Building a Distribution of LLVM

Introduction

This document is geared toward people who want to build and package LLVM and any combination of LLVM sub-project tools for distribution. This document covers useful features of the LLVM build system as well as best practices and general information about packaging LLVM.

If you are new to CMake you may find the Building LLVM with CMake or CMake Primer documentation useful. Some of the things covered in this document are the inner workings of the builds described in the Advanced Build Configurations document.

General Distribution Guidance

When building a distribution of a compiler it is generally advised to perform a bootstrap build of the compiler. That means building a “stage 1” compiler with your host toolchain, then building the “stage 2” compiler using the “stage 1” compiler. This is done so that the compiler you distribute benefits from all the bug fixes, performance optimizations and general improvements provided by the new compiler.

In deciding how to build your distribution there are a few trade-offs that you will need to evaluate. The big two are:

1. Compile time of the distribution against performance of the built compiler
2. Binary size of the distribution against performance of the built compiler

The guidance for maximizing performance of the generated compiler is to use LTO, PGO, and statically link everything. This will result in an overall larger distribution, and it will take longer to generate, but it provides the most opportunity for the compiler to optimize.

The guidance for minimizing distribution size is to dynamically link LLVM and Clang libraries into the tools to reduce code duplication. This will come at a substantial performance penalty to the generated binary both because it reduces optimization opportunity, and because dynamic linking requires resolving symbols at process launch time, which can be very slow for C++ code.

Warning

One very important note: Distributions should never be built using the BUILD_SHARED_LIBS CMake option. That option exists for optimizing developer workflow only. Due to design and implementation decisions, LLVM relies on global data which can end up being duplicated across shared libraries resulting in bugs. As such this is not a safe way to distribute LLVM or LLVM-based tools.

The simplest example of building a distribution with reasonable performance is captured in the DistributionExample CMake cache file located at clang/cmake/caches/DistributionExample.cmake. The following command will perform and install the distribution build:

$ cmake -G Ninja -C <path to clang>/cmake/caches/DistributionExample.cmake <path to LLVM source>
$ ninja stage2-distribution
$ ninja stage2-install-distribution

Difference between install and install-distribution

One subtle but important thing to note is the difference between the install and install-distribution targets. The install target is expected to install every part of LLVM that your build is configured to generate except the LLVM testing tools. Alternatively the install-distribution target, which is recommended for building distributions, only installs specific parts of LLVM as specified at configuration time by LLVM_DISTRIBUTION_COMPONENTS.

Additionally by default the install target will install the LLVM testing tools as the public tools. This can be changed well by setting LLVM_INSTALL_TOOLCHAIN_ONLY to On. The LLVM tools are intended for development and testing of LLVM, and should only be included in distributions that support LLVM development.

When building with LLVM_DISTRIBUTION_COMPONENTS the build system also generates a distribution target which builds all the components specified in the list. This is a convenience build target to allow building just the distributed pieces without needing to build all configured targets.

Special Notes for Library-only Distributions

One of the most powerful features of LLVM is its library-first design mentality and the way you can compose a wide variety of tools using different portions of LLVM. Even in this situation using BUILD_SHARED_LIBS is not supported. If you want to distribute LLVM as a shared library for use in a tool, the recommended method is using LLVM_BUILD_LLVM_DYLIB, and you can use LLVM_DYLIB_COMPONENTS to configure which LLVM components are part of libLLVM.

Options for Optimizing LLVM

There are four main build optimizations that our CMake build system supports. When performing a bootstrap build it is not beneficial to do anything other than setting CMAKE_BUILD_TYPE to Release for the stage-1 compiler. This is because the more intensive optimizations are expensive to perform and the stage-1 compiler is thrown away. All of the further options described should be set on the stage-2 compiler either using a CMake cache file, or by prefixing the option with BOOTSTRAP_.

The first and simplest to use is the compiler optimization level by setting the CMAKE_BUILD_TYPE option. The main values of interest are Release or RelWithDebInfo. By default the Release option uses the -O3 optimization level, and RelWithDebInfo uses -O2. If you want to generate debug information and use -O3 you can override the CMAKE_<LANG>_FLAGS_RELWITHDEBINFO option for C and CXX. DistributionExample.cmake does this.

