Getting Started With LLVM (1)

您所在的位置：网站首页 › llvm download page › Getting Started With LLVM (1)

Getting Started With LLVM (1)

2023-03-08 09:18| 来源: 网络整理| 查看: 265

Why should I learn LLVM?

According to ChatGPT, there are several reasons why you should learn the LLVM compiler framework:

Popularity: LLVM is widely used in the industry, making it an important skill for those working in software development or computer science.

Versatility: LLVM can be used for a wide range of programming languages, making it a valuable tool for those who work with multiple languages.

Performance: LLVM is known for its high-performance code generation, making it a valuable tool for optimizing software applications.

Open-source: LLVM is an open-source framework, which means that you have access to the source code and can make modifications to fit your needs.

Community: LLVM has a large and active community of developers and users, making it easier to find support and resources when working with the framework.

Career opportunities: Knowledge of LLVM can help you stand out in the job market and open up new career opportunities in areas such as compiler development, software engineering, and optimization.

How to get started according to ChatGPT?

According to ChatGPT:

To get started with the LLVM compiler framework, follow these steps:

Familiarize yourself with the basics of compilers and how they work.

Read the LLVM documentation: The LLVM documentation is a great resource for learning about the framework, its features, and how to use them.

Set up a development environment: You will need to have a development environment set up with the LLVM compiler framework installed. There are many tutorials available online to help you get started with the setup process.

Practice writing simple programs: Start by writing simple programs and experimenting with the LLVM compiler framework to get a feel for how it works.

Participate in online communities: Join online forums or communities dedicated to LLVM to ask questions, learn from others, and stay up-to-date with the latest developments.

Read LLVM source code: Reading LLVM source code is a great way to understand how the framework works and learn best practices for using it.

Get involved in development: Consider contributing to the LLVM project by fixing bugs or adding new features. This is a great way to get hands-on experience with the framework and make a meaningful contribution to the community.

A concrete way to get started

Well, that sounds abstract. Let us be more concrete. LLVM has an excellent tutorial. However, there are challenges to get started:

The LLVM tutorial started by guiding us to build LLVM from source. It takes hours and LLVM is very demanding of the host C++ compiler, and as such tends to expose bugs in the compiler.

The LLVM tutorial quickly jumps to more advanced compiler concepts, such Static Single Assignment(SSA). A blog recommended a large number of materials to learn before even starting.

So the objective of this blog series is to provide some concrete steps to start with LLVM. We will solve the 2 issues above with:

Instead of building LLVM from the source, we will download prebuilt binaries as well as how to use these binaries.

I will try to explain more advanced compiler concepts, assuming you already have some basic ideas about a compiler, such as lexer, parser, syntactic and semantic analysis, optimization and code-generation.

Getting the LLVM binaries

The LLVM download page provides prebuilt binaries for a variety of operating systems and CPU architectures. On Linux or MacOS, we can download clang+llvm* for our platform. For Windows, download llvm*-win64.exe. If the download is a compressed archive, just expand into any directory.

On some platforms, it is possible package manager. For example, on macOS, we can install it with Homebrew.

The installers/package managers usually do not add Clang/LLVM to PATH and that is fine. It is likely we will need to work with multiple versions of LLVM. Some OS, such as macOS, has a system Clang, so we do not want to override it. I strongly recommend that we add Clang/LLVM in terminal sessions when we need them.

Add Clang/LLVM to path

To add LLVM to path, just run:

export PATH=PATH_TO_LLVM/bin:$PATH

If we install LLVM on macOS with Brew, we can run

export PATH="$(brew --prefix llvm)/bin:$PATH" Other environment variables

To build C++ with LLVM libraries, we can configure the LDFLAGS environment variable to point to the libraries. This can be done with:

export LDFLAGS="-LPATH_TO_LLVM/lib"

Or with Brew installed LLVM:

export LDFLAGS="-L$(brew --prefix llvm)/lib"

Alternatively, we can also use the llvm-config utility that comes with LLVM:

clang++ `llvm-config --cxxflags` my_source.cpp Building LLVM tutorial code

Once setting up, we can now build our first LLVM source code. Let us save the tutorial code into a file, said, toy.cpp. Assuming Clang is already in the path, we can run:

# Compile clang++ -g -O3 toy.cpp `llvm-config --cxxflags` # Run ./a.out

Here -g means to generate debug symbol and -O3 is optimization level 3. We can now run Kaleidoscope Read-Eval-Print Loop (REPL):

$ ./a.out ready> def foo(x y) x+foo(y, 4.0); Parsed a function definition. ready> def foo(x y) x+y y; Parsed a function definition. Parsed a top-level expr ready> def foo(x y) x+y ); Parsed a function definition. Error: unknown token when expecting an expression ready> extern sin(a); ready> Parsed an extern ready> ^D $

The first 2 chapters of LLVM tutorial are relatively straigh-forward. You should read through them if you are not already familiar with the material.

Summary

In the first blog post of this series, we walk through how to compile and run the first LLVM tutorial.

Last modified on 2023-02-04

【本文地址】

Getting Started With LLVM (1)

Getting Started With LLVM (1)

今日新闻

推荐新闻