Concurrent Process

Presented By
Mr. Ajay Maurya
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 1
Technology, Varanasi
 Process Synchrnization

• Process Synchrnization is a technique used in an operating

system to Control the execution of multiple processes so that
they can safely shared data and resources without causing
errors.
• Shared Resources (Variable, Memory, Code, Resources(CPU,
Printer, Scanner.))
• It is mainly required in Concurrent(Co-operative)
Concurrent(Co Processes.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 2
Technology, Varanasi
 Why Synchronization Needed?
• When Two or More Processes access shared data at the same time, It
may cause:
• Data Inconsistency
• Race Condition
• Unexpected Result

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 3
Technology, Varanasi
 Example
Process P1 Process P2
Int shared = 5
Int x = shared Int y = shared
x++ Y--
sleep(1) sleep(1)
shared = x shared = y

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 4
Technology, Varanasi
 Result
• Final output will be based on execution order:
• If firstly Process P1 is running then output will be 4.
• If firstly Process P2 is running then output wiil be 6.
• This condition in known as Race Condition. Processes are doing race
in between to get the results based on execution order.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 5
Technology, Varanasi
 Code Segment(Problem)

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 6
Technology, Varanasi
 Code Segment(Solution)

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 7
Technology, Varanasi
 Problem in Concurrency
Race Condition : When two processes access shared data at
the same time and results depends on execution order.
Deadlock : Two or more processes are waiting for each other
to release resources.
Starvation : A process never CPU Time.
Critical Section : A critical section is the part of program
where a process accesses shared resource or shared data.
• Only one process at a time should execute the Critical Section to
avoid errors or data inconsistency.
inconsistency
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 8
Technology, Varanasi
 Why Critical Section Needed?
• When multiple processes are running concurrently
and access shared resources, following problem may
occure:
• Race Condition
• Data Inconsistency
• Unexpected Result
• Therefore, the Operating System must control access
to the Critical Section.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 9
Technology, Varanasi
 Solutions/Requirement for Critical Section

A correct solution must satisfy three condition:

1. Mutual Exclusion
• Only one process can execute the critical section at a time.
2. Progress
• If no process in Critical Section, another process should be
allowed to enter.
3. Bound Waiting
• A process should not wait forever to enter critical section.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 10
Technology, Varanasi
 Critical Section
Structure of Critical Section Example of Critical Section

A typical process using a do{

critical section has 4-Parts Entry Section
1. Entry Section Critical Section
2. Critical Section
Exit Section
3. Exiti Section
Remainder Section
4. Remainder Section
}while(true)
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 11
Technology, Varanasi
 Method to solve Critical Section Problem
Operating System use several Techniques:
• Lock Variable
• Peterson’s Solutions
• Semaphores
• Mutex
• Monitors

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 12
Technology, Varanasi
Producer Consumer Problem / Bounded Buffer
Problem
• The Producer Consumer Problem is a classical Process
Synchronization problem in an operating system.
• It occurs when:
• A Producer process produces data/items.
• A Consumer process consumes those data/items.
• Both are shared a common buffer(Shared Memory)
• The buffer has limited size, so proper coordination is
required.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 13
Technology, Varanasi
 Main issue in Producer Consumer Problem
Without synchronization, the following problems may occur :
• Buffer Overflow
• Producer tries to add an item when the buffer is full.
• Buffer Underflow
• Consumer tries to remove an item when the buffer is
empty.
• Race Condition
• Producer and Consumer access the buffer simultaneously,
causing incorrect data.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 14
Technology, Varanasi
 Pseudo Code for Producer Consumer Problem

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 15
Technology, Varanasi
 Pseudo Code for Printer Spooler Problem
Spooler Directory
0

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 16
Technology, Varanasi
 Critical Section Problem
• The Critical Section Problem occurs in a multitasking operating system
when multiple processes or threads share common resources (like
variable, files, files or memory) and try to access or modify them
simultaneously.
• What is Critical Section?
• A critical section is a part of program where shared resources are
accessed.
• “If two processes execute this at the same time, incorrect results
may occure due to race conditions.”