Another easy to use option is Link-Time-Optimization. You can set the LLVM_ENABLE_LTO option on your stage-2 build to Thin or Full to enable building LLVM with LTO. These options will significantly increase link time of the binaries in the distribution, but it will create much faster binaries. This option should not be used if your distribution includes static archives, as the objects inside the archive will be LLVM bitcode, which is not portable.

The Advanced Build Configurations documentation describes the built-in tooling for generating LLVM profiling information to drive Profile-Guided-Optimization. The in-tree profiling tests are very limited, and generating the profile takes a significant amount of time, but it can result in a significant improvement in the performance of the generated binaries.

In addition to PGO profiling we also have limited support in-tree for generating linker order files. These files provide the linker with a suggested ordering for functions in the final binary layout. This can measurably speed up clang by physically grouping functions that are called temporally close to eachother. The current tooling is only available on Darwin systems with dtrace(1). It is worth noting that dtrace is non-deterministic, and so the order file generation using dtrace is also non-deterministic.

Options for Reducing Size

Warning

Any steps taken to reduce the binary size will come at a cost of runtime performance in the generated binaries.

The simplest and least significant way to reduce binary size is to set the CMAKE_BUILD_TYPE variable to MinSizeRel, which will set the compiler optimization level to -Os which optimizes for binary size. This will have both the least benefit to size and the least impact on performance.

The most impactful way to reduce binary size is to dynamically link LLVM into all the tools. This reduces code size by decreasing duplication of common code between the LLVM-based tools. This can be done by setting the following two CMake options to On: LLVM_BUILD_LLVM_DYLIB and LLVM_LINK_LLVM_DYLIB.

Warning

Distributions should never be built using the BUILD_SHARED_LIBS CMake option. (See the warning above for more explanation.).

Relevant CMake Options

This section provides documentation of the CMake options that are intended to help construct distributions. This is not an exhaustive list, and many additional options are documented in the Building LLVM with CMake page. Some key options that are already documented include: LLVM_TARGETS_TO_BUILD, LLVM_ENABLE_PROJECTS, LLVM_BUILD_LLVM_DYLIB, and LLVM_LINK_LLVM_DYLIB.

LLVM_ENABLE_RUNTIMES:STRING

When building a distribution that includes LLVM runtime projects (i.e. libcxx, compiler-rt, libcxxabi, libunwind…), it is important to build those projects with the just-built compiler.

LLVM_DISTRIBUTION_COMPONENTS:STRING

This variable can be set to a semi-colon separated list of LLVM build system components to install. All LLVM-based tools are components, as well as most of the libraries and runtimes. Component names match the names of the build system targets.

LLVM_RUNTIME_DISTRIBUTION_COMPONENTS:STRING

This variable can be set to a semi-colon separated list of runtime library components. This is used in conjunction with LLVM_ENABLE_RUNTIMES to specify components of runtime libraries that you want to include in your distribution. Just like with LLVM_DISTRIBUTION_COMPONENTS, component names match the names of the build system targets.

LLVM_DYLIB_COMPONENTS:STRING

This variable can be set to a semi-colon separated name of LLVM library components. LLVM library components are either library names with the LLVM prefix removed (i.e. Support, Demangle…), LLVM target names, or special purpose component names. The special purpose component names are:

1. all - All LLVM available component libraries
2. Native - The LLVM target for the Native system
3. AllTargetsAsmPrinters - All the included target ASM printers libraries
4. AllTargetsAsmParsers - All the included target ASM parsers libraries
5. AllTargetsDescs - All the included target descriptions libraries
6. AllTargetsDisassemblers - All the included target dissassemblers libraries
7. AllTargetsInfos - All the included target info libraries

LLVM_INSTALL_TOOLCHAIN_ONLY:BOOL

This option defaults to Off: when set to On it removes many of the LLVM development and testing tools as well as component libraries from the default install target. Including the development tools is not recommended for distributions as many of the LLVM tools are only intended for development and testing use.