Distributed run-time WCET controller for concurrent critical tasks in mixed-critical systems

Abstract

International audienceWhen integrating mixed critical systems on a multi/many-core, one challenge is to ensure predictability for high crit-icality tasks and an increased utilization for low criticality tasks. In this paper, we address this problem when several high criticality tasks with different deadlines, periods and offsets are concurrently executed on the system. We pro-pose a distributed run-time WCET controller that works as follows: (1) locally, each critical task regularly checks if the interferences due to the low criticality tasks can be toler-ated, otherwise it decides their suspension; (2) globally, a master suspends and restarts the low criticality tasks based on the received requests from the critical tasks. Our ap-proach has been implemented as a software controller on a real multi-core COTS system with significant gains 1

Timely Fine-Grained Interference-Sensitive Run-Time Adaptation of Time-Triggered Schedules

“…Following similar reasoning to the ready phase, the response time of the controller for the update phase is the cost of executing one iteration of the control-loop plus the waiting times to gain access to the required critical sections. The cost of one loop-iteration is C [11−12] +|K| * C [13][14][15]19] +deg + (v) * C [16−18] . The worst-case waiting time to gain access to the required critical sections is when all cores execute a ready phase or update phase (whichever is greater) that tries to gain access to the same critical sections.…”

“…For the update phase, the worst case for C T A U is equal to the execution of one iteration of the for-loop without taking the branch at line 16, i.e. |K| * C [13][14][15][16]19] . This time delay is sufficient because: i) the waiting times to gain access to the required critical sections have already been accounted for in the response time of the update phase, and ii) the branch is considered to be taken once, which has also been accounted for in the response time of the update phase.…”

“…The makespan of the isRA-FG is compared with the makespan given by (i) the global synchronisation approach (isRA-GLO), (ii) the time-triggered execution (isTT), where the isWCET computed offline is extended with the cost of the time-triggered mechanism for each task. The time-triggered mechanism cost is computed 270 cycles, since 70 cycles is required to write a control register and at least 200 cycles for serving an interrupt handling routine [15]. Each experiment has been executed ten consecutive iterations.…”

“…To further reduce the impact of the inherent pessimism in any kind of WCET estimations, several run-time mechanisms have been proposed, e.g. [15], [9]. In [9], the authors provide a run-time approach suitable for interference-sensitive WCET estimation.…”

Deterministic platform software for hard real-time systems using multi-core COTS

“…However, it involves speculation, reducing partitioning guarantees, as summarized in Figure 9. In [33] Kritikakou et al propose a distributed run-time WCET controller, another regulation software solution enabling the user to run a critical application at the same time as non-critical applications.…”