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 17
Technology, Varanasi
 Why is it a Problem?
• If two or more processes enter their critical sections at the same time
and manipulated shared data, it may lead to:
• Race conditions
• Data Inconsistencey
• Unexpected results

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 18
Technology, Varanasi
 Requirements for Good Solutions
Any solution to the critical section problem must satisfy four
conditions:
• Mutual Exclusion
Only one process can enter the critical section at a
Only
time.
• Progress
If no process is in the critical section, the decision of
If
who enters next cannot be postponed indefinitely.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 19
Technology, Varanasi
 Requirements for Good Solutions
• Bounded Waiting
AA process should not wait forever to enter the critical
section (no starvation)
• No assumption related to hardware speed
To achieve synchronization in Critical Section I have
To
given a solution and said that this solution will run on a
32 bit system, on Operating System, would not run on
64. This is not a condition, means there should not be
an assumption like this.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 20
Technology, Varanasi
 Method to solve Critical Section Problem
Operating System use several Techniques:
• Lock Variable
• Peterson’s Solutions
• Semaphores
• Mutex
• Monitors

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 21
Technology, Varanasi
 Lock Variable
A Lock Variable is a simple method used to solve the critical section
problem in an Operating System. It ensures that only one process can
enter the critical section at a time.
time
• A lock variable is a shared variable used to control access to the
critical section.
• If lock = 0 → Cri cal sec on is free
• If lock = 1 → Cri cal sec on is busy (locked)

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 22
Technology, Varanasi
 Working of Lock Variable
Steps:
• A process checks the value of lock.
• If lock = 0, the process sets lock = 1 and enters the critical section.
• After finishing, the process sets lock = 0 to release the resource.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 23
Technology, Varanasi
 Simple Example (Pseudo Code)
while(lock == 1); // wait until lock becomes 0
lock = 1; // acquire lock
/* Critical Section */
printf("Process is using shared resource");
lock = 0; // release lock

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 24
Technology, Varanasi
 Problem with Lock Variable
• The lock variable solution is not perfect because it may fail in
multiprocessing environments.
environments
• Example problem:
• Two processes check lock at the same time when it is 0.
Both may set it to 1 simultaneously and enter the critical
section, causing a race condition.
condition
• So it does not guarantee mutual exclusion.
exclusion

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 25
Technology, Varanasi
 Better Solutions
Because of the limitations of lock variables, better
mechanisms are used:
• Test and Set Instruction
• Peterson's Algorithm
• Semaphores
• Mutex locks
“A lock variable is a shared variable that indicates whether a
critical section is free or busy,, but it is not a reliable solution
for process synchronization.”
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 26
Technology, Varanasi
 Test and Set Instruction
• Test-and-Set is a instruction used in process synchronization
to solve the critical section problem. It ensures that only
one process can enter the critical section at a time.
• Test-and-Set is an atomic (indivisible) instruction that:
1. Tests the value of a lock variable.
variable
2. Sets it to true (locked) in one single operation.
• Because it is atomic, no other process can interrupt it during
execution.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 27
Technology, Varanasi
 Working idea
A shared variable lock is used.
• lock = 0 → Cri cal sec on is free
• lock = 1 → Cri cal sec on is busy
When a process wants to enter the critical section:
• It executes Test-and-Set(lock)
Set(lock)
• If lock was 0, it becomes 1 and the process enters.
• If lock was 1,, the process keeps waiting.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 28
Technology, Varanasi
 Code Implementation

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 29
Technology, Varanasi
 Simple Example (Real Life)
Imagine a single bathroom with a door lock.
lock
1. lock = 0 → Bathroom free
2. lock = 1 → Someone inside
Steps:
1. Person checks door and locks it (Test-and-Set).
(Test
2. If already locked → waits outside.
3. A er use → unlocks the door.
So only one person can use the bathroom at a time.
time
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 30
Technology, Varanasi
 Turn Variable (Strict Alteration)
In Operating Systems, the Turn Variable is a shared variable
used in process synchronization to control which process is
allowed to enter the critical section.
section
• A turn variable is a shared integer variable that indicates
whose turn it is to enter the critical section.
section
• If turn = 0 → Process P0 can enter the critical section.
• If turn = 1 → Process P1 can enter the critical section.
• This technique is used in software solutions of the critical
section problem.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 31
Technology, Varanasi
 Basic Idea
Two processes P0 and P1 share a variable:

turn = 0
Process P0 Process P1
while(turn!=1);
while(turn!=0);
//Critical Section
//Critical Section
turn = 0
turn = 1
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 32
Technology, Varanasi
 Simple Example (Real Life)
Imagine two students sharing one computer.
computer
• turn = 0 → Student P0 uses the computer.
• After finishing, P0 sets turn = 1.
• Now P1 can use the computer.
They alternate turns to use the resource.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 33
Technology, Varanasi
 Peterson’s Solutions
Peterson’s Solution is a classic algorithm used in Operating Systems to
solve the mutual exclusion (critical section) problem for two
processes.
• It uses two shared variables:
1. flag[2] → indicates whether a process wants to enter the critical
section.
2. turn → indicates whose turn it is to enter.
• So each process:
• Announces its intention to enter the critical section.
• Gives priority to the other process if both want to enter at the
same time.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 34
Technology, Varanasi
 Peterson’s Solutions

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 35
Technology, Varanasi
 Properties of Peterson’s Solution
Peterson’s solution satisfies the three requirements of the critical
section problem:
1. Mutual Exclusion
Only one process enters the critical section.
2. Progress
If no process is in the critical section, one waiting process will
enter.
3. Bounded Waiting
A process will not wait forever.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 36
Technology, Varanasi
 Mutext
A Mutex is a locking mechanism that allows only one process/thread
to access a shared resource at a time. Other processes must wait until
the lock is released.
A mutex has two states:
1. Locked → Resource is being used
2. Unlocked → Resource is free
Steps:
1. Process requests the mutex lock.
lock
2. If the lock is free → process enters the critical section.
3. A er finishing → process releases the lock.
lock
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 37
Technology, Varanasi
 Example
Imagine two processes sharing one printer.
printer
• Process P1 locks the mutex → uses printer.
• Process P2 tries to print → must wait.
• When P1 unlocks the mutex,, P2 can use the printer.
Thus only one process prints at a time.
time
• Advantages
✔ Prevents race conditions
✔ Ensures mutual exclusion
✔ Simple locking mechanism

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 38
Technology, Varanasi
 Advantage & Disadvantage
• Advantages
✔ Prevents race conditions
✔ Ensures mutual exclusion
✔ Simple locking mechanism

• Disadvantages
❌ If not released properly → deadlock may occur
❌ Other processes may have to wait

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 39
Technology, Varanasi
 Mutex (Code Implementation)

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 40
Technology, Varanasi
 Semaphore
• A Semaphore is a synchronization mechanism used in an Operating
System to control access to shared resources when multiple
processes or threads are executing concurrently.
• It helps prevent race conditions and ensures proper process
synchronization.
• A semaphore is basically an integer variable that is accessed only
through two atomic operations:
• wait() (P operation) P() Down() Wait()
• signal() (V operation)
V() Up() Signal()
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 41
Technology, Varanasi
 Operations on Semaphore
1. wait() Operation
Executed when a process want to use a resource.
wait(S)
{
S = S - 1;
if(S < 0)
wait until resource becomes available
}
Meaning :
• Decrease the semaphore value.
• If the value become negative, the process must wait.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 42
Technology, Varanasi
 Operations on Semaphore
2. signal() Operation
Executed when a process releases the resource.
signal(S)
{
S = S + 1;
if(S <= 0)
wake up one waiting process
}
Meaning:
• Increases the semaphore value.
• Wakes up one waiting process.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 43
Technology, Varanasi
 Types of Semaphore
There are two main types of semaphores.
semaphores
1. Binary Semaphore
• A Binary Semaphore can have only two values: 0 or 1.
• Used for mutual exclusion.
• Similar to a lock mechanism.
mechanism

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 44
Technology, Varanasi
 Types of Semaphore
Example
Suppose one printer is shared by multiple processes.
Initial value:
S=1
Steps:
• Process P1 executes wait(S) → S = 0 → uses printer.
• Process P2 executes wait(S) → must wait.
• After printing, P1 executes signal(S) → S = 1.
• Now P2 can use the printer.
✔ Only one process uses the printer at a time.
time
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 45
Technology, Varanasi
 Types of Semaphore
2. Counting Semaphore
• A Counting Semaphore can have multiple integer values
(0,1,2,3...).
• Used when multiple instances of a resource are available.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 46
Technology, Varanasi
 Types of Semaphore
Example
Suppose there are 3 printers in a lab.
Initial value:
S=3
Steps:
1. Three processes can execute wait(S) and use printers.
2. If a fourth process requests a printer → it must wait.
3. When any process finishes → it executes signal(S) → another
process can use the printer.
✔ Thus multiple resources can be managed efficiently.
efficiently
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 47
Technology, Varanasi
 Difference Between Binary and Counting Semaphore

