path: root/documentation/poky-ref-manual/extendpoky.xml

<!DOCTYPE chapter PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
"http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">

<chapter id='extendpoky'>

<title>Extending Poky</title>
    <para>
        This chapter provides information about how to extend the functionality
        already present in Poky.
        The chapter also documents standard tasks such as adding new
        software packages, extending or customizing images or porting Poky to
        new hardware (adding a new machine). 
        Finally, the chapter contains advice about how to make changes to Poky to achieve the best results.
    </para>

    <section id='usingpoky-extend-addpkg'>
        <title>Adding a Package</title>
        <para>
            To add a package into Poky you need to write a recipe for it. 
            Writing a recipe means creating a <filename>.bb</filename> file that sets some
            variables.
            For information on variables that are useful for recipes and for information about recipe naming
            issues, see the <link linkend='ref-varlocality-recipe-required'>Recipe Variables - Required</link>
            appendix.
        </para>
        <para>
            Before writing a recipe from scratch it is often useful to check
            whether someone else has written one already. 
            OpenEmbedded is a good place to look as it has a wider scope and range of packages.
            Because Poky aims to be compatible with OpenEmbedded, most recipes should
            simply work in Poky.
        </para>
        <para>
            For new packages, the simplest way to add a recipe is to base it on a similar
            pre-existing recipe. 
            Following are some examples showing how to add standard types of packages:
        </para>

        <section id='usingpoky-extend-addpkg-singlec'>
            <title>Single .c File Package (Hello World!)</title>
            <para>
                Building an application from a single file that is stored locally (e.g. under 
                <filename>files/</filename>) requires a recipe that has the file listed in 
                the <glossterm><link linkend='var-SRC_URI'>SRC_URI</link></glossterm> variable. 
                Additionally, you need to manually write the "do_compile" and
                "do_install" tasks.
                The <glossterm><link linkend='var-S'>S</link></glossterm> variable defines the 
                directory containing the source code, which is set to <glossterm><link linkend='var-WORKDIR'>
                WORKDIR</link></glossterm> in this case - the directory BitBake uses for the build.
            </para>
            <programlisting>
DESCRIPTION = "Simple helloworld application"
SECTION = "examples"
LICENSE = "MIT"
PR = "r0"

SRC_URI = "file://helloworld.c"

S = "${WORKDIR}"

do_compile() {
	${CC} helloworld.c -o helloworld
}

do_install() {
	install -d ${D}${bindir}
	install -m 0755 helloworld ${D}${bindir}
}
            </programlisting>
            <para>
                By default, the "helloworld", "helloworld-dbg" and "helloworld-dev"
                packages are built. 
                For information on how to customize the packaging process, see
                <link linkend='usingpoky-extend-addpkg-files'>Controlling Package Content</link>.
            </para>
        </section>

        <section id='usingpoky-extend-addpkg-autotools'>
            <title>Autotooled Package</title>
            <para>
                Applications that use autotools such as <filename>autoconf</filename> and 
                <filename>automake</filename> require a recipe that has a source archive listed in 
                <glossterm><link linkend='var-SRC_URI'>SRC_URI</link></glossterm> and 
                also inherits autotools, which instructs BitBake to use the
                <filename>autotools.bbclass</filename> file, which contains the definitions of all the steps
                needed to build an autotooled application.
                The result of the build is automatically packaged. 
                And, if the application uses NLS for localization, packages with local information are 
                generated (one package per language). 
                Following is one example: (<filename>hello_2.2.bb</filename>)
            </para>
            <programlisting>
DESCRIPTION = "GNU Helloworld application"
SECTION = "examples"
LICENSE = "GPLv2+"
LIC_FILES_CHKSUM = "file://COPYING;md5=751419260aa954499f7abaabaa882bbe"
PR = "r0"

SRC_URI = "${GNU_MIRROR}/hello/hello-${PV}.tar.gz"

inherit autotools gettext
            </programlisting>
            <para>
                The variable <glossterm><link linkend='var-LIC_FILES_CHKSUM'>LIC_FILES_CHKSUM</link>
                </glossterm> is used to <link linkend='usingpoky-configuring-LIC_FILES_CHKSUM'>
                track source license changes</link>. 
                You can quickly create autotool-based recipes in a manner similar to the previous example.
            </para>

        </section>

        <section id='usingpoky-extend-addpkg-makefile'>
            <title>Makefile-Based Package</title>
            <para>
                Applications that use GNU <filename>make</filename> also require a recipe that has
                the source archive listed in <glossterm><link linkend='var-SRC_URI'>SRC_URI</link></glossterm>. 
                You do not need to add a "do_compile" step since by default BitBake 
                starts the <filename>make</filename> command to compile the application. 
                If you need additional <filename>make</filename> options you should store them in the 
                <glossterm><link linkend='var-EXTRA_OEMAKE'>EXTRA_OEMAKE</link></glossterm> variable.
                BitBake passes these options into the <filename>make</filename> GNU invocation. 
                Note that a "do_install" task is still required.
                Otherwise BitBake runs an empty "do_install" task by default. 
            </para>
            <para>
                Some applications might require extra parameters to be passed to the compiler.
                For example the application might need an additional header path. 
                You can accomplish this by adding to the <glossterm><link linkend='var-CFLAGS'>CFLAGS</link>
                </glossterm> variable.
                The following example shows this:
            </para>
            <programlisting>
CFLAGS_prepend = "-I ${S}/include "
            </programlisting>
            <para>
            In the following example <filename>mtd-utils</filename> is a makefile-based package:
            </para>
            <programlisting>
DESCRIPTION = "Tools for managing memory technology devices."
SECTION = "base"
DEPENDS = "zlib lzo e2fsprogs util-linux"
HOMEPAGE = "http://www.linux-mtd.infradead.org/"
LICENSE = "GPLv2"

