES:

1.“OPERATING SYSTEM CONCEPTS” 9TH EDITION BY ABRAHAM SILBERSCHATZ, PETER BAER GALVIN AND GREG GAGNE
2.“OPERATING SYSTEMS: INTERNALS AND DESIGN PRINCIPLES”, 7TH EDITION BY WILLIAM STALLINGS
[Link] lectures on “Introduction to Operating Systems” by Prof. Chester Rebeiro, IIT Madras [Link]
 WHY WE NEED OPERATING SYSTEMS

• OS act as secure and efficient interface

• Programmers develop applications to
perform some task over computer. between the applications and the hardware
resources.
• It is not efficient to write applications
for a particular hardware platform. • OS provides a uniform access to resources
when requested by applications
 WHAT DOES OS MANAGES?

• It provides resources to multiple applications

• It helps multiple applications to use processor simultaneously by switching

processor among applications.

• All applications make progress concurrently.

• It helps in efficiently sharing Processor and I/O devices among applications.

Define PROCESS?

• It is an instance or a portion of a program which is getting executed in the processor

at the moment
• Executing a process means executing instructions in a sequence.
• Each process has a current state, and a set of system resources acquired by it
 PROCESS CONCEPT
• Types of systems
• Batch system: Where computational tasks are completed by processors in the form of Jobs.
• Time-shared systems: Processor is shared among applications.

• process can also be termed as job

• A program is not a process.
• A program is a passive entity, for example a file with a list of instructions stored on disk.
• Whereas, a process is an active part of a process.
• A program counter specifies the next instruction to execute and a set of associated resources.
• When an executable file is loaded into RAM, the program becomes a process.
 A Process in Memory

• A Process consists of following sections in the memory:

• The code or instructions, also called text section
• Program counter and registers maintain current activities of a process
• Stack stores temporary data of a process.
• For example local variables, function arguments or return addresses
• Data section stores global variables
• Memory allocated dynamically during run time is stored in Heap

A process in memory
Figure 3.1 in Ref.1
 STATES OF A PROCESS
• During execution, a process is in one of the mentioned states:
• New: When a process is being created.
• Waiting: When the process is waiting for some I/O resource or an event to occur like
getting input from keyboard, a message from another process, to access disk storage or an
another process to finish a task, and so on.
• Ready: The process is in queue having all the required resources and waiting to get a
processor
• Running: When CPU is executing process instructions
• Terminated: When process finishes execution of all its instructions
 WHAT IS A PROCESS CONTROL BLOCK (PCB)

• Each process has a PCB or a Task Control Block.

• Each process can be uniquely identified by a no. of parameters
• A process execution can be interrupted and later resumed
• It stores process specific information, like
• State: State of a process can be ready, waiting, running, etc
• Program counter: It stores location of next instruction to execute
• CPU scheduling: it provides relative priorities of process and pointers to queue
• CPU registers: It stores information/ data associated with process

Fig. 3.1 from Ref. 2

PROCESS CONTROL BLOCK (CONTD.)

• Memory-management information: Size of memory and starting address allocated to the

process
• Usage/Accounting information: CPU utilization, turn around time, etc
• I/O status: resources or devices allocated to a process
• Identifier: Each process has a unique id
• Memory pointers: It stores the address of the program code and data related to process
• Priority: Priority of a process compared to other
TRACE AND DISPATCHER
• TRACE monitors and notes execution sequence of instructions. It also keeps record of
process interleaving while executing.

