An interrupt is a hardware mechanism used to inform the CPU that an asynchronous event occurred. When an interrupt is recognized, the CPU saves part (or all) of its context (i.e., registers) and jumps to a special subroutine called an Interrupt Service Routine (ISR). The ISR processes the event, and – upon completion of the ISR – the program either returns to the interrupted task, or the highest priority task, if the ISR made a higher priority task ready-to-run.
Interrupts allow a microprocessor to process events when they occur (i.e., asynchronously), which prevents the microprocessor from continuously polling (looking at) an event to see if it occurred. Task level response to events is typically better using interrupt mode as opposed to polling mode. Microprocessors allow interrupts to be ignored or recognized through the use of two special instructions: disable interrupts and enable interrupts, respectively.
In a real-time environment, interrupts should be disabled as little as possible. Disabling interrupts affects interrupt latency possibly causing interrupts to be missed.
Processors generally allow interrupts to be nested, which means that while servicing an interrupt, the processor recognizes and services other (more important) interrupts.
One of the most important specifications of a real-time kernel is the maximum amount of time that interrupts are disabled. This is called interrupt disable time. All real-time systems disable interrupts to manipulate critical sections of code and re-enable interrupts when critical sections are completed. The longer interrupts are disabled, the higher the interrupt latency.
Interrupt response is defined as the time between the reception of the interrupt and the start of the user code that handles the interrupt. Interrupt response time accounts for the entire overhead involved in handling an interrupt. Typically, the processor’s context (CPU registers) is saved on the stack before the user code is executed.
Interrupt recovery is de?ned as the time required for the processor to return to the interrupted code or to a higher priority task if the ISR made such a task ready-to-run.
Task latency is defined as the time it takes from the time the interrupt occurs to the time task level code resumes.