An executable specification and verifier for relaxed memory order

被引:20
作者
Park, SJ [1 ]
Dill, DL
机构
[1] NASA, RIACS, Ames Res Ctr, Moffett Field, CA 94035 USA
[2] Stanford Univ, Dept Comp Sci, Stanford, CA 94305 USA
关键词
multiprocessors; memory models; formal method; executable specification; automatic verification;
D O I
10.1109/12.752664
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
The Mur phi description language and verification system for finite-state concurrent systems is applied to the problem of specifying a family of multiprocessor memory models described in the SPARC Version 9 architecture manual. The description language allows for a straightforward operational description of the memory model which can be used as a specification for programmers and machine architects. The automatic verifier can be used to generate all possible outcomes of small assembly language multiprocessor programs in a given memory model, which is very helpful for understanding the subtleties of the model. The verifier can also check the correctness of assembly language programs including synchronization routines. This paper describes the memory models and their encoding in the Mur phi description language. We describe how synchronization routines can be verified and how finite stale programs can be analyzed. We also present some interesting findings from the verification and the analysis.
引用
收藏
页码:227 / 235
页数:9
相关论文
共 22 条
[1]  
Adve S.V., 1990, P INT C PARALLEL PRO, P47
[2]  
Adve Sarita V., 1990, P ISCA, P2, DOI [10.1145/325164.325100, DOI 10.1145/325164.325100]
[3]  
[Anonymous], 1994, POWERPC ARCHITECTURE
[4]  
[Anonymous], 1996, LNCS, DOI DOI 10.1007/3-540-61474-586
[5]  
[Anonymous], 1992, SPARC ARCHITECTURE M
[6]  
Ben-Ari M., 1990, Principles of Concurrent and Distributed Programming
[7]  
COLLIER WW, 1992, REASONING PARALLEL A
[8]  
DILL DL, 1993, RESEARCH ON INTEGRATED SYSTEMS, P38
[9]  
DUBOIS M, 1986, P 13 ANN INT S COMP, P434, DOI DOI 10.1145/17407.17406
[10]   PROGRAMMING FOR DIFFERENT MEMORY CONSISTENCY MODELS [J].
GHARACHORLOO, K ;
ADVE, SV ;
GUPTA, A ;
HENNESSY, JL ;
HILL, MD .
JOURNAL OF PARALLEL AND DISTRIBUTED COMPUTING, 1992, 15 (04) :399-407