Feature Binary Semaphore Counting Semaphore

Values 0 or 1 0,1,2,3…

Resource Single resource Multiple resources

Purpose Mutual exclusion Resource management

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 48
Technology, Varanasi
 Counting Semaphore
down(Semaphore s){ up(Semaphore s){
up
[Link] = [Link] – 1; [Link] = [Link] + 1;
If([Link] < 0){ If([Link] <= 0){
//put process in suspended
list //select a process from suspended
//sleep(); //list and wake up();
}else{ }
//Critical Section }
}
}
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 49
Technology, Varanasi
 Binary Semaphore
down(Semaphore s){ up(Semaphore s){
up
If([Link] == 1){ If(suspended list is empty){
//cs
[Link] = 0; [Link] = 1;
}else{ }else{
//Block this process and //Select a proces from
place in suspended list, suspended list and wake up();
sleep();
} }
} }

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 50
Technology, Varanasi
 Reader–Writer
Writer Problem
The Reader–Writer Problem is a classical synchronization problem in
Operating Systems. It deals with a situation where multiple processes
want to read and write shared data (such as a file or database).

The processes are divided into two types:

• Readers → Only read the data
• Writers → Modify (write) the data

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 51
Technology, Varanasi
 Basic Rule
• Multiple readers can read the data at the same time.
• Only one writer can write at a time.
• When a writer is writing, no reader should read.
This ensures data consistency.
Example
• Suppose there is a database file.
file
• Many users are reading the file → Allowed simultaneously.
• If one user wants to update the file → No other reader or writer
should access it until writing finishes.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 52
Technology, Varanasi
 Example
Reader1 → Reading database
Reader2 → Reading database
Reader3 → Reading database
Writer1 → Must wait
When all readers finish:
Writer1 → Starts wri ng
During writing:
Reader1, Reader2 → Must wait

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 53
Technology, Varanasi
 Solution Using Semaphores
We use:
• mutex → Protects reader count variable
• wrt → Controls access to shared data
Variable:
• readcount → Number of readers currently reading

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 54
Technology, Varanasi
 Reader – Writer Problem
Int rc = 0;
Binary Semaphore mutex =
Down(mutex); Void writer(){
Rc = rc – 1;
1
If(rc == 0){
While(true){
Binary Semaphore db = 1
Void reader(){
Up(db) Down(db)
}
While(true){ Db // CS
Up(mutex)
Down(mutex);
Process_data; Up(db)
Rc = rc +1
If(rc == 1) then down(db);
} }
}
Up(mutex) }
Db // CS Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 55
Technology, Varanasi
How It Works
1. First reader locks writer access.
2. Other readers can read simultaneously.
simultaneously
3. Writer must wait until all readers finish.
finish
4. A er the last reader exits → writer gets access.
access
Real-Life Example
Think of a library 📚 :
• Many students can read books at the same time.
time
• If a librarian wants to update the book record,
record nobody can read
until the update finishes.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 56
Technology, Varanasi
 Sleeping Barber Problem
• The Sleeping Barber Problem is a classical synchronization
problem in Operating Systems that demonstrates how
processes coordinate using shared resources. It was
introduced by Edsger W. Dijkstra.
Dijkstra
• The problem models a barber shop where a barber serves
customers, and synchronization is needed between the
barber process and customer processes.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 57
Technology, Varanasi
 Problem Description
Consider a barber shop with:
• 1 Barber
• 1 Barber chair
• N waiting chairs for customers
The challenge is to synchronized the barber and customer so that-
that
• The barber does not sleep when there are customer waiting.
• Customer do not leave unnecessarily.
• No two person occupy the barber chair at same time.
• Customer are served in FIFO order.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 58
Technology, Varanasi
 Rules
1. If no customers are present,, the barber sleeps.
2. When a customer arrives:
• If the barber is sleeping → the customer wakes the barber.
• If the barber is busy → the customer sits in a waiting chair.
3. If all waiting chairs are full,, the customer leaves the shop.
4. When the barber finishes a haircut:
• If there are customers wai ng → he takes the next customer.
• If no customers are wai ng → he goes to sleep again.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 59
Technology, Varanasi
 Goal of the Problem
The goal is to avoid:
• Race conditions
• Deadlock
• Busy waiting
and ensure proper process synchronization between barber and
customers.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 60
Technology, Varanasi
 Sleeping Barber Solution using Semaphore
We use three semaphores:
• customers → number of customers wai ng
• barbers → number of barbers available
• mutex → protects access to wai ng chairs
Also a variable:
• waiting → number of wai ng customers
• The Sleeping Barber problem is solved using a combination
of counting semaphores (to track available seats and customers)
and binary semaphores (for mutual exclusion, such as updating the
seat count or locking the barber chair).

