Haiku Git Repositories

Haiku uses Git for source control, combined with Gerrit for review of code changes.

Most of the operating system sources are stored in a single repository at

http://cgit.haiku-os.org/haiku

.

Another repository at

http://cgit.haiku-os.org/buildtools

contains the build tools, that is, gcc, binutils, and Jam, which are maintained by Haiku developers.

Additional repositories

are hosed on GitHub.

Finally, some pre-compiled packages are downloaded during the build, these are built using Haikuporter from

recipes available here

.

Getting the sourcecode

https://www.haiku-os.org/guides/building/get-source-git

Sending change reviews

https://dev.haiku-os.org/wiki/CodingGuidelines/SubmittingPatches

https://review.haiku-os.org/Documentation/user-upload.html

Source tree organization

The source tree is organized so you can easily find what you look for. If you’re already familiar with Haiku, you will notice that the source directory generally mirrors the way the filesystem in Haiku is organized.

At the top level, things that need to be “built” in some way are put in the "src" directory. For example, the “data” folder at the root contains files that are used as-is in the disk image, while “src/data” contain files that need to be compild or converted to different formats, such as the MIME database.

• src - All files that have to be built

• add-ons - Everything that will be installed to /boot/system/add-ons: kernel drivers, media codecs, translators, …

• apps - GUI applications that are not preferences

• bin - Command-line applications

• build - Files to allow using the Haiku buildtools on non-Haiku platforms

• data - Data files of any type: icons, MIME database, …

• kits - The public C++ API of Haiku: libbe, libmedia, libgame, …

• libs - Static and shared libraries used by Haiku applications

• preferences - The preference applications

• servers - The system servers: app_server, input_server, net_server, …

• system - The low-level system that makes Haiku tick

• boot - The bootloaders for all supported platforms

• glue - The “glue code” that makes shared libraries execute their constructors and destructors, and programs start their execution at "main()"

• kernel - The kernel and all of its core services

• ldscripts - Linker scripts for building various parts of Haiku

• libnetwork - Files for building libnetwork.so, including the POSIX/BSD socket implementation and some extensions to it, as well as the DNS resolver

• libroot - Files for building libroot.so, including the standard C and POSIX library implementation

• runtime_loader: The special application that knows how to load and run other applications from ELF executable files

• tests - This more or less mirrors the main source tree layout, and contains tests and debugging tools for each component. Some of the tests are run using cppunit, other can be run manually.

• tools - Tools that can be built on non-Haiku platforms. Either needed for compiling Haiku itself, or otherwise useful outside of Haiku

• headers - All shared, private and public headers

• build - Compatibility headers for building Haiku code on non-Haiku systems

• compatibility - Compatbility layers allowing to build BSD and GNU code to run on Haiku

• config - Platform-specific configuration, definition of standard types that can be used in other places

• cpp - The C++ standard library headers (only for gcc2, for later gcc versions this is provided by the gcc package)

• glibc - Headers from glibc, for configuration and definition of some glibc specific functions

• libs - Headers for the libraries found in src/libs

• os - The public headers that define the Haiku API

• posix - The POSIX APIs supported by Haiku

• private - Private headers that are shared between haiku components, including work-in-progress APIs that may become public in the future

• tools - Headers for various tools and utilities

• docs - Documentation

• develop - Internal documentation for developers working on Haiku itself (this is what you are reading now)

• user - *API reference <https://api.haiku-os.org>*

• Some other miscellaneous documentation

• build - Build files

• jam - Jam rules used by the Haiku build, defining how to build an application, a library, a disk image, …

• config_headers - Configurable headers for enabling various debug features

• scripts - Various scripts used by the Haiku build process

• 3rd_party - Developers custom files. Used for various side projects from Haiku developers, useful personal scripts, and integration with other tools and projects such as virtualization software

Managing 3rd-party code

GCC and binutils updates using vendor branches

The buidtools repository uses vendor branches. This concept originates from

the SVN Book

but applies just as well to Git. This organization allows to clearly separate the imported code from upstream, and the changes we have made to it.

