Shared cache partitioning is a well-known technique used in multicore real-time systems to isolate task workloads and improve system predictability. Presently, the state-of-the-art studies that evaluate shared cache partitioning on multicore processors lack two key issues. First, the cache partitioning mechanism is typically implemented either in a simulation environment or in a general-purpose OS, and so the impact of kernel activities, such as interrupt handlers and context switching, on the task partitions tend to be overlooked.