Combinatorial Register Allocation and Instruction Scheduling

TL;DR

This paper presents Unison, a practical combinatorial optimization approach for register allocation and instruction scheduling that improves code quality and scales to medium-sized functions, outperforming traditional compiler methods.

Contribution

It is the first to practically leverage combinatorial optimization for complete program transformations in compiler design, integrated with LLVM and capable of handling larger functions.

Findings

Unison generates better code than LLVM across multiple architectures.

It scales to functions with up to 946 instructions, surpassing previous methods.

Estimated speedups range from 1.1% to 10%, with code size reductions of 1.3% to 3.8%.

Abstract

This paper introduces a combinatorial optimization approach to register allocation and instruction scheduling, two central compiler problems. Combinatorial optimization has the potential to solve these problems optimally and to exploit processor-specific features readily. Our approach is the first to leverage this potential in practice: it captures the complete set of program transformations used in state-of-the-art compilers, scales to medium-sized functions of up to 1000 instructions, and generates executable code. This level of practicality is reached by using constraint programming, a particularly suitable combinatorial optimization technique. Unison, the implementation of our approach, is open source, used in industry, and integrated with the LLVM toolchain. An extensive evaluation confirms that Unison generates better code than LLVM while scaling to medium-sized functions. The evaluation uses systematically selected benchmarks from MediaBench and SPEC CPU2006 and different processor architectures (Hexagon, ARM, MIPS). Mean estimated speedup ranges from 1.1% to 10% and mean code size reduction ranges from 1.3% to 3.8% for the different architectures. A significant part of this improvement is due to the integrated nature of the approach. Executing the generated code on Hexagon confirms that the estimated speedup results in actual speedup. Given a fixed time limit, Unison solves optimally functions of up to 946 instructions, nearly an order of magnitude larger than previous approaches. The results show that our combinatorial approach can be applied in practice to trade compilation time for code quality beyond the usual compiler optimization levels, identify improvement opportunities in heuristic algorithms, and fully exploit processor-specific features.