Copyright © 2010-2020 Linux Foundation
Permission is granted to copy, distribute and/or modify this document under the terms of the Creative Commons Attribution-Share Alike 2.0 UK: England & Wales as published by Creative Commons.
This version of the Yocto Project Mega-Manual is for the 3.1.2 release of the Yocto Project. To be sure you have the latest version of the manual for this release, go to the Yocto Project documentation page and select the manual from that site. Manuals from the site are more up-to-date than manuals derived from the Yocto Project released TAR files.
If you located this manual through a web search, the version of the manual might not be the one you want (e.g. the search might have returned a manual much older than the Yocto Project version with which you are working). You can see all Yocto Project major releases by visiting the Releases page. If you need a version of this manual for a different Yocto Project release, visit the Yocto Project documentation page and select the manual set by using the "ACTIVE RELEASES DOCUMENTATION" or "DOCUMENTS ARCHIVE" pull-down menus.
To report any inaccuracies or problems with this
(or any other Yocto Project) manual, send an email to
the Yocto Project documentation mailing list at
docs@lists.yoctoproject.org
or
log into the freenode #yocto
channel.
Revision History | |
---|---|
Revision 1.8 | April 2015 |
The initial document released with the Yocto Project 1.8 Release. | |
Revision 2.0 | October 2015 |
Released with the Yocto Project 2.0 Release. | |
Revision 2.1 | April 2016 |
Released with the Yocto Project 2.1 Release. | |
Revision 2.2 | October 2016 |
Released with the Yocto Project 2.2 Release. | |
Revision 2.3 | May 2017 |
Released with the Yocto Project 2.3 Release. | |
Revision 2.4 | October 2017 |
Released with the Yocto Project 2.4 Release. | |
Revision 2.5 | May 2018 |
Released with the Yocto Project 2.5 Release. | |
Revision 2.6 | November 2018 |
Released with the Yocto Project 2.6 Release. | |
Revision 2.7 | May 2019 |
Released with the Yocto Project 2.7 Release. | |
Revision 3.0 | October 2019 |
Released with the Yocto Project 3.0 Release. | |
Revision 3.1 | April 2020 |
Released with the Yocto Project 3.1 Release. | |
Revision 3.1.1 | June 2020 |
Released with the Yocto Project 3.1.1 Release. | |
Revision 3.1.2 | July 2020 |
Released with the Yocto Project 3.1.2 Release. |
Abstract¶
The Yocto Project Mega-Manual is a concatenation of the published Yocto Project HTML manuals along with the corresponding BitBake User Manual for the given release. The Mega-Manual exists to help users efficiently search for strings across the entire Yocto Project documentation set inclusive of the BitBake User Manual.
Copyright © 2010-2020 Linux Foundation
Permission is granted to copy, distribute and/or modify this document under the terms of the Creative Commons Attribution-Share Alike 2.0 UK: England & Wales as published by Creative Commons.
Abstract¶
Welcome! This short document steps you through the process for a typical image build using the Yocto Project. The document also introduces how to configure a build for specific hardware. You will use Yocto Project to build a reference embedded OS called Poky.
The examples in this paper assume you are using a native Linux system running a recent Ubuntu Linux distribution. If the machine you want to use Yocto Project on to build an image (build host) is not a native Linux system, you can still perform these steps by using CROss PlatformS (CROPS) and setting up a Poky container. See the Setting Up to Use CROss PlatformS (CROPS)" section in the Yocto Project Development Tasks Manual for more information.
You may use Windows Subsystem For Linux v2 to set up a build host using Windows 10.
See the Setting Up to Use Windows Subsystem For Linux" section in the Yocto Project Development Tasks Manual for more information.
If you want more conceptual or background information on the Yocto Project, see the Yocto project Overview and Concepts Manual.
Make sure your build host meets the following requirements:
50 Gbytes of free disk space
Runs a supported Linux distribution (i.e. recent releases of Fedora, openSUSE, CentOS, Debian, or Ubuntu). For a list of Linux distributions that support the Yocto Project, see the "Supported Linux Distributions" section in the Yocto Project Reference Manual. For detailed information on preparing your build host, see the "Preparing the Build Host" section in the Yocto Project Development Tasks Manual.
Git 1.8.3.1 or greater
tar 1.28 or greater
Python 3.5.0 or greater.
gcc 5.0 or greater.
If your build host does not meet any of these three listed version requirements, you can take steps to prepare the system so that you can still use the Yocto Project. See the "Required Git, tar, Python and gcc Versions" section in the Yocto Project Reference Manual for information.
You must install essential host packages on your build host. The following command installs the host packages based on an Ubuntu distribution:
$ sudo apt-get install gawk wget git-core diffstat unzip texinfo gcc-multilib \ build-essential chrpath socat cpio python3 python3-pip python3-pexpect \ xz-utils debianutils iputils-ping python3-git python3-jinja2 libegl1-mesa libsdl1.2-dev \ pylint3 xterm
Once you complete the setup instructions for your machine, you need to get a copy of the Poky repository on your build host. Use the following commands to clone the Poky repository.
$ git clone git://git.yoctoproject.org/poky Cloning into 'poky'... remote: Counting objects: 432160, done. remote: Compressing objects: 100% (102056/102056), done. remote: Total 432160 (delta 323116), reused 432037 (delta 323000) Receiving objects: 100% (432160/432160), 153.81 MiB | 8.54 MiB/s, done. Resolving deltas: 100% (323116/323116), done. Checking connectivity... done.
Move to the poky
directory and take a look
at the tags:
$ cd poky $ git fetch --tags $ git tag 1.1_M1.final 1.1_M1.rc1 1.1_M1.rc2 1.1_M2.final 1.1_M2.rc1 . . . yocto-2.5 yocto-2.5.1 yocto-2.5.2 yocto-2.6 yocto-2.6.1 yocto-2.6.2 yocto-2.7 yocto_1.5_M5.rc8
For this example, check out the branch based on the yocto-3.1.2 release:
$ git checkout tags/yocto-3.1.2 -b my-yocto-3.1.2 Switched to a new branch 'my-yocto-3.1.2'
The previous Git checkout command creates a local branch named my-yocto-3.1.2. The files available to you in that branch exactly match the repository's files in the "dunfell" development branch at the time of the Yocto Project yocto-3.1.2 release.
For more options and information about accessing Yocto Project related repositories, see the "Locating Yocto Project Source Files" section in the Yocto Project Development Tasks Manual.
Use the following steps to build your image. The build process creates an entire Linux distribution, including the toolchain, from source.
If you are working behind a firewall and your build host is not set up for proxies, you could encounter problems with the build process when fetching source code (e.g. fetcher failures or Git failures).
If you do not know your proxy settings, consult your local network infrastructure resources and get that information. A good starting point could also be to check your web browser settings. Finally, you can find more information on the "Working Behind a Network Proxy" page of the Yocto Project Wiki.
Initialize the Build Environment:
From within the poky
directory, run the
oe-init-build-env
environment setup script to define Yocto Project's
build environment on your build host.
$ cd ~/poky $ source oe-init-build-env You had no conf/local.conf file. This configuration file has therefore been created for you with some default values. You may wish to edit it to, for example, select a different MACHINE (target hardware). See conf/local.conf for more information as common configuration options are commented. You had no conf/bblayers.conf file. This configuration file has therefore been created for you with some default values. To add additional metadata layers into your configuration please add entries to conf/bblayers.conf. The Yocto Project has extensive documentation about OE including a reference manual which can be found at: http://yoctoproject.org/documentation For more information about OpenEmbedded see their website: http://www.openembedded.org/ ### Shell environment set up for builds. ### You can now run 'bitbake <target>' Common targets are: core-image-minimal core-image-sato meta-toolchain meta-ide-support You can also run generated qemu images with a command like 'runqemu qemux86-64'
Among other things, the script creates the
Build Directory,
which is build
in this case
and is located in the
Source Directory.
After the script runs, your current working directory
is set to the Build Directory.
Later, when the build completes, the Build Directory
contains all the files created during the build.
Examine Your Local Configuration File:
When you set up the build environment, a local
configuration file named
local.conf
becomes available in
a conf
subdirectory of the
Build Directory.
For this example, the defaults are set to build
for a qemux86
target, which is
suitable for emulation.
The package manager used is set to the RPM package
manager.
local.conf
file in the Build
directory:
SSTATE_MIRRORS = "\ file://.* http://sstate.yoctoproject.org/dev/PATH;downloadfilename=PATH \n \ file://.* http://sstate.yoctoproject.org/3.0.2/PATH;downloadfilename=PATH \n \ file://.* http://sstate.yoctoproject.org/3.1.2/PATH;downloadfilename=PATH \n \ "The previous examples showed how to add sstate paths for Yocto Project 3.0.2, 3.1.2, and a development area. For a complete index of sstate locations, see http://sstate.yoctoproject.org/.
Start the Build:
Continue with the following command to build an OS image
for the target, which is
core-image-sato
in this example:
$ bitbake core-image-sato
For information on using the
bitbake
command, see the
"BitBake"
section in the Yocto Project Overview and Concepts Manual,
or see the
"BitBake Command"
section in the BitBake User Manual.
Simulate Your Image Using QEMU: Once this particular image is built, you can start QEMU, which is a Quick EMUlator that ships with the Yocto Project:
$ runqemu qemux86-64
If you want to learn more about running QEMU, see the "Using the Quick EMUlator (QEMU)" chapter in the Yocto Project Development Tasks Manual.
Exit QEMU:
Exit QEMU by either clicking on the shutdown icon or by
typing Ctrl-C
in the QEMU
transcript window from which you evoked QEMU.
So far, all you have done is quickly built an image suitable for emulation only. This section shows you how to customize your build for specific hardware by adding a hardware layer into the Yocto Project development environment.
In general, layers are repositories that contain related sets of instructions and configurations that tell the Yocto Project what to do. Isolating related metadata into functionally specific layers facilitates modular development and makes it easier to reuse the layer metadata.
Follow these steps to add a hardware layer:
Find a Layer: Lots of hardware layers exist. The Yocto Project Source Repositories has many hardware layers. This example adds the meta-altera hardware layer.
Clone the Layer Use Git to make a local copy of the layer on your machine. You can put the copy in the top level of the copy of the Poky repository created earlier:
$ cd ~/poky $ git clone https://github.com/kraj/meta-altera.git Cloning into 'meta-altera'... remote: Counting objects: 25170, done. remote: Compressing objects: 100% (350/350), done. remote: Total 25170 (delta 645), reused 719 (delta 538), pack-reused 24219 Receiving objects: 100% (25170/25170), 41.02 MiB | 1.64 MiB/s, done. Resolving deltas: 100% (13385/13385), done. Checking connectivity... done.
The hardware layer now exists with other layers inside
the Poky reference repository on your build host as
meta-altera
and contains all the
metadata needed to support hardware from Altera, which
is owned by Intel.
Change the Configuration to Build for a Specific Machine:
The
MACHINE
variable in the local.conf
file
specifies the machine for the build.
For this example, set the MACHINE
variable to "cyclone5".
These configurations are used:
https://github.com/kraj/meta-altera/blob/master/conf/machine/cyclone5.conf.
Add Your Layer to the Layer Configuration File:
Before you can use a layer during a build, you must add it
to your bblayers.conf
file, which
is found in the
Build Directory's
conf
directory.
Use the bitbake-layers add-layer
command to add the layer to the configuration file:
$ cd ~/poky/build $ bitbake-layers add-layer ../meta-altera NOTE: Starting bitbake server... Parsing recipes: 100% |##################################################################| Time: 0:00:32 Parsing of 918 .bb files complete (0 cached, 918 parsed). 1401 targets, 123 skipped, 0 masked, 0 errors.
You can find more information on adding layers in the
"Adding a Layer Using the bitbake-layers
Script"
section.
Completing these steps has added the
meta-altera
layer to your Yocto Project
development environment and configured it to build for the
"cyclone5" machine.
README
.
Maybe you have an application or specific set of behaviors you
need to isolate.
You can create your own general layer using the
bitbake-layers create-layer
command.
The tool automates layer creation by setting up a
subdirectory with a layer.conf
configuration file, a recipes-example
subdirectory that contains an example.bb
recipe, a licensing file, and a README
.
The following commands run the tool to create a layer named
meta-mylayer
in the
poky
directory:
$ cd ~/poky $ bitbake-layers create-layer meta-mylayer NOTE: Starting bitbake server... Add your new layer with 'bitbake-layers add-layer meta-mylayer'
For more information on layers and how to create them, see the
"Creating a General Layer Using the bitbake-layers
Script"
section in the Yocto Project Development Tasks Manual.
Now that you have experienced using the Yocto Project, you might be asking yourself "What now?" The Yocto Project has many sources of information including the website, wiki pages, and user manuals:
Website: The Yocto Project Website provides background information, the latest builds, breaking news, full development documentation, and access to a rich Yocto Project Development Community into which you can tap.
Developer Screencast: The Getting Started with the Yocto Project - New Developer Screencast Tutorial provides a 30-minute video created for users unfamiliar with the Yocto Project but familiar with Linux build hosts. While this screencast is somewhat dated, the introductory and fundamental concepts are useful for the beginner.
Yocto Project Overview and Concepts Manual: The Yocto project Overview and Concepts Manual is a great place to start to learn about the Yocto Project. This manual introduces you to the Yocto Project and its development environment. The manual also provides conceptual information for various aspects of the Yocto Project.
Yocto Project Wiki: The Yocto Project Wiki provides additional information on where to go next when ramping up with the Yocto Project, release information, project planning, and QA information.
Yocto Project Mailing Lists: Related mailing lists provide a forum for discussion, patch submission and announcements. Several mailing lists exist and are grouped according to areas of concern. See the "Mailing lists" section in the Yocto Project Reference Manual for a complete list of Yocto Project mailing lists.
Comprehensive List of Links and Other Documentation: The "Links and Related Documentation" section in the Yocto Project Reference Manual provides a comprehensive list of all related links and other user documentation.
Welcome to the Yocto Project Overview and Concepts Manual! This manual introduces the Yocto Project by providing concepts, software overviews, best-known-methods (BKMs), and any other high-level introductory information suitable for a new Yocto Project user.
The following list describes what you can get from this manual:
Introducing the Yocto Project: This chapter provides an introduction to the Yocto Project. You will learn about features and challenges of the Yocto Project, the layer model, components and tools, development methods, the Poky reference distribution, the OpenEmbedded build system workflow, and some basic Yocto terms.
The Yocto Project Development Environment: This chapter helps you get started understanding the Yocto Project development environment. You will learn about open source, development hosts, Yocto Project source repositories, workflows using Git and the Yocto Project, a Git primer, and information about licensing.
Yocto Project Concepts: This chapter presents various concepts regarding the Yocto Project. You can find conceptual information about components, development, cross-toolchains, and so forth.
This manual does not give you the following:
Step-by-step Instructions for Development Tasks: Instructional procedures reside in other manuals within the Yocto Project documentation set. For example, the Yocto Project Development Tasks Manual provides examples on how to perform various development tasks. As another example, the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual contains detailed instructions on how to install an SDK, which is used to develop applications for target hardware.
Reference Material: This type of material resides in an appropriate reference manual. For example, system variables are documented in the Yocto Project Reference Manual. As another example, the Yocto Project Board Support Package (BSP) Developer's Guide contains reference information on BSPs.
Detailed Public Information Not Specific to the Yocto Project: For example, exhaustive information on how to use the Source Control Manager Git is better covered with Internet searches and official Git Documentation than through the Yocto Project documentation.
Because this manual presents information for many different topics, supplemental information is recommended for full comprehension. For additional introductory information on the Yocto Project, see the Yocto Project Website. If you want to build an image with no knowledge of Yocto Project as a way of quickly testing it out, see the Yocto Project Quick Build document. For a comprehensive list of links and other documentation, see the "Links and Related Documentation" section in the Yocto Project Reference Manual.
The Yocto Project is an open source collaboration project that helps developers create custom Linux-based systems that are designed for embedded products regardless of the product's hardware architecture. Yocto Project provides a flexible toolset and a development environment that allows embedded device developers across the world to collaborate through shared technologies, software stacks, configurations, and best practices used to create these tailored Linux images.
Thousands of developers worldwide have discovered that Yocto Project provides advantages in both systems and applications development, archival and management benefits, and customizations used for speed, footprint, and memory utilization. The project is a standard when it comes to delivering embedded software stacks. The project allows software customizations and build interchange for multiple hardware platforms as well as software stacks that can be maintained and scaled.
For further introductory information on the Yocto Project, you might be interested in this article by Drew Moseley and in this short introductory video.
The remainder of this section overviews advantages and challenges tied to the Yocto Project.
The following list describes features and advantages of the Yocto Project:
Widely Adopted Across the Industry: Semiconductor, operating system, software, and service vendors exist whose products and services adopt and support the Yocto Project. For a look at the Yocto Project community and the companies involved with the Yocto Project, see the "COMMUNITY" and "ECOSYSTEM" tabs on the Yocto Project home page.
Architecture Agnostic: Yocto Project supports Intel, ARM, MIPS, AMD, PPC and other architectures. Most ODMs, OSVs, and chip vendors create and supply BSPs that support their hardware. If you have custom silicon, you can create a BSP that supports that architecture.
Aside from lots of architecture support, the Yocto Project fully supports a wide range of device emulation through the Quick EMUlator (QEMU).
Images and Code Transfer Easily: Yocto Project output can easily move between architectures without moving to new development environments. Additionally, if you have used the Yocto Project to create an image or application and you find yourself not able to support it, commercial Linux vendors such as Wind River, Mentor Graphics, Timesys, and ENEA could take it and provide ongoing support. These vendors have offerings that are built using the Yocto Project.
Flexibility: Corporations use the Yocto Project many different ways. One example is to create an internal Linux distribution as a code base the corporation can use across multiple product groups. Through customization and layering, a project group can leverage the base Linux distribution to create a distribution that works for their product needs.
Ideal for Constrained Embedded and IoT devices: Unlike a full Linux distribution, you can use the Yocto Project to create exactly what you need for embedded devices. You only add the feature support or packages that you absolutely need for the device. For devices that have display hardware, you can use available system components such as X11, GTK+, Qt, Clutter, and SDL (among others) to create a rich user experience. For devices that do not have a display or where you want to use alternative UI frameworks, you can choose to not install these components.
Comprehensive Toolchain Capabilities: Toolchains for supported architectures satisfy most use cases. However, if your hardware supports features that are not part of a standard toolchain, you can easily customize that toolchain through specification of platform-specific tuning parameters. And, should you need to use a third-party toolchain, mechanisms built into the Yocto Project allow for that.
Mechanism Rules Over Policy: Focusing on mechanism rather than policy ensures that you are free to set policies based on the needs of your design instead of adopting decisions enforced by some system software provider.
Uses a Layer Model: The Yocto Project layer infrastructure groups related functionality into separate bundles. You can incrementally add these grouped functionalities to your project as needed. Using layers to isolate and group functionality reduces project complexity and redundancy, allows you to easily extend the system, make customizations, and keep functionality organized.
Supports Partial Builds: You can build and rebuild individual packages as needed. Yocto Project accomplishes this through its shared-state cache (sstate) scheme. Being able to build and debug components individually eases project development.
Releases According to a Strict Schedule: Major releases occur on a six-month cycle predictably in October and April. The most recent two releases support point releases to address common vulnerabilities and exposures. This predictability is crucial for projects based on the Yocto Project and allows development teams to plan activities.
Rich Ecosystem of Individuals and Organizations: For open source projects, the value of community is very important. Support forums, expertise, and active developers who continue to push the Yocto Project forward are readily available.
Binary Reproducibility:
The Yocto Project allows you to be very specific about
dependencies and achieves very high percentages of
binary reproducibility (e.g. 99.8% for
core-image-minimal
).
When distributions are not specific about which
packages are pulled in and in what order to support
dependencies, other build systems can arbitrarily
include packages.
License Manifest: The Yocto Project provides a license manifest for review by people who need to track the use of open source licenses (e.g.legal teams).
The following list presents challenges you might encounter when developing using the Yocto Project:
Steep Learning Curve: The Yocto Project has a steep learning curve and has many different ways to accomplish similar tasks. It can be difficult to choose how to proceed when varying methods exist by which to accomplish a given task.
Understanding What Changes You Need to Make For Your Design Requires Some Research: Beyond the simple tutorial stage, understanding what changes need to be made for your particular design can require a significant amount of research and investigation. For information that helps you transition from trying out the Yocto Project to using it for your project, see the "What I wish I'd Known" and "Transitioning to a Custom Environment for Systems Development" documents on the Yocto Project website.
Project Workflow Could Be Confusing: The Yocto Project workflow could be confusing if you are used to traditional desktop and server software development. In a desktop development environment, mechanisms exist to easily pull and install new packages, which are typically pre-compiled binaries from servers accessible over the Internet. Using the Yocto Project, you must modify your configuration and rebuild to add additional packages.
Working in a Cross-Build Environment Can Feel Unfamiliar: When developing code to run on a target, compilation, execution, and testing done on the actual target can be faster than running a BitBake build on a development host and then deploying binaries to the target for test. While the Yocto Project does support development tools on the target, the additional step of integrating your changes back into the Yocto Project build environment would be required. Yocto Project supports an intermediate approach that involves making changes on the development system within the BitBake environment and then deploying only the updated packages to the target.
The Yocto Project
OpenEmbedded build system
produces packages in standard formats (i.e. RPM,
DEB, IPK, and TAR).
You can deploy these packages into the running system
on the target by using utilities on the target such
as rpm
or
ipk
.
Initial Build Times Can be Significant: Long initial build times are unfortunately unavoidable due to the large number of packages initially built from scratch for a fully functioning Linux system. Once that initial build is completed, however, the shared-state (sstate) cache mechanism Yocto Project uses keeps the system from rebuilding packages that have not been "touched" since the last build. The sstate mechanism significantly reduces times for successive builds.
The Yocto Project's "Layer Model" is a development model for embedded and IoT Linux creation that distinguishes the Yocto Project from other simple build systems. The Layer Model simultaneously supports collaboration and customization. Layers are repositories that contain related sets of instructions that tell the OpenEmbedded build system what to do. You can collaborate, share, and reuse layers.
Layers can contain changes to previous instructions or settings at any time. This powerful override capability is what allows you to customize previously supplied collaborative or community layers to suit your product requirements.
You use different layers to logically separate information in your build. As an example, you could have BSP, GUI, distro configuration, middleware, or application layers. Putting your entire build into one layer limits and complicates future customization and reuse. Isolating information into layers, on the other hand, helps simplify future customizations and reuse. You might find it tempting to keep everything in one layer when working on a single project. However, the more modular your Metadata, the easier it is to cope with future changes.
Use Board Support Package (BSP) layers from silicon vendors when possible.
Familiarize yourself with the Yocto Project curated layer index or the OpenEmbedded layer index. The latter contains more layers but they are less universally validated.
Layers support the inclusion of technologies, hardware components, and software components. The Yocto Project Compatible designation provides a minimum level of standardization that contributes to a strong ecosystem. "YP Compatible" is applied to appropriate products and software components such as BSPs, other OE-compatible layers, and related open-source projects, allowing the producer to use Yocto Project badges and branding assets.
To illustrate how layers are used to keep things modular, consider
machine customizations.
These types of customizations typically reside in a special layer,
rather than a general layer, called a BSP Layer.
Furthermore, the machine customizations should be isolated from
recipes and Metadata that support a new GUI environment,
for example.
This situation gives you a couple of layers: one for the machine
configurations, and one for the GUI environment.
It is important to understand, however, that the BSP layer can
still make machine-specific additions to recipes within the GUI
environment layer without polluting the GUI layer itself
with those machine-specific changes.
You can accomplish this through a recipe that is a BitBake append
(.bbappend
) file, which is described later
in this section.
The
Source Directory
contains both general layers and BSP layers right out of the box.
You can easily identify layers that ship with a Yocto Project
release in the Source Directory by their names.
Layers typically have names that begin with the string
meta-
.
meta-
, but it is a commonly
accepted standard in the Yocto Project community.
For example, if you were to examine the
tree view
of the poky
repository, you will see several
layers: meta
,
meta-skeleton
,
meta-selftest
,
meta-poky
, and
meta-yocto-bsp
.
Each of these repositories represents a distinct layer.
For procedures on how to create layers, see the "Understanding and Creating Layers" section in the Yocto Project Development Tasks Manual.
The Yocto Project employs a collection of components and tools used by the project itself, by project developers, and by those using the Yocto Project. These components and tools are open source projects and metadata that are separate from the reference distribution (Poky) and the OpenEmbedded build system. Most of the components and tools are downloaded separately.
This section provides brief overviews of the components and tools associated with the Yocto Project.
The following list consists of tools that help you develop images and applications using the Yocto Project:
CROPS: CROPS is an open source, cross-platform development framework that leverages Docker Containers. CROPS provides an easily managed, extensible environment that allows you to build binaries for a variety of architectures on Windows, Linux and Mac OS X hosts.
devtool
:
This command-line tool is available as part of the
extensible SDK (eSDK) and is its cornerstone.
You can use devtool
to help build,
test, and package software within the eSDK.
You can use the tool to optionally integrate what you
build into an image built by the OpenEmbedded build
system.
The devtool
command employs
a number of sub-commands that allow you to add, modify,
and upgrade recipes.
As with the OpenEmbedded build system, “recipes”
represent software packages within
devtool
.
When you use devtool add
, a recipe
is automatically created.
When you use devtool modify
, the
specified existing recipe is used in order to determine
where to get the source code and how to patch it.
In both cases, an environment is set up so that when
you build the recipe a source tree that is under your
control is used in order to allow you to make changes
to the source as desired.
By default, both new recipes and the source go into
a “workspace” directory under the eSDK.
The devtool upgrade
command
updates an existing recipe so that you can build it
for an updated set of source files.
You can read about the
devtool
workflow in the Yocto
Project Application Development and Extensible
Software Development Kit (eSDK) Manual in the
"Using devtool
in Your SDK Workflow'"
section.
Extensible Software Development Kit (eSDK):
The eSDK provides a cross-development toolchain and
libraries tailored to the contents of a specific image.
The eSDK makes it easy to add new applications and
libraries to an image, modify the source for an
existing component, test changes on the target
hardware, and integrate into the rest of the
OpenEmbedded build system.
The eSDK gives you a toolchain experience supplemented
with the powerful set of devtool
commands tailored for the Yocto Project environment.
For information on the eSDK, see the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) Manual.
Toaster: Toaster is a web interface to the Yocto Project OpenEmbedded build system. Toaster allows you to configure, run, and view information about builds. For information on Toaster, see the Toaster User Manual.
The following list consists of tools that help production related activities using the Yocto Project:
Auto Upgrade Helper: This utility when used in conjunction with the OpenEmbedded build system (BitBake and OE-Core) automatically generates upgrades for recipes that are based on new versions of the recipes published upstream.
Recipe Reporting System: The Recipe Reporting System tracks recipe versions available for Yocto Project. The main purpose of the system is to help you manage the recipes you maintain and to offer a dynamic overview of the project. The Recipe Reporting System is built on top of the OpenEmbedded Layer Index, which is a website that indexes OpenEmbedded-Core layers.
Patchwork: Patchwork is a fork of a project originally started by OzLabs. The project is a web-based tracking system designed to streamline the process of bringing contributions into a project. The Yocto Project uses Patchwork as an organizational tool to handle patches, which number in the thousands for every release.
AutoBuilder: AutoBuilder is a project that automates build tests and quality assurance (QA). By using the public AutoBuilder, anyone can determine the status of the current "master" branch of Poky.
A goal of the Yocto Project is to lead the open source industry with a project that automates testing and QA procedures. In doing so, the project encourages a development community that publishes QA and test plans, publicly demonstrates QA and test plans, and encourages development of tools that automate and test and QA procedures for the benefit of the development community.
You can learn more about the AutoBuilder used by the Yocto Project here.
Cross-Prelink: Prelinking is the process of pre-computing the load addresses and link tables generated by the dynamic linker as compared to doing this at runtime. Doing this ahead of time results in performance improvements when the application is launched and reduced memory usage for libraries shared by many applications.
Historically, cross-prelink is a variant of
prelink, which was conceived by
Jakub Jelínek
a number of years ago.
Both prelink and cross-prelink are maintained in the
same repository albeit on separate branches.
By providing an emulated runtime dynamic linker
(i.e. glibc
-derived
ld.so
emulation), the
cross-prelink project extends the prelink software’s
ability to prelink a sysroot environment.
Additionally, the cross-prelink software enables the
ability to work in sysroot style environments.
The dynamic linker determines standard load address calculations based on a variety of factors such as mapping addresses, library usage, and library function conflicts. The prelink tool uses this information, from the dynamic linker, to determine unique load addresses for executable and linkable format (ELF) binaries that are shared libraries and dynamically linked. The prelink tool modifies these ELF binaries with the pre-computed information. The result is faster loading and often lower memory consumption because more of the library code can be re-used from shared Copy-On-Write (COW) pages.
The original upstream prelink project only supports running prelink on the end target device due to the reliance on the target device’s dynamic linker. This restriction causes issues when developing a cross-compiled system. The cross-prelink adds a synthesized dynamic loader that runs on the host, thus permitting cross-prelinking without ever having to run on a read-write target filesystem.
Pseudo: Pseudo is the Yocto Project implementation of fakeroot, which is used to run commands in an environment that seemingly has root privileges.
During a build, it can be necessary to perform
operations that require system administrator
privileges.
For example, file ownership or permissions might need
definition.
Pseudo is a tool that you can either use directly or
through the environment variable
LD_PRELOAD
.
Either method allows these operations to succeed as
if system administrator privileges exist even
when they do not.
You can read more about Pseudo in the "Fakeroot and Pseudo" section.
The following list consists of components associated with the OpenEmbedded build system:
BitBake: BitBake is a core component of the Yocto Project and is used by the OpenEmbedded build system to build images. While BitBake is key to the build system, BitBake is maintained separately from the Yocto Project.
BitBake is a generic task execution engine that allows shell and Python tasks to be run efficiently and in parallel while working within complex inter-task dependency constraints. In short, BitBake is a build engine that works through recipes written in a specific format in order to perform sets of tasks.
You can learn more about BitBake in the BitBake User Manual.
OpenEmbedded-Core: OpenEmbedded-Core (OE-Core) is a common layer of metadata (i.e. recipes, classes, and associated files) used by OpenEmbedded-derived systems, which includes the Yocto Project. The Yocto Project and the OpenEmbedded Project both maintain the OpenEmbedded-Core. You can find the OE-Core metadata in the Yocto Project Source Repositories.
Historically, the Yocto Project integrated the OE-Core metadata throughout the Yocto Project source repository reference system (Poky). After Yocto Project Version 1.0, the Yocto Project and OpenEmbedded agreed to work together and share a common core set of metadata (OE-Core), which contained much of the functionality previously found in Poky. This collaboration achieved a long-standing OpenEmbedded objective for having a more tightly controlled and quality-assured core. The results also fit well with the Yocto Project objective of achieving a smaller number of fully featured tools as compared to many different ones.
Sharing a core set of metadata results in Poky as an integration layer on top of OE-Core. You can see that in this figure. The Yocto Project combines various components such as BitBake, OE-Core, script “glue”, and documentation for its build system.
Poky is the Yocto Project reference distribution. It contains the Open-Embedded build system (BitBake and OE-Core) as well as a set of metadata to get you started building your own distribution. See the figure in "What is the Yocto Project?" section for an illustration that shows Poky and its relationship with other parts of the Yocto Project.
To use the Yocto Project tools and components, you
can download (clone
) Poky and use it
to bootstrap your own distribution.
You can read more about Poky in the "Reference Embedded Distribution (Poky)" section.
The following lists components associated with packages for finished targets:
Matchbox: Matchbox is an Open Source, base environment for the X Window System running on non-desktop, embedded platforms such as handhelds, set-top boxes, kiosks, and anything else for which screen space, input mechanisms, or system resources are limited.
Matchbox consists of a number of interchangeable and optional applications that you can tailor to a specific, non-desktop platform to enhance usability in constrained environments.
You can find the Matchbox source in the Yocto Project Source Repositories.
Opkg Open PacKaGe management (opkg) is a lightweight package management system based on the itsy package (ipkg) management system. Opkg is written in C and resembles Advanced Package Tool (APT) and Debian Package (dpkg) in operation.
Opkg is intended for use on embedded Linux devices and is used in this capacity in the OpenEmbedded and OpenWrt projects, as well as the Yocto Project.
The Build Appliance is a virtual machine image that enables you to build and boot a custom embedded Linux image with the Yocto Project using a non-Linux development system.
Historically, the Build Appliance was the second of three methods by which you could use the Yocto Project on a system that was not native to Linux.
Hob: Hob, which is now deprecated and is no longer available since the 2.1 release of the Yocto Project provided a rudimentary, GUI-based interface to the Yocto Project. Toaster has fully replaced Hob.
Build Appliance: Post Hob, the Build Appliance became available. It was never recommended that you use the Build Appliance as a day-to-day production development environment with the Yocto Project. Build Appliance was useful as a way to try out development in the Yocto Project environment.
CROPS: The final and best solution available now for developing using the Yocto Project on a system not native to Linux is with CROPS.
The Yocto Project development environment usually involves a Build Host and target hardware. You use the Build Host to build images and develop applications, while you use the target hardware to test deployed software.
This section provides an introduction to the choices or development methods you have when setting up your Build Host. Depending on the your particular workflow preference and the type of operating system your Build Host runs, several choices exist that allow you to use the Yocto Project.
Native Linux Host: By far the best option for a Build Host. A system running Linux as its native operating system allows you to develop software by directly using the BitBake tool. You can accomplish all aspects of development from a familiar shell of a supported Linux distribution.
For information on how to set up a Build Host on a system running Linux as its native operating system, see the "Setting Up a Native Linux Host" section in the Yocto Project Development Tasks Manual.
CROss PlatformS (CROPS): Typically, you use CROPS, which leverages Docker Containers, to set up a Build Host that is not running Linux (e.g. Microsoft® Windows™ or macOS®).
CROPS is an open source, cross-platform development framework that provides an easily managed, extensible environment for building binaries targeted for a variety of architectures on Windows, macOS, or Linux hosts. Once the Build Host is set up using CROPS, you can prepare a shell environment to mimic that of a shell being used on a system natively running Linux.
For information on how to set up a Build Host with CROPS, see the "Setting Up to Use CROss PlatformS (CROPS)" section in the Yocto Project Development Tasks Manual.
Windows Subsystem For Linux (WSLv2): You may use Windows Subsystem For Linux v2 to set up a build host using Windows 10.
The Windows Subsystem For Linux allows Windows 10 to run a real Linux kernel inside of a lightweight utility virtual machine (VM) using virtualization technology.
For information on how to set up a Build Host with WSLv2, see the "Setting Up to Use Windows Subsystem For Linux" section in the Yocto Project Development Tasks Manual.
Toaster: Regardless of what your Build Host is running, you can use Toaster to develop software using the Yocto Project. Toaster is a web interface to the Yocto Project's Open-Embedded build system. The interface enables you to configure and run your builds. Information about builds is collected and stored in a database. You can use Toaster to configure and start builds on multiple remote build servers.
For information about and how to use Toaster, see the Toaster User Manual.
"Poky", which is pronounced Pock-ee, is the name of the Yocto Project's reference distribution or Reference OS Kit. Poky contains the OpenEmbedded Build System (BitBake and OpenEmbedded-Core) as well as a set of metadata to get you started building your own distro. In other words, Poky is a base specification of the functionality needed for a typical embedded system as well as the components from the Yocto Project that allow you to build a distribution into a usable binary image.
Poky is a combined repository of BitBake, OpenEmbedded-Core
(which is found in meta
),
meta-poky
,
meta-yocto-bsp
, and documentation provided
all together and known to work well together.
You can view these items that make up the Poky repository in the
Source Repositories.
poky
Git repository, see the
"Top-Level Core Components"
section in the Yocto Project Reference Manual.
The following figure illustrates what generally comprises Poky:
BitBake is a task executor and scheduler that is the heart of the OpenEmbedded build system.
meta-poky
, which is Poky-specific
metadata.
meta-yocto-bsp
, which are Yocto
Project-specific Board Support Packages (BSPs).
OpenEmbedded-Core (OE-Core) metadata, which includes shared configurations, global variable definitions, shared classes, packaging, and recipes. Classes define the encapsulation and inheritance of build logic. Recipes are the logical units of software and images to be built.
Documentation, which contains the Yocto Project source files used to make the set of user manuals.
To use the Yocto Project tools, you can use Git to clone (download) the Poky repository then use your local copy of the reference distribution to bootstrap your own distribution.
Poky has a regular, well established, six-month release cycle under its own version. Major releases occur at the same time major releases (point releases) occur for the Yocto Project, which are typically in the Spring and Fall. For more information on the Yocto Project release schedule and cadence, see the "Yocto Project Releases and the Stable Release Process" chapter in the Yocto Project Reference Manual.
Much has been said about Poky being a "default configuration." A default configuration provides a starting image footprint. You can use Poky out of the box to create an image ranging from a shell-accessible minimal image all the way up to a Linux Standard Base-compliant image that uses a GNOME Mobile and Embedded (GMAE) based reference user interface called Sato.
One of the most powerful properties of Poky is that every aspect of a build is controlled by the metadata. You can use metadata to augment these base image types by adding metadata layers that extend functionality. These layers can provide, for example, an additional software stack for an image type, add a board support package (BSP) for additional hardware, or even create a new image type.
Metadata is loosely grouped into configuration files or package
recipes.
A recipe is a collection of non-executable metadata used by
BitBake to set variables or define additional build-time tasks.
A recipe contains fields such as the recipe description, the recipe
version, the license of the package and the upstream source
repository.
A recipe might also indicate that the build process uses autotools,
make, distutils or any other build process, in which case the basic
functionality can be defined by the classes it inherits from
the OE-Core layer's class definitions in
./meta/classes
.
Within a recipe you can also define additional tasks as well as
task prerequisites.
Recipe syntax through BitBake also supports both
_prepend
and _append
operators as a method of extending task functionality.
These operators inject code into the beginning or end of a task.
For information on these BitBake operators, see the
"Appending and Prepending (Override Style Syntax)"
section in the BitBake User's Manual.
The OpenEmbedded build system uses a "workflow" to accomplish image and SDK generation. The following figure overviews that workflow:
Following is a brief summary of the "workflow":
Developers specify architecture, policies, patches and configuration details.
The build system fetches and downloads the source code from the specified location. The build system supports standard methods such as tarballs or source code repositories systems such as Git.
Once source code is downloaded, the build system extracts the sources into a local work area where patches are applied and common steps for configuring and compiling the software are run.
The build system then installs the software into a temporary staging area where the binary package format you select (DEB, RPM, or IPK) is used to roll up the software.
Different QA and sanity checks run throughout entire build process.
After the binaries are created, the build system generates a binary package feed that is used to create the final root file image.
The build system generates the file system image and a customized Extensible SDK (eSDK) for application development in parallel.
For a very detailed look at this workflow, see the "OpenEmbedded Build System Concepts" section.
It helps to understand some basic fundamental terms when learning the Yocto Project. Although a list of terms exists in the "Yocto Project Terms" section of the Yocto Project Reference Manual, this section provides the definitions of some terms helpful for getting started:
Configuration Files: Files that hold global definitions of variables, user-defined variables, and hardware configuration information. These files tell the Open-Embedded build system what to build and what to put into the image to support a particular platform.
Extensible Software Development Kit (eSDK): A custom SDK for application developers. This eSDK allows developers to incorporate their library and programming changes back into the image to make their code available to other application developers. For information on the eSDK, see the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
Layer: A collection of related recipes. Layers allow you to consolidate related metadata to customize your build. Layers also isolate information used when building for multiple architectures. Layers are hierarchical in their ability to override previous specifications. You can include any number of available layers from the Yocto Project and customize the build by adding your layers after them. You can search the Layer Index for layers used within Yocto Project.
For more detailed information on layers, see the "Understanding and Creating Layers" section in the Yocto Project Development Tasks Manual. For a discussion specifically on BSP Layers, see the "BSP Layers" section in the Yocto Project Board Support Packages (BSP) Developer's Guide.
Metadata: A key element of the Yocto Project is the Metadata that is used to construct a Linux distribution and is contained in the files that the OpenEmbedded build system parses when building an image. In general, Metadata includes recipes, configuration files, and other information that refers to the build instructions themselves, as well as the data used to control what things get built and the effects of the build. Metadata also includes commands and data used to indicate what versions of software are used, from where they are obtained, and changes or additions to the software itself (patches or auxiliary files) that are used to fix bugs or customize the software for use in a particular situation. OpenEmbedded-Core is an important set of validated metadata.
OpenEmbedded Build System: The terms "BitBake" and "build system" are sometimes used for the OpenEmbedded Build System.
BitBake is a task scheduler and execution engine
that parses instructions (i.e. recipes) and configuration
data.
After a parsing phase, BitBake creates a dependency tree
to order the compilation, schedules the compilation of
the included code, and finally executes the building
of the specified custom Linux image (distribution).
BitBake is similar to the make
tool.
During a build process, the build system tracks dependencies and performs a native or cross-compilation of the package. As a first step in a cross-build setup, the framework attempts to create a cross-compiler toolchain (i.e. Extensible SDK) suited for the target platform.
OpenEmbedded-Core (OE-Core): OE-Core is metadata comprised of foundation recipes, classes, and associated files that are meant to be common among many different OpenEmbedded-derived systems, including the Yocto Project. OE-Core is a curated subset of an original repository developed by the OpenEmbedded community that has been pared down into a smaller, core set of continuously validated recipes. The result is a tightly controlled and quality-assured core set of recipes.
You can see the Metadata in the
meta
directory of the Yocto Project
Source Repositories.
Packages: In the context of the Yocto Project, this term refers to a recipe's packaged output produced by BitBake (i.e. a "baked recipe"). A package is generally the compiled binaries produced from the recipe's sources. You "bake" something by running it through BitBake.
It is worth noting that the term "package" can, in general, have subtle meanings. For example, the packages referred to in the "Required Packages for the Build Host" section in the Yocto Project Reference Manual are compiled binaries that, when installed, add functionality to your Linux distribution.
Another point worth noting is that historically within
the Yocto Project, recipes were referred to as packages - thus,
the existence of several BitBake variables that are seemingly
mis-named,
(e.g. PR
,
PV
,
and
PE
).
Poky: Poky is a reference embedded distribution and a reference test configuration. Poky provides the following:
A base-level functional distro used to illustrate how to customize a distribution.
A means by which to test the Yocto Project components (i.e. Poky is used to validate the Yocto Project).
A vehicle through which you can download the Yocto Project.
Poky is not a product level distro. Rather, it is a good starting point for customization.
Recipe: The most common form of metadata. A recipe contains a list of settings and tasks (i.e. instructions) for building packages that are then used to build the binary image. A recipe describes where you get source code and which patches to apply. Recipes describe dependencies for libraries or for other recipes as well as configuration and compilation options. Related recipes are consolidated into a layer.
This chapter takes a look at the Yocto Project development environment. The chapter provides Yocto Project Development environment concepts that help you understand how work is accomplished in an open source environment, which is very different as compared to work accomplished in a closed, proprietary environment.
Specifically, this chapter addresses open source philosophy, source repositories, workflows, Git, and licensing.
Open source philosophy is characterized by software development directed by peer production and collaboration through an active community of developers. Contrast this to the more standard centralized development models used by commercial software companies where a finite set of developers produces a product for sale using a defined set of procedures that ultimately result in an end product whose architecture and source material are closed to the public.
Open source projects conceptually have differing concurrent agendas, approaches, and production. These facets of the development process can come from anyone in the public (community) who has a stake in the software project. The open source environment contains new copyright, licensing, domain, and consumer issues that differ from the more traditional development environment. In an open source environment, the end product, source material, and documentation are all available to the public at no cost.
A benchmark example of an open source project is the Linux kernel, which was initially conceived and created by Finnish computer science student Linus Torvalds in 1991. Conversely, a good example of a non-open source project is the Windows® family of operating systems developed by Microsoft® Corporation.
Wikipedia has a good historical description of the Open Source Philosophy here. You can also find helpful information on how to participate in the Linux Community here.
A development host or build host is key to using the Yocto Project. Because the goal of the Yocto Project is to develop images or applications that run on embedded hardware, development of those images and applications generally takes place on a system not intended to run the software - the development host.
You need to set up a development host in order to use it with the Yocto Project. Most find that it is best to have a native Linux machine function as the development host. However, it is possible to use a system that does not run Linux as its operating system as your development host. When you have a Mac or Windows-based system, you can set it up as the development host by using CROPS, which leverages Docker Containers. Once you take the steps to set up a CROPS machine, you effectively have access to a shell environment that is similar to what you see when using a Linux-based development host. For the steps needed to set up a system using CROPS, see the "Setting Up to Use CROss PlatformS (CROPS)" section in the Yocto Project Development Tasks Manual.
If your development host is going to be a system that runs a Linux distribution, steps still exist that you must take to prepare the system for use with the Yocto Project. You need to be sure that the Linux distribution on the system is one that supports the Yocto Project. You also need to be sure that the correct set of host packages are installed that allow development using the Yocto Project. For the steps needed to set up a development host that runs Linux, see the "Setting Up a Native Linux Host" section in the Yocto Project Development Tasks Manual.
Once your development host is set up to use the Yocto Project, several methods exist for you to do work in the Yocto Project environment:
Command Lines, BitBake, and Shells: Traditional development in the Yocto Project involves using the OpenEmbedded build system, which uses BitBake, in a command-line environment from a shell on your development host. You can accomplish this from a host that is a native Linux machine or from a host that has been set up with CROPS. Either way, you create, modify, and build images and applications all within a shell-based environment using components and tools available through your Linux distribution and the Yocto Project.
For a general flow of the build procedures, see the "Building a Simple Image" section in the Yocto Project Development Tasks Manual.
Board Support Package (BSP) Development: Development of BSPs involves using the Yocto Project to create and test layers that allow easy development of images and applications targeted for specific hardware. To development BSPs, you need to take some additional steps beyond what was described in setting up a development host.
The Yocto Project Board Support Package (BSP) Developer's Guide provides BSP-related development information. For specifics on development host preparation, see the "Preparing Your Build Host to Work With BSP Layers" section in the Yocto Project Board Support Package (BSP) Developer's Guide.
Kernel Development:
If you are going to be developing kernels using the Yocto
Project you likely will be using devtool
.
A workflow using devtool
makes kernel
development quicker by reducing iteration cycle times.
The Yocto Project Linux Kernel Development Manual provides kernel-related development information. For specifics on development host preparation, see the "Preparing the Build Host to Work on the Kernel" section in the Yocto Project Linux Kernel Development Manual.
Using Toaster: The other Yocto Project development method that involves an interface that effectively puts the Yocto Project into the background is Toaster. Toaster provides an interface to the OpenEmbedded build system. The interface enables you to configure and run your builds. Information about builds is collected and stored in a database. You can use Toaster to configure and start builds on multiple remote build servers.
For steps that show you how to set up your development host to use Toaster and on how to use Toaster in general, see the Toaster User Manual.
The Yocto Project team maintains complete source repositories for all Yocto Project files at http://git.yoctoproject.org. This web-based source code browser is organized into categories by function such as IDE Plugins, Matchbox, Poky, Yocto Linux Kernel, and so forth. From the interface, you can click on any particular item in the "Name" column and see the URL at the bottom of the page that you need to clone a Git repository for that particular item. Having a local Git repository of the Source Directory, which is usually named "poky", allows you to make changes, contribute to the history, and ultimately enhance the Yocto Project's tools, Board Support Packages, and so forth.
For any supported release of Yocto Project, you can also go to the
Yocto Project Website and
select the "DOWNLOADS" item from the "SOFTWARE" menu and get a
released tarball of the poky
repository, any
supported BSP tarball, or Yocto Project tools.
Unpacking these tarballs gives you a snapshot of the released
files.
The recommended method for setting up the Yocto Project
Source Directory
and the files for supported BSPs
(e.g., meta-intel
) is to use
Git to create a local copy of
the upstream repositories.
Be sure to always work in matching branches for both
the selected BSP repository and the Source Directory
(i.e. poky
) repository.
For example, if you have checked out the "master" branch
of poky
and you are going to use
meta-intel
, be sure to checkout the
"master" branch of meta-intel
.
In summary, here is where you can get the project files needed for development:
Source Repositories: This area contains IDE Plugins, Matchbox, Poky, Poky Support, Tools, Yocto Linux Kernel, and Yocto Metadata Layers. You can create local copies of Git repositories for each of these areas.
For steps on how to view and access these upstream Git repositories, see the "Accessing Source Repositories" Section in the Yocto Project Development Tasks Manual.
Index of /releases: This is an index of releases such as Poky, Pseudo, installers for cross-development toolchains, miscellaneous support and all released versions of Yocto Project in the form of images or tarballs. Downloading and extracting these files does not produce a local copy of the Git repository but rather a snapshot of a particular release or image.
For steps on how to view and access these files, see the "Accessing Index of Releases" section in the Yocto Project Development Tasks Manual.
"DOWNLOADS" page for the Yocto Project Website:
The Yocto Project website includes a "DOWNLOADS" page accessible through the "SOFTWARE" menu that allows you to download any Yocto Project release, tool, and Board Support Package (BSP) in tarball form. The tarballs are similar to those found in the Index of /releases: area.
For steps on how to use the "DOWNLOADS" page, see the "Using the Downloads Page" section in the Yocto Project Development Tasks Manual.
Developing using the Yocto Project likely requires the use of Git. Git is a free, open source distributed version control system used as part of many collaborative design environments. This section provides workflow concepts using the Yocto Project and Git. In particular, the information covers basic practices that describe roles and actions in a collaborative development environment.
The Yocto Project files are maintained using Git in "branches" whose Git histories track every change and whose structures provide branches for all diverging functionality. Although there is no need to use Git, many open source projects do so.
For the Yocto Project, a key individual called the "maintainer" is responsible for the integrity of the "master" branch of a given Git repository. The "master" branch is the “upstream” repository from which final or most recent builds of a project occur. The maintainer is responsible for accepting changes from other developers and for organizing the underlying branch structure to reflect release strategies and so forth.
The Yocto Project poky
Git repository also has an
upstream contribution Git repository named
poky-contrib
.
You can see all the branches in this repository using the web interface
of the
Source Repositories organized
within the "Poky Support" area.
These branches hold changes (commits) to the project that have been
submitted or committed by the Yocto Project development team and by
community members who contribute to the project.
The maintainer determines if the changes are qualified to be moved
from the "contrib" branches into the "master" branch of the Git
repository.
Developers (including contributing community members) create and maintain cloned repositories of upstream branches. The cloned repositories are local to their development platforms and are used to develop changes. When a developer is satisfied with a particular feature or change, they "push" the change to the appropriate "contrib" repository.
Developers are responsible for keeping their local repository up-to-date with whatever upstream branch they are working against. They are also responsible for straightening out any conflicts that might arise within files that are being worked on simultaneously by more than one person. All this work is done locally on the development host before anything is pushed to a "contrib" area and examined at the maintainer’s level.
A somewhat formal method exists by which developers commit changes and push them into the "contrib" area and subsequently request that the maintainer include them into an upstream branch. This process is called “submitting a patch” or "submitting a change." For information on submitting patches and changes, see the "Submitting a Change to the Yocto Project" section in the Yocto Project Development Tasks Manual.
In summary, a single point of entry exists for changes into a "master" or development branch of the Git repository, which is controlled by the project’s maintainer. And, a set of developers exist who independently develop, test, and submit changes to "contrib" areas for the maintainer to examine. The maintainer then chooses which changes are going to become a permanent part of the project.
While each development environment is unique, there are some best practices or methods that help development run smoothly. The following list describes some of these practices. For more information about Git workflows, see the workflow topics in the Git Community Book.
Make Small Changes: It is best to keep the changes you commit small as compared to bundling many disparate changes into a single commit. This practice not only keeps things manageable but also allows the maintainer to more easily include or refuse changes.
Make Complete Changes: It is also good practice to leave the repository in a state that allows you to still successfully build your project. In other words, do not commit half of a feature, then add the other half as a separate, later commit. Each commit should take you from one buildable project state to another buildable state.
Use Branches Liberally: It is very easy to create, use, and delete local branches in your working Git repository on the development host. You can name these branches anything you like. It is helpful to give them names associated with the particular feature or change on which you are working. Once you are done with a feature or change and have merged it into your local master branch, simply discard the temporary branch.
Merge Changes:
The git merge
command allows you to take
the changes from one branch and fold them into another branch.
This process is especially helpful when more than a single
developer might be working on different parts of the same
feature.
Merging changes also automatically identifies any collisions
or "conflicts" that might happen as a result of the same lines
of code being altered by two different developers.
Manage Branches: Because branches are easy to use, you should use a system where branches indicate varying levels of code readiness. For example, you can have a "work" branch to develop in, a "test" branch where the code or change is tested, a "stage" branch where changes are ready to be committed, and so forth. As your project develops, you can merge code across the branches to reflect ever-increasing stable states of the development.
Use Push and Pull:
The push-pull workflow is based on the concept of developers
"pushing" local commits to a remote repository, which is
usually a contribution repository.
This workflow is also based on developers "pulling" known
states of the project down into their local development
repositories.
The workflow easily allows you to pull changes submitted by
other developers from the upstream repository into your
work area ensuring that you have the most recent software
on which to develop.
The Yocto Project has two scripts named
create-pull-request
and
send-pull-request
that ship with the
release to facilitate this workflow.
You can find these scripts in the scripts
folder of the
Source Directory.
For information on how to use these scripts, see the
"Using Scripts to Push a Change Upstream and Request a Pull"
section in the Yocto Project Development Tasks Manual.
Patch Workflow:
This workflow allows you to notify the maintainer through an
email that you have a change (or patch) you would like
considered for the "master" branch of the Git repository.
To send this type of change, you format the patch and then
send the email using the Git commands
git format-patch
and
git send-email
.
For information on how to use these scripts, see the
"Submitting a Change to the Yocto Project"
section in the Yocto Project Development Tasks Manual.
The Yocto Project makes extensive use of Git, which is a free, open source distributed version control system. Git supports distributed development, non-linear development, and can handle large projects. It is best that you have some fundamental understanding of how Git tracks projects and how to work with Git if you are going to use the Yocto Project for development. This section provides a quick overview of how Git works and provides you with a summary of some essential Git commands.
For more information on Git, see http://git-scm.com/documentation.
If you need to download Git, it is recommended that you add Git to your system through your distribution's "software store" (e.g. for Ubuntu, use the Ubuntu Software feature). For the Git download page, see http://git-scm.com/download.
For information beyond the introductory nature in this section, see the "Locating Yocto Project Source Files" section in the Yocto Project Development Tasks Manual.
As mentioned briefly in the previous section and also in the "Git Workflows and the Yocto Project" section, the Yocto Project maintains source repositories at http://git.yoctoproject.org. If you look at this web-interface of the repositories, each item is a separate Git repository.
Git repositories use branching techniques that track content change (not files) within a project (e.g. a new feature or updated documentation). Creating a tree-like structure based on project divergence allows for excellent historical information over the life of a project. This methodology also allows for an environment from which you can do lots of local experimentation on projects as you develop changes or new features.
A Git repository represents all development efforts for a given
project.
For example, the Git repository poky
contains
all changes and developments for that repository over the course
of its entire life.
That means that all changes that make up all releases are captured.
The repository maintains a complete history of changes.
You can create a local copy of any repository by "cloning" it
with the git clone
command.
When you clone a Git repository, you end up with an identical
copy of the repository on your development system.
Once you have a local copy of a repository, you can take steps to
develop locally.
For examples on how to clone Git repositories, see the
"Locating Yocto Project Source Files"
section in the Yocto Project Development Tasks Manual.
It is important to understand that Git tracks content change and
not files.
Git uses "branches" to organize different development efforts.
For example, the poky
repository has
several branches that include the current "dunfell"
branch, the "master" branch, and many branches for past
Yocto Project releases.
You can see all the branches by going to
http://git.yoctoproject.org/cgit.cgi/poky/ and
clicking on the
[...]
link beneath the "Branch" heading.
Each of these branches represents a specific area of development. The "master" branch represents the current or most recent development. All other branches represent offshoots of the "master" branch.
When you create a local copy of a Git repository, the copy has the same set of branches as the original. This means you can use Git to create a local working area (also called a branch) that tracks a specific development branch from the upstream source Git repository. in other words, you can define your local Git environment to work on any development branch in the repository. To help illustrate, consider the following example Git commands:
$ cd ~ $ git clone git://git.yoctoproject.org/poky $ cd poky $ git checkout -b dunfell origin/dunfell
In the previous example after moving to the home directory, the
git clone
command creates a
local copy of the upstream poky
Git repository.
By default, Git checks out the "master" branch for your work.
After changing the working directory to the new local repository
(i.e. poky
), the
git checkout
command creates
and checks out a local branch named "dunfell", which
tracks the upstream "origin/dunfell" branch.
Changes you make while in this branch would ultimately affect
the upstream "dunfell" branch of the
poky
repository.
It is important to understand that when you create and checkout a local working branch based on a branch name, your local environment matches the "tip" of that particular development branch at the time you created your local branch, which could be different from the files in the "master" branch of the upstream repository. In other words, creating and checking out a local branch based on the "dunfell" branch name is not the same as checking out the "master" branch in the repository. Keep reading to see how you create a local snapshot of a Yocto Project Release.
Git uses "tags" to mark specific changes in a repository branch
structure.
Typically, a tag is used to mark a special point such as the final
change (or commit) before a project is released.
You can see the tags used with the poky
Git
repository by going to
http://git.yoctoproject.org/cgit.cgi/poky/ and
clicking on the
[...]
link beneath the "Tag" heading.
Some key tags for the poky
repository are
jethro-14.0.3
,
morty-16.0.1
,
pyro-17.0.0
, and
dunfell-23.0.2
.
These tags represent Yocto Project releases.
When you create a local copy of the Git repository, you also have access to all the tags in the upstream repository. Similar to branches, you can create and checkout a local working Git branch based on a tag name. When you do this, you get a snapshot of the Git repository that reflects the state of the files when the change was made associated with that tag. The most common use is to checkout a working branch that matches a specific Yocto Project release. Here is an example:
$ cd ~ $ git clone git://git.yoctoproject.org/poky $ cd poky $ git fetch --tags $ git checkout tags/rocko-18.0.0 -b my_rocko-18.0.0
In this example, the name of the top-level directory of your
local Yocto Project repository is poky
.
After moving to the poky
directory, the
git fetch
command makes all the upstream
tags available locally in your repository.
Finally, the git checkout
command
creates and checks out a branch named "my-rocko-18.0.0" that is
based on the upstream branch whose "HEAD" matches the
commit in the repository associated with the "rocko-18.0.0" tag.
The files in your repository now exactly match that particular
Yocto Project release as it is tagged in the upstream Git
repository.
It is important to understand that when you create and
checkout a local working branch based on a tag, your environment
matches a specific point in time and not the entire development
branch (i.e. from the "tip" of the branch backwards).
Git has an extensive set of commands that lets you manage changes and perform collaboration over the life of a project. Conveniently though, you can manage with a small set of basic operations and workflows once you understand the basic philosophy behind Git. You do not have to be an expert in Git to be functional. A good place to look for instruction on a minimal set of Git commands is here.
The following list of Git commands briefly describes some basic Git operations as a way to get started. As with any set of commands, this list (in most cases) simply shows the base command and omits the many arguments it supports. See the Git documentation for complete descriptions and strategies on how to use these commands:
git init
:
Initializes an empty Git repository.
You cannot use Git commands unless you have a
.git
repository.
git clone
:
Creates a local clone of a Git repository that is on
equal footing with a fellow developer’s Git repository
or an upstream repository.
git add
:
Locally stages updated file contents to the index that
Git uses to track changes.
You must stage all files that have changed before you
can commit them.
git commit
:
Creates a local "commit" that documents the changes you
made.
Only changes that have been staged can be committed.
Commits are used for historical purposes, for determining
if a maintainer of a project will allow the change,
and for ultimately pushing the change from your local
Git repository into the project’s upstream repository.
git status
:
Reports any modified files that possibly need to be
staged and gives you a status of where you stand regarding
local commits as compared to the upstream repository.
git checkout
branch-name
:
Changes your local working branch and in this form
assumes the local branch already exists.
This command is analogous to "cd".
git checkout –b
working-branch
upstream-branch
:
Creates and checks out a working branch on your local
machine.
The local branch tracks the upstream branch.
You can use your local branch to isolate your work.
It is a good idea to use local branches when adding
specific features or changes.
Using isolated branches facilitates easy removal of
changes if they do not work out.
git branch
:
Displays the existing local branches associated with your
local repository.
The branch that you have currently checked out is noted
with an asterisk character.
git branch -D
branch-name
:
Deletes an existing local branch.
You need to be in a local branch other than the one you
are deleting in order to delete
branch-name
.
git pull --rebase
:
Retrieves information from an upstream Git repository
and places it in your local Git repository.
You use this command to make sure you are synchronized with
the repository from which you are basing changes
(.e.g. the "master" branch).
The "--rebase" option ensures that any local commits you
have in your branch are preserved at the top of your
local branch.
git push
repo-name
local-branch
:
upstream-branch
:
Sends all your committed local changes to the upstream Git
repository that your local repository is tracking
(e.g. a contribution repository).
The maintainer of the project draws from these repositories
to merge changes (commits) into the appropriate branch
of project's upstream repository.
git merge
:
Combines or adds changes from one
local branch of your repository with another branch.
When you create a local Git repository, the default branch
is named "master".
A typical workflow is to create a temporary branch that is
based off "master" that you would use for isolated work.
You would make your changes in that isolated branch,
stage and commit them locally, switch to the "master"
branch, and then use the git merge
command to apply the changes from your isolated branch
into the currently checked out branch (e.g. "master").
After the merge is complete and if you are done with
working in that isolated branch, you can safely delete
the isolated branch.
git cherry-pick
commits
:
Choose and apply specific commits from one branch
into another branch.
There are times when you might not be able to merge
all the changes in one branch with
another but need to pick out certain ones.
gitk
:
Provides a GUI view of the branches and changes in your
local Git repository.
This command is a good way to graphically see where things
have diverged in your local repository.
gitk
package on your development system to use this
command.
git log
:
Reports a history of your commits to the repository.
This report lists all commits regardless of whether you
have pushed them upstream or not.
git diff
:
Displays line-by-line differences between a local
working file and the same file as understood by Git.
This command is useful to see what you have changed
in any given file.
Because open source projects are open to the public, they have different licensing structures in place. License evolution for both Open Source and Free Software has an interesting history. If you are interested in this history, you can find basic information here:
In general, the Yocto Project is broadly licensed under the Massachusetts Institute of Technology (MIT) License. MIT licensing permits the reuse of software within proprietary software as long as the license is distributed with that software. MIT is also compatible with the GNU General Public License (GPL). Patches to the Yocto Project follow the upstream licensing scheme. You can find information on the MIT license here. You can find information on the GNU GPL here.
When you build an image using the Yocto Project, the build process
uses a known list of licenses to ensure compliance.
You can find this list in the
Source Directory
at meta/files/common-licenses
.
Once the build completes, the list of all licenses found and used
during that build are kept in the
Build Directory
at tmp/deploy/licenses
.
If a module requires a license that is not in the base list, the build process generates a warning during the build. These tools make it easier for a developer to be certain of the licenses with which their shipped products must comply. However, even with these tools it is still up to the developer to resolve potential licensing issues.
The base list of licenses used by the build process is a combination of the Software Package Data Exchange (SPDX) list and the Open Source Initiative (OSI) projects. SPDX Group is a working group of the Linux Foundation that maintains a specification for a standard format for communicating the components, licenses, and copyrights associated with a software package. OSI is a corporation dedicated to the Open Source Definition and the effort for reviewing and approving licenses that conform to the Open Source Definition (OSD).
You can find a list of the combined SPDX and OSI licenses that the
Yocto Project uses in the
meta/files/common-licenses
directory in your
Source Directory.
For information that can help you maintain compliance with various open source licensing during the lifecycle of a product created using the Yocto Project, see the "Maintaining Open Source License Compliance During Your Product's Lifecycle" section in the Yocto Project Development Tasks Manual.
This chapter provides explanations for Yocto Project concepts that go beyond the surface of "how-to" information and reference (or look-up) material. Concepts such as components, the OpenEmbedded build system workflow, cross-development toolchains, shared state cache, and so forth are explained.
The BitBake task executor together with various types of configuration files form the OpenEmbedded-Core. This section overviews these components by describing their use and how they interact.
BitBake handles the parsing and execution of the data files. The data itself is of various types:
Recipes: Provides details about particular pieces of software.
Class Data: Abstracts common build information (e.g. how to build a Linux kernel).
Configuration Data: Defines machine-specific settings, policy decisions, and so forth. Configuration data acts as the glue to bind everything together.
BitBake knows how to combine multiple data sources together and refers to each data source as a layer. For information on layers, see the "Understanding and Creating Layers" section of the Yocto Project Development Tasks Manual.
Following are some brief details on these core components. For additional information on how these components interact during a build, see the "OpenEmbedded Build System Concepts" section.
BitBake is the tool at the heart of the OpenEmbedded build system and is responsible for parsing the Metadata, generating a list of tasks from it, and then executing those tasks.
This section briefly introduces BitBake. If you want more information on BitBake, see the BitBake User Manual.
To see a list of the options BitBake supports, use either of the following commands:
$ bitbake -h $ bitbake --help
The most common usage for BitBake is
bitbake
,
where packagename
packagename
is the name of the
package you want to build (referred to as the "target").
The target often equates to the first part of a recipe's
filename (e.g. "foo" for a recipe named
foo_1.3.0-r0.bb
).
So, to process the
matchbox-desktop_1.2.3.bb
recipe file, you
might type the following:
$ bitbake matchbox-desktop
Several different versions of
matchbox-desktop
might exist.
BitBake chooses the one selected by the distribution
configuration.
You can get more details about how BitBake chooses between
different target versions and providers in the
"Preferences"
section of the BitBake User Manual.
BitBake also tries to execute any dependent tasks first.
So for example, before building
matchbox-desktop
, BitBake would build a
cross compiler and glibc
if they had not
already been built.
A useful BitBake option to consider is the
-k
or --continue
option.
This option instructs BitBake to try and continue processing
the job as long as possible even after encountering an error.
When an error occurs, the target that failed and those that
depend on it cannot be remade.
However, when you use this option other dependencies can
still be processed.
Files that have the .bb
suffix are
"recipes" files.
In general, a recipe contains information about a single piece
of software.
This information includes the location from which to download
the unaltered source, any source patches to be applied to that
source (if needed), which special configuration options to
apply, how to compile the source files, and how to package the
compiled output.
The term "package" is sometimes used to refer to recipes.
However, since the word "package" is used for the packaged
output from the OpenEmbedded build system (i.e.
.ipk
or .deb
files),
this document avoids using the term "package" when referring
to recipes.
Class files (.bbclass
) contain information
that is useful to share between recipes files.
An example is the
autotools
class, which contains common settings for any application that
Autotools uses.
The
"Classes"
chapter in the Yocto Project Reference Manual provides
details about classes and how to use them.
The configuration files (.conf
) define
various configuration variables that govern the OpenEmbedded
build process.
These files fall into several areas that define machine
configuration options, distribution configuration options,
compiler tuning options, general common configuration options,
and user configuration options in
conf/local.conf
, which is found in the
Build Directory.
Layers are repositories that contain related metadata (i.e. sets of instructions) that tell the OpenEmbedded build system how to build a target. Yocto Project's layer model facilitates collaboration, sharing, customization, and reuse within the Yocto Project development environment. Layers logically separate information for your project. For example, you can use a layer to hold all the configurations for a particular piece of hardware. Isolating hardware-specific configurations allows you to share other metadata by using a different layer where that metadata might be common across several pieces of hardware.
Many layers exist that work in the Yocto Project development environment. The Yocto Project Curated Layer Index and OpenEmbedded Layer Index both contain layers from which you can use or leverage.
By convention, layers in the Yocto Project follow a specific form.
Conforming to a known structure allows BitBake to make assumptions
during builds on where to find types of metadata.
You can find procedures and learn about tools (i.e.
bitbake-layers
) for creating layers suitable
for the Yocto Project in the
"Understanding and Creating Layers"
section of the Yocto Project Development Tasks Manual.
This section takes a more detailed look inside the build process used by the OpenEmbedded build system, which is the build system specific to the Yocto Project. At the heart of the build system is BitBake, the task executor.
The following diagram represents the high-level workflow of a build. The remainder of this section expands on the fundamental input, output, process, and metadata logical blocks that make up the workflow.
In general, the build's workflow consists of several functional areas:
User Configuration: metadata you can use to control the build process.
Metadata Layers: Various layers that provide software, machine, and distro metadata.
Source Files: Upstream releases, local projects, and SCMs.
Build System: Processes under the control of BitBake. This block expands on how BitBake fetches source, applies patches, completes compilation, analyzes output for package generation, creates and tests packages, generates images, and generates cross-development tools.
Package Feeds: Directories containing output packages (RPM, DEB or IPK), which are subsequently used in the construction of an image or Software Development Kit (SDK), produced by the build system. These feeds can also be copied and shared using a web server or other means to facilitate extending or updating existing images on devices at runtime if runtime package management is enabled.
Images: Images produced by the workflow.
Application Development SDK: Cross-development tools that are produced along with an image or separately with BitBake.
User configuration helps define the build. Through user configuration, you can tell BitBake the target architecture for which you are building the image, where to store downloaded source, and other build properties.
The following figure shows an expanded representation of the "User Configuration" box of the general workflow figure:
BitBake needs some basic configuration files in order to
complete a build.
These files are *.conf
files.
The minimally necessary ones reside as example files in the
build/conf
directory of the
Source Directory.
For simplicity, this section refers to the Source Directory as
the "Poky Directory."
When you clone the
Poky
Git repository or you download and unpack a Yocto Project
release, you can set up the Source Directory to be named
anything you want.
For this discussion, the cloned repository uses the default
name poky
.
The meta-poky
layer inside Poky contains
a conf
directory that has example
configuration files.
These example files are used as a basis for creating actual
configuration files when you source
oe-init-build-env
,
which is the build environment script.
Sourcing the build environment script creates a
Build Directory
if one does not already exist.
BitBake uses the Build Directory for all its work during
builds.
The Build Directory has a conf
directory
that contains default versions of your
local.conf
and
bblayers.conf
configuration files.
These default configuration files are created only if versions
do not already exist in the Build Directory at the time you
source the build environment setup script.
Because the Poky repository is fundamentally an aggregation of
existing repositories, some users might be familiar with
running the oe-init-build-env
script
in the context of separate
OpenEmbedded-Core
and BitBake repositories rather than a single Poky repository.
This discussion assumes the script is executed from
within a cloned or unpacked version of Poky.
Depending on where the script is sourced, different
sub-scripts are called to set up the Build Directory
(Yocto or OpenEmbedded).
Specifically, the script
scripts/oe-setup-builddir
inside the
poky directory sets up the Build Directory and seeds the
directory (if necessary) with configuration files appropriate
for the Yocto Project development environment.
scripts/oe-setup-builddir
script
uses the $TEMPLATECONF
variable to
determine which sample configuration files to locate.
The local.conf
file provides many
basic variables that define a build environment.
Here is a list of a few.
To see the default configurations in a
local.conf
file created by the build
environment script, see the
local.conf.sample
in the meta-poky
layer:
Target Machine Selection:
Controlled by the
MACHINE
variable.
Download Directory:
Controlled by the
DL_DIR
variable.
Shared State Directory:
Controlled by the
SSTATE_DIR
variable.
Build Output:
Controlled by the
TMPDIR
variable.
Distribution Policy:
Controlled by the
DISTRO
variable.
Packaging Format:
Controlled by the
PACKAGE_CLASSES
variable.
SDK Target Architecture:
Controlled by the
SDKMACHINE
variable.
Extra Image Packages:
Controlled by the
EXTRA_IMAGE_FEATURES
variable.
conf/local.conf
file can also be set
in the conf/site.conf
and
conf/auto.conf
configuration files.
The bblayers.conf
file tells BitBake what
layers you want considered during the build.
By default, the layers listed in this file include layers
minimally needed by the build system.
However, you must manually add any custom layers you have
created.
You can find more information on working with the
bblayers.conf
file in the
"Enabling Your Layer"
section in the Yocto Project Development Tasks Manual.
The files site.conf
and
auto.conf
are not created by the
environment initialization script.
If you want the site.conf
file, you
need to create that yourself.
The auto.conf
file is typically created by
an autobuilder:
site.conf
:
You can use the conf/site.conf
configuration file to configure multiple
build directories.
For example, suppose you had several build environments
and they shared some common features.
You can set these default build properties here.
A good example is perhaps the packaging format to use
through the
PACKAGE_CLASSES
variable.
One useful scenario for using the
conf/site.conf
file is to extend
your
BBPATH
variable to include the path to a
conf/site.conf
.
Then, when BitBake looks for Metadata using
BBPATH
, it finds the
conf/site.conf
file and applies
your common configurations found in the file.
To override configurations in a particular build
directory, alter the similar configurations within
that build directory's
conf/local.conf
file.
auto.conf
:
The file is usually created and written to by
an autobuilder.
The settings put into the file are typically the
same as you would find in the
conf/local.conf
or the
conf/site.conf
files.
You can edit all configuration files to further define any particular build environment. This process is represented by the "User Configuration Edits" box in the figure.
When you launch your build with the
bitbake
command, BitBake sorts out the configurations to ultimately
define your build environment.
It is important to understand that the
OpenEmbedded build system
reads the configuration files in a specific order:
target
site.conf
, auto.conf
,
and local.conf
.
And, the build system applies the normal assignment statement
rules as described in the
"Syntax and Operators"
chapter of the BitBake User Manual.
Because the files are parsed in a specific order, variable
assignments for the same variable could be affected.
For example, if the auto.conf
file and
the local.conf
set
variable1
to different values,
because the build system parses local.conf
after auto.conf
,
variable1
is assigned the value from
the local.conf
file.
The previous section described the user configurations that define BitBake's global behavior. This section takes a closer look at the layers the build system uses to further control the build. These layers provide Metadata for the software, machine, and policies.
In general, three types of layer input exists. You can see them below the "User Configuration" box in the general workflow figure:
Metadata (.bb
+ Patches):
Software layers containing user-supplied recipe files,
patches, and append files.
A good example of a software layer might be the
meta-qt5
layer from the
OpenEmbedded Layer Index.
This layer is for version 5.0 of the popular
Qt
cross-platform application development framework for
desktop, embedded and mobile.
Machine BSP Configuration:
Board Support Package (BSP) layers (i.e. "BSP Layer"
in the following figure) providing machine-specific
configurations.
This type of information is specific to a particular
target architecture.
A good example of a BSP layer from the
Poky Reference Distribution
is the
meta-yocto-bsp
layer.
Policy Configuration:
Distribution Layers (i.e. "Distro Layer" in the
following figure) providing top-level or general
policies for the images or SDKs being built for a
particular distribution.
For example, in the Poky Reference Distribution the
distro layer is the
meta-poky
layer.
Within the distro layer is a
conf/distro
directory that
contains distro configuration files (e.g.
poky.conf
that contain many policy configurations for the
Poky distribution.
The following figure shows an expanded representation of these three layers from the general workflow figure:
In general, all layers have a similar structure.
They all contain a licensing file
(e.g. COPYING.MIT
) if the layer is to be
distributed, a README
file as good
practice and especially if the layer is to be distributed, a
configuration directory, and recipe directories.
You can learn about the general structure for layers used with
the Yocto Project in the
"Creating Your Own Layer"
section in the Yocto Project Development Tasks Manual.
For a general discussion on layers and the many layers from
which you can draw, see the
"Layers" and
"The Yocto Project Layer Model"
sections both earlier in this manual.
If you explored the previous links, you discovered some areas where many layers that work with the Yocto Project exist. The Source Repositories also shows layers categorized under "Yocto Metadata Layers."
BitBake uses the conf/bblayers.conf
file,
which is part of the user configuration, to find what layers it
should be using as part of the build.
The distribution layer provides policy configurations for
your distribution.
Best practices dictate that you isolate these types of
configurations into their own layer.
Settings you provide in
conf/distro/
override
similar settings that BitBake finds in your
distro
.confconf/local.conf
file in the Build
Directory.
The following list provides some explanation and references for what you typically find in the distribution layer:
classes:
Class files (.bbclass
) hold
common functionality that can be shared among
recipes in the distribution.
When your recipes inherit a class, they take on the
settings and functions for that class.
You can read more about class files in the
"Classes"
chapter of the Yocto Reference Manual.
conf:
This area holds configuration files for the
layer (conf/layer.conf
),
the distribution
(conf/distro/
),
and any distribution-wide include files.
distro
.conf
recipes-*:
Recipes and append files that affect common
functionality across the distribution.
This area could include recipes and append files
to add distribution-specific configuration,
initialization scripts, custom image recipes,
and so forth.
Examples of recipes-*
directories are recipes-core
and recipes-extra
.
Hierarchy and contents within a
recipes-*
directory can vary.
Generally, these directories contain recipe files
(*.bb
), recipe append files
(*.bbappend
), directories
that are distro-specific for configuration files,
and so forth.
The BSP Layer provides machine configurations that target specific hardware. Everything in this layer is specific to the machine for which you are building the image or the SDK. A common structure or form is defined for BSP layers. You can learn more about this structure in the Yocto Project Board Support Package (BSP) Developer's Guide.
The BSP Layer's configuration directory contains
configuration files for the machine
(conf/machine/
)
and, of course, the layer
(machine
.confconf/layer.conf
).
The remainder of the layer is dedicated to specific recipes
by function: recipes-bsp
,
recipes-core
,
recipes-graphics
,
recipes-kernel
, and so forth.
Metadata can exist for multiple formfactors, graphics
support systems, and so forth.
recipes-*
directories, not all
these directories appear in all BSP layers.
The software layer provides the Metadata for additional software packages used during the build. This layer does not include Metadata that is specific to the distribution or the machine, which are found in their respective layers.
This layer contains any recipes, append files, and patches, that your project needs.
In order for the OpenEmbedded build system to create an image or any target, it must be able to access source files. The general workflow figure represents source files using the "Upstream Project Releases", "Local Projects", and "SCMs (optional)" boxes. The figure represents mirrors, which also play a role in locating source files, with the "Source Materials" box.
The method by which source files are ultimately organized is a function of the project. For example, for released software, projects tend to use tarballs or other archived files that can capture the state of a release guaranteeing that it is statically represented. On the other hand, for a project that is more dynamic or experimental in nature, a project might keep source files in a repository controlled by a Source Control Manager (SCM) such as Git. Pulling source from a repository allows you to control the point in the repository (the revision) from which you want to build software. Finally, a combination of the two might exist, which would give the consumer a choice when deciding where to get source files.
BitBake uses the
SRC_URI
variable to point to source files regardless of their location.
Each recipe must have a SRC_URI
variable
that points to the source.
Another area that plays a significant role in where source
files come from is pointed to by the
DL_DIR
variable.
This area is a cache that can hold previously downloaded
source.
You can also instruct the OpenEmbedded build system to create
tarballs from Git repositories, which is not the default
behavior, and store them in the DL_DIR
by using the
BB_GENERATE_MIRROR_TARBALLS
variable.
Judicious use of a DL_DIR
directory can
save the build system a trip across the Internet when looking
for files.
A good method for using a download directory is to have
DL_DIR
point to an area outside of your
Build Directory.
Doing so allows you to safely delete the Build Directory
if needed without fear of removing any downloaded source file.
The remainder of this section provides a deeper look into the source files and the mirrors. Here is a more detailed look at the source file area of the general workflow figure:
Upstream project releases exist anywhere in the form of an archived file (e.g. tarball or zip file). These files correspond to individual recipes. For example, the figure uses specific releases each for BusyBox, Qt, and Dbus. An archive file can be for any released product that can be built using a recipe.
Local projects are custom bits of software the user provides. These bits reside somewhere local to a project - perhaps a directory into which the user checks in items (e.g. a local directory containing a development source tree used by the group).
The canonical method through which to include a local
project is to use the
externalsrc
class to include that local project.
You use either the local.conf
or a
recipe's append file to override or set the
recipe to point to the local directory on your disk to pull
in the whole source tree.
Another place from which the build system can get source
files is with
fetchers
employing various Source Control Managers (SCMs) such as
Git or Subversion.
In such cases, a repository is cloned or checked out.
The
do_fetch
task inside BitBake uses
the SRC_URI
variable and the argument's prefix to determine the correct
fetcher module.
DL_DIR
directory, see the
BB_GENERATE_MIRROR_TARBALLS
variable in the Yocto Project Reference Manual.
When fetching a repository, BitBake uses the
SRCREV
variable to determine the specific revision from which to
build.
Two kinds of mirrors exist: pre-mirrors and regular
mirrors.
The
PREMIRRORS
and
MIRRORS
variables point to these, respectively.
BitBake checks pre-mirrors before looking upstream for any
source files.
Pre-mirrors are appropriate when you have a shared
directory that is not a directory defined by the
DL_DIR
variable.
A Pre-mirror typically points to a shared directory that is
local to your organization.
Regular mirrors can be any site across the Internet that is used as an alternative location for source code should the primary site not be functioning for some reason or another.
When the OpenEmbedded build system generates an image or an SDK, it gets the packages from a package feed area located in the Build Directory. The general workflow figure shows this package feeds area in the upper-right corner.
This section looks a little closer into the package feeds area used by the build system. Here is a more detailed look at the area:
Package feeds are an intermediary step in the build process.
The OpenEmbedded build system provides classes to generate
different package types, and you specify which classes to
enable through the
PACKAGE_CLASSES
variable.
Before placing the packages into package feeds,
the build process validates them with generated output quality
assurance checks through the
insane
class.
The package feed area resides in the Build Directory. The directory the build system uses to temporarily store packages is determined by a combination of variables and the particular package manager in use. See the "Package Feeds" box in the illustration and note the information to the right of that area. In particular, the following defines where package files are kept:
DEPLOY_DIR
:
Defined as tmp/deploy
in the Build
Directory.
DEPLOY_DIR_*
:
Depending on the package manager used, the package type
sub-folder.
Given RPM, IPK, or DEB packaging and tarball creation,
the
DEPLOY_DIR_RPM
,
DEPLOY_DIR_IPK
,
DEPLOY_DIR_DEB
,
or
DEPLOY_DIR_TAR
,
variables are used, respectively.
PACKAGE_ARCH
:
Defines architecture-specific sub-folders.
For example, packages could exist for the i586 or
qemux86 architectures.
BitBake uses the
do_package_write_*
tasks to generate packages and place them into the package
holding area (e.g. do_package_write_ipk
for IPK packages).
See the
"do_package_write_deb
",
"do_package_write_ipk
",
"do_package_write_rpm
",
and
"do_package_write_tar
"
sections in the Yocto Project Reference Manual
for additional information.
As an example, consider a scenario where an IPK packaging
manager is being used and package architecture support for
both i586 and qemux86 exist.
Packages for the i586 architecture are placed in
build/tmp/deploy/ipk/i586
, while packages
for the qemux86 architecture are placed in
build/tmp/deploy/ipk/qemux86
.
The OpenEmbedded build system uses BitBake to produce images and Software Development Kits (SDKs). You can see from the general workflow figure, the BitBake area consists of several functional areas. This section takes a closer look at each of those areas.
The first stages of building a recipe are to fetch and unpack the source code:
The
do_fetch
and
do_unpack
tasks fetch the source files and unpack them into the
Build Directory.
file://
)
that is part of a recipe's
SRC_URI
statement, the OpenEmbedded build system takes a
checksum of the file for the recipe and inserts the
checksum into the signature for the
do_fetch
task.
If any local file has been modified, the
do_fetch
task and all tasks that
depend on it are re-executed.
By default, everything is accomplished in the Build
Directory, which has a defined structure.
For additional general information on the Build Directory,
see the
"build/
"
section in the Yocto Project Reference Manual.
Each recipe has an area in the Build Directory where the
unpacked source code resides.
The
S
variable points to this area for a recipe's unpacked source
code.
The name of that directory for any given recipe is defined
from several different variables.
The preceding figure and the following list describe
the Build Directory's hierarchy:
TMPDIR
:
The base directory where the OpenEmbedded build
system performs all its work during the build.
The default base directory is the
tmp
directory.
PACKAGE_ARCH
:
The architecture of the built package or packages.
Depending on the eventual destination of the
package or packages (i.e. machine architecture,
build host,
SDK, or specific machine),
PACKAGE_ARCH
varies.
See the variable's description for details.
TARGET_OS
:
The operating system of the target device.
A typical value would be "linux" (e.g.
"qemux86-poky-linux").
PN
:
The name of the recipe used to build the package.
This variable can have multiple meanings.
However, when used in the context of input files,
PN
represents the the name
of the recipe.
WORKDIR
:
The location where the OpenEmbedded build system
builds a recipe (i.e. does the work to create the
package).
S
:
Contains the unpacked source files for a given
recipe.
PACKAGE_ARCH
) and one based on a
machine (i.e. MACHINE
).
The underlying structures are identical.
The differentiator being what the OpenEmbedded build
system is using as a build target (e.g. general
architecture, a build host, an SDK, or a specific
machine).
Once source code is fetched and unpacked, BitBake locates patch files and applies them to the source files:
The
do_patch
task uses a recipe's
SRC_URI
statements and the
FILESPATH
variable to locate applicable patch files.
Default processing for patch files assumes the files have
either *.patch
or
*.diff
file types.
You can use SRC_URI
parameters to
change the way the build system recognizes patch files.
See the
do_patch
task for more information.
BitBake finds and applies multiple patches for a single
recipe in the order in which it locates the patches.
The FILESPATH
variable defines the
default set of directories that the build system uses to
search for patch files.
Once found, patches are applied to the recipe's source
files, which are located in the
S
directory.
For more information on how the source directories are
created, see the
"Source Fetching"
section.
For more information on how to create patches and how the
build system processes patches, see the
"Patching Code"
section in the Yocto Project Development Tasks Manual.
You can also see the
"Use devtool modify
to Modify the Source of an Existing Component"
section in the Yocto Project Application Development and
the Extensible Software Development Kit (SDK) manual and
the
"Using Traditional Kernel Development to Patch the Kernel"
section in the Yocto Project Linux Kernel Development
Manual.
After source code is patched, BitBake executes tasks that configure and compile the source code. Once compilation occurs, the files are copied to a holding area (staged) in preparation for packaging:
This step in the build process consists of the following tasks:
do_prepare_recipe_sysroot
:
This task sets up the two sysroots in
${
WORKDIR
}
(i.e. recipe-sysroot
and
recipe-sysroot-native
) so that
during the packaging phase the sysroots can contain
the contents of the
do_populate_sysroot
tasks of the recipes on which the recipe
containing the tasks depends.
A sysroot exists for both the target and for the
native binaries, which run on the host system.
do_configure
:
This task configures the source by enabling and
disabling any build-time and configuration options
for the software being built.
Configurations can come from the recipe itself as
well as from an inherited class.
Additionally, the software itself might configure
itself depending on the target for which it is
being built.
The configurations handled by the
do_configure
task are specific to configurations for the source
code being built by the recipe.
If you are using the
autotools
class, you can add additional configuration options
by using the
EXTRA_OECONF
or
PACKAGECONFIG_CONFARGS
variables.
For information on how this variable works within
that class, see the
autotools
class
here.
do_compile
:
Once a configuration task has been satisfied,
BitBake compiles the source using the
do_compile
task.
Compilation occurs in the directory pointed to by
the
B
variable.
Realize that the B
directory
is, by default, the same as the
S
directory.
do_install
:
After compilation completes, BitBake executes the
do_install
task.
This task copies files from the
B
directory and places them
in a holding area pointed to by the
D
variable.
Packaging occurs later using files from this
holding directory.
After source code is configured, compiled, and staged, the build system analyzes the results and splits the output into packages:
The
do_package
and
do_packagedata
tasks combine to analyze the files found in the
D
directory and split them into subsets based on available
packages and files.
Analysis involves the following as well as other items:
splitting out debugging symbols, looking at shared library
dependencies between packages, and looking at package
relationships.
The do_packagedata
task creates
package metadata based on the analysis such that the
build system can generate the final packages.
The
do_populate_sysroot
task stages (copies) a subset of the files installed by
the
do_install
task into the appropriate sysroot.
Working, staged, and intermediate results of the analysis
and package splitting process use several areas:
PKGD
:
The destination directory
(i.e. package
) for packages
before they are split into individual packages.
PKGDESTWORK
:
A temporary work area (i.e.
pkgdata
) used by the
do_package
task to save
package metadata.
PKGDEST
:
The parent directory (i.e.
packages-split
) for packages
after they have been split.
PKGDATA_DIR
:
A shared, global-state directory that holds
packaging metadata generated during the packaging
process.
The packaging process copies metadata from
PKGDESTWORK
to the
PKGDATA_DIR
area where it
becomes globally available.
STAGING_DIR_HOST
:
The path for the sysroot for the system on which
a component is built to run (i.e.
recipe-sysroot
).
STAGING_DIR_NATIVE
:
The path for the sysroot used when building
components for the build host (i.e.
recipe-sysroot-native
).
STAGING_DIR_TARGET
:
The path for the sysroot used when a component that
is built to execute on a system and it generates
code for yet another machine (e.g. cross-canadian
recipes).
The
FILES
variable defines the files that go into each package in
PACKAGES
.
If you want details on how this is accomplished, you can
look at
package.bbclass
.
Depending on the type of packages being created (RPM, DEB,
or IPK), the
do_package_write_*
task creates the actual packages and places them in the
Package Feed area, which is
${TMPDIR}/deploy
.
You can see the
"Package Feeds"
section for more detail on that part of the build process.
deploy/*
directories does not
exist.
Creating such feeds usually requires some kind of feed
maintenance mechanism that would upload the new
packages into an official package feed (e.g. the
Ångström distribution).
This functionality is highly distribution-specific
and thus is not provided out of the box.
Once packages are split and stored in the Package Feeds area, the build system uses BitBake to generate the root filesystem image:
The image generation process consists of several stages and
depends on several tasks and variables.
The
do_rootfs
task creates the root filesystem (file and directory
structure) for an image.
This task uses several key variables to help create the
list of packages to actually install:
IMAGE_INSTALL
:
Lists out the base set of packages from which to
install from the Package Feeds area.
PACKAGE_EXCLUDE
:
Specifies packages that should not be installed
into the image.
IMAGE_FEATURES
:
Specifies features to include in the image.
Most of these features map to additional packages
for installation.
PACKAGE_CLASSES
:
Specifies the package backend (e.g. RPM, DEB, or
IPK) to use and consequently helps determine where
to locate packages within the Package Feeds area.
IMAGE_LINGUAS
:
Determines the language(s) for which additional
language support packages are installed.
PACKAGE_INSTALL
:
The final list of packages passed to the package
manager for installation into the image.
With
IMAGE_ROOTFS
pointing to the location of the filesystem under
construction and the PACKAGE_INSTALL
variable providing the final list of packages to install,
the root file system is created.
Package installation is under control of the package manager (e.g. dnf/rpm, opkg, or apt/dpkg) regardless of whether or not package management is enabled for the target. At the end of the process, if package management is not enabled for the target, the package manager's data files are deleted from the root filesystem. As part of the final stage of package installation, post installation scripts that are part of the packages are run. Any scripts that fail to run on the build host are run on the target when the target system is first booted. If you are using a read-only root filesystem, all the post installation scripts must succeed on the build host during the package installation phase since the root filesystem on the target is read-only.
The final stages of the do_rootfs
task
handle post processing.
Post processing includes creation of a manifest file and
optimizations.
The manifest file (.manifest
) resides
in the same directory as the root filesystem image.
This file lists out, line-by-line, the installed packages.
The manifest file is useful for the
testimage
class, for example, to determine whether or not to run
specific tests.
See the
IMAGE_MANIFEST
variable for additional information.
Optimizing processes that are run across the image include
mklibs
, prelink
,
and any other post-processing commands as defined by the
ROOTFS_POSTPROCESS_COMMAND
variable.
The mklibs
process optimizes the size
of the libraries, while the prelink
process optimizes the dynamic linking of shared libraries
to reduce start up time of executables.
After the root filesystem is built, processing begins on
the image through the
do_image
task.
The build system runs any pre-processing commands as
defined by the
IMAGE_PREPROCESS_COMMAND
variable.
This variable specifies a list of functions to call before
the build system creates the final image output files.
The build system dynamically creates
do_image_*
tasks as needed, based
on the image types specified in the
IMAGE_FSTYPES
variable.
The process turns everything into an image file or a set of
image files and can compress the root filesystem image to
reduce the overall size of the image.
The formats used for the root filesystem depend on the
IMAGE_FSTYPES
variable.
Compression depends on whether the formats support
compression.
As an example, a dynamically created task when creating a
particular image type
would
take the following form:
do_image_type
So, if the type
as specified by
the IMAGE_FSTYPES
were
ext4
, the dynamically generated task
would be as follows:
do_image_ext4
The final task involved in image creation is the
do_image_complete
task.
This task completes the image by applying any image
post processing as defined through the
IMAGE_POSTPROCESS_COMMAND
variable.
The variable specifies a list of functions to call once the
build system has created the final image output files.
The OpenEmbedded build system uses BitBake to generate the Software Development Kit (SDK) installer scripts for both the standard SDK and the extensible SDK (eSDK):
do_populate_sdk
task, see the
"Building an SDK Installer"
section in the Yocto Project Application Development
and the Extensible Software Development Kit (eSDK)
manual.
Like image generation, the SDK script process consists of
several stages and depends on many variables.
The
do_populate_sdk
and
do_populate_sdk_ext
tasks use these key variables to help create the list of
packages to actually install.
For information on the variables listed in the figure,
see the
"Application Development SDK"
section.
The do_populate_sdk
task helps create
the standard SDK and handles two parts: a target part and a
host part.
The target part is the part built for the target hardware
and includes libraries and headers.
The host part is the part of the SDK that runs on the
SDKMACHINE
.
The do_populate_sdk_ext
task helps
create the extensible SDK and handles host and target parts
differently than its counter part does for the standard SDK.
For the extensible SDK, the task encapsulates the build
system, which includes everything needed (host and target)
for the SDK.
Regardless of the type of SDK being constructed, the
tasks perform some cleanup after which a cross-development
environment setup script and any needed configuration files
are created.
The final output is the Cross-development
toolchain installation script (.sh
file), which includes the environment setup script.
For each task that completes successfully, BitBake writes a
stamp file into the
STAMPS_DIR
directory.
The beginning of the stamp file's filename is determined
by the
STAMP
variable, and the end of the name consists of the task's
name and current
input checksum.
BB_SIGNATURE_HANDLER
is "OEBasicHash", which is almost always the case in
current OpenEmbedded.
To determine if a task needs to be rerun, BitBake checks if a stamp file with a matching input checksum exists for the task. If such a stamp file exists, the task's output is assumed to exist and still be valid. If the file does not exist, the task is rerun.
The stamp mechanism is more general than the shared state (sstate) cache mechanism described in the "Setscene Tasks and Shared State" section. BitBake avoids rerunning any task that has a valid stamp file, not just tasks that can be accelerated through the sstate cache.
However, you should realize that stamp files only
serve as a marker that some work has been done and that
these files do not record task output.
The actual task output would usually be somewhere in
TMPDIR
(e.g. in some recipe's
WORKDIR
.)
What the sstate cache mechanism adds is a way to cache
task output that can then be shared between build
machines.
Since STAMPS_DIR
is usually a
subdirectory of TMPDIR
, removing
TMPDIR
will also remove
STAMPS_DIR
, which means tasks will
properly be rerun to repopulate
TMPDIR
.
If you want some task to always be considered "out of
date", you can mark it with the
nostamp
varflag.
If some other task depends on such a task, then that
task will also always be considered out of date, which
might not be what you want.
For details on how to view information about a task's signature, see the "Viewing Task Variable Dependencies" section in the Yocto Project Development Tasks Manual.
The description of tasks so far assumes that BitBake needs to build everything and no available prebuilt objects exist. BitBake does support skipping tasks if prebuilt objects are available. These objects are usually made available in the form of a shared state (sstate) cache.
SSTATE_DIR
and
SSTATE_MIRRORS
variables.
The idea of a setscene task (i.e
do_
taskname
_setscene
)
is a version of the task where
instead of building something, BitBake can skip to the end
result and simply place a set of files into specific
locations as needed.
In some cases, it makes sense to have a setscene task
variant (e.g. generating package files in the
do_package_write_*
task).
In other cases, it does not make sense (e.g. a
do_patch
task or a
do_unpack
task) since the work involved would be equal to or greater
than the underlying task.
In the build system, the common tasks that have setscene
variants are
do_package
,
do_package_write_*
,
do_deploy
,
do_packagedata
,
and
do_populate_sysroot
.
Notice that these tasks represent most of the tasks whose
output is an end result.
The build system has knowledge of the relationship between
these tasks and other preceding tasks.
For example, if BitBake runs
do_populate_sysroot_setscene
for
something, it does not make sense to run any of the
do_fetch
,
do_unpack
,
do_patch
,
do_configure
,
do_compile
, and
do_install
tasks.
However, if do_package
needs to be
run, BitBake needs to run those other tasks.
It becomes more complicated if everything can come
from an sstate cache because some objects are simply
not required at all.
For example, you do not need a compiler or native tools,
such as quilt, if nothing exists to compile or patch.
If the do_package_write_*
packages
are available from sstate, BitBake does not need the
do_package
task data.
To handle all these complexities, BitBake runs in two phases. The first is the "setscene" stage. During this stage, BitBake first checks the sstate cache for any targets it is planning to build. BitBake does a fast check to see if the object exists rather than a complete download. If nothing exists, the second phase, which is the setscene stage, completes and the main build proceeds.
If objects are found in the sstate cache, the build system works backwards from the end targets specified by the user. For example, if an image is being built, the build system first looks for the packages needed for that image and the tools needed to construct an image. If those are available, the compiler is not needed. Thus, the compiler is not even downloaded. If something was found to be unavailable, or the download or setscene task fails, the build system then tries to install dependencies, such as the compiler, from the cache.
The availability of objects in the sstate cache is
handled by the function specified by the
BB_HASHCHECK_FUNCTION
variable and returns a list of available objects.
The function specified by the
BB_SETSCENE_DEPVALID
variable is the function that determines whether a given
dependency needs to be followed, and whether for any given
relationship the function needs to be passed.
The function returns a True or False value.
The images produced by the build system are compressed forms of the root filesystem and are ready to boot on a target device. You can see from the general workflow figure that BitBake output, in part, consists of images. This section takes a closer look at this output:
The build process writes images out to the
Build Directory
inside the
tmp/deploy/images/
folder as shown in the figure.
This folder contains any files expected to be loaded on the
target device.
The
machine
/DEPLOY_DIR
variable points to the deploy
directory,
while the
DEPLOY_DIR_IMAGE
variable points to the appropriate directory containing images
for the current configuration.
kernel-image
:
A kernel binary file.
The
KERNEL_IMAGETYPE
variable determines the naming scheme for the
kernel image file.
Depending on this variable, the file could begin with
a variety of naming strings.
The
deploy/images/
machine
directory can contain multiple image files for the
machine.
root-filesystem-image
:
Root filesystems for the target device (e.g.
*.ext3
or
*.bz2
files).
The
IMAGE_FSTYPES
variable determines the root filesystem image type.
The
deploy/images/
machine
directory can contain multiple root filesystems for the
machine.
kernel-modules
:
Tarballs that contain all the modules built for the
kernel.
Kernel module tarballs exist for legacy purposes and
can be suppressed by setting the
MODULE_TARBALL_DEPLOY
variable to "0".
The
deploy/images/
machine
directory can contain multiple kernel module tarballs
for the machine.
bootloaders
:
If applicable to the target machine, bootloaders
supporting the image.
The deploy/images/
machine
directory can contain multiple bootloaders for the
machine.
symlinks
:
The
deploy/images/
machine
folder contains a symbolic link that points to the
most recently built file for each machine.
These links might be useful for external scripts that
need to obtain the latest version of each file.
In the
general workflow figure,
the output labeled "Application Development SDK" represents an
SDK.
The SDK generation process differs depending on whether you
build an extensible SDK (e.g.
bitbake -c populate_sdk_ext
imagename
)
or a standard SDK (e.g.
bitbake -c populate_sdk
imagename
).
This section takes a closer look at this output:
The specific form of this output is a set of files that
includes a self-extracting SDK installer
(*.sh
), host and target manifest files,
and files used for SDK testing.
When the SDK installer file is run, it installs the SDK.
The SDK consists of a cross-development toolchain, a set of
libraries and headers, and an SDK environment setup script.
Running this installer essentially sets up your
cross-development environment.
You can think of the cross-toolchain as the "host"
part because it runs on the SDK machine.
You can think of the libraries and headers as the "target"
part because they are built for the target hardware.
The environment setup script is added so that you can
initialize the environment before using the tools.
The Yocto Project supports several methods by which you can set up this cross-development environment. These methods include downloading pre-built SDK installers or building and installing your own SDK installer.
For background information on cross-development toolchains in the Yocto Project development environment, see the "Cross-Development Toolchain Generation" section.
For information on setting up a cross-development environment, see the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
All the output files for an SDK are written to the
deploy/sdk
folder inside the
Build Directory
as shown in the previous figure.
Depending on the type of SDK, several variables exist that help
configure these files.
The following list shows the variables associated with an
extensible SDK:
DEPLOY_DIR
:
Points to the deploy
directory.
SDK_EXT_TYPE
:
Controls whether or not shared state artifacts are
copied into the extensible SDK.
By default, all required shared state artifacts are
copied into the SDK.
SDK_INCLUDE_PKGDATA
:
Specifies whether or not packagedata is included in the
extensible SDK for all recipes in the "world" target.
SDK_INCLUDE_TOOLCHAIN
:
Specifies whether or not the toolchain is included
when building the extensible SDK.
SDK_LOCAL_CONF_WHITELIST
:
A list of variables allowed through from the build
system configuration into the extensible SDK
configuration.
SDK_LOCAL_CONF_BLACKLIST
:
A list of variables not allowed through from the build
system configuration into the extensible SDK
configuration.
SDK_INHERIT_BLACKLIST
:
A list of classes to remove from the
INHERIT
value globally within the extensible SDK configuration.
This next list, shows the variables associated with a standard SDK:
DEPLOY_DIR
:
Points to the deploy
directory.
SDKMACHINE
:
Specifies the architecture of the machine on which the
cross-development tools are run to create packages for
the target hardware.
SDKIMAGE_FEATURES
:
Lists the features to include in the "target" part
of the SDK.
TOOLCHAIN_HOST_TASK
:
Lists packages that make up the host part of the SDK
(i.e. the part that runs on the
SDKMACHINE
).
When you use
bitbake -c populate_sdk
to create the SDK, a set of default packages apply.
This variable allows you to add more packages.
imagename
TOOLCHAIN_TARGET_TASK
:
Lists packages that make up the target part of the SDK
(i.e. the part built for the target hardware).
SDKPATH
:
Defines the default SDK installation path offered by
the installation script.
SDK_HOST_MANIFEST
:
Lists all the installed packages that make up the host
part of the SDK.
This variable also plays a minor role for extensible
SDK development as well.
However, it is mainly used for the standard SDK.
SDK_TARGET_MANIFEST
:
Lists all the installed packages that make up the
target part of the SDK.
This variable also plays a minor role for extensible
SDK development as well.
However, it is mainly used for the standard SDK.
The Yocto Project does most of the work for you when it comes to creating cross-development toolchains. This section provides some technical background on how cross-development toolchains are created and used. For more information on toolchains, you can also see the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
In the Yocto Project development environment, cross-development toolchains are used to build images and applications that run on the target hardware. With just a few commands, the OpenEmbedded build system creates these necessary toolchains for you.
The following figure shows a high-level build environment regarding toolchain construction and use.
Most of the work occurs on the Build Host.
This is the machine used to build images and generally work within
the the Yocto Project environment.
When you run
BitBake
to create an image, the OpenEmbedded build system
uses the host gcc
compiler to bootstrap a
cross-compiler named gcc-cross
.
The gcc-cross
compiler is what BitBake uses to
compile source files when creating the target image.
You can think of gcc-cross
simply as an
automatically generated cross-compiler that is used internally
within BitBake only.
gcc-cross-canadian
since this SDK
ships a copy of the OpenEmbedded build system and the sysroot
within it contains gcc-cross
.
The chain of events that occurs when gcc-cross
is
bootstrapped is as follows:
gcc -> binutils-cross -> gcc-cross-initial -> linux-libc-headers -> glibc-initial -> glibc -> gcc-cross -> gcc-runtime
gcc
:
The build host's GNU Compiler Collection (GCC).
binutils-cross
:
The bare minimum binary utilities needed in order to run
the gcc-cross-initial
phase of the
bootstrap operation.
gcc-cross-initial
:
An early stage of the bootstrap process for creating
the cross-compiler.
This stage builds enough of the gcc-cross
,
the C library, and other pieces needed to finish building the
final cross-compiler in later stages.
This tool is a "native" package (i.e. it is designed to run on
the build host).
linux-libc-headers
:
Headers needed for the cross-compiler.
glibc-initial
:
An initial version of the Embedded GNU C Library
(GLIBC) needed to bootstrap glibc
.
glibc
:
The GNU C Library.
gcc-cross
:
The final stage of the bootstrap process for the
cross-compiler.
This stage results in the actual cross-compiler that
BitBake uses when it builds an image for a targeted
device.
gcc-cross
.
This tool is also a "native" package (i.e. it is designed to run on the build host).
gcc-runtime
:
Runtime libraries resulting from the toolchain bootstrapping
process.
This tool produces a binary that consists of the
runtime libraries need for the targeted device.
You can use the OpenEmbedded build system to build an installer for
the relocatable SDK used to develop applications.
When you run the installer, it installs the toolchain, which
contains the development tools (e.g.,
gcc-cross-canadian
,
binutils-cross-canadian
, and other
nativesdk-*
tools),
which are tools native to the SDK (i.e. native to
SDK_ARCH
),
you need to cross-compile and test your software.
The figure shows the commands you use to easily build out this
toolchain.
This cross-development toolchain is built to execute on the
SDKMACHINE
,
which might or might not be the same
machine as the Build Host.
Here is the bootstrap process for the relocatable toolchain:
gcc -> binutils-crosssdk -> gcc-crosssdk-initial -> linux-libc-headers -> glibc-initial -> nativesdk-glibc -> gcc-crosssdk -> gcc-cross-canadian
gcc
:
The build host's GNU Compiler Collection (GCC).
binutils-crosssdk
:
The bare minimum binary utilities needed in order to run
the gcc-crosssdk-initial
phase of the
bootstrap operation.
gcc-crosssdk-initial
:
An early stage of the bootstrap process for creating
the cross-compiler.
This stage builds enough of the
gcc-crosssdk
and supporting pieces so that
the final stage of the bootstrap process can produce the
finished cross-compiler.
This tool is a "native" binary that runs on the build host.
linux-libc-headers
:
Headers needed for the cross-compiler.
glibc-initial
:
An initial version of the Embedded GLIBC needed to bootstrap
nativesdk-glibc
.
nativesdk-glibc
:
The Embedded GLIBC needed to bootstrap the
gcc-crosssdk
.
gcc-crosssdk
:
The final stage of the bootstrap process for the
relocatable cross-compiler.
The gcc-crosssdk
is a transitory
compiler and never leaves the build host.
Its purpose is to help in the bootstrap process to create
the eventual gcc-cross-canadian
compiler, which is relocatable.
This tool is also a "native" package (i.e. it is
designed to run on the build host).
gcc-cross-canadian
:
The final relocatable cross-compiler.
When run on the
SDKMACHINE
,
this tool
produces executable code that runs on the target device.
Only one cross-canadian compiler is produced per architecture
since they can be targeted at different processor optimizations
using configurations passed to the compiler through the
compile commands.
This circumvents the need for multiple compilers and thus
reduces the size of the toolchains.
By design, the OpenEmbedded build system builds everything from scratch unless BitBake can determine that parts do not need to be rebuilt. Fundamentally, building from scratch is attractive as it means all parts are built fresh and no possibility of stale data exists that can cause problems. When developers hit problems, they typically default back to building from scratch so they have a know state from the start.
Building an image from scratch is both an advantage and a disadvantage to the process. As mentioned in the previous paragraph, building from scratch ensures that everything is current and starts from a known state. However, building from scratch also takes much longer as it generally means rebuilding things that do not necessarily need to be rebuilt.
The Yocto Project implements shared state code that supports incremental builds. The implementation of the shared state code answers the following questions that were fundamental roadblocks within the OpenEmbedded incremental build support system:
What pieces of the system have changed and what pieces have not changed?
How are changed pieces of software removed and replaced?
How are pre-built components that do not need to be rebuilt from scratch used when they are available?
For the first question, the build system detects changes in the "inputs" to a given task by creating a checksum (or signature) of the task's inputs. If the checksum changes, the system assumes the inputs have changed and the task needs to be rerun. For the second question, the shared state (sstate) code tracks which tasks add which output to the build process. This means the output from a given task can be removed, upgraded or otherwise manipulated. The third question is partly addressed by the solution for the second question assuming the build system can fetch the sstate objects from remote locations and install them if they are deemed to be valid.
The build system does not maintain
PR
information as part of the shared state packages.
Consequently, considerations exist that affect
maintaining shared state feeds.
For information on how the build system works with
packages and can track incrementing
PR
information, see the
"Automatically Incrementing a Binary Package Revision Number"
section in the Yocto Project Development Tasks Manual.
The code in the build system that supports incremental builds is not simple code. For techniques that help you work around issues related to shared state code, see the "Viewing Metadata Used to Create the Input Signature of a Shared State Task" and "Invalidating Shared State to Force a Task to Run" sections both in the Yocto Project Development Tasks Manual.
The rest of this section goes into detail about the overall incremental build architecture, the checksums (signatures), and shared state.
When determining what parts of the system need to be built,
BitBake works on a per-task basis rather than a per-recipe
basis.
You might wonder why using a per-task basis is preferred over
a per-recipe basis.
To help explain, consider having the IPK packaging backend
enabled and then switching to DEB.
In this case, the
do_install
and
do_package
task outputs are still valid.
However, with a per-recipe approach, the build would not
include the .deb
files.
Consequently, you would have to invalidate the whole build and
rerun it.
Rerunning everything is not the best solution.
Also, in this case, the core must be "taught" much about
specific tasks.
This methodology does not scale well and does not allow users
to easily add new tasks in layers or as external recipes
without touching the packaged-staging core.
The shared state code uses a checksum, which is a unique signature of a task's inputs, to determine if a task needs to be run again. Because it is a change in a task's inputs that triggers a rerun, the process needs to detect all the inputs to a given task. For shell tasks, this turns out to be fairly easy because the build process generates a "run" shell script for each task and it is possible to create a checksum that gives you a good idea of when the task's data changes.
To complicate the problem, there are things that should not be
included in the checksum.
First, there is the actual specific build path of a given
task - the
WORKDIR
.
It does not matter if the work directory changes because it
should not affect the output for target packages.
Also, the build process has the objective of making native
or cross packages relocatable.
The checksum therefore needs to exclude
WORKDIR
.
The simplistic approach for excluding the work directory is to
set WORKDIR
to some fixed value and
create the checksum for the "run" script.
Another problem results from the "run" scripts containing functions that might or might not get called. The incremental build solution contains code that figures out dependencies between shell functions. This code is used to prune the "run" scripts down to the minimum set, thereby alleviating this problem and making the "run" scripts much more readable as a bonus.
So far, solutions for shell scripts exist. What about Python tasks? The same approach applies even though these tasks are more difficult. The process needs to figure out what variables a Python function accesses and what functions it calls. Again, the incremental build solution contains code that first figures out the variable and function dependencies, and then creates a checksum for the data used as the input to the task.
Like the WORKDIR
case, situations exist
where dependencies should be ignored.
For these situations, you can instruct the build process to
ignore a dependency by using a line like the following:
PACKAGE_ARCHS[vardepsexclude] = "MACHINE"
This example ensures that the
PACKAGE_ARCHS
variable does not depend on the value of
MACHINE
,
even if it does reference it.
Equally, there are cases where you need to add dependencies BitBake is not able to find. You can accomplish this by using a line like the following:
PACKAGE_ARCHS[vardeps] = "MACHINE"
This example explicitly adds the MACHINE
variable as a dependency for
PACKAGE_ARCHS
.
As an example, consider a case with in-line Python where
BitBake is not able to figure out dependencies.
When running in debug mode (i.e. using
-DDD
), BitBake produces output when it
discovers something for which it cannot figure out dependencies.
The Yocto Project team has currently not managed to cover
those dependencies in detail and is aware of the need to fix
this situation.
Thus far, this section has limited discussion to the direct inputs into a task. Information based on direct inputs is referred to as the "basehash" in the code. However, the question of a task's indirect inputs still exits - items already built and present in the Build Directory. The checksum (or signature) for a particular task needs to add the hashes of all the tasks on which the particular task depends. Choosing which dependencies to add is a policy decision. However, the effect is to generate a master checksum that combines the basehash and the hashes of the task's dependencies.
At the code level, a variety of ways exist by which both the basehash and the dependent task hashes can be influenced. Within the BitBake configuration file, you can give BitBake some extra information to help it construct the basehash. The following statement effectively results in a list of global variable dependency excludes (i.e. variables never included in any checksum):
BB_HASHBASE_WHITELIST ?= "TMPDIR FILE PATH PWD BB_TASKHASH BBPATH DL_DIR \ SSTATE_DIR THISDIR FILESEXTRAPATHS FILE_DIRNAME HOME LOGNAME SHELL TERM \ USER FILESPATH STAGING_DIR_HOST STAGING_DIR_TARGET COREBASE PRSERV_HOST \ PRSERV_DUMPDIR PRSERV_DUMPFILE PRSERV_LOCKDOWN PARALLEL_MAKE \ CCACHE_DIR EXTERNAL_TOOLCHAIN CCACHE CCACHE_DISABLE LICENSE_PATH SDKPKGSUFFIX"
The previous example excludes
WORKDIR
since that variable is actually constructed as a path within
TMPDIR
,
which is on the whitelist.
The rules for deciding which hashes of dependent tasks to
include through dependency chains are more complex and are
generally accomplished with a Python function.
The code in meta/lib/oe/sstatesig.py
shows
two examples of this and also illustrates how you can insert
your own policy into the system if so desired.
This file defines the two basic signature generators
OE-Core
uses: "OEBasic" and "OEBasicHash".
By default, a dummy "noop" signature handler is enabled
in BitBake.
This means that behavior is unchanged from previous versions.
OE-Core uses the "OEBasicHash" signature handler by default
through this setting in the bitbake.conf
file:
BB_SIGNATURE_HANDLER ?= "OEBasicHash"
The "OEBasicHash" BB_SIGNATURE_HANDLER
is the same as the "OEBasic" version but adds the task hash to
the
stamp files.
This results in any metadata change that changes the task hash,
automatically causing the task to be run again.
This removes the need to bump
PR
values, and changes to metadata automatically ripple across
the build.
It is also worth noting that the end result of these signature generators is to make some dependency and hash information available to the build. This information includes:
BB_BASEHASH_task-
taskname
:
The base hashes for each task in the recipe.
BB_BASEHASH_
filename
:
taskname
:
The base hashes for each dependent task.
BBHASHDEPS_
filename
:
taskname
:
The task dependencies for each task.
BB_TASKHASH
:
The hash of the currently running task.
Checksums and dependencies, as discussed in the previous section, solve half the problem of supporting a shared state. The other half of the problem is being able to use checksum information during the build and being able to reuse or rebuild specific components.
The
sstate
class is a relatively generic implementation of how to
"capture" a snapshot of a given task.
The idea is that the build process does not care about the
source of a task's output.
Output could be freshly built or it could be downloaded and
unpacked from somewhere.
In other words, the build process does not need to worry about
its origin.
Two types of output exist.
One type is just about creating a directory in
WORKDIR
.
A good example is the output of either
do_install
or
do_package
.
The other type of output occurs when a set of data is merged
into a shared directory tree such as the sysroot.
The Yocto Project team has tried to keep the details of the
implementation hidden in sstate
class.
From a user's perspective, adding shared state wrapping to a
task is as simple as this
do_deploy
example taken from the
deploy
class:
DEPLOYDIR = "${WORKDIR}/deploy-${PN}" SSTATETASKS += "do_deploy" do_deploy[sstate-inputdirs] = "${DEPLOYDIR}" do_deploy[sstate-outputdirs] = "${DEPLOY_DIR_IMAGE}" python do_deploy_setscene () { sstate_setscene(d) } addtask do_deploy_setscene do_deploy[dirs] = "${DEPLOYDIR} ${B}" do_deploy[stamp-extra-info] = "${MACHINE_ARCH}"
The following list explains the previous example:
Adding "do_deploy" to SSTATETASKS
adds some required sstate-related processing, which is
implemented in the
sstate
class, to before and after the
do_deploy
task.
The
do_deploy[sstate-inputdirs] = "${DEPLOYDIR}"
declares that do_deploy
places its
output in ${DEPLOYDIR}
when run
normally (i.e. when not using the sstate cache).
This output becomes the input to the shared state cache.
The
do_deploy[sstate-outputdirs] = "${DEPLOY_DIR_IMAGE}"
line causes the contents of the shared state cache to be
copied to ${DEPLOY_DIR_IMAGE}
.
do_deploy
is not already in
the shared state cache or if its input checksum
(signature) has changed from when the output was
cached, the task runs to populate the shared
state cache, after which the contents of the shared
state cache is copied to
${DEPLOY_DIR_IMAGE}
.
If do_deploy
is in the shared
state cache and its signature indicates that the
cached output is still valid (i.e. if no
relevant task inputs have changed), then the
contents of the shared state cache copies
directly to
${DEPLOY_DIR_IMAGE}
by the
do_deploy_setscene
task
instead, skipping the
do_deploy
task.
The following task definition is glue logic needed to make the previous settings effective:
python do_deploy_setscene () { sstate_setscene(d) } addtask do_deploy_setscene
sstate_setscene()
takes the flags
above as input and accelerates the
do_deploy
task through the
shared state cache if possible.
If the task was accelerated,
sstate_setscene()
returns True.
Otherwise, it returns False, and the normal
do_deploy
task runs.
For more information, see the
"setscene"
section in the BitBake User Manual.
The do_deploy[dirs] = "${DEPLOYDIR} ${B}"
line creates ${DEPLOYDIR}
and
${B}
before the
do_deploy
task runs, and also sets
the current working directory of
do_deploy
to
${B}
.
For more information, see the
"Variable Flags"
section in the BitBake User Manual.
sstate-inputdirs
and
sstate-outputdirs
would be the
same, you can use
sstate-plaindirs
.
For example, to preserve the
${PKGD}
and
${PKGDEST}
output from the
do_package
task, use the following:
do_package[sstate-plaindirs] = "${PKGD} ${PKGDEST}"
The do_deploy[stamp-extra-info] = "${MACHINE_ARCH}"
line appends extra metadata to the
stamp file.
In this case, the metadata makes the task specific
to a machine's architecture.
See
"The Task List"
section in the BitBake User Manual for more
information on the stamp-extra-info
flag.
sstate-inputdirs
and
sstate-outputdirs
can also be used
with multiple directories.
For example, the following declares
PKGDESTWORK
and
SHLIBWORK
as shared state
input directories, which populates the shared state
cache, and PKGDATA_DIR
and
SHLIBSDIR
as the corresponding
shared state output directories:
do_package[sstate-inputdirs] = "${PKGDESTWORK} ${SHLIBSWORKDIR}" do_package[sstate-outputdirs] = "${PKGDATA_DIR} ${SHLIBSDIR}"
These methods also include the ability to take a lockfile when manipulating shared state directory structures, for cases where file additions or removals are sensitive:
do_package[sstate-lockfile] = "${PACKAGELOCK}"
Behind the scenes, the shared state code works by looking in
SSTATE_DIR
and
SSTATE_MIRRORS
for shared state files.
Here is an example:
SSTATE_MIRRORS ?= "\ file://.* http://someserver.tld/share/sstate/PATH;downloadfilename=PATH \n \ file://.* file:///some/local/dir/sstate/PATH"
SSTATE_DIR
) is organized into
two-character subdirectories, where the subdirectory
names are based on the first two characters of the hash.
If the shared state directory structure for a mirror has the
same structure as SSTATE_DIR
, you must
specify "PATH" as part of the URI to enable the build system
to map to the appropriate subdirectory.
The shared state package validity can be detected just by looking at the filename since the filename contains the task checksum (or signature) as described earlier in this section. If a valid shared state package is found, the build process downloads it and uses it to accelerate the task.
The build processes use the *_setscene
tasks for the task acceleration phase.
BitBake goes through this phase before the main execution
code and tries to accelerate any tasks for which it can find
shared state packages.
If a shared state package for a task is available, the
shared state package is used.
This means the task and any tasks on which it is dependent
are not executed.
As a real world example, the aim is when building an IPK-based
image, only the
do_package_write_ipk
tasks would have their shared state packages fetched and
extracted.
Since the sysroot is not used, it would never get extracted.
This is another reason why a task-based approach is preferred
over a recipe-based approach, which would have to install the
output from every task.
The OpenEmbedded build system automatically adds common types of
runtime dependencies between packages, which means that you do not
need to explicitly declare the packages using
RDEPENDS
.
Three automatic mechanisms exist (shlibdeps
,
pcdeps
, and depchains
)
that handle shared libraries, package configuration (pkg-config)
modules, and -dev
and
-dbg
packages, respectively.
For other types of runtime dependencies, you must manually declare
the dependencies.
shlibdeps
:
During the
do_package
task of each recipe, all shared libraries installed by the
recipe are located.
For each shared library, the package that contains the
shared library is registered as providing the shared
library.
More specifically, the package is registered as providing
the
soname
of the library.
The resulting shared-library-to-package mapping
is saved globally in
PKGDATA_DIR
by the
do_packagedata
task.
Simultaneously, all executables and shared libraries
installed by the recipe are inspected to see what shared
libraries they link against.
For each shared library dependency that is found,
PKGDATA_DIR
is queried to
see if some package (likely from a different recipe)
contains the shared library.
If such a package is found, a runtime dependency is added
from the package that depends on the shared library to the
package that contains the library.
The automatically added runtime dependency also
includes a version restriction.
This version restriction specifies that at least the
current version of the package that provides the shared
library must be used, as if
"package
(>= version
)"
had been added to RDEPENDS
.
This forces an upgrade of the package containing the shared
library when installing the package that depends on the
library, if needed.
If you want to avoid a package being registered as
providing a particular shared library (e.g. because the library
is for internal use only), then add the library to
PRIVATE_LIBS
inside the package's recipe.
pcdeps
:
During the do_package
task of each
recipe, all pkg-config modules
(*.pc
files) installed by the recipe
are located.
For each module, the package that contains the module is
registered as providing the module.
The resulting module-to-package mapping is saved globally in
PKGDATA_DIR
by the
do_packagedata
task.
Simultaneously, all pkg-config modules installed by
the recipe are inspected to see what other pkg-config
modules they depend on.
A module is seen as depending on another module if it
contains a "Requires:" line that specifies the other module.
For each module dependency,
PKGDATA_DIR
is queried to see if some
package contains the module.
If such a package is found, a runtime dependency is added
from the package that depends on the module to the package
that contains the module.
pcdeps
mechanism most often
infers dependencies between -dev
packages.
depchains
:
If a package foo
depends on a package
bar
, then foo-dev
and foo-dbg
are also made to depend on
bar-dev
and
bar-dbg
, respectively.
Taking the -dev
packages as an
example, the bar-dev
package might
provide headers and shared library symlinks needed by
foo-dev
, which shows the need
for a dependency between the packages.
The dependencies added by
depchains
are in the form of
RRECOMMENDS
.
foo-dev
also has an
RDEPENDS
-style dependency on
foo
, because the default value of
RDEPENDS_${PN}-dev
(set in
bitbake.conf
) includes
"${PN}".
To ensure that the dependency chain is never broken,
-dev
and -dbg
packages are always generated by default, even if the
packages turn out to be empty.
See the
ALLOW_EMPTY
variable for more information.
The do_package
task depends on the
do_packagedata
task of each recipe in
DEPENDS
through use of a
[
deptask
]
declaration, which guarantees that the required
shared-library/module-to-package mapping information will be available
when needed as long as DEPENDS
has been
correctly set.
Some tasks are easier to implement when allowed to perform certain
operations that are normally reserved for the root user (e.g.
do_install
,
do_package_write*
,
do_rootfs
,
and
do_image*
).
For example, the do_install
task benefits
from being able to set the UID and GID of installed files to
arbitrary values.
One approach to allowing tasks to perform root-only operations would be to require BitBake to run as root. However, this method is cumbersome and has security issues. The approach that is actually used is to run tasks that benefit from root privileges in a "fake" root environment. Within this environment, the task and its child processes believe that they are running as the root user, and see an internally consistent view of the filesystem. As long as generating the final output (e.g. a package or an image) does not require root privileges, the fact that some earlier steps ran in a fake root environment does not cause problems.
The capability to run tasks in a fake root environment is known as "fakeroot", which is derived from the BitBake keyword/variable flag that requests a fake root environment for a task.
In the
OpenEmbedded build system,
the program that implements fakeroot is known as
Pseudo.
Pseudo overrides system calls by using the environment variable
LD_PRELOAD
, which results in the illusion
of running as root.
To keep track of "fake" file ownership and permissions resulting
from operations that require root permissions, Pseudo uses
an SQLite 3 database.
This database is stored in
${
WORKDIR
}/pseudo/files.db
for individual recipes.
Storing the database in a file as opposed to in memory
gives persistence between tasks and builds, which is not
accomplished using fakeroot.
virtual/fakeroot-native:do_populate_sysroot
,
giving the following:
fakeroot do_mytask () { ... } do_mytask[depends] += "virtual/fakeroot-native:do_populate_sysroot"
For more information, see the
FAKEROOT*
variables in the BitBake User Manual.
You can also reference the
"Why Not Fakeroot?"
article for background information on Fakeroot and Pseudo.
Welcome to the Yocto Project Development Tasks Manual! This manual provides relevant procedures necessary for developing in the Yocto Project environment (i.e. developing embedded Linux images and user-space applications that run on targeted devices). The manual groups related procedures into higher-level sections. Procedures can consist of high-level steps or low-level steps depending on the topic.
This manual provides the following:
Procedures that help you get going with the Yocto Project. For example, procedures that show you how to set up a build host and work with the Yocto Project source repositories.
Procedures that show you how to submit changes to the Yocto Project. Changes can be improvements, new features, or bug fixes.
Procedures related to "everyday" tasks you perform while developing images and applications using the Yocto Project. For example, procedures to create a layer, customize an image, write a new recipe, and so forth.
This manual does not provide the following:
Redundant Step-by-step Instructions: For example, the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual contains detailed instructions on how to install an SDK, which is used to develop applications for target hardware.
Reference or Conceptual Material: This type of material resides in an appropriate reference manual. For example, system variables are documented in the Yocto Project Reference Manual.
Detailed Public Information Not Specific to the Yocto Project: For example, exhaustive information on how to use the Source Control Manager Git is better covered with Internet searches and official Git Documentation than through the Yocto Project documentation.
Because this manual presents information for many different topics, supplemental information is recommended for full comprehension. For introductory information on the Yocto Project, see the Yocto Project Website. If you want to build an image with no knowledge of Yocto Project as a way of quickly testing it out, see the Yocto Project Quick Build document.
For a comprehensive list of links and other documentation, see the "Links and Related Documentation" section in the Yocto Project Reference Manual.
This chapter provides guidance on how to prepare to use the Yocto Project. You can learn about creating a team environment that develops using the Yocto Project, how to set up a build host, how to locate Yocto Project source repositories, and how to create local Git repositories.
It might not be immediately clear how you can use the Yocto Project in a team development environment, or how to scale it for a large team of developers. You can adapt the Yocto Project to many different use cases and scenarios; however, this flexibility could cause difficulties if you are trying to create a working setup that scales effectively.
To help you understand how to set up this type of environment, this section presents a procedure that gives you information that can help you get the results you want. The procedure is high-level and presents some of the project's most successful experiences, practices, solutions, and available technologies that have proved to work well in the past; however, keep in mind, the procedure here is simply a starting point. You can build off these steps and customize the procedure to fit any particular working environment and set of practices.
Determine Who is Going to be Developing: You first need to understand who is going to be doing anything related to the Yocto Project and determine their roles. Making this determination is essential to completing subsequent steps, which are to get your equipment together and set up your development environment's hardware topology.
The following roles exist:
Application Developer: This type of developer does application level work on top of an existing software stack.
Core System Developer: This type of developer works on the contents of the operating system image itself.
Build Engineer: This type of developer manages Autobuilders and releases. Depending on the specifics of the environment, not all situations might need a Build Engineer.
Test Engineer: This type of developer creates and manages automated tests that are used to ensure all application and core system development meets desired quality standards.
Gather the Hardware: Based on the size and make-up of the team, get the hardware together. Ideally, any development, build, or test engineer uses a system that runs a supported Linux distribution. These systems, in general, should be high performance (e.g. dual, six-core Xeons with 24 Gbytes of RAM and plenty of disk space). You can help ensure efficiency by having any machines used for testing or that run Autobuilders be as high performance as possible.
Understand the Hardware Topology of the Environment: Once you understand the hardware involved and the make-up of the team, you can understand the hardware topology of the development environment. You can get a visual idea of the machines and their roles across the development environment.
Use Git as Your Source Control Manager (SCM): Keeping your Metadata (i.e. recipes, configuration files, classes, and so forth) and any software you are developing under the control of an SCM system that is compatible with the OpenEmbedded build system is advisable. Of all of the SCMs supported by BitBake, the Yocto Project team strongly recommends using Git. Git is a distributed system that is easy to back up, allows you to work remotely, and then connects back to the infrastructure.
It is relatively easy to set up Git services and create
infrastructure like
http://git.yoctoproject.org,
which is based on server software called
gitolite
with cgit
being used to generate the web interface that lets you view the
repositories.
The gitolite
software identifies users
using SSH keys and allows branch-based access controls to
repositories that you can control as little or as much as
necessary.
Git documentation:
Describes how to install
gitolite
on the server.
Gitolite:
Information for gitolite
.
Interfaces, frontends, and tools: Documentation on how to create interfaces and frontends for Git.
Set up the Application Development Machines: As mentioned earlier, application developers are creating applications on top of existing software stacks. Following are some best practices for setting up machines used for application development:
Use a pre-built toolchain that contains the software stack itself. Then, develop the application code on top of the stack. This method works well for small numbers of relatively isolated applications.
Keep your cross-development toolchains updated.
You can do this through provisioning either as new
toolchain downloads or as updates through a package
update mechanism using opkg
to provide updates to an existing toolchain.
The exact mechanics of how and when to do this depend
on local policy.
Use multiple toolchains installed locally into different locations to allow development across versions.
Set up the Core Development Machines: As mentioned earlier, core developers work on the contents of the operating system itself. Following are some best practices for setting up machines used for developing images:
Have the OpenEmbedded build system available on the developer workstations so developers can run their own builds and directly rebuild the software stack.
Keep the core system unchanged as much as possible and do your work in layers on top of the core system. Doing so gives you a greater level of portability when upgrading to new versions of the core system or Board Support Packages (BSPs).
Share layers amongst the developers of a particular project and contain the policy configuration that defines the project.
Set up an Autobuilder: Autobuilders are often the core of the development environment. It is here that changes from individual developers are brought together and centrally tested. Based on this automated build and test environment, subsequent decisions about releases can be made. Autobuilders also allow for "continuous integration" style testing of software components and regression identification and tracking.
See "Yocto Project Autobuilder" for more information and links to buildbot. The Yocto Project team has found this implementation works well in this role. A public example of this is the Yocto Project Autobuilders, which the Yocto Project team uses to test the overall health of the project.
The features of this system are:
Highlights when commits break the build.
Populates an sstate cache from which developers can pull rather than requiring local builds.
Allows commit hook triggers, which trigger builds when commits are made.
Allows triggering of automated image booting and testing under the QuickEMUlator (QEMU).
Supports incremental build testing and from-scratch builds.
Shares output that allows developer testing and historical regression investigation.
Creates output that can be used for releases.
Allows scheduling of builds so that resources can be used efficiently.
Set up Test Machines: Use a small number of shared, high performance systems for testing purposes. Developers can use these systems for wider, more extensive testing while they continue to develop locally using their primary development system.
Document Policies and Change Flow:
The Yocto Project uses a hierarchical structure and a
pull model.
Scripts exist to create and send pull requests
(i.e. create-pull-request
and
send-pull-request
).
This model is in line with other open source projects where
maintainers are responsible for specific areas of the project
and a single maintainer handles the final "top-of-tree" merges.
gitolite
software supports both the
push and pull models quite easily.
As with any development environment, it is important to document the policy used as well as any main project guidelines so they are understood by everyone. It is also a good idea to have well-structured commit messages, which are usually a part of a project's guidelines. Good commit messages are essential when looking back in time and trying to understand why changes were made.
If you discover that changes are needed to the core layer of the project, it is worth sharing those with the community as soon as possible. Chances are if you have discovered the need for changes, someone else in the community needs them also.
Development Environment Summary: Aside from the previous steps, some best practices exist within the Yocto Project development environment. Consider the following:
Use Git as the source control system.
Maintain your Metadata in layers that make sense for your situation. See the "The Yocto Project Layer Model" section in the Yocto Project Overview and Concepts Manual and the "Understanding and Creating Layers" section for more information on layers.
Separate the project's Metadata and code by using separate Git repositories. See the "Yocto Project Source Repositories" section in the Yocto Project Overview and Concepts Manual for information on these repositories. See the "Locating Yocto Project Source Files" section for information on how to set up local Git repositories for related upstream Yocto Project Git repositories.
Set up the directory for the shared state cache
(SSTATE_DIR
)
where it makes sense.
For example, set up the sstate cache on a system used
by developers in the same organization and share the
same source directories on their machines.
Set up an Autobuilder and have it populate the sstate cache and source directories.
The Yocto Project community encourages you to send patches to the project to fix bugs or add features. If you do submit patches, follow the project commit guidelines for writing good commit messages. See the "Submitting a Change to the Yocto Project" section.
Send changes to the core sooner than later as others are likely to run into the same issues. For some guidance on mailing lists to use, see the list in the "Submitting a Change to the Yocto Project" section. For a description of the available mailing lists, see the "Mailing Lists" section in the Yocto Project Reference Manual.
This section provides procedures to set up a system to be used as your build host for development using the Yocto Project. Your build host can be a native Linux machine (recommended), it can be a machine (Linux, Mac, or Windows) that uses CROPS, which leverages Docker Containers or it can be a Windows machine capable of running Windows Subsystem For Linux v2 (WSL).
Once your build host is set up to use the Yocto Project, further steps are necessary depending on what you want to accomplish. See the following references for information on how to prepare for Board Support Package (BSP) development and kernel development:
BSP Development: See the "Preparing Your Build Host to Work With BSP Layers" section in the Yocto Project Board Support Package (BSP) Developer's Guide.
Kernel Development: See the "Preparing the Build Host to Work on the Kernel" section in the Yocto Project Linux Kernel Development Manual.
Follow these steps to prepare a native Linux machine as your Yocto Project Build Host:
Use a Supported Linux Distribution: You should have a reasonably current Linux-based host system. You will have the best results with a recent release of Fedora, openSUSE, Debian, Ubuntu, RHEL or CentOS as these releases are frequently tested against the Yocto Project and officially supported. For a list of the distributions under validation and their status, see the "Supported Linux Distributions" section in the Yocto Project Reference Manual and the wiki page at Distribution Support.
Have Enough Free Memory: Your system should have at least 50 Gbytes of free disk space for building images.
Meet Minimal Version Requirements: The OpenEmbedded build system should be able to run on any modern distribution that has the following versions for Git, tar, Python and gcc.
Git 1.8.3.1 or greater
tar 1.28 or greater
Python 3.5.0 or greater.
gcc 5.0 or greater.
If your build host does not meet any of these three listed version requirements, you can take steps to prepare the system so that you can still use the Yocto Project. See the "Required Git, tar, Python and gcc Versions" section in the Yocto Project Reference Manual for information.
Install Development Host Packages: Required development host packages vary depending on your build host and what you want to do with the Yocto Project. Collectively, the number of required packages is large if you want to be able to cover all cases.
For lists of required packages for all scenarios, see the "Required Packages for the Build Host" section in the Yocto Project Reference Manual.
Once you have completed the previous steps, you are ready to
continue using a given development path on your native Linux
machine.
If you are going to use BitBake, see the
"Cloning the poky
Repository"
section.
If you are going to use the Extensible SDK, see the
"Using the Extensible SDK"
Chapter in the Yocto Project Application Development and the
Extensible Software Development Kit (eSDK) manual.
If you want to work on the kernel, see the
Yocto Project Linux Kernel Development Manual.
If you are going to use Toaster, see the
"Setting Up and Using Toaster"
section in the Toaster User Manual.
With CROPS, which leverages Docker Containers, you can create a Yocto Project development environment that is operating system agnostic. You can set up a container in which you can develop using the Yocto Project on a Windows, Mac, or Linux machine.
Follow these general steps to prepare a Windows, Mac, or Linux machine as your Yocto Project build host:
Determine What Your Build Host Needs: Docker is a software container platform that you need to install on the build host. Depending on your build host, you might have to install different software to support Docker containers. Go to the Docker installation page and read about the platform requirements in "Supported Platforms" your build host needs to run containers.
Choose What To Install: Depending on whether or not your build host meets system requirements, you need to install "Docker CE Stable" or the "Docker Toolbox". Most situations call for Docker CE. However, if you have a build host that does not meet requirements (e.g. Pre-Windows 10 or Windows 10 "Home" version), you must install Docker Toolbox instead.
Go to the Install Site for Your Platform: Click the link for the Docker edition associated with your build host's native software. For example, if your build host is running Microsoft Windows Version 10 and you want the Docker CE Stable edition, click that link under "Supported Platforms".
Install the Software: Once you have understood all the pre-requisites, you can download and install the appropriate software. Follow the instructions for your specific machine and the type of the software you need to install:
Install Docker CE for Windows for Windows build hosts that meet requirements.
Install Docker CE for Macs for Mac build hosts that meet requirements.
Install Docker Toolbox for Windows for Windows build hosts that do not meet Docker requirements.
Install Docker Toolbox for MacOS for Mac build hosts that do not meet Docker requirements.
Install Docker CE for CentOS for Linux build hosts running the CentOS distribution.
Install Docker CE for Debian for Linux build hosts running the Debian distribution.
Install Docker CE for Fedora for Linux build hosts running the Fedora distribution.
Install Docker CE for Ubuntu for Linux build hosts running the Ubuntu distribution.
Optionally Orient Yourself With Docker: If you are unfamiliar with Docker and the container concept, you can learn more here - https://docs.docker.com/get-started/.
Launch Docker or Docker Toolbox: You should be able to launch Docker or the Docker Toolbox and have a terminal shell on your development host.
Set Up the Containers to Use the Yocto Project: Go to https://github.com/crops/docker-win-mac-docs/wiki and follow the directions for your particular build host (i.e. Linux, Mac, or Windows).
Once you complete the setup instructions for your machine, you have the Poky, Extensible SDK, and Toaster containers available. You can click those links from the page and learn more about using each of those containers.
Once you have a container set up, everything is in place to
develop just as if you were running on a native Linux machine.
If you are going to use the Poky container, see the
"Cloning the poky
Repository"
section.
If you are going to use the Extensible SDK container, see the
"Using the Extensible SDK"
Chapter in the Yocto Project Application Development and the
Extensible Software Development Kit (eSDK) manual.
If you are going to use the Toaster container, see the
"Setting Up and Using Toaster"
section in the Toaster User Manual.
With Windows Subsystem for Linux (WSLv2), you can create a Yocto Project development environment that allows you to build on Windows. You can set up a Linux distribution inside Windows in which you can develop using the Yocto Project.
Follow these general steps to prepare a Windows machine using WSLv2 as your Yocto Project build host:
Make sure your Windows 10 machine is capable of running WSLv2: WSLv2 is only available for Windows 10 builds > 18917. To check which build version you are running, you may open a command prompt on Windows and execute the command "ver".
C:\Users\myuser> ver Microsoft Windows [Version 10.0.19041.153]
If your build is capable of running WSLv2 you may continue, for more information on this subject or instructions on how to upgrade to WSLv2 visit Windows 10 WSLv2
Install the Linux distribution of your choice inside Windows 10: Once you know your version of Windows 10 supports WSLv2, you can install the distribution of your choice from the Microsoft Store. Open the Microsoft Store and search for Linux. While there are several Linux distributions available, the assumption is that your pick will be one of the distributions supported by the Yocto Project as stated on the instructions for using a native Linux host. After making your selection, simply click "Get" to download and install the distribution.
Check your Linux distribution is using WSLv2: Open a Windows PowerShell and run:
C:\WINDOWS\system32> wsl -l -v NAME STATE VERSION *Ubuntu Running 2
Note the version column which says the WSL version being used by your distribution, on compatible systems, this can be changed back at any point in time.
Optionally Orient Yourself on WSL: If you are unfamiliar with WSL, you can learn more here - https://docs.microsoft.com/en-us/windows/wsl/wsl2-about.
Launch your WSL Distibution: From the Windows start menu simply launch your WSL distribution just like any other application.
Optimize your WSLv2 storage often: Due to the way storage is handled on WSLv2, the storage space used by the undelying Linux distribution is not reflected immedately, and since bitbake heavily uses storage, after several builds, you may be unaware you are running out of space. WSLv2 uses a VHDX file for storage, this issue can be easily avoided by manually optimizing this file often, this can be done in the following way:
Find the location of your VHDX file: First you need to find the distro app package directory, to achieve this open a Windows Powershell as Administrator and run:
C:\WINDOWS\system32> Get-AppxPackage -Name "*Ubuntu*" | Select PackageFamilyName PackageFamilyName ----------------- CanonicalGroupLimited.UbuntuonWindows_79abcdefgh
You should now replace the PackageFamilyName
and your user
on the following
path to find your VHDX file: C:\Users\user\AppData\Local\Packages\PackageFamilyName\LocalState\
For example:
ls C:\Users\myuser\AppData\Local\Packages\CanonicalGroupLimited.UbuntuonWindows_79abcdefgh\LocalState\ Mode LastWriteTime Length Name -a---- 3/14/2020 9:52 PM 57418973184 ext4.vhdx
Your VHDX file path is: C:\Users\myuser\AppData\Local\Packages\CanonicalGroupLimited.UbuntuonWindows_79abcdefgh\LocalState\ext4.vhdx
Optimize your VHDX file: Open a Windows Powershell as Administrator to optimize your VHDX file, shutting down WSL first:
C:\WINDOWS\system32> wsl --shutdown C:\WINDOWS\system32> optimize-vhd -Path C:\Users\myuser\AppData\Local\Packages\CanonicalGroupLimited.UbuntuonWindows_79abcdefgh\LocalState\ext4.vhdx -Mode full
A progress bar should be shown while optimizing the VHDX file, and storage should now be reflected correctly on the Windows Explorer.
C:
drive on /mnt/c/
(and others), which
you can use to share deploy artifacts to be later flashed on
hardware through Windows, but your build directory should not
reside inside this mountpoint.
Once you have WSLv2 set up, everything is in place to develop just as if you were running on a native Linux machine. If you are going to use the Extensible SDK container, see the "Using the Extensible SDK" Chapter in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual. If you are going to use the Toaster container, see the "Setting Up and Using Toaster" section in the Toaster User Manual.
This section shows you how to locate, fetch and configure the source files you'll need to work with the Yocto Project.
For concepts and introductory information about Git as it is used in the Yocto Project, see the "Git" section in the Yocto Project Overview and Concepts Manual.
For concepts on Yocto Project source repositories, see the "Yocto Project Source Repositories" section in the Yocto Project Overview and Concepts Manual."
Working from a copy of the upstream Yocto Project
Source Repositories
is the preferred method for obtaining and using a Yocto Project
release.
You can view the Yocto Project Source Repositories at
http://git.yoctoproject.org.
In particular, you can find the
poky
repository at
http://git.yoctoproject.org/cgit/cgit.cgi/poky/.
Use the following procedure to locate the latest upstream copy of
the poky
Git repository:
Access Repositories: Open a browser and go to http://git.yoctoproject.org to access the GUI-based interface into the Yocto Project source repositories.
Select the Repository:
Click on the repository in which you are interested (e.g.
poky
).
Find the URL Used to Clone the Repository:
At the bottom of the page, note the URL used to
clone
that repository (e.g.
http://git.yoctoproject.org/poky
).
poky
Repository"
section.
Yocto Project maintains an Index of Releases area that contains related files that contribute to the Yocto Project. Rather than Git repositories, these files are tarballs that represent snapshots in time of a given component.
Follow these steps to locate and download a particular tarball:
Access the Index of Releases:
Open a browser and go to
http://downloads.yoctoproject.org/releases to access the
Index of Releases.
The list represents released components (e.g.
bitbake
,
sato
, and so on).
yocto
directory contains the
full array of released Poky tarballs.
The poky
directory in the
Index of Releases was historically used for very
early releases and exists now only for retroactive
completeness.
Select a Component:
Click on any released component in which you are interested
(e.g. yocto
).
Find the Tarball:
Drill down to find the associated tarball.
For example, click on yocto-3.1.2
to
view files associated with the Yocto Project 3.1.2
release (e.g. poky-dunfell-23.0.2.tar.bz2
,
which is the released Poky tarball).
Download the Tarball: Click the tarball to download and save a snapshot of the given component.
The Yocto Project Website uses a "DOWNLOADS" page from which you can locate and download tarballs of any Yocto Project release. Rather than Git repositories, these files represent snapshot tarballs similar to the tarballs located in the Index of Releases described in the "Accessing Index of Releases" section.
Go to the Yocto Project Website: Open The Yocto Project Website in your browser.
Get to the Downloads Area: Select the "DOWNLOADS" item from the pull-down "SOFTWARE" tab menu near the top of the page.
Select a Yocto Project Release: Use the menu next to "RELEASE" to display and choose a recent or past supported Yocto Project release (e.g. dunfell, zeus, and so forth).
You can use the "RELEASE ARCHIVE" link to reveal a menu of all Yocto Project releases.
Download Tools or Board Support Packages (BSPs): From the "DOWNLOADS" page, you can download tools or BSPs as well. Just scroll down the page and look for what you need.
Yocto Project maintains an area for nightly builds that contains tarball releases at https://autobuilder.yocto.io//pub/nightly/. These builds include Yocto Project releases ("poky"), toolchains, and builds for supported machines.
Should you ever want to access a nightly build of a particular Yocto Project component, use the following procedure:
Locate the Index of Nightly Builds: Open a browser and go to https://autobuilder.yocto.io//pub/nightly/ to access the Nightly Builds.
Select a Date: Click on the date in which you are interested. If you want the latest builds, use "CURRENT".
Select a Build: Choose the area in which you are interested. For example, if you are looking for the most recent toolchains, select the "toolchain" link.
Find the Tarball: Drill down to find the associated tarball.
Download the Tarball: Click the tarball to download and save a snapshot of the given component.
To use the Yocto Project for development, you need a release locally installed on your development system. This locally installed set of files is referred to as the Source Directory in the Yocto Project documentation.
The preferred method of creating your Source Directory is by using
Git to clone a local
copy of the upstream poky
repository.
Working from a cloned copy of the upstream repository allows you
to contribute back into the Yocto Project or to simply work with
the latest software on a development branch.
Because Git maintains and creates an upstream repository with
a complete history of changes and you are working with a local
clone of that repository, you have access to all the Yocto
Project development branches and tag names used in the upstream
repository.
poky
Repository¶
Follow these steps to create a local version of the
upstream
poky
Git repository.
Set Your Directory:
Change your working directory to where you want to
create your local copy of
poky
.
Clone the Repository:
The following example command clones the
poky
repository and uses
the default name "poky" for your local repository:
$ git clone git://git.yoctoproject.org/poky Cloning into 'poky'... remote: Counting objects: 432160, done. remote: Compressing objects: 100% (102056/102056), done. remote: Total 432160 (delta 323116), reused 432037 (delta 323000) Receiving objects: 100% (432160/432160), 153.81 MiB | 8.54 MiB/s, done. Resolving deltas: 100% (323116/323116), done. Checking connectivity... done.
Unless you specify a specific development branch or tag name, Git clones the "master" branch, which results in a snapshot of the latest development changes for "master". For information on how to check out a specific development branch or on how to check out a local branch based on a tag name, see the "Checking Out By Branch in Poky" and Checking Out By Tag in Poky" sections, respectively.
Once the local repository is created, you can change to that directory and check its status. Here, the single "master" branch exists on your system and by default, it is checked out:
$ cd ~/poky $ git status On branch master Your branch is up-to-date with 'origin/master'. nothing to commit, working directory clean $ git branch * master
Your local repository of poky is identical to the
upstream poky repository at the time from which it was
cloned.
As you work with the local branch, you can periodically
use the git pull ‐‐rebase
command to be sure you are up-to-date with the upstream
branch.
When you clone the upstream poky repository, you have access to all its development branches. Each development branch in a repository is unique as it forks off the "master" branch. To see and use the files of a particular development branch locally, you need to know the branch name and then specifically check out that development branch.
Switch to the Poky Directory:
If you have a local poky Git repository, switch to that
directory.
If you do not have the local copy of poky, see the
"Cloning the poky
Repository"
section.
Determine Existing Branch Names:
$ git branch -a * master remotes/origin/1.1_M1 remotes/origin/1.1_M2 remotes/origin/1.1_M3 remotes/origin/1.1_M4 remotes/origin/1.2_M1 remotes/origin/1.2_M2 remotes/origin/1.2_M3 . . . remotes/origin/thud remotes/origin/thud-next remotes/origin/warrior remotes/origin/warrior-next remotes/origin/zeus remotes/origin/zeus-next ... and so on ...
Check out the Branch: Check out the development branch in which you want to work. For example, to access the files for the Yocto Project 3.1.2 Release (Dunfell), use the following command:
$ git checkout -b dunfell origin/dunfell Branch dunfell set up to track remote branch dunfell from origin. Switched to a new branch 'dunfell'
The previous command checks out the "dunfell" development branch and reports that the branch is tracking the upstream "origin/dunfell" branch.
The following command displays the branches that are now part of your local poky repository. The asterisk character indicates the branch that is currently checked out for work:
$ git branch master * dunfell
Similar to branches, the upstream repository uses tags to mark specific commits associated with significant points in a development branch (i.e. a release point or stage of a release). You might want to set up a local branch based on one of those points in the repository. The process is similar to checking out by branch name except you use tag names.
Switch to the Poky Directory:
If you have a local poky Git repository, switch to that
directory.
If you do not have the local copy of poky, see the
"Cloning the poky
Repository"
section.
Fetch the Tag Names: To checkout the branch based on a tag name, you need to fetch the upstream tags into your local repository:
$ git fetch --tags $
List the Tag Names: You can list the tag names now:
$ git tag 1.1_M1.final 1.1_M1.rc1 1.1_M1.rc2 1.1_M2.final 1.1_M2.rc1 . . . yocto-2.5 yocto-2.5.1 yocto-2.5.2 yocto-2.5.3 yocto-2.6 yocto-2.6.1 yocto-2.6.2 yocto-2.7 yocto_1.5_M5.rc8
Check out the Branch:
$ git checkout tags/yocto-3.1.2 -b my_yocto_3.1.2 Switched to a new branch 'my_yocto_3.1.2' $ git branch master * my_yocto_3.1.2
The previous command creates and checks out a local
branch named "my_yocto_3.1.2", which is based on
the commit in the upstream poky repository that has
the same tag.
In this example, the files you have available locally
as a result of the checkout
command are a snapshot of the
"dunfell" development branch at the point
where Yocto Project 3.1.2 was released.
bitbake-layers
Scriptbitbake-layers
Scriptbmaptool
oe-pkgdata-util
This chapter describes fundamental procedures such as creating layers, adding new software packages, extending or customizing images, porting work to new hardware (adding a new machine), and so forth. You will find that the procedures documented here occur often in the development cycle using the Yocto Project.
The OpenEmbedded build system supports organizing Metadata into multiple layers. Layers allow you to isolate different types of customizations from each other. For introductory information on the Yocto Project Layer Model, see the "The Yocto Project Layer Model" section in the Yocto Project Overview and Concepts Manual.
It is very easy to create your own layers to use with the
OpenEmbedded build system.
The Yocto Project ships with tools that speed up creating
layers.
This section describes the steps you perform by hand to create
layers so that you can better understand them.
For information about the layer-creation tools, see the
"Creating a New BSP Layer Using the bitbake-layers
Script"
section in the Yocto Project Board Support Package (BSP)
Developer's Guide and the
"Creating a General Layer Using the bitbake-layers
Script"
section further down in this manual.
Follow these general steps to create your layer without using tools:
Check Existing Layers: Before creating a new layer, you should be sure someone has not already created a layer containing the Metadata you need. You can see the OpenEmbedded Metadata Index for a list of layers from the OpenEmbedded community that can be used in the Yocto Project. You could find a layer that is identical or close to what you need.
Create a Directory:
Create the directory for your layer.
When you create the layer, be sure to create the
directory in an area not associated with the
Yocto Project
Source Directory
(e.g. the cloned poky
repository).
While not strictly required, prepend the name of the directory with the string "meta-". For example:
meta-mylayer meta-GUI_xyz meta-mymachine
With rare exceptions, a layer's name follows this form:
meta-root_name
Following this layer naming convention can save you trouble later when tools, components, or variables "assume" your layer name begins with "meta-". A notable example is in configuration files as shown in the following step where layer names without the "meta-" string are appended to several variables used in the configuration.
Create a Layer Configuration File:
Inside your new layer folder, you need to create a
conf/layer.conf
file.
It is easiest to take an existing layer configuration
file and copy that to your layer's
conf
directory and then modify the
file as needed.
The
meta-yocto-bsp/conf/layer.conf
file
in the Yocto Project
Source Repositories
demonstrates the required syntax.
For your layer, you need to replace "yoctobsp" with
a unique identifier for your layer (e.g. "machinexyz"
for a layer named "meta-machinexyz"):
# We have a conf and classes directory, add to BBPATH BBPATH .= ":${LAYERDIR}" # We have recipes-* directories, add to BBFILES BBFILES += "${LAYERDIR}/recipes-*/*/*.bb \ ${LAYERDIR}/recipes-*/*/*.bbappend" BBFILE_COLLECTIONS += "yoctobsp" BBFILE_PATTERN_yoctobsp = "^${LAYERDIR}/" BBFILE_PRIORITY_yoctobsp = "5" LAYERVERSION_yoctobsp = "4" LAYERSERIES_COMPAT_yoctobsp = "dunfell"
Following is an explanation of the layer configuration file:
BBPATH
:
Adds the layer's root directory to BitBake's
search path.
Through the use of the
BBPATH
variable, BitBake
locates class files
(.bbclass
),
configuration files, and files that are
included with include
and
require
statements.
For these cases, BitBake uses the first file
that matches the name found in
BBPATH
.
This is similar to the way the
PATH
variable is used for
binaries.
It is recommended, therefore, that you use
unique class and configuration filenames in
your custom layer.
BBFILES
:
Defines the location for all recipes in the
layer.
BBFILE_COLLECTIONS
:
Establishes the current layer through a
unique identifier that is used throughout the
OpenEmbedded build system to refer to the layer.
In this example, the identifier "yoctobsp" is
the representation for the container layer
named "meta-yocto-bsp".
BBFILE_PATTERN
:
Expands immediately during parsing to
provide the directory of the layer.
BBFILE_PRIORITY
:
Establishes a priority to use for
recipes in the layer when the OpenEmbedded build
finds recipes of the same name in different
layers.
LAYERVERSION
:
Establishes a version number for the layer.
You can use this version number to specify this
exact version of the layer as a dependency when
using the
LAYERDEPENDS
variable.
LAYERDEPENDS
:
Lists all layers on which this layer depends (if any).
LAYERSERIES_COMPAT
:
Lists the
Yocto Project
releases for which the current version is
compatible.
This variable is a good way to indicate if
your particular layer is current.
Add Content:
Depending on the type of layer, add the content.
If the layer adds support for a machine, add the machine
configuration in a conf/machine/
file within the layer.
If the layer adds distro policy, add the distro
configuration in a conf/distro/
file within the layer.
If the layer introduces new recipes, put the recipes
you need in recipes-*
subdirectories within the layer.
Optionally Test for Compatibility: If you want permission to use the Yocto Project Compatibility logo with your layer or application that uses your layer, perform the steps to apply for compatibility. See the "Making Sure Your Layer is Compatible With Yocto Project" section for more information.
To create layers that are easier to maintain and that will not impact builds for other machines, you should consider the information in the following list:
Avoid "Overlaying" Entire Recipes from Other Layers in Your Configuration:
In other words, do not copy an entire recipe into your
layer and then modify it.
Rather, use an append file
(.bbappend
) to override only those
parts of the original recipe you need to modify.
Avoid Duplicating Include Files:
Use append files (.bbappend
)
for each recipe that uses an include file.
Or, if you are introducing a new recipe that requires
the included file, use the path relative to the
original layer directory to refer to the file.
For example, use
require recipes-core/
package
/
file
.inc
instead of
require
file
.inc
.
If you're finding you have to overlay the include file,
it could indicate a deficiency in the include file in
the layer to which it originally belongs.
If this is the case, you should try to address that
deficiency instead of overlaying the include file.
For example, you could address this by getting the
maintainer of the include file to add a variable or
variables to make it easy to override the parts needing
to be overridden.
Structure Your Layers: Proper use of overrides within append files and placement of machine-specific files within your layer can ensure that a build is not using the wrong Metadata and negatively impacting a build for a different machine. Following are some examples:
Modify Variables to Support a
Different Machine:
Suppose you have a layer named
meta-one
that adds support
for building machine "one".
To do so, you use an append file named
base-files.bbappend
and
create a dependency on "foo" by altering the
DEPENDS
variable:
DEPENDS = "foo"
The dependency is created during any build that
includes the layer
meta-one
.
However, you might not want this dependency
for all machines.
For example, suppose you are building for
machine "two" but your
bblayers.conf
file has the
meta-one
layer included.
During the build, the
base-files
for machine
"two" will also have the dependency on
foo
.
To make sure your changes apply only when
building machine "one", use a machine override
with the DEPENDS
statement:
DEPENDS_one = "foo"
You should follow the same strategy when using
_append
and
_prepend
operations:
DEPENDS_append_one = " foo" DEPENDS_prepend_one = "foo "
As an actual example, here's a snippet from the
generic kernel include file
linux-yocto.inc
,
wherein the kernel compile and link options are
adjusted in the case of a subset of the supported
architectures:
DEPENDS_append_aarch64 = " libgcc" KERNEL_CC_append_aarch64 = " ${TOOLCHAIN_OPTIONS}" KERNEL_LD_append_aarch64 = " ${TOOLCHAIN_OPTIONS}" DEPENDS_append_nios2 = " libgcc" KERNEL_CC_append_nios2 = " ${TOOLCHAIN_OPTIONS}" KERNEL_LD_append_nios2 = " ${TOOLCHAIN_OPTIONS}" DEPENDS_append_arc = " libgcc" KERNEL_CC_append_arc = " ${TOOLCHAIN_OPTIONS}" KERNEL_LD_append_arc = " ${TOOLCHAIN_OPTIONS}" KERNEL_FEATURES_append_qemuall=" features/debug/printk.scc"
_append
and _prepend
operations
is recommended as well.
Place Machine-Specific Files in
Machine-Specific Locations:
When you have a base recipe, such as
base-files.bb
, that
contains a
SRC_URI
statement to a file, you can use an append file
to cause the build to use your own version of
the file.
For example, an append file in your layer at
meta-one/recipes-core/base-files/base-files.bbappend
could extend
FILESPATH
using
FILESEXTRAPATHS
as follows:
FILESEXTRAPATHS_prepend := "${THISDIR}/${BPN}:"
The build for machine "one" will pick up your
machine-specific file as long as you have the
file in
meta-one/recipes-core/base-files/base-files/
.
However, if you are building for a different
machine and the
bblayers.conf
file includes
the meta-one
layer and
the location of your machine-specific file is
the first location where that file is found
according to FILESPATH
,
builds for all machines will also use that
machine-specific file.
You can make sure that a machine-specific
file is used for a particular machine by putting
the file in a subdirectory specific to the
machine.
For example, rather than placing the file in
meta-one/recipes-core/base-files/base-files/
as shown above, put it in
meta-one/recipes-core/base-files/base-files/one/
.
Not only does this make sure the file is used
only when building for machine "one", but the
build process locates the file more quickly.
In summary, you need to place all files
referenced from SRC_URI
in a machine-specific subdirectory within the
layer in order to restrict those files to
machine-specific builds.
Perform Steps to Apply for Yocto Project Compatibility: If you want permission to use the Yocto Project Compatibility logo with your layer or application that uses your layer, perform the steps to apply for compatibility. See the "Making Sure Your Layer is Compatible With Yocto Project" section for more information.
Follow the Layer Naming Convention:
Store custom layers in a Git repository that use the
meta-
format.
layer_name
Group Your Layers Locally:
Clone your repository alongside other cloned
meta
directories from the
Source Directory.
When you create a layer used with the Yocto Project, it is advantageous to make sure that the layer interacts well with existing Yocto Project layers (i.e. the layer is compatible with the Yocto Project). Ensuring compatibility makes the layer easy to be consumed by others in the Yocto Project community and could allow you permission to use the Yocto Project Compatible Logo.
The Yocto Project Compatibility Program consists of a layer application process that requests permission to use the Yocto Project Compatibility Logo for your layer and application. The process consists of two parts:
Successfully passing a script
(yocto-check-layer
) that
when run against your layer, tests it against
constraints based on experiences of how layers have
worked in the real world and where pitfalls have been
found.
Getting a "PASS" result from the script is required for
successful compatibility registration.
Completion of an application acceptance form, which you can find at https://www.yoctoproject.org/webform/yocto-project-compatible-registration.
To be granted permission to use the logo, you need to satisfy the following:
Be able to check the box indicating that you got a "PASS" when running the script against your layer.
Answer "Yes" to the questions on the form or have an acceptable explanation for any questions answered "No".
Be a Yocto Project Member Organization.
The remainder of this section presents information on the
registration form and on the
yocto-check-layer
script.
Use the form to apply for your layer's approval. Upon successful application, you can use the Yocto Project Compatibility Logo with your layer and the application that uses your layer.
To access the form, use this link: https://www.yoctoproject.org/webform/yocto-project-compatible-registration. Follow the instructions on the form to complete your application.
The application consists of the following sections:
Contact Information: Provide your contact information as the fields require. Along with your information, provide the released versions of the Yocto Project for which your layer is compatible.
Acceptance Criteria: Provide "Yes" or "No" answers for each of the items in the checklist. Space exists at the bottom of the form for any explanations for items for which you answered "No".
Recommendations: Provide answers for the questions regarding Linux kernel use and build success.
yocto-check-layer
Script¶
The yocto-check-layer
script
provides you a way to assess how compatible your layer is
with the Yocto Project.
You should run this script prior to using the form to
apply for compatibility as described in the previous
section.
You need to achieve a "PASS" result in order to have
your application form successfully processed.
The script divides tests into three areas: COMMON, BSP, and DISTRO. For example, given a distribution layer (DISTRO), the layer must pass both the COMMON and DISTRO related tests. Furthermore, if your layer is a BSP layer, the layer must pass the COMMON and BSP set of tests.
To execute the script, enter the following commands from your build directory:
$ source oe-init-build-env
$ yocto-check-layer your_layer_directory
Be sure to provide the actual directory for your layer as part of the command.
Entering the command causes the script to determine the type of layer and then to execute a set of specific tests against the layer. The following list overviews the test:
common.test_readme
:
Tests if a README
file
exists in the layer and the file is not empty.
common.test_parse
:
Tests to make sure that BitBake can parse the
files without error (i.e.
bitbake -p
).
common.test_show_environment
:
Tests that the global or per-recipe environment
is in order without errors (i.e.
bitbake -e
).
common.test_world
:
Verifies that bitbake world
works.
common.test_signatures
:
Tests to be sure that BSP and DISTRO layers do not
come with recipes that change signatures.
common.test_layerseries_compat
:
Verifies layer compatibility is set properly.
bsp.test_bsp_defines_machines
:
Tests if a BSP layer has machine configurations.
bsp.test_bsp_no_set_machine
:
Tests to ensure a BSP layer does not set the
machine when the layer is added.
bsp.test_machine_world
:
Verifies that bitbake world
works regardless of which machine is selected.
bsp.test_machine_signatures
:
Verifies that building for a particular machine
affects only the signature of tasks specific to that
machine.
distro.test_distro_defines_distros
:
Tests if a DISTRO layer has distro configurations.
distro.test_distro_no_set_distros
:
Tests to ensure a DISTRO layer does not set the
distribution when the layer is added.
Before the OpenEmbedded build system can use your new layer,
you need to enable it.
To enable your layer, simply add your layer's path to the
BBLAYERS
variable in your conf/bblayers.conf
file,
which is found in the
Build Directory.
The following example shows how to enable a layer named
meta-mylayer
:
# POKY_BBLAYERS_CONF_VERSION is increased each time build/conf/bblayers.conf # changes incompatibly POKY_BBLAYERS_CONF_VERSION = "2" BBPATH = "${TOPDIR}" BBFILES ?= "" BBLAYERS ?= " \ /home/user
/poky/meta \ /home/user
/poky/meta-poky \ /home/user
/poky/meta-yocto-bsp \ /home/user
/poky/meta-mylayer \ "
BitBake parses each conf/layer.conf
file
from the top down as specified in the
BBLAYERS
variable
within the conf/bblayers.conf
file.
During the processing of each
conf/layer.conf
file, BitBake adds the
recipes, classes and configurations contained within the
particular layer to the source directory.
A recipe that appends Metadata to another recipe is called a
BitBake append file.
A BitBake append file uses the .bbappend
file type suffix, while the corresponding recipe to which
Metadata is being appended uses the .bb
file type suffix.
You can use a .bbappend
file in your
layer to make additions or changes to the content of another
layer's recipe without having to copy the other layer's
recipe into your layer.
Your .bbappend
file resides in your layer,
while the main .bb
recipe file to
which you are appending Metadata resides in a different layer.
Being able to append information to an existing recipe not only avoids duplication, but also automatically applies recipe changes from a different layer into your layer. If you were copying recipes, you would have to manually merge changes as they occur.
When you create an append file, you must use the same root
name as the corresponding recipe file.
For example, the append file
someapp_3.1.2.bbappend
must apply to
someapp_3.1.2.bb
.
This means the original recipe and append file names are
version number-specific.
If the corresponding recipe is renamed to update to a newer
version, you must also rename and possibly update
the corresponding .bbappend
as well.
During the build process, BitBake displays an error on starting
if it detects a .bbappend
file that does
not have a corresponding recipe with a matching name.
See the
BB_DANGLINGAPPENDS_WARNONLY
variable for information on how to handle this error.
As an example, consider the main formfactor recipe and a
corresponding formfactor append file both from the
Source Directory.
Here is the main formfactor recipe, which is named
formfactor_0.0.bb
and located in the
"meta" layer at
meta/recipes-bsp/formfactor
:
SUMMARY = "Device formfactor information" SECTION = "base" LICENSE = "MIT" LIC_FILES_CHKSUM = "file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420" PR = "r45" SRC_URI = "file://config file://machconfig" S = "${WORKDIR}" PACKAGE_ARCH = "${MACHINE_ARCH}" INHIBIT_DEFAULT_DEPS = "1" do_install() { # Install file only if it has contents install -d ${D}${sysconfdir}/formfactor/ install -m 0644 ${S}/config ${D}${sysconfdir}/formfactor/ if [ -s "${S}/machconfig" ]; then install -m 0644 ${S}/machconfig ${D}${sysconfdir}/formfactor/ fi }
In the main recipe, note the
SRC_URI
variable, which tells the OpenEmbedded build system where to
find files during the build.
Following is the append file, which is named
formfactor_0.0.bbappend
and is from the
Raspberry Pi BSP Layer named
meta-raspberrypi
.
The file is in the layer at
recipes-bsp/formfactor
:
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
By default, the build system uses the
FILESPATH
variable to locate files.
This append file extends the locations by setting the
FILESEXTRAPATHS
variable.
Setting this variable in the .bbappend
file is the most reliable and recommended method for adding
directories to the search path used by the build system
to find files.
The statement in this example extends the directories to
include
${
THISDIR
}/${
PN
}
,
which resolves to a directory named
formfactor
in the same directory
in which the append file resides (i.e.
meta-raspberrypi/recipes-bsp/formfactor
.
This implies that you must have the supporting directory
structure set up that will contain any files or patches you
will be including from the layer.
Using the immediate expansion assignment operator
:=
is important because of the reference
to THISDIR
.
The trailing colon character is important as it ensures that
items in the list remain colon-separated.
BitBake automatically defines the
THISDIR
variable.
You should never set this variable yourself.
Using "_prepend" as part of the
FILESEXTRAPATHS
ensures your path
will be searched prior to other paths in the final
list.
Also, not all append files add extra files.
Many append files simply exist to add build options
(e.g. systemd
).
For these cases, your append file would not even
use the FILESEXTRAPATHS
statement.
Each layer is assigned a priority value.
Priority values control which layer takes precedence if there
are recipe files with the same name in multiple layers.
For these cases, the recipe file from the layer with a higher
priority number takes precedence.
Priority values also affect the order in which multiple
.bbappend
files for the same recipe are
applied.
You can either specify the priority manually, or allow the
build system to calculate it based on the layer's dependencies.
To specify the layer's priority manually, use the
BBFILE_PRIORITY
variable and append the layer's root name:
BBFILE_PRIORITY_mylayer = "1"
It is possible for a recipe with a lower version number
PV
in a layer that has a higher priority to take precedence.
Also, the layer priority does not currently affect the
precedence order of .conf
or .bbclass
files.
Future versions of BitBake might address this.
You can use the BitBake layer management tool
bitbake-layers
to provide a view
into the structure of recipes across a multi-layer project.
Being able to generate output that reports on configured layers
with their paths and priorities and on
.bbappend
files and their applicable
recipes can help to reveal potential problems.
For help on the BitBake layer management tool, use the following command:
$ bitbake-layers --help NOTE: Starting bitbake server... usage: bitbake-layers [-d] [-q] [-F] [--color COLOR] [-h] <subcommand> ... BitBake layers utility optional arguments: -d, --debug Enable debug output -q, --quiet Print only errors -F, --force Force add without recipe parse verification --color COLOR Colorize output (where COLOR is auto, always, never) -h, --help show this help message and exit subcommands: <subcommand> show-layers show current configured layers. show-overlayed list overlayed recipes (where the same recipe exists in another layer) show-recipes list available recipes, showing the layer they are provided by show-appends list bbappend files and recipe files they apply to show-cross-depends Show dependencies between recipes that cross layer boundaries. add-layer Add one or more layers to bblayers.conf. remove-layer Remove one or more layers from bblayers.conf. flatten flatten layer configuration into a separate output directory. layerindex-fetch Fetches a layer from a layer index along with its dependent layers, and adds them to conf/bblayers.conf. layerindex-show-depends Find layer dependencies from layer index. create-layer Create a basic layer Use bitbake-layers <subcommand> --help to get help on a specific command
The following list describes the available commands:
help:
Displays general help or help on a specified command.
show-layers:
Shows the current configured layers.
show-overlayed:
Lists overlayed recipes.
A recipe is overlayed when a recipe with the same name
exists in another layer that has a higher layer
priority.
show-recipes:
Lists available recipes and the layers that provide them.
show-appends:
Lists .bbappend
files and the
recipe files to which they apply.
show-cross-depends:
Lists dependency relationships between recipes that
cross layer boundaries.
add-layer:
Adds a layer to bblayers.conf
.
remove-layer:
Removes a layer from bblayers.conf
flatten:
Flattens the layer configuration into a separate output
directory.
Flattening your layer configuration builds a "flattened"
directory that contains the contents of all layers,
with any overlayed recipes removed and any
.bbappend
files appended to the
corresponding recipes.
You might have to perform some manual cleanup of the
flattened layer as follows:
Non-recipe files (such as patches) are overwritten. The flatten command shows a warning for these files.
Anything beyond the normal layer
setup has been added to the
layer.conf
file.
Only the lowest priority layer's
layer.conf
is used.
Overridden and appended items from
.bbappend
files need to be
cleaned up.
The contents of each
.bbappend
end up in the
flattened recipe.
However, if there are appended or changed
variable values, you need to tidy these up
yourself.
Consider the following example.
Here, the bitbake-layers
command adds the line
#### bbappended ...
so that
you know where the following lines originate:
... DESCRIPTION = "A useful utility" ... EXTRA_OECONF = "--enable-something" ... #### bbappended from meta-anotherlayer #### DESCRIPTION = "Customized utility" EXTRA_OECONF += "--enable-somethingelse"
Ideally, you would tidy up these utilities as follows:
... DESCRIPTION = "Customized utility" ... EXTRA_OECONF = "--enable-something --enable-somethingelse" ...
layerindex-fetch
:
Fetches a layer from a layer index, along with its
dependent layers, and adds the layers to the
conf/bblayers.conf
file.
layerindex-show-depends
:
Finds layer dependencies from the layer index.
create-layer
:
Creates a basic layer.
bitbake-layers
Script¶
The bitbake-layers
script with the
create-layer
subcommand simplifies
creating a new general layer.
For information on BSP layers, see the "BSP Layers" section in the Yocto Project Board Specific (BSP) Developer's Guide.
In order to use a layer with the OpenEmbedded
build system, you need to add the layer to your
bblayers.conf
configuration
file.
See the
"Adding a Layer Using the bitbake-layers
Script"
section for more information.
The default mode of the script's operation with this subcommand is to create a layer with the following:
A layer priority of 6.
A conf
subdirectory that contains a
layer.conf
file.
A recipes-example
subdirectory
that contains a further subdirectory named
example
, which contains
an example.bb
recipe file.
A COPYING.MIT
,
which is the license statement for the layer.
The script assumes you want to use the MIT license,
which is typical for most layers, for the contents of
the layer itself.
A README
file, which is a file
describing the contents of your new layer.
In its simplest form, you can use the following command form
to create a layer.
The command creates a layer whose name corresponds to
your_layer_name
in the current
directory:
$ bitbake-layers create-layer your_layer_name
As an example, the following command creates a layer named
meta-scottrif
in your home directory:
$ cd /usr/home $ bitbake-layers create-layer meta-scottrif NOTE: Starting bitbake server... Add your new layer with 'bitbake-layers add-layer meta-scottrif'
If you want to set the priority of the layer to other than the
default value of "6", you can either use the
‐‐priority
option or you can
edit the
BBFILE_PRIORITY
value in the conf/layer.conf
after the
script creates it.
Furthermore, if you want to give the example recipe file
some name other than the default, you can
use the
‐‐example-recipe-name
option.
The easiest way to see how the
bitbake-layers create-layer
command
works is to experiment with the script.
You can also read the usage information by entering the
following:
$ bitbake-layers create-layer --help NOTE: Starting bitbake server... usage: bitbake-layers create-layer [-h] [--priority PRIORITY] [--example-recipe-name EXAMPLERECIPE] layerdir Create a basic layer positional arguments: layerdir Layer directory to create optional arguments: -h, --help show this help message and exit --priority PRIORITY, -p PRIORITY Layer directory to create --example-recipe-name EXAMPLERECIPE, -e EXAMPLERECIPE Filename of the example recipe
bitbake-layers
Script¶
Once you create your general layer, you must add it to your
bblayers.conf
file.