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 61
Technology, Varanasi
 Pseudocode
Customer = 0, Barber = 0, Mutex = 1, Waiting = 0
Chair = n Barber Process
Customer Process while(true){
down(mutex);
down(customers);// sleep if no customers
if(waiting < N){
down(mutex); // enter critical section
waiting = waiting + 1;
up(customers); // notify barber waiting = waiting - 1;
up(mutex); up(barbers); // barber ready
down(barbers); // wait until barber ready up(mutex); // leave critical section
get_haircut(); cut_hair();
}
}
else{
up(mutex);
leave_shop();
}
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 62
Technology, Varanasi
 Simple Working Example
Suppose:
• Waiting chairs = 3
Scenario
1. No customers → barber sleeps.
sleeps
2. Customer1 arrives → wakes barber → haircut starts.
3. Customer2 and Customer3 arrive → sit in wai ng chairs.
4. Customer4 arrives → waits if chair available.
5. If all chairs are full → customer leaves.
leaves

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 63
Technology, Varanasi
 Diagram (Concept)
Customers → Wai ng Chairs → Barber Chair → Barber

If no customer → Barber Sleeps

If chairs full → Customer Leaves

Key Concepts Demonstrated

• Mutual Exclusion
• Process Synchronization
• Semaphore usage
• Resource sharing
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 64
Technology, Varanasi
 Dinning Philosopher Problem
The Dining Philosopher Problem is one of the most famous problems in
Operating Systems related to process synchronization and deadlock.
• Problem Statement
• There are 5 philosophers sitting around a circular table.
• Between each pair of philosophers, there is 1 fork.
• So, total 5 forks for 5 philosophers.
• A philosopher alternates between:
• Thinking 🤔
• Eating 🍝
Rule:
• To eat, a philosopher needs 2 forks:
• One from the left
• One from the right
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 65
Technology, Varanasi
 Problem
1. Deadlock
• If all philosophers pick up their left fork at the same time, then:
• Each is waiting for the right fork
• No one can proceed
👉 System gets stuck (Deadlock)
2. Starvation
• Some philosophers may never get both forks
👉 They remain hungry forever

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 66
Technology, Varanasi
 Dinning Philosopher Problem
void philosopher() P0 f0
f1
{
while (true)
{
P4
thinking(); P1
Rice Bowl
take_fork(i); f4
take_fork((i + 1) % n)
eat(); f2

put_fork(i); P3
put_fork((i + 1) % n) P2
} f3
}
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 67
Technology, Varanasi
 Dinning Philosopher Solution Using
Semaphore
oid philosopher() P0
f1 f0

while (true)
{ P4
thinking(); P1
wait take_fork(i); Rice Bowl
f4
wait take_fork((i + 1) % n)
eat(); f2
signal put_fork(i); P3
signal put_fork((i + 1) % n) P2
} f3

