Revisiting the Cache Interference Costs of Context Switching

The high cost of context switching is one reason that operating system performance is not keeping pace with hardware improvements. Besides the cost of saving and restoring registers, another component of context switch cost is the cache interference between multiple processes sharing the same cache. We measured kernel on user, user on kernel, and user on user cache interference from context switching using a complete machine simulator. While the resulting cache interference is noticeable, the time between process switches is generally sufficient to amortize this cost over a large number of instructions. Completely eliminating all context-switch cache interference would therefore have a minimal impact on total execution time.

[1]  Brian N. Bershad,et al.  The impact of operating system structure on memory system performance , 1994, SOSP '93.

[2]  Katherine Yelick,et al.  A Case for Intelligent DRAM: IRAM , 1998 .

[3]  John K. Ousterhout,et al.  Why Aren't Operating Systems Getting Faster As Fast as Hardware? , 1990, USENIX Summer.

[4]  K. Yelick,et al.  Intelligent RAM (IRAM): chips that remember and compute , 1997, 1997 IEEE International Solids-State Circuits Conference. Digest of Technical Papers.

[5]  Carl Staelin,et al.  lmbench: Portable Tools for Performance Analysis , 1996, USENIX Annual Technical Conference.

[6]  Sharon E. Perl,et al.  Studies of Windows NT performance using dynamic execution traces , 1996, OSDI '96.

[7]  Jeffrey C. Mogul,et al.  The effect of context switches on cache performance , 1991, ASPLOS IV.

[8]  Brian N. Bershad,et al.  The interaction of architecture and operating system design , 1991, ASPLOS IV.

[9]  Brian N. Bershad,et al.  The interaction of architecture and operating system design , 1991, ASPLOS IV.

[10]  Zarka Cvetanovic,et al.  Performance characterization of the Alpha 21164 microprocessor using TP and SPEC workloads , 1996, Proceedings. Second International Symposium on High-Performance Computer Architecture.

[11]  Ann Marie Grizzaffi Maynard,et al.  Contrasting characteristics and cache performance of technical and multi-user commercial workloads , 1994, ASPLOS VI.

[12]  Anoop Gupta,et al.  Complete computer system simulation: the SimOS approach , 1995, IEEE Parallel Distributed Technol. Syst. Appl..

[13]  David A. Patterson,et al.  Computer Architecture: A Quantitative Approach , 1969 .

[14]  Mark Horowitz,et al.  Cache performance of operating system and multiprogramming workloads , 1988, TOCS.