Answer: A
Correct answer: Paging to disk is often avoided so that timing is predictable
A real-time system is judged by whether it meets deadlines, not by average throughput, so predictability matters more than efficiency. Demand paging introduces a page fault whose cost is several milliseconds and whose occurrence cannot be predicted, which is intolerable where a response must be guaranteed within a bounded time. Such systems therefore either avoid paging altogether or allow critical pages and code to be pinned in memory - locked so the replacement algorithm cannot evict them. Many use fixed contiguous allocation, accepting the fragmentation in exchange for determinism. The same reasoning runs through the rest of the design: preemptive priority scheduling with bounded latency, priority inheritance to avoid unbounded priority inversion, and interrupt handlers kept short. The distinction is between hard real-time systems, where a missed deadline is a failure - flight controls, medical devices - and soft ones, where it merely degrades quality, as in video playback. Only the former forbids paging outright.