SRC_URI = "git://git.infradead.org/mtd-utils.git;protocol=git;tag=v${PV}"

S = "${WORKDIR}/git/"

EXTRA_OEMAKE = "'CC=${CC}' 'CFLAGS=${CFLAGS} -I${S}/include -DWITHOUT_XATTR' \
                'BUILDDIR=${S}'"

do_install () {
        oe_runmake install DESTDIR=${D} SBINDIR=${sbindir} MANDIR=${mandir} \
                           INCLUDEDIR=${includedir}
        install -d ${D}${includedir}/mtd/
        for f in ${S}/include/mtd/*.h; do
                install -m 0644 $f ${D}${includedir}/mtd/
        done
}
            </programlisting>

        </section>

        <section id='usingpoky-extend-addpkg-files'>
            <title>Controlling Package Content</title>
            <para>                        
                You can use the variables <glossterm><link linkend='var-PACKAGES'>PACKAGES</link></glossterm> and 
                <glossterm><link linkend='var-FILES'>FILES</link></glossterm> to split an application into 
                multiple packages.
            </para>
            <para>
                Following is an example that uses the "libXpm" recipe (<filename>libxpm_3.5.7.bb</filename>).
                By default, the "libXpm" recipe generates a single package that contains the library along 
                with a few binaries.  
                You can modify the recipe to split the binaries into separate packages:
            </para>
            <programlisting>
require xorg-lib-common.inc

DESCRIPTION = "X11 Pixmap library"
LICENSE = "X-BSD"
DEPENDS += "libxext libsm libxt"
PR = "r3"
PE = "1"

XORG_PN = "libXpm"

PACKAGES =+ "sxpm cxpm"
FILES_cxpm = "${bindir}/cxpm"
FILES_sxpm = "${bindir}/sxpm"
            </programlisting>
            <para>
                In the previous example we want to ship the "sxpm" and "cxpm" binaries 
                in separate packages. 
                Since "bindir" would be packaged into the main 
                <glossterm><link linkend='var-PN'>PN</link></glossterm> 
                package by default, we prepend the <glossterm><link linkend='var-PACKAGES'>PACKAGES</link>
                </glossterm> variable so additional package names are added to the start of list. 
                This results in the extra <glossterm><link linkend='var-FILES'>FILES</link></glossterm>_*
                variables then containing information that define which files and
                directories go into which packages. 
                Files included by earlier packages are skipped by latter packages.
                Thus, the main <glossterm><link linkend='var-PN'>PN</link></glossterm> package does not include 
                the above listed files.
            </para>
        </section>

        <section id='usingpoky-extend-addpkg-postinstalls'>
            <title>Post Install Scripts</title>

            <para>
                To add a post-installation script to a package, add a <filename>pkg_postinst_PACKAGENAME()
                </filename> function to the <filename>.bb</filename> file and use 
                <filename>PACKAGENAME</filename> as the name of the package you want to attach to the 
                <filename>postinst</filename> script.
                Normally <glossterm><link linkend='var-PN'>PN</link></glossterm> can be used, which 
                automatically expands to PACKAGENAME.
                A post-installation function has the following structure:
            </para>
            <programlisting>
pkg_postinst_PACKAGENAME () {
#!/bin/sh -e
# Commands to carry out
}
            </programlisting>
            <para>
                The script defined in the post-installation function is called when the rootfs is made. 
                If the script succeeds, the package is marked as installed. 
                If the script fails, the package is marked as unpacked and the script is
                executed when the image boots again.
            </para>
            <para>
                Sometimes it is necessary for the execution of a post-installation
                script to be delayed until the first boot.  
                For example, the script might need to be executed on the device itself. 
                To delay script execution until boot time, use the following structure in the 
                post-installation script:
            </para>
            <programlisting>
pkg_postinst_PACKAGENAME () {
#!/bin/sh -e
if [ x"$D" = "x" ]; then
	# Actions to carry out on the device go here
else
	exit 1
fi
}
            </programlisting>
            <para>
                The previous example delays execution until the image boots again because the 
                <glossterm><link linkend='var-D'>D</link></glossterm> variable points
                to the 'image' directory when the rootfs is being made at build time but
                is unset when executed on the first boot. 
            </para>
        </section>
    </section>

    <section id='usingpoky-extend-customimage'>
        <title>Customizing Images</title>
        <para>
            You can customize Poky images to satisfy particular requirements. 
            This section describes several methods and provides guidelines for each.
        </para>

