Zhide Zhou, Zhilei Ren, Guojun Gao, He Jiang,
... optimization bugs in two mainstream compilers, GCC and LLVM. We collect about 57K and 22K bugs of GCC and LLVM, and then exhaustively examine 8,771 and 1, ... optimization are the buggiest optimizations in GCC and LLVM, respectively; the loop optimizations in both GCC and LLVM are more bug-prone than other optimizations; (3) ... of the optimization bugs in both GCC and LLVM are misoptimization bugs, accounting for 57.21% and ... months for GCC and 13.55 months for LLVM to fix an optimization bug; in both GCC ...
Tópico(s): Security and Verification in Computing
2020 - Elsevier BV | Journal of Systems and Software
Tao B. Schardl, William S. Moses, Charles E. Leiserson,
... only minor modifications. For example, to implement Tapir/LLVM, a prototype of Tapir in the LLVM compiler, we added or modified less than 3,000 lines of LLVM’s half-million-line core middle-end functionality. These changes sufficed to enable LLVM’s existing compiler optimizations for serial code—including ... of the program. By making use of existing LLVM optimizations and new parallel optimizations, Tapir/LLVM can optimize recursive fork-join programs more effectively ... a suite of 35 Cilk application benchmarks, Tapir/LLVM produces more efficient executables for 30 benchmarks, with ...
Tópico(s): Distributed systems and fault tolerance
2019 - Association for Computing Machinery | ACM Transactions on Parallel Computing
Yun Zhao, Santosh Nagarakatte, Milo M. K. Martin, Steve Zdancewic,
This paper presents Vellvm (verified LLVM), a framework for reasoning about programs expressed in LLVM's intermediate representation and transformations that operate on it. Vellvm provides a mechanized formal semantics of LLVM's intermediate representation, its type system, and properties ... Coq formal semantics that can execute programs from LLVM test suite and thus be compared against LLVM reference implementations. To demonstrate Vellvm's practicality, we ... the transformation pass that plugs into the real LLVM infrastructure; its performance is competitive with the non- ...
Tópico(s): Parallel Computing and Optimization Techniques
2012 - Association for Computing Machinery | ACM SIGPLAN Notices
Juneyoung Lee, Yoonseung Kim, Youngju Song, Chung-Kil Hur, Sanjoy Das, David Majnemer, John Regehr, Nuno P. Lopes,
... optimizing compiler we have looked at, including GCC, LLVM, Intel's, and Microsoft's, supports one or ... desirable transformations become illegal. The current semantics of LLVM's IR fails to justify some cases of ... solutions to the problems we have identified in LLVM's IR and show that most optimizations currently in LLVM remain sound, and that some desirable new transformations ...
Tópico(s): Security and Verification in Computing
2017 - Association for Computing Machinery | ACM SIGPLAN Notices
IEEE Xplore
Umer Liqat, Kyriakos Georgiou, Steve Kerrison, Pedro López-García, John P. Gallagher, Manuel V. Hermenegildo, Kerstin Eder,
... instruction set architecture (ISA) and the intermediate code (LLVM IR) levels, and reflects it upwards to the ... This required the development of a translation from LLVM IR to an intermediate representation and its integration ... model, and a mapping from this model to LLVM IR. The approach has been applied to programs ... to other languages. Experimental results show that our LLVM IR level analysis is reasonably accurate (less than $$ ...
Tópico(s): Real-Time Systems Scheduling
2016 - Springer Science+Business Media | Lecture notes in computer science
Yun Zhao, Santosh Nagarakatte, Milo M. K. Martin, Steve Zdancewic,
Modern compilers, such as LLVM and GCC, use a static single assignment (SSA) intermediate representation (IR) to simplify and enable many advanced optimizations. However, formally ... this approach, we formally verify a variant of LLVM's mem2reg transformation in Vellvm, a Coq-based formal semantics of the LLVM IR. The extracted implementation generates code with performance comparable to that of LLVM's unverified implementation.
Tópico(s): Formal Methods in Verification
2013 - Association for Computing Machinery | ACM SIGPLAN Notices
Ghassan Shobaki, V. Scott Gordon, Paul McHugh, Théodore Dubois, Austin Kerbow,
... The proposed ACO algorithm is implemented in the LLVM open-source compiler, and its performance is evaluated ... and ARM, both algorithms are evaluated relative to LLVM's generic scheduler, while on the AMD GPU, ... on x86 and ARM, respectively, relative to the LLVM scheduler. Using PlaidML on an AMD GPU, it ... on. Both ACO and B&B outperform the LLVM algorithm on the CPU and the AMD algorithm ...
