“…Till recently, this classical scheduling problem is deemed to be practically resolved as a heuristic-based algorithm is shown to be able to achieve a rather high runtime schedulability considering real system overheads [10]. Besides this basic system model (i.e., sporadic workloads preemptively scheduled on homogeneous multiprocessor), extensive efforts have been made to handle increasing complexities seen in the task model, such as non-preemptive execution [2], [3], [4], precedence constraints [11], [12], [13], [14], selfsuspensions [15], [16], and mixed-criticality [17], [18], [19], [20], as well as those seen in the hardware platform, such as heterogeneous multiprocessors [21], [22], computing accelerators [23], [24], [25], and shared resources [26], [27], [28]. Unfortunately, although promising progress has been made for dealing with various complexities, many such complexities yield rather pessimistic results.…”