<?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>OpenWrt 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>OpenWrt Buildroot</h1> </div> <p>Usage and documentation by Felix Fietkau, based on uClibc Buildroot documentation by Thomas Petazzoni. Contributions from Karsten Kruse, Ned Ludd, Martin Herren.</p> <p><small>Last modification : $Id$</small></p> <ul> <li><a href="#about">About OpenWrt Buildroot</a></li> <li><a href="#download">Obtaining OpenWrt Buildroot</a></li> <li><a href="#using">Using OpenWrt 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 OpenWrt 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 OpenWrt with more Software</a></li> <li><a href="#links">Ressources</a></li> </ul> <h2><a name="about" id="about"></a>About OpenWrt Buildroot</h2> <p>OpenWrt Buildroot is a set of Makefiles and patches that allows to easily generate both a cross-compilation toolchain and a root filesystem for your Wireless Router. The cross-compilation toolchain uses uClibc (<a href= "http://www.uclibc.org/">http://www.uclibc.org/</a>), a tiny C standard library.</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 "host compilation toolchain", and more generally, the machine on which it is running, and on which you're working is called the "host system". The compilation toolchain is provided by your distribution, and OpenWrt 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 MIPS, 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 MIPS.</p> <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. OpenWrt 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 the MIPS architecture of most Broadcom based Wireless Routers.</p> <h2><a name="download" id="download"></a>Obtaining OpenWrt Buildroot</h2> <p>OpenWrt Buildroot is currently available as experimental snapshots</p> <p>The latest snapshot is always available at <a href="http://openwrt.org/downloads/experimental/">http://openwrt.org/downloads/experimental/</a>, <h2><a name="using" id="using"></a>Using OpenWrt Buildroot</h2> <p>OpenWrt 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 run everything as a normal user. There is no need to be root to configure and use the 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>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 target firmware images and additional packages (depending on your selections in <code>make menuconfig</code>. All the target files can be found in the <code>bin/</code> subdirectory. You can compile firmware images containing two different filesystem types: <ul> <li>jffs2</li> <li>squashfs</li> </ul> <p><code>jffs2</code> contains a writable root filesystem, which will expand to the size of your flash image. Note that you if you use the generic firmware Image, you need to pick the right image for your Flash size, because of different eraseblock sizes.</p> <p><code>squashfs</code> contains a read-only root filesystem using a modified <code>squashfs</code> filesystem with LZMA compression. When booting it, you can create a writable second filesystem, which will contain your modifications to the root filesystem, including the packages you install. <h2><a name="custom_targetfs" id="custom_targetfs"></a>Customizing the target filesystem</h2> <p>There are two 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>build_ARCH/root/</code> where <code>ARCH</code> is the chosen target architecture, usually mipsel. You can simply make your changes here, and run make target_install afterwards, which will rebuild the target filesystem image. This method allows to do everything on the target filesystem, but if you decide to rebuild your toolchain, tools or packages, these changes will be lost.</li> <li>Customize the target filesystem skeleton, available under <code>target/default/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 remains even when you completely rebuild the cross-compilation toolchain and the tools.<br /> </ul> <h2><a name="custom_busybox" id="custom_busybox"></a>Customizing the Busybox configuration</h2> <p>Busybox is very configurable, and you may want to customize it. Its configuration is completely integrated into the main menuconfig system. You can find it under "OpenWrt Package Selection" => "Busybox Configuration"</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>Go into the directory <code>toolchain_build_ARCH/uClibc/</code> and run <code>make menuconfig</code>. The nice configuration assistant, similar to the one used in the Linux Kernel 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 the Buildroot configuration, and the latter is used if you have selected locale support.