        <section id='usingpoky-extend-customimage-custombb'>
            <title>Customizing Images Using Custom .bb Files</title>
            <para>
                One way to get additional software into an image is to create a custom image. 
                The following example shows the form for the two lines you need:
            </para>
            <programlisting>
IMAGE_INSTALL = "task-core-x11-base package1 package2"

inherit poky-image
            </programlisting>
            <para>
                By creating a custom image, a developer has total control
                over the contents of the image. 
                It is important to use the correct names of packages in the 
                <glossterm><link linkend='var-IMAGE_INSTALL'>IMAGE_INSTALL</link></glossterm> variable. 
                You must use the OpenEmbedded notation and not the Debian notation for the names 
                (e.g. "glibc-dev" instead of "libc6-dev").
            </para>
            <para>
                The other method for creating a custom image is to modify an existing image. 
                For example, if a developer wants to add "strace" into "core-image-sato", they can use 
                the following recipe:
            </para>
            <programlisting>
require core-image-sato.bb

IMAGE_INSTALL += "strace"
            </programlisting>
        </section>

        <section id='usingpoky-extend-customimage-customtasks'>
            <title>Customizing Images Using Custom Tasks</title>
            <para>
                For complex custom images, the best approach is to create a custom task package
                that is used to build the image or images. 
                A good example of a tasks package is 
                <filename>meta/recipes-sato/tasks/task-poky.bb</filename>. 
                The <glossterm><link linkend='var-PACKAGES'>PACKAGES</link></glossterm> 
                variable lists the task packages to build along with the complementary
                -dbg and -dev packages. 
                For each package added, you can use 
                <glossterm><link linkend='var-RDEPENDS'>RDEPENDS</link></glossterm>
                and <glossterm><link linkend='var-RRECOMMENDS'>RRECOMMENDS</link></glossterm> 
                entries to provide a list of packages the parent task package should contain. 
                Following is an example:
            </para>
            <para>
            <programlisting>
DESCRIPTION = "My Custom Tasks"

PACKAGES = "\
    task-custom-apps \
    task-custom-apps-dbg \
    task-custom-apps-dev \
    task-custom-tools \
    task-custom-tools-dbg \
    task-custom-tools-dev \
    "

RDEPENDS_task-custom-apps = "\
    dropbear \
    portmap \
    psplash"

RDEPENDS_task-custom-tools = "\
    oprofile \
    oprofileui-server \
    lttng-control \
    lttng-viewer"

RRECOMMENDS_task-custom-tools = "\
    kernel-module-oprofile"
            </programlisting>
            </para>
            <para>
                In the previous example, two task packages are created with their dependencies and their
                recommended package dependencies listed: <filename>task-custom-apps</filename>, and 
                <filename>task-custom-tools</filename>. 
                To build an image using these task packages, you need to add 
                "task-custom-apps" and/or "task-custom-tools" to <glossterm><link 
                linkend='var-IMAGE_INSTALL'>IMAGE_INSTALL</link></glossterm>.
                For other forms of image dependencies see the other areas of this section.
            </para>
        </section>

        <section id='usingpoky-extend-customimage-imagefeatures'>
            <title>Customizing Images Using Custom IMAGE_FEATURES</title>
            <para>
                Ultimately users might want to add extra image "features" as used by Poky with the 
                <glossterm><link linkend='var-IMAGE_FEATURES'>IMAGE_FEATURES</link></glossterm>
                variable. 
                To create these features, the best reference is 
                <filename>meta/classes/poky-image.bbclass</filename>, which shows how poky achieves this. 
                In summary, the file looks at the contents of the 
                <glossterm><link linkend='var-IMAGE_FEATURES'>IMAGE_FEATURES</link></glossterm>
                variable and then maps that into a set of tasks or packages. 
                Based on this information the <glossterm><link linkend='var-IMAGE_INSTALL'> IMAGE_INSTALL</link>
                </glossterm> variable is generated automatically. 
                Users can add extra features by extending the class or creating a custom class for use 
                with specialized image <filename>.bb</filename> files.
            </para>
            <para>
                Poky ships with two SSH servers you can use in your images: Dropbear and OpenSSH. 
                Dropbear is a minimal SSH server appropriate for resource-constrained environments,
                while OpenSSH is a well-known standard SSH server implementation.
                By default, core-image-sato is configured to use Dropbear.
                The core-image-basic and core-image-lsb images both include OpenSSH.
                To change these defaults, edit the <filename>IMAGE_FEATURES</filename> variable
                so that it sets the image you are working with to include ssh-server-dropbear 
                or ssh-server-openssh.
            </para>
        </section>

        <section id='usingpoky-extend-customimage-localconf'>
            <title>Customizing Images Using local.conf</title>
            <para>
                It is possible to customize image contents by using variables used by distribution 
                maintainers in the <filename>local.conf</filename>.
                This method only allows the addition of packages and is not recommended.
            </para>
            <para>
                For example, to add the "strace" package into the image you would add this package to the 
                <filename>local.conf</filename> file:
            </para>
            <programlisting>
DISTRO_EXTRA_RDEPENDS += "strace"
            </programlisting>
            <para>
                However, since the <glossterm><link linkend='var-DISTRO_EXTRA_RDEPENDS'>
                DISTRO_EXTRA_RDEPENDS</link></glossterm> variable is for
                distribution maintainers, adding packages using this method is not as simple as adding 
                them using a custom <filename>.bb</filename> file. 
                Using the <filename>local.conf</filename> file method could result in some packages 
                needing to be recreated.
                For example, if packages were previously created and the image was rebuilt then the packages
                would need to be recreated.
            </para>
            <para>
                Cleaning task-* packages are required because they use the
                <glossterm><link linkend='var-DISTRO_EXTRA_RDEPENDS'>
                DISTRO_EXTRA_RDEPENDS</link></glossterm> variable. 
                You do not have to build them by hand because Poky images depend on the packages they contain.
                This means dependencies are automatically built when the image builds. 
                For this reason we don't use the "rebuild" task.
                In this case the "rebuild" task does not care about
                dependencies - it only rebuilds the specified package.
            </para>
            <programlisting>
$ bitbake -c clean task-boot task-base task-poky
$ bitbake core-image-sato
            </programlisting>
        </section>