Adding the layer to this configuration file makes the
OpenEmbedded build system aware of your layer so that it can
search it for metadata.
Add your layer by using the
bitbake-layers add-layer
command:
$ bitbake-layers add-layer your_layer_name
Here is an example that adds a layer named
meta-scottrif
to the configuration file.
Following the command that adds the layer is another
bitbake-layers
command that shows the
layers that are in your bblayers.conf
file:
$ bitbake-layers add-layer meta-scottrif NOTE: Starting bitbake server... Parsing recipes: 100% |##########################################################| Time: 0:00:49 Parsing of 1441 .bb files complete (0 cached, 1441 parsed). 2055 targets, 56 skipped, 0 masked, 0 errors. $ bitbake-layers show-layers NOTE: Starting bitbake server... layer path priority ========================================================================== meta /home/scottrif/poky/meta 5 meta-poky /home/scottrif/poky/meta-poky 5 meta-yocto-bsp /home/scottrif/poky/meta-yocto-bsp 5 workspace /home/scottrif/poky/build/workspace 99 meta-scottrif /home/scottrif/poky/build/meta-scottrif 6
Adding the layer to this file enables the build system to locate the layer during the build.
You can customize images to satisfy particular requirements. This section describes several methods and provides guidelines for each.
local.conf
¶
Probably the easiest way to customize an image is to add a
package by way of the local.conf
configuration file.
Because it is limited to local use, this method generally only
allows you to add packages and is not as flexible as creating
your own customized image.
When you add packages using local variables this way, you need
to realize that these variable changes are in effect for every
build and consequently affect all images, which might not
be what you require.
To add a package to your image using the local configuration
file, use the
IMAGE_INSTALL
variable with the _append
operator:
IMAGE_INSTALL_append = " strace"
Use of the syntax is important - specifically, the space between
the quote and the package name, which is
strace
in this example.
This space is required since the _append
operator does not add the space.
Furthermore, you must use _append
instead
of the +=
operator if you want to avoid
ordering issues.
The reason for this is because doing so unconditionally appends
to the variable and avoids ordering problems due to the
variable being set in image recipes and
.bbclass
files with operators like
?=
.
Using _append
ensures the operation takes
affect.
As shown in its simplest use,
IMAGE_INSTALL_append
affects all images.
It is possible to extend the syntax so that the variable
applies to a specific image only.
Here is an example:
IMAGE_INSTALL_append_pn-core-image-minimal = " strace"
This example adds strace
to the
core-image-minimal
image only.
You can add packages using a similar approach through the
CORE_IMAGE_EXTRA_INSTALL
variable.
If you use this variable, only
core-image-*
images are affected.
IMAGE_FEATURES
and
EXTRA_IMAGE_FEATURES
¶
Another method for customizing your image is to enable or
disable high-level image features by using the
IMAGE_FEATURES
and EXTRA_IMAGE_FEATURES
variables.
Although the functions for both variables are nearly equivalent,
best practices dictate using IMAGE_FEATURES
from within a recipe and using
EXTRA_IMAGE_FEATURES
from within
your local.conf
file, which is found in the
Build Directory.
To understand how these features work, the best reference is
meta/classes/core-image.bbclass
.
This class lists out the available
IMAGE_FEATURES
of which most map to
package groups while some, such as
debug-tweaks
and
read-only-rootfs
, resolve as general
configuration settings.
In summary, the file looks at the contents of the
IMAGE_FEATURES
variable and then maps
or configures the feature accordingly.
Based on this information, the build system automatically
adds the appropriate packages or configurations to the
IMAGE_INSTALL
variable.
Effectively, you are enabling extra features by extending the
class or creating a custom class for use with specialized image
.bb
files.
Use the EXTRA_IMAGE_FEATURES
variable
from within your local configuration file.
Using a separate area from which to enable features with
this variable helps you avoid overwriting the features in the
image recipe that are enabled with
IMAGE_FEATURES
.
The value of EXTRA_IMAGE_FEATURES
is added
to IMAGE_FEATURES
within
meta/conf/bitbake.conf
.
To illustrate how you can use these variables to modify your
image, consider an example that selects the SSH server.
The Yocto Project ships with two SSH servers you can use
with 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, the core-image-sato
image
is configured to use Dropbear.
The core-image-full-cmdline
and
core-image-lsb
images both
include OpenSSH.
The core-image-minimal
image does not
contain an SSH server.
You can customize your image and change these defaults.
Edit the IMAGE_FEATURES
variable
in your recipe or use the
EXTRA_IMAGE_FEATURES
in your
local.conf
file so that it configures the
image you are working with to include
ssh-server-dropbear
or
ssh-server-openssh
.
You can also customize an image by creating a custom recipe that defines additional software as part of the image. The following example shows the form for the two lines you need:
IMAGE_INSTALL = "packagegroup-core-x11-base package1 package2" inherit core-image
Defining the software using a custom recipe gives you total
control over the contents of the image.
It is important to use the correct names of packages in the
IMAGE_INSTALL
variable.
You must use the OpenEmbedded notation and not the Debian notation for the names
(e.g. glibc-dev
instead of libc6-dev
).
The other method for creating a custom image is to base it on an existing image.
For example, if you want to create an image based on core-image-sato
but add the additional package strace
to the image,
copy the meta/recipes-sato/images/core-image-sato.bb
to a
new .bb
and add the following line to the end of the copy:
IMAGE_INSTALL += "strace"
For complex custom images, the best approach for customizing
an image is to create a custom package group recipe that is
used to build the image or images.
A good example of a package group recipe is
meta/recipes-core/packagegroups/packagegroup-base.bb
.
If you examine that recipe, you see that the
PACKAGES
variable lists the package group packages to produce.
The inherit packagegroup
statement
sets appropriate default values and automatically adds
-dev
, -dbg
, and
-ptest
complementary packages for each
package specified in the PACKAGES
statement.
inherit packagegroup
line should be
located near the top of the recipe, certainly before
the PACKAGES
statement.
For each package you specify in PACKAGES
,
you can use
RDEPENDS
and
RRECOMMENDS
entries to provide a list of packages the parent task package
should contain.
You can see examples of these further down in the
packagegroup-base.bb
recipe.
Here is a short, fabricated example showing the same basic
pieces for a hypothetical packagegroup defined in
packagegroup-custom.bb
, where the
variable PN
is the standard way to
abbreviate the reference to the full packagegroup name
packagegroup-custom
:
DESCRIPTION = "My Custom Package Groups" inherit packagegroup PACKAGES = "\ ${PN}-apps \ ${PN}-tools \ " RDEPENDS_${PN}-apps = "\ dropbear \ portmap \ psplash" RDEPENDS_${PN}-tools = "\ oprofile \ oprofileui-server \ lttng-tools" RRECOMMENDS_${PN}-tools = "\ kernel-module-oprofile"
In the previous example, two package group packages are created with their dependencies and their
recommended package dependencies listed: packagegroup-custom-apps
, and
packagegroup-custom-tools
.
To build an image using these package group packages, you need to add
packagegroup-custom-apps
and/or
packagegroup-custom-tools
to
IMAGE_INSTALL
.
For other forms of image dependencies see the other areas of this section.
By default, the configured hostname (i.e.
/etc/hostname
) in an image is the
same as the machine name.
For example, if
MACHINE
equals "qemux86", the configured hostname written to
/etc/hostname
is "qemux86".
You can customize this name by altering the value of the
"hostname" variable in the
base-files
recipe using either
an append file or a configuration file.
Use the following in an append file:
hostname="myhostname"
Use the following in a configuration file:
hostname_pn-base-files = "myhostname"
Changing the default value of the variable "hostname" can be useful in certain situations. For example, suppose you need to do extensive testing on an image and you would like to easily identify the image under test from existing images with typical default hostnames. In this situation, you could change the default hostname to "testme", which results in all the images using the name "testme". Once testing is complete and you do not need to rebuild the image for test any longer, you can easily reset the default hostname.
Another point of interest is that if you unset the variable, the image will have no default hostname in the filesystem. Here is an example that unsets the variable in a configuration file:
hostname_pn-base-files = ""
Having no default hostname in the filesystem is suitable for environments that use dynamic hostnames such as virtual machines.
Recipes (.bb
files) are fundamental components
in the Yocto Project environment.
Each software component built by the OpenEmbedded build system
requires a recipe to define the component.
This section describes how to create, write, and test a new
recipe.
The following figure shows the basic process for creating a new recipe. The remainder of the section provides details for the steps.
You can always write a recipe from scratch. However, three choices exist that can help you quickly get a start on a new recipe:
devtool add
:
A command that assists in creating a recipe and
an environment conducive to development.
recipetool create
:
A command provided by the Yocto Project that automates
creation of a base recipe based on the source
files.
Existing Recipes: Location and modification of an existing recipe that is similar in function to the recipe you need.
devtool add
¶
The devtool add
command uses the same
logic for auto-creating the recipe as
recipetool create
, which is listed
below.
Additionally, however, devtool add
sets up an environment that makes it easy for you to
patch the source and to make changes to the recipe as
is often necessary when adding a recipe to build a new
piece of software to be included in a build.
You can find a complete description of the
devtool add
command in the
"A Closer Look at devtool
add"
section in the Yocto Project Application Development
and the Extensible Software Development Kit (eSDK) manual.
recipetool create
¶
recipetool create
automates creation
of a base recipe given a set of source code files.
As long as you can extract or point to the source files,
the tool will construct a recipe and automatically
configure all pre-build information into the recipe.
For example, suppose you have an application that builds
using Autotools.
Creating the base recipe using
recipetool
results in a recipe
that has the pre-build dependencies, license requirements,
and checksums configured.
To run the tool, you just need to be in your
Build Directory
and have sourced the build environment setup script
(i.e.
oe-init-build-env
).
To get help on the tool, use the following command:
$ recipetool -h NOTE: Starting bitbake server... usage: recipetool [-d] [-q] [--color COLOR] [-h] <subcommand> ... OpenEmbedded recipe tool options: -d, --debug Enable debug output -q, --quiet Print only errors --color COLOR Colorize output (where COLOR is auto, always, never) -h, --help show this help message and exit subcommands: create Create a new recipe newappend Create a bbappend for the specified target in the specified layer setvar Set a variable within a recipe appendfile Create/update a bbappend to replace a target file appendsrcfiles Create/update a bbappend to add or replace source files appendsrcfile Create/update a bbappend to add or replace a source file Use recipetool <subcommand> --help to get help on a specific command
Running
recipetool create -o
OUTFILE
creates the base recipe and locates it properly in the
layer that contains your source files.
Following are some syntax examples:
Use this syntax to generate a recipe based on
source
.
Once generated, the recipe resides in the existing source
code layer:
recipetool create -oOUTFILE
source
Use this syntax to generate a recipe using code that you
extract from source
.
The extracted code is placed in its own layer defined
by EXTERNALSRC
.
recipetool create -oOUTFILE
-xEXTERNALSRC
source
Use this syntax to generate a recipe based on
source
.
The options direct recipetool
to
generate debugging information.
Once generated, the recipe resides in the existing source
code layer:
recipetool create -d -oOUTFILE
source
Before writing a recipe from scratch, it is often useful to discover whether someone else has already written one that meets (or comes close to meeting) your needs. The Yocto Project and OpenEmbedded communities maintain many recipes that might be candidates for what you are doing. You can find a good central index of these recipes in the OpenEmbedded Layer Index.
Working from an existing recipe or a skeleton recipe is the best way to get started. Here are some points on both methods:
Locate and modify a recipe that is close to what you want to do: This method works when you are familiar with the current recipe space. The method does not work so well for those new to the Yocto Project or writing recipes.
Some risks associated with this method are using a recipe that has areas totally unrelated to what you are trying to accomplish with your recipe, not recognizing areas of the recipe that you might have to add from scratch, and so forth. All these risks stem from unfamiliarity with the existing recipe space.
Use and modify the following
skeleton recipe:
If for some reason you do not want to use
recipetool
and you cannot
find an existing recipe that is close to meeting
your needs, you can use the following structure to
provide the fundamental areas of a new recipe.
DESCRIPTION = "" HOMEPAGE = "" LICENSE = "" SECTION = "" DEPENDS = "" LIC_FILES_CHKSUM = "" SRC_URI = ""
Once you have your base recipe, you should put it in your own layer and name it appropriately. Locating it correctly ensures that the OpenEmbedded build system can find it when you use BitBake to process the recipe.
Storing Your Recipe:
The OpenEmbedded build system locates your recipe
through the layer's conf/layer.conf
file and the
BBFILES
variable.
This variable sets up a path from which the build system can
locate recipes.
Here is the typical use:
BBFILES += "${LAYERDIR}/recipes-*/*/*.bb \ ${LAYERDIR}/recipes-*/*/*.bbappend"
Consequently, you need to be sure you locate your new recipe inside your layer such that it can be found.
You can find more information on how layers are structured in the "Understanding and Creating Layers" section.
Naming Your Recipe: When you name your recipe, you need to follow this naming convention:
basename
_version
.bb
Use lower-cased characters and do not include the reserved
suffixes -native
,
-cross
, -initial
,
or -dev
casually (i.e. do not use them
as part of your recipe name unless the string applies).
Here are some examples:
cups_1.7.0.bb gawk_4.0.2.bb irssi_0.8.16-rc1.bb
Creating a new recipe is usually an iterative process that requires using BitBake to process the recipe multiple times in order to progressively discover and add information to the recipe file.
Assuming you have sourced the build environment setup script (i.e.
oe-init-build-env
)
and you are in the
Build Directory,
use BitBake to process your recipe.
All you need to provide is the
of the recipe as described
in the previous section:
basename
$ bitbake basename
During the build, the OpenEmbedded build system creates a
temporary work directory for each recipe
(${
WORKDIR
}
)
where it keeps extracted source files, log files, intermediate
compilation and packaging files, and so forth.
The path to the per-recipe temporary work directory depends on the context in which it is being built. The quickest way to find this path is to have BitBake return it by running the following:
$ bitbake -e basename
| grep ^WORKDIR=
As an example, assume a Source Directory top-level folder named
poky
, a default Build Directory at
poky/build
, and a
qemux86-poky-linux
machine target system.
Furthermore, suppose your recipe is named
foo_1.3.0.bb
.
In this case, the work directory the build system uses to
build the package would be as follows:
poky/build/tmp/work/qemux86-poky-linux/foo/1.3.0-r0
Inside this directory you can find sub-directories such as
image
, packages-split
,
and temp
.
After the build, you can examine these to determine how well
the build went.
temp
directory (e.g.
poky/build/tmp/work/qemux86-poky-linux/foo/1.3.0-r0/temp
).
Log files are named log.taskname
(e.g. log.do_configure
,
log.do_fetch
, and
log.do_compile
).
You can find more information about the build process in "The Yocto Project Development Environment" chapter of the Yocto Project Overview and Concepts Manual.
The first thing your recipe must do is specify how to fetch
the source files.
Fetching is controlled mainly through the
SRC_URI
variable.
Your recipe must have a SRC_URI
variable
that points to where the source is located.
For a graphical representation of source locations, see the
"Sources"
section in the Yocto Project Overview and Concepts Manual.
The
do_fetch
task uses the prefix of each entry in the
SRC_URI
variable value to determine which
fetcher
to use to get your source files.
It is the SRC_URI
variable that triggers
the fetcher.
The
do_patch
task uses the variable after source is fetched to apply
patches.
The OpenEmbedded build system uses
FILESOVERRIDES
for scanning directory locations for local files in
SRC_URI
.
The SRC_URI
variable in your recipe must
define each unique location for your source files.
It is good practice to not hard-code version numbers in a URL used
in SRC_URI
.
Rather than hard-code these values, use
${
PV
}
,
which causes the fetch process to use the version specified in
the recipe filename.
Specifying the version in this manner means that upgrading the
recipe to a future version is as simple as renaming the recipe
to match the new version.
Here is a simple example from the
meta/recipes-devtools/strace/strace_5.5.bb
recipe where the source comes from a single tarball.
Notice the use of the
PV
variable:
SRC_URI = "https://strace.io/files/${PV}/strace-${PV}.tar.xz \
Files mentioned in SRC_URI
whose names end
in a typical archive extension (e.g. .tar
,
.tar.gz
, .tar.bz2
,
.zip
, and so forth), are automatically
extracted during the
do_unpack
task.
For another example that specifies these types of files, see
the
"Autotooled Package"
section.
Another way of specifying source is from an SCM.
For Git repositories, you must specify
SRCREV
and you should specify
PV
to include the revision with
SRCPV
.
Here is an example from the recipe
meta/recipes-kernel/blktrace/blktrace_git.bb
:
SRCREV = "d6918c8832793b4205ed3bfede78c2f915c23385" PR = "r6" PV = "1.0.5+git${SRCPV}" SRC_URI = "git://git.kernel.dk/blktrace.git \ file://ldflags.patch"
If your SRC_URI
statement includes
URLs pointing to individual files fetched from a remote server
other than a version control system, BitBake attempts to
verify the files against checksums defined in your recipe to
ensure they have not been tampered with or otherwise modified
since the recipe was written.
Two checksums are used:
SRC_URI[md5sum]
and
SRC_URI[sha256sum]
.
If your SRC_URI
variable points to
more than a single URL (excluding SCM URLs), you need to
provide the md5
and
sha256
checksums for each URL.
For these cases, you provide a name for each URL as part of
the SRC_URI
and then reference that name
in the subsequent checksum statements.
Here is an example combining lines from the files
git.inc
and
git_2.24.1.bb
:
SRC_URI = "${KERNELORG_MIRROR}/software/scm/git/git-${PV}.tar.gz;name=tarball \ ${KERNELORG_MIRROR}/software/scm/git/git-manpages-${PV}.tar.gz;name=manpages" SRC_URI[tarball.md5sum] = "166bde96adbbc11c8843d4f8f4f9811b" SRC_URI[tarball.sha256sum] = "ad5334956301c86841eb1e5b1bb20884a6bad89a10a6762c958220c7cf64da02" SRC_URI[manpages.md5sum] = "31c2272a8979022497ba3d4202df145d" SRC_URI[manpages.sha256sum] = "9a7ae3a093bea39770eb96ca3e5b40bff7af0b9f6123f089d7821d0e5b8e1230"
Proper values for md5
and
sha256
checksums might be available
with other signatures on the download page for the upstream
source (e.g. md5
,
sha1
, sha256
,
GPG
, and so forth).
Because the OpenEmbedded build system only deals with
sha256sum
and md5sum
,
you should verify all the signatures you find by hand.
If no SRC_URI
checksums are specified
when you attempt to build the recipe, or you provide an
incorrect checksum, the build will produce an error for each
missing or incorrect checksum.
As part of the error message, the build system provides
the checksum string corresponding to the fetched file.
Once you have the correct checksums, you can copy and paste
them into your recipe and then run the build again to continue.
This final example is a bit more complicated and is from the
meta/recipes-sato/rxvt-unicode/rxvt-unicode_9.20.bb
recipe.
The example's SRC_URI
statement identifies
multiple files as the source files for the recipe: a tarball, a
patch file, a desktop file, and an icon.
SRC_URI = "http://dist.schmorp.de/rxvt-unicode/Attic/rxvt-unicode-${PV}.tar.bz2 \ file://xwc.patch \ file://rxvt.desktop \ file://rxvt.png"
When you specify local files using the
file://
URI protocol, the build system
fetches files from the local machine.
The path is relative to the
FILESPATH
variable and searches specific directories in a certain order:
${
BP
}
,
${
BPN
}
,
and files
.
The directories are assumed to be subdirectories of the
directory in which the recipe or append file resides.
For another example that specifies these types of files, see the
"Single .c File Package (Hello World!)"
section.
The previous example also specifies a patch file.
Patch files are files whose names usually end in
.patch
or .diff
but
can end with compressed suffixes such as
diff.gz
and
patch.bz2
, for example.
The build system automatically applies patches as described
in the
"Patching Code" section.
During the build, the
do_unpack
task unpacks the source with
${
S
}
pointing to where it is unpacked.
If you are fetching your source files from an upstream source
archived tarball and the tarball's internal structure matches
the common convention of a top-level subdirectory named
${
BPN
}-${
PV
}
,
then you do not need to set S
.
However, if SRC_URI
specifies to fetch
source from an archive that does not use this convention,
or from an SCM like Git or Subversion, your recipe needs to
define S
.
If processing your recipe using BitBake successfully unpacks
the source files, you need to be sure that the directory
pointed to by ${S}
matches the structure
of the source.
Sometimes it is necessary to patch code after it has been
fetched.
Any files mentioned in SRC_URI
whose
names end in .patch
or
.diff
or compressed versions of these
suffixes (e.g. diff.gz
are treated as
patches.
The
do_patch
task automatically applies these patches.
The build system should be able to apply patches with the "-p1"
option (i.e. one directory level in the path will be stripped
off).
If your patch needs to have more directory levels stripped off,
specify the number of levels using the "striplevel" option in
the SRC_URI
entry for the patch.
Alternatively, if your patch needs to be applied in a specific
subdirectory that is not specified in the patch file, use the
"patchdir" option in the entry.
As with all local files referenced in
SRC_URI
using file://
, you should place
patch files in a directory next to the recipe either
named the same as the base name of the recipe
(BP
and
BPN
)
or "files".
Your recipe needs to have both the
LICENSE
and
LIC_FILES_CHKSUM
variables:
LICENSE
:
This variable specifies the license for the software.
If you do not know the license under which the software
you are building is distributed, you should go to the
source code and look for that information.
Typical files containing this information include
COPYING
,
LICENSE
, and
README
files.
You could also find the information near the top of
a source file.
For example, given a piece of software licensed under
the GNU General Public License version 2, you would
set LICENSE
as follows:
LICENSE = "GPLv2"
The licenses you specify within
LICENSE
can have any name as long
as you do not use spaces, since spaces are used as
separators between license names.
For standard licenses, use the names of the files in
meta/files/common-licenses/
or the SPDXLICENSEMAP
flag names
defined in meta/conf/licenses.conf
.
LIC_FILES_CHKSUM
:
The OpenEmbedded build system uses this variable to
make sure the license text has not changed.
If it has, the build produces an error and it affords
you the chance to figure it out and correct the problem.
You need to specify all applicable licensing
files for the software.
At the end of the configuration step, the build process
will compare the checksums of the files to be sure
the text has not changed.
Any differences result in an error with the message
containing the current checksum.
For more explanation and examples of how to set the
LIC_FILES_CHKSUM
variable, see the
"Tracking License Changes"
section.
To determine the correct checksum string, you
can list the appropriate files in the
LIC_FILES_CHKSUM
variable with
incorrect md5 strings, attempt to build the software,
and then note the resulting error messages that will
report the correct md5 strings.
See the
"Fetching Code"
section for additional information.
Here is an example that assumes the software has a
COPYING
file:
LIC_FILES_CHKSUM = "file://COPYING;md5=xxx"
When you try to build the software, the build system will produce an error and give you the correct string that you can substitute into the recipe file for a subsequent build.
Most software packages have a short list of other packages that they require, which are called dependencies. These dependencies fall into two main categories: build-time dependencies, which are required when the software is built; and runtime dependencies, which are required to be installed on the target in order for the software to run.
Within a recipe, you specify build-time dependencies using the
DEPENDS
variable.
Although nuances exist, items specified in
DEPENDS
should be names of other recipes.
It is important that you specify all build-time dependencies
explicitly.
If you do not, due to the parallel nature of BitBake's
execution, you can end up with a race condition where the
dependency is present for one task of a recipe (e.g.
do_configure
)
and then gone when the next task runs (e.g.
do_compile
).
Another consideration is that configure scripts might
automatically check for optional dependencies and enable
corresponding functionality if those dependencies are found.
This behavior means that to ensure deterministic results and
thus avoid more race conditions, you need to either explicitly
specify these dependencies as well, or tell the configure
script explicitly to disable the functionality.
If you wish to make a recipe that is more generally useful
(e.g. publish the recipe in a layer for others to use),
instead of hard-disabling the functionality, you can use the
PACKAGECONFIG
variable to allow functionality and the corresponding
dependencies to be enabled and disabled easily by other
users of the recipe.
Similar to build-time dependencies, you specify runtime
dependencies through a variable -
RDEPENDS
,
which is package-specific.
All variables that are package-specific need to have the name
of the package added to the end as an override.
Since the main package for a recipe has the same name as the
recipe, and the recipe's name can be found through the
${
PN
}
variable, then you specify the dependencies for the main
package by setting RDEPENDS_${PN}
.
If the package were named ${PN}-tools
,
then you would set RDEPENDS_${PN}-tools
,
and so forth.
Some runtime dependencies will be set automatically at packaging time. These dependencies include any shared library dependencies (i.e. if a package "example" contains "libexample" and another package "mypackage" contains a binary that links to "libexample" then the OpenEmbedded build system will automatically add a runtime dependency to "mypackage" on "example"). See the "Automatically Added Runtime Dependencies" section in the Yocto Project Overview and Concepts Manual for further details.
Most software provides some means of setting build-time configuration options before compilation. Typically, setting these options is accomplished by running a configure script with options, or by modifying a build configuration file.
pkg-config
now, which is much more
robust.
You can find a list of the *-config
scripts that are disabled list in the
"Binary Configuration Scripts Disabled"
section in the Yocto Project Reference Manual.
A major part of build-time configuration is about checking for
build-time dependencies and possibly enabling optional
functionality as a result.
You need to specify any build-time dependencies for the
software you are building in your recipe's
DEPENDS
value, in terms of other recipes that satisfy those
dependencies.
You can often find build-time or runtime
dependencies described in the software's documentation.
The following list provides configuration items of note based on how your software is built:
Autotools:
If your source files have a
configure.ac
file, then your
software is built using Autotools.
If this is the case, you just need to worry about
modifying the configuration.
When using Autotools, your recipe needs to inherit
the
autotools
class and your recipe does not have to contain a
do_configure
task.
However, you might still want to make some adjustments.
For example, you can set
EXTRA_OECONF
or
PACKAGECONFIG_CONFARGS
to pass any needed configure options that are specific
to the recipe.
CMake:
If your source files have a
CMakeLists.txt
file, then your
software is built using CMake.
If this is the case, you just need to worry about
modifying the configuration.
When you use CMake, your recipe needs to inherit
the
cmake
class and your recipe does not have to contain a
do_configure
task.
You can make some adjustments by setting
EXTRA_OECMAKE
to pass any needed configure options that are specific
to the recipe.
${D}${datadir}/cmake/
Modules
during
do_install
.
Other:
If your source files do not have a
configure.ac
or
CMakeLists.txt
file, then your
software is built using some method other than Autotools
or CMake.
If this is the case, you normally need to provide a
do_configure
task in your recipe
unless, of course, there is nothing to configure.
Even if your software is not being built by Autotools or CMake, you still might not need to deal with any configuration issues. You need to determine if configuration is even a required step. You might need to modify a Makefile or some configuration file used for the build to specify necessary build options. Or, perhaps you might need to run a provided, custom configure script with the appropriate options.
For the case involving a custom configure
script, you would run
./configure --help
and look for
the options you need to set.
Once configuration succeeds, it is always good practice to
look at the log.do_configure
file to
ensure that the appropriate options have been enabled and no
additional build-time dependencies need to be added to
DEPENDS
.
For example, if the configure script reports that it found
something not mentioned in DEPENDS
, or
that it did not find something that it needed for some
desired optional functionality, then you would need to add
those to DEPENDS
.
Looking at the log might also reveal items being checked for,
enabled, or both that you do not want, or items not being found
that are in DEPENDS
, in which case
you would need to look at passing extra options to the
configure script as needed.
For reference information on configure options specific to the
software you are building, you can consult the output of the
./configure --help
command within
${S}
or consult the software's upstream
documentation.
If your recipe builds an application that needs to
communicate with some device or needs an API into a custom
kernel, you will need to provide appropriate header files.
Under no circumstances should you ever modify the existing
meta/recipes-kernel/linux-libc-headers/linux-libc-headers.inc
file.
These headers are used to build libc
and
must not be compromised with custom or machine-specific
header information.
If you customize libc
through modified
headers all other applications that use
libc
thus become affected.
libc
header file (i.e.
meta/recipes-kernel/linux-libc-headers/linux-libc-headers.inc
).
The correct way to interface to a device or custom kernel is to use a separate package that provides the additional headers for the driver or other unique interfaces. When doing so, your application also becomes responsible for creating a dependency on that specific provider.
Consider the following:
Never modify
linux-libc-headers.inc
.
Consider that file to be part of the
libc
system, and not something
you use to access the kernel directly.
You should access libc
through
specific libc
calls.
Applications that must talk directly to devices should either provide necessary headers themselves, or establish a dependency on a special headers package that is specific to that driver.
For example, suppose you want to modify an existing header that adds I/O control or network support. If the modifications are used by a small number programs, providing a unique version of a header is easy and has little impact. When doing so, bear in mind the guidelines in the previous list.
libc
, and subsequently all
other applications on the system, use a
.bbappend
to modify the
linux-kernel-headers.inc
file.
However, take care to not make the changes
machine specific.
Consider a case where your kernel is older and you need
an older libc
ABI.
The headers installed by your recipe should still be a
standard mainline kernel, not your own custom one.
When you use custom kernel headers you need to get them from
STAGING_KERNEL_DIR
,
which is the directory with kernel headers that are
required to build out-of-tree modules.
Your recipe will also need the following:
do_configure[depends] += "virtual/kernel:do_shared_workdir"
During a build, the do_compile
task
happens after source is fetched, unpacked, and configured.
If the recipe passes through do_compile
successfully, nothing needs to be done.
However, if the compile step fails, you need to diagnose the failure. Here are some common issues that cause failures.
pkg-config
.
See the note in section
"Configuring the Recipe"
for additional information.
Parallel build failures: These failures manifest themselves as intermittent errors, or errors reporting that a file or directory that should be created by some other part of the build process could not be found. This type of failure can occur even if, upon inspection, the file or directory does exist after the build has failed, because that part of the build process happened in the wrong order.
To fix the problem, you need to either satisfy
the missing dependency in the Makefile or whatever
script produced the Makefile, or (as a workaround)
set
PARALLEL_MAKE
to an empty string:
PARALLEL_MAKE = ""
For information on parallel Makefile issues, see the "Debugging Parallel Make Races" section.
Improper host path usage:
This failure applies to recipes building for the target
or nativesdk
only.
The failure occurs when the compilation process uses
improper headers, libraries, or other files from the
host system when cross-compiling for the target.
To fix the problem, examine the
log.do_compile
file to identify
the host paths being used (e.g.
/usr/include
,
/usr/lib
, and so forth) and then
either add configure options, apply a patch, or do both.
Failure to find required
libraries/headers:
If a build-time dependency is missing because it has
not been declared in
DEPENDS
,
or because the dependency exists but the path used by
the build process to find the file is incorrect and the
configure step did not detect it, the compilation
process could fail.
For either of these failures, the compilation process
notes that files could not be found.
In these cases, you need to go back and add additional
options to the configure script as well as possibly
add additional build-time dependencies to
DEPENDS
.
Occasionally, it is necessary to apply a patch
to the source to ensure the correct paths are used.
If you need to specify paths to find files staged
into the sysroot from other recipes, use the variables
that the OpenEmbedded build system provides
(e.g.
STAGING_BINDIR
,
STAGING_INCDIR
,
STAGING_DATADIR
, and so forth).
During do_install
, the task copies the
built files along with their hierarchy to locations that
would mirror their locations on the target device.
The installation process copies files from the
${
S
}
,
${
B
}
,
and
${
WORKDIR
}
directories to the
${
D
}
directory to create the structure as it should appear on the
target system.
How your software is built affects what you must do to be sure your software is installed correctly. The following list describes what you must do for installation depending on the type of build system used by the software being built:
Autotools and CMake:
If the software your recipe is building uses Autotools
or CMake, the OpenEmbedded build
system understands how to install the software.
Consequently, you do not have to have a
do_install
task as part of your
recipe.
You just need to make sure the install portion of the
build completes with no issues.
However, if you wish to install additional files not
already being installed by
make install
, you should do this
using a do_install_append
function
using the install command as described in
the "Manual" bulleted item later in this list.
Other (using
make install
):
You need to define a
do_install
function in your
recipe.
The function should call
oe_runmake install
and will likely
need to pass in the destination directory as well.
How you pass that path is dependent on how the
Makefile
being run is written
(e.g. DESTDIR=${D}
,
PREFIX=${D}
,
INSTALLROOT=${D}
, and so forth).
For an example recipe using
make install
, see the
"Makefile-Based Package"
section.
Manual:
You need to define a
do_install
function in your
recipe.
The function must first use
install -d
to create the
directories under
${
D
}
.
Once the directories exist, your function can use
install
to manually install the
built software into the directories.
You can find more information on
install
at
http://www.gnu.org/software/coreutils/manual/html_node/install-invocation.html.
For the scenarios that do not use Autotools or
CMake, you need to track the installation
and diagnose and fix any issues until everything installs
correctly.
You need to look in the default location of
${D}
, which is
${WORKDIR}/image
, to be sure your
files have been installed correctly.
During the installation process, you might need to
modify some of the installed files to suit the target
layout.
For example, you might need to replace hard-coded paths
in an initscript with values of variables provided by
the build system, such as replacing
/usr/bin/
with
${bindir}
.
If you do perform such modifications during
do_install
, be sure to modify the
destination file after copying rather than before
copying.
Modifying after copying ensures that the build system
can re-execute do_install
if
needed.
oe_runmake install
, which can be
run directly or can be run indirectly by the
autotools
and
cmake
classes, runs make install
in
parallel.
Sometimes, a Makefile can have missing dependencies
between targets that can result in race conditions.
If you experience intermittent failures during
do_install
, you might be able to
work around them by disabling parallel Makefile
installs by adding the following to the recipe:
PARALLEL_MAKEINST = ""
See
PARALLEL_MAKEINST
for additional information.
If you need to install one or more custom CMake
toolchain files that are supplied by the
application you are building, install the files to
${D}${datadir}/cmake/
Modules
during
do_install
.
If you want to install a service, which is a process that usually starts on boot and runs in the background, then you must include some additional definitions in your recipe.
If you are adding services and the service initialization
script or the service file itself is not installed, you must
provide for that installation in your recipe using a
do_install_append
function.
If your recipe already has a do_install
function, update the function near its end rather than
adding an additional do_install_append
function.
When you create the installation for your services, you need
to accomplish what is normally done by
make install
.
In other words, make sure your installation arranges the output
similar to how it is arranged on the target system.
The OpenEmbedded build system provides support for starting services two different ways:
SysVinit: SysVinit is a system and service manager that manages the init system used to control the very basic functions of your system. The init program is the first program started by the Linux kernel when the system boots. Init then controls the startup, running and shutdown of all other programs.
To enable a service using SysVinit, your recipe
needs to inherit the
update-rc.d
class.
The class helps facilitate safely installing the
package on the target.
You will need to set the
INITSCRIPT_PACKAGES
,
INITSCRIPT_NAME
,
and
INITSCRIPT_PARAMS
variables within your recipe.
systemd: System Management Daemon (systemd) was designed to replace SysVinit and to provide enhanced management of services. For more information on systemd, see the systemd homepage at http://freedesktop.org/wiki/Software/systemd/.
To enable a service using systemd, your recipe
needs to inherit the
systemd
class.
See the systemd.bbclass
file
located in your
Source Directory.
section for more information.
Successful packaging is a combination of automated processes performed by the OpenEmbedded build system and some specific steps you need to take. The following list describes the process:
Splitting Files:
The do_package
task splits the
files produced by the recipe into logical components.
Even software that produces a single binary might
still have debug symbols, documentation, and other
logical components that should be split out.
The do_package
task ensures
that files are split up and packaged correctly.
Running QA Checks:
The
insane
class adds a step to
the package generation process so that output quality
assurance checks are generated by the OpenEmbedded
build system.
This step performs a range of checks to be sure the
build's output is free of common problems that show
up during runtime.
For information on these checks, see the
insane
class and the
"QA Error and Warning Messages"
chapter in the Yocto Project Reference Manual.
Hand-Checking Your Packages:
After you build your software, you need to be sure
your packages are correct.
Examine the
${
WORKDIR
}/packages-split
directory and make sure files are where you expect
them to be.
If you discover problems, you can set
PACKAGES
,
FILES
,
do_install(_append)
, and so forth as
needed.
Splitting an Application into Multiple Packages: If you need to split an application into several packages, see the "Splitting an Application into Multiple Packages" section for an example.
Installing a Post-Installation Script: For an example showing how to install a post-installation script, see the "Post-Installation Scripts" section.
Marking Package Architecture: Depending on what your recipe is building and how it is configured, it might be important to mark the packages produced as being specific to a particular machine, or to mark them as not being specific to a particular machine or architecture at all.
By default, packages apply to any machine with the
same architecture as the target machine.
When a recipe produces packages that are
machine-specific (e.g. the
MACHINE
value is passed into the configure script or a patch
is applied only for a particular machine), you should
mark them as such by adding the following to the
recipe:
PACKAGE_ARCH = "${MACHINE_ARCH}"
On the other hand, if the recipe produces packages
that do not contain anything specific to the target
machine or architecture at all (e.g. recipes
that simply package script files or configuration
files), you should use the
allarch
class to do this for you by adding this to your
recipe:
inherit allarch
Ensuring that the package architecture is correct is not critical while you are doing the first few builds of your recipe. However, it is important in order to ensure that your recipe rebuilds (or does not rebuild) appropriately in response to changes in configuration, and to ensure that you get the appropriate packages installed on the target machine, particularly if you run separate builds for more than one target machine.
Recipes often need to use files provided by other recipes on
the build host.
For example, an application linking to a common library needs
access to the library itself and its associated headers.
The way this access is accomplished is by populating a sysroot
with files.
Each recipe has two sysroots in its work directory, one for
target files
(recipe-sysroot
) and one for files that
are native to the build host
(recipe-sysroot-native
).
STAGING_DIR
variable).
Recipes should never populate the sysroot directly (i.e. write
files into sysroot).
Instead, files should be installed into standard locations
during the
do_install
task within the
${
D
}
directory.
The reason for this limitation is that almost all files that
populate the sysroot are cataloged in manifests in order to
ensure the files can be removed later when a recipe is either
modified or removed.
Thus, the sysroot is able to remain free from stale files.
A subset of the files installed by the
do_install
task are used by the
do_populate_sysroot
task as defined by the the
SYSROOT_DIRS
variable to automatically populate the sysroot.
It is possible to modify the list of directories that populate
the sysroot.
The following example shows how you could add the
/opt
directory to the list of
directories within a recipe:
SYSROOT_DIRS += "/opt"
For a more complete description of the
do_populate_sysroot
task and its associated functions, see the
staging
class.
Prior to a build, if you know that several different recipes
provide the same functionality, you can use a virtual provider
(i.e. virtual/*
) as a placeholder for the
actual provider.
The actual provider is determined at build-time.
A common scenario where a virtual provider is used would be
for the kernel recipe.
Suppose you have three kernel recipes whose
PN
values map to kernel-big
,
kernel-mid
, and
kernel-small
.
Furthermore, each of these recipes in some way uses a
PROVIDES
statement that essentially identifies itself as being able
to provide virtual/kernel
.
Here is one way through the
kernel
class:
PROVIDES += "${@ "virtual/kernel" if (d.getVar("KERNEL_PACKAGE_NAME") == "kernel") else "" }"
Any recipe that inherits the kernel
class
is going to utilize a PROVIDES
statement
that identifies that recipe as being able to provide the
virtual/kernel
item.
Now comes the time to actually build an image and you need a
kernel recipe, but which one?
You can configure your build to call out the kernel recipe
you want by using the
PREFERRED_PROVIDER
variable.
As an example, consider the
x86-base.inc
include file, which is a machine
(i.e. MACHINE
)
configuration file.
This include file is the reason all x86-based machines use the
linux-yocto
kernel.
Here are the relevant lines from the include file:
PREFERRED_PROVIDER_virtual/kernel ??= "linux-yocto" PREFERRED_VERSION_linux-yocto ??= "4.15%"
When you use a virtual provider, you do not have to
"hard code" a recipe name as a build dependency.
You can use the
DEPENDS
variable to state the build is dependent on
virtual/kernel
for example:
DEPENDS = "virtual/kernel"
During the build, the OpenEmbedded build system picks
the correct recipe needed for the
virtual/kernel
dependency based on the
PREFERRED_PROVIDER
variable.
If you want to use the small kernel mentioned at the beginning
of this section, configure your build as follows:
PREFERRED_PROVIDER_virtual/kernel ??= "kernel-small"
PROVIDES
a virtual/*
item that is ultimately
not selected through
PREFERRED_PROVIDER
does not get built.
Preventing these recipes from building is usually the
desired behavior since this mechanism's purpose is to
select between mutually exclusive alternative providers.
The following lists specific examples of virtual providers:
virtual/kernel
:
Provides the name of the kernel recipe to use when
building a kernel image.
virtual/bootloader
:
Provides the name of the bootloader to use when
building an image.
virtual/mesa
:
Provides gbm.pc
.
virtual/egl
:
Provides egl.pc
and possibly
wayland-egl.pc
.
virtual/libgl
:
Provides gl.pc
(i.e. libGL).
virtual/libgles1
:
Provides glesv1_cm.pc
(i.e. libGLESv1_CM).
virtual/libgles2
:
Provides glesv2.pc
(i.e. libGLESv2).
Sometimes the name of a recipe can lead to versioning
problems when the recipe is upgraded to a final release.
For example, consider the
irssi_0.8.16-rc1.bb
recipe file in
the list of example recipes in the
"Storing and Naming the Recipe"
section.
This recipe is at a release candidate stage (i.e.
"rc1").
When the recipe is released, the recipe filename becomes
irssi_0.8.16.bb
.
The version change from 0.8.16-rc1
to 0.8.16
is seen as a decrease by the
build system and package managers, so the resulting packages
will not correctly trigger an upgrade.
In order to ensure the versions compare properly, the
recommended convention is to set
PV
within the recipe to
"previous_version
+current_version
".
You can use an additional variable so that you can use the
current version elsewhere.
Here is an example:
REALPV = "0.8.16-rc1" PV = "0.8.15+${REALPV}"
Post-installation scripts run immediately after installing
a package on the target or during image creation when a
package is included in an image.
To add a post-installation script to a package, add a
pkg_postinst_
PACKAGENAME
()
function to
the recipe file (.bb
) and replace
PACKAGENAME
with the name of the package
you want to attach to the postinst
script.
To apply the post-installation script to the main package
for the recipe, which is usually what is required, specify
${
PN
}
in place of PACKAGENAME
.
A post-installation function has the following structure:
pkg_postinst_PACKAGENAME
() {
# Commands to carry out
}
The script defined in the post-installation function is called when the root filesystem is created. If the script succeeds, the package is marked as installed.
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, you must explicitly
mark post installs to defer to the target.
You can use pkg_postinst_ontarget()
or
call
postinst_intercept delay_to_first_boot
from pkg_postinst()
.
Any failure of a pkg_postinst()
script
(including exit 1) triggers an error during the
do_rootfs
task.
If you have recipes that use
pkg_postinst
function
and they require the use of non-standard native
tools that have dependencies during rootfs construction, you
need to use the
PACKAGE_WRITE_DEPS
variable in your recipe to list these tools.
If you do not use this variable, the tools might be missing and
execution of the post-installation script is deferred until
first boot.
Deferring the script to first boot is undesirable and for
read-only rootfs impossible.
pkg_preinst
,
pkg_prerm
, and
pkg_postrm
, respectively.
These scrips work in exactly the same way as does
pkg_postinst
with the exception
that they run at different times.
Also, because of when they run, they are not applicable to
being run at image creation time like
pkg_postinst
.
The final step for completing your recipe is to be sure that the software you built runs correctly. To accomplish runtime testing, add the build's output packages to your image and test them on the target.
For information on how to customize your image by adding specific packages, see the "Customizing Images" section.
To help summarize how to write a recipe, this section provides some examples given various scenarios:
Recipes that use local files
Using an Autotooled package
Using a Makefile-based package
Splitting an application into multiple packages
Adding binaries to an image
Building an application from a single file that is stored
locally (e.g. under files
) requires
a recipe that has the file listed in the
SRC_URI
variable.
Additionally, you need to manually write the
do_compile
and
do_install
tasks.
The S
variable defines the directory containing the source code,
which is set to
WORKDIR
in this case - the directory BitBake uses for the build.
SUMMARY = "Simple helloworld application" SECTION = "examples" LICENSE = "MIT" LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302" 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} }
By default, the helloworld
,
helloworld-dbg
, and
helloworld-dev
packages are built.
For information on how to customize the packaging process,
see the
"Splitting an Application into Multiple Packages"
section.
Applications that use Autotools such as autoconf
and
automake
require a recipe that has a source archive listed in
SRC_URI
and
also inherit the
autotools
class, which contains the definitions of all the steps
needed to build an Autotool-based 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: (hello_2.3.bb
)
SUMMARY = "GNU Helloworld application" SECTION = "examples" LICENSE = "GPLv2+" LIC_FILES_CHKSUM = "file://COPYING;md5=751419260aa954499f7abaabaa882bbe" SRC_URI = "${GNU_MIRROR}/hello/hello-${PV}.tar.gz" inherit autotools gettext
The variable
LIC_FILES_CHKSUM
is used to track source license changes as described in the
"Tracking License Changes"
section in the Yocto Project Overview and Concepts Manual.
You can quickly create Autotool-based recipes in a manner
similar to the previous example.
Applications that use GNU make
also require a recipe that has
the source archive listed in
SRC_URI
.
You do not need to add a do_compile
step since by default BitBake
starts the make
command to compile the application.
If you need additional make
options, you should store them in the
EXTRA_OEMAKE
or
PACKAGECONFIG_CONFARGS
variables.
BitBake passes these options into the GNU make
invocation.
Note that a do_install
task is still required.
Otherwise, BitBake runs an empty do_install
task by default.
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
CFLAGS
variable.
The following example shows this:
CFLAGS_prepend = "-I ${S}/include "
In the following example, mtd-utils
is a makefile-based package:
SUMMARY = "Tools for managing memory technology devices" SECTION = "base" DEPENDS = "zlib lzo e2fsprogs util-linux" HOMEPAGE = "http://www.linux-mtd.infradead.org/" LICENSE = "GPLv2+" LIC_FILES_CHKSUM = "file://COPYING;md5=0636e73ff0215e8d672dc4c32c317bb3 \ file://include/common.h;beginline=1;endline=17;md5=ba05b07912a44ea2bf81ce409380049c" # Use the latest version at 26 Oct, 2013 SRCREV = "9f107132a6a073cce37434ca9cda6917dd8d866b" SRC_URI = "git://git.infradead.org/mtd-utils.git \ file://add-exclusion-to-mkfs-jffs2-git-2.patch \ " PV = "1.5.1+git${SRCPV}" S = "${WORKDIR}/git" EXTRA_OEMAKE = "'CC=${CC}' 'RANLIB=${RANLIB}' 'AR=${AR}' 'CFLAGS=${CFLAGS} -I${S}/include -DWITHOUT_XATTR' 'BUILDDIR=${S}'" do_install () { oe_runmake install DESTDIR=${D} SBINDIR=${sbindir} MANDIR=${mandir} INCLUDEDIR=${includedir} } PACKAGES =+ "mtd-utils-jffs2 mtd-utils-ubifs mtd-utils-misc" FILES_mtd-utils-jffs2 = "${sbindir}/mkfs.jffs2 ${sbindir}/jffs2dump ${sbindir}/jffs2reader ${sbindir}/sumtool" FILES_mtd-utils-ubifs = "${sbindir}/mkfs.ubifs ${sbindir}/ubi*" FILES_mtd-utils-misc = "${sbindir}/nftl* ${sbindir}/ftl* ${sbindir}/rfd* ${sbindir}/doc* ${sbindir}/serve_image ${sbindir}/recv_image" PARALLEL_MAKE = "" BBCLASSEXTEND = "native"
You can use the variables
PACKAGES
and
FILES
to split an application into multiple packages.
Following is an example that uses the libxpm
recipe.
By default, this recipe generates a single package that contains the library along
with a few binaries.
You can modify the recipe to split the binaries into separate packages:
require xorg-lib-common.inc SUMMARY = "Xpm: X Pixmap extension library" LICENSE = "BSD" LIC_FILES_CHKSUM = "file://COPYING;md5=51f4270b012ecd4ab1a164f5f4ed6cf7" DEPENDS += "libxext libsm libxt" PE = "1" XORG_PN = "libXpm" PACKAGES =+ "sxpm cxpm" FILES_cxpm = "${bindir}/cxpm" FILES_sxpm = "${bindir}/sxpm"
In the previous example, we want to ship the sxpm
and cxpm
binaries in separate packages.
Since bindir
would be packaged into the main
PN
package by default, we prepend the PACKAGES
variable so additional package names are added to the start of list.
This results in the extra FILES_*
variables then containing information that define which files and
directories go into which packages.
Files included by earlier packages are skipped by latter packages.
Thus, the main PN
package
does not include the above listed files.
Sometimes, you need to add pre-compiled binaries to an
image.
For example, suppose that binaries for proprietary code
exist, which are created by a particular division of a
company.
Your part of the company needs to use those binaries as
part of an image that you are building using the
OpenEmbedded build system.
Since you only have the binaries and not the source code,
you cannot use a typical recipe that expects to fetch the
source specified in
SRC_URI
and then compile it.
One method is to package the binaries and then install them as part of the image. Generally, it is not a good idea to package binaries since, among other things, it can hinder the ability to reproduce builds and could lead to compatibility problems with ABI in the future. However, sometimes you have no choice.
The easiest solution is to create a recipe that uses
the
bin_package
class and to be sure that you are using default locations
for build artifacts.
In most cases, the bin_package
class
handles "skipping" the configure and compile steps as well
as sets things up to grab packages from the appropriate
area.
In particular, this class sets noexec
on both the
do_configure
and
do_compile
tasks, sets
FILES_${PN}
to "/" so that it picks
up all files, and sets up a
do_install
task, which effectively copies all files from
${S}
to ${D}
.
The bin_package
class works well when
the files extracted into ${S}
are
already laid out in the way they should be laid out
on the target.
For more information on these variables, see the
FILES
,
PN
,
S
,
and
D
variables in the Yocto Project Reference Manual's variable
glossary.
Using
DEPENDS
is a good idea even for components distributed
in binary form, and is often necessary for
shared libraries.
For a shared library, listing the library
dependencies in
DEPENDS
makes sure that
the libraries are available in the staging
sysroot when other recipes link against the
library, which might be necessary for
successful linking.
Using DEPENDS
also
allows runtime dependencies between packages
to be added automatically.
See the
"Automatically Added Runtime Dependencies"
section in the Yocto Project Overview and
Concepts Manual for more information.
If you cannot use the bin_package
class, you need to be sure you are doing the following:
Create a recipe where the
do_configure
and
do_compile
tasks do nothing:
It is usually sufficient to just not define these
tasks in the recipe, because the default
implementations do nothing unless a Makefile is
found in
${
S
}
.
If
${S}
might contain a Makefile,
or if you inherit some class that replaces
do_configure
and
do_compile
with custom
versions, then you can use the
[
noexec
]
flag to turn the tasks into no-ops, as follows:
do_configure[noexec] = "1" do_compile[noexec] = "1"
Unlike
deleting the tasks
,
using the flag preserves the dependency chain from
the
do_fetch
, do_unpack
,
and
do_patch
tasks to the
do_install
task.
Make sure your
do_install
task installs the
binaries appropriately.
Ensure that you set up
FILES
(usually
FILES_${
PN
}
)
to point to the files you have installed, which of
course depends on where you have installed them
and whether those files are in different locations
than the defaults.
When writing recipes, it is good to conform to existing style guidelines. The OpenEmbedded Styleguide wiki page provides rough guidelines for preferred recipe style.
It is common for existing recipes to deviate a bit from this
style.
However, aiming for at least a consistent style is a good idea.
Some practices, such as omitting spaces around
=
operators in assignments or ordering
recipe components in an erratic way, are widely seen as poor
style.
Understanding recipe file syntax is important for writing recipes. The following list overviews the basic items that make up a BitBake recipe file. For more complete BitBake syntax descriptions, see the "Syntax and Operators" chapter of the BitBake User Manual.
Variable Assignments and Manipulations: Variable assignments allow a value to be assigned to a variable. The assignment can be static text or might include the contents of other variables. In addition to the assignment, appending and prepending operations are also supported.
The following example shows some of the ways you can use variables in recipes:
S = "${WORKDIR}/postfix-${PV}" CFLAGS += "-DNO_ASM" SRC_URI_append = " file://fixup.patch"
Functions:
Functions provide a series of actions to be performed.
You usually use functions to override the default
implementation of a task function or to complement
a default function (i.e. append or prepend to an
existing function).
Standard functions use sh
shell
syntax, although access to OpenEmbedded variables and
internal methods are also available.
The following is an example function from the
sed
recipe:
do_install () { autotools_do_install install -d ${D}${base_bindir} mv ${D}${bindir}/sed ${D}${base_bindir}/sed rmdir ${D}${bindir}/ }
It is also possible to implement new functions that are called between existing tasks as long as the new functions are not replacing or complementing the default functions. You can implement functions in Python instead of shell. Both of these options are not seen in the majority of recipes.
Keywords:
BitBake recipes use only a few keywords.
You use keywords to include common
functions (inherit
), load parts
of a recipe from other files
(include
and
require
) and export variables
to the environment (export
).
The following example shows the use of some of these keywords:
export POSTCONF = "${STAGING_BINDIR}/postconf" inherit autoconf require otherfile.inc
Comments (#):
Any lines that begin with the hash character
(#
) are treated as comment lines
and are ignored:
# This is a comment
This next list summarizes the most important and most commonly used parts of the recipe syntax. For more information on these parts of the syntax, you can reference the Syntax and Operators chapter in the BitBake User Manual.
Line Continuation (\):
Use the backward slash (\
)
character to split a statement over multiple lines.
Place the slash character at the end of the line that
is to be continued on the next line:
VAR = "A really long \ line"
Using Variables (${VARNAME
}):
Use the ${
syntax to access the contents of a variable:
VARNAME
}
SRC_URI = "${SOURCEFORGE_MIRROR}/libpng/zlib-${PV}.tar.gz"
:=
operator instead of
=
when you make the
assignment, but this is not generally needed.
Quote All Assignments ("value
"):
Use double quotes around values in all variable
assignments (e.g.
"
).
Following is an example:
value
"
VAR1 = "${OTHERVAR}" VAR2 = "The version is ${PV}"
Conditional Assignment (?=):
Conditional assignment is used to assign a
value to a variable, but only when the variable is
currently unset.
Use the question mark followed by the equal sign
(?=
) to make a "soft" assignment
used for conditional assignment.
Typically, "soft" assignments are used in the
local.conf
file for variables
that are allowed to come through from the external
environment.
Here is an example where
VAR1
is set to "New value" if
it is currently empty.
However, if VAR1
has already been
set, it remains unchanged:
VAR1 ?= "New value"
In this next example, VAR1
is left with the value "Original value":
VAR1 = "Original value" VAR1 ?= "New value"
Appending (+=):
Use the plus character followed by the equals sign
(+=
) to append values to existing
variables.
Here is an example:
SRC_URI += "file://fix-makefile.patch"
Prepending (=+):
Use the equals sign followed by the plus character
(=+
) to prepend values to existing
variables.
Here is an example:
VAR =+ "Starts"
Appending (_append):
Use the _append
operator to
append values to existing variables.
This operator does not add any additional space.
Also, the operator is applied after all the
+=
, and
=+
operators have been applied and
after all =
assignments have
occurred.
The following example shows the space being explicitly added to the start to ensure the appended value is not merged with the existing value:
SRC_URI_append = " file://fix-makefile.patch"
You can also use the _append
operator with overrides, which results in the actions
only being performed for the specified target or
machine:
SRC_URI_append_sh4 = " file://fix-makefile.patch"
Prepending (_prepend):
Use the _prepend
operator to
prepend values to existing variables.
This operator does not add any additional space.
Also, the operator is applied after all the
+=
, and
=+
operators have been applied and
after all =
assignments have
occurred.
The following example shows the space being explicitly added to the end to ensure the prepended value is not merged with the existing value:
CFLAGS_prepend = "-I${S}/myincludes "
You can also use the _prepend
operator with overrides, which results in the actions
only being performed for the specified target or
machine:
CFLAGS_prepend_sh4 = "-I${S}/myincludes "
Overrides:
You can use overrides to set a value conditionally,
typically based on how the recipe is being built.
For example, to set the
KBRANCH
variable's value to "standard/base" for any target
MACHINE
,
except for qemuarm where it should be set to
"standard/arm-versatile-926ejs", you would do the
following:
KBRANCH = "standard/base" KBRANCH_qemuarm = "standard/arm-versatile-926ejs"
Overrides are also used to separate alternate values
of a variable in other situations.
For example, when setting variables such as
FILES
and
RDEPENDS
that are specific to individual packages produced by
a recipe, you should always use an override that
specifies the name of the package.
Indentation: Use spaces for indentation rather than than tabs. For shell functions, both currently work. However, it is a policy decision of the Yocto Project to use tabs in shell functions. Realize that some layers have a policy to use spaces for all indentation.
Using Python for Complex Operations: For more advanced processing, it is possible to use Python code during variable assignments (e.g. search and replacement on a variable).
You indicate Python code using the
${@
syntax for the variable assignment:
python_code
}
SRC_URI = "ftp://ftp.info-zip.org/pub/infozip/src/zip${@d.getVar('PV',1).replace('.', '')}.tgz
Shell Function Syntax:
Write shell functions as if you were writing a shell
script when you describe a list of actions to take.
You should ensure that your script works with a generic
sh
and that it does not require
any bash
or other shell-specific
functionality.
The same considerations apply to various system
utilities (e.g. sed
,
grep
, awk
,
and so forth) that you might wish to use.
If in doubt, you should check with multiple
implementations - including those from BusyBox.
Adding a new machine to the Yocto Project is a straightforward process. This section describes how to add machines that are similar to those that the Yocto Project already supports.
gcc/glibc
and to the site
information, which is beyond the scope of this manual.
For a complete example that shows how to add a new machine,
see the
"Creating a New BSP Layer Using the bitbake-layers
Script"
section in the Yocto Project Board Support Package (BSP)
Developer's Guide.
To add a new machine, you need to add a new machine
configuration file to the layer's
conf/machine
directory.
This configuration file provides details about the device
you are adding.
The OpenEmbedded build system uses the root name of the
machine configuration file to reference the new machine.
For example, given a machine configuration file named
crownbay.conf
, the build system
recognizes the machine as "crownbay".
The most important variables you must set in your machine configuration file or include from a lower-level configuration file are as follows:
TARGET_ARCH
(e.g. "arm")
PREFERRED_PROVIDER_virtual/kernel
MACHINE_FEATURES
(e.g. "apm screen wifi")
You might also need these variables:
SERIAL_CONSOLES
(e.g. "115200;ttyS0 115200;ttyS1")
KERNEL_IMAGETYPE
(e.g. "zImage")
IMAGE_FSTYPES
(e.g. "tar.gz jffs2")
You can find full details on these variables in the reference
section.
You can leverage existing machine .conf
files from meta-yocto-bsp/conf/machine/
.
The OpenEmbedded build system 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 kernel recipe.
You can find several kernel recipe examples in the
Source Directory at
meta/recipes-kernel/linux
that you can use as references.
If you are creating a new kernel recipe, normal recipe-writing
rules apply for setting up a
SRC_URI
.
Thus, you need to specify any necessary patches and set
S
to point at the source code.
You need to create a do_configure
task that
configures the unpacked kernel with a
defconfig
file.
You can do this by using a make defconfig
command or, more commonly, by copying in a suitable
defconfig
file and then running
make oldconfig
.
By making use of inherit kernel
and
potentially some of the linux-*.inc
files,
most other functionality is centralized and the defaults of the
class normally work well.
If you are extending an existing kernel recipe, it is usually
a matter of adding a suitable defconfig
file.
The file needs to be added into a location similar to
defconfig
files used for other machines
in a given kernel recipe.
A possible way to do this is by listing the file in the
SRC_URI
and adding the machine to the
expression in
COMPATIBLE_MACHINE
:
COMPATIBLE_MACHINE = '(qemux86|qemumips)'
For more information on defconfig
files,
see the
"Changing the Configuration"
section in the Yocto Project Linux Kernel Development Manual.
A formfactor configuration file provides information about the target hardware for which the image is being built and information that the build system 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.
The build system uses reasonable defaults in most cases.
However, if customization is
necessary, you need to create a machconfig
file
in the meta/recipes-bsp/formfactor/files
directory.
This directory contains directories for specific machines such as
qemuarm
and qemux86
.
For information about the settings available and the defaults, see the
meta/recipes-bsp/formfactor/files/config
file found in the
same area.
Following is an example for "qemuarm" machine:
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
Over time, upstream developers publish new versions for software built by layer recipes. It is recommended to keep recipes up-to-date with upstream version releases.
While several methods exist that allow you upgrade a recipe,
you might consider checking on the upgrade status of a recipe
first.
You can do so using the
devtool check-upgrade-status
command.
See the
"Checking on the Upgrade Status of a Recipe"
section in the Yocto Project Reference Manual for more information.
The remainder of this section describes three ways you can
upgrade a recipe.
You can use the Automated Upgrade Helper (AUH) to set up
automatic version upgrades.
Alternatively, you can use devtool upgrade
to set up semi-automatic version upgrades.
Finally, you can manually upgrade a recipe by editing the
recipe itself.
The AUH utility works in conjunction with the OpenEmbedded build system in order to automatically generate upgrades for recipes based on new versions being published upstream. Use AUH when you want to create a service that performs the upgrades automatically and optionally sends you an email with the results.
AUH allows you to update several recipes with a single use. You can also optionally perform build and integration tests using images with the results saved to your hard drive and emails of results optionally sent to recipe maintainers. Finally, AUH creates Git commits with appropriate commit messages in the layer's tree for the changes made to recipes.
devtool upgrade
or upgrade your
recipes manually:
When AUH cannot complete the upgrade sequence.
This situation usually results because custom
patches carried by the recipe cannot be
automatically rebased to the new version.
In this case, devtool upgrade
allows you to manually resolve conflicts.
When for any reason you want fuller control over the upgrade process. For example, when you want special arrangements for testing.
The following steps describe how to set up the AUH utility:
Be Sure the Development Host is Set Up: You need to be sure that your development host is set up to use the Yocto Project. For information on how to set up your host, see the "Preparing the Build Host" section.
Make Sure Git is Configured: The AUH utility requires Git to be configured because AUH uses Git to save upgrades. Thus, you must have Git user and email configured. The following command shows your configurations:
$ git config --list
If you do not have the user and email configured, you can use the following commands to do so:
$ git config --global user.namesome_name
$ git config --global user.emailusername
@domain
.com
Clone the AUH Repository: To use AUH, you must clone the repository onto your development host. The following command uses Git to create a local copy of the repository on your system:
$ git clone git://git.yoctoproject.org/auto-upgrade-helper Cloning into 'auto-upgrade-helper'... remote: Counting objects: 768, done. remote: Compressing objects: 100% (300/300), done. remote: Total 768 (delta 499), reused 703 (delta 434) Receiving objects: 100% (768/768), 191.47 KiB | 98.00 KiB/s, done. Resolving deltas: 100% (499/499), done. Checking connectivity... done.
AUH is not part of the OpenEmbedded-Core (OE-Core) or Poky repositories.
Create a Dedicated Build Directory:
Run the
oe-init-build-env
script to create a fresh build directory that you
use exclusively for running the AUH utility:
$ cd ~/poky
$ source oe-init-build-env your_AUH_build_directory
Re-using an existing build directory and its configurations is not recommended as existing settings could cause AUH to fail or behave undesirably.
Make Configurations in Your Local Configuration File:
Several settings need to exist in the
local.conf
file in the build
directory you just created for AUH.
Make these following configurations:
If you want to enable
Build History,
which is optional, you need the following
lines in the
conf/local.conf
file:
INHERIT =+ "buildhistory" BUILDHISTORY_COMMIT = "1"
With this configuration and a successful
upgrade, a build history "diff" file appears in
the
upgrade-helper/work/recipe/buildhistory-diff.txt
file found in your build directory.
If you want to enable testing through the
testimage
class, which is optional, you need to have the
following set in your
conf/local.conf
file:
INHERIT += "testimage"
local.conf
file:
DISTRO_FEATURES_append = " ptest"
Optionally Start a vncserver: If you are running in a server without an X11 session, you need to start a vncserver:
$ vncserver :1 $ export DISPLAY=:1
Create and Edit an AUH Configuration File:
You need to have the
upgrade-helper/upgrade-helper.conf
configuration file in your build directory.
You can find a sample configuration file in the
AUH source repository.
Read through the sample file and make
configurations as needed.
For example, if you enabled build history in your
local.conf
as described earlier,
you must enable it in
upgrade-helper.conf
.
Also, if you are using the default
maintainers.inc
file supplied
with Poky and located in
meta-yocto
and you do not set a
"maintainers_whitelist" or "global_maintainer_override"
in the upgrade-helper.conf
configuration, and you specify "-e all" on the
AUH command-line, the utility automatically sends out
emails to all the default maintainers.
Please avoid this.
This next set of examples describes how to use the AUH:
Upgrading a Specific Recipe: To upgrade a specific recipe, use the following form:
$ upgrade-helper.py recipe_name
For example, this command upgrades the
xmodmap
recipe:
$ upgrade-helper.py xmodmap
Upgrading a Specific Recipe to a Particular Version: To upgrade a specific recipe to a particular version, use the following form:
$ upgrade-helper.pyrecipe_name
-tversion
For example, this command upgrades the
xmodmap
recipe to version
1.2.3:
$ upgrade-helper.py xmodmap -t 1.2.3
Upgrading all Recipes to the Latest Versions and Suppressing Email Notifications: To upgrade all recipes to their most recent versions and suppress the email notifications, use the following command:
$ upgrade-helper.py all
Upgrading all Recipes to the Latest Versions and Send Email Notifications: To upgrade all recipes to their most recent versions and send email messages to maintainers for each attempted recipe as well as a status email, use the following command:
$ upgrade-helper.py -e all
Once you have run the AUH utility, you can find the results in the AUH build directory:
${BUILDDIR}/upgrade-helper/timestamp
The AUH utility also creates recipe update commits from successful upgrade attempts in the layer tree.
You can easily set up to run the AUH utility on a regular
basis by using a cron job.
See the
weeklyjob.sh
file distributed with the utility for an example.
devtool upgrade
¶
As mentioned earlier, an alternative method for upgrading
recipes to newer versions is to use
devtool upgrade
.
You can read about devtool upgrade
in
general in the
"Use devtool upgrade
to Create a Version of the Recipe that Supports a Newer Version of the Software"
section in the Yocto Project Application Development and the
Extensible Software Development Kit (eSDK) Manual.
To see all the command-line options available with
devtool upgrade
, use the following help
command:
$ devtool upgrade -h
If you want to find out what version a recipe is currently at upstream without any attempt to upgrade your local version of the recipe, you can use the following command:
$ devtool latest-version recipe_name
As mentioned in the previous section describing AUH,
devtool upgrade
works in a
less-automated manner than AUH.
Specifically, devtool upgrade
only
works on a single recipe that you name on the command line,
cannot perform build and integration testing using images,
and does not automatically generate commits for changes in
the source tree.
Despite all these "limitations",
devtool upgrade
updates the recipe file
to the new upstream version and attempts to rebase custom
patches contained by the recipe as needed.
devtool upgrade
behind the scenes making AUH somewhat of a "wrapper"
application for devtool upgrade
.
A typical scenario involves having used Git to clone an
upstream repository that you use during build operations.
Because you are (or have) built the recipe in the past, the
layer is likely added to your configuration already.
If for some reason, the layer is not added, you could add
it easily using the
bitbake-layers
script.
For example, suppose you use the nano.bb
recipe from the meta-oe
layer in the
meta-openembedded
repository.
For this example, assume that the layer has been cloned into
following area:
/home/scottrif/meta-openembedded
The following command from your
Build Directory
adds the layer to your build configuration (i.e.
${BUILDDIR}/conf/bblayers.conf
):
$ bitbake-layers add-layer /home/scottrif/meta-openembedded/meta-oe NOTE: Starting bitbake server... Parsing recipes: 100% |##########################################| Time: 0:00:55 Parsing of 1431 .bb files complete (0 cached, 1431 parsed). 2040 targets, 56 skipped, 0 masked, 0 errors. Removing 12 recipes from the x86_64 sysroot: 100% |##############| Time: 0:00:00 Removing 1 recipes from the x86_64_i586 sysroot: 100% |##########| Time: 0:00:00 Removing 5 recipes from the i586 sysroot: 100% |#################| Time: 0:00:00 Removing 5 recipes from the qemux86 sysroot: 100% |##############| Time: 0:00:00
For this example, assume that the nano.bb
recipe that is upstream has a 2.9.3 version number.
However, the version in the local repository is 2.7.4.
The following command from your build directory automatically
upgrades the recipe for you:
-V
option is not necessary.
Omitting the version number causes
devtool upgrade
to upgrade the recipe
to the most recent version.
$ devtool upgrade nano -V 2.9.3 NOTE: Starting bitbake server... NOTE: Creating workspace layer in /home/scottrif/poky/build/workspace Parsing recipes: 100% |##########################################| Time: 0:00:46 Parsing of 1431 .bb files complete (0 cached, 1431 parsed). 2040 targets, 56 skipped, 0 masked, 0 errors. NOTE: Extracting current version source... NOTE: Resolving any missing task queue dependencies . . . NOTE: Executing SetScene Tasks NOTE: Executing RunQueue Tasks NOTE: Tasks Summary: Attempted 74 tasks of which 72 didn't need to be rerun and all succeeded. Adding changed files: 100% |#####################################| Time: 0:00:00 NOTE: Upgraded source extracted to /home/scottrif/poky/build/workspace/sources/nano NOTE: New recipe is /home/scottrif/poky/build/workspace/recipes/nano/nano_2.9.3.bb
Continuing with this example, you can use
devtool build
to build the newly upgraded
recipe:
$ devtool build nano NOTE: Starting bitbake server... Loading cache: 100% |################################################################################################| Time: 0:00:01 Loaded 2040 entries from dependency cache. Parsing recipes: 100% |##############################################################################################| Time: 0:00:00 Parsing of 1432 .bb files complete (1431 cached, 1 parsed). 2041 targets, 56 skipped, 0 masked, 0 errors. NOTE: Resolving any missing task queue dependencies . . . NOTE: Executing SetScene Tasks NOTE: Executing RunQueue Tasks NOTE: nano: compiling from external source tree /home/scottrif/poky/build/workspace/sources/nano NOTE: Tasks Summary: Attempted 520 tasks of which 304 didn't need to be rerun and all succeeded.
Within the devtool upgrade
workflow,
opportunity exists to deploy and test your rebuilt software.
For this example, however, running
devtool finish
cleans up the workspace
once the source in your workspace is clean.
This usually means using Git to stage and submit commits
for the changes generated by the upgrade process.
Once the tree is clean, you can clean things up in this
example with the following command from the
${BUILDDIR}/workspace/sources/nano
directory:
$ devtool finish nano meta-oe NOTE: Starting bitbake server... Loading cache: 100% |################################################################################################| Time: 0:00:00 Loaded 2040 entries from dependency cache. Parsing recipes: 100% |##############################################################################################| Time: 0:00:01 Parsing of 1432 .bb files complete (1431 cached, 1 parsed). 2041 targets, 56 skipped, 0 masked, 0 errors. NOTE: Adding new patch 0001-nano.bb-Stuff-I-changed-when-upgrading-nano.bb.patch NOTE: Updating recipe nano_2.9.3.bb NOTE: Removing file /home/scottrif/meta-openembedded/meta-oe/recipes-support/nano/nano_2.7.4.bb NOTE: Moving recipe file to /home/scottrif/meta-openembedded/meta-oe/recipes-support/nano NOTE: Leaving source tree /home/scottrif/poky/build/workspace/sources/nano as-is; if you no longer need it then please delete it manually
Using the devtool finish
command cleans
up the workspace and creates a patch file based on your
commits.
The tool puts all patch files back into the source directory
in a sub-directory named nano
in this
case.
If for some reason you choose not to upgrade recipes using the
Auto Upgrade Helper (AUH)
or by using
devtool upgrade
,
you can manually edit the recipe files to upgrade the versions.
devtool upgrade
, both of which
automate some steps and provide guidance for others needed
for the manual process.
To manually upgrade recipe versions, follow these general steps:
Change the Version:
Rename the recipe such that the version (i.e. the
PV
part of the recipe name) changes appropriately.
If the version is not part of the recipe name, change
the value as it is set for PV
within the recipe itself.
Update SRCREV
if Needed:
If the source code your recipe builds is fetched from
Git or some other version control system, update
SRCREV
to point to the commit hash that matches the new
version.
Build the Software: Try to build the recipe using BitBake. Typical build failures include the following:
License statements were updated for the new
version.
For this case, you need to review any changes
to the license and update the values of
LICENSE
and
LIC_FILES_CHKSUM
as needed.
Custom patches carried by the older version of the recipe might fail to apply to the new version. For these cases, you need to review the failures. Patches might not be necessary for the new version of the software if the upgraded version has fixed those issues. If a patch is necessary and failing, you need to rebase it into the new version.
Optionally Attempt to Build for Several Architectures:
Once you successfully build the new software for a
given architecture, you could test the build for
other architectures by changing the
MACHINE
variable and rebuilding the software.
This optional step is especially important if the
recipe is to be released publicly.
Check the Upstream Change Log or Release Notes: Checking both these reveals if new features exist that could break backwards-compatibility. If so, you need to take steps to mitigate or eliminate that situation.
Optionally Create a Bootable Image and Test: If you want, you can test the new software by booting it onto actual hardware.
Create a Commit with the Change in the Layer Repository: After all builds work and any testing is successful, you can create commits for any changes in the layer holding your upgraded recipe.
You might find it helpful during development to modify the temporary source code used by recipes to build packages. For example, suppose you are developing a patch and you need to experiment a bit to figure out your solution. After you have initially built the package, you can iteratively tweak the source code, which is located in the Build Directory, and then you can force a re-compile and quickly test your altered code. Once you settle on a solution, you can then preserve your changes in the form of patches.
During a build, the unpacked temporary source code used by recipes
to build packages is available in the Build Directory as
defined by the
S
variable.
Below is the default value for the S
variable
as defined in the
meta/conf/bitbake.conf
configuration file
in the
Source Directory:
S = "${WORKDIR}/${BP}"
You should be aware that many recipes override the
S
variable.
For example, recipes that fetch their source from Git usually set
S
to ${WORKDIR}/git
.
BP
represents the base recipe name, which consists of the name
and version:
BP = "${BPN}-${PV}"
The path to the work directory for the recipe
(WORKDIR
)
is defined as follows:
${TMPDIR}/work/${MULTIMACH_TARGET_SYS}/${PN}/${EXTENDPE}${PV}-${PR}
The actual directory depends on several things:
As an example, assume a Source Directory top-level folder
named poky
, a default Build Directory at
poky/build
, and a
qemux86-poky-linux
machine target
system.
Furthermore, suppose your recipe is named
foo_1.3.0.bb
.
In this case, the work directory the build system uses to
build the package would be as follows:
poky/build/tmp/work/qemux86-poky-linux/foo/1.3.0-r0
Quilt is a powerful tool that allows you to capture source code changes without having a clean source tree. This section outlines the typical workflow you can use to modify source code, test changes, and then preserve the changes in the form of a patch all using Quilt.
rm_work
enabled,
the
devtool
workflow
as described in the Yocto Project Application Development
and the Extensible Software Development Kit (eSDK) manual
is a safer development flow than the flow that uses Quilt.
Follow these general steps:
Find the Source Code: Temporary source code used by the OpenEmbedded build system is kept in the Build Directory. See the "Finding Temporary Source Code" section to learn how to locate the directory that has the temporary source code for a particular package.
Change Your Working Directory:
You need to be in the directory that has the temporary
source code.
That directory is defined by the
S
variable.
Create a New Patch:
Before modifying source code, you need to create a new
patch.
To create a new patch file, use
quilt new
as below:
$ quilt new my_changes.patch
Notify Quilt and Add Files: After creating the patch, you need to notify Quilt about the files you plan to edit. You notify Quilt by adding the files to the patch you just created:
$ quilt add file1.c file2.c file3.c
Edit the Files: Make your changes in the source code to the files you added to the patch.
Test Your Changes:
Once you have modified the source code, the easiest way to
test your changes is by calling the
do_compile
task as shown in the
following example:
$ bitbake -c compile -f package
The -f
or --force
option forces the specified task to execute.
If you find problems with your code, you can just keep
editing and re-testing iteratively until things work
as expected.
do_clean
or
do_cleanall
tasks using BitBake (i.e.
bitbake -c clean package
and
bitbake -c cleanall package
).
Modifications will also disappear if you use the
rm_work
feature as described
in the
"Conserving Disk Space During Builds"
section.
Generate the Patch: Once your changes work as expected, you need to use Quilt to generate the final patch that contains all your modifications.
$ quilt refresh
At this point, the my_changes.patch
file has all your edits made to the
file1.c
, file2.c
,
and file3.c
files.
You can find the resulting patch file in the
patches/
subdirectory of the source
(S
) directory.
Copy the Patch File:
For simplicity, copy the patch file into a directory
named files
, which you can create
in the same directory that holds the recipe
(.bb
) file or the append
(.bbappend
) file.
Placing the patch here guarantees that the OpenEmbedded
build system will find the patch.
Next, add the patch into the
SRC_URI
of the recipe.
Here is an example:
SRC_URI += "file://my_changes.patch"
When debugging certain commands or even when just editing packages,
devshell
can be a useful tool.
When you invoke devshell
, all tasks up to and
including
do_patch
are run for the specified target.
Then, a new terminal is opened and you are placed in
${
S
}
,
the source directory.
In the new terminal, all the OpenEmbedded build-related environment variables are
still defined so you can use commands such as configure
and
make
.
The commands execute just as if the OpenEmbedded build system were executing them.
Consequently, working this way can be helpful when debugging a build or preparing
software to be used with the OpenEmbedded build system.
Following is an example that uses devshell
on a target named
matchbox-desktop
:
$ bitbake matchbox-desktop -c devshell
This command spawns a terminal with a shell prompt within the OpenEmbedded build environment.
The OE_TERMINAL
variable controls what type of shell is opened.
For spawned terminals, the following occurs:
The PATH
variable includes the
cross-toolchain.
The pkgconfig
variables find the correct
.pc
files.
The configure
command finds the
Yocto Project site files as well as any other necessary files.
Within this environment, you can run configure or compile
commands as if they were being run by
the OpenEmbedded build system itself.
As noted earlier, the working directory also automatically changes to the
Source Directory (S
).
To manually run a specific task using devshell
,
run the corresponding run.*
script in
the
${
WORKDIR
}/temp
directory (e.g.,
run.do_configure.
pid
).
If a task's script does not exist, which would be the case if the task was
skipped by way of the sstate cache, you can create the task by first running
it outside of the devshell
:
$ bitbake -c task
Execution of a task's run.*
script and BitBake's execution of a task are identical.
In other words, running the script re-runs the task
just as it would be run using the
bitbake -c
command.
Any run.*
file that does not
have a .pid
extension is a
symbolic link (symlink) to the most recent version of that
file.
Remember, that the devshell
is a mechanism that allows
you to get into the BitBake task execution environment.
And as such, all commands must be called just as BitBake would call them.
That means you need to provide the appropriate options for
cross-compilation and so forth as applicable.
When you are finished using devshell
, exit the shell
or close the terminal window.
It is worth remembering that when using devshell
you need to use the full compiler name such as arm-poky-linux-gnueabi-gcc
instead of just using gcc
.
The same applies to other applications such as binutils
,
libtool
and so forth.
BitBake sets up environment variables such as CC
to assist applications, such as make
to find the correct tools.
It is also worth noting that devshell
still works over
X11 forwarding and similar situations.
Similar to working within a development shell as described in
the previous section, you can also spawn and work within an
interactive Python development shell.
When debugging certain commands or even when just editing packages,
devpyshell
can be a useful tool.
When you invoke devpyshell
, all tasks up to and
including
do_patch
are run for the specified target.
Then a new terminal is opened.
Additionally, key Python objects and code are available in the same
way they are to BitBake tasks, in particular, the data store 'd'.
So, commands such as the following are useful when exploring the data
store and running functions:
pydevshell> d.getVar("STAGING_DIR") '/media/build1/poky/build/tmp/sysroots' pydevshell> d.getVar("STAGING_DIR") '${TMPDIR}/sysroots' pydevshell> d.setVar("FOO", "bar") pydevshell> d.getVar("FOO") 'bar' pydevshell> d.delVar("FOO") pydevshell> d.getVar("FOO") pydevshell> bb.build.exec_func("do_unpack", d) pydevshell>
The commands execute just as if the OpenEmbedded build system were executing them. Consequently, working this way can be helpful when debugging a build or preparing software to be used with the OpenEmbedded build system.
Following is an example that uses devpyshell
on a target named
matchbox-desktop
:
$ bitbake matchbox-desktop -c devpyshell
This command spawns a terminal and places you in an interactive
Python interpreter within the OpenEmbedded build environment.
The OE_TERMINAL
variable controls what type of shell is opened.
When you are finished using devpyshell
, you
can exit the shell either by using Ctrl+d or closing the terminal
window.
This section describes various build procedures. For example, the steps needed for a simple build, a target that uses multiple configurations, building an image for more than one machine, and so forth.
In the development environment, you need to build an image whenever you change hardware support, add or change system libraries, or add or change services that have dependencies. Several methods exist that allow you to build an image within the Yocto Project. This section presents the basic steps you need to build a simple image using BitBake from a build host running Linux.
For information on how to build an image using Toaster, see the Toaster User Manual.
For information on how to use
devtool
to build images, see
the
"Using devtool
in Your SDK Workflow"
section in the Yocto Project Application
Development and the Extensible Software Development
Kit (eSDK) manual.
For a quick example on how to build an image using the OpenEmbedded build system, see the Yocto Project Quick Build document.
The build process creates an entire Linux distribution from
source and places it in your
Build Directory
under tmp/deploy/images
.
For detailed information on the build process using BitBake,
see the
"Images"
section in the Yocto Project Overview and Concepts Manual.
The following figure and list overviews the build process:
Set up Your Host Development System to Support Development Using the Yocto Project: See the "Setting Up to Use the Yocto Project" section for options on how to get a build host ready to use the Yocto Project.
Initialize the Build Environment:
Initialize the build environment by sourcing the build
environment script (i.e.
oe-init-build-env
):
$ source oe-init-build-env [build_dir
]
When you use the initialization script, the
OpenEmbedded build system uses
build
as the default Build
Directory in your current work directory.
You can use a build_dir
argument with the script to specify a different build
directory.
~/build/x86
for a
qemux86
target, and
~/build/arm
for a
qemuarm
target.
Make Sure Your local.conf
File is Correct:
Ensure the conf/local.conf
configuration file, which is found in the Build
Directory, is set up how you want it.
This file defines many aspects of the build environment
including the target machine architecture through the
MACHINE
variable,
the packaging format used during the build
(PACKAGE_CLASSES
),
and a centralized tarball download directory through the
DL_DIR
variable.
Build the Image:
Build the image using the bitbake
command:
$ bitbake target
The target
is the name of the
recipe you want to build.
Common targets are the images in
meta/recipes-core/images
,
meta/recipes-sato/images
, and so
forth all found in the
Source Directory.
Or, the target can be the name of a recipe for a
specific piece of software such as BusyBox.
For more details about the images the OpenEmbedded build
system supports, see the
"Images"
chapter in the Yocto Project Reference Manual.
As an example, the following command builds the
core-image-minimal
image:
$ bitbake core-image-minimal
Once an image has been built, it often needs to be
installed.
The images and kernels built by the OpenEmbedded
build system are placed in the Build Directory in
tmp/deploy/images
.
For information on how to run pre-built images such as
qemux86
and qemuarm
,
see the
Yocto Project Application Development and the Extensible Software Development Kit (eSDK)
manual.
For information about how to install these images,
see the documentation for your particular board or
machine.
You can use a single bitbake
command
to build multiple images or packages for different targets
where each image or package requires a different configuration
(multiple configuration builds).
The builds, in this scenario, are sometimes referred to as
"multiconfigs", and this section uses that term throughout.
This section describes how to set up for multiple configuration builds and how to account for cross-build dependencies between the multiconfigs.
To accomplish a multiple configuration build, you must
define each target's configuration separately using
a parallel configuration file in the
Build Directory,
and you must follow a required file hierarchy.
Additionally, you must enable the multiple configuration
builds in your local.conf
file.
Follow these steps to set up and execute multiple configuration builds:
Create Separate Configuration Files:
You need to create a single configuration file for
each build target (each multiconfig).
Minimally, each configuration file must define the
machine and the temporary directory BitBake uses
for the build.
Suggested practice dictates that you do not
overlap the temporary directories
used during the builds.
However, it is possible that you can share the
temporary directory
(TMPDIR
).
For example, consider a scenario with two
different multiconfigs for the same
MACHINE
: "qemux86" built for
two distributions such as "poky" and "poky-lsb".
In this case, you might want to use the same
TMPDIR
.
Here is an example showing the minimal
statements needed in a configuration file for
a "qemux86" target whose temporary build directory
is tmpmultix86
:
MACHINE="qemux86" TMPDIR="${TOPDIR}/tmpmultix86"
The location for these multiconfig
configuration files is specific.
They must reside in the current build directory in
a sub-directory of conf
named
multiconfig
.
Following is an example that defines two
configuration files for the "x86" and "arm"
multiconfigs:
The reason for this required file hierarchy
is because the BBPATH
variable
is not constructed until the layers are parsed.
Consequently, using the configuration file as a
pre-configuration file is not possible unless it is
located in the current working directory.
Add the BitBake Multi-configuration Variable to the Local Configuration File:
Use the
BBMULTICONFIG
variable in your
conf/local.conf
configuration
file to specify each multiconfig.
Continuing with the example from the previous
figure, the BBMULTICONFIG
variable needs to enable two multiconfigs: "x86"
and "arm" by specifying each configuration file:
BBMULTICONFIG = "x86 arm"
local.conf
file.
Consequently, the previous example actually
adds two additional configurations to your
build: "arm" and "x86" along with "".
Launch BitBake: Use the following BitBake command form to launch the multiple configuration build:
$ bitbake [mc:multiconfigname
:]target
[[[mc:multiconfigname
:]target
] ... ]
For the example in this section, the following command applies:
$ bitbake mc:x86:core-image-minimal mc:arm:core-image-sato mc::core-image-base
The previous BitBake command builds a
core-image-minimal
image that
is configured through the
x86.conf
configuration file,
a core-image-sato
image that is configured through the
arm.conf
configuration file
and a core-image-base
that is
configured through your
local.conf
configuration file.
TMPDIR
directories, the build
either loads from an existing sstate cache for that
build at the start or builds the object fresh.
Sometimes dependencies can exist between targets
(multiconfigs) in a multiple configuration build.
For example, suppose that in order to build a
core-image-sato
image for an "x86"
multiconfig, the root filesystem of an "arm"
multiconfig must exist.
This dependency is essentially that the
do_image
task in the core-image-sato
recipe
depends on the completion of the
do_rootfs
task of the core-image-minimal
recipe.
To enable dependencies in a multiple configuration build, you must declare the dependencies in the recipe using the following statement form:
task_or_package
[mcdepends] = "mc:from_multiconfig
:to_multiconfig
:recipe_name
:task_on_which_to_depend
"
To better show how to use this statement, consider the
example scenario from the first paragraph of this section.
The following statement needs to be added to the recipe
that builds the core-image-sato
image:
do_image[mcdepends] = "mc:x86:arm:core-image-minimal:do_rootfs"
In this example, the
from_multiconfig
is "x86".
The to_multiconfig
is "arm".
The task on which the do_image
task
in the recipe depends is the do_rootfs
task from the core-image-minimal
recipe associated with the "arm" multiconfig.
Once you set up this dependency, you can build the "x86" multiconfig using a BitBake command as follows:
$ bitbake mc:x86:core-image-sato
This command executes all the tasks needed to create
the core-image-sato
image for the
"x86" multiconfig.
Because of the dependency, BitBake also executes through
the do_rootfs
task for the "arm"
multiconfig build.
Having a recipe depend on the root filesystem of another
build might not seem that useful.
Consider this change to the statement in the
core-image-sato
recipe:
do_image[mcdepends] = "mc:x86:arm:core-image-minimal:do_image"
In this case, BitBake must create the
core-image-minimal
image for the
"arm" build since the "x86" build depends on it.
Because "x86" and "arm" are enabled for multiple
configuration builds and have separate configuration
files, BitBake places the artifacts for each build in the
respective temporary build directories (i.e.
TMPDIR
).
An initial RAM filesystem (initramfs) image provides a temporary root filesystem used for early system initialization (e.g. loading of modules needed to locate and mount the "real" root filesystem).
Follow these steps to create an initramfs image:
Create the initramfs Image Recipe:
You can reference the
core-image-minimal-initramfs.bb
recipe found in the meta/recipes-core
directory of the
Source Directory
as an example from which to work.
Decide if You Need to Bundle the initramfs Image
Into the Kernel Image:
If you want the initramfs image that is built to be
bundled in with the kernel image, set the
INITRAMFS_IMAGE_BUNDLE
variable to "1" in your local.conf
configuration file and set the
INITRAMFS_IMAGE
variable in the recipe that builds the kernel image.
Setting the INITRAMFS_IMAGE_BUNDLE
flag causes the initramfs image to be unpacked
into the ${B}/usr/
directory.
The unpacked initramfs image is then passed to the kernel's
Makefile
using the
CONFIG_INITRAMFS_SOURCE
variable, allowing the initramfs image to be built into
the kernel normally.
INITRD_IMAGE
,
INITRD_LIVE
, and
INITRD_IMAGE_LIVE
variables.
For more information, see the
image-live.bbclass
file.
Optionally Add Items to the initramfs Image
Through the initramfs Image Recipe:
If you add items to the initramfs image by way of its
recipe, you should use
PACKAGE_INSTALL
rather than
IMAGE_INSTALL
.
PACKAGE_INSTALL
gives more direct
control of what is added to the image as compared to
the defaults you might not necessarily want that are
set by the
image
or
core-image
classes.
Build the Kernel Image and the initramfs
Image:
Build your kernel image using BitBake.
Because the initramfs image recipe is a dependency of the
kernel image, the initramfs image is built as well and
bundled with the kernel image if you used the
INITRAMFS_IMAGE_BUNDLE
variable described earlier.
Very small distributions have some significant advantages such as requiring less on-die or in-package memory (cheaper), better performance through efficient cache usage, lower power requirements due to less memory, faster boot times, and reduced development overhead. Some real-world examples where a very small distribution gives you distinct advantages are digital cameras, medical devices, and small headless systems.
This section presents information that shows you how you can
trim your distribution to even smaller sizes than the
poky-tiny
distribution, which is around
5 Mbytes, that can be built out-of-the-box using the Yocto Project.
The following list presents the overall steps you need to consider and perform to create distributions with smaller root filesystems, achieve faster boot times, maintain your critical functionality, and avoid initial RAM disks:
Before you can reach your destination, you need to know where you are going. Here is an example list that you can use as a guide when creating very small distributions:
Determine how much space you need (e.g. a kernel that is 1 Mbyte or less and a root filesystem that is 3 Mbytes or less).
Find the areas that are currently taking 90% of the space and concentrate on reducing those areas.
Do not create any difficult "hacks" to achieve your goals.
Leverage the device-specific options.
Work in a separate layer so that you keep changes isolated. For information on how to create layers, see the "Understanding and Creating Layers" section.
It is easiest to have something to start with when creating
your own distribution.
You can use the Yocto Project out-of-the-box to create the
poky-tiny
distribution.
Ultimately, you will want to make changes in your own
distribution that are likely modeled after
poky-tiny
.
poky-tiny
in your build,
set the
DISTRO
variable in your
local.conf
file to "poky-tiny"
as described in the
"Creating Your Own Distribution"
section.
Understanding some memory concepts will help you reduce the
system size.
Memory consists of static, dynamic, and temporary memory.
Static memory is the TEXT (code), DATA (initialized data
in the code), and BSS (uninitialized data) sections.
Dynamic memory represents memory that is allocated at runtime:
stacks, hash tables, and so forth.
Temporary memory is recovered after the boot process.
This memory consists of memory used for decompressing
the kernel and for the __init__
functions.
To help you see where you currently are with kernel and root
filesystem sizes, you can use two tools found in the
Source Directory in
the scripts/tiny/
directory:
ksize.py
: Reports
component sizes for the kernel build objects.
dirsize.py
: Reports
component sizes for the root filesystem.
This next tool and command help you organize configuration fragments and view file dependencies in a human-readable form:
merge_config.sh
:
Helps you manage configuration files and fragments
within the kernel.
With this tool, you can merge individual configuration
fragments together.
The tool allows you to make overrides and warns you
of any missing configuration options.
The tool is ideal for allowing you to iterate on
configurations, create minimal configurations, and
create configuration files for different machines
without having to duplicate your process.
The merge_config.sh
script is
part of the Linux Yocto kernel Git repositories
(i.e. linux-yocto-3.14
,
linux-yocto-3.10
,
linux-yocto-3.8
, and so forth)
in the
scripts/kconfig
directory.
For more information on configuration fragments, see the "Creating Configuration Fragments" section in the Yocto Project Linux Kernel Development Manual.
bitbake -u taskexp -g
:
Using the BitBake command with these options brings up
a Dependency Explorer from which you can view file
dependencies.
Understanding these dependencies allows you to make
informed decisions when cutting out various pieces of the
kernel and root filesystem.bitbake_target
The root filesystem is made up of packages for booting, libraries, and applications. To change things, you can configure how the packaging happens, which changes the way you build them. You can also modify the filesystem itself or select a different filesystem.
First, find out what is hogging your root filesystem by running the
dirsize.py
script from your root directory:
$ cd root-directory-of-image
$ dirsize.py 100000 > dirsize-100k.log
$ cat dirsize-100k.log
You can apply a filter to the script to ignore files under a certain size. The previous example filters out any files below 100 Kbytes. The sizes reported by the tool are uncompressed, and thus will be smaller by a relatively constant factor in a compressed root filesystem. When you examine your log file, you can focus on areas of the root filesystem that take up large amounts of memory.
You need to be sure that what you eliminate does not cripple the functionality you need. One way to see how packages relate to each other is by using the Dependency Explorer UI with the BitBake command:
$ cdimage-directory
$ bitbake -u taskexp -gimage
Use the interface to select potential packages you wish to eliminate and see their dependency relationships.
When deciding how to reduce the size, get rid of packages that
result in minimal impact on the feature set.
For example, you might not need a VGA display.
Or, you might be able to get by with devtmpfs
and mdev
instead of
udev
.
Use your local.conf
file to make changes.
For example, to eliminate udev
and
glib
, set the following in the
local configuration file:
VIRTUAL-RUNTIME_dev_manager = ""
Finally, you should consider exactly the type of root
filesystem you need to meet your needs while also reducing
its size.
For example, consider cramfs
,
squashfs
, ubifs
,
ext2
, or an initramfs
using initramfs
.
Be aware that ext3
requires a 1 Mbyte
journal.
If you are okay with running read-only, you do not need this
journal.
The kernel is built by including policies for hardware-independent aspects. What subsystems do you enable? For what architecture are you building? Which drivers do you build by default?
Run the ksize.py
script from the top-level
Linux build directory to get an idea of what is making up
the kernel:
$ cd top-level-linux-build-directory
$ ksize.py > ksize.log
$ cat ksize.log
When you examine the log, you will see how much space is
taken up with the built-in .o
files for
drivers, networking, core kernel files, filesystem, sound,
and so forth.
The sizes reported by the tool are uncompressed, and thus
will be smaller by a relatively constant factor in a compressed
kernel image.
Look to reduce the areas that are large and taking up around
the "90% rule."
To examine, or drill down, into any particular area, use the
-d
option with the script:
$ ksize.py -d > ksize.log
Using this option breaks out the individual file information for each area of the kernel (e.g. drivers, networking, and so forth).
Use your log file to see what you can eliminate from the kernel based on features you can let go. For example, if you are not going to need sound, you do not need any drivers that support sound.
After figuring out what to eliminate, you need to reconfigure
the kernel to reflect those changes during the next build.
You could run menuconfig
and make all your
changes at once.
However, that makes it difficult to see the effects of your
individual eliminations and also makes it difficult to replicate
the changes for perhaps another target device.
A better method is to start with no configurations using
allnoconfig
, create configuration
fragments for individual changes, and then manage the
fragments into a single configuration file using
merge_config.sh
.
The tool makes it easy for you to iterate using the
configuration change and build cycle.
Each time you make configuration changes, you need to rebuild the kernel and check to see what impact your changes had on the overall size.
Packaging requirements add size to the image. One way to reduce the size of the image is to remove all the packaging requirements from the image. This reduction includes both removing the package manager and its unique dependencies as well as removing the package management data itself.
To eliminate all the packaging requirements for an image,
be sure that "package-management" is not part of your
IMAGE_FEATURES
statement for the image.
When you remove this feature, you are removing the package
manager as well as its dependencies from the root filesystem.
Depending on your particular circumstances, other areas that you can trim likely exist. The key to finding these areas is through tools and methods described here combined with experimentation and iteration. Here are a couple of areas to experiment with:
glibc
:
In general, follow this process:
Remove glibc
features from
DISTRO_FEATURES
that you think you do not need.
Build your distribution.
If the build fails due to missing
symbols in a package, determine if you can
reconfigure the package to not need those
features.
For example, change the configuration to not
support wide character support as is done for
ncurses
.
Or, if support for those characters is needed,
determine what glibc
features provide the support and restore the
configuration.
Rebuild and repeat the process.
busybox
:
For BusyBox, use a process similar as described for
glibc
.
A difference is you will need to boot the resulting
system to see if you are able to do everything you
expect from the running system.
You need to be sure to integrate configuration fragments
into Busybox because BusyBox handles its own core
features and then allows you to add configuration
fragments on top.
If you have not reached your goals on system size, you need to iterate on the process. The process is the same. Use the tools and see just what is taking up 90% of the root filesystem and the kernel. Decide what you can eliminate without limiting your device beyond what you need.
Depending on your system, a good place to look might be Busybox, which provides a stripped down version of Unix tools in a single, executable file. You might be able to drop virtual terminal services or perhaps ipv6.
A common scenario developers face is creating images for several
different machines that use the same software environment.
In this situation, it is tempting to set the
tunings and optimization flags for each build specifically for
the targeted hardware (i.e. "maxing out" the tunings).
Doing so can considerably add to build times and package feed
maintenance collectively for the machines.
For example, selecting tunes that are extremely specific to a
CPU core used in a system might enable some micro optimizations
in GCC for that particular system but would otherwise not gain
you much of a performance difference across the other systems
as compared to using a more general tuning across all the builds
(e.g. setting
DEFAULTTUNE
specifically for each machine's build).
Rather than "max out" each build's tunings, you can take steps that
cause the OpenEmbedded build system to reuse software across the
various machines where it makes sense.
If build speed and package feed maintenance are considerations, you should consider the points in this section that can help you optimize your tunings to best consider build times and package feed maintenance.
Share the Build Directory:
If at all possible, share the
TMPDIR
across builds.
The Yocto Project supports switching between different
MACHINE
values in the same TMPDIR
.
This practice is well supported and regularly used by
developers when building for multiple machines.
When you use the same TMPDIR
for
multiple machine builds, the OpenEmbedded build system can
reuse the existing native and often cross-recipes for
multiple machines.
Thus, build time decreases.
DISTRO
settings change or fundamental configuration settings
such as the filesystem layout, you need to work with
a clean TMPDIR
.
Sharing TMPDIR
under these
circumstances might work but since it is not
guaranteed, you should use a clean
TMPDIR
.
Enable the Appropriate Package Architecture: By default, the OpenEmbedded build system enables three levels of package architectures: "all", "tune" or "package", and "machine". Any given recipe usually selects one of these package architectures (types) for its output. Depending for what a given recipe creates packages, making sure you enable the appropriate package architecture can directly impact the build time.
A recipe that just generates scripts can enable
"all" architecture because there are no binaries to build.
To specifically enable "all" architecture, be sure your
recipe inherits the
allarch
class.
This class is useful for "all" architectures because it
configures many variables so packages can be used across
multiple architectures.
If your recipe needs to generate packages that are
machine-specific or when one of the build or runtime
dependencies is already machine-architecture dependent,
which makes your recipe also machine-architecture dependent,
make sure your recipe enables the "machine" package
architecture through the
MACHINE_ARCH
variable:
PACKAGE_ARCH = "${MACHINE_ARCH}"
When you do not specifically enable a package
architecture through the
PACKAGE_ARCH
,
The OpenEmbedded build system defaults to the
TUNE_PKGARCH
setting:
PACKAGE_ARCH = "${TUNE_PKGARCH}"
Choose a Generic Tuning File if Possible:
Some tunes are more generic and can run on multiple targets
(e.g. an armv5
set of packages could
run on armv6
and
armv7
processors in most cases).
Similarly, i486
binaries could work
on i586
and higher processors.
You should realize, however, that advances on newer
processor versions would not be used.
If you select the same tune for several different machines, the OpenEmbedded build system reuses software previously built, thus speeding up the overall build time. Realize that even though a new sysroot for each machine is generated, the software is not recompiled and only one package feed exists.
Manage Granular Level Packaging:
Sometimes cases exist where injecting another level of
package architecture beyond the three higher levels noted
earlier can be useful.
For example, consider how NXP (formerly Freescale) allows
for the easy reuse of binary packages in their layer
meta-freescale
.
In this example, the
fsl-dynamic-packagearch
class shares GPU packages for i.MX53 boards because
all boards share the AMD GPU.
The i.MX6-based boards can do the same because all boards
share the Vivante GPU.
This class inspects the BitBake datastore to identify if
the package provides or depends on one of the
sub-architecture values.
If so, the class sets the
PACKAGE_ARCH
value based on the MACHINE_SUBARCH
value.
If the package does not provide or depend on one of the
sub-architecture values but it matches a value in the
machine-specific filter, it sets
MACHINE_ARCH
.
This behavior reduces the number of packages built and
saves build time by reusing binaries.
Use Tools to Debug Issues:
Sometimes you can run into situations where software is
being rebuilt when you think it should not be.
For example, the OpenEmbedded build system might not be
using shared state between machines when you think it
should be.
These types of situations are usually due to references
to machine-specific variables such as
MACHINE
,
SERIAL_CONSOLES
,
XSERVER
,
MACHINE_FEATURES
,
and so forth in code that is supposed to only be
tune-specific or when the recipe depends
(DEPENDS
,
RDEPENDS
,
RRECOMMENDS
,
RSUGGESTS
,
and so forth) on some other recipe that already has
PACKAGE_ARCH
defined as "${MACHINE_ARCH}".
For such cases, you can use some tools to help you sort out the situation:
sstate-diff-machines.sh
:
You can find this tool in the
scripts
directory of the
Source Repositories.
See the comments in the script for information on
how to use the tool.
BitBake's "-S printdiff" Option:
Using this option causes BitBake to try to
establish the closest signature match it can
(e.g. in the shared state cache) and then run
bitbake-diffsigs
over the
matches to determine the stamps and delta where
these two stamp trees diverge.
By default, the OpenEmbedded build system uses the Build Directory when building source code. The build process involves fetching the source files, unpacking them, and then patching them if necessary before the build takes place.
Situations exist where you might want to build software from source
files that are external to and thus outside of the
OpenEmbedded build system.
For example, suppose you have a project that includes a new BSP with
a heavily customized kernel.
And, you want to minimize exposing the build system to the
development team so that they can focus on their project and
maintain everyone's workflow as much as possible.
In this case, you want a kernel source directory on the development
machine where the development occurs.
You want the recipe's
SRC_URI
variable to point to the external directory and use it as is, not
copy it.
To build from software that comes from an external source, all you
need to do is inherit the
externalsrc
class and then set the
EXTERNALSRC
variable to point to your external source code.
Here are the statements to put in your
local.conf
file:
INHERIT += "externalsrc" EXTERNALSRC_pn-myrecipe
= "path-to-your-source-tree
"
This next example shows how to accomplish the same thing by setting
EXTERNALSRC
in the recipe itself or in the
recipe's append file:
EXTERNALSRC = "path
" EXTERNALSRC_BUILD = "path
"
externalsrc
class.
By default, externalsrc.bbclass
builds
the source code in a directory separate from the external source
directory as specified by
EXTERNALSRC
.
If you need to have the source built in the same directory in
which it resides, or some other nominated directory, you can set
EXTERNALSRC_BUILD
to point to that directory:
EXTERNALSRC_BUILD_pn-myrecipe
= "path-to-your-source-tree
"
It can be useful to take a "snapshot" of upstream sources used in a build and then use that "snapshot" later to replicate the build offline. To do so, you need to first prepare and populate your downloads directory your "snapshot" of files. Once your downloads directory is ready, you can use it at any time and from any machine to replicate your build.
Follow these steps to populate your Downloads directory:
Create a Clean Downloads Directory:
Start with an empty downloads directory
(DL_DIR
).
You start with an empty downloads directory by either
removing the files in the existing directory or by
setting
DL_DIR
to point to either an
empty location or one that does not yet exist.
Generate Tarballs of the Source Git Repositories:
Edit your local.conf
configuration
file as follows:
DL_DIR = "/home/your-download-dir
/"
BB_GENERATE_MIRROR_TARBALLS = "1"
During the fetch process in the next step, BitBake
gathers the source files and creates tarballs in
the directory pointed to by DL_DIR
.
See the
BB_GENERATE_MIRROR_TARBALLS
variable for more information.
Populate Your Downloads Directory Without Building: Use BitBake to fetch your sources but inhibit the build:
$ bitbake target
--runonly=fetch
The downloads directory (i.e.
${DL_DIR}
) now has a "snapshot" of
the source files in the form of tarballs, which can
be used for the build.
Optionally Remove Any Git or other SCM Subdirectories From the Downloads Directory:
If you want, you can clean up your downloads directory
by removing any Git or other Source Control Management
(SCM) subdirectories such as
${DL_DIR}/git2/*
.
The tarballs already contain these subdirectories.
Once your downloads directory has everything it needs regarding source files, you can create your "own-mirror" and build your target. Understand that you can use the files to build the target offline from any machine and at any time.
Follow these steps to build your target using the files in the downloads directory:
Using Local Files Only:
Inside your local.conf
file, add
the
SOURCE_MIRROR_URL
variable,
inherit the own-mirrors
class, and use the
BB_NO_NETWORK
variable to your local.conf
.
SOURCE_MIRROR_URL ?= "file:///home/your-download-dir
/"
INHERIT += "own-mirrors"
BB_NO_NETWORK = "1"
The SOURCE_MIRROR_URL
and
own-mirror
class set up the system
to use the downloads directory as your "own mirror".
Using the BB_NO_NETWORK
variable makes sure that BitBake's fetching process
in step 3 stays local, which means files from
your "own-mirror" are used.
Start With a Clean Build:
You can start with a clean build by removing the
${
TMPDIR
}
directory or using a new
Build Directory.
Build Your Target: Use BitBake to build your target:
$ bitbake target
The build completes using the known local "snapshot" of source files from your mirror. The resulting tarballs for your "snapshot" of source files are in the downloads directory.
The offline build does not work if recipes
attempt to find the latest version of software
by setting
SRCREV
to
${
AUTOREV
}
:
SRCREV = "${AUTOREV}"
When a recipe sets
SRCREV
to
${AUTOREV}
, the build system
accesses the network in an attempt to determine the
latest version of software from the SCM.
Typically, recipes that use
AUTOREV
are custom or
modified recipes.
Recipes that reside in public repositories
usually do not use AUTOREV
.
If you do have recipes that use
AUTOREV
, you can take steps to
still use the recipes in an offline build.
Do the following:
Use a configuration generated by enabling build history.
Use the
buildhistory-collect-srcrevs
command to collect the stored
SRCREV
values from
the build's history.
For more information on collecting these
values, see the
"Build History Package Information"
section.
Once you have the correct source
revisions, you can modify those recipes
to to set SRCREV
to specific versions of the software.
Build time can be an issue. By default, the build system uses simple controls to try and maximize build efficiency. In general, the default settings for all the following variables result in the most efficient build times when dealing with single socket systems (i.e. a single CPU). If you have multiple CPUs, you might try increasing the default values to gain more speed. See the descriptions in the glossary for each variable for more information:
BB_NUMBER_THREADS
:
The maximum number of threads BitBake simultaneously executes.
BB_NUMBER_PARSE_THREADS
:
The number of threads BitBake uses during parsing.
PARALLEL_MAKE
:
Extra options passed to the make
command
during the
do_compile
task in order to specify parallel compilation on the
local build host.
PARALLEL_MAKEINST
:
Extra options passed to the make
command
during the
do_install
task in order to specify parallel installation on the
local build host.
As mentioned, these variables all scale to the number of processor cores available on the build system. For single socket systems, this auto-scaling ensures that the build system fundamentally takes advantage of potential parallel operations during the build based on the build machine's capabilities.
Following are additional factors that can affect build speed:
File system type:
The file system type that the build is being performed on can
also influence performance.
Using ext4
is recommended as compared
to ext2
and ext3
due to ext4
improved features
such as extents.
Disabling the updating of access time using
noatime
:
The noatime
mount option prevents the
build system from updating file and directory access times.
Setting a longer commit: Using the "commit=" mount option increases the interval in seconds between disk cache writes. Changing this interval from the five second default to something longer increases the risk of data loss but decreases the need to write to the disk, thus increasing the build performance.
Choosing the packaging backend: Of the available packaging backends, IPK is the fastest. Additionally, selecting a singular packaging backend also helps.
Using tmpfs
for
TMPDIR
as a temporary file system:
While this can help speed up the build, the benefits are
limited due to the compiler using
-pipe
.
The build system goes to some lengths to avoid
sync()
calls into the
file system on the principle that if there was a significant
failure, the
Build Directory
contents could easily be rebuilt.
Inheriting the
rm_work
class:
Inheriting this class has shown to speed up builds due to
significantly lower amounts of data stored in the data
cache as well as on disk.
Inheriting this class also makes cleanup of
TMPDIR
faster, at the expense of being easily able to dive into the
source code.
File system maintainers have recommended that the fastest way
to clean up large numbers of files is to reformat partitions
rather than delete files due to the linear nature of
partitions.
This, of course, assumes you structure the disk partitions and
file systems in a way that this is practical.
Aside from the previous list, you should keep some trade offs in mind that can help you speed up the build:
Remove items from
DISTRO_FEATURES
that you might not need.
Exclude debug symbols and other debug information:
If you do not need these symbols and other debug information,
disabling the *-dbg
package generation
can speed up the build.
You can disable this generation by setting the
INHIBIT_PACKAGE_DEBUG_SPLIT
variable to "1".
Disable static library generation for recipes derived from
autoconf
or libtool
:
Following is an example showing how to disable static
libraries and still provide an override to handle exceptions:
STATICLIBCONF = "--disable-static" STATICLIBCONF_sqlite3-native = "" EXTRA_OECONF += "${STATICLIBCONF}"
Some recipes need static libraries in order to work
correctly (e.g. pseudo-native
needs sqlite3-native
).
Overrides, as in the previous example, account for
these kinds of exceptions.
Some packages have packaging code that assumes the presence of the static libraries. If so, you might need to exclude them as well.
Libraries are an integral part of your system. This section describes some common practices you might find helpful when working with libraries to build your system:
If you are building a library and the library offers static linking, you can control
which static library files (*.a
files) get included in the
built library.
The PACKAGES
and FILES_*
variables in the
meta/conf/bitbake.conf
configuration file define how files installed
by the do_install
task are packaged.
By default, the PACKAGES
variable includes
${PN}-staticdev
, which represents all static library files.
${PN}-dev
.
Following is part of the BitBake configuration file, where you can see how the static library files are defined:
PACKAGE_BEFORE_PN ?= "" PACKAGES = "${PN}-dbg ${PN}-staticdev ${PN}-dev ${PN}-doc ${PN}-locale ${PACKAGE_BEFORE_PN} ${PN}" PACKAGES_DYNAMIC = "^${PN}-locale-.*" FILES = "" FILES_${PN} = "${bindir}/* ${sbindir}/* ${libexecdir}/* ${libdir}/lib*${SOLIBS} \ ${sysconfdir} ${sharedstatedir} ${localstatedir} \ ${base_bindir}/* ${base_sbindir}/* \ ${base_libdir}/*${SOLIBS} \ ${base_prefix}/lib/udev/rules.d ${prefix}/lib/udev/rules.d \ ${datadir}/${BPN} ${libdir}/${BPN}/* \ ${datadir}/pixmaps ${datadir}/applications \ ${datadir}/idl ${datadir}/omf ${datadir}/sounds \ ${libdir}/bonobo/servers" FILES_${PN}-bin = "${bindir}/* ${sbindir}/*" FILES_${PN}-doc = "${docdir} ${mandir} ${infodir} ${datadir}/gtk-doc \ ${datadir}/gnome/help" SECTION_${PN}-doc = "doc" FILES_SOLIBSDEV ?= "${base_libdir}/lib*${SOLIBSDEV} ${libdir}/lib*${SOLIBSDEV}" FILES_${PN}-dev = "${includedir} ${FILES_SOLIBSDEV} ${libdir}/*.la \ ${libdir}/*.o ${libdir}/pkgconfig ${datadir}/pkgconfig \ ${datadir}/aclocal ${base_libdir}/*.o \ ${libdir}/${BPN}/*.la ${base_libdir}/*.la" SECTION_${PN}-dev = "devel" ALLOW_EMPTY_${PN}-dev = "1" RDEPENDS_${PN}-dev = "${PN} (= ${EXTENDPKGV})" FILES_${PN}-staticdev = "${libdir}/*.a ${base_libdir}/*.a ${libdir}/${BPN}/*.a" SECTION_${PN}-staticdev = "devel" RDEPENDS_${PN}-staticdev = "${PN}-dev (= ${EXTENDPKGV})"
The build system offers the ability to build libraries with different target optimizations or architecture formats and combine these together into one system image. You can link different binaries in the image against the different libraries as needed for specific use cases. This feature is called "Multilib."
An example would be where you have most of a system compiled in 32-bit mode using 32-bit libraries, but you have something large, like a database engine, that needs to be a 64-bit application and uses 64-bit libraries. Multilib allows you to get the best of both 32-bit and 64-bit libraries.
While the Multilib feature is most commonly used for 32 and 64-bit differences, the approach the build system uses facilitates different target optimizations. You could compile some binaries to use one set of libraries and other binaries to use a different set of libraries. The libraries could differ in architecture, compiler options, or other optimizations.
Several examples exist in the
meta-skeleton
layer found in the
Source Directory:
conf/multilib-example.conf
configuration file
conf/multilib-example2.conf
configuration file
recipes-multilib/images/core-image-multilib-example.bb
recipe
User-specific requirements drive the Multilib feature. Consequently, there is no one "out-of-the-box" configuration that likely exists to meet your needs.
In order to enable Multilib, you first need to ensure your recipe is
extended to support multiple libraries.
Many standard recipes are already extended and support multiple libraries.
You can check in the meta/conf/multilib.conf
configuration file in the
Source Directory to see how this is
done using the
BBCLASSEXTEND
variable.
Eventually, all recipes will be covered and this list will
not be needed.
For the most part, the Multilib class extension works automatically to
extend the package name from ${PN}
to
${MLPREFIX}${PN}
, where MLPREFIX
is the particular multilib (e.g. "lib32-" or "lib64-").
Standard variables such as
DEPENDS
,
RDEPENDS
,
RPROVIDES
,
RRECOMMENDS
,
PACKAGES
, and
PACKAGES_DYNAMIC
are automatically extended by the system.
If you are extending any manual code in the recipe, you can use the
${MLPREFIX}
variable to ensure those names are extended
correctly.
This automatic extension code resides in multilib.bbclass
.
After you have set up the recipes, you need to define the actual
combination of multiple libraries you want to build.
You accomplish this through your local.conf
configuration file in the
Build Directory.
An example configuration would be as follows:
MACHINE = "qemux86-64" require conf/multilib.conf MULTILIBS = "multilib:lib32" DEFAULTTUNE_virtclass-multilib-lib32 = "x86" IMAGE_INSTALL_append = " lib32-glib-2.0"
This example enables an
additional library named lib32
alongside the
normal target packages.
When combining these "lib32" alternatives, the example uses "x86" for tuning.
For information on this particular tuning, see
meta/conf/machine/include/ia32/arch-ia32.inc
.
The example then includes lib32-glib-2.0
in all the images, which illustrates one method of including a
multiple library dependency.
You can use a normal image build to include this dependency,
for example:
$ bitbake core-image-sato
You can also build Multilib packages specifically with a command like this:
$ bitbake lib32-glib-2.0
Generic implementation details as well as details that are specific to package management systems exist. Following are implementation details that exist regardless of the package management system:
The typical convention used for the
class extension code as used by
Multilib assumes that all package names specified
in
PACKAGES
that contain ${PN}
have
${PN}
at the start of the name.
When that convention is not followed and
${PN}
appears at
the middle or the end of a name, problems occur.
The
TARGET_VENDOR
value under Multilib will be extended to
"-vendor
mlmultilib
"
(e.g. "-pokymllib32" for a "lib32" Multilib with
Poky).
The reason for this slightly unwieldy contraction
is that any "-" characters in the vendor
string presently break Autoconf's
config.sub
, and
other separators are problematic for different
reasons.
For the RPM Package Management System, the following implementation details exist:
A unique architecture is defined for the Multilib packages,
along with creating a unique deploy folder under
tmp/deploy/rpm
in the
Build Directory.
For example, consider lib32
in a
qemux86-64
image.
The possible architectures in the system are "all", "qemux86_64",
"lib32_qemux86_64", and "lib32_x86".
The ${MLPREFIX}
variable is stripped from
${PN}
during RPM packaging.
The naming for a normal RPM package and a Multilib RPM package in a
qemux86-64
system resolves to something similar to
bash-4.1-r2.x86_64.rpm
and
bash-4.1.r2.lib32_x86.rpm
, respectively.
When installing a Multilib image, the RPM backend first installs the base image and then installs the Multilib libraries.
The build system relies on RPM to resolve the identical files in the two (or more) Multilib packages.
For the IPK Package Management System, the following implementation details exist:
The ${MLPREFIX}
is not stripped from
${PN}
during IPK packaging.
The naming for a normal RPM package and a Multilib IPK package in a
qemux86-64
system resolves to something like
bash_4.1-r2.x86_64.ipk
and
lib32-bash_4.1-rw_x86.ipk
, respectively.
The IPK deploy folder is not modified with
${MLPREFIX}
because packages with and without
the Multilib feature can exist in the same folder due to the
${PN}
differences.
IPK defines a sanity check for Multilib installation using certain rules for file comparison, overridden, etc.
Situations can exist where you need to install and use multiple versions of the same library on the same system at the same time. These situations almost always exist when a library API changes and you have multiple pieces of software that depend on the separate versions of the library. To accommodate these situations, you can install multiple versions of the same library in parallel on the same system.
The process is straightforward as long as the libraries use
proper versioning.
With properly versioned libraries, all you need to do to
individually specify the libraries is create separate,
appropriately named recipes where the
PN
part of the
name includes a portion that differentiates each library version
(e.g.the major part of the version number).
Thus, instead of having a single recipe that loads one version
of a library (e.g. clutter
), you provide
multiple recipes that result in different versions
of the libraries you want.
As an example, the following two recipes would allow the
two separate versions of the clutter
library to co-exist on the same system:
clutter-1.6_1.6.20.bb clutter-1.8_1.8.4.bb
Additionally, if you have other recipes that depend on a given
library, you need to use the
DEPENDS
variable to create the dependency.
Continuing with the same example, if you want to have a recipe
depend on the 1.8 version of the clutter
library, use the following in your recipe:
DEPENDS = "clutter-1.8"
x32 processor-specific Application Binary Interface (x32 psABI) is a native 32-bit processor-specific ABI for Intel® 64 (x86-64) architectures. An ABI defines the calling conventions between functions in a processing environment. The interface determines what registers are used and what the sizes are for various C data types.
Some processing environments prefer using 32-bit applications even when running on Intel 64-bit platforms. Consider the i386 psABI, which is a very old 32-bit ABI for Intel 64-bit platforms. The i386 psABI does not provide efficient use and access of the Intel 64-bit processor resources, leaving the system underutilized. Now consider the x86_64 psABI. This ABI is newer and uses 64-bits for data sizes and program pointers. The extra bits increase the footprint size of the programs, libraries, and also increases the memory and file system size requirements. Executing under the x32 psABI enables user programs to utilize CPU and system resources more efficiently while keeping the memory footprint of the applications low. Extra bits are used for registers but not for addressing mechanisms.
The Yocto Project supports the final specifications of x32 psABI as follows:
You can create packages and images in x32 psABI format on x86_64 architecture targets.
You can successfully build recipes with the x32 toolchain.
You can create and boot
core-image-minimal
and
core-image-sato
images.
RPM Package Manager (RPM) support exists for x32 binaries.
Support for large images exists.
To use the x32 psABI, you need to edit your
conf/local.conf
configuration file as
follows:
MACHINE = "qemux86-64" DEFAULTTUNE = "x86-64-x32" baselib = "${@d.getVar('BASE_LIB_tune-' + (d.getVar('DEFAULTTUNE') \ or 'INVALID')) or 'lib'}"
Once you have set up your configuration file, use BitBake to build an image that supports the x32 psABI. Here is an example:
$ bitbake core-image-sato
GObject introspection is the standard mechanism for accessing GObject-based software from runtime environments. GObject is a feature of the GLib library that provides an object framework for the GNOME desktop and related software. GObject Introspection adds information to GObject that allows objects created within it to be represented across different programming languages. If you want to construct GStreamer pipelines using Python, or control UPnP infrastructure using Javascript and GUPnP, GObject introspection is the only way to do it.
This section describes the Yocto Project support for generating
and packaging GObject introspection data.
GObject introspection data is a description of the
API provided by libraries built on top of GLib framework,
and, in particular, that framework's GObject mechanism.
GObject Introspection Repository (GIR) files go to
-dev
packages,
typelib
files go to main packages as they
are packaged together with libraries that are introspected.
The data is generated when building such a library, by linking the library with a small executable binary that asks the library to describe itself, and then executing the binary and processing its output.
Generating this data in a cross-compilation environment is difficult because the library is produced for the target architecture, but its code needs to be executed on the build host. This problem is solved with the OpenEmbedded build system by running the code through QEMU, which allows precisely that. Unfortunately, QEMU does not always work perfectly as mentioned in the "Known Issues" section.
Enabling the generation of introspection data (GIR files) in your library package involves the following:
Inherit the
gobject-introspection
class.
Make sure introspection is not disabled anywhere in
the recipe or from anything the recipe includes.
Also, make sure that "gobject-introspection-data" is
not in
DISTRO_FEATURES_BACKFILL_CONSIDERED
and that "qemu-usermode" is not in
MACHINE_FEATURES_BACKFILL_CONSIDERED
.
If either of these conditions exist, nothing will
happen.
Try to build the recipe.
If you encounter build errors that look like
something is unable to find
.so
libraries, check where these
libraries are located in the source tree and add
the following to the recipe:
GIR_EXTRA_LIBS_PATH = "${B}/something
/.libs"
oe-core
repository that use that
GIR_EXTRA_LIBS_PATH
variable
as an example.
Look for any other errors, which probably mean that introspection support in a package is not entirely standard, and thus breaks down in a cross-compilation environment. For such cases, custom-made fixes are needed. A good place to ask and receive help in these cases is the Yocto Project mailing lists.
You might find that you do not want to generate introspection data. Or, perhaps QEMU does not work on your build host and target architecture combination. If so, you can use either of the following methods to disable GIR file generations:
Add the following to your distro configuration:
DISTRO_FEATURES_BACKFILL_CONSIDERED = "gobject-introspection-data"
Adding this statement disables generating introspection data using QEMU but will still enable building introspection tools and libraries (i.e. building them does not require the use of QEMU).
Add the following to your machine configuration:
MACHINE_FEATURES_BACKFILL_CONSIDERED = "qemu-usermode"
Adding this statement disables the use of QEMU when building packages for your machine. Currently, this feature is used only by introspection recipes and has the same effect as the previously described option.
If you disable introspection data, you can still obtain it through other means such as copying the data from a suitable sysroot, or by generating it on the target hardware. The OpenEmbedded build system does not currently provide specific support for these techniques.
Use the following procedure to test if generating introspection data is working in an image:
Make sure that "gobject-introspection-data" is not in
DISTRO_FEATURES_BACKFILL_CONSIDERED
and that "qemu-usermode" is not in
MACHINE_FEATURES_BACKFILL_CONSIDERED
.
Build core-image-sato
.
Launch a Terminal and then start Python in the terminal.
Enter the following in the terminal:
>>> from gi.repository import GLib >>> GLib.get_host_name()
For something a little more advanced, enter the following:
http://python-gtk-3-tutorial.readthedocs.org/en/latest/introduction.html
The following know issues exist for GObject Introspection Support:
qemu-ppc64
immediately crashes.
Consequently, you cannot build introspection data on
that architecture.
x32 is not supported by QEMU. Consequently, introspection data is disabled.
musl causes transient GLib binaries to crash on assertion failures. Consequently, generating introspection data is disabled.
Because QEMU is not able to run the binaries correctly,
introspection is disabled for some specific packages
under specific architectures (e.g.
gcr
,
libsecret
, and
webkit
).
QEMU usermode might not work properly when running 64-bit binaries under 32-bit host machines. In particular, "qemumips64" is known to not work under i686.
You might want to use an external toolchain as part of your development. If this is the case, the fundamental steps you need to accomplish are as follows:
Understand where the installed toolchain resides. For cases where you need to build the external toolchain, you would need to take separate steps to build and install the toolchain.
Make sure you add the layer that contains the toolchain to
your bblayers.conf
file through the
BBLAYERS
variable.
Set the EXTERNAL_TOOLCHAIN
variable in your local.conf
file
to the location in which you installed the toolchain.
A good example of an external toolchain used with the Yocto Project
is Mentor Graphics®
Sourcery G++ Toolchain.
You can see information on how to use that particular layer in the
README
file at
http://github.com/MentorEmbedded/meta-sourcery/.
You can find further information by reading about the
TCMODE
variable in the Yocto Project Reference Manual's variable glossary.
Creating an image for a particular hardware target using the OpenEmbedded build system does not necessarily mean you can boot that image as is on your device. Physical devices accept and boot images in various ways depending on the specifics of the device. Usually, information about the hardware can tell you what image format the device requires. Should your device require multiple partitions on an SD card, flash, or an HDD, you can use the OpenEmbedded Image Creator, Wic, to create the properly partitioned image.
The wic
command generates partitioned
images from existing OpenEmbedded build artifacts.
Image generation is driven by partitioning commands
contained in an Openembedded kickstart file
(.wks
) specified either directly on
the command line or as one of a selection of canned
kickstart files as shown with the
wic list images
command in the
"Using an Existing Kickstart File"
section.
When you apply the command to a given set of build
artifacts, the result is an image or set of images that
can be directly written onto media and used on a particular
system.
.wks
) Reference"
Chapter in the Yocto Project Reference Manual.
The wic
command and the infrastructure
it is based on is by definition incomplete.
The purpose of the command is to allow the generation of
customized images, and as such, was designed to be
completely extensible through a plugin interface.
See the
"Using the Wic PlugIn Interface"
section for information on these plugins.
This section provides some background information on Wic, describes what you need to have in place to run the tool, provides instruction on how to use the Wic utility, provides information on using the Wic plugins interface, and provides several examples that show how to use Wic.
This section provides some background on the Wic utility. While none of this information is required to use Wic, you might find it interesting.
The name "Wic" is derived from OpenEmbedded Image Creator (oeic). The "oe" diphthong in "oeic" was promoted to the letter "w", because "oeic" is both difficult to remember and to pronounce.
Wic is loosely based on the
Meego Image Creator (mic
)
framework.
The Wic implementation has been
heavily modified to make direct use of OpenEmbedded
build artifacts instead of package installation and
configuration, which are already incorporated within
the OpenEmbedded artifacts.
Wic is a completely independent
standalone utility that initially provides
easier-to-use and more flexible replacements for an
existing functionality in OE-Core's
image-live
class.
The difference between Wic and those examples is
that with Wic the functionality of those scripts is
implemented by a general-purpose partitioning language,
which is based on Redhat kickstart syntax.
In order to use the Wic utility with the OpenEmbedded Build system, your system needs to meet the following requirements:
The Linux distribution on your development host must support the Yocto Project. See the "Supported Linux Distributions" section in the Yocto Project Reference Manual for the list of distributions that support the Yocto Project.
The standard system utilities, such as
cp
, must be installed on your
development host system.
You must have sourced the build environment
setup script (i.e.
oe-init-build-env
)
found in the
Build Directory.
You need to have the build artifacts already
available, which typically means that you must
have already created an image using the
Openembedded build system (e.g.
core-image-minimal
).
While it might seem redundant to generate an image
in order to create an image using
Wic, the current version of
Wic requires the artifacts
in the form generated by the OpenEmbedded build
system.
You must build several native tools, which are built to run on the build system:
$ bitbake parted-native dosfstools-native mtools-native
Include "wic" as part of the
IMAGE_FSTYPES
variable.
Include the name of the
wic kickstart file
as part of the
WKS_FILE
variable
You can get general help for the wic
command by entering the wic
command
by itself or by entering the command with a help argument
as follows:
$ wic -h $ wic --help $ wic help
Currently, Wic supports seven commands:
cp
, create
,
help
, list
,
ls
, rm
, and
write
.
You can get help for all these commands except "help" by
using the following form:
$ wic help command
For example, the following command returns help for the
write
command:
$ wic help write
Wic supports help for three topics:
overview
,
plugins
, and
kickstart
.
You can get help for any topic using the following form:
$ wic help topic
For example, the following returns overview help for Wic:
$ wic help overview
One additional level of help exists for Wic.
You can get help on individual images through the
list
command.
You can use the list
command to return the
available Wic images as follows:
$ wic list images genericx86 Create an EFI disk image for genericx86* beaglebone-yocto Create SD card image for Beaglebone edgerouter Create SD card image for Edgerouter qemux86-directdisk Create a qemu machine 'pcbios' direct disk image directdisk-gpt Create a 'pcbios' direct disk image mkefidisk Create an EFI disk image directdisk Create a 'pcbios' direct disk image systemd-bootdisk Create an EFI disk image with systemd-boot mkhybridiso Create a hybrid ISO image sdimage-bootpart Create SD card image with a boot partition directdisk-multi-rootfs Create multi rootfs image using rootfs plugin directdisk-bootloader-config Create a 'pcbios' direct disk image with custom bootloader config
Once you know the list of available Wic images, you can use
help
with the command to get help on a
particular image.
For example, the following command returns help on the
"beaglebone-yocto" image:
$ wic list beaglebone-yocto help Creates a partitioned SD card image for Beaglebone. Boot files are located in the first vfat partition.
You can use Wic in two different modes, depending on how much control you need for specifying the Openembedded build artifacts that are used for creating the image: Raw and Cooked:
Raw Mode: You explicitly specify build artifacts through Wic command-line arguments.
Cooked Mode:
The current
MACHINE
setting and image name are used to automatically
locate and provide the build artifacts.
You just supply a kickstart file and the name
of the image from which to use artifacts.
Regardless of the mode you use, you need to have the build artifacts ready and available.
Running Wic in raw mode allows you to specify all the
partitions through the wic
command line.
The primary use for raw mode is if you have built
your kernel outside of the Yocto Project
Build Directory.
In other words, you can point to arbitrary kernel,
root filesystem locations, and so forth.
Contrast this behavior with cooked mode where Wic
looks in the Build Directory (e.g.
tmp/deploy/images/
machine
).
The general form of the
wic
command in raw mode is:
$ wic createwks_file
options
... Where:wks_file
: An OpenEmbedded kickstart file. You can provide your own custom file or use a file from a set of existing files as described by further options. optional arguments: -h, --help show this help message and exit -oOUTDIR
, --outdirOUTDIR
name of directory to create image in -eIMAGE_NAME
, --image-nameIMAGE_NAME
name of the image to use the artifacts from e.g. core- image-sato -rROOTFS_DIR
, --rootfs-dirROOTFS_DIR
path to the /rootfs dir to use as the .wks rootfs source -bBOOTIMG_DIR
, --bootimg-dirBOOTIMG_DIR
path to the dir containing the boot artifacts (e.g. /EFI or /syslinux dirs) to use as the .wks bootimg source -kKERNEL_DIR
, --kernel-dirKERNEL_DIR
path to the dir containing the kernel to use in the .wks bootimg -nNATIVE_SYSROOT
, --native-sysrootNATIVE_SYSROOT
path to the native sysroot containing the tools to use to build the image -s, --skip-build-check skip the build check -f, --build-rootfs build rootfs -c {gzip,bzip2,xz}, --compress-with {gzip,bzip2,xz} compress image with specified compressor -m, --bmap generate .bmap --no-fstab-update Do not change fstab file. -vVARS_DIR
, --varsVARS_DIR
directory with <image>.env files that store bitbake variables -D, --debug output debug information
Running Wic in cooked mode leverages off artifacts in
the Build Directory.
In other words, you do not have to specify kernel or
root filesystem locations as part of the command.
All you need to provide is a kickstart file and the
name of the image from which to use artifacts by using
the "-e" option.
Wic looks in the Build Directory (e.g.
tmp/deploy/images/
machine
)
for artifacts.
The general form of the wic
command using Cooked Mode is as follows:
$ wic createwks_file
-eIMAGE_NAME
Where:wks_file
: An OpenEmbedded kickstart file. You can provide your own custom file or use a file from a set of existing files provided with the Yocto Project release. required argument: -eIMAGE_NAME
, --image-nameIMAGE_NAME
name of the image to use the artifacts from e.g. core- image-sato
If you do not want to create your own kickstart file, you can use an existing file provided by the Wic installation. As shipped, kickstart files can be found in the Yocto Project Source Repositories in the following two locations:
poky/meta-yocto-bsp/wic poky/scripts/lib/wic/canned-wks
Use the following command to list the available kickstart files:
$ wic list images genericx86 Create an EFI disk image for genericx86* beaglebone-yocto Create SD card image for Beaglebone edgerouter Create SD card image for Edgerouter qemux86-directdisk Create a qemu machine 'pcbios' direct disk image directdisk-gpt Create a 'pcbios' direct disk image mkefidisk Create an EFI disk image directdisk Create a 'pcbios' direct disk image systemd-bootdisk Create an EFI disk image with systemd-boot mkhybridiso Create a hybrid ISO image sdimage-bootpart Create SD card image with a boot partition directdisk-multi-rootfs Create multi rootfs image using rootfs plugin directdisk-bootloader-config Create a 'pcbios' direct disk image with custom bootloader config
When you use an existing file, you do not have to use the
.wks
extension.
Here is an example in Raw Mode that uses the
directdisk
file:
$ wic create directdisk -rrootfs_dir
-bbootimg_dir
\ -kkernel_dir
-nnative_sysroot
Here are the actual partition language commands
used in the genericx86.wks
file to
generate an image:
# short-description: Create an EFI disk image for genericx86* # long-description: Creates a partitioned EFI disk image for genericx86* machines part /boot --source bootimg-efi --sourceparams="loader=grub-efi" --ondisk sda --label msdos --active --align 1024 part / --source rootfs --ondisk sda --fstype=ext4 --label platform --align 1024 --use-uuid part swap --ondisk sda --size 44 --label swap1 --fstype=swap bootloader --ptable gpt --timeout=5 --append="rootfstype=ext4 console=ttyS0,115200 console=tty0"
You can extend and specialize Wic functionality by using Wic plugins. This section explains the Wic plugin interface.
Source plugins provide a mechanism to customize partition
content during the Wic image generation process.
You can use source plugins to map values that you specify
using --source
commands in kickstart
files (i.e. *.wks
) to a plugin
implementation used to populate a given partition.
WKS_FILE_DEPENDS
variable.
Source plugins are subclasses defined in plugin files. As shipped, the Yocto Project provides several plugin files. You can see the source plugin files that ship with the Yocto Project here. Each of these plugin files contains source plugins that are designed to populate a specific Wic image partition.
Source plugins are subclasses of the
SourcePlugin
class, which is
defined in the
poky/scripts/lib/wic/pluginbase.py
file.
For example, the BootimgEFIPlugin
source plugin found in the
bootimg-efi.py
file is a subclass of
the SourcePlugin
class, which is found
in the pluginbase.py
file.
You can also implement source plugins in a layer outside
of the Source Repositories (external layer).
To do so, be sure that your plugin files are located in
a directory whose path is
scripts/lib/wic/plugins/source/
within your external layer.
When the plugin files are located there, the source
plugins they contain are made available to Wic.
When the Wic implementation needs to invoke a
partition-specific implementation, it looks for the plugin
with the same name as the --source
parameter used in the kickstart file given to that
partition.
For example, if the partition is set up using the following
command in a kickstart file:
part /boot --source bootimg-pcbios --ondisk sda --label boot --active --align 1024
The methods defined as class members of the matching
source plugin (i.e. bootimg-pcbios
)
in the bootimg-pcbios.py
plugin file
are used.
To be more concrete, here is the corresponding plugin
definition from the bootimg-pcbios.py
file for the previous command along with an example
method called by the Wic implementation when it needs to
prepare a partition using an implementation-specific
function:
. . . class BootimgPcbiosPlugin(SourcePlugin): """ Create MBR boot partition and install syslinux on it. """ name = 'bootimg-pcbios' . . . @classmethod def do_prepare_partition(cls, part, source_params, creator, cr_workdir, oe_builddir, bootimg_dir, kernel_dir, rootfs_dir, native_sysroot): """ Called to do the actual content population for a partition i.e. it 'prepares' the partition to be incorporated into the image. In this case, prepare content for legacy bios boot partition. """ . . .
If a subclass (plugin) itself does not implement a
particular function, Wic locates and uses the default
version in the superclass.
It is for this reason that all source plugins are derived
from the SourcePlugin
class.
The SourcePlugin
class defined in
the pluginbase.py
file defines
a set of methods that source plugins can implement or
override.
Any plugins (subclass of
SourcePlugin
) that do not implement
a particular method inherit the implementation of the
method from the SourcePlugin
class.
For more information, see the
SourcePlugin
class in the
pluginbase.py
file for details:
The following list describes the methods implemented in the
SourcePlugin
class:
do_prepare_partition()
:
Called to populate a partition with actual content.
In other words, the method prepares the final
partition image that is incorporated into the
disk image.
do_configure_partition()
:
Called before
do_prepare_partition()
to
create custom configuration files for a partition
(e.g. syslinux or grub configuration files).
do_install_disk()
:
Called after all partitions have been prepared and
assembled into a disk image.
This method provides a hook to allow finalization
of a disk image (e.g. writing an MBR).
do_stage_partition()
:
Special content-staging hook called before
do_prepare_partition()
.
This method is normally empty.
Typically, a partition just uses the passed-in
parameters (e.g. the unmodified value of
bootimg_dir
).
However, in some cases, things might need to be
more tailored.
As an example, certain files might additionally
need to be taken from
bootimg_dir + /boot
.
This hook allows those files to be staged in a
customized fashion.
get_bitbake_var()
allows you to access non-standard variables
that you might want to use for this
behavior.
You can extend the source plugin mechanism.
To add more hooks, create more source plugin methods
within SourcePlugin
and the
corresponding derived subclasses.
The code that calls the plugin methods uses the
plugin.get_source_plugin_methods()
function to find the method or methods needed by the call.
Retrieval of those methods is accomplished by filling up
a dict with keys that contain the method names of interest.
On success, these will be filled in with the actual
methods.
See the Wic implementation for examples and details.
This section provides several examples that show how to use
the Wic utility.
All the examples assume the list of requirements in the
"Requirements"
section have been met.
The examples assume the previously generated image is
core-image-minimal
.
This example runs in Cooked Mode and uses the
mkefidisk
kickstart file:
$ wic create mkefidisk -e core-image-minimal INFO: Building wic-tools... . . . INFO: The new image(s) can be found here: ./mkefidisk-201804191017-sda.direct The following build artifacts were used to create the image(s): ROOTFS_DIR: /home/stephano/build/master/build/tmp-glibc/work/qemux86-oe-linux/core-image-minimal/1.0-r0/rootfs BOOTIMG_DIR: /home/stephano/build/master/build/tmp-glibc/work/qemux86-oe-linux/core-image-minimal/1.0-r0/recipe-sysroot/usr/share KERNEL_DIR: /home/stephano/build/master/build/tmp-glibc/deploy/images/qemux86 NATIVE_SYSROOT: /home/stephano/build/master/build/tmp-glibc/work/i586-oe-linux/wic-tools/1.0-r0/recipe-sysroot-native INFO: The image(s) were created using OE kickstart file: /home/stephano/build/master/openembedded-core/scripts/lib/wic/canned-wks/mkefidisk.wks
The previous example shows the easiest way to create
an image by running in cooked mode and supplying
a kickstart file and the "-e" option to point to the
existing build artifacts.
Your local.conf
file needs to have
the
MACHINE
variable set to the machine you are using, which is
"qemux86" in this example.
Once the image builds, the output provides image location, artifact use, and kickstart file information.
Continuing with the example, you can now write the
image from the Build Directory onto a USB stick, or
whatever media for which you built your image, and boot
from the media.
You can write the image by using
bmaptool
or
dd
:
$ oe-run-native bmaptool copy mkefidisk-201804191017-sda.direct /dev/sdX
or
$ sudo dd if=mkefidisk-201804191017-sda.direct of=/dev/sdX
bmaptool
to flash a device
with an image, see the
"Flashing Images Using bmaptool
"
section.
Because partitioned image creation is driven by the
kickstart file, it is easy to affect image creation by
changing the parameters in the file.
This next example demonstrates that through modification
of the directdisk-gpt
kickstart
file.
As mentioned earlier, you can use the command
wic list images
to show the list
of existing kickstart files.
The directory in which the
directdisk-gpt.wks
file resides is
scripts/lib/image/canned-wks/
,
which is located in the
Source Directory
(e.g. poky
).
Because available files reside in this directory,
you can create and add your own custom files to the
directory.
Subsequent use of the
wic list images
command would then
include your kickstart files.
In this example, the existing
directdisk-gpt
file already does
most of what is needed.
However, for the hardware in this example, the image
will need to boot from sdb
instead
of sda
, which is what the
directdisk-gpt
kickstart file
uses.
The example begins by making a copy of the
directdisk-gpt.wks
file in the
scripts/lib/image/canned-wks
directory and then by changing the lines that specify
the target disk from which to boot.
$ cp /home/stephano/poky/scripts/lib/wic/canned-wks/directdisk-gpt.wks \ /home/stephano/poky/scripts/lib/wic/canned-wks/directdisksdb-gpt.wks
Next, the example modifies the
directdisksdb-gpt.wks
file and
changes all instances of
"--ondisk sda
" to
"--ondisk sdb
".
The example changes the following two lines and leaves
the remaining lines untouched:
part /boot --source bootimg-pcbios --ondisk sdb --label boot --active --align 1024 part / --source rootfs --ondisk sdb --fstype=ext4 --label platform --align 1024 --use-uuid
Once the lines are changed, the example generates the
directdisksdb-gpt
image.
The command points the process at the
core-image-minimal
artifacts for
the Next Unit of Computing (nuc)
MACHINE
the local.conf
.
$ wic create directdisksdb-gpt -e core-image-minimal INFO: Building wic-tools... . . . Initialising tasks: 100% |#######################################| Time: 0:00:01 NOTE: Executing SetScene Tasks NOTE: Executing RunQueue Tasks NOTE: Tasks Summary: Attempted 1161 tasks of which 1157 didn't need to be rerun and all succeeded. INFO: Creating image(s)... INFO: The new image(s) can be found here: ./directdisksdb-gpt-201710090938-sdb.direct The following build artifacts were used to create the image(s): ROOTFS_DIR: /home/stephano/build/master/build/tmp-glibc/work/qemux86-oe-linux/core-image-minimal/1.0-r0/rootfs BOOTIMG_DIR: /home/stephano/build/master/build/tmp-glibc/work/qemux86-oe-linux/core-image-minimal/1.0-r0/recipe-sysroot/usr/share KERNEL_DIR: /home/stephano/build/master/build/tmp-glibc/deploy/images/qemux86 NATIVE_SYSROOT: /home/stephano/build/master/build/tmp-glibc/work/i586-oe-linux/wic-tools/1.0-r0/recipe-sysroot-native INFO: The image(s) were created using OE kickstart file: /home/stephano/poky/scripts/lib/wic/canned-wks/directdisksdb-gpt.wks
Continuing with the example, you can now directly
dd
the image to a USB stick, or
whatever media for which you built your image,
and boot the resulting media:
$ sudo dd if=directdisksdb-gpt-201710090938-sdb.direct of=/dev/sdb 140966+0 records in 140966+0 records out 72174592 bytes (72 MB, 69 MiB) copied, 78.0282 s, 925 kB/s $ sudo eject /dev/sdb
This next example manually specifies each build artifact
(runs in Raw Mode) and uses a modified kickstart file.
The example also uses the -o
option
to cause Wic to create the output
somewhere other than the default output directory,
which is the current directory:
$ wic create /home/stephano/my_yocto/test.wks -o /home/stephano/testwic \ --rootfs-dir /home/stephano/build/master/build/tmp/work/qemux86-poky-linux/core-image-minimal/1.0-r0/rootfs \ --bootimg-dir /home/stephano/build/master/build/tmp/work/qemux86-poky-linux/core-image-minimal/1.0-r0/recipe-sysroot/usr/share \ --kernel-dir /home/stephano/build/master/build/tmp/deploy/images/qemux86 \ --native-sysroot /home/stephano/build/master/build/tmp/work/i586-poky-linux/wic-tools/1.0-r0/recipe-sysroot-native INFO: Creating image(s)... INFO: The new image(s) can be found here: /home/stephano/testwic/test-201710091445-sdb.direct The following build artifacts were used to create the image(s): ROOTFS_DIR: /home/stephano/build/master/build/tmp-glibc/work/qemux86-oe-linux/core-image-minimal/1.0-r0/rootfs BOOTIMG_DIR: /home/stephano/build/master/build/tmp-glibc/work/qemux86-oe-linux/core-image-minimal/1.0-r0/recipe-sysroot/usr/share KERNEL_DIR: /home/stephano/build/master/build/tmp-glibc/deploy/images/qemux86 NATIVE_SYSROOT: /home/stephano/build/master/build/tmp-glibc/work/i586-oe-linux/wic-tools/1.0-r0/recipe-sysroot-native INFO: The image(s) were created using OE kickstart file: /home/stephano/my_yocto/test.wks
For this example,
MACHINE
did not have to be specified in the
local.conf
file since the
artifact is manually specified.
Wic image manipulation allows you to shorten turnaround time during image development. For example, you can use Wic to delete the kernel partition of a Wic image and then insert a newly built kernel. This saves you time from having to rebuild the entire image each time you modify the kernel.
mtools
package installed.
The following example examines the contents of the Wic image, deletes the existing kernel, and then inserts a new kernel:
List the Partitions:
Use the wic ls
command to list
all the partitions in the Wic image:
$ wic ls tmp/deploy/images/qemux86/core-image-minimal-qemux86.wic Num Start End Size Fstype 1 1048576 25041919 23993344 fat16 2 25165824 72157183 46991360 ext4
The previous output shows two partitions in the
core-image-minimal-qemux86.wic
image.
Examine a Particular Partition:
Use the wic ls
command again
but in a different form to examine a particular
partition.
$ wic help command
For example, the following command shows you
the various ways to use the
wic ls
command:
$ wic help ls
The following command shows what is in Partition one:
$ wic ls tmp/deploy/images/qemux86/core-image-minimal-qemux86.wic:1 Volume in drive : is boot Volume Serial Number is E894-1809 Directory for ::/ libcom32 c32 186500 2017-10-09 16:06 libutil c32 24148 2017-10-09 16:06 syslinux cfg 220 2017-10-09 16:06 vesamenu c32 27104 2017-10-09 16:06 vmlinuz 6904608 2017-10-09 16:06 5 files 7 142 580 bytes 16 582 656 bytes free
The previous output shows five files, with the
vmlinuz
being the kernel.
~/.mtoolsrc
file and
be sure to have the line “mtools_skip_check=1“
in the file.
Then, run the Wic command again:
ERROR: _exec_cmd: /usr/bin/mdir -i /tmp/wic-parttfokuwra ::/ returned '1' instead of 0 output: Total number of sectors (47824) not a multiple of sectors per track (32)! Add mtools_skip_check=1 to your .mtoolsrc file to skip this test
Remove the Old Kernel:
Use the wic rm
command to
remove the vmlinuz
file
(kernel):
$ wic rm tmp/deploy/images/qemux86/core-image-minimal-qemux86.wic:1/vmlinuz
Add In the New Kernel:
Use the wic cp
command to
add the updated kernel to the Wic image.
Depending on how you built your kernel, it could
be in different places.
If you used devtool
and
an SDK to build your kernel, it resides in the
tmp/work
directory of the
extensible SDK.
If you used make
to build the
kernel, the kernel will be in the
workspace/sources
area.
The following example assumes
devtool
was used to build
the kernel:
cp ~/poky_sdk/tmp/work/qemux86-poky-linux/linux-yocto/4.12.12+git999-r0/linux-yocto-4.12.12+git999/arch/x86/boot/bzImage \ ~/poky/build/tmp/deploy/images/qemux86/core-image-minimal-qemux86.wic:1/vmlinuz
Once the new kernel is added back into the image,
you can use the dd
command or
bmaptool
to flash your wic image onto an SD card
or USB stick and test your target.
bmaptool
is
generally 10 to 20 times faster than using
dd
.
bmaptool
¶A fast and easy way to flash an image to a bootable device is to use Bmaptool, which is integrated into the OpenEmbedded build system. Bmaptool is a generic tool that creates a file's block map (bmap) and then uses that map to copy the file. As compared to traditional tools such as dd or cp, Bmaptool can copy (or flash) large files like raw system image files much faster.
If you are using Ubuntu or Debian distributions, you
can install the bmap-tools
package
using the following command and then use the tool
without specifying PATH
even from
the root account:
$ sudo apt-get install bmap-tools
If you are unable to install the
bmap-tools
package, you will
need to build Bmaptool before using it.
Use the following command:
$ bitbake bmap-tools-native
Following, is an example that shows how to flash a Wic image. Realize that while this example uses a Wic image, you can use Bmaptool to flash any type of image. Use these steps to flash an image using Bmaptool:
Update your local.conf
File:
You need to have the following set in your
local.conf
file before building
your image:
IMAGE_FSTYPES += "wic wic.bmap"
Get Your Image:
Either have your image ready (pre-built with the
IMAGE_FSTYPES
setting previously mentioned) or take the step to build
the image:
$ bitbake image
Flash the Device:
Flash the device with the image by using Bmaptool
depending on your particular setup.
The following commands assume the image resides in the
Build Directory's deploy/images/
area:
If you have write access to the media, use this command form:
$ oe-run-native bmap-tools-native bmaptool copybuild-directory
/tmp/deploy/images/machine
/image
.wic /dev/sdX
If you do not have write access to the media, set your permissions first and then use the same command form:
$ sudo chmod 666 /dev/sdX
$ oe-run-native bmap-tools-native bmaptool copybuild-directory
/tmp/deploy/images/machine
/image
.wic /dev/sdX
For help on the bmaptool
command, use the
following command:
$ bmaptool --help
Security is of increasing concern for embedded devices. Consider the issues and problems discussed in just this sampling of work found across the Internet:
"Security Risks of Embedded Systems" by Bruce Schneier
"Internet Census 2012" by Carna Botnet
"Security Issues for Embedded Devices" by Jake Edge
When securing your image is of concern, there are steps, tools, and variables that you can consider to help you reach the security goals you need for your particular device. Not all situations are identical when it comes to making an image secure. Consequently, this section provides some guidance and suggestions for consideration when you want to make your image more secure.
General considerations exist that help you create more secure images. You should consider the following suggestions to help make your device more secure:
Scan additional code you are adding to the system (e.g. application code) by using static analysis tools. Look for buffer overflows and other potential security problems.
Pay particular attention to the security for any web-based administration interface.
Web interfaces typically need to perform administrative functions and tend to need to run with elevated privileges. Thus, the consequences resulting from the interface's security becoming compromised can be serious. Look for common web vulnerabilities such as cross-site-scripting (XSS), unvalidated inputs, and so forth.
As with system passwords, the default credentials for accessing a web-based interface should not be the same across all devices. This is particularly true if the interface is enabled by default as it can be assumed that many end-users will not change the credentials.
Ensure you can update the software on the device to mitigate vulnerabilities discovered in the future. This consideration especially applies when your device is network-enabled.
Ensure you remove or disable debugging functionality before producing the final image. For information on how to do this, see the "Considerations Specific to the OpenEmbedded Build System" section.
Ensure you have no network services listening that are not needed.
Remove any software from the image that is not needed.
Enable hardware support for secure boot functionality when your device supports this functionality.
The Yocto Project has security flags that you can enable that
help make your build output more secure.
The security flags are in the
meta/conf/distro/include/security_flags.inc
file in your
Source Directory
(e.g. poky
).
Use the following line in your
local.conf
file or in your custom
distribution configuration file to enable the security
compiler and linker flags for your build:
require conf/distro/include/security_flags.inc
You can take some steps that are specific to the OpenEmbedded build system to make your images more secure:
Ensure "debug-tweaks" is not one of your selected
IMAGE_FEATURES
.
When creating a new project, the default is to provide you
with an initial local.conf
file that
enables this feature using the
EXTRA_IMAGE_FEATURES
variable with the line:
EXTRA_IMAGE_FEATURES = "debug-tweaks"
To disable that feature, simply comment out that line in your
local.conf
file, or
make sure IMAGE_FEATURES
does not contain
"debug-tweaks" before producing your final image.
Among other things, leaving this in place sets the
root password as blank, which makes logging in for
debugging or inspection easy during
development but also means anyone can easily log in
during production.
It is possible to set a root password for the image and also to set passwords for any extra users you might add (e.g. administrative or service type users). When you set up passwords for multiple images or users, you should not duplicate passwords.
To set up passwords, use the
extrausers
class, which is the preferred method.
For an example on how to set up both root and user
passwords, see the
"extrausers.bbclass
"
section.
Consider enabling a Mandatory Access Control (MAC)
framework such as SMACK or SELinux and tuning it
appropriately for your device's usage.
You can find more information in the
meta-selinux
layer.
The Yocto Project provides tools for making your image
more secure.
You can find these tools in the
meta-security
layer of the
Yocto Project Source Repositories.
When you build an image using the Yocto Project and do not alter any distribution Metadata, you are creating a Poky distribution. If you wish to gain more control over package alternative selections, compile-time options, and other low-level configurations, you can create your own distribution.
To create your own distribution, the basic steps consist of creating your own distribution layer, creating your own distribution configuration file, and then adding any needed code and Metadata to the layer. The following steps provide some more detail:
Create a layer for your new distro:
Create your distribution layer so that you can keep your
Metadata and code for the distribution separate.
It is strongly recommended that you create and use your own
layer for configuration and code.
Using your own layer as compared to just placing
configurations in a local.conf
configuration file makes it easier to reproduce the same
build configuration when using multiple build machines.
See the
"Creating a General Layer Using the bitbake-layers
Script"
section for information on how to quickly set up a layer.
Create the distribution configuration file:
The distribution configuration file needs to be created in
the conf/distro
directory of your
layer.
You need to name it using your distribution name
(e.g. mydistro.conf
).
You can split out parts of your configuration file
into include files and then "require" them from within
your distribution configuration file.
Be sure to place the include files in the
conf/distro/include
directory of
your layer.
A common example usage of include files would be to
separate out the selection of desired version and revisions
for individual recipes.
Your configuration file needs to set the following required variables:
DISTRO_NAME
DISTRO_VERSION
These following variables are optional and you typically set them from the distribution configuration file:
DISTRO_FEATURES
DISTRO_EXTRA_RDEPENDS
DISTRO_EXTRA_RRECOMMENDS
TCLIBC
conf/distro/defaultsetup.conf
as
a reference and just include variables that differ
as compared to defaultsetup.conf
.
Alternatively, you can create a distribution
configuration file from scratch using the
defaultsetup.conf
file
or configuration files from other distributions
such as Poky or Angstrom as references.
Provide miscellaneous variables:
Be sure to define any other variables for which you want to
create a default or enforce as part of the distribution
configuration.
You can include nearly any variable from the
local.conf
file.
The variables you use are not limited to the list in the
previous bulleted item.
Point to Your distribution configuration file:
In your local.conf
file in the
Build Directory,
set your
DISTRO
variable to point to your distribution's configuration file.
For example, if your distribution's configuration file is
named mydistro.conf
, then you point
to it as follows:
DISTRO = "mydistro"
Add more to the layer if necessary: Use your layer to hold other information needed for the distribution:
Add recipes for installing
distro-specific configuration files that are not
already installed by another recipe.
If you have distro-specific configuration files
that are included by an existing recipe, you should
add an append file (.bbappend
)
for those.
For general information and recommendations
on how to add recipes to your layer, see the
"Creating Your Own Layer"
and
"Following Best Practices When Creating Layers"
sections.
Add any image recipes that are specific to your distribution.
Add a psplash
append file for a branded splash screen.
For information on append files, see the
"Using .bbappend Files in Your Layer"
section.
Add any other append files to make custom changes that are specific to individual recipes.
If you are producing your own customized version
of the build system for use by other users, you might
want to customize the message shown by the setup script or
you might want to change the template configuration files (i.e.
local.conf
and
bblayers.conf
) that are created in
a new build directory.
The OpenEmbedded build system uses the environment variable
TEMPLATECONF
to locate the directory
from which it gathers configuration information that ultimately
ends up in the
Build Directory
conf
directory.
By default, TEMPLATECONF
is set as
follows in the poky
repository:
TEMPLATECONF=${TEMPLATECONF:-meta-poky/conf}
This is the directory used by the build system to find templates
from which to build some key configuration files.
If you look at this directory, you will see the
bblayers.conf.sample
,
local.conf.sample
, and
conf-notes.txt
files.
The build system uses these files to form the respective
bblayers.conf
file,
local.conf
file, and display the list of
BitBake targets when running the setup script.
To override these default configuration files with
configurations you want used within every new
Build Directory, simply set the
TEMPLATECONF
variable to your directory.
The TEMPLATECONF
variable is set in the
.templateconf
file, which is in the
top-level
Source Directory
folder (e.g. poky
).
Edit the .templateconf
so that it can locate
your directory.
Best practices dictate that you should keep your
template configuration directory in your custom distribution layer.
For example, suppose you have a layer named
meta-mylayer
located in your home directory
and you want your template configuration directory named
myconf
.
Changing the .templateconf
as follows
causes the OpenEmbedded build system to look in your directory
and base its configuration files on the
*.sample
configuration files it finds.
The final configuration files (i.e.
local.conf
and
bblayers.conf
ultimately still end up in
your Build Directory, but they are based on your
*.sample
files.
TEMPLATECONF=${TEMPLATECONF:-meta-mylayer/myconf}
Aside from the *.sample
configuration files,
the conf-notes.txt
also resides in the
default meta-poky/conf
directory.
The script that sets up the build environment
(i.e.
oe-init-build-env
)
uses this file to display BitBake targets as part of the script
output.
Customizing this conf-notes.txt
file is a
good way to make sure your list of custom targets appears
as part of the script's output.
Here is the default list of targets displayed as a result of running either of the setup scripts:
You can now run 'bitbake <target>' Common targets are: core-image-minimal core-image-sato meta-toolchain meta-ide-support
Changing the listed common targets is as easy as editing your
version of conf-notes.txt
in your
custom template configuration directory and making sure you
have TEMPLATECONF
set to your directory.
To help conserve disk space during builds, you can add the
following statement to your project's
local.conf
configuration file found in the
Build Directory:
INHERIT += "rm_work"
Adding this statement deletes the work directory used for building
a recipe once the recipe is built.
For more information on "rm_work", see the
rm_work
class in the Yocto Project Reference Manual.
This section describes a few tasks that involve packages:
You might find it necessary to prevent specific packages from being installed into an image. If so, you can use several variables to direct the build system to essentially ignore installing recommended packages or to not install a package at all.
The following list introduces variables you can use to
prevent packages from being installed into your image.
Each of these variables only works with IPK and RPM
package types.
Support for Debian packages does not exist.
Also, you can use these variables from your
local.conf
file or attach them to a
specific image recipe by using a recipe name override.
For more detail on the variables, see the descriptions in the
Yocto Project Reference Manual's glossary chapter.
BAD_RECOMMENDATIONS
:
Use this variable to specify "recommended-only"
packages that you do not want installed.
NO_RECOMMENDATIONS
:
Use this variable to prevent all "recommended-only"
packages from being installed.
PACKAGE_EXCLUDE
:
Use this variable to prevent specific packages from
being installed regardless of whether they are
"recommended-only" or not.
You need to realize that the build process could
fail with an error when you
prevent the installation of a package whose presence
is required by an installed package.
This section provides some background on how binary package versioning is accomplished and presents some of the services, variables, and terminology involved.
In order to understand binary package versioning, you need to consider the following:
Binary Package: The binary package that is eventually built and installed into an image.
Binary Package Version: The binary package version is composed of two components - a version and a revision.
PE
)
is involved but this discussion for the most part
ignores PE
.
The version and revision are taken from the
PV
and
PR
variables, respectively.
PV
: The recipe version.
PV
represents the version of the
software being packaged.
Do not confuse PV
with the
binary package version.
PR
: The recipe revision.
SRCPV
:
The OpenEmbedded build system uses this string
to help define the value of PV
when the source code revision needs to be included
in it.
PR Service: A network-based service that helps automate keeping package feeds compatible with existing package manager applications such as RPM, APT, and OPKG.
Whenever the binary package content changes, the binary package
version must change.
Changing the binary package version is accomplished by changing
or "bumping" the PR
and/or
PV
values.
Increasing these values occurs one of two ways:
Automatically using a Package Revision Service (PR Service).
Manually incrementing the
PR
and/or
PV
variables.
Given a primary challenge of any build system and its users is how to maintain a package feed that is compatible with existing package manager applications such as RPM, APT, and OPKG, using an automated system is much preferred over a manual system. In either system, the main requirement is that binary package version numbering increases in a linear fashion and that a number of version components exist that support that linear progression. For information on how to ensure package revisioning remains linear, see the "Automatically Incrementing a Binary Package Revision Number" section.
The following three sections provide related information on the
PR Service, the manual method for "bumping"
PR
and/or PV
, and
on how to ensure binary package revisioning remains linear.
As mentioned, attempting to maintain revision numbers in the Metadata is error prone, inaccurate, and causes problems for people submitting recipes. Conversely, the PR Service automatically generates increasing numbers, particularly the revision field, which removes the human element.
The Yocto Project uses variables in order of
decreasing priority to facilitate revision numbering (i.e.
PE
,
PV
, and
PR
for epoch, version, and revision, respectively).
The values are highly dependent on the policies and
procedures of a given distribution and package feed.
Because the OpenEmbedded build system uses
"signatures",
which are unique to a given build, the build system
knows when to rebuild packages.
All the inputs into a given task are represented by a
signature, which can trigger a rebuild when different.
Thus, the build system itself does not rely on the
PR
, PV
, and
PE
numbers to trigger a rebuild.
The signatures, however, can be used to generate
these values.
The PR Service works with both
OEBasic
and
OEBasicHash
generators.
The value of PR
bumps when the
checksum changes and the different generator mechanisms
change signatures under different circumstances.
As implemented, the build system includes values from
the PR Service into the PR
field as
an addition using the form ".x
" so
r0
becomes r0.1
,
r0.2
and so forth.
This scheme allows existing PR
values
to be used for whatever reasons, which include manual
PR
bumps, should it be necessary.
By default, the PR Service is not enabled or running. Thus, the packages generated are just "self consistent". The build system adds and removes packages and there are no guarantees about upgrade paths but images will be consistent and correct with the latest changes.
The simplest form for a PR Service is for it to exist
for a single host development system that builds the
package feed (building system).
For this scenario, you can enable a local PR Service by
setting
PRSERV_HOST
in your local.conf
file in the
Build Directory:
PRSERV_HOST = "localhost:0"
Once the service is started, packages will automatically
get increasing PR
values and
BitBake takes care of starting and stopping the server.
If you have a more complex setup where multiple host
development systems work against a common, shared package
feed, you have a single PR Service running and it is
connected to each building system.
For this scenario, you need to start the PR Service using
the bitbake-prserv
command:
bitbake-prserv --hostip
--portport
--start
In addition to hand-starting the service, you need to
update the local.conf
file of each
building system as described earlier so each system
points to the server and port.
It is also recommended you use build history, which adds
some sanity checks to binary package versions, in
conjunction with the server that is running the PR Service.
To enable build history, add the following to each building
system's local.conf
file:
# It is recommended to activate "buildhistory" for testing the PR service INHERIT += "buildhistory" BUILDHISTORY_COMMIT = "1"
For information on build history, see the "Maintaining Build Output Quality" section.
The OpenEmbedded build system does not maintain
PR
information as part of the
shared state (sstate) packages.
If you maintain an sstate feed, its expected that either
all your building systems that contribute to the sstate
feed use a shared PR Service, or you do not run a PR
Service on any of your building systems.
Having some systems use a PR Service while others do
not leads to obvious problems.
For more information on shared state, see the "Shared State Cache" section in the Yocto Project Overview and Concepts Manual.
The alternative to setting up a PR Service is to manually
"bump" the
PR
variable.
If a committed change results in changing the package
output, then the value of the PR variable needs to be
increased (or "bumped") as part of that commit.
For new recipes you should add the PR
variable and set its initial value equal to "r0", which is
the default.
Even though the default value is "r0", the practice of
adding it to a new recipe makes it harder to forget to bump
the variable when you make changes to the recipe in future.
If you are sharing a common .inc
file
with multiple recipes, you can also use the
INC_PR
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 "${INC_PR}.0" initially, incrementing the last number
when the recipe is changed.
If the .inc
file is changed then its
INC_PR
should be incremented.
When upgrading the version of a binary package, assuming the
PV
changes, the PR
variable should be
reset to "r0" (or "${INC_PR}.0" if you are using
INC_PR
).
Usually, version increases occur only to binary packages.
However, if for some reason PV
changes
but does not increase, you can increase the
PE
variable (Package Epoch).
The PE
variable defaults to "0".
Binary package version numbering strives to follow the Debian Version Field Policy Guidelines. These guidelines define how versions are compared and what "increasing" a version means.
When fetching a repository, BitBake uses the
SRCREV
variable to determine the specific source code revision
from which to build.
You set the SRCREV
variable to
AUTOREV
to cause the OpenEmbedded build system to automatically use the
latest revision of the software:
SRCREV = "${AUTOREV}"
Furthermore, you need to reference SRCPV
in PV
in order to automatically update
the version whenever the revision of the source code
changes.
Here is an example:
PV = "1.0+git${SRCPV}"
The OpenEmbedded build system substitutes
SRCPV
with the following:
AUTOINC+source_code_revision
The build system replaces the AUTOINC
with
a number.
The number used depends on the state of the PR Service:
If PR Service is enabled, the build system increments
the number, which is similar to the behavior of
PR
.
This behavior results in linearly increasing package
versions, which is desirable.
Here is an example:
hello-world-git_0.0+git0+b6558dd387-r0.0_armv7a-neon.ipk hello-world-git_0.0+git1+dd2f5c3565-r0.0_armv7a-neon.ipk
If PR Service is not enabled, the build system
replaces the AUTOINC
placeholder with zero (i.e. "0").
This results in changing the package version since
the source revision is included.
However, package versions are not increased linearly.
Here is an example:
hello-world-git_0.0+git0+b6558dd387-r0.0_armv7a-neon.ipk hello-world-git_0.0+git0+dd2f5c3565-r0.0_armv7a-neon.ipk
In summary, the OpenEmbedded build system does not track the
history of binary package versions for this purpose.
AUTOINC
, in this case, is comparable to
PR
.
If PR server is not enabled, AUTOINC
in the package version is simply replaced by "0".
If PR server is enabled, the build system keeps track of the
package versions and bumps the number when the package
revision changes.
Many pieces of software split functionality into optional modules (or plugins) and the plugins that are built might depend on configuration options. To avoid having to duplicate the logic that determines what modules are available in your recipe or to avoid having to package each module by hand, the OpenEmbedded build system provides functionality to handle module packaging dynamically.
To handle optional module packaging, you need to do two things:
Ensure the module packaging is actually done.
Ensure that any dependencies on optional modules from other recipes are satisfied by your recipe.
To ensure the module packaging actually gets done, you use
the do_split_packages
function within
the populate_packages
Python function
in your recipe.
The do_split_packages
function
searches for a pattern of files or directories under a
specified path and creates a package for each one it finds
by appending to the
PACKAGES
variable and setting the appropriate values for
FILES_packagename
,
RDEPENDS_packagename
,
DESCRIPTION_packagename
, and so forth.
Here is an example from the lighttpd
recipe:
python populate_packages_prepend () { lighttpd_libdir = d.expand('${libdir}') do_split_packages(d, lighttpd_libdir, '^mod_(.*)\.so$', 'lighttpd-module-%s', 'Lighttpd module for %s', extra_depends='') }
The previous example specifies a number of things in the
call to do_split_packages
.
A directory within the files installed
by your recipe through do_install
in which to search.
A regular expression used to match module files in that directory. In the example, note the parentheses () that mark the part of the expression from which the module name should be derived.
A pattern to use for the package names.
A description for each package.
An empty string for
extra_depends
, which disables
the default dependency on the main
lighttpd
package.
Thus, if a file in ${libdir}
called mod_alias.so
is found,
a package called lighttpd-module-alias
is created for it and the
DESCRIPTION
is set to "Lighttpd module for alias".
Often, packaging modules is as simple as the previous
example.
However, more advanced options exist that you can use
within do_split_packages
to modify its
behavior.
And, if you need to, you can add more logic by specifying
a hook function that is called for each package.
It is also perfectly acceptable to call
do_split_packages
multiple times if
you have more than one set of modules to package.
For more examples that show how to use
do_split_packages
, see the
connman.inc
file in the
meta/recipes-connectivity/connman/
directory of the poky
source repository.
You can also find examples in
meta/classes/kernel.bbclass
.
Following is a reference that shows
do_split_packages
mandatory and
optional arguments:
Mandatory arguments root The path in which to search file_regex Regular expression to match searched files. Use parentheses () to mark the part of this expression that should be used to derive the module name (to be substituted where %s is used in other function arguments as noted below) output_pattern Pattern to use for the package names. Must include %s. description Description to set for each package. Must include %s. Optional arguments postinst Postinstall script to use for all packages (as a string) recursive True to perform a recursive search - default False hook A hook function to be called for every match. The function will be called with the following arguments (in the order listed): f Full path to the file/directory match pkg The package name file_regex As above output_pattern As above modulename The module name derived using file_regex extra_depends Extra runtime dependencies (RDEPENDS) to be set for all packages. The default value of None causes a dependency on the main package (${PN}) - if you do not want this, pass empty string '' for this parameter. aux_files_pattern Extra item(s) to be added to FILES for each package. Can be a single string item or a list of strings for multiple items. Must include %s. postrm postrm script to use for all packages (as a string) allow_dirs True to allow directories to be matched - default False prepend If True, prepend created packages to PACKAGES instead of the default False which appends them match_path match file_regex on the whole relative path to the root rather than just the file name aux_files_pattern_verbatim Extra item(s) to be added to FILES for each package, using the actual derived module name rather than converting it to something legal for a package name. Can be a single string item or a list of strings for multiple items. Must include %s. allow_links True to allow symlinks to be matched - default False summary Summary to set for each package. Must include %s; defaults to description if not set.
The second part for handling optional module packaging
is to ensure that any dependencies on optional modules
from other recipes are satisfied by your recipe.
You can be sure these dependencies are satisfied by
using the
PACKAGES_DYNAMIC
variable.
Here is an example that continues with the
lighttpd
recipe shown earlier:
PACKAGES_DYNAMIC = "lighttpd-module-.*"
The name specified in the regular expression can of
course be anything.
In this example, it is lighttpd-module-
and is specified as the prefix to ensure that any
RDEPENDS
and RRECOMMENDS
on a package name starting with the prefix are satisfied
during build time.
If you are using do_split_packages
as described in the previous section, the value you put in
PACKAGES_DYNAMIC
should correspond to
the name pattern specified in the call to
do_split_packages
.
During a build, BitBake always transforms a recipe into one or
more packages.
For example, BitBake takes the bash
recipe
and produces a number of packages (e.g.
bash
, bash-bashbug
,
bash-completion
,
bash-completion-dbg
,
bash-completion-dev
,
bash-completion-extra
,
bash-dbg
, and so forth).
Not all generated packages are included in an image.
In several situations, you might need to update, add, remove, or query the packages on a target device at runtime (i.e. without having to generate a new image). Examples of such situations include:
You want to provide in-the-field updates to deployed devices (e.g. security updates).
You want to have a fast turn-around development cycle for one or more applications that run on your device.
You want to temporarily install the "debug" packages of various applications on your device so that debugging can be greatly improved by allowing access to symbols and source debugging.
You want to deploy a more minimal package selection of your device but allow in-the-field updates to add a larger selection for customization.
In all these situations, you have something similar to a more traditional Linux distribution in that in-field devices are able to receive pre-compiled packages from a server for installation or update. Being able to install these packages on a running, in-field device is what is termed "runtime package management".
In order to use runtime package management, you need a host or server machine that serves up the pre-compiled packages plus the required metadata. You also need package manipulation tools on the target. The build machine is a likely candidate to act as the server. However, that machine does not necessarily have to be the package server. The build machine could push its artifacts to another machine that acts as the server (e.g. Internet-facing). In fact, doing so is advantageous for a production environment as getting the packages away from the development system's build directory prevents accidental overwrites.
A simple build that targets just one device produces
more than one package database.
In other words, the packages produced by a build are separated
out into a couple of different package groupings based on
criteria such as the target's CPU architecture, the target
board, or the C library used on the target.
For example, a build targeting the qemux86
device produces the following three package databases:
noarch
, i586
, and
qemux86
.
If you wanted your qemux86
device to be
aware of all the packages that were available to it,
you would need to point it to each of these databases
individually.
In a similar way, a traditional Linux distribution usually is
configured to be aware of a number of software repositories
from which it retrieves packages.
Using runtime package management is completely optional and not required for a successful build or deployment in any way. But if you want to make use of runtime package management, you need to do a couple things above and beyond the basics. The remainder of this section describes what you need to do.
This section describes build considerations of which you need to be aware in order to provide support for runtime package management.
When BitBake generates packages, it needs to know
what format or formats to use.
In your configuration, you use the
PACKAGE_CLASSES
variable to specify the format:
Open the local.conf
file
inside your
Build Directory
(e.g. ~/poky/build/conf/local.conf
).
Select the desired package format as follows:
PACKAGE_CLASSES ?= “package_packageformat
”
where packageformat
can be "ipk", "rpm", "deb", or "tar" which are the
supported package formats.
If you would like your image to start off with a basic
package database containing the packages in your current
build as well as to have the relevant tools available on the
target for runtime package management, you can include
"package-management" in the
IMAGE_FEATURES
variable.
Including "package-management" in this configuration
variable ensures that when the image is assembled for your
target, the image includes the currently-known package
databases as well as the target-specific tools required
for runtime package management to be performed on the
target.
However, this is not strictly necessary.
You could start your image off without any databases
but only include the required on-target package
tool(s).
As an example, you could include "opkg" in your
IMAGE_INSTALL
variable if you are using the IPK package format.
