Michael McCool, Stefanus Du Toit · 2009
This chapter details the intermediate representation (IR) and related concepts. The IR encapsulates both the statements and the control-flow contained by a program. The chapter provides an overview of the reference-counting mechanism used by Sh and explains the way basic blocks are stored. Basic blocks are represented in Sh by the ShBasicBlock class. Most Sh operations are arithmetic or logic in nature. Sh programs can include more than just basic blocks: they can contain arbitrary data-dependent control flow. Many types in Sh are actually reference-counted. This means that for each object declared of such a type, a count is kept of how many pointer variables refer to it.
Cited by 1
Vivek Sarkar · 2013
Cited by 0
Christian Gerth · Lecture notes in computer science · 2013
Cited by 1
Pedersen, Jens Egholm, Abreu, Steven, Jobst, Matthias et al. · Zenodo (CERN European Organization for Nuclear Research) · 2024
NIR defines neuron models and connectivity for arbitrary networks that include spiking neurons. Neuron models are defined as dynamical system equations because time is an essential component of neuromorphic systems. The goal is to provide a common format that different spiking neural network (SNN) frameworks can convert to. That allows a user to train an SNN in framework X and convert it to framework Y. Framework X might offer particularly fast training, while framework Y might offer deployment to neuromorphic hardware.
Cited by 1Open access
Simone Campanoni · SpringerBriefs in computer science · 2011
Cited by 1
Pedersen, Jens Egholm, Abreu, Steven, Jobst, Matthias et al. · Zenodo (CERN European Organization for Nuclear Research) · 2024
NIR defines neuron models and connectivity for arbitrary networks that include spiking neurons. Neuron models are defined as dynamical system equations because time is an essential component of neuromorphic systems. The goal is to provide a common format that different spiking neural network (SNN) frameworks can convert to. That allows a user to train an SNN in framework X and convert it to framework Y. Framework X might offer particularly fast training, while framework Y might offer deployment to neuromorphic hardware.
Cited by 0Open access
Abreu, Steven, Bauer, Felix, Eshraghian, Jason et al. · Zenodo (CERN European Organization for Nuclear Research) · 2023
NIR defines neuron models and connectivity for arbitrary networks that include spiking neurons. Neuron models are defined as dynamical system equations because time is an essential component of neuromorphic systems. The goal is to provide a common format that different spiking neural network (SNN) frameworks can convert to. That allows a user to train an SNN in framework X and convert it to framework Y. Framework X might offer particularly fast training, while framework Y might offer deployment to neuromorphic hardware.
Cited by 1Open access
Abreu, Steven, Bauer, Felix, Eshraghian, Jason et al. · Zenodo (CERN European Organization for Nuclear Research) · 2023
NIR defines neuron models and connectivity for arbitrary networks that include spiking neurons. Neuron models are defined as dynamical system equations because time is an essential component of neuromorphic systems. The goal is to provide a common format that different spiking neural network (SNN) frameworks can convert to. That allows a user to train an SNN in framework X and convert it to framework Y. Framework X might offer particularly fast training, while framework Y might offer deployment to neuromorphic hardware.
Cited by 0Open access
Florent Jacquemard, Clément Poncelet Sanchez · arXiv (Cornell University) · 2014
We describe an intermediate language designed as a medium-level internal\nrepresentation of programs of the interactive music system Antescofo. This\nrepresentation is independent both of the Antescofo source language and of the\narchitecture of the execution platform. It is used in tasks such as\nverification of timings, model-based conformance testing, static control-flow\nanalysis or simulation. This language is essentially a flat representation of\nAntescofo's code, as a finite state machine extended with local and global\nvariables, with delays and with concurrent threads creation. It features a\nsmall number of simple instructions which are either blocking (wait for\nexternal event, signal or duration) or not (variable assignment, message\nemission and control).\n
Cited by 0Open access
V. А. Vasenin, M. A. Krivchikov · PROGRAMMNAYA INGENERIA · 2017
Cited by 1
Cerny, Tomas · Zenodo (CERN European Organization for Nuclear Research) · 2024
This dataset contains samples of the Intermediate representation (IR) and Delta represenation generated using ISAR metholdogy opertaing on 8 open source microservices-based projects.
