The JR24 implementation reduces compilation times, mainly due to reducing the number of files generated during JR program translation, without noticeably impacting execution times.
Moreover, we provide a security analysis and performance evaluation to estimate the impact on execution time and space consumption, and a conceptual implementation that validates the framework.
MBPTA also requires that those events with high impact on execution time are properly captured in the (R) runs made at analysis time.
However, those solutions only work for programs for which the impact on execution time of mapping any subset, bigger than W, of program addresses to a given set is the same.
This requires being aware of the actual access pattern of the program under analysis so that only those cache placements producing a high impact on execution time are considered.
The Representativeness Validation by Simulation (ReVS) method identifies the (conflictive) sets of (cache line) addresses, a C i, with high impact on execution time when they are randomly mapped to the same cache set.
The modelling framework allows the designer to verify the impact of execution platform and application mapping on the schedulability (meeting hard real-time requirements), power consumption and memory utilization, while taking communication into account.
Part of our future work consists of considering, in a first step, all (cache line) addresses, and quickly discarding those combinations that cannot be the most conflictive ones, i.e. those that if mapped to the same set cause a low impact on execution time.
Impacted Impacted Not impacted Yes Variation of the application Impacted Impacted Impacted Yes Variation of the execution time precision Not impacted Impacted Impacted Yes.