You can then initialize your target's package database(s)
later once your image is up and running.
Whenever you perform any sort of build step that can potentially generate a package or modify existing package, it is always a good idea to re-generate the package index after the build by using the following command:
$ bitbake package-index
It might be tempting to build the package and the package index at the same time with a command such as the following:
$ bitbake some-package
package-index
Do not do this as BitBake does not schedule the package index for after the completion of the package you are building. Consequently, you cannot be sure of the package index including information for the package you just built. Thus, be sure to run the package update step separately after building any packages.
You can use the
PACKAGE_FEED_ARCHS
,
PACKAGE_FEED_BASE_PATHS
,
and
PACKAGE_FEED_URIS
variables to pre-configure target images to use a package
feed.
If you do not define these variables, then manual steps
as described in the subsequent sections are necessary to
configure the target.
You should set these variables before building the image
in order to produce a correctly configured image.
When your build is complete, your packages reside in the
${TMPDIR}/deploy/
directory.
For example, if
packageformat
${
TMPDIR
}
is tmp
and your selected package type
is RPM, then your RPM packages are available in
tmp/deploy/rpm
.
Although other protocols are possible, a server using HTTP typically serves packages. If you want to use HTTP, then set up and configure a web server such as Apache 2, lighttpd, or SimpleHTTPServer on the machine serving the packages.
To keep things simple, this section describes how to set up a SimpleHTTPServer web server to share package feeds from the developer's machine. Although this server might not be the best for a production environment, the setup is simple and straight forward. Should you want to use a different server more suited for production (e.g. Apache 2, Lighttpd, or Nginx), take the appropriate steps to do so.
From within the build directory where you have built an
image based on your packaging choice (i.e. the
PACKAGE_CLASSES
setting), simply start the server.
The following example assumes a build directory of
~/poky/build/tmp/deploy/rpm
and a
PACKAGE_CLASSES
setting of
"package_rpm":
$ cd ~/poky/build/tmp/deploy/rpm $ python -m SimpleHTTPServer
Setting up the target differs depending on the package management system. This section provides information for RPM, IPK, and DEB.
The
Dandified Packaging Tool
(DNF) performs runtime package management of RPM
packages.
In order to use DNF for runtime package management,
you must perform an initial setup on the target
machine for cases where the
PACKAGE_FEED_*
variables were not
set as part of the image that is running on the
target.
This means if you built your image and did not not use
these variables as part of the build and your image is
now running on the target, you need to perform the
steps in this section if you want to use runtime
package management.
PACKAGE_FEED_*
variables, see
PACKAGE_FEED_ARCHS
,
PACKAGE_FEED_BASE_PATHS
,
and
PACKAGE_FEED_URIS
in the Yocto Project Reference Manual variables
glossary.
On the target, you must inform DNF that package
databases are available.
You do this by creating a file named
/etc/yum.repos.d/oe-packages.repo
and defining the oe-packages
.
As an example, assume the target is able to use the
following package databases:
all
, i586
,
and qemux86
from a server named
my.server
.
The specifics for setting up the web server are up to
you.
The critical requirement is that the URIs in the
target repository configuration point to the
correct remote location for the feeds.
deploy
directory.
However, for production use, it is better to copy
the package directories to a location outside of
the build area and use that location.
Doing so avoids situations where the build system
overwrites or changes the
deploy
directory.
When telling DNF where to look for the package databases, you must declare individual locations per architecture or a single location used for all architectures. You cannot do both:
Create an Explicit List of Architectures: Define individual base URLs to identify where each package database is located:
[oe-packages] baseurl=http://my.server/rpm/i586 http://my.server/rpm/qemux86 http://my.server/rpm/all
This example informs DNF about individual package databases for all three architectures.
Create a Single (Full) Package Index: Define a single base URL that identifies where a full package database is located:
[oe-packages] baseurl=http://my.server/rpm
This example informs DNF about a single package database that contains all the package index information for all supported architectures.
Once you have informed DNF where to find the package databases, you need to fetch them:
# dnf makecache
DNF is now able to find, install, and upgrade packages from the specified repository or repositories.
The opkg
application performs
runtime package management of IPK packages.
You must perform an initial setup for
opkg
on the target machine
if the
PACKAGE_FEED_ARCHS
,
PACKAGE_FEED_BASE_PATHS
, and
PACKAGE_FEED_URIS
variables have not been set or the target image was
built before the variables were set.
The opkg
application uses
configuration files to find available package
databases.
Thus, you need to create a configuration file inside
the /etc/opkg/
direction, which
informs opkg
of any repository
you want to use.
As an example, suppose you are serving packages from a
ipk/
directory containing the
i586
,
all
, and
qemux86
databases through an
HTTP server named my.server
.
On the target, create a configuration file
(e.g. my_repo.conf
) inside the
/etc/opkg/
directory containing
the following:
src/gz all http://my.server/ipk/all src/gz i586 http://my.server/ipk/i586 src/gz qemux86 http://my.server/ipk/qemux86
Next, instruct opkg
to fetch
the repository information:
# opkg update
The opkg
application is now able
to find, install, and upgrade packages from the
specified repository.
The apt
application performs
runtime package management of DEB packages.
This application uses a source list file to find
available package databases.
You must perform an initial setup for
apt
on the target machine
if the
PACKAGE_FEED_ARCHS
,
PACKAGE_FEED_BASE_PATHS
, and
PACKAGE_FEED_URIS
variables have not been set or the target image was
built before the variables were set.
To inform apt
of the repository
you want to use, you might create a list file (e.g.
my_repo.list
) inside the
/etc/apt/sources.list.d/
directory.
As an example, suppose you are serving packages from a
deb/
directory containing the
i586
,
all
, and
qemux86
databases through an
HTTP server named my.server
.
The list file should contain:
deb http://my.server/deb/all ./ deb http://my.server/deb/i586 ./ deb http://my.server/deb/qemux86 ./
Next, instruct the apt
application to fetch the repository information:
# apt-get update
After this step, apt
is able
to find, install, and upgrade packages from the
specified repository.
In order to add security to RPM packages used during a build, you can take steps to securely sign them. Once a signature is verified, the OpenEmbedded build system can use the package in the build. If security fails for a signed package, the build system aborts the build.
This section describes how to sign RPM packages during a build and how to use signed package feeds (repositories) when doing a build.
To enable signing RPM packages, you must set up the
following configurations in either your
local.config
or
distro.config
file:
# Inherit sign_rpm.bbclass to enable signing functionality INHERIT += " sign_rpm" # Define the GPG key that will be used for signing. RPM_GPG_NAME = "key_name
" # Provide passphrase for the key RPM_GPG_PASSPHRASE = "passphrase
"
key_name
and
passphrase
Aside from the
RPM_GPG_NAME
and
RPM_GPG_PASSPHRASE
variables in the
previous example, two optional variables related to signing
exist:
GPG_BIN
:
Specifies a gpg
binary/wrapper
that is executed when the package is signed.
GPG_PATH
:
Specifies the gpg
home
directory used when the package is signed.
In addition to being able to sign RPM packages, you can also enable signed package feeds for IPK and RPM packages.
The steps you need to take to enable signed package feed
use are similar to the steps used to sign RPM packages.
You must define the following in your
local.config
or
distro.config
file:
INHERIT += "sign_package_feed" PACKAGE_FEED_GPG_NAME = "key_name
" PACKAGE_FEED_GPG_PASSPHRASE_FILE = "path_to_file_containing_passphrase
"
For signed package feeds, the passphrase must exist in a
separate file, which is pointed to by the
PACKAGE_FEED_GPG_PASSPHRASE_FILE
variable.
Regarding security, keeping a plain text passphrase out of
the configuration is more secure.
Aside from the
PACKAGE_FEED_GPG_NAME
and
PACKAGE_FEED_GPG_PASSPHRASE_FILE
variables, three optional variables related to signed
package feeds exist:
GPG_BIN
:
Specifies a gpg
binary/wrapper
that is executed when the package is signed.
GPG_PATH
:
Specifies the gpg
home
directory used when the package is signed.
PACKAGE_FEED_GPG_SIGNATURE_TYPE
:
Specifies the type of gpg
signature.
This variable applies only to RPM and IPK package
feeds.
Allowable values for the
PACKAGE_FEED_GPG_SIGNATURE_TYPE
are "ASC", which is the default and specifies ascii
armored, and "BIN", which specifies binary.
A Package Test (ptest) runs tests against packages built
by the OpenEmbedded build system on the target machine.
A ptest contains at least two items: the actual test, and
a shell script (run-ptest
) that starts
the test.
The shell script that starts the test must not contain
the actual test - the script only starts the test.
On the other hand, the test can be anything from a simple
shell script that runs a binary and checks the output to
an elaborate system of test binaries and data files.
The test generates output in the format used by Automake:
result
:testname
where the result can be PASS
,
FAIL
, or SKIP
,
and the testname can be any identifying string.
For a list of Yocto Project recipes that are already enabled with ptest, see the Ptest wiki page.
ptest
class.
To add package testing to your build, add the
DISTRO_FEATURES
and EXTRA_IMAGE_FEATURES
variables to your local.conf
file,
which is found in the
Build Directory:
DISTRO_FEATURES_append = " ptest" EXTRA_IMAGE_FEATURES += "ptest-pkgs"
Once your build is complete, the ptest files are installed
into the
/usr/lib/
directory within the image, where
package
/ptest
is the name of the package.
package
The ptest-runner
package installs a
shell script that loops through all installed ptest test
suites and runs them in sequence.
Consequently, you might want to add this package to
your image.
In order to enable a recipe to run installed ptests on target hardware, you need to prepare the recipes that build the packages you want to test. Here is what you have to do for each recipe:
Be sure the recipe
inherits the
ptest
class:
Include the following line in each recipe:
inherit ptest
Create run-ptest
:
This script starts your test.
Locate the script where you will refer to it
using
SRC_URI
.
Here is an example that starts a test for
dbus
:
#!/bin/sh cd test make -k runtest-TESTS
Ensure dependencies are
met:
If the test adds build or runtime dependencies
that normally do not exist for the package
(such as requiring "make" to run the test suite),
use the
DEPENDS
and
RDEPENDS
variables in your recipe in order for the package
to meet the dependencies.
Here is an example where the package has a runtime
dependency on "make":
RDEPENDS_${PN}-ptest += "make"
Add a function to build the test suite: Not many packages support cross-compilation of their test suites. Consequently, you usually need to add a cross-compilation function to the package.
Many packages based on Automake compile and
run the test suite by using a single command
such as make check
.
However, the host make check
builds and runs on the same computer, while
cross-compiling requires that the package is built
on the host but executed for the target
architecture (though often, as in the case for
ptest, the execution occurs on the host).
The built version of Automake that ships with the
Yocto Project includes a patch that separates
building and execution.
Consequently, packages that use the unaltered,
patched version of make check
automatically cross-compiles.
Regardless, you still must add a
do_compile_ptest
function to
build the test suite.
Add a function similar to the following to your
recipe:
do_compile_ptest() { oe_runmake buildtest-TESTS }
Ensure special configurations
are set:
If the package requires special configurations
prior to compiling the test code, you must
insert a do_configure_ptest
function into the recipe.
Install the test
suite:
The ptest
class
automatically copies the file
run-ptest
to the target and
then runs make install-ptest
to run the tests.
If this is not enough, you need to create a
do_install_ptest
function and
make sure it gets called after the
"make install-ptest" completes.
NPM
is a package manager for the JavaScript programming
language.
The Yocto Project supports the NPM
fetcher.
You can use this fetcher in combination with
devtool
to create recipes that produce NPM packages.
Two workflows exist that allow you to create NPM packages
using devtool
: the NPM registry modules
method and the NPM project code method.
devtool
is far simpler.
Additionally, some requirements and caveats exist.
You need to be aware of the following before using
devtool
to create NPM packages:
Of the two methods that you can use
devtool
to create NPM
packages, the registry approach is slightly
simpler.
However, you might consider the project
approach because you do not have to publish
your module in the NPM registry
(npm-registry
),
which is NPM's public registry.
Be familiar with
devtool
.
The NPM host tools need the native
nodejs-npm
package, which
is part of the OpenEmbedded environment.
You need to get the package by cloning the
https://github.com/openembedded/meta-openembedded
repository out of GitHub.
Be sure to add the path to your local copy to
your bblayers.conf
file.
devtool
cannot detect
native libraries in module dependencies.
Consequently, you must manually add packages
to your recipe.
While deploying NPM packages,
devtool
cannot determine
which dependent packages are missing on the
target (e.g. the node runtime
nodejs
).
Consequently, you need to find out what
files are missing and be sure they are on the
target.
Although you might not need NPM to run your
node package, it is useful to have NPM on your
target.
The NPM package name is
nodejs-npm
.
This section presents an example that uses the
cute-files
module, which is a
file browser web application.
cute-files
module version.
The first thing you need to do is use
devtool
and the NPM fetcher to
create the recipe:
$ devtool add "npm://registry.npmjs.org;package=cute-files;version=1.0.2"
The devtool add
command runs
recipetool create
and uses the
same fetch URI to download each dependency and capture
license details where possible.
The result is a generated recipe.
The recipe file is fairly simple and contains every
license that recipetool
finds
and includes the licenses in the recipe's
LIC_FILES_CHKSUM
variables.
You need to examine the variables and look for those
with "unknown" in the
LICENSE
field.
You need to track down the license information for
"unknown" modules and manually add the information to the
recipe.
recipetool
creates a "shrinkwrap" file
for your recipe.
Shrinkwrap files capture the version of all dependent
modules.
Many packages do not provide shrinkwrap files.
recipetool
create a shrinkwrap
file as it runs.
The devtool edit-recipe
command
lets you take a look at the recipe:
$ devtool edit-recipe cute-files SUMMARY = "Turn any folder on your computer into a cute file browser, available on the local network." LICENSE = "MIT & ISC & Unknown" LIC_FILES_CHKSUM = "file://LICENSE;md5=71d98c0a1db42956787b1909c74a86ca \ file://node_modules/toidentifier/LICENSE;md5=1a261071a044d02eb6f2bb47f51a3502 \ file://node_modules/debug/LICENSE;md5=ddd815a475e7338b0be7a14d8ee35a99 \ ... SRC_URI = " \ npm://registry.npmjs.org/;package=cute-files;version=${PV} \ npmsw://${THISDIR}/${BPN}/npm-shrinkwrap.json \ " S = "${WORKDIR}/npm" inherit npm LICENSE_${PN} = "MIT" LICENSE_${PN}-accepts = "MIT" LICENSE_${PN}-array-flatten = "MIT" ... LICENSE_${PN}-vary = "MIT"
Three key points exist in the previous example:
You can run the following command to build the
cute-files
package:
$ devtool build cute-files
Remember that nodejs
must be
installed on the target before your package.
Assuming 192.168.7.2 for the target's IP address, use the following command to deploy your package:
$ devtool deploy-target -s cute-files root@192.168.7.2
Once the package is installed on the target, you can test the application:
cute-files
as you would if you
had run npm install
.
$ cd /usr/lib/node_modules/cute-files $ node cute-files.js
On a browser, go to
http://192.168.7.2:3000
and you
see the following:
You can find the recipe in
workspace/recipes/cute-files
.
You can use the recipe in any layer you choose.
Although it is useful to package modules already in the
NPM registry, adding node.js
projects
under development is a more common developer use case.
This section covers the NPM projects code method, which is
very similar to the "registry" approach described in the
previous section.
In the NPM projects method, you provide
devtool
with an URL that points to the
source files.
Replicating the same example, (i.e.
cute-files
) use the following command:
$ devtool add https://github.com/martinaglv/cute-files.git
The recipe this command generates is very similar to the
recipe created in the previous section.
However, the SRC_URI
looks like the
following:
SRC_URI = " \ git://github.com/martinaglv/cute-files.git;protocol=https \ npmsw://${THISDIR}/${BPN}/npm-shrinkwrap.json \ "
In this example, the main module is taken from the Git repository and dependents are taken from the NPM registry. Other than those differences, the recipe is basically the same between the two methods. You can build and deploy the package exactly as described in the previous section that uses the registry modules method.
The OpenEmbedded build system works with source files located
through the
SRC_URI
variable.
When you build something using BitBake, a big part of the operation
is locating and downloading all the source tarballs.
For images, downloading all the source for various packages can
take a significant amount of time.
This section shows you how you can use mirrors to speed up fetching source files and how you can pre-fetch files all of which leads to more efficient use of resources and time.
A good deal that goes into a Yocto Project build is simply downloading all of the source tarballs. Maybe you have been working with another build system (OpenEmbedded or Angstrom) for which you have built up a sizable directory of source tarballs. Or, perhaps someone else has such a directory for which you have read access. If so, you can save time by adding statements to your configuration file so that the build process checks local directories first for existing tarballs before checking the Internet.
Here is an efficient way to set it up in your
local.conf
file:
SOURCE_MIRROR_URL ?= "file:///home/you/your-download-dir/" INHERIT += "own-mirrors" BB_GENERATE_MIRROR_TARBALLS = "1" # BB_NO_NETWORK = "1"
In the previous example, the
BB_GENERATE_MIRROR_TARBALLS
variable causes the OpenEmbedded build system to generate
tarballs of the Git repositories and store them in the
DL_DIR
directory.
Due to performance reasons, generating and storing these
tarballs is not the build system's default behavior.
You can also use the
PREMIRRORS
variable.
For an example, see the variable's glossary entry in the
Yocto Project Reference Manual.
Another technique you can use to ready yourself for a
successive string of build operations, is to pre-fetch
all the source files without actually starting a build.
This technique lets you work through any download issues
and ultimately gathers all the source files into your
download directory
build/downloads
,
which is located with
DL_DIR
.
Use the following BitBake command form to fetch all the necessary sources without starting the build:
$ bitbake target
--runall=fetch
This variation of the BitBake command guarantees that you have all the sources for that BitBake target should you disconnect from the Internet and want to do the build later offline.
By default, the Yocto Project uses SysVinit as the initialization manager. However, support also exists for systemd, which is a full replacement for init with parallel starting of services, reduced shell overhead and other features that are used by many distributions.
Within the system, SysVinit treats system components as services.
These services are maintained as shell scripts stored in the
/etc/init.d/
directory.
Services organize into different run levels.
This organization is maintained by putting links to the services
in the /etc/rcN.d/
directories, where
N/
is one of the following options:
"S", "0", "1", "2", "3", "4", "5", or "6".
In comparison, systemd treats components as units. Using units is a broader concept as compared to using a service. A unit includes several different types of entities. Service is one of the types of entities. The runlevel concept in SysVinit corresponds to the concept of a target in systemd, where target is also a type of supported unit.
In a SysVinit-based system, services load sequentially (i.e. one by one) during and parallelization is not supported. With systemd, services start in parallel. Needless to say, the method can have an impact on system startup performance.
If you want to use SysVinit, you do not have to do anything. But, if you want to use systemd, you must take some steps as described in the following sections.
Set these variables in your distribution configuration file as follows:
DISTRO_FEATURES_append = " systemd" VIRTUAL-RUNTIME_init_manager = "systemd"
You can also prevent the SysVinit distribution feature from being automatically enabled as follows:
DISTRO_FEATURES_BACKFILL_CONSIDERED = "sysvinit"
Doing so removes any redundant SysVinit scripts.
To remove initscripts from your image altogether, set this variable also:
VIRTUAL-RUNTIME_initscripts = ""
For information on the backfill variable, see
DISTRO_FEATURES_BACKFILL_CONSIDERED
.
Set these variables in your distribution configuration file as follows:
DISTRO_FEATURES_append = " systemd" VIRTUAL-RUNTIME_init_manager = "systemd"
Doing so causes your main image to use the
packagegroup-core-boot.bb
recipe and
systemd.
The rescue/minimal image cannot use this package group.
However, it can install SysVinit
and the appropriate packages will have support for both
systemd and SysVinit.
The Yocto Project provides multiple ways to manage the device
manager (/dev
):
Persistent and Pre-Populated/dev
:
For this case, the /dev
directory
is persistent and the required device nodes are created
during the build.
Use devtmpfs
with a Device Manager:
For this case, the /dev
directory
is provided by the kernel as an in-memory file system and
is automatically populated by the kernel at runtime.
Additional configuration of device nodes is done in user
space by a device manager like
udev
or
busybox-mdev
.
/dev
¶
To use the static method for device population, you need to
set the
USE_DEVFS
variable to "0" as follows:
USE_DEVFS = "0"
The content of the resulting /dev
directory is defined in a Device Table file.
The
IMAGE_DEVICE_TABLES
variable defines the Device Table to use and should be set
in the machine or distro configuration file.
Alternatively, you can set this variable in your
local.conf
configuration file.
If you do not define the
IMAGE_DEVICE_TABLES
variable, the default
device_table-minimal.txt
is used:
IMAGE_DEVICE_TABLES = "device_table-mymachine.txt"
The population is handled by the makedevs
utility during image creation:
devtmpfs
and a Device Manager¶
To use the dynamic method for device population, you need to
use (or be sure to set) the
USE_DEVFS
variable to "1", which is the default:
USE_DEVFS = "1"
With this setting, the resulting /dev
directory is populated by the kernel using
devtmpfs
.
Make sure the corresponding kernel configuration variable
CONFIG_DEVTMPFS
is set when building
you build a Linux kernel.
All devices created by devtmpfs
will be
owned by root
and have permissions
0600
.
To have more control over the device nodes, you can use a
device manager like udev
or
busybox-mdev
.
You choose the device manager by defining the
VIRTUAL-RUNTIME_dev_manager
variable
in your machine or distro configuration file.
Alternatively, you can set this variable in your
local.conf
configuration file:
VIRTUAL-RUNTIME_dev_manager = "udev" # Some alternative values # VIRTUAL-RUNTIME_dev_manager = "busybox-mdev" # VIRTUAL-RUNTIME_dev_manager = "systemd"
If you're working on a recipe that pulls from an external Source Code Manager (SCM), it is possible to have the OpenEmbedded build system notice new recipe changes added to the SCM and then build the resulting packages that depend on the new recipes by using the latest versions. This only works for SCMs from which it is possible to get a sensible revision number for changes. Currently, you can do this with Apache Subversion (SVN), Git, and Bazaar (BZR) repositories.
To enable this behavior, the
PV
of the recipe needs to reference
SRCPV
.
Here is an example:
PV = "1.2.3+git${SRCPV}"
Then, you can add the following to your
local.conf
:
SRCREV_pn-PN
= "${AUTOREV}"
PN
is the name of the recipe for which you want to enable automatic source
revision updating.
If you do not want to update your local configuration file, you can add the following directly to the recipe to finish enabling the feature:
SRCREV = "${AUTOREV}"
The Yocto Project provides a distribution named
poky-bleeding
, whose configuration
file contains the line:
require conf/distro/include/poky-floating-revisions.inc
This line pulls in the listed include file that contains numerous lines of exactly that form:
#SRCREV_pn-opkg-native ?= "${AUTOREV}" #SRCREV_pn-opkg-sdk ?= "${AUTOREV}" #SRCREV_pn-opkg ?= "${AUTOREV}" #SRCREV_pn-opkg-utils-native ?= "${AUTOREV}" #SRCREV_pn-opkg-utils ?= "${AUTOREV}" SRCREV_pn-gconf-dbus ?= "${AUTOREV}" SRCREV_pn-matchbox-common ?= "${AUTOREV}" SRCREV_pn-matchbox-config-gtk ?= "${AUTOREV}" SRCREV_pn-matchbox-desktop ?= "${AUTOREV}" SRCREV_pn-matchbox-keyboard ?= "${AUTOREV}" SRCREV_pn-matchbox-panel-2 ?= "${AUTOREV}" SRCREV_pn-matchbox-themes-extra ?= "${AUTOREV}" SRCREV_pn-matchbox-terminal ?= "${AUTOREV}" SRCREV_pn-matchbox-wm ?= "${AUTOREV}" SRCREV_pn-settings-daemon ?= "${AUTOREV}" SRCREV_pn-screenshot ?= "${AUTOREV}" . . .
These lines allow you to experiment with building a distribution that tracks the latest development source for numerous packages.
poky-bleeding
distribution
is not tested on a regular basis.
Keep this in mind if you use it.
Suppose, for security reasons, you need to disable your target device's root filesystem's write permissions (i.e. you need a read-only root filesystem). Or, perhaps you are running the device's operating system from a read-only storage device. For either case, you can customize your image for that behavior.
To create the read-only root filesystem, simply add the
"read-only-rootfs" feature to your image, normally in one of two ways.
The first way is to add the "read-only-rootfs" image feature
in the image's recipe file via the
IMAGE_FEATURES
variable:
IMAGE_FEATURES += "read-only-rootfs"
As an alternative, you can add the same feature from within your
build directory's local.conf
file with the
associated EXTRA_IMAGE_FEATURES
variable, as in:
EXTRA_IMAGE_FEATURES = "read-only-rootfs"
For more information on how to use these variables, see the
"Customizing Images Using Custom IMAGE_FEATURES
and EXTRA_IMAGE_FEATURES
"
section.
For information on the variables, see
IMAGE_FEATURES
and EXTRA_IMAGE_FEATURES
.
It is very important that you make sure all
post-Installation (pkg_postinst
) scripts
for packages that are installed into the image can be run
at the time when the root filesystem is created during the
build on the host system.
These scripts cannot attempt to run during first-boot on the
target device.
With the "read-only-rootfs" feature enabled,
the build system checks during root filesystem creation to make
sure all post-installation scripts succeed.
If any of these scripts still need to be run after the root
filesystem is created, the build immediately fails.
These build-time checks ensure that the build fails
rather than the target device fails later during its
initial boot operation.
Most of the common post-installation scripts generated by the build system for the out-of-the-box Yocto Project are engineered so that they can run during root filesystem creation (e.g. post-installation scripts for caching fonts). However, if you create and add custom scripts, you need to be sure they can be run during this file system creation.
Here are some common problems that prevent post-installation scripts from running during root filesystem creation:
Not using $D in front of absolute
paths:
The build system defines
$
D
when the root filesystem is created.
Furthermore, $D
is blank when the
script is run on the target device.
This implies two purposes for $D
:
ensuring paths are valid in both the host and target
environments, and checking to determine which
environment is being used as a method for taking
appropriate actions.
Attempting to run processes that are
specific to or dependent on the target
architecture:
You can work around these attempts by using native
tools, which run on the host system,
to accomplish the same tasks, or
by alternatively running the processes under QEMU,
which has the qemu_run_binary
function.
For more information, see the
qemu
class.
With the "read-only-rootfs" feature enabled,
any attempt by the target to write to the root filesystem at
runtime fails.
Consequently, you must make sure that you configure processes
and applications that attempt these types of writes do so
to directories with write access (e.g.
/tmp
or /var/run
).
Many factors can influence the quality of a build. For example, if you upgrade a recipe to use a new version of an upstream software package or you experiment with some new configuration options, subtle changes can occur that you might not detect until later. Consider the case where your recipe is using a newer version of an upstream package. In this case, a new version of a piece of software might introduce an optional dependency on another library, which is auto-detected. If that library has already been built when the software is building, the software will link to the built library and that library will be pulled into your image along with the new software even if you did not want the library.
The
buildhistory
class exists to help you maintain the quality of your build
output.
You can use the class to highlight unexpected and possibly
unwanted changes in the build output.
When you enable build history, it records information about the
contents of each package and image and then commits that
information to a local Git repository where you can examine
the information.
The remainder of this section describes the following:
How you can enable and disable build history
How to understand what the build history contains
How to limit the information used for build history
How to examine the build history from both a command-line and web interface
Build history is disabled by default.
To enable it, add the following INHERIT
statement and set the
BUILDHISTORY_COMMIT
variable to "1" at the end of your
conf/local.conf
file found in the
Build Directory:
INHERIT += "buildhistory" BUILDHISTORY_COMMIT = "1"
Enabling build history as previously described causes the OpenEmbedded build system to collect build output information and commit it as a single commit to a local Git repository.
You can disable build history by removing the previous
statements from your conf/local.conf
file.
Build history information is kept in
${
TOPDIR
}/buildhistory
in the Build Directory as defined by the
BUILDHISTORY_DIR
variable.
The following is an example abbreviated listing:
At the top level, a metadata-revs
file exists that lists the revisions of the repositories for
the enabled layers when the build was produced.
The rest of the data splits into separate
packages
, images
and sdk
directories, the contents of
which are described as follows.
The history for each package contains a text file that has
name-value pairs with information about the package.
For example,
buildhistory/packages/i586-poky-linux/busybox/busybox/latest
contains the following:
PV = 1.22.1 PR = r32 RPROVIDES = RDEPENDS = glibc (>= 2.20) update-alternatives-opkg RRECOMMENDS = busybox-syslog busybox-udhcpc update-rc.d PKGSIZE = 540168 FILES = /usr/bin/* /usr/sbin/* /usr/lib/busybox/* /usr/lib/lib*.so.* \ /etc /com /var /bin/* /sbin/* /lib/*.so.* /lib/udev/rules.d \ /usr/lib/udev/rules.d /usr/share/busybox /usr/lib/busybox/* \ /usr/share/pixmaps /usr/share/applications /usr/share/idl \ /usr/share/omf /usr/share/sounds /usr/lib/bonobo/servers FILELIST = /bin/busybox /bin/busybox.nosuid /bin/busybox.suid /bin/sh \ /etc/busybox.links.nosuid /etc/busybox.links.suid
Most of these name-value pairs correspond to variables
used to produce the package.
The exceptions are FILELIST
, which
is the actual list of files in the package, and
PKGSIZE
, which is the total size of
files in the package in bytes.
A file also exists that corresponds to the recipe from
which the package came (e.g.
buildhistory/packages/i586-poky-linux/busybox/latest
):
PV = 1.22.1 PR = r32 DEPENDS = initscripts kern-tools-native update-rc.d-native \ virtual/i586-poky-linux-compilerlibs virtual/i586-poky-linux-gcc \ virtual/libc virtual/update-alternatives PACKAGES = busybox-ptest busybox-httpd busybox-udhcpd busybox-udhcpc \ busybox-syslog busybox-mdev busybox-hwclock busybox-dbg \ busybox-staticdev busybox-dev busybox-doc busybox-locale busybox
Finally, for those recipes fetched from a version control
system (e.g., Git), a file exists that lists source
revisions that are specified in the recipe and lists
the actual revisions used during the build.
Listed and actual revisions might differ when
SRCREV
is set to
${AUTOREV
}.
Here is an example assuming
buildhistory/packages/qemux86-poky-linux/linux-yocto/latest_srcrev
):
# SRCREV_machine = "38cd560d5022ed2dbd1ab0dca9642e47c98a0aa1" SRCREV_machine = "38cd560d5022ed2dbd1ab0dca9642e47c98a0aa1" # SRCREV_meta = "a227f20eff056e511d504b2e490f3774ab260d6f" SRCREV_meta = "a227f20eff056e511d504b2e490f3774ab260d6f"
You can use the
buildhistory-collect-srcrevs
command with the -a
option to
collect the stored SRCREV
values
from build history and report them in a format suitable for
use in global configuration (e.g.,
local.conf
or a distro include file)
to override floating AUTOREV
values
to a fixed set of revisions.
Here is some example output from this command:
$ buildhistory-collect-srcrevs -a # i586-poky-linux SRCREV_pn-glibc = "b8079dd0d360648e4e8de48656c5c38972621072" SRCREV_pn-glibc-initial = "b8079dd0d360648e4e8de48656c5c38972621072" SRCREV_pn-opkg-utils = "53274f087565fd45d8452c5367997ba6a682a37a" SRCREV_pn-kmod = "fd56638aed3fe147015bfa10ed4a5f7491303cb4" # x86_64-linux SRCREV_pn-gtk-doc-stub-native = "1dea266593edb766d6d898c79451ef193eb17cfa" SRCREV_pn-dtc-native = "65cc4d2748a2c2e6f27f1cf39e07a5dbabd80ebf" SRCREV_pn-update-rc.d-native = "eca680ddf28d024954895f59a241a622dd575c11" SRCREV_glibc_pn-cross-localedef-native = "b8079dd0d360648e4e8de48656c5c38972621072" SRCREV_localedef_pn-cross-localedef-native = "c833367348d39dad7ba018990bfdaffaec8e9ed3" SRCREV_pn-prelink-native = "faa069deec99bf61418d0bab831c83d7c1b797ca" SRCREV_pn-opkg-utils-native = "53274f087565fd45d8452c5367997ba6a682a37a" SRCREV_pn-kern-tools-native = "23345b8846fe4bd167efdf1bd8a1224b2ba9a5ff" SRCREV_pn-kmod-native = "fd56638aed3fe147015bfa10ed4a5f7491303cb4" # qemux86-poky-linux SRCREV_machine_pn-linux-yocto = "38cd560d5022ed2dbd1ab0dca9642e47c98a0aa1" SRCREV_meta_pn-linux-yocto = "a227f20eff056e511d504b2e490f3774ab260d6f" # all-poky-linux SRCREV_pn-update-rc.d = "eca680ddf28d024954895f59a241a622dd575c11"
buildhistory-collect-srcrevs
command:
By default, only values where the
SRCREV
was not hardcoded
(usually when AUTOREV
is used) are reported.
Use the -a
option to
see all SRCREV
values.
The output statements might not have any effect
if overrides are applied elsewhere in the
build system configuration.
Use the -f
option to add
the forcevariable
override
to each output line if you need to work around
this restriction.
The script does apply special handling when
building for multiple machines.
However, the script does place a comment before
each set of values that specifies which
triplet to which they belong as previously
shown (e.g.,
i586-poky-linux
).
The files produced for each image are as follows:
image-files:
A directory containing selected files from the root
filesystem.
The files are defined by
BUILDHISTORY_IMAGE_FILES
.
build-id.txt:
Human-readable information about the build
configuration and metadata source revisions.
This file contains the full build header as printed
by BitBake.
*.dot:
Dependency graphs for the image that are
compatible with graphviz
.
files-in-image.txt:
A list of files in the image with permissions,
owner, group, size, and symlink information.
image-info.txt:
A text file containing name-value pairs with
information about the image.
See the following listing example for more
information.
installed-package-names.txt:
A list of installed packages by name only.
installed-package-sizes.txt:
A list of installed packages ordered by size.
installed-packages.txt:
A list of installed packages with full package
filenames.
Here is an example of image-info.txt
:
DISTRO = poky DISTRO_VERSION = 1.7 USER_CLASSES = buildstats image-mklibs image-prelink IMAGE_CLASSES = image_types IMAGE_FEATURES = debug-tweaks IMAGE_LINGUAS = IMAGE_INSTALL = packagegroup-core-boot run-postinsts BAD_RECOMMENDATIONS = NO_RECOMMENDATIONS = PACKAGE_EXCLUDE = ROOTFS_POSTPROCESS_COMMAND = write_package_manifest; license_create_manifest; \ write_image_manifest ; buildhistory_list_installed_image ; \ buildhistory_get_image_installed ; ssh_allow_empty_password; \ postinst_enable_logging; rootfs_update_timestamp ; ssh_disable_dns_lookup ; IMAGE_POSTPROCESS_COMMAND = buildhistory_get_imageinfo ; IMAGESIZE = 6900
Other than IMAGESIZE
, which is the
total size of the files in the image in Kbytes, the
name-value pairs are variables that may have influenced the
content of the image.
This information is often useful when you are trying to
determine why a change in the package or file
listings has occurred.
As you can see, build history produces image information,
including dependency graphs, so you can see why something
was pulled into the image.
If you are just interested in this information and not
interested in collecting specific package or SDK
information, you can enable writing only image information
without any history by adding the following to your
conf/local.conf
file found in the
Build Directory:
INHERIT += "buildhistory" BUILDHISTORY_COMMIT = "0" BUILDHISTORY_FEATURES = "image"
Here, you set the
BUILDHISTORY_FEATURES
variable to use the image feature only.
Build history collects similar information on the contents
of SDKs
(e.g. bitbake -c populate_sdk imagename
)
as compared to information it collects for images.
Furthermore, this information differs depending on whether
an extensible or standard SDK is being produced.
The following list shows the files produced for SDKs:
files-in-sdk.txt:
A list of files in the SDK with permissions,
owner, group, size, and symlink information.
This list includes both the host and target parts
of the SDK.
sdk-info.txt:
A text file containing name-value pairs with
information about the SDK.
See the following listing example for more
information.
sstate-task-sizes.txt:
A text file containing name-value pairs with
information about task group sizes
(e.g. do_populate_sysroot
tasks have a total size).
The sstate-task-sizes.txt
file
exists only when an extensible SDK is created.
sstate-package-sizes.txt:
A text file containing name-value pairs with
information for the shared-state packages and
sizes in the SDK.
The sstate-package-sizes.txt
file exists only when an extensible SDK is created.
sdk-files:
A folder that contains copies of the files
mentioned in
BUILDHISTORY_SDK_FILES
if the
files are present in the output.
Additionally, the default value of
BUILDHISTORY_SDK_FILES
is
specific to the extensible SDK although you can
set it differently if you would like to pull in
specific files from the standard SDK.
The default files are
conf/local.conf
,
conf/bblayers.conf
,
conf/auto.conf
,
conf/locked-sigs.inc
, and
conf/devtool.conf
.
Thus, for an extensible SDK, these files get
copied into the sdk-files
directory.
The following information appears under
each of the host
and target
directories
for the portions of the SDK that run on the host
and on the target, respectively:
depends.dot:
Dependency graph for the SDK that is
compatible with
graphviz
.
installed-package-names.txt:
A list of installed packages by name only.
installed-package-sizes.txt:
A list of installed packages ordered by size.
installed-packages.txt:
A list of installed packages with full
package filenames.
Here is an example of sdk-info.txt
:
DISTRO = poky DISTRO_VERSION = 1.3+snapshot-20130327 SDK_NAME = poky-glibc-i686-arm SDK_VERSION = 1.3+snapshot SDKMACHINE = SDKIMAGE_FEATURES = dev-pkgs dbg-pkgs BAD_RECOMMENDATIONS = SDKSIZE = 352712
Other than SDKSIZE
, which is the
total size of the files in the SDK in Kbytes, the
name-value pairs are variables that might have influenced
the content of the SDK.
This information is often useful when you are trying to
determine why a change in the package or file listings
has occurred.
You can examine build history output from the command line or from a web interface.
To see any changes that have occurred (assuming you have
BUILDHISTORY_COMMIT
= "1"
),
you can simply use any Git command that allows you to
view the history of a repository.
Here is one method:
$ git log -p
You need to realize, however, that this method does show changes that are not significant (e.g. a package's size changing by a few bytes).
A command-line tool called
buildhistory-diff
does exist, though,
that queries the Git repository and prints just the
differences that might be significant in human-readable
form.
Here is an example:
$ ~/poky/poky/scripts/buildhistory-diff . HEAD^ Changes to images/qemux86_64/glibc/core-image-minimal (files-in-image.txt): /etc/anotherpkg.conf was added /sbin/anotherpkg was added * (installed-package-names.txt): * anotherpkg was added Changes to images/qemux86_64/glibc/core-image-minimal (installed-package-names.txt): anotherpkg was added packages/qemux86_64-poky-linux/v86d: PACKAGES: added "v86d-extras" * PR changed from "r0" to "r1" * PV changed from "0.1.10" to "0.1.12" packages/qemux86_64-poky-linux/v86d/v86d: PKGSIZE changed from 110579 to 144381 (+30%) * PR changed from "r0" to "r1" * PV changed from "0.1.10" to "0.1.12"
buildhistory-diff
tool
requires the GitPython
package.
Be sure to install it using Pip3 as follows:
$ pip3 install GitPython --userAlternatively, you can install
python3-git
using the appropriate
distribution package manager (e.g.
apt-get
, dnf
,
or zipper
).
To see changes to the build history using a web interface,
follow the instruction in the README
file here.
http://git.yoctoproject.org/cgit/cgit.cgi/buildhistory-web/.
Here is a sample screenshot of the interface:
The OpenEmbedded build system makes available a series of automated
tests for images to verify runtime functionality.
You can run these tests on either QEMU or actual target hardware.
Tests are written in Python making use of the
unittest
module, and the majority of them
run commands on the target system over SSH.
This section describes how you set up the environment to use these
tests, run available tests, and write and add your own tests.
For information on the test and QA infrastructure available within the Yocto Project, see the "Testing and Quality Assurance" section in the Yocto Project Reference Manual.
Depending on whether you are planning to run tests using QEMU or on the hardware, you have to take different steps to enable the tests. See the following subsections for information on how to enable both types of tests.
In order to run tests, you need to do the following:
Set up to avoid interaction
with sudo
for networking:
To accomplish this, you must do one of the
following:
Add
NOPASSWD
for your user
in /etc/sudoers
either for
all commands or just for
runqemu-ifup
.
You must provide the full path as that can
change if you are using multiple clones of the
source repository.
/etc/sudoers
.
Manually configure a tap interface for your system.
Run as root the script in
scripts/runqemu-gen-tapdevs
,
which should generate a list of tap devices.
This is the option typically chosen for
Autobuilder-type environments.
Be sure to use an absolute path when calling this script with sudo.
The package recipe
qemu-helper-native
is required to run this script.
Build the package using the
following command:
$ bitbake qemu-helper-native
Set the
DISPLAY
variable:
You need to set this variable so that you have an X
server available (e.g. start
vncserver
for a headless machine).
Be sure your host's firewall
accepts incoming connections from
192.168.7.0/24:
Some of the tests (in particular DNF tests) start
an HTTP server on a random high number port,
which is used to serve files to the target.
The DNF module serves
${WORKDIR}/oe-rootfs-repo
so it can run DNF channel commands.
That means your host's firewall
must accept incoming connections from 192.168.7.0/24,
which is the default IP range used for tap devices
by runqemu
.
Be sure your host has the correct packages installed: Depending your host's distribution, you need to have the following packages installed:
Ubuntu and Debian:
sysstat
and
iproute2
OpenSUSE:
sysstat
and
iproute2
Fedora:
sysstat
and
iproute
CentOS:
sysstat
and
iproute
Once you start running the tests, the following happens:
A copy of the root filesystem is written
to ${WORKDIR}/testimage
.
The image is booted under QEMU using the
standard runqemu
script.
A default timeout of 500 seconds occurs
to allow for the boot process to reach the login prompt.
You can change the timeout period by setting
TEST_QEMUBOOT_TIMEOUT
in the local.conf
file.
Once the boot process is reached and the
login prompt appears, the tests run.
The full boot log is written to
${WORKDIR}/testimage/qemu_boot_log
.
Each test module loads in the order found
in TEST_SUITES
.
You can find the full output of the commands run over
SSH in
${WORKDIR}/testimgage/ssh_target_log
.
If no failures occur, the task running the
tests ends successfully.
You can find the output from the
unittest
in the task log at
${WORKDIR}/temp/log.do_testimage
.
The OpenEmbedded build system can run tests on real hardware, and for certain devices it can also deploy the image to be tested onto the device beforehand.
For automated deployment, a "master image" is installed onto the hardware once as part of setup. Then, each time tests are to be run, the following occurs:
The master image is booted into and used to write the image to be tested to a second partition.
The device is then rebooted using an external script that you need to provide.
The device boots into the image to be tested.
When running tests (independent of whether the image has been deployed automatically or not), the device is expected to be connected to a network on a pre-determined IP address. You can either use static IP addresses written into the image, or set the image to use DHCP and have your DHCP server on the test network assign a known IP address based on the MAC address of the device.
In order to run tests on hardware, you need to set
TEST_TARGET
to an appropriate value.
For QEMU, you do not have to change anything, the default
value is "qemu".
For running tests on hardware, the following options exist:
"simpleremote": Choose "simpleremote" if you are going to run tests on a target system that is already running the image to be tested and is available on the network. You can use "simpleremote" in conjunction with either real hardware or an image running within a separately started QEMU or any other virtual machine manager.
"SystemdbootTarget":
Choose "SystemdbootTarget" if your hardware is
an EFI-based machine with
systemd-boot
as bootloader and
core-image-testmaster
(or something similar) is installed.
Also, your hardware under test must be in a
DHCP-enabled network that gives it the same IP
address for each reboot.
If you choose "SystemdbootTarget", there are additional requirements and considerations. See the "Selecting SystemdbootTarget" section, which follows, for more information.
"BeagleBoneTarget":
Choose "BeagleBoneTarget" if you are deploying
images and running tests on the BeagleBone
"Black" or original "White" hardware.
For information on how to use these tests, see the
comments at the top of the BeagleBoneTarget
meta-yocto-bsp/lib/oeqa/controllers/beaglebonetarget.py
file.
"EdgeRouterTarget":
Choose "EdgeRouterTarget" is you are deploying
images and running tests on the Ubiquiti Networks
EdgeRouter Lite.
For information on how to use these tests, see the
comments at the top of the EdgeRouterTarget
meta-yocto-bsp/lib/oeqa/controllers/edgeroutertarget.py
file.
"GrubTarget":
Choose the "supports deploying images and running
tests on any generic PC that boots using GRUB.
For information on how to use these tests, see the
comments at the top of the GrubTarget
meta-yocto-bsp/lib/oeqa/controllers/grubtarget.py
file.
"your-target
":
Create your own custom target if you want to run
tests when you are deploying images and running
tests on a custom machine within your BSP layer.
To do this, you need to add a Python unit that
defines the target class under
lib/oeqa/controllers/
within
your layer.
You must also provide an empty
__init__.py
.
For examples, see files in
meta-yocto-bsp/lib/oeqa/controllers/
.
If you did not set TEST_TARGET
to
"SystemdbootTarget", then you do not need any information
in this section.
You can skip down to the
"Running Tests"
section.
If you did set TEST_TARGET
to
"SystemdbootTarget", you also need to perform a one-time
setup of your master image by doing the following:
Set EFI_PROVIDER
:
Be sure that EFI_PROVIDER
is as follows:
EFI_PROVIDER = "systemd-boot"
Build the master image:
Build the core-image-testmaster
image.
The core-image-testmaster
recipe is provided as an example for a
"master" image and you can customize the image
recipe as you would any other recipe.
Here are the image recipe requirements:
Inherits
core-image
so that kernel modules are installed.
Installs normal linux utilities
not busybox ones (e.g.
bash
,
coreutils
,
tar
,
gzip
, and
kmod
).
Uses a custom Initial RAM Disk (initramfs) image with a custom installer. A normal image that you can install usually creates a single rootfs partition. This image uses another installer that creates a specific partition layout. Not all Board Support Packages (BSPs) can use an installer. For such cases, you need to manually create the following partition layout on the target:
First partition mounted
under /boot
,
labeled "boot".
The main rootfs
partition where this image gets
installed, which is mounted under
/
.
Another partition labeled "testrootfs" where test images get deployed.
Install image: Install the image that you just built on the target system.
The final thing you need to do when setting
TEST_TARGET
to "SystemdbootTarget" is
to set up the test image:
Set up your local.conf
file:
Make sure you have the following statements in
your local.conf
file:
IMAGE_FSTYPES += "tar.gz" INHERIT += "testimage" TEST_TARGET = "SystemdbootTarget" TEST_TARGET_IP = "192.168.2.3"
Build your test image: Use BitBake to build the image:
$ bitbake core-image-sato
For most hardware targets other than "simpleremote", you can control power:
You can use
TEST_POWERCONTROL_CMD
together with
TEST_POWERCONTROL_EXTRA_ARGS
as a command that runs on the host and does power
cycling.
The test code passes one argument to that command:
off, on or cycle (off then on).
Here is an example that could appear in your
local.conf
file:
TEST_POWERCONTROL_CMD = "powercontrol.exp test 10.11.12.1 nuc1"
In this example, the expect script does the following:
ssh test@10.11.12.1 "pyctl nuc1 arg
"
It then runs a Python script that controls power
for a label called nuc1
.
TEST_POWERCONTROL_CMD
and
TEST_POWERCONTROL_EXTRA_ARGS
for your own setup.
The one requirement is that it accepts
"on", "off", and "cycle" as the last argument.
When no command is defined, it connects to the device over SSH and uses the classic reboot command to reboot the device. Classic reboot is fine as long as the machine actually reboots (i.e. the SSH test has not failed). It is useful for scenarios where you have a simple setup, typically with a single board, and where some manual interaction is okay from time to time.
If you have no hardware to automatically perform power
control but still wish to experiment with automated
hardware testing, you can use the dialog-power-control
script that shows a dialog prompting you to perform the
required power action.
This script requires either KDialog or Zenity to be
installed.
To use this script, set the
TEST_POWERCONTROL_CMD
variable as follows:
TEST_POWERCONTROL_CMD = "${COREBASE}/scripts/contrib/dialog-power-control"
For test target classes requiring a serial console
to interact with the bootloader (e.g. BeagleBoneTarget,
EdgeRouterTarget, and GrubTarget), you need to
specify a command to use to connect to the serial console
of the target machine by using the
TEST_SERIALCONTROL_CMD
variable and optionally the
TEST_SERIALCONTROL_EXTRA_ARGS
variable.
These cases could be a serial terminal program if the
machine is connected to a local serial port, or a
telnet
or
ssh
command connecting to a remote
console server.
Regardless of the case, the command simply needs to
connect to the serial console and forward that connection
to standard input and output as any normal terminal
program does.
For example, to use the picocom terminal program on
serial device /dev/ttyUSB0
at 115200bps, you would set the variable as follows:
TEST_SERIALCONTROL_CMD = "picocom /dev/ttyUSB0 -b 115200"
For local devices where the serial port device disappears
when the device reboots, an additional "serdevtry" wrapper
script is provided.
To use this wrapper, simply prefix the terminal command
with
${COREBASE}/scripts/contrib/serdevtry
:
TEST_SERIALCONTROL_CMD = "${COREBASE}/scripts/contrib/serdevtry picocom -b 115200 /dev/ttyUSB0"
You can start the tests automatically or manually:
Automatically running tests:
To run the tests automatically after the
OpenEmbedded build system successfully creates an image,
first set the
TESTIMAGE_AUTO
variable to "1" in your local.conf
file in the
Build Directory:
TESTIMAGE_AUTO = "1"
Next, build your image. If the image successfully builds, the tests run:
bitbake core-image-sato
Manually running tests:
To manually run the tests, first globally inherit the
testimage
class by editing your local.conf
file:
INHERIT += "testimage"
Next, use BitBake to run the tests:
bitbake -c testimage image
All test files reside in
meta/lib/oeqa/runtime
in the
Source Directory.
A test name maps directly to a Python module.
Each test module may contain a number of individual tests.
Tests are usually grouped together by the area
tested (e.g tests for systemd reside in
meta/lib/oeqa/runtime/systemd.py
).
You can add tests to any layer provided you place them in the
proper area and you extend
BBPATH
in the local.conf
file as normal.
Be sure that tests reside in
.
layer
/lib/oeqa/runtime
meta/lib/oeqa/runtime
.
You can change the set of tests run by appending or overriding
TEST_SUITES
variable in local.conf
.
Each name in TEST_SUITES
represents a
required test for the image.
Test modules named within TEST_SUITES
cannot be skipped even if a test is not suitable for an image
(e.g. running the RPM tests on an image without
rpm
).
Appending "auto" to TEST_SUITES
causes the
build system to try to run all tests that are suitable for the
image (i.e. each test module may elect to skip itself).
The order you list tests in TEST_SUITES
is important and influences test dependencies.
Consequently, tests that depend on other tests should be added
after the test on which they depend.
For example, since the ssh
test
depends on the
ping
test, "ssh" needs to come after
"ping" in the list.
The test class provides no re-ordering or dependency handling.
unittest
rules apply.
Here are some things to keep in mind when running tests:
The default tests for the image are defined as:
DEFAULT_TEST_SUITES_pn-image
= "ping ssh df connman syslog xorg scp vnc date rpm dnf dmesg"
Add your own test to the list of the by using the following:
TEST_SUITES_append = " mytest"
Run a specific list of tests as follows:
TEST_SUITES = "test1 test2 test3"
Remember, order is important. Be sure to place a test that is dependent on another test later in the order.
You can export tests so that they can run independently of
the build system.
Exporting tests is required if you want to be able to hand
the test execution off to a scheduler.
You can only export tests that are defined in
TEST_SUITES
.
If your image is already built, make sure the following are set
in your local.conf
file:
INHERIT +="testexport" TEST_TARGET_IP = "IP-address-for-the-test-target
" TEST_SERVER_IP = "IP-address-for-the-test-server
"
You can then export the tests with the following BitBake command form:
$ bitbake image
-c testexport
Exporting the tests places them in the
Build Directory
in
tmp/testexport/
image
,
which is controlled by the
TEST_EXPORT_DIR
variable.
You can now run the tests outside of the build environment:
$ cd tmp/testexport/image
$ ./runexported.py testdata.json
Here is a complete example that shows IP addresses and uses
the core-image-sato
image:
INHERIT +="testexport" TEST_TARGET_IP = "192.168.7.2" TEST_SERVER_IP = "192.168.7.1"
Use BitBake to export the tests:
$ bitbake core-image-sato -c testexport
Run the tests outside of the build environment using the following:
$ cd tmp/testexport/core-image-sato $ ./runexported.py testdata.json
As mentioned previously, all new test files need to be in the
proper place for the build system to find them.
New tests for additional functionality outside of the core
should be added to the layer that adds the functionality, in
(as long as
layer
/lib/oeqa/runtimeBBPATH
is extended in the layer's
layer.conf
file as normal).
Just remember the following:
Filenames need to map directly to test (module) names.
Do not use module names that collide with existing core tests.
Minimally, an empty
__init__.py
file must exist
in the runtime directory.
To create a new test, start by copying an existing module
(e.g. syslog.py
or
gcc.py
are good ones to use).
Test modules can use code from
meta/lib/oeqa/utils
, which are helper
classes.
df.py
and
date.py
modules for examples.
You will notice that all test classes inherit
oeRuntimeTest
, which is found in
meta/lib/oetest.py
.
This base class offers some helper attributes, which are
described in the following sections:
Class methods are as follows:
hasPackage(pkg)
:
Returns "True" if pkg
is in the
installed package list of the image, which is based
on the manifest file that is generated during the
do_rootfs
task.
hasFeature(feature)
:
Returns "True" if the feature is in
IMAGE_FEATURES
or
DISTRO_FEATURES
.
Class attributes are as follows:
pscmd
:
Equals "ps -ef" if procps
is
installed in the image.
Otherwise, pscmd
equals
"ps" (busybox).
tc
:
The called test context, which gives access to the
following attributes:
d
:
The BitBake datastore, which allows you to
use stuff such as
oeRuntimeTest.tc.d.getVar("VIRTUAL-RUNTIME_init_manager")
.
testslist
and testsrequired
:
Used internally.
The tests do not need these.
filesdir
:
The absolute path to
meta/lib/oeqa/runtime/files
,
which contains helper files for tests meant
for copying on the target such as small
files written in C for compilation.
target
:
The target controller object used to deploy
and start an image on a particular target
(e.g. Qemu, SimpleRemote, and
SystemdbootTarget).
Tests usually use the following:
ip
:
The target's IP address.
server_ip
:
The host's IP address, which is
usually used by the DNF test
suite.
run(cmd, timeout=None)
:
The single, most used method.
This command is a wrapper for:
ssh root@host "cmd"
.
The command returns a tuple:
(status, output), which are what
their names imply - the return code
of "cmd" and whatever output
it produces.
The optional timeout argument
represents the number of seconds the
test should wait for "cmd" to
return.
If the argument is "None", the
test uses the default instance's
timeout period, which is 300
seconds.
If the argument is "0", the test
runs until the command returns.
copy_to(localpath, remotepath)
:
scp localpath root@ip:remotepath
.
copy_from(remotepath, localpath)
:
scp root@host:remotepath localpath
.
A single instance attribute exists, which is
target
.
The target
instance attribute is
identical to the class attribute of the same name, which
is described in the previous section.
This attribute exists as both an instance and class
attribute so tests can use
self.target.run(cmd)
in instance
methods instead of
oeRuntimeTest.tc.target.run(cmd)
.
When a test requires a package built by BitBake, it is possible
to install that package.
Installing the package does not require a package manager be
installed in the device under test (DUT).
It does, however, require an SSH connection and the target must
be using the sshcontrol
class.
scp
to copy files
from the host to the target, which causes permissions and
special attributes to be lost.
A JSON file is used to define the packages needed by a test.
This file must be in the same path as the file used to define
the tests.
Furthermore, the filename must map directly to the test
module name with a .json
extension.
The JSON file must include an object with the test name as keys of an object or an array. This object (or array of objects) uses the following data:
"pkg" - A mandatory string that is the name of the package to be installed.
"rm" - An optional boolean, which defaults to "false", that specifies to remove the package after the test.
"extract" - An optional boolean, which defaults to "false", that specifies if the package must be extracted from the package format. When set to "true", the package is not automatically installed into the DUT.
Following is an example JSON file that handles test "foo" installing package "bar" and test "foobar" installing packages "foo" and "bar". Once the test is complete, the packages are removed from the DUT.
{ "foo": { "pkg": "bar" }, "foobar": [ { "pkg": "foo", "rm": true }, { "pkg": "bar", "rm": true } ] }
The exact method for debugging build failures depends on the nature of the problem and on the system's area from which the bug originates. Standard debugging practices such as comparison against the last known working version with examination of the changes and the re-application of steps to identify the one causing the problem are valid for the Yocto Project just as they are for any other system. Even though it is impossible to detail every possible potential failure, this section provides some general tips to aid in debugging given a variety of situations.
The following list shows the debugging topics in the remainder of this section:
"Viewing Logs from Failed Tasks" describes how to find and view logs from tasks that failed during the build process.
"Viewing Variable Values"
describes how to use the BitBake -e
option to examine variable values after a recipe has been
parsed.
"Viewing Package Information with oe-pkgdata-util
"
describes how to use the
oe-pkgdata-util
utility to query
PKGDATA_DIR
and display package-related information for built
packages.
"Viewing Dependencies Between Recipes and Tasks"
describes how to use the BitBake -g
option to display recipe dependency information used
during the build.
"Viewing Task Variable Dependencies"
describes how to use the
bitbake-dumpsig
command in
conjunction with key subdirectories in the
Build Directory
to determine variable dependencies.
"Running Specific Tasks"
describes how to use several BitBake options (e.g.
-c
, -C
, and
-f
) to run specific tasks in the
build chain.
It can be useful to run tasks "out-of-order" when trying
isolate build issues.
"General BitBake Problems"
describes how to use BitBake's -D
debug output option to reveal more about what BitBake is
doing during the build.
"Building with No Dependencies"
describes how to use the BitBake -b
option to build a recipe while ignoring dependencies.
"Recipe Logging Mechanisms" describes how to use the many recipe logging functions to produce debugging output and report errors and warnings.
"Debugging Parallel Make Races" describes how to debug situations where the build consists of several parts that are run simultaneously and when the output or result of one part is not ready for use with a different part of the build that depends on that output.
"Debugging With the GNU Project Debugger (GDB) Remotely" describes how to use GDB to allow you to examine running programs, which can help you fix problems.
"Debugging with the GNU Project Debugger (GDB) on the Target" describes how to use GDB directly on target hardware for debugging.
"Other Debugging Tips" describes miscellaneous debugging tips that can be useful.
You can find the log for a task in the file
${
WORKDIR
}/temp/log.do_
taskname
.
For example, the log for the
do_compile
task of the QEMU minimal image for the x86 machine
(qemux86
) might be in
tmp/work/qemux86-poky-linux/core-image-minimal/1.0-r0/temp/log.do_compile
.
To see the commands
BitBake
ran to generate a log, look at the corresponding
run.do_
taskname
file in the same directory.
log.do_
taskname
and
run.do_
taskname
are actually symbolic links to
log.do_
taskname
.
pid
and
log.run_
taskname
.
pid
,
where pid
is the PID the task had
when it ran.
The symlinks always point to the files corresponding to the most
recent run.
Sometimes you need to know the value of a variable as a result of BitBake's parsing step. This could be because some unexpected behavior occurred in your project. Perhaps an attempt to modify a variable did not work out as expected.
BitBake's -e
option is used to display
variable values after parsing.
The following command displays the variable values after the
configuration files (i.e. local.conf
,
bblayers.conf
,
bitbake.conf
and so forth) have been
parsed:
$ bitbake -e
The following command displays variable values after a specific recipe has been parsed. The variables include those from the configuration as well:
$ bitbake -e recipename
Each recipe has its own private set of variables (datastore). Internally, after parsing the configuration, a copy of the resulting datastore is made prior to parsing each recipe. This copying implies that variables set in one recipe will not be visible to other recipes.
Likewise, each task within a recipe gets a private datastore based on the recipe datastore, which means that variables set within one task will not be visible to other tasks.
In the output of bitbake -e
, each
variable is preceded by a description of how the variable
got its value, including temporary values that were later
overriden.
This description also includes variable flags (varflags) set on
the variable.
The output can be very helpful during debugging.
Variables that are exported to the environment are preceded by
export
in the output of
bitbake -e
.
See the following example:
export CC="i586-poky-linux-gcc -m32 -march=i586 --sysroot=/home/ulf/poky/build/tmp/sysroots/qemux86"
In addition to variable values, the output of the
bitbake -e
and
bitbake -e
recipe
commands includes the following information:
The output starts with a tree listing all configuration
files and classes included globally, recursively listing
the files they include or inherit in turn.
Much of the behavior of the OpenEmbedded build system
(including the behavior of the
normal recipe build tasks)
is implemented in the
base
class and the classes it inherits, rather than being
built into BitBake itself.
After the variable values, all functions appear in the
output.
For shell functions, variables referenced within the
function body are expanded.
If a function has been modified using overrides or
using override-style operators like
_append
and
_prepend
, then the final assembled
function body appears in the output.
oe-pkgdata-util
¶
You can use the oe-pkgdata-util
command-line utility to query
PKGDATA_DIR
and display various package-related information.
When you use the utility, you must use it to view information
on packages that have already been built.
Following are a few of the available
oe-pkgdata-util
subcommands.
oe-pkgdata-util list-pkgs [
pattern
]
:
Lists all packages that have been built, optionally
limiting the match to packages that match
pattern
.
oe-pkgdata-util list-pkg-files
package
...
:
Lists the files and directories contained in the given
packages.
A different way to view the contents of a package is
to look at the
${
WORKDIR
}/packages-split
directory of the recipe that generates the
package.
This directory is created by the
do_package
task and has one subdirectory for each package the
recipe generates, which contains the files stored in
that package.
If you want to inspect the
${WORKDIR}/packages-split
directory, make sure that
rm_work
is not enabled when you build the recipe.
oe-pkgdata-util find-path
path
...
:
Lists the names of the packages that contain the given
paths.
For example, the following tells us that
/usr/share/man/man1/make.1
is contained in the make-doc
package:
$ oe-pkgdata-util find-path /usr/share/man/man1/make.1 make-doc: /usr/share/man/man1/make.1
oe-pkgdata-util lookup-recipe
package
...
:
Lists the name of the recipes that
produce the given packages.
For more information on the oe-pkgdata-util
command, use the help facility:
$ oe-pkgdata-util ‐‐help
$ oe-pkgdata-util subcommand
--help
Sometimes it can be hard to see why BitBake wants to build other recipes before the one you have specified. Dependency information can help you understand why a recipe is built.
To generate dependency information for a recipe, run the following command:
$ bitbake -g recipename
This command writes the following files in the current directory:
pn-buildlist
: A list of
recipes/targets involved in building
recipename
.
"Involved" here means that at least one task from the
recipe needs to run when building
recipename
from scratch.
Targets that are in
ASSUME_PROVIDED
are not listed.
task-depends.dot
: A graph showing
dependencies between tasks.
The graphs are in
DOT
format and can be converted to images (e.g. using the
dot
tool from
Graphviz).
DOT files use a plain text format.
The graphs generated using the
bitbake -g
command are often so
large as to be difficult to read without special
pruning (e.g. with Bitbake's
-I
option) and processing.
Despite the form and size of the graphs, the
corresponding .dot
files can
still be possible to read and provide useful
information.
As an example, the
task-depends.dot
file contains
lines such as the following:
"libxslt.do_configure" -> "libxml2.do_populate_sysroot"
The above example line reveals that the
do_configure
task in libxslt
depends on the
do_populate_sysroot
task in libxml2
, which is a
normal
DEPENDS
dependency between the two recipes.
For an example of how .dot
files can be processed, see the
scripts/contrib/graph-tool
Python script, which finds and displays paths
between graph nodes.
You can use a different method to view dependency information by using the following command:
$ bitbake -g -u taskexp recipename
This command displays a GUI window from which you can view
build-time and runtime dependencies for the recipes involved in
building recipename
.
As mentioned in the
"Checksums (Signatures)"
section of the BitBake User Manual, BitBake tries to
automatically determine what variables a task depends on so
that it can rerun the task if any values of the variables
change.
This determination is usually reliable.
However, if you do things like construct variable names at
runtime, then you might have to manually declare dependencies
on those variables using vardeps
as
described in the
"Variable Flags"
section of the BitBake User Manual.
If you are unsure whether a variable dependency is being picked up automatically for a given task, you can list the variable dependencies BitBake has determined by doing the following:
Build the recipe containing the task:
$ bitbake recipename
Inside the
STAMPS_DIR
directory, find the signature data
(sigdata
) file that corresponds
to the task.
The sigdata
files contain a pickled
Python database of all the metadata that went into
creating the input checksum for the task.
As an example, for the
do_fetch
task of the db
recipe, the
sigdata
file might be found in the
following location:
${BUILDDIR}/tmp/stamps/i586-poky-linux/db/6.0.30-r1.do_fetch.sigdata.7c048c18222b16ff0bcee2000ef648b1
For tasks that are accelerated through the shared state
(sstate)
cache, an additional siginfo
file
is written into
SSTATE_DIR
along with the cached task output.
The siginfo
files contain exactly
the same information as sigdata
files.
Run bitbake-dumpsig
on the
sigdata
or
siginfo
file.
Here is an example:
$ bitbake-dumpsig ${BUILDDIR}/tmp/stamps/i586-poky-linux/db/6.0.30-r1.do_fetch.sigdata.7c048c18222b16ff0bcee2000ef648b1
In the output of the above command, you will find a line like the following, which lists all the (inferred) variable dependencies for the task. This list also includes indirect dependencies from variables depending on other variables, recursively.
Task dependencies: ['PV', 'SRCREV', 'SRC_URI', 'SRC_URI[md5sum]', 'SRC_URI[sha256sum]', 'base_do_fetch']
base_do_fetch
)
also count as variable dependencies.
These functions in turn depend on the variables they
reference.
The output of bitbake-dumpsig
also
includes the value each variable had, a list of
dependencies for each variable, and
BB_HASHBASE_WHITELIST
information.
There is also a bitbake-diffsigs
command
for comparing two siginfo
or
sigdata
files.
This command can be helpful when trying to figure out what
changed between two versions of a task.
If you call bitbake-diffsigs
with just one
file, the command behaves like
bitbake-dumpsig
.
You can also use BitBake to dump out the signature construction information without executing tasks by using either of the following BitBake command-line options:
‐‐dump-signatures=SIGNATURE_HANDLER
-SSIGNATURE_HANDLER
SIGNATURE_HANDLER
are "none" and
"printdiff", which dump only the signature or compare the
dumped signature with the cached one, respectively.
Using BitBake with either of these options causes BitBake to
dump out sigdata
files in the
stamps
directory for every task it would
have executed instead of building the specified target package.
Seeing what metadata went into creating the input signature
of a shared state (sstate) task can be a useful debugging
aid.
This information is available in signature information
(siginfo
) files in
SSTATE_DIR
.
For information on how to view and interpret information in
siginfo
files, see the
"Viewing Task Variable Dependencies"
section.
For conceptual information on shared state, see the "Shared State" section in the Yocto Project Overview and Concepts Manual.
The OpenEmbedded build system uses checksums and shared state cache to avoid unnecessarily rebuilding tasks. Collectively, this scheme is known as "shared state code."
As with all schemes, this one has some drawbacks.
It is possible that you could make implicit changes to your
code that the checksum calculations do not take into
account.
These implicit changes affect a task's output but do not
trigger the shared state code into rebuilding a recipe.
Consider an example during which a tool changes its output.
Assume that the output of rpmdeps
changes.
The result of the change should be that all the
package
and
package_write_rpm
shared state cache
items become invalid.
However, because the change to the output is
external to the code and therefore implicit,
the associated shared state cache items do not become
invalidated.
In this case, the build process uses the cached items
rather than running the task again.
Obviously, these types of implicit changes can cause
problems.
To avoid these problems during the build, you need to understand the effects of any changes you make. Realize that changes you make directly to a function are automatically factored into the checksum calculation. Thus, these explicit changes invalidate the associated area of shared state cache. However, you need to be aware of any implicit changes that are not obvious changes to the code and could affect the output of a given task.
When you identify an implicit change, you can easily
take steps to invalidate the cache and force the tasks
to run.
The steps you can take are as simple as changing a
function's comments in the source code.
For example, to invalidate package shared state files,
change the comment statements of
do_package
or the comments of one of the functions it calls.
Even though the change is purely cosmetic, it causes the
checksum to be recalculated and forces the build system to
run the task again.
Any given recipe consists of a set of tasks.
The standard BitBake behavior in most cases is:
do_fetch
,
do_unpack
,
do_patch
,
do_configure
,
do_compile
,
do_install
,
do_package
,
do_package_write_*
, and
do_build
.
The default task is do_build
and any tasks
on which it depends build first.
Some tasks, such as do_devshell
, are not
part of the default build chain.
If you wish to run a task that is not part of the default build
chain, you can use the -c
option in
BitBake.
Here is an example:
$ bitbake matchbox-desktop -c devshell
The -c
option respects task dependencies,
which means that all other tasks (including tasks from other
recipes) that the specified task depends on will be run before
the task.
Even when you manually specify a task to run with
-c
, BitBake will only run the task if it
considers it "out of date".
See the
"Stamp Files and the Rerunning of Tasks"
section in the Yocto Project Overview and Concepts Manual for
how BitBake determines whether a task is "out of date".
If you want to force an up-to-date task to be rerun (e.g.
because you made manual modifications to the recipe's
WORKDIR
that you want to try out), then you can use the
-f
option.
-f
is never required when
running the
do_devshell
task is because the
[
nostamp
]
variable flag is already set for the task.
The following example shows one way you can use the
-f
option:
$ bitbake matchbox-desktop . . make some changes to the source code in the work directory . . $ bitbake matchbox-desktop -c compile -f $ bitbake matchbox-desktop
This sequence first builds and then recompiles
matchbox-desktop
.
The last command reruns all tasks (basically the packaging
tasks) after the compile.
BitBake recognizes that the do_compile
task was rerun and therefore understands that the other tasks
also need to be run again.
Another, shorter way to rerun a task and all
normal recipe build tasks
that depend on it is to use the -C
option.
-c
option, which is lower-cased.
Using this option invalidates the given task and then runs the
do_build
task, which is the default task if no task is given, and the
tasks on which it depends.
You could replace the final two commands in the previous example
with the following single command:
$ bitbake matchbox-desktop -C compile
Internally, the -f
and
-C
options work by tainting (modifying) the
input checksum of the specified task.
This tainting indirectly causes the task and its
dependent tasks to be rerun through the normal task dependency
mechanisms.
WARNING: /home/ulf/poky/meta/recipes-sato/matchbox-desktop/matchbox-desktop_2.1.bb.do_compile is tainted from a forced runThe purpose of the warning is to let you know that the work directory and build output might not be in the clean state they would be in for a "normal" build, depending on what actions you took. To get rid of such warnings, you can remove the work directory and rebuild the recipe, as follows:
$ bitbake matchbox-desktop -c clean $ bitbake matchbox-desktop
You can view a list of tasks in a given package by running the
do_listtasks
task as follows:
$ bitbake matchbox-desktop -c listtasks
The results appear as output to the console and are also in the
file ${WORKDIR}/temp/log.do_listtasks
.
You can see debug output from BitBake by using the
-D
option.
The debug output gives more information about what BitBake
is doing and the reason behind it.
Each -D
option you use increases the
logging level.
The most common usage is -DDD
.
The output from
bitbake -DDD -v
targetname
can reveal why BitBake chose a certain version of a package or
why BitBake picked a certain provider.
This command could also help you in a situation where you think
BitBake did something unexpected.
To build a specific recipe (.bb
file),
you can use the following command form:
$ bitbake -bsomepath
/somerecipe
.bb
This command form does not check for dependencies. Consequently, you should use it only when you know existing dependencies have been met.
The Yocto Project provides several logging functions for
producing debugging output and reporting errors and warnings.
For Python functions, the following logging functions exist.
All of these functions log to
${T}/log.do_
task
,
and can also log to standard output (stdout) with the right
settings:
bb.plain(
msg
)
:
Writes msg
as is to the
log while also logging to stdout.
bb.note(
msg
)
:
Writes "NOTE: msg
" to the
log.
Also logs to stdout if BitBake is called with "-v".
bb.debug(
level
,
msg
)
:
Writes "DEBUG: msg
" to the
log.
Also logs to stdout if the log level is greater than or
equal to level
.
See the
"-D"
option in the BitBake User Manual for more information.
bb.warn(
msg
)
:
Writes "WARNING: msg
" to the
log while also logging to stdout.
bb.error(
msg
)
:
Writes "ERROR: msg
" to the
log while also logging to standard out (stdout).
bb.fatal(
msg
)
:
This logging function is similar to
bb.error(
msg
)
but also causes the calling task to fail.
bb.fatal()
raises an exception,
which means you do not need to put a "return"
statement after the function.
The same logging functions are also available in shell
functions, under the names
bbplain
, bbnote
,
bbdebug
, bbwarn
,
bberror
, and bbfatal
.
The
logging
class implements these functions.
See that class in the
meta/classes
folder of the
Source Directory
for information.
When creating recipes using Python and inserting code that handles build logs, keep in mind the goal is to have informative logs while keeping the console as "silent" as possible. Also, if you want status messages in the log, use the "debug" loglevel.
Following is an example written in Python.
The code handles logging for a function that determines the
number of tasks needed to be run.
See the
"do_listtasks
"
section for additional information:
python do_listtasks() { bb.debug(2, "Starting to figure out the task list") if noteworthy_condition: bb.note("There are 47 tasks to run") bb.debug(2, "Got to point xyz") if warning_trigger: bb.warn("Detected warning_trigger, this might be a problem later.") if recoverable_error: bb.error("Hit recoverable_error, you really need to fix this!") if fatal_error: bb.fatal("fatal_error detected, unable to print the task list") bb.plain("The tasks present are abc") bb.debug(2, "Finished figuring out the tasklist") }
When creating recipes using Bash and inserting code that handles build logs, you have the same goals - informative with minimal console output. The syntax you use for recipes written in Bash is similar to that of recipes written in Python described in the previous section.
Following is an example written in Bash.
The code logs the progress of the do_my_function
function.
do_my_function() { bbdebug 2 "Running do_my_function" if [ exceptional_condition ]; then bbnote "Hit exceptional_condition" fi bbdebug 2 "Got to point xyz" if [ warning_trigger ]; then bbwarn "Detected warning_trigger, this might cause a problem later." fi if [ recoverable_error ]; then bberror "Hit recoverable_error, correcting" fi if [ fatal_error ]; then bbfatal "fatal_error detected" fi bbdebug 2 "Completed do_my_function" }
A parallel make
race occurs when the build
consists of several parts that are run simultaneously and
a situation occurs when the output or result of one
part is not ready for use with a different part of the build
that depends on that output.
Parallel make races are annoying and can sometimes be difficult
to reproduce and fix.
However, some simple tips and tricks exist that can help
you debug and fix them.
This section presents a real-world example of an error
encountered on the Yocto Project autobuilder and the process
used to fix it.
make
race
condition, you can work around it by clearing either the
PARALLEL_MAKE
or
PARALLEL_MAKEINST
variables.
For this example, assume that you are building an image that depends on the "neard" package. And, during the build, BitBake runs into problems and creates the following output.
If you examine the output or the log file, you see the
failure during make
:
| DEBUG: SITE files ['endian-little', 'bit-32', 'ix86-common', 'common-linux', 'common-glibc', 'i586-linux', 'common'] | DEBUG: Executing shell function do_compile | NOTE: make -j 16 | make --no-print-directory all-am | /bin/mkdir -p include/near | /bin/mkdir -p include/near | /bin/mkdir -p include/near | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/types.h include/near/types.h | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/log.h include/near/log.h | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/plugin.h include/near/plugin.h | /bin/mkdir -p include/near | /bin/mkdir -p include/near | /bin/mkdir -p include/near | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/tag.h include/near/tag.h | /bin/mkdir -p include/near | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/adapter.h include/near/adapter.h | /bin/mkdir -p include/near | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/ndef.h include/near/ndef.h | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/tlv.h include/near/tlv.h | /bin/mkdir -p include/near | /bin/mkdir -p include/near | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/setting.h include/near/setting.h | /bin/mkdir -p include/near | /bin/mkdir -p include/near | /bin/mkdir -p include/near | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/device.h include/near/device.h | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/nfc_copy.h include/near/nfc_copy.h | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/snep.h include/near/snep.h | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/version.h include/near/version.h | ln -s /home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/work/i586-poky-linux/neard/ 0.14-r0/neard-0.14/include/dbus.h include/near/dbus.h | ./src/genbuiltin nfctype1 nfctype2 nfctype3 nfctype4 p2p > src/builtin.h | i586-poky-linux-gcc -m32 -march=i586 --sysroot=/home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/ build/build/tmp/sysroots/qemux86 -DHAVE_CONFIG_H -I. -I./include -I./src -I./gdbus -I/home/pokybuild/ yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/sysroots/qemux86/usr/include/glib-2.0 -I/home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/tmp/sysroots/qemux86/usr/ lib/glib-2.0/include -I/home/pokybuild/yocto-autobuilder/yocto-slave/nightly-x86/build/build/ tmp/sysroots/qemux86/usr/include/dbus-1.0 -I/home/pokybuild/yocto-autobuilder/yocto-slave/ nightly-x86/build/build/tmp/sysroots/qemux86/usr/lib/dbus-1.0/include -I/home/pokybuild/yocto-autobuilder/ yocto-slave/nightly-x86/build/build/tmp/sysroots/qemux86/usr/include/libnl3 -DNEAR_PLUGIN_BUILTIN -DPLUGINDIR=\""/usr/lib/near/plugins"\" -DCONFIGDIR=\""/etc/neard\"" -O2 -pipe -g -feliminate-unused-debug-types -c -o tools/snep-send.o tools/snep-send.c | In file included from tools/snep-send.c:16:0: | tools/../src/near.h:41:23: fatal error: near/dbus.h: No such file or directory | #include <near/dbus.h> | ^ | compilation terminated. | make[1]: *** [tools/snep-send.o] Error 1 | make[1]: *** Waiting for unfinished jobs.... | make: *** [all] Error 2 | ERROR: oe_runmake failed
Because race conditions are intermittent, they do not manifest themselves every time you do the build. In fact, most times the build will complete without problems even though the potential race condition exists. Thus, once the error surfaces, you need a way to reproduce it.
In this example, compiling the "neard" package is causing
the problem.
So the first thing to do is build "neard" locally.
Before you start the build, set the
PARALLEL_MAKE
variable in your local.conf
file to
a high number (e.g. "-j 20").
Using a high value for PARALLEL_MAKE
increases the chances of the race condition showing up:
$ bitbake neard
Once the local build for "neard" completes, start a
devshell
build:
$ bitbake neard -c devshell
For information on how to use a
devshell
, see the
"Using a Development Shell"
section.
In the devshell
, do the following:
$ make clean $ make tools/snep-send.o
The devshell
commands cause the failure
to clearly be visible.
In this case, a missing dependency exists for the "neard"
Makefile target.
Here is some abbreviated, sample output with the
missing dependency clearly visible at the end:
i586-poky-linux-gcc -m32 -march=i586 --sysroot=/home/scott-lenovo/...... . . . tools/snep-send.c In file included from tools/snep-send.c:16:0: tools/../src/near.h:41:23: fatal error: near/dbus.h: No such file or directory #include <near/dbus.h> ^ compilation terminated. make: *** [tools/snep-send.o] Error 1 $
Because there is a missing dependency for the Makefile
target, you need to patch the
Makefile.am
file, which is generated
from Makefile.in
.
You can use Quilt to create the patch:
$ quilt new parallelmake.patch Patch patches/parallelmake.patch is now on top $ quilt add Makefile.am File Makefile.am added to patch patches/parallelmake.patch
For more information on using Quilt, see the "Using Quilt in Your Workflow" section.
At this point you need to make the edits to
Makefile.am
to add the missing
dependency.
For our example, you have to add the following line
to the file:
tools/snep-send.$(OBJEXT): include/near/dbus.h
Once you have edited the file, use the
refresh
command to create the patch:
$ quilt refresh Refreshed patch patches/parallelmake.patch
Once the patch file exists, you need to add it back to the
originating recipe folder.
Here is an example assuming a top-level
Source Directory
named poky
:
$ cp patches/parallelmake.patch poky/meta/recipes-connectivity/neard/neard
The final thing you need to do to implement the fix in the
build is to update the "neard" recipe (i.e.
neard-0.14.bb
) so that the
SRC_URI
statement includes the patch file.
The recipe file is in the folder above the patch.
Here is what the edited SRC_URI
statement would look like:
SRC_URI = "${KERNELORG_MIRROR}/linux/network/nfc/${BPN}-${PV}.tar.xz \ file://neard.in \ file://neard.service.in \ file://parallelmake.patch \ "
With the patch complete and moved to the correct folder and
the SRC_URI
statement updated, you can
exit the devshell
:
$ exit
With everything in place, you can get back to trying the build again locally:
$ bitbake neard
This build should succeed.
Now you can open up a devshell
again
and repeat the clean and make operations as follows:
$ bitbake neard -c devshell $ make clean $ make tools/snep-send.o
The build should work without issue.
As with all solved problems, if they originated upstream, you need to submit the fix for the recipe in OE-Core and upstream so that the problem is taken care of at its source. See the "Submitting a Change to the Yocto Project" section for more information.
GDB allows you to examine running programs, which in turn helps you to understand and fix problems. It also allows you to perform post-mortem style analysis of program crashes. GDB is available as a package within the Yocto Project and is installed in SDK images by default. See the "Images" chapter in the Yocto Project Reference Manual for a description of these images. You can find information on GDB at http://sourceware.org/gdb/.
-dbg
)
packages for the applications you are going to debug.
Doing so makes extra debug symbols available that give you
more meaningful output.
Sometimes, due to memory or disk space constraints, it is not possible to use GDB directly on the remote target to debug applications. These constraints arise because GDB needs to load the debugging information and the binaries of the process being debugged. Additionally, GDB needs to perform many computations to locate information such as function names, variable names and values, stack traces and so forth - even before starting the debugging process. These extra computations place more load on the target system and can alter the characteristics of the program being debugged.
To help get past the previously mentioned constraints, you can use gdbserver, which runs on the remote target and does not load any debugging information from the debugged process. Instead, a GDB instance processes the debugging information that is run on a remote computer - the host GDB. The host GDB then sends control commands to gdbserver to make it stop or start the debugged program, as well as read or write memory regions of that debugged program. All the debugging information loaded and processed as well as all the heavy debugging is done by the host GDB. Offloading these processes gives the gdbserver running on the target a chance to remain small and fast.
Because the host GDB is responsible for loading the debugging information and for doing the necessary processing to make actual debugging happen, you have to make sure the host can access the unstripped binaries complete with their debugging information and also be sure the target is compiled with no optimizations. The host GDB must also have local access to all the libraries used by the debugged program. Because gdbserver does not need any local debugging information, the binaries on the remote target can remain stripped. However, the binaries must also be compiled without optimization so they match the host's binaries.
To remain consistent with GDB documentation and terminology, the binary being debugged on the remote target machine is referred to as the "inferior" binary. For documentation on GDB see the GDB site.
The following steps show you how to debug using the GNU project debugger.
Configure your build system to construct the companion debug filesystem:
In your local.conf
file, set
the following:
IMAGE_GEN_DEBUGFS = "1" IMAGE_FSTYPES_DEBUGFS = "tar.bz2"
These options cause the OpenEmbedded build system
to generate a special companion filesystem fragment,
which contains the matching source and debug symbols to
your deployable filesystem.
The build system does this by looking at what is in the
deployed filesystem, and pulling the corresponding
-dbg
packages.
The companion debug filesystem is not a complete filesystem, but only contains the debug fragments. This filesystem must be combined with the full filesystem for debugging. Subsequent steps in this procedure show how to combine the partial filesystem with the full filesystem.
Configure the system to include gdbserver in the target filesystem:
Make the following addition in either your
local.conf
file or in an image
recipe:
IMAGE_INSTALL_append = “ gdbserver"
The change makes sure the gdbserver
package is included.
Build the environment:
Use the following command to construct the image and the companion Debug Filesystem:
$ bitbake image
Build the cross GDB component and make it available for debugging. Build the SDK that matches the image. Building the SDK is best for a production build that can be used later for debugging, especially during long term maintenance:
$ bitbake -c populate_sdk image
Alternatively, you can build the minimal toolchain components that match the target. Doing so creates a smaller than typical SDK and only contains a minimal set of components with which to build simple test applications, as well as run the debugger:
$ bitbake meta-toolchain
A final method is to build Gdb itself within the build system:
$ bitbake gdb-cross-architecture
Doing so produces a temporary copy of
cross-gdb
you can use for
debugging during development.
While this is the quickest approach, the two previous
methods in this step are better when considering
long-term maintenance strategies.
bitbake gdb-cross
, the
OpenEmbedded build system suggests the actual
image (e.g. gdb-cross-i586
).
The suggestion is usually the actual name you want
to use.
Set up the debugfs
Run the following commands to set up the
debugfs
:
$ mkdir debugfs $ cd debugfs $ tar xvfjbuild-dir
/tmp-glibc/deploy/images/machine
/image
.rootfs.tar.bz2 $ tar xvfjbuild-dir
/tmp-glibc/deploy/images/machine
/image
-dbg.rootfs.tar.bz2
Set up GDB
Install the SDK (if you built one) and then
source the correct environment file.
Sourcing the environment file puts the SDK in your
PATH
environment variable.
If you are using the build system, Gdb is
located in
build-dir
/tmp/sysroots/host
/usr/bin/architecture
/architecture
-gdb
Boot the target:
For information on how to run QEMU, see the QEMU Documentation.
Debug a program:
Debugging a program involves running gdbserver
on the target and then running Gdb on the host.
The example in this step debugs
gzip
:
root@qemux86:~# gdbserver localhost:1234 /bin/gzip —help
For additional gdbserver options, see the GDB Server Documentation.
After running gdbserver on the target, you need to run Gdb on the host and configure it and connect to the target. Use these commands:
$ cddirectory-holding-the-debugfs-directory
$arch
-gdb (gdb) set sysroot debugfs (gdb) set substitute-path /usr/src/debug debugfs/usr/src/debug (gdb) target remoteIP-of-target
:1234
At this point, everything should automatically load (i.e. matching binaries, symbols and headers).
set
commands in the
previous example can be placed into the users
~/.gdbinit
file.
Upon starting, Gdb automatically runs whatever
commands are in that file.
Deploying without a full image rebuild:
In many cases, during development you want a quick method to deploy a new binary to the target and debug it, without waiting for a full image build.
One approach to solving this situation is to
just build the component you want to debug.
Once you have built the component, copy the
executable directly to both the target and the
host debugfs
.
If the binary is processed through the debug
splitting in OpenEmbedded, you should also
copy the debug items (i.e. .debug
contents and corresponding
/usr/src/debug
files)
from the work directory.
Here is an example:
$ bitbake bash $ bitbake -c devshell bash $ cd .. $ scp packages-split/bash/bin/bashtarget
:/bin/bash $ cp -a packages-split/bash-dbg/*path
/debugfs
The previous section addressed using GDB remotely for debugging purposes, which is the most usual case due to the inherent hardware limitations on many embedded devices. However, debugging in the target hardware itself is also possible with more powerful devices. This section describes what you need to do in order to support using GDB to debug on the target hardware.
To support this kind of debugging, you need do the following:
Ensure that GDB is on the target.
You can do this by adding "gdb" to
IMAGE_INSTALL
:
IMAGE_INSTALL_append = " gdb"
Alternatively, you can add "tools-debug" to
IMAGE_FEATURES
:
IMAGE_FEATURES_append = " tools-debug"
Ensure that debug symbols are present.
You can make sure these symbols are present by
installing -dbg
:
IMAGE_INSTALL_append = " packagename
-dbg"
Alternatively, you can do the following to include all the debug symbols:
IMAGE_FEATURES_append = " dbg-pkgs"
local.conf
file, you will reduce
optimization from
FULL_OPTIMIZATION
of "-O2" to
DEBUG_OPTIMIZATION
of "-O -fno-omit-frame-pointer":
DEBUG_BUILD = "1"Consider that this will reduce the application's performance and is recommended only for debugging purposes.
Here are some other tips that you might find useful:
When adding new packages, it is worth watching for
undesirable items making their way into compiler command
lines.
For example, you do not want references to local system
files like
/usr/lib/
or
/usr/include/
.
If you want to remove the psplash
boot splashscreen,
add psplash=false
to the kernel
command line.
Doing so prevents psplash
from
loading and thus allows you to see the console.
It is also possible to switch out of the splashscreen by
switching the virtual console (e.g. Fn+Left or Fn+Right
on a Zaurus).
Removing
TMPDIR
(usually tmp/
, within the
Build Directory)
can often fix temporary build issues.
Removing TMPDIR
is usually a
relatively cheap operation, because task output will be
cached in
SSTATE_DIR
(usually sstate-cache/
, which is
also in the Build Directory).
TMPDIR
might be a
workaround rather than a fix.
Consequently, trying to determine the underlying
cause of an issue before removing the directory is
a good idea.
Understanding how a feature is used in practice within existing recipes can be very helpful. It is recommended that you configure some method that allows you to quickly search through files.
Using GNU Grep, you can use the following shell
function to recursively search through common
recipe-related files, skipping binary files,
.git
directories, and the
Build Directory (assuming its name starts with
"build"):
g() { grep -Ir \ --exclude-dir=.git \ --exclude-dir='build*' \ --include='*.bb*' \ --include='*.inc*' \ --include='*.conf*' \ --include='*.py*' \ "$@" }
Following are some usage examples:
$ g FOO # Search recursively for "FOO" $ g -i foo # Search recursively for "foo", ignoring case $ g -w FOO # Search recursively for "FOO" as a word, ignoring e.g. "FOOBAR"
If figuring out how some feature works requires a lot of searching, it might indicate that the documentation should be extended or improved. In such cases, consider filing a documentation bug using the Yocto Project implementation of Bugzilla. For information on how to submit a bug against the Yocto Project, see the Yocto Project Bugzilla wiki page and the "Submitting a Defect Against the Yocto Project" section.
.bbclass
file).
Because the Yocto Project is an open-source, community-based project, you can effect changes to the project. This section presents procedures that show you how to submit a defect against the project and how to submit a change.
Use the Yocto Project implementation of Bugzilla to submit a defect (bug) against the Yocto Project. For additional information on this implementation of Bugzilla see the "Yocto Project Bugzilla" section in the Yocto Project Reference Manual. For more detail on any of the following steps, see the Yocto Project Bugzilla wiki page.
Use the following general steps to submit a bug"
Open the Yocto Project implementation of Bugzilla.
Click "File a Bug" to enter a new bug.
Choose the appropriate "Classification", "Product", and
"Component" for which the bug was found.
Bugs for the Yocto Project fall into one of several
classifications, which in turn break down into several
products and components.
For example, for a bug against the
meta-intel
layer, you would choose
"Build System, Metadata & Runtime", "BSPs", and
"bsps-meta-intel", respectively.
Choose the "Version" of the Yocto Project for which you found the bug (e.g. 3.1.2).
Determine and select the "Severity" of the bug. The severity indicates how the bug impacted your work.
Choose the "Hardware" that the bug impacts.
Choose the "Architecture" that the bug impacts.
Choose a "Documentation change" item for the bug. Fixing a bug might or might not affect the Yocto Project documentation. If you are unsure of the impact to the documentation, select "Don't Know".
Provide a brief "Summary" of the bug. Try to limit your summary to just a line or two and be sure to capture the essence of the bug.
Provide a detailed "Description" of the bug. You should provide as much detail as you can about the context, behavior, output, and so forth that surrounds the bug. You can even attach supporting files for output from logs by using the "Add an attachment" button.
Click the "Submit Bug" button submit the bug. A new Bugzilla number is assigned to the bug and the defect is logged in the bug tracking system.
Once you file a bug, the bug is processed by the Yocto Project Bug Triage Team and further details concerning the bug are assigned (e.g. priority and owner). You are the "Submitter" of the bug and any further categorization, progress, or comments on the bug result in Bugzilla sending you an automated email concerning the particular change or progress to the bug.
Contributions to the Yocto Project and OpenEmbedded are very welcome. Because the system is extremely configurable and flexible, we recognize that developers will want to extend, configure or optimize it for their specific uses.
The Yocto Project uses a mailing list and a patch-based workflow
that is similar to the Linux kernel but contains important
differences.
In general, a mailing list exists through which you can submit
patches.
You should send patches to the appropriate mailing list so that they
can be reviewed and merged by the appropriate maintainer.
The specific mailing list you need to use depends on the
location of the code you are changing.
Each component (e.g. layer) should have a
README
file that indicates where to send
the changes and which process to follow.
You can send the patch to the mailing list using whichever approach you feel comfortable with to generate the patch. Once sent, the patch is usually reviewed by the community at large. If somebody has concerns with the patch, they will usually voice their concern over the mailing list. If a patch does not receive any negative reviews, the maintainer of the affected layer typically takes the patch, tests it, and then based on successful testing, merges the patch.
The "poky" repository, which is the Yocto Project's reference build environment, is a hybrid repository that contains several individual pieces (e.g. BitBake, Metadata, documentation, and so forth) built using the combo-layer tool. The upstream location used for submitting changes varies by component:
Core Metadata:
Send your patch to the
openembedded-core
mailing list. For example, a change to anything under
the meta
or
scripts
directories should be sent
to this mailing list.
BitBake:
For changes to BitBake (i.e. anything under the
bitbake
directory), send your patch
to the
bitbake-devel
mailing list.
"meta-*" trees: These trees contain Metadata. Use the poky mailing list.
For changes to other layers hosted in the Yocto Project source
repositories (i.e. yoctoproject.org
), tools,
and the Yocto Project documentation, use the
Yocto Project
general mailing list.
For additional recipes that do not fit into the core Metadata, you
should determine which layer the recipe should go into and submit
the change in the manner recommended by the documentation (e.g.
the README
file) supplied with the layer.
If in doubt, please ask on the Yocto general mailing list or on
the openembedded-devel mailing list.
You can also push a change upstream and request a maintainer to pull the change into the component's upstream repository. You do this by pushing to a contribution repository that is upstream. See the "Git Workflows and the Yocto Project" section in the Yocto Project Overview and Concepts Manual for additional concepts on working in the Yocto Project development environment.
Two commonly used testing repositories exist for OpenEmbedded-Core:
"ross/mut" branch:
The "mut" (master-under-test) tree
exists in the poky-contrib
repository
in the
Yocto Project source repositories.
"master-next" branch: This branch is part of the main "poky" repository in the Yocto Project source repositories.
Maintainers use these branches to test submissions prior to merging patches. Thus, you can get an idea of the status of a patch based on whether the patch has been merged into one of these branches.
The following sections provide procedures for submitting a change.
Follow this procedure to push a change to an upstream "contrib" Git repository:
Make Your Changes Locally: Make your changes in your local Git repository. You should make small, controlled, isolated changes. Keeping changes small and isolated aids review, makes merging/rebasing easier and keeps the change history clean should anyone need to refer to it in future.
Stage Your Changes:
Stage your changes by using the git add
command on each file you changed.
Commit Your Changes:
Commit the change by using the
git commit
command.
Make sure your commit information follows standards by
following these accepted conventions:
Be sure to include a "Signed-off-by:" line in the same style as required by the Linux kernel. Adding this line signifies that you, the submitter, have agreed to the Developer's Certificate of Origin 1.1 as follows:
Developer's Certificate of Origin 1.1 By making a contribution to this project, I certify that: (a) The contribution was created in whole or in part by me and I have the right to submit it under the open source license indicated in the file; or (b) The contribution is based upon previous work that, to the best of my knowledge, is covered under an appropriate open source license and I have the right under that license to submit that work with modifications, whether created in whole or in part by me, under the same open source license (unless I am permitted to submit under a different license), as indicated in the file; or (c) The contribution was provided directly to me by some other person who certified (a), (b) or (c) and I have not modified it. (d) I understand and agree that this project and the contribution are public and that a record of the contribution (including all personal information I submit with it, including my sign-off) is maintained indefinitely and may be redistributed consistent with this project or the open source license(s) involved.
Provide a single-line summary of the change. and, if more explanation is needed, provide more detail in the body of the commit. This summary is typically viewable in the "shortlist" of changes. Thus, providing something short and descriptive that gives the reader a summary of the change is useful when viewing a list of many commits. You should prefix this short description with the recipe name (if changing a recipe), or else with the short form path to the file being changed.
For the body of the commit message, provide detailed information that describes what you changed, why you made the change, and the approach you used. It might also be helpful if you mention how you tested the change. Provide as much detail as you can in the body of the commit message.
If the change addresses a specific bug or issue
that is associated with a bug-tracking ID,
include a reference to that ID in your detailed
description.
For example, the Yocto Project uses a specific
convention for bug references - any commit that
addresses a specific bug should use the following
form for the detailed description.
Be sure to use the actual bug-tracking ID from
Bugzilla for
bug-id
:
Fixes [YOCTO #bug-id
]detailed description of change
Push Your Commits to a "Contrib" Upstream: If you have arranged for permissions to push to an upstream contrib repository, push the change to that repository:
$ git pushupstream_remote_repo
local_branch_name
For example, suppose you have permissions to push into the
upstream meta-intel-contrib
repository and you are working in a local branch named
your_name
/README
.
The following command pushes your local commits to the
meta-intel-contrib
upstream
repository and puts the commit in a branch named
your_name
/README
:
$ git push meta-intel-contrib your_name
/README
Determine Who to Notify: Determine the maintainer or the mailing list that you need to notify for the change.
Before submitting any change, you need to be sure who the maintainer is or what mailing list that you need to notify. Use either these methods to find out:
Maintenance File:
Examine the maintainers.inc
file, which is located in the
Source Directory
at
meta/conf/distro/include
,
to see who is responsible for code.
Search by File: Using Git, you can enter the following command to bring up a short list of all commits against a specific file:
git shortlog -- filename
Just provide the name of the file for which you are interested. The information returned is not ordered by history but does include a list of everyone who has committed grouped by name. From the list, you can see who is responsible for the bulk of the changes against the file.
Examine the List of Mailing Lists: For a list of the Yocto Project and related mailing lists, see the "Mailing lists" section in the Yocto Project Reference Manual.
Make a Pull Request: Notify the maintainer or the mailing list that you have pushed a change by making a pull request.
The Yocto Project provides two scripts that
conveniently let you generate and send pull requests to the
Yocto Project.
These scripts are create-pull-request
and send-pull-request
.
You can find these scripts in the
scripts
directory within the
Source Directory
(e.g. ~/poky/scripts
).
Using these scripts correctly formats the requests without introducing any whitespace or HTML formatting. The maintainer that receives your patches either directly or through the mailing list needs to be able to save and apply them directly from your emails. Using these scripts is the preferred method for sending patches.
First, create the pull request. For example, the following command runs the script, specifies the upstream repository in the contrib directory into which you pushed the change, and provides a subject line in the created patch files:
$ ~/poky/scripts/create-pull-request -u meta-intel-contrib -s "Updated Manual Section Reference in README"
Running this script forms
*.patch
files in a folder named
pull-
PID
in the current directory.
One of the patch files is a cover letter.
Before running the
send-pull-request
script, you must
edit the cover letter patch to insert information about
your change.
After editing the cover letter, send the pull request.
For example, the following command runs the script and
specifies the patch directory and email address.
In this example, the email address is a mailing list:
$ ~/poky/scripts/send-pull-request -p ~/meta-intel/pull-10565 -t meta-intel@yoctoproject.org
You need to follow the prompts as the script is interactive.
-h
argument as follows:
$ poky/scripts/create-pull-request -h $ poky/scripts/send-pull-request -h
You can submit patches without using the
create-pull-request
and
send-pull-request
scripts described in the
previous section.
However, keep in mind, the preferred method is to use the scripts.
Depending on the components changed, you need to submit the email to a specific mailing list. For some guidance on which mailing list to use, see the list at the beginning of this section. For a description of all the available mailing lists, see the "Mailing Lists" section in the Yocto Project Reference Manual.
Here is the general procedure on how to submit a patch through email without using the scripts:
Make Your Changes Locally: Make your changes in your local Git repository. You should make small, controlled, isolated changes. Keeping changes small and isolated aids review, makes merging/rebasing easier and keeps the change history clean should anyone need to refer to it in future.
Stage Your Changes:
Stage your changes by using the git add
command on each file you changed.
Commit Your Changes:
Commit the change by using the
git commit --signoff
command.
Using the --signoff
option identifies
you as the person making the change and also satisfies
the Developer's Certificate of Origin (DCO) shown earlier.
When you form a commit, you must follow certain standards established by the Yocto Project development team. See Step 3 in the previous section for information on how to provide commit information that meets Yocto Project commit message standards.
Format the Commit:
Format the commit into an email message.
To format commits, use the
git format-patch
command.
When you provide the command, you must include a revision
list or a number of patches as part of the command.
For example, either of these two commands takes your most
recent single commit and formats it as an email message in
the current directory:
$ git format-patch -1
or
$ git format-patch HEAD~
After the command is run, the current directory
contains a numbered .patch
file for
the commit.
If you provide several commits as part of the
command, the git format-patch
command
produces a series of numbered files in the current
directory – one for each commit.
If you have more than one patch, you should also use the
--cover
option with the command,
which generates a cover letter as the first "patch" in
the series.
You can then edit the cover letter to provide a
description for the series of patches.
For information on the
git format-patch
command,
see GIT_FORMAT_PATCH(1)
displayed
using the man git-format-patch
command.
Import the Files Into Your Mail Client:
Import the files into your mail client by using the
git send-email
command.
git send-email
,
you must have the proper Git packages installed on
your host.
For Ubuntu, Debian, and Fedora the package is
git-email
.
The git send-email
command
sends email by using a local or remote Mail Transport Agent
(MTA) such as msmtp
,
sendmail
, or through a direct
smtp
configuration in your Git
~/.gitconfig
file.
If you are submitting patches through email only, it is
very important that you submit them without any whitespace
or HTML formatting that either you or your mailer
introduces.
The maintainer that receives your patches needs to be able
to save and apply them directly from your emails.
A good way to verify that what you are sending will be
applicable by the maintainer is to do a dry run and send
them to yourself and then save and apply them as the
maintainer would.
The git send-email
command is
the preferred method for sending your patches using
email since there is no risk of compromising whitespace
in the body of the message, which can occur when you use
your own mail client.
The command also has several options that let you
specify recipients and perform further editing of the
email message.
For information on how to use the
git send-email
command,
see GIT-SEND-EMAIL(1)
displayed using
the man git-send-email
command.
As mentioned in the "Licensing" section in the Yocto Project Overview and Concepts Manual, open source projects are open to the public and they consequently have different licensing structures in place. This section describes the mechanism by which the OpenEmbedded build system tracks changes to licensing text and covers how to maintain open source license compliance during your project's lifecycle. The section also describes how to enable commercially licensed recipes, which by default are disabled.
The license of an upstream project might change in the future.
In order to prevent these changes going unnoticed, the
LIC_FILES_CHKSUM
variable tracks changes to the license text. The checksums are
validated at the end of the configure step, and if the
checksums do not match, the build will fail.
LIC_FILES_CHKSUM
Variable¶
The LIC_FILES_CHKSUM
variable contains checksums of the license text in the
source code for the recipe.
Following is an example of how to specify
LIC_FILES_CHKSUM
:
LIC_FILES_CHKSUM = "file://COPYING;md5=xxxx \ file://licfile1.txt;beginline=5;endline=29;md5=yyyy \ file://licfile2.txt;endline=50;md5=zzzz \ ..."
When using "beginline" and "endline", realize
that line numbering begins with one and not
zero.
Also, the included lines are inclusive (i.e.
lines five through and including 29 in the
previous example for
licfile1.txt
).
When a license check fails, the selected license text is included as part of the QA message. Using this output, you can determine the exact start and finish for the needed license text.
The build system uses the
S
variable as the default directory when searching files
listed in LIC_FILES_CHKSUM
.
The previous example employs the default directory.
Consider this next example:
LIC_FILES_CHKSUM = "file://src/ls.c;beginline=5;endline=16;\ md5=bb14ed3c4cda583abc85401304b5cd4e" LIC_FILES_CHKSUM = "file://${WORKDIR}/license.html;md5=5c94767cedb5d6987c902ac850ded2c6"
The first line locates a file in
${S}/src/ls.c
and isolates lines five
through 16 as license text.
The second line refers to a file in
WORKDIR
.
Note that LIC_FILES_CHKSUM
variable is
mandatory for all recipes, unless the
LICENSE
variable is set to "CLOSED".
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.
It is possible to specify a checksum for an entire file,
or a specific section of a file (specified by beginning and
ending line numbers with the "beginline" and "endline"
parameters, respectively).
The latter is useful for source files with a license
notice header, README documents, and so forth.
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 with the last
line of the file.
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 mismatch 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 mismatch occurs during the build, the correct md5 checksum is placed in the build log and can be easily copied to the recipe.
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 allows you to just track the "COPYING"
file as long as it is kept up to date.
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.
If the whole file contains only license text, you do not need to use the "beginline" and "endline" parameters.
By default, the OpenEmbedded build system disables
components that have commercial or other special licensing
requirements.
Such requirements are defined on a
recipe-by-recipe basis through the
LICENSE_FLAGS
variable definition in the affected recipe.
For instance, the
poky/meta/recipes-multimedia/gstreamer/gst-plugins-ugly
recipe contains the following statement:
LICENSE_FLAGS = "commercial"
Here is a slightly more complicated example that contains both an explicit recipe name and version (after variable expansion):
LICENSE_FLAGS = "license_${PN}_${PV}"
In order for a component restricted by a
LICENSE_FLAGS
definition to be enabled and
included in an image, it needs to have a matching entry in the
global
LICENSE_FLAGS_WHITELIST
variable, which is a variable typically defined in your
local.conf
file.
For example, to enable the
poky/meta/recipes-multimedia/gstreamer/gst-plugins-ugly
package, you could add either the string
"commercial_gst-plugins-ugly" or the more general string
"commercial" to LICENSE_FLAGS_WHITELIST
.
See the
"License Flag Matching"
section for a full
explanation of how LICENSE_FLAGS
matching
works.
Here is the example:
LICENSE_FLAGS_WHITELIST = "commercial_gst-plugins-ugly"
Likewise, to additionally enable the package built from the
recipe containing
LICENSE_FLAGS = "license_${PN}_${PV}"
,
and assuming that the actual recipe name was
emgd_1.10.bb
, the following string would
enable that package as well as the original
gst-plugins-ugly
package:
LICENSE_FLAGS_WHITELIST = "commercial_gst-plugins-ugly license_emgd_1.10"
As a convenience, you do not need to specify the complete license string in the whitelist for every package. You can use an abbreviated form, which consists of just the first portion or portions of the license string before the initial underscore character or characters. A partial string will match any license that contains the given string as the first portion of its license. For example, the following whitelist string will also match both of the packages previously mentioned as well as any other packages that have licenses starting with "commercial" or "license".
LICENSE_FLAGS_WHITELIST = "commercial license"
License flag matching allows you to control what recipes
the OpenEmbedded build system includes in the build.
Fundamentally, the build system attempts to match
LICENSE_FLAGS
strings found in recipes
against LICENSE_FLAGS_WHITELIST
strings found in the whitelist.
A match causes the build system to include a recipe in the
build, while failure to find a match causes the build
system to exclude a recipe.
In general, license flag matching is simple. However, understanding some concepts will help you correctly and effectively use matching.
Before a flag
defined by a particular recipe is tested against the
contents of the whitelist, the expanded string
_${PN}
is appended to the flag.
This expansion makes each
LICENSE_FLAGS
value recipe-specific.
After expansion, the string is then matched against the
whitelist.
Thus, specifying
LICENSE_FLAGS = "commercial"
in recipe "foo", for example, results in the string
"commercial_foo"
.
And, to create a match, that string must appear in the
whitelist.
Judicious use of the LICENSE_FLAGS
strings and the contents of the
LICENSE_FLAGS_WHITELIST
variable
allows you a lot of flexibility for including or excluding
recipes based on licensing.
For example, you can broaden the matching capabilities by
using license flags string subsets in the whitelist.
usethispart_1.3
,
usethispart_1.4
, and so forth).
For example, simply specifying the string "commercial" in
the whitelist matches any expanded
LICENSE_FLAGS
definition that starts
with the string "commercial" such as "commercial_foo" and
"commercial_bar", which are the strings the build system
automatically generates for hypothetical recipes named
"foo" and "bar" assuming those recipes simply specify the
following:
LICENSE_FLAGS = "commercial"
Thus, you can choose to exhaustively enumerate each license flag in the whitelist and allow only specific recipes into the image, or you can use a string subset that causes a broader range of matches to allow a range of recipes into the image.
This scheme works even if the
LICENSE_FLAGS
string already
has _${PN}
appended.
For example, the build system turns the license flag
"commercial_1.2_foo" into "commercial_1.2_foo_foo" and
would match both the general "commercial" and the specific
"commercial_1.2_foo" strings found in the whitelist, as
expected.
Here are some other scenarios:
You can specify a versioned string in the recipe such as "commercial_foo_1.2" in a "foo" recipe. The build system expands this string to "commercial_foo_1.2_foo". Combine this license flag with a whitelist that has the string "commercial" and you match the flag along with any other flag that starts with the string "commercial".
Under the same circumstances, you can use "commercial_foo" in the whitelist and the build system not only matches "commercial_foo_1.2" but also matches any license flag with the string "commercial_foo", regardless of the version.
You can be very specific and use both the package and version parts in the whitelist (e.g. "commercial_foo_1.2") to specifically match a versioned recipe.
Other helpful variables related to commercial
license handling exist and are defined in the
poky/meta/conf/distro/include/default-distrovars.inc
file:
COMMERCIAL_AUDIO_PLUGINS ?= "" COMMERCIAL_VIDEO_PLUGINS ?= ""
If you want to enable these components, you can do so by
making sure you have statements similar to the following
in your local.conf
configuration file:
COMMERCIAL_AUDIO_PLUGINS = "gst-plugins-ugly-mad \ gst-plugins-ugly-mpegaudioparse" COMMERCIAL_VIDEO_PLUGINS = "gst-plugins-ugly-mpeg2dec \ gst-plugins-ugly-mpegstream gst-plugins-bad-mpegvideoparse" LICENSE_FLAGS_WHITELIST = "commercial_gst-plugins-ugly commercial_gst-plugins-bad commercial_qmmp"
Of course, you could also create a matching whitelist
for those components using the more general "commercial"
in the whitelist, but that would also enable all the
other packages with LICENSE_FLAGS
containing "commercial", which you may or may not want:
LICENSE_FLAGS_WHITELIST = "commercial"
Specifying audio and video plugins as part of the
COMMERCIAL_AUDIO_PLUGINS
and
COMMERCIAL_VIDEO_PLUGINS
statements
(along with the enabling
LICENSE_FLAGS_WHITELIST
) includes the
plugins or components into built images, thus adding
support for media formats or components.
One of the concerns for a development organization using open source software is how to maintain compliance with various open source licensing during the lifecycle of the product. While this section does not provide legal advice or comprehensively cover all scenarios, it does present methods that you can use to assist you in meeting the compliance requirements during a software release.
With hundreds of different open source licenses that the Yocto Project tracks, it is difficult to know the requirements of each and every license. However, the requirements of the major FLOSS licenses can begin to be covered by assuming that three main areas of concern exist:
Source code must be provided.
License text for the software must be provided.
Compilation scripts and modifications to the source code must be provided.
There are other requirements beyond the scope of these three and the methods described in this section (e.g. the mechanism through which source code is distributed).
As different organizations have different methods of complying with open source licensing, this section is not meant to imply that there is only one single way to meet your compliance obligations, but rather to describe one method of achieving compliance. The remainder of this section describes methods supported to meet the previously mentioned three requirements. Once you take steps to meet these requirements, and prior to releasing images, sources, and the build system, you should audit all artifacts to ensure completeness.
${DEPLOY_DIR}/licenses/image_name-datestamp
to assist with any audits.
Compliance activities should begin before you generate the final image. The first thing you should look at is the requirement that tops the list for most compliance groups - providing the source. The Yocto Project has a few ways of meeting this requirement.
One of the easiest ways to meet this requirement is
to provide the entire
DL_DIR
used by the build.
This method, however, has a few issues.
The most obvious is the size of the directory since it includes
all sources used in the build and not just the source used in
the released image.
It will include toolchain source, and other artifacts, which
you would not generally release.
However, the more serious issue for most companies is accidental
release of proprietary software.
The Yocto Project provides an
archiver
class to help avoid some of these concerns.
Before you employ DL_DIR
or the
archiver
class, you need to decide how
you choose to provide source.
The source archiver
class can generate
tarballs and SRPMs and can create them with various levels of
compliance in mind.
One way of doing this (but certainly not the only way) is to
release just the source as a tarball.
You can do this by adding the following to the
local.conf
file found in the
Build Directory:
INHERIT += "archiver" ARCHIVER_MODE[src] = "original"
During the creation of your image, the source from all
recipes that deploy packages to the image is placed within
subdirectories of
DEPLOY_DIR/sources
based on the
LICENSE
for each recipe.
Releasing the entire directory enables you to comply with
requirements concerning providing the unmodified source.
It is important to note that the size of the directory can
get large.
A way to help mitigate the size issue is to only release tarballs for licenses that require the release of source. Let us assume you are only concerned with GPL code as identified by running the following script:
# Script to archive a subset of packages matching specific license(s) # Source and license files are copied into sub folders of package folder # Must be run from build folder #!/bin/bash src_release_dir="source-release" mkdir -p $src_release_dir for a in tmp/deploy/sources/*; do for d in $a/*; do # Get package name from path p=`basename $d` p=${p%-*} p=${p%-*} # Only archive GPL packages (update *GPL* regex for your license check) numfiles=`ls tmp/deploy/licenses/$p/*GPL* 2> /dev/null | wc -l` if [ $numfiles -gt 1 ]; then echo Archiving $p mkdir -p $src_release_dir/$p/source cp $d/* $src_release_dir/$p/source 2> /dev/null mkdir -p $src_release_dir/$p/license cp tmp/deploy/licenses/$p/* $src_release_dir/$p/license 2> /dev/null fi done done
At this point, you could create a tarball from the
gpl_source_release
directory and
provide that to the end user.
This method would be a step toward achieving compliance
with section 3a of GPLv2 and with section 6 of GPLv3.
One requirement that is often overlooked is inclusion
of license text.
This requirement also needs to be dealt with prior to
generating the final image.
Some licenses require the license text to accompany
the binary.
You can achieve this by adding the following to your
local.conf
file:
COPY_LIC_MANIFEST = "1" COPY_LIC_DIRS = "1" LICENSE_CREATE_PACKAGE = "1"
Adding these statements to the configuration file ensures that the licenses collected during package generation are included on your image.
Setting all three variables to "1" results in the
image having two copies of the same license file.
One copy resides in
/usr/share/common-licenses
and
the other resides in
/usr/share/license
.
The reason for this behavior is because
COPY_LIC_DIRS
and
COPY_LIC_MANIFEST
add a copy of the license when the image is built but do
not offer a path for adding licenses for newly installed
packages to an image.
LICENSE_CREATE_PACKAGE
adds a separate package and an upgrade path for adding
licenses to an image.
As the source archiver
class has already
archived the original
unmodified source that contains the license files,
you would have already met the requirements for inclusion
of the license information with source as defined by the GPL
and other open source licenses.
At this point, we have addressed all we need to prior to generating the image. The next two requirements are addressed during the final packaging of the release.
By releasing the version of the OpenEmbedded build system and the layers used during the build, you will be providing both compilation scripts and the source code modifications in one step.
If the deployment team has a BSP layer and a distro layer, and those those layers are used to patch, compile, package, or modify (in any way) any open source software included in your released images, you might be required to release those layers under section 3 of GPLv2 or section 1 of GPLv3. One way of doing that is with a clean checkout of the version of the Yocto Project and layers used during your build. Here is an example:
# We built using the dunfell branch of the poky repo $ git clone -b dunfell git://git.yoctoproject.org/poky $ cd poky # We built using the release_branch for our layers $ git clone -b release_branch git://git.mycompany.com/meta-my-bsp-layer $ git clone -b release_branch git://git.mycompany.com/meta-my-software-layer # clean up the .git repos $ find . -name ".git" -type d -exec rm -rf {} \;
One thing a development organization might want to consider
for end-user convenience is to modify
meta-poky/conf/bblayers.conf.sample
to
ensure that when the end user utilizes the released build
system to build an image, the development organization's
layers are included in the bblayers.conf
file automatically:
# POKY_BBLAYERS_CONF_VERSION is increased each time build/conf/bblayers.conf # changes incompatibly POKY_BBLAYERS_CONF_VERSION = "2" BBPATH = "${TOPDIR}" BBFILES ?= "" BBLAYERS ?= " \ ##OEROOT##/meta \ ##OEROOT##/meta-poky \ ##OEROOT##/meta-yocto-bsp \ ##OEROOT##/meta-mylayer \ "
Creating and providing an archive of the Metadata layers (recipes, configuration files, and so forth) enables you to meet your requirements to include the scripts to control compilation as well as any modifications to the original source.
Some packages, such as the linux-firmware package, have many
licenses that are not in any way common.
You can avoid adding a lot of these types of common license
files, which are only applicable to a specific package, by using
the
NO_GENERIC_LICENSE
variable.
Using this variable also avoids QA errors when you use a
non-common, non-CLOSED license in a recipe.
The following is an example that uses the
LICENSE.Abilis.txt
file as the license from the fetched source:
NO_GENERIC_LICENSE[Firmware-Abilis] = "LICENSE.Abilis.txt"
The error reporting tool allows you to
submit errors encountered during builds to a central database.
Outside of the build environment, you can use a web interface to
browse errors, view statistics, and query for errors.
The tool works using a client-server system where the client
portion is integrated with the installed Yocto Project
Source Directory
(e.g. poky
).
The server receives the information collected and saves it in a
database.
A live instance of the error reporting server exists at http://errors.yoctoproject.org. This server exists so that when you want to get help with build failures, you can submit all of the information on the failure easily and then point to the URL in your bug report or send an email to the mailing list.
By default, the error reporting tool is disabled.
You can enable it by inheriting the
report-error
class by adding the following statement to the end of
your local.conf
file in your
Build Directory.
INHERIT += "report-error"
By default, the error reporting feature stores information in
${
LOG_DIR
}/error-report
.
However, you can specify a directory to use by adding the following
to your local.conf
file:
ERR_REPORT_DIR = "path"
Enabling error reporting causes the build process to collect the errors and store them in a file as previously described. When the build system encounters an error, it includes a command as part of the console output. You can run the command to send the error file to the server. For example, the following command sends the errors to an upstream server:
$ send-error-report /home/brandusa/project/poky/build/tmp/log/error-report/error_report_201403141617.txt
In the previous example, the errors are sent to a public database available at http://errors.yoctoproject.org, which is used by the entire community. If you specify a particular server, you can send the errors to a different database. Use the following command for more information on available options:
$ send-error-report --help
When sending the error file, you are prompted to review the data being sent as well as to provide a name and optional email address. Once you satisfy these prompts, the command returns a link from the server that corresponds to your entry in the database. For example, here is a typical link:
http://errors.yoctoproject.org/Errors/Details/9522/
Following the link takes you to a web interface where you can browse, query the errors, and view statistics.
To disable the error reporting feature, simply remove or comment
out the following statement from the end of your
local.conf
file in your
Build Directory.
INHERIT += "report-error"
If you want to set up your own error reporting server, you can obtain the code from the Git repository at http://git.yoctoproject.org/cgit/cgit.cgi/error-report-web/. Instructions on how to set it up are in the README document.
Wayland is a computer display server protocol that provides a method for compositing window managers to communicate directly with applications and video hardware and expects them to communicate with input hardware using other libraries. Using Wayland with supporting targets can result in better control over graphics frame rendering than an application might otherwise achieve.
The Yocto Project provides the Wayland protocol libraries and the
reference
Weston
compositor as part of its release.
You can find the integrated packages in the
meta
layer of the
Source Directory.
Specifically, you can find the recipes that build both Wayland
and Weston at meta/recipes-graphics/wayland
.
You can build both the Wayland and Weston packages for use only with targets that accept the Mesa 3D and Direct Rendering Infrastructure, which is also known as Mesa DRI. This implies that you cannot build and use the packages if your target uses, for example, the Intel® Embedded Media and Graphics Driver (Intel® EMGD) that overrides Mesa DRI.
This section describes what you need to do to implement Wayland and use the Weston compositor when building an image for a supporting target.
To enable Wayland, you need to enable it to be built and enable it to be included (installed) in the image.
To cause Mesa to build the wayland-egl
platform and Weston to build Wayland with Kernel Mode
Setting
(KMS)
support, include the "wayland" flag in the
DISTRO_FEATURES
statement in your local.conf
file:
DISTRO_FEATURES_append = " wayland"
To install the Wayland feature into an image, you must
include the following
CORE_IMAGE_EXTRA_INSTALL
statement in your local.conf
file:
CORE_IMAGE_EXTRA_INSTALL += "wayland weston"
To run Weston inside X11, enabling it as described earlier and building a Sato image is sufficient. If you are running your image under Sato, a Weston Launcher appears in the "Utility" category.
Alternatively, you can run Weston through the command-line interpretor (CLI), which is better suited for development work. To run Weston under the CLI, you need to do the following after your image is built:
Run these commands to export
XDG_RUNTIME_DIR
:
mkdir -p /tmp/$USER-weston chmod 0700 /tmp/$USER-weston export XDG_RUNTIME_DIR=/tmp/$USER-weston
Launch Weston in the shell:
weston
The Yocto Project uses an implementation of the Quick EMUlator (QEMU) Open Source project as part of the Yocto Project development "tool set". This chapter provides both procedures that show you how to use the Quick EMUlator (QEMU) and other QEMU information helpful for development purposes.
Within the context of the Yocto Project, QEMU is an emulator and virtualization machine that allows you to run a complete image you have built using the Yocto Project as just another task on your build system. QEMU is useful for running and testing images and applications on supported Yocto Project architectures without having actual hardware. Among other things, the Yocto Project uses QEMU to run automated Quality Assurance (QA) tests on final images shipped with each release.
This section provides a brief reference for the Yocto Project implementation of QEMU.
For official information and documentation on QEMU in general, see the following references:
QEMU Website: The official website for the QEMU Open Source project.
Documentation: The QEMU user manual.
To use QEMU, you need to have QEMU installed and initialized as well as have the proper artifacts (i.e. image files and root filesystems) available. Follow these general steps to run QEMU:
Install QEMU: QEMU is made available with the Yocto Project a number of ways. One method is to install a Software Development Kit (SDK). See "The QEMU Emulator" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual for information on how to install QEMU.
Setting Up the Environment: How you set up the QEMU environment depends on how you installed QEMU:
If you cloned the poky
repository or you downloaded and unpacked a
Yocto Project release tarball, you can source
the build environment script (i.e.
oe-init-build-env
):
$ cd ~/poky $ source oe-init-build-env
If you installed a cross-toolchain, you can
run the script that initializes the toolchain.
For example, the following commands run the
initialization script from the default
poky_sdk
directory:
. ~/poky_sdk/environment-setup-core2-64-poky-linux
Ensure the Artifacts are in Place: You need to be sure you have a pre-built kernel that will boot in QEMU. You also need the target root filesystem for your target machine’s architecture:
If you have previously built an image for QEMU
(e.g. qemux86
,
qemuarm
, and so forth),
then the artifacts are in place in your
Build Directory.
If you have not built an image, you can go to the machines/qemu area and download a pre-built image that matches your architecture and can be run on QEMU.
See the "Extracting the Root Filesystem" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual for information on how to extract a root filesystem.
Run QEMU:
The basic runqemu
command syntax is as
follows:
$ runqemu [option
] [...]
Based on what you provide on the command line,
runqemu
does a good job of figuring
out what you are trying to do.
For example, by default, QEMU looks for the most recently
built image according to the timestamp when it needs to
look for an image.
Minimally, through the use of options, you must provide
either a machine name, a virtual machine image
(*wic.vmdk
), or a kernel image
(*.bin
).
Here are some additional examples to help illustrate further QEMU:
This example starts QEMU with
MACHINE
set to "qemux86-64".
Assuming a standard
Build Directory,
runqemu
automatically finds the
bzImage-qemux86-64.bin
image file and
the
core-image-minimal-qemux86-64-20200218002850.rootfs.ext4
(assuming the current build created a
core-image-minimal
image).
$ runqemu qemux86-64
This example produces the exact same results as the previous example. This command, however, specifically provides the image and root filesystem type.
$ runqemu qemux86-64 core-image-minimal ext4
This example specifies to boot an initial RAM disk image
and to enable audio in QEMU.
For this case, runqemu
set the
internal variable FSTYPE
to
"cpio.gz".
Also, for audio to be enabled, an appropriate driver must
be installed (see the previous description for the
audio
option for more information).
$ runqemu qemux86-64 ramfs audio
This example does not provide enough information for
QEMU to launch.
While the command does provide a root filesystem type, it
must also minimally provide a
MACHINE
,
KERNEL
, or
VM
option.
$ runqemu ext4
This example specifies to boot a virtual machine
image (.wic.vmdk
file).
From the .wic.vmdk
,
runqemu
determines the QEMU
architecture (MACHINE
) to be
"qemux86-64" and the root filesystem type to be "vmdk".
$ runqemu /home/scott-lenovo/vm/core-image-minimal-qemux86-64.wic.vmdk
When booting or running QEMU, you can switch between
supported consoles by using
Ctrl+Alt+number
.
For example, Ctrl+Alt+3 switches you to the serial console
as long as that console is enabled.
Being able to switch consoles is helpful, for example, if
the main QEMU console breaks for some reason.
You can remove the splash screen when QEMU is booting by using Alt+left. Removing the splash screen allows you to see what is happening in the background.
The default QEMU integration captures the cursor within the main window. It does this since standard mouse devices only provide relative input and not absolute coordinates. You then have to break out of the grab using the "Ctrl+Alt" key combination. However, the Yocto Project's integration of QEMU enables the wacom USB touch pad driver by default to allow input of absolute coordinates. This default means that the mouse can enter and leave the main window without the grab taking effect leading to a better user experience.
One method for running QEMU is to run it on an NFS server. This is useful when you need to access the same file system from both the build and the emulated system at the same time. It is also worth noting that the system does not need root privileges to run. It uses a user space NFS server to avoid that. Follow these steps to set up for running QEMU using an NFS server.
Extract a Root Filesystem:
Once you are able to run QEMU in your environment, you can
use the runqemu-extract-sdk
script,
which is located in the scripts
directory along with the runqemu
script.
The runqemu-extract-sdk
takes a
root filesystem tarball and extracts it into a location
that you specify.
Here is an example that takes a file system and
extracts it to a directory named
test-nfs
:
runqemu-extract-sdk ./tmp/deploy/images/qemux86-64/core-image-sato-qemux86-64.tar.bz2 test-nfs
Start QEMU:
Once you have extracted the file system, you can run
runqemu
normally with the additional
location of the file system.
You can then also make changes to the files within
./test-nfs
and see those changes
appear in the image in real time.
Here is an example using the qemux86
image:
runqemu qemux86-64 ./test-nfs
Should you need to start, stop, or restart the NFS share, you can use the following commands:
The following command starts the NFS share:
runqemu-export-rootfs start file-system-location
The following command stops the NFS share:
runqemu-export-rootfs stop file-system-location
The following command restarts the NFS share:
runqemu-export-rootfs restart file-system-location
By default, the QEMU build compiles for and targets 64-bit and x86 Intel® Core™2 Duo processors and 32-bit x86 Intel® Pentium® II processors. QEMU builds for and targets these CPU types because they display a broad range of CPU feature compatibility with many commonly used CPUs.
Despite this broad range of compatibility, the CPUs could support
a feature that your host CPU does not support.
Although this situation is not a problem when QEMU uses software
emulation of the feature, it can be a problem when QEMU is
running with KVM enabled.
Specifically, software compiled with a certain CPU feature crashes
when run on a CPU under KVM that does not support that feature.
To work around this problem, you can override QEMU's runtime CPU
setting by changing the QB_CPU_KVM
variable in qemuboot.conf
in the
Build Directory's
deploy/image
directory.
This setting specifies a -cpu
option
passed into QEMU in the runqemu
script.
Running qemu -cpu help
returns a list of
available supported CPU types.
Using QEMU to emulate your hardware can result in speed issues
depending on the target and host architecture mix.
For example, using the qemux86
image in the
emulator on an Intel-based 32-bit (x86) host machine is fast
because the target and host architectures match.
On the other hand, using the qemuarm
image
on the same Intel-based host can be slower.
But, you still achieve faithful emulation of ARM-specific issues.
To speed things up, the QEMU images support using
distcc
to call a cross-compiler outside the
emulated system.
If you used runqemu
to start QEMU, and the
distccd
application is present on the host
system, any BitBake cross-compiling toolchain available from the
build system is automatically used from within QEMU simply by
calling distcc
.
You can accomplish this by defining the cross-compiler variable
(e.g. export CC="distcc"
).
Alternatively, if you are using a suitable SDK image or the
appropriate stand-alone toolchain is present, the toolchain is
also automatically used.
QEMU provides a framebuffer interface that makes standard consoles available.
Generally, headless embedded devices have a serial port. If so, you can configure the operating system of the running image to use that port to run a console. The connection uses standard IP networking.
SSH servers exist in some QEMU images.
The core-image-sato
QEMU image
has a Dropbear secure shell (SSH) server that runs
with the root password disabled.
The core-image-full-cmdline
and
core-image-lsb
QEMU images
have OpenSSH instead of Dropbear.
Including these SSH servers allow you to use standard
ssh
and scp
commands.
The core-image-minimal
QEMU image,
however, contains no SSH server.
You can use a provided, user-space NFS server to boot
the QEMU session using a local copy of the root
filesystem on the host.
In order to make this connection, you must extract a
root filesystem tarball by using the
runqemu-extract-sdk
command.
After running the command, you must then point the
runqemu
script to the extracted directory instead of a root
filesystem image file.
See the
"Running Under a Network File System (NFS) Server"
section for more information.
The basic runqemu
command syntax is as
follows:
$ runqemu [option
] [...]
Based on what you provide on the command line,
runqemu
does a good job of figuring out what
you are trying to do.
For example, by default, QEMU looks for the most recently built
image according to the timestamp when it needs to look for an
image.
Minimally, through the use of options, you must provide either
a machine name, a virtual machine image
(*wic.vmdk
), or a kernel image
(*.bin
).
Following is the command-line help output for the
runqemu
command:
$ runqemu --help Usage: you can run this script with any valid combination of the following environment variables (in any order): KERNEL - the kernel image file to use ROOTFS - the rootfs image file or nfsroot directory to use MACHINE - the machine name (optional, autodetected from KERNEL filename if unspecified) Simplified QEMU command-line options can be passed with: nographic - disable video console serial - enable a serial console on /dev/ttyS0 slirp - enable user networking, no root privileges is required kvm - enable KVM when running x86/x86_64 (VT-capable CPU required) kvm-vhost - enable KVM with vhost when running x86/x86_64 (VT-capable CPU required) publicvnc - enable a VNC server open to all hosts audio - enable audio [*/]ovmf* - OVMF firmware file or base name for booting with UEFI tcpserial=<port> - specify tcp serial port number biosdir=<dir> - specify custom bios dir biosfilename=<filename> - specify bios filename qemuparams=<xyz> - specify custom parameters to QEMU bootparams=<xyz> - specify custom kernel parameters during boot help, -h, --help: print this text Examples: runqemu runqemu qemuarm runqemu tmp/deploy/images/qemuarm runqemu tmp/deploy/images/qemux86/<qemuboot.conf> runqemu qemux86-64 core-image-sato ext4 runqemu qemux86-64 wic-image-minimal wic runqemu path/to/bzImage-qemux86.bin path/to/nfsrootdir/ serial runqemu qemux86 iso/hddimg/wic.vmdk/wic.qcow2/wic.vdi/ramfs/cpio.gz... runqemu qemux86 qemuparams="-m 256" runqemu qemux86 bootparams="psplash=false" runqemu path/to/<image>-<machine>.wic runqemu path/to/<image>-<machine>.wic.vmdk
runqemu
Command-Line Options¶
Following is a description of runqemu
options you can provide on the command line:
runqemu
provides appropriate error
messaging to help you correct the problem.
QEMUARCH
:
The QEMU machine architecture, which must be "qemuarm",
"qemuarm64", "qemumips", "qemumips64", "qemuppc",
"qemux86", or "qemux86-64".
:
The virtual machine image, which must be a
VM
.wic.vmdk
file.
Use this option when you want to boot a
.wic.vmdk
image.
The image filename you provide must contain one of the
following strings: "qemux86-64", "qemux86", "qemuarm",
"qemumips64", "qemumips", "qemuppc", or "qemush4".
ROOTFS
:
A root filesystem that has one of the following
filetype extensions: "ext2", "ext3", "ext4", "jffs2",
"nfs", or "btrfs".
If the filename you provide for this option uses “nfs”, it
must provide an explicit root filesystem path.
KERNEL
:
A kernel image, which is a .bin
file.
When you provide a .bin
file,
runqemu
detects it and assumes the
file is a kernel image.
MACHINE
:
The architecture of the QEMU machine, which must be one
of the following: "qemux86", "qemux86-64", "qemuarm",
"qemuarm64", "qemumips", “qemumips64", or "qemuppc".
The MACHINE
and
QEMUARCH
options are basically
identical.
If you do not provide a MACHINE
option, runqemu
tries to determine
it based on other options.
ramfs
:
Indicates you are booting an initial RAM disk (initramfs)
image, which means the FSTYPE
is
cpio.gz
.
iso
:
Indicates you are booting an ISO image, which means the
FSTYPE
is
.iso
.
nographic
:
Disables the video console, which sets the console to
"ttys0".
This option is useful when you have logged into a server
and you do not want to disable forwarding from the
X Window System (X11) to your workstation or laptop.
serial
:
Enables a serial console on
/dev/ttyS0
.
biosdir
:
Establishes a custom directory for BIOS, VGA BIOS and
keymaps.
biosfilename
:
Establishes a custom BIOS name.
qemuparams=\"
:
Specifies custom QEMU parameters.
Use this option to pass options other than the simple
"kvm" and "serial" options.
xyz
\"
bootparams=\"
:
Specifies custom boot parameters for the kernel.
xyz
\"
audio
:
Enables audio in QEMU.
The MACHINE
option must be
either "qemux86" or "qemux86-64" in order for audio to be
enabled.
Additionally, the snd_intel8x0
or snd_ens1370
driver must be
installed in linux guest.
slirp
:
Enables "slirp" networking, which is a different way
of networking that does not need root access
but also is not as easy to use or comprehensive
as the default.
kvm
:
Enables KVM when running "qemux86" or "qemux86-64"
QEMU architectures.
For KVM to work, all the following conditions must be met:
Your MACHINE
must be either
qemux86" or "qemux86-64".
Your build host has to have the KVM modules
installed, which are
/dev/kvm
.
The build host /dev/kvm
directory has to be both writable and readable.
kvm-vhost
:
Enables KVM with VHOST support when running "qemux86"
or "qemux86-64" QEMU architectures.
For KVM with VHOST to work, the following conditions must
be met:
kvm option conditions must be met.
Your build host has to have virtio net device, which
are /dev/vhost-net
.
The build host /dev/vhost-net
directory has to be either readable or writable
and “slirp-enabled”.
publicvnc
:
Enables a VNC server open to all hosts.
Welcome to the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual. This manual provides information that explains how to use both the Yocto Project extensible and standard SDKs to develop applications and images.
All SDKs consist of the following:
Cross-Development Toolchain: This toolchain contains a compiler, debugger, and various miscellaneous tools.
Libraries, Headers, and Symbols: The libraries, headers, and symbols are specific to the image (i.e. they match the image).
Environment Setup Script:
This *.sh
file, once run, sets up the
cross-development environment by defining variables and
preparing for SDK use.
Additionally, an extensible SDK has tools that allow you to easily add new applications and libraries to an image, modify the source of an existing component, test changes on the target hardware, and easily integrate an application into the OpenEmbedded build system.
You can use an SDK to independently develop and test code
that is destined to run on some target machine.
SDKs are completely self-contained.
The binaries are linked against their own copy of
libc
, which results in no dependencies
on the target system.
To achieve this, the pointer to the dynamic loader is
configured at install time since that path cannot be dynamically
altered.
This is the reason for a wrapper around the
populate_sdk
and
populate_sdk_ext
archives.
Another feature for the SDKs is that only one set of cross-compiler
toolchain binaries are produced for any given architecture.
This feature takes advantage of the fact that the target hardware can
be passed to gcc
as a set of compiler options.
Those options are set up by the environment script and contained in
variables such as
CC
and
LD
.
This reduces the space needed for the tools.
Understand, however, that every target still needs a sysroot because
those binaries are target-specific.
The SDK development environment consists of the following:
The self-contained SDK, which is an
architecture-specific cross-toolchain and
matching sysroots (target and native) all built by the
OpenEmbedded build system (e.g. the SDK).
The toolchain and sysroots are based on a
Metadata
configuration and extensions,
which allows you to cross-develop on the host machine for the
target hardware.
Additionally, the extensible SDK contains the
devtool
functionality.
The Quick EMUlator (QEMU), which lets you simulate target hardware. QEMU is not literally part of the SDK. You must build and include this emulator separately. However, QEMU plays an important role in the development process that revolves around use of the SDK.
In summary, the extensible and standard SDK share many features. However, the extensible SDK has powerful development tools to help you more quickly develop applications. Following is a table that summarizes the primary differences between the standard and extensible SDK types when considering which to build:
Feature | Standard SDK | Extensible SDK |
---|---|---|
Toolchain | Yes | Yes* |
Debugger | Yes | Yes* |
Size | 100+ MBytes | 1+ GBytes (or 300+ MBytes for minimal w/toolchain) |
devtool | No | Yes |
Build Images | No | Yes |
Updateable | No | Yes |
Managed Sysroot** | No | Yes |
Installed Packages | No*** | Yes**** |
Construction | Packages | Shared State |
* Extensible SDK contains the toolchain and debugger ifSDK_EXT_TYPE
is "full" orSDK_INCLUDE_TOOLCHAIN
is "1", which is the default. ** Sysroot is managed through the use ofdevtool
. Thus, it is less likely that you will corrupt your SDK sysroot when you try to add additional libraries. *** You can add runtime package management to the standard SDK but it is not supported by default. **** You must build and make the shared state available to extensible SDK users for "packages" you want to enable users to install.
The
Cross-Development Toolchain
consists of a cross-compiler, cross-linker, and cross-debugger
that are used to develop user-space applications for targeted
hardware.
Additionally, for an extensible SDK, the toolchain also has
built-in devtool
functionality.
This toolchain is created by running a SDK installer script
or through a
Build Directory
that is based on your metadata configuration or extension for
your targeted device.
The cross-toolchain works with a matching target sysroot.
The native and target sysroots contain needed headers and libraries for generating binaries that run on the target architecture. The target sysroot is based on the target root filesystem image that is built by the OpenEmbedded build system and uses the same metadata configuration used to build the cross-toolchain.
The QEMU emulator allows you to simulate your hardware while running your application or image. QEMU is not part of the SDK but is made available a number of different ways:
If you have cloned the poky
Git
repository to create a
Source Directory
and you have sourced the environment setup script, QEMU is
installed and automatically available.
If you have downloaded a Yocto Project release and unpacked it to create a Source Directory and you have sourced the environment setup script, QEMU is installed and automatically available.
If you have installed the cross-toolchain tarball and you have sourced the toolchain's setup environment script, QEMU is also installed and automatically available.
Fundamentally, the SDK fits into the development process as follows:
The SDK is installed on any machine and can be used to develop applications, images, and kernels. An SDK can even be used by a QA Engineer or Release Engineer. The fundamental concept is that the machine that has the SDK installed does not have to be associated with the machine that has the Yocto Project installed. A developer can independently compile and test an object on their machine and then, when the object is ready for integration into an image, they can simply make it available to the machine that has the Yocto Project. Once the object is available, the image can be rebuilt using the Yocto Project to produce the modified image.
You just need to follow these general steps:
Install the SDK for your target hardware: For information on how to install the SDK, see the "Installing the SDK" section.
Download or Build the Target Image: The Yocto Project supports several target architectures and has many pre-built kernel images and root filesystem images.
If you are going to develop your application on
hardware, go to the
machines
download area and choose a target machine area
from which to download the kernel image and root filesystem.
This download area could have several files in it that
support development using actual hardware.
For example, the area might contain
.hddimg
files that combine the
kernel image with the filesystem, boot loaders, and
so forth.
Be sure to get the files you need for your particular
development process.
If you are going to develop your application and
then run and test it using the QEMU emulator, go to the
machines/qemu
download area.
From this area, go down into the directory for your
target architecture (e.g. qemux86_64
for an Intel®-based
64-bit architecture).
Download the kernel, root filesystem, and any other files you
need for your process.
Develop and Test your Application: At this point, you have the tools to develop your application. If you need to separately install and use the QEMU emulator, you can go to QEMU Home Page to download and learn about the emulator. See the "Using the Quick EMUlator (QEMU)" chapter in the Yocto Project Development Tasks Manual for information on using QEMU within the Yocto Project.
The remainder of this manual describes how to use the extensible and standard SDKs. Information also exists in appendix form that describes how you can build, install, and modify an SDK.
devtool
in Your SDK Workflowdevtool add
This chapter describes the extensible SDK and how to install it.
Information covers the pieces of the SDK, how to install it, and
presents a look at using the devtool
functionality.
The extensible SDK makes it easy to add new applications and libraries
to an image, modify the source for an existing component, test
changes on the target hardware, and ease integration into the rest of
the
OpenEmbedded build system.
In addition to the functionality available through
devtool
, you can alternatively make use of the
toolchain directly, for example from Makefile and Autotools.
See the
"Using the SDK Toolchain Directly"
chapter for more information.
The extensible SDK provides a cross-development toolchain and
libraries tailored to the contents of a specific image.
You would use the Extensible SDK if you want a toolchain experience
supplemented with the powerful set of devtool
commands tailored for the Yocto Project environment.
The installed extensible SDK consists of several files and
directories.
Basically, it contains an SDK environment setup script, some
configuration files, an internal build system, and the
devtool
functionality.
The first thing you need to do is install the SDK on your
Build Host
by running the *.sh
installation script.
You can download a tarball installer, which includes the
pre-built toolchain, the runqemu
script, the internal build system, devtool
,
and support files from the appropriate
toolchain
directory within the Index of Releases.
Toolchains are available for several 32-bit and 64-bit
architectures with the x86_64
directories,
respectively.
The toolchains the Yocto Project provides are based off the
core-image-sato
and
core-image-minimal
images and contain
libraries appropriate for developing against that image.
The names of the tarball installer scripts are such that a string representing the host system appears first in the filename and then is immediately followed by a string representing the target architecture. An extensible SDK has the string "-ext" as part of the name. Following is the general form:
poky-glibc-host_system
-image_type
-arch
-toolchain-ext-release_version
.sh Where:host_system
is a string representing your development system: i686 or x86_64.image_type
is the image for which the SDK was built: core-image-sato or core-image-minimalarch
is a string representing the tuned target architecture: aarch64, armv5e, core2-64, i586, mips32r2, mips64, ppc7400, or cortexa8hf-neonrelease_version
is a string representing the release number of the Yocto Project: 3.1.2, 3.1.2+snapshot
For example, the following SDK installer is for a 64-bit
development host system and a i586-tuned target architecture
based off the SDK for core-image-sato
and
using the current 3.1.2 snapshot:
poky-glibc-x86_64-core-image-sato-i586-toolchain-ext-3.1.2.sh
The SDK and toolchains are self-contained and by default are
installed into the poky_sdk
folder in your
home directory.
You can choose to install the extensible SDK in any location when
you run the installer.
However, because files need to be written under that directory
during the normal course of operation, the location you choose
for installation must be writable for whichever
users need to use the SDK.
The following command shows how to run the installer given a
toolchain tarball for a 64-bit x86 development host system and
a 64-bit x86 target architecture.
The example assumes the SDK installer is located in
~/Downloads/
and has execution rights.
$ ./Downloads/poky-glibc-x86_64-core-image-minimal-core2-64-toolchain-ext-2.5.sh Poky (Yocto Project Reference Distro) Extensible SDK installer version 2.5 ========================================================================== Enter target directory for SDK (default: ~/poky_sdk): You are about to install the SDK to "/home/scottrif/poky_sdk". Proceed [Y/n]? Y Extracting SDK..............done Setting it up... Extracting buildtools... Preparing build system... Parsing recipes: 100% |##################################################################| Time: 0:00:52 Initialising tasks: 100% |###############################################################| Time: 0:00:00 Checking sstate mirror object availability: 100% |#######################################| Time: 0:00:00 Loading cache: 100% |####################################################################| Time: 0:00:00 Initialising tasks: 100% |###############################################################| Time: 0:00:00 done SDK has been successfully set up and is ready to be used. Each time you wish to use the SDK in a new shell session, you need to source the environment setup script e.g. $ . /home/scottrif/poky_sdk/environment-setup-core2-64-poky-linux
Once you have the SDK installed, you must run the SDK environment
setup script before you can actually use the SDK.
This setup script resides in the directory you chose when you
installed the SDK, which is either the default
poky_sdk
directory or the directory you
chose during installation.
Before running the script, be sure it is the one that matches the
architecture for which you are developing.
Environment setup scripts begin with the string
"environment-setup
" and include as part of
their name the tuned target architecture.
As an example, the following commands set the working directory
to where the SDK was installed and then source the environment
setup script.
In this example, the setup script is for an IA-based
target machine using i586 tuning:
$ cd /home/scottrif/poky_sdk $ source environment-setup-core2-64-poky-linux SDK environment now set up; additionally you may now run devtool to perform development tasks. Run devtool --help for further details.
Running the setup script defines many environment variables needed
in order to use the SDK (e.g. PATH
,
CC
,
LD
,
and so forth).
If you want to see all the environment variables the script
exports, examine the installation file itself.
devtool
in Your SDK Workflow¶
The cornerstone of the extensible SDK is a command-line tool
called devtool
.
This tool provides a number of features that help
you build, test and package software within the extensible SDK, and
optionally integrate it into an image built by the OpenEmbedded
build system.
devtool
is not limited to
the extensible SDK.
You can use devtool
to help you easily
develop any project whose build output must be part of an
image built using the build system.
The devtool
command line is organized
similarly to
Git in that it
has a number of sub-commands for each function.
You can run devtool --help
to see all the
commands.
devtool
Quick Reference"
in the Yocto Project Reference Manual for a
devtool
quick reference.
Three devtool
subcommands exist that provide
entry-points into development:
devtool add
:
Assists in adding new software to be built.
devtool modify
:
Sets up an environment to enable you to modify the source of
an existing component.
devtool upgrade
:
Updates an existing recipe so that you can build it for
an updated set of source files.
As with the build system, "recipes" represent software packages
within devtool
.
When you use devtool add
, a recipe is
automatically created.
When you use devtool modify
, the specified
existing recipe is used in order to determine where to get the
source code and how to patch it.
In both cases, an environment is set up so that when you build the
recipe a source tree that is under your control is used in order to
allow you to make changes to the source as desired.
By default, new recipes and the source go into a "workspace"
directory under the SDK.
The remainder of this section presents the
devtool add
,
devtool modify
, and
devtool upgrade
workflows.
devtool add
to Add an Application¶
The devtool add
command generates
a new recipe based on existing source code.
This command takes advantage of the
workspace
layer that many devtool
commands
use.
The command is flexible enough to allow you to extract source
code into both the workspace or a separate local Git repository
and to use existing code that does not need to be extracted.
Depending on your particular scenario, the arguments and options
you use with devtool add
form different
combinations.
The following diagram shows common development flows
you would use with the devtool add
command:
Generating the New Recipe:
The top part of the flow shows three scenarios by which
you could use devtool add
to
generate a recipe based on existing source code.
In a shared development environment, it is typical for other developers to be responsible for various areas of source code. As a developer, you are probably interested in using that source code as part of your development within the Yocto Project. All you need is access to the code, a recipe, and a controlled area in which to do your work.
Within the diagram, three possible scenarios
feed into the devtool add
workflow:
Left: The left scenario in the figure represents a common situation where the source code does not exist locally and needs to be extracted. In this situation, the source code is extracted to the default workspace - you do not want the files in some specific location outside of the workspace. Thus, everything you need will be located in the workspace:
$ devtool add recipe fetchuri
With this command, devtool
extracts the upstream source files into a local
Git repository within the
sources
folder.
The command then creates a recipe named
recipe
and a
corresponding append file in the workspace.
If you do not provide
recipe
, the command
makes an attempt to determine the recipe name.
Middle: The middle scenario in the figure also represents a situation where the source code does not exist locally. In this case, the code is again upstream and needs to be extracted to some local area - this time outside of the default workspace.
devtool
always creates
a Git repository locally during the
extraction.
Furthermore, the first positional argument
srctree
in this
case identifies where the
devtool add
command
will locate the extracted code outside of the
workspace.
You need to specify an empty directory:
$ devtool add recipe srctree fetchuri
In summary, the source code is pulled from
fetchuri
and
extracted into the location defined by
srctree
as a local
Git repository.
Within workspace,
devtool
creates a
recipe named recipe
along with an associated append file.
Right:
The right scenario in the figure represents a
situation where the
srctree
has been
previously prepared outside of the
devtool
workspace.
The following command provides a new recipe name and identifies the existing source tree location:
$ devtool add recipe srctree
The command examines the source code and
creates a recipe named
recipe
for the code
and places the recipe into the workspace.
Because the extracted source code already
exists, devtool
does not
try to relocate the source code into the
workspace - only the new recipe is placed
in the workspace.
Aside from a recipe folder, the command
also creates an associated append folder and
places an initial
*.bbappend
file within.
Edit the Recipe:
You can use devtool edit-recipe
to open up the editor as defined by the
$EDITOR
environment variable
and modify the file:
$ devtool edit-recipe recipe
From within the editor, you can make modifications to the recipe that take affect when you build it later.
Build the Recipe or Rebuild the Image: The next step you take depends on what you are going to do with the new code.
If you need to eventually move the build output
to the target hardware, use the following
devtool
command:
$ devtool build recipe
On the other hand, if you want an image to
contain the recipe's packages from the workspace
for immediate deployment onto a device (e.g. for
testing purposes), you can use
the devtool build-image
command:
$ devtool build-image image
Deploy the Build Output:
When you use the devtool build
command to build out your recipe, you probably want to
see if the resulting build output works as expected
on the target hardware.
You can deploy your build output to that target
hardware by using the
devtool deploy-target
command:
$ devtool deploy-target recipe target
The target
is a live target
machine running as an SSH server.
You can, of course, also deploy the image you
build to actual hardware by using the
devtool build-image
command.
However, devtool
does not provide
a specific command that allows you to deploy the
image to actual hardware.
Finish Your Work With the Recipe:
The devtool finish
command creates
any patches corresponding to commits in the local
Git repository, moves the new recipe to a more permanent
layer, and then resets the recipe so that the recipe is
built normally rather than from the workspace.
$ devtool finish recipe layer
As mentioned, the
devtool finish
command moves the
final recipe to its permanent layer.
As a final process of the
devtool finish
command, the state
of the standard layers and the upstream source is
restored so that you can build the recipe from those
areas rather than the workspace.
devtool reset
command to put things back should you decide you
do not want to proceed with your work.
If you do use this command, realize that the source
tree is preserved.
devtool modify
to Modify the Source of an Existing Component¶
The devtool modify
command prepares the
way to work on existing code that already has a local recipe in
place that is used to build the software.
The command is flexible enough to allow you to extract code
from an upstream source, specify the existing recipe, and
keep track of and gather any patch files from other developers
that are associated with the code.
Depending on your particular scenario, the arguments and options
you use with devtool modify
form different
combinations.
The following diagram shows common development flows for the
devtool modify
command:
Preparing to Modify the Code:
The top part of the flow shows three scenarios by which
you could use devtool modify
to
prepare to work on source files.
Each scenario assumes the following:
The recipe exists locally in a layer external
to the devtool
workspace.
The source files exist either upstream in an un-extracted state or locally in a previously extracted state.
The typical situation is where another developer has created a layer for use with the Yocto Project and their recipe already resides in that layer. Furthermore, their source code is readily available either upstream or locally.
Left:
The left scenario in the figure represents a
common situation where the source code does
not exist locally and it needs to be extracted
from an upstream source.
In this situation, the source is extracted
into the default devtool
workspace location.
The recipe, in this scenario, is in its own
layer outside the workspace
(i.e.
meta-
layername
).
The following command identifies the recipe and, by default, extracts the source files:
$ devtool modify recipe
Once devtool
locates the
recipe, devtool
uses the
recipe's
SRC_URI
statements to locate the source code and any
local patch files from other developers.
With this scenario, no
srctree
argument
exists.
Consequently, the default behavior of the
devtool modify
command is
to extract the source files pointed to by the
SRC_URI
statements into a
local Git structure.
Furthermore, the location for the extracted
source is the default area within the
devtool
workspace.
The result is that the command sets up both
the source code and an append file within the
workspace while the recipe remains in its
original location.
Additionally, if you have any non-patch
local files (i.e. files referred to with
file://
entries in
SRC_URI
statement excluding
*.patch/
or
*.diff
), these files are
copied to an
oe-local-files
folder
under the newly created source tree.
Copying the files here gives you a convenient
area from which you can modify the files.
Any changes or additions you make to those
files are incorporated into the build the next
time you build the software just as are other
changes you might have made to the source.
Middle: The middle scenario in the figure represents a situation where the source code also does not exist locally. In this case, the code is again upstream and needs to be extracted to some local area as a Git repository. The recipe, in this scenario, is again local and in its own layer outside the workspace.
The following command tells
devtool
the recipe with
which to work and, in this case, identifies a
local area for the extracted source files that
exists outside of the default
devtool
workspace:
$ devtool modify recipe srctree
srctree
using
the devtool
command.
As with all extractions, the command uses
the recipe's SRC_URI
statements to locate the source files and any
associated patch files.
Non-patch files are copied to an
oe-local-files
folder
under the newly created source tree.
Once the files are located, the command
by default extracts them into
srctree
.
Within workspace,
devtool
creates an append
file for the recipe.
The recipe remains in its original location but
the source files are extracted to the location
you provide with
srctree
.
Right:
The right scenario in the figure represents a
situation where the source tree
(srctree
) already
exists locally as a previously extracted Git
structure outside of the
devtool
workspace.
In this example, the recipe also exists
elsewhere locally in its own layer.
The following command tells
devtool
the recipe
with which to work, uses the "-n" option to
indicate source does not need to be extracted,
and uses srctree
to
point to the previously extracted source files:
$ devtool modify -n recipe srctree
If an oe-local-files
subdirectory happens to exist and it contains
non-patch files, the files are used.
However, if the subdirectory does not exist and
you run the devtool finish
command, any non-patch files that might exist
next to the recipe are removed because it
appears to devtool
that
you have deleted those files.
Once the
devtool modify
command
finishes, it creates only an append file for
the recipe in the devtool
workspace.
The recipe and the source code remain in their
original locations.
Edit the Source:
Once you have used the
devtool modify
command, you are
free to make changes to the source files.
You can use any editor you like to make and save
your source code modifications.
Build the Recipe or Rebuild the Image: The next step you take depends on what you are going to do with the new code.
If you need to eventually move the build output
to the target hardware, use the following
devtool
command:
$ devtool build recipe
On the other hand, if you want an image to
contain the recipe's packages from the workspace
for immediate deployment onto a device (e.g. for
testing purposes), you can use
the devtool build-image
command:
$ devtool build-image image
Deploy the Build Output:
When you use the devtool build
command to build out your recipe, you probably want to
see if the resulting build output works as expected
on target hardware.
You can deploy your build output to that target
hardware by using the
devtool deploy-target
command:
$ devtool deploy-target recipe target
The target
is a live target
machine running as an SSH server.
You can, of course, use other methods to deploy
the image you built using the
devtool build-image
command to
actual hardware.
devtool
does not provide
a specific command to deploy the image to actual
hardware.
Finish Your Work With the Recipe:
The devtool finish
command creates
any patches corresponding to commits in the local
Git repository, updates the recipe to point to them
(or creates a .bbappend
file to do
so, depending on the specified destination layer), and
then resets the recipe so that the recipe is built
normally rather than from the workspace.
$ devtool finish recipe layer
devtool finish
command.
Because there is no need to move the recipe,
devtool finish
either updates the
original recipe in the original layer or the command
creates a .bbappend
file in a
different layer as provided by
layer
.
Any work you did in the
oe-local-files
directory is
preserved in the original files next to the recipe
during the devtool finish
command.
As a final process of the
devtool finish
command, the state
of the standard layers and the upstream source is
restored so that you can build the recipe from those
areas rather than from the workspace.
devtool reset
command to put things back should you decide you
do not want to proceed with your work.
If you do use this command, realize that the source
tree is preserved.
devtool upgrade
to Create a Version of the Recipe that Supports a Newer Version of the Software¶
The devtool upgrade
command upgrades
an existing recipe to that of a more up-to-date version
found upstream.
Throughout the life of software, recipes continually undergo
version upgrades by their upstream publishers.
You can use the devtool upgrade
workflow to make sure your recipes you are using for builds
are up-to-date with their upstream counterparts.
devtool upgrade
happens to be one.
You can read about all the methods by which you can
upgrade recipes in the
"Upgrading Recipes"
section of the Yocto Project Development Tasks Manual.
The devtool upgrade
command is flexible
enough to allow you to specify source code revision and
versioning schemes, extract code into or out of the
devtool
workspace,
and work with any source file forms that the
fetchers
support.
The following diagram shows the common development flow
used with the devtool upgrade
command:
Initiate the Upgrade:
The top part of the flow shows the typical scenario by
which you use the devtool upgrade
command.
The following conditions exist:
The recipe exists in a local layer external
to the devtool
workspace.
The source files for the new release
exist in the same location pointed to by
SRC_URI
in the recipe (e.g. a tarball with the new
version number in the name, or as a different
revision in the upstream Git repository).
A common situation is where third-party software has undergone a revision so that it has been upgraded. The recipe you have access to is likely in your own layer. Thus, you need to upgrade the recipe to use the newer version of the software:
$ devtool upgrade -V version recipe
By default, the devtool upgrade
command extracts source code into the
sources
directory in the
workspace.
If you want the code extracted to any other location,
you need to provide the
srctree
positional argument
with the command as follows:
$ devtool upgrade -V version recipe srctree
If the source files pointed to by the
SRC_URI
statement in the recipe
are in a Git repository, you must provide the "-S"
option and specify a revision for the software.
Once devtool
locates the
recipe, it uses the SRC_URI
variable to locate the source code and any local patch
files from other developers.
The result is that the command sets up the source
code, the new version of the recipe, and an append file
all within the workspace.
Additionally, if you have any non-patch
local files (i.e. files referred to with
file://
entries in
SRC_URI
statement excluding
*.patch/
or
*.diff
), these files are
copied to an
oe-local-files
folder
under the newly created source tree.
Copying the files here gives you a convenient
area from which you can modify the files.
Any changes or additions you make to those
files are incorporated into the build the next
time you build the software just as are other
changes you might have made to the source.
Resolve any Conflicts created by the Upgrade:
Conflicts could exist due to the software being
upgraded to a new version.
Conflicts occur if your recipe specifies some patch
files in SRC_URI
that conflict
with changes made in the new version of the software.
For such cases, you need to resolve the conflicts
by editing the source and following the normal
git rebase
conflict resolution
process.
Before moving onto the next step, be sure to resolve any such conflicts created through use of a newer or different version of the software.
Build the Recipe or Rebuild the Image: The next step you take depends on what you are going to do with the new code.
If you need to eventually move the build output
to the target hardware, use the following
devtool
command:
$ devtool build recipe
On the other hand, if you want an image to
contain the recipe's packages from the workspace
for immediate deployment onto a device (e.g. for
testing purposes), you can use
the devtool build-image
command:
$ devtool build-image image
Deploy the Build Output:
When you use the devtool build
command or bitbake
to build
your recipe, you probably want to see if the resulting
build output works as expected on target hardware.
You can deploy your build output to that target
hardware by using the
devtool deploy-target
command:
$ devtool deploy-target recipe target
The target
is a live target
machine running as an SSH server.
You can, of course, also deploy the image you
build using the
devtool build-image
command
to actual hardware.
However, devtool
does not provide
a specific command that allows you to do this.
Finish Your Work With the Recipe:
The devtool finish
command creates
any patches corresponding to commits in the local
Git repository, moves the new recipe to a more
permanent layer, and then resets the recipe so that
the recipe is built normally rather than from the
workspace.
Any work you did in the
oe-local-files
directory is
preserved in the original files next to the recipe
during the devtool finish
command.
If you specify a destination layer that is the same as the original source, then the old version of the recipe and associated files are removed prior to adding the new version.
$ devtool finish recipe layer
As a final process of the
devtool finish
command, the state
of the standard layers and the upstream source is
restored so that you can build the recipe from those
areas rather than the workspace.
devtool reset
command to put things back should you decide you
do not want to proceed with your work.
If you do use this command, realize that the source
tree is preserved.
devtool add
¶
The devtool add
command automatically creates
a recipe based on the source tree you provide with the command.
Currently, the command has support for the following:
Autotools (autoconf
and
automake
)
CMake
Scons
qmake
Plain Makefile
Out-of-tree kernel module
Binary package (i.e. "-b" option)
Node.js module
Python modules that use setuptools
or distutils
Apart from binary packages, the determination of how a source tree
should be treated is automatic based on the files present within
that source tree.
For example, if a CMakeLists.txt
file is found,
then the source tree is assumed to be using
CMake and is treated accordingly.
The remainder of this section covers specifics regarding how parts of the recipe are generated.
If you do not specify a name and version on the command
line, devtool add
uses various metadata
within the source tree in an attempt to determine
the name and version of the software being built.
Based on what the tool determines, devtool
sets the name of the created recipe file accordingly.
If devtool
cannot determine the name and
version, the command prints an error.
For such cases, you must re-run the command and provide
the name and version, just the name, or just the version as
part of the command line.
Sometimes the name or version determined from the source tree might be incorrect. For such a case, you must reset the recipe:
$ devtool reset -n recipename
After running the devtool reset
command,
you need to run devtool add
again and
provide the name or the version.
The devtool add
command attempts to
detect build-time dependencies and map them to other recipes
in the system.
During this mapping, the command fills in the names of those
recipes as part of the
DEPENDS
variable within the recipe.
If a dependency cannot be mapped, devtool
places a comment in the recipe indicating such.
The inability to map a dependency can result from naming not
being recognized or because the dependency simply is not
available.
For cases where the dependency is not available, you must use
the devtool add
command to add an
additional recipe that satisfies the dependency.
Once you add that recipe, you need to update the
DEPENDS
variable in the original recipe
to include the new recipe.
If you need to add runtime dependencies, you can do so by adding the following to your recipe:
RDEPENDS_${PN} += "dependency1 dependency2 ...
"
devtool add
command often cannot
distinguish between mandatory and optional dependencies.
Consequently, some of the detected dependencies might
in fact be optional.
When in doubt, consult the documentation or the configure
script for the software the recipe is building for further
details.
In some cases, you might find you can substitute the
dependency with an option that disables the associated
functionality passed to the configure script.
The devtool add
command attempts to
determine if the software you are adding is able to be
distributed under a common, open-source license.
If so, the command sets the
LICENSE
value accordingly.
You should double-check the value added by the command against
the documentation or source files for the software you are
building and, if necessary, update that
LICENSE
value.
The devtool add
command also sets the
LIC_FILES_CHKSUM
value to point to all files that appear to be license-related.
Realize that license statements often appear in comments at
the top of source files or within the documentation.
In such cases, the command does not recognize those license
statements.
Consequently, you might need to amend the
LIC_FILES_CHKSUM
variable to point to one
or more of those comments if present.
Setting LIC_FILES_CHKSUM
is particularly
important for third-party software.
The mechanism attempts to ensure correct licensing should you
upgrade the recipe to a newer upstream version in future.
Any change in licensing is detected and you receive an error
prompting you to check the license text again.
If the devtool add
command cannot
determine licensing information, devtool
sets the LICENSE
value to "CLOSED" and
leaves the LIC_FILES_CHKSUM
value unset.
This behavior allows you to continue with development even
though the settings are unlikely to be correct in all cases.
You should check the documentation or source files for the
software you are building to determine the actual license.
The use of Make by itself is very common in both proprietary
and open-source software.
Unfortunately, Makefiles are often not written with
cross-compilation in mind.
Thus, devtool add
often cannot do very
much to ensure that these Makefiles build correctly.
It is very common, for example, to explicitly call
gcc
instead of using the
CC
variable.
Usually, in a cross-compilation environment,
gcc
is the compiler for the build host
and the cross-compiler is named something similar to
arm-poky-linux-gnueabi-gcc
and might
require arguments (e.g. to point to the associated sysroot
for the target machine).
When writing a recipe for Makefile-only software, keep the following in mind:
You probably need to patch the Makefile to use
variables instead of hardcoding tools within the
toolchain such as gcc
and
g++
.
The environment in which Make runs is set up with
various standard variables for compilation (e.g.
CC
, CXX
, and
so forth) in a similar manner to the environment set
up by the SDK's environment setup script.
One easy way to see these variables is to run the
devtool build
command on the
recipe and then look in
oe-logs/run.do_compile
.
Towards the top of this file, a list of environment
variables exists that are being set.
You can take advantage of these variables within the
Makefile.
If the Makefile sets a default for a variable using "=",
that default overrides the value set in the environment,
which is usually not desirable.
For this case, you can either patch the Makefile
so it sets the default using the "?=" operator, or
you can alternatively force the value on the
make
command line.
To force the value on the command line, add the
variable setting to
EXTRA_OEMAKE
or
PACKAGECONFIG_CONFARGS
within the recipe.
Here is an example using EXTRA_OEMAKE
:
EXTRA_OEMAKE += "'CC=${CC}' 'CXX=${CXX}'"
In the above example, single quotes are used around the variable settings as the values are likely to contain spaces because required default options are passed to the compiler.
Hardcoding paths inside Makefiles is often problematic in a cross-compilation environment. This is particularly true because those hardcoded paths often point to locations on the build host and thus will either be read-only or will introduce contamination into the cross-compilation because they are specific to the build host rather than the target. Patching the Makefile to use prefix variables or other path variables is usually the way to handle this situation.
Sometimes a Makefile runs target-specific commands such
as ldconfig
.
For such cases, you might be able to apply patches that
remove these commands from the Makefile.
Often, you need to build additional tools that run on the
build host
as opposed to the target.
You should indicate this requirement by using one of the
following methods when you run
devtool add
:
Specify the name of the recipe such that it ends with "-native". Specifying the name like this produces a recipe that only builds for the build host.
Specify the "‐‐also-native" option with the
devtool add
command.
Specifying this option creates a recipe file that still
builds for the target but also creates a variant with
a "-native" suffix that builds for the build host.
You can use the devtool add
command two
different ways to add Node.js modules: 1) Through
npm
and, 2) from a repository or local
source.
Use the following form to add Node.js modules through
npm
:
$ devtool add "npm://registry.npmjs.org;name=forever;version=0.15.1"
The name and version parameters are mandatory. Lockdown and shrinkwrap files are generated and pointed to by the recipe in order to freeze the version that is fetched for the dependencies according to the first time. This also saves checksums that are verified on future fetches. Together, these behaviors ensure the reproducibility and integrity of the build.
You must use quotes around the URL.
The devtool add
does not require
the quotes, but the shell considers ";" as a splitter
between multiple commands.
Thus, without the quotes,
devtool add
does not receive the
other parts, which results in several "command not
found" errors.
In order to support adding Node.js modules, a
nodejs
recipe must be part
of your SDK.
As mentioned earlier, you can also add Node.js modules
directly from a repository or local source tree.
To add modules this way, use devtool add
in the following form:
$ devtool add https://github.com/diversario/node-ssdp
In this example, devtool
fetches the
specified Git repository, detects the code as Node.js
code, fetches dependencies using npm
, and
sets
SRC_URI
accordingly.
When building a recipe using the
devtool build
command, the typical build
progresses as follows:
Fetch the source
Unpack the source
Configure the source
Compile the source
Install the build output
Package the installed output
For recipes in the workspace, fetching and unpacking is disabled
as the source tree has already been prepared and is persistent.
Each of these build steps is defined as a function (task), usually
with a "do_" prefix (e.g.
do_fetch
,
do_unpack
,
and so forth).
These functions are typically shell scripts but can instead be
written in Python.
If you look at the contents of a recipe, you will see that the
recipe does not include complete instructions for building the
software.
Instead, common functionality is encapsulated in classes inherited
with the inherit
directive.
This technique leaves the recipe to describe just the things that
are specific to the software being built.
A
base
class exists that is implicitly inherited by all recipes and
provides the functionality that most recipes typically need.
The remainder of this section presents information useful when working with recipes.
After the first run of the devtool build
command, recipes that were previously created using the
devtool add
command or whose sources were
modified using the devtool modify
command contain symbolic links created within the source tree:
oe-logs
:
This link points to the directory in which log files
and run scripts for each build step are created.
oe-workdir
:
This link points to the temporary work area for the
recipe.
The following locations under
oe-workdir
are particularly
useful:
image/
:
Contains all of the files installed during
the
do_install
stage.
Within a recipe, this directory is referred
to by the expression
${
D
}
.
sysroot-destdir/
:
Contains a subset of files installed within
do_install
that have
been put into the shared sysroot.
For more information, see the
"Sharing Files Between Recipes"
section.
packages-split/
:
Contains subdirectories for each package
produced by the recipe.
For more information, see the
"Packaging"
section.
You can use these links to get more information on what is happening at each build step.
If the software your recipe is building uses GNU autoconf,
then a fixed set of arguments is passed to it to enable
cross-compilation plus any extras specified by
EXTRA_OECONF
or
PACKAGECONFIG_CONFARGS
set within the recipe.
If you wish to pass additional options, add them to
EXTRA_OECONF
or
PACKAGECONFIG_CONFARGS
.
Other supported build tools have similar variables
(e.g.
EXTRA_OECMAKE
for CMake,
EXTRA_OESCONS
for Scons, and so forth).
If you need to pass anything on the make
command line, you can use EXTRA_OEMAKE
or the
PACKAGECONFIG_CONFARGS
variables to do so.
You can use the devtool configure-help
command
to help you set the arguments listed in the previous paragraph.
The command determines the exact options being passed, and shows
them to you along with any custom arguments specified through
EXTRA_OECONF
or
PACKAGECONFIG_CONFARGS
.
If applicable, the command also shows you the output of the
configure script's "‐‐help" option as a reference.
Recipes often need to use files provided by other recipes on the build host. For example, an application linking to a common library needs access to the library itself and its associated headers. The way this access is accomplished within the extensible SDK is through the sysroot. One sysroot exists per "machine" for which the SDK is being built. In practical terms, this means a sysroot exists for the target machine, and a sysroot exists for the build host.
Recipes should never write files directly into the sysroot.
Instead, files should be installed into standard locations
during the
do_install
task within the
${
D
}
directory.
A subset of these files automatically goes into the sysroot.
The reason for this limitation is that almost all files that go
into the sysroot are cataloged in manifests in order to ensure
they can be removed later when a recipe is modified or removed.
Thus, the sysroot is able to remain free from stale files.
Packaging is not always particularly relevant within the extensible SDK. However, if you examine how build output gets into the final image on the target device, it is important to understand packaging because the contents of the image are expressed in terms of packages and not recipes.
During the
do_package
task, files installed during the
do_install
task are split into one main package, which is almost always
named the same as the recipe, and into several other packages.
This separation exists because not all of those installed files
are useful in every image.
For example, you probably do not need any of the documentation
installed in a production image.
Consequently, for each recipe the documentation files are
separated into a -doc
package.
Recipes that package software containing optional modules or
plugins might undergo additional package splitting as well.
After building a recipe, you can see where files have gone by
looking in the oe-workdir/packages-split
directory, which contains a subdirectory for each package.
Apart from some advanced cases, the
PACKAGES
and
FILES
variables controls splitting.
The PACKAGES
variable lists all of the
packages to be produced, while the FILES
variable specifies which files to include in each package by
using an override to specify the package.
For example, FILES_${PN}
specifies the
files to go into the main package (i.e. the main package has
the same name as the recipe and
${
PN
}
evaluates to the recipe name).
The order of the PACKAGES
value is
significant.
For each installed file, the first package whose
FILES
value matches the file is the
package into which the file goes.
Defaults exist for both the PACKAGES
and
FILES
variables.
Consequently, you might find you do not even need to set these
variables in your recipe unless the software the recipe is
building installs files into non-standard locations.
If you use the devtool deploy-target
command to write a recipe's build output to the target, and
you are working on an existing component of the system, then you
might find yourself in a situation where you need to restore the
original files that existed prior to running the
devtool deploy-target
command.
Because the devtool deploy-target
command
backs up any files it overwrites, you can use the
devtool undeploy-target
command to restore
those files and remove any other files the recipe deployed.
Consider the following example:
$ devtool undeploy-target lighttpd root@192.168.7.2
If you have deployed multiple applications, you can remove them all using the "-a" option thus restoring the target device to its original state:
$ devtool undeploy-target -a root@192.168.7.2
Information about files deployed to the target as well as any backed up files are stored on the target itself. This storage, of course, requires some additional space on the target machine.
devtool deploy-target
and
devtool undeploy-target
commands do not
currently interact with any package management system on the
target device (e.g. RPM or OPKG).
Consequently, you should not intermingle
devtool deploy-target
and package
manager operations on the target device.
Doing so could result in a conflicting set of files.
Out of the box the extensible SDK typically only comes with a small
number of tools and libraries.
A minimal SDK starts mostly empty and is populated on-demand.
Sometimes you must explicitly install extra items into the SDK.
If you need these extra items, you can first search for the items
using the devtool search
command.
For example, suppose you need to link to libGL but you are not sure
which recipe provides libGL.
You can use the following command to find out:
$ devtool search libGL mesa A free implementation of the OpenGL API
Once you know the recipe (i.e. mesa
in this
example), you can install it:
$ devtool sdk-install mesa
By default, the devtool sdk-install
command
assumes the item is available in pre-built form from your SDK
provider.
If the item is not available and it is acceptable to build the item
from source, you can add the "-s" option as follows:
$ devtool sdk-install -s mesa
It is important to remember that building the item from source
takes significantly longer than installing the pre-built artifact.
Also, if no recipe exists for the item you want to add to the SDK,
you must instead add the item using the
devtool add
command.
If you are working with an installed extensible SDK that gets occasionally updated (e.g. a third-party SDK), then you will need to manually "pull down" the updates into the installed SDK.
To update your installed SDK, use devtool
as
follows:
$ devtool sdk-update
The previous command assumes your SDK provider has set the default
update URL for you through the
SDK_UPDATE_URL
variable as described in the
"Providing Updates to the Extensible SDK After Installation"
section.
