Abstract

We discuss the issues about the interdependency between code scheduling and register allocation. We present two methods as solutions: (1) an integrated code scheduling technique; and (2) a DAGdriven register allocator. The integrated code scheduling method combines two scheduling techniques- one to reduce pipeline delays and the other to minimize register usage- into a single phase. By keeping track of the number of available registers, the scheduler can choose the appropriate scheduling technique to schedule a better code sequence. The DAG-driven register allocator uses a dependency graph to assist in assigning registers; it introduces much less extra dependency than does an ordinary register allocator. For large basic blocks, both approaches were shown to generate more efficient code sequences than conventional techniques in the simulations. 1.

Citations

312	An efficient algorithm for exploiting multiple arithmetic units - Tomasulo - 1967
273	Bulldog: A compiler for VLIW architectures - Ellis - 1986
143	The Architecture of Pipelined Computers - Kogge - 1981
95	T.: Postpass Code Optimization of Pipeline Constraints - Hennessy, Gross - 1983
49	A Study of Scalar Compilation Techniques for Pipelined Supercomputers - Weiss, Smith - 1987
43	VLSI processor architecture - Hennessy - 1984
20	A.R.: Unrolling Loops - Dongarra, Jinds - 1979
16	A Retargetable Instruction Reorganizer - Davidson - 1986
1	Register Allocation aud Code Scheduling for Load/Store Architectures - Hsu - 1987
1	GaAs Microprocessor - Milutinovic - 1986
1	Advanced Compiler dptimizations for Supercomputers,” Communication of the ACM - Padua, M - 1986
1	Detection and Parallel Execution of Independent Iswtructions - Tjaden, Flynn - 1970
1	Evaluation of a Decoupled Computer Atchitecture and the Design of A Vector Extension,” Computer Sciences - Young - 1985