Abstract | ||
---|---|---|
In multi-tasking real-time systems,inter-task cache interference due to preemptions degrades schedulabilityas well as performance. To address this problem, we propose anovel scheduling scheme, called limited preemptive scheduling(LPS), that limits preemptions to execution points with smallcache-related preemption costs. Limiting preemptions decreasesthe cache-related preemption costs of tasks but increases blockingdelay of higher priority tasks. The proposed scheme makes anoptimal trade-off between these two factors to maximize the schedulabilityof a given task set while minimizing cache-related preemptiondelay of tasks. Experimental results show that the LPS schemeimproves the maximum schedulable utilization by up to 40\% comparedwith the traditional fully preemptive scheduling (FPS) scheme.The results also show that up to 20\% of processor time is savedby the LPS scheme due to reduction in the cache-related preemptioncosts. Finally, the results show that both the improvement ofschedulability and the saving of processor time by the LPS schemeincrease as the speed gap between the processor and main memorywidens. |
Year | DOI | Venue |
---|---|---|
1999 | 10.1023/A:1008194508956 | Realtime systems |
Keywords | Field | DocType |
Real-time systems,cache memory,fixed-priority scheduling,preemption | Fixed-priority pre-emptive scheduling,Preemption,CPU cache,Computer science,Cache,Scheduling (computing),Parallel computing,Real-time computing,Rate-monotonic scheduling,Dynamic priority scheduling,Limiting,Distributed computing | Journal |
Volume | Issue | ISSN |
17 | 2-3 | 1573-1383 |
Citations | PageRank | References |
5 | 0.50 | 11 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Sheayun Lee | 1 | 56 | 5.79 |
Sang Lyul Min | 2 | 2212 | 208.29 |
Chong Sang Kim | 3 | 629 | 57.96 |
Chang-Gun Lee | 4 | 1033 | 66.13 |
Minsuk Lee | 5 | 477 | 32.71 |