If the SDK provider has not set that default URL, you need to
specify it yourself in the command as follows:
$ devtool sdk-update path_to_update_directory
You might need to produce an SDK that contains your own custom libraries. A good example would be if you were a vendor with customers that use your SDK to build their own platform-specific software and those customers need an SDK that has custom libraries. In such a case, you can produce a derivative SDK based on the currently installed SDK fairly easily by following these steps:
If necessary, install an extensible SDK that you want to use as a base for your derivative SDK.
Source the environment script for the SDK.
Add the extra libraries or other components you want by
using the devtool add
command.
Run the devtool build-sdk
command.
The previous steps take the recipes added to the workspace and construct a new SDK installer that contains those recipes and the resulting binary artifacts. The recipes go into their own separate layer in the constructed derivative SDK, which leaves the workspace clean and ready for users to add their own recipes.
This chapter describes the standard SDK and how to install it. Information includes unique installation and setup aspects for the standard SDK.
You can use a standard SDK to work on Makefile and Autotools-based projects. See the "Using the SDK Toolchain Directly" chapter for more information.
The Standard SDK provides a cross-development toolchain and
libraries tailored to the contents of a specific image.
You would use the Standard SDK if you want a more traditional
toolchain experience as compared to the extensible SDK, which
provides an internal build system and the
devtool
functionality.
The installed Standard SDK consists of several files and directories. Basically, it contains an SDK environment setup script, some configuration files, and host and target root filesystems to support usage. You can see the directory structure in the "Installed Standard SDK Directory Structure" section.
The first thing you need to do is install the SDK on your
Build Host
by running the *.sh
installation script.
You can download a tarball installer, which includes the
pre-built toolchain, the runqemu
script, and support files from the appropriate
toolchain
directory within the Index of Releases.
Toolchains are available for several 32-bit and 64-bit
architectures with the x86_64
directories,
respectively.
The toolchains the Yocto Project provides are based off the
core-image-sato
and
core-image-minimal
images and contain
libraries appropriate for developing against that image.
The names of the tarball installer scripts are such that a string representing the host system appears first in the filename and then is immediately followed by a string representing the target architecture.
poky-glibc-host_system
-image_type
-arch
-toolchain-release_version
.sh Where:host_system
is a string representing your development system: i686 or x86_64.image_type
is the image for which the SDK was built: core-image-minimal or core-image-sato.arch
is a string representing the tuned target architecture: aarch64, armv5e, core2-64, i586, mips32r2, mips64, ppc7400, or cortexa8hf-neon.release_version
is a string representing the release number of the Yocto Project: 3.1.2, 3.1.2+snapshot
For example, the following SDK installer is for a 64-bit
development host system and a i586-tuned target architecture
based off the SDK for core-image-sato
and
using the current 3.1.2 snapshot:
poky-glibc-x86_64-core-image-sato-i586-toolchain-3.1.2.sh
The SDK and toolchains are self-contained and by default are
installed into the poky_sdk
folder in your
home directory.
You can choose to install the extensible SDK in any location when
you run the installer.
However, because files need to be written under that directory
during the normal course of operation, the location you choose
for installation must be writable for whichever
users need to use the SDK.
The following command shows how to run the installer given a
toolchain tarball for a 64-bit x86 development host system and
a 64-bit x86 target architecture.
The example assumes the SDK installer is located in
~/Downloads/
and has execution rights.
$ ./Downloads/poky-glibc-x86_64-core-image-sato-i586-toolchain-3.1.2.sh Poky (Yocto Project Reference Distro) SDK installer version 3.1.2 =============================================================== Enter target directory for SDK (default: /opt/poky/3.1.2): You are about to install the SDK to "/opt/poky/3.1.2". Proceed [Y/n]? Y Extracting SDK........................................ ..............................done Setting it up...done SDK has been successfully set up and is ready to be used. Each time you wish to use the SDK in a new shell session, you need to source the environment setup script e.g. $ . /opt/poky/3.1.2/environment-setup-i586-poky-linux
Again, reference the "Installed Standard SDK Directory Structure" section for more details on the resulting directory structure of the installed SDK.
Once you have the SDK installed, you must run the SDK environment
setup script before you can actually use the SDK.
This setup script resides in the directory you chose when you
installed the SDK, which is either the default
/opt/poky/3.1.2
directory or the directory
you chose during installation.
Before running the script, be sure it is the one that matches the
architecture for which you are developing.
Environment setup scripts begin with the string
"environment-setup
" and include as part of
their name the tuned target architecture.
As an example, the following commands set the working directory
to where the SDK was installed and then source the environment
setup script.
In this example, the setup script is for an IA-based
target machine using i586 tuning:
$ source /opt/poky/3.1.2/environment-setup-i586-poky-linux
When you run the setup script, the same environment variables are defined as are when you run the setup script for an extensible SDK. See the "Running the Extensible SDK Environment Setup Script" section for more information.
You can use the SDK toolchain directly with Makefile and Autotools-based projects.
Once you have a suitable cross-development toolchain installed, it is very easy to develop a project using the GNU Autotools-based workflow, which is outside of the OpenEmbedded build system.
The following figure presents a simple Autotools workflow.
Follow these steps to create a simple Autotools-based "Hello World" project:
Create a Working Directory and Populate It: Create a clean directory for your project and then make that directory your working location.
$ mkdir $HOME/helloworld $ cd $HOME/helloworld
After setting up the directory, populate it with files
needed for the flow.
You need a project source file, a file to help with
configuration, and a file to help create the Makefile,
and a README file:
hello.c
,
configure.ac
,
Makefile.am
, and
README
, respectively.
Use the following command to create an empty README file, which is required by GNU Coding Standards:
$ touch README
Create the remaining three files as follows:
hello.c
:
#include <stdio.h> main() { printf("Hello World!\n"); }
configure.ac
:
AC_INIT(hello,0.1) AM_INIT_AUTOMAKE([foreign]) AC_PROG_CC AC_CONFIG_FILES(Makefile) AC_OUTPUT
Makefile.am
:
bin_PROGRAMS = hello hello_SOURCES = hello.c
Source the Cross-Toolchain Environment Setup File: As described earlier in the manual, installing the cross-toolchain creates a cross-toolchain environment setup script in the directory that the SDK was installed. Before you can use the tools to develop your project, you must source this setup script. The script begins with the string "environment-setup" and contains the machine architecture, which is followed by the string "poky-linux". For this example, the command sources a script from the default SDK installation directory that uses the 32-bit Intel x86 Architecture and the Dunfell Yocto Project release:
$ source /opt/poky/3.1.2/environment-setup-i586-poky-linux
Create the configure
Script:
Use the autoreconf
command to
generate the configure
script.
$ autoreconf
The autoreconf
tool takes care
of running the other Autotools such as
aclocal
,
autoconf
, and
automake
.
configure.ac
, which
autoreconf
runs, that indicate
missing files, you can use the "-i" option, which
ensures missing auxiliary files are copied to the build
host.
Cross-Compile the Project:
This command compiles the project using the
cross-compiler.
The
CONFIGURE_FLAGS
environment variable provides the minimal arguments for
GNU configure:
$ ./configure ${CONFIGURE_FLAGS}
For an Autotools-based project, you can use the
cross-toolchain by just passing the appropriate host
option to configure.sh
.
The host option you use is derived from the name of the
environment setup script found in the directory in which
you installed the cross-toolchain.
For example, the host option for an ARM-based target that
uses the GNU EABI is
armv5te-poky-linux-gnueabi
.
You will notice that the name of the script is
environment-setup-armv5te-poky-linux-gnueabi
.
Thus, the following command works to update your project
and rebuild it using the appropriate cross-toolchain tools:
$ ./configure --host=armv5te-poky-linux-gnueabi --with-libtool-sysroot=sysroot_dir
Make and Install the Project: These two commands generate and install the project into the destination directory:
$ make $ make install DESTDIR=./tmp
This next command is a simple way to verify the installation of your project. Running the command prints the architecture on which the binary file can run. This architecture should be the same architecture that the installed cross-toolchain supports.
$ file ./tmp/usr/local/bin/hello
Execute Your Project: To execute the project, you would need to run it on your target hardware. If your target hardware happens to be your build host, you could run the project as follows:
$ ./tmp/usr/local/bin/hello
As expected, the project displays the "Hello World!" message.
Simple Makefile-based projects use and interact with the
cross-toolchain environment variables established when you run
the cross-toolchain environment setup script.
The environment variables are subject to general
make
rules.
This section presents a simple Makefile development flow and provides an example that lets you see how you can use cross-toolchain environment variables and Makefile variables during development.
The main point of this section is to explain the following three cases regarding variable behavior:
Case 1 - No Variables Set in the
Makefile
Map to Equivalent
Environment Variables Set in the SDK Setup Script:
Because matching variables are not specifically set in the
Makefile
, the variables retain their
values based on the environment setup script.
Case 2 - Variables Are Set in the Makefile that
Map to Equivalent Environment Variables from the SDK
Setup Script:
Specifically setting matching variables in the
Makefile
during the build results in
the environment settings of the variables being
overwritten.
In this case, the variables you set in the
Makefile
are used.
Case 3 - Variables Are Set Using the Command Line
that Map to Equivalent Environment Variables from the
SDK Setup Script:
Executing the Makefile
from the
command line results in the environment variables being
overwritten.
In this case, the command-line content is used.
make
, the
variables from the SDK setup script take precedence:
$ make -e target
The remainder of this section presents a simple Makefile example that demonstrates these variable behaviors.
In a new shell environment variables are not established for the
SDK until you run the setup script.
For example, the following commands show a null value for the
compiler variable (i.e.
CC
).
$ echo ${CC} $
Running the SDK setup script for a 64-bit build host and an
i586-tuned target architecture for a
core-image-sato
image using the current
3.1.2 Yocto Project release and then echoing that variable
shows the value established through the script:
$ source /opt/poky/3.1.2/environment-setup-i586-poky-linux $ echo ${CC} i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux
To illustrate variable use, work through this simple "Hello World!" example:
Create a Working Directory and Populate It: Create a clean directory for your project and then make that directory your working location.
$ mkdir $HOME/helloworld $ cd $HOME/helloworld
After setting up the directory, populate it with files
needed for the flow.
You need a main.c
file from which you
call your function, a module.h
file
to contain headers, and a module.c
that defines your function.
Create the three files as follows:
main.c
:
#include "module.h" void sample_func(); int main() { sample_func(); return 0; }
module.h
:
#include <stdio.h> void sample_func();
module.c
:
#include "module.h" void sample_func() { printf("Hello World!"); printf("\n"); }
Source the Cross-Toolchain Environment Setup File: As described earlier in the manual, installing the cross-toolchain creates a cross-toolchain environment setup script in the directory that the SDK was installed. Before you can use the tools to develop your project, you must source this setup script. The script begins with the string "environment-setup" and contains the machine architecture, which is followed by the string "poky-linux". For this example, the command sources a script from the default SDK installation directory that uses the 32-bit Intel x86 Architecture and the Dunfell Yocto Project release:
$ source /opt/poky/3.1.2/environment-setup-i586-poky-linux
Create the Makefile
:
For this example, the Makefile contains two lines that
can be used to set the CC
variable.
One line is identical to the value that is set when you
run the SDK environment setup script, and the other line
sets CC
to "gcc", the default GNU
compiler on the build host:
# CC=i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux # CC="gcc" all: main.o module.o ${CC} main.o module.o -o target_bin main.o: main.c module.h ${CC} -I . -c main.c module.o: module.c module.h ${CC} -I . -c module.c clean: rm -rf *.o rm target_bin
Make the Project:
Use the make
command to create the
binary output file.
Because variables are commented out in the Makefile,
the value used for CC
is the value
set when the SDK environment setup file was run:
$ make i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux -I . -c main.c i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux -I . -c module.c i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux main.o module.o -o target_bin
From the results of the previous command, you can see that
the compiler used was the compiler established through
the CC
variable defined in the
setup script.
You can override the CC
environment variable with the same variable as set from
the Makefile by uncommenting the line in the Makefile
and running make
again.
$ make clean rm -rf *.o rm target_bin # # Edit the Makefile by uncommenting the line that sets CC to "gcc" # $ make gcc -I . -c main.c gcc -I . -c module.c gcc main.o module.o -o target_bin
As shown in the previous example, the cross-toolchain compiler is not used. Rather, the default compiler is used.
This next case shows how to override a variable
by providing the variable as part of the command line.
Go into the Makefile and re-insert the comment character
so that running make
uses
the established SDK compiler.
However, when you run make
, use a
command-line argument to set CC
to "gcc":
$ make clean rm -rf *.o rm target_bin # # Edit the Makefile to comment out the line setting CC to "gcc" # $ make i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux -I . -c main.c i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux -I . -c module.c i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux main.o module.o -o target_bin $ make clean rm -rf *.o rm target_bin $ make CC="gcc" gcc -I . -c main.c gcc -I . -c module.c gcc main.o module.o -o target_bin
In the previous case, the command-line argument overrides the SDK environment variable.
In this last case, edit Makefile again to use the
"gcc" compiler but then use the "-e" option on the
make
command line:
$ make clean rm -rf *.o rm target_bin # # Edit the Makefile to use "gcc" # $ make gcc -I . -c main.c gcc -I . -c module.c gcc main.o module.o -o target_bin $ make clean rm -rf *.o rm target_bin $ make -e i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux -I . -c main.c i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux -I . -c module.c i586-poky-linux-gcc -m32 -march=i586 --sysroot=/opt/poky/2.5/sysroots/i586-poky-linux main.o module.o -o target_bin
In the previous case, the "-e" option forces
make
to use the SDK environment
variables regardless of the values in the Makefile.
Execute Your Project:
To execute the project (i.e.
target_bin
), use the following
command:
$ ./target_bin Hello World!
target_bin
and your build host
differs in architecture from that of the target
machine, you need to run your project on the target
device.
As expected, the project displays the "Hello World!" message.
You can use existing, pre-built toolchains by locating and running an SDK installer script that ships with the Yocto Project. Using this method, you select and download an architecture-specific SDK installer and then run the script to hand-install the toolchain.
Follow these steps to locate and hand-install the toolchain:
Go to the Installers Directory: Go to http://downloads.yoctoproject.org/releases/yocto/yocto-3.1.2/toolchain/
Open the Folder for Your Build Host:
Open the folder that matches your
build host
(i.e. i686
for 32-bit machines or
x86_64
for 64-bit machines).
Locate and Download the SDK Installer: You need to find and download the installer appropriate for your build host, target hardware, and image type.
The installer files (*.sh
) follow
this naming convention:
poky-glibc-host_system
-core-image-type
-arch
-toolchain[-ext]-release
.sh Where:host_system
is a string representing your development system: "i686" or "x86_64"type
is a string representing the image: "sato" or "minimal"arch
is a string representing the target architecture: "aarch64", "armv5e", "core2-64", "coretexa8hf-neon", "i586", "mips32r2", "mips64", or "ppc7400"release
is the version of Yocto Project. NOTE: The standard SDK installer does not have the "-ext" string as part of the filename.
The toolchains provided by the Yocto Project are based off of
the core-image-sato
and
core-image-minimal
images and contain
libraries appropriate for developing against those images.
For example, if your build host is a 64-bit x86 system
and you need an extended SDK for a 64-bit core2 target, go
into the x86_64
folder and download the
following installer:
poky-glibc-x86_64-core-image-sato-core2-64-toolchain-ext-3.1.2.sh
Run the Installer:
Be sure you have execution privileges and run the installer.
Following is an example from the Downloads
directory:
$ ~/Downloads/poky-glibc-x86_64-core-image-sato-core2-64-toolchain-ext-3.1.2.sh
During execution of the script, you choose the root location for the toolchain. See the "Installed Standard SDK Directory Structure" section and the "Installed Extensible SDK Directory Structure" section for more information.
As an alternative to locating and downloading an SDK installer, you can build the SDK installer. Follow these steps:
Set Up the Build Environment: Be sure you are set up to use BitBake in a shell. See the "Preparing the Build Host" section in the Yocto Project Development Tasks Manual for information on how to get a build host ready that is either a native Linux machine or a machine that uses CROPS.
Clone the poky
Repository:
You need to have a local copy of the Yocto Project
Source Directory
(i.e. a local poky
repository).
See the
"Cloning the poky
Repository"
and possibly the
"Checking Out by Branch in Poky"
and
"Checking Out by Tag in Poky"
sections all in the Yocto Project Development Tasks Manual for
information on how to clone the poky
repository and check out the appropriate branch for your work.
Initialize the Build Environment:
While in the root directory of the Source Directory (i.e.
poky
), run the
oe-init-build-env
environment setup script to define the OpenEmbedded
build environment on your build host.
$ source oe-init-build-env
Among other things, the script creates the
Build Directory,
which is build
in this case
and is located in the Source Directory.
After the script runs, your current working directory
is set to the build
directory.
Make Sure You Are Building an Installer for the Correct Machine:
Check to be sure that your
MACHINE
variable in the local.conf
file in your
Build Directory matches the architecture for which you are
building.
Make Sure Your SDK Machine is Correctly Set:
If you are building a toolchain designed to run on an
architecture that differs from your current development host
machine (i.e. the build host), be sure that the
SDKMACHINE
variable in the local.conf
file in your
Build Directory is correctly set.
SDKMACHINE
value must be
set for the architecture of the machine you are using to
build the installer.
If SDKMACHINE
is not set appropriately,
the build fails and provides an error message similar to
the following:
The extensible SDK can currently only be built for the same architecture as the machine being built on - SDK_ARCH is set to i686 (likely via setting SDKMACHINE) which is different from the architecture of the build machine (x86_64). Unable to continue.
Build the SDK Installer:
To build the SDK installer for a standard SDK and populate
the SDK image, use the following command form.
Be sure to replace image
with
an image (e.g. "core-image-sato"):
$ bitbake image
-c populate_sdk
You can do the same for the extensible SDK using this command form:
$ bitbake image
-c populate_sdk_ext
These commands produce an SDK installer that contains the sysroot that matches your target root filesystem.
When the bitbake
command completes,
the SDK installer will be in
tmp/deploy/sdk
in the Build Directory.
By default, the previous BitBake command does not
build static binaries.
If you want to use the toolchain to build these
types of libraries, you need to be sure your SDK
has the appropriate static development libraries.
Use the
TOOLCHAIN_TARGET_TASK
variable inside your local.conf
file before building the SDK installer.
Doing so ensures that the eventual SDK installation
process installs the appropriate library packages
as part of the SDK.
Following is an example using
libc
static development
libraries:
TOOLCHAIN_TARGET_TASK_append = " libc-staticdev"
Run the Installer:
You can now run the SDK installer from
tmp/deploy/sdk
in the Build Directory.
Following is an example:
$ cd ~/poky/build/tmp/deploy/sdk $ ./poky-glibc-x86_64-core-image-sato-core2-64-toolchain-ext-3.1.2.sh
During execution of the script, you choose the root location for the toolchain. See the "Installed Standard SDK Directory Structure" section and the "Installed Extensible SDK Directory Structure" section for more information.
After installing the toolchain, for some use cases you might need to separately extract a root filesystem:
You want to boot the image using NFS.
You want to use the root filesystem as the target sysroot.
You want to develop your target application using the root filesystem as the target sysroot.
Follow these steps to extract the root filesystem:
Locate and Download the Tarball for the Pre-Built Root Filesystem Image File: You need to find and download the root filesystem image file that is appropriate for your target system. These files are kept in machine-specific folders in the Index of Releases in the "machines" directory.
The machine-specific folders of the "machines" directory
contain tarballs (*.tar.bz2
) for supported
machines.
These directories also contain flattened root filesystem
image files (*.ext4
), which you can use
with QEMU directly.
The pre-built root filesystem image files follow these naming conventions:
core-image-profile
-arch
.tar.bz2 Where:profile
is the filesystem image's profile: lsb, lsb-dev, lsb-sdk, minimal, minimal-dev, minimal-initramfs, sato, sato-dev, sato-sdk, sato-sdk-ptest. For information on these types of image profiles, see the "Images" chapter in the Yocto Project Reference Manual.arch
is a string representing the target architecture: beaglebone-yocto, beaglebone-yocto-lsb, edgerouter, edgerouter-lsb, genericx86, genericx86-64, genericx86-64-lsb, genericx86-lsb and qemu*.
The root filesystems provided by the Yocto Project are based
off of the core-image-sato
and
core-image-minimal
images.
For example, if you plan on using a BeagleBone device
as your target hardware and your image is a
core-image-sato-sdk
image, you can download the following file:
core-image-sato-sdk-beaglebone-yocto.tar.bz2
Initialize the Cross-Development Environment:
You must source
the cross-development
environment setup script to establish necessary environment
variables.
This script is located in the top-level directory in
which you installed the toolchain (e.g.
poky_sdk
).
Following is an example based on the toolchain installed in the "Locating Pre-Built SDK Installers" section:
$ source ~/poky_sdk/environment-setup-core2-64-poky-linux
Extract the Root Filesystem:
Use the runqemu-extract-sdk
command
and provide the root filesystem image.
Following is an example command that extracts the root
filesystem from a previously built root filesystem image that
was downloaded from the
Index of Releases.
This command extracts the root filesystem into the
core2-64-sato
directory:
$ runqemu-extract-sdk ~/Downloads/core-image-sato-sdk-beaglebone-yocto.tar.bz2 ~/beaglebone-sato
You could now point to the target sysroot at
beablebone-sato
.
The following figure shows the resulting directory structure after
you install the Standard SDK by running the *.sh
SDK installation script:
The installed SDK consists of an environment setup script for the SDK,
a configuration file for the target, a version file for the target,
and the root filesystem (sysroots
) needed to
develop objects for the target system.
Within the figure, italicized text is used to indicate replaceable
portions of the file or directory name.
For example,
install_dir
/version
is the directory where the SDK is installed.
By default, this directory is /opt/poky/
.
And, version
represents the specific
snapshot of the SDK (e.g. 3.1.2
).
Furthermore, target
represents the target
architecture (e.g. i586
) and
host
represents the development system's
architecture (e.g. x86_64
).
Thus, the complete names of the two directories within the
sysroots
could be
i586-poky-linux
and
x86_64-pokysdk-linux
for the target and host,
respectively.
The following figure shows the resulting directory structure after
you install the Extensible SDK by running the *.sh
SDK installation script:
The installed directory structure for the extensible SDK is quite different than the installed structure for the standard SDK. The extensible SDK does not separate host and target parts in the same manner as does the standard SDK. The extensible SDK uses an embedded copy of the OpenEmbedded build system, which has its own sysroots.
Of note in the directory structure are an environment setup script for the SDK, a configuration file for the target, a version file for the target, and log files for the OpenEmbedded build system preparation script run by the installer and BitBake.
Within the figure, italicized text is used to indicate replaceable
portions of the file or directory name.
For example,
install_dir
is the directory where the SDK
is installed, which is poky_sdk
by default, and
target
represents the target
architecture (e.g. i586
).
This appendix describes customizations you can apply to the extensible SDK.
The extensible SDK primarily consists of a pre-configured copy of
the OpenEmbedded build system from which it was produced.
Thus, the SDK's configuration is derived using that build system and
the filters shown in the following list.
When these filters are present, the OpenEmbedded build system applies
them against local.conf
and
auto.conf
:
Variables whose values start with "/" are excluded since the assumption is that those values are paths that are likely to be specific to the build host.
Variables listed in
SDK_LOCAL_CONF_BLACKLIST
are excluded.
These variables are not allowed through from the OpenEmbedded
build system configuration into the extensible SDK
configuration.
Typically, these variables are specific to the machine on
which the build system is running and could be problematic
as part of the extensible SDK configuration.
For a list of the variables excluded by default, see the
SDK_LOCAL_CONF_BLACKLIST
in the glossary of the Yocto Project Reference Manual.
Variables listed in
SDK_LOCAL_CONF_WHITELIST
are included.
Including a variable in the value of
SDK_LOCAL_CONF_WHITELIST
overrides either
of the previous two filters.
The default value is blank.
Classes inherited globally with
INHERIT
that are listed in
SDK_INHERIT_BLACKLIST
are disabled.
Using SDK_INHERIT_BLACKLIST
to disable
these classes is the typical method to disable classes that
are problematic or unnecessary in the SDK context.
The default value blacklists the
buildhistory
and
icecc
classes.
Additionally, the contents of conf/sdk-extra.conf
,
when present, are appended to the end of
conf/local.conf
within the produced SDK, without
any filtering.
The sdk-extra.conf
file is particularly useful
if you want to set a variable value just for the SDK and not the
OpenEmbedded build system used to create the SDK.
In most cases, the extensible SDK defaults should work with your build host's setup. However, some cases exist for which you might consider making adjustments:
If your SDK configuration inherits additional classes
using the
INHERIT
variable and you do not need or want those classes enabled in
the SDK, you can blacklist them by adding them to the
SDK_INHERIT_BLACKLIST
variable as described in the fourth bullet of the previous
section.
SDK_INHERIT_BLACKLIST
is set using
the "?=" operator.
Consequently, you will need to either define the entire
list by using the "=" operator, or you will need to append
a value using either "_append" or the "+=" operator.
You can learn more about these operators in the
"Basic Syntax"
section of the BitBake User Manual.
.
If you have classes or recipes that add additional tasks to the standard build flow (i.e. the tasks execute as the recipe builds as opposed to being called explicitly), then you need to do one of the following:
After ensuring the tasks are
shared state
tasks (i.e. the output of the task is saved to and
can be restored from the shared state cache) or
ensuring the tasks are able to be produced quickly from
a task that is a shared state task, add the task name
to the value of
SDK_RECRDEP_TASKS
.
Disable the tasks if they are added by a class and
you do not need the functionality the class provides
in the extensible SDK.
To disable the tasks, add the class to the
SDK_INHERIT_BLACKLIST
variable
as described in the previous section.
Generally, you want to have a shared state mirror set up so users of the SDK can add additional items to the SDK after installation without needing to build the items from source. See the "Providing Additional Installable Extensible SDK Content" section for information.
If you want users of the SDK to be able to easily update the
SDK, you need to set the
SDK_UPDATE_URL
variable.
For more information, see the
"Providing Updates to the Extensible SDK After Installation"
section.
If you have adjusted the list of files and directories that
appear in
COREBASE
(other than layers that are enabled through
bblayers.conf
), then you must list these
files in
COREBASE_FILES
so that the files are copied into the SDK.
If your OpenEmbedded build system setup uses a different
environment setup script other than
oe-init-build-env
,
then you must set
OE_INIT_ENV_SCRIPT
to point to the environment setup script you use.
COREBASE_FILES
variable as previously
described.
You can change the displayed title for the SDK installer by setting
the
SDK_TITLE
variable and then rebuilding the the SDK installer.
For information on how to build an SDK installer, see the
"Building an SDK Installer"
section.
By default, this title is derived from
DISTRO_NAME
when it is set.
If the DISTRO_NAME
variable is not set, the title
is derived from the
DISTRO
variable.
The
populate_sdk_base
class defines the default value of the SDK_TITLE
variable as follows:
SDK_TITLE ??= "${@d.getVar('DISTRO_NAME') or d.getVar('DISTRO')} SDK"
While several ways exist to change this variable, an efficient method
is to set the variable in your distribution's configuration file.
Doing so creates an SDK installer title that applies across your
distribution.
As an example, assume you have your own layer for your distribution
named "meta-mydistro" and you are using the same type of file
hierarchy as does the default "poky" distribution.
If so, you could update the SDK_TITLE
variable
in the
~/meta-mydistro/conf/distro/mydistro.conf
file
using the following form:
SDK_TITLE = "your_title
"
When you make changes to your configuration or to the metadata and
if you want those changes to be reflected in installed SDKs, you need
to perform additional steps.
These steps make it possible for anyone using the installed SDKs to
update the installed SDKs by using the
devtool sdk-update
command:
Create a directory that can be shared over HTTP or HTTPS. You can do this by setting up a web server such as an Apache HTTP Server or Nginx server in the cloud to host the directory. This directory must contain the published SDK.
Set the
SDK_UPDATE_URL
variable to point to the corresponding HTTP or HTTPS URL.
Setting this variable causes any SDK built to default to that
URL and thus, the user does not have to pass the URL to the
devtool sdk-update
command as described
in the
"Applying Updates to an Installed Extensible SDK"
section.
Build the extensible SDK normally (i.e., use the
bitbake -c populate_sdk_ext
imagename
command).
Publish the SDK using the following command:
$ oe-publish-sdksome_path
/sdk-installer.shpath_to_shared_http_directory
You must repeat this step each time you rebuild the SDK with changes that you want to make available through the update mechanism.
Completing the above steps allows users of the existing installed
SDKs to simply run devtool sdk-update
to
retrieve and apply the latest updates.
See the
"Applying Updates to an Installed Extensible SDK"
section for further information.
When you build the installer for the Extensible SDK, the default
installation directory for the SDK is based on the
DISTRO
and
SDKEXTPATH
variables from within the
populate_sdk_base
class as follows:
SDKEXTPATH ??= "~/${@d.getVar('DISTRO')}_sdk"
You can change this default installation directory by specifically
setting the SDKEXTPATH
variable.
While a number of ways exist through which you can set this variable,
the method that makes the most sense is to set the variable in your
distribution's configuration file.
Doing so creates an SDK installer default directory that applies
across your distribution.
As an example, assume you have your own layer for your distribution
named "meta-mydistro" and you are using the same type of file
hierarchy as does the default "poky" distribution.
If so, you could update the SDKEXTPATH
variable
in the
~/meta-mydistro/conf/distro/mydistro.conf
file
using the following form:
SDKEXTPATH = "some_path_for_your_installed_sdk
"
After building your installer, running it prompts the user for
acceptance of the
some_path_for_your_installed_sdk
directory
as the default location to install the Extensible SDK.
If you want the users of an extensible SDK you build to be able to add items to the SDK without requiring the users to build the items from source, you need to do a number of things:
Ensure the additional items you want the user to be able to install are already built:
Build the items explicitly. You could use one or more "meta" recipes that depend on lists of other recipes.
Build the "world" target and set
EXCLUDE_FROM_WORLD_pn-
recipename
for the recipes you do not want built.
See the
EXCLUDE_FROM_WORLD
variable for additional information.
Expose the sstate-cache
directory
produced by the build.
Typically, you expose this directory by making it available
through an
Apache HTTP Server
or
Nginx
server.
Set the appropriate configuration so that the produced SDK
knows how to find the configuration.
The variable you need to set is
SSTATE_MIRRORS
:
SSTATE_MIRRORS = "file://.* http://example
.com/some_path
/sstate-cache/PATH"
You can set the SSTATE_MIRRORS
variable
in two different places:
If the mirror value you are setting is appropriate to
be set for both the OpenEmbedded build system that is
actually building the SDK and the SDK itself (i.e. the
mirror is accessible in both places or it will fail
quickly on the OpenEmbedded build system side, and its
contents will not interfere with the build), then you
can set the variable in your
local.conf
or custom distro
configuration file.
You can then "whitelist" the variable through
to the SDK by adding the following:
SDK_LOCAL_CONF_WHITELIST = "SSTATE_MIRRORS"
Alternatively, if you just want to set the
SSTATE_MIRRORS
variable's value
for the SDK alone, create a
conf/sdk-extra.conf
file either in
your
Build Directory
or within any layer and put your
SSTATE_MIRRORS
setting within
that file.
SSTATE_MIRRORS
.
By default, the extensible SDK bundles the shared state artifacts for
everything needed to reconstruct the image for which the SDK was built.
This bundling can lead to an SDK installer file that is a Gigabyte or
more in size.
If the size of this file causes a problem, you can build an SDK that
has just enough in it to install and provide access to the
devtool command
by setting the following in your
configuration:
SDK_EXT_TYPE = "minimal"
Setting
SDK_EXT_TYPE
to "minimal" produces an SDK installer that is around 35 Mbytes in
size, which downloads and installs quickly.
You need to realize, though, that the minimal installer does not
install any libraries or tools out of the box.
These libraries and tools must be installed either "on the fly" or
through actions you perform using devtool
or
explicitly with the devtool sdk-install
command.
In most cases, when building a minimal SDK you need to also enable
bringing in the information on a wider range of packages produced by
the system.
Requiring this wider range of information is particularly true
so that devtool add
is able to effectively map
dependencies it discovers in a source tree to the appropriate recipes.
Additionally, the information enables the
devtool search
command to return useful results.
To facilitate this wider range of information, you would need to set the following:
SDK_INCLUDE_PKGDATA = "1"
See the
SDK_INCLUDE_PKGDATA
variable for additional information.
Setting the SDK_INCLUDE_PKGDATA
variable as
shown causes the "world" target to be built so that information
for all of the recipes included within it are available.
Having these recipes available increases build time significantly and
increases the size of the SDK installer by 30-80 Mbytes depending on
how many recipes are included in your configuration.
You can use
EXCLUDE_FROM_WORLD_pn-
recipename
for recipes you want to exclude.
However, it is assumed that you would need to be building the "world"
target if you want to provide additional items to the SDK.
Consequently, building for "world" should not represent undue
overhead in most cases.
SDK_EXT_TYPE
to "minimal",
then providing a shared state mirror is mandatory so that items
can be installed as needed.
See the
"Providing Additional Installable Extensible SDK Content"
section for more information.
You can explicitly control whether or not to include the toolchain
when you build an SDK by setting the
SDK_INCLUDE_TOOLCHAIN
variable to "1".
In particular, it is useful to include the toolchain when you
have set SDK_EXT_TYPE
to "minimal", which by
default, excludes the toolchain.
Also, it is helpful if you are building a small SDK for use with
an IDE or some
other tool where you do not want to take extra steps to install a
toolchain.
This appendix presents customizations you can apply to the standard SDK.
When you build a standard SDK using the
bitbake -c populate_sdk
, a default set of
packages is included in the resulting SDK.
The
TOOLCHAIN_HOST_TASK
and
TOOLCHAIN_TARGET_TASK
variables control the set of packages adding to the SDK.
If you want to add individual packages to the toolchain that runs on
the host, simply add those packages to the
TOOLCHAIN_HOST_TASK
variable.
Similarly, if you want to add packages to the default set that is
part of the toolchain that runs on the target, add the packages to the
TOOLCHAIN_TARGET_TASK
variable.
You can include API documentation as well as any other
documentation provided by recipes with the standard SDK by
adding "api-documentation" to the
DISTRO_FEATURES
variable:
DISTRO_FEATURES_append = " api-documentation"
Setting this variable as shown here causes the OpenEmbedded build system to build the documentation and then include it in the standard SDK.
bitbake-layers
ScriptA Board Support Package (BSP) is a collection of information that defines how to support a particular hardware device, set of devices, or hardware platform. The BSP includes information about the hardware features present on the device and kernel configuration information along with any additional hardware drivers required. The BSP also lists any additional software components required in addition to a generic Linux software stack for both essential and optional platform features.
This guide presents information about BSP layers, defines a structure for components
so that BSPs follow a commonly understood layout, discusses how to customize
a recipe for a BSP, addresses BSP licensing, and provides information that
shows you how to create a
BSP Layer using the
bitbake-layers
tool.
A BSP consists of a file structure inside a base directory. Collectively, you can think of the base directory, its file structure, and the contents as a BSP layer. Although not a strict requirement, BSP layers in the Yocto Project use the following well-established naming convention:
meta-bsp_root_name
The string "meta-" is prepended to the machine or platform name, which is
bsp_root_name
in the above form.
meta-
.
However, various scripts and tools in the Yocto Project
development environment assume this convention.
To help understand the BSP layer concept, consider the BSPs that the Yocto Project supports and provides with each release. You can see the layers in the Yocto Project Source Repositories through a web interface at http://git.yoctoproject.org. If you go to that interface, you will find a list of repositories under "Yocto Metadata Layers".
Each repository is a BSP layer supported by the Yocto Project
(e.g. meta-raspberrypi
and
meta-intel
).
Each of these layers is a repository unto itself and clicking on
the layer name displays two URLs from which you can
clone the layer's repository to your local system.
Here is an example that clones the Raspberry Pi BSP layer:
$ git clone git://git.yoctoproject.org/meta-raspberrypi
In addition to BSP layers, the
meta-yocto-bsp
layer is part of the
shipped poky
repository.
The meta-yocto-bsp
layer maintains several
"reference" BSPs including the ARM-based Beaglebone, MIPS-based
EdgeRouter, and generic versions of
both 32-bit and 64-bit IA machines.
For information on typical BSP development workflow, see the "Developing a Board Support Package (BSP)" section. For more information on how to set up a local copy of source files from a Git repository, see the "Locating Yocto Project Source Files" section in the Yocto Project Development Tasks Manual.
The BSP layer's base directory
(meta-
)
is the root directory of that Layer.
This directory is what you add to the
bsp_root_name
BBLAYERS
variable in the conf/bblayers.conf
file found in your
Build Directory,
which is established after you run the OpenEmbedded build environment
setup script (i.e.
oe-init-build-env
).
Adding the root directory allows the
OpenEmbedded build system
to recognize the BSP layer and from it build an image.
Here is an example:
BBLAYERS ?= " \ /usr/local/src/yocto/meta \ /usr/local/src/yocto/meta-poky \ /usr/local/src/yocto/meta-yocto-bsp \ /usr/local/src/yocto/meta-mylayer \ "
BBFILE_PRIORITY
for the layers listed in BBLAYERS
matter.
For example, if multiple layers define a machine
configuration, the OpenEmbedded build system uses
the last layer searched given similar layer
priorities.
The build system works from the top-down through
the layers listed in BBLAYERS
.
Some BSPs require or depend on additional layers
beyond the BSP's root layer in order to be functional.
In this case, you need to specify these layers in the
README
"Dependencies" section of the
BSP's root layer.
Additionally, if any build instructions exist for the
BSP, you must add them to the "Dependencies" section.
Some layers function as a layer to hold other BSP layers.
These layers are knows as
"container layers".
An example of this type of layer is OpenEmbedded's
meta-openembedded
layer.
The meta-openembedded
layer contains
many meta-*
layers.
In cases like this, you need to include the names of the actual
layers you want to work with, such as:
BBLAYERS ?= " \ /usr/local/src/yocto/meta \ /usr/local/src/yocto/meta-poky \ /usr/local/src/yocto/meta-yocto-bsp \ /usr/local/src/yocto/meta-mylayer \ .../meta-openembedded/meta-oe \ .../meta-openembedded/meta-perl \ .../meta-openembedded/meta-networking \ "
and so on.
For more information on layers, see the "Understanding and Creating Layers" section of the Yocto Project Development Tasks Manual.
This section describes how to get your build host ready
to work with BSP layers.
Once you have the host set up, you can create the layer
as described in the
"Creating a new BSP Layer Using the bitbake-layers
Script"
section.
Set Up the Build Environment: Be sure you are set up to use BitBake in a shell. See the "Preparing the Build Host" section in the Yocto Project Development Tasks Manual for information on how to get a build host ready that is either a native Linux machine or a machine that uses CROPS.
Clone the poky
Repository:
You need to have a local copy of the Yocto Project
Source Directory
(i.e. a local poky
repository).
See the
"Cloning the poky
Repository"
and possibly the
"Checking Out by Branch in Poky"
or
"Checking Out by Tag in Poky"
sections all in the Yocto Project Development Tasks Manual for
information on how to clone the poky
repository and check out the appropriate branch for your work.
Determine the BSP Layer You Want: The Yocto Project supports many BSPs, which are maintained in their own layers or in layers designed to contain several BSPs. To get an idea of machine support through BSP layers, you can look at the index of machines for the release.
Optionally Clone the
meta-intel
BSP Layer:
If your hardware is based on current Intel CPUs and devices,
you can leverage this BSP layer.
For details on the meta-intel
BSP layer,
see the layer's
README
file.
Navigate to Your Source Directory:
Typically, you set up the
meta-intel
Git repository
inside the
Source Directory
(e.g. poky
).
$ cd /home/you
/poky
Clone the Layer:
$ git clone git://git.yoctoproject.org/meta-intel.git Cloning into 'meta-intel'... remote: Counting objects: 15585, done. remote: Compressing objects: 100% (5056/5056), done. remote: Total 15585 (delta 9123), reused 15329 (delta 8867) Receiving objects: 100% (15585/15585), 4.51 MiB | 3.19 MiB/s, done. Resolving deltas: 100% (9123/9123), done. Checking connectivity... done.
Check Out the Proper Branch:
The branch you check out for
meta-intel
must match the same
branch you are using for the Yocto Project release
(e.g. dunfell):
$ cd meta-intel $ git checkout -b dunfell remotes/origin/dunfell Branch dunfell set up to track remote branch dunfell from origin. Switched to a new branch 'dunfell'
git branch -al
command.
See the
"Checking Out By Branch in Poky"
section in the Yocto Project Development Tasks
Manual for more information.
Optionally Set Up an Alternative BSP Layer:
If your hardware can be more closely leveraged to an
existing BSP not within the meta-intel
BSP layer, you can clone that BSP layer.
The process is identical to the process used for the
meta-intel
layer except for the layer's
name.
For example, if you determine that your hardware most
closely matches the meta-raspberrypi
,
clone that layer:
$ git clone git://git.yoctoproject.org/meta-raspberrypi Cloning into 'meta-raspberrypi'... remote: Counting objects: 4743, done. remote: Compressing objects: 100% (2185/2185), done. remote: Total 4743 (delta 2447), reused 4496 (delta 2258) Receiving objects: 100% (4743/4743), 1.18 MiB | 0 bytes/s, done. Resolving deltas: 100% (2447/2447), done. Checking connectivity... done.
Initialize the Build Environment:
While in the root directory of the Source Directory (i.e.
poky
), run the
oe-init-build-env
environment setup script to define the OpenEmbedded
build environment on your build host.
$ source oe-init-build-env
Among other things, the script creates the
Build Directory,
which is build
in this case
and is located in the
Source Directory.
After the script runs, your current working directory
is set to the build
directory.
Defining a common BSP directory structure allows end-users to understand and become familiar with that standard. A common format also encourages standardization of software support for hardware.
The proposed form described in this section does have elements that are specific to the OpenEmbedded build system. It is intended that developers can use this structure with other build systems besides the OpenEmbedded build system. It is also intended that it will be be simple to extract information and convert it to other formats if required. The OpenEmbedded build system, through its standard layers mechanism, can directly accept the format described as a layer. The BSP layer captures all the hardware-specific details in one place using a standard format, which is useful for any person wishing to use the hardware platform regardless of the build system they are using.
The BSP specification does not include a build system or other tools - the specification is concerned with the hardware-specific components only. At the end-distribution point, you can ship the BSP layer combined with a build system and other tools. Realize that it is important to maintain the distinction that the BSP layer, a build system, and tools are separate components that could be combined in certain end products.
Before looking at the recommended form for the directory structure inside a BSP layer, you should be aware that some requirements do exist in order for a BSP layer to be considered compliant with the Yocto Project. For that list of requirements, see the "Released BSP Requirements" section.
Below is the typical directory structure for a BSP layer. While this basic form represents the standard, realize that the actual layout for individual BSPs could differ.
meta-bsp_root_name
/ meta-bsp_root_name
/bsp_license_file
meta-bsp_root_name
/README meta-bsp_root_name
/README.sources meta-bsp_root_name
/binary/bootable_images
meta-bsp_root_name
/conf/layer.conf meta-bsp_root_name
/conf/machine/*.conf meta-bsp_root_name
/recipes-bsp/* meta-bsp_root_name
/recipes-core/* meta-bsp_root_name
/recipes-graphics/* meta-bsp_root_name
/recipes-kernel/linux/linux-yocto_kernel_rev
.bbappend
Below is an example of the Raspberry Pi BSP layer that is available from the Source Respositories:
meta-raspberrypi/COPYING.MIT meta-raspberrypi/README.md meta-raspberrypi/classes meta-raspberrypi/classes/sdcard_image-rpi.bbclass meta-raspberrypi/conf/ meta-raspberrypi/conf/layer.conf meta-raspberrypi/conf/machine/ meta-raspberrypi/conf/machine/raspberrypi-cm.conf meta-raspberrypi/conf/machine/raspberrypi-cm3.conf meta-raspberrypi/conf/machine/raspberrypi.conf meta-raspberrypi/conf/machine/raspberrypi0-wifi.conf meta-raspberrypi/conf/machine/raspberrypi0.conf meta-raspberrypi/conf/machine/raspberrypi2.conf meta-raspberrypi/conf/machine/raspberrypi3-64.conf meta-raspberrypi/conf/machine/raspberrypi3.conf meta-raspberrypi/conf/machine/include meta-raspberrypi/conf/machine/include/rpi-base.inc meta-raspberrypi/conf/machine/include/rpi-default-providers.inc meta-raspberrypi/conf/machine/include/rpi-default-settings.inc meta-raspberrypi/conf/machine/include/rpi-default-versions.inc meta-raspberrypi/conf/machine/include/tune-arm1176jzf-s.inc meta-raspberrypi/docs meta-raspberrypi/docs/Makefile meta-raspberrypi/docs/conf.py meta-raspberrypi/docs/contributing.md meta-raspberrypi/docs/extra-apps.md meta-raspberrypi/docs/extra-build-config.md meta-raspberrypi/docs/index.rst meta-raspberrypi/docs/layer-contents.md meta-raspberrypi/docs/readme.md meta-raspberrypi/files meta-raspberrypi/files/custom-licenses meta-raspberrypi/files/custom-licenses/Broadcom meta-raspberrypi/recipes-bsp meta-raspberrypi/recipes-bsp/bootfiles meta-raspberrypi/recipes-bsp/bootfiles/bcm2835-bootfiles.bb meta-raspberrypi/recipes-bsp/bootfiles/rpi-config_git.bb meta-raspberrypi/recipes-bsp/common meta-raspberrypi/recipes-bsp/common/firmware.inc meta-raspberrypi/recipes-bsp/formfactor meta-raspberrypi/recipes-bsp/formfactor/formfactor meta-raspberrypi/recipes-bsp/formfactor/formfactor/raspberrypi meta-raspberrypi/recipes-bsp/formfactor/formfactor/raspberrypi/machconfig meta-raspberrypi/recipes-bsp/formfactor/formfactor_0.0.bbappend meta-raspberrypi/recipes-bsp/rpi-u-boot-src meta-raspberrypi/recipes-bsp/rpi-u-boot-src/files meta-raspberrypi/recipes-bsp/rpi-u-boot-src/files/boot.cmd.in meta-raspberrypi/recipes-bsp/rpi-u-boot-src/rpi-u-boot-scr.bb meta-raspberrypi/recipes-bsp/u-boot meta-raspberrypi/recipes-bsp/u-boot/u-boot meta-raspberrypi/recipes-bsp/u-boot/u-boot/*.patch meta-raspberrypi/recipes-bsp/u-boot/u-boot_%.bbappend meta-raspberrypi/recipes-connectivity meta-raspberrypi/recipes-connectivity/bluez5 meta-raspberrypi/recipes-connectivity/bluez5/bluez5 meta-raspberrypi/recipes-connectivity/bluez5/bluez5/*.patch meta-raspberrypi/recipes-connectivity/bluez5/bluez5/BCM43430A1.hcd meta-raspberrypi/recipes-connectivity/bluez5/bluez5brcm43438.service meta-raspberrypi/recipes-connectivity/bluez5/bluez5_%.bbappend meta-raspberrypi/recipes-core meta-raspberrypi/recipes-core/images meta-raspberrypi/recipes-core/images/rpi-basic-image.bb meta-raspberrypi/recipes-core/images/rpi-hwup-image.bb meta-raspberrypi/recipes-core/images/rpi-test-image.bb meta-raspberrypi/recipes-core/packagegroups meta-raspberrypi/recipes-core/packagegroups/packagegroup-rpi-test.bb meta-raspberrypi/recipes-core/psplash meta-raspberrypi/recipes-core/psplash/files meta-raspberrypi/recipes-core/psplash/files/psplash-raspberrypi-img.h meta-raspberrypi/recipes-core/psplash/psplash_git.bbappend meta-raspberrypi/recipes-core/udev meta-raspberrypi/recipes-core/udev/udev-rules-rpi meta-raspberrypi/recipes-core/udev/udev-rules-rpi/99-com.rules meta-raspberrypi/recipes-core/udev/udev-rules-rpi.bb meta-raspberrypi/recipes-devtools meta-raspberrypi/recipes-devtools/bcm2835 meta-raspberrypi/recipes-devtools/bcm2835/bcm2835_1.52.bb meta-raspberrypi/recipes-devtools/pi-blaster meta-raspberrypi/recipes-devtools/pi-blaster/files meta-raspberrypi/recipes-devtools/pi-blaster/files/*.patch meta-raspberrypi/recipes-devtools/pi-blaster/pi-blaster_git.bb meta-raspberrypi/recipes-devtools/python meta-raspberrypi/recipes-devtools/python/python-rtimu meta-raspberrypi/recipes-devtools/python/python-rtimu/*.patch meta-raspberrypi/recipes-devtools/python/python-rtimu_git.bb meta-raspberrypi/recipes-devtools/python/python-sense-hat_2.2.0.bb meta-raspberrypi/recipes-devtools/python/rpi-gpio meta-raspberrypi/recipes-devtools/python/rpi-gpio/*.patch meta-raspberrypi/recipes-devtools/python/rpi-gpio_0.6.3.bb meta-raspberrypi/recipes-devtools/python/rpio meta-raspberrypi/recipes-devtools/python/rpio/*.patch meta-raspberrypi/recipes-devtools/python/rpio_0.10.0.bb meta-raspberrypi/recipes-devtools/wiringPi meta-raspberrypi/recipes-devtools/wiringPi/files meta-raspberrypi/recipes-devtools/wiringPi/files/*.patch meta-raspberrypi/recipes-devtools/wiringPi/wiringpi_git.bb meta-raspberrypi/recipes-graphics meta-raspberrypi/recipes-graphics/eglinfo meta-raspberrypi/recipes-graphics/eglinfo/eglinfo-fb_%.bbappend meta-raspberrypi/recipes-graphics/eglinfo/eglinfo-x11_%.bbappend meta-raspberrypi/recipes-graphics/mesa meta-raspberrypi/recipes-graphics/mesa/mesa-gl_%.bbappend meta-raspberrypi/recipes-graphics/mesa/mesa_%.bbappend meta-raspberrypi/recipes-graphics/userland meta-raspberrypi/recipes-graphics/userland/userland meta-raspberrypi/recipes-graphics/userland/userland/*.patch meta-raspberrypi/recipes-graphics/userland/userland_git.bb meta-raspberrypi/recipes-graphics/vc-graphics meta-raspberrypi/recipes-graphics/vc-graphics/files meta-raspberrypi/recipes-graphics/vc-graphics/files/egl.pc meta-raspberrypi/recipes-graphics/vc-graphics/files/vchiq.sh meta-raspberrypi/recipes-graphics/vc-graphics/vc-graphics-hardfp.bb meta-raspberrypi/recipes-graphics/vc-graphics/vc-graphics.bb meta-raspberrypi/recipes-graphics/vc-graphics/vc-graphics.inc meta-raspberrypi/recipes-graphics/wayland meta-raspberrypi/recipes-graphics/wayland/weston_%.bbappend meta-raspberrypi/recipes-graphics/xorg-xserver meta-raspberrypi/recipes-graphics/xorg-xserver/xserver-xf86-config meta-raspberrypi/recipes-graphics/xorg-xserver/xserver-xf86-config/rpi meta-raspberrypi/recipes-graphics/xorg-xserver/xserver-xf86-config/rpi/xorg.conf meta-raspberrypi/recipes-graphics/xorg-xserver/xserver-xf86-config/rpi/xorg.conf.d meta-raspberrypi/recipes-graphics/xorg-xserver/xserver-xf86-config/rpi/xorg.conf.d/10-evdev.conf meta-raspberrypi/recipes-graphics/xorg-xserver/xserver-xf86-config/rpi/xorg.conf.d/98-pitft.conf meta-raspberrypi/recipes-graphics/xorg-xserver/xserver-xf86-config/rpi/xorg.conf.d/99-calibration.conf meta-raspberrypi/recipes-graphics/xorg-xserver/xserver-xf86-config_0.1.bbappend meta-raspberrypi/recipes-graphics/xorg-xserver/xserver-xorg_%.bbappend meta-raspberrypi/recipes-kernel meta-raspberrypi/recipes-kernel/linux-firmware meta-raspberrypi/recipes-kernel/linux-firmware/files meta-raspberrypi/recipes-kernel/linux-firmware/files/brcmfmac43430-sdio.bin meta-raspberrypi/recipes-kernel/linux-firmware/files/brcfmac43430-sdio.txt meta-raspberrypi/recipes-kernel/linux-firmware/linux-firmware_%.bbappend meta-raspberrypi/recipes-kernel/linux meta-raspberrypi/recipes-kernel/linux/linux-raspberrypi-dev.bb meta-raspberrypi/recipes-kernel/linux/linux-raspberrypi.inc meta-raspberrypi/recipes-kernel/linux/linux-raspberrypi_4.14.bb meta-raspberrypi/recipes-kernel/linux/linux-raspberrypi_4.9.bb meta-raspberrypi/recipes-multimedia meta-raspberrypi/recipes-multimedia/gstreamer meta-raspberrypi/recipes-multimedia/gstreamer/gstreamer1.0-omx meta-raspberrypi/recipes-multimedia/gstreamer/gstreamer1.0-omx/*.patch meta-raspberrypi/recipes-multimedia/gstreamer/gstreamer1.0-omx_%.bbappend meta-raspberrypi/recipes-multimedia/gstreamer/gstreamer1.0-plugins-bad_%.bbappend meta-raspberrypi/recipes-multimedia/gstreamer/gstreamer1.0-omx-1.12 meta-raspberrypi/recipes-multimedia/gstreamer/gstreamer1.0-omx-1.12/*.patch meta-raspberrypi/recipes-multimedia/omxplayer meta-raspberrypi/recipes-multimedia/omxplayer/omxplayer meta-raspberrypi/recipes-multimedia/omxplayer/omxplayer/*.patch meta-raspberrypi/recipes-multimedia/omxplayer/omxplayer_git.bb meta-raspberrypi/recipes-multimedia/x264 meta-raspberrypi/recipes-multimedia/x264/x264_git.bbappend meta-raspberrypi/wic meta-raspberrypi/wic/sdimage-raspberrypi.wks
The following sections describe each part of the proposed BSP format.
You can find these files in the BSP Layer at:
meta-bsp_root_name
/bsp_license_file
These optional files satisfy licensing requirements
for the BSP.
The type or types of files here can vary depending
on the licensing requirements.
For example, in the Raspberry Pi BSP, all licensing
requirements are handled with the
COPYING.MIT
file.
Licensing files can be MIT, BSD, GPLv*, and so forth. These files are recommended for the BSP but are optional and totally up to the BSP developer. For information on how to maintain license compliance, see the "Maintaining Open Source License Compliance During Your Product's Lifecycle" section in the Yocto Project Development Tasks Manual.
You can find this file in the BSP Layer at:
meta-bsp_root_name
/README
This file provides information on how to boot the live
images that are optionally included in the
binary/
directory.
The README
file also provides
information needed for building the image.
At a minimum, the README
file must
contain a list of dependencies, such as the names of
any other layers on which the BSP depends and the name of
the BSP maintainer with his or her contact information.
You can find this file in the BSP Layer at:
meta-bsp_root_name
/README.sources
This file provides information on where to locate the BSP
source files used to build the images (if any) that
reside in
meta-
.
Images in the bsp_root_name
/binarybinary
would be images
released with the BSP.
The information in the README.sources
file also helps you find the
Metadata
used to generate the images that ship with the BSP.
binary
directory is
missing or the directory has no images, an existing
README.sources
file is
meaningless and usually does not exist.
You can find these files in the BSP Layer at:
meta-bsp_root_name
/binary/bootable_images
This optional area contains useful pre-built kernels and user-space filesystem images released with the BSP that are appropriate to the target system. This directory typically contains graphical (e.g. Sato) and minimal live images when the BSP tarball has been created and made available in the Yocto Project website. You can use these kernels and images to get a system running and quickly get started on development tasks.
The exact types of binaries present are highly
hardware-dependent.
The
README
file should be present in the BSP Layer and it
explains how to use the images with the target hardware.
Additionally, the
README.sources
file should be present to locate the sources used to
build the images and provide information on the
Metadata.
You can find this file in the BSP Layer at:
meta-bsp_root_name
/conf/layer.conf
The conf/layer.conf
file
identifies the file structure as a layer,
identifies the contents of the layer, and
contains information about how the build system should
use it.
Generally, a standard boilerplate file such as the
following works.
In the following example, you would replace
bsp
with the actual
name of the BSP (i.e.
bsp_root_name
from the example
template).
# We have a conf and classes directory, add to BBPATH BBPATH .= ":${LAYERDIR}" # We have a recipes directory, add to BBFILES BBFILES += "${LAYERDIR}/recipes-*/*/*.bb \ ${LAYERDIR}/recipes-*/*/*.bbappend" BBFILE_COLLECTIONS += "bsp
" BBFILE_PATTERN_bsp
= "^${LAYERDIR}/" BBFILE_PRIORITY_bsp
= "6" LAYERDEPENDS_bsp
= "intel"
To illustrate the string substitutions, here are
the corresponding statements from the Raspberry
Pi conf/layer.conf
file:
# We have a conf and classes directory, append to BBPATH BBPATH .= ":${LAYERDIR}" # We have a recipes directory containing .bb and .bbappend files, add to BBFILES BBFILES += "${LAYERDIR}/recipes*/*/*.bb \ ${LAYERDIR}/recipes*/*/*.bbappend" BBFILE_COLLECTIONS += "raspberrypi" BBFILE_PATTERN_raspberrypi := "^${LAYERDIR}/" BBFILE_PRIORITY_raspberrypi = "9" # Additional license directories. LICENSE_PATH += "${LAYERDIR}/files/custom-licenses" . . .
This file simply makes BitBake aware of the recipes and configuration directories. The file must exist so that the OpenEmbedded build system can recognize the BSP.
You can find these files in the BSP Layer at:
meta-bsp_root_name
/conf/machine/*.conf
The machine files bind together all the information
contained elsewhere in the BSP into a format that
the build system can understand.
Each BSP Layer requires at least one machine file.
If the BSP supports multiple machines, multiple
machine configuration files can exist.
These filenames correspond to the values to which
users have set the
MACHINE
variable.
These files define things such as the kernel package
to use
(PREFERRED_PROVIDER
of
virtual/kernel),
the hardware drivers to include in different types
of images, any special software components that are
needed, any bootloader information, and also any
special image format requirements.
This configuration file could also include a hardware "tuning" file that is commonly used to define the package architecture and specify optimization flags, which are carefully chosen to give best performance on a given processor.
Tuning files are found in the
meta/conf/machine/include
directory within the
Source Directory.
For example, many tune-*
files
(e.g. tune-arm1136jf-s.inc
,
tune-1586-nlp.inc
, and so forth)
reside in the
poky/meta/conf/machine/include
directory.
To use an include file, you simply include them in the
machine configuration file.
For example, the Raspberry Pi BSP
raspberrypi3.conf
contains the
following statement:
include conf/machine/include/rpi-base.inc
You can find these files in the BSP Layer at:
meta-bsp_root_name
/recipes-bsp/*
This optional directory contains miscellaneous recipe
files for the BSP.
Most notably would be the formfactor files.
For example, in the Raspberry Pi BSP, there is the
formfactor_0.0.bbappend
file,
which is an append file used to augment the recipe
that starts the build.
Furthermore, there are machine-specific settings used
during the build that are defined by the
machconfig
file further down in
the directory.
Here is the machconfig
file for
the Raspberry Pi BSP:
HAVE_TOUCHSCREEN=0 HAVE_KEYBOARD=1 DISPLAY_CAN_ROTATE=0 DISPLAY_ORIENTATION=0 DISPLAY_DPI=133
If a BSP does not have a formfactor entry, defaults
are established according to the formfactor
configuration file that is installed by the main
formfactor recipe
meta/recipes-bsp/formfactor/formfactor_0.0.bb
,
which is found in the
Source Directory.
You can find these files in the BSP Layer at:
meta-bsp_root_name
/recipes-graphics/*
This optional directory contains recipes for the BSP if it has special requirements for graphics support. All files that are needed for the BSP to support a display are kept here.
You can find these files in the BSP Layer at:
meta-bsp_root_name
/recipes-kernel/linux/linux*.bbappend meta-bsp_root_name
/recipes-kernel/linux/*.bb
Append files (*.bbappend
) modify
the main kernel recipe being used to build the image.
The *.bb
files would be a
developer-supplied kernel recipe.
This area of the BSP hierarchy can contain both these
types of files although, in practice, it is likely that
you would have one or the other.
For your BSP, you typically want to use an existing Yocto
Project kernel recipe found in the
Source Directory
at meta/recipes-kernel/linux
.
You can append machine-specific changes to the
kernel recipe by using a similarly named append
file, which is located in the BSP Layer for your
target device (e.g. the
meta-
directory).
bsp_root_name
/recipes-kernel/linux
Suppose you are using the
linux-yocto_4.4.bb
recipe to
build the kernel.
In other words, you have selected the kernel in your
bsp_root_name
.conf
file by adding
PREFERRED_PROVIDER
and
PREFERRED_VERSION
statements as follows:
PREFERRED_PROVIDER_virtual/kernel ?= "linux-yocto" PREFERRED_VERSION_linux-yocto ?= "4.4%"
PREFERRED_PROVIDER
statement does not appear in the
bsp_root_name
.conf
file.
You would use the
linux-yocto_4.4.bbappend
file to append specific BSP settings to the kernel,
thus configuring the kernel for your particular BSP.
You can find more information on what your append file should contain in the "Creating the Append File" section in the Yocto Project Linux Kernel Development Manual.
An alternate scenario is when you create your own
kernel recipe for the BSP.
A good example of this is the Raspberry Pi BSP.
If you examine the
recipes-kernel/linux
directory
you see the following:
linux-raspberrypi-dev.bb linux-raspberrypi.inc linux-raspberrypi_4.14.bb linux-raspberrypi_4.9.bb
The directory contains three kernel recipes and a common include file.
This section describes the high-level procedure you can
follow to create a BSP.
Although not required for BSP creation, the
meta-intel
repository, which
contains many BSPs supported by the Yocto Project,
is part of the example.
For an example that shows how to create a new
layer using the tools, see the
"Creating a New BSP Layer Using the bitbake-layers
Script"
section.
The following illustration and list summarize the BSP creation general workflow.
Set up Your Host Development System to Support Development Using the Yocto Project: See the "Preparing the Build Host" section in the Yocto Project Development Tasks Manual for options on how to get a system ready to use the Yocto Project.
Establish the
meta-intel
Repository on Your System:
Having local copies of these supported BSP layers
on your system gives you access to layers you
might be able to leverage when creating your BSP.
For information on how to get these files, see the
"Preparing Your Build Host to Work with BSP Layers"
section.
Create Your Own BSP Layer Using the
bitbake-layers
Script:
Layers are ideal for isolating and storing work
for a given piece of hardware.
A layer is really just a location or area in which you
place the recipes and configurations for your BSP.
In fact, a BSP is, in itself, a special type of layer.
The simplest way to create a new BSP layer that is
compliant with the Yocto Project is to use the
bitbake-layers
script.
For information about that script, see the
"Creating a New BSP Layer Using the bitbake-layers
Script"
section.
Another example that illustrates a layer is an application. Suppose you are creating an application that has library or other dependencies in order for it to compile and run. The layer, in this case, would be where all the recipes that define those dependencies are kept. The key point for a layer is that it is an isolated area that contains all the relevant information for the project that the OpenEmbedded build system knows about. For more information on layers, see the "The Yocto Project Layer Model" section in the Yocto Project Overview and Concepts Manual. You can also reference the "Understanding and Creating Layers" section in the Yocto Project Development Tasks Manual. For more information on BSP layers, see the "BSP Layers" section.
Five hardware reference BSPs exist
that are part of the Yocto Project release
and are located in the
poky/meta-yocto-bsp
BSP
layer:
Texas Instruments Beaglebone
(beaglebone-yocto
)
Ubiquiti Networks EdgeRouter Lite
(edgerouter
)
Two general IA platforms
(genericx86
and
genericx86-64
)
Three core Intel BSPs exist as part of
the Yocto Project release in the
meta-intel
layer:
intel-core2-32
,
which is a BSP optimized for the Core2
family of CPUs as well as all CPUs
prior to the Silvermont core.
intel-corei7-64
,
which is a BSP optimized for Nehalem
and later Core and Xeon CPUs as well
as Silvermont and later Atom CPUs,
such as the Baytrail SoCs.
intel-quark
,
which is a BSP optimized for the
Intel Galileo gen1 & gen2
development boards.
When you set up a layer for a new BSP,
you should follow a standard layout.
This layout is described in the
"Example Filesystem Layout"
section.
In the standard layout, notice the suggested
structure for recipes and configuration
information.
You can see the standard layout for a BSP
by examining any supported BSP found in the
meta-intel
layer inside
the Source Directory.
Make Configuration Changes to Your New
BSP Layer:
The standard BSP layer structure organizes the
files you need to edit in
conf
and several
recipes-*
directories
within the BSP layer.
Configuration changes identify where your new
layer is on the local system and identifies the
kernel you are going to use.
When you run the
bitbake-layers
script,
you are able to interactively configure many
things for the BSP (e.g. keyboard, touchscreen,
and so forth).
Make Recipe Changes to Your New BSP
Layer:
Recipe changes include altering recipes
(*.bb
files), removing
recipes you do not use, and adding new recipes
or append files (.bbappend
)
that support your hardware.
Prepare for the Build:
Once you have made all the changes to your BSP
layer, there remains a few things you need to
do for the OpenEmbedded build system in order
for it to create your image.
You need to get the build environment ready by
sourcing an environment setup script
(i.e. oe-init-build-env
)
and you need to be sure two key configuration
files are configured appropriately: the
conf/local.conf
and the
conf/bblayers.conf
file.
You must make the OpenEmbedded build system aware
of your new layer.
See the
"Enabling Your Layer"
section in the Yocto Project Development Tasks Manual
for information on how to let the build system
know about your new layer.
Build the Image: The OpenEmbedded build system uses the BitBake tool to build images based on the type of image you want to create. You can find more information about BitBake in the BitBake User Manual.
The build process supports several types of images to satisfy different needs. See the "Images" chapter in the Yocto Project Reference Manual for information on supported images.
Certain requirements exist for a released BSP to be considered compliant with the Yocto Project. Additionally, recommendations also exist. This section describes the requirements and recommendations for released BSPs.
Before looking at BSP requirements, you should consider the following:
The requirements here assume the BSP layer is a well-formed, "legal" layer that can be added to the Yocto Project. For guidelines on creating a layer that meets these base requirements, see the "BSP Layers" section in this manual and the "Understanding and Creating Layers"" section in the Yocto Project Development Tasks Manual.
The requirements in this section apply regardless of how you package a BSP. You should consult the packaging and distribution guidelines for your specific release process. For an example of packaging and distribution requirements, see the "Third Party BSP Release Process" wiki page.
The requirements for the BSP as it is made available to a developer are completely independent of the released form of the BSP. For example, the BSP Metadata can be contained within a Git repository and could have a directory structure completely different from what appears in the officially released BSP layer.
It is not required that specific packages or package modifications exist in the BSP layer, beyond the requirements for general compliance with the Yocto Project. For example, no requirement exists dictating that a specific kernel or kernel version be used in a given BSP.
Following are the requirements for a released BSP that conform to the Yocto Project:
Layer Name: The BSP must have a layer name that follows the Yocto Project standards. For information on BSP layer names, see the "BSP Layers" section.
File System Layout:
When possible, use the same directory names
in your BSP layer as listed in the
recipes.txt
file, which
is found in poky/meta
directory of the
Source Directory
or in the OpenEmbedded-Core Layer
(openembedded-core
) at
http://git.openembedded.org/openembedded-core/tree/meta.
You should place recipes
(*.bb
files) and recipe
modifications (*.bbappend
files) into recipes-*
subdirectories by functional area as outlined
in recipes.txt
.
If you cannot find a category in
recipes.txt
to fit a
particular recipe, you can make up your own
recipes-*
subdirectory.
Within any particular
recipes-*
category, the
layout should match what is found in the
OpenEmbedded-Core Git repository
(openembedded-core
)
or the Source Directory (poky
).
In other words, make sure you place related
files in appropriately-related
recipes-*
subdirectories
specific to the recipe's function, or within
a subdirectory containing a set of closely-related
recipes.
The recipes themselves should follow the general
guidelines for recipes used in the Yocto Project
found in the
"OpenEmbedded Style Guide".
License File:
You must include a license file in the
meta-
bsp_root_name
directory.
This license covers the BSP Metadata as a whole.
You must specify which license to use since no
default license exists when one is not specified.
See the
COPYING.MIT
file for the Raspberry Pi BSP in the
meta-raspberrypi
BSP layer
as an example.
README File:
You must include a README
file in the
meta-
bsp_root_name
directory.
See the
README.md
file for the Raspberry Pi BSP in the
meta-raspberrypi
BSP layer
as an example.
At a minimum, the README
file should contain the following:
A brief description of the target hardware.
A list of all the dependencies of the BSP. These dependencies are typically a list of required layers needed to build the BSP. However, the dependencies should also contain information regarding any other dependencies the BSP might have.
Any required special licensing information. For example, this information includes information on special variables needed to satisfy a EULA, or instructions on information needed to build or distribute binaries built from the BSP Metadata.
The name and contact information for the BSP layer maintainer. This is the person to whom patches and questions should be sent. For information on how to find the right person, see the "Submitting a Change to the Yocto Project" section in the Yocto Project Development Tasks Manual.
Instructions on how to build the BSP using the BSP layer.
Instructions on how to boot the BSP build from the BSP layer.
Instructions on how to boot the binary
images contained in the
binary
directory,
if present.
Information on any known bugs or issues that users should know about when either building or booting the BSP binaries.
README.sources File:
If you BSP contains binary images in the
binary
directory, you must
include a README.sources
file in the
meta-
bsp_root_name
directory.
This file specifies exactly where you can find
the sources used to generate the binary images.
Layer Configuration File:
You must include a
conf/layer.conf
file in
the
meta-
bsp_root_name
directory.
This file identifies the
meta-
bsp_root_name
BSP layer as a layer to the build system.
Machine Configuration File:
You must include one or more
conf/machine/
bsp_root_name
.conf
files in the
meta-
bsp_root_name
directory.
These configuration files define machine targets
that can be built using the BSP layer.
Multiple machine configuration files define
variations of machine configurations that the
BSP supports.
If a BSP supports multiple machine variations,
you need to adequately describe each variation
in the BSP README
file.
Do not use multiple machine configuration files
to describe disparate hardware.
If you do have very different targets, you should
create separate BSP layers for each target.
meta-yocto-bsp
layer).
Such considerations are outside the scope of
this document.
Following are recommendations for released BSPs that conform to the Yocto Project:
Bootable Images: Released BSPs can contain one or more bootable images. Including bootable images allows users to easily try out the BSP using their own hardware.
In some cases, it might not be convenient to include a bootable image. If so, you might want to make two versions of the BSP available: one that contains binary images, and one that does not. The version that does not contain bootable images avoids unnecessary download times for users not interested in the images.
If you need to distribute a BSP and include
bootable images or build kernel and filesystems
meant to allow users to boot the BSP for evaluation
purposes, you should put the images and artifacts
within a
binary/
subdirectory located
in the
meta-
bsp_root_name
directory.
Use a Yocto Linux Kernel:
Kernel recipes in the BSP should be based on a
Yocto Linux kernel.
Basing your recipes on these kernels reduces
the costs for maintaining the BSP and increases
its scalability.
See the Yocto Linux Kernel
category in the
Source Repositories
for these kernels.
If you plan on customizing a recipe for a particular BSP, you need to do the following:
Create a *.bbappend
file for
the modified recipe.
For information on using append files, see the
"Using .bbappend Files in Your Layer"
section in the Yocto Project Development Tasks
Manual.
Ensure your directory structure in the BSP layer that supports your machine is such that the OpenEmbedded build system can find it. See the example later in this section for more information.
Put the append file in a directory whose name matches
the machine's name and is located in an appropriate
sub-directory inside the BSP layer (i.e.
recipes-bsp
,
recipes-graphics
,
recipes-core
, and so forth).
Place the BSP-specific files in the proper directory inside the BSP layer. How expansive the layer is affects where you must place these files. For example, if your layer supports several different machine types, you need to be sure your layer's directory structure includes hierarchy that separates the files according to machine. If your layer does not support multiple machines, the layer would not have that additional hierarchy and the files would obviously not be able to reside in a machine-specific directory.
Following is a specific example to help you better understand
the process.
This example customizes customizes a recipe by adding a
BSP-specific configuration file named
interfaces
to the
init-ifupdown_1.0.bb
recipe for machine
"xyz" where the BSP layer also supports several other
machines:
Edit the
init-ifupdown_1.0.bbappend
file
so that it contains the following:
FILESEXTRAPATHS_prepend := "${THISDIR}/files:"
The append file needs to be in the
meta-xyz/recipes-core/init-ifupdown
directory.
Create and place the new
interfaces
configuration file in
the BSP's layer here:
meta-xyz/recipes-core/init-ifupdown/files/xyz-machine-one/interfaces
meta-xyz
layer did
not support multiple machines, you would place
the interfaces
configuration
file in the layer here:
meta-xyz/recipes-core/init-ifupdown/files/interfaces
The
FILESEXTRAPATHS
variable in the append files extends the search path
the build system uses to find files during the build.
Consequently, for this example you need to have the
files
directory in the same
location as your append file.
In some cases, a BSP contains separately-licensed Intellectual Property (IP) for a component or components. For these cases, you are required to accept the terms of a commercial or other type of license that requires some kind of explicit End User License Agreement (EULA). Once you accept the license, the OpenEmbedded build system can then build and include the corresponding component in the final BSP image. If the BSP is available as a pre-built image, you can download the image after agreeing to the license or EULA.
You could find that some separately-licensed components that are essential for normal operation of the system might not have an unencumbered (or free) substitute. Without these essential components, the system would be non-functional. Then again, you might find that other licensed components that are simply 'good-to-have' or purely elective do have an unencumbered, free replacement component that you can use rather than agreeing to the separately-licensed component. Even for components essential to the system, you might find an unencumbered component that is not identical but will work as a less-capable version of the licensed version in the BSP recipe.
For cases where you can substitute a free component and still maintain the system's functionality, the "DOWNLOADS" selection from the "SOFTWARE" tab on the Yocto Project website makes available de-featured BSPs that are completely free of any IP encumbrances. For these cases, you can use the substitution directly and without any further licensing requirements. If present, these fully de-featured BSPs are named appropriately different as compared to the names of their respective encumbered BSPs. If available, these substitutions are your simplest and most preferred options. Obviously, use of these substitutions assumes the resulting functionality meets system requirements.
A couple different methods exist within the OpenEmbedded build system to satisfy the licensing requirements for an encumbered BSP. The following list describes them in order of preference:
Use the
LICENSE_FLAGS
Variable to Define the Recipes that Have Commercial
or Other Types of Specially-Licensed Packages:
For each of those recipes, you can specify a
matching license string in a
local.conf
variable named
LICENSE_FLAGS_WHITELIST
.
Specifying the matching license string signifies
that you agree to the license.
Thus, the build system can build the corresponding
recipe and include the component in the image.
See the
"Enabling Commercially Licensed Recipes"
section in the Yocto Project Development Tasks
Manual for details on how to use these variables.
If you build as you normally would, without
specifying any recipes in the
LICENSE_FLAGS_WHITELIST
, the
build stops and provides you with the list of recipes
that you have tried to include in the image that
need entries in the
LICENSE_FLAGS_WHITELIST
.
Once you enter the appropriate license flags into
the whitelist, restart the build to continue where
it left off.
During the build, the prompt will not appear again
since you have satisfied the requirement.
Once the appropriate license flags are on the
white list in the
LICENSE_FLAGS_WHITELIST
variable,
you can build the encumbered image with no change
at all to the normal build process.
Get a Pre-Built Version of the BSP:
You can get this type of BSP by selecting the
"DOWNLOADS" item from the "SOFTWARE" tab on the
Yocto Project website.
You can download BSP tarballs that contain
proprietary components after agreeing to the
licensing requirements of each of the individually
encumbered packages as part of the download process.
Obtaining the BSP this way allows you to access an
encumbered image immediately after agreeing to the
click-through license agreements presented by the
website.
If you want to build the image yourself using
the recipes contained within the BSP tarball,
you will still need to create an appropriate
LICENSE_FLAGS_WHITELIST
to match the encumbered recipes in the BSP.
bitbake-layers
Script¶
The bitbake-layers create-layer
script
automates creating a BSP layer.
What makes a layer a "BSP layer" is the presence of at least one machine
configuration file.
Additionally, a BSP layer usually has a kernel recipe
or an append file that leverages off an existing kernel recipe.
The primary requirement, however, is the machine configuration.
Use these steps to create a BSP layer:
Create a General Layer:
Use the bitbake-layers
script with the
create-layer
subcommand to create a
new general layer.
For instructions on how to create a general layer using the
bitbake-layers
script, see the
"Creating a General Layer Using the bitbake-layers
Script"
section in the Yocto Project Development Tasks Manual.
Create a Layer Configuration File:
Every layer needs a layer configuration file.
This configuration file establishes locations for the
layer's recipes, priorities for the layer, and so forth.
You can find examples of layer.conf
files in the Yocto Project
Source Repositories.
To get examples of what you need in your configuration
file, locate a layer (e.g. "meta-ti") and examine the
http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/tree/conf/layer.conf
file.
Create a Machine Configuration File:
Create a conf/machine/
bsp_root_name
.conf
file.
See
meta-yocto-bsp/conf/machine
for sample
bsp_root_name
.conf
files.
Other samples such as
meta-ti
and
meta-freescale
exist from other vendors that have more specific machine
and tuning examples.
Create a Kernel Recipe:
Create a kernel recipe in recipes-kernel/linux
by either using a kernel append file or a new custom kernel
recipe file (e.g. yocto-linux_4.12.bb
).
The BSP layers mentioned in the previous step also contain different
kernel examples.
See the
"Modifying an Existing Recipe"
section in the Yocto Project Linux Kernel Development Manual
for information on how to create a custom kernel.
The remainder of this section provides a description of
the Yocto Project reference BSP for Beaglebone, which
resides in the
meta-yocto-bsp
layer.
The layer's conf
directory
contains the layer.conf
configuration file.
In this example, the
conf/layer.conf
is the
following:
# We have a conf and classes directory, add to BBPATH BBPATH .= ":${LAYERDIR}" # We have recipes-* directories, add to BBFILES BBFILES += "${LAYERDIR}/recipes-*/*/*.bb \ ${LAYERDIR}/recipes-*/*/*.bbappend" BBFILE_COLLECTIONS += "yoctobsp" BBFILE_PATTERN_yoctobsp = "^${LAYERDIR}/" BBFILE_PRIORITY_yoctobsp = "5" LAYERVERSION_yoctobsp = "4" LAYERSERIES_COMPAT_yoctobsp = "dunfell"
The variables used in this file configure the layer. A good way to learn about layer configuration files is to examine various files for BSP from the Source Repositories.
For a detailed description of this particular layer configuration file, see "step 3 in the discussion that describes how to create layers in the Yocto Project Development Tasks Manual.
As mentioned earlier in this section, the existence of a machine configuration file is what makes a layer a BSP layer as compared to a general or kernel layer.
One or more machine configuration files exist in the
bsp_layer
/conf/machine/
directory of the layer:
bsp_layer
/conf/machine/
machine1
.conf
bsp_layer
/conf/machine/
machine2
.conf
bsp_layer
/conf/machine/
machine3
.conf
... more ...
For example, the machine configuration file for the
BeagleBone and BeagleBone Black development boards
is located in the layer
poky/meta-yocto-bsp/conf/machine
and is named beaglebone-yocto.conf
:
#@TYPE: Machine #@NAME: Beaglebone-yocto machine #@DESCRIPTION: Reference machine configuration for http://beagleboard.org/bone and http://beagleboard.org/black boards PREFERRED_PROVIDER_virtual/xserver ?= "xserver-xorg" XSERVER ?= "xserver-xorg \ xf86-video-modesetting \ " MACHINE_EXTRA_RRECOMMENDS = "kernel-modules kernel-devicetree" EXTRA_IMAGEDEPENDS += "u-boot" DEFAULTTUNE ?= "cortexa8hf-neon" include conf/machine/include/tune-cortexa8.inc IMAGE_FSTYPES += "tar.bz2 jffs2 wic wic.bmap" EXTRA_IMAGECMD_jffs2 = "-lnp " WKS_FILE ?= "beaglebone-yocto.wks" IMAGE_INSTALL_append = " kernel-devicetree kernel-image-zimage" do_image_wic[depends] += "mtools-native:do_populate_sysroot dosfstools-native:do_populate_sysroot" SERIAL_CONSOLES ?= "115200;ttyS0 115200;ttyO0" SERIAL_CONSOLES_CHECK = "${SERIAL_CONSOLES}" PREFERRED_PROVIDER_virtual/kernel ?= "linux-yocto" PREFERRED_VERSION_linux-yocto ?= "5.0%" KERNEL_IMAGETYPE = "zImage" KERNEL_DEVICETREE = "am335x-bone.dtb am335x-boneblack.dtb am335x-bonegreen.dtb" KERNEL_EXTRA_ARGS += "LOADADDR=${UBOOT_ENTRYPOINT}" SPL_BINARY = "MLO" UBOOT_SUFFIX = "img" UBOOT_MACHINE = "am335x_evm_defconfig" UBOOT_ENTRYPOINT = "0x80008000" UBOOT_LOADADDRESS = "0x80008000" MACHINE_FEATURES = "usbgadget usbhost vfat alsa" IMAGE_BOOT_FILES ?= "u-boot.${UBOOT_SUFFIX} MLO zImage am335x-bone.dtb am335x-boneblack.dtb am335x-bonegreen.dtb"
The variables used to configure the machine define machine-specific properties; for example, machine-dependent packages, machine tunings, the type of kernel to build, and U-Boot configurations.
The following list provides some explanation for the statements found in the example reference machine configuration file for the BeagleBone development boards. Realize that much more can be defined as part of a machine's configuration file. In general, you can learn about related variables that this example does not have by locating the variables in the "Yocto Project Variables Glossary" in the Yocto Project Reference Manual.
PREFERRED_PROVIDER_virtual/xserver
:
The recipe that provides "virtual/xserver" when
more than one provider is found.
In this case, the recipe that provides
"virtual/xserver" is "xserver-xorg", which
exists in
poky/meta/recipes-graphics/xorg-xserver
.
XSERVER
:
The packages that should be installed to provide
an X server and drivers for the machine.
In this example, the "xserver-xorg" and
"xf86-video-modesetting" are installed.
MACHINE_EXTRA_RRECOMMENDS
:
A list of machine-dependent packages
not essential for booting the image.
Thus, the build does not fail if the packages
do not exist.
However, the packages are required for a
fully-featured image.
MACHINE*
variables
exist that help you configure a particular
piece of hardware.
EXTRA_IMAGEDEPENDS
:
Recipes to build that do not provide packages
for installing into the root filesystem
but building the image depends on the
recipes.
Sometimes a recipe is required to build
the final image but is not needed in the
root filesystem.
In this case, the U-Boot recipe must be
built for the image.
DEFAULTTUNE
:
Machines use tunings to optimize machine,
CPU, and application performance.
These features, which are collectively known
as "tuning features", exist in the
OpenEmbedded-Core (OE-Core)
layer (e.g.
poky/meta/conf/machine/include
).
In this example, the default tunning file is
"cortexa8hf-neon".
include
statement
that pulls in the
conf/machine/include/tune-cortexa8.inc
file provides many tuning possibilities.
IMAGE_FSTYPES
:
The formats the OpenEmbedded build system
uses during the build when creating the
root filesystem.
In this example, four types of images are
supported.
EXTRA_IMAGECMD
:
Specifies additional options for image
creation commands.
In this example, the "-lnp " option is used
when creating the
JFFS2
image.
WKS_FILE
:
The location of the
Wic kickstart
file used by the OpenEmbedded build system to
create a partitioned image (image.wic).
IMAGE_INSTALL
:
Specifies packages to install into an image
through the
image
class.
Recipes use the IMAGE_INSTALL
variable.
do_image_wic[depends]
:
A task that is constructed during the build.
In this example, the task depends on specific tools
in order to create the sysroot when buiding a Wic
image.
SERIAL_CONSOLES
:
Defines a serial console (TTY) to enable using
getty.
In this case, the baud rate is "115200" and the
device name is "ttyO0".
PREFERRED_PROVIDER_virtual/kernel
:
Specifies the recipe that provides
"virtual/kernel" when more than one provider
is found.
In this case, the recipe that provides
"virtual/kernel" is "linux-yocto", which
exists in the layer's
recipes-kernel/linux
directory.
PREFERRED_VERSION_linux-yocto
:
Defines the version of the recipe used
to build the kernel, which is "5.0" in this
case.
KERNEL_IMAGETYPE
:
The type of kernel to build for the device.
In this case, the OpenEmbedded build system
creates a "zImage" image type.
KERNEL_DEVICETREE
:
The names of the generated Linux kernel device
trees (i.e. the *.dtb
) files.
All the device trees for the various BeagleBone
devices are included.
KERNEL_EXTRA_ARGS
:
Additional make
command-line arguments the OpenEmbedded build
system passes on when compiling the kernel.
In this example, "LOADADDR=${UBOOT_ENTRYPOINT}"
is passed as a command-line argument.
SPL_BINARY
:
Defines the Secondary Program Loader (SPL) binary
type.
In this case, the SPL binary is set to
"MLO", which stands for Multimedia card LOader.
The BeagleBone development board requires an
SPL to boot and that SPL file type must be MLO.
Consequently, the machine configuration needs to
define SPL_BINARY
as "MLO".
u-boot.inc
include file.
UBOOT_*
:
Defines various U-Boot configurations needed
to build a U-Boot image.
In this example, a U-Boot image is required
to boot the BeagleBone device.
See the following variables for more information:
UBOOT_SUFFIX
:
Points to the generated U-Boot extension.
UBOOT_MACHINE
:
Specifies the value passed on the make command line when building a U-Boot image.
UBOOT_ENTRYPOINT
:
Specifies the entry point for the U-Boot image.
UBOOT_LOADADDRESS
:
Specifies the load address for the U-Boot image.
MACHINE_FEATURES
:
Specifies the list of hardware features the
BeagleBone device is capable of supporting.
In this case, the device supports
"usbgadget usbhost vfat alsa".
IMAGE_BOOT_FILES
:
Files installed into the device's boot partition
when preparing the image using the Wic tool
with the bootimg-partition
source plugin.
The kernel recipe used to build the kernel image for the BeagleBone device was established in the machine configuration:
PREFERRED_PROVIDER_virtual/kernel ?= "linux-yocto" PREFERRED_VERSION_linux-yocto ?= "5.0%"
The meta-yocto-bsp/recipes-kernel/linux
directory in the layer contains metadata used
to build the kernel.
In this case, a kernel append file (i.e.
linux-yocto_5.0.bbappend
) is used to
override an established kernel recipe (i.e.
linux-yocto_5.0.bb
), which is
located in
http://git.yoctoproject.org/cgit/cgit.cgi/poky/tree/meta/recipes-kernel/linux.
Following is the contents of the append file:
KBRANCH_genericx86 = "v5.0/standard/base" KBRANCH_genericx86-64 = "v5.0/standard/base" KBRANCH_edgerouter = "v5.0/standard/edgerouter" KBRANCH_beaglebone-yocto = "v5.0/standard/beaglebone" KMACHINE_genericx86 ?= "common-pc" KMACHINE_genericx86-64 ?= "common-pc-64" KMACHINE_beaglebone-yocto ?= "beaglebone" SRCREV_machine_genericx86 ?= "3df4aae6074e94e794e27fe7f17451d9353cdf3d" SRCREV_machine_genericx86-64 ?= "3df4aae6074e94e794e27fe7f17451d9353cdf3d" SRCREV_machine_edgerouter ?= "3df4aae6074e94e794e27fe7f17451d9353cdf3d" SRCREV_machine_beaglebone-yocto ?= "3df4aae6074e94e794e27fe7f17451d9353cdf3d" COMPATIBLE_MACHINE_genericx86 = "genericx86" COMPATIBLE_MACHINE_genericx86-64 = "genericx86-64" COMPATIBLE_MACHINE_edgerouter = "edgerouter" COMPATIBLE_MACHINE_beaglebone-yocto = "beaglebone-yocto" LINUX_VERSION_genericx86 = "5.0.3" LINUX_VERSION_genericx86-64 = "5.0.3" LINUX_VERSION_edgerouter = "5.0.3" LINUX_VERSION_beaglebone-yocto = "5.0.3"
This particular append file works for all the
machines that are part of the
meta-yocto-bsp
layer.
The relevant statements are appended with
the "beaglebone-yocto" string.
The OpenEmbedded build system uses these
statements to override similar statements
in the kernel recipe:
KBRANCH
:
Identifies the kernel branch that is validated,
patched, and configured during the build.
KMACHINE
:
Identifies the machine name as known by the
kernel, which is sometimes a different name
than what is known by the OpenEmbedded build
system.
SRCREV
:
Identifies the revision of the source code used
to build the image.
COMPATIBLE_MACHINE
:
A regular expression that resolves to one or
more target machines with which the recipe
is compatible.
LINUX_VERSION
:
The Linux version from kernel.org used by
the OpenEmbedded build system to build the
kernel image.
Regardless of how you intend to make use of the Yocto Project, chances are you will work with the Linux kernel. This manual describes how to set up your build host to support kernel development, introduces the kernel development process, provides background information on the Yocto Linux kernel Metadata, describes common tasks you can perform using the kernel tools, shows you how to use the kernel Metadata needed to work with the kernel inside the Yocto Project, and provides insight into how the Yocto Project team develops and maintains Yocto Linux kernel Git repositories and Metadata.
Each Yocto Project release has a set of Yocto Linux kernel recipes, whose Git repositories you can view in the Yocto Source Repositories under the "Yocto Linux Kernel" heading. New recipes for the release track the latest Linux kernel upstream developments from http://www.kernel.org and introduce newly-supported platforms. Previous recipes in the release are refreshed and supported for at least one additional Yocto Project release. As they align, these previous releases are updated to include the latest from the Long Term Support Initiative (LTSI) project. You can learn more about Yocto Linux kernels and LTSI in the "Yocto Project Kernel Development and Maintenance" section.
Also included is a Yocto Linux kernel development recipe
(linux-yocto-dev.bb
) should you want to work
with the very latest in upstream Yocto Linux kernel development and
kernel Metadata development.
The Yocto Project also provides a powerful set of kernel tools for managing Yocto Linux kernel sources and configuration data. You can use these tools to make a single configuration change, apply multiple patches, or work with your own kernel sources.
In particular, the kernel tools allow you to generate configuration
fragments that specify only what you must, and nothing more.
Configuration fragments only need to contain the highest level
visible CONFIG
options as presented by the
Yocto Linux kernel menuconfig
system.
Contrast this against a complete Yocto Linux kernel
.config
file, which includes all the automatically
selected CONFIG
options.
This efficiency reduces your maintenance effort and allows you
to further separate your configuration in ways that make sense for
your project.
A common split separates policy and hardware.
For example, all your kernels might support the
proc
and sys
filesystems,
but only specific boards require sound, USB, or specific drivers.
Specifying these configurations individually allows you to aggregate
them together as needed, but maintains them in only one place.
Similar logic applies to separating source changes.
If you do not maintain your own kernel sources and need to make only minimal changes to the sources, the released recipes provide a vetted base upon which to layer your changes. Doing so allows you to benefit from the continual kernel integration and testing performed during development of the Yocto Project.
If, instead, you have a very specific Linux kernel source tree and are unable to align with one of the official Yocto Linux kernel recipes, an alternative exists by which you can use the Yocto Project Linux kernel tools with your own kernel sources.
The remainder of this manual provides instructions for completing specific Linux kernel development tasks. These instructions assume you are comfortable working with BitBake recipes and basic open-source development tools. Understanding these concepts will facilitate the process of working with the kernel recipes. If you find you need some additional background, please be sure to review and understand the following documentation:
Yocto Project Quick Build document.
Yocto project Overview and Concepts Manual.
devtool
workflow
as described in the Yocto Project Application Development and
the Extensible Software Development Kit (eSDK) manual.
The "Understanding and Creating Layers" section in the Yocto Project Development Tasks Manual.
The "Kernel Modification Workflow" section.
Kernel modification involves changing the Yocto Project kernel,
which could involve changing configuration options as well as adding
new kernel recipes.
Configuration changes can be added in the form of configuration
fragments, while recipe modification comes through the kernel's
recipes-kernel
area in a kernel layer you create.
This section presents a high-level overview of the Yocto Project kernel modification workflow. The illustration and accompanying list provide general information and references for further information.
Set up Your Host Development System to Support Development Using the Yocto Project: See the "Setting Up the Development Host to Use the Yocto Project" section in the Yocto Project Development Tasks Manual for options on how to get a build host ready to use the Yocto Project.
Set Up Your Host Development System for Kernel Development:
It is recommended that you use devtool
and an extensible SDK for kernel development.
Alternatively, you can use traditional kernel development
methods with the Yocto Project.
Either way, there are steps you need to take to get the
development environment ready.
Using devtool
and the eSDK requires
that you have a clean build of the image and that you are
set up with the appropriate eSDK.
For more information, see the
"Getting Ready to Develop Using devtool
"
section.
Using traditional kernel development requires that you have the kernel source available in an isolated local Git repository. For more information, see the "Getting Ready for Traditional Kernel Development" section.
Make Changes to the Kernel Source Code if
applicable:
Modifying the kernel does not always mean directly
changing source files.
However, if you have to do this, you make the changes to the
files in the eSDK's Build Directory if you are using
devtool
.
For more information, see the
"Using devtool
to Patch the Kernel"
section.
If you are using traditional kernel development, you edit the source files in the kernel's local Git repository. For more information, see the "Using Traditional Kernel Development to Patch the Kernel" section.
Make Kernel Configuration Changes if
Applicable:
If your situation calls for changing the kernel's
configuration, you can use
menuconfig
,
which allows you to interactively develop and test the
configuration changes you are making to the kernel.
Saving changes you make with menuconfig
updates the kernel's .config
file.
.config
file, which is
found in the Build Directory among the source code
used for the build.
Doing so, can produce unexpected results when the
OpenEmbedded build system regenerates the configuration
file.
Once you are satisfied with the configuration
changes made using menuconfig
and you have saved them, you can directly compare the
resulting .config
file against an
existing original and gather those changes into a
configuration fragment file
to be referenced from within the kernel's
.bbappend
file.
Additionally, if you are working in a BSP layer
and need to modify the BSP's kernel's configuration,
you can use menuconfig
.
Rebuild the Kernel Image With Your Changes: Rebuilding the kernel image applies your changes. Depending on your target hardware, you can verify your changes on actual hardware or perhaps QEMU.
The remainder of this developer's guide covers common tasks typically used during kernel development, advanced Metadata usage, and Yocto Linux kernel maintenance concepts.
devtool
to Patch the KernelThis chapter presents several common tasks you perform when you work with the Yocto Project Linux kernel. These tasks include preparing your host development system for kernel development, preparing a layer, modifying an existing recipe, patching the kernel, configuring the kernel, iterative development, working with your own sources, and incorporating out-of-tree modules.
Before you can do any kernel development, you need to be
sure your build host is set up to use the Yocto Project.
For information on how to get set up, see the
"Preparing the Build Host"
section in the Yocto Project Development Tasks Manual.
Part of preparing the system is creating a local Git
repository of the
Source Directory
(poky
) on your system.
Follow the steps in the
"Cloning the poky
Repository"
section in the Yocto Project Development Tasks Manual to set up your
Source Directory.
Kernel development is best accomplished using
devtool
and not through traditional kernel workflow methods.
The remainder of this section provides information for both
scenarios.
devtool
¶
Follow these steps to prepare to update the kernel image using
devtool
.
Completing this procedure leaves you with a clean kernel image
and ready to make modifications as described in the
"Using devtool
to Patch the Kernel"
section:
Initialize the BitBake Environment:
Before building an extensible SDK, you need to
initialize the BitBake build environment by sourcing the
build environment script
(i.e. oe-init-build-env
):
$ cd ~/poky $ source oe-init-build-env
poky
) have been cloned
using Git and the local repository is named
"poky".
Prepare Your local.conf
File:
By default, the
MACHINE
variable is set to "qemux86-64", which is fine if you are
building for the QEMU emulator in 64-bit mode.
However, if you are not, you need to set the
MACHINE
variable appropriately in
your conf/local.conf
file found in
the
Build Directory
(i.e. ~/poky/build
in this
example).
Also, since you are preparing to work on the
kernel image, you need to set the
MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS
variable to include kernel modules.
In this example we wish to build for qemux86 so
we must set the MACHINE
variable
to "qemux86" and also add the "kernel-modules". As described
we do this by appending to conf/local.conf
:
MACHINE = "qemux86" MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS += "kernel-modules"
Create a Layer for Patches:
You need to create a layer to hold patches created
for the kernel image.
You can use the
bitbake-layers create-layer
command as follows:
$ cd ~/poky/build $ bitbake-layers create-layer ../../meta-mylayer NOTE: Starting bitbake server... Add your new layer with 'bitbake-layers add-layer ../../meta-mylayer' $
bitbake-layers create-layer
command to quickly set up a layer, see the
"Creating a General Layer Using the bitbake-layers
Script"
section in the Yocto Project Development Tasks
Manual.
Inform the BitBake Build Environment About
Your Layer:
As directed when you created your layer, you need to
add the layer to the
BBLAYERS
variable in the bblayers.conf
file
as follows:
$ cd ~/poky/build $ bitbake-layers add-layer ../../meta-mylayer NOTE: Starting bitbake server... $
Build the Extensible SDK: Use BitBake to build the extensible SDK specifically for use with images to be run using QEMU:
$ cd ~/poky/build $ bitbake core-image-minimal -c populate_sdk_ext
Once the build finishes, you can find the SDK installer
file (i.e. *.sh
file) in the
following directory:
~/poky/build/tmp/deploy/sdk
For this example, the installer file is named
poky-glibc-x86_64-core-image-minimal-i586-toolchain-ext-3.1.2.sh
Install the Extensible SDK:
Use the following command to install the SDK.
For this example, install the SDK in the default
~/poky_sdk
directory:
$ cd ~/poky/build/tmp/deploy/sdk $ ./poky-glibc-x86_64-core-image-minimal-i586-toolchain-ext-3.1.2.sh Poky (Yocto Project Reference Distro) Extensible SDK installer version 3.1.2 ============================================================================ Enter target directory for SDK (default: ~/poky_sdk): You are about to install the SDK to "/home/scottrif/poky_sdk". Proceed [Y/n]? Y Extracting SDK......................................done Setting it up... Extracting buildtools... Preparing build system... Parsing recipes: 100% |#################################################################| Time: 0:00:52 Initializing tasks: 100% |############## ###############################################| Time: 0:00:04 Checking sstate mirror object availability: 100% |######################################| Time: 0:00:00 Parsing recipes: 100% |#################################################################| Time: 0:00:33 Initializing tasks: 100% |##############################################################| Time: 0:00:00 done SDK has been successfully set up and is ready to be used. Each time you wish to use the SDK in a new shell session, you need to source the environment setup script e.g. $ . /home/scottrif/poky_sdk/environment-setup-i586-poky-linux
Set Up a New Terminal to Work With the Extensible SDK: You must set up a new terminal to work with the SDK. You cannot use the same BitBake shell used to build the installer.
After opening a new shell, run the SDK environment setup script as directed by the output from installing the SDK:
$ source ~/poky_sdk/environment-setup-i586-poky-linux "SDK environment now set up; additionally you may now run devtool to perform development tasks. Run devtool --help for further details.
Build the Clean Image:
The final step in preparing to work on the kernel is to
build an initial image using
devtool
in the new terminal you
just set up and initialized for SDK work:
$ devtool build-image Parsing recipes: 100% |##########################################| Time: 0:00:05 Parsing of 830 .bb files complete (0 cached, 830 parsed). 1299 targets, 47 skipped, 0 masked, 0 errors. WARNING: No packages to add, building image core-image-minimal unmodified Loading cache: 100% |############################################| Time: 0:00:00 Loaded 1299 entries from dependency cache. NOTE: Resolving any missing task queue dependencies Initializing tasks: 100% |#######################################| Time: 0:00:07 Checking sstate mirror object availability: 100% |###############| Time: 0:00:00 NOTE: Executing SetScene Tasks NOTE: Executing RunQueue Tasks NOTE: Tasks Summary: Attempted 2866 tasks of which 2604 didn't need to be rerun and all succeeded. NOTE: Successfully built core-image-minimal. You can find output files in /home/scottrif/poky_sdk/tmp/deploy/images/qemux86
If you were building for actual hardware and not for
emulation, you could flash the image to a USB stick
on /dev/sdd
and boot your device.
For an example that uses a Minnowboard, see the
TipsAndTricks/KernelDevelopmentWithEsdk
Wiki page.
At this point you have set up to start making modifications to
the kernel by using the extensible SDK.
For a continued example, see the
"Using devtool
to Patch the Kernel"
section.
Getting ready for traditional kernel development using the Yocto Project involves many of the same steps as described in the previous section. However, you need to establish a local copy of the kernel source since you will be editing these files.
Follow these steps to prepare to update the kernel image using traditional kernel development flow with the Yocto Project. Completing this procedure leaves you ready to make modifications to the kernel source as described in the "Using Traditional Kernel Development to Patch the Kernel" section:
Initialize the BitBake Environment:
Before you can do anything using BitBake, you need to
initialize the BitBake build environment by sourcing the
build environment script
(i.e. oe-init-build-env
).
Also, for this example, be sure that the local branch
you have checked out for poky
is
the Yocto Project Dunfell branch.
If you need to checkout out the Dunfell branch,
see the
"Checking out by Branch in Poky"
section in the Yocto Project Development Tasks Manual.
$ cd ~/poky $ git branch master * Dunfell $ source oe-init-build-env
poky
) have been cloned
using Git and the local repository is named
"poky".
Prepare Your local.conf
File:
By default, the
MACHINE
variable is set to "qemux86-64", which is fine if you are
building for the QEMU emulator in 64-bit mode.
However, if you are not, you need to set the
MACHINE
variable appropriately in
your conf/local.conf
file found
in the
Build Directory
(i.e. ~/poky/build
in this
example).
Also, since you are preparing to work on the
kernel image, you need to set the
MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS
variable to include kernel modules.
In this example we wish to build for qemux86 so
we must set the MACHINE
variable
to "qemux86" and also add the "kernel-modules". As described
we do this by appending to conf/local.conf
:
MACHINE = "qemux86" MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS += "kernel-modules"
Create a Layer for Patches:
You need to create a layer to hold patches created
for the kernel image.
You can use the
bitbake-layers create-layer
command as follows:
$ cd ~/poky/build $ bitbake-layers create-layer ../../meta-mylayer NOTE: Starting bitbake server... Add your new layer with 'bitbake-layers add-layer ../../meta-mylayer'
bitbake-layers create-layer
command to quickly set up a layer, see the
"Creating a General Layer Using the bitbake-layers
Script"
section in the Yocto Project Development Tasks
Manual.
Inform the BitBake Build Environment About
Your Layer:
As directed when you created your layer, you need to add
the layer to the
BBLAYERS
variable in the bblayers.conf
file
as follows:
$ cd ~/poky/build $ bitbake-layers add-layer ../../meta-mylayer NOTE: Starting bitbake server ... $
Create a Local Copy of the Kernel Git Repository: You can find Git repositories of supported Yocto Project kernels organized under "Yocto Linux Kernel" in the Yocto Project Source Repositories at http://git.yoctoproject.org.
For simplicity, it is recommended that you create your
copy of the kernel Git repository outside of the
Source Directory,
which is usually named poky
.
Also, be sure you are in the
standard/base
branch.
The following commands show how to create a local copy
of the linux-yocto-4.12
kernel and
be in the standard/base
branch.
linux-yocto-4.12
kernel
can be used with the Yocto Project 2.4 release
and forward.
You cannot use the
linux-yocto-4.12
kernel with
releases prior to Yocto Project 2.4:
$ cd ~ $ git clone git://git.yoctoproject.org/linux-yocto-4.12 --branch standard/base Cloning into 'linux-yocto-4.12'... remote: Counting objects: 6097195, done. remote: Compressing objects: 100% (901026/901026), done. remote: Total 6097195 (delta 5152604), reused 6096847 (delta 5152256) Receiving objects: 100% (6097195/6097195), 1.24 GiB | 7.81 MiB/s, done. Resolving deltas: 100% (5152604/5152604), done. Checking connectivity... done. Checking out files: 100% (59846/59846), done.
Create a Local Copy of the Kernel Cache Git
Repository:
For simplicity, it is recommended that you create your
copy of the kernel cache Git repository outside of the
Source Directory,
which is usually named poky
.
Also, for this example, be sure you are in the
yocto-4.12
branch.
The following commands show how to create a local copy
of the yocto-kernel-cache
and
be in the yocto-4.12
branch:
$ cd ~ $ git clone git://git.yoctoproject.org/yocto-kernel-cache --branch yocto-4.12 Cloning into 'yocto-kernel-cache'... remote: Counting objects: 22639, done. remote: Compressing objects: 100% (9761/9761), done. remote: Total 22639 (delta 12400), reused 22586 (delta 12347) Receiving objects: 100% (22639/22639), 22.34 MiB | 6.27 MiB/s, done. Resolving deltas: 100% (12400/12400), done. Checking connectivity... done.
At this point, you are ready to start making modifications to the kernel using traditional kernel development steps. For a continued example, see the "Using Traditional Kernel Development to Patch the Kernel" section.
If you are going to be modifying kernel recipes, it is recommended
that you create and prepare your own layer in which to do your
work.
Your layer contains its own
BitBake
append files (.bbappend
) and provides a
convenient mechanism to create your own recipe files
(.bb
) as well as store and use kernel
patch files.
For background information on working with layers, see the
"Understanding and Creating Layers"
section in the Yocto Project Development Tasks Manual.
bitbake-layers create-layer
command, which simplifies creating a new layer.
See the
"Creating a General Layer Using the bitbake-layers
Script"
section in the Yocto Project Development Tasks Manual for
information on how to use this script to quick set up a
new layer.
To better understand the layer you create for kernel development,
the following section describes how to create a layer
without the aid of tools.
These steps assume creation of a layer named
mylayer
in your home directory:
Create Structure: Create the layer's structure:
$ cd $HOME $ mkdir meta-mylayer $ mkdir meta-mylayer/conf $ mkdir meta-mylayer/recipes-kernel $ mkdir meta-mylayer/recipes-kernel/linux $ mkdir meta-mylayer/recipes-kernel/linux/linux-yocto
The conf
directory holds your
configuration files, while the
recipes-kernel
directory holds your
append file and eventual patch files.
Create the Layer Configuration File:
Move to the meta-mylayer/conf
directory and create the layer.conf
file as follows:
# We have a conf and classes directory, add to BBPATH BBPATH .= ":${LAYERDIR}" # We have recipes-* directories, add to BBFILES BBFILES += "${LAYERDIR}/recipes-*/*/*.bb \ ${LAYERDIR}/recipes-*/*/*.bbappend" BBFILE_COLLECTIONS += "mylayer" BBFILE_PATTERN_mylayer = "^${LAYERDIR}/" BBFILE_PRIORITY_mylayer = "5"
Notice mylayer
as part of the last
three statements.
Create the Kernel Recipe Append File:
Move to the
meta-mylayer/recipes-kernel/linux
directory and create the kernel's append file.
This example uses the
linux-yocto-4.12
kernel.
Thus, the name of the append file is
linux-yocto_4.12.bbappend
:
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:" SRC_URI_append = " file://patch-file-one
" SRC_URI_append = " file://patch-file-two
" SRC_URI_append = " file://patch-file-three
"
The
FILESEXTRAPATHS
and
SRC_URI
statements enable the OpenEmbedded build system to find
patch files.
For more information on using append files, see the
"Using .bbappend Files in Your Layer"
section in the Yocto Project Development Tasks Manual.
In many cases, you can customize an existing linux-yocto recipe to
meet the needs of your project.
Each release of the Yocto Project provides a few Linux
kernel recipes from which you can choose.
These are located in the
Source Directory
in meta/recipes-kernel/linux
.
Modifying an existing recipe can consist of the following:
Creating the append file
Applying patches
Changing the configuration
Before modifying an existing recipe, be sure that you have created a minimal, custom layer from which you can work. See the "Creating and Preparing a Layer" section for information.
You create this file in your custom layer.
You also name it accordingly based on the linux-yocto recipe
you are using.
For example, if you are modifying the
meta/recipes-kernel/linux/linux-yocto_4.12.bb
recipe, the append file will typically be located as follows
within your custom layer:
your-layer
/recipes-kernel/linux/linux-yocto_4.12.bbappend
The append file should initially extend the
FILESPATH
search path by prepending the directory that contains your
files to the
FILESEXTRAPATHS
variable as follows:
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
The path ${
THISDIR
}/${
PN
}
expands to "linux-yocto" in the current directory for this
example.
If you add any new files that modify the kernel recipe and you
have extended FILESPATH
as
described above, you must place the files in your layer in the
following area:
your-layer
/recipes-kernel/linux/linux-yocto/
As an example, consider the following append file
used by the BSPs in meta-yocto-bsp
:
meta-yocto-bsp/recipes-kernel/linux/linux-yocto_4.12.bbappend
The following listing shows the file.
Be aware that the actual commit ID strings in this
example listing might be different than the actual strings
in the file from the meta-yocto-bsp
layer upstream.
KBRANCH_genericx86 = "standard/base" KBRANCH_genericx86-64 = "standard/base" KMACHINE_genericx86 ?= "common-pc" KMACHINE_genericx86-64 ?= "common-pc-64" KBRANCH_edgerouter = "standard/edgerouter" KBRANCH_beaglebone = "standard/beaglebone" SRCREV_machine_genericx86 ?= "d09f2ce584d60ecb7890550c22a80c48b83c2e19" SRCREV_machine_genericx86-64 ?= "d09f2ce584d60ecb7890550c22a80c48b83c2e19" SRCREV_machine_edgerouter ?= "b5c8cfda2dfe296410d51e131289fb09c69e1e7d" SRCREV_machine_beaglebone ?= "b5c8cfda2dfe296410d51e131289fb09c69e1e7d" COMPATIBLE_MACHINE_genericx86 = "genericx86" COMPATIBLE_MACHINE_genericx86-64 = "genericx86-64" COMPATIBLE_MACHINE_edgerouter = "edgerouter" COMPATIBLE_MACHINE_beaglebone = "beaglebone" LINUX_VERSION_genericx86 = "4.12.7" LINUX_VERSION_genericx86-64 = "4.12.7" LINUX_VERSION_edgerouter = "4.12.10" LINUX_VERSION_beaglebone = "4.12.10"
This append file contains statements used to support
several BSPs that ship with the Yocto Project.
The file defines machines using the
COMPATIBLE_MACHINE
variable and uses the
KMACHINE
variable to ensure the machine name used by the OpenEmbedded
build system maps to the machine name used by the Linux Yocto
kernel.
The file also uses the optional
KBRANCH
variable to ensure the build process uses the
appropriate kernel branch.
Although this particular example does not use it, the
KERNEL_FEATURES
variable could be used to enable features specific to
the kernel.
The append file points to specific commits in the
Source Directory
Git repository and the meta
Git repository
branches to identify the exact kernel needed to build the
BSP.
One thing missing in this particular BSP, which you will
typically need when developing a BSP, is the kernel
configuration file (.config
) for your BSP.
When developing a BSP, you probably have a kernel configuration
file or a set of kernel configuration files that, when taken
together, define the kernel configuration for your BSP.
You can accomplish this definition by putting the configurations
in a file or a set of files inside a directory located at the
same level as your kernel's append file and having the same
name as the kernel's main recipe file.
With all these conditions met, simply reference those files in
the
SRC_URI
statement in the append file.
For example, suppose you had some configuration options
in a file called network_configs.cfg
.
You can place that file inside a directory named
linux-yocto
and then add
a SRC_URI
statement such as the
following to the append file.
When the OpenEmbedded build system builds the kernel, the
configuration options are picked up and applied.
SRC_URI += "file://network_configs.cfg"
To group related configurations into multiple files, you
perform a similar procedure.
Here is an example that groups separate configurations
specifically for Ethernet and graphics into their own
files and adds the configurations by using a
SRC_URI
statement like the following
in your append file:
SRC_URI += "file://myconfig.cfg \ file://eth.cfg \ file://gfx.cfg"
Another variable you can use in your kernel recipe append
file is the
FILESEXTRAPATHS
variable.
When you use this statement, you are extending the locations
used by the OpenEmbedded system to look for files and
patches as the recipe is processed.
Other methods exist to accomplish grouping and defining
configuration options.
For example, if you are working with a local clone of the
kernel repository, you could checkout the kernel's
meta
branch, make your changes, and
then push the changes to the local bare clone of the
kernel.
The result is that you directly add configuration options
to the meta
branch for your BSP.
The configuration options will likely end up in that
location anyway if the BSP gets added to the Yocto Project.
In general, however, the Yocto Project maintainers take
care of moving the SRC_URI
-specified
configuration options to the kernel's
meta
branch.
Not only is it easier for BSP developers to not have to
worry about putting those configurations in the branch,
but having the maintainers do it allows them to apply
'global' knowledge about the kinds of common configuration
options multiple BSPs in the tree are typically using.
This allows for promotion of common configurations into
common features.
If you have a single patch or a small series of patches
that you want to apply to the Linux kernel source, you
can do so just as you would with any other recipe.
You first copy the patches to the path added to
FILESEXTRAPATHS
in your .bbappend
file as described in
the previous section, and then reference them in
SRC_URI
statements.
For example, you can apply a three-patch series by adding the
following lines to your linux-yocto
.bbappend
file in your layer:
SRC_URI += "file://0001-first-change.patch" SRC_URI += "file://0002-second-change.patch" SRC_URI += "file://0003-third-change.patch"
The next time you run BitBake to build the Linux kernel, BitBake detects the change in the recipe and fetches and applies the patches before building the kernel.
For a detailed example showing how to patch the kernel using
devtool
, see the
"Using devtool
to Patch the Kernel"
and
"Using Traditional Kernel Development to Patch the Kernel"
sections.
You can make wholesale or incremental changes to the final
.config
file used for the eventual
Linux kernel configuration by including a
defconfig
file and by specifying
configuration fragments in the
SRC_URI
to be applied to that file.
If you have a complete, working Linux kernel
.config
file you want to use for the configuration, as before, copy
that file to the appropriate ${PN}
directory in your layer's
recipes-kernel/linux
directory,
and rename the copied file to "defconfig".
Then, add the following lines to the linux-yocto
.bbappend
file in your layer:
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:" SRC_URI += "file://defconfig"
The SRC_URI
tells the build system how to
search for the file, while the
FILESEXTRAPATHS
extends the
FILESPATH
variable (search directories) to include the
${PN}
directory you created to hold the
configuration changes.
defconfig
file before applying any
subsequent configuration fragments.
The final kernel configuration is a combination of the
configurations in the defconfig
file and
any configuration fragments you provide.
You need to realize that if you have any configuration
fragments, the build system applies these on top of and
after applying the existing defconfig
file configurations.
Generally speaking, the preferred approach is to determine the
incremental change you want to make and add that as a
configuration fragment.
For example, if you want to add support for a basic serial
console, create a file named 8250.cfg
in
the ${PN}
directory with the following
content (without indentation):
CONFIG_SERIAL_8250=y CONFIG_SERIAL_8250_CONSOLE=y CONFIG_SERIAL_8250_PCI=y CONFIG_SERIAL_8250_NR_UARTS=4 CONFIG_SERIAL_8250_RUNTIME_UARTS=4 CONFIG_SERIAL_CORE=y CONFIG_SERIAL_CORE_CONSOLE=y
Next, include this configuration fragment and extend the
FILESPATH
variable in your
.bbappend
file:
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:" SRC_URI += "file://8250.cfg"
The next time you run BitBake to build the Linux kernel, BitBake detects the change in the recipe and fetches and applies the new configuration before building the kernel.
For a detailed example showing how to configure the kernel, see the "Configuring the Kernel" section.
defconfig
File¶
It might be desirable to have kernel configuration fragment
support through a defconfig
file that
is pulled from the kernel source tree for the configured
machine.
By default, the OpenEmbedded build system looks for
defconfig
files in the layer used for
Metadata, which is "out-of-tree", and then configures them
using the following:
SRC_URI += "file://defconfig"
If you do not want to maintain copies of
defconfig
files in your layer but would
rather allow users to use the default configuration from the
kernel tree and still be able to add configuration fragments
to the
SRC_URI
through, for example, append files, you can direct the
OpenEmbedded build system to use a
defconfig
file that is "in-tree".
To specify an "in-tree" defconfig
file,
use the following statement form:
KBUILD_DEFCONFIG_KMACHINE
?=defconfig_file
Here is an example that assigns the
KBUILD_DEFCONFIG
variable based on
"raspberrypi2" and provides the path to the "in-tree"
defconfig
file
to be used for a Raspberry Pi 2,
which is based on the Broadcom 2708/2709 chipset:
KBUILD_DEFCONFIG_raspberrypi2 ?= "bcm2709_defconfig"
Aside from modifying your kernel recipe and providing your own
defconfig
file, you need to be sure no
files or statements set SRC_URI
to use a
defconfig
other than your "in-tree"
file (e.g. a kernel's
linux-
machine
.inc
file).
In other words, if the build system detects a statement
that identifies an "out-of-tree"
defconfig
file, that statement
will override your
KBUILD_DEFCONFIG
variable.
See the
KBUILD_DEFCONFIG
variable description for more information.
devtool
to Patch the Kernel¶
The steps in this procedure show you how you can patch the
kernel using the extensible SDK and devtool
.
devtool
"
section.
Patching the kernel involves changing or adding configurations to an existing kernel, changing or adding recipes to the kernel that are needed to support specific hardware features, or even altering the source code itself.
This example creates a simple patch by adding some QEMU emulator
console output at boot time through printk
statements in the kernel's calibrate.c
source
code file.
Applying the patch and booting the modified image causes the added
messages to appear on the emulator's console.
The example is a continuation of the setup procedure found in
the
"Getting Ready to Develop Using devtool
"
Section.
Check Out the Kernel Source Files:
First you must use devtool
to checkout
the kernel source code in its workspace.
Be sure you are in the terminal set up to do work
with the extensible SDK.
Use the following devtool
command
to check out the code:
$ devtool modify linux-yocto
ERROR: Taskhash mismatch 2c793438c2d9f8c3681fd5f7bc819efa versus be3a89ce7c47178880ba7bf6293d7404 for /path/to/esdk/layers/poky/meta/recipes-kernel/linux/linux-yocto_4.10.bb.do_unpackYou can safely ignore these messages. The source code is correctly checked out.
Edit the Source Files Follow these steps to make some simple changes to the source files:
Change the working directory:
In the previous step, the output noted where you can find
the source files (e.g.
~/poky_sdk/workspace/sources/linux-yocto
).
Change to where the kernel source code is before making
your edits to the calibrate.c
file:
$ cd ~/poky_sdk/workspace/sources/linux-yocto
Edit the source file:
Edit the init/calibrate.c
file to have
the following changes:
void calibrate_delay(void) { unsigned long lpj; static bool printed; int this_cpu = smp_processor_id(); printk("*************************************\n"); printk("* *\n"); printk("* HELLO YOCTO KERNEL *\n"); printk("* *\n"); printk("*************************************\n"); if (per_cpu(cpu_loops_per_jiffy, this_cpu)) { . . .
Build the Updated Kernel Source:
To build the updated kernel source, use
devtool
:
$ devtool build linux-yocto
Create the Image With the New Kernel:
Use the devtool build-image
command
to create a new image that has the new kernel.
$ cd ~ $ devtool build-image core-image-minimal
Test the New Image: For this example, you can run the new image using QEMU to verify your changes:
Boot the image: Boot the modified image in the QEMU emulator using this command:
$ runqemu qemux86
Verify the changes:
Log into the machine using root
with no password and then use the following shell
command to scroll through the console's boot output.
# dmesg | less
You should see the results of your
printk
statements
as part of the output when you scroll down the
console window.
Stage and commit your changes:
Within your eSDK terminal, change your working directory to
where you modified the calibrate.c
file and use these Git commands to stage and commit your
changes:
$ cd ~/poky_sdk/workspace/sources/linux-yocto $ git status $ git add init/calibrate.c $ git commit -m "calibrate: Add printk example"
Export the Patches and Create an Append File:
To export your commits as patches and create a
.bbappend
file, use the following
command in the terminal used to work with the extensible
SDK.
This example uses the previously established layer named
meta-mylayer
.
$ devtool finish linux-yocto ~/meta-mylayer
Once the command finishes, the patches and the
.bbappend
file are located in the
~/meta-mylayer/recipes-kernel/linux
directory.
Build the Image With Your Modified Kernel: You can now build an image that includes your kernel patches. Execute the following command from your Build Directory in the terminal set up to run BitBake:
$ cd ~/poky/build $ bitbake core-image-minimal
The steps in this procedure show you how you can patch the
kernel using traditional kernel development (i.e. not using
devtool
and the extensible SDK as
described in the
"Using devtool
to Patch the Kernel"
section).
Patching the kernel involves changing or adding configurations to an existing kernel, changing or adding recipes to the kernel that are needed to support specific hardware features, or even altering the source code itself.
The example in this section creates a simple patch by adding some
QEMU emulator console output at boot time through
printk
statements in the kernel's
calibrate.c
source code file.
Applying the patch and booting the modified image causes the added
messages to appear on the emulator's console.
The example is a continuation of the setup procedure found in
the
"Getting Ready for Traditional Kernel Development"
Section.
Edit the Source Files
Prior to this step, you should have used Git to create a
local copy of the repository for your kernel.
Assuming you created the repository as directed in the
"Getting Ready for Traditional Kernel Development"
section, use the following commands to edit the
calibrate.c
file:
Change the working directory:
You need to locate the source files in the
local copy of the kernel Git repository:
Change to where the kernel source code is before making
your edits to the calibrate.c
file:
$ cd ~/linux-yocto-4.12/init
Edit the source file:
Edit the calibrate.c
file to have
the following changes:
void calibrate_delay(void) { unsigned long lpj; static bool printed; int this_cpu = smp_processor_id(); printk("*************************************\n"); printk("* *\n"); printk("* HELLO YOCTO KERNEL *\n"); printk("* *\n"); printk("*************************************\n"); if (per_cpu(cpu_loops_per_jiffy, this_cpu)) { . . .
Stage and Commit Your Changes: Use standard Git commands to stage and commit the changes you just made:
$ git add calibrate.c $ git commit -m "calibrate.c - Added some printk statements"
If you do not stage and commit your changes, the OpenEmbedded Build System will not pick up the changes.
Update Your local.conf
File
to Point to Your Source Files:
In addition to your local.conf
file
specifying to use "kernel-modules" and the "qemux86"
machine, it must also point to the updated kernel source
files.
Add
SRC_URI
and
SRCREV
statements similar to the following to your
local.conf
:
$ cd ~/poky/build/conf
Add the following to the local.conf
:
SRC_URI_pn-linux-yocto = "git:///path-to
/linux-yocto-4.12;protocol=file;name=machine;branch=standard/base; \ git:///path-to
/yocto-kernel-cache;protocol=file;type=kmeta;name=meta;branch=yocto-4.12;destsuffix=${KMETA}" SRCREV_meta_qemux86 = "${AUTOREV}" SRCREV_machine_qemux86 = "${AUTOREV}"
path-to
with the pathname
to your local Git repositories.
Also, you must be sure to specify the correct branch
and machine types.
For this example, the branch is
standard/base
and the machine is
"qemux86".
Build the Image:
With the source modified, your changes staged and
committed, and the local.conf
file
pointing to the kernel files, you can now use BitBake to
build the image:
$ cd ~/poky/build $ bitbake core-image-minimal
Boot the image: Boot the modified image in the QEMU emulator using this command. When prompted to login to the QEMU console, use "root" with no password:
$ cd ~/poky/build $ runqemu qemux86
Look for Your Changes: As QEMU booted, you might have seen your changes rapidly scroll by. If not, use these commands to see your changes:
# dmesg | less
You should see the results of your
printk
statements
as part of the output when you scroll down the
console window.
Generate the Patch File:
Once you are sure that your patch works correctly, you
can generate a *.patch
file in the
kernel source repository:
$ cd ~/linux-yocto-4.12/init $ git format-patch -1 0001-calibrate.c-Added-some-printk-statements.patch
Move the Patch File to Your Layer:
In order for subsequent builds to pick up patches, you
need to move the patch file you created in the previous
step to your layer meta-mylayer
.
For this example, the layer created earlier is located
in your home directory as meta-mylayer
.
When the layer was created using the
yocto-create
script, no additional
hierarchy was created to support patches.
Before moving the patch file, you need to add additional
structure to your layer using the following commands:
$ cd ~/meta-mylayer $ mkdir recipes-kernel $ mkdir recipes-kernel/linux $ mkdir recipes-kernel/linux/linux-yocto
Once you have created this hierarchy in your layer, you can move the patch file using the following command:
$ mv ~/linux-yocto-4.12/init/0001-calibrate.c-Added-some-printk-statements.patch ~/meta-mylayer/recipes-kernel/linux/linux-yocto
Create the Append File:
Finally, you need to create the
linux-yocto_4.12.bbappend
file and
insert statements that allow the OpenEmbedded build
system to find the patch.
The append file needs to be in your layer's
recipes-kernel/linux
directory and it must be named
linux-yocto_4.12.bbappend
and have
the following contents:
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:" SRC_URI_append = " file://0001-calibrate.c-Added-some-printk-statements.patch"
The
FILESEXTRAPATHS
and
SRC_URI
statements enable the OpenEmbedded build system to find
the patch file.
For more information on append files and patches, see the "Creating the Append File" and "Applying Patches" sections. You can also see the "Using .bbappend Files in Your Layer"" section in the Yocto Project Development Tasks Manual.
core-image-minimal
again and see the effects of your patch, you can
essentially eliminate the temporary source files
saved in poky/build/tmp/work/...
and residual effects of the build by entering the
following sequence of commands:
$ cd ~/poky/build $ bitbake -c cleanall yocto-linux $ bitbake core-image-minimal -c cleanall $ bitbake core-image-minimal $ runqemu qemux86
Configuring the Yocto Project kernel consists of making sure the
.config
file has all the right information
in it for the image you are building.
You can use the menuconfig
tool and
configuration fragments to make sure your
.config
file is just how you need it.
You can also save known configurations in a
defconfig
file that the build system can use
for kernel configuration.
This section describes how to use menuconfig
,
create and use configuration fragments, and how to interactively
modify your .config
file to create the
leanest kernel configuration file possible.
For more information on kernel configuration, see the "Changing the Configuration" section.
menuconfig
¶
The easiest way to define kernel configurations is to set
them through the menuconfig
tool.
This tool provides an interactive method with which
to set kernel configurations.
For general information on menuconfig
, see
http://en.wikipedia.org/wiki/Menuconfig.
To use the menuconfig
tool in the Yocto
Project development environment, you must do the following:
Because you launch menuconfig
using BitBake, you must be sure to set up your
environment by running the
oe-init-build-env
script found in the
Build Directory.
You must be sure of the state of your build's configuration in the Source Directory.
Your build host must have the following two packages installed:
libncurses5-dev libtinfo-dev
The following commands initialize the BitBake environment,
run the
do_kernel_configme
task, and launch menuconfig
.
These commands assume the Source Directory's top-level folder
is ~/poky
:
$ cd poky $ source oe-init-build-env $ bitbake linux-yocto -c kernel_configme -f $ bitbake linux-yocto -c menuconfig
Once menuconfig
comes up, its standard
interface allows you to interactively examine and configure
all the kernel configuration parameters.
After making your changes, simply exit the tool and save your
changes to create an updated version of the
.config
configuration file.
.config
file
as the defconfig
file.
For information on defconfig
files,
see the
"Changing the Configuration",
"Using an In-Tree defconfig
File,
and
"Creating a defconfig
File"
sections.
Consider an example that configures the "CONFIG_SMP" setting
for the linux-yocto-4.12
kernel.
linux-yocto
through Metadata (e.g.
PREFERRED_VERSION
_linux-yocto ?= "12.4%"
).
Once menuconfig
launches, use the
interface to navigate through the selections to find the
configuration settings in which you are interested.
For this example, you deselect "CONFIG_SMP" by clearing the
"Symmetric Multi-Processing Support" option.
Using the interface, you can find the option under
"Processor Type and Features".
To deselect "CONFIG_SMP", use the arrow keys to
highlight "Symmetric Multi-Processing Support" and enter "N"
to clear the asterisk.
When you are finished, exit out and save the change.
Saving the selections updates the .config
configuration file.
This is the file that the OpenEmbedded build system uses to
configure the kernel during the build.
You can find and examine this file in the Build Directory in
tmp/work/
.
The actual .config
is located in the
area where the specific kernel is built.
For example, if you were building a Linux Yocto kernel based
on the linux-yocto-4.12
kernel and you
were building a QEMU image targeted for
x86
architecture, the
.config
file would be:
poky/build/tmp/work/qemux86-poky-linux/linux-yocto/4.12.12+gitAUTOINC+eda4d18... ...967-r0/linux-qemux86-standard-build/.config
Within the .config
file, you can see the
kernel settings.
For example, the following entry shows that symmetric
multi-processor support is not set:
# CONFIG_SMP is not set
A good method to isolate changed configurations is to use a
combination of the menuconfig
tool and
simple shell commands.
Before changing configurations with
menuconfig
, copy the existing
.config
and rename it to something else,
use menuconfig
to make as many changes as
you want and save them, then compare the renamed configuration
file against the newly created file.
You can use the resulting differences as your base to create
configuration fragments to permanently save in your kernel
layer.
.config
file and do not just rename it.
The build system needs an existing
.config
file from which to work.
defconfig
File¶
A defconfig
file in the context of
the Yocto Project is often a .config
file that is copied from a build or a
defconfig
taken from the kernel tree
and moved into recipe space.
You can use a defconfig
file
to retain a known set of kernel configurations from which the
OpenEmbedded build system can draw to create the final
.config
file.
defconfig
or
.config
file.
The OpenEmbedded build system creates the final
.config
file used to configure the
kernel.
To create a defconfig
, start with a
complete, working Linux kernel .config
file.
Copy that file to the appropriate
${
PN
}
directory in your layer's
recipes-kernel/linux
directory, and rename
the copied file to "defconfig" (e.g.
~/meta-mylayer/recipes-kernel/linux/linux-yocto/defconfig
).
Then, add the following lines to the linux-yocto
.bbappend
file in your layer:
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:" SRC_URI += "file://defconfig"
The
SRC_URI
tells the build system how to search for the file, while the
FILESEXTRAPATHS
extends the
FILESPATH
variable (search directories) to include the
${PN}
directory you created to hold the
configuration changes.
defconfig
file before applying any
subsequent configuration fragments.
The final kernel configuration is a combination of the
configurations in the defconfig
file and any configuration fragments you provide.
You need to realize that if you have any configuration
fragments, the build system applies these on top of and
after applying the existing defconfig file configurations.
For more information on configuring the kernel, see the "Changing the Configuration" section.
Configuration fragments are simply kernel options that
appear in a file placed where the OpenEmbedded build system
can find and apply them.
The build system applies configuration fragments after
applying configurations from a defconfig
file.
Thus, the final kernel configuration is a combination of the
configurations in the defconfig
file and then any configuration fragments you provide.
The build system applies fragments on top of and
after applying the existing defconfig file configurations.
Syntactically, the configuration statement is identical to
what would appear in the .config
file,
which is in the
Build Directory.
.config
file is located, see the
example in the
"Using menuconfig
"
section.
It is simple to create a configuration fragment.
One method is to use shell commands.
For example, issuing the following from the shell creates a
configuration fragment file named
my_smp.cfg
that enables multi-processor
support within the kernel:
$ echo "CONFIG_SMP=y" >> my_smp.cfg
.cfg
extension in order for the
OpenEmbedded build system to recognize them as a
configuration fragment.
Another method is to create a configuration fragment using the
differences between two configuration files: one previously
created and saved, and one freshly created using the
menuconfig
tool.
To create a configuration fragment using this method, follow these steps:
Complete a Build Through Kernel Configuration: Complete a build at least through the kernel configuration task as follows:
$ bitbake linux-yocto -c kernel_configme -f
This step ensures that you create a
.config
file from a known state.
Because situations exist where your build state might
become unknown, it is best to run this task prior
to starting menuconfig
.
Launch menuconfig
:
Run the menuconfig
command:
$ bitbake linux-yocto -c menuconfig
Create the Configuration Fragment:
Run the diffconfig
command to prepare a configuration fragment.
The resulting file fragment.cfg
is placed in the
${
WORKDIR
}
directory:
$ bitbake linux-yocto -c diffconfig
The diffconfig
command creates a file
that is a list of Linux kernel CONFIG_
assignments.
See the "Changing the Configuration"
section for additional information on how to use the output
as a configuration fragment.
Where do you put your configuration fragment files?
You can place these files in an area pointed to by
SRC_URI
as directed by your bblayers.conf
file,
which is located in your layer.
The OpenEmbedded build system picks up the configuration and
adds it to the kernel's configuration.
For example, suppose you had a set of configuration options
in a file called myconfig.cfg
.
If you put that file inside a directory named
linux-yocto
that resides in the same
directory as the kernel's append file within your layer
and then add the following statements to the kernel's append
file, those configuration options will be picked up and applied
when the kernel is built:
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:" SRC_URI += "file://myconfig.cfg"
As mentioned earlier, you can group related configurations
into multiple files and name them all in the
SRC_URI
statement as well.
For example, you could group separate configurations
specifically for Ethernet and graphics into their own files
and add those by using a SRC_URI
statement
like the following in your append file:
SRC_URI += "file://myconfig.cfg \ file://eth.cfg \ file://gfx.cfg"
You can use the
do_kernel_configcheck
task to provide configuration validation:
$ bitbake linux-yocto -c kernel_configcheck -f
Running this task produces warnings for when a
requested configuration does not appear in the final
.config
file or when you override a
policy configuration in a hardware configuration fragment.
In order to run this task, you must have an existing
.config
file.
See the
"Using menuconfig
"
section for information on how to create a configuration file.
Following is sample output from the
do_kernel_configcheck
task:
Loading cache: 100% |########################################################| Time: 0:00:00 Loaded 1275 entries from dependency cache. NOTE: Resolving any missing task queue dependencies Build Configuration: . . . NOTE: Executing SetScene Tasks NOTE: Executing RunQueue Tasks WARNING: linux-yocto-4.12.12+gitAUTOINC+eda4d18ce4_16de014967-r0 do_kernel_configcheck: [kernel config]: specified values did not make it into the kernel's final configuration: ---------- CONFIG_X86_TSC ----------------- Config: CONFIG_X86_TSC From: /home/scottrif/poky/build/tmp/work-shared/qemux86/kernel-source/.kernel-meta/configs/standard/bsp/common-pc/common-pc-cpu.cfg Requested value: CONFIG_X86_TSC=y Actual value: ---------- CONFIG_X86_BIGSMP ----------------- Config: CONFIG_X86_BIGSMP From: /home/scottrif/poky/build/tmp/work-shared/qemux86/kernel-source/.kernel-meta/configs/standard/cfg/smp.cfg /home/scottrif/poky/build/tmp/work-shared/qemux86/kernel-source/.kernel-meta/configs/standard/defconfig Requested value: # CONFIG_X86_BIGSMP is not set Actual value: ---------- CONFIG_NR_CPUS ----------------- Config: CONFIG_NR_CPUS From: /home/scottrif/poky/build/tmp/work-shared/qemux86/kernel-source/.kernel-meta/configs/standard/cfg/smp.cfg /home/scottrif/poky/build/tmp/work-shared/qemux86/kernel-source/.kernel-meta/configs/standard/bsp/common-pc/common-pc.cfg /home/scottrif/poky/build/tmp/work-shared/qemux86/kernel-source/.kernel-meta/configs/standard/defconfig Requested value: CONFIG_NR_CPUS=8 Actual value: CONFIG_NR_CPUS=1 ---------- CONFIG_SCHED_SMT ----------------- Config: CONFIG_SCHED_SMT From: /home/scottrif/poky/build/tmp/work-shared/qemux86/kernel-source/.kernel-meta/configs/standard/cfg/smp.cfg /home/scottrif/poky/build/tmp/work-shared/qemux86/kernel-source/.kernel-meta/configs/standard/defconfig Requested value: CONFIG_SCHED_SMT=y Actual value: NOTE: Tasks Summary: Attempted 288 tasks of which 285 didn't need to be rerun and all succeeded. Summary: There were 3 WARNING messages shown.
The output describes the various problems that you can
encounter along with where to find the offending configuration
items.
You can use the information in the logs to adjust your
configuration files and then repeat the
do_kernel_configme
and
do_kernel_configcheck
tasks until they produce no warnings.
For more information on how to use the
menuconfig
tool, see the
"Using menuconfig
"
section.
You can make sure the .config
file is as
lean or efficient as possible by reading the output of the
kernel configuration fragment audit, noting any issues, making
changes to correct the issues, and then repeating.
As part of the kernel build process, the
do_kernel_configcheck
task runs.
This task validates the kernel configuration by checking the
final .config
file against the input
files.
During the check, the task produces warning messages for the
following issues:
Requested options that did not make the final
.config
file.
Configuration items that appear twice in the same configuration fragment.
Configuration items tagged as "required" that were overridden.
A board overrides a non-board specific option.
Listed options not valid for the kernel being processed. In other words, the option does not appear anywhere.
do_kernel_configcheck
task can
also optionally report if an option is overridden during
processing.
For each output warning, a message points to the file that contains a list of the options and a pointer to the configuration fragment that defines them. Collectively, the files are the key to streamlining the configuration.
To streamline the configuration, do the following:
Use a Working Configuration: Start with a full configuration that you know works. Be sure the configuration builds and boots successfully. Use this configuration file as your baseline.
Run Configure and Check Tasks:
Separately run the
do_kernel_configme
and
do_kernel_configcheck
tasks:
$ bitbake linux-yocto -c kernel_configme -f $ bitbake linux-yocto -c kernel_configcheck -f
Process the Results:
Take the resulting list of files from the
do_kernel_configcheck
task
warnings and do the following:
Drop values that are redefined in the fragment
but do not change the final
.config
file.
Analyze and potentially drop values from the
.config
file that override
required configurations.
Analyze and potentially remove non-board specific options.
Remove repeated and invalid options.
Re-Run Configure and Check Tasks:
After you have worked through the output of the kernel
configuration audit, you can re-run the
do_kernel_configme
and
do_kernel_configcheck
tasks to
see the results of your changes.
If you have more issues, you can deal with them as
described in the previous step.
Iteratively working through steps two through four eventually
yields a minimal, streamlined configuration file.
Once you have the best .config
, you can
build the Linux Yocto kernel.
Sometimes it is helpful to determine what a variable expands
to during a build.
You can do examine the values of variables by examining the
output of the bitbake -e
command.
The output is long and is more easily managed in a text file,
which allows for easy searches:
$ bitbake -e virtual/kernel > some_text_file
Within the text file, you can see exactly how each variable is expanded and used by the OpenEmbedded build system.
If you build a kernel image and the version string has a "+" or a "-dirty" at the end, uncommitted modifications exist in the kernel's source directory. Follow these steps to clean up the version string:
Discover the Uncommitted Changes: Go to the kernel's locally cloned Git repository (source directory) and use the following Git command to list the files that have been changed, added, or removed:
$ git status
Commit the Changes: You should commit those changes to the kernel source tree regardless of whether or not you will save, export, or use the changes:
$ git add $ git commit -s -a -m "getting rid of -dirty"
Rebuild the Kernel Image: Once you commit the changes, rebuild the kernel.
Depending on your particular kernel development
workflow, the commands you use to rebuild the
kernel might differ.
For information on building the kernel image when
using devtool
, see the
"Using devtool
to Patch the Kernel"
section.
For information on building the kernel image when
using Bitbake, see the
"Using Traditional Kernel Development to Patch the Kernel"
section.
If you cannot work with one of the Linux kernel versions supported by existing linux-yocto recipes, you can still make use of the Yocto Project Linux kernel tooling by working with your own sources. When you use your own sources, you will not be able to leverage the existing kernel Metadata and stabilization work of the linux-yocto sources. However, you will be able to manage your own Metadata in the same format as the linux-yocto sources. Maintaining format compatibility facilitates converging with linux-yocto on a future, mutually-supported kernel version.
To help you use your own sources, the Yocto Project provides a
linux-yocto custom recipe
(linux-yocto-custom.bb
) that uses
kernel.org
sources
and the Yocto Project Linux kernel tools for managing
kernel Metadata.
You can find this recipe in the
poky
Git repository of the
Yocto Project Source Repository
at:
poky/meta-skeleton/recipes-kernel/linux/linux-yocto-custom.bb
Here are some basic steps you can use to work with your own sources:
Create a Copy of the Kernel Recipe:
Copy the linux-yocto-custom.bb
recipe to your layer and give it a meaningful name.
The name should include the version of the Yocto Linux
kernel you are using (e.g.
linux-yocto-myproject_4.12.bb
,
where "4.12" is the base version of the Linux kernel
with which you would be working).
Create a Directory for Your Patches:
In the same directory inside your layer, create a matching
directory to store your patches and configuration files
(e.g. linux-yocto-myproject
).
Ensure You Have Configurations:
Make sure you have either a defconfig
file or configuration fragment files in your layer.
When you use the linux-yocto-custom.bb
recipe, you must specify a configuration.
If you do not have a defconfig
file,
you can run the following:
$ make defconfig
After running the command, copy the resulting
.config
file to the
files
directory in your layer
as "defconfig" and then add it to the
SRC_URI
variable in the recipe.
Running the make defconfig
command results in the default configuration for your
architecture as defined by your kernel.
However, no guarantee exists that this configuration is
valid for your use case, or that your board will even boot.
This is particularly true for non-x86 architectures.
To use non-x86 defconfig
files,
you need to be more specific and find one that matches your
board (i.e. for arm, you look in
arch/arm/configs
and use the one that
is the best starting point for your board).
Edit the Recipe: Edit the following variables in your recipe as appropriate for your project:
SRC_URI
:
The SRC_URI
should specify
a Git repository that uses one of the supported Git
fetcher protocols (i.e. file
,
git
, http
,
and so forth).
The SRC_URI
variable should
also specify either a defconfig
file or some configuration fragment files.
The skeleton recipe provides an example
SRC_URI
as a syntax reference.
LINUX_VERSION
:
The Linux kernel version you are using (e.g.
"4.12").
LINUX_VERSION_EXTENSION
:
The Linux kernel
CONFIG_LOCALVERSION
that is
compiled into the resulting kernel and visible
through the uname
command.
SRCREV
:
The commit ID from which you want to build.
PR
:
Treat this variable the same as you would in any
other recipe.
Increment the variable to indicate to the
OpenEmbedded build system that the recipe has
changed.
PV
:
The default PV
assignment is
typically adequate.
It combines the LINUX_VERSION
with the Source Control Manager (SCM) revision
as derived from the
SRCPV
variable.
The combined results are a string with the
following form:
3.19.11+git1+68a635bf8dfb64b02263c1ac80c948647cc76d5f_1+218bd8d2022b9852c60d32f0d770931e3cf343e2
While lengthy, the extra verbosity in
PV
helps ensure you are using
the exact sources from which you intend to build.
COMPATIBLE_MACHINE
:
A list of the machines supported by your new recipe.
This variable in the example recipe is set
by default to a regular expression that matches
only the empty string, "(^$)".
This default setting triggers an explicit build
failure.
You must change it to match a list of the machines
that your new recipe supports.
For example, to support the
qemux86
and
qemux86-64
machines, use
the following form:
COMPATIBLE_MACHINE = "qemux86|qemux86-64"
Customize Your Recipe as Needed: Provide further customizations to your recipe as needed just as you would customize an existing linux-yocto recipe. See the "Modifying an Existing Recipe" section for information.
This section describes steps to build out-of-tree modules on your target and describes how to incorporate out-of-tree modules in the build.
While the traditional Yocto Project development model would be to include kernel modules as part of the normal build process, you might find it useful to build modules on the target. This could be the case if your target system is capable and powerful enough to handle the necessary compilation. Before deciding to build on your target, however, you should consider the benefits of using a proper cross-development environment from your build host.
If you want to be able to build out-of-tree modules on
the target, there are some steps you need to take
on the target that is running your SDK image.
Briefly, the kernel-dev
package
is installed by default on all
*.sdk
images and the
kernel-devsrc
package is installed
on many of the *.sdk
images.
However, you need to create some scripts prior to
attempting to build the out-of-tree modules on the target
that is running that image.
Prior to attempting to build the out-of-tree modules,
you need to be on the target as root and you need to
change to the /usr/src/kernel
directory.
Next, make
the scripts:
# cd /usr/src/kernel # make scripts
Because all SDK image recipes include
dev-pkgs
, the
kernel-dev
packages will be installed
as part of the SDK image and the
kernel-devsrc
packages will be installed
as part of applicable SDK images.
The SDK uses the scripts when building out-of-tree
modules.
Once you have switched to that directory and created the
scripts, you should be able to build your out-of-tree modules
on the target.
While it is always preferable to work with sources integrated
into the Linux kernel sources, if you need an external kernel
module, the hello-mod.bb
recipe is
available as a template from which you can create your
own out-of-tree Linux kernel module recipe.
This template recipe is located in the
poky
Git repository of the
Yocto Project Source Repository
at:
poky/meta-skeleton/recipes-kernel/hello-mod/hello-mod_0.1.bb
To get started, copy this recipe to your layer and give it a
meaningful name (e.g. mymodule_1.0.bb
).
In the same directory, create a new directory named
files
where you can store any source files,
patches, or other files necessary for building
the module that do not come with the sources.
Finally, update the recipe as needed for the module.
Typically, you will need to set the following variables:
Depending on the build system used by the module sources,
you might need to make some adjustments.
For example, a typical module Makefile
looks much like the one provided with the
hello-mod
template:
obj-m := hello.o SRC := $(shell pwd) all: $(MAKE) -C $(KERNEL_SRC) M=$(SRC) modules_install: $(MAKE) -C $(KERNEL_SRC) M=$(SRC) modules_install ...
The important point to note here is the
KERNEL_SRC
variable.
The
module
class sets this variable and the
KERNEL_PATH
variable to
${
with the necessary Linux kernel build information to build
modules.
If your module STAGING_KERNEL_DIR
}Makefile
uses a different
variable, you might want to override the
do_compile
step, or create a patch to
the Makefile
to work with the more typical
KERNEL_SRC
or
KERNEL_PATH
variables.
After you have prepared your recipe, you will likely want to include the module in your images. To do this, see the documentation for the following variables in the Yocto Project Reference Manual and set one of them appropriately for your machine configuration file:
Modules are often not required for boot and can be excluded from certain build configurations. The following allows for the most flexibility:
MACHINE_EXTRA_RRECOMMENDS += "kernel-module-mymodule"
The value is derived by appending the module filename without
the .ko
extension to the string
"kernel-module-".
Because the variable is
RRECOMMENDS
and not a
RDEPENDS
variable, the build will not fail if this module is not
available to include in the image.
A common question when working with a kernel is: "What changes have been applied to this tree?" Rather than using "grep" across directories to see what has changed, you can use Git to inspect or search the kernel tree. Using Git is an efficient way to see what has changed in the tree.
Following are a few examples that show how to use Git commands to examine changes. These examples are by no means the only way to see changes.
kernel.org
history is blended
with Yocto Project kernel changes.
You can form ranges by using branch names from the
kernel tree as the upper and lower commit markers with
the Git commands.
You can see the branch names through the web interface
to the Yocto Project source repositories at
http://git.yoctoproject.org.
To see a full range of the changes, use the
git whatchanged
command and specify a
commit range for the branch
(commit
..
commit
).
Here is an example that looks at what has changed in the
emenlow
branch of the
linux-yocto-3.19
kernel.
The lower commit range is the commit associated with the
standard/base
branch, while
the upper commit range is the commit associated with the
standard/emenlow
branch.
$ git whatchanged origin/standard/base..origin/standard/emenlow
To see short, one line summaries of changes use the
git log
command:
$ git log --oneline origin/standard/base..origin/standard/emenlow
Use this command to see code differences for the changes:
$ git diff origin/standard/base..origin/standard/emenlow
Use this command to see the commit log messages and the text differences:
$ git show origin/standard/base..origin/standard/emenlow
Use this command to create individual patches for
each change.
Here is an example that that creates patch files for each
commit and places them in your Documents
directory:
$ git format-patch -o $HOME/Documents origin/standard/base..origin/standard/emenlow
Tags in the Yocto Project kernel tree divide changes for
significant features or branches.
The git show
tag
command shows changes based on a tag.
Here is an example that shows systemtap
changes:
$ git show systemtap
You can use the
git branch --contains
tag
command to show the branches that contain a particular feature.
This command shows the branches that contain the
systemtap
feature:
$ git branch --contains systemtap
You can add kernel features in the
recipe-space by
using the
KERNEL_FEATURES
variable and by specifying the feature's .scc
file path in the
SRC_URI
statement.
When you add features using this method, the OpenEmbedded build
system checks to be sure the features are present.
If the features are not present, the build stops.
Kernel features are the last elements processed for configuring
and patching the kernel.
Therefore, adding features in this manner is a way
to enforce specific features are present and enabled
without needing to do a full audit of any other layer's additions
to the SRC_URI
statement.
You add a kernel feature by providing the feature as part of the
KERNEL_FEATURES
variable and by providing the
path to the feature's .scc
file, which is
relative to the root of the kernel Metadata.
The OpenEmbedded build system searches all forms of kernel
Metadata on the SRC_URI
statement regardless
of whether the Metadata is in the "kernel-cache", system kernel
Metadata, or a recipe-space Metadata (i.e. part of the kernel
recipe).
See the
"Kernel Metadata Location"
section for additional information.
When you specify the feature's .scc
file
on the SRC_URI
statement, the OpenEmbedded
build system adds the directory of that
.scc
file along with all its subdirectories
to the kernel feature search path.
Because subdirectories are searched, you can reference a single
.scc
file in the
SRC_URI
statement to reference multiple kernel
features.
Consider the following example that adds the "test.scc" feature to the build.
Create the Feature File:
Create a .scc
file and locate it
just as you would any other patch file,
.cfg
file, or fetcher item
you specify in the SRC_URI
statement.
You must add the directory of the
.scc
file to the fetcher's
search path in the same manner as you would
add a .patch
file.
You can create additional
.scc
files beneath the
directory that contains the file you are
adding.
All subdirectories are searched during the
build as potential feature directories.
Continuing with the example, suppose the "test.scc"
feature you are adding has a
test.scc
file in the following
directory:
my_recipe
|
+-linux-yocto
|
+-test.cfg
+-test.scc
In this example, the linux-yocto
directory has both the feature
test.scc
file and a similarly
named configuration fragment file
test.cfg
.
Add the Feature File to SRC_URI
:
Add the .scc
file to the
recipe's SRC_URI
statement:
SRC_URI_append = " file://test.scc"
The leading space before the path is important as the path is appended to the existing path.
Specify the Feature as a Kernel Feature:
Use the KERNEL_FEATURES
statement
to specify the feature as a kernel feature:
KERNEL_FEATURES_append = " test.scc"
The OpenEmbedded build system processes the kernel feature when it builds the kernel.
test.scc
file.
yocto-kernel-cache
)¶In addition to supporting configuration fragments and patches, the Yocto Project kernel tools also support rich Metadata that you can use to define complex policies and Board Support Package (BSP) support. The purpose of the Metadata and the tools that manage it is to help you manage the complexity of the configuration and sources used to support multiple BSPs and Linux kernel types.
Kernel Metadata exists in many places.
One area in the Yocto Project
Source Repositories
is the yocto-kernel-cache
Git repository.
You can find this repository grouped under the "Yocto Linux Kernel"
heading in the
Yocto Project Source Repositories.
Kernel development tools ("kern-tools") exist also in the Yocto
Project Source Repositories under the "Yocto Linux Kernel" heading
in the yocto-kernel-tools
Git repository.
The recipe that builds these tools is
meta/recipes-kernel/kern-tools/kern-tools-native_git.bb
in the
Source Directory
(e.g. poky
).
As mentioned in the introduction, the Yocto Project contains kernel
Metadata, which is located in the
yocto-kernel-cache
Git repository.
This Metadata defines Board Support Packages (BSPs) that
correspond to definitions in linux-yocto recipes for corresponding BSPs.
A BSP consists of an aggregation of kernel policy and enabled
hardware-specific features.
The BSP can be influenced from within the linux-yocto recipe.
linux-yocto.inc
file)
is said to be a "linux-yocto style" recipe.
Every linux-yocto style recipe must define the
KMACHINE
variable.
This variable is typically set to the same value as the
MACHINE
variable, which is used by
BitBake.
However, in some cases, the variable might instead refer to the
underlying platform of the MACHINE
.
Multiple BSPs can reuse the same KMACHINE
name if they are built using the same BSP description.
Multiple Corei7-based BSPs could share the same "intel-corei7-64"
value for KMACHINE
.
It is important to realize that KMACHINE
is
just for kernel mapping, while MACHINE
is the machine type within a BSP Layer.
Even with this distinction, however, these two variables can hold
the same value.
See the BSP Descriptions
section for more information.
Every linux-yocto style recipe must also indicate the Linux kernel
source repository branch used to build the Linux kernel.
The KBRANCH
variable must be set to indicate the branch.
KBRANCH
value to define an
alternate branch typically with a machine override as shown here
from the meta-yocto-bsp
layer:
KBRANCH_edgerouter = "standard/edgerouter"
The linux-yocto style recipes can optionally define the following variables:
KERNEL_FEATURES LINUX_KERNEL_TYPE
LINUX_KERNEL_TYPE
defines the kernel type to be
used in assembling the configuration.
If you do not specify a LINUX_KERNEL_TYPE
,
it defaults to "standard".
Together with KMACHINE
,
LINUX_KERNEL_TYPE
defines the search
arguments used by the kernel tools to find the
appropriate description within the kernel Metadata with which to
build out the sources and configuration.
The linux-yocto recipes define "standard", "tiny", and "preempt-rt"
kernel types.
See the "Kernel Types" section
for more information on kernel types.
During the build, the kern-tools search for the BSP description
file that most closely matches the KMACHINE
and LINUX_KERNEL_TYPE
variables passed in from the
recipe.
The tools use the first BSP description it finds that match
both variables.
If the tools cannot find a match, they issue a warning.
The tools first search for the KMACHINE
and
then for the LINUX_KERNEL_TYPE
.
If the tools cannot find a partial match, they will use the
sources from the KBRANCH
and any configuration
specified in the
SRC_URI
.
You can use the
KERNEL_FEATURES
variable
to include features (configuration fragments, patches, or both) that
are not already included by the KMACHINE
and
LINUX_KERNEL_TYPE
variable combination.
For example, to include a feature specified as
"features/netfilter/netfilter.scc",
specify:
KERNEL_FEATURES += "features/netfilter/netfilter.scc"
To include a feature called "cfg/sound.scc" just for the
qemux86
machine, specify:
KERNEL_FEATURES_append_qemux86 = " cfg/sound.scc"
The value of the entries in KERNEL_FEATURES
are dependent on their location within the kernel Metadata itself.
The examples here are taken from the
yocto-kernel-cache
repository.
Each branch of this repository contains "features" and "cfg"
subdirectories at the top-level.
For more information, see the
"Kernel Metadata Syntax"
section.
The kernel Metadata consists of three primary types of files:
scc
[1]
description files, configuration fragments, and patches.
The scc
files define variables and include or
otherwise reference any of the three file types.
The description files are used to aggregate all types of kernel
Metadata into
what ultimately describes the sources and the configuration required
to build a Linux kernel tailored to a specific machine.
The scc
description files are used to define two
fundamental types of kernel Metadata:
Features
Board Support Packages (BSPs)
Features aggregate sources in the form of patches and configuration fragments into a modular reusable unit. You can use features to implement conceptually separate kernel Metadata descriptions such as pure configuration fragments, simple patches, complex features, and kernel types. Kernel types define general kernel features and policy to be reused in the BSPs.
BSPs define hardware-specific features and aggregate them with kernel types to form the final description of what will be assembled and built.
While the kernel Metadata syntax does not enforce any logical separation of configuration fragments, patches, features or kernel types, best practices dictate a logical separation of these types of Metadata. The following Metadata file hierarchy is recommended:
base
/
bsp/
cfg/
features/
ktypes/
patches/
The bsp
directory contains the
BSP descriptions.
The remaining directories all contain "features".
Separating bsp
from the rest of the structure
aids conceptualizing intended usage.
Use these guidelines to help place your scc
description files within the structure:
If your file contains
only configuration fragments, place the file in the
cfg
directory.
If your file contains
only source-code fixes, place the file in the
patches
directory.
If your file encapsulates
a major feature, often combining sources and configurations,
place the file in features
directory.
If your file aggregates
non-hardware configuration and patches in order to define a
base kernel policy or major kernel type to be reused across
multiple BSPs, place the file in ktypes
directory.
These distinctions can easily become blurred - especially as
out-of-tree features slowly merge upstream over time.
Also, remember that how the description files are placed is
a purely logical organization and has no impact on the functionality
of the kernel Metadata.
There is no impact because all of cfg
,
features
, patches
, and
ktypes
, contain "features" as far as the kernel
tools are concerned.
Paths used in kernel Metadata files are relative to
base
, which is either
FILESEXTRAPATHS
if you are creating Metadata in
recipe-space,
or the top level of
yocto-kernel-cache
if you are creating
Metadata outside of the recipe-space.
The simplest unit of kernel Metadata is the configuration-only
feature.
This feature consists of one or more Linux kernel configuration
parameters in a configuration fragment file
(.cfg
) and a .scc
file
that describes the fragment.
As an example, consider the Symmetric Multi-Processing (SMP)
fragment used with the linux-yocto-4.12
kernel as defined outside of the recipe space (i.e.
yocto-kernel-cache
).
This Metadata consists of two files: smp.scc
and smp.cfg
.
You can find these files in the cfg
directory
of the yocto-4.12
branch in the
yocto-kernel-cache
Git repository:
cfg/smp.scc: define KFEATURE_DESCRIPTION "Enable SMP for 32 bit builds" define KFEATURE_COMPATIBILITY all kconf hardware smp.cfg cfg/smp.cfg: CONFIG_SMP=y CONFIG_SCHED_SMT=y # Increase default NR_CPUS from 8 to 64 so that platform with # more than 8 processors can be all activated at boot time CONFIG_NR_CPUS=64 # The following is needed when setting NR_CPUS to something # greater than 8 on x86 architectures, it should be automatically # disregarded by Kconfig when using a different arch CONFIG_X86_BIGSMP=y
You can find general information on configuration fragment files in the "Creating Configuration Fragments" section.
Within the smp.scc
file, the
KFEATURE_DESCRIPTION
statement provides a short description of the fragment.
Higher level kernel tools use this description.
Also within the smp.scc
file, the
kconf
command includes the
actual configuration fragment in an .scc
file, and the "hardware" keyword identifies the fragment as
being hardware enabling, as opposed to general policy,
which would use the "non-hardware" keyword.
The distinction is made for the benefit of the configuration
validation tools, which warn you if a hardware fragment
overrides a policy set by a non-hardware fragment.
kconf
statements, one per fragment.
As described in the "Validating Configuration" section, you can use the following BitBake command to audit your configuration:
$ bitbake linux-yocto -c kernel_configcheck -f
Patch descriptions are very similar to configuration fragment
descriptions, which are described in the previous section.
However, instead of a .cfg
file, these
descriptions work with source patches (i.e.
.patch
files).
A typical patch includes a description file and the patch itself.
As an example, consider the build patches used with the
linux-yocto-4.12
kernel as defined outside of
the recipe space (i.e. yocto-kernel-cache
).
This Metadata consists of several files:
build.scc
and a set of
*.patch
files.
You can find these files in the patches/build
directory of the yocto-4.12
branch in the
yocto-kernel-cache
Git repository.
The following listings show the build.scc
file and part of the
modpost-mask-trivial-warnings.patch
file:
patches/build/build.scc: patch arm-serialize-build-targets.patch patch powerpc-serialize-image-targets.patch patch kbuild-exclude-meta-directory-from-distclean-processi.patch # applied by kgit # patch kbuild-add-meta-files-to-the-ignore-li.patch patch modpost-mask-trivial-warnings.patch patch menuconfig-check-lxdiaglog.sh-Allow-specification-of.patch patches/build/modpost-mask-trivial-warnings.patch: From bd48931bc142bdd104668f3a062a1f22600aae61 Mon Sep 17 00:00:00 2001 From: Paul Gortmaker <paul.gortmaker@windriver.com> Date: Sun, 25 Jan 2009 17:58:09 -0500 Subject: [PATCH] modpost: mask trivial warnings Newer HOSTCC will complain about various stdio fcns because . . . char *dump_write = NULL, *files_source = NULL; int opt; -- 2.10.1 generated by cgit v0.10.2 at 2017-09-28 15:23:23 (GMT)
The description file can include multiple patch statements where
each statement handles a single patch.
In the example build.scc
file, five patch
statements exist for the five patches in the directory.
You can create a typical .patch
file using
diff -Nurp
or
git format-patch
commands.
For information on how to create patches, see the
"Using devtool
to Patch the Kernel"
and
"Using Traditional Kernel Development to Patch the Kernel"
sections.
Features are complex kernel Metadata types that consist of configuration fragments, patches, and possibly other feature description files. As an example, consider the following generic listing:
features/myfeature
.scc define KFEATURE_DESCRIPTION "Enablemyfeature
" patch 0001-myfeature
-core.patch patch 0002-myfeature
-interface.patch include cfg/myfeature
_dependency.scc kconf non-hardwaremyfeature
.cfg
This example shows how the patch
and
kconf
commands are used as well as
how an additional feature description file is included with
the include
command.
Typically, features are less granular than configuration
fragments and are more likely than configuration fragments
and patches to be the types of things you want to specify
in the KERNEL_FEATURES
variable of the
Linux kernel recipe.
See the "Using Kernel Metadata in a Recipe"
section earlier in the manual.
A kernel type defines a high-level kernel policy by
aggregating non-hardware configuration fragments with
patches you want to use when building a Linux kernel of a
specific type (e.g. a real-time kernel).
Syntactically, kernel types are no different than features
as described in the "Features"
section.
The
LINUX_KERNEL_TYPE
variable in the kernel recipe selects the kernel type.
For example, in the linux-yocto_4.12.bb
kernel recipe found in
poky/meta/recipes-kernel/linux
, a
require
directive includes the
poky/meta/recipes-kernel/linux/linux-yocto.inc
file, which has the following statement that defines the default
kernel type:
LINUX_KERNEL_TYPE ??= "standard"
Another example would be the real-time kernel (i.e.
linux-yocto-rt_4.12.bb
).
This kernel recipe directly sets the kernel type as follows:
LINUX_KERNEL_TYPE = "preempt-rt"
meta/recipes-kernel/linux
directory of the
Source Directory
(e.g. poky/meta/recipes-kernel/linux/linux-yocto_4.12.bb
).
See the "Using Kernel Metadata in a Recipe"
section for more information.
Three kernel types ("standard", "tiny", and "preempt-rt") are supported for Linux Yocto kernels:
"standard": Includes the generic Linux kernel policy of the Yocto Project linux-yocto kernel recipes. This policy includes, among other things, which file systems, networking options, core kernel features, and debugging and tracing options are supported.
"preempt-rt":
Applies the PREEMPT_RT
patches and the configuration options required to
build a real-time Linux kernel.
This kernel type inherits from the "standard" kernel type.
"tiny": Defines a bare minimum configuration meant to serve as a base for very small Linux kernels. The "tiny" kernel type is independent from the "standard" configuration. Although the "tiny" kernel type does not currently include any source changes, it might in the future.
For any given kernel type, the Metadata is defined by the
.scc
(e.g. standard.scc
).
Here is a partial listing for the standard.scc
file, which is found in the ktypes/standard
directory of the yocto-kernel-cache
Git
repository:
# Include this kernel type fragment to get the standard features and # configuration values. # Note: if only the features are desired, but not the configuration # then this should be included as: # include ktypes/standard/standard.scc nocfg # if no chained configuration is desired, include it as: # include ktypes/standard/standard.scc nocfg inherit include ktypes/base/base.scc branch standard kconf non-hardware standard.cfg include features/kgdb/kgdb.scc . . . include cfg/net/ip6_nf.scc include cfg/net/bridge.scc include cfg/systemd.scc include features/rfkill/rfkill.scc
As with any .scc
file, a
kernel type definition can aggregate other
.scc
files with
include
commands.
These definitions can also directly pull in
configuration fragments and patches with the
kconf
and patch
commands, respectively.
.scc
file.
The Board Support Package (BSP) file can implicitly define
the kernel type using a define
KTYPE myktype
line.
See the "BSP Descriptions"
section for more information.
BSP descriptions (i.e. *.scc
files)
combine kernel types with hardware-specific features.
The hardware-specific Metadata is typically defined
independently in the BSP layer, and then aggregated with each
supported kernel type.
bsp
directory
of the
yocto-kernel-cache
repository organized under the "Yocto Linux Kernel" heading
in the
Yocto Project Source Repositories.
This section overviews the BSP description structure, the aggregation concepts, and presents a detailed example using a BSP supported by the Yocto Project (i.e. BeagleBone Board). For complete information on BSP layer file hierarchy, see the Yocto Project Board Support Package (BSP) Developer's Guide.
For simplicity, consider the following root BSP layer description files for the BeagleBone board. These files employ both a structure and naming convention for consistency. The naming convention for the file is as follows:
bsp_root_name
-kernel_type
.scc
Here are some example root layer BSP filenames for the BeagleBone Board BSP, which is supported by the Yocto Project:
beaglebone-standard.scc beaglebone-preempt-rt.scc
Each file uses the root name (i.e "beaglebone") BSP name followed by the kernel type.
Examine the beaglebone-standard.scc
file:
define KMACHINE beaglebone define KTYPE standard define KARCH arm include ktypes/standard/standard.scc branch beaglebone include beaglebone.scc # default policy for standard kernels include features/latencytop/latencytop.scc include features/profiling/profiling.scc
Every top-level BSP description file should define the
KMACHINE
,
KTYPE
,
and KARCH
variables.
These variables allow the OpenEmbedded build system to identify
the description as meeting the criteria set by the recipe being
built.
This example supports the "beaglebone" machine for the
"standard" kernel and the "arm" architecture.
Be aware that a hard link between the
KTYPE
variable and a kernel type
description file does not exist.
Thus, if you do not have the kernel type defined in your kernel
Metadata as it is here, you only need to ensure that the
LINUX_KERNEL_TYPE
variable in the kernel recipe and the
KTYPE
variable in the BSP description
file match.
To separate your kernel policy from your hardware configuration,
you include a kernel type (ktype
), such as
"standard".
In the previous example, this is done using the following:
include ktypes/standard/standard.scc
This file aggregates all the configuration fragments, patches, and features that make up your standard kernel policy. See the "Kernel Types" section for more information.
To aggregate common configurations and features specific to the
kernel for mybsp
, use the following:
include mybsp
.scc
You can see that in the BeagleBone example with the following:
include beaglebone.scc
For information on how to break a complete
.config
file into the various
configuration fragments, see the
"Creating Configuration Fragments"
section.
Finally, if you have any configurations specific to the
hardware that are not in a *.scc
file,
you can include them as follows:
kconf hardwaremybsp
-extra
.cfg
The BeagleBone example does not include these types of
configurations.
However, the Malta 32-bit board does ("mti-malta32").
Here is the mti-malta32-le-standard.scc
file:
define KMACHINE mti-malta32-le define KMACHINE qemumipsel define KTYPE standard define KARCH mips include ktypes/standard/standard.scc branch mti-malta32 include mti-malta32.scc kconf hardware mti-malta32-le.cfg
Many real-world examples are more complex.
Like any other .scc
file, BSP
descriptions can aggregate features.
Consider the Minnow BSP definition given the
linux-yocto-4.4
branch of the
yocto-kernel-cache
(i.e.
yocto-kernel-cache/bsp/minnow/minnow.scc
):
include cfg/x86.scc include features/eg20t/eg20t.scc include cfg/dmaengine.scc include features/power/intel.scc include cfg/efi.scc include features/usb/ehci-hcd.scc include features/usb/ohci-hcd.scc include features/usb/usb-gadgets.scc include features/usb/touchscreen-composite.scc include cfg/timer/hpet.scc include features/leds/leds.scc include features/spi/spidev.scc include features/i2c/i2cdev.scc include features/mei/mei-txe.scc # Earlyprintk and port debug requires 8250 kconf hardware cfg/8250.cfg kconf hardware minnow.cfg kconf hardware minnow-dev.cfg
The minnow.scc
description file includes
a hardware configuration fragment
(minnow.cfg
) specific to the Minnow
BSP as well as several more general configuration
fragments and features enabling hardware found on the
machine.
This minnow.scc
description file is then
included in each of the three
"minnow" description files for the supported kernel types
(i.e. "standard", "preempt-rt", and "tiny").
Consider the "minnow" description for the "standard" kernel
type (i.e. minnow-standard.scc
:
define KMACHINE minnow define KTYPE standard define KARCH i386 include ktypes/standard include minnow.scc # Extra minnow configs above the minimal defined in minnow.scc include cfg/efi-ext.scc include features/media/media-all.scc include features/sound/snd_hda_intel.scc # The following should really be in standard.scc # USB live-image support include cfg/usb-mass-storage.scc include cfg/boot-live.scc # Basic profiling include features/latencytop/latencytop.scc include features/profiling/profiling.scc # Requested drivers that don't have an existing scc kconf hardware minnow-drivers-extra.cfg
The include
command midway through the file
includes the minnow.scc
description that
defines all enabled hardware for the BSP that is common to
all kernel types.
Using this command significantly reduces duplication.
Now consider the "minnow" description for the "tiny" kernel
type (i.e. minnow-tiny.scc
):
define KMACHINE minnow define KTYPE tiny define KARCH i386 include ktypes/tiny include minnow.scc
As you might expect, the "tiny" description includes quite a bit less. In fact, it includes only the minimal policy defined by the "tiny" kernel type and the hardware-specific configuration required for booting the machine along with the most basic functionality of the system as defined in the base "minnow" description file.
Notice again the three critical variables:
KMACHINE
,
KTYPE
,
and
KARCH
.
Of these variables, only KTYPE
has changed to specify the "tiny" kernel type.
Kernel Metadata always exists outside of the kernel tree either defined in a kernel recipe (recipe-space) or outside of the recipe. Where you choose to define the Metadata depends on what you want to do and how you intend to work. Regardless of where you define the kernel Metadata, the syntax used applies equally.
If you are unfamiliar with the Linux kernel and only wish to apply a configuration and possibly a couple of patches provided to you by others, the recipe-space method is recommended. This method is also a good approach if you are working with Linux kernel sources you do not control or if you just do not want to maintain a Linux kernel Git repository on your own. For partial information on how you can define kernel Metadata in the recipe-space, see the "Modifying an Existing Recipe" section.
Conversely, if you are actively developing a kernel and are already maintaining a Linux kernel Git repository of your own, you might find it more convenient to work with kernel Metadata kept outside the recipe-space. Working with Metadata in this area can make iterative development of the Linux kernel more efficient outside of the BitBake environment.
When stored in recipe-space, the kernel Metadata files reside in a
directory hierarchy below
FILESEXTRAPATHS
.
For a linux-yocto recipe or for a Linux kernel recipe derived
by copying and modifying
oe-core/meta-skeleton/recipes-kernel/linux/linux-yocto-custom.bb
to a recipe in your layer, FILESEXTRAPATHS
is typically set to
${
THISDIR
}/${
PN
}
.
See the "Modifying an Existing Recipe"
section for more information.
Here is an example that shows a trivial tree of kernel Metadata stored in recipe-space within a BSP layer:
meta-my_bsp_layer
/
`-- recipes-kernel
`-- linux
`-- linux-yocto
|-- bsp-standard.scc
|-- bsp.cfg
`-- standard.cfg
When the Metadata is stored in recipe-space, you must take
steps to ensure BitBake has the necessary information to decide
what files to fetch and when they need to be fetched again.
It is only necessary to specify the .scc
files on the
SRC_URI
.
BitBake parses them and fetches any files referenced in the
.scc
files by the include
,
patch
, or kconf
commands.
Because of this, it is necessary to bump the recipe
PR
value when changing the content of files not explicitly listed
in the SRC_URI
.
If the BSP description is in recipe space, you cannot simply list
the *.scc
in the SRC_URI
statement.
You need to use the following form from your kernel append file:
SRC_URI_append_myplatform
= " \ file://myplatform
;type=kmeta;destsuffix=myplatform
\ "
When stored outside of the recipe-space, the kernel Metadata
files reside in a separate repository.
The OpenEmbedded build system adds the Metadata to the build as
a "type=kmeta" repository through the
SRC_URI
variable.
As an example, consider the following SRC_URI
statement from the linux-yocto_4.12.bb
kernel recipe:
SRC_URI = "git://git.yoctoproject.org/linux-yocto-4.12.git;name=machine;branch=${KBRANCH}; \ git://git.yoctoproject.org/yocto-kernel-cache;type=kmeta;name=meta;branch=yocto-4.12;destsuffix=${KMETA}"
${KMETA}
, in this context, is simply used to
name the directory into which the Git fetcher places the Metadata.
This behavior is no different than any multi-repository
SRC_URI
statement used in a recipe (e.g.
see the previous section).
You can keep kernel Metadata in a "kernel-cache", which is a
directory containing configuration fragments.
As with any Metadata kept outside the recipe-space, you simply
need to use the SRC_URI
statement with the
"type=kmeta" attribute.
Doing so makes the kernel Metadata available during the
configuration phase.
If you modify the Metadata, you must not forget to update the
SRCREV
statements in the kernel's recipe.
In particular, you need to update the
SRCREV_meta
variable to match the commit in
the KMETA
branch you wish to use.
Changing the data in these branches and not updating the
SRCREV
statements to match will cause the
build to fetch an older commit.
Many recipes based on the linux-yocto-custom.bb
recipe use Linux kernel sources that have only a single
branch - "master".
This type of repository structure is fine for linear development
supporting a single machine and architecture.
However, if you work with multiple boards and architectures,
a kernel source repository with multiple branches is more
efficient.
For example, suppose you need a series of patches for one board to boot.
Sometimes, these patches are works-in-progress or fundamentally wrong,
yet they are still necessary for specific boards.
In these situations, you most likely do not want to include these
patches in every kernel you build (i.e. have the patches as part of
the lone "master" branch).
It is situations like these that give rise to multiple branches used
within a Linux kernel sources Git repository.
Repository organization strategies exist that maximize source reuse, remove redundancy, and logically order your changes. This section presents strategies for the following cases:
Encapsulating patches in a feature description and only including the patches in the BSP descriptions of the applicable boards.
Creating a machine branch in your kernel source repository and applying the patches on that branch only.
Creating a feature branch in your kernel source repository and merging that branch into your BSP when needed.
The approach you take is entirely up to you and depends on what works best for your development model.
if you are reusing patches from an external tree and are not working on the patches, you might find the encapsulated feature to be appropriate. Given this scenario, you do not need to create any branches in the source repository. Rather, you just take the static patches you need and encapsulate them within a feature description. Once you have the feature description, you simply include that into the BSP description as described in the "BSP Descriptions" section.
You can find information on how to create patches and BSP descriptions in the "Patches" and "BSP Descriptions" sections.