• DISPATCHER is a small program in operating system that helps in switching processor

from one process to another

• Fig. 3.2 and 3.3 from Ref.2

SWITCHING Between PROCESSES

Fig. 3.4 from Ref.1

Fig. 3.4 from Ref.2
THREADS

• A process can be divided into multiple light weight executable entities to

parallelize the task.
• These are called Threads.
• Suppose a process consists of multiple program counters (PC)
• Code of a process can be executed from multiple locations or instructions
PROCESS SCHEDULING

• Multiple processes share same CPU in time sharing systems

• It is required to maximize CPU utilization
• A process should not hold processor while it is busy some other tasks like I/O activity.
• A quick processes switching is required onto CPU
PROCESS SCHEDULING

• A Process scheduler selects the next process for execution on CPU while multiple are
available to run
• It maintains scheduling queues of processes
• I/O or Device queues: consist of processes waiting to get access to an I/O device
• Ready queue: It is a queue of processes which are waiting in main memory to execute in
CPU
• Processes switches from one queue to another
READY QUEUE AND VARIOUS I/O DEVICE QUEUES

Fig. 3.5 from Ref.1

QUEUEING DIAGRAM

Fig. 3.6 from Ref.1

 SCHEDULERS
• Processes are usually stored to a secondary storage device (like a disk) for later execution
• For execution OS brings them to main memory (RAM)
• Processes are classified as:
• I/O-bound process – These processes majorly spends time performing I/O activities than doing
computations using CPU
• CPU-bound process – – These processes majorly perform computations using CPU

• The portion of the OS responsible for selecting processes to be brought to main memory/ ready queue is
Long-term scheduler (or job scheduler).
• It is called less-frequently (like in seconds, minutes)
• It decides the degree of multiprogramming
• It should select a proper combination of I/O bound and CPU bound processes
 SCHEDULERS (CONTD.)
• The portion of the OS responsible for selecting processes which can be executed next in
CPU is called Short-term scheduler (or CPU scheduler)

• Few system have only this scheduler

• It is called frequently (like in milliseconds)

• Medium-term scheduler: It can used to reduce the degree of multiple programming.

• It performs swapping of processes. It swaps processes from memory to disk and vice-versa

Fig. 3.7 from Ref.1

 CONTEXT SWITCH
• Sometimes OS interrupts CPU from performing its current task to run a kernel routine
• Or CPU switches to another process when CPU usage of one process is either complete or interrupted
• Context switching means saving the current state of a running process so that it can be restored.
• It helps in saving the state of the old process and loading the saved state for the new process
• PCB provides the Context/state of a process
• Context-switch time is an overhead; not a useful work for the system
• Complex OS and PCB increases the switching time
• Time also depends on hardware support
OPERATIONS ON PROCESSES

• System must provide mechanisms for:

• process creation,
• process termination,
• and so on
TWO-STATE DIAGRAM

Ref.2
REASONS FOR PROCESS CREATION

Or when a process requests the OS to create another process

Table 3.1 from Ref.2

 PROCESS CREATION
• A process can create another process by requesting OS.
• These processes are called Parent process and children processes, respectively.
• This may result in creating a tree of processes
• Each process has a unique process identifier (pid)

Resource Sharing
• A Parent process has to distribute its resources among its children
• It prevents any process from creating too many child processes as it may
overload the system
• A child process can acquire its resources either from its parent or directly
from the operating system
 PROCESS CREATION

Process Execution

• Parent and children processes may execute simultaneously

• Parent waits to terminate until all its children terminate

Address-space options for the new process

•A child process is a copy of its parent process.
•They both have same program and data
 CREATING PROCESS IN UNIX
• fork(): This system call creates new process
• The parent and the child both continue the execution starting from the instruction just after the
fork()
• Fork returns 0 to the child process and pid of child to the parent

• exec(): After a fork(), the parent or the child process uses the exec() system call to replace the
memory space of process with a new program
• Parent can use wait() to leave the ready queue until its children terminate
C PROGRAM FORKING SEPARATE PROCESS

Figure 3.9 Creating a separate process using the UNIX fork() system call
Ref.1
 PROCESS TERMINATION

Methods available to terminate a process:

HALT instruction or an explicit OS service call can be used for termination
•

Close or Logoff application

•

Unix uses exit() system call

•

Parent process receives a status value from child process using wait() system call
•