Tópico(s): Cloud Computing and Resource Management
2022 - Association for Computing Machinery | ACM Transactions on Architecture and Code Optimization
Shao-Chung Wang, Lin-Ya Yu, Li-An Her, Yuan‐Shin Hwang, Jenq‐Kuen Lee,
... statements. The proposed scheme was implemented in an LLVM compiler and can be applied to OpenCL programs. ... program size of 1,306 instructions in an LLVM intermediate representation, with 885 variables, 108 branches, and ... was always divergent, and 155 for the current LLVM default scheme; the total numbers of divergent variables ... a pointer is always divergent and the current LLVM default scheme.
Tópico(s): Logic, programming, and type systems
2021 - Association for Computing Machinery | ACM Transactions on Parallel Computing
Yannick Zakowski, Calvin Beck, Irene Yoon, Ilia Zaichuk, Vadim Zaliva, Steve Zdancewic,
... Coq, for a large, sequential subset of the LLVM IR. In contrast to previous approaches, which use ... executable interpreter. Our semantics handles many of the LLVM IR's non-trivial language features and is ... semantics by evaluating it on unit tests and LLVM IR programs generated by HELIX.
Tópico(s): Software Engineering Research
2021 - Association for Computing Machinery | Proceedings of the ACM on Programming Languages
Joseph Huber, Weile Wei, Giorgis Georgakoudis, Johannes Doerfert, Óscar Hernández,
This paper presents a methodology for using LLVM-based tools to tune the DCA++ (dynamical cluster approximation) application that targets the new ARM A64FX processor. The goal is to describe the ... set. During manual tuning, the authors used the LLVM tools to improve code parallelization by using OpenMP ... built on top of existing and newly introduced LLVM tooling.
Tópico(s): Distributed and Parallel Computing Systems
2021 - Springer Science+Business Media | Lecture notes in computer science
... a new reliable model with the methodology of LLVM (Low-dimensional Linear Ventilation Models)-based ANN (artificial ... LVM (Linear Ventilation Models) method was used. Next, LLVM method was utilized for reconstruction and expansion of CFD database. Then, the well-trained LLVM-based ANN was employed to predict indoor CO2 ... source strength and positions. With ventilation control using LLVM-based ANN model, both indoor pollutant concentration and ...
Tópico(s): Infection Control and Ventilation
2018 - Elsevier BV | Building and Environment
Juneyoung Lee, Chung-Kil Hur, Ralf Jung, Zhengyang Liu, John Regehr, Nuno P. Lopes,
LLVM miscompiles certain programs in C, C++, and Rust that use low-level language features such as ... deeper problem is that the memory model for LLVM's intermediate representation (IR) is informal and the ... developers. We developed a novel memory model for LLVM IR and formalized it. The new model requires ...
Tópico(s): Formal Methods in Verification
2018 - Association for Computing Machinery | Proceedings of the ACM on Programming Languages
Kyriakos Georgiou, Steve Kerrison, Zbigniew Chamski, Kerstin Eder,
... than the Instruction Set Architecture (ISA), namely the LLVM intermediate representation (IR) level, and therefore introduces energy transparency directly to the LLVM optimizer. We apply our energy estimation techniques to ... patterns: task farms and pipelines. Using SRA, our LLVM IR results demonstrate a high accuracy with a ... technique captures the actual energy consumption at the LLVM IR level with an average error of 3%.
Tópico(s): Real-Time Systems Scheduling
2017 - Association for Computing Machinery | ACM Transactions on Architecture and Code Optimization
Franck Cassez, Anthony M. Sloane, Matthew Roberts, Matthew Pigram, Pongsak Suvanpong, Pablo González de Aledo,
... is a static analysis tool that analyses the LLVM intermediate representation (LLVM-IR) of a program source code. The analysis ... can reach a designated error block in the LLVM-IR. The result of a program analysis is " ...
Tópico(s): Logic, programming, and type systems
2017 - Springer Science+Business Media | Lecture notes in computer science
Nuno P. Lopes, David Menendez, Santosh Nagarakatte, John Regehr,
... perform local rewriting to improve the efficiency of LLVM code. These optimizations are individually difficult to get ... code that is suitable for inclusion in an LLVM optimization pass. Alive is based on an attempt ... of three different kinds of undefined behavior in LLVM. We have translated more than 300 LLVM optimizations into Alive and, in the process, found ...