When you have multiple machines and architectures to support, or you are actively working on board support, it is more efficient to create branches in the repository based on individual machines. Having machine branches allows common source to remain in the "master" branch with any features specific to a machine stored in the appropriate machine branch. This organization method frees you from continually reintegrating your patches into a feature.
Once you have a new branch, you can set up your kernel Metadata
to use the branch a couple different ways.
In the recipe, you can specify the new branch as the
KBRANCH
to use for the board as
follows:
KBRANCH = "mynewbranch"
Another method is to use the branch
command
in the BSP description:
mybsp.scc: define KMACHINE mybsp define KTYPE standard define KARCH i386 include standard.scc branch mynewbranch include mybsp-hw.scc
If you find yourself with numerous branches, you might consider using a hierarchical branching system similar to what the Yocto Linux Kernel Git repositories use:
common
/kernel_type
/machine
If you had two kernel types, "standard" and "small" for
instance, three machines, and common
as mydir
, the branches in your
Git repository might look like this:
mydir/base mydir/standard/base mydir/standard/machine_a mydir/standard/machine_b mydir/standard/machine_c mydir/small/base mydir/small/machine_a
This organization can help clarify the branch relationships.
In this case, mydir/standard/machine_a
includes everything in mydir/base
and
mydir/standard/base
.
The "standard" and "small" branches add sources specific to those
kernel types that for whatever reason are not appropriate for the
other branches.
mydir/standard
and
mydir/standard/machine_a
because it
would have to create a file and a directory named "standard".
When you are actively developing new features, it can be more
efficient to work with that feature as a branch, rather than
as a set of patches that have to be regularly updated.
The Yocto Project Linux kernel tools provide for this with
the git merge
command.
To merge a feature branch into a BSP, insert the
git merge
command after any
branch
commands:
mybsp.scc: define KMACHINE mybsp define KTYPE standard define KARCH i386 include standard.scc branch mynewbranch git merge myfeature include mybsp-hw.scc
This section provides a brief reference for the commands you can use
within an SCC description file (.scc
):
branch [ref]
:
Creates a new branch relative to the current branch
(typically ${KTYPE}
) using
the currently checked-out branch, or "ref" if specified.
define
:
Defines variables, such as
KMACHINE
,
KTYPE
,
KARCH
,
and
KFEATURE_DESCRIPTION
.
include SCC_FILE
:
Includes an SCC file in the current file.
The file is parsed as if you had inserted it inline.
kconf [hardware|non-hardware] CFG_FILE
:
Queues a configuration fragment for merging into the final
Linux .config
file.
git merge GIT_BRANCH
:
Merges the feature branch into the current branch.
patch PATCH_FILE
:
Applies the patch to the current Git branch.
[1]
scc
stands for Series Configuration
Control, but the naming has less significance in the
current implementation of the tooling than it had in the
past.
Consider scc
files to be description files.
Kernels available through the Yocto Project (Yocto Linux kernels),
like other kernels, are based off the Linux kernel releases from
http://www.kernel.org.
At the beginning of a major Linux kernel development cycle, the
Yocto Project team chooses a Linux kernel based on factors such as
release timing, the anticipated release timing of final upstream
kernel.org
versions, and Yocto Project
feature requirements.
Typically, the Linux kernel chosen is in the final stages of
development by the Linux community.
In other words, the Linux kernel is in the release candidate
or "rc" phase and has yet to reach final release.
But, by being in the final stages of external development, the
team knows that the kernel.org
final release
will clearly be within the early stages of the Yocto Project
development window.
This balance allows the Yocto Project team to deliver the most up-to-date Yocto Linux kernel possible, while still ensuring that the team has a stable official release for the baseline Linux kernel version.
As implied earlier, the ultimate source for Yocto Linux kernels
are released kernels from kernel.org
.
In addition to a foundational kernel from
kernel.org
, the available Yocto Linux kernels
contain a mix of important new mainline developments, non-mainline
developments (when no alternative exists), Board Support Package
(BSP) developments, and custom features.
These additions result in a commercially released Yocto
Project Linux kernel that caters to specific embedded designer
needs for targeted hardware.
You can find a web interface to the Yocto Linux kernels in the Source Repositories at http://git.yoctoproject.org. If you look at the interface, you will see to the left a grouping of Git repositories titled "Yocto Linux Kernel". Within this group, you will find several Linux Yocto kernels developed and included with Yocto Project releases:
linux-yocto-4.1
:
The stable Yocto Project kernel to use with the Yocto
Project Release 2.0.
This kernel is based on the Linux 4.1 released kernel.
linux-yocto-4.4
:
The stable Yocto Project kernel to use with the Yocto
Project Release 2.1.
This kernel is based on the Linux 4.4 released kernel.
linux-yocto-4.6
:
A temporary kernel that is not tied to any Yocto Project
release.
linux-yocto-4.8
:
The stable yocto Project kernel to use with the Yocto
Project Release 2.2.
linux-yocto-4.9
:
The stable Yocto Project kernel to use with the Yocto
Project Release 2.3.
This kernel is based on the Linux 4.9 released kernel.
linux-yocto-4.10
:
The default stable Yocto Project kernel to use with the
Yocto Project Release 2.3.
This kernel is based on the Linux 4.10 released kernel.
linux-yocto-4.12
:
The default stable Yocto Project kernel to use with the
Yocto Project Release 2.4.
This kernel is based on the Linux 4.12 released kernel.
yocto-kernel-cache
:
The linux-yocto-cache
contains
patches and configurations for the linux-yocto kernel
tree.
This repository is useful when working on the linux-yocto
kernel.
For more information on this "Advanced Kernel Metadata",
see the
"Working With Advanced Metadata (yocto-kernel-cache
)"
Chapter.
linux-yocto-dev
:
A development kernel based on the latest upstream release
candidate available.
For Yocto Project releases 1.7, 1.8, and 2.0,
the LTSI kernel is
linux-yocto-3.14
.
For Yocto Project releases 2.1, 2.2, and 2.3,
the LTSI kernel is linux-yocto-4.1
.
For Yocto Project release 2.4, the LTSI kernel is
linux-yocto-4.9
linux-yocto-4.4
is an LTS
kernel.
Once a Yocto Linux kernel is officially released, the Yocto Project team goes into their next development cycle, or upward revision (uprev) cycle, while still continuing maintenance on the released kernel. It is important to note that the most sustainable and stable way to include feature development upstream is through a kernel uprev process. Back-porting hundreds of individual fixes and minor features from various kernel versions is not sustainable and can easily compromise quality.
During the uprev cycle, the Yocto Project team uses an ongoing
analysis of Linux kernel development, BSP support, and release
timing to select the best possible kernel.org
Linux kernel version on which to base subsequent Yocto Linux
kernel development.
The team continually monitors Linux community kernel development
to look for significant features of interest.
The team does consider back-porting large features if they have a
significant advantage.
User or community demand can also trigger a back-port or creation
of new functionality in the Yocto Project baseline kernel during
the uprev cycle.
Generally speaking, every new Linux kernel both adds features and introduces new bugs. These consequences are the basic properties of upstream Linux kernel development and are managed by the Yocto Project team's Yocto Linux kernel development strategy. It is the Yocto Project team's policy to not back-port minor features to the released Yocto Linux kernel. They only consider back-porting significant technological jumps ‐ and, that is done after a complete gap analysis. The reason for this policy is that back-porting any small to medium sized change from an evolving Linux kernel can easily create mismatches, incompatibilities and very subtle errors.
The policies described in this section result in both a stable and a cutting edge Yocto Linux kernel that mixes forward ports of existing Linux kernel features and significant and critical new functionality. Forward porting Linux kernel functionality into the Yocto Linux kernels available through the Yocto Project can be thought of as a "micro uprev." The many “micro uprevs” produce a Yocto Linux kernel version with a mix of important new mainline, non-mainline, BSP developments and feature integrations. This Yocto Linux kernel gives insight into new features and allows focused amounts of testing to be done on the kernel, which prevents surprises when selecting the next major uprev. The quality of these cutting edge Yocto Linux kernels is evolving and the kernels are used in leading edge feature and BSP development.
As mentioned earlier, a key goal of the Yocto Project is
to present the developer with a kernel that has a clear and
continuous history that is visible to the user.
The architecture and mechanisms, in particular the branching
strategies, used achieve that goal in a manner similar to
upstream Linux kernel development in
kernel.org
.
You can think of a Yocto Linux kernel as consisting of a baseline Linux kernel with added features logically structured on top of the baseline. The features are tagged and organized by way of a branching strategy implemented by the Yocto Project team using the Source Code Manager (SCM) Git.
Git is the obvious SCM for meeting the Yocto Linux
kernel organizational and structural goals described
in this section.
Not only is Git the SCM for Linux kernel development in
kernel.org
but, Git continues to
grow in popularity and supports many different work
flows, front-ends and management techniques.
You can find documentation on Git at http://git-scm.com/documentation. You can also get an introduction to Git as it applies to the Yocto Project in the "Git" section in the Yocto Project Overview and Concepts Manual. The latter reference provides an overview of Git and presents a minimal set of Git commands that allows you to be functional using Git. You can use as much, or as little, of what Git has to offer to accomplish what you need for your project. You do not have to be a "Git Expert" in order to use it with the Yocto Project.
Using Git's tagging and branching features, the Yocto Project team creates kernel branches at points where functionality is no longer shared and thus, needs to be isolated. For example, board-specific incompatibilities would require different functionality and would require a branch to separate the features. Likewise, for specific kernel features, the same branching strategy is used.
This "tree-like" architecture results in a structure that has features organized to be specific for particular functionality, single kernel types, or a subset of kernel types. Thus, the user has the ability to see the added features and the commits that make up those features. In addition to being able to see added features, the user can also view the history of what made up the baseline Linux kernel.
Another consequence of this strategy results in not having to store the same feature twice internally in the tree. Rather, the kernel team stores the unique differences required to apply the feature onto the kernel type in question.
BSP-specific code additions are handled in a similar manner to kernel-specific additions. Some BSPs only make sense given certain kernel types. So, for these types, the team creates branches off the end of that kernel type for all of the BSPs that are supported on that kernel type. From the perspective of the tools that create the BSP branch, the BSP is really no different than a feature. Consequently, the same branching strategy applies to BSPs as it does to kernel features. So again, rather than store the BSP twice, the team only stores the unique differences for the BSP across the supported multiple kernels.
While this strategy can result in a tree with a significant number
of branches, it is important to realize that from the developer's
point of view, there is a linear path that travels from the
baseline kernel.org
, through a select
group of features and ends with their BSP-specific commits.
In other words, the divisions of the kernel are transparent and
are not relevant to the developer on a day-to-day basis.
From the developer's perspective, this path is the "master" branch
in Git terms.
The developer does not need to be aware of the existence of any
other branches at all.
Of course, value exists in the having these branches in the tree,
should a person decide to explore them.
For example, a comparison between two BSPs at either the commit
level or at the line-by-line code diff
level
is now a trivial operation.
The following illustration shows the conceptual Yocto Linux kernel.
In the illustration, the "Kernel.org Branch Point" marks the specific spot (or Linux kernel release) from which the Yocto Linux kernel is created. From this point forward in the tree, features and differences are organized and tagged.
The "Yocto Project Baseline Kernel" contains functionality that is common to every kernel type and BSP that is organized further along in the tree. Placing these common features in the tree this way means features do not have to be duplicated along individual branches of the tree structure.
From the "Yocto Project Baseline Kernel", branch points represent specific functionality for individual Board Support Packages (BSPs) as well as real-time kernels. The illustration represents this through three BSP-specific branches and a real-time kernel branch. Each branch represents some unique functionality for the BSP or for a real-time Yocto Linux kernel.
In this example structure, the "Real-time (rt) Kernel" branch has common features for all real-time Yocto Linux kernels and contains more branches for individual BSP-specific real-time kernels. The illustration shows three branches as an example. Each branch points the way to specific, unique features for a respective real-time kernel as they apply to a given BSP.
The resulting tree structure presents a clear path of markers (or branches) to the developer that, for all practical purposes, is the Yocto Linux kernel needed for any given set of requirements.
Working with the kernel as a structured tree follows recognized community best practices. In particular, the kernel as shipped with the product, should be considered an "upstream source" and viewed as a series of historical and documented modifications (commits). These modifications represent the development and stabilization done by the Yocto Project kernel development team.
Because commits only change at significant release points in the product life cycle, developers can work on a branch created from the last relevant commit in the shipped Yocto Project Linux kernel. As mentioned previously, the structure is transparent to the developer because the kernel tree is left in this state after cloning and building the kernel.
Upstream storage of all the available kernel source code is one thing, while representing and using the code on your host development system is another. Conceptually, you can think of the kernel source repositories as all the source files necessary for all the supported Yocto Linux kernels. As a developer, you are just interested in the source files for the kernel on which you are working. And, furthermore, you need them available on your host system.
Kernel source code is available on your host system several different ways:
Files Accessed While using devtool
:
devtool
, which is available with the
Yocto Project, is the preferred method by which to
modify the kernel.
See the
"Kernel Modification Workflow"
section.
Cloned Repository: If you are working in the kernel all the time, you probably would want to set up your own local Git repository of the Yocto Linux kernel tree. For information on how to clone a Yocto Linux kernel Git repository, see the "Preparing the Build Host to Work on the Kernel" section.
Temporary Source Files from a Build:
If you just need to make some patches to the kernel using
a traditional BitBake workflow (i.e. not using the
devtool
), you can access temporary
kernel source files that were extracted and used during
a kernel build.
The temporary kernel source files resulting from a build using
BitBake have a particular hierarchy.
When you build the kernel on your development system, all files
needed for the build are taken from the source repositories
pointed to by the
SRC_URI
variable and gathered in a temporary work area where they are
subsequently used to create the unique kernel.
Thus, in a sense, the process constructs a local source tree
specific to your kernel from which to generate the new kernel
image.
The following figure shows the temporary file structure created on your host system when you build the kernel using Bitbake. This Build Directory contains all the source files used during the build.
Again, for additional information on the Yocto Project kernel's
architecture and its branching strategy, see the
"Yocto Linux Kernel Architecture and Branching Strategies"
section.
You can also reference the
"Using devtool
to Patch the Kernel"
and
"Using Traditional Kernel Development to Patch the Kernel"
sections for detailed example that modifies the kernel.
This section describes part of the kernel configuration audit
phase that most developers can ignore.
For general information on kernel configuration including
menuconfig
, defconfig
files, and configuration fragments, see the
"Configuring the Kernel"
section.
During this part of the audit phase, the contents of the final
.config
file are compared against the
fragments specified by the system.
These fragments can be system fragments, distro fragments,
or user-specified configuration elements.
Regardless of their origin, the OpenEmbedded build system
warns the user if a specific option is not included in the
final kernel configuration.
By default, in order to not overwhelm the user with configuration warnings, the system only reports missing "hardware" options as they could result in a boot failure or indicate that important hardware is not available.
To determine whether or not a given option is "hardware" or
"non-hardware", the kernel Metadata in
yocto-kernel-cache
contains files that
classify individual or groups of options as either hardware
or non-hardware.
To better show this, consider a situation where the
yocto-kernel-cache
contains the following
files:
yocto-kernel-cache/features/drm-psb/hardware.cfg yocto-kernel-cache/features/kgdb/hardware.cfg yocto-kernel-cache/ktypes/base/hardware.cfg yocto-kernel-cache/bsp/mti-malta32/hardware.cfg yocto-kernel-cache/bsp/qemu-ppc32/hardware.cfg yocto-kernel-cache/bsp/qemuarma9/hardware.cfg yocto-kernel-cache/bsp/mti-malta64/hardware.cfg yocto-kernel-cache/bsp/arm-versatile-926ejs/hardware.cfg yocto-kernel-cache/bsp/common-pc/hardware.cfg yocto-kernel-cache/bsp/common-pc-64/hardware.cfg yocto-kernel-cache/features/rfkill/non-hardware.cfg yocto-kernel-cache/ktypes/base/non-hardware.cfg yocto-kernel-cache/features/aufs/non-hardware.kcf yocto-kernel-cache/features/ocf/non-hardware.kcf yocto-kernel-cache/ktypes/base/non-hardware.kcf yocto-kernel-cache/ktypes/base/hardware.kcf yocto-kernel-cache/bsp/qemu-ppc32/hardware.kcf
The following list provides explanations for the various files:
hardware.kcf
:
Specifies a list of kernel Kconfig files that contain
hardware options only.
non-hardware.kcf
:
Specifies a list of kernel Kconfig files that contain
non-hardware options only.
hardware.cfg
:
Specifies a list of kernel CONFIG_
options that are hardware, regardless of whether or not
they are within a Kconfig file specified by a hardware
or non-hardware Kconfig file (i.e.
hardware.kcf
or
non-hardware.kcf
).
non-hardware.cfg
:
Specifies a list of kernel CONFIG_
options that are not hardware, regardless of whether or
not they are within a Kconfig file specified by a
hardware or non-hardware Kconfig file (i.e.
hardware.kcf
or
non-hardware.kcf
).
Here is a specific example using the
kernel-cache/bsp/mti-malta32/hardware.cfg
:
CONFIG_SERIAL_8250 CONFIG_SERIAL_8250_CONSOLE CONFIG_SERIAL_8250_NR_UARTS CONFIG_SERIAL_8250_PCI CONFIG_SERIAL_CORE CONFIG_SERIAL_CORE_CONSOLE CONFIG_VGA_ARB
The kernel configuration audit automatically detects these
files (hence the names must be exactly the ones discussed here),
and uses them as inputs when generating warnings about the
final .config
file.
A user-specified kernel Metadata repository, or recipe space
feature, can use these same files to classify options that are
found within its .cfg
files as hardware
or non-hardware, to prevent the OpenEmbedded build system from
producing an error or warning when an option is not in the
final .config
file.
This section describes construction of the Yocto Project kernel
source repositories as accomplished by the Yocto Project team to
create Yocto Linux kernel repositories.
These kernel repositories are found under the heading "Yocto Linux
Kernel" at
http://git.yoctoproject.org
and are shipped as part of a Yocto Project release.
The team creates these repositories by compiling and executing the
set of feature descriptions for every BSP and feature in the
product.
Those feature descriptions list all necessary patches,
configurations, branches, tags, and feature divisions found in a
Yocto Linux kernel.
Thus, the Yocto Project Linux kernel repository (or tree) and
accompanying Metadata in the
yocto-kernel-cache
are built.
The existence of these repositories allow you to access and clone a particular Yocto Project Linux kernel repository and use it to build images based on their configurations and features.
You can find the files used to describe all the valid features and
BSPs in the Yocto Project Linux kernel in any clone of the Yocto
Project Linux kernel source repository and
yocto-kernel-cache
Git trees.
For example, the following commands clone the Yocto Project
baseline Linux kernel that branches off
linux.org
version 4.12 and the
yocto-kernel-cache
, which contains stores of
kernel Metadata:
$ git clone git://git.yoctoproject.org/linux-yocto-4.12 $ git clone git://git.yoctoproject.org/linux-kernel-cache
For more information on how to set up a local Git repository of the Yocto Project Linux kernel files, see the "Preparing the Build Host to Work on the Kernel" section.
Once you have cloned the kernel Git repository and the cache of Metadata on your local machine, you can discover the branches that are available in the repository using the following Git command:
$ git branch -a
Checking out a branch allows you to work with a particular
Yocto Linux kernel.
For example, the following commands check out the
"standard/beagleboard" branch of the Yocto Linux kernel repository
and the "yocto-4.12" branch of the
yocto-kernel-cache
repository:
$ cd ~/linux-yocto-4.12 $ git checkout -b my-kernel-4.12 remotes/origin/standard/beagleboard $ cd ~/linux-kernel-cache $ git checkout -b my-4.12-metadata remotes/origin/yocto-4.12
yocto-kernel-cache
repository correspond to Yocto Linux kernel versions
(e.g. "yocto-4.12", "yocto-4.10", "yocto-4.9", and so forth).
Once you have checked out and switched to appropriate branches, you can see a snapshot of all the kernel source files used to used to build that particular Yocto Linux kernel for a particular board.
To see the features and configurations for a particular Yocto
Linux kernel, you need to examine the
yocto-kernel-cache
Git repository.
As mentioned, branches in the
yocto-kernel-cache
repository correspond to
Yocto Linux kernel versions (e.g. yocto-4.12
).
Branches contain descriptions in the form of
.scc
and .cfg
files.
You should realize, however, that browsing your local
yocto-kernel-cache
repository for feature
descriptions and patches is not an effective way to determine what
is in a particular kernel branch.
Instead, you should use Git directly to discover the changes in
a branch.
Using Git is an efficient and flexible way to inspect changes to
the kernel.
The following steps describe what happens when the Yocto Project Team constructs the Yocto Project kernel source Git repository (or tree) found at http://git.yoctoproject.org given the introduction of a new top-level kernel feature or BSP. The following actions effectively provide the Metadata and create the tree that includes the new feature, patch, or BSP:
Pass Feature to the OpenEmbedded Build System: A top-level kernel feature is passed to the kernel build subsystem. Normally, this feature is a BSP for a particular kernel type.
Locate Feature: The file that describes the top-level feature is located by searching these system directories:
The in-tree kernel-cache directories, which are
located in the
yocto-kernel-cache
repository organized under the "Yocto Linux Kernel"
heading in the
Yocto Project Source Repositories.
Areas pointed to by SRC_URI
statements found in kernel recipes
For a typical build, the target of the search is a
feature description in an .scc
file
whose name follows this format (e.g.
beaglebone-standard.scc
and
beaglebone-preempt-rt.scc
):
bsp_root_name
-kernel_type
.scc
Expand Feature: Once located, the feature description is either expanded into a simple script of actions, or into an existing equivalent script that is already part of the shipped kernel.
Append Extra Features:
Extra features are appended to the top-level feature
description.
These features can come from the
KERNEL_FEATURES
variable in recipes.
Locate, Expand, and Append Each Feature: Each extra feature is located, expanded and appended to the script as described in step three.
Execute the Script:
The script is executed to produce files
.scc
and .cfg
files in appropriate directories of the
yocto-kernel-cache
repository.
These files are descriptions of all the branches, tags,
patches and configurations that need to be applied to the
base Git repository to completely create the
source (build) branch for the new BSP or feature.
Clone Base Repository:
The base repository is cloned, and the actions
listed in the yocto-kernel-cache
directories are applied to the tree.
Perform Cleanup: The Git repositories are left with the desired branches checked out and any required branching, patching and tagging has been performed.
The kernel tree and cache are ready for developer consumption to be locally cloned, configured, and built into a Yocto Project kernel specific to some target hardware.
The generated yocto-kernel-cache
repository adds to the kernel as shipped with the Yocto
Project release.
Any add-ons and configuration data are applied to the
end of an existing branch.
The full repository generation that is found in the
official Yocto Project kernel repositories at
http://git.yoctoproject.org
is the combination of all supported boards and
configurations.
The technique the Yocto Project team uses is flexible and allows for seamless blending of an immutable history with additional patches specific to a deployment. Any additions to the kernel become an integrated part of the branches.
The full kernel tree that you see on
http://git.yoctoproject.org is
generated through repeating the above steps for all
valid BSPs.
The end result is a branched, clean history tree that
makes up the kernel for a given release.
You can see the script (kgit-scc
)
responsible for this in the
yocto-kernel-tools
repository.
The steps used to construct the full kernel tree are the same steps that BitBake uses when it builds a kernel image.
Once you have cloned a Yocto Linux kernel repository and the
cache repository (yocto-kernel-cache
) onto
your development system, you can consider the compilation phase
of kernel development, which is building a kernel image.
Some prerequisites exist that are validated by the build process
before compilation starts:
The
SRC_URI
points to the kernel Git repository.
A BSP build branch with Metadata exists in the
yocto-kernel-cache
repository.
The branch is based on the Yocto Linux kernel version and
has configurations and features grouped under the
yocto-kernel-cache/bsp
directory.
For example, features and configurations for the
BeagleBone Board assuming a
linux-yocto_4.12
kernel reside in the
following area of the yocto-kernel-cache
repository:
yocto-kernel-cache/bsp/beaglebone
yocto-kernel-cache
repository.
The OpenEmbedded build system makes sure these conditions exist before attempting compilation. Other means, however, do exist, such as as bootstrapping a BSP.
Before building a kernel, the build process verifies the tree
and configures the kernel by processing all of the
configuration "fragments" specified by feature descriptions
in the .scc
files.
As the features are compiled, associated kernel configuration
fragments are noted and recorded in the series of directories
in their compilation order.
The fragments are migrated, pre-processed and passed to the
Linux Kernel Configuration subsystem (lkc
) as
raw input in the form of a .config
file.
The lkc
uses its own internal dependency
constraints to do the final processing of that information and
generates the final .config
file that is used
during compilation.
Using the board's architecture and other relevant values from the board's template, kernel compilation is started and a kernel image is produced.
The other thing that you notice once you configure a kernel is that
the build process generates a build tree that is separate from
your kernel's local Git source repository tree.
This build tree has a name that uses the following form, where
${MACHINE}
is the metadata name of the
machine (BSP) and "kernel_type" is one of the Yocto Project
supported kernel types (e.g. "standard"):
linux-${MACHINE}-kernel_type
-build
The existing support in the kernel.org
tree
achieves this default functionality.
This behavior means that all the generated files for a particular
machine or BSP are now in the build tree directory.
The files include the final .config
file,
all the .o
files, the .a
files, and so forth.
Since each machine or BSP has its own separate
Build Directory
in its own separate branch of the Git repository, you can easily
switch between different builds.
Yocto bundles a number of tracing and profiling tools - this 'HOWTO' describes their basic usage and shows by example how to make use of them to examine application and system behavior.
The tools presented are for the most part completely open-ended and have quite good and/or extensive documentation of their own which can be used to solve just about any problem you might come across in Linux. Each section that describes a particular tool has links to that tool's documentation and website.
The purpose of this 'HOWTO' is to present a set of common and generally useful tracing and profiling idioms along with their application (as appropriate) to each tool, in the context of a general-purpose 'drill-down' methodology that can be applied to solving a large number (90%?) of problems. For help with more advanced usages and problems, please see the documentation and/or websites listed for each tool.
The final section of this 'HOWTO' is a collection of real-world examples which we'll be continually adding to as we solve more problems using the tools - feel free to add your own examples to the list!
Most of the tools are available only in 'sdk' images or in images built after adding 'tools-profile' to your local.conf. So, in order to be able to access all of the tools described here, please first build and boot an 'sdk' image e.g.
$ bitbake core-image-sato-sdk
or alternatively by adding 'tools-profile' to the EXTRA_IMAGE_FEATURES line in your local.conf:
EXTRA_IMAGE_FEATURES = "debug-tweaks tools-profile"
If you use the 'tools-profile' method, you don't need to build an sdk image - the tracing and profiling tools will be included in non-sdk images as well e.g.:
$ bitbake core-image-sato
By default, the Yocto build system strips symbols from the binaries it packages, which makes it difficult to use some of the tools.
You can prevent that by setting the
INHIBIT_PACKAGE_STRIP
variable to "1" in your
local.conf
when you build the image:
INHIBIT_PACKAGE_STRIP = "1"
The above setting will noticeably increase the size of your image.
If you've already built a stripped image, you can generate debug packages (xxx-dbg) which you can manually install as needed.
To generate debug info for packages, you can add dbg-pkgs to EXTRA_IMAGE_FEATURES in local.conf. For example:
EXTRA_IMAGE_FEATURES = "debug-tweaks tools-profile dbg-pkgs"
Additionally, in order to generate the right type of
debuginfo, we also need to set
PACKAGE_DEBUG_SPLIT_STYLE
in the local.conf
file:
PACKAGE_DEBUG_SPLIT_STYLE = 'debug-file-directory'
It may seem surprising to see a section covering an 'overall architecture' for what seems to be a random collection of tracing tools that together make up the Linux tracing and profiling space. The fact is, however, that in recent years this seemingly disparate set of tools has started to converge on a 'core' set of underlying mechanisms:
This chapter presents basic usage examples for each of the tracing tools.
The 'perf' tool is the profiling and tracing tool that comes bundled with the Linux kernel.
Don't let the fact that it's part of the kernel fool you into thinking that it's only for tracing and profiling the kernel - you can indeed use it to trace and profile just the kernel, but you can also use it to profile specific applications separately (with or without kernel context), and you can also use it to trace and profile the kernel and all applications on the system simultaneously to gain a system-wide view of what's going on.
In many ways, perf aims to be a superset of all the tracing and profiling tools available in Linux today, including all the other tools covered in this HOWTO. The past couple of years have seen perf subsume a lot of the functionality of those other tools and, at the same time, those other tools have removed large portions of their previous functionality and replaced it with calls to the equivalent functionality now implemented by the perf subsystem. Extrapolation suggests that at some point those other tools will simply become completely redundant and go away; until then, we'll cover those other tools in these pages and in many cases show how the same things can be accomplished in perf and the other tools when it seems useful to do so.
The coverage below details some of the most common ways you'll likely want to apply the tool; full documentation can be found either within the tool itself or in the man pages at perf(1).
For this section, we'll assume you've already performed the basic setup outlined in the General Setup section.
In particular, you'll get the most mileage out of perf if you
profile an image built with the following in your
local.conf
file:
INHIBIT_PACKAGE_STRIP = "1"
perf runs on the target system for the most part. You can archive profile data and copy it to the host for analysis, but for the rest of this document we assume you've ssh'ed to the host and will be running the perf commands on the target.
The perf tool is pretty much self-documenting. To remind yourself of the available commands, simply type 'perf', which will show you basic usage along with the available perf subcommands:
root@crownbay:~# perf usage: perf [--version] [--help] COMMAND [ARGS] The most commonly used perf commands are: annotate Read perf.data (created by perf record) and display annotated code archive Create archive with object files with build-ids found in perf.data file bench General framework for benchmark suites buildid-cache Manage build-id cache. buildid-list List the buildids in a perf.data file diff Read two perf.data files and display the differential profile evlist List the event names in a perf.data file inject Filter to augment the events stream with additional information kmem Tool to trace/measure kernel memory(slab) properties kvm Tool to trace/measure kvm guest os list List all symbolic event types lock Analyze lock events probe Define new dynamic tracepoints record Run a command and record its profile into perf.data report Read perf.data (created by perf record) and display the profile sched Tool to trace/measure scheduler properties (latencies) script Read perf.data (created by perf record) and display trace output stat Run a command and gather performance counter statistics test Runs sanity tests. timechart Tool to visualize total system behavior during a workload top System profiling tool. See 'perf help COMMAND' for more information on a specific command.
As a simple test case, we'll profile the 'wget' of a fairly large file, which is a minimally interesting case because it has both file and network I/O aspects, and at least in the case of standard Yocto images, it's implemented as part of busybox, so the methods we use to analyze it can be used in a very similar way to the whole host of supported busybox applets in Yocto.
root@crownbay:~# rm linux-2.6.19.2.tar.bz2; \ wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2
The quickest and easiest way to get some basic overall data about what's going on for a particular workload is to profile it using 'perf stat'. 'perf stat' basically profiles using a few default counters and displays the summed counts at the end of the run:
root@crownbay:~# perf stat wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2 Connecting to downloads.yoctoproject.org (140.211.169.59:80) linux-2.6.19.2.tar.b 100% |***************************************************| 41727k 0:00:00 ETA Performance counter stats for 'wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2': 4597.223902 task-clock # 0.077 CPUs utilized 23568 context-switches # 0.005 M/sec 68 CPU-migrations # 0.015 K/sec 241 page-faults # 0.052 K/sec 3045817293 cycles # 0.663 GHz <not supported> stalled-cycles-frontend <not supported> stalled-cycles-backend 858909167 instructions # 0.28 insns per cycle 165441165 branches # 35.987 M/sec 19550329 branch-misses # 11.82% of all branches 59.836627620 seconds time elapsed
Many times such a simple-minded test doesn't yield much of interest, but sometimes it does (see Real-world Yocto bug (slow loop-mounted write speed)).
Also, note that 'perf stat' isn't restricted to a fixed set of counters - basically any event listed in the output of 'perf list' can be tallied by 'perf stat'. For example, suppose we wanted to see a summary of all the events related to kernel memory allocation/freeing along with cache hits and misses:
root@crownbay:~# perf stat -e kmem:* -e cache-references -e cache-misses wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2 Connecting to downloads.yoctoproject.org (140.211.169.59:80) linux-2.6.19.2.tar.b 100% |***************************************************| 41727k 0:00:00 ETA Performance counter stats for 'wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2': 5566 kmem:kmalloc 125517 kmem:kmem_cache_alloc 0 kmem:kmalloc_node 0 kmem:kmem_cache_alloc_node 34401 kmem:kfree 69920 kmem:kmem_cache_free 133 kmem:mm_page_free 41 kmem:mm_page_free_batched 11502 kmem:mm_page_alloc 11375 kmem:mm_page_alloc_zone_locked 0 kmem:mm_page_pcpu_drain 0 kmem:mm_page_alloc_extfrag 66848602 cache-references 2917740 cache-misses # 4.365 % of all cache refs 44.831023415 seconds time elapsed
So 'perf stat' gives us a nice easy way to get a quick overview of what might be happening for a set of events, but normally we'd need a little more detail in order to understand what's going on in a way that we can act on in a useful way.
To dive down into a next level of detail, we can use 'perf record'/'perf report' which will collect profiling data and present it to use using an interactive text-based UI (or simply as text if we specify --stdio to 'perf report').
As our first attempt at profiling this workload, we'll simply run 'perf record', handing it the workload we want to profile (everything after 'perf record' and any perf options we hand it - here none - will be executed in a new shell). perf collects samples until the process exits and records them in a file named 'perf.data' in the current working directory.
root@crownbay:~# perf record wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2 Connecting to downloads.yoctoproject.org (140.211.169.59:80) linux-2.6.19.2.tar.b 100% |************************************************| 41727k 0:00:00 ETA [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.176 MB perf.data (~7700 samples) ]
To see the results in a 'text-based UI' (tui), simply run 'perf report', which will read the perf.data file in the current working directory and display the results in an interactive UI:
root@crownbay:~# perf report
The above screenshot displays a 'flat' profile, one entry for each 'bucket' corresponding to the functions that were profiled during the profiling run, ordered from the most popular to the least (perf has options to sort in various orders and keys as well as display entries only above a certain threshold and so on - see the perf documentation for details). Note that this includes both userspace functions (entries containing a [.]) and kernel functions accounted to the process (entries containing a [k]). (perf has command-line modifiers that can be used to restrict the profiling to kernel or userspace, among others).
Notice also that the above report shows an entry for 'busybox', which is the executable that implements 'wget' in Yocto, but that instead of a useful function name in that entry, it displays a not-so-friendly hex value instead. The steps below will show how to fix that problem.
Before we do that, however, let's try running a different profile, one which shows something a little more interesting. The only difference between the new profile and the previous one is that we'll add the -g option, which will record not just the address of a sampled function, but the entire callchain to the sampled function as well:
root@crownbay:~# perf record -g wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2 Connecting to downloads.yoctoproject.org (140.211.169.59:80) linux-2.6.19.2.tar.b 100% |************************************************| 41727k 0:00:00 ETA [ perf record: Woken up 3 times to write data ] [ perf record: Captured and wrote 0.652 MB perf.data (~28476 samples) ] root@crownbay:~# perf report
Using the callgraph view, we can actually see not only which functions took the most time, but we can also see a summary of how those functions were called and learn something about how the program interacts with the kernel in the process.
Notice that each entry in the above screenshot now contains a '+' on the left-hand side. This means that we can expand the entry and drill down into the callchains that feed into that entry. Pressing 'enter' on any one of them will expand the callchain (you can also press 'E' to expand them all at the same time or 'C' to collapse them all).
In the screenshot above, we've toggled the __copy_to_user_ll() entry and several subnodes all the way down. This lets us see which callchains contributed to the profiled __copy_to_user_ll() function which contributed 1.77% to the total profile.
As a bit of background explanation for these callchains, think about what happens at a high level when you run wget to get a file out on the network. Basically what happens is that the data comes into the kernel via the network connection (socket) and is passed to the userspace program 'wget' (which is actually a part of busybox, but that's not important for now), which takes the buffers the kernel passes to it and writes it to a disk file to save it.
The part of this process that we're looking at in the above call stacks is the part where the kernel passes the data it's read from the socket down to wget i.e. a copy-to-user.
Notice also that here there's also a case where the hex value is displayed in the callstack, here in the expanded sys_clock_gettime() function. Later we'll see it resolve to a userspace function call in busybox.
The above screenshot shows the other half of the journey for the data - from the wget program's userspace buffers to disk. To get the buffers to disk, the wget program issues a write(2), which does a copy-from-user to the kernel, which then takes care via some circuitous path (probably also present somewhere in the profile data), to get it safely to disk.
Now that we've seen the basic layout of the profile data and the basics of how to extract useful information out of it, let's get back to the task at hand and see if we can get some basic idea about where the time is spent in the program we're profiling, wget. Remember that wget is actually implemented as an applet in busybox, so while the process name is 'wget', the executable we're actually interested in is busybox. So let's expand the first entry containing busybox:
Again, before we expanded we saw that the function was labeled with a hex value instead of a symbol as with most of the kernel entries. Expanding the busybox entry doesn't make it any better.
The problem is that perf can't find the symbol information for the busybox binary, which is actually stripped out by the Yocto build system.
One way around that is to put the following in your
local.conf
file when you build the image:
INHIBIT_PACKAGE_STRIP = "1"
However, we already have an image with the binaries stripped, so what can we do to get perf to resolve the symbols? Basically we need to install the debuginfo for the busybox package.
To generate the debug info for the packages in the image, we can add dbg-pkgs to EXTRA_IMAGE_FEATURES in local.conf. For example:
EXTRA_IMAGE_FEATURES = "debug-tweaks tools-profile dbg-pkgs"
Additionally, in order to generate the type of debuginfo that
perf understands, we also need to set
PACKAGE_DEBUG_SPLIT_STYLE
in the local.conf
file:
PACKAGE_DEBUG_SPLIT_STYLE = 'debug-file-directory'
Once we've done that, we can install the debuginfo for busybox. The debug packages once built can be found in build/tmp/deploy/rpm/* on the host system. Find the busybox-dbg-...rpm file and copy it to the target. For example:
[trz@empanada core2]$ scp /home/trz/yocto/crownbay-tracing-dbg/build/tmp/deploy/rpm/core2_32/busybox-dbg-1.20.2-r2.core2_32.rpm root@192.168.1.31: root@192.168.1.31's password: busybox-dbg-1.20.2-r2.core2_32.rpm 100% 1826KB 1.8MB/s 00:01
Now install the debug rpm on the target:
root@crownbay:~# rpm -i busybox-dbg-1.20.2-r2.core2_32.rpm
Now that the debuginfo is installed, we see that the busybox entries now display their functions symbolically:
If we expand one of the entries and press 'enter' on a leaf node, we're presented with a menu of actions we can take to get more information related to that entry:
One of these actions allows us to show a view that displays a busybox-centric view of the profiled functions (in this case we've also expanded all the nodes using the 'E' key):
Finally, we can see that now that the busybox debuginfo is installed, the previously unresolved symbol in the sys_clock_gettime() entry mentioned previously is now resolved, and shows that the sys_clock_gettime system call that was the source of 6.75% of the copy-to-user overhead was initiated by the handle_input() busybox function:
At the lowest level of detail, we can dive down to the assembly level and see which instructions caused the most overhead in a function. Pressing 'enter' on the 'udhcpc_main' function, we're again presented with a menu:
Selecting 'Annotate udhcpc_main', we get a detailed listing of percentages by instruction for the udhcpc_main function. From the display, we can see that over 50% of the time spent in this function is taken up by a couple tests and the move of a constant (1) to a register:
As a segue into tracing, let's try another profile using a different counter, something other than the default 'cycles'.
The tracing and profiling infrastructure in Linux has become unified in a way that allows us to use the same tool with a completely different set of counters, not just the standard hardware counters that traditional tools have had to restrict themselves to (of course the traditional tools can also make use of the expanded possibilities now available to them, and in some cases have, as mentioned previously).
We can get a list of the available events that can be used to profile a workload via 'perf list':
root@crownbay:~# perf list List of pre-defined events (to be used in -e): cpu-cycles OR cycles [Hardware event] stalled-cycles-frontend OR idle-cycles-frontend [Hardware event] stalled-cycles-backend OR idle-cycles-backend [Hardware event] instructions [Hardware event] cache-references [Hardware event] cache-misses [Hardware event] branch-instructions OR branches [Hardware event] branch-misses [Hardware event] bus-cycles [Hardware event] ref-cycles [Hardware event] cpu-clock [Software event] task-clock [Software event] page-faults OR faults [Software event] minor-faults [Software event] major-faults [Software event] context-switches OR cs [Software event] cpu-migrations OR migrations [Software event] alignment-faults [Software event] emulation-faults [Software event] L1-dcache-loads [Hardware cache event] L1-dcache-load-misses [Hardware cache event] L1-dcache-prefetch-misses [Hardware cache event] L1-icache-loads [Hardware cache event] L1-icache-load-misses [Hardware cache event] . . . rNNN [Raw hardware event descriptor] cpu/t1=v1[,t2=v2,t3 ...]/modifier [Raw hardware event descriptor] (see 'perf list --help' on how to encode it) mem:<addr>[:access] [Hardware breakpoint] sunrpc:rpc_call_status [Tracepoint event] sunrpc:rpc_bind_status [Tracepoint event] sunrpc:rpc_connect_status [Tracepoint event] sunrpc:rpc_task_begin [Tracepoint event] skb:kfree_skb [Tracepoint event] skb:consume_skb [Tracepoint event] skb:skb_copy_datagram_iovec [Tracepoint event] net:net_dev_xmit [Tracepoint event] net:net_dev_queue [Tracepoint event] net:netif_receive_skb [Tracepoint event] net:netif_rx [Tracepoint event] napi:napi_poll [Tracepoint event] sock:sock_rcvqueue_full [Tracepoint event] sock:sock_exceed_buf_limit [Tracepoint event] udp:udp_fail_queue_rcv_skb [Tracepoint event] hda:hda_send_cmd [Tracepoint event] hda:hda_get_response [Tracepoint event] hda:hda_bus_reset [Tracepoint event] scsi:scsi_dispatch_cmd_start [Tracepoint event] scsi:scsi_dispatch_cmd_error [Tracepoint event] scsi:scsi_eh_wakeup [Tracepoint event] drm:drm_vblank_event [Tracepoint event] drm:drm_vblank_event_queued [Tracepoint event] drm:drm_vblank_event_delivered [Tracepoint event] random:mix_pool_bytes [Tracepoint event] random:mix_pool_bytes_nolock [Tracepoint event] random:credit_entropy_bits [Tracepoint event] gpio:gpio_direction [Tracepoint event] gpio:gpio_value [Tracepoint event] block:block_rq_abort [Tracepoint event] block:block_rq_requeue [Tracepoint event] block:block_rq_issue [Tracepoint event] block:block_bio_bounce [Tracepoint event] block:block_bio_complete [Tracepoint event] block:block_bio_backmerge [Tracepoint event] . . writeback:writeback_wake_thread [Tracepoint event] writeback:writeback_wake_forker_thread [Tracepoint event] writeback:writeback_bdi_register [Tracepoint event] . . writeback:writeback_single_inode_requeue [Tracepoint event] writeback:writeback_single_inode [Tracepoint event] kmem:kmalloc [Tracepoint event] kmem:kmem_cache_alloc [Tracepoint event] kmem:mm_page_alloc [Tracepoint event] kmem:mm_page_alloc_zone_locked [Tracepoint event] kmem:mm_page_pcpu_drain [Tracepoint event] kmem:mm_page_alloc_extfrag [Tracepoint event] vmscan:mm_vmscan_kswapd_sleep [Tracepoint event] vmscan:mm_vmscan_kswapd_wake [Tracepoint event] vmscan:mm_vmscan_wakeup_kswapd [Tracepoint event] vmscan:mm_vmscan_direct_reclaim_begin [Tracepoint event] . . module:module_get [Tracepoint event] module:module_put [Tracepoint event] module:module_request [Tracepoint event] sched:sched_kthread_stop [Tracepoint event] sched:sched_wakeup [Tracepoint event] sched:sched_wakeup_new [Tracepoint event] sched:sched_process_fork [Tracepoint event] sched:sched_process_exec [Tracepoint event] sched:sched_stat_runtime [Tracepoint event] rcu:rcu_utilization [Tracepoint event] workqueue:workqueue_queue_work [Tracepoint event] workqueue:workqueue_execute_end [Tracepoint event] signal:signal_generate [Tracepoint event] signal:signal_deliver [Tracepoint event] timer:timer_init [Tracepoint event] timer:timer_start [Tracepoint event] timer:hrtimer_cancel [Tracepoint event] timer:itimer_state [Tracepoint event] timer:itimer_expire [Tracepoint event] irq:irq_handler_entry [Tracepoint event] irq:irq_handler_exit [Tracepoint event] irq:softirq_entry [Tracepoint event] irq:softirq_exit [Tracepoint event] irq:softirq_raise [Tracepoint event] printk:console [Tracepoint event] task:task_newtask [Tracepoint event] task:task_rename [Tracepoint event] syscalls:sys_enter_socketcall [Tracepoint event] syscalls:sys_exit_socketcall [Tracepoint event] . . . syscalls:sys_enter_unshare [Tracepoint event] syscalls:sys_exit_unshare [Tracepoint event] raw_syscalls:sys_enter [Tracepoint event] raw_syscalls:sys_exit [Tracepoint event]
Only a subset of these would be of interest to us when looking at this workload, so let's choose the most likely subsystems (identified by the string before the colon in the Tracepoint events) and do a 'perf stat' run using only those wildcarded subsystems:
root@crownbay:~# perf stat -e skb:* -e net:* -e napi:* -e sched:* -e workqueue:* -e irq:* -e syscalls:* wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2 Performance counter stats for 'wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2': 23323 skb:kfree_skb 0 skb:consume_skb 49897 skb:skb_copy_datagram_iovec 6217 net:net_dev_xmit 6217 net:net_dev_queue 7962 net:netif_receive_skb 2 net:netif_rx 8340 napi:napi_poll 0 sched:sched_kthread_stop 0 sched:sched_kthread_stop_ret 3749 sched:sched_wakeup 0 sched:sched_wakeup_new 0 sched:sched_switch 29 sched:sched_migrate_task 0 sched:sched_process_free 1 sched:sched_process_exit 0 sched:sched_wait_task 0 sched:sched_process_wait 0 sched:sched_process_fork 1 sched:sched_process_exec 0 sched:sched_stat_wait 2106519415641 sched:sched_stat_sleep 0 sched:sched_stat_iowait 147453613 sched:sched_stat_blocked 12903026955 sched:sched_stat_runtime 0 sched:sched_pi_setprio 3574 workqueue:workqueue_queue_work 3574 workqueue:workqueue_activate_work 0 workqueue:workqueue_execute_start 0 workqueue:workqueue_execute_end 16631 irq:irq_handler_entry 16631 irq:irq_handler_exit 28521 irq:softirq_entry 28521 irq:softirq_exit 28728 irq:softirq_raise 1 syscalls:sys_enter_sendmmsg 1 syscalls:sys_exit_sendmmsg 0 syscalls:sys_enter_recvmmsg 0 syscalls:sys_exit_recvmmsg 14 syscalls:sys_enter_socketcall 14 syscalls:sys_exit_socketcall . . . 16965 syscalls:sys_enter_read 16965 syscalls:sys_exit_read 12854 syscalls:sys_enter_write 12854 syscalls:sys_exit_write . . . 58.029710972 seconds time elapsed
Let's pick one of these tracepoints and tell perf to do a profile using it as the sampling event:
root@crownbay:~# perf record -g -e sched:sched_wakeup wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2
The screenshot above shows the results of running a profile using sched:sched_switch tracepoint, which shows the relative costs of various paths to sched_wakeup (note that sched_wakeup is the name of the tracepoint - it's actually defined just inside ttwu_do_wakeup(), which accounts for the function name actually displayed in the profile:
/* * Mark the task runnable and perform wakeup-preemption. */ static void ttwu_do_wakeup(struct rq *rq, struct task_struct *p, int wake_flags) { trace_sched_wakeup(p, true); . . . }
A couple of the more interesting callchains are expanded and displayed above, basically some network receive paths that presumably end up waking up wget (busybox) when network data is ready.
Note that because tracepoints are normally used for tracing, the default sampling period for tracepoints is 1 i.e. for tracepoints perf will sample on every event occurrence (this can be changed using the -c option). This is in contrast to hardware counters such as for example the default 'cycles' hardware counter used for normal profiling, where sampling periods are much higher (in the thousands) because profiling should have as low an overhead as possible and sampling on every cycle would be prohibitively expensive.
Profiling is a great tool for solving many problems or for getting a high-level view of what's going on with a workload or across the system. It is however by definition an approximation, as suggested by the most prominent word associated with it, 'sampling'. On the one hand, it allows a representative picture of what's going on in the system to be cheaply taken, but on the other hand, that cheapness limits its utility when that data suggests a need to 'dive down' more deeply to discover what's really going on. In such cases, the only way to see what's really going on is to be able to look at (or summarize more intelligently) the individual steps that go into the higher-level behavior exposed by the coarse-grained profiling data.
As a concrete example, we can trace all the events we think might be applicable to our workload:
root@crownbay:~# perf record -g -e skb:* -e net:* -e napi:* -e sched:sched_switch -e sched:sched_wakeup -e irq:* -e syscalls:sys_enter_read -e syscalls:sys_exit_read -e syscalls:sys_enter_write -e syscalls:sys_exit_write wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2
We can look at the raw trace output using 'perf script' with no arguments:
root@crownbay:~# perf script perf 1262 [000] 11624.857082: sys_exit_read: 0x0 perf 1262 [000] 11624.857193: sched_wakeup: comm=migration/0 pid=6 prio=0 success=1 target_cpu=000 wget 1262 [001] 11624.858021: softirq_raise: vec=1 [action=TIMER] wget 1262 [001] 11624.858074: softirq_entry: vec=1 [action=TIMER] wget 1262 [001] 11624.858081: softirq_exit: vec=1 [action=TIMER] wget 1262 [001] 11624.858166: sys_enter_read: fd: 0x0003, buf: 0xbf82c940, count: 0x0200 wget 1262 [001] 11624.858177: sys_exit_read: 0x200 wget 1262 [001] 11624.858878: kfree_skb: skbaddr=0xeb248d80 protocol=0 location=0xc15a5308 wget 1262 [001] 11624.858945: kfree_skb: skbaddr=0xeb248000 protocol=0 location=0xc15a5308 wget 1262 [001] 11624.859020: softirq_raise: vec=1 [action=TIMER] wget 1262 [001] 11624.859076: softirq_entry: vec=1 [action=TIMER] wget 1262 [001] 11624.859083: softirq_exit: vec=1 [action=TIMER] wget 1262 [001] 11624.859167: sys_enter_read: fd: 0x0003, buf: 0xb7720000, count: 0x0400 wget 1262 [001] 11624.859192: sys_exit_read: 0x1d7 wget 1262 [001] 11624.859228: sys_enter_read: fd: 0x0003, buf: 0xb7720000, count: 0x0400 wget 1262 [001] 11624.859233: sys_exit_read: 0x0 wget 1262 [001] 11624.859573: sys_enter_read: fd: 0x0003, buf: 0xbf82c580, count: 0x0200 wget 1262 [001] 11624.859584: sys_exit_read: 0x200 wget 1262 [001] 11624.859864: sys_enter_read: fd: 0x0003, buf: 0xb7720000, count: 0x0400 wget 1262 [001] 11624.859888: sys_exit_read: 0x400 wget 1262 [001] 11624.859935: sys_enter_read: fd: 0x0003, buf: 0xb7720000, count: 0x0400 wget 1262 [001] 11624.859944: sys_exit_read: 0x400
This gives us a detailed timestamped sequence of events that occurred within the workload with respect to those events.
In many ways, profiling can be viewed as a subset of tracing - theoretically, if you have a set of trace events that's sufficient to capture all the important aspects of a workload, you can derive any of the results or views that a profiling run can.
Another aspect of traditional profiling is that while powerful in many ways, it's limited by the granularity of the underlying data. Profiling tools offer various ways of sorting and presenting the sample data, which make it much more useful and amenable to user experimentation, but in the end it can't be used in an open-ended way to extract data that just isn't present as a consequence of the fact that conceptually, most of it has been thrown away.
Full-blown detailed tracing data does however offer the opportunity to manipulate and present the information collected during a tracing run in an infinite variety of ways.
Another way to look at it is that there are only so many ways that the 'primitive' counters can be used on their own to generate interesting output; to get anything more complicated than simple counts requires some amount of additional logic, which is typically very specific to the problem at hand. For example, if we wanted to make use of a 'counter' that maps to the value of the time difference between when a process was scheduled to run on a processor and the time it actually ran, we wouldn't expect such a counter to exist on its own, but we could derive one called say 'wakeup_latency' and use it to extract a useful view of that metric from trace data. Likewise, we really can't figure out from standard profiling tools how much data every process on the system reads and writes, along with how many of those reads and writes fail completely. If we have sufficient trace data, however, we could with the right tools easily extract and present that information, but we'd need something other than pre-canned profiling tools to do that.
Luckily, there is a general-purpose way to handle such needs, called 'programming languages'. Making programming languages easily available to apply to such problems given the specific format of data is called a 'programming language binding' for that data and language. Perf supports two programming language bindings, one for Python and one for Perl.
Now that we have the trace data in perf.data, we can use 'perf script -g' to generate a skeleton script with handlers for the read/write entry/exit events we recorded:
root@crownbay:~# perf script -g python generated Python script: perf-script.py
The skeleton script simply creates a python function for each event type in the perf.data file. The body of each function simply prints the event name along with its parameters. For example:
def net__netif_rx(event_name, context, common_cpu, common_secs, common_nsecs, common_pid, common_comm, skbaddr, len, name): print_header(event_name, common_cpu, common_secs, common_nsecs, common_pid, common_comm) print "skbaddr=%u, len=%u, name=%s\n" % (skbaddr, len, name),
We can run that script directly to print all of the events contained in the perf.data file:
root@crownbay:~# perf script -s perf-script.py in trace_begin syscalls__sys_exit_read 0 11624.857082795 1262 perf nr=3, ret=0 sched__sched_wakeup 0 11624.857193498 1262 perf comm=migration/0, pid=6, prio=0, success=1, target_cpu=0 irq__softirq_raise 1 11624.858021635 1262 wget vec=TIMER irq__softirq_entry 1 11624.858074075 1262 wget vec=TIMER irq__softirq_exit 1 11624.858081389 1262 wget vec=TIMER syscalls__sys_enter_read 1 11624.858166434 1262 wget nr=3, fd=3, buf=3213019456, count=512 syscalls__sys_exit_read 1 11624.858177924 1262 wget nr=3, ret=512 skb__kfree_skb 1 11624.858878188 1262 wget skbaddr=3945041280, location=3243922184, protocol=0 skb__kfree_skb 1 11624.858945608 1262 wget skbaddr=3945037824, location=3243922184, protocol=0 irq__softirq_raise 1 11624.859020942 1262 wget vec=TIMER irq__softirq_entry 1 11624.859076935 1262 wget vec=TIMER irq__softirq_exit 1 11624.859083469 1262 wget vec=TIMER syscalls__sys_enter_read 1 11624.859167565 1262 wget nr=3, fd=3, buf=3077701632, count=1024 syscalls__sys_exit_read 1 11624.859192533 1262 wget nr=3, ret=471 syscalls__sys_enter_read 1 11624.859228072 1262 wget nr=3, fd=3, buf=3077701632, count=1024 syscalls__sys_exit_read 1 11624.859233707 1262 wget nr=3, ret=0 syscalls__sys_enter_read 1 11624.859573008 1262 wget nr=3, fd=3, buf=3213018496, count=512 syscalls__sys_exit_read 1 11624.859584818 1262 wget nr=3, ret=512 syscalls__sys_enter_read 1 11624.859864562 1262 wget nr=3, fd=3, buf=3077701632, count=1024 syscalls__sys_exit_read 1 11624.859888770 1262 wget nr=3, ret=1024 syscalls__sys_enter_read 1 11624.859935140 1262 wget nr=3, fd=3, buf=3077701632, count=1024 syscalls__sys_exit_read 1 11624.859944032 1262 wget nr=3, ret=1024
That in itself isn't very useful; after all, we can accomplish pretty much the same thing by simply running 'perf script' without arguments in the same directory as the perf.data file.
We can however replace the print statements in the generated function bodies with whatever we want, and thereby make it infinitely more useful.
As a simple example, let's just replace the print statements in the function bodies with a simple function that does nothing but increment a per-event count. When the program is run against a perf.data file, each time a particular event is encountered, a tally is incremented for that event. For example:
def net__netif_rx(event_name, context, common_cpu, common_secs, common_nsecs, common_pid, common_comm, skbaddr, len, name): inc_counts(event_name)
Each event handler function in the generated code is modified to do this. For convenience, we define a common function called inc_counts() that each handler calls; inc_counts() simply tallies a count for each event using the 'counts' hash, which is a specialized hash function that does Perl-like autovivification, a capability that's extremely useful for kinds of multi-level aggregation commonly used in processing traces (see perf's documentation on the Python language binding for details):
counts = autodict() def inc_counts(event_name): try: counts[event_name] += 1 except TypeError: counts[event_name] = 1
Finally, at the end of the trace processing run, we want to print the result of all the per-event tallies. For that, we use the special 'trace_end()' function:
def trace_end(): for event_name, count in counts.iteritems(): print "%-40s %10s\n" % (event_name, count)
The end result is a summary of all the events recorded in the trace:
skb__skb_copy_datagram_iovec 13148 irq__softirq_entry 4796 irq__irq_handler_exit 3805 irq__softirq_exit 4795 syscalls__sys_enter_write 8990 net__net_dev_xmit 652 skb__kfree_skb 4047 sched__sched_wakeup 1155 irq__irq_handler_entry 3804 irq__softirq_raise 4799 net__net_dev_queue 652 syscalls__sys_enter_read 17599 net__netif_receive_skb 1743 syscalls__sys_exit_read 17598 net__netif_rx 2 napi__napi_poll 1877 syscalls__sys_exit_write 8990
Note that this is pretty much exactly the same information we get from 'perf stat', which goes a little way to support the idea mentioned previously that given the right kind of trace data, higher-level profiling-type summaries can be derived from it.
Documentation on using the 'perf script' python binding.
The examples so far have focused on tracing a particular program or workload - in other words, every profiling run has specified the program to profile in the command-line e.g. 'perf record wget ...'.
It's also possible, and more interesting in many cases, to run a system-wide profile or trace while running the workload in a separate shell.
To do system-wide profiling or tracing, you typically use the -a flag to 'perf record'.
To demonstrate this, open up one window and start the profile using the -a flag (press Ctrl-C to stop tracing):
root@crownbay:~# perf record -g -a ^C[ perf record: Woken up 6 times to write data ] [ perf record: Captured and wrote 1.400 MB perf.data (~61172 samples) ]
In another window, run the wget test:
root@crownbay:~# wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2 Connecting to downloads.yoctoproject.org (140.211.169.59:80) linux-2.6.19.2.tar.b 100% |*******************************| 41727k 0:00:00 ETA
Here we see entries not only for our wget load, but for other processes running on the system as well:
In the snapshot above, we can see callchains that originate in libc, and a callchain from Xorg that demonstrates that we're using a proprietary X driver in userspace (notice the presence of 'PVR' and some other unresolvable symbols in the expanded Xorg callchain).
Note also that we have both kernel and userspace entries in the above snapshot. We can also tell perf to focus on userspace but providing a modifier, in this case 'u', to the 'cycles' hardware counter when we record a profile:
root@crownbay:~# perf record -g -a -e cycles:u ^C[ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 0.376 MB perf.data (~16443 samples) ]
Notice in the screenshot above, we see only userspace entries ([.])
Finally, we can press 'enter' on a leaf node and select the 'Zoom into DSO' menu item to show only entries associated with a specific DSO. In the screenshot below, we've zoomed into the 'libc' DSO which shows all the entries associated with the libc-xxx.so DSO.
We can also use the system-wide -a switch to do system-wide tracing. Here we'll trace a couple of scheduler events:
root@crownbay:~# perf record -a -e sched:sched_switch -e sched:sched_wakeup ^C[ perf record: Woken up 38 times to write data ] [ perf record: Captured and wrote 9.780 MB perf.data (~427299 samples) ]
We can look at the raw output using 'perf script' with no arguments:
root@crownbay:~# perf script perf 1383 [001] 6171.460045: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001 perf 1383 [001] 6171.460066: sched_switch: prev_comm=perf prev_pid=1383 prev_prio=120 prev_state=R+ ==> next_comm=kworker/1:1 next_pid=21 next_prio=120 kworker/1:1 21 [001] 6171.460093: sched_switch: prev_comm=kworker/1:1 prev_pid=21 prev_prio=120 prev_state=S ==> next_comm=perf next_pid=1383 next_prio=120 swapper 0 [000] 6171.468063: sched_wakeup: comm=kworker/0:3 pid=1209 prio=120 success=1 target_cpu=000 swapper 0 [000] 6171.468107: sched_switch: prev_comm=swapper/0 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=kworker/0:3 next_pid=1209 next_prio=120 kworker/0:3 1209 [000] 6171.468143: sched_switch: prev_comm=kworker/0:3 prev_pid=1209 prev_prio=120 prev_state=S ==> next_comm=swapper/0 next_pid=0 next_prio=120 perf 1383 [001] 6171.470039: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001 perf 1383 [001] 6171.470058: sched_switch: prev_comm=perf prev_pid=1383 prev_prio=120 prev_state=R+ ==> next_comm=kworker/1:1 next_pid=21 next_prio=120 kworker/1:1 21 [001] 6171.470082: sched_switch: prev_comm=kworker/1:1 prev_pid=21 prev_prio=120 prev_state=S ==> next_comm=perf next_pid=1383 next_prio=120 perf 1383 [001] 6171.480035: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001
Notice that there are a lot of events that don't really have anything to do with what we're interested in, namely events that schedule 'perf' itself in and out or that wake perf up. We can get rid of those by using the '--filter' option - for each event we specify using -e, we can add a --filter after that to filter out trace events that contain fields with specific values:
root@crownbay:~# perf record -a -e sched:sched_switch --filter 'next_comm != perf && prev_comm != perf' -e sched:sched_wakeup --filter 'comm != perf' ^C[ perf record: Woken up 38 times to write data ] [ perf record: Captured and wrote 9.688 MB perf.data (~423279 samples) ] root@crownbay:~# perf script swapper 0 [000] 7932.162180: sched_switch: prev_comm=swapper/0 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=kworker/0:3 next_pid=1209 next_prio=120 kworker/0:3 1209 [000] 7932.162236: sched_switch: prev_comm=kworker/0:3 prev_pid=1209 prev_prio=120 prev_state=S ==> next_comm=swapper/0 next_pid=0 next_prio=120 perf 1407 [001] 7932.170048: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001 perf 1407 [001] 7932.180044: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001 perf 1407 [001] 7932.190038: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001 perf 1407 [001] 7932.200044: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001 perf 1407 [001] 7932.210044: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001 perf 1407 [001] 7932.220044: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001 swapper 0 [001] 7932.230111: sched_wakeup: comm=kworker/1:1 pid=21 prio=120 success=1 target_cpu=001 swapper 0 [001] 7932.230146: sched_switch: prev_comm=swapper/1 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=kworker/1:1 next_pid=21 next_prio=120 kworker/1:1 21 [001] 7932.230205: sched_switch: prev_comm=kworker/1:1 prev_pid=21 prev_prio=120 prev_state=S ==> next_comm=swapper/1 next_pid=0 next_prio=120 swapper 0 [000] 7932.326109: sched_wakeup: comm=kworker/0:3 pid=1209 prio=120 success=1 target_cpu=000 swapper 0 [000] 7932.326171: sched_switch: prev_comm=swapper/0 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=kworker/0:3 next_pid=1209 next_prio=120 kworker/0:3 1209 [000] 7932.326214: sched_switch: prev_comm=kworker/0:3 prev_pid=1209 prev_prio=120 prev_state=S ==> next_comm=swapper/0 next_pid=0 next_prio=120
In this case, we've filtered out all events that have 'perf' in their 'comm' or 'comm_prev' or 'comm_next' fields. Notice that there are still events recorded for perf, but notice that those events don't have values of 'perf' for the filtered fields. To completely filter out anything from perf will require a bit more work, but for the purpose of demonstrating how to use filters, it's close enough.
perf isn't restricted to the fixed set of static tracepoints listed by 'perf list'. Users can also add their own 'dynamic' tracepoints anywhere in the kernel. For instance, suppose we want to define our own tracepoint on do_fork(). We can do that using the 'perf probe' perf subcommand:
root@crownbay:~# perf probe do_fork Added new event: probe:do_fork (on do_fork) You can now use it in all perf tools, such as: perf record -e probe:do_fork -aR sleep 1
Adding a new tracepoint via 'perf probe' results in an event with all the expected files and format in /sys/kernel/debug/tracing/events, just the same as for static tracepoints (as discussed in more detail in the trace events subsystem section:
root@crownbay:/sys/kernel/debug/tracing/events/probe/do_fork# ls -al drwxr-xr-x 2 root root 0 Oct 28 11:42 . drwxr-xr-x 3 root root 0 Oct 28 11:42 .. -rw-r--r-- 1 root root 0 Oct 28 11:42 enable -rw-r--r-- 1 root root 0 Oct 28 11:42 filter -r--r--r-- 1 root root 0 Oct 28 11:42 format -r--r--r-- 1 root root 0 Oct 28 11:42 id root@crownbay:/sys/kernel/debug/tracing/events/probe/do_fork# cat format name: do_fork ID: 944 format: field:unsigned short common_type; offset:0; size:2; signed:0; field:unsigned char common_flags; offset:2; size:1; signed:0; field:unsigned char common_preempt_count; offset:3; size:1; signed:0; field:int common_pid; offset:4; size:4; signed:1; field:int common_padding; offset:8; size:4; signed:1; field:unsigned long __probe_ip; offset:12; size:4; signed:0; print fmt: "(%lx)", REC->__probe_ip
We can list all dynamic tracepoints currently in existence:
root@crownbay:~# perf probe -l probe:do_fork (on do_fork) probe:schedule (on schedule)
Let's record system-wide ('sleep 30' is a trick for recording system-wide but basically do nothing and then wake up after 30 seconds):
root@crownbay:~# perf record -g -a -e probe:do_fork sleep 30 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.087 MB perf.data (~3812 samples) ]
Using 'perf script' we can see each do_fork event that fired:
root@crownbay:~# perf script # ======== # captured on: Sun Oct 28 11:55:18 2012 # hostname : crownbay # os release : 3.4.11-yocto-standard # perf version : 3.4.11 # arch : i686 # nrcpus online : 2 # nrcpus avail : 2 # cpudesc : Intel(R) Atom(TM) CPU E660 @ 1.30GHz # cpuid : GenuineIntel,6,38,1 # total memory : 1017184 kB # cmdline : /usr/bin/perf record -g -a -e probe:do_fork sleep 30 # event : name = probe:do_fork, type = 2, config = 0x3b0, config1 = 0x0, config2 = 0x0, excl_usr = 0, excl_kern = 0, id = { 5, 6 } # HEADER_CPU_TOPOLOGY info available, use -I to display # ======== # matchbox-deskto 1197 [001] 34211.378318: do_fork: (c1028460) matchbox-deskto 1295 [001] 34211.380388: do_fork: (c1028460) pcmanfm 1296 [000] 34211.632350: do_fork: (c1028460) pcmanfm 1296 [000] 34211.639917: do_fork: (c1028460) matchbox-deskto 1197 [001] 34217.541603: do_fork: (c1028460) matchbox-deskto 1299 [001] 34217.543584: do_fork: (c1028460) gthumb 1300 [001] 34217.697451: do_fork: (c1028460) gthumb 1300 [001] 34219.085734: do_fork: (c1028460) gthumb 1300 [000] 34219.121351: do_fork: (c1028460) gthumb 1300 [001] 34219.264551: do_fork: (c1028460) pcmanfm 1296 [000] 34219.590380: do_fork: (c1028460) matchbox-deskto 1197 [001] 34224.955965: do_fork: (c1028460) matchbox-deskto 1306 [001] 34224.957972: do_fork: (c1028460) matchbox-termin 1307 [000] 34225.038214: do_fork: (c1028460) matchbox-termin 1307 [001] 34225.044218: do_fork: (c1028460) matchbox-termin 1307 [000] 34225.046442: do_fork: (c1028460) matchbox-deskto 1197 [001] 34237.112138: do_fork: (c1028460) matchbox-deskto 1311 [001] 34237.114106: do_fork: (c1028460) gaku 1312 [000] 34237.202388: do_fork: (c1028460)
And using 'perf report' on the same file, we can see the callgraphs from starting a few programs during those 30 seconds:
Online versions of the man pages for the commands discussed in this section can be found here:
The 'perf stat' manpage.
The 'perf probe' manpage.
Documentation on using the 'perf script' python binding.
The top-level perf(1) manpage.
Normally, you should be able to invoke the man pages via perf itself e.g. 'perf help' or 'perf help record'.
However, by default Yocto doesn't install man pages, but perf invokes the man pages for most help functionality. This is a bug and is being addressed by a Yocto bug: Bug 3388 - perf: enable man pages for basic 'help' functionality.
The man pages in text form, along with some other files, such as a set of examples, can be found in the 'perf' directory of the kernel tree:
tools/perf/Documentation
There's also a nice perf tutorial on the perf wiki that goes into more detail than we do here in certain areas: Perf Tutorial
'ftrace' literally refers to the 'ftrace function tracer' but in reality this encompasses a number of related tracers along with the infrastructure that they all make use of.
For this section, we'll assume you've already performed the basic setup outlined in the General Setup section.
ftrace, trace-cmd, and kernelshark run on the target system, and are ready to go out-of-the-box - no additional setup is necessary. For the rest of this section we assume you've ssh'ed to the host and will be running ftrace on the target. kernelshark is a GUI application and if you use the '-X' option to ssh you can have the kernelshark GUI run on the target but display remotely on the host if you want.
'ftrace' essentially refers to everything included in the /tracing directory of the mounted debugfs filesystem (Yocto follows the standard convention and mounts it at /sys/kernel/debug). Here's a listing of all the files found in /sys/kernel/debug/tracing on a Yocto system:
root@sugarbay:/sys/kernel/debug/tracing# ls README kprobe_events trace available_events kprobe_profile trace_clock available_filter_functions options trace_marker available_tracers per_cpu trace_options buffer_size_kb printk_formats trace_pipe buffer_total_size_kb saved_cmdlines tracing_cpumask current_tracer set_event tracing_enabled dyn_ftrace_total_info set_ftrace_filter tracing_on enabled_functions set_ftrace_notrace tracing_thresh events set_ftrace_pid free_buffer set_graph_function
The files listed above are used for various purposes - some relate directly to the tracers themselves, others are used to set tracing options, and yet others actually contain the tracing output when a tracer is in effect. Some of the functions can be guessed from their names, others need explanation; in any case, we'll cover some of the files we see here below but for an explanation of the others, please see the ftrace documentation.
We'll start by looking at some of the available built-in tracers.
cat'ing the 'available_tracers' file lists the set of available tracers:
root@sugarbay:/sys/kernel/debug/tracing# cat available_tracers blk function_graph function nop
The 'current_tracer' file contains the tracer currently in effect:
root@sugarbay:/sys/kernel/debug/tracing# cat current_tracer nop
The above listing of current_tracer shows that the 'nop' tracer is in effect, which is just another way of saying that there's actually no tracer currently in effect.
echo'ing one of the available_tracers into current_tracer makes the specified tracer the current tracer:
root@sugarbay:/sys/kernel/debug/tracing# echo function > current_tracer root@sugarbay:/sys/kernel/debug/tracing# cat current_tracer function
The above sets the current tracer to be the 'function tracer'. This tracer traces every function call in the kernel and makes it available as the contents of the 'trace' file. Reading the 'trace' file lists the currently buffered function calls that have been traced by the function tracer:
root@sugarbay:/sys/kernel/debug/tracing# cat trace | less # tracer: function # # entries-in-buffer/entries-written: 310629/766471 #P:8 # # _-----=> irqs-off # / _----=> need-resched # | / _---=> hardirq/softirq # || / _--=> preempt-depth # ||| / delay # TASK-PID CPU# |||| TIMESTAMP FUNCTION # | | | |||| | | <idle>-0 [004] d..1 470.867169: ktime_get_real <-intel_idle <idle>-0 [004] d..1 470.867170: getnstimeofday <-ktime_get_real <idle>-0 [004] d..1 470.867171: ns_to_timeval <-intel_idle <idle>-0 [004] d..1 470.867171: ns_to_timespec <-ns_to_timeval <idle>-0 [004] d..1 470.867172: smp_apic_timer_interrupt <-apic_timer_interrupt <idle>-0 [004] d..1 470.867172: native_apic_mem_write <-smp_apic_timer_interrupt <idle>-0 [004] d..1 470.867172: irq_enter <-smp_apic_timer_interrupt <idle>-0 [004] d..1 470.867172: rcu_irq_enter <-irq_enter <idle>-0 [004] d..1 470.867173: rcu_idle_exit_common.isra.33 <-rcu_irq_enter <idle>-0 [004] d..1 470.867173: local_bh_disable <-irq_enter <idle>-0 [004] d..1 470.867173: add_preempt_count <-local_bh_disable <idle>-0 [004] d.s1 470.867174: tick_check_idle <-irq_enter <idle>-0 [004] d.s1 470.867174: tick_check_oneshot_broadcast <-tick_check_idle <idle>-0 [004] d.s1 470.867174: ktime_get <-tick_check_idle <idle>-0 [004] d.s1 470.867174: tick_nohz_stop_idle <-tick_check_idle <idle>-0 [004] d.s1 470.867175: update_ts_time_stats <-tick_nohz_stop_idle <idle>-0 [004] d.s1 470.867175: nr_iowait_cpu <-update_ts_time_stats <idle>-0 [004] d.s1 470.867175: tick_do_update_jiffies64 <-tick_check_idle <idle>-0 [004] d.s1 470.867175: _raw_spin_lock <-tick_do_update_jiffies64 <idle>-0 [004] d.s1 470.867176: add_preempt_count <-_raw_spin_lock <idle>-0 [004] d.s2 470.867176: do_timer <-tick_do_update_jiffies64 <idle>-0 [004] d.s2 470.867176: _raw_spin_lock <-do_timer <idle>-0 [004] d.s2 470.867176: add_preempt_count <-_raw_spin_lock <idle>-0 [004] d.s3 470.867177: ntp_tick_length <-do_timer <idle>-0 [004] d.s3 470.867177: _raw_spin_lock_irqsave <-ntp_tick_length . . .
Each line in the trace above shows what was happening in the kernel on a given cpu, to the level of detail of function calls. Each entry shows the function called, followed by its caller (after the arrow).
The function tracer gives you an extremely detailed idea of what the kernel was doing at the point in time the trace was taken, and is a great way to learn about how the kernel code works in a dynamic sense.
It is a little more difficult to follow the call chains than it needs to be - luckily there's a variant of the function tracer that displays the callchains explicitly, called the 'function_graph' tracer:
root@sugarbay:/sys/kernel/debug/tracing# echo function_graph > current_tracer root@sugarbay:/sys/kernel/debug/tracing# cat trace | less tracer: function_graph CPU DURATION FUNCTION CALLS | | | | | | | 7) 0.046 us | pick_next_task_fair(); 7) 0.043 us | pick_next_task_stop(); 7) 0.042 us | pick_next_task_rt(); 7) 0.032 us | pick_next_task_fair(); 7) 0.030 us | pick_next_task_idle(); 7) | _raw_spin_unlock_irq() { 7) 0.033 us | sub_preempt_count(); 7) 0.258 us | } 7) 0.032 us | sub_preempt_count(); 7) + 13.341 us | } /* __schedule */ 7) 0.095 us | } /* sub_preempt_count */ 7) | schedule() { 7) | __schedule() { 7) 0.060 us | add_preempt_count(); 7) 0.044 us | rcu_note_context_switch(); 7) | _raw_spin_lock_irq() { 7) 0.033 us | add_preempt_count(); 7) 0.247 us | } 7) | idle_balance() { 7) | _raw_spin_unlock() { 7) 0.031 us | sub_preempt_count(); 7) 0.246 us | } 7) | update_shares() { 7) 0.030 us | __rcu_read_lock(); 7) 0.029 us | __rcu_read_unlock(); 7) 0.484 us | } 7) 0.030 us | __rcu_read_lock(); 7) | load_balance() { 7) | find_busiest_group() { 7) 0.031 us | idle_cpu(); 7) 0.029 us | idle_cpu(); 7) 0.035 us | idle_cpu(); 7) 0.906 us | } 7) 1.141 us | } 7) 0.022 us | msecs_to_jiffies(); 7) | load_balance() { 7) | find_busiest_group() { 7) 0.031 us | idle_cpu(); . . . 4) 0.062 us | msecs_to_jiffies(); 4) 0.062 us | __rcu_read_unlock(); 4) | _raw_spin_lock() { 4) 0.073 us | add_preempt_count(); 4) 0.562 us | } 4) + 17.452 us | } 4) 0.108 us | put_prev_task_fair(); 4) 0.102 us | pick_next_task_fair(); 4) 0.084 us | pick_next_task_stop(); 4) 0.075 us | pick_next_task_rt(); 4) 0.062 us | pick_next_task_fair(); 4) 0.066 us | pick_next_task_idle(); ------------------------------------------ 4) kworker-74 => <idle>-0 ------------------------------------------ 4) | finish_task_switch() { 4) | _raw_spin_unlock_irq() { 4) 0.100 us | sub_preempt_count(); 4) 0.582 us | } 4) 1.105 us | } 4) 0.088 us | sub_preempt_count(); 4) ! 100.066 us | } . . . 3) | sys_ioctl() { 3) 0.083 us | fget_light(); 3) | security_file_ioctl() { 3) 0.066 us | cap_file_ioctl(); 3) 0.562 us | } 3) | do_vfs_ioctl() { 3) | drm_ioctl() { 3) 0.075 us | drm_ut_debug_printk(); 3) | i915_gem_pwrite_ioctl() { 3) | i915_mutex_lock_interruptible() { 3) 0.070 us | mutex_lock_interruptible(); 3) 0.570 us | } 3) | drm_gem_object_lookup() { 3) | _raw_spin_lock() { 3) 0.080 us | add_preempt_count(); 3) 0.620 us | } 3) | _raw_spin_unlock() { 3) 0.085 us | sub_preempt_count(); 3) 0.562 us | } 3) 2.149 us | } 3) 0.133 us | i915_gem_object_pin(); 3) | i915_gem_object_set_to_gtt_domain() { 3) 0.065 us | i915_gem_object_flush_gpu_write_domain(); 3) 0.065 us | i915_gem_object_wait_rendering(); 3) 0.062 us | i915_gem_object_flush_cpu_write_domain(); 3) 1.612 us | } 3) | i915_gem_object_put_fence() { 3) 0.097 us | i915_gem_object_flush_fence.constprop.36(); 3) 0.645 us | } 3) 0.070 us | add_preempt_count(); 3) 0.070 us | sub_preempt_count(); 3) 0.073 us | i915_gem_object_unpin(); 3) 0.068 us | mutex_unlock(); 3) 9.924 us | } 3) + 11.236 us | } 3) + 11.770 us | } 3) + 13.784 us | } 3) | sys_ioctl() {
As you can see, the function_graph display is much easier to follow. Also note that in addition to the function calls and associated braces, other events such as scheduler events are displayed in context. In fact, you can freely include any tracepoint available in the trace events subsystem described in the next section by simply enabling those events, and they'll appear in context in the function graph display. Quite a powerful tool for understanding kernel dynamics.
Also notice that there are various annotations on the left hand side of the display. For example if the total time it took for a given function to execute is above a certain threshold, an exclamation point or plus sign appears on the left hand side. Please see the ftrace documentation for details on all these fields.
One especially important directory contained within the /sys/kernel/debug/tracing directory is the 'events' subdirectory, which contains representations of every tracepoint in the system. Listing out the contents of the 'events' subdirectory, we see mainly another set of subdirectories:
root@sugarbay:/sys/kernel/debug/tracing# cd events root@sugarbay:/sys/kernel/debug/tracing/events# ls -al drwxr-xr-x 38 root root 0 Nov 14 23:19 . drwxr-xr-x 5 root root 0 Nov 14 23:19 .. drwxr-xr-x 19 root root 0 Nov 14 23:19 block drwxr-xr-x 32 root root 0 Nov 14 23:19 btrfs drwxr-xr-x 5 root root 0 Nov 14 23:19 drm -rw-r--r-- 1 root root 0 Nov 14 23:19 enable drwxr-xr-x 40 root root 0 Nov 14 23:19 ext3 drwxr-xr-x 79 root root 0 Nov 14 23:19 ext4 drwxr-xr-x 14 root root 0 Nov 14 23:19 ftrace drwxr-xr-x 8 root root 0 Nov 14 23:19 hda -r--r--r-- 1 root root 0 Nov 14 23:19 header_event -r--r--r-- 1 root root 0 Nov 14 23:19 header_page drwxr-xr-x 25 root root 0 Nov 14 23:19 i915 drwxr-xr-x 7 root root 0 Nov 14 23:19 irq drwxr-xr-x 12 root root 0 Nov 14 23:19 jbd drwxr-xr-x 14 root root 0 Nov 14 23:19 jbd2 drwxr-xr-x 14 root root 0 Nov 14 23:19 kmem drwxr-xr-x 7 root root 0 Nov 14 23:19 module drwxr-xr-x 3 root root 0 Nov 14 23:19 napi drwxr-xr-x 6 root root 0 Nov 14 23:19 net drwxr-xr-x 3 root root 0 Nov 14 23:19 oom drwxr-xr-x 12 root root 0 Nov 14 23:19 power drwxr-xr-x 3 root root 0 Nov 14 23:19 printk drwxr-xr-x 8 root root 0 Nov 14 23:19 random drwxr-xr-x 4 root root 0 Nov 14 23:19 raw_syscalls drwxr-xr-x 3 root root 0 Nov 14 23:19 rcu drwxr-xr-x 6 root root 0 Nov 14 23:19 rpm drwxr-xr-x 20 root root 0 Nov 14 23:19 sched drwxr-xr-x 7 root root 0 Nov 14 23:19 scsi drwxr-xr-x 4 root root 0 Nov 14 23:19 signal drwxr-xr-x 5 root root 0 Nov 14 23:19 skb drwxr-xr-x 4 root root 0 Nov 14 23:19 sock drwxr-xr-x 10 root root 0 Nov 14 23:19 sunrpc drwxr-xr-x 538 root root 0 Nov 14 23:19 syscalls drwxr-xr-x 4 root root 0 Nov 14 23:19 task drwxr-xr-x 14 root root 0 Nov 14 23:19 timer drwxr-xr-x 3 root root 0 Nov 14 23:19 udp drwxr-xr-x 21 root root 0 Nov 14 23:19 vmscan drwxr-xr-x 3 root root 0 Nov 14 23:19 vsyscall drwxr-xr-x 6 root root 0 Nov 14 23:19 workqueue drwxr-xr-x 26 root root 0 Nov 14 23:19 writeback
Each one of these subdirectories corresponds to a 'subsystem' and contains yet again more subdirectories, each one of those finally corresponding to a tracepoint. For example, here are the contents of the 'kmem' subsystem:
root@sugarbay:/sys/kernel/debug/tracing/events# cd kmem root@sugarbay:/sys/kernel/debug/tracing/events/kmem# ls -al drwxr-xr-x 14 root root 0 Nov 14 23:19 . drwxr-xr-x 38 root root 0 Nov 14 23:19 .. -rw-r--r-- 1 root root 0 Nov 14 23:19 enable -rw-r--r-- 1 root root 0 Nov 14 23:19 filter drwxr-xr-x 2 root root 0 Nov 14 23:19 kfree drwxr-xr-x 2 root root 0 Nov 14 23:19 kmalloc drwxr-xr-x 2 root root 0 Nov 14 23:19 kmalloc_node drwxr-xr-x 2 root root 0 Nov 14 23:19 kmem_cache_alloc drwxr-xr-x 2 root root 0 Nov 14 23:19 kmem_cache_alloc_node drwxr-xr-x 2 root root 0 Nov 14 23:19 kmem_cache_free drwxr-xr-x 2 root root 0 Nov 14 23:19 mm_page_alloc drwxr-xr-x 2 root root 0 Nov 14 23:19 mm_page_alloc_extfrag drwxr-xr-x 2 root root 0 Nov 14 23:19 mm_page_alloc_zone_locked drwxr-xr-x 2 root root 0 Nov 14 23:19 mm_page_free drwxr-xr-x 2 root root 0 Nov 14 23:19 mm_page_free_batched drwxr-xr-x 2 root root 0 Nov 14 23:19 mm_page_pcpu_drain
Let's see what's inside the subdirectory for a specific tracepoint, in this case the one for kmalloc:
root@sugarbay:/sys/kernel/debug/tracing/events/kmem# cd kmalloc root@sugarbay:/sys/kernel/debug/tracing/events/kmem/kmalloc# ls -al drwxr-xr-x 2 root root 0 Nov 14 23:19 . drwxr-xr-x 14 root root 0 Nov 14 23:19 .. -rw-r--r-- 1 root root 0 Nov 14 23:19 enable -rw-r--r-- 1 root root 0 Nov 14 23:19 filter -r--r--r-- 1 root root 0 Nov 14 23:19 format -r--r--r-- 1 root root 0 Nov 14 23:19 id
The 'format' file for the tracepoint describes the event in memory, which is used by the various tracing tools that now make use of these tracepoint to parse the event and make sense of it, along with a 'print fmt' field that allows tools like ftrace to display the event as text. Here's what the format of the kmalloc event looks like:
root@sugarbay:/sys/kernel/debug/tracing/events/kmem/kmalloc# cat format name: kmalloc ID: 313 format: field:unsigned short common_type; offset:0; size:2; signed:0; field:unsigned char common_flags; offset:2; size:1; signed:0; field:unsigned char common_preempt_count; offset:3; size:1; signed:0; field:int common_pid; offset:4; size:4; signed:1; field:int common_padding; offset:8; size:4; signed:1; field:unsigned long call_site; offset:16; size:8; signed:0; field:const void * ptr; offset:24; size:8; signed:0; field:size_t bytes_req; offset:32; size:8; signed:0; field:size_t bytes_alloc; offset:40; size:8; signed:0; field:gfp_t gfp_flags; offset:48; size:4; signed:0; print fmt: "call_site=%lx ptr=%p bytes_req=%zu bytes_alloc=%zu gfp_flags=%s", REC->call_site, REC->ptr, REC->bytes_req, REC->bytes_alloc, (REC->gfp_flags) ? __print_flags(REC->gfp_flags, "|", {(unsigned long)(((( gfp_t)0x10u) | (( gfp_t)0x40u) | (( gfp_t)0x80u) | (( gfp_t)0x20000u) | (( gfp_t)0x02u) | (( gfp_t)0x08u)) | (( gfp_t)0x4000u) | (( gfp_t)0x10000u) | (( gfp_t)0x1000u) | (( gfp_t)0x200u) | (( gfp_t)0x400000u)), "GFP_TRANSHUGE"}, {(unsigned long)((( gfp_t)0x10u) | (( gfp_t)0x40u) | (( gfp_t)0x80u) | (( gfp_t)0x20000u) | (( gfp_t)0x02u) | (( gfp_t)0x08u)), "GFP_HIGHUSER_MOVABLE"}, {(unsigned long)((( gfp_t)0x10u) | (( gfp_t)0x40u) | (( gfp_t)0x80u) | (( gfp_t)0x20000u) | (( gfp_t)0x02u)), "GFP_HIGHUSER"}, {(unsigned long)((( gfp_t)0x10u) | (( gfp_t)0x40u) | (( gfp_t)0x80u) | (( gfp_t)0x20000u)), "GFP_USER"}, {(unsigned long)((( gfp_t)0x10u) | (( gfp_t)0x40u) | (( gfp_t)0x80u) | (( gfp_t)0x80000u)), GFP_TEMPORARY"}, {(unsigned long)((( gfp_t)0x10u) | (( gfp_t)0x40u) | (( gfp_t)0x80u)), "GFP_KERNEL"}, {(unsigned long)((( gfp_t)0x10u) | (( gfp_t)0x40u)), "GFP_NOFS"}, {(unsigned long)((( gfp_t)0x20u)), "GFP_ATOMIC"}, {(unsigned long)((( gfp_t)0x10u)), "GFP_NOIO"}, {(unsigned long)(( gfp_t)0x20u), "GFP_HIGH"}, {(unsigned long)(( gfp_t)0x10u), "GFP_WAIT"}, {(unsigned long)(( gfp_t)0x40u), "GFP_IO"}, {(unsigned long)(( gfp_t)0x100u), "GFP_COLD"}, {(unsigned long)(( gfp_t)0x200u), "GFP_NOWARN"}, {(unsigned long)(( gfp_t)0x400u), "GFP_REPEAT"}, {(unsigned long)(( gfp_t)0x800u), "GFP_NOFAIL"}, {(unsigned long)(( gfp_t)0x1000u), "GFP_NORETRY"}, {(unsigned long)(( gfp_t)0x4000u), "GFP_COMP"}, {(unsigned long)(( gfp_t)0x8000u), "GFP_ZERO"}, {(unsigned long)(( gfp_t)0x10000u), "GFP_NOMEMALLOC"}, {(unsigned long)(( gfp_t)0x20000u), "GFP_HARDWALL"}, {(unsigned long)(( gfp_t)0x40000u), "GFP_THISNODE"}, {(unsigned long)(( gfp_t)0x80000u), "GFP_RECLAIMABLE"}, {(unsigned long)(( gfp_t)0x08u), "GFP_MOVABLE"}, {(unsigned long)(( gfp_t)0), "GFP_NOTRACK"}, {(unsigned long)(( gfp_t)0x400000u), "GFP_NO_KSWAPD"}, {(unsigned long)(( gfp_t)0x800000u), "GFP_OTHER_NODE"} ) : "GFP_NOWAIT"
The 'enable' file in the tracepoint directory is what allows the user (or tools such as trace-cmd) to actually turn the tracepoint on and off. When enabled, the corresponding tracepoint will start appearing in the ftrace 'trace' file described previously. For example, this turns on the kmalloc tracepoint:
root@sugarbay:/sys/kernel/debug/tracing/events/kmem/kmalloc# echo 1 > enable
At the moment, we're not interested in the function tracer or some other tracer that might be in effect, so we first turn it off, but if we do that, we still need to turn tracing on in order to see the events in the output buffer:
root@sugarbay:/sys/kernel/debug/tracing# echo nop > current_tracer root@sugarbay:/sys/kernel/debug/tracing# echo 1 > tracing_on
Now, if we look at the the 'trace' file, we see nothing but the kmalloc events we just turned on:
root@sugarbay:/sys/kernel/debug/tracing# cat trace | less # tracer: nop # # entries-in-buffer/entries-written: 1897/1897 #P:8 # # _-----=> irqs-off # / _----=> need-resched # | / _---=> hardirq/softirq # || / _--=> preempt-depth # ||| / delay # TASK-PID CPU# |||| TIMESTAMP FUNCTION # | | | |||| | | dropbear-1465 [000] ...1 18154.620753: kmalloc: call_site=ffffffff816650d4 ptr=ffff8800729c3000 bytes_req=2048 bytes_alloc=2048 gfp_flags=GFP_KERNEL <idle>-0 [000] ..s3 18154.621640: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d555800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC <idle>-0 [000] ..s3 18154.621656: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d555800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC matchbox-termin-1361 [001] ...1 18154.755472: kmalloc: call_site=ffffffff81614050 ptr=ffff88006d5f0e00 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_KERNEL|GFP_REPEAT Xorg-1264 [002] ...1 18154.755581: kmalloc: call_site=ffffffff8141abe8 ptr=ffff8800734f4cc0 bytes_req=168 bytes_alloc=192 gfp_flags=GFP_KERNEL|GFP_NOWARN|GFP_NORETRY Xorg-1264 [002] ...1 18154.755583: kmalloc: call_site=ffffffff814192a3 ptr=ffff88001f822520 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL|GFP_ZERO Xorg-1264 [002] ...1 18154.755589: kmalloc: call_site=ffffffff81419edb ptr=ffff8800721a2f00 bytes_req=64 bytes_alloc=64 gfp_flags=GFP_KERNEL|GFP_ZERO matchbox-termin-1361 [001] ...1 18155.354594: kmalloc: call_site=ffffffff81614050 ptr=ffff88006db35400 bytes_req=576 bytes_alloc=1024 gfp_flags=GFP_KERNEL|GFP_REPEAT Xorg-1264 [002] ...1 18155.354703: kmalloc: call_site=ffffffff8141abe8 ptr=ffff8800734f4cc0 bytes_req=168 bytes_alloc=192 gfp_flags=GFP_KERNEL|GFP_NOWARN|GFP_NORETRY Xorg-1264 [002] ...1 18155.354705: kmalloc: call_site=ffffffff814192a3 ptr=ffff88001f822520 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL|GFP_ZERO Xorg-1264 [002] ...1 18155.354711: kmalloc: call_site=ffffffff81419edb ptr=ffff8800721a2f00 bytes_req=64 bytes_alloc=64 gfp_flags=GFP_KERNEL|GFP_ZERO <idle>-0 [000] ..s3 18155.673319: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d555800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC dropbear-1465 [000] ...1 18155.673525: kmalloc: call_site=ffffffff816650d4 ptr=ffff8800729c3000 bytes_req=2048 bytes_alloc=2048 gfp_flags=GFP_KERNEL <idle>-0 [000] ..s3 18155.674821: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d554800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC <idle>-0 [000] ..s3 18155.793014: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d554800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC dropbear-1465 [000] ...1 18155.793219: kmalloc: call_site=ffffffff816650d4 ptr=ffff8800729c3000 bytes_req=2048 bytes_alloc=2048 gfp_flags=GFP_KERNEL <idle>-0 [000] ..s3 18155.794147: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d555800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC <idle>-0 [000] ..s3 18155.936705: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d555800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC dropbear-1465 [000] ...1 18155.936910: kmalloc: call_site=ffffffff816650d4 ptr=ffff8800729c3000 bytes_req=2048 bytes_alloc=2048 gfp_flags=GFP_KERNEL <idle>-0 [000] ..s3 18155.937869: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d554800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC matchbox-termin-1361 [001] ...1 18155.953667: kmalloc: call_site=ffffffff81614050 ptr=ffff88006d5f2000 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_KERNEL|GFP_REPEAT Xorg-1264 [002] ...1 18155.953775: kmalloc: call_site=ffffffff8141abe8 ptr=ffff8800734f4cc0 bytes_req=168 bytes_alloc=192 gfp_flags=GFP_KERNEL|GFP_NOWARN|GFP_NORETRY Xorg-1264 [002] ...1 18155.953777: kmalloc: call_site=ffffffff814192a3 ptr=ffff88001f822520 bytes_req=24 bytes_alloc=32 gfp_flags=GFP_KERNEL|GFP_ZERO Xorg-1264 [002] ...1 18155.953783: kmalloc: call_site=ffffffff81419edb ptr=ffff8800721a2f00 bytes_req=64 bytes_alloc=64 gfp_flags=GFP_KERNEL|GFP_ZERO <idle>-0 [000] ..s3 18156.176053: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d554800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC dropbear-1465 [000] ...1 18156.176257: kmalloc: call_site=ffffffff816650d4 ptr=ffff8800729c3000 bytes_req=2048 bytes_alloc=2048 gfp_flags=GFP_KERNEL <idle>-0 [000] ..s3 18156.177717: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d555800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC <idle>-0 [000] ..s3 18156.399229: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d555800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC dropbear-1465 [000] ...1 18156.399434: kmalloc: call_site=ffffffff816650d4 ptr=ffff8800729c3000 bytes_http://rostedt.homelinux.com/kernelshark/req=2048 bytes_alloc=2048 gfp_flags=GFP_KERNEL <idle>-0 [000] ..s3 18156.400660: kmalloc: call_site=ffffffff81619b36 ptr=ffff88006d554800 bytes_req=512 bytes_alloc=512 gfp_flags=GFP_ATOMIC matchbox-termin-1361 [001] ...1 18156.552800: kmalloc: call_site=ffffffff81614050 ptr=ffff88006db34800 bytes_req=576 bytes_alloc=1024 gfp_flags=GFP_KERNEL|GFP_REPEAT
To again disable the kmalloc event, we need to send 0 to the enable file:
root@sugarbay:/sys/kernel/debug/tracing/events/kmem/kmalloc# echo 0 > enable
You can enable any number of events or complete subsystems (by using the 'enable' file in the subsystem directory) and get an arbitrarily fine-grained idea of what's going on in the system by enabling as many of the appropriate tracepoints as applicable.
A number of the tools described in this HOWTO do just that, including trace-cmd and kernelshark in the next section.
trace-cmd is essentially an extensive command-line 'wrapper' interface that hides the details of all the individual files in /sys/kernel/debug/tracing, allowing users to specify specific particular events within the /sys/kernel/debug/tracing/events/ subdirectory and to collect traces and avoid having to deal with those details directly.
As yet another layer on top of that, kernelshark provides a GUI that allows users to start and stop traces and specify sets of events using an intuitive interface, and view the output as both trace events and as a per-CPU graphical display. It directly uses 'trace-cmd' as the plumbing that accomplishes all that underneath the covers (and actually displays the trace-cmd command it uses, as we'll see).
To start a trace using kernelshark, first start kernelshark:
root@sugarbay:~# kernelshark
Then bring up the 'Capture' dialog by choosing from the kernelshark menu:
Capture | Record
That will display the following dialog, which allows you to choose one or more events (or even one or more complete subsystems) to trace:
Note that these are exactly the same sets of events described in the previous trace events subsystem section, and in fact is where trace-cmd gets them for kernelshark.
In the above screenshot, we've decided to explore the graphics subsystem a bit and so have chosen to trace all the tracepoints contained within the 'i915' and 'drm' subsystems.
After doing that, we can start and stop the trace using the 'Run' and 'Stop' button on the lower right corner of the dialog (the same button will turn into the 'Stop' button after the trace has started):
Notice that the right-hand pane shows the exact trace-cmd command-line that's used to run the trace, along with the results of the trace-cmd run.
Once the 'Stop' button is pressed, the graphical view magically fills up with a colorful per-cpu display of the trace data, along with the detailed event listing below that:
Here's another example, this time a display resulting from tracing 'all events':
The tool is pretty self-explanatory, but for more detailed information on navigating through the data, see the kernelshark website.
The documentation for ftrace can be found in the kernel Documentation directory:
Documentation/trace/ftrace.txt
The documentation for the trace event subsystem can also be found in the kernel Documentation directory:
Documentation/trace/events.txt
There is a nice series of articles on using ftrace and trace-cmd at LWN:
There's more detailed documentation kernelshark usage here: KernelShark
An amusing yet useful README (a tracing mini-HOWTO) can be found in /sys/kernel/debug/tracing/README.
SystemTap is a system-wide script-based tracing and profiling tool.
SystemTap scripts are C-like programs that are executed in the kernel to gather/print/aggregate data extracted from the context they end up being invoked under.
For example, this probe from the SystemTap tutorial simply prints a line every time any process on the system open()s a file. For each line, it prints the executable name of the program that opened the file, along with its PID, and the name of the file it opened (or tried to open), which it extracts from the open syscall's argstr.
probe syscall.open { printf ("%s(%d) open (%s)\n", execname(), pid(), argstr) } probe timer.ms(4000) # after 4 seconds { exit () }
Normally, to execute this probe, you'd simply install systemtap on the system you want to probe, and directly run the probe on that system e.g. assuming the name of the file containing the above text is trace_open.stp:
# stap trace_open.stp
What systemtap does under the covers to run this probe is 1) parse and convert the probe to an equivalent 'C' form, 2) compile the 'C' form into a kernel module, 3) insert the module into the kernel, which arms it, and 4) collect the data generated by the probe and display it to the user.
In order to accomplish steps 1 and 2, the 'stap' program needs access to the kernel build system that produced the kernel that the probed system is running. In the case of a typical embedded system (the 'target'), the kernel build system unfortunately isn't typically part of the image running on the target. It is normally available on the 'host' system that produced the target image however; in such cases, steps 1 and 2 are executed on the host system, and steps 3 and 4 are executed on the target system, using only the systemtap 'runtime'.
The systemtap support in Yocto assumes that only steps 3 and 4 are run on the target; it is possible to do everything on the target, but this section assumes only the typical embedded use-case.
So basically what you need to do in order to run a systemtap script on the target is to 1) on the host system, compile the probe into a kernel module that makes sense to the target, 2) copy the module onto the target system and 3) insert the module into the target kernel, which arms it, and 4) collect the data generated by the probe and display it to the user.
Those are a lot of steps and a lot of details, but fortunately Yocto includes a script called 'crosstap' that will take care of those details, allowing you to simply execute a systemtap script on the remote target, with arguments if necessary.
In order to do this from a remote host, however, you need to have access to the build for the image you booted. The 'crosstap' script provides details on how to do this if you run the script on the host without having done a build:
$ crosstap root@192.168.1.88 trace_open.stp Error: No target kernel build found. Did you forget to create a local build of your image? 'crosstap' requires a local sdk build of the target system (or a build that includes 'tools-profile') in order to build kernel modules that can probe the target system. Practically speaking, that means you need to do the following: - If you're running a pre-built image, download the release and/or BSP tarballs used to build the image. - If you're working from git sources, just clone the metadata and BSP layers needed to build the image you'll be booting. - Make sure you're properly set up to build a new image (see the BSP README and/or the widely available basic documentation that discusses how to build images). - Build an -sdk version of the image e.g.: $ bitbake core-image-sato-sdk OR - Build a non-sdk image but include the profiling tools: [ edit local.conf and add 'tools-profile' to the end of the EXTRA_IMAGE_FEATURES variable ] $ bitbake core-image-sato Once you've build the image on the host system, you're ready to boot it (or the equivalent pre-built image) and use 'crosstap' to probe it (you need to source the environment as usual first): $ source oe-init-build-env $ cd ~/my/systemtap/scripts $ crosstap root@192.168.1.xxx myscript.stp
So essentially what you need to do is build an SDK image or image with 'tools-profile' as detailed in the "General Setup" section of this manual, and boot the resulting target image.
Once you've done that, you should be able to run a systemtap script on the target:
$ cd /path/to/yocto $ source oe-init-build-env ### Shell environment set up for builds. ### You can now run 'bitbake <target>' Common targets are: core-image-minimal core-image-sato meta-toolchain meta-ide-support You can also run generated qemu images with a command like 'runqemu qemux86-64'
Once you've done that, you can cd to whatever directory contains your scripts and use 'crosstap' to run the script:
$ cd /path/to/my/systemap/script $ crosstap root@192.168.7.2 trace_open.stp
If you get an error connecting to the target e.g.:
$ crosstap root@192.168.7.2 trace_open.stp error establishing ssh connection on remote 'root@192.168.7.2'
Try ssh'ing to the target and see what happens:
$ ssh root@192.168.7.2
A lot of the time, connection problems are due specifying a wrong IP address or having a 'host key verification error'.
If everything worked as planned, you should see something like this (enter the password when prompted, or press enter if it's set up to use no password):
$ crosstap root@192.168.7.2 trace_open.stp root@192.168.7.2's password: matchbox-termin(1036) open ("/tmp/vte3FS2LW", O_RDWR|O_CREAT|O_EXCL|O_LARGEFILE, 0600) matchbox-termin(1036) open ("/tmp/vteJMC7LW", O_RDWR|O_CREAT|O_EXCL|O_LARGEFILE, 0600)
The SystemTap language reference can be found here: SystemTap Language Reference
Links to other SystemTap documents, tutorials, and examples can be found here: SystemTap documentation page
Sysprof is a very easy to use system-wide profiler that consists of a single window with three panes and a few buttons which allow you to start, stop, and view the profile from one place.
For this section, we'll assume you've already performed the basic setup outlined in the General Setup section.
Sysprof is a GUI-based application that runs on the target system. For the rest of this document we assume you've ssh'ed to the host and will be running Sysprof on the target (you can use the '-X' option to ssh and have the Sysprof GUI run on the target but display remotely on the host if you want).
To start profiling the system, you simply press the 'Start' button. To stop profiling and to start viewing the profile data in one easy step, press the 'Profile' button.
Once you've pressed the profile button, the three panes will fill up with profiling data:
The left pane shows a list of functions and processes. Selecting one of those expands that function in the right pane, showing all its callees. Note that this caller-oriented display is essentially the inverse of perf's default callee-oriented callchain display.
In the screenshot above, we're focusing on __copy_to_user_ll() and looking up the callchain we can see that one of the callers of __copy_to_user_ll is sys_read() and the complete callpath between them. Notice that this is essentially a portion of the same information we saw in the perf display shown in the perf section of this page.
Similarly, the above is a snapshot of the Sysprof display of a copy-from-user callchain.
Finally, looking at the third Sysprof pane in the lower left, we can see a list of all the callers of a particular function selected in the top left pane. In this case, the lower pane is showing all the callers of __mark_inode_dirty:
Double-clicking on one of those functions will in turn change the focus to the selected function, and so on.
There doesn't seem to be any documentation for Sysprof, but maybe that's because it's pretty self-explanatory. The Sysprof website, however, is here: Sysprof, System-wide Performance Profiler for Linux
For this section, we'll assume you've already performed the basic setup outlined in the General Setup section. LTTng is run on the target system by ssh'ing to it.
Once you've applied the above commits and built and booted your image (you need to build the core-image-sato-sdk image or use one of the other methods described in the General Setup section), you're ready to start tracing.
First, from the host, ssh to the target:
$ ssh -l root 192.168.1.47 The authenticity of host '192.168.1.47 (192.168.1.47)' can't be established. RSA key fingerprint is 23:bd:c8:b1:a8:71:52:00:ee:00:4f:64:9e:10:b9:7e. Are you sure you want to continue connecting (yes/no)? yes Warning: Permanently added '192.168.1.47' (RSA) to the list of known hosts. root@192.168.1.47's password:
Once on the target, use these steps to create a trace:
root@crownbay:~# lttng create Spawning a session daemon Session auto-20121015-232120 created. Traces will be written in /home/root/lttng-traces/auto-20121015-232120
Enable the events you want to trace (in this case all kernel events):
root@crownbay:~# lttng enable-event --kernel --all All kernel events are enabled in channel channel0
Start the trace:
root@crownbay:~# lttng start Tracing started for session auto-20121015-232120
And then stop the trace after awhile or after running a particular workload that you want to trace:
root@crownbay:~# lttng stop Tracing stopped for session auto-20121015-232120
You can now view the trace in text form on the target:
root@crownbay:~# lttng view [23:21:56.989270399] (+?.?????????) sys_geteuid: { 1 }, { } [23:21:56.989278081] (+0.000007682) exit_syscall: { 1 }, { ret = 0 } [23:21:56.989286043] (+0.000007962) sys_pipe: { 1 }, { fildes = 0xB77B9E8C } [23:21:56.989321802] (+0.000035759) exit_syscall: { 1 }, { ret = 0 } [23:21:56.989329345] (+0.000007543) sys_mmap_pgoff: { 1 }, { addr = 0x0, len = 10485760, prot = 3, flags = 131362, fd = 4294967295, pgoff = 0 } [23:21:56.989351694] (+0.000022349) exit_syscall: { 1 }, { ret = -1247805440 } [23:21:56.989432989] (+0.000081295) sys_clone: { 1 }, { clone_flags = 0x411, newsp = 0xB5EFFFE4, parent_tid = 0xFFFFFFFF, child_tid = 0x0 } [23:21:56.989477129] (+0.000044140) sched_stat_runtime: { 1 }, { comm = "lttng-consumerd", tid = 1193, runtime = 681660, vruntime = 43367983388 } [23:21:56.989486697] (+0.000009568) sched_migrate_task: { 1 }, { comm = "lttng-consumerd", tid = 1193, prio = 20, orig_cpu = 1, dest_cpu = 1 } [23:21:56.989508418] (+0.000021721) hrtimer_init: { 1 }, { hrtimer = 3970832076, clockid = 1, mode = 1 } [23:21:56.989770462] (+0.000262044) hrtimer_cancel: { 1 }, { hrtimer = 3993865440 } [23:21:56.989771580] (+0.000001118) hrtimer_cancel: { 0 }, { hrtimer = 3993812192 } [23:21:56.989776957] (+0.000005377) hrtimer_expire_entry: { 1 }, { hrtimer = 3993865440, now = 79815980007057, function = 3238465232 } [23:21:56.989778145] (+0.000001188) hrtimer_expire_entry: { 0 }, { hrtimer = 3993812192, now = 79815980008174, function = 3238465232 } [23:21:56.989791695] (+0.000013550) softirq_raise: { 1 }, { vec = 1 } [23:21:56.989795396] (+0.000003701) softirq_raise: { 0 }, { vec = 1 } [23:21:56.989800635] (+0.000005239) softirq_raise: { 0 }, { vec = 9 } [23:21:56.989807130] (+0.000006495) sched_stat_runtime: { 1 }, { comm = "lttng-consumerd", tid = 1193, runtime = 330710, vruntime = 43368314098 } [23:21:56.989809993] (+0.000002863) sched_stat_runtime: { 0 }, { comm = "lttng-sessiond", tid = 1181, runtime = 1015313, vruntime = 36976733240 } [23:21:56.989818514] (+0.000008521) hrtimer_expire_exit: { 0 }, { hrtimer = 3993812192 } [23:21:56.989819631] (+0.000001117) hrtimer_expire_exit: { 1 }, { hrtimer = 3993865440 } [23:21:56.989821866] (+0.000002235) hrtimer_start: { 0 }, { hrtimer = 3993812192, function = 3238465232, expires = 79815981000000, softexpires = 79815981000000 } [23:21:56.989822984] (+0.000001118) hrtimer_start: { 1 }, { hrtimer = 3993865440, function = 3238465232, expires = 79815981000000, softexpires = 79815981000000 } [23:21:56.989832762] (+0.000009778) softirq_entry: { 1 }, { vec = 1 } [23:21:56.989833879] (+0.000001117) softirq_entry: { 0 }, { vec = 1 } [23:21:56.989838069] (+0.000004190) timer_cancel: { 1 }, { timer = 3993871956 } [23:21:56.989839187] (+0.000001118) timer_cancel: { 0 }, { timer = 3993818708 } [23:21:56.989841492] (+0.000002305) timer_expire_entry: { 1 }, { timer = 3993871956, now = 79515980, function = 3238277552 } [23:21:56.989842819] (+0.000001327) timer_expire_entry: { 0 }, { timer = 3993818708, now = 79515980, function = 3238277552 } [23:21:56.989854831] (+0.000012012) sched_stat_runtime: { 1 }, { comm = "lttng-consumerd", tid = 1193, runtime = 49237, vruntime = 43368363335 } [23:21:56.989855949] (+0.000001118) sched_stat_runtime: { 0 }, { comm = "lttng-sessiond", tid = 1181, runtime = 45121, vruntime = 36976778361 } [23:21:56.989861257] (+0.000005308) sched_stat_sleep: { 1 }, { comm = "kworker/1:1", tid = 21, delay = 9451318 } [23:21:56.989862374] (+0.000001117) sched_stat_sleep: { 0 }, { comm = "kworker/0:0", tid = 4, delay = 9958820 } [23:21:56.989868241] (+0.000005867) sched_wakeup: { 0 }, { comm = "kworker/0:0", tid = 4, prio = 120, success = 1, target_cpu = 0 } [23:21:56.989869358] (+0.000001117) sched_wakeup: { 1 }, { comm = "kworker/1:1", tid = 21, prio = 120, success = 1, target_cpu = 1 } [23:21:56.989877460] (+0.000008102) timer_expire_exit: { 1 }, { timer = 3993871956 } [23:21:56.989878577] (+0.000001117) timer_expire_exit: { 0 }, { timer = 3993818708 } . . .
You can now safely destroy the trace session (note that this doesn't delete the trace - it's still there in ~/lttng-traces):
root@crownbay:~# lttng destroy Session auto-20121015-232120 destroyed at /home/root
Note that the trace is saved in a directory of the same name as returned by 'lttng create', under the ~/lttng-traces directory (note that you can change this by supplying your own name to 'lttng create'):
root@crownbay:~# ls -al ~/lttng-traces drwxrwx--- 3 root root 1024 Oct 15 23:21 . drwxr-xr-x 5 root root 1024 Oct 15 23:57 .. drwxrwx--- 3 root root 1024 Oct 15 23:21 auto-20121015-232120
For LTTng userspace tracing, you need to have a properly instrumented userspace program. For this example, we'll use the 'hello' test program generated by the lttng-ust build.
The 'hello' test program isn't installed on the rootfs by the lttng-ust build, so we need to copy it over manually. First cd into the build directory that contains the hello executable:
$ cd build/tmp/work/core2_32-poky-linux/lttng-ust/2.0.5-r0/git/tests/hello/.libs
Copy that over to the target machine:
$ scp hello root@192.168.1.20:
You now have the instrumented lttng 'hello world' test program on the target, ready to test.
First, from the host, ssh to the target:
$ ssh -l root 192.168.1.47 The authenticity of host '192.168.1.47 (192.168.1.47)' can't be established. RSA key fingerprint is 23:bd:c8:b1:a8:71:52:00:ee:00:4f:64:9e:10:b9:7e. Are you sure you want to continue connecting (yes/no)? yes Warning: Permanently added '192.168.1.47' (RSA) to the list of known hosts. root@192.168.1.47's password:
Once on the target, use these steps to create a trace:
root@crownbay:~# lttng create Session auto-20190303-021943 created. Traces will be written in /home/root/lttng-traces/auto-20190303-021943
Enable the events you want to trace (in this case all userspace events):
root@crownbay:~# lttng enable-event --userspace --all All UST events are enabled in channel channel0
Start the trace:
root@crownbay:~# lttng start Tracing started for session auto-20190303-021943
Run the instrumented hello world program:
root@crownbay:~# ./hello Hello, World! Tracing... done.
And then stop the trace after awhile or after running a particular workload that you want to trace:
root@crownbay:~# lttng stop Tracing stopped for session auto-20190303-021943
You can now view the trace in text form on the target:
root@crownbay:~# lttng view [02:31:14.906146544] (+?.?????????) hello:1424 ust_tests_hello:tptest: { cpu_id = 1 }, { intfield = 0, intfield2 = 0x0, longfield = 0, netintfield = 0, netintfieldhex = 0x0, arrfield1 = [ [0] = 1, [1] = 2, [2] = 3 ], arrfield2 = "test", _seqfield1_length = 4, seqfield1 = [ [0] = 116, [1] = 101, [2] = 115, [3] = 116 ], _seqfield2_length = 4, seqfield2 = "test", stringfield = "test", floatfield = 2222, doublefield = 2, boolfield = 1 } [02:31:14.906170360] (+0.000023816) hello:1424 ust_tests_hello:tptest: { cpu_id = 1 }, { intfield = 1, intfield2 = 0x1, longfield = 1, netintfield = 1, netintfieldhex = 0x1, arrfield1 = [ [0] = 1, [1] = 2, [2] = 3 ], arrfield2 = "test", _seqfield1_length = 4, seqfield1 = [ [0] = 116, [1] = 101, [2] = 115, [3] = 116 ], _seqfield2_length = 4, seqfield2 = "test", stringfield = "test", floatfield = 2222, doublefield = 2, boolfield = 1 } [02:31:14.906183140] (+0.000012780) hello:1424 ust_tests_hello:tptest: { cpu_id = 1 }, { intfield = 2, intfield2 = 0x2, longfield = 2, netintfield = 2, netintfieldhex = 0x2, arrfield1 = [ [0] = 1, [1] = 2, [2] = 3 ], arrfield2 = "test", _seqfield1_length = 4, seqfield1 = [ [0] = 116, [1] = 101, [2] = 115, [3] = 116 ], _seqfield2_length = 4, seqfield2 = "test", stringfield = "test", floatfield = 2222, doublefield = 2, boolfield = 1 } [02:31:14.906194385] (+0.000011245) hello:1424 ust_tests_hello:tptest: { cpu_id = 1 }, { intfield = 3, intfield2 = 0x3, longfield = 3, netintfield = 3, netintfieldhex = 0x3, arrfield1 = [ [0] = 1, [1] = 2, [2] = 3 ], arrfield2 = "test", _seqfield1_length = 4, seqfield1 = [ [0] = 116, [1] = 101, [2] = 115, [3] = 116 ], _seqfield2_length = 4, seqfield2 = "test", stringfield = "test", floatfield = 2222, doublefield = 2, boolfield = 1 } . . .
You can now safely destroy the trace session (note that this doesn't delete the trace - it's still there in ~/lttng-traces):
root@crownbay:~# lttng destroy Session auto-20190303-021943 destroyed at /home/root
You can find the primary LTTng Documentation on the LTTng Documentation site. The documentation on this site is appropriate for intermediate to advanced software developers who are working in a Linux environment and are interested in efficient software tracing.
For information on LTTng in general, visit the LTTng Project site. You can find a "Getting Started" link on this site that takes you to an LTTng Quick Start.
blktrace is a tool for tracing and reporting low-level disk I/O. blktrace provides the tracing half of the equation; its output can be piped into the blkparse program, which renders the data in a human-readable form and does some basic analysis:
For this section, we'll assume you've already performed the basic setup outlined in the "General Setup" section.
blktrace is an application that runs on the target system. You can run the entire blktrace and blkparse pipeline on the target, or you can run blktrace in 'listen' mode on the target and have blktrace and blkparse collect and analyze the data on the host (see the "Using blktrace Remotely" section below). For the rest of this section we assume you've ssh'ed to the host and will be running blkrace on the target.
To record a trace, simply run the 'blktrace' command, giving it the name of the block device you want to trace activity on:
root@crownbay:~# blktrace /dev/sdc
In another shell, execute a workload you want to trace.
root@crownbay:/media/sdc# rm linux-2.6.19.2.tar.bz2; wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2; sync Connecting to downloads.yoctoproject.org (140.211.169.59:80) linux-2.6.19.2.tar.b 100% |*******************************| 41727k 0:00:00 ETA
Press Ctrl-C in the blktrace shell to stop the trace. It will display how many events were logged, along with the per-cpu file sizes (blktrace records traces in per-cpu kernel buffers and simply dumps them to userspace for blkparse to merge and sort later).
^C=== sdc === CPU 0: 7082 events, 332 KiB data CPU 1: 1578 events, 74 KiB data Total: 8660 events (dropped 0), 406 KiB data
If you examine the files saved to disk, you see multiple files, one per CPU and with the device name as the first part of the filename:
root@crownbay:~# ls -al drwxr-xr-x 6 root root 1024 Oct 27 22:39 . drwxr-sr-x 4 root root 1024 Oct 26 18:24 .. -rw-r--r-- 1 root root 339938 Oct 27 22:40 sdc.blktrace.0 -rw-r--r-- 1 root root 75753 Oct 27 22:40 sdc.blktrace.1
To view the trace events, simply invoke 'blkparse' in the directory containing the trace files, giving it the device name that forms the first part of the filenames:
root@crownbay:~# blkparse sdc 8,32 1 1 0.000000000 1225 Q WS 3417048 + 8 [jbd2/sdc-8] 8,32 1 2 0.000025213 1225 G WS 3417048 + 8 [jbd2/sdc-8] 8,32 1 3 0.000033384 1225 P N [jbd2/sdc-8] 8,32 1 4 0.000043301 1225 I WS 3417048 + 8 [jbd2/sdc-8] 8,32 1 0 0.000057270 0 m N cfq1225 insert_request 8,32 1 0 0.000064813 0 m N cfq1225 add_to_rr 8,32 1 5 0.000076336 1225 U N [jbd2/sdc-8] 1 8,32 1 0 0.000088559 0 m N cfq workload slice:150 8,32 1 0 0.000097359 0 m N cfq1225 set_active wl_prio:0 wl_type:1 8,32 1 0 0.000104063 0 m N cfq1225 Not idling. st->count:1 8,32 1 0 0.000112584 0 m N cfq1225 fifo= (null) 8,32 1 0 0.000118730 0 m N cfq1225 dispatch_insert 8,32 1 0 0.000127390 0 m N cfq1225 dispatched a request 8,32 1 0 0.000133536 0 m N cfq1225 activate rq, drv=1 8,32 1 6 0.000136889 1225 D WS 3417048 + 8 [jbd2/sdc-8] 8,32 1 7 0.000360381 1225 Q WS 3417056 + 8 [jbd2/sdc-8] 8,32 1 8 0.000377422 1225 G WS 3417056 + 8 [jbd2/sdc-8] 8,32 1 9 0.000388876 1225 P N [jbd2/sdc-8] 8,32 1 10 0.000397886 1225 Q WS 3417064 + 8 [jbd2/sdc-8] 8,32 1 11 0.000404800 1225 M WS 3417064 + 8 [jbd2/sdc-8] 8,32 1 12 0.000412343 1225 Q WS 3417072 + 8 [jbd2/sdc-8] 8,32 1 13 0.000416533 1225 M WS 3417072 + 8 [jbd2/sdc-8] 8,32 1 14 0.000422121 1225 Q WS 3417080 + 8 [jbd2/sdc-8] 8,32 1 15 0.000425194 1225 M WS 3417080 + 8 [jbd2/sdc-8] 8,32 1 16 0.000431968 1225 Q WS 3417088 + 8 [jbd2/sdc-8] 8,32 1 17 0.000435251 1225 M WS 3417088 + 8 [jbd2/sdc-8] 8,32 1 18 0.000440279 1225 Q WS 3417096 + 8 [jbd2/sdc-8] 8,32 1 19 0.000443911 1225 M WS 3417096 + 8 [jbd2/sdc-8] 8,32 1 20 0.000450336 1225 Q WS 3417104 + 8 [jbd2/sdc-8] 8,32 1 21 0.000454038 1225 M WS 3417104 + 8 [jbd2/sdc-8] 8,32 1 22 0.000462070 1225 Q WS 3417112 + 8 [jbd2/sdc-8] 8,32 1 23 0.000465422 1225 M WS 3417112 + 8 [jbd2/sdc-8] 8,32 1 24 0.000474222 1225 I WS 3417056 + 64 [jbd2/sdc-8] 8,32 1 0 0.000483022 0 m N cfq1225 insert_request 8,32 1 25 0.000489727 1225 U N [jbd2/sdc-8] 1 8,32 1 0 0.000498457 0 m N cfq1225 Not idling. st->count:1 8,32 1 0 0.000503765 0 m N cfq1225 dispatch_insert 8,32 1 0 0.000512914 0 m N cfq1225 dispatched a request 8,32 1 0 0.000518851 0 m N cfq1225 activate rq, drv=2 . . . 8,32 0 0 58.515006138 0 m N cfq3551 complete rqnoidle 1 8,32 0 2024 58.516603269 3 C WS 3156992 + 16 [0] 8,32 0 0 58.516626736 0 m N cfq3551 complete rqnoidle 1 8,32 0 0 58.516634558 0 m N cfq3551 arm_idle: 8 group_idle: 0 8,32 0 0 58.516636933 0 m N cfq schedule dispatch 8,32 1 0 58.516971613 0 m N cfq3551 slice expired t=0 8,32 1 0 58.516982089 0 m N cfq3551 sl_used=13 disp=6 charge=13 iops=0 sect=80 8,32 1 0 58.516985511 0 m N cfq3551 del_from_rr 8,32 1 0 58.516990819 0 m N cfq3551 put_queue CPU0 (sdc): Reads Queued: 0, 0KiB Writes Queued: 331, 26,284KiB Read Dispatches: 0, 0KiB Write Dispatches: 485, 40,484KiB Reads Requeued: 0 Writes Requeued: 0 Reads Completed: 0, 0KiB Writes Completed: 511, 41,000KiB Read Merges: 0, 0KiB Write Merges: 13, 160KiB Read depth: 0 Write depth: 2 IO unplugs: 23 Timer unplugs: 0 CPU1 (sdc): Reads Queued: 0, 0KiB Writes Queued: 249, 15,800KiB Read Dispatches: 0, 0KiB Write Dispatches: 42, 1,600KiB Reads Requeued: 0 Writes Requeued: 0 Reads Completed: 0, 0KiB Writes Completed: 16, 1,084KiB Read Merges: 0, 0KiB Write Merges: 40, 276KiB Read depth: 0 Write depth: 2 IO unplugs: 30 Timer unplugs: 1 Total (sdc): Reads Queued: 0, 0KiB Writes Queued: 580, 42,084KiB Read Dispatches: 0, 0KiB Write Dispatches: 527, 42,084KiB Reads Requeued: 0 Writes Requeued: 0 Reads Completed: 0, 0KiB Writes Completed: 527, 42,084KiB Read Merges: 0, 0KiB Write Merges: 53, 436KiB IO unplugs: 53 Timer unplugs: 1 Throughput (R/W): 0KiB/s / 719KiB/s Events (sdc): 6,592 entries Skips: 0 forward (0 - 0.0%) Input file sdc.blktrace.0 added Input file sdc.blktrace.1 added
The report shows each event that was found in the blktrace data, along with a summary of the overall block I/O traffic during the run. You can look at the blkparse manpage to learn the meaning of each field displayed in the trace listing.
blktrace and blkparse are designed from the ground up to be able to operate together in a 'pipe mode' where the stdout of blktrace can be fed directly into the stdin of blkparse:
root@crownbay:~# blktrace /dev/sdc -o - | blkparse -i -
This enables long-lived tracing sessions to run without writing anything to disk, and allows the user to look for certain conditions in the trace data in 'real-time' by viewing the trace output as it scrolls by on the screen or by passing it along to yet another program in the pipeline such as grep which can be used to identify and capture conditions of interest.
There's actually another blktrace command that implements the above pipeline as a single command, so the user doesn't have to bother typing in the above command sequence:
root@crownbay:~# btrace /dev/sdc
Because blktrace traces block I/O and at the same time normally writes its trace data to a block device, and in general because it's not really a great idea to make the device being traced the same as the device the tracer writes to, blktrace provides a way to trace without perturbing the traced device at all by providing native support for sending all trace data over the network.
To have blktrace operate in this mode, start blktrace on the target system being traced with the -l option, along with the device to trace:
root@crownbay:~# blktrace -l /dev/sdc server: waiting for connections...
On the host system, use the -h option to connect to the target system, also passing it the device to trace:
$ blktrace -d /dev/sdc -h 192.168.1.43 blktrace: connecting to 192.168.1.43 blktrace: connected!
On the target system, you should see this:
server: connection from 192.168.1.43
In another shell, execute a workload you want to trace.
root@crownbay:/media/sdc# rm linux-2.6.19.2.tar.bz2; wget http://downloads.yoctoproject.org/mirror/sources/linux-2.6.19.2.tar.bz2; sync Connecting to downloads.yoctoproject.org (140.211.169.59:80) linux-2.6.19.2.tar.b 100% |*******************************| 41727k 0:00:00 ETA
When it's done, do a Ctrl-C on the host system to stop the trace:
^C=== sdc === CPU 0: 7691 events, 361 KiB data CPU 1: 4109 events, 193 KiB data Total: 11800 events (dropped 0), 554 KiB data
On the target system, you should also see a trace summary for the trace just ended:
server: end of run for 192.168.1.43:sdc === sdc === CPU 0: 7691 events, 361 KiB data CPU 1: 4109 events, 193 KiB data Total: 11800 events (dropped 0), 554 KiB data
The blktrace instance on the host will save the target output inside a hostname-timestamp directory:
$ ls -al drwxr-xr-x 10 root root 1024 Oct 28 02:40 . drwxr-sr-x 4 root root 1024 Oct 26 18:24 .. drwxr-xr-x 2 root root 1024 Oct 28 02:40 192.168.1.43-2012-10-28-02:40:56
cd into that directory to see the output files:
$ ls -l -rw-r--r-- 1 root root 369193 Oct 28 02:44 sdc.blktrace.0 -rw-r--r-- 1 root root 197278 Oct 28 02:44 sdc.blktrace.1
And run blkparse on the host system using the device name:
$ blkparse sdc 8,32 1 1 0.000000000 1263 Q RM 6016 + 8 [ls] 8,32 1 0 0.000036038 0 m N cfq1263 alloced 8,32 1 2 0.000039390 1263 G RM 6016 + 8 [ls] 8,32 1 3 0.000049168 1263 I RM 6016 + 8 [ls] 8,32 1 0 0.000056152 0 m N cfq1263 insert_request 8,32 1 0 0.000061600 0 m N cfq1263 add_to_rr 8,32 1 0 0.000075498 0 m N cfq workload slice:300 . . . 8,32 0 0 177.266385696 0 m N cfq1267 arm_idle: 8 group_idle: 0 8,32 0 0 177.266388140 0 m N cfq schedule dispatch 8,32 1 0 177.266679239 0 m N cfq1267 slice expired t=0 8,32 1 0 177.266689297 0 m N cfq1267 sl_used=9 disp=6 charge=9 iops=0 sect=56 8,32 1 0 177.266692649 0 m N cfq1267 del_from_rr 8,32 1 0 177.266696560 0 m N cfq1267 put_queue CPU0 (sdc): Reads Queued: 0, 0KiB Writes Queued: 270, 21,708KiB Read Dispatches: 59, 2,628KiB Write Dispatches: 495, 39,964KiB Reads Requeued: 0 Writes Requeued: 0 Reads Completed: 90, 2,752KiB Writes Completed: 543, 41,596KiB Read Merges: 0, 0KiB Write Merges: 9, 344KiB Read depth: 2 Write depth: 2 IO unplugs: 20 Timer unplugs: 1 CPU1 (sdc): Reads Queued: 688, 2,752KiB Writes Queued: 381, 20,652KiB Read Dispatches: 31, 124KiB Write Dispatches: 59, 2,396KiB Reads Requeued: 0 Writes Requeued: 0 Reads Completed: 0, 0KiB Writes Completed: 11, 764KiB Read Merges: 598, 2,392KiB Write Merges: 88, 448KiB Read depth: 2 Write depth: 2 IO unplugs: 52 Timer unplugs: 0 Total (sdc): Reads Queued: 688, 2,752KiB Writes Queued: 651, 42,360KiB Read Dispatches: 90, 2,752KiB Write Dispatches: 554, 42,360KiB Reads Requeued: 0 Writes Requeued: 0 Reads Completed: 90, 2,752KiB Writes Completed: 554, 42,360KiB Read Merges: 598, 2,392KiB Write Merges: 97, 792KiB IO unplugs: 72 Timer unplugs: 1 Throughput (R/W): 15KiB/s / 238KiB/s Events (sdc): 9,301 entries Skips: 0 forward (0 - 0.0%)
You should see the trace events and summary just as you would have if you'd run the same command on the target.
It's also possible to trace block I/O using only trace events subsystem, which can be useful for casual tracing if you don't want to bother dealing with the userspace tools.
To enable tracing for a given device, use /sys/block/xxx/trace/enable, where xxx is the device name. This for example enables tracing for /dev/sdc:
root@crownbay:/sys/kernel/debug/tracing# echo 1 > /sys/block/sdc/trace/enable
Once you've selected the device(s) you want to trace, selecting the 'blk' tracer will turn the blk tracer on:
root@crownbay:/sys/kernel/debug/tracing# cat available_tracers blk function_graph function nop root@crownbay:/sys/kernel/debug/tracing# echo blk > current_tracer
Execute the workload you're interested in:
root@crownbay:/sys/kernel/debug/tracing# cat /media/sdc/testfile.txt
And look at the output (note here that we're using 'trace_pipe' instead of trace to capture this trace - this allows us to wait around on the pipe for data to appear):
root@crownbay:/sys/kernel/debug/tracing# cat trace_pipe cat-3587 [001] d..1 3023.276361: 8,32 Q R 1699848 + 8 [cat] cat-3587 [001] d..1 3023.276410: 8,32 m N cfq3587 alloced cat-3587 [001] d..1 3023.276415: 8,32 G R 1699848 + 8 [cat] cat-3587 [001] d..1 3023.276424: 8,32 P N [cat] cat-3587 [001] d..2 3023.276432: 8,32 I R 1699848 + 8 [cat] cat-3587 [001] d..1 3023.276439: 8,32 m N cfq3587 insert_request cat-3587 [001] d..1 3023.276445: 8,32 m N cfq3587 add_to_rr cat-3587 [001] d..2 3023.276454: 8,32 U N [cat] 1 cat-3587 [001] d..1 3023.276464: 8,32 m N cfq workload slice:150 cat-3587 [001] d..1 3023.276471: 8,32 m N cfq3587 set_active wl_prio:0 wl_type:2 cat-3587 [001] d..1 3023.276478: 8,32 m N cfq3587 fifo= (null) cat-3587 [001] d..1 3023.276483: 8,32 m N cfq3587 dispatch_insert cat-3587 [001] d..1 3023.276490: 8,32 m N cfq3587 dispatched a request cat-3587 [001] d..1 3023.276497: 8,32 m N cfq3587 activate rq, drv=1 cat-3587 [001] d..2 3023.276500: 8,32 D R 1699848 + 8 [cat]
And this turns off tracing for the specified device:
root@crownbay:/sys/kernel/debug/tracing# echo 0 > /sys/block/sdc/trace/enable
Online versions of the man pages for the commands discussed in this section can be found here:
The above manpages, along with manpages for the other blktrace utilities (btt, blkiomon, etc) can be found in the /doc directory of the blktrace tools git repo:
$ git clone git://git.kernel.dk/blktrace.git
This chapter contains real-world examples.
In one of our previous releases (denzil), users noticed that booting off of a live image and writing to disk was noticeably slower. This included the boot itself, especially the first one, since first boots tend to do a significant amount of writing due to certain post-install scripts.
The problem (and solution) was discovered by using the Yocto tracing tools, in this case 'perf stat', 'perf script', 'perf record' and 'perf report'.
See all the unvarnished details of how this bug was diagnosed and solved here: Yocto Bug #3049
Welcome to the Yocto Project Reference Manual! This manual provides reference information for the current release of the Yocto Project, and is most effectively used after you have an understanding of the basics of the Yocto Project. The manual is neither meant to be read as a starting point to the Yocto Project, nor read from start to finish. Rather, use this manual to find variable definitions, class descriptions, and so forth as needed during the course of using the Yocto Project.
For introductory information on the Yocto Project, see the Yocto Project Website and the "Yocto Project Development Environment" chapter in the Yocto Project Overview and Concepts Manual.
If you want to use the Yocto Project to quickly build an image without having to understand concepts, work through the Yocto Project Quick Build document. You can find "how-to" information in the Yocto Project Development Tasks Manual. You can find Yocto Project overview and conceptual information in the Yocto project Overview and Concepts Manual.
Currently, the Yocto Project is supported on the following distributions:
Yocto Project releases are tested against the stable Linux distributions in the following list. The Yocto Project should work on other distributions but validation is not performed against them.
In particular, the Yocto Project does not support and currently has no plans to support rolling-releases or development distributions due to their constantly changing nature. We welcome patches and bug reports, but keep in mind that our priority is on the supported platforms listed below.
You may use Windows Subsystem For Linux v2 to set up a build host using Windows 10, but validation is not performed against build hosts using WSLv2.
If you encounter problems, please go to Yocto Project Bugzilla and submit a bug. We are interested in hearing about your experience. For information on how to submit a bug, see the Yocto Project Bugzilla wiki page and the "Submitting a Defect Against the Yocto Project" section in the Yocto Project Development Tasks Manual.
Ubuntu 16.04 (LTS)
Ubuntu 18.04 (LTS)
Ubuntu 19.04
Ubuntu 20.04
Fedora 28
Fedora 29
Fedora 30
Fedora 31
Fedora 32
CentOS 7.x
Debian GNU/Linux 8.x (Jessie)
Debian GNU/Linux 9.x (Stretch)
Debian GNU/Linux 10.x (Buster)
OpenSUSE Leap 15.1
The list of packages you need on the host development system can be large when covering all build scenarios using the Yocto Project. This section describes required packages according to Linux distribution and function.
The following list shows the required packages by function given a supported Ubuntu or Debian Linux distribution:
If your build system has the
oss4-dev
package installed, you
might experience QEMU build failures due to the package
installing its own custom
/usr/include/linux/soundcard.h
on
the Debian system.
If you run into this situation, either of the following
solutions exist:
$ sudo apt-get build-dep qemu $ sudo apt-get remove oss4-dev
For Debian-8, python3-git
and pylint3
are no longer available via apt-get
.
$ sudo pip3 install GitPython pylint==1.9.5
Essentials: Packages needed to build an image on a headless system:
$ sudo apt-get install gawk wget git-core diffstat unzip texinfo gcc-multilib \ build-essential chrpath socat cpio python3 python3-pip python3-pexpect \ xz-utils debianutils iputils-ping python3-git python3-jinja2 libegl1-mesa libsdl1.2-dev \ pylint3 xterm
Documentation: Packages needed if you are going to build out the Yocto Project documentation manuals:
$ sudo apt-get install make xsltproc docbook-utils fop dblatex xmlto
The following list shows the required packages by function given a supported Fedora Linux distribution:
Essentials: Packages needed to build an image for a headless system:
$ sudo dnf install gawk make wget tar bzip2 gzip python3 unzip perl patch \ diffutils diffstat git cpp gcc gcc-c++ glibc-devel texinfo chrpath \ ccache perl-Data-Dumper perl-Text-ParseWords perl-Thread-Queue perl-bignum socat \ python3-pexpect findutils which file cpio python python3-pip xz python3-GitPython \ python3-jinja2 SDL-devel xterm rpcgen
Documentation: Packages needed if you are going to build out the Yocto Project documentation manuals:
$ sudo dnf install docbook-style-dsssl docbook-style-xsl \ docbook-dtds docbook-utils fop libxslt dblatex xmlto
The following list shows the required packages by function given a supported openSUSE Linux distribution:
Essentials: Packages needed to build an image for a headless system:
$ sudo zypper install python gcc gcc-c++ git chrpath make wget python-xml \ diffstat makeinfo python-curses patch socat python3 python3-curses tar python3-pip \ python3-pexpect xz which python3-Jinja2 Mesa-libEGL1 libSDL-devel xterm rpcgen $ sudo pip3 install GitPython
Documentation: Packages needed if you are going to build out the Yocto Project documentation manuals:
$ sudo zypper install dblatex xmlto
The following list shows the required packages by function given a supported CentOS-7 Linux distribution:
Essentials: Packages needed to build an image for a headless system:
$ sudo yum install -y epel-release $ sudo yum makecache $ sudo yum install gawk make wget tar bzip2 gzip python3 unzip perl patch \ diffutils diffstat git cpp gcc gcc-c++ glibc-devel texinfo chrpath socat \ perl-Data-Dumper perl-Text-ParseWords perl-Thread-Queue python36-pip xz \ which SDL-devel xterm $ sudo pip3 install GitPython jinja2
Extra Packages for Enterprise Linux
(i.e. epel-release
)
is a collection of packages from Fedora
built on RHEL/CentOS for easy installation
of packages not included in enterprise
Linux by default.
You need to install these packages
separately.
The makecache
command
consumes additional Metadata from
epel-release
.
Documentation: Packages needed if you are going to build out the Yocto Project documentation manuals:
$ sudo yum install docbook-style-dsssl docbook-style-xsl \ docbook-dtds docbook-utils fop libxslt dblatex xmlto
The following list shows the required packages by function given a supported CentOS-8 Linux distribution:
Essentials: Packages needed to build an image for a headless system:
$ sudo dnf install -y epel-release $ sudo dnf config-manager --set-enabled PowerTools $ sudo dnf makecache $ sudo dnf install gawk make wget tar bzip2 gzip python3 unzip perl patch \ diffutils diffstat git cpp gcc gcc-c++ glibc-devel texinfo chrpath ccache \ socat perl-Data-Dumper perl-Text-ParseWords perl-Thread-Queue python3-pip \ python3-GitPython python3-jinja2 python3-pexpect xz which SDL-devel xterm \ rpcgen
Extra Packages for Enterprise Linux
(i.e. epel-release
)
is a collection of packages from Fedora
built on RHEL/CentOS for easy installation
of packages not included in enterprise
Linux by default.
You need to install these packages
separately.
The PowerTools
repo
provides additional packages such as
rpcgen
and
texinfo
.
The makecache
command
consumes additional Metadata from
epel-release
.
Documentation: Packages needed if you are going to build out the Yocto Project documentation manuals:
$ sudo dnf install docbook-style-dsssl docbook-style-xsl \ docbook-dtds docbook-utils fop libxslt dblatex xmlto
In order to use the build system, your host development system must meet the following version requirements for Git, tar, and Python:
Git 1.8.3.1 or greater
tar 1.28 or greater
Python 3.5.0 or greater
If your host development system does not meet all these requirements,
you can resolve this by installing a buildtools
tarball that contains these tools.
You can get the tarball one of two ways: download a pre-built
tarball or use BitBake to build the tarball.
In addition, your host development system must meet the following version requirement for gcc:
gcc 5.0 or greater
If your host development system does not meet this requirement,
you can resolve this by installing a buildtools-extended
tarball that contains additional tools, the equivalent of buildtools-essential
.
buildtools
Tarball with install-buildtools
script¶
The install-buildtools
script is the easiest
of the three methods by which you can get these tools. It downloads
a pre-built buildtools installer and automatically installs the tools
for you:
Execute the install-buildtools
script.
Here is an example:
$ cd poky $ scripts/install-buildtools --without-extended-buildtools \ --base-url http://downloads.yoctoproject.org/releases/yocto \ --release yocto-3.1.2 \ --installer-version 3.1.2
During execution, the buildtools tarball will be downloaded, the checksum of the download will be verified, the installer will be run for you, and some basic checks will be run to to make sure the installation is functional.
To avoid the need of sudo
privileges,
the install-buildtools
script will
by default tell the installer to install in:
/path/to/
poky/buildtools
If your host development system needs the additional tools
provided in the buildtools-extended
tarball, you can instead execute the
install-buildtools
script with the
default parameters:
$ cd poky $ scripts/install-buildtools
Source the tools environment setup script by using a command like the following:
$ source /path/to/
poky/buildtools/environment-setup-x86_64-pokysdk-linux
Of course, you need to supply your installation directory and be sure to use the right file (i.e. i586 or x86_64).
After you have sourced the setup script,
the tools are added to PATH
and any other environment variables required to run the
tools are initialized.
The results are working versions versions of Git, tar,
Python and chrpath
. And in the case of
the buildtools-extended
tarball, additional
working versions of tools including gcc
,
make
and the other tools included in
packagegroup-core-buildessential
.
buildtools
Tarball¶Downloading and running a pre-built buildtools installer is the easiest of the two methods by which you can get these tools:
Locate and download the *.sh
at
http://downloads.yoctoproject.org/releases/yocto/yocto-3.1.2/buildtools/.
Execute the installation script. Here is an example for the traditional installer:
$ sh ~/Downloads/x86_64-buildtools-nativesdk-standalone-3.1.2.sh
Here is an example for the extended installer:
$ sh ~/Downloads/x86_64-buildtools-extended-nativesdk-standalone-3.1.2.sh
During execution, a prompt appears that allows you to choose the installation directory. For example, you could choose the following:
/home/your-username
/buildtools
Source the tools environment setup script by using a command like the following:
$ source /home/your_username
/buildtools/environment-setup-i586-poky-linux
Of course, you need to supply your installation directory and be sure to use the right file (i.e. i585 or x86-64).
After you have sourced the setup script,
the tools are added to PATH
and any other environment variables required to run the
tools are initialized.
The results are working versions versions of Git, tar,
Python and chrpath
. And in the case of
the buildtools-extended
tarball, additional
working versions of tools including gcc
,
make
and the other tools included in
packagegroup-core-buildessential
.
buildtools
Tarball¶
Building and running your own buildtools installer applies
only when you have a build host that can already run BitBake.
In this case, you use that machine to build the
.sh
file and then
take steps to transfer and run it on a
machine that does not meet the minimal Git, tar, and Python
(or gcc) requirements.
Here are the steps to take to build and run your own buildtools installer:
On the machine that is able to run BitBake,
be sure you have set up your build environment with
the setup script
(oe-init-build-env
).
Run the BitBake command to build the tarball:
$ bitbake buildtools-tarball
or run the BitBake command to build the extended tarball:
$ bitbake buildtools-extended-tarball
SDKMACHINE
variable in your local.conf
file
determines whether you build tools for a 32-bit
or 64-bit system.
Once the build completes, you can find the
.sh
file that installs
the tools in the tmp/deploy/sdk
subdirectory of the
Build Directory.
The installer file has the string "buildtools"
(or "buildtools-extended") in the name.
Transfer the .sh
file from the
build host to the machine that does not meet the
Git, tar, or Python (or gcc) requirements.
On the machine that does not meet the requirements,
run the .sh
file
to install the tools.
Here is an example for the traditional installer:
$ sh ~/Downloads/x86_64-buildtools-nativesdk-standalone-3.1.2.sh
Here is an example for the extended installer:
$ sh ~/Downloads/x86_64-buildtools-extended-nativesdk-standalone-3.1.2.sh
During execution, a prompt appears that allows you to choose the installation directory. For example, you could choose the following:
/home/your_username
/buildtools
Source the tools environment setup script by using a command like the following:
$ source /home/your_username
/buildtools/environment-setup-x86_64-poky-linux
Of course, you need to supply your installation directory and be sure to use the right file (i.e. i586 or x86_64).
After you have sourced the setup script,
the tools are added to PATH
and any other environment variables required to run the
tools are initialized.
The results are working versions versions of Git, tar,
Python and chrpath
. And in the case of
the buildtools-extended
tarball, additional
working versions of tools including gcc
,
make
and the other tools included in
packagegroup-core-buildessential
.
Following is a list of terms and definitions users new to the Yocto Project development environment might find helpful. While some of these terms are universal, the list includes them just in case:
Append Files:
Files that append build information to a recipe file.
Append files are known as BitBake append files and
.bbappend
files.
The OpenEmbedded build system expects every append file to have
a corresponding recipe (.bb
) file.
Furthermore, the append file and corresponding recipe file
must use the same root filename.
The filenames can differ only in the file type suffix used
(e.g.
formfactor_0.0.bb
and
formfactor_0.0.bbappend
).
Information in append files extends or overrides the information in the similarly-named recipe file. For an example of an append file in use, see the "Using .bbappend Files in Your Layer" section in the Yocto Project Development Tasks Manual.
When you name an append file, you can use the
"%
" wildcard character to allow for
matching recipe names.
For example, suppose you have an append file named as follows:
busybox_1.21.%.bbappend
That append file would match any
busybox_1.21.
x
.bb
version of the recipe.
So, the append file would match any of the following recipe names:
busybox_1.21.1.bb busybox_1.21.2.bb busybox_1.21.3.bb busybox_1.21.10.bb busybox_1.21.25.bb
%
" character
is limited in that it only works directly in front of the
.bbappend
portion of the append file's
name.
You cannot use the wildcard character in any other
location of the name.
BitBake: The task executor and scheduler used by the OpenEmbedded build system to build images. For more information on BitBake, see the BitBake User Manual.
Board Support Package (BSP): A group of drivers, definitions, and other components that provide support for a specific hardware configuration. For more information on BSPs, see the Yocto Project Board Support Package (BSP) Developer's Guide.
Build Directory:
This term refers to the area used by the OpenEmbedded build
system for builds.
The area is created when you source
the
setup environment script that is found in the Source Directory
(i.e. oe-init-build-env
).
The
TOPDIR
variable points to the Build Directory.
You have a lot of flexibility when creating the Build
Directory.
Following are some examples that show how to create the
directory.
The examples assume your
Source Directory is
named poky
:
Create the Build Directory inside your
Source Directory and let the name of the Build
Directory default to build
:
$ cd $HOME/poky $ source oe-init-build-env
Create the Build Directory inside your
home directory and specifically name it
test-builds
:
$ cd $HOME $ source poky/oe-init-build-env test-builds
Provide a directory path and specifically name the
Build Directory.
Any intermediate folders in the pathname must exist.
This next example creates a Build Directory named
YP-23.0.2
in your home directory within the existing
directory mybuilds
:
$ cd $HOME $ source $HOME/poky/oe-init-build-env $HOME/mybuilds/YP-23.0.2
TMPDIR
,
which is a temporary directory the build system uses for
its work.
TMPDIR
cannot be under NFS.
Thus, by default, the Build Directory cannot be under NFS.
However, if you need the Build Directory to be under NFS,
you can set this up by setting TMPDIR
in your local.conf
file
to use a local drive.
Doing so effectively separates TMPDIR
from TOPDIR
, which is the Build
Directory.
Build Host: The system used to build images in a Yocto Project Development environment. The build system is sometimes referred to as the development host.
Classes:
Files that provide for logic encapsulation and inheritance so
that commonly used patterns can be defined once and then
easily used in multiple recipes.
For reference information on the Yocto Project classes, see the
"Classes" chapter.
Class files end with the .bbclass
filename extension.
Configuration File: Files that hold global definitions of variables, user-defined variables, and hardware configuration information. These files tell the OpenEmbedded build system what to build and what to put into the image to support a particular platform.
Configuration files end with a .conf
filename extension.
The conf/local.conf
configuration file in
the
Build Directory
contains user-defined variables that affect every build.
The meta-poky/conf/distro/poky.conf
configuration file defines Yocto "distro" configuration
variables used only when building with this policy.
Machine configuration files, which
are located throughout the
Source Directory, define
variables for specific hardware and are only used when building
for that target (e.g. the
machine/beaglebone.conf
configuration
file defines variables for the Texas Instruments ARM Cortex-A8
development board).
Container Layer:
Layers that hold other layers.
An example of a container layer is OpenEmbedded's
meta-openembedded
layer.
The meta-openembedded
layer contains
many meta-*
layers.
Cross-Development Toolchain: In general, a cross-development toolchain is a collection of software development tools and utilities that run on one architecture and allow you to develop software for a different, or targeted, architecture. These toolchains contain cross-compilers, linkers, and debuggers that are specific to the target architecture.
The Yocto Project supports two different cross-development toolchains:
A toolchain only used by and within BitBake when building an image for a target architecture.
A relocatable toolchain used outside of BitBake by developers when developing applications that will run on a targeted device.
Creation of these toolchains is simple and automated. For information on toolchain concepts as they apply to the Yocto Project, see the "Cross-Development Toolchain Generation" section in the Yocto Project Overview and Concepts Manual. You can also find more information on using the relocatable toolchain in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
Extensible Software Development Kit (eSDK): A custom SDK for application developers. This eSDK allows developers to incorporate their library and programming changes back into the image to make their code available to other application developers.
For information on the eSDK, see the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
Image: An image is an artifact of the BitBake build process given a collection of recipes and related Metadata. Images are the binary output that run on specific hardware or QEMU and are used for specific use-cases. For a list of the supported image types that the Yocto Project provides, see the "Images" chapter.
Layer: A collection of related recipes. Layers allow you to consolidate related metadata to customize your build. Layers also isolate information used when building for multiple architectures. Layers are hierarchical in their ability to override previous specifications. You can include any number of available layers from the Yocto Project and customize the build by adding your layers after them. You can search the Layer Index for layers used within Yocto Project.
For introductory information on layers, see the "The Yocto Project Layer Model" section in the Yocto Project Overview and Concepts Manual. For more detailed information on layers, see the "Understanding and Creating Layers" section in the Yocto Project Development Tasks Manual. For a discussion specifically on BSP Layers, see the "BSP Layers" section in the Yocto Project Board Support Packages (BSP) Developer's Guide.
Metadata: A key element of the Yocto Project is the Metadata that is used to construct a Linux distribution and is contained in the files that the OpenEmbedded build system parses when building an image. In general, Metadata includes recipes, configuration files, and other information that refers to the build instructions themselves, as well as the data used to control what things get built and the effects of the build. Metadata also includes commands and data used to indicate what versions of software are used, from where they are obtained, and changes or additions to the software itself (patches or auxiliary files) that are used to fix bugs or customize the software for use in a particular situation. OpenEmbedded-Core is an important set of validated metadata.
In the context of the kernel ("kernel Metadata"), the
term refers to the kernel config fragments and features
contained in the
yocto-kernel-cache
Git repository.
OpenEmbedded-Core (OE-Core): OE-Core is metadata comprised of foundational recipes, classes, and associated files that are meant to be common among many different OpenEmbedded-derived systems, including the Yocto Project. OE-Core is a curated subset of an original repository developed by the OpenEmbedded community that has been pared down into a smaller, core set of continuously validated recipes. The result is a tightly controlled and an quality-assured core set of recipes.
You can see the Metadata in the
meta
directory of the Yocto Project
Source Repositories.
OpenEmbedded Build System: The build system specific to the Yocto Project. The OpenEmbedded build system is based on another project known as "Poky", which uses BitBake as the task executor. Throughout the Yocto Project documentation set, the OpenEmbedded build system is sometimes referred to simply as "the build system". If other build systems, such as a host or target build system are referenced, the documentation clearly states the difference.
Package: In the context of the Yocto Project, this term refers to a recipe's packaged output produced by BitBake (i.e. a "baked recipe"). A package is generally the compiled binaries produced from the recipe's sources. You "bake" something by running it through BitBake.
It is worth noting that the term "package" can, in general, have subtle meanings. For example, the packages referred to in the "Required Packages for the Build Host" section are compiled binaries that, when installed, add functionality to your Linux distribution.
Another point worth noting is that historically within
the Yocto Project, recipes were referred to as packages - thus,
the existence of several BitBake variables that are seemingly
mis-named,
(e.g. PR
,
PV
, and
PE
).
Package Groups:
Arbitrary groups of software Recipes.
You use package groups to hold recipes that, when built,
usually accomplish a single task.
For example, a package group could contain the recipes for a
company’s proprietary or value-add software.
Or, the package group could contain the recipes that enable
graphics.
A package group is really just another recipe.
Because package group files are recipes, they end with the
.bb
filename extension.
Poky: Poky, which is pronounced Pock-ee, is a reference embedded distribution and a reference test configuration. Poky provides the following:
A base-level functional distro used to illustrate how to customize a distribution.
A means by which to test the Yocto Project components (i.e. Poky is used to validate the Yocto Project).
A vehicle through which you can download the Yocto Project.
Poky is not a product level distro. Rather, it is a good starting point for customization.
Recipe:
A set of instructions for building packages.
A recipe describes where you get source code, which patches
to apply, how to configure the source, how to compile it and so on.
Recipes also describe dependencies for libraries or for other
recipes.
Recipes represent the logical unit of execution, the software
to build, the images to build, and use the
.bb
file extension.
Reference Kit: A working example of a system, which includes a BSP as well as a build host and other components, that can work on specific hardware.
Source Directory:
This term refers to the directory structure created as a result
of creating a local copy of the poky
Git
repository git://git.yoctoproject.org/poky
or expanding a released poky
tarball.
poky
Git repository is the recommended method for setting up
your Source Directory.
Sometimes you might hear the term "poky directory" used to refer to this directory structure.
The Source Directory contains BitBake, Documentation, Metadata and other files that all support the Yocto Project. Consequently, you must have the Source Directory in place on your development system in order to do any development using the Yocto Project.
When you create a local copy of the Git repository, you
can name the repository anything you like.
Throughout much of the documentation, "poky"
is used as the name of the top-level folder of the local copy of
the poky Git repository.
So, for example, cloning the poky
Git
repository results in a local Git repository whose top-level
folder is also named "poky".
While it is not recommended that you use tarball expansion
to set up the Source Directory, if you do, the top-level
directory name of the Source Directory is derived from the
Yocto Project release tarball.
For example, downloading and unpacking
poky-dunfell-23.0.2.tar.bz2
results in a
Source Directory whose root folder is named
poky-dunfell-23.0.2
.
It is important to understand the differences between the
Source Directory created by unpacking a released tarball as
compared to cloning
git://git.yoctoproject.org/poky
.
When you unpack a tarball, you have an exact copy of the files
based on the time of release - a fixed release point.
Any changes you make to your local files in the Source Directory
are on top of the release and will remain local only.
On the other hand, when you clone the poky
Git repository, you have an active development repository with
access to the upstream repository's branches and tags.
In this case, any local changes you make to the local
Source Directory can be later applied to active development
branches of the upstream poky
Git
repository.
For more information on concepts related to Git repositories, branches, and tags, see the "Repositories, Tags, and Branches" section in the Yocto Project Overview and Concepts Manual.
Task:
A unit of execution for BitBake (e.g.
do_compile
,
do_fetch
,
do_patch
,
and so forth).
Toaster: A web interface to the Yocto Project's OpenEmbedded Build System. The interface enables you to configure and run your builds. Information about builds is collected and stored in a database. For information on Toaster, see the Toaster User Manual.
Upstream: A reference to source code or repositories that are not local to the development system but located in a master area that is controlled by the maintainer of the source code. For example, in order for a developer to work on a particular piece of code, they need to first get a copy of it from an "upstream" source.
The Yocto Project release process is predictable and consists of both major and minor (point) releases. This brief chapter provides information on how releases are named, their life cycle, and their stability.
The Yocto Project delivers major releases (e.g. 3.1.2) using a six month cadence roughly timed each April and October of the year. Following are examples of some major YP releases with their codenames also shown. See the "Major Release Codenames" section for information on codenames used with major releases.
2.2 (Morty) 2.1 (Krogoth) 2.0 (Jethro)
While the cadence is never perfect, this timescale facilitates regular releases that have strong QA cycles while not overwhelming users with too many new releases. The cadence is predictable and avoids many major holidays in various geographies.
The Yocto project delivers minor (point) releases on an unscheduled basis and are usually driven by the accumulation of enough significant fixes or enhancements to the associated major release. Following are some example past point releases:
2.1.1 2.1.2 2.2.1
The point release indicates a point in the major release branch where a full QA cycle and release process validates the content of the new branch.
Each major release receives a codename that identifies the release in the Yocto Project Source Repositories. The concept is that branches of Metadata with the same codename are likely to be compatible and thus work together.
Releases are given a nominal release version as well but the codename is used in repositories for this reason. You can find information on Yocto Project releases and codenames at https://wiki.yoctoproject.org/wiki/Releases.
Once released, the release enters the stable release process at which time a person is assigned as the maintainer for that stable release. This maintainer monitors activity for the release by investigating and handling nominated patches and backport activity. Only fixes and enhancements that have first been applied on the "master" branch (i.e. the current, in-development branch) are considered for backporting to a stable release.
Stable release branches have strong maintenance for about a year after their initial release. Should significant issues be found for any release regardless of its age, fixes could be backported to older releases. For issues that are not backported given an older release, Community LTS trees and branches exist where community members share patches for older releases. However, these types of patches do not go through the same release process as do point releases. You can find more information about stable branch maintenance at https://wiki.yoctoproject.org/wiki/Stable_branch_maintenance.
Part of the Yocto Project development and release process is quality assurance through the execution of test strategies. Test strategies provide the Yocto Project team a way to ensure a release is validated. Additionally, because the test strategies are visible to you as a developer, you can validate your projects. This section overviews the available test infrastructure used in the Yocto Project. For information on how to run available tests on your projects, see the "Performing Automated Runtime Testing" section in the Yocto Project Development Tasks Manual.
The QA/testing infrastructure is woven into the project to the point where core developers take some of it for granted. The infrastructure consists of the following pieces:
bitbake-selftest
:
A standalone command that runs unit tests on key pieces of
BitBake and its fetchers.
sanity.bbclass
:
This automatically included class checks the build environment
for missing tools (e.g. gcc
) or common
misconfigurations such as
MACHINE
set incorrectly.
insane.bbclass
:
This class checks the generated output from builds for sanity.
For example, if building for an ARM target, did the build
produce ARM binaries.
If, for example, the build produced PPC binaries then there
is a problem.
testimage.bbclass
:
This class performs runtime testing of images after they are
built.
The tests are usually used with
QEMU
to boot the images and check the combined runtime result
boot operation and functions.
However, the test can also use the IP address of a machine to
test.
ptest
:
Runs tests against packages produced during the build for a
given piece of software.
The test allows the packages to be be run within a target
image.
oe-selftest
:
Tests combination BitBake invocations.
These tests operate outside the OpenEmbedded build system
itself.
The oe-selftest
can run all tests by
default or can run selected tests or test suites.
oe-selftest
requires
host packages beyond the "Essential" grouping.
See the
"Required Packages for the Build Host"
section for more information.
Originally, much of this testing was done manually. However, significant effort has been made to automate the tests so that more people can use them and the Yocto Project development team can run them faster and more efficiently.
The Yocto Project's main Autobuilder
(autobuilder.yoctoproject.org
) publicly tests
each Yocto Project release's code in the
OE-Core, Poky, and BitBake
repositories.
The testing occurs for both the current state of the
"master" branch and also for submitted patches.
Testing for submitted patches usually occurs in the
"ross/mut" branch in the poky-contrib
repository
(i.e. the master-under-test branch) or in the "master-next" branch
in the poky
repository.
Testing within these public branches ensures in a publicly visible way that all of the main supposed architectures and recipes in OE-Core successfully build and behave properly.
Various features such as multilib
, sub
architectures (e.g. x32
,
poky-tiny
, musl
,
no-x11
and and so forth),
bitbake-selftest
, and
oe-selftest
are tested as part of
the QA process of a release.
Complete testing and validation for a release takes the Autobuilder
workers several hours.
Finally, in addition to the Autobuilder's tests, the Yocto Project QA team also performs testing on a variety of platforms, which includes actual hardware, to ensure expected results.
atom-pc
Board Support Package (BSP)SRCREV
ValuesIMAGE_FEATURES
/run
udev
archiver
Classqemu-native
core-image-basic
CFLAGS
Optionsupdate-alternative
Providervirtclass
OverridesPACKAGECONFIG
Options in local.conf
eglibc 2.19
Replaced with glibc 2.20
linux-yocto
Kernel Metadata Repository Now Split from SourcegetVar()
and
getVarFlag()
is Now Mandatorylibexecdir
Reverted to ${prefix}/libexec
ac_cv_sizeof_off_t
is No Longer Cached in Site Filestmp/deploy
Now Enabled${B}
No Longer Default Working Directory for Tasksrunqemu
Ported to PythonKERNEL_IMAGE_BASE_NAME
no Longer Uses KERNEL_IMAGETYPE
PATH
Variabledistutils
and distutils3
Now Prevent Fetching Dependencies During the do_configure
Tasklinux-yocto
Configuration Audit Issues Now Correctly ReportedSERIAL_CONSOLE
Deprecatedsystemd
Configuration is Now Split Into systemd-conf
qemu-native
Splits the System and User-Mode Partsupstream-tracking.inc
File Has Been RemovedDISTRO_FEATURES_LIBC
Variable Has Been Removedwic
image type now used instead of live
by default for x86This chapter provides information you can use to migrate work to a newer Yocto Project release. You can find the same information in the release notes for a given release.
Some considerations are not tied to a specific Yocto Project release. This section presents information you should consider when migrating to any new Yocto Project release.
Dealing with Customized Recipes: Issues could arise if you take older recipes that contain customizations and simply copy them forward expecting them to work after you migrate to new Yocto Project metadata. For example, suppose you have a recipe in your layer that is a customized version of a core recipe copied from the earlier release, rather than through the use of an append file. When you migrate to a newer version of Yocto Project, the metadata (e.g. perhaps an include file used by the recipe) could have changed in a way that would break the build. Say, for example, a function is removed from an include file and the customized recipe tries to call that function.
You could "forward-port" all your customizations in your recipe so that everything works for the new release. However, this is not the optimal solution as you would have to repeat this process with each new release if changes occur that give rise to problems.
The better solution (where practical) is to use append
files (*.bbappend
) to capture any
customizations you want to make to a recipe.
Doing so, isolates your changes from the main recipe making
them much more manageable.
However, sometimes it is not practical to use an append
file.
A good example of this is when introducing a newer or older
version of a recipe in another layer.
Updating Append Files:
Since append files generally only contain your customizations,
they often do not need to be adjusted for new releases.
However, if the .bbappend
file is
specific to a particular version of the recipe (i.e. its
name does not use the % wildcard) and the version of the
recipe to which it is appending has changed, then you will
at a minimum need to rename the append file to match the
name of the recipe file.
A mismatch between an append file and its corresponding
recipe file (.bb
) will
trigger an error during parsing.
Depending on the type of customization the append file applies, other incompatibilities might occur when you upgrade. For example, if your append file applies a patch and the recipe to which it is appending is updated to a newer version, the patch might no longer apply. If this is the case and assuming the patch is still needed, you must modify the patch file so that it does apply.
This section provides migration information for moving to the Yocto Project 1.3 Release from the prior release.
Differences include changes for
SSTATE_MIRRORS
and bblayers.conf
.
The shared state cache (sstate-cache), as pointed to by
SSTATE_DIR
,
by default now has two-character subdirectories to prevent
issues arising from too many files in the same directory.
Also, native sstate-cache packages, which are built to run
on the host system, will go into a subdirectory named using
the distro ID string.
If you copy the newly structured sstate-cache to a mirror
location (either local or remote) and then point to it in
SSTATE_MIRRORS
,
you need to append "PATH" to the end of the mirror URL so that
the path used by BitBake before the mirror substitution is
appended to the path used to access the mirror.
Here is an example:
SSTATE_MIRRORS = "file://.* http://someserver
.tld/share/sstate/PATH"
The meta-yocto
layer consists of two parts
that correspond to the Poky reference distribution and the
reference hardware Board Support Packages (BSPs), respectively:
meta-yocto
and
meta-yocto-bsp
.
When running BitBake for the first time after upgrading,
your conf/bblayers.conf
file will be
updated to handle this change and you will be asked to
re-run or restart for the changes to take effect.
Differences include changes for the following:
Python function whitespace
proto=
in SRC_URI
nativesdk
Task recipes
IMAGE_FEATURES
Removed recipes
All Python functions must now use four spaces for indentation.
Previously, an inconsistent mix of spaces and tabs existed,
which made extending these functions using
_append
or _prepend
complicated given that Python treats whitespace as
syntactically significant.
If you are defining or extending any Python functions (e.g.
populate_packages
, do_unpack
,
do_patch
and so forth) in custom recipes
or classes, you need to ensure you are using consistent
four-space indentation.
Any use of proto=
in
SRC_URI
needs to be changed to protocol=
.
In particular, this applies to the following URIs:
svn://
bzr://
hg://
osc://
Other URIs were already using protocol=
.
This change improves consistency.
The suffix nativesdk
is now implemented
as a prefix, which simplifies a lot of the packaging code for
nativesdk
recipes.
All custom nativesdk
recipes, which are
relocatable packages that are native to
SDK_ARCH
,
and any references need to be updated to use
nativesdk-*
instead of
*-nativesdk
.
"Task" recipes are now known as "Package groups" and have
been renamed from task-*.bb
to
packagegroup-*.bb
.
Existing references to the previous task-*
names should work in most cases as there is an automatic
upgrade path for most packages.
However, you should update references in your own recipes and
configurations as they could be removed in future releases.
You should also rename any custom task-*
recipes to packagegroup-*
, and change
them to inherit packagegroup
instead of
task
, as well as taking the opportunity
to remove anything now handled by
packagegroup.bbclass
, such as providing
-dev
and -dbg
packages, setting
LIC_FILES_CHKSUM
,
and so forth.
See the
"packagegroup.bbclass
"
section for further details.
Image recipes that previously included "apps-console-core"
in IMAGE_FEATURES
should now include "splash" instead to enable the boot-up
splash screen.
Retaining "apps-console-core" will still include the splash
screen but generates a warning.
The "apps-x11-core" and "apps-x11-games"
IMAGE_FEATURES
features have been removed.
The following recipes have been removed. For most of them, it is unlikely that you would have any references to them in your own Metadata. However, you should check your metadata against this list to be sure:
libx11-trim
:
Replaced by libx11
, which has a negligible
size difference with modern Xorg.
xserver-xorg-lite
:
Use xserver-xorg
, which has a negligible
size difference when DRI and GLX modules are not installed.
xserver-kdrive
:
Effectively unmaintained for many years.
mesa-xlib
:
No longer serves any purpose.
galago
:
Replaced by telepathy.
gail
:
Functionality was integrated into GTK+ 2.13.
eggdbus
:
No longer needed.
gcc-*-intermediate
:
The build has been restructured to avoid the need for
this step.
libgsmd
:
Unmaintained for many years.
Functionality now provided by
ofono
instead.
contacts, dates, tasks, eds-tools:
Largely unmaintained PIM application suite.
It has been moved to meta-gnome
in meta-openembedded
.
In addition to the previously listed changes, the
meta-demoapps
directory has also been removed
because the recipes in it were not being maintained and many
had become obsolete or broken.
Additionally, these recipes were not parsed in the default configuration.
Many of these recipes are already provided in an updated and
maintained form within the OpenEmbedded community layers such as
meta-oe
and meta-gnome
.
For the remainder, you can now find them in the
meta-extras
repository, which is in the
Yocto Project
Source Repositories.
The naming scheme for kernel output binaries has been changed to
now include
PE
as part of the
filename:
KERNEL_IMAGE_BASE_NAME ?= "${KERNEL_IMAGETYPE}-${PE}-${PV}-${PR}-${MACHINE}-${DATETIME}"
Because the PE
variable is not set by default,
these binary files could result with names that include two dash
characters.
Here is an example:
bzImage--3.10.9+git0+cd502a8814_7144bcc4b8-r0-qemux86-64-20130830085431.bin
This section provides migration information for moving to the Yocto Project 1.4 Release from the prior release.
Differences include the following:
Comment Continuation: If a comment ends with a line continuation (\) character, then the next line must also be a comment. Any instance where this is not the case, now triggers a warning. You must either remove the continuation character, or be sure the next line is a comment.
Package Name Overrides:
The runtime package specific variables
RDEPENDS
,
RRECOMMENDS
,
RSUGGESTS
,
RPROVIDES
,
RCONFLICTS
,
RREPLACES
,
FILES
,
ALLOW_EMPTY
,
and the pre, post, install, and uninstall script functions
pkg_preinst
,
pkg_postinst
,
pkg_prerm
, and
pkg_postrm
should always have a
package name override.
For example, use RDEPENDS_${PN}
for
the main package instead of RDEPENDS
.
BitBake uses more strict checks when it parses recipes.
Differences include the following:
Shared State Code: The shared state code has been optimized to avoid running unnecessary tasks. For example, the following no longer populates the target sysroot since that is not necessary:
$ bitbake -c rootfs some-image
Instead, the system just needs to extract the output package contents, re-create the packages, and construct the root filesystem. This change is unlikely to cause any problems unless you have missing declared dependencies.
Scanning Directory Names:
When scanning for files in
SRC_URI
,
the build system now uses
FILESOVERRIDES
instead of OVERRIDES
for the directory names.
In general, the values previously in
OVERRIDES
are now in
FILESOVERRIDES
as well.
However, if you relied upon an additional value
you previously added to OVERRIDES
,
you might now need to add it to
FILESOVERRIDES
unless you are already
adding it through the
MACHINEOVERRIDES
or DISTROOVERRIDES
variables, as appropriate.
For more related changes, see the
"Variables"
section.
A new oe-git-proxy
script has been added to
replace previous methods of handling proxies and fetching source
from Git.
See the meta-yocto/conf/site.conf.sample
file
for information on how to use this script.
If you have created your own custom
etc/network/interfaces
file by creating
an append file for the netbase
recipe,
you now need to create an append file for the
init-ifupdown
recipe instead, which you can
find in the
Source Directory
at meta/recipes-core/init-ifupdown
.
For information on how to use append files, see the
"Using .bbappend Files"
section in the Yocto Project Development Tasks Manual.
Support for remote debugging with the Eclipse IDE is now
separated into an image feature
(eclipse-debug
) that corresponds to the
packagegroup-core-eclipse-debug
package group.
Previously, the debugging feature was included through the
tools-debug
image feature, which corresponds
to the packagegroup-core-tools-debug
package group.
The following variables have changed:
SANITY_TESTED_DISTROS
:
This variable now uses a distribution ID, which is composed
of the host distributor ID followed by the release.
Previously,
SANITY_TESTED_DISTROS
was composed of the description field.
For example, "Ubuntu 12.10" becomes "Ubuntu-12.10".
You do not need to worry about this change if you are not
specifically setting this variable, or if you are
specifically setting it to "".
SRC_URI
:
The ${
PN
}
,
${
PF
}
,
${
P
}
,
and FILE_DIRNAME
directories have been
dropped from the default value of the
FILESPATH
variable, which is used as the search path for finding files
referred to in
SRC_URI
.
If you have a recipe that relied upon these directories,
which would be unusual, then you will need to add the
appropriate paths within the recipe or, alternatively,
rearrange the files.
The most common locations are still covered by
${BP}
, ${BPN}
,
and "files", which all remain in the default value of
FILESPATH
.
If runtime package management is enabled and the RPM backend is selected, Smart is now installed for package download, dependency resolution, and upgrades instead of Zypper. For more information on how to use Smart, run the following command on the target:
smart --help
The following recipes were moved from their previous locations because they are no longer used by anything in the OpenEmbedded-Core:
clutter-box2d
:
Now resides in the meta-oe
layer.
evolution-data-server
:
Now resides in the meta-gnome
layer.
gthumb
:
Now resides in the meta-gnome
layer.
gtkhtml2
:
Now resides in the meta-oe
layer.
gupnp
:
Now resides in the meta-multimedia
layer.
gypsy
:
Now resides in the meta-oe
layer.
libcanberra
:
Now resides in the meta-gnome
layer.
libgdata
:
Now resides in the meta-gnome
layer.
libmusicbrainz
:
Now resides in the meta-multimedia
layer.
metacity
:
Now resides in the meta-gnome
layer.
polkit
:
Now resides in the meta-oe
layer.
zeroconf
:
Now resides in the meta-networking
layer.
The following list shows what has been removed or renamed:
evieext
:
Removed because it has been removed from
xserver
since 2008.
Gtk+ DirectFB: Removed support because upstream Gtk+ no longer supports it as of version 2.18.
libxfontcache / xfontcacheproto
:
Removed because they were removed from the Xorg server in 2008.
libxp / libxprintapputil / libxprintutil / printproto
:
Removed because the XPrint server was removed from
Xorg in 2008.
libxtrap / xtrapproto
:
Removed because their functionality was broken upstream.
linux-yocto 3.0 kernel: Removed with linux-yocto 3.8 kernel being added. The linux-yocto 3.2 and linux-yocto 3.4 kernels remain as part of the release.
lsbsetup
:
Removed with functionality now provided by
lsbtest
.
matchbox-stroke
:
Removed because it was never more than a proof-of-concept.
matchbox-wm-2 / matchbox-theme-sato-2
:
Removed because they are not maintained.
However, matchbox-wm
and
matchbox-theme-sato
are still
provided.
mesa-dri
:
Renamed to mesa
.
mesa-xlib
:
Removed because it was no longer useful.
mutter
:
Removed because nothing ever uses it and the recipe is
very old.
orinoco-conf
:
Removed because it has become obsolete.
update-modules
:
Removed because it is no longer used.
The kernel module postinstall
and
postrm
scripts can now do the same
task without the use of this script.
web
:
Removed because it is not maintained. Superseded by
web-webkit
.
xf86bigfontproto
:
Removed because upstream it has been disabled by default
since 2007.
Nothing uses xf86bigfontproto
.
xf86rushproto
:
Removed because its dependency in
xserver
was spurious and it was
removed in 2005.
zypper / libzypp / sat-solver
:
Removed and been functionally replaced with Smart
(python-smartpm
) when RPM packaging
is used and package management is enabled on the target.
This section provides migration information for moving to the Yocto Project 1.5 Release from the prior release.
The OpenEmbedded build system now has some additional requirements on the host system:
Python 2.7.3+
Tar 1.24+
Git 1.7.8+
Patched version of Make if you are using 3.82. Most distributions that provide Make 3.82 use the patched version.
