CPU Scheduling concepts in OS: Preemptive and Non-Preemptive

 

CPU Scheduling:

 

The CPU scheduling serves as a basis for multi programmed OS. The idea of multiprogramming is simple; if a process or job is waiting for an IO request then the CPU switches from that job to another job. So, the CPU is always kept busy in multiprogramming as it has something to execute at all times. However, in case of a simple computer system, CPU often sits idle until the IO request is granted. Thus, by switching the CPU among processes, an OS can increase productivity and utilization of a computer system.

 

Scheduling is a fundamental operating system function. Managing the CPU resources is commonly termed as Scheduling. All of the computer resources are scheduled before their use. The CPU is one of the primary resources amongst all resources. Hence, the CPU is also scheduled before use.

 

An allocation of the CPU to processes is determined using various CPU scheduling algorithms. There are two types of CPU scheduling algorithms:

1.    Non- preemptive scheduling algorithms

2.    Preemptive scheduling algorithms

 

A CPU has to make scheduling decisions under the following circumstances when a process switches from:

1.    running to waiting state

2.    running to ready state

3.    waiting to ready state or;

4.    when a process terminates.

 

In non-preemptive scheduling, once the CPU is assigned to a process, the processor do not release until the completion of that process. The CPU will be assigned to some other job only after the previous job has finished. Non – preemptive scheduling takes place for circumstances 1 and 4 mentioned above.  MS Windows 3.1 and Apple’s Mac uses this method of CPU scheduling. It is relatively very simple to implement.

 

However, in preemptive scheduling, the CPU can release the processes even in the middle of its execution. Consider an example where the CPU is executing process P1. While it is executing, the CPU receives a request signal from process P2.  Then, in this case an OS compares the priorities of processes P1 and P2. If the priority of P1 > P2, then the CPU continues with the execution of process P1. Otherwise, CPU preempts process P1 and assigns it to process P2. Preemptive scheduling takes place for circumstances 2 and 3 mentioned above.

The preemptive scheduling techniques are used to ensure quick response to high priority processes. But, they incur a cost. It is more responsive but it imposes higher overhead because it requires a process switching.

Differences between Preemptive and Non-Preemptive Scheduling:

Preemptive Scheduling	Non-Preemptive Scheduling
It allows a process to be interrupted in the midst of its execution by taking the CPU and allocating it to another process.	It ensures that process gives control of the CPU only when it finishes with its current CPU burst.
It incurs cost associated with access to shared data.	Does not incur cost.
It also affects design of OS kernel.	It does not affect design of OS kernel.
More complex to implement.	Simple to implement but, very inefficient.
Ex: Round Robin (RR) method.	Ex: First Come First Serve (FCFS) method.