The integration of virtual memory management and interprocess communication in Accent

Fitzgerald, Robert P.; Rashid, Richard F.

doi:10.1145/214419.214422

Cited by 69 publications

(16 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In support of the second claim, Table 5 shows performance figures for message-passing in DASH and other current systems: Accent [2], LYNX [13], Mach [14], Quicksilver [9], Topaz [15], and the V system [16,17].…”

Section: The Effectiveness Of the Dash Designmentioning

confidence: 90%

“…For example, Accent takes 1.15 milliseconds to send a short message without using VM remapping, and 10 milliseconds for a message with 1KB of data remapped [2]. The sources of overhead include 1) buffer allocation, 2) manipulating VM system data structures representing remapped objects, 3) adjusting the representation of a message to reflect address changes due to remapping, and 4) remapping objects twice (from a user VAS to the kernel, and from the kernel to another user VAS).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Performance Evaluation of the DASH Message-Passing System

Tzou¹,

Anderson²

1988

View full text Add to dashboard Cite

DASH is a distributed operating system kernel. Message-passing (MP) is used for local communication, and the MP system uses virtual memory remapping instead of software copying for moving large amounts of data between virtual address spaces. This design eliminates a bottleneck in high-performance communication, and increases the feasibility of moving services such as file services into user spaces.Other systems that have used VM remapping for message transfer have suffered from high per-operation overhead, limiting the use of the technique. The DASH design is intended to reduce this overhead. To evaluate our design, we measured the performance of the DASH kernel implementation on Sun 3/50 workstations. Our throughput measurements show that large messages can be moved between user spaces at a rate of more than 30 MB/sec, an order of magnitude higher than with software copying. Furthermore, the per-operation overhead is low, so performance for small messages is not sacrificed.To further understand the performance of the DASH MP system, we then broke an MP operation into short code segments and timed them with microsecond precision. The results show the relative costs of data movement and the other components of MP operations, and allow us to evaluate several specific design decisions. Report Documentation PageForm Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. DASH is a distributed operating system kernel. Message-passing (MP) is used for local communication, and the MP system uses virtual memory remapping instead of software copying for moving large amounts of data between virtual address spaces. This design eliminates a bottleneck in high-performance communication, and increases the feasibility of moving services such as file services into user spaces. Other systems that have used VM remapping for message transfer have suffered from high per-operation overhead, limiting the use of the technique. The DASH design is intended to reduce this overhead. To evaluate our design, we measured the performance of the DASH kernel implementation on Sun 3/50 workstations. Our throughput measurements show that large messages can be moved between user spaces at a rate of more than 30 MB/sec, an order of magnitude higher than with software copying. Furthermor...

show abstract

Section: The Effectiveness Of the Dash Designmentioning

confidence: 90%

Section: Introductionmentioning

confidence: 99%

A Performance Evaluation of the DASH Message-Passing System

Tzou¹,

Anderson²

1988

View full text Add to dashboard Cite

show abstract

“…This has proven adequate, however, for a range of systems tasks including transaction processing, window management, and even speech recognition. Overall performance of Accent, for example, is in the range of comparable traditional operating systems such as UNIX [8].…”

Section: Discussionmentioning

confidence: 99%

Mach and Matchmaker: kernel and language support for object-oriented distributed systems

Jones

Rashid

1986

SIGPLAN Not.

Self Cite

View full text Add to dashboard Cite

Mach, a multiprocessor operating system kernel providing capability-based interprocess communication, and Matchmaker, a language for specifying and automating the generation of multi-lingual interprocess communication interfaces, are presented. Their usage together providing a heterogeneous, distributed, object-oriented programming environment is described. Performance Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Association for Computing Machinery. To copy otherwise, or to republish, requires a fee and/or specific permission. THE VIEWS AND CONCLUSIONS CONTAINED IN THIS DOCUMENT ARE THOSE OF THE AUTHORS AND SHOULD NOT BE INTERPRETED AS REPRESENTING THE OFFICIAL POLICIES, EITHER EXPRESSED OR IMPLIED, OF THE DEFENSE ADVANCED RESEARCH PROJECTS AGENCY OR THE U.S. GOVERNMENT.© 1986 ACM 0-89791-204-7/86/0900-0067 IntroductionMany different object-oriented systems have been built and proposed.Some are inapplicable to distributed environments. Design and implementation considerations have often been driven by abstract issues rather than the concrete requirements of a distributed system. Many systems assume that all code will be written in a single programming language, ignoring the diversity of existing applications and programming tools already available. Some do provide multiple languages, but allow inter-language and inter-machine interfaces to be coded by hand in an ad-hoc manner. Many systems ignore protection issues, particularly for distributed applications. Others fail to provide efficient mechanisms for passing objects, and invoking operations. Finally, while an object-oriented programming style is supported for "applications code," the operating system itself is often not extensible, or implemented using the same abstractions.

show abstract

“…Internal manipulation of nodes and links allows multiple perspectives to share large information sources without unnecessary duplication. The use of "virtual copying" or "delta storage" is well-known in system software [7], but was not previously used in CSCW hypermedia systems. We have chosen to implement our own hypermedia substrate -rather then use something like Lotus Notes -for reasons of granularity, control, and speed.…”

Section: Hypertext and Hypermediamentioning

confidence: 99%