/cmake-git-version-tracking

Embed Git metadata in a C/C++ compatible static library using CMake (e.g. commit ID, author, date, and message).

Primary LanguageCMakeMIT LicenseMIT

Regression Tests

Embed Git metadata in C/C++ projects via CMake

This project embeds up-to-date git metadata in a standalone C/C++ static library via CMake. It's written responsibly to only trigger rebuilds if git metadata changes (e.g. a new commit is added). The core capability is baked into single self-contained script.

Requirements

  • CMake >= 3.2
  • C Compiler (with C99 standard support)
  • Git

Quickstart via FetchContent

You can use CMake's FetchContent module to build the static library cmake_git_version_tracking:

include(FetchContent)
FetchContent_Declare(cmake_git_version_tracking                   
  GIT_REPOSITORY https://github.com/andrew-hardin/cmake-git-version-tracking.git
  GIT_TAG 904dbda1336ba4b9a1415a68d5f203f576b696bb
)
FetchContent_MakeAvailable(cmake_git_version_tracking)

target_link_libraries(your_target
  cmake_git_version_tracking
)

Then #include git.h and use the provided functions to retrieve git metadata.

Intended use case

You're continuously shipping prebuilt binaries for an application. A user discovers a bug and files a bug report. By embedding up-to-date versioning information, the user can include this information in their report, e.g.:

Commit SHA1: 46a396e (46a396e6c1eb3d)
Dirty: false (there were no uncommitted changes at time of build)

This allows you to investigate the precise version of the application that the bug was reported in.

Q: What if I want to track $special_git_field?

Fork the project and modify git_watcher.cmake to track new additional fields (e.g. kernel version or build hostname). Sections that need to be modified are marked with >>>.

Q: Doesn't this already exist?

It depends on your specific requirements. Before writing this, I found two categories of existing solutions:

  • Write the commit ID to the header at configure time (e.g. cmake <source_dir>). This works well for automated build processes (e.g. check-in code and build artifacts). However, any changes made after running cmake (e.g. git commit -am "Changed X") aren't reflected in the header.

  • Every time a build is started (e.g. make), write the commit ID to a header. The major drawback of this method is that any object file that includes the new header will be recompiled -- even if the state of the git repo hasn't changed.

Q: What's the better solution?

We check Git every time a build is started (e.g. make) to see if anything has changed, like a new commit to the current branch. If nothing has changed, then we don't touch anything- no recompiling or linking is triggered. If something has changed, then we reconfigure the header and CMake rebuilds any downstream dependencies.