</li> <li>Run the compilation 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> <h2><a name="buildroot_innards" id="buildroot_innards"></a>How OpenWrt Buildroot works</h2> <p>As said above, OpenWrt 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 <code>Makefile</code>. Makefiles are split into three sections:</p> <ul> <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>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>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. Two types of filesystems are supported : jffs2 and squashfs. </ul> <p>Each directory contains at least 2 files :</p> <ul> <li><code>Makefile</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 build directory (<code>build_ARCH/</code> by default, where <code>ARCH</code> is your architecture). This is where all user-space tools while be compiled.</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/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>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 copies the skeleton available in <code>target/default/target_skeleton</code> and then removes useless <code>CVS/</code> directories.</li> <li>Call the <code>prepare</code>, <code>compile</code> and <code>install</code> targets for the subdirectories <code>toolchain</code>, <code>package</code> and <code>target</code></li> </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 OpenWrt. In order to do this, you can use the toolchain that was generated by the Buildroot.</p> <p>The toolchain generated by the 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/bin/</code> to your PATH environment 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=$PATH:~/buildroot/build_mipsel/staging_dir/bin/ </pre> <p>Then you can simply do :</p> <pre> mipsel-linux-uclibc-gcc -o foo foo.c </pre> <p><b>Important</b> : do not try to move the toolchain to an other directory, it won't work. There are some hard-coded paths in the <i>gcc</i> configuration. If the default toolchain directory doesn't suit your needs, please refer to the <a href="#toolchain_standalone">Using the uClibc toolchain outside of buildroot</a> section.</p> <h2><a name="toolchain_standalone" id="toolchain_standalone"></a>Using the uClibc toolchain outside of buildroot</h2> <p>By default, the cross-compilation toolchain is generated inside <code>build_ARCH/staging_dir/</code>. But sometimes, it may be useful to install it somewhere else, so that it can be used to compile other programs or by other users. Moving the <code>build_ARCH/staging_dir/</code> directory elsewhere is <b>not possible</b>, because they are some hardcoded paths in the toolchain configuration.</p> <p>If you want to use the generated toolchain for other purposes, you can configure Buildroot to generate it elsewhere using the option of the configuration tool : <code>Build options -> Toolchain and header file location</code>, which defaults to <code>$(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> <h2><a name="add_software" id="add_software"></a>Extending OpenWrt 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 tristate "foo" default n help This is a comment that explains what foo is. </pre> <p>Of course, you can add other options to configure particular things in your software.</p> <h3><code>Makefile</code> in the package directory</h3> <p>To add your package to the build process, you need to edit the Makefile in the <code>package/</code> directory. Locate the lines that look like the following:</p> <pre> package-$(BR2_PACKAGE_FOO) += foo </pre> <p>As you can see, this short line simply adds the target <code>foo</code> to the list of targets handled by OpenWrt Buildroot.</p> <p>In addition to the default dependencies, you make your package depend on another package (e.g. a library) by adding a line: <pre> foo-compile: bar-compile </pre> <h3>The <i>.control</i> file</h3> <p>Additionally, you need to create a control file which contains information about your package, readable by the <i>ipkg</i> package utility.</p> <p>The file looks like this</p> <pre> 1 Package: foo 2 Priority: optional 3 Section: net 4 Maintainer: Foo Software <foo@foosoftware.com> 5 Source: http://foosoftware.com 6 Description: Your Package Description </pre> <p>You can skip the usual <code>Version:</code> and <code>Architecture</code> fields, as they will be generated by the <code>make-ipkg-dir.sh</code> script called from your Makefile</p> <h3>The real <i>Makefile</i></h3> <p>Finally, here's the hardest part. Create a file named <code>Makefile</code>. It will contain the <i>Makefile</i> rules that are in charge of downloading, configuring, compiling and installing the software. Below is an example that we will comment afterwards.