If the Linux distribution you are using on your build host does not provide packages for these, you can install and use the Buildtools tarball, which provides an SDK-like environment containing them.
For more information on this requirement, see the "Required Git, tar, Python and gcc Versions" section.
atom-pc
Board Support Package (BSP)¶
The atom-pc
hardware reference BSP has been
replaced by a genericx86
BSP.
This BSP is not necessarily guaranteed to work on all x86
hardware, but it will run on a wider range of systems than the
atom-pc
did.
genericx86-64
BSP has
been added for 64-bit Atom systems.
The following changes have been made that relate to BitBake:
BitBake now supports a _remove
operator.
The addition of this operator means you will have to
rename any items in recipe space (functions, variables)
whose names currently contain
_remove_
or end with
_remove
to avoid unexpected behavior.
BitBake's global method pool has been removed. This method is not particularly useful and led to clashes between recipes containing functions that had the same name.
The "none" server backend has been removed. The "process" server backend has been serving well as the default for a long time now.
The bitbake-runtask
script has been
removed.
${
P
}
and
${
PF
}
are no longer added to
PROVIDES
by default in bitbake.conf
.
These version-specific PROVIDES
items were seldom used.
Attempting to use them could result in two versions being
built simultaneously rather than just one version due to
the way BitBake resolves dependencies.
The following changes have been made to the package QA checks:
If you have customized
ERROR_QA
or WARN_QA
values in your configuration, check that they contain all of
the issues that you wish to be reported.
Previous Yocto Project versions contained a bug that meant
that any item not mentioned in ERROR_QA
or WARN_QA
would be treated as a
warning.
Consequently, several important items were not already in
the default value of WARN_QA
.
All of the possible QA checks are now documented in the
"insane.bbclass
"
section.
An additional QA check has been added to check if
/usr/share/info/dir
is being installed.
Your recipe should delete this file within
do_install
if "make install" is installing it.
If you are using the buildhistory class, the check for the
package version going backwards is now controlled using a
standard QA check.
Thus, if you have customized your
ERROR_QA
or
WARN_QA
values and still wish to have
this check performed, you should add
"version-going-backwards" to your value for one or the
other variables depending on how you wish it to be handled.
See the documented QA checks in the
"insane.bbclass
"
section.
The following directory changes exist:
Output SDK installer files are now named to include the
image name and tuning architecture through the
SDK_NAME
variable.
Images and related files are now installed into a directory
that is specific to the machine, instead of a parent
directory containing output files for multiple machines.
The
DEPLOY_DIR_IMAGE
variable continues to point to the directory containing
images for the current
MACHINE
and should be used anywhere there is a need to refer to
this directory.
The runqemu
script now uses this
variable to find images and kernel binaries and will use
BitBake to determine the directory.
Alternatively, you can set the
DEPLOY_DIR_IMAGE
variable in the
external environment.
When buildhistory is enabled, its output is now written
under the
Build Directory
rather than
TMPDIR
.
Doing so makes it easier to delete
TMPDIR
and preserve the build history.
Additionally, data for produced SDKs is now split by
IMAGE_NAME
.
The pkgdata
directory produced as
part of the packaging process has been collapsed into a
single machine-specific directory.
This directory is located under
sysroots
and uses a machine-specific
name (i.e.
tmp/sysroots/
).
machine
/pkgdata
SRCREV
Values¶
BitBake will now shorten revisions from Git repositories from the
normal 40 characters down to 10 characters within
SRCPV
for improved usability in path and file names.
This change should be safe within contexts where these revisions
are used because the chances of spatially close collisions
is very low.
Distant collisions are not a major issue in the way
the values are used.
IMAGE_FEATURES
¶
The following changes have been made that relate to
IMAGE_FEATURES
:
The value of IMAGE_FEATURES
is now
validated to ensure invalid feature items are not added.
Some users mistakenly add package names to this variable
instead of using
IMAGE_INSTALL
in order to have the package added to the image, which does
not work.
This change is intended to catch those kinds of situations.
Valid IMAGE_FEATURES
are drawn from
PACKAGE_GROUP
definitions,
COMPLEMENTARY_GLOB
and a new "validitems" varflag on
IMAGE_FEATURES
.
The "validitems" varflag change allows additional features
to be added if they are not provided using the previous
two mechanisms.
The previously deprecated "apps-console-core"
IMAGE_FEATURES
item is no longer
supported.
Add "splash" to IMAGE_FEATURES
if you
wish to have the splash screen enabled, since this is
all that apps-console-core was doing.
/run
¶
The /run
directory from the Filesystem
Hierarchy Standard 3.0 has been introduced.
You can find some of the implications for this change
here.
The change also means that recipes that install files to
/var/run
must be changed.
You can find a guide on how to make these changes
here.
The image core-image-minimal
no longer adds
remove_packaging_data_files
to
ROOTFS_POSTPROCESS_COMMAND
.
This addition is now handled automatically when "package-management"
is not in
IMAGE_FEATURES
.
If you have custom image recipes that make this addition,
you should remove the lines, as they are not needed and might
interfere with correct operation of postinstall scripts.
The
do_rootfs
and other related image
construction tasks are no longer marked as "nostamp".
Consequently, they will only be re-executed when their inputs have
changed.
Previous versions of the OpenEmbedded build system always rebuilt
the image when requested rather when necessary.
The previously deprecated task.bbclass
has
now been dropped.
For recipes that previously inherited from this class, you should
rename them from task-*
to
packagegroup-*
and inherit packagegroup
instead.
For more information, see the
"packagegroup.bbclass
"
section.
By default, we now split BusyBox into two binaries:
one that is suid root for those components that need it, and
another for the rest of the components.
Splitting BusyBox allows for optimization that eliminates the
tinylogin
recipe as recommended by upstream.
You can disable this split by setting
BUSYBOX_SPLIT_SUID
to "0".
A new automated image testing framework has been added
through the
testimage.bbclass
class.
This framework replaces the older
imagetest-qemu
framework.
You can learn more about performing automated image tests in the "Performing Automated Runtime Testing" section in the Yocto Project Development Tasks Manual.
Following are changes to Build History:
Installed package sizes:
installed-package-sizes.txt
for an
image now records the size of the files installed by each
package instead of the size of each compressed package
archive file.
The dependency graphs (depends*.dot
)
now use the actual package names instead of replacing
dashes, dots and plus signs with underscores.
The buildhistory-diff
and
buildhistory-collect-srcrevs
utilities have improved command-line handling.
Use the --help
option for
each utility for more information on the new syntax.
For more information on Build History, see the "Maintaining Build Output Quality" section in the Yocto Project Development Tasks Manual.
udev
¶
Following are changes to udev
:
udev
no longer brings in
udev-extraconf
automatically
through
RRECOMMENDS
,
since this was originally intended to be optional.
If you need the extra rules, then add
udev-extraconf
to your image.
udev
no longer brings in
pciutils-ids
or
usbutils-ids
through
RRECOMMENDS
.
These are not needed by udev
itself
and removing them saves around 350KB.
The linux-yocto
3.2 kernel has been
removed.
libtool-nativesdk
has been renamed to
nativesdk-libtool
.
tinylogin
has been removed.
It has been replaced by a suid portion of Busybox.
See the
"BusyBox" section
for more information.
external-python-tarball
has been renamed
to buildtools-tarball
.
web-webkit
has been removed.
It has been functionally replaced by
midori
.
imake
has been removed.
It is no longer needed by any other recipe.
transfig-native
has been removed.
It is no longer needed by any other recipe.
anjuta-remote-run
has been removed.
Anjuta IDE integration has not been officially supported for
several releases.
Following is a list of short entries describing other changes:
run-postinsts
: Make this generic.
base-files
: Remove the unnecessary
media/
xxx
directories.
alsa-state
: Provide an empty
asound.conf
by default.
classes/image
: Ensure
BAD_RECOMMENDATIONS
supports pre-renamed package names.
classes/rootfs_rpm
: Implement
BAD_RECOMMENDATIONS
for RPM.
systemd
: Remove
systemd_unitdir
if
systemd
is not in
DISTRO_FEATURES
.
systemd
: Remove
init.d
dir if
systemd
unit file is present and
sysvinit
is not a distro feature.
libpam
: Deny all services for the
OTHER
entries.
image.bbclass
: Move
runtime_mapping_rename
to avoid
conflict with multilib
.
See
YOCTO #4993
in Bugzilla for more information.
linux-dtb
: Use kernel build system
to generate the dtb
files.
kern-tools
: Switch from guilt to
new kgit-s2q
tool.
This section provides migration information for moving to the Yocto Project 1.6 Release from the prior release.
archiver
Class¶
The
archiver
class has been rewritten and its configuration has been simplified.
For more details on the source archiver, see the
"Maintaining Open Source License Compliance During Your Product's Lifecycle"
section in the Yocto Project Development Tasks Manual.
The following packaging changes have been made:
The binutils
recipe no longer produces
a binutils-symlinks
package.
update-alternatives
is now used to
handle the preferred binutils
variant on the target instead.
The tc (traffic control) utilities have been split out of
the main iproute2
package and put
into the iproute2-tc
package.
The gtk-engines
schemas have been
moved to a dedicated
gtk-engines-schemas
package.
The armv7a
with thumb package
architecture suffix has changed.
The suffix for these packages with the thumb
optimization enabled is "t2" as it should be.
Use of this suffix was not the case in the 1.5 release.
Architecture names will change within package feeds as a
result.
The following changes have been made to BitBake.
When fetching source from a Git repository using
SRC_URI
,
BitBake will now validate the
SRCREV
value against the branch.
You can specify the branch using the following form:
SRC_URI = "git://server.name/repository;branch=branchname
"
If you do not specify a branch, BitBake looks in the default "master" branch.
Alternatively, if you need to bypass this check (e.g.
if you are fetching a revision corresponding to a tag that
is not on any branch), you can add ";nobranch=1" to
the end of the URL within SRC_URI
.
BitBake had some previously deprecated Python definitions
within its bb
module removed.
You should use their sub-module counterparts instead:
bb.MalformedUrl
:
Use bb.fetch.MalformedUrl
.
bb.encodeurl
:
Use bb.fetch.encodeurl
.
bb.decodeurl
:
Use bb.fetch.decodeurl
bb.mkdirhier
:
Use bb.utils.mkdirhier
.
bb.movefile
:
Use bb.utils.movefile
.
bb.copyfile
:
Use bb.utils.copyfile
.
bb.which
:
Use bb.utils.which
.
bb.vercmp_string
:
Use bb.utils.vercmp_string
.
bb.vercmp
:
Use bb.utils.vercmp
.
The SVK fetcher has been removed from BitBake.
The BitBake console UI will now output errors to
stderr
instead of
stdout
.
Consequently, if you are piping or redirecting the output of
bitbake
to somewhere else, and you wish
to retain the errors, you will need to add
2>&1
(or something similar) to the
end of your bitbake
command line.
task-
taskname
Overrides¶
task-
taskname
overrides have been
adjusted so that tasks whose names contain underscores have the
underscores replaced by hyphens for the override so that they
now function properly.
For example, the task override for
do_populate_sdk
is task-populate-sdk
.
The following variables have changed. For information on the OpenEmbedded build system variables, see the "Variables Glossary" Chapter.
TMPDIR
¶
TMPDIR
can no longer be on an NFS mount.
NFS does not offer full POSIX locking and inode consistency
and can cause unexpected issues if used to store
TMPDIR
.
The check for this occurs on startup.
If TMPDIR
is detected on an NFS mount,
an error occurs.
PRINC
¶
The PRINC
variable has been deprecated and triggers a warning if
detected during a build.
For
PR
increments on changes, use the PR service instead.
You can find out more about this service in the
"Working With a PR Service"
section in the Yocto Project Development Tasks Manual.
IMAGE_TYPES
¶
The "sum.jffs2" option for
IMAGE_TYPES
has been replaced by the "jffs2.sum" option, which fits the
processing order.
COPY_LIC_MANIFEST
¶
The
COPY_LIC_MANIFEST
variable must
now be set to "1" rather than any value in order to enable
it.
COPY_LIC_DIRS
¶
The
COPY_LIC_DIRS
variable must
now be set to "1" rather than any value in order to enable
it.
PACKAGE_GROUP
¶
The
PACKAGE_GROUP
variable has been renamed to
FEATURE_PACKAGES
to more accurately reflect its purpose.
You can still use PACKAGE_GROUP
but
the OpenEmbedded build system produces a warning message when
it encounters the variable.
The following variables now expect a semicolon separated list of functions to call and not arbitrary shell commands:
ROOTFS_PREPROCESS_COMMAND ROOTFS_POSTPROCESS_COMMAND SDK_POSTPROCESS_COMMAND POPULATE_SDK_POST_TARGET_COMMAND POPULATE_SDK_POST_HOST_COMMAND IMAGE_POSTPROCESS_COMMAND IMAGE_PREPROCESS_COMMAND ROOTFS_POSTUNINSTALL_COMMAND ROOTFS_POSTINSTALL_COMMAND
For migration purposes, you can simply wrap shell commands in a shell function and then call the function. Here is an example:
my_postprocess_function() { echo "hello" > ${IMAGE_ROOTFS}/hello.txt } ROOTFS_POSTPROCESS_COMMAND += "my_postprocess_function; "
Package Tests (ptest) are built but not installed by default.
For information on using Package Tests, see the
"Testing Packages with ptest"
section in the Yocto Project Development Tasks Manual.
For information on the ptest
class, see the
"ptest.bbclass
"
section.
Separate build and source directories have been enabled
by default for selected recipes where it is known to work
(a whitelist) and for all recipes that inherit the
cmake
class.
In future releases the
autotools
class will enable a separate build directory by default as
well.
Recipes building Autotools-based
software that fails to build with a separate build directory
should be changed to inherit from the
autotools-brokensep
class instead of the autotools
or
autotools_stage
classes.
qemu-native
¶
qemu-native
now builds without
SDL-based graphical output support by default.
The following additional lines are needed in your
local.conf
to enable it:
PACKAGECONFIG_pn-qemu-native = "sdl" ASSUME_PROVIDED += "libsdl-native"
local.conf
contains these statements.
Consequently, if you are building a headless system and using
a default local.conf
file, you will need
comment these two lines out.
core-image-basic
¶
core-image-basic
has been renamed to
core-image-full-cmdline
.
In addition to core-image-basic
being renamed,
packagegroup-core-basic
has been renamed to
packagegroup-core-full-cmdline
to match.
The top-level LICENSE
file has been changed
to better describe the license of the various components of
OE-Core.
However, the licensing itself remains unchanged.
Normally, this change would not cause any side-effects.
However, some recipes point to this file within
LIC_FILES_CHKSUM
(as ${COREBASE}/LICENSE
) and thus the
accompanying checksum must be changed from
3f40d7994397109285ec7b81fdeb3b58 to
4d92cd373abda3937c2bc47fbc49d690.
A better alternative is to have
LIC_FILES_CHKSUM
point to a file
describing the license that is distributed with the source
that the recipe is building, if possible, rather than pointing
to ${COREBASE}/LICENSE
.
CFLAGS
Options¶
The "-fpermissive" option has been removed from the default
CFLAGS
value.
You need to take action on individual recipes that fail when
building with this option.
You need to either patch the recipes to fix the issues reported by
the compiler, or you need to add "-fpermissive" to
CFLAGS
in the recipes.
Custom image output types, as selected using
IMAGE_FSTYPES
,
must declare their dependencies on other image types (if any) using
a new
IMAGE_TYPEDEP
variable.
The do_package_write
task has been removed.
The task is no longer needed.
update-alternative
Provider¶
The default update-alternatives
provider has
been changed from opkg
to
opkg-utils
.
This change resolves some troublesome circular dependencies.
The runtime package has also been renamed from
update-alternatives-cworth
to update-alternatives-opkg
.
virtclass
Overrides¶
The virtclass
overrides are now deprecated.
Use the equivalent class overrides instead (e.g.
virtclass-native
becomes
class-native
.)
The following recipes have been removed:
packagegroup-toolset-native
-
This recipe is largely unused.
linux-yocto-3.8
-
Support for the Linux yocto 3.8 kernel has been dropped.
Support for the 3.10 and 3.14 kernels have been added
with the linux-yocto-3.10
and
linux-yocto-3.14
recipes.
ocf-linux
-
This recipe has been functionally replaced using
cryptodev-linux
.
genext2fs
-
genext2fs
is no longer used by the
build system and is unmaintained upstream.
js
-
This provided an ancient version of Mozilla's javascript
engine that is no longer needed.
zaurusd
-
The recipe has been moved to the
meta-handheld
layer.
eglibc 2.17
-
Replaced by the eglibc 2.19
recipe.
gcc 4.7.2
-
Replaced by the now stable
gcc 4.8.2
.
external-sourcery-toolchain
-
this recipe is now maintained in the
meta-sourcery
layer.
linux-libc-headers-yocto 3.4+git
-
Now using version 3.10 of the
linux-libc-headers
by default.
meta-toolchain-gmae
-
This recipe is obsolete.
packagegroup-core-sdk-gmae
-
This recipe is obsolete.
packagegroup-core-standalone-gmae-sdk-target
-
This recipe is obsolete.
The following classes have become obsolete and have been removed:
module_strip
pkg_metainfo
pkg_distribute
image-empty
The following reference BSPs changes occurred:
The BeagleBoard
(beagleboard
) ARM reference hardware
has been replaced by the BeagleBone
(beaglebone
) hardware.
The RouterStation Pro
(routerstationpro
) MIPS reference
hardware has been replaced by the EdgeRouter Lite
(edgerouter
) hardware.
The previous reference BSPs for the
beagleboard
and
routerstationpro
machines are still available
in a new meta-yocto-bsp-old
layer in the
Source Repositories
at
http://git.yoctoproject.org/cgit/cgit.cgi/meta-yocto-bsp-old/.
This section provides migration information for moving to the Yocto Project 1.7 Release from the prior release.
PACKAGECONFIG
Options in local.conf
¶
The QEMU recipe now uses a number of
PACKAGECONFIG
options to enable various optional features.
The method used to set defaults for these options means that
existing
local.conf
files will need to be be
modified to append to PACKAGECONFIG
for
qemu-native
and
nativesdk-qemu
instead of setting it.
In other words, to enable graphical output for QEMU, you should
now have these lines in local.conf
:
PACKAGECONFIG_append_pn-qemu-native = " sdl" PACKAGECONFIG_append_pn-nativesdk-qemu = " sdl"
The minimum
Git version
required on the build host is now 1.7.8 because the
--list
option is now required by
BitBake's Git fetcher.
As always, if your host distribution does not provide a version of
Git that meets this requirement, you can use the
buildtools-tarball
that does.
See the
"Required Git, tar, Python and gcc Versions"
section for more information.
The following
autotools
class changes occurred:
A separate build directory is now used by default:
The autotools
class has been changed
to use a directory for building
(B
),
which is separate from the source directory
(S
).
This is commonly referred to as
B != S
, or an out-of-tree build.
If the software being built is already capable of
building in a directory separate from the source, you
do not need to do anything.
However, if the software is not capable of being built
in this manner, you will
need to either patch the software so that it can build
separately, or you will need to change the recipe to
inherit the
autotools-brokensep
class instead of the autotools
or
autotools_stage
classes.
The --foreign
option is
no longer passed to automake
when
running autoconf
:
This option tells automake
that a
particular software package does not follow the GNU
standards and therefore should not be expected
to distribute certain files such as
ChangeLog
,
AUTHORS
, and so forth.
Because the majority of upstream software packages already
tell automake
to enable foreign mode
themselves, the option is mostly superfluous.
However, some recipes will need patches for this change.
You can easily make the change by patching
configure.ac
so that it passes
"foreign" to AM_INIT_AUTOMAKE()
.
See
this commit
for an example showing how to make the patch.
Some of the core recipes that package binary configuration scripts
now disable the scripts due to the
scripts previously requiring error-prone path substitution.
Software that links against these libraries using these scripts
should use the much more robust pkg-config
instead.
The list of recipes changed in this version (and their
configuration scripts) is as follows:
directfb (directfb-config) freetype (freetype-config) gpgme (gpgme-config) libassuan (libassuan-config) libcroco (croco-6.0-config) libgcrypt (libgcrypt-config) libgpg-error (gpg-error-config) libksba (ksba-config) libpcap (pcap-config) libpcre (pcre-config) libpng (libpng-config, libpng16-config) libsdl (sdl-config) libusb-compat (libusb-config) libxml2 (xml2-config) libxslt (xslt-config) ncurses (ncurses-config) neon (neon-config) npth (npth-config) pth (pth-config) taglib (taglib-config)
Additionally, support for pkg-config
has been
added to some recipes in the previous list in the rare cases
where the upstream software package does not already provide
it.
eglibc 2.19
Replaced with glibc 2.20
¶
Because eglibc
and
glibc
were already fairly close, this
replacement should not require any significant changes to other
software that links to eglibc
.
However, there were a number of minor changes in
glibc 2.20
upstream that could require
patching some software (e.g. the removal of the
_BSD_SOURCE
feature test macro).
glibc 2.20
requires version 2.6.32 or greater
of the Linux kernel.
Thus, older kernels will no longer be usable in conjunction with it.
For full details on the changes in glibc 2.20
,
see the upstream release notes
here.
The
module_autoload_*
variable is now deprecated and a new
KERNEL_MODULE_AUTOLOAD
variable should be used instead.
Also,
module_conf_*
must now be used in conjunction with a new
KERNEL_MODULE_PROBECONF
variable.
The new variables no longer require you to specify the module name
as part of the variable name.
This change not only simplifies usage but also allows the values
of these variables to be appropriately incorporated into task
signatures and thus trigger the appropriate tasks to re-execute
when changed.
You should replace any references to
module_autoload_*
with
KERNEL_MODULE_AUTOLOAD
, and add any modules
for which module_conf_*
is specified to
KERNEL_MODULE_PROBECONF
.
The following changes have occurred to the QA check process:
Additional QA checks file-rdeps
and build-deps
have been added in
order to verify that file dependencies are satisfied
(e.g. package contains a script requiring
/bin/bash
) and build-time dependencies
are declared, respectively.
For more information, please see the
"QA Error and Warning Messages"
chapter.
Package QA checks are now performed during a new
do_package_qa
task rather than being part of the
do_package
task.
This allows more parallel execution.
This change is unlikely to be an issue except for highly
customized recipes that disable packaging tasks themselves
by marking them as noexec
.
For those packages, you will need to disable the
do_package_qa
task as well.
Files being overwritten during the
do_populate_sysroot
task now trigger an error instead of a warning.
Recipes should not be overwriting files written to the
sysroot by other recipes.
If you have these types of recipes, you need to alter them
so that they do not overwrite these files.
You might now receive this error after changes in
configuration or metadata resulting in orphaned files
being left in the sysroot.
If you do receive this error, the way to resolve the issue
is to delete your
TMPDIR
or to move it out of the way and then re-start the build.
Anything that has been fully built up to that point and
does not need rebuilding will be restored from the shared
state cache and the rest of the build will be able to
proceed as normal.
The following recipes have been removed:
x-load
:
This recipe has been superseded by
U-boot SPL for all Cortex-based TI SoCs.
For legacy boards, the meta-ti
layer, which contains a maintained recipe, should be used
instead.
ubootchart
:
This recipe is obsolete.
A bootchart2
recipe has been added
to functionally replace it.
linux-yocto 3.4
:
Support for the linux-yocto 3.4 kernel has been dropped.
Support for the 3.10 and 3.14 kernels remains, while
support for version 3.17 has been added.
eglibc
has been removed in favor of
glibc
.
See the
"eglibc 2.19
Replaced with glibc 2.20
"
section for more information.
The following miscellaneous change occurred:
The build history feature now writes
build-id.txt
instead of
build-id
.
Additionally, build-id.txt
now contains the full build header as printed by
BitBake upon starting the build.
You should manually remove old "build-id" files from your
existing build history repositories to avoid confusion.
For information on the build history feature, see the
"Maintaining Build Output Quality"
section in the Yocto Project Development Tasks Manual.
This section provides migration information for moving to the Yocto Project 1.8 Release from the prior release.
The following recipes have been removed:
owl-video
:
Functionality replaced by gst-player
.
gaku
:
Functionality replaced by gst-player
.
gnome-desktop
:
This recipe is now available in
meta-gnome
and is no longer needed.
gsettings-desktop-schemas
:
This recipe is now available in
meta-gnome
and is no longer needed.
python-argparse
:
The argparse
module is already
provided in the default Python distribution in a
package named python-argparse
.
Consequently, the separate
python-argparse
recipe is no
longer needed.
telepathy-python, libtelepathy, telepathy-glib, telepathy-idle, telepathy-mission-control
:
All these recipes have moved to
meta-oe
and are consequently no
longer needed by any recipes in OpenEmbedded-Core.
linux-yocto_3.10
and linux-yocto_3.17
:
Support for the linux-yocto 3.10 and 3.17 kernels has been
dropped.
Support for the 3.14 kernel remains, while support for
3.19 kernel has been added.
poky-feed-config-opkg
:
This recipe has become obsolete and is no longer needed.
Use distro-feed-config
from
meta-oe
instead.
libav 0.8.x
:
libav 9.x
is now used.
sed-native
:
No longer needed.
A working version of sed
is expected
to be provided by the host distribution.
Proper built-in support for selecting BlueZ 5.x in preference
to the default of 4.x now exists.
To use BlueZ 5.x, simply add "bluez5" to your
DISTRO_FEATURES
value.
If you had previously added append files
(*.bbappend
) to make this selection, you can
now remove them.
Additionally, a bluetooth
class has been added
to make selection of the appropriate bluetooth support within a
recipe a little easier.
If you wish to make use of this class in a recipe, add something
such as the following:
inherit bluetooth PACKAGECONFIG ??= "${@bb.utils.contains('DISTRO_FEATURES', 'bluetooth', '${BLUEZ}', '', d)}" PACKAGECONFIG[bluez4] = "--enable-bluetooth,--disable-bluetooth,bluez4" PACKAGECONFIG[bluez5] = "--enable-bluez5,--disable-bluez5,bluez5"
The kernel build process was changed to place the source
in a common shared work area and to place build artifacts
separately in the source code tree.
In theory, migration paths have been provided for most common
usages in kernel recipes but this might not work in all cases.
In particular, users need to ensure that
${S}
(source files) and
${B}
(build artifacts) are used
correctly in functions such as
do_configure
and
do_install
.
For kernel recipes that do not inherit from
kernel-yocto
or include
linux-yocto.inc
, you might wish to
refer to the linux.inc
file in the
meta-oe
layer for the kinds of changes you
need to make.
For reference, here is the
commit
where the linux.inc
file in
meta-oe
was updated.
Recipes that rely on the kernel source code and do not inherit
the module classes might need to add explicit dependencies on
the do_shared_workdir
kernel task, for example:
do_configure[depends] += "virtual/kernel:do_shared_workdir"
SSL 3.0 is now disabled when building OpenSSL.
Disabling SSL 3.0 avoids any lingering instances of the POODLE
vulnerability.
If you feel you must re-enable SSL 3.0, then you can add an
append file (*.bbappend
) for the
openssl
recipe to remove "-no-ssl3"
from
EXTRA_OECONF
.
gcc's
default sysroot and include directories
are now "poisoned".
In other words, the sysroot and include directories are being
redirected to a non-existent location in order to catch when
host directories are being used due to the correct options not
being passed.
This poisoning applies both to the cross-compiler used within the
build and to the cross-compiler produced in the SDK.
If this change causes something in the build to fail, it almost certainly means the various compiler flags and commands are not being passed correctly to the underlying piece of software. In such cases, you need to take corrective steps.
Changes have been made to the
base
,
autotools
,
and
cmake
classes to clean out generated files when the
do_configure
task needs to be re-executed.
One of the improvements is to attempt to run "make clean" during
the do_configure
task if a
Makefile
exists.
Some software packages do not provide a working clean target
within their make files.
If you have such recipes, you need to set
CLEANBROKEN
to "1" within the recipe, for example:
CLEANBROKEN = "1"
The following QA Check and Validation Changes have occurred:
Usage of PRINC
previously triggered a warning.
It now triggers an error.
You should remove any remaining usage of
PRINC
in any recipe or append file.
An additional QA check has been added to detect usage of
${D}
in
FILES
values where
D
values
should not be used at all.
The same check ensures that $D
is used
in
pkg_preinst/pkg_postinst/pkg_prerm/pkg_postrm
functions instead of ${D}
.
S
now
needs to be set to a valid value within a recipe.
If S
is not set in the recipe, the
directory is not automatically created.
If S
does not point to a directory
that exists at the time the
do_unpack
task finishes, a warning will be shown.
LICENSE
is now validated for correct formatting of multiple
licenses.
If the format is invalid (e.g. multiple licenses are
specified with no operators to specify how the multiple
licenses interact), then a warning will be shown.
The following miscellaneous changes have occurred:
The send-error-report
script now
expects a "-s" option to be specified before the server
address.
This assumes a server address is being specified.
The oe-pkgdata-util
script now
expects a "-p" option to be specified before the
pkgdata
directory, which is now
optional.
If the pkgdata
directory is not
specified, the script will run BitBake to query
PKGDATA_DIR
from the build environment.
This section provides migration information for moving to the Yocto Project 2.0 Release from the prior release.
The default compiler is now GCC 5.2. This change has required fixes for compilation errors in a number of other recipes.
One important example is a fix for when the Linux kernel freezes at
boot time on ARM when built with GCC 5.
If you are using your own kernel recipe or source tree and
building for ARM, you will likely need to apply this
patch.
The standard linux-yocto
kernel source tree
already has a workaround for the same issue.
For further details, see https://gcc.gnu.org/gcc-5/changes.html and the porting guide at https://gcc.gnu.org/gcc-5/porting_to.html.
Alternatively, you can switch back to GCC 4.9 or 4.8 by
setting GCCVERSION
in your configuration,
as follows:
GCCVERSION = "4.9%"
Gstreamer 0.10 has been removed in favor of Gstreamer 1.x.
As part of the change, recipes for Gstreamer 0.10 and related
software are now located
in meta-multimedia
.
This change results in Qt4 having Phonon and Gstreamer
support in QtWebkit disabled by default.
The following recipes have been moved or removed:
bluez4
: The recipe is obsolete and
has been moved due to bluez5
becoming fully integrated.
The bluez4
recipe now resides in
meta-oe
.
gamin
: The recipe is obsolete and
has been removed.
gnome-icon-theme
: The recipe's
functionally has been replaced by
adwaita-icon-theme
.
Gstreamer 0.10 Recipes: Recipes for Gstreamer 0.10 have been removed in favor of the recipes for Gstreamer 1.x.
insserv
: The recipe is obsolete and
has been removed.
libunique
: The recipe is no longer
used and has been moved to meta-oe
.
midori
: The recipe's functionally
has been replaced by epiphany
.
python-gst
: The recipe is obsolete
and has been removed since it only contains bindings for
Gstreamer 0.10.
qt-mobility
: The recipe is obsolete and
has been removed since it requires
Gstreamer 0.10
, which has been
replaced.
subversion
: All 1.6.x versions of this
recipe have been removed.
webkit-gtk
: The older 1.8.3 version
of this recipe has been removed in favor of
webkitgtk
.
The method by which BitBake's datastore handles overrides has
changed.
Overrides are now applied dynamically and
bb.data.update_data()
is now a no-op.
Thus, bb.data.update_data()
is no longer
required in order to apply the correct overrides.
In practice, this change is unlikely to require any changes to
Metadata.
However, these minor changes in behavior exist:
All potential overrides are now visible in the variable history as seen when you run the following:
$ bitbake -e
d.delVar('
VARNAME
')
and
d.setVar('
VARNAME
', None)
result in the variable and all of its overrides being
cleared out.
Before the change, only the non-overridden values
were cleared.
The shell versions of the BitBake message functions (i.e.
bbdebug
, bbnote
,
bbwarn
, bbplain
,
bberror
, and bbfatal
)
are now connected through to their BitBake equivalents
bb.debug()
, bb.note()
,
bb.warn()
, bb.plain()
,
bb.error()
, and
bb.fatal()
, respectively.
Thus, those message functions that you would expect to be printed
by the BitBake UI are now actually printed.
In practice, this change means two things:
If you now see messages on the console that you did not
previously see as a result of this change, you might
need to clean up the calls to
bbwarn
, bberror
,
and so forth.
Or, you might want to simply remove the calls.
The bbfatal
message function now
suppresses the full error log in the UI, which means any
calls to bbfatal
where you still
wish to see the full error log should be replaced by
die
or
bbfatal_log
.
The following recipes have had extra
dev/dbg
packages removed:
acl
apmd
aspell
attr
augeas
bzip2
cogl
curl
elfutils
gcc-target
libgcc
libtool
libxmu
opkg
pciutils
rpm
sysfsutils
tiff
xz
All of the above recipes now conform to the standard packaging
scheme where a single -dev
,
-dbg
, and -staticdev
package exists per recipe.
Maintenance tracking data for recipes that was previously part
of meta-yocto
has been moved to
OE-Core.
The change includes package_regex.inc
and
distro_alias.inc
, which are typically enabled
when using the distrodata
class.
Additionally, the contents of
upstream_tracking.inc
has now been split out
to the relevant recipes.
Stale files from recipes that no longer exist in the current
configuration are now automatically removed from
sysroot as well as removed from
any other place managed by shared state.
This automatic cleanup means that the build system now properly
handles situations such as renaming the build system side of
recipes, removal of layers from
bblayers.conf
, and
DISTRO_FEATURES
changes.
Additionally, work directories for old versions of recipes are now pruned. If you wish to disable pruning old work directories, you can set the following variable in your configuration:
SSTATE_PRUNE_OBSOLETEWORKDIR = "0"
linux-yocto
Kernel Metadata Repository Now Split from Source¶
The linux-yocto
tree has up to now been a
combined set of kernel changes and configuration (meta) data
carried in a single tree.
While this format is effective at keeping kernel configuration and
source modifications synchronized, it is not always obvious to
developers how to manipulate the Metadata as compared to the
source.
Metadata processing has now been removed from the
kernel-yocto
class and the external Metadata repository
yocto-kernel-cache
, which has always been used
to seed the linux-yocto
"meta" branch.
This separate linux-yocto
cache repository
is now the primary location for this data.
Due to this change, linux-yocto
is no longer
able to process combined trees.
Thus, if you need to have your own combined kernel repository,
you must do the split there as well and update your recipes
accordingly.
See the meta/recipes-kernel/linux/linux-yocto_4.1.bb
recipe for an example.
The following QA checks have been added:
Added a "host-user-contaminated" check for ownership
issues for packaged files outside of
/home
.
The check looks for files that are incorrectly owned by the
user that ran BitBake instead of owned by a valid user in
the target system.
Added an "invalid-chars" check for invalid (non-UTF8)
characters in recipe metadata variable values
(i.e.
DESCRIPTION
,
SUMMARY
,
LICENSE
,
and
SECTION
).
Some package managers do not support these characters.
Added an "invalid-packageconfig" check for any options
specified in
PACKAGECONFIG
that do not match any PACKAGECONFIG
option defined for the recipe.
These additional changes exist:
gtk-update-icon-cache
has been
renamed to gtk-icon-utils
.
The tools-profile
IMAGE_FEATURES
item as well as its corresponding packagegroup and
packagegroup-core-tools-profile
no
longer bring in oprofile
.
Bringing in oprofile
was originally
added to aid compilation on resource-constrained
targets.
However, this aid has not been widely used and is not
likely to be used going forward due to the more powerful
target platforms and the existence of better
cross-compilation tools.
The
IMAGE_FSTYPES
variable's default value now specifies
ext4
instead of
ext3
.
All support for the PRINC
variable has been removed.
The packagegroup-core-full-cmdline
packagegroup no longer brings in
lighttpd
due to the fact that
bringing in lighttpd
is not really in
line with the packagegroup's purpose, which is to add full
versions of command-line tools that by default are
provided by busybox
.
This section provides migration information for moving to the Yocto Project 2.1 Release from the prior release.
Variable expressions, such as
${
VARNAME
}
no longer expand automatically within Python functions.
Suppressing expansion was done to allow Python functions to
construct shell scripts or other code for situations in which you
do not want such expressions expanded.
For any existing code that relies on these expansions, you need to
change the expansions to expand the value of individual
variables through d.getVar()
.
To alternatively expand more complex expressions,
use d.expand()
.
The convention for overrides has always been for them to be
lower-case characters.
This practice is now a requirement as BitBake's datastore now
assumes lower-case characters in order to give a slight performance
boost during parsing.
In practical terms, this requirement means that anything that ends
up in
OVERRIDES
must now appear in lower-case characters (e.g. values for
MACHINE
, TARGET_ARCH
,
DISTRO
, and also recipe names if
_pn-
recipename
overrides are to be effective).
getVar()
and
getVarFlag()
is Now Mandatory¶
The expand parameter to getVar()
and
getVarFlag()
previously defaulted to
False if not specified.
Now, however, no default exists so one must be specified.
You must change any getVar()
calls that
do not specify the final expand parameter to calls that do specify
the parameter.
You can run the following sed
command at the
base of a layer to make this change:
sed -e 's:\(\.getVar([^,()]*\)):\1, False):g' -i `grep -ril getVar *` sed -e 's:\(\.getVarFlag([^,()]*, [^,()]*\)):\1, False):g' -i `grep -ril getVarFlag *`
EXTRA_OEMAKE
now defaults to "" instead of "-e MAKEFLAGS=".
Setting EXTRA_OEMAKE
to "-e MAKEFLAGS=" by
default was a historical accident that has required many classes
(e.g. autotools
, module
)
and recipes to override this default in order to work with
sensible build systems.
When upgrading to the release, you must edit any recipe that
relies upon this old default by either setting
EXTRA_OEMAKE
back to "-e MAKEFLAGS=" or by
explicitly setting any required variable value overrides using
EXTRA_OEMAKE
, which is typically only needed
when a Makefile sets a default value for a variable that is
inappropriate for cross-compilation using the "=" operator rather
than the "?=" operator.
libexecdir
Reverted to ${prefix}/libexec
¶
The use of ${libdir}/${BPN}
as
libexecdir
is different as compared to all
other mainstream distributions, which either uses
${prefix}/libexec
or
${libdir}
.
The use is also contrary to the GNU Coding Standards
(i.e. https://www.gnu.org/prep/standards/html_node/Directory-Variables.html)
that suggest ${prefix}/libexec
and also
notes that any package-specific nesting should be done by the
package itself.
Finally, having libexecdir
change between
recipes makes it very difficult for different recipes to invoke
binaries that have been installed into
libexecdir
.
The Filesystem Hierarchy Standard
(i.e. http://refspecs.linuxfoundation.org/FHS_3.0/fhs/ch04s07.html)
now recognizes the use of ${prefix}/libexec/
,
giving distributions the choice between
${prefix}/lib
or
${prefix}/libexec
without breaking FHS.
ac_cv_sizeof_off_t
is No Longer Cached in Site Files¶
For recipes inheriting the
autotools
class, ac_cv_sizeof_off_t
is no longer cached
in the site files for autoconf
.
The reason for this change is because the
ac_cv_sizeof_off_t
value is not necessarily
static per architecture as was previously assumed.
Rather, the value changes based on whether large file support is
enabled.
For most software that uses autoconf
, this
change should not be a problem.
However, if you have a recipe that bypasses the standard
do_configure
task from the autotools
class and the software
the recipe is building uses a very old version of
autoconf
, the recipe might be incapable of
determining the correct size of off_t
during
do_configure
.
The best course of action is to patch the software as necessary
to allow the default implementation from the
autotools
class to work such that
autoreconf
succeeds and produces a working
configure script, and to remove the
overridden do_configure
task such that the
default implementation does get used.
Previously, for image recipes the
do_rootfs
task assembled the filesystem and then from that filesystem
generated images.
With this Yocto Project release, image generation is split into
separate
do_image_*
tasks for clarity both in operation and in the code.
For most cases, this change does not present any problems.
However, if you have made customizations that directly modify the
do_rootfs
task or that mention
do_rootfs
, you might need to update those
changes.
In particular, if you had added any tasks after
do_rootfs
, you should make edits so that
those tasks are after the
do_image_complete
task rather than after do_rootfs
so that the your added tasks
run at the correct time.
A minor part of this restructuring is that the post-processing
definitions and functions have been moved from the
image
class to the
rootfs-postcommands
class.
Functionally, however, they remain unchanged.
The following recipes have been removed in the 2.1 release:
gcc
version 4.8:
Versions 4.9 and 5.3 remain.
qt4
:
All support for Qt 4.x has been moved out to a separate
meta-qt4
layer because Qt 4 is no
longer supported upstream.
x11vnc
:
Moved to the meta-oe
layer.
linux-yocto-3.14
:
No longer supported.
linux-yocto-3.19
:
No longer supported.
libjpeg
:
Replaced by the libjpeg-turbo
recipe.
pth
:
Became obsolete.
liboil
:
Recipe is no longer needed and has been moved to the
meta-multimedia
layer.
gtk-theme-torturer
:
Recipe is no longer needed and has been moved to the
meta-gnome
layer.
gnome-mime-data
:
Recipe is no longer needed and has been moved to the
meta-gnome
layer.
udev
:
Replaced by the eudev
recipe for
compatibility when using sysvinit
with newer kernels.
python-pygtk
:
Recipe became obsolete.
adt-installer
:
Recipe became obsolete.
See the
"ADT Removed"
section for more information.
The following classes have changed:
autotools_stage
:
Removed because the
autotools
class now provides its functionality.
Recipes that inherited from
autotools_stage
should now inherit
from autotools
instead.
boot-directdisk
:
Merged into the image-vm
class.
The boot-directdisk
class was rarely
directly used.
Consequently, this change should not cause any issues.
bootimg
:
Merged into the
image-live
class.
The bootimg
class was rarely
directly used.
Consequently, this change should not cause any issues.
packageinfo
:
Removed due to its limited use by the Hob UI, which has
itself been removed.
The following changes have been made to the build system user interface:
Hob GTK+-based UI: Removed because it is unmaintained and based on the outdated GTK+ 2 library. The Toaster web-based UI is much more capable and is actively maintained. See the "Using the Toaster Web Interface" section in the Toaster User Manual for more information on this interface.
"puccho" BitBake UI: Removed because is unmaintained and no longer useful.
The Application Development Toolkit (ADT) has been removed because its functionality almost completely overlapped with the standard SDK and the extensible SDK. For information on these SDKs and how to build and use them, see the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
The following changes have been made for the Poky distribution:
The meta-yocto
layer has been renamed
to meta-poky
to better match its
purpose, which is to provide the Poky reference
distribution.
The meta-yocto-bsp
layer retains its
original name since it provides reference machines for
the Yocto Project and it is otherwise unrelated to Poky.
References to meta-yocto
in your
conf/bblayers.conf
should
automatically be updated, so you should not need to change
anything unless you are relying on this naming elsewhere.
The
uninative
class is now enabled by default in Poky.
This class attempts to isolate the build system from the
host distribution's C library and makes re-use of native
shared state artifacts across different host distributions
practical.
With this class enabled, a tarball containing a pre-built
C library is downloaded at the start of the build.
The uninative
class is enabled
through the
meta/conf/distro/include/yocto-uninative.inc
file, which for those not using the Poky distribution, can
include to easily enable the same functionality.
Alternatively, if you wish to build your own
uninative
tarball, you can do so by
building the uninative-tarball
recipe,
making it available to your build machines
(e.g. over HTTP/HTTPS) and setting a similar configuration
as the one set by yocto-uninative.inc
.
Static library generation, for most cases, is now disabled by default in the Poky distribution. Disabling this generation saves some build time as well as the size used for build output artifacts.
Disabling this library generation is accomplished
through a
meta/conf/distro/include/no-static-libs.inc
,
which for those not using the Poky distribution can
easily include to enable the same functionality.
Any recipe that needs to opt-out of having the "--disable-static" option specified on the configure command line either because it is not a supported option for the configure script or because static libraries are needed should set the following variable:
DISABLE_STATIC = ""
The separate poky-tiny
distribution
now uses the musl C library instead of a heavily pared
down glibc
.
Using musl results in a smaller
distribution and facilitates much greater maintainability
because musl is designed to have a small footprint.
If you have used poky-tiny
and
have customized the glibc
configuration you will need to redo those customizations
with musl when upgrading to the new release.
The following changes have been made to packaging:
The runuser
and
mountpoint
binaries, which were
previously in the main util-linux
package, have been split out into the
util-linux-runuser
and
util-linux-mountpoint
packages,
respectively.
The python-elementtree
package has
been merged into the python-xml
package.
The following changes have been made to the tuning files:
The "no-thumb-interwork" tuning feature has been dropped from the ARM tune include files. Because interworking is required for ARM EABI, attempting to disable it through a tuning feature no longer makes sense.
The tune-cortexm*.inc
and
tune-cortexr4.inc
files have been
removed because they are poorly tested.
Until the OpenEmbedded build system officially gains
support for CPUs without an MMU, these tuning files would
probably be better maintained in a separate layer
if needed.
This release supports generation of GLib Introspective Repository (GIR) files through GObject introspection, which is the standard mechanism for accessing GObject-based software from runtime environments. You can enable, disable, and test the generation of this data. See the "Enabling GObject Introspection Support" section in the Yocto Project Development Tasks Manual for more information.
These additional changes exist:
The minimum Git version has been increased to 1.8.3.1. If your host distribution does not provide a sufficiently recent version, you can install the buildtools, which will provide it. See the "Required Git, tar, Python and gcc Versions" section for more information on the buildtools tarball.
The buggy and incomplete support for the RPM version 4 package manager has been removed. The well-tested and maintained support for RPM version 5 remains.
Previously, the following list of packages were removed
if package-management was not in
IMAGE_FEATURES
,
regardless of any dependencies:
update-rc.d base-passwd shadow update-alternatives run-postinsts
With the Yocto Project 2.1 release, these packages are only
removed if "read-only-rootfs" is in
IMAGE_FEATURES
, since they might
still be needed for a read-write image even in the absence
of a package manager (e.g. if users need to be added,
modified, or removed at runtime).
The
devtool modify
command now defaults to extracting the source since that
is most commonly expected.
The "-x" or "--extract" options are now no-ops.
If you wish to provide your own existing source tree, you
will now need to specify either the "-n" or
"--no-extract" options when running
devtool modify
.
If the formfactor for a machine is either not supplied or does not specify whether a keyboard is attached, then the default is to assume a keyboard is attached rather than assume no keyboard. This change primarily affects the Sato UI.
The .debug
directory packaging is
now automatic.
If your recipe builds software that installs binaries into
directories other than the standard ones, you no longer
need to take care of setting
FILES_${PN}-dbg
to pick up the
resulting .debug
directories as these
directories are automatically found and added.
Inaccurate disk and CPU percentage data has been dropped
from buildstats
output.
This data has been replaced with
getrusage()
data and corrected IO
statistics.
You will probably need to update any custom code that reads
the buildstats
data.
The
meta/conf/distro/include/package_regex.inc
is now deprecated.
The contents of this file have been moved to individual
recipes.
The v86d/uvesafb
has been removed from
the genericx86
and
genericx86-64
reference machines,
which are provided by the
meta-yocto-bsp
layer.
Most modern x86 boards do not rely on this file and it only
adds kernel error messages during startup.
If you do still need to support
uvesafb
, you can
simply add v86d
to your image.
Build sysroot paths are now removed from debug symbol
files.
Removing these paths means that remote GDB using an
unstripped build system sysroot will no longer work
(although this was never documented to work).
The supported method to accomplish something similar is
to set IMAGE_GEN_DEBUGFS
to "1",
which will generate a companion debug image
containing unstripped binaries and associated debug
sources alongside the image.
This section provides migration information for moving to the Yocto Project 2.2 Release from the prior release.
The minimum kernel version for the target system and for SDK
is now 3.2.0, due to the upgrade
to glibc 2.24
.
Specifically, for AArch64-based targets the version is
3.14.
For Nios II-based targets, the minimum kernel version is 3.19.
OLDEST_KERNEL
to anything down to 2.6.32 if desired.
The way directories are staged in sysroot has been simplified and
introduces the new
SYSROOT_DIRS
,
SYSROOT_DIRS_NATIVE
,
and
SYSROOT_DIRS_BLACKLIST
.
See the
v2 patch series on the OE-Core Mailing List
for additional information.
tmp/deploy
Now Enabled¶
Removal of old images and other files in
tmp/deploy/
is now enabled by default due
to a new staging method used for those files.
As a result of this change, the
RM_OLD_IMAGE
variable is now redundant.
The following changes for Python occurred:
BitBake requires Python 3.4 or greater.
A UTF-8 locale is required on the build host due to Python 3. Since C.UTF-8 is not a standard, the default is en_US.UTF-8.
The metadata is now required to use Python 3 syntax. For help preparing metadata, see any of the many Python 3 porting guides available. Alternatively, you can reference the conversion commits for Bitbake and you can use OE-Core as a guide for changes. Following are particular areas of interest:
* subprocess command-line pipes needing locale decoding
* the syntax for octal values changed
* the iter*()
functions changed name
* iterators now return views, not lists
* changed names for Python modules
Most target Python recipes have now been switched to Python 3. Unfortunately, systems using RPM as a package manager and providing online package-manager support through SMART still require Python 2.
buildtools-tarball
Includes Python 3¶
buildtools-tarball
now includes Python 3.
uClibc has been removed in favor of musl. Musl has matured, is better maintained, and is compatible with a wider range of applications as compared to uClibc.
${B}
No Longer Default Working Directory for Tasks¶
${
B
}
is no longer the default working directory for tasks.
Consequently, any custom tasks you define now need to either
have the
[
dirs
]
flag set, or the task needs to change into the
appropriate working directory manually (e.g using
cd
for a shell task).
[dirs]
flag.
runqemu
Ported to Python¶
runqemu
has been ported to Python and has
changed behavior in some cases.
Previous usage patterns continue to be supported.
The new runqemu
is a Python script.
Machine knowledge is no longer hardcoded into
runqemu
.
You can choose to use the qemuboot
configuration file to define the BSP's own arguments and to make
it bootable with runqemu
.
If you use a configuration file, use the following form:
image-name
-machine
.qemuboot.conf
The configuration file enables fine-grained tuning of options
passed to QEMU without the runqemu
script
hard-coding any knowledge about different machines.
Using a configuration file is particularly convenient when trying
to use QEMU with machines other than the
qemu*
machines in
OE-Core.
The qemuboot.conf
file is generated by the
qemuboot
class when the root filesystem is being build (i.e.
build rootfs).
QEMU boot arguments can be set in BSP's configuration file and
the qemuboot
class will save them to
qemuboot.conf
.
If you want to use runqemu
without a
configuration file, use the following command form:
$ runqemumachine
rootfs
kernel
[options
]
Supported machines
are as follows:
qemuarm qemuarm64 qemux86 qemux86-64 qemuppc qemumips qemumips64 qemumipsel qemumips64el
Consider the following example, which uses the
qemux86-64
machine,
provides a root filesystem, provides an image, and uses
the nographic
option:
$ runqemu qemux86-64 tmp/deploy/images/qemux86-64/core-image-minimal-qemux86-64.ext4 tmp/deploy/images/qemux86-64/bzImage nographic
Following is a list of variables that can be set in configuration
files such as bsp.conf
to enable the BSP
to be booted by runqemu
:
QB_SYSTEM_NAME: QEMU name (e.g. "qemu-system-i386") QB_OPT_APPEND: Options to append to QEMU (e.g. "-show-cursor") QB_DEFAULT_KERNEL: Default kernel to boot (e.g. "bzImage") QB_DEFAULT_FSTYPE: Default FSTYPE to boot (e.g. "ext4") QB_MEM: Memory (e.g. "-m 512") QB_MACHINE: QEMU machine (e.g. "-machine virt") QB_CPU: QEMU cpu (e.g. "-cpu qemu32") QB_CPU_KVM: Similar to QB_CPU except used for kvm support (e.g. "-cpu kvm64") QB_KERNEL_CMDLINE_APPEND: Options to append to the kernel's -append option (e.g. "console=ttyS0 console=tty") QB_DTB: QEMU dtb name QB_AUDIO_DRV: QEMU audio driver (e.g. "alsa", set it when support audio) QB_AUDIO_OPT: QEMU audio option (e.g. "-soundhw ac97,es1370"), which is used when QB_AUDIO_DRV is set. QB_KERNEL_ROOT: Kernel's root (e.g. /dev/vda) QB_TAP_OPT: Network option for 'tap' mode (e.g. "-netdev tap,id=net0,ifname=@TAP@,script=no,downscript=no -device virtio-net-device,netdev=net0"). runqemu will replace "@TAP@" with the one that is used, such as tap0, tap1 ... QB_SLIRP_OPT: Network option for SLIRP mode (e.g. "-netdev user,id=net0 -device virtio-net-device,netdev=net0") QB_ROOTFS_OPT: Used as rootfs (e.g. "-drive id=disk0,file=@ROOTFS@,if=none,format=raw -device virtio-blk-device,drive=disk0"). runqemu will replace "@ROOTFS@" with the one which is used, such as core-image-minimal-qemuarm64.ext4. QB_SERIAL_OPT: Serial port (e.g. "-serial mon:stdio") QB_TCPSERIAL_OPT: tcp serial port option (e.g. " -device virtio-serial-device -chardev socket,id=virtcon,port=@PORT@,host=127.0.0.1 -device virtconsole,chardev=virtcon" runqemu will replace "@PORT@" with the port number which is used.
To use runqemu
, set
IMAGE_CLASSES
as follows and run runqemu
:
runqemu help
.
IMAGE_CLASSES += "qemuboot"
The default linker hash style for gcc-cross
is now "sysv" in order to catch recipes that are building software
without using the OpenEmbedded
LDFLAGS
.
This change could result in seeing some "No GNU_HASH in the elf
binary" QA issues when building such recipes.
You need to fix these recipes so that they use the expected
LDFLAGS
.
Depending on how the software is built, the build system used by
the software (e.g. a Makefile) might need to be patched.
However, sometimes making this fix is as simple as adding the
following to the recipe:
TARGET_CC_ARCH += "${LDFLAGS}"
KERNEL_IMAGE_BASE_NAME
no Longer Uses KERNEL_IMAGETYPE
¶
The
KERNEL_IMAGE_BASE_NAME
variable no longer uses the
KERNEL_IMAGETYPE
variable to create the image's base name.
Because the OpenEmbedded build system can now build multiple kernel
image types, this part of the kernel image base name as been
removed leaving only the following:
KERNEL_IMAGE_BASE_NAME ?= "${PKGE}-${PKGV}-${PKGR}-${MACHINE}-${DATETIME}"
If you have recipes or classes that use
KERNEL_IMAGE_BASE_NAME
directly, you might
need to update the references to ensure they continue to work.
The following changes took place for BitBake:
The "goggle" UI and standalone image-writer tool have been removed as they both require GTK+ 2.0 and were not being maintained.
The Perforce fetcher now supports
SRCREV
for specifying the source revision to use, be it
${
AUTOREV
}
,
changelist number, p4date, or label, in preference to
separate
SRC_URI
parameters to specify these.
This change is more in-line with how the other fetchers
work for source control systems.
Recipes that fetch from Perforce will need to be updated
to use SRCREV
in place of specifying
the source revision within
SRC_URI
.
Some of BitBake's internal code structures for accessing the recipe cache needed to be changed to support the new multi-configuration functionality. These changes will affect external tools that use BitBake's tinfoil module. For information on these changes, see the changes made to the scripts supplied with OpenEmbedded-Core: 1 and 2.
The task management code has been rewritten to avoid using
ID indirection in order to improve performance.
This change is unlikely to cause any problems for most
users.
However, the setscene verification function as pointed to
by BB_SETSCENE_VERIFY_FUNCTION
needed to change signature.
Consequently, a new variable named
BB_SETSCENE_VERIFY_FUNCTION2
has been added allowing multiple versions of BitBake
to work with suitably written metadata, which includes
OpenEmbedded-Core and Poky.
Anyone with custom BitBake task scheduler code might also
need to update the code to handle the new structure.
Swabber, a tool that was intended to detect host contamination in the build process, has been removed, as it has been unmaintained and unused for some time and was never particularly effective. The OpenEmbedded build system has since incorporated a number of mechanisms including enhanced QA checks that mean that there is less of a need for such a tool.
The following recipes have been removed:
augeas
:
No longer needed and has been moved to
meta-oe
.
directfb
:
Unmaintained and has been moved to
meta-oe
.
gcc
:
Removed 4.9 version.
Versions 5.4 and 6.2 are still present.
gnome-doc-utils
:
No longer needed.
gtk-doc-stub
:
Replaced by gtk-doc
.
gtk-engines
:
No longer needed and has been moved to
meta-gnome
.
gtk-sato-engine
:
Became obsolete.
libglade
:
No longer needed and has been moved to
meta-oe
.
libmad
:
Unmaintained and functionally replaced by
libmpg123
.
libmad
has been moved to
meta-oe
.
libowl
:
Became obsolete.
libxsettings-client
:
No longer needed.
oh-puzzles
:
Functionally replaced by
puzzles
.
oprofileui
:
Became obsolete.
OProfile has been largely supplanted by perf.
packagegroup-core-directfb.bb
:
Removed.
core-image-directfb.bb
:
Removed.
pointercal
:
No longer needed and has been moved to
meta-oe
.
python-imaging
:
No longer needed and moved to
meta-python
python-pyrex
:
No longer needed and moved to
meta-python
.
sato-icon-theme
:
Became obsolete.
swabber-native
:
Swabber has been removed.
See the
entry on Swabber.
tslib
:
No longer needed and has been moved to
meta-oe
.
uclibc
:
Removed in favor of musl.
xtscal
:
No longer needed and moved to
meta-oe
The following classes have been removed:
distutils-native-base
:
No longer needed.
distutils3-native-base
:
No longer needed.
sdl
:
Only set
DEPENDS
and
SECTION
,
which are better set within the recipe instead.
sip
:
Mostly unused.
swabber
:
See the
entry on Swabber.
The following minor packaging changes have occurred:
grub
:
Split grub-editenv
into its own
package.
systemd
:
Split container and vm related units into a new package,
systemd-container.
util-linux
:
Moved prlimit
to a separate
util-linux-prlimit
package.
The following miscellaneous changes have occurred:
package_regex.inc
:
Removed because the definitions
package_regex.inc
previously contained
have been moved to their respective recipes.
Both devtool add
and
recipetool create
now use a fixed
SRCREV
by default when fetching from a Git repository.
You can override this in either case to use
${
AUTOREV
}
instead by using the -a
or
‐‐autorev
command-line
option
distcc
:
GTK+ UI is now disabled by default.
packagegroup-core-tools-testapps
:
Removed Piglit.
image.bbclass
:
Renamed COMPRESS(ION) to CONVERSION.
This change means that
COMPRESSIONTYPES
,
COMPRESS_DEPENDS
and
COMPRESS_CMD
are deprecated in favor
of CONVERSIONTYPES
,
CONVERSION_DEPENDS
and
CONVERSION_CMD
.
The COMPRESS*
variable names will
still work in the 2.2 release but metadata that does not
need to be backwards-compatible should be changed to
use the new names as the COMPRESS*
ones will be removed in a future release.
gtk-doc
:
A full version of gtk-doc
is now
made available.
However, some old software might not be capable of using
the current version of gtk-doc
to build documentation.
You need to change recipes that build such software so that
they explicitly disable building documentation with
gtk-doc
.
This section provides migration information for moving to the Yocto Project 2.3 Release from the prior release.
The OpenEmbedded build system now uses one sysroot per recipe to resolve long-standing issues with configuration script auto-detection of undeclared dependencies. Consequently, you might find that some of your previously written custom recipes are missing declared dependencies, particularly those dependencies that are incidentally built earlier in a typical build process and thus are already likely to be present in the shared sysroot in previous releases.
Consider the following:
Declare Build-Time Dependencies: Because of this new feature, you must explicitly declare all build-time dependencies for your recipe. If you do not declare these dependencies, they are not populated into the sysroot for the recipe.
Specify Pre-Installation and Post-Installation
Native Tool Dependencies:
You must specifically specify any special native tool
dependencies of pkg_preinst
and
pkg_postinst
scripts by using the
PACKAGE_WRITE_DEPS
variable.
Specifying these dependencies ensures that these tools
are available if these scripts need to be run on the
build host during the
do_rootfs
task.
As an example, see the dbus
recipe.
You will see that this recipe has a
pkg_postinst
that calls
systemctl
if "systemd" is in
DISTRO_FEATURES
.
In the example,
systemd-systemctl-native
is added to
PACKAGE_WRITE_DEPS
, which is also
conditional on "systemd" being in
DISTRO_FEATURES
.
Examine Recipes that Use
SSTATEPOSTINSTFUNCS
:
You need to examine any recipe that uses
SSTATEPOSTINSTFUNCS
and determine
steps to take.
Functions added to
SSTATEPOSTINSTFUNCS
are still
called as they were in previous Yocto Project releases.
However, since a separate sysroot is now being populated
for every recipe and if existing functions being called
through SSTATEPOSTINSTFUNCS
are
doing relocation, then you will need to change these
to use a post-installation script that is installed by a
function added to
SYSROOT_PREPROCESS_FUNCS
.
For an example, see the
pixbufcache
class in
meta/classes/
in the Yocto Project
Source Repositories.
SSTATEPOSTINSTFUNCS
variable
itself is now deprecated in favor of the
do_populate_sysroot[postfuncs]
task.
Consequently, if you do still have any function or
functions that need to be called after the sysroot
component is created for a recipe, then you would be
well advised to take steps to use a post installation
script as described previously.
Taking these steps prepares your code for when
SSTATEPOSTINSTFUNCS
is
removed in a future Yocto Project release.
Specify the Sysroot when Using Certain
External Scripts:
Because the shared sysroot is now gone, the scripts
oe-find-native-sysroot
and
oe-run-native
have been changed such
that you need to specify which recipe's
STAGING_DIR_NATIVE
is used.
staging.bbclass
"
section.
PATH
Variable¶
Within the environment used to run build tasks, the environment
variable PATH
is now sanitized such that
the normal native binary paths
(/bin
, /sbin
,
/usr/bin
and so forth) are
removed and a directory containing symbolic links linking only
to the binaries from the host mentioned in the
HOSTTOOLS
and
HOSTTOOLS_NONFATAL
variables is added to PATH
.
Consequently, any native binaries provided by the host that you need to call needs to be in one of these two variables at the configuration level.
Alternatively, you can add a native recipe (i.e.
-native
) that provides the
binary to the recipe's
DEPENDS
value.
PATH
is not sanitized in the same way
within devshell
.
If it were, you would have difficulty running host tools for
development and debugging within the shell.
The following changes to scripts took place:
oe-find-native-sysroot
:
The usage for the
oe-find-native-sysroot
script has
changed to the following:
$ . oe-find-native-sysroot recipe
You must now supply a recipe for
recipe
as part of the command.
Prior to the Yocto Project 3.1.2 release, it was not
necessary to provide the script with the command.
oe-run-native
:
The usage for the
oe-run-native
script has changed
to the following:
$ oe-run-nativenative_recipe
tool
You must supply the name of the native recipe and the tool you want to run as part of the command. Prior to the Yocto Project 3.1.2 release, it was not necessary to provide the native recipe with the command.
cleanup-workdir
:
The cleanup-workdir
script has been
removed because the script was found to be deleting
files it should not have, which lead to broken build
trees.
Rather than trying to delete portions of
TMPDIR
and getting it wrong, it is recommended that you
delete TMPDIR
and have it restored
from shared state (sstate) on subsequent builds.
wipe-sysroot
:
The wipe-sysroot
script has been
removed as it is no longer needed with recipe-specific
sysroots.
The previously deprecated
bb.data.getVar()
,
bb.data.setVar()
, and
related functions have been removed in favor of
d.getVar()
,
d.setVar()
, and so forth.
You need to fix any references to these old functions.
The following changes took place for BitBake:
BitBake's Graphical Dependency Explorer UI Replaced:
BitBake's graphical dependency explorer UI
depexp
was replaced by
taskexp
("Task Explorer"), which
provides a graphical way of exploring the
task-depends.dot
file.
The data presented by Task Explorer is much more
accurate than the data that was presented by
depexp
.
Being able to visualize the data is an often requested
feature as standard *.dot
file
viewers cannot usual cope with the size of
the task-depends.dot
file.
BitBake "-g" Output Changes:
The package-depends.dot
and
pn-depends.dot
files as previously
generated using the bitbake -g
command
have been removed.
A recipe-depends.dot
file
is now generated as a collapsed version of
task-depends.dot
instead.
The reason for this change is because
package-depends.dot
and
pn-depends.dot
largely date back
to a time before task-based execution and do not take
into account task-level dependencies between recipes,
which could be misleading.
Mirror Variable Splitting Changes:
Mirror variables including
MIRRORS
,
PREMIRRORS
,
and
SSTATE_MIRRORS
can now separate values entirely with spaces.
Consequently, you no longer need "\\n".
BitBake looks for pairs of values, which simplifies usage.
There should be no change required to existing mirror
variable values themselves.
The Subversion (SVN) Fetcher Uses an "ssh" Parameter and Not an "rsh" Parameter:
The SVN fetcher now takes an "ssh" parameter instead of an
"rsh" parameter.
This new optional parameter is used when the "protocol"
parameter is set to "svn+ssh".
You can only use the new parameter to specify the
ssh
program used by SVN.
The SVN fetcher passes the new parameter through the
SVN_SSH
environment variable during
the
do_fetch
task.
See the "Subversion (SVN) Fetcher (svn://)" section in the BitBake User Manual for additional information.
BB_SETSCENE_VERIFY_FUNCTION
and BB_SETSCENE_VERIFY_FUNCTION2
Removed:
Because the mechanism they were part of is no longer
necessary with recipe-specific sysroots, the
BB_SETSCENE_VERIFY_FUNCTION
and
BB_SETSCENE_VERIFY_FUNCTION2
variables have been removed.
Absolute symbolic links (symlinks) within staged files are no
longer permitted and now trigger an error.
Any explicit creation of symlinks can use the
lnr
script, which is a replacement for
ln -r
.
If the build scripts in the software that the recipe is building
are creating a number of absolute symlinks that need to be
corrected, you can inherit
relative_symlinks
within the recipe to turn
those absolute symlinks into relative symlinks.
Older GPLv2 versions of GPLv3 recipes have moved to a
separate meta-gplv2
layer.
If you use
INCOMPATIBLE_LICENSE
to exclude GPLv3 or set
PREFERRED_VERSION
to substitute a GPLv2 version of a GPLv3 recipe, then you must add
the meta-gplv2
layer to your configuration.
meta-gplv2
layer in the
OpenEmbedded layer index at
https://layers.openembedded.org/layerindex/branch/master/layer/meta-gplv2/.
These relocated GPLv2 recipes do not receive the same level of maintenance as other core recipes. The recipes do not get security fixes and upstream no longer maintains them. In fact, the upstream community is actively hostile towards people that use the old versions of the recipes. Moving these recipes into a separate layer both makes the different needs of the recipes clearer and clearly identifies the number of these recipes.
The following package management changes took place:
Smart package manager is replaced by DNF package manager. Smart has become unmaintained upstream, is not ported to Python 3.x. Consequently, Smart needed to be replaced. DNF is the only feasible candidate.
The change in functionality is that the on-target runtime package management from remote package feeds is now done with a different tool that has a different set of command-line options. If you have scripts that call the tool directly, or use its API, they need to be fixed.
For more information, see the DNF Documentation.
Rpm 5.x is replaced with Rpm 4.x. This is done for two major reasons:
DNF is API-incompatible with Rpm 5.x and porting it and maintaining the port is non-trivial.
Rpm 5.x itself has limited maintenance upstream, and the Yocto Project is one of the very few remaining users.
Berkeley DB 6.x is removed and Berkeley DB 5.x becomes the default:
Version 6.x of Berkeley DB has largely been rejected by the open source community due to its AGPLv3 license. As a result, most mainstream open source projects that require DB are still developed and tested with DB 5.x.
In OE-core, the only thing that was requiring DB 6.x was Rpm 5.x. Thus, no reason exists to continue carrying DB 6.x in OE-core.
createrepo
is replaced with
createrepo_c
.
createrepo_c
is the current
incarnation of the tool that generates remote repository
metadata.
It is written in C as compared to
createrepo
, which is written in
Python.
createrepo_c
is faster and is
maintained.
Architecture-independent RPM packages are "noarch" instead of "all".
This change was made because too many places in
DNF/RPM4 stack already make that assumption.
Only the filenames and the architecture tag has changed.
Nothing else has changed in OE-core system, particularly
in the
allarch.bbclass
class.
Signing of remote package feeds using
PACKAGE_FEED_SIGN
is not currently supported.
This issue will be fully addressed in a future
Yocto Project release.
See defect 11209
for more information on a solution to package feed
signing with RPM in the Yocto Project 2.3 release.
OPKG now uses the libsolv backend for resolving package dependencies by default. This is vastly superior to OPKG's internal ad-hoc solver that was previously used. This change does have a small impact on disk (around 500 KB) and memory footprint.
The following recipes have been removed:
linux-yocto 4.8:
Version 4.8 has been removed.
Versions 4.1 (LTSI), 4.4 (LTS), 4.9 (LTS/LTSI) and 4.10
are now present.
python-smartpm:
Functionally replaced by dnf
.
createrepo:
Replaced by the createrepo-c
recipe.
rpmresolve:
No longer needed with the move to RPM 4 as RPM itself is
used instead.
gstreamer:
Removed the GStreamer Git version recipes as they have
been stale.
1.10.
x
recipes are still present.
alsa-conf-base:
Merged into alsa-conf
since
libasound
depended on both.
Essentially, no way existed to install only one of these.
tremor:
Moved to meta-multimedia
.
Fixed-integer Vorbis decoding is not
needed by current hardware.
Thus, GStreamer's ivorbis plugin has been disabled
by default eliminating the need for the
tremor
recipe in
OE-Core.
gummiboot:
Replaced by systemd-boot
.
The following changes have been made to Wic:
Default Output Directory Changed:
Wic's default output directory is now the current directory
by default instead of the unusual
/var/tmp/wic
.
The "-o" and "--outdir" options remain unchanged and are used to specify your preferred output directory if you do not want to use the default directory.
fsimage Plug-in Removed: The Wic fsimage plugin has been removed as it duplicates functionality of the rawcopy plugin.
The following QA checks have changed:
unsafe-references-in-binaries
:
The unsafe-references-in-binaries
QA check, which was disabled by default, has now been
removed.
This check was intended to detect binaries in
/bin
that link to libraries in
/usr/lib
and have the case where
the user has /usr
on a separate
filesystem to /
.
The removed QA check was buggy.
Additionally, /usr
residing on a
separate partition from /
is now
a rare configuration.
Consequently,
unsafe-references-in-binaries
was
removed.
file-rdeps
:
The file-rdeps
QA check is now an
error by default instead of a warning.
Because it is an error instead of a warning, you need to
address missing runtime dependencies.
For additional information, see the
insane
class and the
"Errors and Warnings"
section.
The following miscellaneous changes have occurred:
In this release, a number of recipes have been changed to
ignore the largefile
DISTRO_FEATURES
item, enabling large file support unconditionally.
This feature has always been enabled by default.
Disabling the feature has not been widely tested.
largefile
feature,
which would make it unconditionally enabled everywhere.
If the
DISTRO_VERSION
value contains the value of the
DATE
variable, which is the default between Poky releases,
the DATE
value is explicitly excluded
from /etc/issue
and
/etc/issue.net
, which is displayed at
the login prompt, in order to avoid conflicts with
Multilib enabled.
Regardless, the DATE
value is
inaccurate if the base-files
recipe is restored from shared state (sstate) rather
than rebuilt.
If you need the build date recorded in
/etc/issue*
or anywhere else in your
image, a better method is to define a post-processing
function to do it and have the function called from
ROOTFS_POSTPROCESS_COMMAND
.
Doing so ensures the value is always up-to-date with the
created image.
Dropbear's init
script now disables
DSA host keys by default.
This change is in line with the systemd service
file, which supports RSA keys only, and with recent
versions of OpenSSH, which deprecates DSA host keys.
The
buildhistory
class now correctly uses tabs as separators between all
columns in installed-package-sizes.txt
in order to aid import into other tools.
The USE_LDCONFIG
variable has been
replaced with the "ldconfig"
DISTRO_FEATURES
feature.
Distributions that previously set:
USE_LDCONFIG = "0"
should now instead use the following:
DISTRO_FEATURES_BACKFILL_CONSIDERED_append = " ldconfig"
The default value of
COPYLEFT_LICENSE_INCLUDE
now includes all versions of AGPL licenses in addition
to GPL and LGPL.
Kernel module packages are now suffixed with the kernel version in order to allow module packages from multiple kernel versions to co-exist on a target system. If you wish to return to the previous naming scheme that does not include the version suffix, use the following:
KERNEL_MODULE_PACKAGE_SUFFIX to ""
Removal of libtool
*.la
files is now enabled by default.
The *.la
files are not actually
needed on Linux and relocating them is an unnecessary
burden.
If you need to preserve these
.la
files (e.g. in a custom
distribution), you must change
INHERIT_DISTRO
such that "remove-libtool" is not included in the value.
Extensible SDKs built for GCC 5+ now refuse to install on a
distribution where the host GCC version is 4.8 or 4.9.
This change resulted from the fact that the installation
is known to fail due to the way the
uninative
shared state (sstate)
package is built.
See the
uninative
class for additional information.
All native and nativesdk recipes now use a separate
DISTRO_FEATURES
value instead of
sharing the value used by recipes for the target, in order
to avoid unnecessary rebuilds.
The DISTRO_FEATURES
for
native
recipes is
DISTRO_FEATURES_NATIVE
added to an intersection of
DISTRO_FEATURES
and
DISTRO_FEATURES_FILTER_NATIVE
.
For nativesdk recipes, the
corresponding variables are
DISTRO_FEATURES_NATIVESDK
and
DISTRO_FEATURES_FILTER_NATIVESDK
.
The FILESDIR
variable, which was previously deprecated and rarely used,
has now been removed.
You should change any recipes that set
FILESDIR
to set
FILESPATH
instead.
The MULTIMACH_HOST_SYS
variable has been removed as it is no longer needed
with recipe-specific sysroots.
This section provides migration information for moving to the Yocto Project 2.4 Release from the prior release.
A persistent mode is now available in BitBake's default operation,
replacing its previous "memory resident mode" (i.e.
oe-init-build-env-memres
).
Now you only need to set
BB_SERVER_TIMEOUT
to a timeout (in seconds) and BitBake's server stays resident for
that amount of time between invocations.
The oe-init-build-env-memres
script has been
removed since a separate environment setup script is no longer
needed.
This section provides information about packaging changes that have occurred:
python3
Changes:
The main "python3" package now brings in all of the
standard Python 3 distribution rather than a subset.
This behavior matches what is expected based on
traditional Linux distributions.
If you wish to install a subset of Python 3, specify
python-core
plus one or more of
the individual packages that are still produced.
python3
:
The bz2.py
,
lzma.py
, and
_compression.py
scripts have
been moved from the
python3-misc
package to
the python3-compression
package.
binutils
:
The libbfd
library is now packaged in
a separate "libbfd" package.
This packaging saves space when certain tools
(e.g. perf
) are installed.
In such cases, the tools only need
libbfd
rather than all the packages in
binutils
.
util-linux
Changes:
The su
program is now packaged
in a separate "util-linux-su" package, which is only
built when "pam" is listed in the
DISTRO_FEATURES
variable.
util-linux
should not be
installed unless it is needed because
su
is normally provided through
the shadow file format.
The main util-linux
package has
runtime dependencies (i.e.
RDEPENDS
)
on the util-linux-su
package
when "pam" is in
DISTRO_FEATURES
.
The switch_root
program is now
packaged in a separate "util-linux-switch-root"
package for small initramfs images that do not need
the whole util-linux
package or
the busybox binary, which are both much larger than
switch_root
.
The main util-linux
package has
a recommended runtime dependency (i.e.
RRECOMMENDS
)
on the util-linux-switch-root
package.
The ionice
program is now
packaged in a separate "util-linux-ionice" package.
The main util-linux
package has
a recommended runtime dependency (i.e.
RRECOMMENDS
)
on the util-linux-ionice
package.
initscripts
:
The sushell
program is now packaged in
a separate "initscripts-sushell" package.
This packaging change allows systems to pull
sushell
in when
selinux
is enabled.
The change also eliminates needing to pull in the entire
initscripts
package.
The main initscripts
package has a
runtime dependency (i.e. RDEPENDS
)
on the sushell
package when
"selinux" is in DISTRO_FEATURES
.
glib-2.0
:
The glib-2.0
package now has a
recommended runtime dependency (i.e.
RRECOMMENDS
) on the
shared-mime-info
package, since large
portions of GIO are not useful without the MIME database.
You can remove the dependency by using the
BAD_RECOMMENDATIONS
variable if shared-mime-info
is too
large and is not required.
Go Standard Runtime:
The Go standard runtime has been split out from the main
go
recipe into a separate
go-runtime
recipe.
The following recipes have been removed:
acpitests
:
This recipe is not maintained.
autogen-native
:
No longer required by Grub, oe-core, or meta-oe.
bdwgc
:
Nothing in OpenEmbedded-Core requires this recipe.
It has moved to meta-oe.
byacc
:
This recipe was only needed by rpm 5.x and has moved to
meta-oe.
gcc (5.4)
:
The 5.4 series dropped the recipe in favor of 6.3 / 7.2.
gnome-common
:
Deprecated upstream and no longer needed.
go-bootstrap-native
:
Go 1.9 does its own bootstrapping so this recipe has been
removed.
guile
:
This recipe was only needed by
autogen-native
and
remake
.
The recipe is no longer needed by either of these programs.
libclass-isa-perl
:
This recipe was previously needed for LSB 4, no longer
needed.
libdumpvalue-perl
:
This recipe was previously needed for LSB 4, no longer
needed.
libenv-perl
:
This recipe was previously needed for LSB 4, no longer
needed.
libfile-checktree-perl
:
This recipe was previously needed for LSB 4, no longer
needed.
libi18n-collate-perl
:
This recipe was previously needed for LSB 4, no longer
needed.
libiconv
:
This recipe was only needed for uclibc
,
which was removed in the previous release.
glibc
and musl
have their own implementations.
meta-mingw
still needs
libiconv
, so it has
been moved to meta-mingw
.
libpng12
:
This recipe was previously needed for LSB. The current
libpng
is 1.6.x.
libpod-plainer-perl
:
This recipe was previously needed for LSB 4, no longer
needed.
linux-yocto (4.1)
:
This recipe was removed in favor of 4.4, 4.9, 4.10 and 4.12.
mailx
:
This recipe was previously only needed for LSB
compatibility, and upstream is defunct.
mesa (git version only)
:
The git version recipe was stale with respect to the release
version.
ofono (git version only)
:
The git version recipe was stale with respect to the release
version.
portmap
:
This recipe is obsolete and is superseded by
rpcbind
.
python3-pygpgme
:
This recipe is old and unmaintained. It was previously
required by dnf
, which has switched
to official gpgme
Python bindings.
python-async
:
This recipe has been removed in favor of the Python 3
version.
python-gitdb
:
This recipe has been removed in favor of the Python 3
version.
python-git
:
This recipe was removed in favor of the Python 3
version.
python-mako
:
This recipe was removed in favor of the Python 3
version.
python-pexpect
:
This recipe was removed in favor of the Python 3 version.
python-ptyprocess
:
This recipe was removed in favor of Python the 3 version.
python-pycurl
:
Nothing is using this recipe in OpenEmbedded-Core
(i.e. meta-oe
).
python-six
:
This recipe was removed in favor of the Python 3 version.
python-smmap
:
This recipe was removed in favor of the Python 3 version.
remake
:
Using remake
as the provider of
virtual/make
is broken.
Consequently, this recipe is not needed in OpenEmbedded-Core.
Kernel Device Tree support is now easier to enable in a kernel
recipe.
The Device Tree code has moved to a
kernel-devicetree
class.
Functionality is automatically enabled for any recipe that inherits
the
kernel
class and sets the
KERNEL_DEVICETREE
variable.
The previous mechanism for doing this,
meta/recipes-kernel/linux/linux-dtb.inc
,
is still available to avoid breakage, but triggers a
deprecation warning.
Future releases of the Yocto Project will remove
meta/recipes-kernel/linux/linux-dtb.inc
.
It is advisable to remove any require
statements that request
meta/recipes-kernel/linux/linux-dtb.inc
from any custom kernel recipes you might have.
This will avoid breakage in post 2.4 releases.
The following package QA changes took place:
The "unsafe-references-in-scripts" QA check has been removed.
If you refer to ${COREBASE}/LICENSE
within
LIC_FILES_CHKSUM
you receive a warning because this file is a description of
the license for OE-Core.
Use ${COMMON_LICENSE_DIR}/MIT
if your recipe is MIT-licensed and you cannot use the
preferred method of referring to a file within the source
tree.
README
File Changes¶
The following are changes to README
files:
The main Poky README
file has been
moved to the meta-poky
layer and
has been renamed README.poky
.
A symlink has been created so that references to the old
location work.
The README.hardware
file has been moved
to meta-yocto-bsp
.
A symlink has been created so that references to the old
location work.
A README.qemu
file has been created
with coverage of the qemu*
machines.
The following are additional changes:
The ROOTFS_PKGMANAGE_BOOTSTRAP
variable and any references to it have been removed.
You should remove this variable from any custom recipes.
The meta-yocto
directory has been
removed.
meta-yocto
was renamed to
meta-poky
and the
meta-yocto
subdirectory remained
to avoid breaking existing configurations.
The maintainers.inc
file, which tracks
maintainers by listing a primary person responsible for each
recipe in OE-Core, has been moved from
meta-poky
to OE-Core (i.e. from
meta-poky/conf/distro/include
to
meta/conf/distro/include
).
The
buildhistory
class now makes a single commit per build rather than one
commit per subdirectory in the repository.
This behavior assumes the commits are enabled with
BUILDHISTORY_COMMIT
= "1", which is typical.
Previously, the buildhistory
class made
one commit per subdirectory in the repository in order to
make it easier to see the changes for a particular
subdirectory.
To view a particular change, specify that subdirectory as
the last parameter on the git show
or git diff
commands.
The x86-base.inc
file, which is
included by all x86-based machine configurations, now sets
IMAGE_FSTYPES
using ?=
to "live" rather than
appending with +=
.
This change makes the default easier to override.
BitBake fires multiple "BuildStarted" events when multiconfig is enabled (one per configuration). For more information, see the "Events" section in the BitBake User Manual.
By default, the security_flags.inc
file
sets a
GCCPIE
variable with an option to enable Position Independent
Executables (PIE) within gcc
.
Enabling PIE in the GNU C Compiler (GCC), makes Return
Oriented Programming (ROP) attacks much more difficult to
execute.
OE-Core now provides a
bitbake-layers
plugin that implements
a "create-layer" subcommand.
The implementation of this subcommand has resulted in the
yocto-layer
script being deprecated and
will likely be removed in the next Yocto Project release.
The vmdk
, vdi
,
and qcow2
image file types are now
used in conjunction with the "wic" image type through
CONVERSION_CMD
.
Consequently, the equivalent image types are now
wic.vmdk
, wic.vdi
,
and wic.qcow2
, respectively.
do_image_<type>[depends]
has
replaced IMAGE_DEPENDS_<type>
.
If you have your own classes that implement custom image
types, then you need to update them.
OpenSSL 1.1 has been introduced.
However, the default is still 1.0.x through the
PREFERRED_VERSION
variable.
This preference is set is due to the remaining compatibility
issues with other software.
The
PROVIDES
variable in the openssl 1.0 recipe now includes "openssl10"
as a marker that can be used in
DEPENDS
within recipes that build software that still depend on
OpenSSL 1.0.
To ensure consistent behavior, BitBake's "-r" and "-R" options (i.e. prefile and postfile), which are used to read or post-read additional configuration files from the command line, now only affect the current BitBake command. Before these BitBake changes, these options would "stick" for future executions.
This section provides migration information for moving to the Yocto Project 2.5 Release from the prior release.
This section provides information about packaging changes that have occurred:
bind-libs
:
The libraries packaged by the bind recipe are in a
separate bind-libs
package.
libfm-gtk
:
The libfm
GTK+ bindings are split into
a separate libfm-gtk
package.
flex-libfl
:
The flex recipe splits out libfl into a separate
flex-libfl
package to avoid too many
dependencies being pulled in where only the library is
needed.
grub-efi
:
The grub-efi
configuration is split
into a separate grub-bootconf
recipe.
However, the dependency relationship from
grub-efi
is through a
virtual/grub-bootconf provider making it possible to have
your own recipe provide the dependency.
Alternatively, you can use a BitBake append file to bring
the configuration back into the
grub-efi
recipe.
armv7a Legacy Package Feed Support:
Legacy support is removed for transitioning from
armv7a
to
armv7a-vfp-neon
in package feeds,
which was previously enabled by setting
PKGARCHCOMPAT_ARMV7A
.
This transition occurred in 2011 and active package feeds
should by now be updated to the new naming.
The following recipes have been removed:
gcc
:
The version 6.4 recipes are replaced by 7.x.
gst-player
:
Renamed to gst-examples
as per
upstream.
hostap-utils
:
This software package is obsolete.
latencytop
:
This recipe is no longer maintained upstream.
The last release was in 2009.
libpfm4
:
The only file that requires this recipe is
oprofile
, which has been removed.
linux-yocto
:
The version 4.4, 4.9, and 4.10 recipes have been removed.
Versions 4.12, 4.14, and 4.15 remain.
man
:
This recipe has been replaced by modern
man-db
mkelfimage
:
This tool has been removed in the upstream coreboot project,
and is no longer needed with the removal of the ELF image
type.
nativesdk-postinst-intercept
:
This recipe is not maintained.
neon
:
This software package is no longer maintained upstream and
is no longer needed by anything in OpenEmbedded-Core.
oprofile
:
The functionality of this recipe is replaced by
perf
and keeping compatibility on
an ongoing basis with musl
is
difficult.
pax
:
This software package is obsolete.
stat
:
This software package is not maintained upstream.
coreutils
provides a modern stat binary.
zisofs-tools-native
:
This recipe is no longer needed because the compressed
ISO image feature has been removed.
The following are changes to scripts and tools:
yocto-bsp
,
yocto-kernel
, and
yocto-layer
:
The yocto-bsp
,
yocto-kernel
, and
yocto-layer
scripts previously shipped
with poky but not in OpenEmbedded-Core have been removed.
These scripts are not maintained and are outdated.
In many cases, they are also limited in scope.
The bitbake-layers create-layer
command
is a direct replacement for yocto-layer
.
See the documentation to create a BSP or kernel recipe in
the
"BSP Kernel Recipe Example"
section.
devtool finish
:
devtool finish
now exits with an error
if there are uncommitted changes or a rebase/am in progress
in the recipe's source repository.
If this error occurs, there might be uncommitted changes
that will not be included in updates to the patches applied
by the recipe.
A -f/--force option is provided for situations that the
uncommitted changes are inconsequential and you want to
proceed regardless.
scripts/oe-setup-rpmrepo
script:
The functionality of
scripts/oe-setup-rpmrepo
is replaced by
bitbake package-index
.
scripts/test-dependencies.sh
script:
The script is largely made obsolete by the
recipe-specific sysroots functionality introduced in the
previous release.
The following are BitBake changes:
The --runall
option has changed.
There are two different behaviors people might want:
Behavior A: For a given target (or set of targets) look through the task graph and run task X only if it is present and will be built.
Behavior B: For a given target (or set of targets) look through the task graph and run task X if any recipe in the taskgraph has such a target, even if it is not in the original task graph.
The --runall
option now performs
"Behavior B".
Previously --runall
behaved like
"Behavior A".
A --runonly
option has been added to
retain the ability to perform "Behavior A".
Several explicit "run this task for all recipes in the
dependency tree" tasks have been removed (e.g.
fetchall
,
checkuriall
, and the
*all
tasks provided by the
distrodata
and
archiver
classes).
There is a BitBake option to complete this for any arbitrary
task. For example:
bitbake <target> -c fetchall
should now be replaced with:
bitbake <target> --runall=fetch
The following are auto-packaging changes to Python and Python 3:
The script-managed python-*-manifest.inc
files
that were previously used to generate Python and Python 3
packages have been replaced with a JSON-based file that is
easier to read and maintain.
A new task is available for maintainers of the Python recipes to
update the JSON file when upgrading to new Python versions.
You can now edit the file directly instead of having to edit a
script and run it to update the file.
One particular change to note is that the Python recipes no longer
have build-time provides for their packages.
This assumes python-foo
is one of the packages
provided by the Python recipe.
You can no longer run bitbake python-foo
or
have a DEPENDS
on
python-foo
, but doing either of the following
causes the package to work as expected:
IMAGE_INSTALL_append = " python-foo"
or
RDEPENDS_${PN} = "python-foo"
The earlier build-time provides behavior was a quirk of the way the Python manifest file was created. For more information on this change please see this commit.
The following are additional changes:
The kernel
class supports building
packages for multiple kernels.
If your kernel recipe or .bbappend
file
mentions packaging at all, you should replace references to
the kernel in package names with
${KERNEL_PACKAGE_NAME}
.
For example, if you disable automatic installation of the
kernel image using
RDEPENDS_kernel-base = ""
you can avoid
warnings using
RDEPENDS_${KERNEL_PACKAGE_NAME}-base = ""
instead.
The buildhistory
class commits changes
to the repository by default so you no longer need to set
BUILDHISTORY_COMMIT = "1"
.
If you want to disable commits you need to set
BUILDHISTORY_COMMIT = "0"
in your
configuration.
The beaglebone
reference machine has
been renamed to beaglebone-yocto
.
The beaglebone-yocto
BSP is a reference
implementation using only mainline components available in
OpenEmbedded-Core and meta-yocto-bsp
,
whereas Texas Instruments maintains a full-featured BSP in
the meta-ti
layer.
This rename avoids the previous name clash that existed
between the two BSPs.
The update-alternatives
class no longer
works with SysV init
scripts because
this usage has been problematic.
Also, the sysklogd
recipe no longer
uses update-alternatives
because it is
incompatible with other implementations.
By default, the
cmake
class uses ninja
instead of
make
for building.
This improves build performance.
If a recipe is broken with ninja
, then
the recipe can set
OECMAKE_GENERATOR = "Unix Makefiles"
to change back to make
.
The previously deprecated base_*
functions have been removed in favor of their replacements
in meta/lib/oe
and
bitbake/lib/bb
.
These are typically used from recipes and classes.
Any references to the old functions must be updated.
The following table shows the removed functions and their
replacements:
Removed Replacement ============================ ============================ base_path_join() oe.path.join() base_path_relative() oe.path.relative() base_path_out() oe.path.format_display() base_read_file() oe.utils.read_file() base_ifelse() oe.utils.ifelse() base_conditional() oe.utils.conditional() base_less_or_equal() oe.utils.less_or_equal() base_version_less_or_equal() oe.utils.version_less_or_equal() base_contains() bb.utils.contains() base_both_contain() oe.utils.both_contain() base_prune_suffix() oe.utils.prune_suffix() oe_filter() oe.utils.str_filter() oe_filter_out() oe.utils.str_filter_out() (or use the _remove operator).
Using exit 1
to explicitly defer a
postinstall script until first boot is now deprecated since
it is not an obvious mechanism and can mask actual errors.
If you want to explicitly defer a postinstall to first boot
on the target rather than at rootfs
creation time, use
pkg_postinst_ontarget()
or call
postinst_intercept delay_to_first_boot
from pkg_postinst()
.
Any failure of a pkg_postinst()
script (including exit 1
)
will trigger a warning during
do_rootfs
.
For more information, see the "Post-Installation Scripts" section in the Yocto Project Development Tasks Manual.
The elf
image type has been removed.
This image type was removed because the
mkelfimage
tool
that was required to create it is no longer provided by
coreboot upstream and required updating every time
binutils
updated.
Support for .iso image compression (previously enabled
through COMPRESSISO = "1"
) has been
removed.
The userspace tools (zisofs-tools
) are
unmaintained and squashfs
provides
better performance and compression.
In order to build a live image with squashfs+lz4 compression
enabled you should now set
LIVE_ROOTFS_TYPE = "squashfs-lz4"
and ensure that live
is in IMAGE_FSTYPES
.
Recipes with an unconditional dependency on
libpam
are only buildable with
pam
in
DISTRO_FEATURES
.
If the dependency is truly optional then it is recommended
that the dependency be conditional upon
pam
being in
DISTRO_FEATURES
.
For EFI-based machines, the bootloader
(grub-efi
by default) is installed into
the image at /boot.
Wic can be used to split the bootloader into separate boot
and rootfs partitions if necessary.
Patches whose context does not match exactly (i.e. where patch reports "fuzz" when applying) will generate a warning. For an example of this see this commit.
Layers are expected to set
LAYERSERIES_COMPAT_layername
to match the version(s) of OpenEmbedded-Core they are
compatible with.
This is specified as codenames using spaces to separate
multiple values (e.g. "rocko sumo").
If a layer does not set
LAYERSERIES_COMPAT_layername
, a warning
will is shown.
If a layer sets a value that does not include the current
version ("sumo" for the 2.5 release), then an error will be
produced.
The TZ
environment variable is set to
"UTC" within the build environment in order to fix
reproducibility problems in some recipes.
This section provides migration information for moving to the Yocto Project 2.6 Release from the prior release.
The GNU Compiler Collection version 8.2 is now used by default for compilation. For more information on what has changed in the GCC 8.x release, see https://gcc.gnu.org/gcc-8/changes.html.
If you still need to compile with version 7.x, GCC 7.3 is
also provided.
You can select this version by setting the
and can be selected by setting the
GCCVERSION
variable to "7.%" in your configuration.
The following recipes have been removed:
beecrypt
: No longer needed since moving to RPM 4.bigreqsproto
: Replaced byxorgproto
.calibrateproto
: Removed in favor ofxinput
.compositeproto
: Replaced byxorgproto
.damageproto
: Replaced byxorgproto
.dmxproto
: Replaced byxorgproto
.dri2proto
: Replaced byxorgproto
.dri3proto
: Replaced byxorgproto
.eee-acpi-scripts
: Became obsolete.fixesproto
: Replaced byxorgproto
.fontsproto
: Replaced byxorgproto
.fstests
: Became obsolete.gccmakedep
: No longer used.glproto
: Replaced byxorgproto
.gnome-desktop3
: No longer needed. This recipe has moved tometa-oe
.icon-naming-utils
: No longer used since the Sato theme was removed in 2016.inputproto
: Replaced byxorgproto
.kbproto
: Replaced byxorgproto
.libusb-compat
: Became obsolete.libuser
: Became obsolete.libnfsidmap
: No longer an external requirement sincenfs-utils
2.2.1.libnfsidmap
is now integrated.libxcalibrate
: No longer needed withxinput
mktemp
: Became obsolete. Themktemp
command is provided by bothbusybox
andcoreutils
.ossp-uuid
: Is not being maintained and has mostly been replaced byuuid.h
inutil-linux
.pax-utils
: No longer needed. Previous QA tests that did use this recipe are now done at build time.pcmciautils
: Became obsolete.pixz
: No longer needed.xz
now supports multi-threaded compression.presentproto
: Replaced byxorgproto
.randrproto
: Replaced byxorgproto
.recordproto
: Replaced byxorgproto
.renderproto
: Replaced byxorgproto
.resourceproto
: Replaced byxorgproto
.scrnsaverproto
: Replaced byxorgproto
.trace-cmd
: Became obsolete.perf
replaced this recipe's functionally.videoproto
: Replaced byxorgproto
.wireless-tools
: Became obsolete. Superseded byiw
.xcmiscproto
: Replaced byxorgproto
.xextproto
: Replaced byxorgproto
.xf86dgaproto
: Replaced byxorgproto
.xf86driproto
: Replaced byxorgproto
.xf86miscproto
: Replaced byxorgproto
.xf86-video-omapfb
: Became obsolete. Use kernel modesetting driver instead.xf86-video-omap
: Became obsolete. Use kernel modesetting driver instead.xf86vidmodeproto
: Replaced byxorgproto
.xineramaproto
: Replaced byxorgproto
.xproto
: Replaced byxorgproto
.yasm
: No longer needed since previous usages are now satisfied bynasm
.
The following packaging changes have been made:
cmake
:
cmake.m4
and
toolchain
files have been moved to the
main package.
iptables
:
The iptables
modules have been split
into separate packages.
alsa-lib
:
libasound
is now in the main
alsa-lib
package instead of
libasound
.
glibc
:
libnss-db
is now in its own package
along with a /var/db/makedbs.sh
script to update databases.
python
and python3
:
The main package has been removed from the recipe.
You must install specific packages or
python-modules
/
python3-modules
for everything.
systemtap
:
Moved systemtap-exporter
into its own
package.
The "*proto" upstream repositories have been combined into one
"xorgproto" repository.
Thus, the corresponding recipes have also been combined into a
single xorgproto
recipe.
Any recipes that depend upon the older *proto
recipes need to be changed to depend on the newer
xorgproto
recipe instead.
For names of recipes removed because of this repository change, see the Removed Recipes section.
distutils
and distutils3
Now Prevent Fetching Dependencies During the do_configure
Task¶
Previously, it was possible for Python recipes that inherited
the
distutils
and
distutils3
classes to fetch code during the
do_configure
task to satisfy dependencies mentioned in
setup.py
if those dependencies were not
provided in the sysroot (i.e. recipes providing the dependencies
were missing from
DEPENDS
).
distutils
and
distutils3
).
Any recipe that inherits distutils*
classes are affected.
For example, the setuptools
and
setuptools3
recipes are affected since
they inherit the distutils*
classes.
Fetching these types of dependencies that are not provided in the
sysroot negatively affects the ability to reproduce builds.
This type of fetching is now explicitly disabled.
Consequently, any missing dependencies in Python recipes that
use these classes now result in an error during the
do_configure
task.
linux-yocto
Configuration Audit Issues Now Correctly Reported¶
Due to a bug, the kernel configuration audit functionality was
not writing out any resulting warnings during the build.
This issue is now corrected.
You might notice these warnings now if you have a custom kernel
configuration with a linux-yocto
style
kernel recipe.
The following changes have been made:
Name variables (e.g.
IMAGE_NAME
)
use a new IMAGE_VERSION_SUFFIX
variable instead of
DATETIME
.
Using IMAGE_VERSION_SUFFIX
allows
easier and more direct changes.
The IMAGE_VERSION_SUFFIX
variable is set in the
bitbake.conf
configuration file as
follows:
IMAGE_VERSION_SUFFIX = "-${DATETIME}"
Several variables have changed names for consistency:
Old Variable Name New Variable Name ======================================================== KERNEL_IMAGE_BASE_NAME KERNEL_IMAGE_NAME KERNEL_IMAGE_SYMLINK_NAME KERNEL_IMAGE_LINK_NAME MODULE_TARBALL_BASE_NAME MODULE_TARBALL_NAME MODULE_TARBALL_SYMLINK_NAME MODULE_TARBALL_LINK_NAME INITRAMFS_BASE_NAME INITRAMFS_NAME
The MODULE_IMAGE_BASE_NAME
variable
has been removed.
The module tarball name is now controlled directly with the
MODULE_TARBALL_NAME
variable.
The
KERNEL_DTB_NAME
and
KERNEL_DTB_LINK_NAME
variables have been introduced to control kernel Device
Tree Binary (DTB) artifact names instead of mangling
KERNEL_IMAGE_*
variables.
The
KERNEL_FIT_NAME
and
KERNEL_FIT_LINK_NAME
variables have been introduced to specify the name of
flattened image tree (FIT) kernel images similar to other
deployed artifacts.
The
MODULE_TARBALL_NAME
and
MODULE_TARBALL_LINK_NAME
variable values no longer include the "module-" prefix or
".tgz" suffix.
These parts are now hardcoded so that the values are
consistent with other artifact naming variables.
Added the
INITRAMFS_LINK_NAME
variable so that the symlink can be controlled similarly
to other artifact types.
INITRAMFS_NAME
now uses
"${PKGE}-${PKGV}-${PKGR}-${MACHINE}${IMAGE_VERSION_SUFFIX}"
instead of
"${PV}-${PR}-${MACHINE}-${DATETIME}", which
makes it consistent with other variables.
SERIAL_CONSOLE
Deprecated¶
The
SERIAL_CONSOLE
variable has been functionally replaced by the
SERIAL_CONSOLES
variable for some time.
With the Yocto Project 2.6 release,
SERIAL_CONSOLE
has been officially deprecated.
SERIAL_CONSOLE
will continue to work as
before for the 2.6 release.
However, for the sake of future compatibility, it is recommended
that you replace all instances of
SERIAL_CONSOLE
with
SERIAL_CONSOLES
.
SERIAL_CONSOLES
expects entries to be
separated using semicolons as compared to
SERIAL_CONSOLE
, which expects spaces.
If the configure script reports an unknown option, this now triggers a QA error instead of a warning. Any recipes that previously got away with specifying such unknown options now need to be fixed.
The following changes have occurred:
The virtclass-native
and
virtclass-nativesdk
Overrides Have
Been Removed:
The virtclass-native
and
virtclass-nativesdk
overrides have
been deprecated since 2012 in favor of
class-native
and
class-nativesdk
, respectively.
Both virtclass-native
and
virtclass-nativesdk
are now dropped.
virtclass-multilib-
overrides
for multilib are still valid.
The forcevariable
Override Now Has a Higher Priority Than
libc
Overrides:
The forcevariable
override is
documented to be the highest priority override.
However, due to a long-standing quirk of how
OVERRIDES
is set, the libc
overrides (e.g.
libc-glibc
,
libc-musl
, and so forth) erroneously
had a higher priority.
This issue is now corrected.
It is likely this change will not cause any
problems.
However, it is possible with some unusual configurations
that you might see a change in behavior if you were
relying on the previous behavior.
Be sure to check how you use
forcevariable
and
libc-*
overrides in your custom
layers and configuration files to ensure they make sense.
The build-${BUILD_OS}
Override Has Been Removed:
The build-${BUILD_OS}
, which is
typically build-linux
, override has
been removed because building on a host operating system
other than a recent version of Linux is neither supported
nor recommended.
Dropping the override avoids giving the impression that
other host operating systems might be supported.
The "_remove" operator now preserves whitespace. Consequently, when specifying list items to remove, be aware that leading and trailing whitespace resulting from the removal is retained.
See the "Removal (Override Style Syntax)" section in the BitBake User Manual for a detailed example.
systemd
Configuration is Now Split Into systemd-conf
¶
The configuration for the systemd
recipe
has been moved into a system-conf
recipe.
Moving this configuration to a separate recipe avoids the
systemd
recipe from becoming machine-specific
for cases where machine-specific configurations need to be applied
(e.g. for qemu*
machines).
Currently, the new recipe packages the following files:
${sysconfdir}/machine-id ${sysconfdir}/systemd/coredump.conf ${sysconfdir}/systemd/journald.conf ${sysconfdir}/systemd/logind.conf ${sysconfdir}/systemd/system.conf ${sysconfdir}/systemd/user.conf
If you previously used bbappend files to append the
systemd
recipe to change any of the
listed files, you must do so for the
systemd-conf
recipe instead.
This section provides information about automatic testing changes:
TEST_IMAGE
Variable Removed:
Prior to this release, you set the
TEST_IMAGE
variable to "1" to
enable automatic testing for successfully built images.
The TEST_IMAGE
variable no longer
exists and has been replaced by the
TESTIMAGE_AUTO
variable.
Inheriting the testimage
and
testsdk
Classes:
Best practices now dictate that you use the
IMAGE_CLASSES
variable rather than the
INHERIT
variable when you inherit the
testimage
and
testsdk
classes used for automatic testing.
OpenSSL has been upgraded from 1.0 to 1.1. By default, this upgrade could cause problems for recipes that have both versions in their dependency chains. The problem is that both versions cannot be installed together at build time.
The server logfile bitbake-cookerdaemon.log
is
now always placed in the
Build Directory
instead of the current directory.
The Poky distribution now uses security compiler flags by default. Inclusion of these flags could cause new failures due to stricter checking for various potential security issues in code.
You must explicitly mark post installs to defer to the target.
If you want to explicitly defer a postinstall to first boot on
the target rather than at rootfs creation time, use
pkg_postinst_ontarget()
or call
postinst_intercept delay_to_first_boot
from
pkg_postinst()
.
Any failure of a pkg_postinst()
script
(including exit 1) triggers an error during the
do_rootfs
task.
For more information on post-installation behavior, see the "Post-Installation Scripts" section in the Yocto Project Development Tasks Manual.
The python3
recipe now enables profile-guided
optimization.
Using this optimization requires a little extra build time in
exchange for improved performance on the target at runtime.
Additionally, the optimization is only enabled if the current
MACHINE
has support for user-mode emulation in QEMU (i.e. "qemu-usermode"
is in
MACHINE_FEATURES
,
which it is by default).
If you wish to disable Python profile-guided optimization
regardless of the value of
MACHINE_FEATURES
, then ensure that
PACKAGECONFIG
for the python3
recipe does not contain "pgo".
You could accomplish the latter using the following at the
configuration level:
PACKAGECONFIG_remove_pn-python3 = "pgo"
Alternatively, you can set
PACKAGECONFIG
using an append file for the
python3
recipe.
The following miscellaneous changes occurred:
Default to using the Thumb-2 instruction set for armv7a and above. If you have any custom recipes that build software that needs to be built with the ARM instruction set, change the recipe to set the instruction set as follows:
ARM_INSTRUCTION_SET = "arm"
run-postinsts
no longer uses
/etc/*-postinsts
for
dpkg/opkg
in favor of built-in
postinst support.
RPM behavior remains unchanged.
The NOISO
and
NOHDD
variables are no longer used.
You now control building *.iso
and
*.hddimg
image types directly
by using the
IMAGE_FSTYPES
variable.
The scripts/contrib/mkefidisk.sh
has been removed in favor of Wic.
kernel-modules
has been removed from
RRECOMMENDS
for qemumips
and
qemumips64
machines.
Removal also impacts the x86-base.inc
file.
genericx86
and
genericx86-64
retain
kernel-modules
as part of the
RRECOMMENDS
variable setting.
The LGPLv2_WHITELIST_GPL-3.0
variable has been removed.
If you are setting this variable in your configuration,
set or append it to the
WHITELIST_GPL-3.0
variable instead.
${ASNEEDED}
is now included in
the
TARGET_LDFLAGS
variable directly.
The remaining definitions from
meta/conf/distro/include/as-needed.inc
have been moved to corresponding recipes.
Support for DSA host keys has been dropped from the OpenSSH recipes. If you are still using DSA keys, you must switch over to a more secure algorithm as recommended by OpenSSH upstream.
The dhcp
recipe now uses the
dhcpd6.conf
configuration file in
dhcpd6.service
for IPv6 DHCP rather
than re-using dhcpd.conf
, which is
now reserved for IPv4.
This section provides migration information for moving to the Yocto Project 2.7 Release from the prior release.
The following changes have been made to BitBake:
BitBake now checks anonymous Python functions and pure
Python functions (e.g. def funcname:
)
in the metadata for tab indentation.
If found, BitBake produces a warning.
Bitbake now checks
BBFILE_COLLECTIONS
for duplicate entries and triggers an error if any are
found.
Support for the Eclipse IDE has been removed. Support continues for those releases prior to 2.7 that did include support. The 2.7 release does not include the Eclipse Yocto plugin.
qemu-native
Splits the System and User-Mode Parts¶
The system and user-mode parts of qemu-native
are now split.
qemu-native
provides the user-mode components
and qemu-system-native
provides the system
components.
If you have recipes that depend on QEMU's system emulation
functionality at build time, they should now depend upon
qemu-system-native
instead of
qemu-native
.
upstream-tracking.inc
File Has Been Removed¶
The previously deprecated upstream-tracking.inc
file is now removed.
Any UPSTREAM_TRACKING*
variables are now set
in the corresponding recipes instead.
Remove any references you have to the
upstream-tracking.inc
file in your
configuration.
DISTRO_FEATURES_LIBC
Variable Has Been Removed¶
The DISTRO_FEATURES_LIBC
variable is no
longer used.
The ability to configure glibc using kconfig has been removed
for quite some time making the libc-*
features
set no longer effective.
Remove any references you have to
DISTRO_FEATURES_LIBC
in your own layers.
The following corrections have been made to the
LICENSE
values set by recipes:
socat: Corrected LICENSE
to be "GPLv2" rather than
"GPLv2+".
libgfortran: Set license to "GPL-3.0-with-GCC-exception".
elfutils: Removed "Elfutils-Exception" and set to "GPLv2" for shared
libraries
This section provides information about packaging changes.
bind
: The
nsupdate
binary has been moved to
the bind-utils
package.
Debug split: The default debug split has been changed to
create separate source packages (i.e.
package_name
-dbg
and
package_name
-src
).
If you are currently using dbg-pkgs
in
IMAGE_FEATURES
to bring in debug symbols and you still need the sources,
you must now also add src-pkgs
to
IMAGE_FEATURES
.
Source packages remain in the target portion of the SDK
by default, unless you have set your own value for
SDKIMAGE_FEATURES
that does not include src-pkgs
.
Mount all using util-linux
:
/etc/default/mountall
has
moved into the -mount sub-package.
Splitting binaries using util-linux
:
util-linux
now splits each binary into
its own package for fine-grained control.
The main util-linux
package pulls in
the individual binary packages using the
RRECOMMENDS
and
RDEPENDS
variables.
As a result, existing images should not see any changes
assuming
NO_RECOMMENDATIONS
is not set.
netbase/base-files
:
/etc/hosts
has moved from
netbase
to
base-files
.
tzdata
: The main package has been
converted to an empty meta package that pulls in all
tzdata
packages by default.
lrzsz
: This package has been removed
from packagegroup-self-hosted
and
packagegroup-core-tools-testapps
.
The X/Y/ZModem support is less likely to be needed on
modern systems.
If you are relying on these packagegroups to include the
lrzsz
package in your image, you
now need to explicitly add the package.
The following recipes have been removed:
gcc: Drop version 7.3 recipes. Version 8.3 now remains. linux-yocto: Drop versions 4.14 and 4.18 recipes. Versions 4.19 and 5.0 remain. go: Drop version 1.9 recipes. Versions 1.11 and 1.12 remain. xvideo-tests: Became obsolete. libart-lgpl: Became obsolete. gtk-icon-utils-native: These tools are now provided by gtk+3-native gcc-cross-initial: No longer needed. gcc-cross/gcc-crosssdk is now used instead. gcc-crosssdk-initial: No longer needed. gcc-cross/gcc-crosssdk is now used instead. glibc-initial: Removed because the benefits of having it for site_config are currently outweighed by the cost of building the recipe.
The following classes have been removed:
distutils-tools: This class was never used. bugzilla.bbclass: Became obsolete. distrodata: This functionally has been replaced by a more modern tinfoil-based implementation.
The following miscellaneous changes occurred:
The distro
subdirectory of the Poky
repository has been removed from the top-level
scripts
directory.
Perl now builds for the target using
perl-cross
for better maintainability and improved build performance.
This change should not present any problems unless you have
heavily customized your Perl recipe.
arm-tunes
: Removed the "-march"
option if mcpu is already added.
update-alternatives
: Convert file
renames to
PACKAGE_PREPROCESS_FUNCS
base/pixbufcache
: Obsolete
sstatecompletions
code has been
removed.
inetutils
: This recipe has rsh
disabled.
This section provides migration information for moving to the Yocto Project 3.0 Release from the prior release.
Changing the init system manager previously required setting a
number of different variables.
You can now change the manager by setting the
INIT_MANAGER
variable and the corresponding
include files
(i.e. conf/distro/include/init-manager-*.conf
).
Include files are provided for four values: "none", "sysvinit",
"systemd", and "mdev-busybox".
The default value, "none", for INIT_MANAGER
should allow your current settings to continue working.
However, it is advisable to explicitly set
INIT_MANAGER
.
Linux Standard Base (LSB) as a standard is not current, and is not well suited for embedded applications. Support can be continued in a separate layer if needed. However, presently LSB support has been removed from the core.
As a result of this change, the poky-lsb
derivative distribution configuration that was also used for
testing alternative configurations has been replaced with a
poky-altcfg
distribution that has LSB
parts removed.
The following recipes have been removed.
core-image-lsb-dev
: Part of removed
LSB support.
core-image-lsb
: Part of removed
LSB support.
core-image-lsb-sdk
: Part of removed
LSB support.
cve-check-tool
: Functionally replaced
by the cve-update-db
recipe and
cve-check
class.
eglinfo
: No longer maintained.
eglinfo
from
mesa-demos
is an adequate and
maintained alternative.
gcc-8.3
: Version 8.3 removed.
Replaced by 9.2.
gnome-themes-standard
: Only needed
by gtk+ 2.x, which has been removed.
gtk+
: GTK+ 2 is obsolete and has been
replaced by gtk+3.
irda-utils
: Has become obsolete.
IrDA support has been removed from the Linux kernel in
version 4.17 and later.
libnewt-python
:
libnewt
Python support merged into
main libnewt
recipe.
libsdl
: Replaced by newer
libsdl2
.
libx11-diet
: Became obsolete.
libxx86dga
: Removed obsolete client
library.
libxx86misc
: Removed. Library is
redundant.
linux-yocto
: Version 5.0 removed,
which is now redundant (5.2 / 4.19 present).
lsbinitscripts
: Part of removed LSB
support.
lsb
: Part of removed LSB support.
lsbtest
: Part of removed LSB support.
openssl10
: Replaced by newer
openssl
version 1.1.
packagegroup-core-lsb
: Part of removed
LSB support.
python-nose
: Removed the Python 2.x
version of the recipe.
python-numpy
: Removed the Python 2.x
version of the recipe.
python-scons
: Removed the Python 2.x
version of the recipe.
source-highlight
: No longer needed.
stress
: Replaced by
stress-ng
.
vulkan
: Split into
vulkan-loader
,
vulkan-headers
, and
vulkan-tools
.
weston-conf
: Functionality moved to
weston-init
.
The following packaging changes have occurred.
The
Epiphany
browser has been dropped from
packagegroup-self-hosted
as it has
not been needed inside
build-appliance-image
for
quite some time and was causing resource problems.
libcap-ng
Python support has been
moved to a separate libcap-ng-python
recipe to streamline the build process when the Python
bindings are not needed.
libdrm
now packages the file
amdgpu.ids
into a separate
libdrm-amdgpu
package.
python3
: The
runpy
module is now in the
python3-core
package as it is
required to support the common "python3 -m" command usage.
distcc
now provides separate
distcc-client
and
distcc-server
packages as typically
one or the other are needed, rather than both.
python*-setuptools
recipes now
separately package the pkg_resources
module in a python-pkg-resources
/
python3-pkg-resources
package as
the module is useful independent of the rest of the
setuptools package.
The main python-setuptools
/
python3-setuptools
package depends
on this new package so you should only need to update
dependencies unless you want to take advantage of the
increased granularity.
cve-check-tool
has been functionally replaced
by a new cve-update-db
recipe and
functionality built into the cve-check
class.
The result uses NVD JSON data feeds rather than the deprecated
XML feeds that cve-check-tool
was using,
supports CVSSv3 scoring, and makes other improvements.
Additionally, the CVE_CHECK_CVE_WHITELIST
variable has been replaced by
CVE_CHECK_WHITELIST
.
The following BitBake changes have occurred.
addtask
statements now properly
validate dependent tasks.
Previously, an invalid task was silently ignored.
With this change, the invalid task generates a warning.
Other invalid addtask
and
deltask
usages now trigger these
warnings: "multiple target tasks arguments with
addtask / deltask", and "multiple before/after clauses".
The "multiconfig" prefix is now shortened to "mc". "multiconfig" will continue to work, however it may be removed in a future release.
The bitbake -g
command no longer
generates a recipe-depends.dot
file
as the contents (i.e. a reprocessed version of
task-depends.dot
) were confusing.
The bb.build.FuncFailed
exception,
previously raised by
bb.build.exec_func()
when certain
other exceptions have occurred, has been removed.
The real underlying exceptions will be raised instead.
If you have calls to
bb.build.exec_func()
in custom classes
or tinfoil-using
scripts, any
references to bb.build.FuncFailed
should be cleaned up.
Additionally, the
bb.build.exec_func()
no longer accepts
the "pythonexception" parameter.
The function now always raises exceptions.
Remove this argument in any calls to
bb.build.exec_func()
in custom classes
or scripts.
The
BB_SETSCENE_VERIFY_FUNCTION2
is no longer used.
In the unlikely event that you have any references to it,
they should be removed.
The RunQueueExecuteScenequeue
and
RunQueueExecuteTasks
events have been
removed since setscene tasks are now executed as part of
the normal runqueue.
Any event handling code in custom classes or scripts that
handles these two events need to be updated.
The arguments passed to functions used with
BB_HASHCHECK_FUNCTION
have changed.
If you are using your own custom hash check function, see
http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=40a5e193c4ba45c928fccd899415ea56b5417725
for details.
Task specifications in BB_TASKDEPDATA
and class implementations used in signature generator
classes now use "<fn>:<task>" everywhere rather than
the "." delimiter that was being used in some places.
This change makes it consistent with all areas in the code.
Custom signature generator classes and code that reads
BB_TASKDEPDATA
need to be updated to
use ':' as a separator rather than '.'.
The following sanity check changes occurred.
SRC_URI
is now checked for usage of two problematic items:
"${PN}" prefix/suffix use - Warnings always appear if ${PN} is used. You must fix the issue regardless of whether multiconfig or anything else that would cause prefixing/suffixing to happen.
Github archive tarballs - these are not guaranteed to be stable. Consequently, it is likely that the tarballs will be refreshed and thus the SRC_URI checksums will fail to apply. It is recommended that you fetch either an official release tarball or a specific revision from the actual Git repository instead.
Either one of these items now trigger a warning by default.
If you wish to disable this check, remove
src-uri-bad
from
WARN_QA
.
The file-rdeps
runtime dependency
check no longer expands
RDEPENDS
recursively as there is no mechanism to ensure they can be
fully computed, and thus races sometimes result in errors
either showing up or not.
Thus, you might now see errors for missing runtime
dependencies that were previously satisfied recursively.
Here is an example: package A contains a shell script
starting with #!/bin/bash
but has no
dependency on bash.
However, package A depends on package B, which does depend
on bash.
You need to add the missing dependency or dependencies to
resolve the warning.
Setting DEPENDS_${PN}
anywhere
(i.e. typically in a recipe) now triggers an error.
The error is triggered because
DEPENDS
is not a package-specific variable unlike RDEPENDS.
You should set DEPENDS
instead.
systemd currently does not work well with the musl C library because only upstream officially supports linking the library with glibc. Thus, a warning is shown when building systemd in conjunction with musl.
The following miscellaneous changes have occurred.
The gnome
class has been removed because it now does very little.
You should update recipes that previously inherited this
class to do the following:
inherit gnomebase gtk-icon-cache gconf mime
The
meta/recipes-kernel/linux/linux-dtb.inc
file has been removed.
This file was previously deprecated in favor of setting
KERNEL_DEVICETREE
in any kernel recipe and only produced a warning.
Remove any include
or
require
statements pointing to this
file.
TARGET_CFLAGS
,
TARGET_CPPFLAGS
,
TARGET_CXXFLAGS
,
and
TARGET_LDFLAGS
are no longer exported to the external environment.
This change did not require any changes to core recipes,
which is a good indicator that no changes will be
required.
However, if for some reason the software being built by one
of your recipes is expecting these variables to be set,
then building the recipe will fail.
In such cases, you must either export the variable or
variables in the recipe or change the scripts so that
exporting is not necessary.
You must change the host distro identifier used in
NATIVELSBSTRING
to use all lowercase characters even if it does not contain
a version number.
This change is necessary only if you are not using
uninative
and
SANITY_TESTED_DISTROS
.
In the base-files
recipe, writing the
hostname into /etc/hosts
and
/etc/hostname
is now done within the
main
do_install
function rather than in the
do_install_basefilesissue
function.
The reason for the change is because
do_install_basefilesissue
is more
easily overridden without having to duplicate the hostname
functionality.
If you have done the latter (e.g. in a
base-files
bbappend), then you should
remove it from your customized
do_install_basefilesissue
function.
The wic --expand
command now uses
commas to separate "key:value" pairs rather than hyphens.
You must update any scripts or commands where you use
wic --expand
with multiple
"key:value" pairs.
UEFI image variable settings have been moved from various
places to a central
conf/image-uefi.conf
.
This change should not influence any existing configuration
as the meta/conf/image-uefi.conf
in the core metadata sets defaults that can be overridden
in the same manner as before.
conf/distro/include/world-broken.inc
has been removed.
For cases where certain recipes need to be disabled when
using the musl C library, these recipes now have
COMPATIBLE_HOST_libc-musl
set with a
comment that explains why.
This section provides migration information for moving to the Yocto Project 3.1 Release from the prior release.
The following versions / requirements of build host components have been updated:
gcc 5.0
python 3.5
tar 1.28
rpcgen
is now required on the host (part of the libc-dev-bin
package on Ubuntu, Debian and related distributions, and the glibc
package on RPM-based distributions).
Additionally, the makeinfo
and pod2man
tools are no longer required on the host.
The MPC8315E-RDB machine is old/obsolete and unobtainable, thus given the maintenance burden
the mpc8315e-rdb
machine configuration that supported it has been removed
in this release. The removal does leave a gap in official PowerPC reference hardware
support; this may change in future if a suitable machine with accompanying support resources
is found.
Due to the expiration of upstream support in January 2020, support for Python 2 has now been removed; it is recommended that you use Python 3 instead. If absolutely needed there is a meta-python2 community layer containing Python 2, related classes and various Python 2-based modules, however it should not be considered as supported.
In order to avoid unnecessary differences in output files (aiding binary reproducibility), the Poky distribution configuration (DISTRO = "poky"
) now inherits the reproducible_build
class by default.
The Poky distribution configuration (DISTRO = "poky"
) enables ptests by default to enable runtime testing of various components. In this release, a dependency needed to be added that has resulted in a significant increase in the number of components that will be built just when building a simple image such as core-image-minimal. If you do not need runtime tests enabled for core components, then it is recommended that you remove "ptest" from DISTRO_FEATURES
to save a significant amount of build time e.g. by adding the following in your configuration:
DISTRO_FEATURES_remove = "ptest"
The following recipes have been removed:
chkconfig
: obsolete
console-tools
: obsolete
enchant
: replaced by enchant2
foomatic-filters
: obsolete
libidn
: no longer needed, moved to meta-oe
libmodulemd
: replaced by libmodulemd-v1
linux-yocto
: drop 4.19, 5.2 version recipes (5.4 now provided)
nspr
: no longer needed, moved to meta-oe
nss
: no longer needed, moved to meta-oe
python
: Python 2 removed (Python 3 preferred)
python-setuptools
: Python 2 version removed (python3-setuptools preferred)
sysprof
: no longer needed, moved to meta-oe
texi2html
: obsolete
u-boot-fw-utils
: functionally replaced by libubootenv
The distro_features_check
class has had its functionality expanded, now supporting ANY_OF_MACHINE_FEATURES
, REQUIRED_MACHINE_FEATURES
, CONFLICT_MACHINE_FEATURES
, ANY_OF_COMBINED_FEATURES
, REQUIRED_COMBINED_FEATURES
, CONFLICT_COMBINED_FEATURES
. As a result the class has now been renamed to features_check
; the distro_features_check
class still exists but generates a warning and redirects to the new class. In preparation for a future removal of the old class it is recommended that you update recipes currently inheriting distro_features_check
to inherit features_check
instead.
The following classes have been removed:
distutils-base
: moved to meta-python2
distutils
: moved to meta-python2
libc-common
: merged into the glibc recipe as nothing else used it.
python-dir
: moved to meta-python2
pythonnative
: moved to meta-python2
setuptools
: moved to meta-python2
tinderclient
: dropped as it was obsolete.
Previously, recipes by tradition included both SHA256 and MD5 checksums for remotely fetched files in SRC_URI
, even though only one is actually mandated. However, the MD5 checksum does not add much given its inherent weakness; thus when a checksum fails only the SHA256 sum will now be printed. The md5sum will still be verified if it is specified.
The npm fetcher has been completely reworked in this release. The npm fetcher now only fetches the package source itself and no longer the dependencies; there is now also an npmsw fetcher which explicitly fetches the shrinkwrap file and the dependencies. This removes the slightly awkward NPM_LOCKDOWN
and NPM_SHRINKWRAP
variables which pointed to local files; the lockdown file is no longer needed at all. Additionally, the package name in npm://
entries in SRC_URI
is now specified using a package
parameter instead of the earlier name
which overlapped with the generic name
parameter. All recipes using the npm fetcher will need to be changed as a result.
An example of the new scheme:
SRC_URI = "npm://registry.npmjs.org;package=array-flatten;version=1.1.1 \ npmsw://${THISDIR}/npm-shrinkwrap.json"
Another example where the sources are fetched from git rather than an npm repository:
SRC_URI = "git://github.com/foo/bar.git;protocol=https \ npmsw://${THISDIR}/npm-shrinkwrap.json"
devtool and recipetool have also been updated to match with the npm fetcher changes. Other than producing working and more complete recipes for npm sources, there is also a minor change to the command line for devtool: the --fetch-dev
option has been renamed to --npm-dev
as it is npm-specific.
intltool
has been removed from packagegroup-core-sdk
as it is rarely needed to build modern software - gettext can do most of the things it used to be needed for. intltool
has also been removed from packagegroup-core-self-hosted
as it is not needed to for standard builds.
git: git-am
, git-difftool
, git-submodule
, and git-request-pull
are no longer perl-based, so are now installed with the main git
package instead of within git-perltools
.
The ldconfig
binary built as part of glibc has now been moved to its own ldconfig
package (note no glibc-
prefix). This package is in the RRECOMMENDS
of the main glibc
package if ldconfig
is present in DISTRO_FEATURES
.
libevent
now splits each shared library into its own package (as Debian does). Since these are shared libraries and will be pulled in through the normal shared library dependency handling, there should be no impact to existing configurations other than less unnecessary libraries being installed in some cases.
linux-firmware now has a new package for bcm4366c
and includes available NVRAM config files into the bcm43340
, bcm43362
, bcm43430
and bcm4356-pcie
packages.
harfbuzz
now splits the new libharfbuzz-subset.so
library into its own package to reduce the main package size in cases where libharfbuzz-subset.so
is not needed.
Warnings will now be shown at do_package_qa
time in the following circumstances:
A recipe installs .desktop
files containing MimeType
keys but does not inherit the new mime-xdg
class
A recipe installs .xml
files into ${datadir}/mime/packages
but does not inherit the mime
class
wic
image type now used instead of live
by default for x86¶
conf/machine/include/x86-base.inc
(inherited by most x86 machine configurations) now specifies wic
instead of live
by default in IMAGE_FSTYPES
. The live
image type will likely be removed in a future release so it is recommended that you use wic
instead.
The undocumented SRC_DISTRIBUTE_LICENSES
variable has now been removed in favour of a new AVAILABLE_LICENSES
variable which is dynamically set based upon license files found in ${COMMON_LICENSE_DIR}
and ${LICENSE_PATH}
.
The tune definition for big-endian microblaze machines is now microblaze
instead of microblazeeb
.
newlib
no longer has built-in syscalls. libgloss
should then provide the syscalls, crt0.o
and other functions that are no longer part of newlib
itself. If you are using TCLIBC = "newlib"
this now means that you must link applications with both newlib
and libgloss
, whereas before newlib
would run in many configurations by itself.
build/
build/buildhistory/
build/conf/local.conf
build/conf/bblayers.conf
build/cache/sanity_info
build/downloads/
build/sstate-cache/
build/tmp/
build/tmp/buildstats/
build/tmp/cache/
build/tmp/deploy/
build/tmp/deploy/deb/
build/tmp/deploy/rpm/
build/tmp/deploy/ipk/
build/tmp/deploy/licenses/
build/tmp/deploy/images/
build/tmp/deploy/sdk/
build/tmp/sstate-control/
build/tmp/sysroots-components/
build/tmp/sysroots/
build/tmp/stamps/
build/tmp/log/
build/tmp/work/
build/tmp/work/tunearch
/recipename
/version
/
build/tmp/work-shared/
meta/
meta/classes/
meta/conf/
meta/conf/machine/
meta/conf/distro/
meta/conf/machine-sdk/
meta/files/
meta/lib/
meta/recipes-bsp/
meta/recipes-connectivity/
meta/recipes-core/
meta/recipes-devtools/
meta/recipes-extended/
meta/recipes-gnome/
meta/recipes-graphics/
meta/recipes-kernel/
meta/recipes-lsb4/
meta/recipes-multimedia/
meta/recipes-rt/
meta/recipes-sato/
meta/recipes-support/
meta/site/
meta/recipes.txt
The Source Directory consists of numerous files, directories and subdirectories; understanding their locations and contents is key to using the Yocto Project effectively. This chapter describes the Source Directory and gives information about those files and directories.
For information on how to establish a local Source Directory on your development system, see the "Locating Yocto Project Source Files" section in the Yocto Project Development Tasks Manual.
This section describes the top-level components of the Source Directory.
bitbake/
¶This directory includes a copy of BitBake for ease of use. The copy usually matches the current stable BitBake release from the BitBake project. BitBake, a Metadata interpreter, reads the Yocto Project Metadata and runs the tasks defined by that data. Failures are usually caused by errors in your Metadata and not from BitBake itself; consequently, most users do not need to worry about BitBake.
When you run the bitbake
command, the
main BitBake executable (which resides in the
bitbake/bin/
directory) starts.
Sourcing the environment setup script (i.e.
oe-init-build-env
)
places the scripts/
and
bitbake/bin/
directories (in that order) into
the shell's PATH
environment variable.
For more information on BitBake, see the BitBake User Manual.
build/
¶
This directory contains user configuration files and the output
generated by the OpenEmbedded build system in its standard configuration where
the source tree is combined with the output.
The
Build Directory
is created initially when you source
the OpenEmbedded build environment setup script
(i.e.
oe-init-build-env
).
It is also possible to place output and configuration
files in a directory separate from the
Source Directory
by providing a directory name when you source
the setup script.
For information on separating output from your local
Source Directory files (commonly described as an "out of tree" build), see the
"oe-init-build-env
"
section.
documentation/
¶
This directory holds the source for the Yocto Project documentation
as well as templates and tools that allow you to generate PDF and HTML
versions of the manuals.
Each manual is contained in its own sub-folder;
for example, the files for this reference manual reside in
the ref-manual/
directory.
meta/
¶
This directory contains the minimal, underlying OpenEmbedded-Core metadata.
The directory holds recipes, common classes, and machine
configuration for strictly emulated targets (qemux86
,
qemuarm
, and so forth.)
meta-poky/
¶
Designed above the meta/
content, this directory
adds just enough metadata to define the Poky reference distribution.
meta-yocto-bsp/
¶This directory contains the Yocto Project reference hardware Board Support Packages (BSPs). For more information on BSPs, see the Yocto Project Board Support Package (BSP) Developer's Guide.
meta-selftest/
¶
This directory adds additional recipes and append files
used by the OpenEmbedded selftests to verify the behavior
of the build system.
You do not have to add this layer to your
bblayers.conf
file unless you want to run the
selftests.
meta-skeleton/
¶This directory contains template recipes for BSP and kernel development.
scripts/
¶
This directory contains various integration scripts that implement
extra functionality in the Yocto Project environment (e.g. QEMU scripts).
The oe-init-build-env
script prepends this directory to the shell's
PATH
environment variable.
The scripts
directory has useful scripts that assist in contributing
back to the Yocto Project, such as create-pull-request
and
send-pull-request
.
oe-init-build-env
¶
This script sets up the OpenEmbedded build environment.
Running this script with the source
command in
a shell makes changes to PATH
and sets other
core BitBake variables based on the current working directory.
You need to run an environment setup script before running BitBake
commands.
The script uses other scripts within the
scripts
directory to do the bulk of the work.
When you run this script, your Yocto Project environment is set up, a Build Directory is created, your working directory becomes the Build Directory, and you are presented with some simple suggestions as to what to do next, including a list of some possible targets to build. Here is an example:
$ source oe-init-build-env ### Shell environment set up for builds. ### You can now run 'bitbake <target>' Common targets are: core-image-minimal core-image-sato meta-toolchain meta-ide-support You can also run generated qemu images with a command like 'runqemu qemux86-64'
The default output of the oe-init-build-env
script
is from the conf-notes.txt
file, which is found in the
meta-poky
directory within the
Source Directory.
If you design a custom distribution, you can include your own version
of this configuration file to mention the targets defined by your
distribution.
See the
"Creating a Custom Template Configuration Directory"
section in the Yocto Project Development Tasks Manual for more
information.
By default, running this script without a Build Directory
argument creates the build/
directory
in your current working directory.
If you provide a Build Directory argument when you
source
the script, you direct the OpenEmbedded
build system to create a Build Directory of your choice.
For example, the following command creates a Build Directory named
mybuilds/
that is outside of the
Source Directory:
$ source oe-init-build-env ~/mybuilds
The OpenEmbedded build system uses the template configuration
files, which are found by default in the
meta-poky/conf/
directory in the
Source Directory.
See the
"Creating a Custom Template Configuration Directory"
section in the Yocto Project Development Tasks Manual for more
information.
oe-init-build-env
script
from a Source Directory that contains spaces in either the filenames
or directory names, the script returns an error indicating no such
file or directory.
Be sure to use a Source Directory free of names containing spaces.
LICENSE, README, and README.hardware
¶These files are standard top-level files.
build/
¶
The OpenEmbedded build system creates the
Build Directory
when you run the build environment setup script
oe-init-build-env
.
If you do not give the Build Directory a specific name when you run
the setup script, the name defaults to build/
.
For subsequent parsing and processing, the name of the Build
directory is available via the
TOPDIR
variable.
build/buildhistory/
¶
The OpenEmbedded build system creates this directory when you
enable build history via the buildhistory
class file.
The directory organizes build information into image, packages, and
SDK subdirectories.
For information on the build history feature, see the
"Maintaining Build Output Quality"
section in the Yocto Project Development Tasks Manual.
build/conf/local.conf
¶
This configuration file contains all the local user configurations
for your build environment.
The local.conf
file contains documentation on
the various configuration options.
Any variable set here overrides any variable set elsewhere within
the environment unless that variable is hard-coded within a file
(e.g. by using '=' instead of '?=').
Some variables are hard-coded for various reasons but such
variables are relatively rare.
At a minimum, you would normally edit this file to select the target
MACHINE
,
which package types you wish to use
(PACKAGE_CLASSES
),
and the location from which you want to access downloaded files
(DL_DIR
).
If local.conf
is not present when you
start the build, the OpenEmbedded build system creates it from
local.conf.sample
when
you source
the top-level build environment
setup script
oe-init-build-env
.
The source local.conf.sample
file used
depends on the $TEMPLATECONF
script variable,
which defaults to meta-poky/conf/
when you are building from the Yocto Project development
environment, and to meta/conf/
when
you are building from the OpenEmbedded-Core environment.
Because the script variable points to the source of the
local.conf.sample
file, this implies that
you can configure your build environment from any layer by setting
the variable in the top-level build environment setup script as
follows:
TEMPLATECONF=your_layer
/conf
Once the build process gets the sample file, it uses
sed
to substitute final
${
OEROOT
}
values for all ##OEROOT##
values.
TEMPLATECONF
variable
is used by looking at the
scripts/oe-setup-builddir
script in the
Source Directory.
You can find the Yocto Project version of the
local.conf.sample
file in the
meta-poky/conf
directory.
build/conf/bblayers.conf
¶
This configuration file defines
layers,
which are directory trees, traversed (or walked) by BitBake.
The bblayers.conf
file uses the
BBLAYERS
variable to list the layers BitBake tries to find.
If bblayers.conf
is not present when you
start the build, the OpenEmbedded build system creates it from
bblayers.conf.sample
when
you source
the top-level build environment
setup script (i.e.
oe-init-build-env
).
As with the local.conf
file,
the source bblayers.conf.sample
file used
depends on the $TEMPLATECONF
script variable,
which defaults to meta-poky/conf/
when you are building from the Yocto Project development
environment, and to meta/conf/
when
you are building from the OpenEmbedded-Core environment.
Because the script variable points to the source of the
bblayers.conf.sample
file, this implies that
you can base your build from any layer by setting the variable in
the top-level build environment setup script as follows:
TEMPLATECONF=your_layer
/conf
Once the build process gets the sample file, it uses
sed
to substitute final
${
OEROOT
}
values for all ##OEROOT##
values.
TEMPLATECONF
variable
scripts/oe-setup-builddir
script in the
Source Directory.
You can find the Yocto Project version of the
bblayers.conf.sample
file in the
meta-poky/conf/
directory.
build/cache/sanity_info
¶This file indicates the state of the sanity checks and is created during the build.
build/downloads/
¶
This directory contains downloaded upstream source tarballs.
You can reuse the directory for multiple builds or move
the directory to another location.
You can control the location of this directory through the
DL_DIR
variable.
build/sstate-cache/
¶
This directory contains the shared state cache.
You can reuse the directory for multiple builds or move
the directory to another location.
You can control the location of this directory through the
SSTATE_DIR
variable.
build/tmp/
¶
The OpenEmbedded build system creates and uses this directory
for all the build system's output.
The
TMPDIR
variable points to this directory.
BitBake creates this directory if it does not exist.
As a last resort, to clean up a build and start it from scratch
(other than the downloads), you can remove everything in the
tmp
directory or get rid of the
directory completely.
If you do, you should also completely remove the
build/sstate-cache
directory.
build/tmp/buildstats/
¶This directory stores the build statistics.
build/tmp/cache/
¶
When BitBake parses the metadata (recipes and configuration files),
it caches the results in build/tmp/cache/
to speed up future builds.
The results are stored on a per-machine basis.
During subsequent builds, BitBake checks each recipe (together with, for example, any files included or appended to it) to see if they have been modified. Changes can be detected, for example, through file modification time (mtime) changes and hashing of file contents. If no changes to the file are detected, then the parsed result stored in the cache is reused. If the file has changed, it is reparsed.
build/tmp/deploy/
¶
This directory contains any "end result" output from the
OpenEmbedded build process.
The DEPLOY_DIR
variable points to this directory.
For more detail on the contents of the deploy
directory, see the
"Images"
and
"Application Development SDK"
sections in the Yocto Project Overview and Concepts Manual.
build/tmp/deploy/deb/
¶
This directory receives any .deb
packages produced by
the build process.
The packages are sorted into feeds for different architecture types.
build/tmp/deploy/rpm/
¶
This directory receives any .rpm
packages produced by
the build process.
The packages are sorted into feeds for different architecture types.
build/tmp/deploy/ipk/
¶
This directory receives .ipk
packages produced by
the build process.
build/tmp/deploy/licenses/
¶
This directory receives package licensing information.
For example, the directory contains sub-directories for bash
,
busybox
, and glibc
(among others) that in turn
contain appropriate COPYING
license files with other licensing information.
For information on licensing, see the
"Maintaining Open Source License Compliance During Your Product's Lifecycle"
section in the Yocto Project Development Tasks Manual.
build/tmp/deploy/images/
¶This directory is populated with the basic output objects of the build (think of them as the "generated artifacts" of the build process), including things like the boot loader image, kernel, root filesystem and more. If you want to flash the resulting image from a build onto a device, look here for the necessary components.
Be careful when deleting files in this directory.
You can safely delete old images from this directory (e.g.
core-image-*
).
However, the kernel (*zImage*
, *uImage*
, etc.),
bootloader and other supplementary files might be deployed here prior to building an
image.
Because these files are not directly produced from the image, if you
delete them they will not be automatically re-created when you build the image again.
If you do accidentally delete files here, you will need to force them to be re-created. In order to do that, you will need to know the target that produced them. For example, these commands rebuild and re-create the kernel files:
$ bitbake -c clean virtual/kernel $ bitbake virtual/kernel
build/tmp/deploy/sdk/
¶The OpenEmbedded build system creates this directory to hold toolchain installer scripts which, when executed, install the sysroot that matches your target hardware. You can find out more about these installers in the "Building an SDK Installer" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
build/tmp/sstate-control/
¶The OpenEmbedded build system uses this directory for the shared state manifest files. The shared state code uses these files to record the files installed by each sstate task so that the files can be removed when cleaning the recipe or when a newer version is about to be installed. The build system also uses the manifests to detect and produce a warning when files from one task are overwriting those from another.
build/tmp/sysroots-components/
¶
This directory is the location of the sysroot contents that the
task
do_prepare_recipe_sysroot
links or copies into the recipe-specific sysroot for each
recipe listed in
DEPENDS
.
Population of this directory is handled through shared state, while
the path is specified by the
COMPONENTS_DIR
variable. Apart from a few unusual circumstances, handling of the
sysroots-components
directory should be
automatic, and recipes should not directly reference
build/tmp/sysroots-components
.
build/tmp/sysroots/
¶
Previous versions of the OpenEmbedded build system used to
create a global shared sysroot per machine along with a native
sysroot.
Beginning with the 3.1.2 version of the Yocto Project,
sysroots exist in recipe-specific
WORKDIR
directories.
Thus, the build/tmp/sysroots/
directory
is unused.
build/tmp/sysroots/
directory
can still be populated using the
bitbake build-sysroots
command and can
be used for compatibility in some cases.
However, in general it is not recommended to populate
this directory.
Individual recipe-specific sysroots should be used.
build/tmp/stamps/
¶This directory holds information that BitBake uses for accounting purposes to track what tasks have run and when they have run. The directory is sub-divided by architecture, package name, and version. Following is an example:
stamps/all-poky-linux/distcc-config/1.0-r0.do_build-2fdd....2do
Although the files in the directory are empty of data, BitBake uses the filenames and timestamps for tracking purposes.
For information on how BitBake uses stamp files to determine if a task should be rerun, see the "Stamp Files and the Rerunning of Tasks" section in the Yocto Project Overview and Concepts Manual.
build/tmp/log/
¶
This directory contains general logs that are not otherwise placed using the
package's WORKDIR
.
Examples of logs are the output from the
do_check_pkg
or
do_distro_check
tasks.
Running a build does not necessarily mean this directory is created.
build/tmp/work/
¶
This directory contains architecture-specific work sub-directories
for packages built by BitBake.
All tasks execute from the appropriate work directory.
For example, the source for a particular package is unpacked,
patched, configured and compiled all within its own work directory.
Within the work directory, organization is based on the package group
and version for which the source is being compiled
as defined by the
WORKDIR
.
It is worth considering the structure of a typical work directory.
As an example, consider linux-yocto-kernel-3.0
on the machine qemux86
built within the Yocto Project.
For this package, a work directory of
tmp/work/qemux86-poky-linux/linux-yocto/3.0+git1+<.....>
,
referred to as the WORKDIR
, is created.
Within this directory, the source is unpacked to
linux-qemux86-standard-build
and then patched by Quilt.
(See the
"Using Quilt in Your Workflow"
section in the Yocto Project Development Tasks Manual for more
information.)
Within the linux-qemux86-standard-build
directory,
standard Quilt directories linux-3.0/patches
and linux-3.0/.pc
are created,
and standard Quilt commands can be used.
There are other directories generated within WORKDIR
.
The most important directory is WORKDIR/temp/
,
which has log files for each task (log.do_*.pid
)
and contains the scripts BitBake runs for each task
(run.do_*.pid
).
The WORKDIR/image/
directory is where "make
install" places its output that is then split into sub-packages
within WORKDIR/packages-split/
.
build/tmp/work/tunearch
/recipename
/version
/
¶
The recipe work directory - ${WORKDIR}
.
As described earlier in the
"build/tmp/sysroots/
"
section, beginning with the 3.1.2 release of the Yocto
Project, the OpenEmbedded build system builds each recipe in its
own work directory (i.e.
WORKDIR
).
The path to the work directory is constructed using the
architecture of the given build (e.g.
TUNE_PKGARCH
,
MACHINE_ARCH
,
or "allarch"), the recipe name, and the version of the recipe (i.e.
PE
:
PV
-
PR
).
A number of key subdirectories exist within each recipe work directory:
${WORKDIR}/temp
:
Contains the log files of each task executed for this
recipe, the "run" files for each executed task, which
contain the code run, and a
log.task_order
file, which lists the
order in which tasks were executed.
${WORKDIR}/image
:
Contains the output of the
do_install
task, which corresponds to the
${
D
}
variable in that task.
${WORKDIR}/pseudo
:
Contains the pseudo database and log for any tasks executed
under pseudo for the recipe.
${WORKDIR}/sysroot-destdir
:
Contains the output of the
do_populate_sysroot
task.
${WORKDIR}/package
:
Contains the output of the
do_package
task before the output is split into individual packages.
${WORKDIR}/packages-split
:
Contains the output of the do_package
task after the output has been split into individual
packages.
Subdirectories exist for each individual package created
by the recipe.
${WORKDIR}/recipe-sysroot
:
A directory populated with the target dependencies of the
recipe.
This directory looks like the target filesystem and
contains libraries that the recipe might need to link
against (e.g. the C library).
${WORKDIR}/recipe-sysroot-native
:
A directory populated with the native dependencies of the
recipe.
This directory contains the tools the recipe needs to build
(e.g. the compiler, Autoconf, libtool, and so forth).
${WORKDIR}/build
:
This subdirectory applies only to recipes that support
builds where the source is separate from the
build artifacts.
The OpenEmbedded build system uses this directory as a
separate build directory (i.e.
${
B
}
).
build/tmp/work-shared/
¶
For efficiency, the OpenEmbedded build system creates and uses
this directory to hold recipes that share a work directory with
other recipes.
In practice, this is only used for gcc
and its variants (e.g. gcc-cross
,
libgcc
, gcc-runtime
,
and so forth).
meta/
¶As mentioned previously, Metadata is the core of the Yocto Project. Metadata has several important subdivisions:
meta/classes/
¶
This directory contains the *.bbclass
files.
Class files are used to abstract common code so it can be reused by multiple
packages.
Every package inherits the base.bbclass
file.
Examples of other important classes are autotools.bbclass
, which
in theory allows any Autotool-enabled package to work with the Yocto Project with minimal effort.
Another example is kernel.bbclass
that contains common code and functions
for working with the Linux kernel.
Functions like image generation or packaging also have their specific class files
such as image.bbclass
, rootfs_*.bbclass
and
package*.bbclass
.
For reference information on classes, see the "Classes" chapter.
meta/conf/
¶
This directory contains the core set of configuration files that start from
bitbake.conf
and from which all other configuration
files are included.
See the include statements at the end of the
bitbake.conf
file and you will note that even
local.conf
is loaded from there.
While bitbake.conf
sets up the defaults, you can often override
these by using the (local.conf
) file, machine file or
the distribution configuration file.
meta/conf/machine/
¶
This directory contains all the machine configuration files.
If you set MACHINE = "qemux86"
,
the OpenEmbedded build system looks for a qemux86.conf
file in this
directory.
The include
directory contains various data common to multiple machines.
If you want to add support for a new machine to the Yocto Project, look in this directory.
meta/conf/distro/
¶
The contents of this directory controls any distribution-specific
configurations.
For the Yocto Project, the defaultsetup.conf
is the main file here.
This directory includes the versions and the
SRCDATE
definitions for applications that are configured here.
An example of an alternative configuration might be poky-bleeding.conf
.
Although this file mainly inherits its configuration from Poky.
meta/conf/machine-sdk/
¶
The OpenEmbedded build system searches this directory for
configuration files that correspond to the value of
SDKMACHINE
.
By default, 32-bit and 64-bit x86 files ship with the Yocto
Project that support some SDK hosts.
However, it is possible to extend that support to other SDK hosts
by adding additional configuration files in this subdirectory
within another layer.
meta/files/
¶This directory contains common license files and several text files used by the build system. The text files contain minimal device information and lists of files and directories with known permissions.
meta/lib/
¶This directory contains OpenEmbedded Python library code used during the build process.
meta/recipes-bsp/
¶This directory contains anything linking to specific hardware or hardware configuration information such as "u-boot" and "grub".
meta/recipes-connectivity/
¶This directory contains libraries and applications related to communication with other devices.
meta/recipes-core/
¶This directory contains what is needed to build a basic working Linux image including commonly used dependencies.
meta/recipes-devtools/
¶This directory contains tools that are primarily used by the build system. The tools, however, can also be used on targets.
meta/recipes-extended/
¶This directory contains non-essential applications that add features compared to the alternatives in core. You might need this directory for full tool functionality or for Linux Standard Base (LSB) compliance.
meta/recipes-gnome/
¶This directory contains all things related to the GTK+ application framework.
meta/recipes-graphics/
¶This directory contains X and other graphically related system libraries.
meta/recipes-kernel/
¶This directory contains the kernel and generic applications and libraries that have strong kernel dependencies.
meta/recipes-lsb4/
¶This directory contains recipes specifically added to support the Linux Standard Base (LSB) version 4.x.
meta/recipes-multimedia/
¶This directory contains codecs and support utilities for audio, images and video.
meta/recipes-rt/
¶
This directory contains package and image recipes for using and testing
the PREEMPT_RT
kernel.
meta/recipes-sato/
¶This directory contains the Sato demo/reference UI/UX and its associated applications and configuration data.
meta/recipes-support/
¶This directory contains recipes used by other recipes, but that are not directly included in images (i.e. dependencies of other recipes).
meta/site/
¶This directory contains a list of cached results for various architectures. Because certain "autoconf" test results cannot be determined when cross-compiling due to the tests not able to run on a live system, the information in this directory is passed to "autoconf" for the various architectures.
meta/recipes.txt
¶
This file is a description of the contents of recipes-*
.
allarch.bbclass
archiver.bbclass
autotools*.bbclass
base.bbclass
bash-completion.bbclass
bin_package.bbclass
binconfig.bbclass
binconfig-disabled.bbclass
blacklist.bbclass
buildhistory.bbclass
buildstats.bbclass
buildstats-summary.bbclass
ccache.bbclass
chrpath.bbclass
clutter.bbclass
cmake.bbclass
cml1.bbclass
compress_doc.bbclass
copyleft_compliance.bbclass
copyleft_filter.bbclass
core-image.bbclass
cpan*.bbclass
cross.bbclass
cross-canadian.bbclass
crosssdk.bbclass
debian.bbclass
deploy.bbclass
devshell.bbclass
devupstream.bbclass
distro_features_check.bbclass
distutils*.bbclass
distutils3*.bbclass
externalsrc.bbclass
extrausers.bbclass
fontcache.bbclass
fs-uuid.bbclass
gconf.bbclass
gettext.bbclass
gnomebase.bbclass
gobject-introspection.bbclass
grub-efi.bbclass
gsettings.bbclass
gtk-doc.bbclass
gtk-icon-cache.bbclass
gtk-immodules-cache.bbclass
gzipnative.bbclass
icecc.bbclass
image.bbclass
image-buildinfo.bbclass
image_types.bbclass
image-live.bbclass
image-mklibs.bbclass
image-prelink.bbclass
insane.bbclass
insserv.bbclass
kernel.bbclass
kernel-arch.bbclass
kernel-devicetree.bbclass
kernel-fitimage.bbclass
kernel-grub.bbclass
kernel-module-split.bbclass
kernel-uboot.bbclass
kernel-uimage.bbclass
kernel-yocto.bbclass
kernelsrc.bbclass
lib_package.bbclass
libc*.bbclass
license.bbclass
linux-kernel-base.bbclass
linuxloader.bbclass
logging.bbclass
meta.bbclass
metadata_scm.bbclass
migrate_localcount.bbclass
mime.bbclass
mirrors.bbclass
module.bbclass
module-base.bbclass
multilib*.bbclass
native.bbclass
nativesdk.bbclass
nopackages.bbclass
npm.bbclass
oelint.bbclass
own-mirrors.bbclass
package.bbclass
package_deb.bbclass
package_ipk.bbclass
package_rpm.bbclass
package_tar.bbclass
packagedata.bbclass
packagegroup.bbclass
patch.bbclass
perlnative.bbclass
pixbufcache.bbclass
pkgconfig.bbclass
populate_sdk.bbclass
populate_sdk_*.bbclass
prexport.bbclass
primport.bbclass
prserv.bbclass
ptest.bbclass
ptest-gnome.bbclass
python-dir.bbclass
python3native.bbclass
pythonnative.bbclass
qemu.bbclass
recipe_sanity.bbclass
relocatable.bbclass
remove-libtool.bbclass
report-error.bbclass
rm_work.bbclass
rootfs*.bbclass
sanity.bbclass
scons.bbclass
sdl.bbclass
setuptools.bbclass
setuptools3.bbclass
sign_rpm.bbclass
sip.bbclass
siteconfig.bbclass
siteinfo.bbclass
spdx.bbclass
sstate.bbclass
staging.bbclass
syslinux.bbclass
systemd.bbclass
systemd-boot.bbclass
terminal.bbclass
testimage*.bbclass
testsdk.bbclass
texinfo.bbclass
tinderclient.bbclass
toaster.bbclass
toolchain-scripts.bbclass
typecheck.bbclass
uboot-config.bbclass
uninative.bbclass
update-alternatives.bbclass
update-rc.d.bbclass
useradd*.bbclass
utility-tasks.bbclass
utils.bbclass
vala.bbclass
waf.bbclass
Class files are used to abstract common functionality and share it amongst
multiple recipe (.bb
) files.
To use a class file, you simply make sure the recipe inherits the class.
In most cases, when a recipe inherits a class it is enough to enable its
features.
There are cases, however, where in the recipe you might need to set
variables or override some default behavior.
Any Metadata usually
found in a recipe can also be placed in a class file.
Class files are identified by the extension .bbclass
and are usually placed in a classes/
directory beneath
the meta*/
directory found in the
Source Directory.
Class files can also be pointed to by
BUILDDIR
(e.g. build/
) in the same way as
.conf
files in the conf
directory.
Class files are searched for in
BBPATH
using the same method by which .conf
files are
searched.
This chapter discusses only the most useful and important classes.
Other classes do exist within the meta/classes
directory in the Source Directory.
You can reference the .bbclass
files directly
for more information.
allarch.bbclass
¶
The allarch
class is inherited
by recipes that do not produce architecture-specific output.
The class disables functionality that is normally needed for recipes
that produce executable binaries (such as building the cross-compiler
and a C library as pre-requisites, and splitting out of debug symbols
during packaging).
Unlike some distro recipes (e.g. Debian), OpenEmbedded recipes
that produce packages that depend on tunings through use of the
RDEPENDS
and
TUNE_PKGARCH
variables, should never be configured for all architectures
using allarch
.
This is the case even if the recipes do not produce
architecture-specific output.
Configuring such recipes for all architectures causes the
do_package_write_*
tasks to have different signatures for the machines with different
tunings.
Additionally, unnecessary rebuilds occur every time an
image for a different MACHINE
is built
even when the recipe never changes.
By default, all recipes inherit the
base
and
package
classes, which enable functionality
needed for recipes that produce executable output.
If your recipe, for example, only produces packages that contain
configuration files, media files, or scripts (e.g. Python and Perl),
then it should inherit the allarch
class.
archiver.bbclass
¶
The archiver
class supports releasing
source code and other materials with the binaries.
For more details on the source archiver, see the
"Maintaining Open Source License Compliance During Your Product's Lifecycle"
section in the Yocto Project Development Tasks Manual.
You can also see the
ARCHIVER_MODE
variable for information about the variable flags (varflags)
that help control archive creation.
autotools*.bbclass
¶
The autotools*
classes support Autotooled
packages.
The autoconf
, automake
,
and libtool
packages bring standardization.
This class defines a set of tasks (e.g.
configure
, compile
and
so forth) that
work for all Autotooled packages.
It should usually be enough to define a few standard variables
and then simply inherit autotools
.
These classes can also work with software that emulates Autotools.
For more information, see the
"Autotooled Package"
section in the Yocto Project Development Tasks Manual.
By default, the autotools*
classes
use out-of-tree builds (i.e.
autotools.bbclass
building with
B != S
).
If the software being built by a recipe does not support
using out-of-tree builds, you should have the recipe inherit the
autotools-brokensep
class.
The autotools-brokensep
class behaves the same
as the autotools
class but builds with
B
==
S
.
This method is useful when out-of-tree build support is either not
present or is broken.
It's useful to have some idea of how the tasks defined by
the autotools*
classes work and what they do
behind the scenes.
do_configure
-
Regenerates the
configure script (using autoreconf
) and
then launches it with a standard set of arguments used during
cross-compilation.
You can pass additional parameters to
configure
through the
EXTRA_OECONF
or
PACKAGECONFIG_CONFARGS
variables.
do_compile
-
Runs make
with arguments that specify the
compiler and linker.
You can pass additional arguments through
the EXTRA_OEMAKE
variable.
do_install
-
Runs make install
and passes in
${
D
}
as DESTDIR
.
base.bbclass
¶
The base
class is special in that every
.bb
file implicitly inherits the class.
This class contains definitions for standard basic
tasks such as fetching, unpacking, configuring (empty by default),
compiling (runs any Makefile
present), installing
(empty by default) and packaging (empty by default).
These classes are often overridden or extended by other classes
such as the
autotools
class or the
package
class.
The class also contains some commonly used functions such as
oe_runmake
, which runs
make
with the arguments specified in
EXTRA_OEMAKE
variable as well as the arguments passed directly to
oe_runmake
.
bash-completion.bbclass
¶Sets up packaging and dependencies appropriate for recipes that build software that includes bash-completion data.
bin_package.bbclass
¶
The bin_package
class is a
helper class for recipes that extract the contents of a binary package
(e.g. an RPM) and install those contents rather than building the
binary from source.
The binary package is extracted and new packages in the configured
output package format are created.
Extraction and installation of proprietary binaries is a good example
use for this class.
git://
), the "subpath" parameter limits
the checkout to a specific subpath of the tree.
Here is an example where ${BP}
is used so that
the files are extracted into the subdirectory expected by the
default value of
S
:
SRC_URI = "git://example.com/downloads/somepackage.rpm;subpath=${BP}"See the "Fetchers" section in the BitBake User Manual for more information on supported BitBake Fetchers.
binconfig.bbclass
¶
The binconfig
class helps to correct paths in
shell scripts.
Before pkg-config
had become widespread, libraries
shipped shell scripts to give information about the libraries and
include paths needed to build software (usually named
LIBNAME-config
).
This class assists any recipe using such scripts.
During staging, the OpenEmbedded build system installs such scripts
into the sysroots/
directory.
Inheriting this class results in all paths in these scripts being
changed to point into the sysroots/
directory so
that all builds that use the script use the correct directories
for the cross compiling layout.
See the
BINCONFIG_GLOB
variable for more information.
binconfig-disabled.bbclass
¶
An alternative version of the
binconfig
class, which disables binary configuration scripts by making them
return an error in favor of using pkg-config
to query the information.
The scripts to be disabled should be specified using the
BINCONFIG
variable within the recipe inheriting the class.
blacklist.bbclass
¶
The blacklist
class prevents
the OpenEmbedded build system from building specific recipes
(blacklists them).
To use this class, inherit the class globally and set
PNBLACKLIST
for each recipe you wish to blacklist.
Specify the PN
value as a variable flag (varflag) and provide a reason, which is
reported, if the package is requested to be built as the value.
For example, if you want to blacklist a recipe called "exoticware",
you add the following to your local.conf
or distribution configuration:
INHERIT += "blacklist" PNBLACKLIST[exoticware] = "Not supported by our organization."
buildhistory.bbclass
¶
The buildhistory
class records a
history of build output metadata, which can be used to detect possible
regressions as well as used for analysis of the build output.
For more information on using Build History, see the
"Maintaining Build Output Quality"
section in the Yocto Project Development Tasks Manual.
buildstats.bbclass
¶
The buildstats
class records
performance statistics about each task executed during the build
(e.g. elapsed time, CPU usage, and I/O usage).
When you use this class, the output goes into the
BUILDSTATS_BASE
directory, which defaults to ${TMPDIR}/buildstats/
.
You can analyze the elapsed time using
scripts/pybootchartgui/pybootchartgui.py
, which
produces a cascading chart of the entire build process and can be
useful for highlighting bottlenecks.
Collecting build statistics is enabled by default through the
USER_CLASSES
variable from your local.conf
file.
Consequently, you do not have to do anything to enable the class.
However, if you want to disable the class, simply remove "buildstats"
from the USER_CLASSES
list.
buildstats-summary.bbclass
¶
When inherited globally, prints statistics at the end of the build
on sstate re-use.
In order to function, this class requires the
buildstats
class be enabled.
ccache.bbclass
¶
The ccache
class enables the C/C++ Compiler Cache
for the build.
This class is used to give a minor performance boost during the build.
However, using the class can lead to unexpected side-effects.
Thus, it is recommended that you do not use this class.
See http://ccache.samba.org/ for information on
the C/C++ Compiler Cache.
chrpath.bbclass
¶
The chrpath
class
is a wrapper around the "chrpath" utility, which is used during the
build process for nativesdk
,
cross
, and
cross-canadian
recipes to change
RPATH
records within binaries in order to make
them relocatable.
clutter.bbclass
¶
The clutter
class consolidates the
major and minor version naming and other common items used by Clutter
and related recipes.
cmake.bbclass
¶
The cmake
class allows for recipes that need to
build software using the
CMake build system.
You can use the
EXTRA_OECMAKE
variable to specify additional configuration options to be passed
using the cmake
command line.
On the occasion that you would be installing custom CMake toolchain
files supplied by the application being built, you should install them
to the preferred CMake Module directory:
${D}${datadir}/cmake/
Modules during
do_install
.
cml1.bbclass
¶
The cml1
class provides basic support for the
Linux kernel style build configuration system.
compress_doc.bbclass
¶
Enables compression for man pages and info pages.
This class is intended to be inherited globally.
The default compression mechanism is gz (gzip) but you can
select an alternative mechanism by setting the
DOC_COMPRESS
variable.
copyleft_compliance.bbclass
¶
The copyleft_compliance
class
preserves source code for the purposes of license compliance.
This class is an alternative to the archiver
class and is still used by some users even though it has been
deprecated in favor of the
archiver
class.
copyleft_filter.bbclass
¶
A class used by the
archiver
and
copyleft_compliance
classes for filtering licenses.
The copyleft_filter
class is an internal class
and is not intended to be used directly.
core-image.bbclass
¶
The core-image
class
provides common definitions for the
core-image-*
image recipes, such as support for
additional
IMAGE_FEATURES
.
cpan*.bbclass
¶
The cpan*
classes support Perl modules.
Recipes for Perl modules are simple. These recipes usually only need to point to the source's archive and then inherit the proper class file. Building is split into two methods depending on which method the module authors used.
Modules that use old
Makefile.PL
-based build system require
cpan.bbclass
in their recipes.
Modules that use
Build.PL
-based build system require
using cpan_build.bbclass
in their recipes.
Both build methods inherit the cpan-base
class
for basic Perl support.
cross.bbclass
¶
The cross
class provides support for the recipes
that build the cross-compilation tools.
cross-canadian.bbclass
¶
The cross-canadian
class
provides support for the recipes that build the Canadian
Cross-compilation tools for SDKs.
See the
"Cross-Development Toolchain Generation"
section in the Yocto Project Overview and Concepts Manual for more
discussion on these cross-compilation tools.
crosssdk.bbclass
¶
The crosssdk
class
provides support for the recipes that build the cross-compilation
tools used for building SDKs.
See the
"Cross-Development Toolchain Generation"
section in the Yocto Project Overview and Concepts Manual for more
discussion on these cross-compilation tools.
debian.bbclass
¶
The debian
class renames output packages so that
they follow the Debian naming policy (i.e. glibc
becomes libc6
and glibc-devel
becomes libc6-dev
.)
Renaming includes the library name and version as part of the package
name.
If a recipe creates packages for multiple libraries
(shared object files of .so
type), use the
LEAD_SONAME
variable in the recipe to specify the library on which to apply the
naming scheme.
deploy.bbclass
¶
The deploy
class handles deploying files
to the
DEPLOY_DIR_IMAGE
directory.
The main function of this class is to allow the deploy step to be
accelerated by shared state.
Recipes that inherit this class should define their own
do_deploy
function to copy the files to be deployed to
DEPLOYDIR
,
and use addtask
to add the task at the appropriate
place, which is usually after
do_compile
or
do_install
.
The class then takes care of staging the files from
DEPLOYDIR
to
DEPLOY_DIR_IMAGE
.
devshell.bbclass
¶
The devshell
class adds the
do_devshell
task.
Distribution policy dictates whether to include this class.
See the
"Using a Development Shell" section
in the Yocto Project Development Tasks Manual for more information about
using devshell
.
devupstream.bbclass
¶
The devupstream
class uses
BBCLASSEXTEND
to add a variant of the recipe that fetches from an alternative URI
(e.g. Git) instead of a tarball.
Following is an example:
BBCLASSEXTEND = "devupstream:target" SRC_URI_class-devupstream = "git://git.example.com/example" SRCREV_class-devupstream = "abcd1234"
Adding the above statements to your recipe creates a variant that has
DEFAULT_PREFERENCE
set to "-1".
Consequently, you need to select the variant of the recipe to use it.
Any development-specific adjustments can be done by using the
class-devupstream
override.
Here is an example:
DEPENDS_append_class-devupstream = " gperf-native" do_configure_prepend_class-devupstream() { touch ${S}/README }
The class currently only supports creating a development variant of
the target recipe, not native
or
nativesdk
variants.
The BBCLASSEXTEND
syntax
(i.e. devupstream:target
) provides support for
native
and nativesdk
variants.
Consequently, this functionality can be added in a future release.
Support for other version control systems such as Subversion is
limited due to BitBake's automatic fetch dependencies (e.g.
subversion-native
).
distro_features_check.bbclass
¶
The distro_features_check
class
allows individual recipes to check for required and conflicting
DISTRO_FEATURES
.
This class provides support for the
REQUIRED_DISTRO_FEATURES
and
CONFLICT_DISTRO_FEATURES
variables.
If any conditions specified in the recipe using the above variables are
not met, the recipe will be skipped.
distutils*.bbclass
¶
The distutils*
classes support recipes for Python
version 2.x extensions, which are simple.
These recipes usually only need to point to the source's archive and
then inherit the proper class.
Building is split into two methods depending on which method the
module authors used.
Extensions that use an Autotools-based build system
require Autotools and the classes based on
distutils
in their recipes.
Extensions that use build systems based on
distutils
require
the distutils
class in their recipes.
Extensions that use build systems based on
setuptools
require the
setuptools
class in their recipes.
The distutils-common-base
class is required by
some of the distutils*
classes to provide common
Python2 support.
distutils3*.bbclass
¶
The distutils3*
classes support recipes for Python
version 3.x extensions, which are simple.
These recipes usually only need to point to the source's archive and
then inherit the proper class.
Building is split into three methods depending on which method the
module authors used.
Extensions that use an Autotools-based build system
require Autotools and
distutils
-based classes in their recipes.
Extensions that use
distutils
-based build systems require
the distutils
class in their recipes.
Extensions that use build systems based on
setuptools3
require the
setuptools3
class in their recipes.
The distutils3*
classes either inherit their
corresponding distutils*
class or replicate them
using a Python3 version instead (e.g.
distutils3-base
inherits
distutils-common-base
, which is the same as
distutils-base
but inherits
python3native
instead of
pythonnative
).
externalsrc.bbclass
¶
The externalsrc
class supports building software
from source code that is external to the OpenEmbedded build system.
Building software from an external source tree means that the build
system's normal fetch, unpack, and patch process is not used.
By default, the OpenEmbedded build system uses the
S
and
B
variables to
locate unpacked recipe source code and to build it, respectively.
When your recipe inherits the externalsrc
class,
you use the
EXTERNALSRC
and
EXTERNALSRC_BUILD
variables to ultimately define S
and
B
.
By default, this class expects the source code to support recipe builds
that use the B
variable to point to the directory in which the OpenEmbedded build
system places the generated objects built from the recipes.
By default, the B
directory is set to the
following, which is separate from the source directory
(S
):
${WORKDIR}/${BPN}/{PV}/
See these variables for more information:
WORKDIR
,
BPN
, and
PV
,
For more information on the
externalsrc
class, see the comments in
meta/classes/externalsrc.bbclass
in the
Source Directory.
For information on how to use the externalsrc
class, see the
"Building Software from an External Source"
section in the Yocto Project Development Tasks Manual.
extrausers.bbclass
¶
The extrausers
class allows
additional user and group configuration to be applied at the image
level.
Inheriting this class either globally or from an image recipe allows
additional user and group operations to be performed using the
EXTRA_USERS_PARAMS
variable.
extrausers
class are not tied to a specific
recipe outside of the recipe for the image.
Thus, the operations can be performed across the image as a whole.
Use the
useradd
class to add user and group configuration to a specific recipe.
Here is an example that uses this class in an image recipe:
inherit extrausers EXTRA_USERS_PARAMS = "\ useradd -p '' tester; \ groupadd developers; \ userdel nobody; \ groupdel -g video; \ groupmod -g 1020 developers; \ usermod -s /bin/sh tester; \ "
Here is an example that adds two users named "tester-jim" and "tester-sue" and assigns passwords:
inherit extrausers EXTRA_USERS_PARAMS = "\ useradd -P tester01 tester-jim; \ useradd -P tester01 tester-sue; \ "
Finally, here is an example that sets the root password to "1876*18":
inherit extrausers EXTRA_USERS_PARAMS = "\ usermod -P 1876*18 root; \ "
fontcache.bbclass
¶
The fontcache
class generates the
proper post-install and post-remove (postinst and postrm)
scriptlets for font packages.
These scriptlets call fc-cache
(part of
Fontconfig
) to add the fonts to the font
information cache.
Since the cache files are architecture-specific,
fc-cache
runs using QEMU if the postinst
scriptlets need to be run on the build host during image creation.
If the fonts being installed are in packages other than the main
package, set
FONT_PACKAGES
to specify the packages containing the fonts.
fs-uuid.bbclass
¶
The fs-uuid
class extracts UUID from
${
ROOTFS
}
,
which must have been built by the time that this function gets called.
The fs-uuid
class only works on
ext
file systems and depends on
tune2fs
.
gconf.bbclass
¶
The gconf
class provides common
functionality for recipes that need to install GConf schemas.
The schemas will be put into a separate package
(${
PN
}-gconf
)
that is created automatically when this class is inherited.
This package uses the appropriate post-install and post-remove
(postinst/postrm) scriptlets to register and unregister the schemas
in the target image.
gettext.bbclass
¶
The gettext
class provides support for
building software that uses the GNU gettext
internationalization and localization system.
All recipes building software that use
gettext
should inherit this class.
gnomebase.bbclass
¶
The gnomebase
class is the base
class for recipes that build software from the GNOME stack.
This class sets
SRC_URI
to
download the source from the GNOME mirrors as well as extending
FILES
with the typical GNOME installation paths.
gobject-introspection.bbclass
¶
Provides support for recipes building software that
supports GObject introspection.
This functionality is only enabled if the
"gobject-introspection-data" feature is in
DISTRO_FEATURES
as well as "qemu-usermode" being in
MACHINE_FEATURES
.
DISTRO_FEATURES_BACKFILL_CONSIDERED
or
MACHINE_FEATURES_BACKFILL_CONSIDERED
,
respectively.
grub-efi.bbclass
¶
The grub-efi
class provides grub-efi
-specific functions for
building bootable images.
This class supports several variables:
INITRD
:
Indicates list of filesystem images to concatenate and use
as an initial RAM disk (initrd) (optional).
ROOTFS
:
Indicates a filesystem image to include as the root filesystem
(optional).
GRUB_GFXSERIAL
:
Set this to "1" to have graphics and serial in the boot menu.
LABELS
:
A list of targets for the automatic configuration.
APPEND
:
An override list of append strings for each
LABEL
.
GRUB_OPTS
:
Additional options to add to the configuration (optional).
Options are delimited using semi-colon characters
(;
).
GRUB_TIMEOUT
:
Timeout before executing the default LABEL
(optional).
gsettings.bbclass
¶
The gsettings
class
provides common functionality for recipes that need to install
GSettings (glib) schemas.
The schemas are assumed to be part of the main package.
Appropriate post-install and post-remove (postinst/postrm)
scriptlets are added to register and unregister the schemas in the
target image.
gtk-doc.bbclass
¶
The gtk-doc
class
is a helper class to pull in the appropriate
gtk-doc
dependencies and disable
gtk-doc
.
gtk-icon-cache.bbclass
¶
The gtk-icon-cache
class
generates the proper post-install and post-remove (postinst/postrm)
scriptlets for packages that use GTK+ and install icons.
These scriptlets call gtk-update-icon-cache
to add
the fonts to GTK+'s icon cache.
Since the cache files are architecture-specific,
gtk-update-icon-cache
is run using QEMU if the
postinst scriptlets need to be run on the build host during image
creation.
gtk-immodules-cache.bbclass
¶
The gtk-immodules-cache
class
generates the proper post-install and post-remove (postinst/postrm)
scriptlets for packages that install GTK+ input method modules for
virtual keyboards.
These scriptlets call gtk-update-icon-cache
to add
the input method modules to the cache.
Since the cache files are architecture-specific,
gtk-update-icon-cache
is run using QEMU if the
postinst scriptlets need to be run on the build host during image
creation.
If the input method modules being installed are in packages other than
the main package, set
GTKIMMODULES_PACKAGES
to specify the packages containing the modules.
gzipnative.bbclass
¶
The gzipnative
class enables the use of
different native versions of gzip
and pigz
rather than the versions of these tools
from the build host.
icecc.bbclass
¶
The icecc
class supports
Icecream, which
facilitates taking compile jobs and distributing them among remote
machines.
The class stages directories with symlinks from gcc
and g++
to icecc
, for both
native and cross compilers.
Depending on each configure or compile, the OpenEmbedded build system
adds the directories at the head of the PATH
list
and then sets the ICECC_CXX
and
ICEC_CC
variables, which are the paths to the
g++
and gcc
compilers,
respectively.
For the cross compiler, the class creates a tar.gz
file that contains the Yocto Project toolchain and sets
ICECC_VERSION
, which is the version of the
cross-compiler used in the cross-development toolchain, accordingly.
The class handles all three different compile stages
(i.e native ,cross-kernel and target) and creates the necessary
environment tar.gz
file to be used by the remote
machines.
The class also supports SDK generation.
If ICECC_PATH
is not set in your local.conf
file, then the
class tries to locate the icecc
binary
using which
.
If
ICECC_ENV_EXEC
is set in your local.conf
file, the variable should
point to the icecc-create-env
script
provided by the user.
If you do not point to a user-provided script, the build system
uses the default script provided by the recipe
icecc-create-env-native.bb
.
icecc
.
If you do not want the Icecream distributed compile support to apply
to specific recipes or classes, you can effectively "blacklist" them
by listing the recipes and classes using the
ICECC_USER_PACKAGE_BL
and
ICECC_USER_CLASS_BL
,
variables, respectively, in your local.conf
file.
Doing so causes the OpenEmbedded build system to handle these
compilations locally.
Additionally, you can list recipes using the
ICECC_USER_PACKAGE_WL
variable in your local.conf
file to force
icecc
to be enabled for recipes using an empty
PARALLEL_MAKE
variable.
Inheriting the icecc
class changes all sstate
signatures.
Consequently, if a development team has a dedicated build system
that populates
STATE_MIRRORS
and they want to reuse sstate from
STATE_MIRRORS
, then all developers and the
build system need to either inherit the icecc
class or nobody should.
At the distribution level, you can inherit the
icecc
class to be sure that all builders start
with the same sstate signatures.
After inheriting the class, you can then disable the feature by setting
the
ICECC_DISABLED
variable to "1" as follows:
INHERIT_DISTRO_append = " icecc" ICECC_DISABLED ??= "1"
This practice makes sure everyone is using the same signatures but also
requires individuals that do want to use Icecream to enable the feature
individually as follows in your local.conf
file:
ICECC_DISABLED = ""
image.bbclass
¶
The image
class helps support creating images
in different formats.