Resources of the terminating process are deallocated like I/O buffers, files,
•

physical and virtual memory

PROCESS TERMINATION
abort(): Parents can use it to terminate the execution of children processes
Reasons:
• When child exceeds allocated resources
• No utility of child and it assigned task
• Parent wants to terminate and the operating systems does not allow a child to continue if its parent
terminates
Cascading termination: Termination of a process leads to the termination of all its children and
grandchildren
Exit() returns status information and the pid of the terminated process
•

pid = wait(&status);
If no parent waiting (i.e. did not invoke wait()) then child process becomes Zombie
•

If parent terminated without invoking wait , process becomes Orphan

•
FIVE-STATE DIAGRAM

Figure 3.6 and 3.7 from Ref.2

 SUSPENDED PROCESSES
Swapping

Scenario: Ready queue is full. Processes in ready queue are blocked and waiting to
complete an I/O activity. Processor is idle.

Solution: No processes in main memory are in the Ready state, the OS swaps one of
the blocked processes out on to disk into a suspend queue

After swapping-out operation, OS has two choices:

• Bring a newly created process into main memory

• Or bring in a previously suspended process.

from Ref.2
• Running to Ready/Suspend: Time allocation expires OR a higher-priority process on the
Blocked/Suspend queue gets unblocked

• Blocked to Blocked/Suspend: No ready processes available OR currently running process

or a ready process about to dispatch requires more main memory

• Blocked/Suspend to Blocked: A process in the (Blocked/Suspend) queue is having higher

priority over the processes in the (Ready/Suspend) queue and OS assumes unblocking event
will occur soon.

• Blocked/Suspend to Ready/Suspend: The event for which it has been waiting occurs

• Ready/Suspend to Ready: there is no ready process in main memory OR a process in the

Ready/Suspend state has higher priority than the processes in the Ready state.

• Ready to Ready/Suspend: Normally, not preferred. Needed to free up a sufficiently large

block of main memory. Might also suspend a lower-priority ready process rather than a
higher priority blocked process
• New to Ready/Suspend: Insufficient room in main memory for a new process
 INTERPROCESS COMMUNICATION
• Types of Processes: Independent or Cooperating
• Cooperating: It can affect or be affected by the other processes and share data with other process
• Need for cooperating processes:
• Modularity
• Convenience
• Computation speedup
• Information sharing
• Cooperating processes need Interprocess Communication (IPC)
Why Cooperation?
• Information sharing: Several users may be interested in the same information
• Computation speedup
• Modularity: dividing the system functions into separate processes or threads
• Convenience
• IPC Mechanisms
• Shared memory
• Message passing
• Signals

(a) Message passing (b) shared memory

Many systems implement both shared memory and message

passing
SHARED MEMORY
• One process reserves an area in the RAM that is accessible by another process.
• Advantage: Fast access- similar to regular reading and writing in the memory
• Disadvantage: Needs process synchronization
 Problems with Shared memory:

Normally, OS tries to prevent one process from accessing another process’s memory.
Shared memory requires to remove this restriction.
Processes are also responsible for ensuring that they are not writing to the same location, simultaneously.
Producer –Consumer Problem Example: A web server produces HTML files and images, which are
consumed by the client browser.
Solution: A shared buffer that can be filled by the producer and emptied by the consumer
 SHARED MEMORY: PRODUCER-CONSUMER
PROBLEM

• A producer process produces information which is consumed by a consumer process.

