Operating System Tutorial

What is Operating System Evolution of Operating System Types of Operating System Functions of Operating System What is Kernel and Types of Kernel Operating System Properties Operating System Services Components of Operating System Needs of the Operating System Linux Operating System Unix Operating System Ubuntu Operating System What is DOS Operating System Difference Between Multi-programming and Multitasking What is Thread and Types of Thread Process Management Process State What is Process Scheduler and Process Queue What is Context Switching What is CPU Scheduling Scheduling Algorithm FCFS (First-come-First-Serve) Scheduling SJF (Shortest Job First) Scheduling Round-Robin CPU Scheduling Priority Based Scheduling HRRN (Highest Response Ratio Next) Scheduling Process Synchronization Lock Variable Mechanism TSL Mechanism Turn Variable Mechanism Interested Variable Mechanism What is Producer-Consumer Problem What is Semaphore in Operating System Monitors in Operating System What is Deadlock Deadlock Avoidance Strategies for Handling Deadlock Deadlock Prevention Deadlock Detection and Recovery Resource Allocation Graph Banker’s Algorithm in Operating System Fixed Partitioning and Dynamic Partitioning Partitioning Algorithms What is Paging and Segmentation What is Demand Paging What is Virtual Memory Disk Scheduling Algorithms FCFS and SSTF Disk Scheduling Algorithm SCAN and C-SCAN Disk Scheduling Algorithm Look and C-Look Disk Scheduling Algorithm File in Operating System File Access Methods in Operating System File Allocation Method Directory Structure in Operating System Difference between C-LOOK and C-SCAN Difference between Rotational Latency and Disk Assess Time Trap vs Interrupt How to implement Monitors using Semaphores N-Step-SCAN Disk Scheduling Why is it critical for the Scheduler to distinguish between I/O-bound and CPU-bound programs Difference between C-SCAN and SSTF Difference between SCAN and FCFS Difference between Seek Time and Disk Access Time Difference between SSTF and LOOK

Deadlock Prevention in Operating System

Deadlock Prevention

The conditions which we used for deadlock prevention are:

  1. Mutual Exclusion
  2. Hold and Wait
  3. No Preemption
  4. Circular Wait

1. Mutual Exclusion: - From the resource point of view, the mutual exclusion means that simultaneously more than one process cannot use the same resource. However, this is fair enough, but due to this, a deadlock occurs. If there is a possibility that we can use the same resource for more than one process at a time, there will be no process that will be waiting for the resource.

Although, if we can stop the resources performing in a mutually exclusive manner, then in this way, we can prevent the system from the deadlock.

 Spooling: - Spooling stands for Simultaneous Peripheral Operations On-line. We use spooling for devices such as Printer, Mouse, Keyboard.  Let’s understand the working spooling process of a printer:

  • In a printer, memory is associated with it.
  • With the help of this memory, we can store the jobs of all the processes.
  • Then all the jobs are collected, and after gathering all the jobs, the printer prints each of the jobs in FCFS (First Come First Serve) manner. If we use this procedure, then no process will need to wait for the printer.
  • Finally, when the outputs are produced, the printer collects them.

So, with the help of Spooling, we can stop mutual exclusion, but we may suffer two kinds of problems:

  1. We cannot use spooling for every resource.
  2. In spooling, a rare condition between the processes may occur in which the processes want some space in the spool. At the same time, we cannot force to use the same resources for more than one process. Because this approach fails here and if we are doing this, then there may be a possibility that some serious problems may arise related to the process. So, we can say that practically the violation of mutual exclusion is not possible.

2. Hold and Wait: - Hold and Wait is a condition where a process holds a resource and still waits for other resources to finish its job. In this situation, there is a possibility of deadlock because more than one process holding one resource each and cyclically wait for other resources for their execution. So, we need to find some procedure through which either process does not hold any resource or does not wait for any resource. This means we should assign all the resources which the process needs before starting its execution. Then, the execution of the process starts without waiting for any resource.

Practically we can implement this only if in starting the process, determine all the resources which the process needs. Although it sounds very practical, in the computer system, we cannot do this; the reason is at the start, no process can define the necessary resources.

Process means a set of instructions that are executed by the CPU. At multiple times each instruction requests for different resources. But the demand for the resources cannot be fixed by the operating system.

The approach has the following issues:

It is not possible, practically.

There may be a chance of starvation because, in some cases, the process can hold a resource for a long time.

3. No Preemption: - The reason for the deadlock occurrence is that once the process starts its execution, then it cannot be halt. But we can prevent deadlock if we take away those resources from the process that may cause deadlock. But this is not a suitable approach because if we bring out a resource that is being used by the process, then this will be inconsistent with the work we have done until now.

For example: Suppose we have a printer that a process is using, and we take the printer away from that process and assign the printer to some other process. So in this way, the data which is printed by the printer become inconsistent as well as ineffective. Also, the process will not start printing further from where it left the process so, due to this performance can be inefficient.

4. Circular Wait: - Circular wait is the condition in which one or more process waits for the required resources in a circular order. By assigning the priority number to every resource, we can solve the problem of a circular wait. The process cannot request a resource that has lesser priority value. It assures that no process should demand the resource that is being used by the other process. Hence, no cycle will be formed.

So, by examining all these conditions, we found that there is only one approach that we can implement practically for preventing deadlock, and that approach is a violation of circular wait.



ADVERTISEMENT
ADVERTISEMENT