A timer provides software with a means of being asynchronously notified when a designated amount of time has passed. All kinds of software, both in user and kernel space, require timers; this commonly includes network protocol implementations, block layer code, device drivers, and various kernel subsystems. This timer provides a means of asynchronous notification, thus allowing the driver to execute work in parallel with the running timer. An important question that arises is, how will I know when the timer expires? In user space apps, typically, the kernel sends a signal to the relevant process (the signal is typically SIGALRM).
In kernel space, it's a bit nuanced. As you will know from our discussion on top and bottom halves for hardware interrupts (see Chapter 4, Handling Hardware Interrupts, the Understanding and using top and bottom halves section), after the timer interrupt's top half (or ISR) completes, the kernel will ensure...
