buildroot/docs/buildroot.html
Bernhard Reutner-Fischer 53cc639eef - trivial style fixes.
TODO: document HAVE_MANPAGES and HAVE_INFOPAGES (and adjust line-numbers in the text accordingly)
2007-09-19 09:08:10 +00:00

1092 lines
48 KiB
HTML

<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<title>Buildroot - Usage and documentation</title>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
<link rel="stylesheet" type="text/css" href="stylesheet.css" />
</head>
<body>
<div class="main">
<div class="titre">
<h1>Buildroot</h1>
</div>
<p><a href="http://buildroot.uclibc.org/">Buildroot</a>
usage and documentation by Thomas Petazzoni. Contributions from
Karsten Kruse, Ned Ludd, Martin Herren and others. </p>
<p><small>$LastChangedDate$</small></p>
<ul>
<li><a href="#about">About Buildroot</a></li>
<li><a href="#download">Obtaining Buildroot</a></li>
<li><a href="#using">Using Buildroot</a></li>
<li><a href="#custom_targetfs">Customizing the target filesystem</a></li>
<li><a href="#custom_busybox">Customizing the Busybox
configuration</a></li>
<li><a href="#custom_uclibc">Customizing the uClibc
configuration</a></li>
<li><a href="#buildroot_innards">How Buildroot works</a></li>
<li><a href="#using_toolchain">Using the uClibc toolchain</a></li>
<li><a href="#toolchain_standalone">Using the uClibc toolchain
outside of Buildroot</a></li>
<li><a href="#downloaded_packages">Location of downloaded packages</a>
</li>
<li><a href="#add_software">Extending Buildroot with more
Software</a></li>
<li><a href="#links">Resources</a></li>
</ul>
<h2><a name="about" id="about"></a>About Buildroot</h2>
<p>Buildroot is a set of Makefiles and patches that allow to easily
generate both a cross-compilation toolchain and a root filesystem for your
target. The cross-compilation toolchain uses uClibc (<a href=
"http://www.uclibc.org/">http://www.uclibc.org/</a>), a tiny C standard
library. </p>
<p>Buildroot is useful mainly for people working with embedded systems.
Embedded systems often use processors that are not the regular x86
processors everyone is used to have on his PC. It can be PowerPC
processors, MIPS processors, ARM processors, etc. </p>
<p>A compilation toolchain is the set of tools that allows to
compile code for your system. It consists of a compiler (in our
case, <code>gcc</code>), binary utils like assembler and linker
(in our case, <code>binutils</code>) and a C standard library (for
example <a href="http://www.gnu.org/software/libc/libc.html">GNU
Libc</a>, <a href="http://www.uclibc.org/">uClibc</a> or <a
href="http://www.fefe.de/dietlibc/">dietlibc</a>). The system
installed on your development station certainly already has a
compilation toolchain that you can use to compile application that
runs on your system. If you're using a PC, your compilation
toolchain runs on an x86 processor and generates code for a x86
processor. Under most Linux systems, the compilation toolchain
uses the GNU libc as C standard library. This compilation
toolchain is called the &quot;host compilation toolchain&quot;, and more
generally, the machine on which it is running, and on which you're
working is called the &quot;host system&quot;. The compilation toolchain
is provided by your distribution, and Buildroot has nothing to do
with it. </p>
<p>As said above, the compilation toolchain that comes with your system
runs and generates code for the processor of your host system. As your
embedded system has a different processor, you need a cross-compilation
toolchain: it's a compilation toolchain that runs on your host system but
that generates code for your target system (and target processor). For
example, if your host system uses x86 and your target system uses ARM, the
regular compilation toolchain of your host runs on x86 and generates code
for x86, while the cross-compilation toolchain runs on x86 and generates
code for ARM. </p>
<p>Even if your embedded system uses a x86 processor, you might interested
in Buildroot, for two reasons:</p>
<ul>
<li>The compilation toolchain of your host certainly uses the GNU Libc
which is a complete but huge C standard library. Instead of using GNU
Libc on your target system, you can use uClibc which is a tiny C standard
library. If you want to use this C library, then you need a compilation
toolchain to generate binaries linked with it. Buildroot can do it for
you. </li>
<li>Buildroot automates the building of a root filesystem with all needed
tools like busybox. It makes it much easier than doing it by hand. </li>
</ul>
<p>You might wonder why such a tool is needed when you can compile
<code>gcc</code>, <code>binutils</code>, uClibc and all the tools by hand.