The idea is to import all upstream changes in a dedicated branch (there are currently two, called vendor-gcc and vendor-binutils). These branches contains the sources of gcc and binutils as distributed by the GNU project, without any Haiku changes.

The master branch can then merge new versions from the vendor branches. This allows to use Git conflict resolution to make sure our patches are correctly ported from one version to the next.

It also makes it easy to compare the current state of our sourcecode with the upstream code, for example to extract patches that could be upstreamed.

# How to import upstream binutils changes

Here is an example of the process used to update to a new version of binutils:

git checkout vendor-binutils          # Move to the branch containing binutils
git rm -rf binutils ; rm -rf binutils # Delete the existing version of binutils
wget http://.../binutils-2.36.tar.xz  # Download the latest version
tar xf binutils-2.36.tar.xz           # Extract the new binutils version
mv binutils-2.36 binutils             # Move the extracted files to the right place
git add -f binutils                   # Add the new files to git
git commit -m "import binutils 2.36"  # Commit the files in the vendor branch
git push origin vendor-binutils       # You can push this directly to the branch

Now this can easily be merged into the master branch:

git checkout master
git merge vendor-binutils

Review and fix the conflicts, if any, then push the changes for review on Gerrit.

# How to import upstream gcc changes

Here is an example of the process used to update to a new version of binutils:

git checkout vendor-gcc               # Move to the branch containing binutils
git rm -rf gcc ; rm -rf gcc           # Delete the existing version of binutils
wget http://.../gcc-13.2.0.tar.xz     # Download the latest version
tar xf gcc-13.2.0.tar.xz              # Extract the new binutils version
mv gcc-13.2.0 gcc                     # Move the extracted files to the right place
pushd gcc
./contrib/download_prerequisites      # Download the required gmp, isl, mpfr and mpc dependencies
rm gmp gmp-6.2.1.tar.bz2              # Remove gmp download and symbolic link
mv gmp-6.2.1 gmp                      # Move the downloaded gmp dependency in place
rm isl isl-0.24.tar.bz2
mv isl-0.24 isl
rm mpc mpc-1.2.1.tar.gz
mv mpc-1.2.1 mpc
rm mpfr mpfr-4.1.0.tar.bz2
mv mpfr-4.1.0 mpfr
popd
git add -f gcc                        # Add the new files to git
git commit -m "import gcc 13.2.0"     # Commit the files in the vendor branch
git push origin vendor-binutils       # You can push this directly to the branch

Now this can easily be merged into the master branch:

git checkout master
git merge vendor-binutils

Review and fix the conflicts, if any, then push the changes for review on Gerrit.

# Comparing our code with upstream

Comparing the two versions is easy because you can refer to them by branch names:

git diff vendor-binutils master -- binutils

Third party code in Haiku without vendor branches

The Haiku main repository contains third party code that does not use vendor branches. This can be for various reasons, such as:

• Imported files are only a very small subset of another project

• The upstream project is not developped actively anymore

• The import is not done from a specific release of the upstream project, but done piecewise.

Some examples of this are:

• All the code in src/libs (often not maintained upstream or very simple),

• The NetBSD DNS resolution code (one version was packaged as a separate “netresolv” package, but that approach was abandoned by NetBSD)

• Parts of glibc, musl and BSD C library that are reused in libroot (each file imported independently as needed)

• FreeBSD and OpenBSD network drivers (each driver synchronized as needed)

In this case, since there is no branch to track where the upstream code came from and what changes were made, a few specific steps should be taken to avoid confusion and possibly losing some of our changes and fixes:

• Mark all changes made on Haiku side with a __HAIKU__ preprocessor guard. This makes them easy to identify, and allows upstream to integrate these changes if they want to

• When importing newer versions, note in the commit message where you got the files and which version you used (either a release number, or the commit identifier if using files directly from upstream source control)

• Keep the changes to a minimum to ease future sycnhronizations with upstream. In particular, it is often not possible to build these files with -Werror due to type declaration differences

When importing new version of the files from upstream, make sure to review all the places where there are __HAIKU__ guards, and consider wether they still apply in the new version, or if the upstream code has been changed so they can be removed.