An implementation of storage management in capability environments

Abstract

The exploitation of the salient features of capability‐based addressing environments leads to a high number of small objects existing in memory at the same time. It is thus necessary to enhance the efficiency of the mechanisms for object relocation, and to avoid congestion of input/output devices due to swapping. In this paper, we present an approach to the management of a large virtual memory space aimed at solving these problems. We insert partial information concerning the physical allocation of each object into the virtual identifier of this object. Objects are grouped into large swapping units, called pages. The page size is independent of the average object size. This results in enhanced efficiency in managing the relocation information both with regard to memory requirements and access times. The allocation of objects into pages, and the movement of pages through the memory hierarchy, are controlled by user processes. This means that programs which have knowledge of their own use of virtual memory can increase their locality of reference, diminish the number of swap operations and reduce fragmentation.

Single address space implementation in distributed systems

“…We have analysed this overhead taking a node configuration into consideration, that is representative of a typical workstation. Forms of software control over the storage hierarchy have been proposed in the past with reference to single-processor architectures [39,[42][43][44]. The cost of distribution is low even if compared with a single-processor machine.…”

“…The form of manual memory management that can be carried out by using these operations is in contrast with the demandpaged memory management often adopted in single address space systems, implemented at the level of the operating system kernel [11][12][13]. Forms of software control over the storage hierarchy have been proposed in the past with reference to single-processor architectures [39,[42][43][44]. The system we have presented in this paper extends the application of this concept to a distributed framework.…”