Of course, doing so is possible. But dealing with all configure options,
with all problems of every <code>gcc</code> or <code>binutils</code>
version it very time-consuming and uninteresting. Buildroot automates this
process through the use of Makefiles, and has a collection of patches for
each <code>gcc</code> and <code>binutils</code> version to make them work
on most architectures. </p>
<h2><a name="download" id="download"></a>Obtaining Buildroot</h2>
<p>Buildroot is available as daily SVN snapshots or directly using
SVN. </p>
<p>The latest snapshot is always available at <a
href="http://buildroot.uclibc.org/downloads/snapshots/buildroot-snapshot.tar.bz2">http://buildroot.uclibc.org/downloads/snapshots/buildroot-snapshot.tar.bz2</a>,
and previous snapshots are also available at <a
href="http://buildroot.uclibc.org/downloads/snapshots/">http://buildroot.uclibc.org/downloads/snapshots/</a>. </p>
<p>To download Buildroot using SVN, you can simply follow
the rules described on the &quot;Accessing SVN&quot;-page (<a href=
"http://buildroot.uclibc.org/subversion.html">http://buildroot.uclibc.org/subversion.html</a>)
of the uClibc buildroot website (<a href=
"http://buildroot.uclibc.org">http://buildroot.uclibc.org</a>), and download the
<code>buildroot</code> SVN module. For the impatient, here's a quick
recipe:</p>
<pre>
$ svn co svn://uclibc.org/trunk/buildroot
</pre>
<h2><a name="using" id="using"></a>Using Buildroot</h2>
<p>Buildroot has a nice configuration tool similar to the one you can find
in the Linux Kernel (<a href=
"http://www.kernel.org/">http://www.kernel.org/</a>) or in Busybox
(<a href="http://www.busybox.org/">http://www.busybox.org/</a>). Note that
you can build everything as a normal user. There is no need to be root to
configure and use Buildroot. The first step is to run the configuration
assistant:</p>
<pre>
$ make menuconfig
</pre>
<p>For each entry of the configuration tool, you can find associated help
that describes the purpose of the entry. </p>
<p>One of the key configuration items is the <code>PROJECT</code> which
determines where some board specific packages are built and where the
results are stored. </p>
<p>Once everything is configured, the configuration tool has generated a
<code>.config</code> file that contains the description of your
configuration. It will be used by the Makefiles to do what's needed. </p>
<p>Let's go:</p>
<pre>
$ make
</pre>
<p>This command will download, configure and compile all the selected
tools, and finally generate a target filesystem. The target filesystem will
be named <code>root_fs_ARCH.EXT</code> where <code>ARCH</code> is your
architecture and <code>EXT</code> depends on the type of target filesystem
selected in the <code>Target options</code> section of the configuration
tool.
The file is stored in the "binaries/<code>$(PROJECT)</code>/" directory</p>
<h3><a name="local_board_support" id="local_board_support"></a>
Creating your own board support</h3>
<p>Once a package has been unpacked, it is possible to manually update
configuration files. Buildroot can automatically save the configuration
of buildroot, linux, busybox, uclibc and u-boot in "local/$(PROJECT) by
using the command:
</p>
<pre>
$ make saveconfig
</pre>
<p>Once a buildroot configuration has been created by saveconfig,
the default "$(TOPDIR)/.config" file can be overridden by</p>
<pre>
$ make BOARD=&lt;project&gt;
</pre>
<p>Buildroot will then use "local/&lt;project&gt;/&lt;project&gt;.config"
instead of ".config". </p>
<p>If you want to modify your board, you can copy the project configuration
file to ".config" by using the command:</p>
<pre>
$ make BOARD=&lt;project&gt; getconfig
</pre>
<p>You can share your custom board support directory between several buildroot trees
by setting the environment variable <code>BUILDROOT_LOCAL</code> to this directory,
</p>
<h3><a name="offline_builds" id="offline_builds"></a>
Offline builds</h3>
<p>If you intend to do an offline-build and just want to download all
sources that you previously selected in &quot;make menuconfig&quot; then
issue:</p>
<pre>
$ make source
</pre>
<p>You can now disconnect or copy the content of your <code>dl</code>
directory to the build-host. </p>
<h3><a name="environment_variables" id="environment_variables"></a>
Environment variables</h3>
<p>Buildroot optionally honors some environment variables that are passed
to <code>make</code> :</p>
<ul>
<li>HOSTCXX</li>
<li>HOSTCC</li>
<li>UCLIBC_CONFIG_FILE=&lt;path/to/.config&gt;</li>
<li>BUSYBOX_CONFIG_FILE=&lt;path/to/.config&gt;</li>
</ul>
<p>An example that uses config files located in the toplevel directory and
in your $HOME:</p>
<pre>
$ make UCLIBC_CONFIG_FILE=uClibc.config BUSYBOX_CONFIG_FILE=$HOME/bb.config
</pre>
<p>If you want to use a compiler other than the default <code>gcc</code>
or <code>g++</code> for building helper-binaries on your host, then do</p>
<pre>
$ make HOSTCXX=g++-4.3-HEAD HOSTCC=gcc-4.3-HEAD
</pre>
<h3><a name="helper_completion" id="helper_completion"></a>
Using auto-completion</h3>
<p>If you are lazy enough that you don't want to type the entire <i>make
menuconfig</i> command line, you can enable auto-completion in your shell.
Here is how you can do that using <i>bash</i>:</p>
<pre>
$ complete -W menuconfig make
</pre>
<p>Then just enter the beginning of the line, and ask <i>bash</i> to
complete it for you by pressing the <i>TAB</i> key:</p>
<pre>
$ make me&lt;TAB&gt;
</pre>
<p>will result in <i>bash</i> to append <i>nuconfig</i> for you!</p>
<p>Alternatively, some distributions (of which Debian and Mandriva are but
an example) have more powerful make completion. Depending on you
distribution, you may have to install a package to enable completion. Under
Mandriva, this is <i>bash-completion</i>, while Debian ships it as part of
the <i>bash</i> package.</p>
<p>Other shells, such as <i>zsh</i>, also have completion facilities. See
the documentation for your shell.</p>
<h2><a name="custom_targetfs" id="custom_targetfs"></a>Customizing the
target filesystem</h2>
<p>There are a few ways to customize the resulting target filesystem:</p>
<ul>
<li>Customize the target filesystem directly, and rebuild the image. The
target filesystem is available under <code>project_build_ARCH/root/</code>
where <code>ARCH</code> is the chosen target architecture.