    </section>

    <section id="platdev-newmachine">
        <title>Porting Poky to a New Machine</title>
        <para>
            Adding a new machine to Poky is a straightforward process. 
            This section provides information that gives you an idea of the changes you must make.
            The information covers adding machines similar to those Poky already supports. 
            Although well within the capabilities of Poky, adding a totally new architecture might require 
            changes to <filename>gcc/glibc</filename> and to the site information, which is 
            beyond the scope of this manual.
        </para>

        <section id="platdev-newmachine-conffile">
            <title>Adding the Machine Configuration File</title>
            <para>
                To add a machine configuration you need to add a <filename>.conf</filename> file
                with details of the device being added to the <filename>conf/machine/</filename> file.
                The name of the file determines the name Poky uses to reference the new machine.
            </para>
            <para>
                The most important variables to set in this file are <glossterm>
                <link linkend='var-TARGET_ARCH'>TARGET_ARCH</link></glossterm> (e.g. "arm"), 
                <glossterm><link linkend='var-PREFERRED_PROVIDER'>PREFERRED_PROVIDER</link></glossterm>_virtual/kernel 
               (see below) and <glossterm><link linkend='var-MACHINE_FEATURES'>MACHINE_FEATURES</link></glossterm> 
               (e.g. "kernel26 apm screen wifi"). 
                You might also need other variables like <glossterm><link linkend='var-SERIAL_CONSOLE'>SERIAL_CONSOLE
                </link></glossterm> (e.g. "115200 ttyS0"), <glossterm> 
                <link linkend='var-KERNEL_IMAGETYPE'>KERNEL_IMAGETYPE</link>
                </glossterm> (e.g. "zImage") and <glossterm><link linkend='var-IMAGE_FSTYPES'>
                IMAGE_FSTYPES</link></glossterm> (e.g. "tar.gz jffs2"). 
                You can find full details on these variables in the reference section. 
                You can leverage many existing machine <filename>.conf</filename> files from 
                <filename>meta/conf/machine/</filename>.
            </para>
        </section>

        <section id="platdev-newmachine-kernel">
            <title>Adding a Kernel for the Machine</title>
            <para>
                Poky needs to be able to build a kernel for the machine. 
                You need to either create a new kernel recipe for this machine, or extend an 
                existing recipe. 
                You can find several kernel examples in the <filename>meta/recipes-kernel/linux</filename>
                directory that can be used as references.
            </para>
            <para>
                If you are creating a new recipe, the "normal" recipe-writing rules apply for setting 
                up a <glossterm><link linkend='var-SRC_URI'>SRC_URI</link></glossterm>. 
                This means specifying any necessary patches and setting <glossterm>
                <link linkend='var-S'>S</link></glossterm> to point at the source code. 
                You need to create a "configure" task that configures the unpacked kernel with a defconfig.
                You can do this by using a <filename>make defconfig</filename> command or
                more commonly by copying in a suitable <filename>defconfig</filename> file and and then running 
                <filename>make oldconfig</filename>. 
                By making use of "inherit kernel" and potentially some of the 
                <filename>linux-*.inc</filename> files, most other functionality is 
                centralized and the the defaults of the class normally work well.
            </para>
            <para>
                If you are extending an existing kernel, it is usually a matter of adding a 
                suitable <filename>defconfig</filename> file.
                The file needs to be added into a location similar to <filename>defconfig</filename> files
                used for other machines in a given kernel. 
                A possible way to do this is by listing the file in the 
                <glossterm><link linkend='var-SRC_URI'>SRC_URI</link></glossterm>
                and adding the machine to the expression in 
                <glossterm><link linkend='var-COMPATIBLE_MACHINE'>COMPATIBLE_MACHINE</link></glossterm>:
            </para>
            <programlisting>
COMPATIBLE_MACHINE = '(qemux86|qemumips)'
            </programlisting>
        </section>

        <section id="platdev-newmachine-formfactor">
            <title>Adding a Formfactor Configuration File</title>
            <para>
                A formfactor configuration file provides information about the 
                target hardware on which Poky is running and information that Poky cannot 
                obtain from other sources such as the kernel.  
                Some examples of information contained in a formfactor configuration file include 
                framebuffer orientation, whether or not the system has a keyboard, 
                the positioning of the keyboard in relation to the screen, and 
                the screen resolution.
            </para>
            <para>
                Reasonable defaults are used in most cases, but if customization is 
                necessary you need to create a <filename>machconfig</filename> file 
                under <filename>meta/packages/formfactor/files/MACHINENAME/</filename>,
                where <filename>MACHINENAME</filename> is the name for which this information
                applies. 
                For information about the settings available and the defaults, see 
                <filename>meta/recipes-bsp/formfactor/files/config</filename>. 
                Following is an example for qemuarm:
            </para>
            <programlisting>
HAVE_TOUCHSCREEN=1
HAVE_KEYBOARD=1