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 68
Technology, Varanasi
 Explaination Line by Line
Dining Philosophers Problem
• This code represents the Dining Philosophers Problem — a classic OS
synchronization problem where n philosophers sit at a round table,
alternating between thinking and eating, with one fork between each pair.
• Line-by-Line Explanation
• void philosopher() Defines the function for a philosopher's behavior. Each
philosopher runs this same function (typically as a separate
process/thread), identified by index i..
• while (true) The philosopher lives in an infinite loop — they never stop
thinking and eating. This simulates a long-running
long concurrent process.
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 69
Technology, Varanasi
 Explaination Line by Line
Dining Philosophers Problem
• thinking() The philosopher is thinking — not using any shared resources.
This is the "safe" phase where no forks are held.
• wait take_fork(i) The philosopher tries to pick up the left fork (fork i). wait
is a semaphore operation (P operation) — if the fork is unavailable (held by
a neighbor), the philosopher blocks here until it's free.
• wait take_fork((i + 1) % n) The philosopher tries to pick up the right fork
(fork (i+1) % n). The modulo % n wraps around so the last philosopher
shares fork 0 with the first, forming a circle.
• eat() The philosopher eats — now holding both forks (left and right). This is
the critical section where two shared resources are in use simultaneously.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 70
Technology, Varanasi
 Explaination Line by Line
Dining Philosophers Problem
• signal put_fork(i) The philosopher puts down the left fork. signal is a
semaphore operation (V operation) — it releases fork i and wakes up
any neighbor that was blocked waiting for it.
• signal put_fork((i + 1) % n) The philosopher puts down the right fork,
releasing fork (i+1) % n for the neighboring philosopher.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 71
Technology, Varanasi
 Explaination Line by Line
⚠ The Problem With This Code
• This naive implementation can cause a deadlock! If all philosophers
pick up their left fork at the same time, everyone waits for their right
fork forever — and nobody eats.
• Classic solutions include:
• Allow only n-1 philosophers to sit at once
• Make one philosopher pick up forks in reverse order
• Use a waiter/mutex to allow only one philosopher to pick up forks at
a time
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 72
Technology, Varanasi
 Inter-Process
Process Communication (IPC)
Inter-Process Communication (IPC) refers to the mechanisms provided
by an operating system that allow processes to communicate with each
other and synchronize their actions. Since processes run in isolated
memory spaces, they need special OS-supported
supported mechanisms to share data
or coordinate.
Why is IPC Needed?
• Processes are isolated by default (each has its own address space)
• Sometimes processes need to share data,
data coordinate tasks, or send signals
to each other
• IPC enables cooperation between concurrent or parallel processes
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 73
Technology, Varanasi
 Main IPC Mechanisms
1. Pipes A unidirectional channel where one process writes and
another reads. There are two types:
types
• Anonymous pipes – between related processes (parent-child)
• Named pipes (FIFOs) – between unrelated processes via a named
file
2. Message Queues Processes send and receive structured messages
via a queue managed by the OS. Messages are stored until the
receiver reads them, allowing asynchronous communication.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 74
Technology, Varanasi
 Main IPC Mechanisms
3. Shared Memory The fastest IPC method. Two or more processes
map the same region of memory into their address space and
read/write directly. Requires synchronization (e.g., semaphores) to
avoid race conditions.
4. Semaphores Not used for data transfer, but for synchronization. A
semaphore is a counter that controls access to a shared resource —
preventing race conditions between processes.
5. Sockets Allow communication between processes on the same
machine or across a network. Used extensively in client-server
architectures (e.g., web servers).
Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 75
Technology, Varanasi
 Main IPC Mechanisms
6. Signals Lightweight notifications sent to a process to indicate an
event (e.g., SIGKILL, SIGTERM). They carry no data, just an event ID.
7. Memory-Mapped Files A file is mapped into the virtual memory of a
process. Multiple processes mapping the same file can communicate
through it — similar to shared memory but backed by a file.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 76
Technology, Varanasi
 IPC Models: Two Core Approaches
Feature Shared Memory Message Passing

Fast (no kernel

Speed Slower (kernel mediates)
involvement after setup)

Higher (needs Lower (OS handles

Complexity
synchronization) coordination)

High-performance
performance data Distributed/networked
Use case
sharing systems

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 77
Technology, Varanasi
 Real-World
World Examples
• Pipes → Shell commands like ls | grep .txt
• Sockets → Browser talking to a web server
• Shared Memory → Database buffer pools
• Signals → Ctrl+C sending SIGINT to terminate a process
• Message Queues → Producer-consumer
consumer systems, job schedulers
IPC is fundamental to building multitasking, concurrent, and
distributed systems, and every major OS (Linux, Windows, macOS)
provides its own implementations of these mechanisms.

Mr. Ajay Maurya || Assistant Professor || Kashi Institute of

27-03-2026 78
Technology, Varanasi
 Mr. Ajay Maurya || Assistant Professor || Kashi Institute of
27-03-2026 79
Technology, Varanasi