Improving register allocation for subscripted variables
David P. Callahan, Steve Carr, Ken Kennedy · ACM SIGPLAN Notices · 1990
Most conventional compilers fail to allocate array elements to registers because standard data-flow analysis treats arrays like scalars, making it impossible to analyze the definitions and uses of individual array elements. This deficiency is particularly troublesome for floating-point registers, which are most often used as temporary repositories for subscripted variables. In this paper, we present a source-to-source transformation, called scalar replacement , that finds opportunities for reuse of subscripted variables and replaces the references involved by references to temporary scalar variables. The objective is to increase the likelihood that these elements will be assigned to registers by the coloring-based register allocators found in most compilers. In addition, we present transformations to improve the overall effectiveness of scalar replacement and show how these transformations can be applied in a variety of loop nest types. Finally, we present experimental results showing that these techniques are extremely effective—capable of achieving integer factor speedups over code generated by good optimizing compilers of conventional design.
Cited by 33 works. doi:10.1145/93548.93553
Authors
References
- LINPACK Users' Guide
- Department of Computer Science
- Register allocation via coloring
- PFC: A Program to Convert Fortran to Parallel Form
- Supernode partitioning
- More iteration space tiling
- Proceedings of the 1999 International Conference on Parallel Processing
- Automatic loop interchange
- Blocking Linear Algebra Codes for Memory Hierarchies
- Supercomputers - Design and Applications