Binding-time analysis applied to mathematical algorithms

Glück, Robert; Nakashige, Ryo; Zöchling, Robert

doi:10.1007/978-0-387-34897-1_14

Cited by 15 publications

(9 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our overall approach can best be described as a combination of two approaches: hand-writing a code generator (cogen) suitable for multi-level specialization [49,50] and creating an embedded domain-specific language (EDSL) [51,52]. Our approach, using staging, monads and functors, seems to permit an extensible set of aspects and appears flexible enough for iterative improvement in the design.…”

Section: Methodsmentioning

confidence: 99%

“…Even though the Aldor compiler can frequently optimize away a lot of abstraction overhead, it does not provide any guarantees that it will do so, unlike our approach. We should also mention early work [49] on automatic specialization of mathematical algorithms. Although it can eliminate some overhead from a very generic implementation (e.g.…”

Section: Related and Future Workmentioning

confidence: 99%

See 1 more Smart Citation

Multi-stage Programming with Functors and Monads: Eliminating Abstraction Overhead from Generic Code

Carette

Kiselyov²

2005

Lecture Notes in Computer Science

View full text Add to dashboard Cite

We use multi-stage programming, monads and Ocaml's advanced module system to demonstrate how to eliminate all abstraction overhead from generic programs while avoiding any inspection of the resulting code. We demonstrate this clearly with Gaussian Elimination as a representative family of symbolic and numeric algorithms. We parameterize our code to a great extent -over domain, input and permutation matrix representations, determinant and rank tracking, pivoting policies, result types, etc. -at no run-time cost. Because the resulting code is generated just right and not changed afterward, MetaOCaml guarantees that the generated code is well-typed. We further demonstrate that various abstraction parameters (aspects) can be made orthogonal and compositional, even in the presence of name-generation for temporaries, and "interleaving" of aspects. We also show how to encode some domain-specific knowledge so that "clearly wrong" compositions can be rejected at or before generation time, rather than during the compilation or running of the generated code.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Related and Future Workmentioning

confidence: 99%

Multi-stage Programming with Functors and Monads: Eliminating Abstraction Overhead from Generic Code

Carette

Kiselyov²

2005

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…(6) Computer-placed program annotations to guide transformation, for example, binding-time annotation [Bondorf 1991;Consel 1993;Consel and Danvy 1993;Glenstrup and Jones 1996;Glück et al 1995;Jones et al 1993;Hatcliff et al 1999;Thiemann 1997]. After annotation, transformation is fully automatic, requiring no human interaction.…”

Section: On the Meaning Of "Automatic" In Program Transformationmentioning

confidence: 99%

“…-Scientific computing [Berlin and Weise 1990;Glück et al 1995]; -Extending functionality of existing languages by adding binding-time options [Grant et al 2000]; -Functional languages [Bondorf 1991;Consel 1993;Hughes 1996;Launchbury 1991;Thiemann 1999]; -Logic programming languages [Gallagher and Bruynooghe 1990;Gallagher 1993;Leuschel and Bruynooghe 2002;Lloyd and Shepherdson 1991]; -Compiling or other transformation by specializing interpreters [Birkedal and Welinder 1994;Bondorf 1991;Glenstrup et al 1999;Jones et al 1993;Jones 1996;Mogensen 1988]; -Optimization of operating systems, for example, remote procedure calls, device drivers, etc. [Consel and Noël 1996;McNamee et al 2001;Poletto et al 1999];…”

Section: Partial Evaluation and Program Generationmentioning

confidence: 99%

Termination analysis and specialization-point insertion in offline partial evaluation

Glenstrup

Jones

2005

ACM Trans. Program. Lang. Syst.

View full text Add to dashboard Cite

Recent research suggests that the goal of fully automatic and reliable program generation for a broad range of applications is coming nearer to feasibility. However, several interesting and challenging problems remain to be solved before it becomes a reality. Solving them is also necessary, if we hope ever to elevate software engineering from its current state (a highly developed handiwork) into a successful branch of engineering, capable of solving a wide range of new problems by systematic, well-automated and well-founded methods.A key problem in all program generation is termination of the generation process. This article focuses on off-line partial evaluation and describes recent progress towards automatically solving the termination problem, first for individual programs, and then for specializers and "generating extensions," the program generators that most offline partial evaluators produce.The technique is based on size-change graphs that approximate the changes in parameter sizes at function calls. We formulate a criterion, bounded anchoring, for detecting parameters known to be bounded during specialization: a bounded parameter can act as an anchor for other parameters. Specialization points necessary for termination are computed by adding a parameter that tracks call depth, and then selecting a specialization point in every call loop where it is unanchored. By generalizing all unbounded parameters, we compute a binding-time division which together with the set of specialization points guarantees termination.Contributions of this article include a proof, based on the operational semantics of partial evaluation with memoization, that the analysis guarantees termination; and an in-depth description of safety of the increasing size approximation operator required for termination analysis in partial evaluation.Initial experiments with a prototype shows that the analysis overall yields binding-time divisions that can achieve a high degree of specialization, while still guaranteeing termination.The article ends with a list of challenging problems whose solution would bring the community closer to the goal of broad-spectrum, fully automatic and reliable program generation.

show abstract

“…PE [11] is an automatic program transformation technique which allows the partial execution of a program when only some of its input data are available, and it also "specializes" the program with respect to partial knowledge of the data. In [9], PE is incorporated in a library of general-purpose mathematical algorithms in order to allow the automatic generation of fast, special-purpose programs from generic algorithms. The results for the FFT show speedup factors between 1.83 and 5.05 if the size N of the input is available, where N ranges from 16 and 512.…”

Section: Introductionmentioning

confidence: 99%

Non-strict Evaluation of the FFT Algorithm in Distributed Memory Systems

Cristóbal-Salas

Tchernykh

Gaudiot

2003

Recent Advances in Parallel Virtual Machine and Message Passing Interface

View full text Add to dashboard Cite

Abstract. This paper focuses on the partial evaluation of local and remote memory accesses of distributed applications, not only to remove much of the excess overhead of message passing implementations, but also to reduce the number of messages, when some information about the input data set is known. The use of split-phase memory operations, the exploitation of spatial data locality, and non-strict information processing are described. Through a detailed performance analysis, we establish conditions under which the technique is beneficial. We show that by incorporating non-strict information processing to FFT MPI, a significant reduction of the number of messages can be archievd, and the overall system performance can be improved.

show abstract

Binding-time analysis applied to mathematical algorithms

Cited by 15 publications

References 5 publications

Multi-stage Programming with Functors and Monads: Eliminating Abstraction Overhead from Generic Code

Multi-stage Programming with Functors and Monads: Eliminating Abstraction Overhead from Generic Code

Termination analysis and specialization-point insertion in offline partial evaluation

Non-strict Evaluation of the FFT Algorithm in Distributed Memory Systems

Contact Info

Product

Resources

About