• Both share a common buffer which can be of two types:
• unbounded-buffer: Unlimited size of buffer
• bounded-buffer: limited buffer size
 SOLUTION WITH BOUNDED-BUFFER

Producer Consumer

Figure 3.13 and 3.14 from Ref.1

MESSAGE PASSING

• A shared memory is created in the kernel

• System calls used for operations:
• send(message)
• receive(message)
• Each send should have a corresponding receive: Process Cooperation

• Message can be of either fixed or variable size

• Processes need to establish a communication link between them and use provided
operations
• Each call requires marshalling and de-marshalling of information.
• Message passing is useful for exchanging smaller amounts of data
MESSAGE PASSING

• Implementing communication link:

• Physical:
• Shared memory
• Hardware bus
• Network

• Logical:
• Direct or indirect
• Synchronous or asynchronous
• Automatic or explicit buffering
MESSAGE PASSING : DIRECT COMMUNICATION

• Each process must have a name and used in the following manner:
• 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
• Exactly one link is associated with each pair of communicating processes
• Usually the link is bi-directional, but can be made unidirectional
MESSAGE PASSING : INDIRECT COMMUNICATION

• Processes share a mailbox, also called as ports

• Each mailbox has a unique id
• mailboxes() are used to send or receive Messages
• Properties of communication link
• Share a common mailbox between processes establishes a link
• Many processes can share a link
• Each pair of processes may share several communication links
• It can be unidirectional or bi-directional
MESSAGE PASSING : INDIRECT COMMUNICATION

Procedure
create a new mailbox (port)
•

send and receive messages through mailbox

•

• send(A, message)—Send a message to mailbox A.

• receive(A, message)—Receive a message from mailbox A.
destroy a mailbox
•

Example :
A, B, and C share a mailbox M1
•

A sends; B and C receive

•
CLIENT-SERVER SYSTEMS: Communication

• Pipes
• Remote Method Invocation (Java)
• Sockets
• Remote Procedure Calls
 SOCKETS

• Sockets: an endpoint for communication

• One socket for each process
• Identified by an IP address concatenated with a port number
• Port no. helps in differentiating applications on a host
• The socket [Link]:1428 refers to port 1428 on host [Link]

• Port no.s below 1024 are well known and used for standard services

• [Link] is a Special IP address (loopback) used to refer to the system on which

process is running
REMOTE PROCEDURE CALLS

• The calls between processes which are separated by networked systems

• Uses ports to differentiate between services
• Each message is sent to a port on the remote system.
• Each message contains an identifier specifying the function to execute and the parameters
to pass to that function.
• The function is executed and output is sent back in a separate message.
• It allows a client to invoke a procedure on a remote host and gives an illusion of calling
the procedure locally
PIPES

Following issues need to be considered while creating pipes:

• There exists a relationship (i.e., parent-child ) between the communicating processes?
• Pipes are going to be used within a machine or over a network?
• Unidirectional or bidirectional communication?
• Half or full-duplex communication?

Ordinary pipes: it is created by a parent process to facilitate communication with its child
•

process. It cannot be accessed from outside the process

Named pipes: it can be accessed without a parent-child relationship.
•
ORDINARY PIPES
pipe() creates a pipe, a
unidirectional data channel
that can be used for
interprocess communication. The
array pipefd is used to return
two file descriptors referring
to the ends of the pipe.
• Producer writes to one end (the write-end of the pipe) pipefd[0] refers to the read
end of the pipe. pipefd[1]
• Consumer reads from the other end (the read-end of the pipe) refers to the write end of the
pipe. Data written to the write
end of the pipe is buffered by
• Unidirectional the kernel until it is read
from the read end of the pipe.

• parent-child relationship is mandatory between communicating processes

• In Windows, these are called anonymous pipes

Figure 3.24 from Ref.1

 NAMED PIPES
• Better than ordinary pipes
• Bidirectional
• Parent-child relationship is not necessary
• Many processes can use it for communication
• Available with both Windows systems and UNIX
• Continue to exist after communicating processes have finished.
• Named pipes are referred to as FIFOs in UNIX systems.
• A FIFO is created with the mkfifo() system call and manipulated with the ordinary open(), read(), write(), and close()
system calls.
• UNIX use only half-duplex transmission
• Intermachine communication needs sockets
• Named pipes on Windows systems provide full-duplex communication and the communicating processes may reside
on either the same or different machines.
• Only byte-oriented data transmitted across a UNIX FIFO, whereas Windows systems allow either byte- or
message-oriented data.