MTM2: Scalable Memory Management for Multi-tasking Managed Runtime Environments

Multi-tasking, managed runtime environments (MREs) for modern type-safe, object-oriented programming languages enable isolated, concurrent execution of multiple applications within a single operating system process. Multi-tasking MREs can potentially extract high-performance on modern desktop and hand-held systems through aggressive sharing of classes and compiled code, and by exploiting high-level dynamic program information for optimization across program executions. In this paper, we investigate the performance of a state-of-the-art multi-taking MRE for concurrent program execution and find that due to limited support for multi-tasking and performance isolation in the memory management subsystem, multi-tasking performs poorly compared to a production-quality, singletasking MRE. To address this limitation, we present MTM: a comprehensive memory management system for concurrent multi-tasking. MTM facilitates performance isolation and efficient heap space usage through on-demand allocation of application-private regions. Moreover, MTM mitigates fragmentation using a novel hybrid garbage collector that combines mark-sweep with opportunistic copying. Our empirical evaluation shows that MTM improves overall performance, scalability, and footprint for concurrent workloads over state-of-the-art, multiand single-tasking MREs.