First, the root filesystem is created from packages using
one of the rootfs*.bbclass
files (depending on the package format used) and then one or more image
files are created.
The
IMAGE_FSTYPES
variable controls the types of images to generate.
The
IMAGE_INSTALL
variable controls the list of packages to install into the
image.
For information on customizing images, see the "Customizing Images" section in the Yocto Project Development Tasks Manual. For information on how images are created, see the "Images" section in the Yocto Project Overview and Concpets Manual.
image-buildinfo.bbclass
¶
The image-buildinfo
class writes information
to the target filesystem on /etc/build
.
image_types.bbclass
¶
The image_types
class defines all of the
standard image output types that you can enable through the
IMAGE_FSTYPES
variable.
You can use this class as a reference on how to add support for
custom image output types.
By default, the
image
class automatically enables the image_types
class.
The image
class uses the
IMGCLASSES
variable as follows:
IMGCLASSES = "rootfs_${IMAGE_PKGTYPE} image_types ${IMAGE_CLASSES}" IMGCLASSES += "${@['populate_sdk_base', 'populate_sdk_ext']['linux' in d.getVar("SDK_OS")]}" IMGCLASSES += "${@bb.utils.contains_any('IMAGE_FSTYPES', 'live iso hddimg', 'image-live', '', d)}" IMGCLASSES += "${@bb.utils.contains('IMAGE_FSTYPES', 'container', 'image-container', '', d)}" IMGCLASSES += "image_types_wic" IMGCLASSES += "rootfs-postcommands" IMGCLASSES += "image-postinst-intercepts" inherit ${IMGCLASSES}
The image_types
class also handles conversion and
compression of images.
IMAGE_FSTYPES
.
This would also be similar for Virtual Box Virtual Disk Image
("vdi") and QEMU Copy On Write Version 2 ("qcow2") images.
image-live.bbclass
¶
This class controls building "live" (i.e. HDDIMG and ISO) images.
Live images contain syslinux for legacy booting, as well as the
bootloader specified by
EFI_PROVIDER
if
MACHINE_FEATURES
contains "efi".
Normally, you do not use this class directly.
Instead, you add "live" to
IMAGE_FSTYPES
.
image-mklibs.bbclass
¶
The image-mklibs
class
enables the use of the mklibs
utility during the
do_rootfs
task, which optimizes the size of
libraries contained in the image.
By default, the class is enabled in the
local.conf.template
using the
USER_CLASSES
variable as follows:
USER_CLASSES ?= "buildstats image-mklibs image-prelink"
image-prelink.bbclass
¶
The image-prelink
class
enables the use of the prelink
utility during
the
do_rootfs
task, which optimizes the dynamic
linking of shared libraries to reduce executable startup time.
By default, the class is enabled in the
local.conf.template
using the
USER_CLASSES
variable as follows:
USER_CLASSES ?= "buildstats image-mklibs image-prelink"
insane.bbclass
¶
The insane
class adds a step to the package
generation process so that output quality assurance checks are
generated by the OpenEmbedded build system.
A range of checks are performed that check the build's output
for common problems that show up during runtime.
Distribution policy usually dictates whether to include this class.
You can configure the sanity checks so that specific test failures either raise a warning or an error message. Typically, failures for new tests generate a warning. Subsequent failures for the same test would then generate an error message once the metadata is in a known and good condition. See the "QA Error and Warning Messages" Chapter for a list of all the warning and error messages you might encounter using a default configuration.
Use the
WARN_QA
and
ERROR_QA
variables to control the behavior of
these checks at the global level (i.e. in your custom distro
configuration).
However, to skip one or more checks in recipes, you should use
INSANE_SKIP
.
For example, to skip the check for symbolic link
.so
files in the main package of a recipe,
add the following to the recipe.
You need to realize that the package name override, in this example
${PN}
, must be used:
INSANE_SKIP_${PN} += "dev-so"
Please keep in mind that the QA checks exist in order to detect real or potential problems in the packaged output. So exercise caution when disabling these checks.
The following list shows the tests you can list with the
WARN_QA
and ERROR_QA
variables:
already-stripped:
Checks that produced binaries have not already been
stripped prior to the build system extracting debug symbols.
It is common for upstream software projects to default to
stripping debug symbols for output binaries.
In order for debugging to work on the target using
-dbg
packages, this stripping must be
disabled.
arch:
Checks the Executable and Linkable Format (ELF) type, bit size,
and endianness of any binaries to ensure they match the target
architecture.
This test fails if any binaries do not match the type since
there would be an incompatibility.
The test could indicate that the
wrong compiler or compiler options have been used.
Sometimes software, like bootloaders, might need to bypass
this check.
buildpaths:
Checks for paths to locations on the build host inside the
output files.
Currently, this test triggers too many false positives and
thus is not normally enabled.
build-deps:
Determines if a build-time dependency that is specified through
DEPENDS
,
explicit
RDEPENDS
,
or task-level dependencies exists to match any runtime
dependency.
This determination is particularly useful to discover where
runtime dependencies are detected and added during packaging.
If no explicit dependency has been specified within the
metadata, at the packaging stage it is too late to ensure that
the dependency is built, and thus you can end up with an
error when the package is installed into the image during the
do_rootfs
task because the auto-detected dependency was not satisfied.
An example of this would be where the
update-rc.d
class automatically adds a dependency on the
initscripts-functions
package to packages
that install an initscript that refers to
/etc/init.d/functions
.
The recipe should really have an explicit
RDEPENDS
for the package in question on
initscripts-functions
so that the
OpenEmbedded build system is able to ensure that the
initscripts
recipe is actually built and
thus the initscripts-functions
package is
made available.
compile-host-path:
Checks the
do_compile
log for indications
that paths to locations on the build host were used.
Using such paths might result in host contamination of the
build output.
debug-deps:
Checks that all packages except -dbg
packages do not depend on -dbg
packages, which would cause a packaging bug.
debug-files:
Checks for .debug
directories in anything but the
-dbg
package.
The debug files should all be in the -dbg
package.
Thus, anything packaged elsewhere is incorrect packaging.
dep-cmp:
Checks for invalid version comparison statements in runtime
dependency relationships between packages (i.e. in
RDEPENDS
,
RRECOMMENDS
,
RSUGGESTS
,
RPROVIDES
,
RREPLACES
,
and
RCONFLICTS
variable values).
Any invalid comparisons might trigger failures or undesirable
behavior when passed to the package manager.
desktop:
Runs the desktop-file-validate
program
against any .desktop
files to validate
their contents against the specification for
.desktop
files.
dev-deps:
Checks that all packages except -dev
or -staticdev
packages do not depend on
-dev
packages, which would be a
packaging bug.
dev-so:
Checks that the .so
symbolic links are in the
-dev
package and not in any of the other packages.
In general, these symlinks are only useful for development purposes.
Thus, the -dev
package is the correct location for
them.
Some very rare cases do exist for dynamically loaded modules where
these symlinks are needed instead in the main package.
file-rdeps:
Checks that file-level dependencies identified by the
OpenEmbedded build system at packaging time are satisfied.
For example, a shell script might start with the line
#!/bin/bash
.
This line would translate to a file dependency on
/bin/bash
.
Of the three package managers that the OpenEmbedded build
system supports, only RPM directly handles file-level
dependencies, resolving them automatically to packages
providing the files.
However, the lack of that functionality in the other two
package managers does not mean the dependencies do not still
need resolving.
This QA check attempts to ensure that explicitly declared
RDEPENDS
exist to handle any file-level dependency detected in
packaged files.
files-invalid:
Checks for
FILES
variable values that contain "//", which is invalid.
host-user-contaminated:
Checks that no package produced by the recipe contains any
files outside of /home
with a user or
group ID that matches the user running BitBake.
A match usually indicates that the files are being installed
with an incorrect UID/GID, since target IDs are independent
from host IDs.
For additional information, see the section describing the
do_install
task.
incompatible-license:
Report when packages are excluded from being created due to
being marked with a license that is in
INCOMPATIBLE_LICENSE
.
install-host-path:
Checks the
do_install
log for indications
that paths to locations on the build host were used.
Using such paths might result in host contamination of the
build output.
installed-vs-shipped:
Reports when files have been installed within
do_install
but have not been included in
any package by way of the
FILES
variable.
Files that do not appear in any package cannot be present in
an image later on in the build process.
Ideally, all installed files should be packaged or not
installed at all.
These files can be deleted at the end of
do_install
if the files are not
needed in any package.
invalid-chars:
Checks that the recipe metadata variables
DESCRIPTION
,
SUMMARY
,
LICENSE
,
and
SECTION
do not contain non-UTF-8 characters.
Some package managers do not support such characters.
invalid-packageconfig:
Checks that no undefined features are being added to
PACKAGECONFIG
.
For example, any name "foo" for which the following form
does not exist:
PACKAGECONFIG[foo] = "..."
la:
Checks .la
files for any TMPDIR
paths.
Any .la
file containing these paths is incorrect since
libtool
adds the correct sysroot prefix when using the
files automatically itself.
ldflags:
Ensures that the binaries were linked with the
LDFLAGS
options provided by the build system.
If this test fails, check that the LDFLAGS
variable
is being passed to the linker command.
libdir:
Checks for libraries being installed into incorrect
(possibly hardcoded) installation paths.
For example, this test will catch recipes that install
/lib/bar.so
when
${base_libdir}
is "lib32".
Another example is when recipes install
/usr/lib64/foo.so
when
${libdir}
is "/usr/lib".
libexec:
Checks if a package contains files in
/usr/libexec
.
This check is not performed if the
libexecdir
variable has been set
explicitly to /usr/libexec
.
packages-list:
Checks for the same package being listed multiple times through
the PACKAGES
variable value.
Installing the package in this manner can cause errors during
packaging.
perm-config:
Reports lines in fs-perms.txt
that have
an invalid format.
perm-line:
Reports lines in fs-perms.txt
that have
an invalid format.
perm-link:
Reports lines in fs-perms.txt
that
specify 'link' where the specified target already exists.
perms:
Currently, this check is unused but reserved.
pkgconfig:
Checks .pc
files for any
TMPDIR
/WORKDIR
paths.
Any .pc
file containing these paths is incorrect
since pkg-config
itself adds the correct sysroot prefix
when the files are accessed.
pkgname:
Checks that all packages in
PACKAGES
have names that do not contain invalid characters (i.e.
characters other than 0-9, a-z, ., +, and -).
pkgv-undefined:
Checks to see if the PKGV
variable
is undefined during
do_package
.
pkgvarcheck:
Checks through the variables
RDEPENDS
,
RRECOMMENDS
,
RSUGGESTS
,
RCONFLICTS
,
RPROVIDES
,
RREPLACES
,
FILES
,
ALLOW_EMPTY
,
pkg_preinst
,
pkg_postinst
,
pkg_prerm
and pkg_postrm
, and reports if there are
variable sets that are not package-specific.
Using these variables without a package suffix is bad practice,
and might unnecessarily complicate dependencies of other packages
within the same recipe or have other unintended consequences.
pn-overrides:
Checks that a recipe does not have a name
(PN
) value
that appears in
OVERRIDES
.
If a recipe is named such that its PN
value matches something already in
OVERRIDES
(e.g. PN
happens to be the same as
MACHINE
or
DISTRO
),
it can have unexpected consequences.
For example, assignments such as
FILES_${PN} = "xyz"
effectively turn into
FILES = "xyz"
.
rpaths:
Checks for rpaths in the binaries that contain build system paths such
as TMPDIR
.
If this test fails, bad -rpath
options are being
passed to the linker commands and your binaries have potential security
issues.
split-strip:
Reports that splitting or stripping debug symbols from binaries
has failed.
staticdev:
Checks for static library files (*.a
) in
non-staticdev
packages.
symlink-to-sysroot:
Checks for symlinks in packages that point into
TMPDIR
on the host.
Such symlinks will work on the host, but are clearly invalid
when running on the target.
textrel:
Checks for ELF binaries that contain relocations in their
.text
sections, which can result in a
performance impact at runtime.
See the explanation for the
ELF binary
message for more information regarding runtime performance issues.
useless-rpaths:
Checks for dynamic library load paths (rpaths) in the binaries that
by default on a standard system are searched by the linker (e.g.
/lib
and /usr/lib
).
While these paths will not cause any breakage, they do waste space and
are unnecessary.
var-undefined:
Reports when variables fundamental to packaging (i.e.
WORKDIR
,
DEPLOY_DIR
,
D
,
PN
, and
PKGD
) are
undefined during
do_package
.
version-going-backwards:
If Build History is enabled, reports when a package
being written out has a lower version than the previously
written package under the same name.
If you are placing output packages into a feed and
upgrading packages on a target system using that feed, the
version of a package going backwards can result in the target
system not correctly upgrading to the "new" version of the
package.
xorg-driver-abi:
Checks that all packages containing Xorg drivers have ABI
dependencies.
The xserver-xorg
recipe provides driver
ABI names.
All drivers should depend on the ABI versions that they have
been built against.
Driver recipes that include
xorg-driver-input.inc
or xorg-driver-video.inc
will
automatically get these versions.
Consequently, you should only need to explicitly add
dependencies to binary driver recipes.
insserv.bbclass
¶
The insserv
class
uses the insserv
utility to update the order of
symbolic links in /etc/rc?.d/
within an image
based on dependencies specified by LSB headers in the
init.d
scripts themselves.
kernel.bbclass
¶
The kernel
class handles building Linux kernels.
The class contains code to build all kernel trees.
All needed headers are staged into the
STAGING_KERNEL_DIR
directory to allow out-of-tree module builds using
the
module
class.
This means that each built kernel module is packaged separately and
inter-module dependencies are created by parsing the
modinfo
output.
If all modules are required, then installing the
kernel-modules
package installs all packages with
modules and various other kernel packages such as
kernel-vmlinux
.
The kernel
class contains logic that allows
you to embed an initial RAM filesystem (initramfs) image when
you build the kernel image.
For information on how to build an initramfs, see the
"Building an Initial RAM Filesystem (initramfs) Image"
section in the Yocto Project Development Tasks Manual.
Various other classes are used by the kernel
and module
classes internally including the
kernel-arch
,
module-base
,
and
linux-kernel-base
classes.
kernel-arch.bbclass
¶
The kernel-arch
class
sets the ARCH
environment variable for Linux
kernel compilation (including modules).
kernel-devicetree.bbclass
¶
The kernel-devicetree
class, which is inherited by
the
kernel
class, supports device tree generation.
kernel-fitimage.bbclass
¶
The kernel-fitimage
class provides support to
pack zImages.
kernel-grub.bbclass
¶
The kernel-grub
class updates the boot area and
the boot menu with the kernel as the priority boot mechanism while
installing a RPM to update the kernel on a deployed target.
kernel-module-split.bbclass
¶
The kernel-module-split
class
provides common functionality for splitting Linux kernel modules into
separate packages.
kernel-uboot.bbclass
¶
The kernel-uboot
class provides support for
building from vmlinux-style kernel sources.
kernel-uimage.bbclass
¶
The kernel-uimage
class provides support to
pack uImage.
kernel-yocto.bbclass
¶
The kernel-yocto
class
provides common functionality for building from linux-yocto style
kernel source repositories.
kernelsrc.bbclass
¶
The kernelsrc
class sets the Linux kernel
source and version.
lib_package.bbclass
¶
The lib_package
class
supports recipes that build libraries and produce executable
binaries, where those binaries should not be installed by default
along with the library.
Instead, the binaries are added to a separate
${
PN
}-bin
package to make their installation optional.
libc*.bbclass
¶
The libc*
classes support recipes that build
packages with libc
:
The libc-common
class
provides common support for building with
libc
.
The libc-package
class
supports packaging up glibc
and
eglibc
.
license.bbclass
¶
The license
class provides license
manifest creation and license exclusion.
This class is enabled by default using the default value for the
INHERIT_DISTRO
variable.
linux-kernel-base.bbclass
¶
The linux-kernel-base
class
provides common functionality for recipes that build out of the Linux
kernel source tree.
These builds goes beyond the kernel itself.
For example, the Perf recipe also inherits this class.
linuxloader.bbclass
¶
Provides the function linuxloader()
, which gives
the value of the dynamic loader/linker provided on the platform.
This value is used by a number of other classes.
logging.bbclass
¶
The logging
class provides the standard
shell functions used to log messages for various BitBake severity levels
(i.e. bbplain
, bbnote
,
bbwarn
, bberror
,
bbfatal
, and bbdebug
).
This class is enabled by default since it is inherited by
the base
class.
meta.bbclass
¶
The meta
class is inherited by recipes
that do not build any output packages themselves, but act as a "meta"
target for building other recipes.
metadata_scm.bbclass
¶
The metadata_scm
class provides functionality for
querying the branch and revision of a Source Code Manager (SCM)
repository.
The base
class uses this class to print the revisions of each layer before
starting every build.
The metadata_scm
class is enabled by default
because it is inherited by the base
class.
migrate_localcount.bbclass
¶
The migrate_localcount
class verifies a recipe's
localcount data and increments it appropriately.
mime.bbclass
¶
The mime
class generates the proper
post-install and post-remove (postinst/postrm) scriptlets for packages
that install MIME type files.
These scriptlets call update-mime-database
to add
the MIME types to the shared database.
mirrors.bbclass
¶
The mirrors
class sets up some standard
MIRRORS
entries
for source code mirrors.
These mirrors provide a fall-back path in case the upstream source
specified in
SRC_URI
within recipes is unavailable.
This class is enabled by default since it is inherited by the
base
class.
module.bbclass
¶
The module
class provides support for building
out-of-tree Linux kernel modules.
The class inherits the
module-base
and
kernel-module-split
classes, and implements the
do_compile
and
do_install
tasks.
The class provides everything needed to build and package a kernel
module.
For general information on out-of-tree Linux kernel modules, see the "Incorporating Out-of-Tree Modules" section in the Yocto Project Linux Kernel Development Manual.
module-base.bbclass
¶
The module-base
class provides the base
functionality for building Linux kernel modules.
Typically, a recipe that builds software that includes one or
more kernel modules and has its own means of building
the module inherits this class as opposed to inheriting the
module
class.
multilib*.bbclass
¶
The multilib*
classes provide support
for building libraries with different target optimizations or target
architectures and installing them side-by-side in the same image.
For more information on using the Multilib feature, see the "Combining Multiple Versions of Library Files into One Image" section in the Yocto Project Development Tasks Manual.
native.bbclass
¶
The native
class provides common
functionality for recipes that build tools to run on the
build host
(i.e. tools that use the compiler or other tools from the
build host).
You can create a recipe that builds tools that run natively on the host a couple different ways:
Create a
myrecipe
-native.bb
recipe that inherits the native
class.
If you use this method, you must order the inherit statement
in the recipe after all other inherit statements so that the
native
class is inherited last.
myrecipe
-native.bb
Not using this naming convention can lead to subtle
problems caused by existing code that depends on that
naming convention.
Create or modify a target recipe that contains the following:
BBCLASSEXTEND
= "native"
Inside the recipe, use _class-native
and
_class-target
overrides to specify any
functionality specific to the respective native or target
case.
Although applied differently, the native
class is
used with both methods.
The advantage of the second method is that you do not need to have two
separate recipes (assuming you need both) for native and target.
All common parts of the recipe are automatically shared.
nativesdk.bbclass
¶
The nativesdk
class provides common
functionality for recipes that wish to build tools to run as part of
an SDK (i.e. tools that run on
SDKMACHINE
).
You can create a recipe that builds tools that run on the SDK machine a couple different ways:
Create a
nativesdk-
myrecipe
.bb
recipe that inherits the nativesdk
class.
If you use this method, you must order the inherit statement
in the recipe after all other inherit statements so that the
nativesdk
class is inherited last.
Create a nativesdk
variant
of any recipe by adding the following:
BBCLASSEXTEND
= "nativesdk"
Inside the recipe, use _class-nativesdk
and
_class-target
overrides to specify any
functionality specific to the respective SDK machine or target
case.
nativesdk-myrecipe
.bb
Not doing so can lead to subtle problems because code exists
that depends on the naming convention.
Although applied differently, the nativesdk
class
is used with both methods.
The advantage of the second method is that you do not need to have two
separate recipes (assuming you need both) for the SDK machine and the
target.
All common parts of the recipe are automatically shared.
nopackages.bbclass
¶Disables packaging tasks for those recipes and classes where packaging is not needed.
npm.bbclass
¶Provides support for building Node.js software fetched using the node package manager (NPM).
npm://
fetcher to have dependencies fetched
and packaged automatically.
For information on how to create NPM packages, see the "Creating Node Package Manager (NPM) Packages" section in the Yocto Project Development Tasks Manual.
oelint.bbclass
¶
The oelint
class is an
obsolete lint checking tool that exists in
meta/classes
in the
Source Directory.
A number of classes exist that could be generally useful in
OE-Core but are never actually used within OE-Core itself.
The oelint
class is one such example.
However, being aware of this class can reduce the proliferation of
different versions of similar classes across multiple layers.
own-mirrors.bbclass
¶
The own-mirrors
class makes it
easier to set up your own
PREMIRRORS
from which to first fetch source before attempting to fetch it from the
upstream specified in
SRC_URI
within each recipe.
To use this class, inherit it globally and specify
SOURCE_MIRROR_URL
.
Here is an example:
INHERIT += "own-mirrors" SOURCE_MIRROR_URL = "http://example.com/my-source-mirror"
You can specify only a single URL in
SOURCE_MIRROR_URL
.
package.bbclass
¶
The package
class supports generating
packages from a build's output.
The core generic functionality is in
package.bbclass
.
The code specific to particular package types resides in these
package-specific classes:
package_deb
,
package_rpm
,
package_ipk
,
and
package_tar
.
package_tar
class is broken and not
supported.
It is recommended that you do not use this class.
You can control the list of resulting package formats by using the
PACKAGE_CLASSES
variable defined in your conf/local.conf
configuration file, which is located in the
Build Directory.
When defining the variable, you can specify one or more package types.
Since images are generated from packages, a packaging class is
needed to enable image generation.
The first class listed in this variable is used for image generation.
If you take the optional step to set up a repository (package feed) on the development host that can be used by DNF, you can install packages from the feed while you are running the image on the target (i.e. runtime installation of packages). For more information, see the "Using Runtime Package Management" section in the Yocto Project Development Tasks Manual.
The package-specific class you choose can affect build-time performance
and has space ramifications.
In general, building a package with IPK takes about thirty percent less
time as compared to using RPM to build the same or similar package.
This comparison takes into account a complete build of the package with
all dependencies previously built.
The reason for this discrepancy is because the RPM package manager
creates and processes more
Metadata than the
IPK package manager.
Consequently, you might consider setting
PACKAGE_CLASSES
to "package_ipk" if you are
building smaller systems.
Before making your package manager decision, however, you should consider some further things about using RPM:
RPM starts to provide more abilities than IPK due to the fact that it processes more Metadata. For example, this information includes individual file types, file checksum generation and evaluation on install, sparse file support, conflict detection and resolution for Multilib systems, ACID style upgrade, and repackaging abilities for rollbacks.
For smaller systems, the extra space used for the Berkeley Database and the amount of metadata when using RPM can affect your ability to perform on-device upgrades.
You can find additional information on the effects of the package class at these two Yocto Project mailing list links:
package_deb.bbclass
¶
The package_deb
class
provides support for creating packages that use the Debian
(i.e. .deb
) file format.
The class ensures the packages are written out in a
.deb
file format to the
${
DEPLOY_DIR_DEB
}
directory.
This class inherits the
package
class and is enabled through the
PACKAGE_CLASSES
variable in the local.conf
file.
package_ipk.bbclass
¶
The package_ipk
class
provides support for creating packages that use the IPK
(i.e. .ipk
) file format.
The class ensures the packages are written out in a
.ipk
file format to the
${
DEPLOY_DIR_IPK
}
directory.
This class inherits the
package
class and is enabled through the
PACKAGE_CLASSES
variable in the local.conf
file.
package_rpm.bbclass
¶
The package_rpm
class
provides support for creating packages that use the RPM
(i.e. .rpm
) file format.
The class ensures the packages are written out in a
.rpm
file format to the
${
DEPLOY_DIR_RPM
}
directory.
This class inherits the
package
class and is enabled through the
PACKAGE_CLASSES
variable in the local.conf
file.
package_tar.bbclass
¶
The package_tar
class
provides support for creating tarballs.
The class ensures the packages are written out in a
tarball format to the
${
DEPLOY_DIR_TAR
}
directory.
This class inherits the
package
class and is enabled through the
PACKAGE_CLASSES
variable in the local.conf
file.
package_tar
class
first using the PACKAGE_CLASSES
variable.
You must use .deb
,
.ipk
, or .rpm
file
formats for your image or SDK.
packagedata.bbclass
¶
The packagedata
class provides
common functionality for reading pkgdata
files
found in
PKGDATA_DIR
.
These files contain information about each output package produced by
the OpenEmbedded build system.
This class is enabled by default because it is inherited by the
package
class.
packagegroup.bbclass
¶
The packagegroup
class sets default values
appropriate for package group recipes (e.g.
PACKAGES
,
PACKAGE_ARCH
,
ALLOW_EMPTY
,
and so forth).
It is highly recommended that all package group recipes inherit this class.
For information on how to use this class, see the "Customizing Images Using Custom Package Groups" section in the Yocto Project Development Tasks Manual.
Previously, this class was called the task
class.
patch.bbclass
¶
The patch
class provides all functionality for
applying patches during the
do_patch
task.
This class is enabled by default because it is inherited by the
base
class.
perlnative.bbclass
¶
When inherited by a recipe, the perlnative
class
supports using the native version of Perl built by the build system
rather than using the version provided by the build host.
pixbufcache.bbclass
¶
The pixbufcache
class generates the proper
post-install and post-remove (postinst/postrm) scriptlets for packages
that install pixbuf loaders, which are used with
gdk-pixbuf
.
These scriptlets call update_pixbuf_cache
to add the pixbuf loaders to the cache.
Since the cache files are architecture-specific,
update_pixbuf_cache
is run using QEMU if the
postinst scriptlets need to be run on the build host during image
creation.
If the pixbuf loaders being installed are in packages other
than the recipe's main package, set
PIXBUF_PACKAGES
to specify the packages containing the loaders.
pkgconfig.bbclass
¶
The pkgconfig
class provides a standard way to get
header and library information by using pkg-config
.
This class aims to smooth integration of
pkg-config
into libraries that use it.
During staging, BitBake installs pkg-config
data into the sysroots/
directory.
By making use of sysroot functionality within
pkg-config
, the pkgconfig
class no longer has to manipulate the files.
populate_sdk.bbclass
¶
The populate_sdk
class provides support for
SDK-only recipes.
For information on advantages gained when building a cross-development
toolchain using the
do_populate_sdk
task, see the
"Building an SDK Installer"
section in the Yocto Project Application Development and the
Extensible Software Development Kit (eSDK) manual.
populate_sdk_*.bbclass
¶
The populate_sdk_*
classes support SDK creation
and consist of the following classes:
populate_sdk_base
:
The base class supporting SDK creation under all package
managers (i.e. DEB, RPM, and opkg).
populate_sdk_deb
:
Supports creation of the SDK given the Debian package manager.
populate_sdk_rpm
:
Supports creation of the SDK given the RPM package manager.
populate_sdk_ipk
:
Supports creation of the SDK given the opkg (IPK format)
package manager.
populate_sdk_ext
:
Supports extensible SDK creation under all package managers.
The populate_sdk_base
class inherits the
appropriate populate_sdk_*
(i.e.
deb
, rpm
, and
ipk
) based on
IMAGE_PKGTYPE
.
The base class ensures all source and destination directories are
established and then populates the SDK.
After populating the SDK, the populate_sdk_base
class constructs two sysroots:
${
SDK_ARCH
}-nativesdk
,
which contains the cross-compiler and associated tooling, and the
target, which contains a target root filesystem that is configured for
the SDK usage.
These two images reside in
SDK_OUTPUT
,
which consists of the following:
${SDK_OUTPUT}/${SDK_ARCH}-nativesdk-pkgs
${SDK_OUTPUT}/${SDKTARGETSYSROOT}/target-pkgs
Finally, the base populate SDK class creates the toolchain environment setup script, the tarball of the SDK, and the installer.
The respective populate_sdk_deb
,
populate_sdk_rpm
, and
populate_sdk_ipk
classes each support the
specific type of SDK.
These classes are inherited by and used with the
populate_sdk_base
class.
For more information on the cross-development toolchain
generation, see the
"Cross-Development Toolchain Generation"
section in the Yocto Project Overview and Concepts Manual.
For information on advantages gained when building a
cross-development toolchain using the
do_populate_sdk
task, see the
"Building an SDK Installer"
section in the Yocto Project Application Development and the
Extensible Software Development Kit (eSDK) manual.
prexport.bbclass
¶
The prexport
class provides functionality for
exporting
PR
values.
bitbake-prserv-tool export
".
primport.bbclass
¶
The primport
class provides functionality for
importing
PR
values.
bitbake-prserv-tool import
".
prserv.bbclass
¶
The prserv
class provides functionality for
using a
PR service
in order to automatically manage the incrementing of the
PR
variable for
each recipe.
This class is enabled by default because it is inherited by the
package
class.
However, the OpenEmbedded build system will not enable the
functionality of this class unless
PRSERV_HOST
has been set.
ptest.bbclass
¶
The ptest
class provides functionality for
packaging and installing runtime tests for recipes that build software
that provides these tests.
This class is intended to be inherited by individual recipes.
However, the class' functionality is largely disabled unless "ptest"
appears in
DISTRO_FEATURES
.
See the
"Testing Packages With ptest"
section in the Yocto Project Development Tasks Manual for more
information on ptest.
ptest-gnome.bbclass
¶
Enables package tests (ptests) specifically for GNOME packages,
which have tests intended to be executed with
gnome-desktop-testing
.
For information on setting up and running ptests, see the "Testing Packages With ptest" section in the Yocto Project Development Tasks Manual.
python-dir.bbclass
¶
The python-dir
class provides the base version,
location, and site package location for Python.
python3native.bbclass
¶
The python3native
class supports using the
native version of Python 3 built by the build system rather than
support of the version provided by the build host.
pythonnative.bbclass
¶
When inherited by a recipe, the pythonnative
class
supports using the native version of Python built by the build system
rather than using the version provided by the build host.
qemu.bbclass
¶
The qemu
class provides functionality for recipes
that either need QEMU or test for the existence of QEMU.
Typically, this class is used to run programs for a target system on
the build host using QEMU's application emulation mode.
recipe_sanity.bbclass
¶
The recipe_sanity
class checks for the presence
of any host system recipe prerequisites that might affect the
build (e.g. variables that are set or software that is present).
relocatable.bbclass
¶
The relocatable
class enables relocation of
binaries when they are installed into the sysroot.
This class makes use of the
chrpath
class and is used by both the
cross
and
native
classes.
remove-libtool.bbclass
¶
The remove-libtool
class adds a post function
to the
do_install
task to remove all .la
files installed by
libtool
.
Removing these files results in them being absent from both the
sysroot and target packages.
If a recipe needs the .la
files to be installed,
then the recipe can override the removal by setting
REMOVE_LIBTOOL_LA
to "0" as follows:
REMOVE_LIBTOOL_LA = "0"
remove-libtool
class is not enabled by
default.
report-error.bbclass
¶
The report-error
class supports enabling the
error reporting tool,
which allows you to submit build error information to a central
database.
The class collects debug information for recipe, recipe version, task,
machine, distro, build system, target system, host distro, branch,
commit, and log.
From the information, report files using a JSON format are created and
stored in
${
LOG_DIR
}/error-report
.
rm_work.bbclass
¶
The rm_work
class supports deletion of temporary
workspace, which can ease your hard drive demands during builds.
The OpenEmbedded build system can use a substantial amount of disk
space during the build process.
A portion of this space is the work files under the
${TMPDIR}/work
directory for each recipe.
Once the build system generates the packages for a recipe, the work
files for that recipe are no longer needed.
However, by default, the build system preserves these files
for inspection and possible debugging purposes.
If you would rather have these files deleted to save disk space
as the build progresses, you can enable rm_work
by adding the following to your local.conf
file,
which is found in the
Build Directory.
INHERIT += "rm_work"
If you are modifying and building source code out of the work directory
for a recipe, enabling rm_work
will potentially
result in your changes to the source being lost.
To exclude some recipes from having their work directories deleted by
rm_work
, you can add the names of the recipe or
recipes you are working on to the RM_WORK_EXCLUDE
variable, which can also be set in your local.conf
file.
Here is an example:
RM_WORK_EXCLUDE += "busybox glibc"
rootfs*.bbclass
¶
The rootfs*
classes support creating
the root filesystem for an image and consist of the following classes:
The rootfs-postcommands
class, which
defines filesystem post-processing functions for image recipes.
The rootfs_deb
class, which supports
creation of root filesystems for images built using
.deb
packages.
The rootfs_rpm
class, which supports
creation of root filesystems for images built using
.rpm
packages.
The rootfs_ipk
class, which supports
creation of root filesystems for images built using
.ipk
packages.
The rootfsdebugfiles
class, which installs
additional files found on the build host directly into the
root filesystem.
The root filesystem is created from packages using one of the
rootfs*.bbclass
files as determined by the
PACKAGE_CLASSES
variable.
For information on how root filesystem images are created, see the "Image Generation" section in the Yocto Project Overview and Concepts Manual.
sanity.bbclass
¶
The sanity
class checks to see if prerequisite
software is present on the host system so that users can be notified
of potential problems that might affect their build.
The class also performs basic user configuration checks from
the local.conf
configuration file to
prevent common mistakes that cause build failures.
Distribution policy usually determines whether to include this class.
scons.bbclass
¶
The scons
class supports recipes that need to
build software that uses the SCons build system.
You can use the
EXTRA_OESCONS
variable to specify additional configuration options you want to pass
SCons command line.
sdl.bbclass
¶
The sdl
class supports recipes that need to build
software that uses the Simple DirectMedia Layer (SDL) library.
setuptools.bbclass
¶
The setuptools
class supports Python
version 2.x extensions that use build systems based on
setuptools
.
If your recipe uses these build systems, the recipe needs to
inherit the setuptools
class.
setuptools3.bbclass
¶
The setuptools3
class supports Python
version 3.x extensions that use build systems based on
setuptools3
.
If your recipe uses these build systems, the recipe needs to
inherit the setuptools3
class.
sign_rpm.bbclass
¶
The sign_rpm
class supports generating signed
RPM packages.
sip.bbclass
¶
The sip
class
supports recipes that build or package SIP-based Python bindings.
siteconfig.bbclass
¶
The siteconfig
class
provides functionality for handling site configuration.
The class is used by the
autotools
class to accelerate the
do_configure
task.
siteinfo.bbclass
¶
The siteinfo
class provides information about
the targets that might be needed by other classes or recipes.
As an example, consider Autotools, which can require tests that must
execute on the target hardware.
Since this is not possible in general when cross compiling, site
information is used to provide cached test results so these tests can
be skipped over but still make the correct values available.
The
meta/site directory
contains test results sorted into different categories such as
architecture, endianness, and the libc
used.
Site information provides a list of files containing data relevant to
the current build in the
CONFIG_SITE
variable
that Autotools automatically picks up.
The class also provides variables like
SITEINFO_ENDIANNESS
and SITEINFO_BITS
that can be used elsewhere in the metadata.
spdx.bbclass
¶
The spdx
class integrates real-time license
scanning, generation of SPDX standard output, and verification
of license information during the build.
sstate.bbclass
¶
The sstate
class provides support for Shared
State (sstate).
By default, the class is enabled through the
INHERIT_DISTRO
variable's default value.
For more information on sstate, see the "Shared State Cache" section in the Yocto Project Overview and Concepts Manual.
staging.bbclass
¶
The staging
class installs files into individual
recipe work directories for sysroots.
The class contains the following key tasks:
The
do_populate_sysroot
task, which is responsible for handing the files that end up
in the recipe sysroots.
The
do_prepare_recipe_sysroot
task (a "partner" task to the
populate_sysroot
task), which installs
the files into the individual recipe work directories (i.e.
WORKDIR
).
The code in the staging
class is complex and
basically works in two stages:
Stage One:
The first stage addresses recipes that have files they want
to share with other recipes that have dependencies on the
originating recipe.
Normally these dependencies are installed through the
do_install
task into
${
D
}
.
The do_populate_sysroot
task copies
a subset of these files into
${SYSROOT_DESTDIR}
.
This subset of files is controlled by the
SYSROOT_DIRS
,
SYSROOT_DIRS_NATIVE
,
and
SYSROOT_DIRS_BLACKLIST
variables.
SYSROOT_PREPROCESS_FUNCS
variable.
A shared state (sstate) object is built from these files
and the files are placed into a subdirectory of
tmp/sysroots-components/
.
The files are scanned for hardcoded paths to the original
installation location.
If the location is found in text files, the hardcoded
locations are replaced by tokens and a list of the files
needing such replacements is created.
These adjustments are referred to as "FIXMEs".
The list of files that are scanned for paths is controlled by
the
SSTATE_SCAN_FILES
variable.
Stage Two:
The second stage addresses recipes that want to use something
from another recipe and declare a dependency on that recipe
through the
DEPENDS
variable.
The recipe will have a
do_prepare_recipe_sysroot
task and when
this task executes, it creates the
recipe-sysroot
and
recipe-sysroot-native
in the recipe
work directory (i.e.
WORKDIR
).
The OpenEmbedded build system creates hard links to copies of the
relevant files from sysroots-components
into the recipe work directory.
The build system then addresses any "FIXMEs" to paths as defined from the list created in the first stage.
Finally, any files in ${bindir}
within the sysroot that have the prefix
"postinst-
" are executed.
Because recipes can have other dependencies outside of
DEPENDS
(e.g.
do_unpack[depends] += "tar-native:do_populate_sysroot"
),
the sysroot creation function
extend_recipe_sysroot
is also added as
a pre-function for those tasks whose dependencies are not
through DEPENDS
but operate similarly.
When installing dependencies into the sysroot, the code
traverses the dependency graph and processes dependencies
in exactly the same way as the dependencies would or would not
be when installed from sstate.
This processing means, for example, a native tool would have
its native dependencies added but a target library would not
have its dependencies traversed or installed.
The same sstate dependency code is used so that
builds should be identical regardless of whether sstate
was used or not.
For a closer look, see the
setscene_depvalid()
function in the
sstate
class.
The build system is careful to maintain manifests of the files it installs so that any given dependency can be installed as needed. The sstate hash of the installed item is also stored so that if it changes, the build system can reinstall it.
syslinux.bbclass
¶
The syslinux
class provides syslinux-specific
functions for building bootable images.
The class supports the following variables:
INITRD
:
Indicates list of filesystem images to concatenate and use as
an initial RAM disk (initrd).
This variable is optional.
ROOTFS
:
Indicates a filesystem image to include as the root filesystem.
This variable is optional.
AUTO_SYSLINUXMENU
:
Enables creating an automatic menu when set to "1".
LABELS
:
Lists targets for automatic configuration.
APPEND
:
Lists append string overrides for each label.
SYSLINUX_OPTS
:
Lists additional options to add to the syslinux file.
Semicolon characters separate multiple options.
SYSLINUX_SPLASH
:
Lists a background for the VGA boot menu when you are using the
boot menu.
SYSLINUX_DEFAULT_CONSOLE
:
Set to "console=ttyX" to change kernel boot default console.
SYSLINUX_SERIAL
:
Sets an alternate serial port.
Or, turns off serial when the variable is set with an
empty string.
SYSLINUX_SERIAL_TTY
:
Sets an alternate "console=tty..." kernel boot argument.
systemd.bbclass
¶
The systemd
class provides support for recipes
that install systemd unit files.
The functionality for this class is disabled unless you have "systemd"
in
DISTRO_FEATURES
.
Under this class, the recipe or Makefile (i.e. whatever the recipe is
calling during the
do_install
task) installs unit files into
${
D
}${systemd_unitdir}/system
.
If the unit files being installed go into packages other than the
main package, you need to set
SYSTEMD_PACKAGES
in your recipe to identify the packages in which the files will be
installed.
You should set
SYSTEMD_SERVICE
to the name of the service file.
You should also use a package name override to indicate the package
to which the value applies.
If the value applies to the recipe's main package, use
${
PN
}
.
Here is an example from the connman recipe:
SYSTEMD_SERVICE_${PN} = "connman.service"
Services are set up to start on boot automatically unless
you have set
SYSTEMD_AUTO_ENABLE
to "disable".
For more information on systemd
, see the
"Selecting an Initialization Manager"
section in the Yocto Project Development Tasks Manual.
systemd-boot.bbclass
¶
The systemd-boot
class provides functions specific
to the systemd-boot bootloader for building bootable images.
This is an internal class and is not intended to be used directly.
systemd-boot
class is a result from
merging the gummiboot
class used in previous
Yocto Project releases with the systemd
project.
Set the
EFI_PROVIDER
variable to "systemd-boot" to use this class.
Doing so creates a standalone EFI bootloader that is not dependent
on systemd.
For information on more variables used and supported in this class,
see the
SYSTEMD_BOOT_CFG
,
SYSTEMD_BOOT_ENTRIES
,
and
SYSTEMD_BOOT_TIMEOUT
variables.
You can also see the Systemd-boot documentation for more information.
terminal.bbclass
¶
The terminal
class provides support for starting
a terminal session.
The
OE_TERMINAL
variable controls which terminal emulator is used for the session.
Other classes use the terminal
class anywhere a
separate terminal session needs to be started.
For example, the
patch
class assuming
PATCHRESOLVE
is set to "user", the
cml1
class, and the
devshell
class all use the terminal
class.
testimage*.bbclass
¶
The testimage*
classes support running
automated tests against images using QEMU and on actual hardware.
The classes handle loading the tests and starting the image.
To use the classes, you need to perform steps to set up the
environment.
IMAGE_CLASSES
rather than
INHERIT
to
inherit the testimage
class for automated
image testing.
The tests are commands that run on the target system over
ssh
.
Each test is written in Python and makes use of the
unittest
module.
The testimage.bbclass
runs tests on an image
when called using the following:
$ bitbake -c testimage image
The testimage-auto
class runs tests on an image
after the image is constructed (i.e.
TESTIMAGE_AUTO
must be set to "1").
For information on how to enable, run, and create new tests, see the "Performing Automated Runtime Testing" section in the Yocto Project Development Tasks Manual.
testsdk.bbclass
¶
This class supports running automated tests against
software development kits (SDKs).
The testsdk
class runs tests on an SDK when
called using the following:
$ bitbake -c testsdk image
IMAGE_CLASSES
rather than
INHERIT
to
inherit the testsdk
class for automated
SDK testing.
texinfo.bbclass
¶
This class should be inherited by recipes whose upstream packages
invoke the texinfo
utilities at build-time.
Native and cross recipes are made to use the dummy scripts provided
by texinfo-dummy-native
, for improved performance.
Target architecture recipes use the genuine
Texinfo utilities.
By default, they use the Texinfo utilities on the host system.
ASSUME_PROVIDED
and makeinfo from
SANITY_REQUIRED_UTILITIES
.
tinderclient.bbclass
¶
The tinderclient
class submits build results to
an external Tinderbox instance.
toaster.bbclass
¶
The toaster
class collects information about
packages and images and sends them as events that the BitBake
user interface can receive.
The class is enabled when the Toaster user interface is running.
This class is not intended to be used directly.
toolchain-scripts.bbclass
¶
The toolchain-scripts
class provides the scripts
used for setting up the environment for installed SDKs.
typecheck.bbclass
¶
The typecheck
class provides support for
validating the values of variables set at the configuration level
against their defined types.
The OpenEmbedded build system allows you to define the type of a
variable using the "type" varflag.
Here is an example:
IMAGE_FEATURES[type] = "list"
uboot-config.bbclass
¶
The uboot-config
class provides support for
U-Boot configuration for a machine.
Specify the machine in your recipe as follows:
UBOOT_CONFIG ??= <default> UBOOT_CONFIG[foo] = "config,images"
You can also specify the machine using this method:
UBOOT_MACHINE = "config"
See the
UBOOT_CONFIG
and
UBOOT_MACHINE
variables for additional information.
uninative.bbclass
¶
Attempts to isolate the build system from the host
distribution's C library in order to make re-use of native shared state
artifacts across different host distributions practical.
With this class enabled, a tarball containing a pre-built C library
is downloaded at the start of the build.
In the Poky reference distribution this is enabled by default
through
meta/conf/distro/include/yocto-uninative.inc
.
Other distributions that do not derive from poky can also
"require conf/distro/include/yocto-uninative.inc
"
to use this.
Alternatively if you prefer, you can build the uninative-tarball recipe
yourself, publish the resulting tarball (e.g. via HTTP) and set
UNINATIVE_URL
and
UNINATIVE_CHECKSUM
appropriately.
For an example, see the
meta/conf/distro/include/yocto-uninative.inc
.
The uninative
class is also used unconditionally
by the extensible SDK.
When building the extensible SDK,
uninative-tarball
is built and the resulting
tarball is included within the SDK.
update-alternatives.bbclass
¶
The update-alternatives
class helps the
alternatives system when multiple sources provide the same command.
This situation occurs when several programs that have the same or
similar function are installed with the same name.
For example, the ar
command is available from the
busybox
, binutils
and
elfutils
packages.
The update-alternatives
class handles
renaming the binaries so that multiple packages can be installed
without conflicts.
The ar
command still works regardless of which
packages are installed or subsequently removed.
The class renames the conflicting binary in each package and symlinks
the highest priority binary during installation or removal of packages.
To use this class, you need to define a number of variables:
These variables list alternative commands needed by a package,
provide pathnames for links, default links for targets, and
so forth.
For details on how to use this class, see the comments in the
update-alternatives.bbclass
file.
update-alternatives
command
directly in your recipes.
However, this class simplifies things in most cases.
update-rc.d.bbclass
¶
The update-rc.d
class uses
update-rc.d
to safely install an
initialization script on behalf of the package.
The OpenEmbedded build system takes care of details such as making
sure the script is stopped before a package is removed and started when
the package is installed.
Three variables control this class:
INITSCRIPT_PACKAGES
,
INITSCRIPT_NAME
and
INITSCRIPT_PARAMS
.
See the variable links for details.
useradd*.bbclass
¶
The useradd*
classes support the addition of users
or groups for usage by the package on the target.
For example, if you have packages that contain system services that
should be run under their own user or group, you can use these classes
to enable creation of the user or group.
The
meta-skeleton/recipes-skeleton/useradd/useradd-example.bb
recipe in the Source Directory
provides a simple example that shows how to add three
users and groups to two packages.
See the useradd-example.bb
recipe for more
information on how to use these classes.
The useradd_base
class provides basic
functionality for user or groups settings.
The useradd*
classes support the
USERADD_PACKAGES
,
USERADD_PARAM
,
GROUPADD_PARAM
,
and
GROUPMEMS_PARAM
variables.
The useradd-staticids
class supports the addition
of users or groups that have static user identification
(uid
) and group identification
(gid
) values.
The default behavior of the OpenEmbedded build system for assigning
uid
and gid
values when
packages add users and groups during package install time is to
add them dynamically.
This works fine for programs that do not care what the values of the
resulting users and groups become.
In these cases, the order of the installation determines the final
uid
and gid
values.
However, if non-deterministic
uid
and gid
values are a
problem, you can override the default, dynamic application of these
values by setting static values.
When you set static values, the OpenEmbedded build system looks in
BBPATH
for
files/passwd
and files/group
files for the values.
To use static uid
and gid
values, you need to set some variables.
See the
USERADDEXTENSION
,
USERADD_UID_TABLES
,
USERADD_GID_TABLES
,
and
USERADD_ERROR_DYNAMIC
variables.
You can also see the
useradd
class for additional information.
useradd-staticids
class directly.
You either enable or disable the class by setting the
USERADDEXTENSION
variable.
If you enable or disable the class in a configured system,
TMPDIR
might contain incorrect uid
and
gid
values.
Deleting the TMPDIR
directory
will correct this condition.
utility-tasks.bbclass
¶
The utility-tasks
class provides support for
various "utility" type tasks that are applicable to all recipes,
such as
do_clean
and
do_listtasks
.
This class is enabled by default because it is inherited by
the
base
class.
utils.bbclass
¶
The utils
class provides some useful Python
functions that are typically used in inline Python expressions
(e.g. ${@...}
).
One example use is for bb.utils.contains()
.
This class is enabled by default because it is inherited by the
base
class.
vala.bbclass
¶
The vala
class supports recipes that need to
build software written using the Vala programming language.
waf.bbclass
¶
The waf
class supports recipes that need to build
software that uses the Waf build system.
You can use the
EXTRA_OECONF
or
PACKAGECONFIG_CONFARGS
variables to specify additional configuration options to be passed on
the Waf command line.
do_build
do_compile
do_compile_ptest_base
do_configure
do_configure_ptest_base
do_deploy
do_fetch
do_image
do_image_complete
do_install
do_install_ptest_base
do_package
do_package_qa
do_package_write_deb
do_package_write_ipk
do_package_write_rpm
do_package_write_tar
do_packagedata
do_patch
do_populate_lic
do_populate_sdk
do_populate_sysroot
do_prepare_recipe_sysroot
do_rm_work
do_unpack
do_compile_kernelmodules
do_diffconfig
do_kernel_checkout
do_kernel_configcheck
do_kernel_configme
do_kernel_menuconfig
do_kernel_metadata
do_menuconfig
do_savedefconfig
do_shared_workdir
do_sizecheck
do_strip
do_validate_branches
Tasks are units of execution for BitBake.
Recipes (.bb
files) use tasks to complete
configuring, compiling, and packaging software.
This chapter provides a reference of the tasks defined in the
OpenEmbedded build system.
The following sections describe normal tasks associated with building a recipe. For more information on tasks and dependencies, see the "Tasks" and "Dependencies" sections in the BitBake User Manual.
do_build
¶The default task for all recipes. This task depends on all other normal tasks required to build a recipe.
do_compile
¶
Compiles the source code.
This task runs with the current working directory set
to
${
B
}
.
The default behavior of this task is to run the
oe_runmake
function if a makefile
(Makefile
, makefile
,
or GNUmakefile
) is found.
If no such file is found, the do_compile
task does nothing.
do_compile_ptest_base
¶Compiles the runtime test suite included in the software being built.
do_configure
¶
Configures the source by enabling and disabling any build-time and
configuration options for the software being built.
The task runs with the current working directory set to
${
B
}
.
The default behavior of this task is to run
oe_runmake clean
if a makefile
(Makefile
, makefile
,
or GNUmakefile
) is found and
CLEANBROKEN
is not set to "1".
If no such file is found or the CLEANBROKEN
variable is set to "1", the do_configure
task does nothing.
do_configure_ptest_base
¶Configures the runtime test suite included in the software being built.
do_deploy
¶
Writes output files that are to be deployed to
${
DEPLOY_DIR_IMAGE
}
.
The task runs with the current working directory set to
${
B
}
.
Recipes implementing this task should inherit the
deploy
class and should write the output to
${
DEPLOYDIR
}
,
which is not to be confused with ${DEPLOY_DIR}
.
The deploy
class sets up
do_deploy
as a shared state (sstate) task that
can be accelerated through sstate use.
The sstate mechanism takes care of copying the output from
${DEPLOYDIR}
to
${DEPLOY_DIR_IMAGE}
.
${DEPLOY_DIR_IMAGE}
, as this causes
the sstate mechanism to malfunction.
The do_deploy
task is not added as a task
by default and consequently needs to be added manually.
If you want the task to run after
do_compile
,
you can add it by doing the following:
addtask deploy after do_compile
Adding do_deploy
after other tasks works the
same way.
before do_build
to the addtask
command (though it is
harmless), because the
base
class contains the following:
do_build[recrdeptask] += "do_deploy"See the "Dependencies" section in the BitBake User Manual for more information.
If the do_deploy
task re-executes, any
previous output is removed (i.e. "cleaned").
do_fetch
¶
Fetches the source code.
This task uses the
SRC_URI
variable and the argument's prefix to determine the correct
fetcher
module.
do_image
¶
Starts the image generation process.
The do_image
task runs after the
OpenEmbedded build system has run the
do_rootfs
task during which packages are identified for installation into
the image and the root filesystem is created, complete with
post-processing.
The do_image
task performs pre-processing
on the image through the
IMAGE_PREPROCESS_COMMAND
and dynamically generates supporting
do_image_*
tasks as needed.
For more information on image creation, see the "Image Generation" section in the Yocto Project Overview and Concepts Manual.
do_image_complete
¶
Completes the image generation process.
The do_image_complete
task runs after the
OpenEmbedded build system has run the
do_image
task during which image pre-processing occurs and through
dynamically generated do_image_*
tasks the
image is constructed.
The do_image_complete
task performs
post-processing on the image through the
IMAGE_POSTPROCESS_COMMAND
.
For more information on image creation, see the "Image Generation" section in the Yocto Project Overview and Concepts Manual.
do_install
¶
Copies files that are to be packaged into the holding area
${
D
}
.
This task runs with the current working directory set to
${
B
}
,
which is the compilation directory.
The do_install
task, as well as other tasks
that either directly or indirectly depend on the installed files
(e.g.
do_package
,
do_package_write_*
,
and
do_rootfs
),
run under
fakeroot.
When installing files, be careful not to set the owner and
group IDs of the installed files to unintended values.
Some methods of copying files, notably when using the
recursive cp
command, can preserve the
UID and/or GID of the original file, which is usually not
what you want.
The
host-user-contaminated
QA check checks for files that probably have the wrong
ownership.
Safe methods for installing files include the following:
The install
utility.
This utility is the preferred method.
The cp
command with the
"--no-preserve=ownership" option.
The tar
command with the
"--no-same-owner" option.
See the bin_package.bbclass
file in the meta/classes
directory of the
Source Directory
for an example.
do_install_ptest_base
¶Copies the runtime test suite files from the compilation directory to a holding area.
do_package
¶
Analyzes the content of the holding area
${
D
}
and splits the content into subsets based on available packages
and files.
This task makes use of the
PACKAGES
and
FILES
variables.
The do_package
task, in conjunction with the
do_packagedata
task, also saves some important package metadata.
For additional information, see the
PKGDESTWORK
variable and the
"Automatically Added Runtime Dependencies"
section in the Yocto Project Overview and Concepts Manual.
do_package_qa
¶
Runs QA checks on packaged files.
For more information on these checks, see the
insane
class.
do_package_write_deb
¶
Creates Debian packages (i.e. *.deb
files) and
places them in the
${
DEPLOY_DIR_DEB
}
directory in the package feeds area.
For more information, see the
"Package Feeds"
section in the Yocto Project Overview and Concepts Manual.
do_package_write_ipk
¶
Creates IPK packages (i.e. *.ipk
files) and
places them in the
${
DEPLOY_DIR_IPK
}
directory in the package feeds area.
For more information, see the
"Package Feeds"
section in the Yocto Project Overview and Concepts Manual.
do_package_write_rpm
¶
Creates RPM packages (i.e. *.rpm
files) and
places them in the
${
DEPLOY_DIR_RPM
}
directory in the package feeds area.
For more information, see the
"Package Feeds"
section in the Yocto Project Overview and Concepts Manual.
do_package_write_tar
¶
Creates tarballs and places them in the
${
DEPLOY_DIR_TAR
}
directory in the package feeds area.
For more information, see the
"Package Feeds"
section in the Yocto Project Overview and Concepts Manual.
do_packagedata
¶
Saves package metadata generated by the
do_package
task in
PKGDATA_DIR
to make it available globally.
do_patch
¶Locates patch files and applies them to the source code.
After fetching and unpacking source files, the build system
uses the recipe's
SRC_URI
statements to locate and apply patch files to the source code.
FILESPATH
variable to determine the default set of directories when
searching for patches.
Patch files, by default, are *.patch
and
*.diff
files created and kept in a
subdirectory of the directory holding the recipe file.
For example, consider the
bluez5
recipe from the OE-Core layer (i.e.
poky/meta
):
poky/meta/recipes-connectivity/bluez5
This recipe has two patch files located here:
poky/meta/recipes-connectivity/bluez5/bluez5
In the bluez5
recipe, the
SRC_URI
statements point to the source and
patch files needed to build the package.
bluez5_5.48.bb
recipe, the SRC_URI
statements are from an
include file bluez5.inc
.
As mentioned earlier, the build system treats files whose file
types are .patch
and
.diff
as patch files.
However, you can use the "apply=yes" parameter with the
SRC_URI
statement to indicate any file as a
patch file:
SRC_URI = " \ git://path_to_repo
/some_package
\ file://file
;apply=yes \ "
Conversely, if you have a directory full of patch files and you
want to exclude some so that the do_patch
task does not apply them during the patch phase, you can use
the "apply=no" parameter with the SRC_URI
statement:
SRC_URI = " \ git://path_to_repo
/some_package
\ file://path_to_lots_of_patch_files
\ file://path_to_lots_of_patch_files
/patch_file5
;apply=no \ "
In the previous example, assuming all the files in the directory
holding the patch files end with either .patch
or .diff
, every file would be applied as a
patch by default except for the
patch_file5
patch.
You can find out more about the patching process in the "Patching" section in the Yocto Project Overview and Concepts Manual and the "Patching Code" section in the Yocto Project Development Tasks Manual.
do_populate_lic
¶Writes license information for the recipe that is collected later when the image is constructed.
do_populate_sdk
¶Creates the file and directory structure for an installable SDK. See the "SDK Generation" section in the Yocto Project Overview and Concepts Manual for more information.
do_populate_sysroot
¶
Stages (copies) a subset of the files installed by the
do_install
task into the appropriate sysroot.
For information on how to access these files from other recipes,
see the
STAGING_DIR*
variables.
Directories that would typically not be needed by other recipes at
build time (e.g. /etc
) are not copied by
default.
For information on what directories are copied by default, see the
SYSROOT_DIRS*
variables.
You can change these variables inside your recipe if you need
to make additional (or fewer) directories available to other
recipes at build time.
The do_populate_sysroot
task is a
shared state (sstate) task, which means that the task can
be accelerated through sstate use.
Realize also that if the task is re-executed, any previous output
is removed (i.e. "cleaned").
do_prepare_recipe_sysroot
¶
Installs the files into the individual recipe specific sysroots
(i.e. recipe-sysroot
and
recipe-sysroot-native
under
${
WORKDIR
}
based upon the dependencies specified by
DEPENDS
).
See the
"staging
"
class for more information.
do_rm_work
¶
Removes work files after the OpenEmbedded build system has
finished with them.
You can learn more by looking at the
"rm_work.bbclass
"
section.
do_unpack
¶
Unpacks the source code into a working directory pointed to
by
${
WORKDIR
}
.
The
S
variable also
plays a role in where unpacked source files ultimately reside.
For more information on how source files are unpacked, see the
"Source Fetching"
section in the Yocto Project Overview and Concepts Manual and also
see the WORKDIR
and
S
variable descriptions.
These tasks are typically manually triggered (e.g. by using the
bitbake -c
command-line option):
do_checkpkg
¶Provides information about the recipe including its upstream version and status. The upstream version and status reveals whether or not a version of the recipe exists upstream and a status of not updated, updated, or unknown.
To check the upstream version and status of a recipe, use the following devtool commands:
$ devtool latest-version $ devtool check-upgrade-status
See the
"devtool
Quick Reference"
chapter for more information on devtool
.
See the
"Checking on the Upgrade Status of a Recipe"
section for information on checking the upgrade status of a recipe.
To build the checkpkg
task, use the
bitbake
command with the "-c" option and
task name:
$ bitbake core-image-minimal -c checkpkg
By default, the results are stored in
$LOG_DIR
(e.g. $BUILD_DIR/tmp/log
).
do_clean
¶
Removes all output files for a target from the
do_unpack
task forward (i.e. do_unpack
,
do_configure
,
do_compile
,
do_install
,
and
do_package
).
You can run this task using BitBake as follows:
$ bitbake -c clean recipe
Running this task does not remove the
sstate
cache files.
Consequently, if no changes have been made and the recipe is
rebuilt after cleaning, output files are simply restored from the
sstate cache.
If you want to remove the sstate cache files for the recipe,
you need to use the
do_cleansstate
task instead (i.e. bitbake -c cleansstate
recipe
).
do_cleanall
¶
Removes all output files, shared state
(sstate)
cache, and downloaded source files for a target (i.e. the contents
of
DL_DIR
).
Essentially, the do_cleanall
task is
identical to the
do_cleansstate
task with the added removal of downloaded source files.
You can run this task using BitBake as follows:
$ bitbake -c cleanall recipe
Typically, you would not normally use the
cleanall
task.
Do so only if you want to start fresh with the
do_fetch
task.
do_cleansstate
¶
Removes all output files and shared state
(sstate)
cache for a target.
Essentially, the do_cleansstate
task is
identical to the
do_clean
task with the added removal of shared state
(sstate)
cache.
You can run this task using BitBake as follows:
$ bitbake -c cleansstate recipe
When you run the do_cleansstate
task,
the OpenEmbedded build system no longer uses any
sstate.
Consequently, building the recipe from scratch is guaranteed.
do_cleansstate
task cannot remove
sstate from a remote sstate mirror.
If you need to build a target from scratch using remote
mirrors, use the "-f" option as follows:
$ bitbake -f -c do_cleansstate target
do_devpyshell
¶
Starts a shell in which an interactive Python interpreter allows
you to interact with the BitBake build environment.
From within this shell, you can directly examine and set
bits from the data store and execute functions as if within
the BitBake environment.
See the
"Using a Development Python Shell"
section in the Yocto Project Development Tasks Manual for more
information about using devpyshell
.
do_devshell
¶
Starts a shell whose environment is set up for
development, debugging, or both.
See the
"Using a Development Shell"
section in the Yocto Project Development Tasks Manual for more
information about using devshell
.
do_listtasks
¶Lists all defined tasks for a target.
do_package_index
¶Creates or updates the index in the Package Feeds area.
bitbake -c
command-line option as
are the other tasks in this section.
Because this task is specifically for the
package-index
recipe,
you run it using
bitbake package-index
.
The following tasks are applicable to image recipes.
do_bootimg
¶
Creates a bootable live image.
See the
IMAGE_FSTYPES
variable for additional information on live image types.
do_bundle_initramfs
¶
Combines an initial RAM disk (initramfs) image and kernel
together to form a single image.
The
CONFIG_INITRAMFS_SOURCE
variable has some more information about these types of images.
do_rootfs
¶Creates the root filesystem (file and directory structure) for an image. See the "Image Generation" section in the Yocto Project Overview and Concepts Manual for more information on how the root filesystem is created.
do_testimage
¶Boots an image and performs runtime tests within the image. For information on automatically testing images, see the "Performing Automated Runtime Testing" section in the Yocto Project Development Tasks Manual.
do_testimage_auto
¶
Boots an image and performs runtime tests within the image
immediately after it has been built.
This task is enabled when you set
TESTIMAGE_AUTO
equal to "1".
For information on automatically testing images, see the "Performing Automated Runtime Testing" section in the Yocto Project Development Tasks Manual.
The following tasks are applicable to kernel recipes.
Some of these tasks (e.g. the
do_menuconfig
task) are also applicable to recipes that use
Linux kernel style configuration such as the BusyBox recipe.
do_compile_kernelmodules
¶
Runs the step that builds the kernel modules (if needed).
Building a kernel consists of two steps: 1) the kernel
(vmlinux
) is built, and 2) the modules
are built (i.e. make modules
).
do_diffconfig
¶
When invoked by the user, this task creates a file containing the
differences between the original config as produced by
do_kernel_configme
task and the changes made by the user with other methods
(i.e. using
(do_kernel_menuconfig
).
Once the file of differences is created, it can be used to create
a config fragment that only contains the differences.
You can invoke this task from the command line as follows:
$ bitbake linux-yocto -c diffconfig
For more information, see the "Creating Configuration Fragments" section in the Yocto Project Linux Kernel Development Manual.
do_kernel_checkout
¶
Converts the newly unpacked kernel source into a form with which
the OpenEmbedded build system can work.
Because the kernel source can be fetched in several different ways,
the do_kernel_checkout
task makes sure that
subsequent tasks are given a clean working tree copy of the kernel
with the correct branches checked out.
do_kernel_configcheck
¶
Validates the configuration produced by the
do_kernel_menuconfig
task.
The do_kernel_configcheck
task produces
warnings when a requested configuration does not appear in the
final .config
file or when you override a
policy configuration in a hardware configuration fragment.
You can run this task explicitly and view the output by using
the following command:
$ bitbake linux-yocto -c kernel_configcheck -f
For more information, see the "Validating Configuration" section in the Yocto Project Linux Kernel Development Manual.
do_kernel_configme
¶
After the kernel is patched by the
do_patch
task, the do_kernel_configme
task assembles
and merges all the kernel config fragments into a merged
configuration that can then be passed to the kernel configuration
phase proper.
This is also the time during which user-specified defconfigs
are applied if present, and where configuration modes such as
--allnoconfig
are applied.
do_kernel_menuconfig
¶
Invoked by the user to manipulate the
.config
file used to build a linux-yocto
recipe.
This task starts the Linux kernel configuration tool, which you
then use to modify the kernel configuration.
$ bitbake linux-yocto -c menuconfig
See the
"Using menuconfig
"
section in the Yocto Project Linux Kernel Development Manual
for more information on this configuration tool.
do_kernel_metadata
¶
Collects all the features required for a given kernel build,
whether the features come from
SRC_URI
or from Git repositories.
After collection, the do_kernel_metadata
task
processes the features into a series of config fragments and
patches, which can then be applied by subsequent tasks such as
do_patch
and
do_kernel_configme
.
do_menuconfig
¶
Runs make menuconfig
for the kernel.
For information on menuconfig
, see the
"Using menuconfig
"
section in the Yocto Project Linux Kernel Development Manual.
do_savedefconfig
¶
When invoked by the user, creates a defconfig file that can be
used instead of the default defconfig.
The saved defconfig contains the differences between the default
defconfig and the changes made by the user using other methods
(i.e. the
do_kernel_menuconfig
task.
You can invoke the task using the following command:
$ bitbake linux-yocto -c savedefconfig
do_shared_workdir
¶
After the kernel has been compiled but before the kernel modules
have been compiled, this task copies files required for module
builds and which are generated from the kernel build into the
shared work directory.
With these copies successfully copied, the
do_compile_kernelmodules
task can successfully build the kernel modules in the next step
of the build.
do_sizecheck
¶
After the kernel has been built, this task checks the size of the
stripped kernel image against
KERNEL_IMAGE_MAXSIZE
.
If that variable was set and the size of the stripped kernel
exceeds that size, the kernel build produces a warning to that
effect.
do_strip
¶
If
KERNEL_IMAGE_STRIP_EXTRA_SECTIONS
is defined,
this task strips the sections named in that variable from
vmlinux
.
This stripping is typically used to remove nonessential sections
such as .comment
sections from a
size-sensitive configuration.
do_validate_branches
¶
After the kernel is unpacked but before it is patched, this task
makes sure that the machine and metadata branches as specified
by the SRCREV
variables actually exist on the specified branches.
If these branches do not exist and
AUTOREV
is not being used, the do_validate_branches
task fails during the build.
devtool
Quick Reference¶
The devtool
command-line tool provides a number
of features that help you build, test, and package software.
This command is available alongside the bitbake
command.
Additionally, the devtool
command is a key
part of the extensible SDK.
This chapter provides a Quick Reference for the
devtool
command.
For more information on how to apply the command when using the
extensible SDK, see the
"Using the Extensible SDK"
chapter in the Yocto Project Application Development and the
Extensible Software Development Kit (eSDK) manual.
The devtool
command line is organized
similarly to Git in that it has a number of sub-commands for
each function.
You can run devtool --help
to see all
the commands:
$ devtool -h NOTE: Starting bitbake server... usage: devtool [--basepath BASEPATH] [--bbpath BBPATH] [-d] [-q] [--color COLOR] [-h] <subcommand> ... OpenEmbedded development tool options: --basepath BASEPATH Base directory of SDK / build directory --bbpath BBPATH Explicitly specify the BBPATH, rather than getting it from the metadata -d, --debug Enable debug output -q, --quiet Print only errors --color COLOR Colorize output (where COLOR is auto, always, never) -h, --help show this help message and exit subcommands: Beginning work on a recipe: add Add a new recipe modify Modify the source for an existing recipe upgrade Upgrade an existing recipe Getting information: status Show workspace status search Search available recipes latest-version Report the latest version of an existing recipe check-upgrade-status Report upgradability for multiple (or all) recipes Working on a recipe in the workspace: build Build a recipe rename Rename a recipe file in the workspace edit-recipe Edit a recipe file find-recipe Find a recipe file configure-help Get help on configure script options update-recipe Apply changes from external source tree to recipe reset Remove a recipe from your workspace finish Finish working on a recipe in your workspace Testing changes on target: deploy-target Deploy recipe output files to live target machine undeploy-target Undeploy recipe output files in live target machine build-image Build image including workspace recipe packages Advanced: create-workspace Set up workspace in an alternative location export Export workspace into a tar archive import Import exported tar archive into workspace extract Extract the source for an existing recipe sync Synchronize the source tree for an existing recipe Use devtool <subcommand> --help to get help on a specific command
As directed in the general help output, you can get more syntax on a specific command by providing the command name and using "--help":
$ devtool add --help NOTE: Starting bitbake server... usage: devtool add [-h] [--same-dir | --no-same-dir] [--fetch URI] [--fetch-dev] [--version VERSION] [--no-git] [--srcrev SRCREV | --autorev] [--srcbranch SRCBRANCH] [--binary] [--also-native] [--src-subdir SUBDIR] [--mirrors] [--provides PROVIDES] [recipename] [srctree] [fetchuri] Adds a new recipe to the workspace to build a specified source tree. Can optionally fetch a remote URI and unpack it to create the source tree. arguments: recipename Name for new recipe to add (just name - no version, path or extension). If not specified, will attempt to auto-detect it. srctree Path to external source tree. If not specified, a subdirectory of /home/scottrif/poky/build/workspace/sources will be used. fetchuri Fetch the specified URI and extract it to create the source tree options: -h, --help show this help message and exit --same-dir, -s Build in same directory as source --no-same-dir Force build in a separate build directory --fetch URI, -f URI Fetch the specified URI and extract it to create the source tree (deprecated - pass as positional argument instead) --fetch-dev For npm, also fetch devDependencies --version VERSION, -V VERSION Version to use within recipe (PV) --no-git, -g If fetching source, do not set up source tree as a git repository --srcrev SRCREV, -S SRCREV Source revision to fetch if fetching from an SCM such as git (default latest) --autorev, -a When fetching from a git repository, set SRCREV in the recipe to a floating revision instead of fixed --srcbranch SRCBRANCH, -B SRCBRANCH Branch in source repository if fetching from an SCM such as git (default master) --binary, -b Treat the source tree as something that should be installed verbatim (no compilation, same directory structure). Useful with binary packages e.g. RPMs. --also-native Also add native variant (i.e. support building recipe for the build host as well as the target machine) --src-subdir SUBDIR Specify subdirectory within source tree to use --mirrors Enable PREMIRRORS and MIRRORS for source tree fetching (disable by default). --provides PROVIDES, -p PROVIDES Specify an alias for the item provided by the recipe. E.g. virtual/libgl
devtool
uses a "Workspace" layer
in which to accomplish builds.
This layer is not specific to any single
devtool
command but is rather a common
working area used across the tool.
The following figure shows the workspace structure:
attic - A directory created if devtool believes it must preserve
anything when you run "devtool reset". For example, if you
run "devtool add", make changes to the recipe, and then
run "devtool reset", devtool takes notice that the file has
been changed and moves it into the attic should you still
want the recipe.
README - Provides information on what is in workspace layer and how to
manage it.
.devtool_md5 - A checksum file used by devtool.
appends - A directory that contains *.bbappend files, which point to
external source.
conf - A configuration directory that contains the layer.conf file.
recipes - A directory containing recipes. This directory contains a
folder for each directory added whose name matches that of the
added recipe. devtool places the recipe
.bb file
within that sub-directory.
sources - A directory containing a working copy of the source files used
when building the recipe. This is the default directory used
as the location of the source tree when you do not provide a
source tree path. This directory contains a folder for each
set of source files matched to a corresponding recipe.
Use the devtool add
command to add a new recipe
to the workspace layer.
The recipe you add should not exist -
devtool
creates it for you.
The source files the recipe uses should exist in an external
area.
The following example creates and adds a new recipe named
jackson
to a workspace layer the tool creates.
The source code built by the recipes resides in
/home/
:
user
/sources/jackson
$ devtool add jackson /home/user
/sources/jackson
If you add a recipe and the workspace layer does not exist, the command creates the layer and populates it as described in "The Workspace Layer Structure" section.
Running devtool add
when the
workspace layer exists causes the tool to add the recipe,
append files, and source files into the existing workspace layer.
The .bbappend
file is created to point
to the external source tree.
By default, devtool add
uses the latest
revision (i.e. master) when unpacking files from a remote URI.
In some cases, you might want to specify a source revision by
branch, tag, or commit hash. You can specify these options when
using the devtool add
command:
To specify a source branch, use the
--srcbranch
option:
$ devtool add --srcbranch dunfell jackson /home/user
/sources/jackson
In the previous example, you are checking out the dunfell branch.
To specify a specific tag or commit hash, use the
--srcrev
option:
$ devtool add --srcrev yocto-3.1.2 jackson /home/user
/sources/jackson $ devtool add --srcrevsome_commit_hash
/home/user
/sources/jackson
The previous examples check out the yocto-3.1.2 tag
and the commit associated with the
some_commit_hash
hash.
--autorev
or -a
.
Use the devtool extract
command to
extract the source for an existing recipe.
When you use this command, you must supply the root name
of the recipe (i.e. no version, paths, or extensions), and
you must supply the directory to which you want the source
extracted.
Additional command options let you control the name of a development branch into which you can checkout the source and whether or not to keep a temporary directory, which is useful for debugging.
Use the devtool sync
command to
synchronize a previously extracted source tree for an
existing recipe.
When you use this command, you must supply the root name
of the recipe (i.e. no version, paths, or extensions), and
you must supply the directory to which you want the source
extracted.
Additional command options let you control the name of a development branch into which you can checkout the source and whether or not to keep a temporary directory, which is useful for debugging.
Use the devtool modify
command to begin
modifying the source of an existing recipe.
This command is very similar to the
add
command except that it does not physically create the
recipe in the workspace layer because the recipe already
exists in an another layer.
The devtool modify
command extracts the
source for a recipe, sets it up as a Git repository if the
source had not already been fetched from Git, checks out a
branch for development, and applies any patches from the recipe
as commits on top.
You can use the following command to checkout the source
files:
$ devtool modify recipe
Using the above command form, devtool
uses
the existing recipe's
SRC_URI
statement to locate the upstream source, extracts the source
into the default sources location in the workspace.
The default development branch used is "devtool".
Use the devtool edit-recipe
command
to run the default editor, which is identified using the
EDITOR
variable, on the specified recipe.
When you use the devtool edit-recipe
command, you must supply the root name of the recipe
(i.e. no version, paths, or extensions).
Also, the recipe file itself must reside in the workspace
as a result of the devtool add
or
devtool upgrade
commands.
However, you can override that requirement by using the
"-a" or "--any-recipe" option.
Using either of these options allows you to edit any recipe
regardless of its location.
Use the devtool update-recipe
command to
update your recipe with patches that reflect changes you make
to the source files.
For example, if you know you are going to work on some
code, you could first use the
devtool modify
command to extract the code and set up the workspace.
After which, you could modify, compile, and test the code.
When you are satisfied with the results and you have committed
your changes to the Git repository, you can then
run the devtool update-recipe
to create the
patches and update the recipe:
$ devtool update-recipe recipe
If you run the devtool update-recipe
without committing your changes, the command ignores the
changes.
Often, you might want to apply customizations made to your
software in your own layer rather than apply them to the
original recipe.
If so, you can use the
-a
or --append
option with the devtool update-recipe
command.
These options allow you to specify the layer into which to
write an append file:
$ devtool update-reciperecipe
-abase-layer-directory
The *.bbappend
file is created at the
appropriate path within the specified layer directory, which
may or may not be in your bblayers.conf
file.
If an append file already exists, the command updates it
appropriately.
Upstream recipes change over time. Consequently, you might find that you need to determine if you can upgrade a recipe to a newer version.
To check on the upgrade status of a recipe, use the
devtool check-upgrade-status
command.
The command displays a table of your current recipe versions,
the latest upstream versions, the email address of the recipe's
maintainer, and any additional information such as commit hash
strings and reasons you might not be able to upgrade a particular
recipe.
For the oe-core
layer, recipe
maintainers come from the
maintainers.inc
file.
If the recipe is using the Git fetcher rather than a tarball, the commit hash points to the commit that matches the recipe's latest version tag.
As with all devtool
commands, you can get
help on the individual command:
$ devtool check-upgrade-status -h NOTE: Starting bitbake server... usage: devtool check-upgrade-status [-h] [--all] [recipe [recipe ...]] Prints a table of recipes together with versions currently provided by recipes, and latest upstream versions, when there is a later version available arguments: recipe Name of the recipe to report (omit to report upgrade info for all recipes) options: -h, --help show this help message and exit --all, -a Show all recipes, not just recipes needing upgrade
Unless you provide a specific recipe name on the command line, the command checks all recipes in all configured layers.
Following is a partial example table that reports on all the
recipes.
Notice the reported reason for not upgrading the
base-passwd
recipe.
In this example, while a new version is available upstream,
you do not want to use it because the dependency on
cdebconf
is not easily satisfied.
RECIPE_NO_UPDATE_REASON
variable.
See the
base-passwd.bb
recipe for an example.
$ devtool check-upgrade-status ... NOTE: acpid 2.0.30 2.0.31 Ross Burton <ross.burton@intel.com> NOTE: u-boot-fw-utils 2018.11 2019.01 Marek Vasut <marek.vasut@gmail.com> d3689267f92c5956e09cc7d1baa4700141662bff NOTE: u-boot-tools 2018.11 2019.01 Marek Vasut <marek.vasut@gmail.com> d3689267f92c5956e09cc7d1baa4700141662bff . . . NOTE: base-passwd 3.5.29 3.5.45 Anuj Mittal <anuj.mittal@intel.com> cannot be updated due to: Version 3.5.38 requires cdebconf for update-passwd utility NOTE: busybox 1.29.2 1.30.0 Andrej Valek <andrej.valek@siemens.com> NOTE: dbus-test 1.12.10 1.12.12 Chen Qi <Qi.Chen@windriver.com>
As software matures, upstream recipes are upgraded to newer
versions.
As a developer, you need to keep your local recipes up-to-date
with the upstream version releases.
Several methods exist by which you can upgrade recipes.
You can read about them in the
"Upgrading Recipes"
section of the Yocto Project Development Tasks Manual.
This section overviews the devtool upgrade
command.
The devtool upgrade
command
upgrades an existing recipe to a more recent version of the
recipe upstream.
The command puts the upgraded recipe file along with any associated
files into a "workspace" and, if necessary, extracts the source
tree to a specified location.
During the upgrade, patches associated with the recipe are
rebased or added as needed.
When you use the devtool upgrade
command,
you must supply the root name of the recipe (i.e. no version,
paths, or extensions), and you must supply the directory
to which you want the source extracted.
Additional command options let you control things such as
the version number to which you want to upgrade (i.e. the
PV
),
the source revision to which you want to upgrade (i.e. the
SRCREV
),
whether or not to apply patches, and so forth.
You can read more on the devtool upgrade
workflow in the
"Use devtool upgrade
to Create a Version of the Recipe that Supports a Newer Version of the Software"
section in the Yocto Project Application Development and the
Extensible Software Development Kit (eSDK) manual.
You can also see an example of how to use
devtool upgrade
in the
"Using devtool upgrade
"
section in the Yocto Project Development Tasks Manual.
Use the devtool reset
command to remove a
recipe and its configuration (e.g. the corresponding
.bbappend
file) from the workspace layer.
Realize that this command deletes the recipe and the
append file.
The command does not physically move them for you.
Consequently, you must be sure to physically relocate your
updated recipe and the append file outside of the workspace
layer before running the devtool reset
command.
If the devtool reset
command detects that
the recipe or the append files have been modified, the
command preserves the modified files in a separate "attic"
subdirectory under the workspace layer.
Here is an example that resets the workspace directory that
contains the mtr
recipe:
$ devtool reset mtr NOTE: Cleaning sysroot for recipe mtr... NOTE: Leaving source tree /home/scottrif/poky/build/workspace/sources/mtr as-is; if you no longer need it then please delete it manually $
Use the devtool build
command to build your
recipe.
The devtool build
command is equivalent to
the bitbake -c populate_sysroot
command.
When you use the devtool build
command,
you must supply the root name of the recipe (i.e. do not provide
versions, paths, or extensions).
You can use either the "-s" or the "--disable-parallel-make"
options to disable parallel makes during the build.
Here is an example:
$ devtool build recipe
Use the devtool build-image
command
to build an image, extending it to include packages from
recipes in the workspace.
Using this command is useful when you want an image that
ready for immediate deployment onto a device for testing.
For proper integration into a final image, you need to
edit your custom image recipe appropriately.
When you use the devtool build-image
command, you must supply the name of the image.
This command has no command line options:
$ devtool build-image image
Use the devtool deploy-target
command to
deploy the recipe's build output to the live target machine:
$ devtool deploy-targetrecipe
target
The target
is the address of the
target machine, which must be running an SSH server (i.e.
user@hostname[:destdir]
).
This command deploys all files installed during the
do_install
task.
Furthermore, you do not need to have package management enabled
within the target machine.
If you do, the package manager is bypassed.
The deploy-target
functionality is for development only.
You should never use it to update an image that will be
used in production.
Some conditions exist that could prevent a deployed application from behaving as expected. When both of the following conditions exist, your application has the potential to not behave correctly when run on the target:
You are deploying a new application to the target and the recipe you used to build the application had correctly defined runtime dependencies.
The target does not physically have the packages on which the application depends installed.
If both of these conditions exist, your application will not
behave as expected.
The reason for this misbehavior is because the
devtool deploy-target
command does not deploy
the packages (e.g. libraries) on which your new application
depends.
The assumption is that the packages are already on the target.
Consequently, when a runtime call is made in the application
for a dependent function (e.g. a library call), the function
cannot be found.
To be sure you have all the dependencies local to the target, you need to be sure that the packages are pre-deployed (installed) on the target before attempting to run your application.
Use the devtool undeploy-target
command to
remove deployed build output from the target machine.
For the devtool undeploy-target
command to
work, you must have previously used the
devtool deploy-target
command.
$ devtool undeploy-targetrecipe
target
The target
is the address of the
target machine, which must be running an SSH server (i.e.
user@hostname
).
Use the devtool create-workspace
command to
create a new workspace layer in your
Build Directory.
When you create a new workspace layer, it is populated with the
README
file and the
conf
directory only.
The following example creates a new workspace layer in your current working and by default names the workspace layer "workspace":
$ devtool create-workspace
You can create a workspace layer anywhere by supplying a pathname with the command. The following command creates a new workspace layer named "new-workspace":
$ devtool create-workspace /home/scottrif/new-workspace
Use the devtool status
command to
list the recipes currently in your workspace.
Information includes the paths to their respective
external source trees.
The devtool status
command has no
command-line options:
$ devtool status
Following is sample output after using
devtool add
to create and add the mtr_0.86.bb
recipe
to the workspace
directory:
$ devtool status mtr: /home/scottrif/poky/build/workspace/sources/mtr (/home/scottrif/poky/build/workspace/recipes/mtr/mtr_0.86.bb) $
Use the devtool search
command to
search for available target recipes.
The command matches the recipe name, package name,
description, and installed files.
The command displays the recipe name as a result of a
match.
When you use the devtool search
command,
you must supply a keyword
.
The command uses the keyword
when
searching for a match.
.wks
) Reference¶
The current Wic implementation supports only the basic kickstart
partitioning commands:
partition
(or part
for short) and bootloader
.
This chapter provides a reference on the available kickstart commands. The information lists the commands, their syntax, and meanings. Kickstart commands are based on the Fedora kickstart versions but with modifications to reflect Wic capabilities. You can see the original documentation for those commands at the following link:
http://pykickstart.readthedocs.io/en/latest/kickstart-docs.html
Either of these commands creates a partition on the system and uses the following syntax:
part [mntpoint
] partition [mntpoint
]
If you do not provide mntpoint
, Wic
creates a partition but does not mount it.
The
is
where the partition is mounted and must be in one of the
following forms:
mntpoint
/
:
For example, "/", "/usr", or "/home"
path
swap
:
The created partition is used as swap space
Specifying a mntpoint
causes the
partition to automatically be mounted.
Wic achieves this by adding entries to the filesystem table (fstab)
during image generation.
In order for Wic to generate a valid fstab, you must also provide
one of the --ondrive
,
--ondisk
, or
--use-uuid
partition options as part of the
command.
--use-uuid
and non-root
mountpoint
, including swap.
The busybox versions of these application are currently
excluded.
Here is an example that uses "/" as the
mountpoint
.
The command uses --ondisk
to force the
partition onto the
sdb
disk:
part / --source rootfs --ondisk sdb --fstype=ext3 --label platform --align 1024
Here is a list that describes other supported options you can use
with the part
and
partition
commands:
--size
:
The minimum partition size in MBytes.
Specify an integer value such as 500.
Do not append the number with "MB".
You do not need this option if you use
--source
.
--fixed-size
:
The exact partition size in MBytes.
You cannot specify with --size
.
An error occurs when assembling the disk image if the
partition data is larger than
--fixed-size
.
--source
:
This option is a Wic-specific option that names the source
of the data that populates the partition.
The most common value for this option is "rootfs", but you
can use any value that maps to a valid source plugin.
For information on the source plugins, see the
"Using the Wic Plugins Interface"
section in the Yocto Project Development Tasks Manual.
If you use --source rootfs
, Wic
creates a partition as large as needed and fills it with
the contents of the root filesystem pointed to by the
-r
command-line option or the
equivalent rootfs derived from the -e
command-line option.
The filesystem type used to create the partition is driven
by the value of the --fstype
option
specified for the partition.
See the entry on --fstype
that follows
for more information.
If you use
--source
,
Wic creates a partition as large as needed and fills it
with the contents of the partition that is generated by the
specified plugin name using the data pointed to by the
plugin-name
-r
command-line option or the
equivalent rootfs derived from the -e
command-line option.
Exactly what those contents are and filesystem type used are
dependent on the given plugin implementation.
If you do not use the --source
option, the wic
command creates an
empty partition.
Consequently, you must use the --size
option to specify the size of the empty partition.
--ondisk
or --ondrive
:
Forces the partition to be created on a particular disk.
--fstype
:
Sets the file system type for the partition.
Valid values are:
ext4
ext3
ext2
btrfs
squashfs
swap
--fsoptions
:
Specifies a free-form string of options to be used when
mounting the filesystem.
This string is copied into the
/etc/fstab
file of the installed
system and should be enclosed in quotes.
If not specified, the default string is "defaults".
--label label
:
Specifies the label to give to the filesystem to be made on
the partition.
If the given label is already in use by another filesystem,
a new label is created for the partition.
--active
:
Marks the partition as active.
--align (in KBytes)
:
This option is a Wic-specific option that says to start
partitions on boundaries given
x
KBytes.
--no-table
:
This option is a Wic-specific option.
Using the option reserves space for the partition and
causes it to become populated.
However, the partition is not added to the partition table.
--exclude-path
:
This option is a Wic-specific option that excludes the given
relative path from the resulting image.
This option is only effective with the rootfs source
plugin.
--extra-space
:
This option is a Wic-specific option that adds extra space
after the space filled by the content of the partition.
The final size can exceed the size specified by the
--size
option.
The default value is 10 Mbytes.
--overhead-factor
:
This option is a Wic-specific option that multiplies the
size of the partition by the option's value.
You must supply a value greater than or equal to "1".
The default value is "1.3".
--part-name
:
This option is a Wic-specific option that specifies a name
for GPT partitions.
--part-type
:
This option is a Wic-specific option that specifies the
partition type globally unique identifier (GUID) for GPT
partitions.
You can find the list of partition type GUIDs at
http://en.wikipedia.org/wiki/GUID_Partition_Table#Partition_type_GUIDs.
--use-uuid
:
This option is a Wic-specific option that causes Wic to
generate a random GUID for the partition.
The generated identifier is used in the bootloader
configuration to specify the root partition.
--uuid
:
This option is a Wic-specific option that specifies the
partition UUID.
--fsuuid
:
This option is a Wic-specific option that specifies the
filesystem UUID.
You can generate or modify
WKS_FILE
with this option if a preconfigured filesystem UUID is
added to the kernel command line in the bootloader
configuration before you run Wic.
--system-id
:
This option is a Wic-specific option that specifies the
partition system ID, which is a one byte long, hexadecimal
parameter with or without the 0x prefix.
--mkfs-extraopts
:
This option specifies additional options to pass to the
mkfs
utility.
Some default options for certain filesystems do not take
effect.
See Wic's help on kickstart
(i.e. wic help kickstart
).
This command specifies how the bootloader should be configured and supports the following options:
--source
plugins that
implement bootloader functionality.
The bootloader command essentially provides a means of
modifying bootloader configuration.
--timeout
:
Specifies the number of seconds before the bootloader times
out and boots the default option.
--append
:
Specifies kernel parameters.
These parameters will be added to the syslinux
APPEND
or grub
kernel command line.
--configfile
:
Specifies a user-defined configuration file for the
bootloader.
You can provide a full pathname for the file or a file that
exists in the canned-wks
folder.
This option overrides all other bootloader options.
When building a recipe, the OpenEmbedded build system performs various QA checks on the output to ensure that common issues are detected and reported. Sometimes when you create a new recipe to build new software, it will build with no problems. When this is not the case, or when you have QA issues building any software, it could take a little time to resolve them.
While it is tempting to ignore a QA message or even to disable QA checks, it is best to try and resolve any reported QA issues. This chapter provides a list of the QA messages and brief explanations of the issues you could encounter so that you can properly resolve problems.
The next section provides a list of all QA error and warning messages based on a default configuration. Each entry provides the message or error form along with an explanation.
At the end of each message, the name of the associated
QA test (as listed in the
"insane.bbclass
"
section) appears within square brackets.
As mentioned, this list of error and warning messages is for QA checks only. The list does not cover all possible build errors or warnings you could encounter.
Because some QA checks are disabled by default, this list does not include all possible QA check errors and warnings.
<packagename>: <path> is using libexec please relocate to <libexecdir> [libexec]
The specified package contains files in
/usr/libexec
when the distro
configuration uses a different path for
<libexecdir>
By default, <libexecdir>
is
$prefix/libexec
.
However, this default can be changed (e.g.
${libdir}
).
package <packagename> contains bad RPATH <rpath> in file <file> [rpaths]
The specified binary produced by the recipe contains dynamic
library load paths (rpaths) that contain build system paths
such as
TMPDIR
,
which are incorrect for the target and could potentially
be a security issue.
Check for bad -rpath
options being
passed to the linker in your
do_compile
log.
Depending on the build system used by the software being
built, there might be a configure option to disable rpath
usage completely within the build of the software.
<packagename>: <file> contains probably-redundant RPATH <rpath> [useless-rpaths]
The specified binary produced by the recipe contains dynamic
library load paths (rpaths) that on a standard system are
searched by default by the linker (e.g.
/lib
and /usr/lib
).
While these paths will not cause any breakage, they do waste
space and are unnecessary.
Depending on the build system used by the software being
built, there might be a configure option to disable rpath
usage completely within the build of the software.
<packagename> requires <files>, but no providers in its RDEPENDS [file-rdeps]
A file-level dependency has been identified from the
specified package on the specified files, but there is
no explicit corresponding entry in
RDEPENDS
.
If particular files are required at runtime then
RDEPENDS
should be declared in the
recipe to ensure the packages providing them are built.
<packagename1> rdepends on <packagename2>, but it isn't a build dependency? [build-deps]
A runtime dependency exists between the two specified
packages, but there is nothing explicit within the recipe
to enable the OpenEmbedded build system to ensure that
dependency is satisfied.
This condition is usually triggered by an
RDEPENDS
value being added at the packaging stage rather than up
front, which is usually automatic based on the contents of
the package.
In most cases, you should change the recipe to add an
explicit RDEPENDS
for the dependency.
non -dev/-dbg/nativesdk- package contains symlink .so: <packagename> path '<path>' [dev-so]
Symlink .so
files are for development
only, and should therefore go into the
-dev
package.
This situation might occur if you add
*.so*
rather than
*.so.*
to a non-dev package.
Change
FILES
(and possibly
PACKAGES
)
such that the specified .so
file goes
into an appropriate -dev
package.
<packagename>: found library in wrong location [libdir]
The specified file may have been installed into an incorrect
(possibly hardcoded) installation path.
For example, this test will catch recipes that install
/lib/bar.so
when
${base_libdir}
is "lib32".
Another example is when recipes install
/usr/lib64/foo.so
when
${libdir}
is "/usr/lib".
False positives occasionally exist.
For these cases add "libdir" to
INSANE_SKIP
for the package.
non debug package contains .debug directory: <packagename> path <path> [debug-files]
The specified package contains a
.debug
directory, which should not
appear in anything but the -dbg
package.
This situation might occur if you add a path which contains
a .debug
directory and do not
explicitly add the .debug
directory
to the -dbg
package.
If this is the case, add the .debug
directory explicitly to
FILES_${PN}-dbg
.
See
FILES
for additional information on FILES
.
Architecture did not match (<machine_arch> to <file_arch>) on <file> [arch]
By default, the OpenEmbedded build system checks the
Executable and Linkable Format (ELF) type, bit size, and
endianness of any binaries to ensure they match the
target architecture.
This test fails if any binaries do not match the type since
there would be an incompatibility.
The test could indicate that the wrong compiler or compiler
options have been used.
Sometimes software, like bootloaders, might need to
bypass this check.
If the file you receive the error for is firmware
that is not intended to be executed within the target
operating system or is intended to run on a separate
processor within the device, you can add "arch" to
INSANE_SKIP
for the package.
Another option is to check the
do_compile
log and verify that the compiler options being used
are correct.
Bit size did not match (<machine_bits> to <file_bits>) <recipe> on <file> [arch]
By default, the OpenEmbedded build system checks
the Executable and Linkable Format (ELF) type,
bit size, and endianness of any binaries to ensure
they match the target architecture.
This test fails if any binaries do not match the type since
there would be an incompatibility.
The test could indicate that the wrong compiler or compiler
options have been used.
Sometimes software, like bootloaders, might need to
bypass this check.
If the file you receive the error for is firmware that
is not intended to be executed within the target
operating system or is intended to run on a separate
processor within the device, you can add "arch" to
INSANE_SKIP
for the package.
Another option is to check the
do_compile
log and verify that the compiler options being used are
correct.
Endianness did not match (<machine_endianness> to <file_endianness>) on <file> [arch]
By default, the OpenEmbedded build system checks
the Executable and Linkable Format (ELF) type, bit
size, and endianness of any binaries to ensure they
match the target architecture.
This test fails if any binaries do not match the type since
there would be an incompatibility.
The test could indicate that the wrong compiler or compiler
options have been used.
Sometimes software, like bootloaders, might need to
bypass this check.
If the file you receive the error for is firmware
that is not intended to be executed within the target
operating system or is intended to run on a separate
processor within the device, you can add "arch" to
INSANE_SKIP
for the package.
Another option is to check the
do_compile
log and verify that the compiler options being used
are correct.
ELF binary '<file>' has relocations in .text [textrel]
The specified ELF binary contains relocations in its
.text
sections.
This situation can result in a performance impact
at runtime.
Typically, the way to solve this performance issue is to
add "-fPIC" or "-fpic" to the compiler command-line
options.
For example, given software that reads
CFLAGS
when you build it, you could add the following to your
recipe:
CFLAGS_append = " -fPIC "
For more information on text relocations at runtime, see http://www.akkadia.org/drepper/textrelocs.html.
No GNU_HASH in the elf binary: '<file>' [ldflags]
This indicates that binaries produced when building the
recipe have not been linked with the
LDFLAGS
options provided by the build system.
Check to be sure that the LDFLAGS
variable is being passed to the linker command.
A common workaround for this situation is to pass in
LDFLAGS
using
TARGET_CC_ARCH
within the recipe as follows:
TARGET_CC_ARCH += "${LDFLAGS}"
Package <packagename> contains Xorg driver (<driver>) but no xorg-abi- dependencies [xorg-driver-abi]
The specified package contains an Xorg driver, but does not
have a corresponding ABI package dependency.
The xserver-xorg recipe provides driver ABI names.
All drivers should depend on the ABI versions that they have
been built against.
Driver recipes that include
xorg-driver-input.inc
or
xorg-driver-video.inc
will
automatically get these versions.
Consequently, you should only need to explicitly add
dependencies to binary driver recipes.
The /usr/share/info/dir file is not meant to be shipped in a particular package. [infodir]
The /usr/share/info/dir
should not be
packaged.
Add the following line to your
do_install
task or to your do_install_append
within the recipe as follows:
rm ${D}${infodir}/dir
Symlink <path> in <packagename> points to TMPDIR [symlink-to-sysroot]
The specified symlink points into
TMPDIR
on the host.
Such symlinks will work on the host.
However, they are clearly invalid when running on
the target.
You should either correct the symlink to use a relative
path or remove the symlink.
<file> failed sanity test (workdir) in path <path> [la]
The specified .la
file contains
TMPDIR
paths.
Any .la
file containing these paths
is incorrect since libtool
adds the
correct sysroot prefix when using the files automatically
itself.
<packagename> rdepends on <debug_packagename> [debug-deps]
A dependency exists between the specified non-dbg package
(i.e. a package whose name does not end in
-dbg
) and a package that is a
dbg
package.
The dbg
packages contain
debug symbols and are brought in using several
different methods:
Using the dbg-pkgs
IMAGE_FEATURES
value.
Using
IMAGE_INSTALL
.
As a dependency of another
dbg
package that was brought
in using one of the above methods.
The dependency might have been automatically added
because the dbg
package erroneously
contains files that it should not contain (e.g. a
non-symlink .so
file) or it might
have been added manually (e.g. by adding to
RDEPENDS
).
<packagename> rdepends on <dev_packagename> [dev-deps]
A dependency exists between the specified non-dev package
(a package whose name does not end in
-dev
) and a package that is a
dev
package.
The dev
packages contain development
headers and are usually brought in using several different
methods:
Using the dev-pkgs
IMAGE_FEATURES
value.
Using
IMAGE_INSTALL
.
As a dependency of another
dev
package that was brought
in using one of the above methods.
The dependency might have been automatically added (because
the dev
package erroneously contains
files that it should not have (e.g. a non-symlink
.so
file) or it might have been added
manually (e.g. by adding to
RDEPENDS
).
<var>_<packagename> is invalid: <comparison> (<value>) only comparisons <, =, >, <=, and >= are allowed [dep-cmp]
If you are adding a versioned dependency relationship to one
of the dependency variables
(RDEPENDS
,
RRECOMMENDS
,
RSUGGESTS
,
RPROVIDES
,
RREPLACES
,
or
RCONFLICTS
),
you must only use the named comparison operators.
Change the versioned dependency values you are adding
to match those listed in the message.
<recipename>: The compile log indicates that host include and/or library paths were used. Please check the log '<logfile>' for more information. [compile-host-path]
The log for the
do_compile
task indicates that paths on the host were searched
for files, which is not appropriate when cross-compiling.
Look for "is unsafe for cross-compilation" or "CROSS COMPILE
Badness" in the specified log file.
<recipename>: The install log indicates that host include and/or library paths were used. Please check the log '<logfile>' for more information. [install-host-path]
The log for the
do_install
task indicates that paths on the host were searched
for files, which is not appropriate when cross-compiling.
Look for "is unsafe for cross-compilation"
or "CROSS COMPILE Badness" in the specified log file.
This autoconf log indicates errors, it looked at host include and/or library paths while determining system capabilities. Rerun configure task after fixing this. The path was '<path>'
The log for the
do_configure
task indicates that paths on the host were searched
for files, which is not appropriate when cross-compiling.
Look for "is unsafe for cross-compilation" or
"CROSS COMPILE Badness" in the specified log file.
<packagename> doesn't match the [a-z0-9.+-]+ regex [pkgname]
The convention within the OpenEmbedded build system
(sometimes enforced by the package manager itself) is to
require that package names are all lower case
and to allow a restricted set of characters.
If your recipe name does not match this, or you add
packages to
PACKAGES
that do not conform to the convention, then you
will receive this error.
Rename your recipe.
Or, if you have added a non-conforming package name to
PACKAGES
, change the package name
appropriately.
<recipe>: configure was passed unrecognized options: <options> [unknown-configure-option]
The configure script is reporting that the specified
options are unrecognized.
This situation could be because the options
were previously valid but have been removed from the
configure script.
Or, there was a mistake when the options were added
and there is another option that should be used instead.
If you are unsure, consult the upstream build
documentation, the
./configure --help
output,
and the upstream change log or release notes.
Once you have worked out what the appropriate
change is, you can update
EXTRA_OECONF
,
PACKAGECONFIG_CONFARGS
,
or the individual
PACKAGECONFIG
option values accordingly.
Recipe <recipefile> has PN of "<recipename>" which is in OVERRIDES, this can result in unexpected behavior. [pn-overrides]
The specified recipe has a name
(PN
)
value that appears in
OVERRIDES
.
If a recipe is named such that its PN
value matches something already in
OVERRIDES
(e.g. PN
happens to be the same as
MACHINE
or
DISTRO
),
it can have unexpected consequences.
For example, assignments such as
FILES_${PN} = "xyz"
effectively
turn into FILES = "xyz"
.
Rename your recipe (or if PN
is being
set explicitly, change the PN
value) so
that the conflict does not occur.
See
FILES
for additional information.
<recipefile>: Variable <variable> is set as not being package specific, please fix this. [pkgvarcheck]
Certain variables
(RDEPENDS
,
RRECOMMENDS
,
RSUGGESTS
,
RCONFLICTS
,
RPROVIDES
,
RREPLACES
,
FILES
,
pkg_preinst
,
pkg_postinst
,
pkg_prerm
,
pkg_postrm
, and
ALLOW_EMPTY
)
should always be set specific to a package (i.e. they
should be set with a package name override such as
RDEPENDS_${PN} = "value"
rather than
RDEPENDS = "value"
).
If you receive this error, correct any assignments to these
variables within your recipe.
File '<file>' from <recipename> was already stripped, this will prevent future debugging! [already-stripped]
Produced binaries have already been stripped prior to the
build system extracting debug symbols.
It is common for upstream software projects to default to
stripping debug symbols for output binaries.
In order for debugging to work on the target using
-dbg
packages, this stripping must be
disabled.
Depending on the build system used by the software being built, disabling this stripping could be as easy as specifying an additional configure option. If not, disabling stripping might involve patching the build scripts. In the latter case, look for references to "strip" or "STRIP", or the "-s" or "-S" command-line options being specified on the linker command line (possibly through the compiler command line if preceded with "-Wl,").
-dbg
package,
it will then strip the symbols from the binaries.
<packagename> is listed in PACKAGES multiple times, this leads to packaging errors. [packages-list]
Package names must appear only once in the
PACKAGES
variable.
You might receive this error if you are attempting to add a
package to PACKAGES
that is
already in the variable's value.
FILES variable for package <packagename> contains '//' which is invalid. Attempting to fix this but you should correct the metadata. [files-invalid]
The string "//" is invalid in a Unix path.
Correct all occurrences where this string appears in a
FILES
variable so that there is only a single "/".
<recipename>: Files/directories were installed but not shipped in any package [installed-vs-shipped]
Files have been installed within the
do_install
task but have not been included in any package by way of the
FILES
variable.
Files that do not appear in any package cannot be present in
an image later on in the build process.
You need to do one of the following:
Add the files to FILES
for the
package you want them to appear in (e.g.
FILES_${
PN
}
for the main
package).
Delete the files at the end of the
do_install
task if the files
are not needed in any package.
<oldpackage>-<oldpkgversion> was registered as shlib provider for <library>, changing it to <newpackage>-<newpkgversion> because it was built later
This message means that both
<oldpackage>
and
<newpackage>
provide the specified
shared library.
You can expect this message when a recipe has been renamed.
However, if that is not the case, the message might indicate
that a private version of a library is being erroneously
picked up as the provider for a common library.
If that is the case, you should add the library's
.so
file name to
PRIVATE_LIBS
in the recipe that provides
the private version of the library.
You can configure the QA checks globally so that specific check
failures either raise a warning or an error message, using the
WARN_QA
and
ERROR_QA
variables, respectively.
You can also disable checks within a particular recipe using
INSANE_SKIP
.
For information on how to work with the QA checks, see the
"insane.bbclass
"
section.
The OpenEmbedded build system provides several example
images to satisfy different needs.
When you issue the bitbake
command you provide a “top-level” recipe
that essentially begins the build for the type of image you want.
local.conf
file before using the BitBake
command to build the minimal or base image:
1. Comment out the EXTRA_IMAGE_FEATURES line 2. Set INCOMPATIBLE_LICENSE = "GPL-3.0 LGPL-3.0 AGPL-3.0"
From within the poky
Git repository, you can use
the following command to display the list of directories within the
Source Directory
that contain image recipe files:
$ ls meta*/recipes*/images/*.bb
Following is a list of supported recipes:
build-appliance-image
:
An example virtual machine that contains all the pieces
required to run builds using the build system as well as the
build system itself.
You can boot and run the image using either the
VMware Player
or
VMware Workstation.
For more information on this image, see the
Build Appliance
page on the Yocto Project website.
core-image-base
:
A console-only image that fully supports the target device hardware.
core-image-clutter
:
An image with support for the Open GL-based toolkit Clutter, which enables development of
rich and animated graphical user interfaces.
core-image-full-cmdline
:
A console-only image with more full-featured Linux system
functionality installed.
core-image-lsb
:
An image that conforms to the Linux Standard Base (LSB)
specification.
This image requires a distribution configuration that
enables LSB compliance (e.g. poky-lsb
).
If you build core-image-lsb
without that
configuration, the image will not be LSB-compliant.
core-image-lsb-dev
:
A core-image-lsb
image that is suitable for development work
using the host.
The image includes headers and libraries you can use in a host development
environment.
This image requires a distribution configuration that
enables LSB compliance (e.g. poky-lsb
).
If you build core-image-lsb-dev
without that
configuration, the image will not be LSB-compliant.
core-image-lsb-sdk
:
A core-image-lsb
that includes everything in
the cross-toolchain but also includes development headers and libraries
to form a complete standalone SDK.
This image requires a distribution configuration that
enables LSB compliance (e.g. poky-lsb
).
If you build core-image-lsb-sdk
without that
configuration, the image will not be LSB-compliant.
This image is suitable for development using the target.
core-image-minimal
:
A small image just capable of allowing a device to boot.
core-image-minimal-dev
:
A core-image-minimal
image suitable for development work
using the host.
The image includes headers and libraries you can use in a host development
environment.
core-image-minimal-initramfs
:
A core-image-minimal
image that has the Minimal RAM-based
Initial Root Filesystem (initramfs) as part of the kernel,
which allows the system to find the first “init” program more efficiently.
See the
PACKAGE_INSTALL
variable for additional information helpful when working with
initramfs images.
core-image-minimal-mtdutils
:
A core-image-minimal
image that has support
for the Minimal MTD Utilities, which let the user interact with the
MTD subsystem in the kernel to perform operations on flash devices.
core-image-rt
:
A core-image-minimal
image plus a real-time test suite and
tools appropriate for real-time use.
core-image-rt-sdk
:
A core-image-rt
image that includes everything in
the cross-toolchain.
The image also includes development headers and libraries to form a complete
stand-alone SDK and is suitable for development using the target.
core-image-sato
:
An image with Sato support, a mobile environment and visual style that works well
with mobile devices.
The image supports X11 with a Sato theme and applications such as
a terminal, editor, file manager, media player, and so forth.
core-image-sato-dev
:
A core-image-sato
image suitable for development
using the host.
The image includes libraries needed to build applications on the device itself,
testing and profiling tools, and debug symbols.
This image was formerly core-image-sdk
.
core-image-sato-sdk
:
A core-image-sato
image that includes everything in
the cross-toolchain.
The image also includes development headers and libraries to form a complete standalone SDK
and is suitable for development using the target.
core-image-testmaster
:
A "master" image designed to be used for automated runtime testing.
Provides a "known good" image that is deployed to a separate
partition so that you can boot into it and use it to deploy a
second image to be tested.
You can find more information about runtime testing in the
"Performing Automated Runtime Testing"
section in the Yocto Project Development Tasks Manual.
core-image-testmaster-initramfs
:
A RAM-based Initial Root Filesystem (initramfs) image tailored for
use with the core-image-testmaster
image.
core-image-weston
:
A very basic Wayland image with a terminal.
This image provides the Wayland protocol libraries and the
reference Weston compositor.
For more information, see the
"Using Wayland and Weston"
section in the Yocto Project Development Tasks Manual.
core-image-x11
:
A very basic X11 image with a terminal.
This chapter provides a reference of shipped machine and distro features you can include as part of your image, a reference on image features you can select, and a reference on feature backfilling.
Features provide a mechanism for working out which packages
should be included in the generated images.
Distributions can select which features they want to support through the
DISTRO_FEATURES
variable, which is set or appended to in a distribution's configuration file such as
poky.conf
,
poky-tiny.conf
,
poky-lsb.conf
and so forth.
Machine features are set in the
MACHINE_FEATURES
variable, which is set in the machine configuration file and
specifies the hardware features for a given machine.
These two variables combine to work out which kernel modules, utilities, and other packages to include. A given distribution can support a selected subset of features so some machine features might not be included if the distribution itself does not support them.
One method you can use to determine which recipes are checking to see if a
particular feature is contained or not is to grep
through
the Metadata
for the feature.
Here is an example that discovers the recipes whose build is potentially
changed based on a given feature:
$ cd poky
$ git grep 'contains.*MACHINE_FEATURES.*feature
'
The items below are features you can use with
MACHINE_FEATURES
.
Features do not have a one-to-one correspondence to packages, and they can
go beyond simply controlling the installation of a package or packages.
Sometimes a feature can influence how certain recipes are built.
For example, a feature might determine whether a particular configure option
is specified within the
do_configure
task for a particular recipe.
This feature list only represents features as shipped with the Yocto Project metadata:
acpi: Hardware has ACPI (x86/x86_64 only)
alsa: Hardware has ALSA audio drivers
apm: Hardware uses APM (or APM emulation)
bluetooth: Hardware has integrated BT
efi: Support for booting through EFI
ext2: Hardware HDD or Microdrive
keyboard: Hardware has a keyboard
pcbios: Support for booting through BIOS
pci: Hardware has a PCI bus
pcmcia: Hardware has PCMCIA or CompactFlash sockets
phone: Mobile phone (voice) support
qvga: Machine has a QVGA (320x240) display
rtc: Machine has a Real-Time Clock
screen: Hardware has a screen
serial: Hardware has serial support (usually RS232)
touchscreen: Hardware has a touchscreen
usbgadget: Hardware is USB gadget device capable
usbhost: Hardware is USB Host capable
vfat: FAT file system support
wifi: Hardware has integrated WiFi
The items below are features you can use with
DISTRO_FEATURES
to enable features across your distribution.
Features do not have a one-to-one correspondence to packages,
and they can go beyond simply controlling the installation of a
package or packages.
In most cases, the presence or absence of a feature translates to
the appropriate option supplied to the configure script during the
do_configure
task for the recipes that optionally
support the feature.
Some distro features are also machine features.
These select features make sense to be controlled both at
the machine and distribution configuration level.
See the
COMBINED_FEATURES
variable for more information.
This list only represents features as shipped with the Yocto Project metadata:
alsa: Include ALSA support (OSS compatibility kernel modules installed if available).
api-documentation:
Enables generation of API documentation during recipe
builds.
The resulting documentation is added to SDK tarballs
when the
bitbake -c populate_sdk
command
is used.
See the
"Adding API Documentation to the Standard SDK"
section in the Yocto Project Application Development and
the Extensible Software Development Kit (eSDK) manual.
bluetooth: Include bluetooth support (integrated BT only).
cramfs: Include CramFS support.
directfb: Include DirectFB support.
ext2: Include tools for supporting for devices with internal HDD/Microdrive for storing files (instead of Flash only devices).
ipsec: Include IPSec support.
ipv6: Include IPv6 support.
keyboard: Include keyboard support (e.g. keymaps will be loaded during boot).
ldconfig:
Include support for ldconfig and
ld.so.conf
on the target.
nfs: Include NFS client support (for mounting NFS exports on device).
opengl: Include the Open Graphics Library, which is a cross-language, multi-platform application programming interface used for rendering two and three-dimensional graphics.
pci: Include PCI bus support.
pcmcia: Include PCMCIA/CompactFlash support.
ppp: Include PPP dialup support.
ptest: Enables building the package tests where supported by individual recipes. For more information on package tests, see the "Testing Packages With ptest" section in the Yocto Project Development Tasks Manual.
smbfs: Include SMB networks client support (for mounting Samba/Microsoft Windows shares on device).
systemd: Include support
for this init
manager, which is a full
replacement of for init
with parallel
starting of services, reduced shell overhead, and other
features.
This init
manager is used by many
distributions.
usbgadget: Include USB Gadget Device support (for USB networking/serial/storage).
usbhost: Include USB Host support (allows to connect external keyboard, mouse, storage, network etc).
usrmerge: Merges the
/bin
, /sbin
,
/lib
, and /lib64
directories into their respective counterparts in the
/usr
directory to provide better package
and application compatibility.
wayland: Include the Wayland display server protocol and the library that supports it.
wifi: Include WiFi support (integrated only).
x11: Include the X server and libraries.
The contents of images generated by the OpenEmbedded build system
can be controlled by the
IMAGE_FEATURES
and
EXTRA_IMAGE_FEATURES
variables that you typically configure in your image recipes.
Through these variables, you can add several different
predefined packages such as development utilities or packages with
debug information needed to investigate application problems or
profile applications.
The following image features are available for all images:
allow-empty-password: Allows Dropbear and OpenSSH to accept root logins and logins from accounts having an empty password string.
dbg-pkgs: Installs debug symbol packages for all packages installed in a given image.
debug-tweaks: Makes an image suitable for development (e.g. allows root logins without passwords and enables post-installation logging). See the 'allow-empty-password', 'empty-root-password', and 'post-install-logging' features in this list for additional information.
dev-pkgs: Installs development packages (headers and extra library links) for all packages installed in a given image.
doc-pkgs: Installs documentation packages for all packages installed in a given image.
empty-root-password: Sets the root password to an empty string, which allows logins with a blank password.
package-management: Installs package management tools and preserves the package manager database.
post-install-logging:
Enables logging postinstall script runs to
the /var/log/postinstall.log
file
on first boot of the image on the target system.
/var/log
directory
on the target persistent, use the
VOLATILE_LOG_DIR
variable by setting it to "no".
ptest-pkgs: Installs ptest packages for all ptest-enabled recipes.
read-only-rootfs: Creates an image whose root filesystem is read-only. See the "Creating a Read-Only Root Filesystem" section in the Yocto Project Development Tasks Manual for more information.
splash:
Enables showing a splash screen during boot.
By default, this screen is provided by
psplash
, which does allow
customization.
If you prefer to use an alternative splash screen package,
you can do so by setting the SPLASH
variable to a different package name (or names) within the
image recipe or at the distro configuration level.
staticdev-pkgs:
Installs static development packages, which are
static libraries (i.e. *.a
files), for
all packages installed in a given image.
Some image features are available only when you inherit the
core-image
class.
The current list of these valid features is as follows:
hwcodecs: Installs hardware acceleration codecs.
nfs-server: Installs an NFS server.
perf:
Installs profiling tools such as
perf
, systemtap
,
and LTTng
.
For general information on user-space tools, see the
Yocto Project Application Development and the Extensible Software Development Kit (eSDK)
manual.
ssh-server-dropbear: Installs the Dropbear minimal SSH server.
ssh-server-openssh:
Installs the OpenSSH SSH server, which is more
full-featured than Dropbear.
Note that if both the OpenSSH SSH server and the Dropbear
minimal SSH server are present in
IMAGE_FEATURES
, then OpenSSH will take
precedence and Dropbear will not be installed.
tools-debug:
Installs debugging tools such as
strace
and gdb
.
For information on GDB, see the
"Debugging With the GNU Project Debugger (GDB) Remotely"
section in the Yocto Project Development Tasks Manual.
For information on tracing and profiling, see the
Yocto Project Profiling and Tracing Manual.
tools-sdk: Installs a full SDK that runs on the device.
tools-testapps: Installs device testing tools (e.g. touchscreen debugging).
x11: Installs the X server.
x11-base: Installs the X server with a minimal environment.
x11-sato: Installs the OpenedHand Sato environment.
Sometimes it is necessary in the OpenEmbedded build system to extend
MACHINE_FEATURES
or DISTRO_FEATURES
to control functionality that was previously enabled and not able
to be disabled.
For these cases, we need to add an
additional feature item to appear in one of these variables,
but we do not want to force developers who have existing values
of the variables in their configuration to add the new feature
in order to retain the same overall level of functionality.
Thus, the OpenEmbedded build system has a mechanism to
automatically "backfill" these added features into existing
distro or machine configurations.
You can see the list of features for which this is done by
finding the
DISTRO_FEATURES_BACKFILL
and MACHINE_FEATURES_BACKFILL
variables in the meta/conf/bitbake.conf
file.
Because such features are backfilled by default into all
configurations as described in the previous paragraph, developers
who wish to disable the new features need to be able to selectively
prevent the backfilling from occurring.
They can do this by adding the undesired feature or features to the
DISTRO_FEATURES_BACKFILL_CONSIDERED
or MACHINE_FEATURES_BACKFILL_CONSIDERED
variables for distro features and machine features respectively.
Here are two examples to help illustrate feature backfilling:
The "pulseaudio" distro feature option:
Previously, PulseAudio support was enabled within the Qt and
GStreamer frameworks.
Because of this, the feature is backfilled and thus
enabled for all distros through the
DISTRO_FEATURES_BACKFILL
variable in the meta/conf/bitbake.conf
file.
However, your distro needs to disable the feature.
You can disable the feature without affecting
other existing distro configurations that need PulseAudio support
by adding "pulseaudio" to
DISTRO_FEATURES_BACKFILL_CONSIDERED
in your distro's .conf
file.
Adding the feature to this variable when it also
exists in the DISTRO_FEATURES_BACKFILL
variable prevents the build system from adding the feature to
your configuration's DISTRO_FEATURES
, effectively disabling
the feature for that particular distro.
The "rtc" machine feature option:
Previously, real time clock (RTC) support was enabled for all
target devices.
Because of this, the feature is backfilled and thus enabled
for all machines through the MACHINE_FEATURES_BACKFILL
variable in the meta/conf/bitbake.conf
file.
However, your target device does not have this capability.
You can disable RTC support for your device without
affecting other machines that need RTC support
by adding the feature to your machine's
MACHINE_FEATURES_BACKFILL_CONSIDERED
list in the machine's .conf
file.
Adding the feature to this variable when it also
exists in the MACHINE_FEATURES_BACKFILL
variable prevents the build system from adding the feature to
your configuration's MACHINE_FEATURES
, effectively
disabling RTC support for that particular machine.
This chapter lists common variables used in the OpenEmbedded build system and gives an overview of their function and contents.
A B C D E F G H I K L M N O P R S T U V W X
Extension to the Application Binary Interface (ABI) field of the GNU canonical architecture name (e.g. "eabi").
ABI extensions are set in the machine include files.
For example, the
meta/conf/machine/include/arm/arch-arm.inc
file sets the following extension:
ABIEXTENSION = "eabi"
Specifies whether to produce an output package even if it is
empty.
By default, BitBake does not produce empty packages.
This default behavior can cause issues when there is an
RDEPENDS
or
some other hard runtime requirement on the existence of the package.
Like all package-controlling variables, you must always use them in conjunction with a package name override, as in:
ALLOW_EMPTY_${PN} = "1" ALLOW_EMPTY_${PN}-dev = "1" ALLOW_EMPTY_${PN}-staticdev = "1"
Lists commands in a package that need an alternative binary naming scheme. Sometimes the same command is provided in multiple packages. When this occurs, the OpenEmbedded build system needs to use the alternatives system to create a different binary naming scheme so the commands can co-exist.
To use the variable, list out the package's commands
that also exist as part of another package.
For example, if the busybox
package
has four commands that also exist as part of another
package, you identify them as follows:
ALTERNATIVE_busybox = "sh sed test bracket"
For more information on the alternatives system, see the
"update-alternatives.bbclass
"
section.
Used by the alternatives system to map duplicated commands
to actual locations.
For example, if the bracket
command
provided by the busybox
package is
duplicated through another package, you must use the
ALTERNATIVE_LINK_NAME
variable to
specify the actual location:
ALTERNATIVE_LINK_NAME[bracket] = "/usr/bin/["
In this example, the binary for the
bracket
command (i.e.
[
) from the
busybox
package resides in
/usr/bin/
.
ALTERNATIVE_LINK_NAME
is not
defined, it defaults to
${bindir}/name
.
For more information on the alternatives system, see the
"update-alternatives.bbclass
"
section.
Used by the alternatives system to create default priorities for duplicated commands. You can use the variable to create a single default regardless of the command name or package, a default for specific duplicated commands regardless of the package, or a default for specific commands tied to particular packages. Here are the available syntax forms:
ALTERNATIVE_PRIORITY = "priority
" ALTERNATIVE_PRIORITY[name
] = "priority
" ALTERNATIVE_PRIORITY_pkg
[name
] = "priority
"
For more information on the alternatives system, see the
"update-alternatives.bbclass
"
section.
Used by the alternatives system to create default link locations for duplicated commands. You can use the variable to create a single default location for all duplicated commands regardless of the command name or package, a default for specific duplicated commands regardless of the package, or a default for specific commands tied to particular packages. Here are the available syntax forms:
ALTERNATIVE_TARGET = "target
" ALTERNATIVE_TARGET[name
] = "target
" ALTERNATIVE_TARGET_pkg
[name
] = "target
"
If ALTERNATIVE_TARGET
is not
defined, it inherits the value from the
ALTERNATIVE_LINK_NAME
variable.
If ALTERNATIVE_LINK_NAME
and
ALTERNATIVE_TARGET
are the
same, the target for
ALTERNATIVE_TARGET
has ".{BPN}
" appended to it.
Finally, if the file referenced has not been
renamed, the alternatives system will rename it to
avoid the need to rename alternative files in the
do_install
task while
retaining support for the command if necessary.
For more information on the alternatives system, see the
"update-alternatives.bbclass
"
section.
An override list of append strings for each target
specified with
LABELS
.
See the
grub-efi
class for more information on how this variable is used.
The minimal command and arguments used to run
ar
.
When used with the
archiver
class, determines the type of information used to create
a released archive.
You can use this variable to create archives of patched
source, original source, configured source, and so forth
by employing the following variable flags (varflags):
ARCHIVER_MODE[src] = "original" # Uses original (unpacked) source # files. ARCHIVER_MODE[src] = "patched" # Uses patched source files. This is # the default. ARCHIVER_MODE[src] = "configured" # Uses configured source files. ARCHIVER_MODE[diff] = "1" # Uses patches between do_unpack and # do_patch. ARCHIVER_MODE[diff-exclude] ?= "file
file
..." # Lists files and directories to # exclude from diff. ARCHIVER_MODE[dumpdata] = "1" # Uses environment data. ARCHIVER_MODE[recipe] = "1" # Uses recipe and include files. ARCHIVER_MODE[srpm] = "1" # Uses RPM package files.
For information on how the variable works, see the
meta/classes/archiver.bbclass
file
in the
Source Directory.
Minimal command and arguments needed to run the assembler.
Lists recipe names
(PN
values) BitBake does not attempt to build.
Instead, BitBake assumes these recipes have already been
built.
In OpenEmbedded-Core, ASSUME_PROVIDED
mostly specifies native tools that should not be built.
An example is git-native
, which when
specified, allows for the Git binary from the host to be
used rather than building git-native
.
Provides additional shlibs
provider
mapping information, which adds to or overwrites the
information provided automatically by the system.
Separate multiple entries using spaces.
As an example, use the following form to add an
shlib
provider of
shlibname
in
packagename
with the optional
version
:
shlibname:packagename
[_version
]
Here is an example that adds a shared library named
libEGL.so.1
as being provided by
the libegl-implementation
package:
ASSUME_SHLIBS = "libEGL.so.1:libegl-implementation"
The email address used to contact the original author or authors in order to send patches and forward bugs.
When the
debian
class is inherited, which is the default behavior,
AUTO_LIBNAME_PKGS
specifies which
packages should be checked for libraries and renamed
according to Debian library package naming.
The default value is "${PACKAGES}", which causes the debian class to act on all packages that are explicitly generated by the recipe.
Enables creating an automatic menu for the syslinux
bootloader.
You must set this variable in your recipe.
The
syslinux
class checks this variable.
When
SRCREV
is set to the value of this variable, it specifies to use
the latest source revision in the repository.
Here is an example:
SRCREV = "${AUTOREV}"
If you use the previous statement to retrieve the latest
version of software, you need to be sure
PV
contains
${
SRCPV
}
.
For example, suppose you have a kernel recipe that
inherits the
kernel class
and you use the previous statement.
In this example, ${SRCPV}
does not
automatically get into PV
.
Consequently, you need to change PV
in your recipe so that it does contain
${SRCPV}
.
For more information see the "Automatically Incrementing a Binary Package Revision Number" section in the Yocto Project Development Tasks Manual.
List of licenses found in the directories specified
by COMMON_LICENSE_DIR
and LICENSE_PATH
.
COMMON_LICENSE_DIR
and LICENSE_PATH
have been done
before AVAILABLE_LICENSES
is
defined
(in license.bbclass).
The list of defined CPU and Application Binary Interface (ABI) tunings (i.e. "tunes") available for use by the OpenEmbedded build system.
The list simply presents the tunes that are available. Not all tunes may be compatible with a particular machine configuration, or with each other in a Multilib configuration.
To add a tune to the list, be sure to append it with spaces using the "+=" BitBake operator. Do not simply replace the list by using the "=" operator. See the "Basic Syntax" section in the BitBake User Manual for more information.
The directory within the
Build Directory
in which the OpenEmbedded build system places generated
objects during a recipe's build process.
By default, this directory is the same as the S
directory, which is defined as:
S = "${WORKDIR}/${BP}"
You can separate the (S
) directory
and the directory pointed to by the B
variable.
Most Autotools-based recipes support separating these
directories.
The build system defaults to using separate directories for
gcc
and some kernel recipes.
Lists "recommended-only" packages to not install.
Recommended-only packages are packages installed only
through the
RRECOMMENDS
variable.
You can prevent any of these "recommended" packages from
being installed by listing them with the
BAD_RECOMMENDATIONS
variable:
BAD_RECOMMENDATIONS = "package_name
package_name
package_name
..."
You can set this variable globally in your
local.conf
file or you can attach it to
a specific image recipe by using the recipe name override:
BAD_RECOMMENDATIONS_pn-target_image
= "package_name
"
It is important to realize that if you choose to not install
packages using this variable and some other packages are
dependent on them (i.e. listed in a recipe's
RDEPENDS
variable), the OpenEmbedded build system ignores your
request and will install the packages to avoid dependency
errors.
Support for this variable exists only when using the IPK and RPM packaging backend. Support does not exist for DEB.
See the
NO_RECOMMENDATIONS
and the
PACKAGE_EXCLUDE
variables for related information.
The library directory name for the CPU or Application
Binary Interface (ABI) tune.
The BASE_LIB
applies only in the
Multilib context.
See the
"Combining Multiple Versions of Library Files into One Image"
section in the Yocto Project Development Tasks Manual for
information on Multilib.
The BASE_LIB
variable is defined in
the machine include files in the
Source Directory.
If Multilib is not being used, the value defaults to "lib".
Points to the base of the work directory for all recipes. The default value is "${TMPDIR}/work".
Specifies a space-delimited list of hosts that the fetcher is allowed to use to obtain the required source code. Following are considerations surrounding this variable:
This host list is only used if
BB_NO_NETWORK
is either not
set or set to "0".
Limited support for wildcard matching against the
beginning of host names exists.
For example, the following setting matches
git.gnu.org
,
ftp.gnu.org
, and
foo.git.gnu.org
.
BB_ALLOWED_NETWORKS = "*.gnu.org"
The use of the "*
"
character only works at the beginning of
a host name and it must be isolated from
the remainder of the host name.
You cannot use the wildcard character in any
other location of the name or combined with
the front part of the name.
For example,
*.foo.bar
is supported,
while *aa.foo.bar
is not.
Mirrors not in the host list are skipped and logged in debug.
Attempts to access networks not in the host list cause a failure.
Using BB_ALLOWED_NETWORKS
in
conjunction with
PREMIRRORS
is very useful.
Adding the host you want to use to
PREMIRRORS
results in the source code
being fetched from an allowed location and avoids raising
an error when a host that is not allowed is in a
SRC_URI
statement.
This is because the fetcher does not attempt to use the
host listed in SRC_URI
after a
successful fetch from the
PREMIRRORS
occurs.
Defines how BitBake handles situations where an append
file (.bbappend
) has no
corresponding recipe file (.bb
).
This condition often occurs when layers get out of sync
(e.g. oe-core
bumps a
recipe version and the old recipe no longer exists and the
other layer has not been updated to the new version
of the recipe yet).
The default fatal behavior is safest because it is the sane reaction given something is out of sync. It is important to realize when your changes are no longer being applied.
You can change the default behavior by setting this
variable to "1", "yes", or "true"
in your local.conf
file, which is
located in the
Build Directory:
Here is an example:
BB_DANGLINGAPPENDS_WARNONLY = "1"
Monitors disk space and available inodes during the build and allows you to control the build based on these parameters.
Disk space monitoring is disabled by default.
To enable monitoring, add the BB_DISKMON_DIRS
variable to your conf/local.conf
file found in the
Build Directory.
Use the following form:
BB_DISKMON_DIRS = "action
,dir
,threshold
[...]" where:action
is: ABORT: Immediately abort the build when a threshold is broken. STOPTASKS: Stop the build after the currently executing tasks have finished when a threshold is broken. WARN: Issue a warning but continue the build when a threshold is broken. Subsequent warnings are issued as defined by the BB_DISKMON_WARNINTERVAL variable, which must be defined in the conf/local.conf file.dir
is: Any directory you choose. You can specify one or more directories to monitor by separating the groupings with a space. If two directories are on the same device, only the first directory is monitored.threshold
is: Either the minimum available disk space, the minimum number of free inodes, or both. You must specify at least one. To omit one or the other, simply omit the value. Specify the threshold using G, M, K for Gbytes, Mbytes, and Kbytes, respectively. If you do not specify G, M, or K, Kbytes is assumed by default. Do not use GB, MB, or KB.
Here are some examples:
BB_DISKMON_DIRS = "ABORT,${TMPDIR},1G,100K WARN,${SSTATE_DIR},1G,100K" BB_DISKMON_DIRS = "STOPTASKS,${TMPDIR},1G" BB_DISKMON_DIRS = "ABORT,${TMPDIR},,100K"
The first example works only if you also provide
the BB_DISKMON_WARNINTERVAL
variable
in the conf/local.conf
.
This example causes the build system to immediately
abort when either the disk space in ${TMPDIR}
drops
below 1 Gbyte or the available free inodes drops below
100 Kbytes.
Because two directories are provided with the variable, the
build system also issue a
warning when the disk space in the
${SSTATE_DIR}
directory drops
below 1 Gbyte or the number of free inodes drops
below 100 Kbytes.
Subsequent warnings are issued during intervals as
defined by the BB_DISKMON_WARNINTERVAL
variable.
The second example stops the build after all currently
executing tasks complete when the minimum disk space
in the ${TMPDIR}
directory drops below 1 Gbyte.
No disk monitoring occurs for the free inodes in this case.
The final example immediately aborts the build when the
number of free inodes in the ${TMPDIR}
directory
drops below 100 Kbytes.
No disk space monitoring for the directory itself occurs
in this case.
Defines the disk space and free inode warning intervals.
To set these intervals, define the variable in your
conf/local.conf
file in the
Build Directory.
If you are going to use the
BB_DISKMON_WARNINTERVAL
variable, you must
also use the
BB_DISKMON_DIRS
variable
and define its action as "WARN".
During the build, subsequent warnings are issued each time
disk space or number of free inodes further reduces by
the respective interval.
If you do not provide a BB_DISKMON_WARNINTERVAL
variable and you do use BB_DISKMON_DIRS
with
the "WARN" action, the disk monitoring interval defaults to
the following:
BB_DISKMON_WARNINTERVAL = "50M,5K"
When specifying the variable in your configuration file, use the following form:
BB_DISKMON_WARNINTERVAL = "disk_space_interval
,disk_inode_interval
" where:disk_space_interval
is: An interval of memory expressed in either G, M, or K for Gbytes, Mbytes, or Kbytes, respectively. You cannot use GB, MB, or KB.disk_inode_interval
is: An interval of free inodes expressed in either G, M, or K for Gbytes, Mbytes, or Kbytes, respectively. You cannot use GB, MB, or KB.
Here is an example:
BB_DISKMON_DIRS = "WARN,${SSTATE_DIR},1G,100K" BB_DISKMON_WARNINTERVAL = "50M,5K"
These variables cause the OpenEmbedded build system to
issue subsequent warnings each time the available
disk space further reduces by 50 Mbytes or the number
of free inodes further reduces by 5 Kbytes in the
${SSTATE_DIR}
directory.
Subsequent warnings based on the interval occur each time
a respective interval is reached beyond the initial warning
(i.e. 1 Gbytes and 100 Kbytes).
Causes tarballs of the source control repositories
(e.g. Git repositories), including metadata, to be placed
in the
DL_DIR
directory.
For performance reasons, creating and placing tarballs of these repositories is not the default action by the OpenEmbedded build system.
BB_GENERATE_MIRROR_TARBALLS = "1"
Set this variable in your local.conf
file in the
Build Directory.
Once you have the tarballs containing your source files,
you can clean up your DL_DIR
directory by deleting any Git or other source control
work directories.
The maximum number of tasks BitBake should run in parallel
at any one time.
The OpenEmbedded build system automatically configures
this variable to be equal to the number of cores on the
build system.
For example, a system with a dual core processor that
also uses hyper-threading causes the
BB_NUMBER_THREADS
variable to default
to "4".
For single socket systems (i.e. one CPU), you should not
have to override this variable to gain optimal parallelism
during builds.
However, if you have very large systems that employ
multiple physical CPUs, you might want to make sure the
BB_NUMBER_THREADS
variable is not
set higher than "20".
For more information on speeding up builds, see the "Speeding Up a Build" section in the Yocto Project Development Tasks Manual.
Specifies the time (in seconds) after which to unload the
BitBake server due to inactivity.
Set BB_SERVER_TIMEOUT
to determine how
long the BitBake server stays resident between invocations.
For example, the following statement in your
local.conf
file instructs the server
to be unloaded after 20 seconds of inactivity:
BB_SERVER_TIMEOUT = "20"
If you want the server to never be unloaded, set
BB_SERVER_TIMEOUT
to "-1".
Allows you to extend a recipe so that it builds variants of the software.
Common variants for recipes exist such as "natives" like quilt-native
,
which is a copy of Quilt built to run on the build system;
"crosses" such as gcc-cross
,
which is a compiler built to run on the build machine but produces binaries
that run on the target MACHINE
;
"nativesdk", which targets the SDK machine instead of MACHINE
;
and "mulitlibs" in the form "multilib:
multilib_name
".
To build a different variant of the recipe with a minimal amount of code, it usually is as simple as adding the following to your recipe:
BBCLASSEXTEND =+ "native nativesdk"
BBCLASSEXTEND =+ "multilib:multilib_name
"
Internally, the BBCLASSEXTEND
mechanism generates recipe variants by rewriting
variable values and applying overrides such as
_class-native
.
For example, to generate a native version of a recipe,
a
DEPENDS
on "foo" is rewritten to a DEPENDS
on "foo-native".
Even when using BBCLASSEXTEND
, the
recipe is only parsed once.
Parsing once adds some limitations.
For example, it is not possible to
include a different file depending on the variant,
since include
statements are
processed when the recipe is parsed.
Lists the names of configured layers.
These names are used to find the other BBFILE_*
variables.
Typically, each layer will append its name to this variable in its
conf/layer.conf
file.
Variable that expands to match files from
BBFILES
in a particular layer.
This variable is used in the conf/layer.conf
file and must
be suffixed with the name of the specific layer (e.g.
BBFILE_PATTERN_emenlow
).
Assigns the priority for recipe files in each layer.
This variable is useful in situations where the same recipe appears in
more than one layer.
Setting this variable allows you to prioritize a
layer against other layers that contain the same recipe - effectively
letting you control the precedence for the multiple layers.
The precedence established through this variable stands regardless of a
recipe's version
(PV
variable).
For example, a layer that has a recipe with a higher PV
value but for
which the BBFILE_PRIORITY
is set to have a lower precedence still has a
lower precedence.
A larger value for the BBFILE_PRIORITY
variable results in a higher
precedence.
For example, the value 6 has a higher precedence than the value 5.
If not specified, the BBFILE_PRIORITY
variable is set based on layer
dependencies (see the
LAYERDEPENDS
variable for
more information.
The default priority, if unspecified
for a layer with no dependencies, is the lowest defined priority + 1
(or 1 if no priorities are defined).
bitbake-layers show-layers
to list
all configured layers along with their priorities.
A space-separated list of recipe files BitBake uses to build software.
When specifying recipe files, you can pattern match using
Python's
glob
syntax.
For details on the syntax, see the documentation by
following the previous link.
Activates content when identified layers are present. You identify the layers by the collections that the layers define.
Use the BBFILES_DYNAMIC
variable to
avoid .bbappend
files whose
corresponding .bb
file is in a layer
that attempts to modify other layers through
.bbappend
but does not want to
introduce a hard dependency on those other layers.
Use the following form for
BBFILES_DYNAMIC
:
collection_name
:filename_pattern
The following example identifies two collection names and two filename patterns:
BBFILES_DYNAMIC += " \ clang-layer:${LAYERDIR}/bbappends/meta-clang/*/*/*.bbappend \ core:${LAYERDIR}/bbappends/openembedded-core/meta/*/*/*.bbappend \ "
This next example shows an error message that occurs because invalid entries are found, which cause parsing to abort:
ERROR: BBFILES_DYNAMIC entries must be of the form <collection name>:<filename pattern>, not: /work/my-layer/bbappends/meta-security-isafw/*/*/*.bbappend /work/my-layer/bbappends/openembedded-core/meta/*/*/*.bbappend
Variable that controls how BitBake displays logs on build failure.
If
BBINCLUDELOGS
is set, specifies the maximum number of lines from the
task log file to print when reporting a failed task.
If you do not set BBINCLUDELOGS_LINES
,
the entire log is printed.
Lists the layers to enable during the build.
This variable is defined in the bblayers.conf
configuration
file in the
Build Directory.
Here is an example:
BBLAYERS = " \ /home/scottrif/poky/meta \ /home/scottrif/poky/meta-poky \ /home/scottrif/poky/meta-yocto-bsp \ /home/scottrif/poky/meta-mykernel \ "
This example enables four layers, one of which is a custom, user-defined layer
named meta-mykernel
.
Prevents BitBake from processing recipes and recipe append files.
You can use the BBMASK
variable
to "hide" these .bb
and
.bbappend
files.
BitBake ignores any recipe or recipe append files that
match any of the expressions.
It is as if BitBake does not see them at all.
Consequently, matching files are not parsed or otherwise
used by BitBake.
The values you provide are passed to Python's regular expression compiler. Consequently, the syntax follows Python's Regular Expression (re) syntax. The expressions are compared against the full paths to the files. For complete syntax information, see Python's documentation at http://docs.python.org/3/library/re.html#re.
The following example uses a complete regular expression
to tell BitBake to ignore all recipe and recipe append
files in the meta-ti/recipes-misc/
directory:
BBMASK = "meta-ti/recipes-misc/"
If you want to mask out multiple directories or recipes, you can specify multiple regular expression fragments. This next example masks out multiple directories and individual recipes:
BBMASK += "/meta-ti/recipes-misc/ meta-ti/recipes-ti/packagegroup/" BBMASK += "/meta-oe/recipes-support/" BBMASK += "/meta-foo/.*/openldap" BBMASK += "opencv.*\.bbappend" BBMASK += "lzma"
Specifies each additional separate configuration when you
are building targets with multiple configurations.
Use this variable in your
conf/local.conf
configuration file.
Specify a multiconfigname
for
each configuration file you are using.
For example, the following line specifies three
configuration files:
BBMULTICONFIG = "configA configB configC"
Each configuration file you use must reside in the
Build Directory
conf/multiconfig
directory
(e.g.
build_directory
/conf/multiconfig/configA.conf
).
For information on how to use
BBMULTICONFIG
in an environment that
supports building targets with multiple configurations,
see the
"Building Images for Multiple Targets Using Multiple Configurations"
section in the Yocto Project Development Tasks Manual.
Used by BitBake to locate
.bbclass
and configuration files.
This variable is analogous to the
PATH
variable.
BBPATH
to point to the
Build Directory.
Set the variable as you would any environment variable
and then run BitBake:
$ BBPATH = "build_directory
" $ export BBPATH $ bitbaketarget
If defined in the BitBake environment,
BBSERVER
points to the BitBake
remote server.
Use the following format to export the variable to the BitBake environment:
export BBSERVER=localhost:$port
By default, BBSERVER
also appears in
BB_HASHBASE_WHITELIST
.
Consequently, BBSERVER
is excluded
from checksum and dependency data.
When inheriting the
binconfig-disabled
class, this variable specifies binary configuration
scripts to disable in favor of using
pkg-config
to query the information.
The binconfig-disabled
class will
modify the specified scripts to return an error so that
calls to them can be easily found and replaced.
To add multiple scripts, separate them by spaces.
Here is an example from the libpng
recipe:
BINCONFIG = "${bindir}/libpng-config ${bindir}/libpng16-config"
When inheriting the
binconfig
class, this variable specifies a wildcard for
configuration scripts that need editing.
The scripts are edited to correct any paths that have been
set up during compilation so that they are correct for
use when installed into the sysroot and called by the
build processes of other recipes.
BINCONFIG_GLOB
variable
uses
shell globbing,
which is recognition and expansion of wildcards during
pattern matching.
Shell globbing is very similar to
fnmatch
and
glob
.
For more information on how this variable works, see
meta/classes/binconfig.bbclass
in the
Source Directory.
You can also find general information on the class in the
"binconfig.bbclass
"
section.
The base recipe name and version but without any special
recipe name suffix (i.e. -native
, lib64-
,
and so forth).
BP
is comprised of the following:
${BPN}-${PV}
This variable is a version of the
PN
variable with common prefixes and suffixes
removed, such as nativesdk-
,
-cross
,
-native
, and multilib's
lib64-
and
lib32-
.
The exact lists of prefixes and suffixes removed are
specified by the
MLPREFIX
and
SPECIAL_PKGSUFFIX
variables, respectively.
Specifies a URL for an upstream bug tracking website for a recipe. The OpenEmbedded build system does not use this variable. Rather, the variable is a useful pointer in case a bug in the software being built needs to be manually reported.
Specifies the architecture of the build host
(e.g. i686
).
The OpenEmbedded build system sets the value of
BUILD_ARCH
from the machine name
reported by the uname
command.
Specifies the architecture-specific assembler flags for
the build host. By default, the value of
BUILD_AS_ARCH
is empty.
Specifies the architecture-specific C compiler flags for
the build host. By default, the value of
BUILD_CC_ARCH
is empty.
Specifies the linker command to be used for the build host
when the C compiler is being used as the linker. By default,
BUILD_CCLD
points to GCC and passes as
arguments the value of
BUILD_CC_ARCH
,
assuming BUILD_CC_ARCH
is set.
Specifies the flags to pass to the C compiler when building
for the build host.
When building in the -native
context,
CFLAGS
is set to the value of this variable by default.
Specifies the flags to pass to the C preprocessor
(i.e. to both the C and the C++ compilers) when building
for the build host.
When building in the -native
context,
CPPFLAGS
is set to the value of this variable by default.
Specifies the flags to pass to the C++ compiler when
building for the build host.
When building in the -native
context,
CXXFLAGS
is set to the value of this variable by default.
Specifies the Fortran compiler command for the build host.
By default, BUILD_FC
points to
Gfortran and passes as arguments the value of
BUILD_CC_ARCH
,
assuming BUILD_CC_ARCH
is set.
Specifies the linker command for the build host. By default,
BUILD_LD
points to the GNU linker (ld)
and passes as arguments the value of
BUILD_LD_ARCH
,
assuming BUILD_LD_ARCH
is set.
Specifies architecture-specific linker flags for the build
host. By default, the value of
BUILD_LD_ARCH
is empty.
Specifies the flags to pass to the linker when building
for the build host.
When building in the -native
context,
LDFLAGS
is set to the value of this variable by default.
Specifies the optimization flags passed to the C compiler
when building for the build host or the SDK.
The flags are passed through the
BUILD_CFLAGS
and
BUILDSDK_CFLAGS
default values.
The default value of the
BUILD_OPTIMIZATION
variable is
"-O2 -pipe".
Specifies the operating system in use on the build
host (e.g. "linux").
The OpenEmbedded build system sets the value of
BUILD_OS
from the OS reported by
the uname
command - the first word,
converted to lower-case characters.
The toolchain binary prefix used for native recipes.
The OpenEmbedded build system uses the
BUILD_PREFIX
value to set the
TARGET_PREFIX
when building for native
recipes.
Specifies the command to be used to strip debugging symbols
from binaries produced for the build host. By default,
BUILD_STRIP
points to
${
BUILD_PREFIX
}strip
.
Specifies the system, including the architecture and
the operating system, to use when building for the build
host (i.e. when building native
recipes).
The OpenEmbedded build system automatically sets this
variable based on
BUILD_ARCH
,
BUILD_VENDOR
,
and
BUILD_OS
.
You do not need to set the BUILD_SYS
variable yourself.
Specifies the vendor name to use when building for the build host. The default value is an empty string ("").
Points to the location of the
Build Directory.
You can define this directory indirectly through the
oe-init-build-env
script by passing in a Build Directory path when you run
the script.
If you run the script and do not provide a Build Directory
path, the BUILDDIR
defaults to
build
in the current directory.
When inheriting the
buildhistory
class, this variable specifies whether or not to commit the
build history output in a local Git repository.
If set to "1", this local repository will be maintained
automatically by the
buildhistory
class and a commit will be created on every
build for changes to each top-level subdirectory of the
build history output (images, packages, and sdk).
If you want to track changes to build history over
time, you should set this value to "1".
By default, the buildhistory
class
does not commit the build history output in a local
Git repository:
BUILDHISTORY_COMMIT ?= "0"
When inheriting the
buildhistory
class, this variable specifies the author to use for each
Git commit.
In order for the BUILDHISTORY_COMMIT_AUTHOR
variable to work, the
BUILDHISTORY_COMMIT
variable must be set to "1".
Git requires that the value you provide for the
BUILDHISTORY_COMMIT_AUTHOR
variable
takes the form of "name email@host
".
Providing an email address or host that is not valid does
not produce an error.
By default, the buildhistory
class
sets the variable as follows:
BUILDHISTORY_COMMIT_AUTHOR ?= "buildhistory <buildhistory@${DISTRO}>"
When inheriting the
buildhistory
class, this variable specifies the directory in which
build history information is kept.
For more information on how the variable works, see the
buildhistory.class
.
By default, the buildhistory
class
sets the directory as follows:
BUILDHISTORY_DIR ?= "${TOPDIR}/buildhistory"
When inheriting the
buildhistory
class, this variable specifies the build history features
to be enabled.
For more information on how build history works, see the
"Maintaining Build Output Quality"
section in the Yocto Project Development Tasks Manual.
You can specify these features in the form of a space-separated list:
image: Analysis of the contents of images, which includes the list of installed packages among other things.
package: Analysis of the contents of individual packages.
sdk: Analysis of the contents of the software development kit (SDK).
task: Save output file signatures for shared state (sstate) tasks. This saves one file per task and lists the SHA-256 checksums for each file staged (i.e. the output of the task).
By default, the buildhistory
class
enables the following features:
BUILDHISTORY_FEATURES ?= "image package sdk"
When inheriting the
buildhistory
class, this variable specifies a list of paths to files
copied from the
image contents into the build history directory under
an "image-files" directory in the directory for
the image, so that you can track the contents of each file.
The default is to copy /etc/passwd
and /etc/group
, which allows you to
monitor for changes in user and group entries.
You can modify the list to include any file.
Specifying an invalid path does not produce an error.
Consequently, you can include files that might
not always be present.
By default, the buildhistory
class
provides paths to the following files:
BUILDHISTORY_IMAGE_FILES ?= "/etc/passwd /etc/group"
When inheriting the
buildhistory
class, this variable optionally specifies a remote
repository to which build history pushes Git changes.
In order for BUILDHISTORY_PUSH_REPO
to work,
BUILDHISTORY_COMMIT
must be set to "1".
The repository should correspond to a remote
address that specifies a repository as understood by
Git, or alternatively to a remote name that you have
set up manually using git remote
within the local repository.
By default, the buildhistory
class
sets the variable as follows:
BUILDHISTORY_PUSH_REPO ?= ""
Specifies the flags to pass to the C compiler when building
for the SDK.
When building in the nativesdk-
context,
CFLAGS
is set to the value of this variable by default.
Specifies the flags to pass to the C pre-processor
(i.e. to both the C and the C++ compilers) when building
for the SDK.
When building in the nativesdk-
context,
CPPFLAGS
is set to the value of this variable by default.
Specifies the flags to pass to the C++ compiler when
building for the SDK.
When building in the nativesdk-
context,
CXXFLAGS
is set to the value of this variable by default.
Specifies the flags to pass to the linker when building
for the SDK.
When building in the nativesdk-
context,
LDFLAGS
is set to the value of this variable by default.
Points to the location of the directory that holds build
statistics when you use and enable the
buildstats
class.
The BUILDSTATS_BASE
directory defaults
to
${
TMPDIR
}/buildstats/
.
For the BusyBox recipe, specifies whether to split the
output executable file into two parts: one for features
that require setuid root
, and one for
the remaining features (i.e. those that do not require
setuid root
).
The BUSYBOX_SPLIT_SUID
variable
defaults to "1", which results in splitting the output
executable file.
Set the variable to "0" to get a single output executable
file.
Specifies the directory BitBake uses to store a cache of the Metadata so it does not need to be parsed every time BitBake is started.
The minimal command and arguments used to run the C compiler.
Specifies the flags to pass to the C compiler. This variable is exported to an environment variable and thus made visible to the software being built during the compilation step.
Default initialization for CFLAGS
varies depending on what is being built:
TARGET_CFLAGS
when building for the target
BUILD_CFLAGS
when building for the build host (i.e.
-native
)
BUILDSDK_CFLAGS
when building for an SDK (i.e.
nativesdk-
)
An internal variable specifying the special class override
that should currently apply (e.g. "class-target",
"class-native", and so forth).
The classes that use this variable (e.g.
native
,
nativesdk
,
and so forth) set the variable to appropriate values.
CLASSOVERRIDE
gets its default
"class-target" value from the
bitbake.conf
file.
As an example, the following override allows you to install extra files, but only when building for the target:
do_install_append_class-target() { install my-extra-file ${D}${sysconfdir} }
Here is an example where FOO
is set to "native" when building for the build host, and
to "other" when not building for the build host:
FOO_class-native = "native" FOO = "other"
The underlying mechanism behind
CLASSOVERRIDE
is simply that it is
included in the default value of
OVERRIDES
.
If set to "1" within a recipe,
CLEANBROKEN
specifies that
the make clean
command does
not work for the software being built.
Consequently, the OpenEmbedded build system will not try
to run make clean
during the
do_configure
task, which is the default behavior.
Provides a list of hardware features that are enabled in
both
MACHINE_FEATURES
and
DISTRO_FEATURES
.
This select list of features contains features that make
sense to be controlled both at the machine and distribution
configuration level.
For example, the "bluetooth" feature requires hardware
support but should also be optional at the distribution
level, in case the hardware supports Bluetooth but you
do not ever intend to use it.
Points to meta/files/common-licenses
in the
Source Directory,
which is where generic license files reside.
A regular expression that resolves to one or more hosts
(when the recipe is native) or one or more targets (when
the recipe is non-native) with which a recipe is compatible.
The regular expression is matched against
HOST_SYS
.
You can use the variable to stop recipes from being built
for classes of systems with which the recipes are not
compatible.
Stopping these builds is particularly useful with kernels.
The variable also helps to increase parsing speed
since the build system skips parsing recipes not
compatible with the current system.
A regular expression that resolves to one or more
target machines with which a recipe is compatible.
The regular expression is matched against
MACHINEOVERRIDES
.
You can use the variable to stop recipes from being built
for machines with which the recipes are not compatible.
Stopping these builds is particularly useful with kernels.
The variable also helps to increase parsing speed
since the build system skips parsing recipes not
compatible with the current machine.
Defines wildcards to match when installing a list of complementary packages for all the packages explicitly (or implicitly) installed in an image.
COMPLEMENTARY_GLOB
variable
uses Unix filename pattern matching
(fnmatch
),
which is similar to the Unix style pathname pattern
expansion
(glob
).
The resulting list of complementary packages is associated
with an item that can be added to
IMAGE_FEATURES
.
An example usage of this is the "dev-pkgs" item that when
added to IMAGE_FEATURES
will
install -dev packages (containing headers and other
development files) for every package in the image.
To add a new feature item pointing to a wildcard, use a variable flag to specify the feature item name and use the value to specify the wildcard. Here is an example:
COMPLEMENTARY_GLOB[dev-pkgs] = '*-dev'
Stores sysroot components for each recipe.
The OpenEmbedded build system uses
COMPONENTS_DIR
when constructing
recipe-specific sysroots for other recipes.
The default is
"${
STAGING_DIR
}-components
."
(i.e. "${
TMPDIR
}/sysroots-components
").
Tracks the version of the local configuration file
(i.e. local.conf
).
The value for CONF_VERSION
increments each time build/conf/
compatibility changes.
Identifies editable or configurable files that are part of a package.
If the Package Management System (PMS) is being used to update
packages on the target system, it is possible that
configuration files you have changed after the original installation
and that you now want to remain unchanged are overwritten.
In other words, editable files might exist in the package that you do not
want reset as part of the package update process.
You can use the CONFFILES
variable to list the files in the
package that you wish to prevent the PMS from overwriting during this update process.
To use the CONFFILES
variable, provide a package name
override that identifies the resulting package.
Then, provide a space-separated list of files.
Here is an example:
CONFFILES_${PN} += "${sysconfdir}/file1 \ ${sysconfdir}/file2 ${sysconfdir}/file3"
A relationship exists between the CONFFILES
and
FILES
variables.
The files listed within CONFFILES
must be a subset of
the files listed within FILES
.
Because the configuration files you provide with CONFFILES
are simply being identified so that the PMS will not overwrite them,
it makes sense that
the files must already be included as part of the package through the
FILES
variable.
CONFFILES
variable,
it is good practice to use appropriate path variables.
For example, ${sysconfdir}
rather than
/etc
or ${bindir}
rather
than /usr/bin
.
You can find a list of these variables at the top of the
meta/conf/bitbake.conf
file in the
Source Directory.
Identifies the initial RAM filesystem (initramfs) source files. The OpenEmbedded build system receives and uses this kernel Kconfig variable as an environment variable. By default, the variable is set to null ("").
The CONFIG_INITRAMFS_SOURCE
can be
either a single cpio archive with a
.cpio
suffix or a
space-separated list of directories and files for building
the initramfs image.
A cpio archive should contain a filesystem archive
to be used as an initramfs image.
Directories should contain a filesystem layout to be
included in the initramfs image.
Files should contain entries according to the format
described by the
usr/gen_init_cpio
program in the
kernel tree.
If you specify multiple directories and files, the initramfs image will be the aggregate of all of them.
For information on creating an initramfs, see the "Building an Initial RAM Filesystem (initramfs) Image" section in the Yocto Project Development Tasks Manual.
A list of files that contains autoconf
test results relevant
to the current build.
This variable is used by the Autotools utilities when running
configure
.
The minimal arguments for GNU configure.
When inheriting the
distro_features_check
class, this
variable identifies distribution features that would
be in conflict should the recipe
be built.
In other words, if the
CONFLICT_DISTRO_FEATURES
variable
lists a feature that also appears in
DISTRO_FEATURES
within the
current configuration, an error occurs and the
build stops.
A space-separated list of licenses to exclude from the
source archived by the
archiver
class.
In other words, if a license in a recipe's
LICENSE
value is in the value of
COPYLEFT_LICENSE_EXCLUDE
, then its
source is not archived by the class.
COPYLEFT_LICENSE_EXCLUDE
variable takes precedence over the
COPYLEFT_LICENSE_INCLUDE
variable.
The default value, which is "CLOSED Proprietary", for
COPYLEFT_LICENSE_EXCLUDE
is set
by the
copyleft_filter
class, which is inherited by the
archiver
class.
A space-separated list of licenses to include in the
source archived by the
archiver
class.
In other words, if a license in a recipe's
LICENSE
value is in the value of
COPYLEFT_LICENSE_INCLUDE
, then its
source is archived by the class.
The default value is set by the
copyleft_filter
class, which is inherited by the
archiver
class.
The default value includes "GPL*", "LGPL*", and "AGPL*".
A list of recipes to exclude in the source archived
by the
archiver
class.
The COPYLEFT_PN_EXCLUDE
variable
overrides the license inclusion and exclusion caused
through the
COPYLEFT_LICENSE_INCLUDE
and
COPYLEFT_LICENSE_EXCLUDE
variables, respectively.
The default value, which is "" indicating to not explicitly
exclude any recipes by name, for
COPYLEFT_PN_EXCLUDE
is set
by the
copyleft_filter
class, which is inherited by the
archiver
class.
A list of recipes to include in the source archived
by the
archiver
class.
The COPYLEFT_PN_INCLUDE
variable
overrides the license inclusion and exclusion caused
through the
COPYLEFT_LICENSE_INCLUDE
and
COPYLEFT_LICENSE_EXCLUDE
variables, respectively.
The default value, which is "" indicating to not explicitly
include any recipes by name, for
COPYLEFT_PN_INCLUDE
is set
by the
copyleft_filter
class, which is inherited by the
archiver
class.
A space-separated list of recipe types to include
in the source archived by the
archiver
class.
Recipe types are target
,
native
,
nativesdk
,
cross
,
crosssdk
, and
cross-canadian
.
The default value, which is "target*", for
COPYLEFT_RECIPE_TYPES
is set
by the
copyleft_filter
class, which is inherited by the
archiver
class.
If set to "1" along with the
COPY_LIC_MANIFEST
variable, the OpenEmbedded build system copies
into the image the license files, which are located in
/usr/share/common-licenses
,
for each package.
The license files are placed
in directories within the image itself during build time.
COPY_LIC_DIRS
does not
offer a path for adding licenses for newly installed
packages to an image, which might be most suitable
for read-only filesystems that cannot be upgraded.
See the
LICENSE_CREATE_PACKAGE
variable for additional information.
You can also reference the
"Providing License Text"
section in the Yocto Project Development Tasks Manual
for information on providing license text.
If set to "1", the OpenEmbedded build system copies
the license manifest for the image to
/usr/share/common-licenses/license.manifest
within the image itself during build time.
COPY_LIC_MANIFEST
does not
offer a path for adding licenses for newly installed
packages to an image, which might be most suitable
for read-only filesystems that cannot be upgraded.
See the
LICENSE_CREATE_PACKAGE
variable for additional information.
You can also reference the
"Providing License Text"
section in the Yocto Project Development Tasks Manual
for information on providing license text.
Specifies the list of packages to be added to the image.
You should only set this variable in the
local.conf
configuration file found
in the
Build Directory.
This variable replaces POKY_EXTRA_INSTALL
, which is no longer supported.
Specifies the parent directory of the OpenEmbedded-Core
Metadata layer (i.e. meta
).
It is an important distinction that
COREBASE
points to the parent of this
layer and not the layer itself.
Consider an example where you have cloned the Poky Git
repository and retained the poky
name for your local copy of the repository.
In this case, COREBASE
points to
the poky
folder because it is the
parent directory of the poky/meta
layer.
Lists files from the
COREBASE
directory that should be copied other than the layers
listed in the bblayers.conf
file.
The COREBASE_FILES
variable exists
for the purpose of copying metadata from the
OpenEmbedded build system into the extensible
SDK.
Explicitly listing files in COREBASE
is needed because it typically contains build
directories and other files that should not normally
be copied into the extensible SDK.
Consequently, the value of
COREBASE_FILES
is used in order to
only copy the files that are actually needed.
The minimal command and arguments used to run the C preprocessor.
Specifies the flags to pass to the C pre-processor (i.e. to both the C and the C++ compilers). This variable is exported to an environment variable and thus made visible to the software being built during the compilation step.
Default initialization for CPPFLAGS
varies depending on what is being built:
TARGET_CPPFLAGS
when building for the target
BUILD_CPPFLAGS
when building for the build host (i.e.
-native
)
BUILDSDK_CPPFLAGS
when building for an SDK (i.e.
nativesdk-
)
The toolchain binary prefix for the target tools.
The CROSS_COMPILE
variable is the
same as the
TARGET_PREFIX
variable.
CROSS_COMPILE
variable only in
certain contexts (e.g. when building for kernel
and kernel module recipes).
The directory in which files checked out under the CVS system are stored.
The minimal command and arguments used to run the C++ compiler.
Specifies the flags to pass to the C++ compiler. This variable is exported to an environment variable and thus made visible to the software being built during the compilation step.
Default initialization for CXXFLAGS
varies depending on what is being built:
TARGET_CXXFLAGS
when building for the target
BUILD_CXXFLAGS
when building for the build host (i.e.
-native
)
BUILDSDK_CXXFLAGS
when building for an SDK (i.e.
nativesdk-
)
The destination directory.
The location in the
Build Directory
where components are installed by the
do_install
task.
This location defaults to:
${WORKDIR}/image
The date the build was started. Dates appear using the year, month, and day (YMD) format (e.g. "20150209" for February 9th, 2015).
The date and time on which the current build started. The format is suitable for timestamps.
When the
debian
class is inherited, which is the default behavior,
DEBIAN_NOAUTONAME
specifies a
particular package should not be renamed according to
Debian library package naming.
You must use the package name as an override when you
set this variable.
Here is an example from the fontconfig
recipe:
DEBIAN_NOAUTONAME_fontconfig-utils = "1"
When the
debian
class is inherited, which is the default behavior,
DEBIANNAME
allows you to override the
library name for an individual package.
Overriding the library name in these cases is rare.
You must use the package name as an override when you
set this variable.
Here is an example from the dbus
recipe:
DEBIANNAME_${PN} = "dbus-1"
Specifies to build packages with debugging information.
This influences the value of the
SELECTED_OPTIMIZATION
variable.
The options to pass in
TARGET_CFLAGS
and CFLAGS
when compiling
a system for debugging.
This variable defaults to "-O -fno-omit-frame-pointer ${DEBUG_FLAGS} -pipe".
Specifies a weak bias for recipe selection priority.
The most common usage of this is variable is to set
it to "-1" within a recipe for a development version of a
piece of software.
Using the variable in this way causes the stable version
of the recipe to build by default in the absence of
PREFERRED_VERSION
being used to build the development version.
DEFAULT_PREFERENCE
is weak and is overridden by
BBFILE_PRIORITY
if that variable is different between two layers
that contain different versions of the same recipe.
The default CPU and Application Binary Interface (ABI)
tunings (i.e. the "tune") used by the OpenEmbedded build
system.
The DEFAULTTUNE
helps define
TUNE_FEATURES
.
The default tune is either implicitly or explicitly set
by the machine
(MACHINE
).
However, you can override the setting using available tunes
as defined with
AVAILTUNES
.
Lists a recipe's build-time dependencies. These are dependencies on other recipes whose contents (e.g. headers and shared libraries) are needed by the recipe at build time.
As an example, consider a recipe foo
that contains the following assignment:
DEPENDS = "bar"
The practical effect of the previous assignment is that
all files installed by bar will be available in the
appropriate staging sysroot, given by the
STAGING_DIR*
variables, by the time the
do_configure
task for foo
runs.
This mechanism is implemented by having
do_configure
depend on the
do_populate_sysroot
task of each recipe listed in DEPENDS
,
through a
[
deptask
]
declaration in the
base
class.
STAGING_DIR_HOST
explicitly.
The standard classes and build-related variables are
configured to automatically use the appropriate staging
sysroots.
As another example, DEPENDS
can also
be used to add utilities that run on the build machine
during the build.
For example, a recipe that makes use of a code generator
built by the recipe codegen
might have
the following:
DEPENDS = "codegen-native"
For more information, see the
native
class and the
EXTRANATIVEPATH
variable.
DEPENDS
is a list of
recipe names.
Or, to be more precise, it is a list of
PROVIDES
names, which usually match recipe names.
Putting a package name such as "foo-dev" in
DEPENDS
does not make
sense.
Use "foo" instead, as this will put files
from all the packages that make up
foo
, which includes
those from foo-dev
, into
the sysroot.
One recipe having another recipe in
DEPENDS
does not by itself
add any runtime dependencies between the
packages produced by the two recipes.
However, as explained in the
"Automatically Added Runtime Dependencies"
section in the Yocto Project Overview and
Concepts Manual, runtime dependencies will
often be added automatically, meaning
DEPENDS
alone is
sufficient for most recipes.
Counterintuitively,
DEPENDS
is often necessary
even for recipes that install precompiled
components.
For example, if libfoo
is a precompiled library that links against
libbar
, then
linking against libfoo
requires both libfoo
and libbar
to be available
in the sysroot.
Without a DEPENDS
from the
recipe that installs libfoo
to the recipe that installs
libbar
, other recipes might
fail to link against
libfoo
.
For information on runtime dependencies, see the
RDEPENDS
variable.
You can also see the
"Tasks" and
"Dependencies"
sections in the BitBake User Manual for additional
information on tasks and dependencies.
Points to the general area that the OpenEmbedded build
system uses to place images, packages, SDKs, and other output
files that are ready to be used outside of the build system.
By default, this directory resides within the
Build Directory
as ${TMPDIR}/deploy
.
For more information on the structure of the Build
Directory, see
"The Build Directory - build/
"
section.
For more detail on the contents of the
deploy
directory, see the
"Images",
"Package Feeds",
and
"Application Development SDK"
sections all in the Yocto Project Overview and Concepts
Manual.
Points to the area that the OpenEmbedded build system uses
to place Debian packages that are ready to be used outside
of the build system.
This variable applies only when
PACKAGE_CLASSES
contains "package_deb".
The BitBake configuration file initially defines the
DEPLOY_DIR_DEB
variable as a
sub-folder of
DEPLOY_DIR
:
DEPLOY_DIR_DEB = "${DEPLOY_DIR}/deb"
The
package_deb
class uses the
DEPLOY_DIR_DEB
variable to make sure
the
do_package_write_deb
task writes Debian packages into the appropriate folder.
For more information on how packaging works, see the
"Package Feeds"
section in the Yocto Project Overview and Concepts Manual.
Points to the area that the OpenEmbedded build system uses
to place images and other associated output files that are
ready to be deployed onto the target machine.
The directory is machine-specific as it contains the
${MACHINE}
name.
By default, this directory resides within the
Build Directory
as ${DEPLOY_DIR}/images/${MACHINE}/
.
For more information on the structure of the Build
Directory, see
"The Build Directory - build/
"
section.
For more detail on the contents of the
deploy
directory, see the
"Images"
and
"Application Development SDK"
sections both in the Yocto Project Overview and Concepts
Manual.
Points to the area that the OpenEmbedded build system uses
to place IPK packages that are ready to be used outside of
the build system.
This variable applies only when
PACKAGE_CLASSES
contains "package_ipk".
The BitBake configuration file initially defines this
variable as a sub-folder of
DEPLOY_DIR
:
DEPLOY_DIR_IPK = "${DEPLOY_DIR}/ipk"
The
package_ipk
class uses the
DEPLOY_DIR_IPK
variable to make sure
the
do_package_write_ipk
task writes IPK packages into the appropriate folder.
For more information on how packaging works, see the
"Package Feeds"
section in the Yocto Project Overview and Concepts Manual.
Points to the area that the OpenEmbedded build system uses
to place RPM packages that are ready to be used outside
of the build system.
This variable applies only when
PACKAGE_CLASSES
contains "package_rpm".
The BitBake configuration file initially defines this
variable as a sub-folder of
DEPLOY_DIR
:
DEPLOY_DIR_RPM = "${DEPLOY_DIR}/rpm"
The
package_rpm
class uses the
DEPLOY_DIR_RPM
variable to make sure
the
do_package_write_rpm
task writes RPM packages into the appropriate folder.
For more information on how packaging works, see the
"Package Feeds"
section in the Yocto Project Overview and Concepts Manual.
Points to the area that the OpenEmbedded build system uses
to place tarballs that are ready to be used outside of
the build system.
This variable applies only when
PACKAGE_CLASSES
contains "package_tar".
The BitBake configuration file initially defines this
variable as a sub-folder of
DEPLOY_DIR
:
DEPLOY_DIR_TAR = "${DEPLOY_DIR}/tar"
The
package_tar
class uses the
DEPLOY_DIR_TAR
variable to make sure
the
do_package_write_tar
task writes TAR packages into the appropriate folder.
For more information on how packaging works, see the
"Package Feeds"
section in the Yocto Project Overview and Concepts Manual.
When inheriting the
deploy
class, the DEPLOYDIR
points to a
temporary work area for deployed files that is set in the
deploy
class as follows:
DEPLOYDIR = "${WORKDIR}/deploy-${PN
}"
Recipes inheriting the deploy
class
should copy files to be deployed into
DEPLOYDIR
, and the class will take
care of copying them into
DEPLOY_DIR_IMAGE
afterwards.
The package description used by package managers.
If not set, DESCRIPTION
takes
the value of the
SUMMARY
variable.
The short name of the distribution.
For information on the long name of the distribution, see
the
DISTRO_NAME
variable.
The DISTRO
variable corresponds to a
distribution configuration file whose root name is the
same as the variable's argument and whose filename
extension is .conf
.
For example, the distribution configuration file for the
Poky distribution is named poky.conf
and resides in the
meta-poky/conf/distro
directory of
the
Source Directory.
Within that poky.conf
file, the
DISTRO
variable is set as follows:
DISTRO = "poky"
Distribution configuration files are located in a
conf/distro
directory within the
Metadata
that contains the distribution configuration.
The value for DISTRO
must not contain
spaces, and is typically all lower-case.
DISTRO
variable is blank,
a set of default configurations are used, which are
specified within
meta/conf/distro/defaultsetup.conf
also in the Source Directory.
Specifies a codename for the distribution being built.
Specifies a list of distro-specific packages to add to all images.
This variable takes affect through
packagegroup-base
so the
variable only really applies to the more full-featured
images that include packagegroup-base
.
You can use this variable to keep distro policy out of
generic images.
As with all other distro variables, you set this variable
in the distro .conf
file.
Specifies a list of distro-specific packages to add to all images if the packages exist. The packages might not exist or be empty (e.g. kernel modules). The list of packages are automatically installed but you can remove them.
The software support you want in your distribution for various features. You define your distribution features in the distribution configuration file.
In most cases, the presence or absence of a feature in
DISTRO_FEATURES
is translated to the
appropriate option supplied to the configure script
during the
do_configure
task for recipes that optionally support the feature.
For example, specifying "x11" in
DISTRO_FEATURES
, causes
every piece of software built for the target that can
optionally support X11 to have its X11 support enabled.
Two more examples are Bluetooth and NFS support. For a more complete list of features that ships with the Yocto Project and that you can provide with this variable, see the "Distro Features" section.
Features to be added to
DISTRO_FEATURES
if not also present in
DISTRO_FEATURES_BACKFILL_CONSIDERED
.
This variable is set in the meta/conf/bitbake.conf
file.
It is not intended to be user-configurable.
It is best to just reference the variable to see which distro features are
being backfilled for all distro configurations.
See the "Feature Backfilling" section for
more information.
Features from
DISTRO_FEATURES_BACKFILL
that should not be backfilled (i.e. added to
DISTRO_FEATURES
)
during the build.
See the "Feature Backfilling" section for
more information.
A convenience variable that gives you the default
list of distro features with the exception of any
features specific to the C library
(libc
).
When creating a custom distribution, you might find it
useful to be able to reuse the default
DISTRO_FEATURES
options without the need to write out the full set.
Here is an example that uses
DISTRO_FEATURES_DEFAULT
from a
custom distro configuration file:
DISTRO_FEATURES ?= "${DISTRO_FEATURES_DEFAULT} myfeature"
Specifies a list of features that if present in
the target
DISTRO_FEATURES
value should be included in
DISTRO_FEATURES
when building native
recipes.
This variable is used in addition to the features
filtered using the
DISTRO_FEATURES_NATIVE
variable.
Specifies a list of features that if present in the target
DISTRO_FEATURES
value should be included in
DISTRO_FEATURES
when building
nativesdk recipes.
This variable is used in addition to the features
filtered using the
DISTRO_FEATURES_NATIVESDK
variable.
Specifies a list of features that should be included in
DISTRO_FEATURES
when building native recipes.
This variable is used in addition to the features
filtered using the
DISTRO_FEATURES_FILTER_NATIVE
variable.
Specifies a list of features that should be included in
DISTRO_FEATURES
when building nativesdk recipes.
This variable is used in addition to the features
filtered using the
DISTRO_FEATURES_FILTER_NATIVESDK
variable.
The long name of the distribution.
For information on the short name of the distribution, see
the
DISTRO
variable.
The DISTRO_NAME
variable corresponds
to a distribution configuration file whose root name is the
same as the variable's argument and whose filename
extension is .conf
.
For example, the distribution configuration file for the
Poky distribution is named poky.conf
and resides in the
meta-poky/conf/distro
directory of
the
Source Directory.
Within that poky.conf
file, the
DISTRO_NAME
variable is set as
follows:
DISTRO_NAME = "Poky (Yocto Project Reference Distro)"
Distribution configuration files are located in a
conf/distro
directory within the
Metadata
that contains the distribution configuration.
DISTRO_NAME
variable is
blank, a set of default configurations are used, which
are specified within
meta/conf/distro/defaultsetup.conf
also in the Source Directory.
The version of the distribution.
A colon-separated list of overrides specific to the
current distribution.
By default, this list includes the value of
DISTRO
.
You can extend DISTROOVERRIDES
to add extra overrides that should apply to
the distribution.
The underlying mechanism behind
DISTROOVERRIDES
is simply that it
is included in the default value of
OVERRIDES
.
The central download directory used by the build process to
store downloads.
By default, DL_DIR
gets files
suitable for mirroring for everything except Git
repositories.
If you want tarballs of Git repositories, use the
BB_GENERATE_MIRROR_TARBALLS
variable.
You can set this directory by defining the
DL_DIR
variable in the
conf/local.conf
file.
This directory is self-maintaining and you should not have
to touch it.
By default, the directory is downloads
in the
Build Directory.
#DL_DIR ?= "${TOPDIR}/downloads"
To specify a different download directory, simply remove the comment from the line and provide your directory.
During a first build, the system downloads many different
source code tarballs from various upstream projects.
Downloading can take a while, particularly if your network
connection is slow.
Tarballs are all stored in the directory defined by
DL_DIR
and the build system looks there
first to find source tarballs.
You can safely share this directory between multiple builds on the same development machine. For additional information on how the build process gets source files when working behind a firewall or proxy server, see this specific question in the "FAQ" chapter. You can also refer to the "Working Behind a Network Proxy" Wiki page.
When inheriting the
compress_doc
class, this variable sets the compression policy used when
the OpenEmbedded build system compresses man pages and info
pages.
