Modeling Universal Instruction Selection

Abstract

Abstract. Instruction selection implements a program under compilation by selecting processor instructions and has tremendous impact on the performance of the code generated by a compiler. This paper introduces a graph-based universal representation that unifies data and control flow for both programs and processor instructions. The representation is the essential prerequisite for a constraint model for instruction selection introduced in this paper. The model is demonstrated to be expressive in that it supports many processor features that are out of reach of state-of-the-art approaches, such as advanced branching instructions, multiple register banks, and SIMD instructions. The resulting model can be solved for small to medium size input programs and sophisticated processor instructions and is competitive with LLVM in code quality. Model and representation are significant due to their expressiveness and their potential to be combined with models for other code generation tasks.

Finding Subgraphs with Side Constraints

“…Finding small "pattern" graphs inside larger "target" graphs is a widely applicable hard problem, with applications including compilers [5], bioinformatics [6,16], chemistry [29], malware detection [8], pattern recognition [17], and the design of mechanical locks [35]. This has led to the development of numerous dedicated algorithms, with the Glasgow Subgraph Solver [24] being the current state of the art [33].…”

Graph Transformation

“…The subgraph isomorphism family of problems involves finding a small "pattern" graph inside a larger "target" graph, or establishing that the pattern does not occur. When the pattern graph is part of the input, these problems are NP-complete; despite this, subgraph isomorphism algorithms are widely used in practice, including for model checking [23], for law enforcement [9], in biological applications [1,6,20], for compiler implementation [5], in designing mechanical locks [27], and inside graph databases [19]. This has encouraged the development of practical subgraph isomorphism algorithms, which fall into two categories: those based upon backtracking and connectivity [8,6,7], and those based upon constraint programming [25,4,18,15,3].…”

The Glasgow Subgraph Solver: Using Constraint Programming to Tackle Hard Subgraph Isomorphism Problem Variants

“…The subgraph isomorphism family of problems involves finding a small "pattern" graph inside a larger "target" graph, or establishing that the pattern does not occur. When the pattern graph is part of the input, these problems are NP-complete; despite this, subgraph isomorphism algorithms are widely used in practice, including for model checking [23], for law enforcement [9], in biological applications [1,6,20], for compiler implementation [5], in designing mechanical locks [27], and inside graph databases [19]. This has encouraged the development of practical subgraph isomorphism algorithms, which fall into two categories: those based upon backtracking and connectivity [6][7][8], and those based upon constraint programming [3,4,15,18,25].…”