You can simply make your changes here, and run make afterwards, which will
rebuild the target filesystem image. This method allows to do everything
on the target filesystem, but if you decide to completely rebuild your
toolchain and tools, these changes will be lost. </li>
<li>Customize the target filesystem skeleton, available under
<code>target/generic/target_skeleton/</code>. You can customize
configuration files or other stuff here. However, the full file hierarchy
is not yet present, because it's created during the compilation process.
So you can't do everything on this target filesystem skeleton, but
changes to it remain even if you completely rebuild the cross-compilation
toolchain and the tools. <br />
You can also customize the <code>target/generic/device_table.txt</code>
file which is used by the tools that generate the target filesystem image
to properly set permissions and create device nodes. The
<code>target/generic/skel.tar.gz</code> file contains the main
directories of a root filesystem and there is no obvious reason for which
it should be changed. These main directories are in an tarball inside of
inside the skeleton because it contains symlinks that would be broken
otherwise. <br />
These customizations are deployed into
<code>project_build_ARCH/root/</code> just before the actual image
is made. So simply rebuilding the image by running
make should propagate any new changes to the image. </li>
<li>When configuring the build system, using <code>make menuconfig</code>,
you can specify the contents of the /etc/hostname and /etc/issue
(the welcome banner) in the <code>PROJECT</code> section</li>
</ul>
<h2><a name="custom_busybox" id="custom_busybox"></a>Customizing the
Busybox configuration</h2>
<p><a href="http://www.busybox.net/">Busybox</a> is very configurable, and
you may want to customize it. You can
follow these simple steps to do it. It's not an optimal way, but it's
simple and it works. </p>
<ol>
<li>Make a first compilation of buildroot with busybox without trying to
customize it. </li>
<li>Invoke <code>make busybox-menuconfig</code>.
The nice configuration tool appears and you can
customize everything. </li>
<li>Run the compilation of buildroot again. </li>
</ol>
<p>Otherwise, you can simply change the
<code>package/busybox/busybox-&lt;version&gt;.config</code> file if you
know the options you want to change without using the configuration tool.
</p>
<p>If you want to use an existing config file for busybox, then see
section <a href="#environment_variables">environment variables</a>. </p>
<h2><a name="custom_uclibc" id="custom_uclibc"></a>Customizing the uClibc
configuration</h2>
<p>Just like <a href="#custom_busybox">BusyBox</a>, <a
href="http://www.uclibc.org/">uClibc</a> offers a lot of
configuration options. They allow to select various
functionalities, depending on your needs and limitations. </p>
<p>The easiest way to modify the configuration of uClibc is to
follow these steps :</p>
<ol>
<li>Make a first compilation of buildroot without trying to
customize uClibc. </li>
<li>Invoke <code>make uclibc-menuconfig</code>.
The nice configuration assistant, similar to
the one used in the Linux Kernel or in Buildroot appears. Make
your configuration as appropriate. </li>
<li>Copy the <code>.config</code> file to
<code>toolchain/uClibc/uClibc.config</code> or
<code>toolchain/uClibc/uClibc.config-locale</code>. The former
is used if you haven't selected locale support in Buildroot
configuration, and the latter is used if you have selected
locale support. </li>
<li>Run the compilation of Buildroot again</li>
</ol>
<p>Otherwise, you can simply change
<code>toolchain/uClibc/uClibc.config</code> or
<code>toolchain/uClibc/uClibc.config-locale</code> without running
the configuration assistant. </p>
<p>If you want to use an existing config file for uclibc, then see
section <a href="#environment_variables">environment variables</a>. </p>
<h2><a name="buildroot_innards" id="buildroot_innards"></a>How Buildroot
works</h2>
<p>As said above, Buildroot is basically a set of Makefiles that download,
configure and compiles software with the correct options. It also includes
some patches for various software, mainly the ones involved in the
cross-compilation tool chain (<code>gcc</code>, <code>binutils</code> and
uClibc). </p>
<p>There is basically one Makefile per software, and they are named with
the <code>.mk</code> extension. Makefiles are split into four
sections:</p>
<ul>
<li><b>project</b> (in the <code>project/</code> directory) contains
the Makefiles and associated files for all software related to the
building several root file systems in the same buildroot tree. </li>
<li><b>toolchain</b> (in the <code>toolchain/</code> directory) contains
the Makefiles and associated files for all software related to the
cross-compilation toolchain : <code>binutils</code>, <code>ccache</code>,
<code>gcc</code>, <code>gdb</code>, <code>kernel-headers</code> and
<code>uClibc</code>. </li>
<li><b>package</b> (in the <code>package/</code> directory) contains the
Makefiles and associated files for all user-space tools that Buildroot
can compile and add to the target root filesystem. There is one
sub-directory per tool. </li>
<li><b>target</b> (in the <code>target</code> directory) contains the
Makefiles and associated files for software related to the generation of
the target root filesystem image. Four types of filesystems are supported
: ext2, jffs2, cramfs and squashfs. For each of them, there's a
sub-directory with the required files. There is also a
<code>default/</code> directory that contains the target filesystem
skeleton. </li>
</ul>
<p>Each directory contains at least 2 files :</p>
<ul>
<li><code>something.mk</code> is the Makefile that downloads, configures,
compiles and installs the software <code>something</code>. </li>
<li><code>Config.in</code> is a part of the configuration tool
description file. It describes the option related to the current
software. </li>
</ul>
<p>The main Makefile do the job through the following steps (once the
configuration is done):</p>
<ol>
<li>Create the download directory (<code>dl/</code> by default). This is
where the tarballs will be downloaded. It is interesting to know that the
tarballs are in this directory because it may be useful to save them
somewhere to avoid further downloads. </li>
<li>Create the shared build directory (<code>build_ARCH/</code> by
default, where <code>ARCH</code> is your architecture). This is where all
non configurable user-space tools will be compiled.When building two or
more targets using the same architecture, the first build will go through
the full download, configure, make process, but the second and later
builds will only copy the result from the first build to its project
specific target directory significantly speeding up the build process</li>
<li>Create the project specific build directory
(<code>project_build_ARCH/$(PROJECT)</code> by default, where
<code>ARCH</code> is your architecture). This is where all configurable
user-space tools will be compiled. The project specific build directory
is neccessary, if two different targets needs to use a specific package,
but the packages have different configuration for both targets. Some
examples of packages built in this directory are busybox and linux.
