Polling and interrupts are two mechanisms that allow the CPU to communicate with peripheral devices such as keyboard, mouse, printer, etc. The main difference between polling and interrupts is that polling is a protocol in which the CPU continuously checks the status of a device to find whether it needs attention, while interrupts are hardware signals that indicate the CPU that a device requires its attention¹.

In polling, the CPU repeatedly checks the status of each device connected to it by reading a special bit called the command-ready bit. This bit indicates whether the device has some command or data to be processed by the CPU or not. If the bit is one, then the CPU reads the command or data from the device and performs the required action. If the bit is zero, then the CPU moves on to check the next device¹.

The advantage of polling is that it is simple and easy to implement. The disadvantage of polling is that it wastes a lot of CPU cycles by constantly checking the status of each device, even if they do not need any attention. Moreover, polling can cause latency and delay in servicing the devices, as the CPU has to wait for its turn to poll each device².

In interrupts, the device notifies the CPU that it needs attention by sending a hardware signal through a special line called the interrupt-request line. When the CPU receives an interrupt signal, it pauses the current process and transfers the control to an interrupt handler, which is a special program that services the device. After completing the service, the CPU resumes the previous process¹.

The advantage of interrupts is that they allow the CPU to respond to devices only when they need attention, thus saving CPU cycles and improving performance. The disadvantage of interrupts is that they require additional hardware and software components to handle them. Moreover, interrupts can cause complexity and confusion when multiple devices send interrupt signals at the same time².

There are two types of interrupts: hardware interrupts and software interrupts. Hardware interrupts are generated by external devices such as keyboard, mouse, printer, etc. Software interrupts are generated by internal events such as division by zero, illegal instruction, system call, etc¹.

Hardware interrupts can be further classified into maskable and non-maskable interrupts. Maskable interrupts are those that can be ignored or disabled by the CPU if they have lower priority than the current process. Non-maskable interrupts are those that cannot be ignored or disabled by the CPU as they have higher priority and urgency than any other process¹.

Here is a code example that demonstrates how to handle a keyboard interrupt in C: