Generation Scavenging
David Ungar
Abstract
David Ungar
Abstract
Many interactive computing environments provide automatic storage reclamation and virtual memory to ease the burden of managing storage. Unfortunately, many storage reclamation algorithms impede interaction with distracting pauses. Generation Scavenging is a reclamation algorithm that has no noticeable pauses, eliminates page faults for transient objects, compacts objects without resorting to indirection, and reclaims circular structures, in one third the time of traditional approaches. We have incorporated Generation Scavenging in Berkeley Smalltalk(BS), our Smalltalk-80 implementation, and instrumented it to obtain performance data. We are also designing a microprocessor with hardware support for Generation Scavenging .
OpenAlex reports 485 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Many interactive computing environments provide automatic storage reclamation and virtual memory to ease the burden of managing storage. Unfortunately, many storage reclamation algorithms impede interaction with distracting pauses. Generation Scavenging is a reclamation algorithm that has no noticeable pauses, eliminates page faults for transient objects, compacts objects without resorting to indirection, and reclaims circular structures, in one third the time of traditional approaches. We have incorporated Generation Scavenging in Berkeley Smalltalk(BS), our Smalltalk-80 implementation, and instrumented it to obtain performance data. We are also designing a microprocessor with hardware support for Generation Scavenging .
Key concepts: Computer science, Smalltalk, Scavenging, Operating system, Embedded system, Object-oriented programming, Chemistry, Antioxidant