Compiler System Research Articles

WhiteFox: White-Box Compiler Fuzzing Empowered by Large Language Models

Compiler correctness is crucial, as miscompilation can falsify program behaviors, leading to serious consequences over the software supply chain. In the literature, fuzzing has been extensively studied to uncover compiler defects. However, compiler fuzzing remains challenging: Existing arts focus on black- and grey-box fuzzing, which generates test programs without sufficient understanding of internal compiler behaviors. As such, they often fail to construct test programs to exercise intricate optimizations. Meanwhile, traditional white-box techniques, such as symbolic execution, are computationally inapplicable to the giant codebase of compiler systems. Recent advances demonstrate that Large Language Models (LLMs) excel in code generation/understanding tasks and even have achieved state-of-the-art performance in black-box fuzzing. Nonetheless, guiding LLMs with compiler source-code information remains a missing piece of research in compiler testing. To this end, we propose WhiteFox, the first white-box compiler fuzzer using LLMs with source-code information to test compiler optimization, with a spotlight on detecting deep logic bugs in the emerging deep learning (DL) compilers. WhiteFox adopts a multi-agent framework: (i) an LLM-based analysis agent examines the low-level optimization source code and produces requirements on the high-level test programs that can trigger the optimization; (ii) an LLM-based generation agent produces test programs based on the summarized requirements. Additionally, optimization-triggering tests are also used as feedback to further enhance the test generation prompt on the fly. Our evaluation on the three most popular DL compilers (i.e., PyTorch Inductor, TensorFlow-XLA, and TensorFlow Lite) shows that WhiteFox can generate high-quality test programs to exercise deep optimizations requiring intricate conditions, practicing up to 8 times more optimizations than state-of-the-art fuzzers. To date, WhiteFox has found in total 101 bugs for the compilers under test, with 92 confirmed as previously unknown and 70 already fixed. Notably, WhiteFox has been recently acknowledged by the PyTorch team, and is in the process of being incorporated into its development workflow. Finally, beyond DL compilers, WhiteFox can also be adapted for compilers in different domains, such as LLVM, where WhiteFox has already found multiple bugs.

Partial program analysis for staged compilation systems

Partial program analysis for staged compilation systems

Design of intelligent controller for obstacle avoidance and navigation of electric patrol mobile robot based on PLC

Currently, the obstacle avoidance control of patrol robots based on intelligent vision lacks professional controller module assistance. Therefore, this paper proposes a design method of intelligent controller for obstacle avoidance and navigation of electrical inspection mobile robot based on PLC control. The controller designs a laser range finder to determine the required position of electrical patrol inspection. Use PLC as the core controller, and combine sensors, actuators, communication module and PLC selection module in the process of hardware design to achieve autonomous navigation and obstacle avoidance functions of the robot. Then design the software including the PLC compiler system and the virtual machine module. Based on the above steps, design the control module of obstacle avoidance navigation, which realizes the key link of robot autonomous navigation. The test results show that the controller can successfully avoid obstacles, improve the efficiency and quality of inspection, and achieve accurate and fast obstacle avoidance navigation for the electrical inspection mobile robot.

The NOMAD mini-apps: A suite of kernels from ab initio electronic structure codes enabling co-design in high-performance computing.

This article introduces a suite of mini-applications (mini-apps) designed to optimise computational kernels in ab initio electronic structure codes. The suite is developed from flagship applications participating in the NOMAD Center of Excellence, such as the ELPA eigensolver library and the GW implementations of the exciting, Abinit, and FHI-aims codes. The mini-apps were identified by targeting functions that significantly contribute to the total execution time in the parent applications. This strategic selection allows for concentrated optimisation efforts. The suite is designed for easy deployment on various High-Performance Computing (HPC) systems, supported by an integrated CMake build system for straightforward compilation and execution. The aim is to harness the capabilities of emerging (post)exascale systems, which necessitate concurrent hardware and software development - a concept known as co-design. The mini-app suite serves as a tool for profiling and benchmarking, providing insights that can guide both software optimisation and hardware design. Ultimately, these developments will enable more accurate and efficient simulations of novel materials, leveraging the full potential of exascale computing in material science research.

Enhancing XML-based Compiler Construction with Large Language Models: A Novel Approach

Considering the prevailing rule of Large Language Models (LLMs) applications and the benefits of XML in a compiler context. This manuscript explores the synergistic integration of Large Language Models with XML-based compiler tools and advanced computing technologies. Marking a significant stride toward redefining compiler construction and data representation paradigms. As computing power and internet proliferation advance, XML emerges as a pivotal technology for representing, exchanging, and transforming documents and data. This study builds on the foundational work of Chomsky's Context-Free Grammar (CFG). Recognized for their critical role in compiler construction, to address and mitigate the speed penalties associated with traditional compiler systems and parser generators through the development of an efficient XML parser generator employing compiler techniques. Our research employs a methodical approach to harness the sophisticated capabilities of LLMs, alongside XML technologies. The key is to automate grammar optimization, facilitate natural language processing capabilities, and pioneer advanced parsing algorithms. To demonstrate their effectiveness, we thoroughly run experiments and compare them to other techniques. This way, we call attention to the efficiency, adaptability, and user-friendliness of the XML-based compiler tools with the help of these integrations. And the target will be the elimination of left-recursive grammars and the development of a global schema for LL(1) grammars, the latter taking advantage of the XML technology, to support the LL(1) grammars construction. The findings in this research not only underscore the signification of these innovations in the field of compilation construction but also indicate a paradigm move towards the use of AI technologies and XML in the context of the resolution of programming traditional issues. The outlined methodology serves as a roadmap for future research and development in compiler technology, which paves the way for open-source software to sweep across all fields. Gradually ushering in a new era of compiler technology featuring better efficiency, adaptability, and all CFGs processed through existing XML utilities on a global basis.

The NOMAD mini-apps: A suite of kernels from ab initio electronic structure codes enabling co-design in high-performance computing

This article introduces a suite of mini-applications (mini-apps) designed to optimise computational kernels in ab initio electronic structure codes. The suite is developed from flagship applications participating in the NOMAD Center of Excellence, such as the ELPA eigensolver library and the GW implementations of the exciting, Abinit, and FHI-aims codes. The mini-apps were identified by targeting functions that significantly contribute to the total execution time in the parent applications. This strategic selection allows for concentrated optimisation efforts. The suite is designed for easy deployment on various High-Performance Computing (HPC) systems, supported by an integrated CMake build system for straightforward compilation and execution. The aim is to harness the capabilities of emerging (post)exascale systems, which necessitate concurrent hardware and software development — a concept known as co-design. The mini-app suite serves as a tool for profiling and benchmarking, providing insights that can guide both software optimisation and hardware design. Ultimately, these developments will enable more accurate and efficient simulations of novel materials, leveraging the full potential of exascale computing in material science research.

Rianú: Multi-tissue tracking software for increased throughput of engineered cardiac tissue screening

Background:The field of tissue engineering has provided valuable three-dimensional species-specific models of the human myocardium in the form of human Engineered Cardiac Tissues (hECTs) and similar constructs. However, hECT systems are often bottlenecked by a lack of openly available software that can collect data from multiple tissues at a time, even in multi-tissue bioreactors, which limits throughput in phenotypic and therapeutic screening applications. Methods:We developed Rianú, an open-source web application capable of simultaneously tracking multiple hECTs on flexible end-posts. This software is operating system agnostic and deployable on a remote server, accessible via a web browser with no local hardware or software requirements. The software incorporates object-tracking capabilities for multiple objects simultaneously, an algorithm for twitch tracing analysis and contractile force calculation, and a data compilation system for comparative analysis within and amongst groups. Validation tests were performed using in-silico and in-vitro experiments for comparison with established methods and interventions. Results:Rianú was able to detect the displacement of the flexible end-posts with a sub-pixel sensitivity of 0.555 px/post (minimum increment in post displacement) and a lower limit of 1.665 px/post (minimum post displacement). Compared to our established reference for contractility assessment, Rianú had a high correlation for all parameters analyzed (ranging from R2=0.7514 to R2=0.9695), demonstrating its high accuracy and reliability. Conclusions:Rianú provides simultaneous tracking of multiple hECTs, expediting the recording and analysis processes, and simplifies time-based intervention studies. It also allows data collection from different formats and has scale-up capabilities proportional to the number of tissues per field of view. These capabilities will enhance throughput of hECTs and similar assays for in-vitro analysis in disease modeling and drug screening applications.

Highly Parallel Multi-FPGA System Compilation from Sequential C/C++ Code in the AWS Cloud

We present a High Level Synthesis compiler that automatically obtains a multi-chip accelerator system from a single-threaded sequential C/C++ application. Invoking the multi-chip accelerator is functionally identical to invoking the single-threaded sequential code the multi-chip accelerator is compiled from. Therefore, software development for using the multi-chip accelerator hardware is simplified, but the multi-chip accelerator can exhibit extremely high parallelism. We have implemented, tested, and verified our push-button system design model on multiple field-programmable gate arrays (FPGAs) of the Amazon Web Services EC2 F1 instances platform, using, as an example, a sequential-natured DES key search application that does not have any DOALL loops and that tries each candidate key in order and stops as soon as a correct key is found. An 8- FPGA accelerator produced by our compiler achieves 44,600 times better performance than an x86 Xeon CPU executing the sequential single-threaded C program the accelerator was compiled from. New features of our compiler system include: an ability to parallelize outer loops with loop-carried control dependences, an ability to pipeline an outer loop without fully unrolling its inner loops, and fully automated deployment, execution and termination of multi-FPGA application-specific accelerators in the AWS cloud, without requiring any manual steps.

ESG and Sustainable Agriculture : Focus on 'True Cost' of Agricultural Operations

In every economy, agriculture is a central activity which has significant impact on environmental, social and economic aspects of the country. This article presents and highlights the ‘true cost of agricultural operation’, a system of compilation and aggregation of all relevant costs, including externalities, which not only limit itself to the financial metrics but go beyond the broader considerations of ESG factors and sustainability. To arrive at the true cost of agricultural operations, global concerns about ESG and sustainability aspects necessitate a transition from “incremental change” to “transformational change.”

Path and Mechanism Optimization of Integrating Tourism Planning into Territorial Spatial Planning System

Tourism planning is a kind of protective development planning for scientific and rational optimization of tourism space and tourism resources. At present, there have been problems in tourism planning, such as unclear role, imperfect content system and immature compilation technology, which have caused difficulties in landing tourism planning and conflicts with other planning. Under the background of territorial spatial planning system reform, tourism planning into the territorial spatial planning system has great theoretical significance and very urgent practical significance. Based on the analysis of the existing problems and causes of tourism planning, this paper puts forward that ' multi-planning ' should include tourism demand and influence in the process of ' integration ', and the optimization path of tourism planning compilation system oriented by integrating into the territorial spatial planning system. And from the census of tourism resources, modify the general rules of tourism planning, the introduction of tourism planning preparation methods, the establishment of tourism sector spatial information subsystem, strengthen the typical tourism destinations spatial planning guidance and tourism planning and other ' multi planning ' department discussion and exchange of six aspects put forward tourism planning into the territorial spatial planning system optimization strategy. The purpose is to standardize the preparation of tourism planning, actively integrate into the territorial spatial planning system, promote the internal cohesion of the ' five levels and three categories ‘territorial spatial planning system, and scientifically guide the planning and construction of tourism destinations.

Simplified High Level Parallelism Expression on Heterogeneous Systems through Data Partition Pattern Description

Abstract With the development of heterogeneous systems, the demand for high-level programming methods that ease heterogeneous programming and produce portable applications has become more urgent. This paper proposes DACL, the data associated computing language. DACL introduces data partition patterns to achieve architecture-independent parallelism expression. Meanwhile, DACL provides simplified language extensions, as well as programming features such as serialization of the computing process, parameterization of data attributes and modularity, thus reducing the difficulty of heterogeneous programming and improving programming productivity. The operational semantics show that DACL enables different levels of parallelism degree calculation and retains data access patterns, reserving optimization potential. To support cross-platform execution, the currently implemented source-to-source compilers employ OpenMP and OpenCL as the backend. We reconstructed multiple benchmarks selected from the Parboil and Rodinia benchmark suits with DACL and conducted a comparison test on CPU, GPU and MIC platforms. The code size of each rebuilt benchmark is roughly equivalent to that of the serial code, which is only 13%–64% of the benchmark OpenCL code. With the support of the compilation system, the reconstructed code can execute on different processors without modification, yielding a competitive or better performance to that of the manually written benchmark code.

Theoretical and Conceptual Approaches to Defining the Concept of Economic Intelligence in the Ukrainian Scientific Practice

This contribution is dedicated to the study of theoretical and conceptual aspects of economic intelligence. The goal of the research is to theoretically define the notion of economic intelligence and categorize the forms of its implementation with regard to the wide range of tools used to obtain, process and analyze information. Since economic intelligence is a separate analytical phenomenon, the author focuses on a series of relevant research tasks that disclose the theoretical essence of such notions as “economic intelligence”, “competitive intelligence” and “industrial espionage” as well as elaborate on their distinctive features in the context of data management process. Moreover, the author singles out the principles of information security within the context of intelligence activities and scientifically substantiates the forms of economic intelligence implementation with respect to the use of data acquisition methods. In the course of the research on economic intelligence, the author emphasizes the existence of various methods to acquire information necessary for the effective functioning of a state, organization or enterprise, which facilitates the successful solving of strategic and operational issues in the policy-making process. Furthermore, the economic intelligence data helps to identify new market opportunities as well as potential risks, and, subsequently, modify economic activities and plans on the development of an enterprise environment. Having studied theoretical aspects and forms of economic intelligence, the author reaches conclusions regarding the notion of economic intelligence as well as its forms – competitive intelligence and industrial espionage. It is determined that economic intelligence represents a system of covert search, gathering, compilation, analysis and goal-oriented use of strategic intelligence. The author notes that economic intelligence can be studied within the contexts of its separate forms, namely the competitive intelligence (as a lawful method to acquire data) and industrial espionage (as an unlawful method of acquiring data). The comparative analysis reveals that the notions of economic and business intelligence can be considered synonyms, although business intelligence in less associated with the illegal gathering of information. The author states that business intelligence is not synonymous with industrial espionage. Even being forms of economic intelligence, the notions of competitive intelligence and industrial espionage cannot be interpreted as one phenomenon, for they differ in the characteristic methods of data acquisition.

Historical Map as a New Description System for a Geographic Compilation

한국의 지지 편찬 전통은 매우 오래되었고 20세기 전반까지 꽤 강고하게 전승되었다. 그러나 대한민국은 공식적으로 아직 순수한 의미에서의 지지를 내놓지 못하고 있다. 그나마 다행인 것은 이러한 전통이 지방지, 대표적으로는 시군지(또는 시군사)로 계승되고 있다는 사실이다. 1990년대 이후에는 시군지가 기존의 인식과 형태에 서 탈피하여 지역사를 기술하는 차원에서 발행되는 추세에 있다. 지역사의 관점에서 시군지는 시군사로 이름을 바꾸기도 하였고, 지방사·미시사·개인사·일상사 등에 대한 관심이 고조되면서 단순 텍스트가 아닌 도표와 사진, 그림, 지도 등이 지역사 기술 체계 안으로 들어왔다. 21세기 지지 편찬에는 ‘현재성’과 ‘시공간성’이 기술 체계의 원칙이 될 필요가 있으며, 그동안 등한시되었던 역사지도가 시급히 이 기술 체계 안에 편입될 필요가 있다. 왜냐하면, 시공간적 현재성을 표현하는데 텍스트와 도표로 부족한 부분을 채워줄 수 있는 기술 체계로 역사지도만 한 것이 없기 때문이다. 이 점에서 시군지 편찬의 전통과 성과가 탁월한 경기도는 ‘경기도지’를 새로 기획하는 시점에서 ‘경기역사지도집’ 펴냄으로써 경기도의 위상을 국내 광역지자체와 중앙 정부, 그리고 세계에 발신할 자격과 당연, 그리고 책무가 있다.

Towards Faster Execution of Ensemble ML Bootstrap Based Techniques.

Algorithms for ensemble methods (EM) based on bootstrap aggregation often perform copious amount of redundant computations (RC) thus limiting their practicality. Given this constraint, we propose a framework that views these algorithms as a collection of computational units (cu), a tightly coupled set of both mathematical operations and data. This view facilitates a reduction in RC (RRC), thereby allowing for faster execution plans. Inspired by the floor tiling approach in VLSI, we look to engineer solutions for RRC while possibly reconfiguring the underlying computing system's compiler technology stack. We start by showing that under the assumption that the computational system has unbounded but finite memory (i.e., the memory is large enough to hold all intermediate values) and that each cu has a uniform cost, our approach reduces to a well-studied directed bandwidth problem for the directed acyclic graphs (DAGs). Next, we consider a more realistic scenario where the computing system has limited memory and concurrent execution while still assuming a uniform cost. Using a new notion of (r,s) set cover of a DAG (nodes representing computational units and edges representing their interdependencies) we formulate the problem of reducing redundant computational steps in EM as a variation of a directed bandwidth problem. We show that the graph's minimum bandwidth is closely related to memory requirements for studying RRC. Finally, our preliminary experimental results are supportive of the proposed approach for RRC and promising that it can be applied to a broader set of algorithms in decision sciences.

QuanTaichi

High-resolution simulations can deliver great visual quality, but they are often limited by available memory, especially on GPUs. We present a compiler for physical simulation that can achieve both high performance and significantly reduced memory costs, by enabling flexible and aggressive quantization. Low-precision ("quantized") numerical data types are used and packed to represent simulation states, leading to reduced memory space and bandwidth consumption. Quantized simulation allows higher resolution simulation with less memory, which is especially attractive on GPUs. Implementing a quantized simulator that has high performance and packs the data tightly for aggressive storage reduction would be extremely labor-intensive and error-prone using a traditional programming language. To make the creation of quantized simulation practical, we have developed a new set of language abstractions and a compilation system. A suite of tailored domain-specific optimizations ensure quantized simulators often run as fast as the full-precision simulators, despite the overhead of encoding-decoding the packed quantized data types. Our programming language and compiler, based on Taichi , allow developers to effortlessly switch between different full-precision and quantized simulators, to explore the full design space of quantization schemes, and ultimately to achieve a good balance between space and precision. The creation of quantized simulation with our system has large benefits in terms of memory consumption and performance, on a variety of hardware, from mobile devices to workstations with high-end GPUs. We can simulate with levels of resolution that were previously only achievable on systems with much more memory, such as multiple GPUs. For example, on a single GPU, we can simulate a Game of Life with 20 billion cells (8× compression per pixel), an Eulerian fluid system with 421 million active voxels (1.6× compression per voxel), and a hybrid Eulerian-Lagrangian elastic object simulation with 235 million particles (1.7× compression per particle). At the same time, quantized simulations create physically plausible results. Our quantization techniques are complementary to existing acceleration approaches of physical simulation: they can be used in combination with these existing approaches, such as sparse data structures, for even higher scalability and performance.

Development of SOA-Based Hydrological Data Compilation Management System

As the results of hydrological data compilation are important fundamental information resources in China, how to efficiently manage them and provide data support for water science research is a major problem that needs to be addressed at present. In this paper, a hydrological database of Haihe River Basin was preliminarily established based on the compilation of the complete electronic data files, and a SOA-based management system of hydrological data compilation results was developed, featuring functions related to the compilation results. The research shows that the system boasts strong practicability and operability, and will be popularized in the later period, in order to improve the capacity and level of hydrological service, support water conservancy information and drive the modernization of hydrological service.

Development and research of rural renewable energy management and ecological management information system under the background of beautiful rural revitalization strategy

At present, rural areas in China have problems such as poor living environment, shortage of facilities, and industrial disappointment. They still use the concept and strategy of urban planning and design to manage the rural areas, which leads to the rural planning results become empty shells, which is time-consuming and laborious, but does not bring practical use to the improvement of rural life. The purpose of this paper is to better implement the rural revitalization strategy, more scientifically guide the beautiful rural construction, and more effectively restore the rural vitality. Based on the strategy of beautiful village planning under the background of rural revitalization strategy, this paper takes Shuanglongqiao Village, Fengming Town, Xichong County, Sichuan Province as the empirical research object, and comprehensively applies research methods such as literature inquiry, on-site investigation and comparative analysis. Firstly, the paper puts forward the background of rural revitalization strategy and the concept of beautiful rural planning. Based on the theory and practice of beautiful rural planning at home and abroad, it analyzes the guiding significance and enlightenment of these theories and practices on beautiful rural planning. And further reflect on the concept and strategy of planning and construction of beautiful villages; then, in combination with national policies, profoundly analyze the scientific connotation of rural revitalization strategy, on this basis, comprehensively analyze the requirements of rural revitalization strategy for beautiful rural construction, and then combine with rural planning The characteristics of the compilation system of urban planning, and the basic strategies of beautiful rural planning under the background of rural revitalization strategy. Finally, combined with the characteristics of the village and the construction needs, the paper summarizes the planning strategies suitable for the development needs and village characteristics of Shuanglongqiao Village under the background of rural revitalization strategy. On the basis of previous studies, this paper comprehensively interprets the scientific connotation of rural revitalization strategy, and conducts summative research on beautiful rural planning. It hopes to refine some beautiful rural planning concepts and strategies, and other similar Shuanglongqiao villages under the background of rural revitalization strategy. The beautiful village planning provides a new perspective and also provides some experience in building practices in other villages.

Implementation and Optimization of Data Prefetching Algorithm Based on LLVM Compilation System

In order to reduce the problem of mismatch between high-performance processors and DRAM speeds, current processors have added a cache structure, but the low cache hit rate also seriously affects the actual performance of the program. Data prefetching technology can alleviate the problems of memory access latency and low hit rate caused by the speed difference between high-performance processors and DRAM. Based on the LLVM open source compiler, this article first implements the data prefetch module on the Shenwei platform. This paper improves the prefetch distance algorithm, proposes a new prefetch scheduling algorithm, introduces a cost model to evaluate the prefetch revenue, and accurately determines the insertion timing of the prefetch instruction to improve the cache hit rate. SPEC2006 performance test results show that after optimization, Shenwei 1621 processor single-core can achieve a maximum performance improvement of 50%, and an average performance improvement of 11%.

A Machine-Learning-Based Framework for Productive Locality Exploitation

Data locality is of extreme importance in programming distributed-memory architectures due to its implications on latency and energy consumption. Automated compiler and runtime system optimization studies have attempted to improve data locality exploitation without burdening the programmer. However, due to the difficulty of static code analysis, conservatism in compiler optimizations to avoid errors, and cost of dynamic analysis, the efficacy of automated optimizations is limited. Therefore, programmers need to spend significant effort in optimizing locality while creating applications for distributed memory parallel systems. We present a machine-learning based framework to automatically exploit locality in distributed memory applications. This framework takes application source whose time-critical blocks are marked by pragmas, and produces optimized source code that uses a regressor for efficient data movement. The regressor is trained with automatically-collected application profiles with very small input data sizes. We integrate our prototype in the Chapel language stack. In our experiments, we show that the Elastic Net model is the ideal regressor for our case and applications that utilize Elastic Net can perform very similarly to programmer-optimized versions. We also show that such regressors can be trained within few minutes on a cluster or within 30 minutes on a workstation, including data collection.

ФОРМУВАННЯ І ПОДАННЯ ФІНАНСОВОЇ ЗВІТНОСТІ В ЕЛЕКТРОННОМУ ВИГЛЯДІ

The article analyzes the composition of financial statements, indicates ways to improve its content and structure. The study outlines and solves the problem of determining the main stages of development of electronic reporting; the theoretical and methodical provisions of preparation and submission to the supervisory authorities of electronic financial reporting are substantiated; developed practical recommendations aimed at improving the methodology of electronic financial reporting. In order to further improve the methodological basis of financial reporting at Ukrainian enterprises, it is proposed to gradually update the accounting policy, use a balanced system of principles, techniques and methods of accounting given the impact of external and internal factors of the enterprise. The necessity of enshrining the concept of "electronic report" at the legislative level has been proved. It is defined that an electronic report is a document prepared by the relevant entities in electronic form in a special program, information in which is displayed in the form of appropriate electronic indicators for the purpose of the report with mandatory details and certified by qualified electronic signatures. The analysis of researches of domestic and foreign scientists concerning essence of the reporting and methodology of its drawing up is carried out. It is substantiated that the use of electronic reporting improves the connection between business entities and state regulatory authorities; improves the procedure for administering taxes and fees; serves to choose the best option for the relationship between control bodies and taxpayers and guarantees their maximum impartiality. The described problems of electronic reporting in Ukraine allow us to determine the prospects for further use of electronic reporting in order to improve the process of electronic document management at the enterprise level and at the state level. It is determined that the results obtained during the study can be used in the development of an effective system of compilation and submission to the control bodies of electronic reporting of domestic enterprises and organizations.