</li>
<li>Create the project specific result directory
(<code>binaries/$(PROJECT)</code> by default, where <code>ARCH</code>
is your architecture). This is where the root filesystem images are
stored, It is also used to store the linux kernel image and any
utilities, boot-loaders etc. needed for a target.
</li>
<li>Create the toolchain build directory
(<code>toolchain_build_ARCH/</code> by default, where <code>ARCH</code>
is your architecture). This is where the cross compilation toolchain will
be compiled. </li>
<li>Setup the staging directory (<code>build_ARCH/staging_dir/</code> by
default). This is where the cross-compilation toolchain will be
installed. If you want to use the same cross-compilation toolchain for
other purposes, such as compiling third-party applications, you can add
<code>build_ARCH/staging_dir/usr/bin</code> to your PATH, and then use
<code>arch-linux-gcc</code> to compile your application. In order to
setup this staging directory, it first removes it, and then it creates
various subdirectories and symlinks inside it. </li>
<li>Create the target directory (<code>project_build_ARCH/root/</code> by
default) and the target filesystem skeleton. This directory will contain
the final root filesystem. To setup it up, it first deletes it, then it
uncompress the <code>target/generic/skel.tar.gz</code> file to create the
main subdirectories and symlinks, copies the skeleton available in
<code>target/generic/target_skeleton</code> and then removes useless
<code>.svn/</code> directories. </li>
<li>Add the <code>TARGETS</code> dependency. This should generally check
if the configuration option for this package is enabled, and if so then
&quot;subscribe&quot; this package to be compiled by adding it to the
TARGETS global variable. </li>
</ol>
<h2><a name="multi_project" id="multi_project"></a>Building several
projects in the same buildroot source tree</h2>
<p><b>BACKGROUND</b></p>
<p>Buildroot has always supported building several projects in the same
tree if each project was for a different architecture. </p>
<p>The root file system has been created in the
<code>&quot;build_&lt;ARCH&gt;/root&quot;</code>
directory which is unique for each architecture.
Toolchains have been built in
<code>&quot;toolchain_build_&lt;ARCH&gt;&quot;</code>. </p>
<p> It the user wanted to build several root file systems for the same
architecture, a prefix or suffix could be added in the configuration file
so the root file system would be built in
<code>&quot;&lt;PREFIX&gt;_build_&lt;ARCH&gt;_&lt;SUFFIX&gt;/root&quot;</code>
By supplying <u>unique</u> combinations of
<code>&quot;&lt;PREFIX&gt;&quot;</code> and
<code>&quot;&lt;SUFFIX&gt;&quot;</code>
each project would get a <u>unique</u> root file system tree. </p>
<p>The disadvantage of this approach is that a new toolchain was
built for each project, adding considerable time to the build
process, even if it was two projects for the same chip. </p>
<p>This drawback has been somewhat lessened with
<code>gcc-4.x.y</code> which allows buildroot to use an external
toolchain. Certain packages requires special
features in the toolchain, and if an external toolchain is selected,
this may lack the neccessary features to complete the build of the root
file system.</p>
<p>A bigger problem was that the
<code>&quot;build_&lt;ARCH&gt;&quot;</code> tree
was also duplicated, so each </code>package</code> would also
be rebuilt once per project, resulting in even longer build times.</p>
<p><b>PROJECT TO SHARE TOOLCHAIN AND PACKAGE BUILDS</b></p>
<p>Work has started on a project which will allow the user to build
multiple root file systems for the same architecture in the same tree.
The toolchain and the package build directory will be shared, but each
project will have a dedicated directory tree for project specific
builds. </p>
<p>With this approach, most, if not all packages will be compiled
when the first project is built.
The process is almost identical to the original process.
Packages are downloaded and extracted to the shared
<code>&quot;build_&lt;ARCH&gt;/&lt;package&gt;&quot;</code>
directory. They are configured and compiled. </p>
<p>Package libraries and headers are installed in the shared $(STAGING_DIR),
and then the project specific root file system &quot;$(TARGET_DIR)&quot;
is populated. </p>
<p>At the end of the build, the root file system will be used
to generate the resulting root file system binaries. </p>
<p>Once the first project has been built, building other projects will
typically involve populating the new project's root file system directory
from the existing binaries generated in the shared
<code>&quot;build_&lt;ARCH&gt;/&lt;&gt;&quot;</code> directory. </p>
<p>Only packages, not used by the first project, will have to go
through the normal extract-configure-compile flow. </p>
<p><b>IMPLEMENTATION</b></p>
<p>The core of the solution is the introduction
of two new directories: </p>
<ul>
<li><code>project_build_&lt;ARCH&gt;</code></li>
<li><code>binaries;</code></li>
</ul>
<p>Each of the directories contain one subdirectory per project.
The name of the subdirectory is configured by the user in the
normal buildroot configuration, using the value of: </p>
<p><code>Project Options ---> Project name</code></p>
<p>The configuration defines the $(PROJECT) variable.</p>
<p>The default project name is <code>&quot;uclibc&quot;</code>.</p>
<p><code>&quot;package/Makefile.in&quot;</code> defines:
<pre>
<code>PROJECT_BUILD_DIR:=project_build_$(ARCH)/$(PROJECT)</code>
<code>BINARIES_DIR:=binaries/$(PROJECT)</code>
</pre>
</p>
<p>It also defines the location for the target root file system:
<pre>
<code>TARGET_DIR:=$(PROJECT_BUILD_DIR)/$(PROJECT)/root</code>
</pre>
</p>
<p>I.E: If the user has choosen
<code>&quot;myproject&quot;</code>
as the $(PROJECT) name:
<ul>
<li><code>&quot;project_build_&lt;ARCH&gt;/myproject&quot;</code></li>
<li><code>&quot;binaries/myproject&quot;</code></li>
</ul>
<p>will be created. </p>
<p>Currently, the <u>root file system</u>, <u>busybox</u> and an Atmel
customized version of
<u><code>U-Boot</code></u>, as well as some Atmel specific
bootloaders like <u>at91-bootstrap</u> and <u>dataflashboot.bin</u>
are built in
<code>&quot;$(PROJECT_BUILD_DIR)&quot;</code>
<p>The resulting binaries for all architectures are stored in the
<code>&quot;$(BINARIES_DIR)&quot;</code> directory. <p>
<p><b>SUMMARY</b></p>
<p>The project will share directories which can be share without
conflicts, but will use unique build directories, where the user
can configure the build. </p>
<p><b>THINGS TO DO</b></p>
<ol>
<li>Linux</li>
<p>The current Linux implementation is flawed. It only works
if the user chooses to use one of the few kernels selected
as base for the kernel-headers. While the Makefile seems to have
hooks, allowing the developer to specify whatever version he/she
wants in the target/device/*/* Makefiles, the build will fail
if another kernel version is choosen.</p>
<p>The reason for this is that the kernel patches are not
applied by the <code>&quot;target/linux/linux.mk&quot;</code>
build script fragment. They are only applied by the
<code>&quot;toolchain/kernel-headers/*.makefile&quot;</code>
build script fragments</p>
<p>If the kernel-header version and the linux version differs,
there will be two <code>&quot;linux-2.6.X.Y&quot;</code>
directories in
<code>&quot;build_&lt;ARCH&gt;/&lt;&gt;&quot;</code>,
each with its own set of patches. </p>
<p>The solution in the works, is to move the build of Linux to
<code>&quot;project_build_&lt;ARCH&gt;/&lt;project name&gt;/linux-2.6.X.Y&quot;</code> combined with method to configure
which patches can be applied. Possibly, the linux source tree
used to generate the kernel headers will be moved to the
<code>&quot;toolchain_build_&lt;ARCH&gt;&quot;</code>
directory
</p>
<p>The user will be able to select from three different
Linux strategies:
<ul>
<li>Conservative Strategy: Only use version ssupported by the kernel headers</li>
<li>Stable Linux Strategy: Allow any 2.6.X.Y combination.
(Minimum 2.6.19)</li>
<li>Power-User Strategy: Allow
<code>&quot;-git&quot;</code>, or
<code>&quot;-mm&quot;</code>, or user downloadable kernels</li>
</ul>
<p>The current kernel patches can be configured to be applied to the
linux source tree even if the version differs from the
kernel header version. </p>
<p>Since the user can select any kernel-patch
he/she will be able to select a non-working combination.
If the patch fails, the user will have to generate a new
proprietary kernel-patch or decide to not apply the kernel
patches</p>
<p>Other optional patches will be <u>board specific</u> or
<u>architecture specific</u> patches. </p>
<p>There will also be a way for the user to supply absolute
or relative paths to patches, possibly outside the main tree.
This can be used to apply custom kernel-header-patches, if
the versions available in buildroot cannot be applied to the
specific linux version used</p>
<p>Maybe, there will also be a possibility to supply an
<code>&quot;URL&quot;</code> to a patch available on Internet. </p>
<li>Configurable packages</li>
<p>Many packages can, on top of the simple
&quot;enable/disable build&quot;,
be further configured using Kconfig.
Currently these packages will be compiled using the
configuration specified in the
<code>&quot;.config&quot;</code> file of the <u>first</u>
project demanding the build of the package.</p>
<p>If <u>another</u> project uses the same packages, but with
a different configuration,these packages will <u>not</u> be rebuilt,
and the root file system for the new project will be populated
with files from the build of the <u>first</u> project</p>
<p>If multiple project are built, and a specific package
needs two different configuration, then the user must
delete the package from the
<code>&quot;build_&lt;ARCH&gt;&quot;</code> directory
before rebuilding the new project.<p>
<p>A long term solution is to edit the package makefile and move
the build of the configurable packages from
<code>&quot;build_&lt;ARCH&gt;&quot;</code> to
<code>&quot;project_build_&lt;ARCH&gt;/&lt;project name&gt;&quot;</code>
and send a patch to the buildroot mailing list.