Tópico(s): Parallel Computing and Optimization Techniques
2015 - Association for Computing Machinery | ACM SIGPLAN Notices
Chanhyun Park, Miseon Han, Hokyoon Lee, Myeongjin Cho, Seon Wook Kim,
... study compared the performance of two major compilers, LLVM and GCC on a32-bit EISC embedded processor. ... evaluated from these compilers with the EEMBC benchmarks.LLVM generally performed better in the ALU intensive benchmarks, ... average 8% and 7% lower than those of LLVM, respectively.
Tópico(s): Embedded Systems and FPGA Design
2014 - | IEIE Transactions on Smart Processing and Computing
Teja Tamboli, Thomas H. Austin, Mark Stamp,
... have created a metamorphic code generator within the LLVM compiler framework. LLVM is a three-phase compiler that supports multiple ... to this IR bytecode as part of the LLVM compilation process. Our metamorphic generator functions at the ...
Tópico(s): Software Testing and Debugging Techniques
2013 - Springer Science+Business Media | Journal of Computer Virology and Hacking Techniques
Stephen K. Dolan, Servesh Muralidharan, David Gregg,
... propose a new language-neutral primitive for the LLVM compiler, which provides efficient context switching and message ... be used by any language implementation based on LLVM to build higher-level language structures such as ... threads. As part of adding the primitives to LLVM, we have also added compiler support for passing parameters across context switches. Our modified LLVM compiler produces highly efficient code through a combination ...
Tópico(s): Real-Time Systems Scheduling
2013 - Association for Computing Machinery | ACM Transactions on Architecture and Code Optimization
The LLVM Compiler Infrastructure (http://llvm.cs. uiuc.edu) is a robust system that is well suited for a wide variety of research and development work. This brief paper introduces the LLVM system and provides pointers to more extensive documentation, ...
Tópico(s): Software Testing and Debugging Techniques
2005 - Springer Science+Business Media | Lecture notes in computer science
Tim Heldmann, Thomas Schneider, Oleksandr Tkachenko, Christian Weinert, Hossein Yalame,
... automating circuit compilation based on the compiler toolchain LLVM. For this, we develop an LLVM optimizer suite consisting of multiple transform passes that operate on the LLVM intermediate representation (IR) and gradually lower functions to ... advantage of powerful source code optimizations built into LLVM. We furthermore make sure to produce circuits that ...
Tópico(s): Complexity and Algorithms in Graphs
2021 - Springer Science+Business Media | Lecture notes in computer science
Zhide Zhou, He Jiang, Zhilei Ren, Yuting Chen, Lei Qiao,
... to automatically localize compiler optimization sequence bugs of LLVM. The core insight of LocSeq is based on ... a benchmark, including 60 optimization sequence bugs of LLVM, and compare LocSeq with the state-of-the- ...
Tópico(s): Software Reliability and Analysis Research
2022 - Institute of Electrical and Electronics Engineers | IEEE Transactions on Reliability
Atmn Patel, Johannes Doerfert,
... of Remote OpenMP Offloading is integrated into the LLVM/OpenMP compiler infrastructure and publicly available (in parts) with LLVM 12 and later. LLVM-based (vendor) compilers are expected to be compatible ...
Tópico(s): Advanced Data Storage Technologies
2022 - Springer Science+Business Media | Lecture notes in computer science
He Jiang, Zhide Zhou, Zhilei Ren, Jingxuan Zhang, Xiaochen Li,
... detecting compiler bugs caused by optimization sequences of LLVM. CTOS first leverages the technique Doc2Vec to transform ... capability on average. Within seven month evaluations on LLVM, we have reported 104 valid bugs within 5 ...
Tópico(s): Parallel Computing and Optimization Techniques
2021 - IEEE Computer Society | IEEE Transactions on Software Engineering
Hao-Cheng Zhu, Chen Ren, Shi-Jie Cao,
... methods by using low-dimensional linear ventilation model (LLVM) based artificial neural network (ANN) and low-dimensional ... and humidity. Low-dimensional linear models (LLM), including LLVM, LLTM and LLHM, were adopted to expand database ... humidity were rapidly predicted on the basis of LLVM-based ANN, LLTM-based CRI (T) and LLHM- ...
Tópico(s): Infection Control and Ventilation
2020 - Springer Science+Business Media | Building Simulation
... that provides several types of program analysis for LLVM bitcode. The main parts of DG are a ... the main one being a static slicer for LLVM bitcode. This paper describes what analyses are implemented ...
Tópico(s): Software Engineering Research
2020 - Springer Science+Business Media | Lecture notes in computer science