DISPLAY_CAN_ROTATE=0
DISPLAY_ORIENTATION=0
#DISPLAY_WIDTH_PIXELS=640
#DISPLAY_HEIGHT_PIXELS=480
#DISPLAY_BPP=16
DISPLAY_DPI=150
DISPLAY_SUBPIXEL_ORDER=vrgb
            </programlisting>
        </section>
    </section>

    <section id="usingpoky-changes">
        <title>Making and Maintaining Changes</title>
        <para>
            Because Poky is extremely configurable and flexible, we recognize that people will want 
            to extend, configure or optimize Poky for their specific uses.
            To best keep pace with future Poky changes we recommend you make controlled changes to Poky.
        </para>
        <para>
            Poky supports the idea of <link linkend='usingpoky-changes-layers'>"layers"</link>.
            If you use layers properly you can ease future upgrades and allow segregation
            between the Poky core and a given developer's changes. 
            The following section provides more advice on managing changes to Poky.
        </para>

        <section id="usingpoky-changes-layers">
          <title>BitBake Layers</title>
          <para>
                Often, people want to extend Poky either by adding packages
                or by overriding files contained within Poky to add their own
                functionality. 
                BitBake has a powerful mechanism called
                "layers", which provides a way to handle this extension in a fully
                supported and non-invasive fashion.
           </para>
           <para>
               The Poky tree includes several additional layers such as meta-emenlow and meta-extras
               that demonstrate this functionality.
               The meta-emenlow layer is an example layer that, by default, is enabled. 
               However, the meta-extras repository is not enabled by default. 
               It is easy though to enable any layer.
               You simply add the layer's path to the 
               <glossterm><link linkend='var-BBLAYERS'>BBLAYERS</link></glossterm> variable in your 
               <filename>bblayers.conf</filename> file. 
               The following example shows how to enable meta-extras in the Poky build:
          </para>
          <para>
               <programlisting>
LCONF_VERSION = "1"

BBFILES ?= ""
BBLAYERS = " \
  /path/to/poky/meta \
  /path/to/poky/meta-emenlow \
  /path/to/poky/meta-extras \
  "
               </programlisting>
           </para>

          <para>
               BitBake parses each <filename>conf/layer.conf</filename> file for each layer in BBLAYERS
               and adds the recipes, classes and configuration contained within the layer to Poky.
               To create your own layer, independent of the main Poky repository,
               simply create a directory with a <filename>conf/layer.conf</filename> file and
               add the directory to your <filename>bblayers.conf</filename> file.
          </para>
          <para>
               The <filename>meta-emenlow/conf/layer.conf</filename> file demonstrates the required syntax:
               <programlisting>
# We have a conf and classes directory, add to BBPATH
BBPATH := "${BBPATH}:${LAYERDIR}"

# We have a recipes directory containing both .bb and .bbappend files, add to BBFILES
BBFILES := "${BBFILES} ${LAYERDIR}/recipes/*/*.bb \
             ${LAYERDIR}/recipes/*/*.bbappend"

BBFILE_COLLECTIONS += "emenlow"
BBFILE_PATTERN_emenlow := "^${LAYERDIR}/"
BBFILE_PRIORITY_emenlow = "6"
               </programlisting>
          </para>
          <para>
                In the previous example, the recipes for the layers are added to 
                <glossterm><link linkend='var-BBFILES'>BBFILES</link></glossterm>. 
                The <glossterm><link linkend='var-BBFILE_COLLECTIONS'>BBFILE_COLLECTIONS</link></glossterm>
                variable is then appended with the layer name. 
                The <glossterm><link linkend='var-BBFILE_PATTERN'>BBFILE_PATTERN</link></glossterm> variable
                immediately expands with a regular expression used to match files from BBFILES into
                a particular layer, in this case by using the base pathname.
                The <glossterm><link linkend='var-BBFILE_PRIORITY'>BBFILE_PRIORITY</link></glossterm> variable 
                then assigns different priorities to the files in different layers. 
                Applying priorities is useful in situations where the same package might appear in multiple
                layers and allows you to choose what layer should take precedence.
            </para>
            <para>
                Note the use of the <glossterm><link linkend='var-LAYERDIR'>LAYERDIR</link></glossterm> 
                variable with the immediate expansion operator.
                The LAYERDIR variable expands to the directory of the current layer and
                requires the immediate expansion operator so that BitBake does not wait to expand the variable 
                when it's parsing a different directory.
            </para>
            <para>
                BitBake can locate where other bbclass and configuration files are applied through 
                the <glossterm><link linkend='var-BBPATH'>BBPATH</link></glossterm>
                environment variable. 
                For these cases, BitBake uses the first file with the matching name found in BBPATH.
                This is similar to the way the PATH variable is used for binaries. 
                We recommend, therefore, that you use unique bbclass and configuration file names in your
                custom layer.
            </para>
            <para>
                We also recommend the following:
                <itemizedlist>
                    <listitem><para>Store custom layers in a git repository that uses the 
                    meta-prvt-XXXX format.</para></listitem>
                    <listitem><para>Clone the repository alongside other meta directories in the Poky 
                    tree.</para></listitem>
                </itemizedlist>
                Following these recommendations keeps your Poky tree and its configuration entirely
                inside COREBASE.
            </para>
        </section>

