book

Learn LLVM 12

by Kai Nacke

May 2021

Beginner to intermediate

392 pages

8h 6m

English

Packt Publishing

Read now

Unlock full access

ContributorsAbout the authorAbout the reviewer
Who this book is forWhat this book coversTo get the most out of this bookDownload the example code filesCode in ActionDownload the color imagesConventions usedGet in touchReviews
Getting the prerequisites readyUbuntuFedora and RedHatFreeBSDOS XWindowsConfiguring GitBuilding with CMakeCloning the repositoryCreating a build directoryGenerating the build system filesCustomizing the build processVariables defined by CMakeVariables defined by LLVMSummary
Technical requirementsContents of the LLVM mono repositoryLLVM core libraries and additionsCompilers and toolsRuntime librariesLayout of an LLVM projectCreating your own project using LLVM librariesCreating the directory structureAdding the CMake filesAdding the C++ source filesCompiling the tinylang applicationTargeting a different CPU architectureSummary
Technical requirementsBuilding blocks of a compilerAn arithmetic expression languageFormalism for specifying the syntax of a programming languageHow grammar helps the compiler writerLexical analysisA handwritten lexerSyntactical analysisA handwritten parserThe abstract syntax treeSemantic analysisGenerating code with the LLVM backendTextual representation of the LLVM IRGenerating the IR from the AST The missing pieces – the driver and the runtime librarySummary
Technical requirementsDefining a real programming languageCreating the project layoutManaging source files and user messagesStructuring the lexerConstructing a recursive descent parserGenerating a parser and lexer with bison and flexPerforming semantic analysisHandling the scope of namesUsing LLVM-style RTTI for the ASTCreating the semantic analyzerSummary
Technical requirementsGenerating IR from the ASTUnderstanding the IR codeKnowing the load-and-store approachMapping the control flow to basic blocksUsing AST numbering to generate IR code in SSA formDefining the data structure to hold valuesReading and writing values local to a basic blockSearching the predecessor blocks for a valueOptimizing the generated phi instructionsSealing a blockCreating IR code for expressionsEmitting the IR code for a functionControlling visibility with linkage and name manglingConverting types from an AST description to LLVM typesCreating the LLVM IR functionEmitting the function bodySetting up the module and the driverWrapping everything in the code generatorInitializing the target machine classEmitting assembler text and object codeSummary
Technical requirementsWorking with arrays, structs, and pointersGetting the application binary interface rightCreating IR code for classes and virtual functionsImplementing single inheritanceExtending single inheritance with interfaces Adding support for multiple inheritanceSummary

Technical requirementsThrowing and catching exceptionsRaising an exceptionCatching an exceptionIntegrating the exception-handling code into the applicationGenerating metadata for type-based alias analysisUnderstanding the need for additional metadataAdding TBAA metadata to tinylangAdding debug metadataUnderstanding the general structure of debug metadataTracking variables and their valuesAdding line numbersAdding debug support to tinylangSummary
Technical requirementsIntroducing the LLVM Pass managerImplementing a Pass using the new Pass managerAdding a Pass to the LLVM source treeAdding a new Pass as a pluginAdapting a Pass for use with the old Pass managerAdding an optimization pipeline to your compilerCreating an optimization pipeline with the new Pass managerExtending the Pass pipelineSummary
Technical requirementsUnderstanding the LLVM target backend structureUsing MIR to test and debug the backendHow instruction selection worksSpecifying the target description in the TableGen languageInstruction selection with the selection DAGFast instruction selection – FastISelThe new global instruction selection – GlobalISelSupporting new machine instructionsAdding a new instruction to the assembler and code generationTesting the new instructionSummary
Technical requirementsGetting an overview of LLVM's JIT implementation and use casesUsing JIT compilation for direct executionExploring the lli toolImplementing our own JIT compiler with LLJITBuilding a JIT compiler class from scratchUtilizing a JIT compiler for code evaluationIdentifying the language semanticsSummary
Technical requirementsInstrumenting an application with sanitizersDetecting memory access problems with the address sanitizerFinding uninitialized memory access with the memory sanitizerPointing out data races with the thread sanitizerFinding bugs with libFuzzerLimitations and alternativesPerformance profiling with XRayChecking the source with the Clang Static AnalyzerAdding a new checker to the Clang Static AnalyzerCreating your own Clang-based toolSummary
Technical requirementsSetting the stage for a new backendAdding the new architecture to the Triple classExtending the ELF file format definition in LLVMCreating the target descriptionImplementing the top-level file of the target descriptionAdding the register definitionDefining the calling conventionCreating the scheduling modelDefining the instruction formats and the instruction informationImplementing the DAG instruction selection classesInitializing the target machineAdding the selection DAG implementationSupporting target-specific operationsConfiguring the target loweringGenerating assembler instructionsEmitting machine codeAdding support for disassemblingPiecing it all togetherSummary
Other Books You May EnjoyPackt is searching for authors like youLeave a review - let other readers know what you think

Overview

Learn LLVM 12 is a comprehensive guide for anyone starting their journey into compiler development using the LLVM framework. This book presents step-by-step lessons on configuring and using LLVM libraries, building a compiler from scratch, and extending LLVM functionalities according to your needs.

What this Book will help me do

Understand the structure and setup of the LLVM framework for development.
Build a basic compiler using LLVM tools and libraries.
Generate and optimize Intermediate Representation (IR) of code.
Extend LLVM with custom passes and backend implementations.
Apply LLVM-based tools to improve code analysis and performance.

Author(s)

The book is authored by Kai Nacke, who is a seasoned developer and enthusiast in compiler technologies. With extensive expertise in LLVM and a rich background in C++, Kai shares practical insights and hands-on instructions to help readers harness the power of LLVM effectively. Kai's approach to writing ensures that complex topics are explained in a clear and accessible manner.

Who is it for?

This book is ideal for developers who are new to LLVM and interested in compiler development. With a focus on practical applications of LLVM tools, it suits intermediate C++ developers who wish to leverage compiler-based optimizations and analysis in their projects. Those curious about the inner workings of compilers and transferring knowledge into building backend systems would find this book rewarding. Individuals with a passion for understanding the nitty-gritty of code compilation frameworks are the ideal readers.