<li>Naming conventions</li>
<p>Names of resulting binaries should reflect the
&quot;project name&quot;
<li>Generating File System binaries</li>
<p>
Packages which needs to be installed with the &quot;root&quot;
as owner, will generate a
<code>&quot;.fakeroot.&lt;package&gt;&quot;</code> file
which will be used for the final build of the root file system binary. </p>
<p>This was previously located in the
<code>&quot;$(STAGING_DIR)&quot;</code> directory, but was
recently moved to the
<code>&quot;$(PROJECT_BUILD_DIR)&quot;</code> directory. </p>
<p>Currently only three packages:
<code>&quot;at&quot;</code>,
<code>&quot;ltp-testsuite&quot;</code> and
<code>&quot;nfs-utils&quot;</code>
requests fakeroot. <p>
<p>The makefile fragments for each file system type like
<code>&quot;ext2&quot;</code>,
<code>&quot;jffs2&quot;</code> or
<code>&quot;squashfs&quot;</code>
will, when the file system binary is generated,
collect all present
<code>&quot;.fakeroot.&lt;package&gt;&quot;</code> files
to a single <code>&quot;_fakeroot.&lt;file system&gt;&quot;</code>
file and call fakeroot.</p>
<code>&quot;.fakeroot.&lt;package&gt;&quot;</code>
files are deleted as the last action of the Buildroot Makefile. </p>
<p>It needs to be evaluated if any further action for the
file system binary build is needed. </p>
</ol>
<h2><a name="using_toolchain" id="using_toolchain"></a>Using the
uClibc toolchain</h2>
<p>You may want to compile your own programs or other software
that are not packaged in Buildroot. In order to do this, you can
use the toolchain that was generated by Buildroot. </p>
<p>The toolchain generated by Buildroot by default is located in
<code>build_ARCH/staging_dir/</code>. The simplest way to use it
is to add <code>build_ARCH/staging_dir/usr/bin/</code> to your PATH
environnement variable, and then to use
<code>arch-linux-gcc</code>, <code>arch-linux-objdump</code>,
<code>arch-linux-ld</code>, etc. </p>
<p>For example, you may add the following to your
<code>.bashrc</code> (considering you're building for the MIPS
architecture and that Buildroot is located in
<code>~/buildroot/</code>) :</p>
<pre>
export PATH=&quot;$PATH:~/buildroot/build_mips/staging_dir/usr/bin/&quot;
</pre>
<p>Then you can simply do :</p>
<pre>
mips-linux-gcc -o foo foo.c
</pre>
<p><b>Important</b> : do not try to move a gcc-3.x toolchain to an other
directory, it won't work. There are some hardcoded paths in the
<i>gcc</i> configuration. If the default toolchain directory
doesn't suit your needs, please refer to the <a
href="#toolchain_standalone">Using the uClibc toolchain outside of
buildroot</a> section. </p>
<p>If you are using a current gcc-4.x, then use --sysroot and -isysroot
since these toolchains have fully functional sysroot support. No
hardcoded paths do exist in these configurations. </p>
<h2><a name="toolchain_standalone" id="toolchain_standalone"></a>Using the
uClibc toolchain outside of buildroot</h2>
<p>By default, the cross-compilation toolchain is generated inside
<code>build_ARCH/staging_dir/</code>. But sometimes, it may be useful to
install it somewhere else, so that it can be used to compile other programs
or by other users. Moving the <code>build_ARCH/staging_dir/</code>
directory elsewhere is <b>not possible if using gcc-3.x</b>, because there
are some hardcoded paths in the toolchain configuration. This works, thanks
to sysroot support, with current, stable gcc-4.x toolchains, of course. </p>
<p>If you want to use the generated gcc-3.x toolchain for other purposes,
you can configure Buildroot to generate it elsewhere using the
option of the configuration tool : <code>Build options -&gt;
Toolchain and header file location</code>, which defaults to
<code>$(BUILD_DIR)/staging_dir/</code>. </p>
<h2><a name="downloaded_packages"
id="downloaded_packages"></a>Location of downloaded packages</h2>
<p>It might be useful to know that the various tarballs that are
downloaded by the <i>Makefiles</i> are all stored in the
<code>DL_DIR</code> which by default is the <code>dl</code>
directory. It's useful for example if you want to keep a complete
version of Buildroot which is know to be working with the
associated tarballs. This will allow you to regenerate the
toolchain and the target filesystem with exactly the same
versions. </p>
<p>If you maintain several buildroot trees, it might be better to have
a shared download location. This can be accessed by creating a symbolic link
from the <code>dl</code> directory to the shared download location. </p>
<p>I.E:</p>
<pre>
ln -s &lt;shared download location&gt; dl
</pre>
<p>Another way of accessing a shared download location is to
create the <code>BUILDROOT_DL_DIR</code> environment variable.
If this is set, then the value of DL_DIR in the project is
overridden. The following line should be added to
<code>&quot;~/.bashrc&quot;</code>. <p>
<pre>
export BUILDROOT_DL_DIR &lt;shared download location&gt;
</pre>
<h2><a name="add_software" id="add_software"></a>Extending Buildroot with
more software</h2>
<p>This section will only consider the case in which you want to
add user-space software. </p>
<h3>Package directory</h3>
<p>First of all, create a directory under the <code>package</code>
directory for your software, for example <code>foo</code>. </p>
<h3><code>Config.in</code> file</h3>
<p>Then, create a file named <code>Config.in</code>. This file
will contain the portion of options description related to our
<code>foo</code> software that will be used and displayed in the
configuration tool. It should basically contain :</p>
<pre>
config BR2_PACKAGE_FOO
bool "foo"
default n
help
This is a comment that explains what foo is.