        <section id="usingpoky-changes-commits">
            <title>Committing Changes</title>
            <para>
                Modifications to Poky are often managed under some kind of source
                revision control system. 
                Because some simple practices can significantly improve usability, policy for committing changes
                is important.
                It helps to use a consistent documentation style when committing changes. 
                We have found the following style works well. 
            </para>
            <para>
                Following are suggestions for committing changes to the Poky core:
            </para>

            <para>
                <itemizedlist>
                    <listitem><para>The first line of the commit summarizes the change and begins with the 
                    name of the affected package or packages.
                    However, not all changes apply to specific packages. 
                    Consequently, the prefix could also be a machine name or class name for 
                    example.</para></listitem>
                    <listitem><para>The second part of the commit (if needed) is a longer more detailed 
                    description of the changes.  Placing a blank line between the first and second parts
                    helps with readability.</para></listitem>
                </itemizedlist>
            </para>
            <para>
                Following is an example commit:
            </para>
            <literallayout class='monospaced'>
    bitbake/data.py: Add emit_func() and generate_dependencies() functions
    
    These functions allow generation of dependency data between functions and
    variables allowing moves to be made towards generating checksums and allowing
    use of the dependency information in other parts of bitbake.
    
    Signed-off-by: Richard Purdie richard.purdie@linuxfoundation.org
            </literallayout>

            <para>
                All commits should be self-contained such that they leave the 
                metadata in a consistent state that builds both before and after the 
                commit is made. 
                Besides being a good policy to follow, this helps ensure the autobuilder test results
                are valid.
            </para>
        </section>

        <section id="usingpoky-changes-prbump">
            <title>Package Revision Incrementing</title>
            <para>
                If a committed change results in changing the package output
                then the value of the <glossterm><link linkend='var-PR'>PR</link>
                </glossterm> variable needs to be increased (or 'bumped') as part of that commit. 
                This means that for new recipes you must be sure to add the PR variable and set its initial value
                equal to "r0".  
                Failing to define PR makes it easy to miss when you bump a package.
                Note that you can only use integer values following the "r" in the PR variable.
            </para>
            <para>
                If you are sharing a common .inc file with multiple recipes, you can also use the 
                <glossterm><link linkend='var-INC_PR'>INC_PR</link></glossterm> variable to ensure that 
                the recipes sharing the .inc file are rebuilt when the .inc file itself is changed. The
                .inc file must set INC_PR (initially to "r0"), and all recipes referring to it should set PR to 
                "$(INC_PR).0" initially, incrementing the last number when the recipe is changed. If the
                .inc file is changed then its INC_PR should be incremented.
            </para>
            <para> 
                When upgrading the version of a package, assuming the <glossterm><link 
                linkend='var-PV'>PV</link></glossterm> changes, the PR variable should be reset to "r0"
                (or "$(INC_PR).0" if you are using INC_PR).
            </para>
            <para>
                Usually, version increases occur only to packages.
                However, if for some reason PV changes but does not increase, you can increase the 
                <glossterm><link linkend='var-PE'>PE</link></glossterm> variable (Package Epoch).
                The PE variable defaults to "0".
            </para>
            <para>
                Version numbering strives to follow the 
                <ulink url='http://www.debian.org/doc/debian-policy/ch-controlfields.html'>
                Debian Version Field Policy Guidelines</ulink>.
                These guidelines define how versions are compared and what "increasing" a version means.
            </para>
            <para>
                There are two reasons for following these guidelines.
                First, to ensure that when a developer updates and rebuilds, they get all the changes to
                the repository and don't have to remember to rebuild any sections.
                Second, to ensure that target users are able to upgrade their
                devices using package manager commands such as <filename>opkg upgrade</filename> 
                (or similar commands for dpkg/apt or rpm-based systems). 
            </para>
            <para>
                The goal is to ensure Poky has upgradeable packages in all cases.
            </para>
        </section>

        <section id="usingpoky-changes-collaborate">
            <title>Using Poky in a Team Environment</title>
            <para>
                It might not be immediately clear how you can use Poky in a team environment, 
                or scale it for a large team of developers. 
                The specifics of any situation determine the best solution.
                Granted that Poky offers immense flexibility regarding this, practices do exist 
                that experience has shown work well.
            </para>
            <para>
                The core component of any development effort with Poky is often an 
                automated build testing framework and an image generation process. 
                You can use these core components to check that the metadata can be built, 
                highlight when commits break the build, and provide up-to-date images that 
                allow people to test the end result and use it as a base platform for further 
                development. 
                Experience shows that buildbot is a good fit for this role. 
                What works well is to configure buildbot to make two types of builds:
                incremental and full (from scratch).  
                See <ulink url='http://autobuilder.pokylinux.org:8010'>poky autobuilder</ulink>
                for an example implementation that uses buildbot.
            </para>
            <para>
                You can tie incremental builds to a commit hook that triggers the build
                each time a commit is made to the metadata.  
                This practice results in useful acid tests that determine whether a given commit 
                breaks the build in some serious way. 
                Associating a build to a commit can catch a lot of simple errors.
                Furthermore, the tests are fast so developers can get quick feedback on changes.
            </para>
            <para>
                Full builds build and test everything from the ground up. 
                They usually happen at predetermined times like during the night when the machine 
                load is low.
            </para>
            <para>
                Most teams have many pieces of software undergoing active development at any given time. 
                You can derive large benefits by putting these pieces under the control of a source 
                control system that is compatible with Poky (i.e. git or svn).
                You can then set the autobuilder to pull the latest revisions of the packages 
                and test the latest commits by the builds.
                This practice quickly highlights issues. 
                Poky easily supports testing configurations that use both a stable known good revision 
                and a floating revision.
                Poky can also take just the changes from specific source control branches.
                This capability allows you to track and test specific changes.
            </para>
            <para>
                Perhaps the hardest part of setting this up is defining the software project or 
                Poky metadata policies that surround the different source control systems.
                Of course circumstances will be different in each case.
                However, this situation reveals one of Poky's advantages - the system itself does not
                force any particular policy on users, unlike a lot of build systems. 
                The system allows the best policies to be chosen for the given circumstances.
            </para>
        </section>

