Performance analysis for complex embedded applications

Abstract

We consider complex embedded systems, where large, heterogeneous sets of communicating tasks are executed on heterogeneous multi-processor / multi-bus architectures with RTOSes and bus-arbitration. Examples include systems-on-chip for mobile communication, multimedia platforms or distributed automotive control systems. In such systems, data-dependent task execution times and preemption lead to data jitter and bursts, and consequently to sophisticated run-time interdependencies between tasks. Reliable validation of timing constraints in such systems can no longer be achieved through simulation due to incomplete corner-case coverage, but instead requires formal performance analysis. Existing system-level performance analysis techniques assume that the completion of one task immediately leads to the activation of a dependent task. However, activation dependencies in realistic embedded applications are often more complex, including multirate data dependencies, multiple activating inputs, cyclic dependencies or conditional communication. In these cases, existing performance analysis techniques are not directly applicable. In this paper, we focus on the calculation of activation timing in the presence of complex task dependencies. These calculations are then integrated into a system-level performance analysis framework for heterogeneous architectures. Using a larger example, we demonstrate how performance analysis allows rapid design-space exploration to optimize the implementation of realistic embedded applications while satisfying all critical timing constraints.

Monotonicity and run-time scheduling

“…schedule independent, and that the bounds on temporal behaviour we obtain from our analysis are independent of the container arrival times and execution times of other applications. Approaches like [11,14,6] have weaker restrictions on the application model and support a larger class of schedulers. However, these approaches both have a more difficult analysis problem and stronger conditions under which the analysis results hold.…”

A Priority-Based Budget Scheduler with Conservative Dataflow Model

“…Third, dataflow analysis techniques accurately compute a guaranteed throughput of the task graph that is independent of the best-case execution times of the tasks. Other analysis techniques [3,7,10] typically have difficulties with arbitrary cyclic dependencies, data-dependent execution rates of tasks, and rely on best-case execution times.…”

Efficient computation of buffer capacities for multi-rate real-time systems with back-pressure

“…Scheduling approaches that include run-time arbitration, as for instance presented by Jersak [6], Goddard [4], or Maxiaguine [10], do not allow feedback cycles that influence the temporal behaviour of the system. Not only do these cycles occur, because of functional constraints, this also means that back-pressure through bounded FIFO buffers cannot be applied in these approaches.…”