Combining loop transformations considering caches and scheduling
Michael E. Wolf, Dror E. Maydan, Ding-Kai Chen · Proceedings of the ... annual International Symposium on Microarchitecture/Proceedings of the annual International Symposium on Microarchitecture · 2002
The performance of modern microprocessors is greatly affected by cache behavior, instruction scheduling, register allocation and loop overhead. High level loop transformations such as fission, fusion, tiling, interchanging and outer loop unrolling (e.g., unroll and jam) are well known to be capable of improving all these aspects of performance. Difficulties arise because these machine characteristics and these optimizations are highly interdependent. Interchanging two loops might, for example, improve cache behavior but make it impossible to allocate registers in the inner loop. Similarly, unrolling or interchanging a loop might individually hurt performance but doing both simultaneously might help performance. Little work has been published on how to combine these transformations into an efficient and effective compiler algorithm. In this paper we present a model that estimates total machine cycle time taking into account cache misses, software pipelining, register pressure and loop overhead. We then develop an algorithm to intelligently search through the various possible transformations, using our machine model to select the set of transformations leading to the best overall performance. We have implemented this algorithm as part of the MIPSPro commercial compiler system. We give experimental results showing that our approach is both effective and efficient in optimizing numerical programs.
Cited by 99 works. doi:10.1109/micro.1996.566468
Authors
References
- Compilers: Principles, Techniques, and Tools
- The cache performance and optimizations of blocked algorithms
- Software pipelining: an effective scheduling technique for VLIW machines
- Strategies for cache and local memory management by global program transformation
- Efficient and exact data dependence analysis
- Tile size selection using cache organization and data layout
- Compiler optimizations for improving data locality
- Estimating interlock and improving balance for pipelined architectures
- Improving the ratio of memory operations to floating-point operations in loops
- Software pipelining showdown