http://foosoftware.org/foo/
</pre>
<p>Of course, you can add other options to configure particular
things in your software. </p>
<h3>The real <i>Makefile</i></h3>
<p>Finally, here's the hardest part. Create a file named
<code>foo.mk</code>. It will contain the <i>Makefile</i> rules that
are in charge of downloading, configuring, compiling and installing
the software. Below is an example that we will comment
afterwards. </p>
<pre>
<a name="line1" id="line1">1</a> #############################################################
<a name="line2" id="line2">2</a> #
<a name="line3" id="line3">3</a> # foo
<a name="line4" id="line4">4</a> #
<a name="line5" id="line5">5</a> #############################################################
<a name="line6" id="line6">6</a> FOO_VERSION:=1.0
<a name="line7" id="line7">7</a> FOO_SOURCE:=foo-$(FOO_VERSION).tar.gz
<a name="line8" id="line8">8</a> FOO_SITE:=http://www.foosoftware.org/downloads
<a name="line9" id="line9">9</a> FOO_DIR:=$(BUILD_DIR)/foo-$(FOO_VERSION)
<a name="line10" id="line10">10</a> FOO_BINARY:=foo
<a name="line11" id="line11">11</a> FOO_TARGET_BINARY:=usr/bin/foo
<a name="line12" id="line12">12</a>
<a name="line13" id="line13">13</a> $(DL_DIR)/$(FOO_SOURCE):
<a name="line14" id="line14">14</a> $(WGET) -P $(DL_DIR) $(FOO_SITE)/$(FOO_SOURCE)
<a name="line15" id="line15">15</a>
<a name="line16" id="line16">16</a> $(FOO_DIR)/.source: $(DL_DIR)/$(FOO_SOURCE)
<a name="line17" id="line17">17</a> $(ZCAT) $(DL_DIR)/$(FOO_SOURCE) | tar -C $(BUILD_DIR) $(TAR_OPTIONS) -
<a name="line18" id="line18">18</a> touch $@
<a name="line19" id="line19">19</a>
<a name="line20" id="line20">20</a> $(FOO_DIR)/.configured: $(FOO_DIR)/.source
<a name="line21" id="line21">21</a> (cd $(FOO_DIR); rm -rf config.cache; \
<a name="line22" id="line22">22</a> $(TARGET_CONFIGURE_OPTS) \
<a name="line23" id="line23">23</a> $(TARGET_CONFIGURE_ARGS) \
<a name="line24" id="line24">24</a> ./configure \
<a name="line25" id="line25">25</a> --target=$(GNU_TARGET_NAME) \
<a name="line26" id="line26">26</a> --host=$(GNU_TARGET_NAME) \
<a name="line27" id="line27">27</a> --build=$(GNU_HOST_NAME) \
<a name="line28" id="line28">28</a> --prefix=/usr \
<a name="line29" id="line29">29</a> --sysconfdir=/etc \
<a name="line30" id="line30">30</a> )
<a name="line31" id="line31">31</a> touch $@
<a name="line32" id="line32">32</a>
<a name="line33" id="line33">33</a> $(FOO_DIR)/$(FOO_BINARY): $(FOO_DIR)/.configured
<a name="line34" id="line34">34</a> $(MAKE) CC=$(TARGET_CC) -C $(FOO_DIR)
<a name="line35" id="line35">35</a>
<a name="line36" id="line36">36</a> $(TARGET_DIR)/$(FOO_TARGET_BINARY): $(FOO_DIR)/$(FOO_BINARY)
<a name="line37" id="line37">37</a> $(MAKE) prefix=$(TARGET_DIR)/usr -C $(FOO_DIR) install
<a name="line38" id="line38">38</a> rm -Rf $(TARGET_DIR)/usr/man
<a name="line39" id="line39">39</a>
<a name="line40" id="line40">40</a> foo: uclibc ncurses $(TARGET_DIR)/$(FOO_TARGET_BINARY)
<a name="line41" id="line41">41</a>
<a name="line42" id="line42">42</a> foo-source: $(DL_DIR)/$(FOO_SOURCE)
<a name="line43" id="line43">43</a>
<a name="line44" id="line44">44</a> foo-clean:
<a name="line45" id="line45">45</a> $(MAKE) prefix=$(TARGET_DIR)/usr -C $(FOO_DIR) uninstall
<a name="line46" id="line46">46</a> -$(MAKE) -C $(FOO_DIR) clean
<a name="line47" id="line47">47</a>
<a name="line48" id="line48">48</a> foo-dirclean:
<a name="line49" id="line49">49</a> rm -rf $(FOO_DIR)
<a name="line50" id="line50">50</a>
<a name="line51" id="line51">51</a> #############################################################
<a name="line52" id="line52">52</a> #
<a name="line53" id="line53">53</a> # Toplevel Makefile options
<a name="line54" id="line54">54</a> #
<a name="line55" id="line55">55</a> #############################################################
<a name="line56" id="line56">56</a> ifeq ($(strip $(BR2_PACKAGE_FOO)),y)
<a name="line57" id="line57">57</a> TARGETS+=foo
<a name="line58" id="line58">58</a> endif
</pre>
<p>First of all, this <i>Makefile</i> example works for a single
binary software. For other software such as libraries or more
complex stuff with multiple binaries, it should be adapted. Look at
the other <code>*.mk</code> files in the <code>package</code>
directory. </p>
<p>At lines <a href="#line6">6-11</a>, a couple of useful variables are
defined :</p>
<ul>
<li><code>FOO_VERSION</code> : The version of <i>foo</i> that
should be downloaded. </li>
<li><code>FOO_SOURCE</code> : The name of the tarball of
<i>foo</i> on the download website of FTP site. As you can see
<code>FOO_VERSION</code> is used. </li>
<li><code>FOO_SITE</code> : The HTTP or FTP site from which
<i>foo</i> archive is downloaded. It must include the complete
path to the directory where <code>FOO_SOURCE</code> can be
found. </li>
<li><code>FOO_DIR</code> : The directory into which the software
will be configured and compiled. Basically, it's a subdirectory
of <code>BUILD_DIR</code> which is created upon decompression of
the tarball. </li>
<li><code>FOO_BINARY</code> : Software binary name. As said
previously, this is an example for a single binary software. </li>
<li><code>FOO_TARGET_BINARY</code> : The full path of the binary
inside the target filesystem. </li>
</ul>
