A theoretical foundation for program transformations to reduce cache thrashing due to true data sharing

Theoretical Computer Science

Volumn 255, Issue 1-2, 2001, Pages 449-481

  Jin, Guohua ^a   Li, Zhiyuan ^b   Chen, Fujie ^c  

^a Rice University   (United States)

^b Purdue University   (United States)

^c Changsha Institute of Technology   (China)

Author keywords

Caches; Complier algorithms; Linear algebra; Multiprocessors; Theory of numbers

Indexed keywords

ALGORITHMS; CACHE MEMORY; COMPUTER SYSTEMS PROGRAMMING; LINEAR ALGEBRA; NUMBER THEORY; PARALLEL PROCESSING SYSTEMS;

CACHE THRASHING; TRUE DATA SHARING;

STORAGE ALLOCATION (COMPUTER);

EID: 0034917963     PISSN: 03043975     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0304-3975(99)00313-8     Document Type: Article

References (38)

1. Compile-time partitioning of iterative parallel loops to reduce cache coherence traffic
2. On the performance enhancement of paging systems through program analysis and transformations
3. Automatic partitioning of parallel loops and data arrays for distributed shared-memory multiprocessors
4. Automatic loop interchange
5. Multilevel cache hierarchies: Organizations, protocols, and performance
6. Improving register allocation for subscripted variables
7. Compiling scientific code for complex memory hierarchies
8. Partitioning and labeling of loops by unimodular transformations
9. Eliminating false sharing
10. A solution of cache ping-pong problem in RISC based parallel processing systems
11. Cache or local memory thrashing and compiler strategy in parallel processing systems
12. An iteration partition approach for cache or local memory thrashing on parallel processing
13. Performance optimizations, implementation, and verification of the SGI Challenge multiprocessor
14. On the problem of optimizing data transfers for complex memory systems
15. Strategies for cache and local memory management by global program transformation
16. Effects of program parallelization and stripmining transformation on cache performance in a multiprocessor
17. Compiler techniques for data partitioning of sequentially iterated parallel loops
18. Reducing false sharing on shared memory multiprocessors through compile-time data transformations
19. The design and the implementation of a knowledge-based parallelizing tool
20. An efficient solution to the cache thrashing problem caused by true data sharing
21. Loop restructuring techniques for the thrashing problem
22. Loop staggering, loop staggering and loop compacting: Restructuring techniques for the thrashing problem
23. Dependence graphs and compiler optimizations
24. A solution of the cache ping-pong problem in multiprocessor systems
25. Minimum distance: A method for partitioning recurrences for multiprocessors
26. Guided self-scheduling: A practical scheduling scheme for parallel supercomputers
27. Time optimal linear schedules for algorithms with uniform dependencies
28. An empirical study of Fortran programs for parallelizing compilers
29. Iteration partitioning for resolving stride conflicts on cache-coherent multiprocessors
30. False sharing and spatial locality in multiprocessor caches
31. A data locality optimizing algorithm
32. More iteration space tiling