        <section id="usingpoky-changes-updatingimages">
            <title>Updating Existing Images</title>
            <para>
                Often, rather than re-flashing a new image you might wish to install updated 
                packages into an existing running system. 
                You can do this by first sharing the 
                <filename class="directory">tmp/deploy/ipk/</filename> directory
                through a web server and then by changing <filename>/etc/opkg/base-feeds.conf</filename> 
                to point at the shared server.
                Following is an example:
            </para>
            <literallayout class='monospaced'>
     src/gz all http://www.mysite.com/somedir/deploy/ipk/all
     src/gz armv7a http://www.mysite.com/somedir/deploy/ipk/armv7a
     src/gz beagleboard http://www.mysite.com/somedir/deploy/ipk/beagleboard
            </literallayout>
        </section>
    </section>

    <section id="usingpoky-modifing-packages">
        <title>Modifying Package Source Code</title>
        <para>
            Although Poky is usually used to build software, you can use it to modify software. 
        </para>
        <para>
            During a build, source is available in the 
            <glossterm><link linkend='var-WORKDIR'>WORKDIR</link></glossterm> directory.
            The actual location depends on the type of package and the architecture of the target device. 
            For a standard recipe not related to 
            <glossterm><link linkend='var-MACHINE'>MACHINE</link></glossterm> the location is
            <filename>tmp/work/PACKAGE_ARCH-poky-TARGET_OS/PN-PV-PR/</filename>.
            For target device-dependent packages you should use the <filename>MACHINE</filename>
            variable instead of 
            <glossterm><link linkend='var-PACKAGE_ARCH'>PACKAGE_ARCH</link></glossterm> 
            in the directory name.
        </para>
        <tip>
            <para>
                Be sure the package recipe sets the 
                <glossterm><link linkend='var-S'>S</link></glossterm> variable to something
                other than the standard <filename>WORKDIR/PN-PV/</filename> value.
            </para>
        </tip>
        <para>
            After building a package, you can modify the package source code without problems. 
            The easiest way to test your changes is by calling the "compile" task as shown in the 
            following example:
        </para>

        <literallayout class='monospaced'>
     $ bitbake -c compile -f NAME_OF_PACKAGE
        </literallayout>

        <para>
            The "-f" or "--force" option forces re-execution of the specified task.
            You can call other tasks this way as well. 
            But note that all the modifications in 
            <glossterm><link linkend='var-WORKDIR'>WORKDIR</link></glossterm>
            are gone once you execute "-c clean" for a package.
        </para>

        <section id="usingpoky-modifying-packages-quilt">
            <title>Modifying Package Source Code with quilt</title>
            <para>
                By default Poky uses <ulink url='http://savannah.nongnu.org/projects/quilt'>quilt</ulink>
                to manage patches in the "do_patch" task. 
                This is a powerful tool that you can use to track all modifications to package sources.
            </para>
            <para>
                Before modifying source code, it is important to notify quilt so it can track the changes 
                into the new patch file:

                <literallayout class='monospaced'>
     quilt new NAME-OF-PATCH.patch
                </literallayout>

                After notifying quilt, add all modified files into that patch:
                <literallayout class='monospaced'>
     quilt add file1 file2 file3
                </literallayout>

                You can now start editing. 
                Once you are done editing, you need to use quilt to generate the final patch that 
                will contain all your modifications.
                <literallayout class='monospaced'>
     quilt refresh
                </literallayout>

                You can find the resulting patch file in the
                <filename>patches/</filename> subdirectory of the source 
                (<glossterm><link linkend='var-S'>S</link></glossterm>) directory. 
                For future builds you should copy the patch into Poky metadata and add it into the 
                <glossterm><link linkend='var-SRC_URI'>SRC_URI</link></glossterm> of a recipe.  
                Here is an example:
                <programlisting>
SRC_URI += "file://NAME-OF-PATCH.patch"
                </programlisting>
                Finally, don't forget to 'bump' the 
                <glossterm><link linkend='var-PR'>PR</link></glossterm> value in the same recipe since 
                the resulting packages have changed.
            </para>
        </section>

    </section>

    <section id="usingpoky-configuring-LIC_FILES_CHKSUM">
        <title>Track License Change</title>
        <para>
        The license of an upstream project might change in the future.
        Poky uses the <glossterm><link linkend='var-LIC_FILES_CHKSUM'>LIC_FILES_CHKSUM</link></glossterm> variable
        to track license changes.
        </para>

