Evolving operating systems and architectures: how do kernel implementors catch up?

The author discusses the trade-off between the conflicting goals of portability and performance in operating system kernel implementations. Using traditional programming techniques, code refinements that yield high performance inevitably lead to low portability. Ideas from the Synthesis kernel implementation are described as possible techniques towards a way out of this trade-off.<<ETX>>

[1]  H. Zimmermann,et al.  Basic Concepts for the Support of Distributed Systems: The Chorus Approach , 1981, ICDCS.

[2]  David R. Cheriton,et al.  The V distributed system , 1988, CACM.

[3]  Calton Pu,et al.  Threads and input/output in the synthesis kernal , 1989, SOSP '89.

[4]  Jonathan Walpole,et al.  Modularity and Interfaces in Micro-kernel Design and Implementation: A Case Study of Chorus on the HP PA-Risc , 1992, USENIX Workshop on Microkernels and Other Kernel Architectures.

[5]  Calton Pu,et al.  Quaject composition in the Synthesis kernel , 1991, Proceedings 1991 International Workshop on Object Orientation in Operating Systems.

[6]  Calton Pu,et al.  The Synthesis Kernel , 1988, Comput. Syst..

[7]  Calton Pu,et al.  Reimplementing the Synthesis Kernel , 1992, USENIX Workshop on Microkernels and Other Kernel Architectures.

[8]  Calton Pu,et al.  Fine-Grain Adaptive Scheduling using Feedback , 1989, Comput. Syst..