Cited by 0Open access
J. Schirmer · Lecture notes in chemistry · 2018
Cited by 1
Mihai Budiu, Seth Copen Goldstein · Research Showcase @ Carnegie Mellon University (Carnegie Mellon University) · 2018
We present Pegasus, a compact and expressive intermediate representation for imperative languages. The representation is suitable for target architectures supporting predicated execution and aggressive speculation. In Pegasus information about the global dataflow of the program is encoded in local structures, enabling compact and efficient algorithms for program optimizations. As a proof of the versatility of Pegasus, we have used it in a compiler translating C programs to hardware implementations.
Cited by 47Open access
Niranjan Hasabnis, R. Sekar · ACM SIGOPS Operating Systems Review · 2016
Translating low-level machine instructions into higher-level intermediate language (IL) is one of the central steps in many binary analysis and instrumentation systems. Existing systems build such translators manually. As a result, it takes a great deal of effort to support new architectures. Even for widely deployed architectures, full instruction sets may not be modeled, e.g., mature systems such as Valgrind still lack support for AVX, FMA4 and SSE4.1 for x86 processors. To overcome these difficulties, we propose a novel approach that leverages knowledge about instruction set semantics that is already embedded into modern compilers such as GCC. In particular, we present a learning-based approach for automating the translation of assembly instructions to a compiler's architecture-neutral IL. We present an experimental evaluation that demonstrates the ability of our approach to easily support many architectures (x86, ARM and AVR), including their advanced instruction sets. Our implementation is available as open-source software.
Cited by 2Open access
Martin Trapp, Götz Lindenmaier, Boris Boesler · 2007
This documentation presents Firm, an intermediate representation (IR) presented first by Armbruster and von Roques in [AvR96], subsequently refined and extended by Trapp ([Tra99]) in his Phd Thesis at the Institute for Program Structures and Data Organization at University of Karlsruhe. Armbruster and von Roques implemented Firm in an experimental compiler, Fiasco, for the language Sather-K ([GS96])
Cited by 0
Krzysztof Palacz, Jason D. Baker, C. Flack et al. · 2003
The Ovm framework is a set of tools and components for building language runtimes. We present the intermediate representation and software design patterns used throughout the framework. One of the main themes in this work has been to support experimentation with new linguistic constructs and implementation techniques. To this end, framework components were designed to be parametric with respect to the instruction set on which they operate. We argue that our approach eases the task of writing new components without sacrificing efficiency.
Cited by 25
Mark Grechanik, Dewayne E Perry, Don Batory et al. · 2008
We present a novel approach that enables automatic generation of code for the front end of a compiler. The proposed approach is based on a Tree Manipulation Language (TML). TML is implemented on top of C++ and adds symbolic computational properties to manipulate hierarchical structures. We introduce an XML-based intermediate representation (XIR) used with TML that dramatically improves various front-end compiler algorithms such as symbol table creation and lookup, and abstract syntax tree (AST) generation and traversal. The proposed approach results in multiple benefits, for example, the significant reduction of code required to build ASTs, symbol table, and inheritance hierarchies. We built a C# parser in order to illustrate the effectiveness of our approach.
Cited by 0
J. H. Hamilton · Mathematical Proceedings of the Cambridge Philosophical Society · 1953
ABSTRACT It is shown that for the quantized φ 3 field the series expansions for intermediate representation field operators are convergent for summation over all graphs except those containing self-energy and vertex parts. The method used does not permit an investigation of the self-energy and vertex parts.
Cited by 0
Ford, Rupert, Porter, Andrew, Siso, Sergi et al. · Figshare · 2019
This talk presented the PSyclone Intermediate representation and the translation of code into this representation by PSyclone. This approach allows the NEMO model to be parallelised for GPUs using PSyclone. This approach is also being used to translate to the Stencil Intermediate Representation (SIR) which is used by the DAWN back-end to generate cuda and gridtools code. The talk was at the PASC 19 conference in Zurich, Switzerland on the 13th June 2019.
Cited by 0Open access
Eric Drew Fritz · UWM Digital Commons (University of Wisconsin–Milwaukee) · 2018
Program transformations that are able to rely on the presence of canonical properties of the program undergoing optimization can be written to be more robust and efficient than an equivalent but generalized transformation that also handles non-canonical programs. If a canonical property is required but broken earlier in an earlier transformation, it must be rebuilt (often from scratch). This additional work can be a dominating factor in compilation time when many transformations are applied over large programs. This dissertation introduces a methodology for constructing program transformations so that the program remains in an always-canonical form as the program is mutated, making only local changes to restore broken properties.
Cited by 1Open access