<p>Lines <a href="#line13">13-14</a> defines a target that downloads the
tarball from the remote site to the download directory
(<code>DL_DIR</code>). </p>
<p>Lines <a href="#line16">16-18</a> defines a target and associated rules
that uncompress the downloaded tarball. As you can see, this target
depends on the tarball file, so that the previous target (line
<a href="#line13">13-14</a>) is called before executing the rules of the
current target. Uncompressing is followed by <i>touching</i> a hidden file
to mark the software has having been uncompressed. This trick is
used everywhere in Buildroot <i>Makefile</i> to split steps
(download, uncompress, configure, compile, install) while still
having correct dependencies. </p>
<p>Lines <a href="#line20">20-31</a> defines a target and associated rules
that configures the software. It depends on the previous target (the
hidden <code>.source</code> file) so that we are sure the software has
been uncompressed. In order to configure it, it basically runs the
well-known <code>./configure</code> script. As we may be doing
cross-compilation, <code>target</code>, <code>host</code> and
<code>build</code> arguments are given. The prefix is also set to
<code>/usr</code>, not because the software will be installed in
<code>/usr</code> on your host system, but in the target
filesystem. Finally it creates a <code>.configured</code> file to
mark the software as configured. </p>
<p>Lines <a href="#line33">33-34</a> defines a target and a rule that
compiles the software. This target will create the binary file in the
compilation directory, and depends on the software being already
configured (hence the reference to the <code>.configured</code>
file). It basically runs <code>make</code> inside the source
directory. </p>
<p>Lines <a href="#line36">36-38</a> defines a target and associated rules
that install the software inside the target filesystem. It depends on the
binary file in the source directory, to make sure the software has
been compiled. It uses the <code>install</code> target of the
software <code>Makefile</code> by passing a <code>prefix</code>
argument, so that the <code>Makefile</code> doesn't try to install
the software inside host <code>/usr</code> but inside target
<code>/usr</code>. After the installation, the
<code>/usr/man</code> directory inside the target filesystem is
removed to save space. </p>
<p>Line <a href="#line40">40</a> defines the main target of the software,
the one that will be eventually be used by the top level
<code>Makefile</code> to download, compile, and then install
this package. This target should first of all depends on all
needed dependecies of the software (in our example,
<i>uclibc</i> and <i>ncurses</i>), and also depend on the
final binary. This last dependency will call all previous
dependencies in the correct order. </p>
<p>Line <a href="#line42">42</a> defines a simple target that only
downloads the code source. This is not used during normal operation of
Buildroot, but is needed if you intend to download all required sources at
once for later offline build. Note that if you add a new package providing
a <code>foo-source</code> target is <i>mandatory</i> to support
users that wish to do offline-builds. Furthermore it eases checking
if all package-sources are downloadable. </p>
<p>Lines <a href="#line44">44-46</a> define a simple target to clean the
software build by calling the <i>Makefiles</i> with the appropriate option.
The <code>-clean</code> target should run <code>make clean</code>
on $(BUILD_DIR)/package-version and MUST uninstall all files of the
package from $(STAGING_DIR) and from $(TARGET_DIR). </p>
<p>Lines <a href="#line48">48-49</a> define a simple target to completely
remove the directory in which the software was uncompressed, configured and
compiled. The <code>-dirclean</code> target MUST completely rm $(BUILD_DIR)/
package-version. </p>
<p>Lines <a href="#line51">51-58</a> adds the target <code>foo</code> to
the list of targets to be compiled by Buildroot by first checking if
the configuration option for this package has been enabled
using the configuration tool, and if so then &quot;subscribes&quot;
this package to be compiled by adding it to the TARGETS
global variable. The name added to the TARGETS global
variable is the name of this package's target, as defined on
line <a href="#line40">40</a>, which is used by Buildroot to download,
compile, and then install this package. </p>
<h3>Conclusion</h3>
<p>As you can see, adding a software to buildroot is simply a
matter of writing a <i>Makefile</i> using an already existing
example and to modify it according to the compilation process of
the software. </p>
<p>If you package software that might be useful for other persons,
don't forget to send a patch to Buildroot developers !</p>
<h2><a name="links" id="links"></a>Resources</h2>
<p>To learn more about Buildroot you can visit these
websites:</p>
<ul>
<li><a href="http://www.uclibc.org/">http://www.uclibc.org/</a></li>
<li><a href="http://www.busybox.net/">http://www.busybox.net/</a></li>
</ul>
</div>
<!--
<a href="http://validator.w3.org/check?uri=referer"><img
border="0" height="31" width="88"
src="images/valid-html401.png"
alt="Valid HTML"></img></a>
-->
</body>
</html>