Optimal code motion

Knoop, Jens; Rüthing, Oliver; Steffen, Bernhard

doi:10.1145/183432.183443

Cited by 159 publications

(117 citation statements)

References 18 publications

Supporting

Mentioning

117

Contrasting

Order By: Relevance

“…As in [17,19], we consider a non-SSA intermediate representation, where each statement has the form v = e such that v is a variable and e a single-operator expression. As is customary, we assume that local common subexpression elimination (LCSE) has already been applied to all basic blocks.…”

Section: Assumptionmentioning

confidence: 99%

“…We adopt the definition of PRE as used in LCM [19] except that we will make use of edge insertions as in [7,23,26] rather than node insertions; these insertions serve to make all the partially redundant computations fully redundant. Therefore, we do not have to split critical edges, i.e., the edges leading from nodes with more than one immediate successor to nodes with more than one immediate predecessor.…”

Section: Assumptionmentioning

confidence: 99%

“…The fundamental assumption in classic PRE as stated clearly in LCM [17,19] is that all control flows in a CFG are nondeterministic. Equivalently, we have:…”

Section: Assumptionmentioning

confidence: 99%

“…This is the key to the main contribution of our paper: to provide a uniform approach for classic and speculative PRE. The insight behind this finding lies in the following assumption made about classic PRE [17,19]: all control flow edges are nondeterministic, or equivalently, have nonzero execution frequencies. We show that finding the optimal transformation for a CFG amounts to finding a unique minimum cut in a flow network derived from the CFG.…”

Section: Introductionmentioning

confidence: 99%

“…In the case of classic PRE, Knoop, Rüthing and Steffen invented an optimal unidirectional bit-vector formulation of the problem [17,19]. This algorithm, known as Lazy Code Motion (LCM), was later recasted to operate in static single assignment (SSA) form [15].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A Fresh Look at PRE as a Maximum Flow Problem

Xue

Knoop

2006

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Abstract. We show that classic PRE is also a maximum flow problem, thereby revealing the missing link between classic and speculative PRE, and more importantly, establishing a common high-level conceptual basis for this important compiler optimisation. To demonstrate this, we formulate a new, simple unidirectional bit-vector algorithm for classic PRE based only on the well-known concepts of availability and anticipatability. Designed to find a unique minimum cut in a flow network derived from a CFG, which is proved simply but rigorously, our algorithm is simple and intuitive, and its optimality is self-evident. This conceptual simplicity also translates into efficiency, as validated by experiments.

show abstract

Section: Assumptionmentioning

confidence: 99%

Section: Assumptionmentioning

confidence: 99%

“…The fundamental assumption in classic PRE as stated clearly in LCM [17,19] is that all control flows in a CFG are nondeterministic. Equivalently, we have:…”

Section: Assumptionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A Fresh Look at PRE as a Maximum Flow Problem

Xue

Knoop

2006

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

Load redundancy elimination on executable code

Fernandez

Espasa

Debray

2003

Concurrency and Computation

View full text Add to dashboard Cite

Optimizations performed at link time or directly applied to final program executables have received increased attention in recent years. This paper discusses the discovery and elimination of redundant load operations in the context of a link‐time optimizer, an optimization that we call Load Redundancy Elimination (LRE). Our experiments show that between 50% and 75% of a program's memory references can be considered redundant because they are accessing memory locations that have been referenced less than 200–400 instructions away. We then present three profile‐based LRE algorithms targeted at optimizing away these redundancies. Our results show that between 5% and 30% of the redundancy detected can indeed be eliminated, which translates into program speedups of around 8%. We also test our algorithm assuming different cache latencies, and show that, if latencies continue to grow, the load redundancy elimination will become more important. Copyright © 2003 John Wiley & Sons, Ltd.

show abstract

Partial redundancy elimination for access expressions by speculative code motion

2004

View full text Add to dashboard Cite

We present a new memory access optimization for Java to perform aggressive code motion for speculatively optimizing memory accesses by applying partial redundancy elimination (PRE) techniques. First, to reduce as many barriers as possible and to enhance code motion, we perform alias analysis to identify all the regions in which each object reference is not aliased. Secondly, we find all the possible barriers. Finally, we perform code motions in three steps. For the first step, we apply a non‐speculative PRE algorithm to move load instructions and their following instructions in the backwards direction of the control flow graph. For the second step, we apply a speculative PRE algorithm to move some of them aggressively before the conditional branches. For the third step, we apply our modified version of a non‐speculative PRE algorithm to move store instructions in the forward direction of the control flow graph and to even move some of them after the merge points. We implemented our new algorithm in our production‐level Java just‐in‐time compiler. Our experimental results show that our speculative algorithm improves the average (maximum) performance by 13.1% (90.7%) for jBYTEmark and 1.4% (4.4%) for SPECjvm98 over the fastest algorithm previously described, while it increases the average (maximum) compilation time by 0.9% (2.9%) for both benchmark suites. Copyright © 2004 John Wiley & Sons, Ltd.

show abstract

Optimal code motion

Cited by 159 publications

References 18 publications

A Fresh Look at PRE as a Maximum Flow Problem

A Fresh Look at PRE as a Maximum Flow Problem

Load redundancy elimination on executable code

Partial redundancy elimination for access expressions by speculative code motion

Contact Info

Product

Resources

About