</p> <pre> 1 ############################################################# 2 # foo 3 ############################################################# 4 PKG_NAME:=foo 5 PKG_VERSION:=1.0 6 PKG_RELEASE:=1 7 PKG_SOURCE:=$(PKG_NAME)-$(PKG_VERSION).tar.gz 8 PKG_SITE:=http://www.foosoftware.org/downloads 9 PKG_DIR:=$(BUILD_DIR)/$(PKG_NAME)-$(PKG_VERSION) 10 PKG_IPK:=$(PACKAGE_DIR)/$(PKG_NAME)_$(PKG_VERSION)-$(PKG_RELEASE)_$(ARCH).ipk 11 PKG_IPK_DIR:=$(PKG_DIR)/ipkg 12 13 $(DL_DIR)/$(PKG_SOURCE): 14 $(WGET) -P $(DL_DIR) $(PKG_SITE)/$(PKG_SOURCE) 15 16 $(PKG_DIR)/.source: $(DL_DIR)/$(PKG_SOURCE) 17 zcat $(DL_DIR)/$(PKG_SOURCE) | tar -C $(BUILD_DIR) $(TAR_OPTIONS) - 18 touch $(PKG_DIR)/.source 19 20 $(PKG_DIR)/.configured: $(PKG_DIR)/.source 21 (cd $(PKG_DIR); \ 22 $(TARGET_CONFIGURE_OPTS) \ 23 CFLAGS="$(TARGET_CFLAGS)" \ 24 ./configure \ 25 --target=$(GNU_TARGET_NAME) \ 26 --host=$(GNU_TARGET_NAME) \ 27 --build=$(GNU_HOST_NAME) \ 28 --prefix=/usr \ 29 --sysconfdir=/etc \ 30 ); 31 touch $(PKG_DIR)/.configured; 32 33 $(PKG_DIR)/foo $(PKG_DIR)/.configured 34 $(MAKE) CC=$(TARGET_CC) -C $(PKG_DIR) 35 36 $(PKG_IPK): $(PKG_DIR)/$(PKG_BINARY) 37 $(SCRIPT_DIR)/make-ipkg-dir.sh $(PKG_IPK_DIR) $(PKG_NAME).control $(PKG_VERSION)-$(PKG_RELEASE) $(ARCH) 38 $(MAKE) prefix=$(PKG_IPK_DIR)/usr -C $(PKG_DIR) install 39 rm -Rf $(PKG_IPK_DIR)/usr/man 40 $(IPKG_BUILD) $(PKG_IPK_DIR) $(PACKAGE_DIR) 41 42 $(IPKG_STATE_DIR)/info/$(PKG_NAME).list: $(PKG_IPK) 43 $(IPKG) install $(PKG_IPK) 44 45 prepare: $(PKG_DIR)/.source 46 compile: $(PKG_IPK) 47 install: $(IPKG_STATE_DIR)/info/$(PKG_NAME).list 48 clean: 49 rm -rf $(PKG_DIR) 50 rm -f $(PKG_IPK) </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>Makefile</code> files in the <code>package</code> directory.</p> <p>At lines 4-11, a couple of useful variables are defined :</p> <ul> <li><code>PKG_NAME</code> : The package name, e.g. <i>foo</i>.</li> <li><code>PKG_VERSION</code> : The version of the package that should be downloaded.</li> <li><code>PKG_RELEASE</code> : The release number that will be appended to the version number of your <i>ipkg</i> package. <li><code>PKG_SOURCE</code> : The name of the tarball of your package on the download website of FTP site. As you can see <code>PKG_NAME</code> and <code>PKG_VERSION</code> are used.</li> <li><code>PKG_SITE</code> : The HTTP or FTP site from which the archive is downloaded. It must include the complete path to the directory where <code>FOO_SOURCE</code> can be found.</li> <li><code>PKG_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>PKG_IPK</code> : The resulting <i>ipkg</i> package </ul> <p>Lines 13-14 defines a target that downloads the tarball from the remote site to the download directory (<code>DL_DIR</code>).</p> <p>Lines 16-18 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 13-14) 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 20-31 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 33-34 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 36-40 defines a target and associated rules that create the <i>ipkg</i> package which can optionally be embedded into the resulting firmware image. It depends on the binary file in the source directory, to make sure the software has been compiled. It uses the make-ipkg-dir.sh script, which will create the ipkg build directory for your package, copy your control file into that directory and add version and architecture information. Then it calls 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. Finally <code>IPKG_BUILD</code> is called to create the package.</p> <p>Line 42 and 43 define the installation target of your package, which will embed the software into the target filesystem.</p> <p>Lines 45-50 define the main targets that the Makefile in the <code>package</code> dir calls. <ul> <li><code>prepare</code> : Download and unpack the source</li> <li><code>compile</code> : Compile the source and create the package</li> <li><code>install</code> : Embed the package into the target filesystem</li> <li><code>clean</code> : Remove all the files created by the build process</li> </ul></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 OpenWrt developers !</p> <h2><a name="links" id="links"></a>Resources</h2> <p>To learn more about OpenWrt Buildroot you can visit this website: <a href="http://openwrt.org/">http://openwrt.org/</a></p> </div> </body> </html>