        <section id="usingpoky-specifying-LIC_FILES_CHKSUM">
            <title>Specifying the LIC_FILES_CHKSUM Variable</title>
            <para>
                The <glossterm><link linkend='var-LIC_FILES_CHKSUM'>LIC_FILES_CHKSUM</link></glossterm>
                variable contains checksums of the license text in the recipe source code. 
                Poky uses this to track changes in the license text of the source code files.
                Following is an example of LIC_FILES_CHKSUM:
            </para>
            <programlisting>
LIC_FILES_CHKSUM = "file://COPYING; md5=xxxx \
                    file://licfile1.txt; beginline=5; endline=29;md5=yyyy \
                    file://licfile2.txt; endline=50;md5=zzzz \
                    ..."
            </programlisting>
            <para>
                Poky uses the <glossterm><link linkend='var-S'>S</link></glossterm> variable as the 
                default directory used when searching files listed in LIC_FILES_CHKSUM.
                The previous example employs the default directory.
            </para>
            <para>
                You can also use relative paths as shown in the following example: 
            </para>
            <programlisting>
LIC_FILES_CHKSUM = "file://src/ls.c;startline=5;endline=16;\
                                    md5=bb14ed3c4cda583abc85401304b5cd4e"
LIC_FILES_CHKSUM = "file://../license.html;md5=5c94767cedb5d6987c902ac850ded2c6"
            </programlisting>
            <para>
                In this example the first line locates a file in 
                <glossterm><link linkend='var-S'>S</link></glossterm><filename>/src/ls.c</filename>. 
                The second line refers to a file in 
                <glossterm><link linkend='var-WORKDIR'>WORKDIR</link></glossterm>, which is the parent
                of <glossterm><link linkend='var-S'>S</link></glossterm>.
            </para>
        </section>

        <section id="usingpoky-LIC_FILES_CHKSUM-explanation-of-syntax">
            <title>Explanation of Syntax</title>
            <para>
                As mentioned in the previous section the LIC_FILES_CHKSUM variable lists all the 
                important files that contain the license text for the source code. 
                Using this variable you can specify the line on which the license text starts and ends
                by supplying "beginline" and "endline" parameters. 
                If you do not use the "beginline" parameter then it is assumed that the text begins on the 
                first line of the file. 
                Similarly, if you do not use the "endline" parameter it is assumed that the license text 
                ends as the last line of the file. 
            </para>
            <para>
                The "md5" parameter stores the md5 checksum of the license text. 
                If the license text changes in any way as compared to this parameter
                then a mis-match occurs. 
                This mismatch triggers a build failure and notifies the developer.
                Notification allows the developer to review and address the license text changes.
                Also note that if a mis-match occurs during the build, the correct md5 
                checksum is placed in the build log and can be easily copied to a .bb file.
            </para>
            <para>
                There is no limit to how many files you can specify using the LIC_FILES_CHKSUM variable.
                Generally, however, every project requires a few specifications for license tracking. 
                Many projects have a "COPYING" file that stores the license information for all the source 
                code files.
                This practice allow you to just track the "COPYING" file as long as it is kept up to date. 
            </para>
            <tip>
                If you specify an empty or invalid "md5" parameter, BitBake returns an md5 mis-match 
                error and displays the correct "md5" parameter value during the build. The correct parameter
                is also captured in the build log. 
            </tip>
            <tip>
                If the whole file contains only license text, you do not need to use the "beginline" and 
                "endline" parameters. 
            </tip>
        </section>
    </section>

    <section id="usingpoky-configuring-DISTRO_PN_ALIAS">
        <title>Handling Package Name Alias</title>
        <para>
            Sometimes a package name you are using might exist under an alias or as a similarly named
            package in a different distribution.
            Poky implements a "distro_check" task that automatically connects to major distributions
            and checks for these situations. 
            If the package exists under a different name in a different distribution you get a 
            distro_check mismatch.  
            You can resolve this problem by defining a per-distro recipe name alias using the 
            <glossterm><link linkend='var-DISTRO_PN_ALIAS'>DISTRO_PN_ALIAS</link></glossterm> variable.
        </para>

        <section id="usingpoky-specifying-DISTRO_PN_ALIAS">
            <title>Specifying the DISTRO_PN_ALIAS Variable</title>
            <para>
                Following is an example that shows how you specify the DISTRO_PN_ALIAS variable:
            <programlisting>
DISTRO_PN_ALIAS_pn-PACKAGENAME = "distro1=package_name_alias1 \
                                  distro2=package_name_alias2 \
                                  distro3=package_name_alias3 \
                                  ..."
            </programlisting>
            </para>
            <para>
                If you have more than one distribution alias, separate them with a space.
                Note that Poky currently automatically checks the Fedora, OpenSuSE, Debian, Ubuntu, 
                and Mandriva distributions for source package recipes without having to specify them 
                using the DISTRO_PN_ALIAS variable.
                For example, the following command generates a report that lists the Linux distributions
                that include the sources for each of the Poky recipes.
             <literallayout class='monospaced'>
     $ bitbake world -f -c distro_check
             </literallayout>
             The results are stored in the <filename>build/tmp/log/distro_check-${DATETIME}.results</filename> 
             file.
            </para>
        </section>
    </section>
</chapter>

<!-- 
vim: expandtab tw=80 ts=4 
-->