By default, the compression method used is gz (gzip).
Other policies available are xz and bz2.
For information on policies and on how to use this
variable, see the comments in the
meta/classes/compress_doc.bbclass
file.
When building bootable images (i.e. where
hddimg
, iso
,
or wic.vmdk
is in
IMAGE_FSTYPES
),
the EFI_PROVIDER
variable specifies
the EFI bootloader to use.
The default is "grub-efi", but "systemd-boot" can be used
instead.
See the
systemd-boot
and
image-live
classes for more information.
Variable that controls which locales for
glibc
are generated during the
build (useful if the target device has 64Mbytes
of RAM or less).
When used with the
report-error
class, specifies the path used for storing the debug files
created by the
error reporting tool,
which allows you to submit build errors you encounter to a
central database.
By default, the value of this variable is
${
LOG_DIR
}/error-report
.
You can set ERR_REPORT_DIR
to the path
you want the error reporting tool to store the debug files
as follows in your local.conf
file:
ERR_REPORT_DIR = "path
"
Specifies the quality assurance checks whose failures are
reported as errors by the OpenEmbedded build system.
You set this variable in your distribution configuration
file.
For a list of the checks you can control with this variable,
see the
"insane.bbclass
"
section.
Triggers the OpenEmbedded build system's shared libraries resolver to exclude an entire package when scanning for shared libraries.
package_do_shlibs
, which is part of
the
do_package
task.
You should be aware that the shared libraries resolver
might implicitly define some dependencies between
packages.
The EXCLUDE_FROM_SHLIBS
variable is
similar to the
PRIVATE_LIBS
variable, which excludes a package's particular libraries
only and not the whole package.
Use the
EXCLUDE_FROM_SHLIBS
variable by
setting it to "1" for a particular package:
EXCLUDE_FROM_SHLIBS = "1"
Directs BitBake to exclude a recipe from world builds (i.e.
bitbake world
).
During world builds, BitBake locates, parses and builds all
recipes found in every layer exposed in the
bblayers.conf
configuration file.
To exclude a recipe from a world build using this variable, set the variable to "1" in the recipe.
EXCLUDE_FROM_WORLD
may still be built during a world build in order to satisfy
dependencies of other recipes.
Adding a recipe to EXCLUDE_FROM_WORLD
only ensures that the recipe is not explicitly added
to the list of build targets in a world build.
Used with file and pathnames to create a prefix for a recipe's
version based on the recipe's
PE
value.
If PE
is set and greater than zero for a recipe,
EXTENDPE
becomes that value (e.g if
PE
is equal to "1" then EXTENDPE
becomes "1_").
If a recipe's PE
is not set (the default) or is equal to
zero, EXTENDPE
becomes "".
See the STAMP
variable for an example.
The full package version specification as it appears on the final packages produced by a recipe. The variable's value is normally used to fix a runtime dependency to the exact same version of another package in the same recipe:
RDEPENDS_${PN}-additional-module = "${PN} (= ${EXTENDPKGV})"
The dependency relationships are intended to force the package manager to upgrade these types of packages in lock-step.
When set, the EXTERNAL_KERNEL_TOOLS
variable indicates that these tools are not in the
source tree.
When kernel tools are available in the tree, they are
preferred over any externally installed tools.
Setting the EXTERNAL_KERNEL_TOOLS
variable tells the OpenEmbedded build system to prefer
the installed external tools.
See the
kernel-yocto
class in meta/classes
to see how
the variable is used.
When inheriting the
externalsrc
class, this variable points to the source tree, which is
outside of the OpenEmbedded build system.
When set, this variable sets the
S
variable, which is what the OpenEmbedded build system uses
to locate unpacked recipe source code.
For more information on
externalsrc.bbclass
, see the
"externalsrc.bbclass
"
section.
You can also find information on how to use this variable
in the
"Building Software from an External Source"
section in the Yocto Project Development Tasks Manual.
When inheriting the
externalsrc
class, this variable points to the directory in which the
recipe's source code is built, which is outside of the
OpenEmbedded build system.
When set, this variable sets the
B
variable, which is what the OpenEmbedded build system uses
to locate the Build Directory.
For more information on
externalsrc.bbclass
, see the
"externalsrc.bbclass
"
section.
You can also find information on how to use this variable
in the
"Building Software from an External Source"
section in the Yocto Project Development Tasks Manual.
For recipes inheriting the
autotools
class, you can use EXTRA_AUTORECONF
to
specify extra options to pass to the
autoreconf
command that is
executed during the
do_configure
task.
The default value is "--exclude=autopoint".
A list of additional features to include in an image. When listing more than one feature, separate them with a space.
Typically, you configure this variable in your
local.conf
file, which is found in the
Build Directory.
Although you can use this variable from within a recipe,
best practices dictate that you do not.
IMAGE_FEATURES
variable.
Here are some examples of features you can add:
"dbg-pkgs" - Adds -dbg packages for all installed packages including symbol information for debugging and profiling. "debug-tweaks" - Makes an image suitable for debugging. For example, allows root logins without passwords and enables post-installation logging. See the 'allow-empty-password' and 'post-install-logging' features in the "Image Features" section for more information. "dev-pkgs" - Adds -dev packages for all installed packages. This is useful if you want to develop against the libraries in the image. "read-only-rootfs" - Creates an image whose root filesystem is read-only. See the "Creating a Read-Only Root Filesystem" section in the Yocto Project Development Tasks Manual for more information "tools-debug" - Adds debugging tools such as gdb and strace. "tools-sdk" - Adds development tools such as gcc, make, pkgconfig and so forth. "tools-testapps" - Adds useful testing tools such as ts_print, aplay, arecord and so forth.
For a complete list of image features that ships with the Yocto Project, see the "Image Features" section.
For an example that shows how to customize your image by
using this variable, see the
"Customizing Images Using Custom IMAGE_FEATURES
and EXTRA_IMAGE_FEATURES
"
section in the Yocto Project Development Tasks Manual.
Specifies additional options for the image
creation command that has been specified in
IMAGE_CMD
.
When setting this variable, use an override for the
associated image type.
Here is an example:
EXTRA_IMAGECMD_ext3 ?= "-i 4096"
A list of recipes to build that do not provide packages for installing into the root filesystem.
Sometimes a recipe is required to build the final image but is not
needed in the root filesystem.
You can use the EXTRA_IMAGEDEPENDS
variable to
list these recipes and thus specify the dependencies.
A typical example is a required bootloader in a machine configuration.
A list of subdirectories of
${
STAGING_BINDIR_NATIVE
}
added to the beginning of the environment variable
PATH
.
As an example, the following prepends
"${STAGING_BINDIR_NATIVE}/foo:${STAGING_BINDIR_NATIVE}/bar:"
to PATH
:
EXTRANATIVEPATH = "foo bar"
Additional
CMake
options.
See the
cmake
class for additional information.
Additional configure
script options.
See
PACKAGECONFIG_CONFARGS
for additional information on passing configure script
options.
Additional GNU make
options.
Because the EXTRA_OEMAKE
defaults to
"", you need to set the variable to specify any required
GNU options.
PARALLEL_MAKE
and
PARALLEL_MAKEINST
also make use of
EXTRA_OEMAKE
to pass the required
flags.
When inheriting the
scons
class, this variable specifies additional configuration
options you want to pass to the
scons
command line.
When inheriting the
extrausers
class, this variable provides image level user and group
operations.
This is a more global method of providing user and group
configuration as compared to using the
useradd
class, which ties user and group configurations to a
specific recipe.
The set list of commands you can configure using the
EXTRA_USERS_PARAMS
is shown in the
extrausers
class.
These commands map to the normal Unix commands of the same
names:
# EXTRA_USERS_PARAMS = "\ # useradd -p '' tester; \ # groupadd developers; \ # userdel nobody; \ # groupdel -g video; \ # groupmod -g 1020 developers; \ # usermod -s /bin/sh tester; \ # "
Defines one or more packages to include in an image when
a specific item is included in
IMAGE_FEATURES
.
When setting the value, FEATURE_PACKAGES
should have the name of the feature item as an override.
Here is an example:
FEATURE_PACKAGES_widget = "package1
package2
"
In this example, if "widget" were added to
IMAGE_FEATURES
, package1
and
package2
would be included in the image.
FEATURE_PACKAGES
are often package
groups.
While similarly named, you should not confuse the
FEATURE_PACKAGES
variable with
package groups, which are discussed elsewhere in the
documentation.
Points to the base URL of the server and location within
the document-root that provides the metadata and
packages required by OPKG to support runtime package
management of IPK packages.
You set this variable in your
local.conf
file.
Consider the following example:
FEED_DEPLOYDIR_BASE_URI = "http://192.168.7.1/BOARD-dir"
This example assumes you are serving your packages over
HTTP and your databases are located in a directory
named BOARD-dir
, which is underneath
your HTTP server's document-root.
In this case, the OpenEmbedded build system generates a set
of configuration files for you in your target that work
with the feed.
The list of files and directories that are placed in a
package.
The
PACKAGES
variable lists the packages generated by a recipe.
To use the FILES
variable, provide a
package name override that identifies the resulting package.
Then, provide a space-separated list of files or paths
that identify the files you want included as part of the
resulting package.
Here is an example:
FILES_${PN} += "${bindir}/mydir1 ${bindir}/mydir2/myfile"
When specifying files or paths, you can pattern
match using Python's
glob
syntax.
For details on the syntax, see the
documentation by following the previous link.
When specifying paths as part of the
FILES
variable, it is
good practice to use appropriate path
variables.
For example, use ${sysconfdir}
rather than /etc
, or
${bindir}
rather than
/usr/bin
.
You can find a list of these variables at the
top of the
meta/conf/bitbake.conf
file in the
Source Directory.
You will also find the default values of the
various FILES_*
variables
in this file.
If some of the files you provide with the
FILES
variable are editable and you
know they should not be overwritten during the package
update process by the Package Management System (PMS), you
can identify these files so that the PMS will not
overwrite them.
See the
CONFFILES
variable for information on how to identify these files to
the PMS.
Defines the file specification to match
SOLIBSDEV
.
In other words, FILES_SOLIBSDEV
defines the full path name of the development symbolic link
(symlink) for shared libraries on the target platform.
The following statement from the
bitbake.conf
shows how it is set:
FILES_SOLIBSDEV ?= "${base_libdir}/lib*${SOLIBSDEV} ${libdir}/lib*${SOLIBSDEV}"
Extends the search path the OpenEmbedded build system uses
when looking for files and patches as it processes recipes
and append files.
The default directories BitBake uses when it processes
recipes are initially defined by the
FILESPATH
variable.
You can extend FILESPATH
variable
by using FILESEXTRAPATHS
.
Best practices dictate that you accomplish this by using
FILESEXTRAPATHS
from within a
.bbappend
file and that you prepend
paths as follows:
FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
In the above example, the build system first looks for files in a directory that has the same name as the corresponding append file.
When extending
FILESEXTRAPATHS
,
be sure to use the immediate expansion
(:=
) operator.
Immediate expansion makes sure that BitBake evaluates
THISDIR
at the time the directive is encountered rather than at
some later time when expansion might result in a
directory that does not contain the files you need.
Also, include the trailing separating colon character if you are prepending. The trailing colon character is necessary because you are directing BitBake to extend the path by prepending directories to the search path.
Here is another common use:
FILESEXTRAPATHS_prepend := "${THISDIR}/files:"
In this example, the build system extends the
FILESPATH
variable to include a
directory named files
that is in the
same directory as the corresponding append file.
This next example specifically adds three paths:
FILESEXTRAPATHS_prepend := "path_1:path_2:path_3:"
A final example shows how you can extend the search path
and include a
MACHINE
-specific
override, which is useful in a BSP layer:
FILESEXTRAPATHS_prepend_intel-x86-common := "${THISDIR}/${PN}:"
The previous statement appears in the
linux-yocto-dev.bbappend
file, which
is found in the Yocto Project
Source Repositories
in
meta-intel/common/recipes-kernel/linux
.
Here, the machine override is a special
PACKAGE_ARCH
definition for multiple meta-intel
machines.
MACHINE
.
By prepending paths in .bbappend
files, you allow multiple append files that reside in
different layers but are used for the same recipe to
correctly extend the path.
A subset of OVERRIDES
used by the OpenEmbedded build system for creating
FILESPATH
.
The FILESOVERRIDES
variable uses
overrides to automatically extend the
FILESPATH
variable.
For an example of how that works, see the
FILESPATH
variable description.
Additionally, you find more information on how overrides
are handled in the
"Conditional Syntax (Overrides)"
section of the BitBake User Manual.
By default, the FILESOVERRIDES
variable is defined as:
FILESOVERRIDES = "${TRANSLATED_TARGET_ARCH}:${MACHINEOVERRIDES}:${DISTROOVERRIDES}"
FILESOVERRIDES
variable.
The values match up with expected overrides and are
used in an expected manner by the build system.
The default set of directories the OpenEmbedded build system uses when searching for patches and files.
During the build process, BitBake searches each directory
in FILESPATH
in the specified order
when looking for files and patches specified by each
file://
URI in a recipe's
SRC_URI
statements.
The default value for the FILESPATH
variable is defined in the base.bbclass
class found in meta/classes
in the
Source Directory:
FILESPATH = "${@base_set_filespath(["${FILE_DIRNAME}/${BP}", \ "${FILE_DIRNAME}/${BPN}", "${FILE_DIRNAME}/files"], d)}"
The FILESPATH
variable is automatically
extended using the overrides from the
FILESOVERRIDES
variable.
Do not hand-edit the
FILESPATH
variable.
If you want the build system to look in
directories other than the defaults, extend the
FILESPATH
variable by
using the
FILESEXTRAPATHS
variable.
Be aware that the default
FILESPATH
directories do
not map to directories in custom layers
where append files
(.bbappend
) are used.
If you want the build system to find patches
or files that reside with your append files,
you need to extend the
FILESPATH
variable by
using the FILESEXTRAPATHS
variable.
You can take advantage of this searching behavior in useful ways. For example, consider a case where the following directory structure exists for general and machine-specific configurations:
files/defconfig files/MACHINEA/defconfig files/MACHINEB/defconfig
Also in the example, the SRC_URI
statement contains "file://defconfig".
Given this scenario, you can set
MACHINE
to "MACHINEA" and cause the build system to use files
from files/MACHINEA
.
Set MACHINE
to "MACHINEB" and the
build system uses files from
files/MACHINEB
.
Finally, for any machine other than "MACHINEA" and
"MACHINEB", the build system uses files from
files/defconfig
.
You can find out more about the patching process in the
"Patching"
section in the Yocto Project Overview and Concepts Manual
and the
"Patching Code"
section in the Yocto Project Development Tasks Manual.
See the
do_patch
task as well.
Allows you to define your own file permissions settings table as part of your configuration for the packaging process. For example, suppose you need a consistent set of custom permissions for a set of groups and users across an entire work project. It is best to do this in the packages themselves but this is not always possible.
By default, the OpenEmbedded build system uses the fs-perms.txt
, which
is located in the meta/files
folder in the
Source Directory.
If you create your own file permissions setting table, you should place it in your
layer or the distro's layer.
You define the FILESYSTEM_PERMS_TABLES
variable in the
conf/local.conf
file, which is found in the
Build Directory, to
point to your custom fs-perms.txt
.
You can specify more than a single file permissions setting table.
The paths you specify to these files must be defined within the
BBPATH
variable.
For guidance on how to create your own file permissions settings table file,
examine the existing fs-perms.txt
.
When inheriting the
fontcache
class, this variable specifies the runtime dependencies
for font packages.
By default, the FONT_EXTRA_RDEPENDS
is set to "fontconfig-utils".
When inheriting the
fontcache
class, this variable identifies packages containing font
files that need to be cached by Fontconfig.
By default, the fontcache
class assumes
that fonts are in the recipe's main package
(i.e. ${
PN
}
).
Use this variable if fonts you need are in a package
other than that main package.
Forces the removal of the packages listed in
ROOTFS_RO_UNNEEDED
during the
generation of the root filesystem.
Set the variable to "1" to force the removal of these packages.
The options to pass in
TARGET_CFLAGS
and CFLAGS
when compiling an optimized system.
This variable defaults to
"-O2 -pipe ${DEBUG_FLAGS}".
Enables Position Independent Executables (PIE) within the GNU C Compiler (GCC). Enabling PIE in the GCC makes Return Oriented Programming (ROP) attacks much more difficult to execute.
By default the security_flags.inc
file enables PIE by setting the variable as follows:
GCCPIE ?= "--enable-default-pie"
Specifies the default version of the GNU C Compiler (GCC)
used for compilation.
By default, GCCVERSION
is set to
"8.x" in the
meta/conf/distro/include/tcmode-default.inc
include file:
GCCVERSION ?= "8.%"
You can override this value by setting it in a configuration
file such as the local.conf
.
The minimal command and arguments to run the GNU Debugger.
The directory in which a local copy of a Git repository is stored when it is cloned.
Specifies the list of GLIBC locales to generate should you not wish to generate all LIBC locals, which can be time consuming.
You can set GLIBC_GENERATE_LOCALES
in your local.conf
file.
By default, all locales are generated.
GLIBC_GENERATE_LOCALES = "en_GB.UTF-8 en_US.UTF-8"
When inheriting the
useradd
class, this variable
specifies for a package what parameters should be passed
to the groupadd
command
if you wish to add a group to the system when the package
is installed.
Here is an example from the dbus
recipe:
GROUPADD_PARAM_${PN} = "-r netdev"
For information on the standard Linux shell command
groupadd
, see
http://linux.die.net/man/8/groupadd.
When inheriting the
useradd
class, this variable
specifies for a package what parameters should be passed
to the groupmems
command
if you wish to modify the members of a group when the
package is installed.
For information on the standard Linux shell command
groupmems
, see
http://linux.die.net/man/8/groupmems.
Configures the GNU GRand Unified Bootloader (GRUB) to have
graphics and serial in the boot menu.
Set this variable to "1" in your
local.conf
or distribution
configuration file to enable graphics and serial
in the menu.
See the
grub-efi
class for more information on how this variable is used.
Additional options to add to the GNU GRand Unified
Bootloader (GRUB) configuration.
Use a semi-colon character (;
) to
separate multiple options.
The GRUB_OPTS
variable is optional.
See the
grub-efi
class for more information on how this variable is used.
Specifies the timeout before executing the default
LABEL
in the GNU GRand Unified
Bootloader (GRUB).
The GRUB_TIMEOUT
variable is optional.
See the
grub-efi
class for more information on how this variable is used.
When inheriting the
gtk-immodules-cache
class, this variable specifies the packages that contain the
GTK+ input method modules being installed when the modules
are in packages other than the main package.
Website where more information about the software the recipe is building can be found.
The name of the target architecture, which is normally
the same as
TARGET_ARCH
.
The OpenEmbedded build system supports many
architectures.
Here is an example list of architectures supported.
This list is by no means complete as the architecture
is configurable:
arm i586 x86_64 powerpc powerpc64 mips mipsel
Specifies architecture-specific compiler flags that are passed to the C compiler.
Default initialization for HOST_CC_ARCH
varies depending on what is being built:
TARGET_CC_ARCH
when building for the target
BUILD_CC_ARCH
when building for the build host (i.e.
-native
)
BUILDSDK_CC_ARCH
when building for an SDK (i.e.
nativesdk-
)
Specifies the name of the target operating system, which
is normally the same as the
TARGET_OS
.
The variable can be set to "linux" for glibc
-based systems and
to "linux-musl" for musl
.
For ARM/EABI targets, there are also "linux-gnueabi" and
"linux-musleabi" values possible.
Specifies the prefix for the cross-compile toolchain.
HOST_PREFIX
is normally the same as
TARGET_PREFIX
.
Specifies the system, including the architecture and the operating system, for which the build is occurring in the context of the current recipe.
The OpenEmbedded build system automatically sets this
variable based on
HOST_ARCH
,
HOST_VENDOR
,
and
HOST_OS
variables.
Consider these two examples:
Given a native recipe on a 32-bit x86 machine running Linux, the value is "i686-linux".
Given a recipe being built for a little-endian MIPS target running Linux, the value might be "mipsel-linux".
A space-separated list (filter) of tools on the build host
that should be allowed to be called from within build tasks.
Using this filter helps reduce the possibility of host
contamination.
If a tool specified in the value of
HOSTTOOLS
is not found on the
build host, the OpenEmbedded build system produces
an error and the build is not started.
For additional information, see
HOSTTOOLS_NONFATAL
.
A space-separated list (filter) of tools on the build host
that should be allowed to be called from within build tasks.
Using this filter helps reduce the possibility of host
contamination.
Unlike
HOSTTOOLS
,
the OpenEmbedded build system does not produce an error
if a tool specified in the value of
HOSTTOOLS_NONFATAL
is not found on the
build host.
Thus, you can use HOSTTOOLS_NONFATAL
to filter optional host tools.
Specifies the name of the vendor.
HOST_VENDOR
is normally the same as
TARGET_VENDOR
.
Disables or enables the icecc
(Icecream) function.
For more information on this function and best practices
for using this variable, see the
"icecc.bbclass
"
section.
Setting this variable to "1" in your
local.conf
disables the function:
ICECC_DISABLED ??= "1"
To enable the function, set the variable as follows:
ICECC_DISABLED = ""
Points to the icecc-create-env
script
that you provide.
This variable is used by the
icecc
class.
You set this variable in your
local.conf
file.
If you do not point to a script that you provide, the
OpenEmbedded build system uses the default script provided
by the icecc-create-env.bb
recipe,
which is a modified version and not the one that comes with
icecc
.
Extra options passed to the make
command during the
do_compile
task that specify parallel compilation.
This variable usually takes the form of
"-j x
", where
x
represents the maximum
number of parallel threads make
can
run.
iceccd
daemon.
If your enabled machines support multiple cores,
coming up with the maximum number of parallel threads
that gives you the best performance could take some
experimentation since machine speed, network lag,
available memory, and existing machine loads can all
affect build time.
Consequently, unlike the
PARALLEL_MAKE
variable, there is no rule-of-thumb for setting
ICECC_PARALLEL_MAKE
to achieve
optimal performance.
If you do not set ICECC_PARALLEL_MAKE
,
the build system does not use it (i.e. the system does
not detect and assign the number of cores as is done with
PARALLEL_MAKE
).
The location of the icecc
binary.
You can set this variable in your
local.conf
file.
If your local.conf
file does not define
this variable, the
icecc
class attempts to define it by locating
icecc
using which
.
Identifies user classes that you do not want the
Icecream distributed compile support to consider.
This variable is used by the
icecc
class.
You set this variable in your
local.conf
file.
When you list classes using this variable, you are "blacklisting" them from distributed compilation across remote hosts. Any classes you list will be distributed and compiled locally.
Identifies user recipes that you do not want the
Icecream distributed compile support to consider.
This variable is used by the
icecc
class.
You set this variable in your
local.conf
file.
When you list packages using this variable, you are "blacklisting" them from distributed compilation across remote hosts. Any packages you list will be distributed and compiled locally.
Identifies user recipes that use an empty
PARALLEL_MAKE
variable that you want to force remote distributed
compilation on using the Icecream distributed compile
support.
This variable is used by the
icecc
class.
You set this variable in your
local.conf
file.
The base name of image output files.
This variable defaults to the recipe name
(${
PN
}
).
A space-separated list of files installed into the
boot partition when preparing an image using the Wic tool
with the bootimg-partition
source
plugin.
By default, the files are installed under the same name as
the source files.
To change the installed name, separate it from the
original name with a semi-colon (;).
Source files need to be located in
DEPLOY_DIR_IMAGE
.
Here are two examples:
IMAGE_BOOT_FILES = "u-boot.img uImage;kernel" IMAGE_BOOT_FILES = "u-boot.${UBOOT_SUFFIX} ${KERNEL_IMAGETYPE}"
Alternatively, source files can be picked up using a glob pattern. In this case, the destination file must have the same name as the base name of the source file path. To install files into a directory within the target location, pass its name after a semi-colon (;). Here are two examples:
IMAGE_BOOT_FILES = "bcm2835-bootfiles/*" IMAGE_BOOT_FILES = "bcm2835-bootfiles/*;boot/"
The first example installs all files from
${DEPLOY_DIR_IMAGE}/bcm2835-bootfiles
into the root of the target partition.
The second example installs the same files into a
boot
directory within the
target partition.
You can find information on how to use the Wic tool in the "Creating Partitioned Images Using Wic" section of the Yocto Project Development Tasks Manual. Reference material for Wic is located in the "OpenEmbedded Kickstart (.wks) Reference" chapter.
A list of classes that all images should inherit. You typically use this variable to specify the list of classes that register the different types of images the OpenEmbedded build system creates.
The default value for IMAGE_CLASSES
is
image_types
.
You can set this variable in your
local.conf
or in a distribution
configuration file.
For more information, see
meta/classes/image_types.bbclass
in the
Source Directory.
Specifies the command to create the image file for a
specific image type, which corresponds to the value set
set in
IMAGE_FSTYPES
,
(e.g. ext3
,
btrfs
, and so forth).
When setting this variable, you should use
an override for the associated type.
Here is an example:
IMAGE_CMD_jffs2 = "mkfs.jffs2 --root=${IMAGE_ROOTFS} \ --faketime --output=${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.rootfs.jffs2 \ ${EXTRA_IMAGECMD}"
You typically do not need to set this variable unless
you are adding support for a new image type.
For more examples on how to set this variable, see the
image_types
class file, which is
meta/classes/image_types.bbclass
.
Specifies one or more files that contain custom device
tables that are passed to the
makedevs
command as part of creating
an image.
These files list basic device nodes that should be
created under /dev
within the image.
If IMAGE_DEVICE_TABLES
is not set,
files/device_table-minimal.txt
is
used, which is located by
BBPATH
.
For details on how you should write device table files,
see meta/files/device_table-minimal.txt
as an example.
The primary list of features to include in an image.
Typically, you configure this variable in an image recipe.
Although you can use this variable from your
local.conf
file, which is found in the
Build Directory,
best practices dictate that you do not.
EXTRA_IMAGE_FEATURES
variable.
For a list of image features that ships with the Yocto Project, see the "Image Features" section.
For an example that shows how to customize your image by
using this variable, see the
"Customizing Images Using Custom IMAGE_FEATURES
and EXTRA_IMAGE_FEATURES
"
section in the Yocto Project Development Tasks Manual.
Specifies the formats the OpenEmbedded build system uses
during the build when creating the root filesystem.
For example, setting IMAGE_FSTYPES
as follows causes the build system to create root
filesystems using two formats: .ext3
and .tar.bz2
:
IMAGE_FSTYPES = "ext3 tar.bz2"
For the complete list of supported image formats from which
you can choose, see
IMAGE_TYPES
.
If an image recipe uses the "inherit image" line
and you are setting
IMAGE_FSTYPES
inside the
recipe, you must set
IMAGE_FSTYPES
prior to
using the "inherit image" line.
Due to the way the OpenEmbedded build system
processes this variable, you cannot update its
contents by using _append
or
_prepend
.
You must use the +=
operator to add one or more options to the
IMAGE_FSTYPES
variable.
Used by recipes to specify the packages to install into an
image through the
image
class.
Use the IMAGE_INSTALL
variable with
care to avoid ordering issues.
Image recipes set IMAGE_INSTALL
to specify the packages to install into an image through
image.bbclass
.
Additionally, "helper" classes such as the
core-image
class exist that can take lists used with
IMAGE_FEATURES
and turn them into auto-generated entries in
IMAGE_INSTALL
in addition to its
default contents.
When you use this variable, it is best to use it as follows:
IMAGE_INSTALL_append = " package-name
"
Be sure to include the space between the quotation character and the start of the package name or names.
When working with a
core-image-minimal-initramfs
image, do not use the
IMAGE_INSTALL
variable to
specify packages for installation.
Instead, use the
PACKAGE_INSTALL
variable, which allows the initial RAM
filesystem (initramfs) recipe to use a fixed
set of packages and not be affected by
IMAGE_INSTALL
.
For information on creating an initramfs, see
the
"Building an Initial RAM Filesystem (initramfs) Image"
section in the Yocto Project Development Tasks
Manual.
Using IMAGE_INSTALL
with
the
+=
BitBake operator within the
/conf/local.conf
file or
from within an image recipe is not recommended.
Use of this operator in these ways can cause
ordering issues.
Since core-image.bbclass
sets IMAGE_INSTALL
to a
default value using the
?=
operator, using a +=
operation against
IMAGE_INSTALL
results in
unexpected behavior when used within
conf/local.conf
.
Furthermore, the same operation from within
an image recipe may or may not succeed
depending on the specific situation.
In both these cases, the behavior is contrary
to how most users expect the
+=
operator to work.
Specifies the list of locales to install into the image
during the root filesystem construction process.
The OpenEmbedded build system automatically splits locale
files, which are used for localization, into separate
packages.
Setting the IMAGE_LINGUAS
variable
ensures that any locale packages that correspond to packages
already selected for installation into the image are also
installed.
Here is an example:
IMAGE_LINGUAS = "pt-br de-de"
In this example, the build system ensures any Brazilian
Portuguese and German locale files that correspond to
packages in the image are installed (i.e.
*-locale-pt-br
and *-locale-de-de
as well as
*-locale-pt
and *-locale-de
, since some software
packages only provide locale files by language and not by
country-specific language).
See the
GLIBC_GENERATE_LOCALES
variable for information on generating GLIBC locales.
The manifest file for the image. This file lists all the installed packages that make up the image. The file contains package information on a line-per-package basis as follows:
packagename
packagearch
version
The
image
class defines the manifest file as follows:
IMAGE_MANIFEST = "${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.rootfs.manifest"
The location is derived using the
DEPLOY_DIR_IMAGE
and
IMAGE_NAME
variables.
You can find information on how the image
is created in the
"Image Generation"
section in the Yocto Project Overview and Concepts Manual.
The name of the output image files minus the extension.
This variable is derived using the
IMAGE_BASENAME
,
MACHINE
,
and
DATETIME
variables:
IMAGE_NAME = "${IMAGE_BASENAME}-${MACHINE}-${DATETIME}"
Defines a multiplier that the build system applies to the initial image
size for cases when the multiplier times the returned disk usage value
for the image is greater than the sum of
IMAGE_ROOTFS_SIZE
and
IMAGE_ROOTFS_EXTRA_SPACE
.
The result of the multiplier applied to the initial image size creates
free disk space in the image as overhead.
By default, the build process uses a multiplier of 1.3 for this variable.
This default value results in 30% free disk space added to the image when this
method is used to determine the final generated image size.
You should be aware that post install scripts and the package management
system uses disk space inside this overhead area.
Consequently, the multiplier does not produce an image with
all the theoretical free disk space.
See IMAGE_ROOTFS_SIZE
for information on how the build system determines the overall image size.
The default 30% free disk space typically gives the image enough room to boot and allows for basic post installs while still leaving a small amount of free disk space. If 30% free space is inadequate, you can increase the default value. For example, the following setting gives you 50% free space added to the image:
IMAGE_OVERHEAD_FACTOR = "1.5"
Alternatively, you can ensure a specific amount of free disk space is added
to the image by using the
IMAGE_ROOTFS_EXTRA_SPACE
variable.
Defines the package type (i.e. DEB, RPM, IPK, or TAR) used
by the OpenEmbedded build system.
The variable is defined appropriately by the
package_deb
,
package_rpm
,
package_ipk
,
or
package_tar
class.
package_tar
class is broken
and is not supported.
It is recommended that you do not use it.
The
populate_sdk_*
and
image
classes use the IMAGE_PKGTYPE
for
packaging up images and SDKs.
You should not set the IMAGE_PKGTYPE
manually.
Rather, the variable is set indirectly through the
appropriate
package_*
class using the
PACKAGE_CLASSES
variable.
The OpenEmbedded build system uses the first package type
(e.g. DEB, RPM, or IPK) that appears with the variable
.tar
format are
never used as a substitute packaging format for DEB,
RPM, and IPK formatted files for your image or SDK.
Specifies a list of functions to call once the OpenEmbedded build system creates the final image output files. You can specify functions separated by semicolons:
IMAGE_POSTPROCESS_COMMAND += "function
; ... "
If you need to pass the root filesystem path to a command
within the function, you can use
${IMAGE_ROOTFS}
, which points to
the directory that becomes the root filesystem image.
See the
IMAGE_ROOTFS
variable for more information.
Specifies a list of functions to call before the OpenEmbedded build system creates the final image output files. You can specify functions separated by semicolons:
IMAGE_PREPROCESS_COMMAND += "function
; ... "
If you need to pass the root filesystem path to a command
within the function, you can use
${IMAGE_ROOTFS}
, which points to
the directory that becomes the root filesystem image.
See the
IMAGE_ROOTFS
variable for more information.
The location of the root filesystem while it is under
construction (i.e. during the
do_rootfs
task).
This variable is not configurable.
Do not change it.
Specifies the alignment for the output image file in
Kbytes.
If the size of the image is not a multiple of
this value, then the size is rounded up to the nearest
multiple of the value.
The default value is "1".
See
IMAGE_ROOTFS_SIZE
for additional information.
Defines additional free disk space created in the image in Kbytes.
By default, this variable is set to "0".
This free disk space is added to the image after the build system determines
the image size as described in
IMAGE_ROOTFS_SIZE
.
This variable is particularly useful when you want to ensure that a specific amount of free disk space is available on a device after an image is installed and running. For example, to be sure 5 Gbytes of free disk space is available, set the variable as follows:
IMAGE_ROOTFS_EXTRA_SPACE = "5242880"
For example, the Yocto Project Build Appliance specifically requests 40 Gbytes of extra space with the line:
IMAGE_ROOTFS_EXTRA_SPACE = "41943040"
Defines the size in Kbytes for the generated image. The OpenEmbedded build system determines the final size for the generated image using an algorithm that takes into account the initial disk space used for the generated image, a requested size for the image, and requested additional free disk space to be added to the image. Programatically, the build system determines the final size of the generated image as follows:
if (image-du * overhead) < rootfs-size: internal-rootfs-size = rootfs-size + xspace else: internal-rootfs-size = (image-du * overhead) + xspace where: image-du = Returned value of the du command on the image. overhead = IMAGE_OVERHEAD_FACTOR rootfs-size = IMAGE_ROOTFS_SIZE internal-rootfs-size = Initial root filesystem size before any modifications. xspace = IMAGE_ROOTFS_EXTRA_SPACE
See the IMAGE_OVERHEAD_FACTOR
and IMAGE_ROOTFS_EXTRA_SPACE
variables for related information.
Specifies a dependency from one image type on another.
Here is an example from the
image-live
class:
IMAGE_TYPEDEP_live = "ext3"
In the previous example, the variable ensures that when
"live" is listed with the
IMAGE_FSTYPES
variable, the OpenEmbedded build system produces an
ext3
image first since one of the
components of the live
image is an ext3
formatted partition containing the root
filesystem.
Specifies the complete list of supported image types by default:
btrfs container cpio cpio.gz cpio.lz4 cpio.lzma cpio.xz cramfs ext2 ext2.bz2 ext2.gz ext2.lzma ext3 ext3.gz ext4 ext4.gz f2fs hddimg iso jffs2 jffs2.sum multiubi squashfs squashfs-lz4 squashfs-lzo squashfs-xz tar tar.bz2 tar.gz tar.lz4 tar.xz tar.zst ubi ubifs wic wic.bz2 wic.gz wic.lzma
For more information about these types of images, see
meta/classes/image_types*.bbclass
in the
Source Directory.
Helps define the recipe revision for recipes that share
a common include
file.
You can think of this variable as part of the recipe revision
as set from within an include file.
Suppose, for example, you have a set of recipes that
are used across several projects.
And, within each of those recipes the revision
(its PR
value) is set accordingly.
In this case, when the revision of those recipes changes,
the burden is on you to find all those recipes and
be sure that they get changed to reflect the updated
version of the recipe.
In this scenario, it can get complicated when recipes
that are used in many places and provide common functionality
are upgraded to a new revision.
A more efficient way of dealing with this situation is
to set the INC_PR
variable inside
the include
files that the recipes
share and then expand the INC_PR
variable within the recipes to help
define the recipe revision.
The following provides an example that shows how to use
the INC_PR
variable
given a common include
file that
defines the variable.
Once the variable is defined in the
include
file, you can use the
variable to set the PR
values in
each recipe.
You will notice that when you set a recipe's
PR
you can provide more granular
revisioning by appending values to the
INC_PR
variable:
recipes-graphics/xorg-font/xorg-font-common.inc:INC_PR = "r2" recipes-graphics/xorg-font/encodings_1.0.4.bb:PR = "${INC_PR}.1" recipes-graphics/xorg-font/font-util_1.3.0.bb:PR = "${INC_PR}.0" recipes-graphics/xorg-font/font-alias_1.0.3.bb:PR = "${INC_PR}.3"
The first line of the example establishes the baseline
revision to be used for all recipes that use the
include
file.
The remaining lines in the example are from individual
recipes and show how the PR
value
is set.
Specifies a space-separated list of license names
(as they would appear in
LICENSE
)
that should be excluded from the build.
Recipes that provide no alternatives to listed incompatible
licenses are not built.
Packages that are individually licensed with the specified
incompatible licenses will be deleted.
INCOMPATIBLE_LICENSE = "GPL-3.0 LGPL-3.0 AGPL-3.0"Although you can use other settings, you might be required to remove dependencies on or provide alternatives to components that are required to produce a functional system image.
COMPATIBLE_LICENSES
with the
names of the licences that are allowed. Then
define INCOMPATIBLE_LICENSE
as:
INCOMPATIBLE_LICENSE = "${@' '.join(sorted(set(d.getVar('AVAILABLE_LICENSES').split()) - set(d.getVar('COMPATIBLE_LICENSES').split())))}"This will result in
INCOMPATIBLE_LICENSE
containing the
names of all licences
from AVAILABLE_LICENSES
except the ones specified
in COMPATIBLE_LICENSES
, thus only
allowing the latter licences to be used.
Causes the named class or classes to be inherited globally.
Anonymous functions in the class or classes
are not executed for the
base configuration and in each individual recipe.
The OpenEmbedded build system ignores changes to
INHERIT
in individual recipes.
For more information on INHERIT
, see
the
"INHERIT
Configuration Directive"
section in the Bitbake User Manual.
Lists classes that will be inherited at the distribution level. It is unlikely that you want to edit this variable.
The default value of the variable is set as follows in the
meta/conf/distro/defaultsetup.conf
file:
INHERIT_DISTRO ?= "debian devshell sstate license"
Prevents the default dependencies, namely the C compiler
and standard C library (libc), from being added to
DEPENDS
.
This variable is usually used within recipes that do not
require any compilation using the C compiler.
Set the variable to "1" to prevent the default dependencies from being added.
Prevents the OpenEmbedded build system from splitting
out debug information during packaging.
By default, the build system splits out debugging
information during the
do_package
task.
For more information on how debug information is split out,
see the
PACKAGE_DEBUG_SPLIT_STYLE
variable.
To prevent the build system from splitting out
debug information during packaging, set the
INHIBIT_PACKAGE_DEBUG_SPLIT
variable
as follows:
INHIBIT_PACKAGE_DEBUG_SPLIT = "1"
If set to "1", causes the build to not strip binaries in
resulting packages and prevents the
-dbg
package from containing the
source files.
By default, the OpenEmbedded build system strips
binaries and puts the debugging symbols into
${
PN
}-dbg
.
Consequently, you should not set
INHIBIT_PACKAGE_STRIP
when you plan
to debug in general.
If set to "1", causes the build to not strip binaries in the resulting sysroot.
By default, the OpenEmbedded build system strips
binaries in the resulting sysroot.
When you specifically set the
INHIBIT_SYSROOT_STRIP
variable to
"1" in your recipe, you inhibit this stripping.
If you want to use this variable, include the
staging
class.
This class uses a sys_strip()
function to test for the variable and acts accordingly.
INHIBIT_SYSROOT_STRIP
variable occurs in rare and special circumstances.
For example, suppose you are building bare-metal
firmware by using an external GCC toolchain.
Furthermore, even if the toolchain's binaries are
strippable, other files exist that are needed for the
build that are not strippable.
Defines the format for the output image of an initial
RAM filesystem (initramfs), which is used during boot.
Supported formats are the same as those supported by the
IMAGE_FSTYPES
variable.
The default value of this variable, which is set in the
meta/conf/bitbake.conf
configuration
file in the
Source Directory,
is "cpio.gz".
The Linux kernel's initramfs mechanism, as opposed to the
initial RAM filesystem
initrd
mechanism, expects an optionally compressed cpio
archive.
Specifies the
PROVIDES
name of an image recipe that is used to build an initial
RAM filesystem (initramfs) image.
In other words, the INITRAMFS_IMAGE
variable causes an additional recipe to be built as
a dependency to whatever root filesystem recipe you
might be using (e.g. core-image-sato
).
The initramfs image recipe you provide should set
IMAGE_FSTYPES
to
INITRAMFS_FSTYPES
.
An initramfs image provides a temporary root filesystem used for early system initialization (e.g. loading of modules needed to locate and mount the "real" root filesystem).
meta/recipes-core/images/core-image-minimal-initramfs.bb
recipe in the
Source Directory
for an example initramfs recipe.
To select this sample recipe as the one built
to provide the initramfs image,
set INITRAMFS_IMAGE
to
"core-image-minimal-initramfs".
You can also find more information by referencing the
meta-poky/conf/local.conf.sample.extended
configuration file in the Source Directory,
the
image
class, and the
kernel
class to see how to use the
INITRAMFS_IMAGE
variable.
If INITRAMFS_IMAGE
is empty, which is
the default, then no initramfs image is built.
For more information, you can also see the
INITRAMFS_IMAGE_BUNDLE
variable, which allows the generated image to be bundled
inside the kernel image.
Additionally, for information on creating an initramfs
image, see the
"Building an Initial RAM Filesystem (initramfs) Image"
section in the Yocto Project Development Tasks Manual.
Controls whether or not the image recipe specified by
INITRAMFS_IMAGE
is run through an extra pass
(do_bundle_initramfs
)
during kernel compilation in order to build a single binary
that contains both the kernel image and the initial RAM
filesystem (initramfs) image.
This makes use of the
CONFIG_INITRAMFS_SOURCE
kernel feature.
The combined binary is deposited into the
tmp/deploy
directory, which is part
of the
Build Directory.
Setting the variable to "1" in a configuration file causes the
OpenEmbedded build system to generate a kernel image with the
initramfs specified in INITRAMFS_IMAGE
bundled within:
INITRAMFS_IMAGE_BUNDLE = "1"
By default, the
kernel
class sets this variable to a null string as follows:
INITRAMFS_IMAGE_BUNDLE ?= ""
INITRAMFS_IMAGE_BUNDLE
variable in
a configuration file.
You cannot set the variable in a recipe file.
See the
local.conf.sample.extended
file for additional information.
Also, for information on creating an initramfs, see the
"Building an Initial RAM Filesystem (initramfs) Image"
section in the Yocto Project Development Tasks Manual.
The link name of the initial RAM filesystem image.
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
INITRAMFS_LINK_NAME ?= "initramfs-${KERNEL_ARTIFACT_LINK_NAME}"
The value of the KERNEL_ARTIFACT_LINK_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_LINK_NAME ?= "${MACHINE}"
See the
MACHINE
variable for additional information.
The base name of the initial RAM filesystem image.
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
INITRAMFS_NAME ?= "initramfs-${KERNEL_ARTIFACT_NAME}"
The value of the
KERNEL_ARTIFACT_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_NAME ?= "${PKGE}-${PKGV}-${PKGR}-${MACHINE}${IMAGE_VERSION_SUFFIX}"
Indicates list of filesystem images to concatenate and use
as an initial RAM disk (initrd
).
The INITRD
variable is an optional
variable used with the
image-live
class.
When building a "live" bootable image (i.e. when
IMAGE_FSTYPES
contains "live"), INITRD_IMAGE
specifies the image recipe that should be built
to provide the initial RAM disk image.
The default value is "core-image-minimal-initramfs".
See the
image-live
class for more information.
The filename of the initialization script as installed to
${sysconfdir}/init.d
.
This variable is used in recipes when using update-rc.d.bbclass
.
The variable is mandatory.
A list of the packages that contain initscripts.
If multiple packages are specified, you need to append the package name
to the other INITSCRIPT_*
as an override.
This variable is used in recipes when using update-rc.d.bbclass
.
The variable is optional and defaults to the
PN
variable.
Specifies the options to pass to update-rc.d
.
Here is an example:
INITSCRIPT_PARAMS = "start 99 5 2 . stop 20 0 1 6 ."
In this example, the script has a runlevel of 99, starts the script in initlevels 2 and 5, and stops the script in levels 0, 1 and 6.
The variable's default value is "defaults", which is
set in the
update-rc.d
class.
The value in
INITSCRIPT_PARAMS
is passed through
to the update-rc.d
command.
For more information on valid parameters, please see the
update-rc.d
manual page at
http://www.tin.org/bin/man.cgi?section=8&topic=update-rc.d.
Specifies the QA checks to skip for a specific package
within a recipe.
For example, to skip the check for symbolic link
.so
files in the main package of a
recipe, add the following to the recipe.
The package name override must be used, which in this
example is ${PN}
:
INSANE_SKIP_${PN} += "dev-so"
See the "insane.bbclass
"
section for a list of the valid QA checks you can
specify using this variable.
By default, the tzdata
recipe packages
an /etc/timezone
file.
Set the INSTALL_TIMEZONE_FILE
variable to "0" at the configuration level to disable this
behavior.
When the IPK backend is in use and package management
is enabled on the target, you can use this variable to
set up opkg
in the target image
to point to package feeds on a nominated server.
Once the feed is established, you can perform
installations or upgrades using the package manager
at runtime.
Defines the kernel architecture used when assembling the configuration. Architectures supported for this release are:
powerpc i386 x86_64 arm qemu mips
You define the KARCH
variable in the
BSP Descriptions.
A regular expression used by the build process to explicitly identify the kernel branch that is validated, patched, and configured during a build. You must set this variable to ensure the exact kernel branch you want is being used by the build process.
Values for this variable are set in the kernel's recipe
file and the kernel's append file.
For example, if you are using the
linux-yocto_4.12
kernel, the kernel
recipe file is the
meta/recipes-kernel/linux/linux-yocto_4.12.bb
file.
KBRANCH
is set as follows in that
kernel recipe file:
KBRANCH ?= "standard/base"
This variable is also used from the kernel's append file
to identify the kernel branch specific to a particular
machine or target hardware.
Continuing with the previous kernel example, the kernel's
append file (i.e.
linux-yocto_4.12.bbappend
) is located
in the BSP layer for a given machine.
For example, the append file for the Beaglebone,
EdgeRouter, and generic versions of both 32 and 64-bit IA
machines (meta-yocto-bsp
) is named
meta-yocto-bsp/recipes-kernel/linux/linux-yocto_4.12.bbappend
.
Here are the related statements from that append file:
KBRANCH_genericx86 = "standard/base" KBRANCH_genericx86-64 = "standard/base" KBRANCH_edgerouter = "standard/edgerouter" KBRANCH_beaglebone = "standard/beaglebone"
The KBRANCH
statements identify
the kernel branch to use when building for each
supported BSP.
When used with the
kernel-yocto
class, specifies an "in-tree" kernel configuration file
for use during a kernel build.
Typically, when using a defconfig
to
configure a kernel during a build, you place the
file in your layer in the same manner as you would
place patch files and configuration fragment files (i.e.
"out-of-tree").
However, if you want to use a defconfig
file that is part of the kernel tree (i.e. "in-tree"),
you can use the
KBUILD_DEFCONFIG
variable and append
the
KMACHINE
variable to point to the defconfig
file.
To use the variable, set it in the append file for your kernel recipe using the following form:
KBUILD_DEFCONFIG_KMACHINE
?=defconfig_file
Here is an example from a "raspberrypi2"
KMACHINE
build that uses a
defconfig
file named
"bcm2709_defconfig":
KBUILD_DEFCONFIG_raspberrypi2 = "bcm2709_defconfig"
As an alternative, you can use the following within your append file:
KBUILD_DEFCONFIG_pn-linux-yocto ?= defconfig_file
For more information on how to use the
KBUILD_DEFCONFIG
variable, see the
"Using an "In-Tree" defconfig
File"
section in the Yocto Project Linux Kernel Development
Manual.
Specifies an alternate kernel image type for creation in
addition to the kernel image type specified using the
KERNEL_IMAGETYPE
variable.
Specifies the name of all of the build artifacts.
You can change the name of the artifacts by changing the
KERNEL_ARTIFACT_NAME
variable.
The value of KERNEL_ARTIFACT_NAME
,
which is set in the
meta/classes/kernel-artifact-names.bbclass
file, has the following default value:
KERNEL_ARTIFACT_NAME ?= "${PKGE}-${PKGV}-${PKGR}-${MACHINE}${IMAGE_VERSION_SUFFIX}"
See the
PKGE
,
PKGV
,
PKGR
,
and
MACHINE
variables for additional information.
A list of classes defining kernel image types that the
kernel
class should inherit.
You typically append this variable to enable extended image
types.
An example is the "kernel-fitimage", which enables
fitImage support and resides in
meta/classes/kernel-fitimage.bbclass
.
You can register custom kernel image types with the
kernel
class using this variable.
Specifies the name of the generated Linux kernel device tree
(i.e. the .dtb
) file.
.dtb
file is preferred.
In order to use this variable, the
kernel-devicetree
class must be inherited.
The link name of the kernel device tree binary (DTB).
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
KERNEL_DTB_LINK_NAME ?= "${KERNEL_ARTIFACT_LINK_NAME}"
The value of the KERNEL_ARTIFACT_LINK_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_LINK_NAME ?= "${MACHINE}"
See the
MACHINE
variable for additional information.
The base name of the kernel device tree binary (DTB).
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
KERNEL_DTB_NAME ?= "${KERNEL_ARTIFACT_NAME}"
The value of the
KERNEL_ARTIFACT_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_NAME ?= "${PKGE}-${PKGV}-${PKGR}-${MACHINE}${IMAGE_VERSION_SUFFIX}"
Specifies additional make
command-line arguments the OpenEmbedded build system
passes on when compiling the kernel.
Includes additional kernel metadata.
In the OpenEmbedded build system, the default Board Support
Packages (BSPs)
Metadata
is provided through
the KMACHINE
and
KBRANCH
variables.
You can use the KERNEL_FEATURES
variable from within the kernel recipe or kernel append
file to further add metadata for all BSPs or specific
BSPs.
The metadata you add through this variable includes config
fragments and features descriptions,
which usually includes patches as well as config fragments.
You typically override the
KERNEL_FEATURES
variable for a
specific machine.
In this way, you can provide validated, but optional,
sets of kernel configurations and features.
For example, the following example from the
linux-yocto-rt_4.12
kernel recipe
adds "netfilter" and "taskstats" features to all BSPs
as well as "virtio" configurations to all QEMU machines.
The last two statements add specific configurations to
targeted machine types:
KERNEL_EXTRA_FEATURES ?= "features/netfilter/netfilter.scc features/taskstats/taskstats.scc" KERNEL_FEATURES_append = " ${KERNEL_EXTRA_FEATURES}" KERNEL_FEATURES_append_qemuall = " cfg/virtio.scc" KERNEL_FEATURES_append_qemux86 = " cfg/sound.scc cfg/paravirt_kvm.scc" KERNEL_FEATURES_append_qemux86-64 = " cfg/sound.scc"
The link name of the kernel flattened image tree (FIT) image.
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
KERNEL_FIT_LINK_NAME ?= "${KERNEL_ARTIFACT_LINK_NAME}"
The value of the KERNEL_ARTIFACT_LINK_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_LINK_NAME ?= "${MACHINE}"
See the
MACHINE
variable for additional information.
The base name of the kernel flattened image tree (FIT) image.
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
KERNEL_FIT_NAME ?= "${KERNEL_ARTIFACT_NAME}"
The value of the
KERNEL_ARTIFACT_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_NAME ?= "${PKGE}-${PKGV}-${PKGR}-${MACHINE}${IMAGE_VERSION_SUFFIX}"
The link name for the kernel image.
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
KERNEL_IMAGE_LINK_NAME ?= "${KERNEL_ARTIFACT_LINK_NAME}"
The value of the KERNEL_ARTIFACT_LINK_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_LINK_NAME ?= "${MACHINE}"
See the
MACHINE
variable for additional information.
Specifies the maximum size of the kernel image file in
kilobytes.
If KERNEL_IMAGE_MAXSIZE
is set,
the size of the kernel image file is checked against
the set value during the
do_sizecheck
task.
The task fails if the kernel image file is larger than
the setting.
KERNEL_IMAGE_MAXSIZE
is useful for
target devices that have a limited amount of space in
which the kernel image must be stored.
By default, this variable is not set, which means the size of the kernel image is not checked.
The base name of the kernel image.
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
KERNEL_IMAGE_NAME ?= "${KERNEL_ARTIFACT_NAME}"
The value of the
KERNEL_ARTIFACT_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_NAME ?= "${PKGE}-${PKGV}-${PKGR}-${MACHINE}${IMAGE_VERSION_SUFFIX}"
The type of kernel to build for a device, usually set by the
machine configuration files and defaults to "zImage".
This variable is used
when building the kernel and is passed to make
as the target to
build.
If you want to build an alternate kernel image type, use the
KERNEL_ALT_IMAGETYPE
variable.
Lists kernel modules that need to be auto-loaded during boot.
module_autoload
variable.
You can use the KERNEL_MODULE_AUTOLOAD
variable anywhere that it can be
recognized by the kernel recipe or by an out-of-tree kernel
module recipe (e.g. a machine configuration file, a
distribution configuration file, an append file for the
recipe, or the recipe itself).
Specify it as follows:
KERNEL_MODULE_AUTOLOAD += "module_name1
module_name2
module_name3
"
Including KERNEL_MODULE_AUTOLOAD
causes
the OpenEmbedded build system to populate the
/etc/modules-load.d/modname.conf
file with the list of modules to be auto-loaded on boot.
The modules appear one-per-line in the file.
Here is an example of the most common use case:
KERNEL_MODULE_AUTOLOAD += "module_name
"
For information on how to populate the
modname.conf
file with
modprobe.d
syntax lines, see the
KERNEL_MODULE_PROBECONF
variable.
Provides a list of modules for which the OpenEmbedded
build system expects to find
module_conf_
modname
values that specify configuration for each of the modules.
For information on how to provide those module
configurations, see the
module_conf_*
variable.
The location of the kernel sources.
This variable is set to the value of the
STAGING_KERNEL_DIR
within the
module
class.
For information on how this variable is used, see the
"Incorporating Out-of-Tree Modules"
section in the Yocto Project Linux Kernel Development
Manual.
To help maximize compatibility with out-of-tree drivers
used to build modules, the OpenEmbedded build system also
recognizes and uses the
KERNEL_SRC
variable, which is identical to the
KERNEL_PATH
variable.
Both variables are common variables used by external
Makefiles to point to the kernel source directory.
The location of the kernel sources.
This variable is set to the value of the
STAGING_KERNEL_DIR
within the
module
class.
For information on how this variable is used, see the
"Incorporating Out-of-Tree Modules"
section in the Yocto Project Linux Kernel Development
Manual.
To help maximize compatibility with out-of-tree drivers
used to build modules, the OpenEmbedded build system also
recognizes and uses the
KERNEL_PATH
variable, which is identical to the
KERNEL_SRC
variable.
Both variables are common variables used by external
Makefiles to point to the kernel source directory.
Specifies the version of the kernel as extracted from
version.h
or
utsrelease.h
within the kernel sources.
Effects of setting this variable do not take affect until
the kernel has been configured.
Consequently, attempting to refer to this variable in
contexts prior to configuration will not work.
Specifies whether the data referenced through
PKGDATA_DIR
is needed or not.
The KERNELDEPMODDEPEND
does not
control whether or not that data exists,
but simply whether or not it is used.
If you do not need to use the data, set the
KERNELDEPMODDEPEND
variable in your
initramfs
recipe.
Setting the variable there when the data is not needed
avoids a potential dependency loop.
Provides a short description of a configuration fragment.
You use this variable in the .scc
file that describes a configuration fragment file.
Here is the variable used in a file named
smp.scc
to describe SMP being
enabled:
define KFEATURE_DESCRIPTION "Enable SMP"
The machine as known by the kernel.
Sometimes the machine name used by the kernel does not
match the machine name used by the OpenEmbedded build
system.
For example, the machine name that the OpenEmbedded build
system understands as
core2-32-intel-common
goes by a
different name in the Linux Yocto kernel.
The kernel understands that machine as
intel-core2-32
.
For cases like these, the KMACHINE
variable maps the kernel machine name to the OpenEmbedded
build system machine name.
These mappings between different names occur in the
Yocto Linux Kernel's meta
branch.
As an example take a look in the
common/recipes-kernel/linux/linux-yocto_3.19.bbappend
file:
LINUX_VERSION_core2-32-intel-common = "3.19.0" COMPATIBLE_MACHINE_core2-32-intel-common = "${MACHINE}" SRCREV_meta_core2-32-intel-common = "8897ef68b30e7426bc1d39895e71fb155d694974" SRCREV_machine_core2-32-intel-common = "43b9eced9ba8a57add36af07736344dcc383f711" KMACHINE_core2-32-intel-common = "intel-core2-32" KBRANCH_core2-32-intel-common = "standard/base" KERNEL_FEATURES_append_core2-32-intel-common = "${KERNEL_FEATURES_INTEL_COMMON}"
The KMACHINE
statement says that
the kernel understands the machine name as
"intel-core2-32".
However, the OpenEmbedded build system understands the
machine as "core2-32-intel-common".
Defines the kernel type to be used in assembling the configuration. The linux-yocto recipes define "standard", "tiny", and "preempt-rt" kernel types. See the "Kernel Types" section in the Yocto Project Linux Kernel Development Manual for more information on kernel types.
You define the KTYPE
variable in the
BSP Descriptions.
The value you use must match the value used for the
LINUX_KERNEL_TYPE
value used by the kernel recipe.
Provides a list of targets for automatic configuration.
See the
grub-efi
class for more information on how this variable is used.
Lists the layers, separated by spaces, on which this recipe depends. Optionally, you can specify a specific layer version for a dependency by adding it to the end of the layer name. Here is an example:
LAYERDEPENDS_mylayer = "anotherlayer (=3)"
In this previous example, version 3 of "anotherlayer"
is compared against
LAYERVERSION
_anotherlayer
.
An error is produced if any dependency is missing or
the version numbers (if specified) do not match exactly.
This variable is used in the
conf/layer.conf
file and must be
suffixed with the name of the specific layer (e.g.
LAYERDEPENDS_mylayer
).
When used inside the layer.conf
configuration
file, this variable provides the path of the current layer.
This variable is not available outside of layer.conf
and references are expanded immediately when parsing of the file completes.
Lists the layers, separated by spaces, recommended for use with this layer.
Optionally, you can specify a specific layer version for a recommendation by adding the version to the end of the layer name. Here is an example:
LAYERRECOMMENDS_mylayer = "anotherlayer (=3)"
In this previous example, version 3 of "anotherlayer" is
compared against
LAYERVERSION_anotherlayer
.
This variable is used in the
conf/layer.conf
file and must be
suffixed with the name of the specific layer (e.g.
LAYERRECOMMENDS_mylayer
).
Lists the versions of the
OpenEmbedded-Core for which
a layer is compatible.
Using the LAYERSERIES_COMPAT
variable
allows the layer maintainer to indicate which combinations
of the layer and OE-Core can be expected to work.
The variable gives the system a way to detect when a layer
has not been tested with new releases of OE-Core (e.g.
the layer is not maintained).
To specify the OE-Core versions for which a layer is
compatible, use this variable in your layer's
conf/layer.conf
configuration file.
For the list, use the Yocto Project
Release Name
(e.g. dunfell).
To specify multiple OE-Core versions for the layer,
use a space-separated list:
LAYERSERIES_COMPAT_layer_root_name
= "dunfell zeus"
LAYERSERIES_COMPAT
is
required by the Yocto Project Compatible version 2
standard.
The OpenEmbedded build system produces a warning if
the variable is not set for any given layer.
See the "Creating Your Own Layer" section in the Yocto Project Development Tasks Manual.
Optionally specifies the version of a layer as a single number.
You can use this within
LAYERDEPENDS
for another layer in order to depend on a specific version
of the layer.
This variable is used in the conf/layer.conf
file
and must be suffixed with the name of the specific layer (e.g.
LAYERVERSION_mylayer
).
The minimal command and arguments used to run the linker.
Specifies the flags to pass to the linker. This variable is exported to an environment variable and thus made visible to the software being built during the compilation step.
Default initialization for LDFLAGS
varies depending on what is being built:
TARGET_LDFLAGS
when building for the target
BUILD_LDFLAGS
when building for the build host (i.e.
-native
)
BUILDSDK_LDFLAGS
when building for an SDK (i.e.
nativesdk-
)
Specifies the lead (or primary) compiled library file
(i.e. .so
) that the
debian
class applies its naming policy to given a recipe that
packages multiple libraries.
This variable works in conjunction with the
debian
class.
Checksums of the license text in the recipe source code.
This variable tracks changes in license text of the source code files. If the license text is changed, it will trigger a build failure, which gives the developer an opportunity to review any license change.
This variable must be defined for all recipes (unless
LICENSE
is set to "CLOSED").
For more information, see the "Tracking License Changes" section in the Yocto Project Development Tasks Manual.
The list of source licenses for the recipe. Follow these rules:
Do not use spaces within individual license names.
Separate license names using | (pipe) when there is a choice between licenses.
Separate license names using & (ampersand) when multiple licenses exist that cover different parts of the source.
You can use spaces between license names.
For standard licenses, use the names
of the files in
meta/files/common-licenses/
or the
SPDXLICENSEMAP
flag names defined in
meta/conf/licenses.conf
.
Here are some examples:
LICENSE = "LGPLv2.1 | GPLv3" LICENSE = "MPL-1 & LGPLv2.1" LICENSE = "GPLv2+"
The first example is from the recipes for Qt, which the user
may choose to distribute under either the LGPL version
2.1 or GPL version 3.
The second example is from Cairo where two licenses cover
different parts of the source code.
The final example is from sysstat
,
which presents a single license.
You can also specify licenses on a per-package basis to handle situations where components of the output have different licenses. For example, a piece of software whose code is licensed under GPLv2 but has accompanying documentation licensed under the GNU Free Documentation License 1.2 could be specified as follows:
LICENSE = "GFDL-1.2 & GPLv2" LICENSE_${PN} = "GPLv2" LICENSE_${PN}-doc = "GFDL-1.2"
Setting LICENSE_CREATE_PACKAGE
to "1" causes the OpenEmbedded build system to create
an extra package (i.e.
${
PN
}-lic
)
for each recipe and to add those packages to the
RRECOMMENDS
_${PN}
.
The ${PN}-lic
package installs a
directory in /usr/share/licenses
named ${PN}
, which is the recipe's
base name, and installs files in that directory that
contain license and copyright information (i.e. copies of
the appropriate license files from
meta/common-licenses
that match the
licenses specified in the
LICENSE
variable of the recipe metadata and copies of files marked
in
LIC_FILES_CHKSUM
as containing license text).
For related information on providing license text, see the
COPY_LIC_DIRS
variable, the
COPY_LIC_MANIFEST
variable, and the
"Providing License Text"
section in the Yocto Project Development Tasks Manual.
Specifies additional flags for a recipe you must
whitelist through
LICENSE_FLAGS_WHITELIST
in order to allow the recipe to be built.
When providing multiple flags, separate them with
spaces.
This value is independent of
LICENSE
and is typically used to mark recipes that might
require additional licenses in order to be used in a
commercial product.
For more information, see the
"Enabling Commercially Licensed Recipes"
section in the Yocto Project Development Tasks Manual.
Lists license flags that when specified in
LICENSE_FLAGS
within a recipe should not prevent that recipe from being
built.
This practice is otherwise known as "whitelisting"
license flags.
For more information, see the
"Enabling Commercially Licensed Recipes"
section in the Yocto Project Development Tasks Manual.
Path to additional licenses used during the build.
By default, the OpenEmbedded build system uses COMMON_LICENSE_DIR
to define the directory that holds common license text used during the build.
The LICENSE_PATH
variable allows you to extend that
location to other areas that have additional licenses:
LICENSE_PATH += "path-to-additional-common-licenses
"
Defines the kernel type to be used in assembling the configuration. The linux-yocto recipes define "standard", "tiny", and "preempt-rt" kernel types. See the "Kernel Types" section in the Yocto Project Linux Kernel Development Manual for more information on kernel types.
If you do not specify a
LINUX_KERNEL_TYPE
, it defaults to
"standard".
Together with
KMACHINE
,
the LINUX_KERNEL_TYPE
variable
defines the search
arguments used by the kernel tools to find the appropriate
description within the kernel
Metadata
with which to build out the sources and configuration.
The Linux version from kernel.org
on which the Linux kernel image being built using the
OpenEmbedded build system is based.
You define this variable in the kernel recipe.
For example, the linux-yocto-3.4.bb
kernel recipe found in
meta/recipes-kernel/linux
defines the variables as follows:
LINUX_VERSION ?= "3.4.24"
The LINUX_VERSION
variable is used to
define PV
for the recipe:
PV = "${LINUX_VERSION}+git${SRCPV}"
A string extension compiled into the version string of the Linux kernel built with the OpenEmbedded build system. You define this variable in the kernel recipe. For example, the linux-yocto kernel recipes all define the variable as follows:
LINUX_VERSION_EXTENSION ?= "-yocto-${LINUX_KERNEL_TYPE}"
Defining this variable essentially sets the
Linux kernel configuration item
CONFIG_LOCALVERSION
, which is visible
through the uname
command.
Here is an example that shows the extension assuming it
was set as previously shown:
$ uname -r 3.7.0-rc8-custom
Specifies the directory to which the OpenEmbedded build
system writes overall log files.
The default directory is ${TMPDIR}/log
.
For the directory containing logs specific to each task,
see the T
variable.
Specifies the target device for which the image is built.
You define MACHINE
in the
local.conf
file found in the
Build Directory.
By default, MACHINE
is set to
"qemux86", which is an x86-based architecture machine to
be emulated using QEMU:
MACHINE ?= "qemux86"
The variable corresponds to a machine configuration file of the
same name, through which machine-specific configurations are set.
Thus, when MACHINE
is set to "qemux86" there
exists the corresponding qemux86.conf
machine
configuration file, which can be found in the
Source Directory
in meta/conf/machine
.
The list of machines supported by the Yocto Project as shipped include the following:
MACHINE ?= "qemuarm" MACHINE ?= "qemuarm64" MACHINE ?= "qemumips" MACHINE ?= "qemumips64" MACHINE ?= "qemuppc" MACHINE ?= "qemux86" MACHINE ?= "qemux86-64" MACHINE ?= "genericx86" MACHINE ?= "genericx86-64" MACHINE ?= "beaglebone" MACHINE ?= "edgerouter"
The last five are Yocto Project reference hardware boards, which
are provided in the meta-yocto-bsp
layer.
MACHINE
.
Specifies the name of the machine-specific architecture.
This variable is set automatically from
MACHINE
or
TUNE_PKGARCH
.
You should not hand-edit the
MACHINE_ARCH
variable.
A list of required machine-specific packages to install as part of
the image being built.
The build process depends on these packages being present.
Furthermore, because this is a "machine-essential" variable, the list of
packages are essential for the machine to boot.
The impact of this variable affects images based on
packagegroup-core-boot
,
including the core-image-minimal
image.
This variable is similar to the
MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS
variable with the exception that the image being built has a build
dependency on the variable's list of packages.
In other words, the image will not build if a file in this list is not found.
As an example, suppose the machine for which you are building requires
example-init
to be run during boot to initialize the hardware.
In this case, you would use the following in the machine's
.conf
configuration file:
MACHINE_ESSENTIAL_EXTRA_RDEPENDS += "example-init"
A list of recommended machine-specific packages to install as part of
the image being built.
The build process does not depend on these packages being present.
However, because this is a "machine-essential" variable, the list of
packages are essential for the machine to boot.
The impact of this variable affects images based on
packagegroup-core-boot
,
including the core-image-minimal
image.
This variable is similar to the
MACHINE_ESSENTIAL_EXTRA_RDEPENDS
variable with the exception that the image being built does not have a build
dependency on the variable's list of packages.
In other words, the image will still build if a package in this list is not found.
Typically, this variable is used to handle essential kernel modules, whose
functionality may be selected to be built into the kernel rather than as a module,
in which case a package will not be produced.
Consider an example where you have a custom kernel where a specific touchscreen
driver is required for the machine to be usable.
However, the driver can be built as a module or
into the kernel depending on the kernel configuration.
If the driver is built as a module, you want it to be installed.
But, when the driver is built into the kernel, you still want the
build to succeed.
This variable sets up a "recommends" relationship so that in the latter case,
the build will not fail due to the missing package.
To accomplish this, assuming the package for the module was called
kernel-module-ab123
, you would use the
following in the machine's .conf
configuration
file:
MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS += "kernel-module-ab123"
kernel-module-ab123
recipe
needs to explicitly set its
PACKAGES
variable to ensure that BitBake does not use the
kernel recipe's
PACKAGES_DYNAMIC
variable to satisfy the dependency.
Some examples of these machine essentials are flash, screen, keyboard, mouse, or touchscreen drivers (depending on the machine).
A list of machine-specific packages to install as part of the image being built that are not essential for the machine to boot. However, the build process for more fully-featured images depends on the packages being present.
This variable affects all images based on
packagegroup-base
, which does not include the
core-image-minimal
or core-image-full-cmdline
images.
The variable is similar to the
MACHINE_EXTRA_RRECOMMENDS
variable with the exception that the image being built has a build
dependency on the variable's list of packages.
In other words, the image will not build if a file in this list is not found.
An example is a machine that has WiFi capability but is not
essential for the machine to boot the image.
However, if you are building a more fully-featured image, you want to enable
the WiFi.
The package containing the firmware for the WiFi hardware is always
expected to exist, so it is acceptable for the build process to depend upon
finding the package.
In this case, assuming the package for the firmware was called
wifidriver-firmware
, you would use the following in the
.conf
file for the machine:
MACHINE_EXTRA_RDEPENDS += "wifidriver-firmware"
A list of machine-specific packages to install as part of the image being built that are not essential for booting the machine. The image being built has no build dependency on this list of packages.
This variable affects only images based on
packagegroup-base
, which does not include the
core-image-minimal
or core-image-full-cmdline
images.
This variable is similar to the
MACHINE_EXTRA_RDEPENDS
variable with the exception that the image being built does not have a build
dependency on the variable's list of packages.
In other words, the image will build if a file in this list is not found.
An example is a machine that has WiFi capability but is not essential
For the machine to boot the image.
However, if you are building a more fully-featured image, you want to enable
WiFi.
In this case, the package containing the WiFi kernel module will not be produced
if the WiFi driver is built into the kernel, in which case you still want the
build to succeed instead of failing as a result of the package not being found.
To accomplish this, assuming the package for the module was called
kernel-module-examplewifi
, you would use the
following in the .conf
file for the machine:
MACHINE_EXTRA_RRECOMMENDS += "kernel-module-examplewifi"
Specifies the list of hardware features the
MACHINE
is capable
of supporting.
For related information on enabling features, see the
DISTRO_FEATURES
,
COMBINED_FEATURES
,
and
IMAGE_FEATURES
variables.
For a list of hardware features supported by the Yocto Project as shipped, see the "Machine Features" section.
Features to be added to
MACHINE_FEATURES
if not also present in
MACHINE_FEATURES_BACKFILL_CONSIDERED
.
This variable is set in the meta/conf/bitbake.conf
file.
It is not intended to be user-configurable.
It is best to just reference the variable to see which machine features are
being backfilled for all machine configurations.
See the "Feature Backfilling" section for
more information.
Features from
MACHINE_FEATURES_BACKFILL
that should not be backfilled (i.e. added to
MACHINE_FEATURES
)
during the build.
See the "Feature Backfilling" section for
more information.
A colon-separated list of overrides that apply to the
current machine.
By default, this list includes the value of
MACHINE
.
You can extend MACHINEOVERRIDES
to add extra overrides that should apply to a machine.
For example, all machines emulated in QEMU (e.g.
qemuarm
, qemux86
,
and so forth) include a file named
meta/conf/machine/include/qemu.inc
that prepends the following override to
MACHINEOVERRIDES
:
MACHINEOVERRIDES =. "qemuall:"
This override allows variables to be overriden for all
machines emulated in QEMU, like in the following example
from the connman-conf
recipe:
SRC_URI_append_qemuall = "file://wired.config \ file://wired-setup \ "
The underlying mechanism behind
MACHINEOVERRIDES
is simply that it is
included in the default value of
OVERRIDES
.
The email address of the distribution maintainer.
Specifies additional paths from which the OpenEmbedded
build system gets source code.
When the build system searches for source code, it first
tries the local download directory.
If that location fails, the build system tries locations
defined by
PREMIRRORS
,
the upstream source, and then locations specified by
MIRRORS
in that order.
Assuming your distribution
(DISTRO
)
is "poky", the default value for
MIRRORS
is defined in the
conf/distro/poky.conf
file in the
meta-poky
Git repository.
Specifies a prefix has been added to
PN
to create a special version
of a recipe or package (i.e. a Multilib version).
The variable is used in places where the prefix needs to be
added to or removed from a the name (e.g. the
BPN
variable).
MLPREFIX
gets set when a prefix has been
added to PN
.
MLPREFIX
stands for
"MultiLib".
This representation is historical and comes from
a time when nativesdk
was a suffix
rather than a prefix on the recipe name.
When nativesdk
was turned into a
prefix, it made sense to set
MLPREFIX
for it as well.
To help understand when MLPREFIX
might be needed, consider when
BBCLASSEXTEND
is used to provide a nativesdk
version
of a recipe in addition to the target version.
If that recipe declares build-time dependencies on tasks in
other recipes by using
DEPENDS
,
then a dependency on "foo" will automatically get rewritten
to a dependency on "nativesdk-foo".
However, dependencies like the following will not get
rewritten automatically:
do_foo[depends] += "recipe
:do_foo"
If you want such a dependency to also get transformed, you can do the following:
do_foo[depends] += "${MLPREFIX}recipe
:do_foo"
This variable has been replaced by the
KERNEL_MODULE_AUTOLOAD
variable.
You should replace all occurrences of
module_autoload
with additions to
KERNEL_MODULE_AUTOLOAD
, for example:
module_autoload_rfcomm = "rfcomm"
should now be replaced with:
KERNEL_MODULE_AUTOLOAD += "rfcomm"
See the
KERNEL_MODULE_AUTOLOAD
variable for more information.
Specifies
modprobe.d
syntax lines for inclusion in the
/etc/modprobe.d/modname.conf
file.
You can use this variable anywhere that it can be
recognized by the kernel recipe or out-of-tree kernel
module recipe (e.g. a machine configuration file, a
distribution configuration file, an append file for the
recipe, or the recipe itself).
If you use this variable, you must also be sure to list
the module name in the
KERNEL_MODULE_AUTOLOAD
variable.
Here is the general syntax:
module_conf_module_name
= "modprobe.d-syntax
"
You must use the kernel module name override.
Run man modprobe.d
in the shell to
find out more information on the exact syntax
you want to provide with module_conf
.
Including module_conf
causes the
OpenEmbedded build system to populate the
/etc/modprobe.d/modname.conf
file with modprobe.d
syntax lines.
Here is an example that adds the options
arg1
and arg2
to a module named mymodule
:
module_conf_mymodule = "options mymodule arg1=val1 arg2=val2"
For information on how to specify kernel modules to
auto-load on boot, see the
KERNEL_MODULE_AUTOLOAD
variable.
Controls creation of the modules-*.tgz
file.
Set this variable to "0" to disable creation of this
file, which contains all of the kernel modules resulting
from a kernel build.
The link name of the kernel module tarball.
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
MODULE_TARBALL_LINK_NAME ?= "${KERNEL_ARTIFACT_LINK_NAME}"
The value of the KERNEL_ARTIFACT_LINK_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_LINK_NAME ?= "${MACHINE}"
See the
MACHINE
variable for additional information.
The base name of the kernel module tarball.
This variable is set in the
meta/classes/kernel-artifact-names.bbclass
file as follows:
MODULE_TARBALL_NAME ?= "${KERNEL_ARTIFACT_NAME}"
The value of the
KERNEL_ARTIFACT_NAME
variable, which is set in the same file, has the following
value:
KERNEL_ARTIFACT_NAME ?= "${PKGE}-${PKGV}-${PKGR}-${MACHINE}${IMAGE_VERSION_SUFFIX}"
Uniquely identifies the type of the target system for which packages are being built. This variable allows output for different types of target systems to be put into different subdirectories of the same output directory.
The default value of this variable is:
${PACKAGE_ARCH}${TARGET_VENDOR}-${TARGET_OS}
Some classes (e.g.
cross-canadian
)
modify the MULTIMACH_TARGET_SYS
value.
See the
STAMP
variable for an example.
See the
STAGING_DIR_TARGET
variable for more information.
A string identifying the host distribution.
Strings consist of the host distributor ID
followed by the release, as reported by the
lsb_release
tool
or as read from /etc/lsb-release
.
For example, when running a build on Ubuntu 12.10, the value
is "Ubuntu-12.10".
If this information is unable to be determined, the value
resolves to "Unknown".
This variable is used by default to isolate native shared
state packages for different distributions (e.g. to avoid
problems with glibc
version
incompatibilities).
Additionally, the variable is checked against
SANITY_TESTED_DISTROS
if that variable is set.
The minimal command and arguments to run
nm
.
Avoids QA errors when you use a non-common, non-CLOSED
license in a recipe.
Packages exist, such as the linux-firmware package, with
many licenses that are not in any way common.
Also, new licenses are added occasionally to avoid
introducing a lot of common license files, which are only
applicable to a specific package.
NO_GENERIC_LICENSE
is used to allow
copying a license that does not exist in common licenses.
The following example shows how to add
NO_GENERIC_LICENSE
to a recipe:
NO_GENERIC_LICENSE[license_name
] = "license_file_in_fetched_source
"
The following is an example that uses the
LICENSE.Abilis.txt
file as the license
from the fetched source:
NO_GENERIC_LICENSE[Firmware-Abilis] = "LICENSE.Abilis.txt"
Prevents installation of all "recommended-only" packages.
Recommended-only packages are packages installed only
through the
RRECOMMENDS
variable).
Setting the NO_RECOMMENDATIONS
variable
to "1" turns this feature on:
NO_RECOMMENDATIONS = "1"
You can set this variable globally in your
local.conf
file or you can attach it to
a specific image recipe by using the recipe name override:
NO_RECOMMENDATIONS_pn-target_image
= "1"
It is important to realize that if you choose to not install
packages using this variable and some other packages are
dependent on them (i.e. listed in a recipe's
RDEPENDS
variable), the OpenEmbedded build system ignores your
request and will install the packages to avoid dependency
errors.
IMAGE_INSTALL
variable.
Support for this variable exists only when using the IPK and RPM packaging backend. Support does not exist for DEB.
See the
BAD_RECOMMENDATIONS
and the
PACKAGE_EXCLUDE
variables for related information.
Disables auto package from splitting
.debug
files. If a recipe requires
FILES_${PN}-dbg
to be set manually,
the NOAUTOPACKAGEDEBUG
can be defined
allowing you to define the content of the debug package.
For example:
NOAUTOPACKAGEDEBUG = "1" FILES_${PN}-dev = "${includedir}/${QT_DIR_NAME}/Qt/*" FILES_${PN}-dbg = "/usr/src/debug/" FILES_${QT_BASE_NAME}-demos-doc = "${docdir}/${QT_DIR_NAME}/qch/qt.qch"
The minimal command and arguments to run
objcopy
.
The minimal command and arguments to run
objdump
.
When inheriting the
binconfig
class, this variable
specifies additional arguments passed to the "sed" command.
The sed command alters any paths in configuration scripts
that have been set up during compilation.
Inheriting this class results in all paths in these scripts
being changed to point into the
sysroots/
directory so that all builds
that use the script will use the correct directories
for the cross compiling layout.
See the meta/classes/binconfig.bbclass
in the
Source Directory
for details on how this class applies these additional
sed command arguments.
For general information on the
binconfig
class, see the
"binconfig.bbclass
"
section.
An internal variable used to tell the OpenEmbedded build system what Python modules to import for every Python function run by the system.
The name of the build environment setup script for the purposes of setting up the environment within the extensible SDK. The default value is "oe-init-build-env".
If you use a custom script to set up your build
environment, set the
OE_INIT_ENV_SCRIPT
variable to its
name.
Controls how the OpenEmbedded build system spawns
interactive terminals on the host development system
(e.g. using the BitBake command with the
-c devshell
command-line option).
For more information, see the
"Using a Development Shell" section
in the Yocto Project Development Tasks Manual.
You can use the following values for the
OE_TERMINAL
variable:
auto gnome xfce rxvt screen konsole none
The directory from which the top-level build environment
setup script is sourced.
The Yocto Project provides a top-level build environment
setup script:
oe-init-build-env
.
When you run this script, the
OEROOT
variable resolves to the
directory that contains the script.
For additional information on how this variable is used, see the initialization script.
Declares the oldest version of the Linux kernel that the
produced binaries must support.
This variable is passed into the build of the Embedded
GNU C Library (glibc
).
The default for this variable comes from the
meta/conf/bitbake.conf
configuration
file.
You can override this default by setting the variable
in a custom distribution configuration file.
A colon-separated list of overrides that currently apply.
Overrides are a BitBake mechanism that allows variables to
be selectively overridden at the end of parsing.
The set of overrides in OVERRIDES
represents the "state" during building, which includes
the current recipe being built, the machine for which
it is being built, and so forth.
As an example, if the string "an-override" appears as an
element in the colon-separated list in
OVERRIDES
, then the following
assignment will override FOO
with the
value "overridden" at the end of parsing:
FOO_an-override = "overridden"
See the "Conditional Syntax (Overrides)" section in the BitBake User Manual for more information on the overrides mechanism.
The default value of OVERRIDES
includes the values of the
CLASSOVERRIDE
,
MACHINEOVERRIDES
,
and
DISTROOVERRIDES
variables.
Another important override included by default is
pn-${PN}
.
This override allows variables to be set for a single
recipe within configuration (.conf
)
files.
Here is an example:
FOO_pn-myrecipe = "myrecipe-specific value"
OVERRIDES
in the output of the
bitbake -e
command.
See the
"Viewing Variable Values"
section in the Yocto Project Development Tasks
Manual for more information.
The recipe name and version.
P
is comprised of the following:
${PN}-${PV}
The architecture of the resulting package or packages.
By default, the value of this variable is set to
TUNE_PKGARCH
when building for the target,
BUILD_ARCH
when building for the
build host, and "${SDK_ARCH}-${SDKPKGSUFFIX}" when building
for the SDK.
SDK_ARCH
for more information.
However, if your recipe's output packages are built
specific to the target machine rather than generally for
the architecture of the machine, you should set
PACKAGE_ARCH
to the value of
MACHINE_ARCH
in the recipe as follows:
PACKAGE_ARCH = "${MACHINE_ARCH}"
Specifies a list of architectures compatible with
the target machine.
This variable is set automatically and should not
normally be hand-edited.
Entries are separated using spaces and listed in order
of priority.
The default value for
PACKAGE_ARCHS
is "all any noarch
${PACKAGE_EXTRA_ARCHS} ${MACHINE_ARCH}".
Enables easily adding packages to
PACKAGES
before ${PN}
so that those added packages can pick up files that would normally be
included in the default package.
This variable, which is set in the
local.conf
configuration file found in
the conf
folder of the
Build Directory,
specifies the package manager the OpenEmbedded build system
uses when packaging data.
You can provide one or more of the following arguments for the variable:
PACKAGE_CLASSES ?= "package_rpm package_deb package_ipk package_tar"
package_tar
class has limited
functionality due to no support for package
dependencies by that backend.
Therefore, it is recommended that you do not use it.
The build system uses only the first argument in the list
as the package manager when creating your image or SDK.
However, packages will be created using any additional
packaging classes you specify.
For example, if you use the following in your
local.conf
file:
PACKAGE_CLASSES ?= "package_ipk"
The OpenEmbedded build system uses the IPK package manager to create your image or SDK.
For information on packaging and build performance effects
as a result of the package manager in use, see the
"package.bbclass
"
section.
Determines how to split up the binary and debug information
when creating *-dbg
packages to be
used with the GNU Project Debugger (GDB).
With the
PACKAGE_DEBUG_SPLIT_STYLE
variable,
you can control where debug information, which can include
or exclude source files, is stored:
".debug": Debug symbol files are placed next
to the binary in a .debug
directory on the target.
For example, if a binary is installed into
/bin
, the corresponding debug
symbol files are installed in
/bin/.debug
.
Source files are placed in
/usr/src/debug
.
"debug-file-directory": Debug symbol files are
placed under /usr/lib/debug
on the target, and separated by the path from where
the binary is installed.
For example, if a binary is installed in
/bin
, the corresponding debug
symbols are installed in
/usr/lib/debug/bin
.
Source files are placed in
/usr/src/debug
.
"debug-without-src": The same behavior as ".debug" previously described with the exception that no source files are installed.
"debug-with-srcpkg": The same behavior as
".debug" previously described with the exception
that all source files are placed in a separate
*-src
pkg.
This is the default behavior.
You can find out more about debugging using GDB by reading the "Debugging With the GNU Project Debugger (GDB) Remotely" section in the Yocto Project Development Tasks Manual.
Prevents specific packages from being installed when you are installing complementary packages.
You might find that you want to prevent installing certain
packages when you are installing complementary packages.
For example, if you are using
IMAGE_FEATURES
to install dev-pkgs
, you might not want
to install all packages from a particular multilib.
If you find yourself in this situation, you can use the
PACKAGE_EXCLUDE_COMPLEMENTARY
variable
to specify regular expressions to match the packages you
want to exclude.
Lists packages that should not be installed into an image. For example:
PACKAGE_EXCLUDE = "package_name
package_name
package_name
..."
You can set this variable globally in your
local.conf
file or you can attach it to
a specific image recipe by using the recipe name override:
PACKAGE_EXCLUDE_pn-target_image
= "package_name
"
If you choose to not install
a package using this variable and some other package is
dependent on it (i.e. listed in a recipe's
RDEPENDS
variable), the OpenEmbedded build system generates a fatal
installation error.
Because the build system halts the process with a fatal
error, you can use the variable with an iterative
development process to remove specific components from a
system.
Support for this variable exists only when using the IPK and RPM packaging backend. Support does not exist for DEB.
See the
NO_RECOMMENDATIONS
and the
BAD_RECOMMENDATIONS
variables for related information.
Specifies the list of architectures compatible with the device CPU. This variable is useful when you build for several different devices that use miscellaneous processors such as XScale and ARM926-EJS.
Optionally specifies the package architectures used as
part of the package feed URIs during the build.
When used, the PACKAGE_FEED_ARCHS
variable is appended to the final package feed URI, which
is constructed using the
PACKAGE_FEED_URIS
and
PACKAGE_FEED_BASE_PATHS
variables.
PACKAGE_FEEDS_ARCHS
variable to whitelist specific package architectures.
If you do not need to whitelist specific architectures,
which is a common case, you can omit this variable.
Omitting the variable results in all available
architectures for the current machine being included
into remote package feeds.
Consider the following example where the
PACKAGE_FEED_URIS
,
PACKAGE_FEED_BASE_PATHS
, and
PACKAGE_FEED_ARCHS
variables are
defined in your local.conf
file:
PACKAGE_FEED_URIS = "https://example.com/packagerepos/release \ https://example.com/packagerepos/updates" PACKAGE_FEED_BASE_PATHS = "rpm rpm-dev" PACKAGE_FEED_ARCHS = "all core2-64"
Given these settings, the resulting package feeds are as follows:
https://example.com/packagerepos/release/rpm/all https://example.com/packagerepos/release/rpm/core2-64 https://example.com/packagerepos/release/rpm-dev/all https://example.com/packagerepos/release/rpm-dev/core2-64 https://example.com/packagerepos/updates/rpm/all https://example.com/packagerepos/updates/rpm/core2-64 https://example.com/packagerepos/updates/rpm-dev/all https://example.com/packagerepos/updates/rpm-dev/core2-64
Specifies the base path used when constructing package feed
URIs.
The PACKAGE_FEED_BASE_PATHS
variable
makes up the middle portion of a package feed URI used
by the OpenEmbedded build system.
The base path lies between the
PACKAGE_FEED_URIS
and
PACKAGE_FEED_ARCHS
variables.
Consider the following example where the
PACKAGE_FEED_URIS
,
PACKAGE_FEED_BASE_PATHS
, and
PACKAGE_FEED_ARCHS
variables are
defined in your local.conf
file:
PACKAGE_FEED_URIS = "https://example.com/packagerepos/release \ https://example.com/packagerepos/updates" PACKAGE_FEED_BASE_PATHS = "rpm rpm-dev" PACKAGE_FEED_ARCHS = "all core2-64"
Given these settings, the resulting package feeds are as follows:
https://example.com/packagerepos/release/rpm/all https://example.com/packagerepos/release/rpm/core2-64 https://example.com/packagerepos/release/rpm-dev/all https://example.com/packagerepos/release/rpm-dev/core2-64 https://example.com/packagerepos/updates/rpm/all https://example.com/packagerepos/updates/rpm/core2-64 https://example.com/packagerepos/updates/rpm-dev/all https://example.com/packagerepos/updates/rpm-dev/core2-64
Specifies the front portion of the package feed URI
used by the OpenEmbedded build system.
Each final package feed URI is comprised of
PACKAGE_FEED_URIS
,
PACKAGE_FEED_BASE_PATHS
,
and
PACKAGE_FEED_ARCHS
variables.
Consider the following example where the
PACKAGE_FEED_URIS
,
PACKAGE_FEED_BASE_PATHS
, and
PACKAGE_FEED_ARCHS
variables are
defined in your local.conf
file:
PACKAGE_FEED_URIS = "https://example.com/packagerepos/release \ https://example.com/packagerepos/updates" PACKAGE_FEED_BASE_PATHS = "rpm rpm-dev" PACKAGE_FEED_ARCHS = "all core2-64"
Given these settings, the resulting package feeds are as follows:
https://example.com/packagerepos/release/rpm/all https://example.com/packagerepos/release/rpm/core2-64 https://example.com/packagerepos/release/rpm-dev/all https://example.com/packagerepos/release/rpm-dev/core2-64 https://example.com/packagerepos/updates/rpm/all https://example.com/packagerepos/updates/rpm/core2-64 https://example.com/packagerepos/updates/rpm-dev/all https://example.com/packagerepos/updates/rpm-dev/core2-64
The final list of packages passed to the package manager for installation into the image.
Because the package manager controls actual installation
of all packages, the list of packages passed using
PACKAGE_INSTALL
is not the final list
of packages that are actually installed.
This variable is internal to the image construction
code.
Consequently, in general, you should use the
IMAGE_INSTALL
variable to specify packages for installation.
The exception to this is when working with
the
core-image-minimal-initramfs
image.
When working with an initial RAM filesystem (initramfs)
image, use the PACKAGE_INSTALL
variable.
For information on creating an initramfs, see the
"Building an Initial RAM Filesystem (initramfs) Image"
section in the Yocto Project Development Tasks Manual.
Specifies a list of packages the OpenEmbedded build system attempts to install when creating an image. If a listed package fails to install, the build system does not generate an error. This variable is generally not user-defined.
Specifies a list of functions run to pre-process the
PKGD
directory prior to splitting the files out to individual
packages.
Specifies a list of dependencies for post-installation and
pre-installation scripts on native/cross tools.
If your post-installation or pre-installation script can
execute at rootfs creation time rather than on the
target but depends on a native tool in order to execute,
you need to list the tools in
PACKAGE_WRITE_DEPS
.
For information on running post-installation scripts, see the "Post-Installation Scripts" section in the Yocto Project Development Tasks Manual.
This variable provides a means of enabling or disabling
features of a recipe on a per-recipe basis.
PACKAGECONFIG
blocks are defined
in recipes when you specify features and then arguments
that define feature behaviors.
Here is the basic block structure (broken over multiple
lines for readability):
PACKAGECONFIG ??= "f1 f2 f3 ..." PACKAGECONFIG[f1] = "\ --with-f1, \ --without-f1, \ build-deps-for-f1, \ runtime-deps-for-f1, \ runtime-recommends-for-f1, \ packageconfig-conflicts-for-f1 \ " PACKAGECONFIG[f2] = "\ ... and so on and so on ...
The PACKAGECONFIG
variable itself specifies a space-separated list of the
features to enable.
Following the features, you can determine the behavior of
each feature by providing up to six order-dependent
arguments, which are separated by commas.
You can omit any argument you like but must retain the
separating commas.
The order is important and specifies the following:
Extra arguments
that should be added to the configure script
argument list
(EXTRA_OECONF
or
PACKAGECONFIG_CONFARGS
)
if the feature is enabled.
Extra arguments
that should be added to EXTRA_OECONF
or PACKAGECONFIG_CONFARGS
if the feature is disabled.
Additional build dependencies
(DEPENDS
)
that should be added if the feature is enabled.
Additional runtime dependencies
(RDEPENDS
)
that should be added if the feature is enabled.
Additional runtime recommendations
(RRECOMMENDS
)
that should be added if the feature is enabled.
Any conflicting (that is, mutually
exclusive) PACKAGECONFIG
settings for this feature.
Consider the following
PACKAGECONFIG
block taken from the
librsvg
recipe.
In this example the feature is gtk
,
which has three arguments that determine the feature's
behavior.
PACKAGECONFIG[gtk] = "--with-gtk3,--without-gtk3,gtk+3"
The --with-gtk3
and
gtk+3
arguments apply only if
the feature is enabled.
In this case, --with-gtk3
is
added to the configure script argument list and
gtk+3
is added to
DEPENDS
.
On the other hand, if the feature is disabled say through
a .bbappend
file in another layer, then
the second argument --without-gtk3
is
added to the configure script instead.
The basic PACKAGECONFIG
structure
previously described holds true regardless of whether you
are creating a block or changing a block.
When creating a block, use the structure inside your
recipe.
If you want to change an existing
PACKAGECONFIG
block, you can do so
one of two ways:
Append file:
Create an append file named
recipename
.bbappend
in your layer and override the value of
PACKAGECONFIG
.
You can either completely override the variable:
PACKAGECONFIG = "f4 f5"
Or, you can just append the variable:
PACKAGECONFIG_append = " f4"
Configuration file:
This method is identical to changing the block
through an append file except you edit your
local.conf
or
file.
As with append files previously described,
you can either completely override the variable:
mydistro
.conf
PACKAGECONFIG_pn-recipename
= "f4 f5"
Or, you can just amend the variable:
PACKAGECONFIG_append_pn-recipename
= " f4"
A space-separated list of configuration options generated
from the
PACKAGECONFIG
setting.
Classes such as
autotools
and
cmake
use PACKAGECONFIG_CONFARGS
to pass
PACKAGECONFIG
options to
configure
and
cmake
, respectively.
If you are using
PACKAGECONFIG
but not a class that
handles the do_configure
task, then
you need to use
PACKAGECONFIG_CONFARGS
appropriately.
For recipes inheriting the
packagegroup
class, setting
PACKAGEGROUP_DISABLE_COMPLEMENTARY
to
"1" specifies that the normal complementary packages
(i.e. -dev
,
-dbg
, and so forth) should not be
automatically created by the
packagegroup
recipe, which is the
default behavior.
The list of packages the recipe creates. The default value is the following:
${PN}-dbg ${PN}-staticdev ${PN}-dev ${PN}-doc ${PN}-locale ${PACKAGE_BEFORE_PN} ${PN}
During packaging, the
do_package
task goes through PACKAGES
and uses
the
FILES
variable corresponding to each package to assign files to
the package.
If a file matches the FILES
variable
for more than one package in PACKAGES
,
it will be assigned to the earliest (leftmost) package.
Packages in the variable's list that are empty (i.e. where
none of the patterns in
FILES_
pkg
match any files installed by the
do_install
task) are not generated, unless generation is forced through
the
ALLOW_EMPTY
variable.
A promise that your recipe satisfies runtime dependencies
for optional modules that are found in other recipes.
PACKAGES_DYNAMIC
does not actually satisfy the dependencies, it only states that
they should be satisfied.
For example, if a hard, runtime dependency
(RDEPENDS
)
of another package is satisfied
at build time through the PACKAGES_DYNAMIC
variable, but a package with the module name is never actually
produced, then the other package will be broken.
Thus, if you attempt to include that package in an image,
you will get a dependency failure from the packaging system
during the
do_rootfs
task.
Typically, if there is a chance that such a situation can
occur and the package that is not created is valid
without the dependency being satisfied, then you should use
RRECOMMENDS
(a soft runtime dependency) instead of
RDEPENDS
.
For an example of how to use the PACKAGES_DYNAMIC
variable when you are splitting packages, see the
"Handling Optional Module Packaging" section
in the Yocto Project Development Tasks Manual.
Specifies a list of functions run to perform additional
splitting of files into individual packages.
Recipes can either prepend to this variable or prepend
to the populate_packages
function
in order to perform additional package splitting.
In either case, the function should set
PACKAGES
,
FILES
,
RDEPENDS
and other packaging variables appropriately in order to
perform the desired splitting.
Extra options passed to the make
command during the
do_compile
task in order to specify parallel compilation on the local
build host.
This variable is usually in the form "-j x
",
where x
represents the maximum
number of parallel threads make
can
run.
PARALLEL_MAKE
to be
effective, make
must be called
with
${
EXTRA_OEMAKE
}
.
An easy way to ensure this is to use the
oe_runmake
function.
By default, the OpenEmbedded build system automatically sets this variable to be equal to the number of cores the build system uses.
do_compile
task that result in race conditions, you can clear
the PARALLEL_MAKE
variable within
the recipe as a workaround.
For information on addressing race conditions, see the
"Debugging Parallel Make Races"
section in the Yocto Project Development Tasks Manual.
For single socket systems (i.e. one CPU), you should not
have to override this variable to gain optimal parallelism
during builds.
However, if you have very large systems that employ
multiple physical CPUs, you might want to make sure the
PARALLEL_MAKE
variable is not
set higher than "-j 20".
For more information on speeding up builds, see the "Speeding Up a Build" section in the Yocto Project Development Tasks Manual.
Extra options passed to the
make install
command during the
do_install
task in order to specify parallel installation.
This variable defaults to the value of
PARALLEL_MAKE
.
In order for PARALLEL_MAKEINST
to be
effective, make
must be called
with
${
EXTRA_OEMAKE
}
.
An easy way to ensure this is to use the
oe_runmake
function.
If the software being built experiences
dependency issues during the
do_install
task that result in
race conditions, you can clear the
PARALLEL_MAKEINST
variable within
the recipe as a workaround.
For information on addressing race conditions, see the
"Debugging Parallel Make Races"
section in the Yocto Project Development Tasks Manual.
Determines the action to take when a patch fails. You can set this variable to one of two values: "noop" and "user".
The default value of "noop" causes the build to simply fail when the OpenEmbedded build system cannot successfully apply a patch. Setting the value to "user" causes the build system to launch a shell and places you in the right location so that you can manually resolve the conflicts.
Set this variable in your
local.conf
file.
Specifies the utility used to apply patches for a recipe
during the
do_patch
task.
You can specify one of three utilities: "patch", "quilt", or
"git".
The default utility used is "quilt" except for the
quilt-native recipe itself.
Because the quilt tool is not available at the
time quilt-native is being patched, it uses "patch".
If you wish to use an alternative patching tool, set the variable in the recipe using one of the following:
PATCHTOOL = "patch" PATCHTOOL = "quilt" PATCHTOOL = "git"
The epoch of the recipe. By default, this variable is unset. The variable is used to make upgrades possible when the versioning scheme changes in some backwards incompatible way.
PE
is the default value of the
PKGE
variable.
Specifies the recipe or package name and includes all version and revision
numbers (i.e. glibc-2.13-r20+svnr15508/
and
bash-4.2-r1/
).
This variable is comprised of the following:
${PN}-${EXTENDPE}${PV}-${PR}
When inheriting the
pixbufcache
class, this variable identifies packages that contain
the pixbuf loaders used with
gdk-pixbuf
.
By default, the pixbufcache
class
assumes that the loaders are in the recipe's main package
(i.e. ${
PN
}
).
Use this variable if the loaders you need are in a package
other than that main package.
The name of the resulting package created by the OpenEmbedded build system.
PKG
variable, you
must use a package name override.
For example, when the
debian
class renames the output package, it does so by setting
PKG_
.
packagename
The path to pkg-config
files for the
current build context.
pkg-config
reads this variable
from the environment.
Points to the destination directory for files to be packaged before they are split into individual packages. This directory defaults to the following:
${WORKDIR}/package
Do not change this default.
Points to a shared, global-state directory that holds data
generated during the packaging process.
During the packaging process, the
do_packagedata
task packages data for each recipe and installs it into
this temporary, shared area.
This directory defaults to the following, which you should
not change:
${STAGING_DIR_HOST}/pkgdata
For examples of how this data is used, see the
"Automatically Added Runtime Dependencies"
section in the Yocto Project Overview and Concepts Manual
and the
"Viewing Package Information with oe-pkgdata-util
"
section in the Yocto Project Development Tasks Manual.
For more information on the shared, global-state directory,
see
STAGING_DIR_HOST
.
Points to the parent directory for files to be packaged after they have been split into individual packages. This directory defaults to the following:
${WORKDIR}/packages-split
Under this directory, the build system creates
directories for each package specified in
PACKAGES
.
Do not change this default.
Points to a temporary work area where the
do_package
task saves package metadata.
The PKGDESTWORK
location defaults to
the following:
${WORKDIR}/pkgdata
Do not change this default.
The
do_packagedata
task copies the package metadata from
PKGDESTWORK
to
PKGDATA_DIR
to make it available globally.
The epoch of the package(s) built by the recipe.
By default, PKGE
is set to
PE
.
The revision of the package(s) built by the recipe.
By default, PKGR
is set to
PR
.
The version of the package(s) built by the
recipe.
By default, PKGV
is set to
PV
.
This variable can have two separate functions depending on the context: a recipe name or a resulting package name.
PN
refers to a recipe name in the context of a file used
by the OpenEmbedded build system as input to create a package.
The name is normally extracted from the recipe file name.
For example, if the recipe is named
expat_2.0.1.bb
, then the default value of PN
will be "expat".
The variable refers to a package name in the context of a file created or produced by the OpenEmbedded build system.
If applicable, the PN
variable also contains any special
suffix or prefix.
For example, using bash
to build packages for the native
machine, PN
is bash-native
.
Using bash
to build packages for the target and for Multilib,
PN
would be bash
and
lib64-bash
, respectively.
Lists recipes you do not want the OpenEmbedded build system
to build.
This variable works in conjunction with the
blacklist
class, which is inherited globally.
To prevent a recipe from being built, use the
PNBLACKLIST
variable in your
local.conf
file.
Here is an example that prevents
myrecipe
from being built:
PNBLACKLIST[myrecipe] = "Not supported by our organization."
Specifies a list of functions to call once the OpenEmbedded build system has created the host part of the SDK. You can specify functions separated by semicolons:
POPULATE_SDK_POST_HOST_COMMAND += "function
; ... "
If you need to pass the SDK path to a command
within a function, you can use
${SDK_DIR}
, which points to
the parent directory used by the OpenEmbedded build
system when creating SDK output.
See the
SDK_DIR
variable for more information.
Specifies a list of functions to call once the OpenEmbedded build system has created the target part of the SDK. You can specify functions separated by semicolons:
POPULATE_SDK_POST_TARGET_COMMAND += "function
; ... "
If you need to pass the SDK path to a command
within a function, you can use
${SDK_DIR}
, which points to
the parent directory used by the OpenEmbedded build
system when creating SDK output.
See the
SDK_DIR
variable for more information.
The revision of the recipe. The default value for this
variable is "r0".
Subsequent revisions of the recipe conventionally have the
values "r1", "r2", and so forth.
When
PV
increases, PR
is conventionally reset
to "r0".
PR
to know when to rebuild a
recipe.
The build system uses the task
input checksums
along with the
stamp
and
shared state cache
mechanisms.
The PR
variable primarily becomes
significant when a package manager dynamically installs
packages on an already built image.
In this case, PR
, which is the default
value of
PKGR
,
helps the package manager distinguish which package is the
most recent one in cases where many packages have the same
PV
(i.e. PKGV
).
A component having many packages with the same
PV
usually means that the packages all
install the same upstream version, but with later
(PR
) version packages including
packaging fixes.
PR
does not need to be increased
for changes that do not change the package contents or
metadata.
Because manually managing PR
can be
cumbersome and error-prone, an automated solution exists.
See the
"Working With a PR Service"
section in the Yocto Project Development Tasks Manual
for more information.
If multiple recipes provide the same item, this variable
determines which recipe is preferred and thus provides
the item (i.e. the preferred provider).
You should always suffix this variable with the name of the
provided item.
And, you should define the variable using the preferred
recipe's name
(PN
).
Here is a common example:
PREFERRED_PROVIDER_virtual/kernel ?= "linux-yocto"
In the previous example, multiple recipes are providing
"virtual/kernel".
The PREFERRED_PROVIDER
variable is
set with the name (PN
) of the recipe
you prefer to provide "virtual/kernel".
Following are more examples:
PREFERRED_PROVIDER_virtual/xserver = "xserver-xf86" PREFERRED_PROVIDER_virtual/libgl ?= "mesa"
For more information, see the "Using Virtual Providers" section in the Yocto Project Development Tasks Manual.
virtual/*
item
with PREFERRED_PROVIDER
, then any
recipe that
PROVIDES
that item but is not selected (defined) by
PREFERRED_PROVIDER
is prevented
from building, which is usually desirable since this
mechanism is designed to select between mutually
exclusive alternative providers.
If multiple versions of recipes exist, this
variable determines which version is given preference.
You must always suffix the variable with the
PN
you want to select, and you should set the
PV
accordingly for precedence.
The PREFERRED_VERSION
variable
supports limited wildcard use through the
"%
" character.
You can use the character to match any number of
characters, which can be useful when specifying versions
that contain long revision numbers that potentially change.
Here are two examples:
PREFERRED_VERSION_python = "3.4.0" PREFERRED_VERSION_linux-yocto = "5.0%"
%
" character
is limited in that it only works at the end of the
string.
You cannot use the wildcard character in any other
location of the string.
The specified version is matched against
PV
,
which does not necessarily match the version part of
the recipe's filename.
For example, consider two recipes
foo_1.2.bb
and
foo_git.bb
where
foo_git.bb
contains the following
assignment:
PV = "1.1+git${SRCPV}"
In this case, the correct way to select
foo_git.bb
is by using an
assignment such as the following:
PREFERRED_VERSION_foo = "1.1+git%"
Compare that previous example against the following incorrect example, which does not work:
PREFERRED_VERSION_foo = "git"
Sometimes the PREFERRED_VERSION
variable can be set by configuration files in a way that
is hard to change.
You can use
OVERRIDES
to set a machine-specific override.
Here is an example:
PREFERRED_VERSION_linux-yocto_qemux86 = "5.0%"
Although not recommended, worst case, you can also use the "forcevariable" override, which is the strongest override possible. Here is an example:
PREFERRED_VERSION_linux-yocto_forcevariable = "5.0%"
_forcevariable
override is
not handled specially.
This override only works because the default value of
OVERRIDES
includes
"forcevariable".
Specifies additional paths from which the OpenEmbedded
build system gets source code.
When the build system searches for source code, it first
tries the local download directory.
If that location fails, the build system tries locations
defined by PREMIRRORS
, the upstream
source, and then locations specified by
MIRRORS
in that order.
Assuming your distribution
(DISTRO
)
is "poky", the default value for
PREMIRRORS
is defined in the
conf/distro/poky.conf
file in the
meta-poky
Git repository.
Typically, you could add a specific server for the
build system to attempt before any others by adding
something like the following to the
local.conf
configuration file in the
Build Directory:
PREMIRRORS_prepend = "\ git://.*/.* http://www.yoctoproject.org/sources/ \n \ ftp://.*/.* http://www.yoctoproject.org/sources/ \n \ http://.*/.* http://www.yoctoproject.org/sources/ \n \ https://.*/.* http://www.yoctoproject.org/sources/ \n"
These changes cause the build system to intercept
Git, FTP, HTTP, and HTTPS requests and direct them to
the http://
sources mirror.
You can use file://
URLs to point
to local directories or network shares as well.
Indicates the importance of a package.
PRIORITY
is considered to be part of
the distribution policy because the importance of any given
recipe depends on the purpose for which the distribution
is being produced.
Thus, PRIORITY
is not normally set
within recipes.
You can set PRIORITY
to "required",
"standard", "extra", and "optional", which is the default.
Specifies libraries installed within a recipe that should be ignored by the OpenEmbedded build system's shared library resolver. This variable is typically used when software being built by a recipe has its own private versions of a library normally provided by another recipe. In this case, you would not want the package containing the private libraries to be set as a dependency on other unrelated packages that should instead depend on the package providing the standard version of the library.
Libraries specified in this variable should be specified by their file name. For example, from the Firefox recipe in meta-browser:
PRIVATE_LIBS = "libmozjs.so \ libxpcom.so \ libnspr4.so \ libxul.so \ libmozalloc.so \ libplc4.so \ libplds4.so"
For more information, see the "Automatically Added Runtime Dependencies" section in the Yocto Project Overview and Concepts Manual.
A list of aliases by which a particular recipe can be
known.
By default, a recipe's own
PN
is implicitly already in its PROVIDES
list and therefore does not need to mention that it provides itself.
If a recipe uses PROVIDES
, the
additional aliases are synonyms for the recipe and can
be useful for satisfying dependencies of other recipes during
the build as specified by
DEPENDS
.
Consider the following example
PROVIDES
statement from the recipe
file eudev_3.2.9.bb
:
PROVIDES = "udev"
The PROVIDES
statement results in
the "eudev" recipe also being available as simply "udev".
PROVIDES
list,
it is unnecessary to add aliases with the "+=" operator;
using a simple assignment will be sufficient. In other
words, while you could write:
PROVIDES += "udev"in the above, the "+=" is overkill and unnecessary.
In addition to providing recipes under alternate names,
the PROVIDES
mechanism is also used
to implement virtual targets.
A virtual target is a name that corresponds to some
particular functionality (e.g. a Linux kernel).
Recipes that provide the functionality in question list the
virtual target in PROVIDES
.
Recipes that depend on the functionality in question can
include the virtual target in DEPENDS
to leave the choice of provider open.
Conventionally, virtual targets have names on the form "virtual/function" (e.g. "virtual/kernel"). The slash is simply part of the name and has no syntactical significance.
The
PREFERRED_PROVIDER
variable is used to select which particular recipe
provides a virtual target.
A corresponding mechanism for virtual runtime
dependencies (packages) exists.
However, the mechanism does not depend on any special
functionality beyond ordinary variable assignments.
For example,
VIRTUAL-RUNTIME_dev_manager
refers to the package of the component that manages
the /dev
directory.
Setting the "preferred provider" for runtime dependencies is as simple as using the following assignment in a configuration file:
VIRTUAL-RUNTIME_dev_manager = "udev"
The network based
PR
service host and port.
The conf/local.conf.sample.extended
configuration file in the
Source Directory
shows how the PRSERV_HOST
variable is
set:
PRSERV_HOST = "localhost:0"
You must set the variable if you want to automatically
start a local
PR service.
You can set PRSERV_HOST
to other
values to use a remote PR service.
Specifies whether or not
Package Test
(ptest) functionality is enabled when building a recipe.
You should not set this variable directly.
Enabling and disabling building Package Tests
at build time should be done by adding "ptest" to (or
removing it from)
DISTRO_FEATURES
.
The version of the recipe.
The version is normally extracted from the recipe filename.
For example, if the recipe is named
expat_2.0.1.bb
, then the default value
of PV
will be "2.0.1".
PV
is generally not overridden within
a recipe unless it is building an unstable (i.e.
development) version from a source code repository
(e.g. Git or Subversion).
PV
is the default value of the
PKGV
variable.
When used by recipes that inherit the
distutils3
,
setuptools3
,
distutils
,
or
setuptools
classes, denotes the Application Binary Interface (ABI)
currently in use for Python.
By default, the ABI is "m".
You do not have to set this variable as the OpenEmbedded
build system sets it for you.
The OpenEmbedded build system uses the ABI to construct
directory names used when installing the Python headers
and libraries in sysroot
(e.g. .../python3.3m/...
).
Recipes that inherit the distutils
class during cross-builds also use this variable to
locate the headers and libraries of the appropriate Python
that the extension is targeting.
When used by recipes that inherit the
distutils3
,
setuptools3
,
distutils
,
or
setuptools
classes, specifies the major Python version being built.
For Python 3.x, PYTHON_PN
would be
"python3".
You do not have to set this variable as the
OpenEmbedded build system automatically sets it for you.
The variable allows recipes to use common infrastructure such as the following:
DEPENDS += "${PYTHON_PN}-native"
In the previous example, the version of the dependency
is PYTHON_PN
.
The minimal command and arguments to run
ranlib
.
The list of packages that conflict with packages. Note that packages will not be installed if conflicting packages are not first removed.
Like all package-controlling variables, you must always use them in conjunction with a package name override. Here is an example:
RCONFLICTS_${PN} = "another_conflicting_package_name
"
BitBake, which the OpenEmbedded build system uses, supports
specifying versioned dependencies.
Although the syntax varies depending on the packaging
format, BitBake hides these differences from you.
Here is the general syntax to specify versions with
the RCONFLICTS
variable:
RCONFLICTS_${PN} = "package
(operator
version
)"
For operator
, you can specify the
following:
= < > <= >=
For example, the following sets up a dependency on version
1.2 or greater of the package foo
:
RCONFLICTS_${PN} = "foo (>= 1.2)"
Lists runtime dependencies of a package.
These dependencies are other packages that must be
installed in order for the package to function correctly.
As an example, the following assignment declares that the
package foo
needs the packages
bar
and baz
to
be installed:
RDEPENDS_foo = "bar baz"
The most common types of package runtime dependencies are
automatically detected and added.
Therefore, most recipes do not need to set
RDEPENDS
.
For more information, see the
"Automatically Added Runtime Dependencies"
section in the Yocto Project Overview and Concepts Manual.
The practical effect of the above
RDEPENDS
assignment is that
bar
and baz
will be declared as dependencies inside the package
foo
when it is written out by one of
the
do_package_write_*
tasks.
Exactly how this is done depends on which package format
is used, which is determined by
PACKAGE_CLASSES
.
When the corresponding package manager installs the
package, it will know to also install the packages on
which it depends.
To ensure that the packages bar
and
baz
get built, the previous
RDEPENDS
assignment also causes a task
dependency to be added.
This dependency is from the recipe's
do_build
(not to be confused with
do_compile
)
task to the do_package_write_*
task of the recipes that build bar
and
baz
.
The names of the packages you list within
RDEPENDS
must be the names of other
packages - they cannot be recipe names.
Although package names and recipe names usually match,
the important point here is that you are
providing package names within the
RDEPENDS
variable.
For an example of the default list of packages created from
a recipe, see the
PACKAGES
variable.
Because the RDEPENDS
variable applies
to packages being built, you should always use the variable
in a form with an attached package name (remember that a
single recipe can build multiple packages).
For example, suppose you are building a development package
that depends on the perl
package.
In this case, you would use the following
RDEPENDS
statement:
RDEPENDS_${PN}-dev += "perl"
In the example, the development package depends on
the perl
package.
Thus, the RDEPENDS
variable has the
${PN}-dev
package name as part of the
variable.
RDEPENDS_${PN}-dev
includes
${
PN
}
by default.
This default is set in the BitBake configuration file
(meta/conf/bitbake.conf
).
Be careful not to accidentally remove
${PN}
when modifying
RDEPENDS_${PN}-dev
.
Use the "+=" operator rather than the "=" operator.
The package names you use with
RDEPENDS
must appear as they would in
the PACKAGES
variable.
The
PKG
variable allows a different name to be used for
the final package (e.g. the
debian
class uses this to rename packages), but this final package
name cannot be used with RDEPENDS
,
which makes sense as RDEPENDS
is meant
to be independent of the package format used.
BitBake, which the OpenEmbedded build system uses, supports
specifying versioned dependencies.
Although the syntax varies depending on the packaging
format, BitBake hides these differences from you.
Here is the general syntax to specify versions with
the RDEPENDS
variable:
RDEPENDS_${PN} = "package
(operator
version
)"
For operator
, you can specify the
following:
= < > <= >=
For version
, provide the version
number.
EXTENDPKGV
to provide a full package version specification.
For example, the following sets up a dependency on version
1.2 or greater of the package foo
:
RDEPENDS_${PN} = "foo (>= 1.2)"
For information on build-time dependencies, see the
DEPENDS
variable.
You can also see the
"Tasks" and
"Dependencies"
sections in the BitBake User Manual for additional
information on tasks and dependencies.
When inheriting the
distro_features_check
class, this
variable identifies distribution features that must
exist in the current configuration in order for the
OpenEmbedded build system to build the recipe.
In other words, if the
REQUIRED_DISTRO_FEATURES
variable
lists a feature that does not appear in
DISTRO_FEATURES
within the
current configuration, an error occurs and the
build stops.
With rm_work
enabled, this
variable specifies a list of recipes whose work directories
should not be removed.
See the "rm_work.bbclass
"
section for more details.
Defines the root home directory. By default, this directory is set as follows in the BitBake configuration file:
ROOT_HOME ??= "/home/root"
You can override the default by setting the variable
in any layer or in the local.conf
file.
Because the default is set using a "weak" assignment
(i.e. "??="), you can use either of the following forms
to define your override:
ROOT_HOME = "/root" ROOT_HOME ?= "/root"
These override examples use /root
,
which is probably the most commonly used override.
Indicates a filesystem image to include as the root filesystem.
The ROOTFS
variable is an optional
variable used with the
image-live
class.
Specifies a list of functions to call after the OpenEmbedded build system has installed packages. You can specify functions separated by semicolons:
ROOTFS_POSTINSTALL_COMMAND += "function
; ... "
If you need to pass the root filesystem path to a command
within a function, you can use
${IMAGE_ROOTFS}
, which points to
the directory that becomes the root filesystem image.
See the
IMAGE_ROOTFS
variable for more information.
Specifies a list of functions to call once the OpenEmbedded build system has created the root filesystem. You can specify functions separated by semicolons:
ROOTFS_POSTPROCESS_COMMAND += "function
; ... "
If you need to pass the root filesystem path to a command
within a function, you can use
${IMAGE_ROOTFS}
, which points to
the directory that becomes the root filesystem image.
See the
IMAGE_ROOTFS
variable for more information.
Specifies a list of functions to call after the
OpenEmbedded build system has removed unnecessary
packages.
When runtime package management is disabled in the
image, several packages are removed including
base-passwd
,
shadow
, and
update-alternatives
.
You can specify functions separated by semicolons:
ROOTFS_POSTUNINSTALL_COMMAND += "function
; ... "
If you need to pass the root filesystem path to a command
within a function, you can use
${IMAGE_ROOTFS}
, which points to
the directory that becomes the root filesystem image.
See the
IMAGE_ROOTFS
variable for more information.
Specifies a list of functions to call before the OpenEmbedded build system has created the root filesystem. You can specify functions separated by semicolons:
ROOTFS_PREPROCESS_COMMAND += "function
; ... "
If you need to pass the root filesystem path to a command
within a function, you can use
${IMAGE_ROOTFS}
, which points to
the directory that becomes the root filesystem image.
See the
IMAGE_ROOTFS
variable for more information.
A list of package name aliases that a package also provides.
These aliases are useful for satisfying runtime dependencies
of other packages both during the build and on the target
(as specified by
RDEPENDS
).
RPROVIDES
list.
As with all package-controlling variables, you must always use the variable in conjunction with a package name override. Here is an example:
RPROVIDES_${PN} = "widget-abi-2"
A list of packages that extends the usability of a package
being built.
The package being built does not depend on this list of
packages in order to successfully build, but rather
uses them for extended usability.
To specify runtime dependencies for packages, see the
RDEPENDS
variable.
The package manager will automatically install the
RRECOMMENDS
list of packages when
installing the built package.
However, you can prevent listed packages from being
installed by using the
BAD_RECOMMENDATIONS
,
NO_RECOMMENDATIONS
,
and
PACKAGE_EXCLUDE
variables.
Packages specified in
RRECOMMENDS
need not actually be
produced.
However, a recipe must exist that provides each package,
either through the
PACKAGES
or
PACKAGES_DYNAMIC
variables or the
RPROVIDES
variable, or an error will occur during the build.
If such a recipe does exist and the package is not produced,
the build continues without error.
Because the RRECOMMENDS
variable
applies to packages being built, you should always attach
an override to the variable to specify the particular
package whose usability is being extended.
For example, suppose you are building a development package
that is extended to support wireless functionality.
In this case, you would use the following:
RRECOMMENDS_${PN}-dev += "wireless_package_name
"
In the example, the package name
(${PN}-dev
)
must appear as it would in the
PACKAGES
namespace before any renaming
of the output package by classes such as
debian.bbclass
.
BitBake, which the OpenEmbedded build system uses, supports
specifying versioned recommends.
Although the syntax varies depending on the packaging
format, BitBake hides these differences from you.
Here is the general syntax to specify versions with
the RRECOMMENDS
variable:
RRECOMMENDS_${PN} = "package
(operator
version
)"
For operator
, you can specify the
following:
= < > <= >=
For example, the following sets up a recommend on version
1.2 or greater of the package foo
:
RRECOMMENDS_${PN} = "foo (>= 1.2)"
A list of packages replaced by a package.
The package manager uses this variable to determine which
package should be installed to replace other package(s)
during an upgrade.
In order to also have the other package(s) removed at the
same time, you must add the name of the other
package to the
RCONFLICTS
variable.
As with all package-controlling variables, you must use this variable in conjunction with a package name override. Here is an example:
RREPLACES_${PN} = "other_package_being_replaced
"
BitBake, which the OpenEmbedded build system uses, supports
specifying versioned replacements.
Although the syntax varies depending on the packaging
format, BitBake hides these differences from you.
Here is the general syntax to specify versions with
the RREPLACES
variable:
RREPLACES_${PN} = "package
(operator
version
)"
For operator
, you can specify the
following:
= < > <= >=
For example, the following sets up a replacement using
version 1.2 or greater of the package
foo
:
RREPLACES_${PN} = "foo (>= 1.2)"
A list of additional packages that you can suggest for installation by the package manager at the time a package is installed. Not all package managers support this functionality.
As with all package-controlling variables, you must always use this variable in conjunction with a package name override. Here is an example:
RSUGGESTS_${PN} = "useful_package
another_package
"
The location in the
Build Directory
where unpacked recipe source code resides.
By default, this directory is
${
WORKDIR
}/${
BPN
}-${
PV
}
,
where ${BPN}
is the base recipe name
and ${PV}
is the recipe version.
If the source tarball extracts the code to a directory
named anything other than ${BPN}-${PV}
,
or if the source code is fetched from an SCM such as
Git or Subversion, then you must set S
in the recipe so that the OpenEmbedded build system
knows where to find the unpacked source.
As an example, assume a
Source Directory
top-level folder named poky
and a
default Build Directory at poky/build
.
In this case, the work directory the build system uses
to keep the unpacked recipe for db
is the following:
poky/build/tmp/work/qemux86-poky-linux/db/5.1.19-r3/db-5.1.19
The unpacked source code resides in the
db-5.1.19
folder.
This next example assumes a Git repository.
By default, Git repositories are cloned to
${WORKDIR}/git
during
do_fetch
.
Since this path is different from the default value of
S
, you must set it specifically
so the source can be located:
SRC_URI = "git://path/to/repo.git" S = "${WORKDIR}/git"
Specifies a list of command-line utilities that should be checked for during the initial sanity checking process when running BitBake. If any of the utilities are not installed on the build host, then BitBake immediately exits with an error.
A list of the host distribution identifiers that the
build system has been tested against.
Identifiers consist of the host distributor ID
followed by the release,
as reported by the lsb_release
tool
or as read from /etc/lsb-release
.
Separate the list items with explicit newline
characters (\n
).
If SANITY_TESTED_DISTROS
is not empty
and the current value of
NATIVELSBSTRING
does not appear in the list, then the build system reports
a warning that indicates the current host distribution has
not been tested as a build host.
The target architecture for the SDK.
Typically, you do not directly set this variable.
Instead, use
SDKMACHINE
.
The directory set up and used by the
populate_sdk_base
class to which the SDK is deployed.
The populate_sdk_base
class defines
SDK_DEPLOY
as follows:
SDK_DEPLOY = "${TMPDIR}/deploy/sdk"
The parent directory used by the OpenEmbedded build system
when creating SDK output.
The
populate_sdk_base
class defines the variable as follows:
SDK_DIR = "${WORKDIR}/sdk"
SDK_DIR
directory is a
temporary directory as it is part of
WORKDIR
.
The final output directory is
SDK_DEPLOY
.
Controls whether or not shared state artifacts are copied into the extensible SDK. The default value of "full" copies all of the required shared state artifacts into the extensible SDK. The value "minimal" leaves these artifacts out of the SDK.
SSTATE_MIRRORS
is set in the SDK's configuration to enable the
artifacts to be fetched as needed.
The manifest file for the host part of the SDK. This file lists all the installed packages that make up the host part of the SDK. The file contains package information on a line-per-package basis as follows:
packagename
packagearch
version
The
populate_sdk_base
class defines the manifest file as follows:
SDK_HOST_MANIFEST = "${SDK_DEPLOY}/${TOOLCHAIN_OUTPUTNAME}.host.manifest"
The location is derived using the
SDK_DEPLOY
and
TOOLCHAIN_OUTPUTNAME
variables.
When set to "1", specifies to include the packagedata for
all recipes in the "world" target in the extensible SDK.
Including this data allows the
devtool search
command to find these
recipes in search results, as well as allows the
devtool add
command to map
dependencies more effectively.
SDK_INCLUDE_PKGDATA
variable significantly increases build time because
all of world needs to be built.
Enabling the variable also slightly increases the size
of the extensible SDK.
When set to "1", specifies to include the toolchain in the
extensible SDK.
Including the toolchain is useful particularly when
SDK_EXT_TYPE
is set to "minimal" to keep the SDK reasonably small
but you still want to provide a usable toolchain.
For example, suppose you want to use the toolchain from an
IDE or from other tools and you do not
want to perform additional steps to install the toolchain.
The SDK_INCLUDE_TOOLCHAIN
variable
defaults to "0" if SDK_EXT_TYPE
is set to "minimal", and defaults to "1" if
SDK_EXT_TYPE
is set to "full".
A list of classes to remove from the
INHERIT
value globally within the extensible SDK configuration.
The
populate-sdk-ext
class sets the default value:
SDK_INHERIT_BLACKLIST ?= "buildhistory icecc"
Some classes are not generally applicable within the extensible SDK context. You can use this variable to disable those classes.
For additional information on how to customize the extensible SDK's configuration, see the "Configuring the Extensible SDK" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
A list of variables not allowed through from the OpenEmbedded build system configuration into the extensible SDK configuration. Usually, these are variables that are specific to the machine on which the build system is running and thus would be potentially problematic within the extensible SDK.
By default,
SDK_LOCAL_CONF_BLACKLIST
is set in the
populate-sdk-ext
class and excludes the following variables:
CONF_VERSION BB_NUMBER_THREADS BB_NUMBER_PARSE_THREADS PARALLEL_MAKE PRSERV_HOST SSTATE_MIRRORS DL_DIR SSTATE_DIR TMPDIR BB_SERVER_TIMEOUT
For additional information on how to customize the extensible SDK's configuration, see the "Configuring the Extensible SDK" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
A list of variables allowed through from the OpenEmbedded
build system configuration into the extensible SDK
configuration.
By default, the list of variables is empty and is set in
the
populate-sdk-ext
class.
This list overrides the variables specified using the
SDK_LOCAL_CONF_BLACKLIST
variable as well as any variables identified by automatic
blacklisting due to the "/" character being found at the
start of the value, which is usually indicative of being a
path and thus might not be valid on the system where the
SDK is installed.
For additional information on how to customize the extensible SDK's configuration, see the "Configuring the Extensible SDK" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
The base name for SDK output files.
The name is derived from the
DISTRO
,
TCLIBC
,
SDK_ARCH
,
IMAGE_BASENAME
,
and
TUNE_PKGARCH
variables:
SDK_NAME = "${DISTRO}-${TCLIBC}-${SDK_ARCH}-${IMAGE_BASENAME}-${TUNE_PKGARCH}"
Specifies the operating system for which the SDK
will be built.
The default value is the value of
BUILD_OS
.
The location used by the OpenEmbedded build system when
creating SDK output.
The
populate_sdk_base
class defines the variable as follows:
SDK_DIR = "${WORKDIR}/sdk" SDK_OUTPUT = "${SDK_DIR}/image" SDK_DEPLOY = "${DEPLOY_DIR}/sdk"
SDK_OUTPUT
directory is a
temporary directory as it is part of
WORKDIR
by way of
SDK_DIR
.
The final output directory is
SDK_DEPLOY
.
Specifies a list of architectures compatible with
the SDK machine.
This variable is set automatically and should not
normally be hand-edited.
Entries are separated using spaces and listed in order
of priority.
The default value for
SDK_PACKAGE_ARCHS
is "all any noarch
${SDK_ARCH}-${SDKPKGSUFFIX}".
Specifies a list of functions to call once the OpenEmbedded build system creates the SDK. You can specify functions separated by semicolons:
SDK_POSTPROCESS_COMMAND += "function
; ... "
If you need to pass an SDK path to a command within a
function, you can use
${SDK_DIR}
, which points to
the parent directory used by the OpenEmbedded build system
when creating SDK output.
See the
SDK_DIR
variable for more information.
The toolchain binary prefix used for
nativesdk
recipes.
The OpenEmbedded build system uses the
SDK_PREFIX
value to set the
TARGET_PREFIX
when building nativesdk
recipes.
The default value is "${SDK_SYS}-".
A list of shared state tasks added to the extensible SDK. By default, the following tasks are added:
do_populate_lic do_package_qa do_populate_sysroot do_deploy
Despite the default value of "" for the
SDK_RECRDEP_TASKS
variable, the
above four tasks are always added to the SDK.
To specify tasks beyond these four, you need to use
the SDK_RECRDEP_TASKS
variable (e.g.
you are defining additional tasks that are needed in
order to build
SDK_TARGETS
).
Specifies the system, including the architecture and the operating system, for which the SDK will be built.
The OpenEmbedded build system automatically sets this
variable based on
SDK_ARCH
,
SDK_VENDOR
,
and
SDK_OS
.
You do not need to set the SDK_SYS
variable yourself.
The manifest file for the target part of the SDK. This file lists all the installed packages that make up the target part of the SDK. The file contains package information on a line-per-package basis as follows:
packagename
packagearch
version
The
populate_sdk_base
class defines the manifest file as follows:
SDK_TARGET_MANIFEST = "${SDK_DEPLOY}/${TOOLCHAIN_OUTPUTNAME}.target.manifest"
The location is derived using the
SDK_DEPLOY
and
TOOLCHAIN_OUTPUTNAME
variables.
A list of targets to install from shared state as part of the standard or extensible SDK installation. The default value is "${PN}" (i.e. the image from which the SDK is built).
The SDK_TARGETS
variable is an
internal variable and typically would not be changed.
The title to be printed when running the SDK installer.
By default, this title is based on the
DISTRO_NAME
or
DISTRO
variable and is set in the
populate_sdk_base
class as follows:
SDK_TITLE ??= "${@d.getVar('DISTRO_NAME') or d.getVar('DISTRO')} SDK"
For the default distribution "poky",
SDK_TITLE
is set to
"Poky (Yocto Project Reference Distro)".
For information on how to change this default title, see the "Changing the Extensible SDK Installer Title" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
An optional URL for an update server for the extensible
SDK.
If set, the value is used as the default update server when
running devtool sdk-update
within the
extensible SDK.
Specifies the name of the SDK vendor.
Specifies the version of the SDK.
The distribution configuration file (e.g.
/meta-poky/conf/distro/poky.conf
)
defines the SDK_VERSION
as follows:
SDK_VERSION = "${@d.getVar('DISTRO_VERSION').replace('snapshot-${DATE}','snapshot')}"
For additional information, see the
DISTRO_VERSION
and
DATE
variables.
The default installation directory for the Extensible SDK.
By default, this directory is based on the
DISTRO
variable and is set in the
populate_sdk_base
class as follows:
SDKEXTPATH ??= "~/${@d.getVar('DISTRO')}_sdk"
For the default distribution "poky", the
SDKEXTPATH
is set to "poky_sdk".
For information on how to change this default directory, see the "Changing the Default SDK Installation Directory" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
Equivalent to
IMAGE_FEATURES
.
However, this variable applies to the SDK generated from an
image using the following command:
$ bitbake -c populate_sdk imagename
The machine for which the SDK is built.
In other words, the SDK is built such that it
runs on the target you specify with the
SDKMACHINE
value.
The value points to a corresponding
.conf
file under
conf/machine-sdk/
.
You can use "i686" and "x86_64" as possible values for this variable. The variable defaults to "i686" and is set in the local.conf file in the Build Directory.
SDKMACHINE ?= "i686"
SDKMACHINE
variable in your distribution configuration file.
If you do, the configuration will not take affect.
Defines the path offered to the user for installation of the SDK that is generated by the OpenEmbedded build system. The path appears as the default location for installing the SDK when you run the SDK's installation script. You can override the offered path when you run the script.
The full path to the sysroot used for cross-compilation
within an SDK as it will be when installed into the
default
SDKPATH
.
The section in which packages should be categorized. Package management utilities can make use of this variable.
Specifies the optimization flags passed to the C compiler
when building for the target.
The flags are passed through the default value of the
TARGET_CFLAGS
variable.
The SELECTED_OPTIMIZATION
variable
takes the value of
FULL_OPTIMIZATION
unless DEBUG_BUILD
= "1".
If that is the case, the value of
DEBUG_OPTIMIZATION
is used.
Defines a serial console (TTY) to enable using getty. Provide a value that specifies the baud rate followed by the TTY device name separated by a space. You cannot specify more than one TTY device:
SERIAL_CONSOLE = "115200 ttyS0"
Defines a serial console (TTY) to enable using getty. Provide a value that specifies the baud rate followed by the TTY device name separated by a semicolon. Use spaces to separate multiple devices:
SERIAL_CONSOLES = "115200;ttyS0 115200;ttyS1"
Specifies serial consoles, which must be listed in
SERIAL_CONSOLES
,
to check against /proc/console
before enabling them using getty.
This variable allows aliasing in the format:
<device>:<alias>.
If a device was listed as "sclp_line0"
in /dev/
and "ttyS0" was listed
in /proc/console
, you would do the
following:
SERIAL_CONSOLES_CHECK = "slcp_line0:ttyS0"
This variable is currently only supported with SysVinit (i.e. not with systemd).
A list of recipe dependencies that should not be used to determine signatures of tasks from one recipe when they depend on tasks from another recipe. For example:
SIGGEN_EXCLUDE_SAFE_RECIPE_DEPS += "intone->mplayer2"
In the previous example, intone
depends on mplayer2
.
You can use the special token "*"
on
the left-hand side of the dependency to match all
recipes except the one on the right-hand side.
Here is an example:
SIGGEN_EXCLUDE_SAFE_RECIPE_DEPS += "*->quilt-native"
In the previous example, all recipes except
quilt-native
ignore task
signatures from the quilt-native
recipe when determining their task signatures.
Use of this variable is one mechanism to remove dependencies that affect task signatures and thus force rebuilds when a recipe changes.
A list of recipes that are completely stable and will never change. The ABI for the recipes in the list are presented by output from the tasks run to build the recipe. Use of this variable is one way to remove dependencies from one recipe on another that affect task signatures and thus force rebuilds when the recipe changes.
Specifies the number of bits for the target system CPU. The value should be either "32" or "64".
Specifies the endian byte order of the target system. The value should be either "le" for little-endian or "be" for big-endian.
Enables removal of all files from the "Provides" section of
an RPM package.
Removal of these files is required for packages containing
prebuilt binaries and libraries such as
libstdc++
and
glibc
.
To enable file removal, set the variable to "1" in your
conf/local.conf
configuration file
in your:
Build Directory.
SKIP_FILEDEPS = "1"
Groups together machines based upon the same family
of SOC (System On Chip).
You typically set this variable in a common
.inc
file that you include in the
configuration files of all the machines.
conf/machine/include/soc-family.inc
for this variable to appear in
MACHINEOVERRIDES
.
Defines the suffix for shared libraries used on the
target platform.
By default, this suffix is ".so.*" for all Linux-based
systems and is defined in the
meta/conf/bitbake.conf
configuration
file.
You will see this variable referenced in the default values
of FILES_${PN}
.
Defines the suffix for the development symbolic link
(symlink) for shared libraries on the target platform.
By default, this suffix is ".so" for Linux-based
systems and is defined in the
meta/conf/bitbake.conf
configuration
file.
You will see this variable referenced in the default values
of FILES_${PN}-dev
.
When you are fetching files to create a mirror of sources
(i.e. creating a source mirror), setting
SOURCE_MIRROR_FETCH
to "1" in your
local.conf
configuration file ensures
the source for all recipes are fetched regardless of
whether or not a recipe is compatible with the
configuration.
A recipe is considered incompatible with the currently
configured machine when either or both the
COMPATIBLE_MACHINE
variable and
COMPATIBLE_HOST
variables specify compatibility with a machine other
than that of the current machine or host.
SOURCE_MIRROR_FETCH
variable
unless you are creating a source mirror.
In other words, do not set the variable during a
normal build.
Defines your own
PREMIRRORS
from which to first fetch source before attempting to fetch
from the upstream specified in
SRC_URI
.
To use this variable, you must globally inherit the
own-mirrors
class and then provide the URL to your mirrors.
Here is the general syntax:
INHERIT += "own-mirrors" SOURCE_MIRROR_URL = "http://example
.com/my_source_mirror
"
SOURCE_MIRROR_URL
.
Maps commonly used license names to their SPDX counterparts
found in meta/files/common-licenses/
.
For the default SPDXLICENSEMAP
mappings, see the
meta/conf/licenses.conf
file.
For additional information, see the
LICENSE
variable.
A list of prefixes for PN
used by the
OpenEmbedded build system to create variants of recipes or packages.
The list specifies the prefixes to strip off during certain circumstances
such as the generation of the BPN
variable.
The file type for the Secondary Program Loader (SPL).
Some devices use an SPL from which to boot (e.g. the
BeagleBone development board).
For such cases, you can declare the file type of the
SPL binary in the u-boot.inc
include
file, which is used in the U-Boot recipe.
The SPL file type is set to "null" by default in the
u-boot.inc
file as follows:
# Some versions of u-boot build an SPL (Second Program Loader) image that # should be packaged along with the u-boot binary as well as placed in the # deploy directory. For those versions they can set the following variables # to allow packaging the SPL. SPL_BINARY ?= "" SPL_BINARYNAME ?= "${@os.path.basename(d.getVar("SPL_BINARY"))}" SPL_IMAGE ?= "${SPL_BINARYNAME}-${MACHINE}-${PV}-${PR}" SPL_SYMLINK ?= "${SPL_BINARYNAME}-${MACHINE}"
The SPL_BINARY
variable helps form
various SPL_*
variables used by
the OpenEmbedded build system.
See the BeagleBone machine configuration example in the
"Creating a new BSP Layer Using the bitbake-layers
Script"
section in the Yocto Project Board Support Package
Developer's Guide for additional information.
The list of source files - local or remote.
This variable tells the OpenEmbedded build system which bits
to pull in for the build and how to pull them in.
For example, if the recipe or append file only needs to
fetch a tarball from the Internet, the recipe or
append file uses a single SRC_URI
entry.
On the other hand, if the recipe or append file needs to
fetch a tarball, apply two patches, and include a custom
file, the recipe or append file would include four
instances of the variable.
The following list explains the available URI protocols. URI protocols are highly dependent on particular BitBake Fetcher submodules. Depending on the fetcher BitBake uses, various URL parameters are employed. For specifics on the supported Fetchers, see the "Fetchers" section in the BitBake User Manual.
file://
-
Fetches files, which are usually files shipped with
the
Metadata,
from the local machine (e.g.
patch
files).
The path is relative to the
FILESPATH
variable.
Thus, the build system searches, in order, from the
following directories, which are assumed to be a
subdirectories of the directory in which the
recipe file (.bb
) or
append file (.bbappend
)
resides:
${BPN}
-
The base recipe name without any special
suffix or version numbers.
${BP}
-
${BPN}-${PV}
.
The base recipe name and version but without
any special package name suffix.
files -
Files within a directory, which is named
files
and is also
alongside the recipe or append file.
SRC_URI
statement from your append file, you need to be
sure to extend the
FILESPATH
variable by also using the
FILESEXTRAPATHS
variable from within your append file.
bzr://
- Fetches files from a
Bazaar revision control repository.
git://
- Fetches files from a
Git revision control repository.
osc://
- Fetches files from
an OSC (OpenSUSE Build service) revision control repository.
repo://
- Fetches files from
a repo (Git) repository.
ccrc://
-
Fetches files from a ClearCase repository.
http://
- Fetches files from
the Internet using http
.
https://
- Fetches files
from the Internet using https
.
ftp://
- Fetches files
from the Internet using ftp
.
cvs://
- Fetches files from
a CVS revision control repository.
hg://
- Fetches files from
a Mercurial (hg
) revision control repository.
p4://
- Fetches files from
a Perforce (p4
) revision control repository.
ssh://
- Fetches files from
a secure shell.
svn://
- Fetches files from
a Subversion (svn
) revision control repository.
npm://
- Fetches JavaScript
modules from a registry.
Standard and recipe-specific options for SRC_URI
exist.
Here are standard options:
apply
- Whether to apply
the patch or not.
The default action is to apply the patch.
striplevel
- Which
striplevel to use when applying the patch.
The default level is 1.
patchdir
- Specifies
the directory in which the patch should be applied.
The default is ${
S
}
.
Here are options specific to recipes building code from a revision control system:
mindate
-
Apply the patch only if
SRCDATE
is equal to or greater than mindate
.
maxdate
-
Apply the patch only if SRCDATE
is not later than maxdate
.
minrev
-
Apply the patch only if SRCREV
is equal to or greater than minrev
.
maxrev
-
Apply the patch only if SRCREV
is not later than maxrev
.
rev
-
Apply the patch only if SRCREV
is equal to rev
.
notrev
-
Apply the patch only if SRCREV
is not equal to rev
.
Here are some additional options worth mentioning:
unpack
- Controls
whether or not to unpack the file if it is an archive.
The default action is to unpack the file.
destsuffix
- Places the file
(or extracts its contents) into the specified
subdirectory of WORKDIR
when the Git fetcher is used.
subdir
- Places the file
(or extracts its contents) into the specified
subdirectory of WORKDIR
when the local (file://
)
fetcher is used.
localdir
- Places the file
(or extracts its contents) into the specified
subdirectory of WORKDIR
when
the CVS fetcher is used.
subpath
-
Limits the checkout to a specific subpath of the
tree when using the Git fetcher is used.
name
- Specifies a
name to be used for association with SRC_URI
checksums
when you have more than one file specified in SRC_URI
.
downloadfilename
- Specifies
the filename used when storing the downloaded file.
By default, the OpenEmbedded build system automatically detects whether
SRC_URI
contains files that are machine-specific.
If so, the build system automatically changes
PACKAGE_ARCH
.
Setting this variable to "0" disables this behavior.
The date of the source code used to build the package. This variable applies only if the source was fetched from a Source Code Manager (SCM).
Returns the version string of the current package.
This string is used to help define the value of
PV
.
The SRCPV
variable is defined in the
meta/conf/bitbake.conf
configuration
file in the
Source Directory
as follows:
SRCPV = "${@bb.fetch2.get_srcrev(d)}"
Recipes that need to define PV
do so
with the help of the SRCPV
.
For example, the ofono
recipe
(ofono_git.bb
) located in
meta/recipes-connectivity
in the
Source Directory defines PV
as
follows:
PV = "0.12-git${SRCPV}"
The revision of the source code used to build the package.
This variable applies to Subversion, Git, Mercurial, and
Bazaar only.
Note that if you want to build a fixed revision and you
want to avoid performing a query on the remote repository
every time BitBake parses your recipe, you should specify
a SRCREV
that is a
full revision identifier and not just a tag.
SRCREV
, see the
AUTOREV
variable description and the
"Automatically Incrementing a Binary Package Revision Number"
section, which is in the Yocto Project Development
Tasks Manual.
The directory for the shared state cache.
If set to "1", allows fetches from
mirrors that are specified in
SSTATE_MIRRORS
to work even when fetching from the network is
disabled by setting BB_NO_NETWORK
to "1".
Using the
SSTATE_MIRROR_ALLOW_NETWORK
variable is useful if you have set
SSTATE_MIRRORS
to point to an
internal server for your shared state cache, but
you want to disable any other fetching from the network.
Configures the OpenEmbedded build system to search other
mirror locations for prebuilt cache data objects before
building out the data.
This variable works like fetcher
MIRRORS
and PREMIRRORS
and points to the cache locations to check for the shared
state (sstate) objects.
You can specify a filesystem directory or a remote URL such as HTTP or FTP. The locations you specify need to contain the shared state cache (sstate-cache) results from previous builds. The sstate-cache you point to can also be from builds on other machines.
When pointing to sstate build artifacts on another machine
that uses a different GCC version for native builds,
you must configure SSTATE_MIRRORS
with a regular expression that maps local search paths
to server paths.
The paths need to take into account
NATIVELSBSTRING
set by the
uninative
class.
For example, the following maps the local search path
universal-4.9
to the server-provided
path server_url_sstate_path
:
SSTATE_MIRRORS ?= file://universal-4.9/(.*) http://server_url_sstate_path
/universal-4.8/\1 \n
If a mirror uses the same structure as
SSTATE_DIR
,
you need to add
"PATH" at the end as shown in the examples below.
The build system substitutes the correct path within the
directory structure.
SSTATE_MIRRORS ?= "\ file://.* http://someserver
.tld/share/sstate/PATH;downloadfilename=PATH \n \ file://.* file:///some-local-dir
/sstate/PATH"
Controls the list of files the OpenEmbedded build system scans for hardcoded installation paths. The variable uses a space-separated list of filenames (not paths) with standard wildcard characters allowed.
During a build, the OpenEmbedded build system creates a
shared state (sstate) object during the first stage of
preparing the sysroots. That object is scanned for
hardcoded paths for original installation locations.
The list of files that are scanned for paths is controlled
by the SSTATE_SCAN_FILES
variable.
Typically, recipes add files they want to be scanned to the
value of SSTATE_SCAN_FILES
rather than
the variable being comprehensively set. The
sstate
class specifies the default list of files.
For details on the process, see the
staging
class.
Specifies the path to the /lib
subdirectory of the sysroot directory for the
build host.
Specifies the path to the /lib
subdirectory of the sysroot directory for the target
for which the current recipe is being built
(STAGING_DIR_HOST
).
Specifies the path to the
/usr/bin
subdirectory of the
sysroot directory for the target for which the current
recipe is being built
(STAGING_DIR_HOST
).
Specifies the path to the directory containing binary configuration scripts. These scripts provide configuration information for other software that wants to make use of libraries or include files provided by the software associated with the script.
pkg-config
.
Consequently, if pkg-config
is supported by the library to which you are linking,
it is recommended you use
pkg-config
instead of a
provided configuration script.
Specifies the path to the
/usr/bin
subdirectory of the
sysroot directory for the build host.
Specifies the path to the /usr/share
subdirectory of the sysroot directory for the target
for which the current recipe is being built
(STAGING_DIR_HOST
).
Specifies the path to the /usr/share
subdirectory of the sysroot directory for the build host.
Helps construct the recipe-sysroots
directory, which is used during packaging.
For information on how staging for recipe-specific
sysroots occurs, see the
do_populate_sysroot
task, the
"Sharing Files Between Recipes"
section in the Yocto Project Development Tasks Manual, the
"Configuration, Compilation, and Staging"
section in the Yocto Project Overview and Concepts Manual,
and the
SYSROOT_DIRS
variable.
STAGING_DIR
directory because
the OpenEmbedded build system
manages the directory automatically.
Instead, files should be installed to
${
D
}
within your recipe's
do_install
task and then the OpenEmbedded build system will
stage a subset of those files into the sysroot.
Specifies the path to the sysroot directory for the system
on which the component is built to run (the system that
hosts the component).
For most recipes, this sysroot is the one in which that
recipe's
do_populate_sysroot
task copies files.
Exceptions include -native
recipes,
where the do_populate_sysroot
task
instead uses
STAGING_DIR_NATIVE
.
Depending on the type of recipe and the build target,
STAGING_DIR_HOST
can have the
following values:
For recipes building for the target machine, the value is "${STAGING_DIR}/${MACHINE}".
For native recipes building for the build host, the value is empty given the assumption that when building for the build host, the build host's own directories should be used.
-native
recipes are
not installed into host paths like such as
/usr
.
Rather, these recipes are installed into
STAGING_DIR_NATIVE
.
When compiling -native
recipes, standard build environment variables
such as
CPPFLAGS
and
CFLAGS
are set up so that both host paths and
STAGING_DIR_NATIVE
are
searched for libraries and headers using, for
example, GCC's -isystem
option.
Thus, the emphasis is that the
STAGING_DIR*
variables
should be viewed as input variables by tasks
such as
do_configure
,
do_compile
,
and
do_install
.
Having the real system root correspond to
STAGING_DIR_HOST
makes
conceptual sense for
-native
recipes, as
they make use of host headers and libraries.
Specifies the path to the sysroot directory used when building components that run on the build host itself.
Specifies the path to the sysroot used for the system for
which the component generates code.
For components that do not generate code, which is the
majority, STAGING_DIR_TARGET
is set
to match
STAGING_DIR_HOST
.
Some recipes build binaries that can run on the target
system but those binaries in turn generate code for
another different system (e.g. cross-canadian recipes).
Using terminology from GNU, the primary system is referred
to as the "HOST" and the secondary, or different, system is
referred to as the "TARGET".
Thus, the binaries run on the "HOST" system
and generate binaries for the "TARGET" system.
The STAGING_DIR_HOST
variable points
to the sysroot used for the "HOST" system, while
STAGING_DIR_TARGET
points to the sysroot used for the "TARGET" system.
Specifies the path to the /etc
subdirectory of the sysroot directory for the
build host.
Specifies the path to the /usr
subdirectory of the sysroot directory for the target
for which the current recipe is being built
(STAGING_DIR_HOST
).
Specifies the path to the
/usr/include
subdirectory of the
sysroot directory for the target for which the current
recipe being built
(STAGING_DIR_HOST
).
Specifies the path to the /usr/include
subdirectory of the sysroot directory for the build host.
Points to the directory containing the kernel build
artifacts.
Recipes building software that needs to access kernel
build artifacts
(e.g. systemtap-uprobes
) can look in
the directory specified with the
STAGING_KERNEL_BUILDDIR
variable to
find these artifacts after the kernel has been built.
The directory with kernel headers that are required to build out-of-tree modules.
Specifies the path to the /usr/lib
subdirectory of the sysroot directory for the target for
which the current recipe is being built
(STAGING_DIR_HOST
).
Specifies the path to the /usr/lib
subdirectory of the sysroot directory for the build host.
Specifies the base path used to create recipe stamp files.
The path to an actual stamp file is constructed by evaluating this
string and then appending additional information.
Currently, the default assignment for STAMP
as set in the meta/conf/bitbake.conf
file
is:
STAMP = "${STAMPS_DIR}/${MULTIMACH_TARGET_SYS}/${PN}/${EXTENDPE}${PV}-${PR}"
For information on how BitBake uses stamp files to determine if a task should be rerun, see the "Stamp Files and the Rerunning of Tasks" section in the Yocto Project Overview and Concepts Manual.
See STAMPS_DIR
,
MULTIMACH_TARGET_SYS
,
PN
,
EXTENDPE
,
PV
, and
PR
for related variable
information.
Specifies the base directory in which the OpenEmbedded
build system places stamps.
The default directory is
${TMPDIR}/stamps
.
The minimal command and arguments to run
strip
, which is used to strip
symbols.
The short (72 characters or less) summary of the binary package for packaging
systems such as opkg
, rpm
, or
dpkg
.
By default, SUMMARY
is used to define
the DESCRIPTION
variable if DESCRIPTION
is not set
in the recipe.
The directory in which files checked out of a Subversion system are stored.
Specifies the kernel boot default console. If you want to use a console other than the default, set this variable in your recipe as follows where "X" is the console number you want to use:
SYSLINUX_DEFAULT_CONSOLE = "console=ttyX"
The
syslinux
class initially sets this variable to null but then checks
for a value later.
Lists additional options to add to the syslinux file.
You need to set this variable in your recipe.
If you want to list multiple options, separate the options
with a semicolon character (;
).
The
syslinux
class uses this variable to create a set of options.
Specifies the alternate serial port or turns it off.
To turn off serial, set this variable to an empty string
in your recipe.
The variable's default value is set in the
syslinux
class as follows:
SYSLINUX_SERIAL ?= "0 115200"
The class checks for and uses the variable as needed.
An .LSS
file used as the background
for the VGA boot menu when you use the boot menu.
You need to set this variable in your recipe.
The
syslinux
class checks for this variable and if found, the
OpenEmbedded build system installs the splash screen.
Specifies the alternate console=tty... kernel boot argument.
The variable's default value is set in the
syslinux
class as follows:
SYSLINUX_SERIAL_TTY ?= "console=ttyS0,115200"
The class checks for and uses the variable as needed.
Points to the temporary directory under the work directory
(default
"${
WORKDIR
}/sysroot-destdir
")
where the files populated into the sysroot are assembled
during the
do_populate_sysroot
task.
Directories that are staged into the sysroot by the
do_populate_sysroot
task.
By default, the following directories are staged:
SYSROOT_DIRS = " \ ${includedir} \ ${libdir} \ ${base_libdir} \ ${nonarch_base_libdir} \ ${datadir} \ "
Directories that are not staged into the sysroot by the
do_populate_sysroot
task.
You can use this variable to exclude certain subdirectories
of directories listed in
SYSROOT_DIRS
from staging.
By default, the following directories are not staged:
SYSROOT_DIRS_BLACKLIST = " \ ${mandir} \ ${docdir} \ ${infodir} \ ${datadir}/locale \ ${datadir}/applications \ ${datadir}/fonts \ ${datadir}/pixmaps \ "
Extra directories staged into the sysroot by the
do_populate_sysroot
task for -native
recipes, in addition
to those specified in
SYSROOT_DIRS
.
By default, the following extra directories are staged:
SYSROOT_DIRS_NATIVE = " \ ${bindir} \ ${sbindir} \ ${base_bindir} \ ${base_sbindir} \ ${libexecdir} \ ${sysconfdir} \ ${localstatedir} \ "
-native
recipes
run directly from the sysroot
(STAGING_DIR_NATIVE
),
which is why additional directories containing program
executables and supporting files need to be staged.
A list of functions to execute after files are staged into the sysroot. These functions are usually used to apply additional processing on the staged files, or to stage additional files.
When inheriting the
systemd
class, this variable specifies whether the specified service
in
SYSTEMD_SERVICE
should start automatically or not.
By default, the service is enabled to automatically start
at boot time.
The default setting is in the
systemd
class as follows:
SYSTEMD_AUTO_ENABLE ??= "enable"
You can disable the service by setting the variable to "disable".
When
EFI_PROVIDER
is set to "systemd-boot", the
SYSTEMD_BOOT_CFG
variable specifies the
configuration file that should be used.
By default, the
systemd-boot
class sets the SYSTEMD_BOOT_CFG
as
follows:
SYSTEMD_BOOT_CFG ?= "${S}/loader.conf"
For information on Systemd-boot, see the Systemd-boot documentation.
When
EFI_PROVIDER
is set to "systemd-boot", the
SYSTEMD_BOOT_ENTRIES
variable specifies
a list of entry files
(*.conf
) to install that contain
one boot entry per file.
By default, the
systemd-boot
class sets the SYSTEMD_BOOT_ENTRIES
as
follows:
SYSTEMD_BOOT_ENTRIES ?= ""
For information on Systemd-boot, see the Systemd-boot documentation.
When
EFI_PROVIDER
is set to "systemd-boot", the
SYSTEMD_BOOT_TIMEOUT
variable specifies
the boot menu timeout in seconds.
By default, the
systemd-boot
class sets the SYSTEMD_BOOT_TIMEOUT
as
follows:
SYSTEMD_BOOT_TIMEOUT ?= "10"
For information on Systemd-boot, see the Systemd-boot documentation.
When inheriting the
systemd
class, this variable locates the systemd unit files when
they are not found in the main recipe's package.
By default, the
SYSTEMD_PACKAGES
variable is set
such that the systemd unit files are assumed to reside in
the recipes main package:
SYSTEMD_PACKAGES ?= "${PN}"
If these unit files are not in this recipe's main
package, you need to use
SYSTEMD_PACKAGES
to list the package
or packages in which the build system can find the systemd
unit files.
When inheriting the
systemd
class, this variable specifies the systemd service name for
a package.
When you specify this file in your recipe, use a package name override to indicate the package to which the value applies. Here is an example from the connman recipe:
SYSTEMD_SERVICE_${PN} = "connman.service"
When using
SysVinit,
specifies a space-separated list of the virtual terminals
that should run a
getty
(allowing login), assuming
USE_VT
is not set to "0".
The default value for
SYSVINIT_ENABLED_GETTYS
is "1"
(i.e. only run a getty on the first virtual terminal).
This variable points to a directory were BitBake places temporary files, which consist mostly of task logs and scripts, when building a particular recipe. The variable is typically set as follows:
T = "${WORKDIR}/temp"
The WORKDIR
is the directory into which BitBake unpacks and builds the
recipe.
The default bitbake.conf
file sets this variable.
The T
variable is not to be confused with
the TMPDIR
variable,
which points to the root of the directory tree where BitBake
places the output of an entire build.
The target machine's architecture. The OpenEmbedded build system supports many architectures. Here is an example list of architectures supported. This list is by no means complete as the architecture is configurable:
arm i586 x86_64 powerpc powerpc64 mips mipsel
For additional information on machine architectures, see
the
TUNE_ARCH
variable.
Specifies architecture-specific assembler flags for the
target system.
TARGET_AS_ARCH
is initialized from
TUNE_ASARGS
by default in the BitBake configuration file
(meta/conf/bitbake.conf
):
TARGET_AS_ARCH = "${TUNE_ASARGS}"
Specifies architecture-specific C compiler flags for the
target system.
TARGET_CC_ARCH
is initialized from
TUNE_CCARGS
by default.
LDFLAGS
to TARGET_CC_ARCH
in recipes that build software for the target that
would not otherwise respect the exported
LDFLAGS
variable.
This is a specific kernel compiler flag for a CPU or
Application Binary Interface (ABI) tune.
The flag is used rarely and only for cases where a
userspace
TUNE_CCARGS
is not compatible with the kernel compilation.
The TARGET_CC_KERNEL_ARCH
variable
allows the kernel (and associated modules) to use a
different configuration.
See the
meta/conf/machine/include/arm/feature-arm-thumb.inc
file in the
Source Directory
for an example.
Specifies the flags to pass to the C compiler when building
for the target.
When building in the target context,
CFLAGS
is set to the value of this variable by default.
Additionally, the SDK's environment setup script sets
the CFLAGS
variable in the environment
to the TARGET_CFLAGS
value so that
executables built using the SDK also have the flags
applied.
Specifies the flags to pass to the C pre-processor
(i.e. to both the C and the C++ compilers) when building
for the target.
When building in the target context,
CPPFLAGS
is set to the value of this variable by default.
Additionally, the SDK's environment setup script sets
the CPPFLAGS
variable in the
environment to the TARGET_CPPFLAGS
value so that executables built using the SDK also have
the flags applied.
Specifies the flags to pass to the C++ compiler when
building for the target.
When building in the target context,
CXXFLAGS
is set to the value of this variable by default.
Additionally, the SDK's environment setup script sets
the CXXFLAGS
variable in the
environment to the TARGET_CXXFLAGS
value so that executables built using the SDK also have
the flags applied.
Specifies the method for handling FPU code. For FPU-less targets, which include most ARM CPUs, the variable must be set to "soft". If not, the kernel emulation gets used, which results in a performance penalty.
Specifies architecture-specific linker flags for the
target system.
TARGET_LD_ARCH
is initialized from
TUNE_LDARGS
by default in the BitBake configuration file
(meta/conf/bitbake.conf
):
TARGET_LD_ARCH = "${TUNE_LDARGS}"
Specifies the flags to pass to the linker when building
for the target.
When building in the target context,
LDFLAGS
is set to the value of this variable by default.
Additionally, the SDK's environment setup script sets
the
LDFLAGS
variable in the environment to the
TARGET_LDFLAGS
value so that
executables built using the SDK also have the flags
applied.
Specifies the target's operating system. The variable can be set to "linux" for glibc-based systems (GNU C Library) and to "linux-musl" for musl libc. For ARM/EABI targets, "linux-gnueabi" and "linux-musleabi" possible values exist.
Specifies the prefix used for the toolchain binary target tools.
Depending on the type of recipe and the build target,
TARGET_PREFIX
is set as follows:
For recipes building for the target machine, the value is "${TARGET_SYS}-".
For native recipes, the build system sets the
variable to the value of
BUILD_PREFIX
.
For native SDK recipes
(nativesdk
), the
build system sets the variable to the value of
SDK_PREFIX
.
Specifies the system, including the architecture and the operating system, for which the build is occurring in the context of the current recipe.
The OpenEmbedded build system automatically sets this
variable based on
TARGET_ARCH
,
TARGET_VENDOR
,
and
TARGET_OS
variables.
TARGET_SYS
variable yourself.
Consider these two examples:
Given a native recipe on a 32-bit, x86 machine running Linux, the value is "i686-linux".
Given a recipe being built for a little-endian, MIPS target running Linux, the value might be "mipsel-linux".
Specifies the name of the target vendor.
Specifies the GNU standard C library
(libc
) variant to use during the
build process.
This variable replaces POKYLIBC
,
which is no longer supported.
You can select "glibc", "musl", "newlib", or "baremetal"
Specifies a suffix to be appended onto the
TMPDIR
value.
The suffix identifies the libc
variant
for building.
When you are building for multiple variants with the same
Build Directory,
this mechanism ensures that output for different
libc
variants is kept separate to
avoid potential conflicts.
In the defaultsetup.conf
file, the
default value of TCLIBCAPPEND
is
"-${TCLIBC}".
However, distros such as poky, which normally only support
one libc
variant, set
TCLIBCAPPEND
to "" in their distro
configuration file resulting in no suffix being applied.
Specifies the toolchain selector.
TCMODE
controls the characteristics
of the generated packages and images by telling the
OpenEmbedded build system which toolchain profile to use.
By default, the OpenEmbedded build system builds its own
internal toolchain.
The variable's default value is "default", which uses
that internal toolchain.
TCMODE
is set to a value
other than "default", then it is your responsibility
to ensure that the toolchain is compatible with the
default toolchain.
Using older or newer versions of these components
might cause build problems.
See the Release Notes for the Yocto Project release
for the specific components with which the toolchain
must be compatible.
To access the Release Notes, go to the
Downloads
page on the Yocto Project website and click on the
"RELEASE INFORMATION" link for the appropriate
release.
The TCMODE
variable is similar to
TCLIBC
,
which controls the variant of the GNU standard C library
(libc
) used during the build process:
glibc
or musl
.
With additional layers, it is possible to use a pre-compiled
external toolchain.
One example is the Sourcery G++ Toolchain.
The support for this toolchain resides in the separate
Mentor Graphics®
meta-sourcery
layer at
http://github.com/MentorEmbedded/meta-sourcery/.
The layer's README
file contains
information on how to use the Sourcery G++ Toolchain as
an external toolchain.
In summary, you must be sure to add the layer to your
bblayers.conf
file in front of the
meta
layer and then set the
EXTERNAL_TOOLCHAIN
variable in your local.conf
file
to the location in which you installed the toolchain.
The fundamentals used for this example apply to any
external toolchain.
You can use meta-sourcery
as a
template for adding support for other external toolchains.
The location the OpenEmbedded build system uses to export
tests when the
TEST_EXPORT_ONLY
variable is set to "1".
The TEST_EXPORT_DIR
variable defaults
to "${TMPDIR}/testimage/${PN}"
.
Specifies to export the tests only. Set this variable to "1" if you do not want to run the tests but you want them to be exported in a manner that you to run them outside of the build system.
Holds the SSH log and the boot log for QEMU machines.
The TEST_LOG_DIR
variable defaults
to "${WORKDIR}/testimage"
.
log.do_testimage
), which is in
the ${WORKDIR}/temp/
directory.
For automated hardware testing, specifies the command to use to control the power of the target machine under test. Typically, this command would point to a script that performs the appropriate action (e.g. interacting with a web-enabled power strip). The specified command should expect to receive as the last argument "off", "on" or "cycle" specifying to power off, on, or cycle (power off and then power on) the device, respectively.
For automated hardware testing, specifies additional
arguments to pass through to the command specified in
TEST_POWERCONTROL_CMD
.
Setting TEST_POWERCONTROL_EXTRA_ARGS
is optional.
You can use it if you wish, for example, to separate the
machine-specific and non-machine-specific parts of the
arguments.
The time in seconds allowed for an image to boot before
automated runtime tests begin to run against an
image.
The default timeout period to allow the boot process to
reach the login prompt is 500 seconds.
You can specify a different value in the
local.conf
file.
For more information on testing images, see the "Performing Automated Runtime Testing" section in the Yocto Project Development Tasks Manual.
For automated hardware testing, specifies the command to use to connect to the serial console of the target machine under test. This command simply needs to connect to the serial console and forward that connection to standard input and output as any normal terminal program does.
For example, to use the Picocom terminal program on
serial device /dev/ttyUSB0
at
115200bps, you would set the variable as follows:
TEST_SERIALCONTROL_CMD = "picocom /dev/ttyUSB0 -b 115200"
For automated hardware testing, specifies additional
arguments to pass through to the command specified in
TEST_SERIALCONTROL_CMD
.
Setting TEST_SERIALCONTROL_EXTRA_ARGS
is optional.
You can use it if you wish, for example, to separate the
machine-specific and non-machine-specific parts of the
command.
The IP address of the build machine (host machine). This IP address is usually automatically detected. However, if detection fails, this variable needs to be set to the IP address of the build machine (i.e. where the build is taking place).
TEST_SERVER_IP
variable
is only used for a small number of tests such as
the "dnf" test suite, which needs to download
packages from
WORKDIR/oe-rootfs-repo
.
Specifies the target controller to use when running tests against a test image. The default controller to use is "qemu":
TEST_TARGET = "qemu"
A target controller is a class that defines how an
image gets deployed on a target and how a target is started.
A layer can extend the controllers by adding a module
in the layer's /lib/oeqa/controllers
directory and by inheriting the
BaseTarget
class, which is an abstract
class that cannot be used as a value of
TEST_TARGET
.
You can provide the following arguments with
TEST_TARGET
:
"qemu": Boots a QEMU image and runs the tests. See the "Enabling Runtime Tests on QEMU" section in the Yocto Project Development Tasks Manual for more information.
"simpleremote":
Runs the tests on target hardware that is already
up and running.
The hardware can be on the network or it can be
a device running an image on QEMU.
You must also set
TEST_TARGET_IP
when you use "simpleremote".
meta/lib/oeqa/controllers/simpleremote.py
.
For information on running tests on hardware, see the "Enabling Runtime Tests on Hardware" section in the Yocto Project Development Tasks Manual.
The IP address of your hardware under test.
The TEST_TARGET_IP
variable has no
effect when
TEST_TARGET
is set to "qemu".
When you specify the IP address, you can also include a port. Here is an example:
TEST_TARGET_IP = "192.168.1.4:2201"
Specifying a port is useful when SSH is started on a non-standard port or in cases when your hardware under test is behind a firewall or network that is not directly accessible from your host and you need to do port address translation.
An ordered list of tests (modules) to run against an image when performing automated runtime testing.
The OpenEmbedded build system provides a core set of tests that can be used against images.
Tests include ping
,
ssh
, df
among
others.
You can add your own tests to the list of tests by
appending TEST_SUITES
as follows:
TEST_SUITES_append = " mytest
"
Alternatively, you can provide the "auto" option to have all applicable tests run against the image.
TEST_SUITES_append = " auto"
Using this option causes the build system to automatically run tests that are applicable to the image. Tests that are not applicable are skipped.
The order in which tests are run is important.
Tests that depend on another test must appear later in the
list than the test on which they depend.
For example, if you append the list of tests with two
tests (test_A
and
test_B
) where
test_B
is dependent on
test_A
, then you must order the tests
as follows:
TEST_SUITES = " test_A test_B"
For more information on testing images, see the "Performing Automated Runtime Testing" section in the Yocto Project Development Tasks Manual.
Automatically runs the series of automated tests for
images when an image is successfully built.
Setting TESTIMAGE_AUTO
to "1"
causes any image that successfully builds to automatically
boot under QEMU.
Using the variable also adds in dependencies so that any
SDK for which testing is requested is automatically built
first.
These tests are written in Python making use of the
unittest
module, and the majority of
them run commands on the target system over
ssh
.
You can set this variable to "1" in your
local.conf
file in the
Build Directory
to have the OpenEmbedded build system automatically run
these tests after an image successfully builds:
TESTIMAGE_AUTO = "1"
For more information on enabling, running, and writing
these tests, see the
"Performing Automated Runtime Testing"
section in the Yocto Project Development Tasks Manual and
the
"testimage*.bbclass
"
section.
The directory in which the file BitBake is currently parsing is located. Do not manually set this variable.
The time the build was started. Times appear using the hour, minute, and second (HMS) format (e.g. "140159" for one minute and fifty-nine seconds past 1400 hours).
This variable is the base directory the OpenEmbedded
build system uses for all build output and intermediate
files (other than the shared state cache).
By default, the TMPDIR
variable points
to tmp
within the
Build Directory.
If you want to establish this directory in a location other
than the default, you can uncomment and edit the following
statement in the
conf/local.conf
file in the
Source Directory:
#TMPDIR = "${TOPDIR}/tmp"
An example use for this scenario is to set
TMPDIR
to a local disk, which does
not use NFS, while having the Build Directory use NFS.
The filesystem used by TMPDIR
must
have standard filesystem semantics (i.e. mixed-case files
are unique, POSIX file locking, and persistent inodes).
Due to various issues with NFS and bugs in some
implementations, NFS does not meet this minimum
requirement.
Consequently, TMPDIR
cannot be on
NFS.
This variable lists packages the OpenEmbedded build system
uses when building an SDK, which contains a
cross-development environment.
The packages specified by this variable are part of the
toolchain set that runs on the
SDKMACHINE
,
and each package should usually have the prefix
nativesdk-
.
For example, consider the following command when
building an SDK:
$ bitbake -c populate_sdk imagename
In this case, a default list of packages is set in this variable, but you can add additional packages to the list. See the "Adding Individual Packages to the Standard SDK" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual for more information.
For background information on cross-development toolchains in the Yocto Project development environment, see the "Cross-Development Toolchain Generation" section in the Yocto Project Overview and Concepts Manual. For information on setting up a cross-development environment, see the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
This variable defines the name used for the toolchain
output.
The
populate_sdk_base
class sets the
TOOLCHAIN_OUTPUTNAME
variable as
follows:
TOOLCHAIN_OUTPUTNAME ?= "${SDK_NAME}-toolchain-${SDK_VERSION}"
See the
SDK_NAME
and
SDK_VERSION
variables for additional information.
This variable lists packages the OpenEmbedded build system uses when it creates the target part of an SDK (i.e. the part built for the target hardware), which includes libraries and headers. Use this variable to add individual packages to the part of the SDK that runs on the target. See the "Adding Individual Packages to the Standard SDK" section in the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual for more information.
For background information on cross-development toolchains in the Yocto Project development environment, see the "Cross-Development Toolchain Generation" section in the Yocto Project Overview and Concepts Manual. For information on setting up a cross-development environment, see the Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual.
The top-level
Build Directory.
BitBake automatically sets this variable when you
initialize your build environment using
oe-init-build-env
.
A sanitized version of
TARGET_ARCH
.
This variable is used where the architecture is needed in
a value where underscores are not allowed, for example
within package filenames.
In this case, dash characters replace any underscore
characters used in TARGET_ARCH
.
Do not edit this variable.
The GNU canonical architecture for a specific architecture
(i.e. arm
,
armeb
,
mips
,
mips64
, and so forth).
BitBake uses this value to setup configuration.
TUNE_ARCH
definitions are specific to
a given architecture.
The definitions can be a single static definition, or
can be dynamically adjusted.
You can see details for a given CPU family by looking at
the architecture's README
file.
For example, the
meta/conf/machine/include/mips/README
file in the
Source Directory
provides information for TUNE_ARCH
specific to the mips
architecture.
TUNE_ARCH
is tied closely to
TARGET_ARCH
,
which defines the target machine's architecture.
The BitBake configuration file
(meta/conf/bitbake.conf
) sets
TARGET_ARCH
as follows:
TARGET_ARCH = "${TUNE_ARCH}"
The following list, which is by no means complete since architectures are configurable, shows supported machine architectures:
arm i586 x86_64 powerpc powerpc64 mips mipsel
Specifies architecture-specific assembler flags for
the target system.
The set of flags is based on the selected tune features.
TUNE_ASARGS
is set using
the tune include files, which are typically under
meta/conf/machine/include/
and are
influenced through
TUNE_FEATURES
.
For example, the
meta/conf/machine/include/x86/arch-x86.inc
file defines the flags for the x86 architecture as follows:
TUNE_ASARGS += "${@bb.utils.contains("TUNE_FEATURES", "mx32", "-x32", "", d)}"
Specifies architecture-specific C compiler flags for
the target system.
The set of flags is based on the selected tune features.
TUNE_CCARGS
is set using
the tune include files, which are typically under
meta/conf/machine/include/
and are
influenced through
TUNE_FEATURES
.
Specifies architecture-specific linker flags for
the target system.
The set of flags is based on the selected tune features.
TUNE_LDARGS
is set using
the tune include files, which are typically under
meta/conf/machine/include/
and are
influenced through
TUNE_FEATURES
.
For example, the
meta/conf/machine/include/x86/arch-x86.inc
file defines the flags for the x86 architecture as follows:
TUNE_LDARGS += "${@bb.utils.contains("TUNE_FEATURES", "mx32", "-m elf32_x86_64", "", d)}"
Features used to "tune" a compiler for optimal use
given a specific processor.
The features are defined within the tune files and allow
arguments (i.e. TUNE_*ARGS
) to be
dynamically generated based on the features.
The OpenEmbedded build system verifies the features to be sure they are not conflicting and that they are supported.
The BitBake configuration file
(meta/conf/bitbake.conf
) defines
TUNE_FEATURES
as follows:
TUNE_FEATURES ??= "${TUNE_FEATURES_tune-${DEFAULTTUNE}}"
See the
DEFAULTTUNE
variable for more information.
The package architecture understood by the packaging system to define the architecture, ABI, and tuning of output packages. The specific tune is defined using the "_tune" override as follows:
TUNE_PKGARCH_tune-tune
= "tune
"
These tune-specific package architectures are defined in
the machine include files.
Here is an example of the "core2-32" tuning as used
in the
meta/conf/machine/include/tune-core2.inc
file:
TUNE_PKGARCH_tune-core2-32 = "core2-32"
An underlying Application Binary Interface (ABI) used by
a particular tuning in a given toolchain layer.
Providers that use prebuilt libraries can use the
TUNEABI
,
TUNEABI_OVERRIDE
,
and
TUNEABI_WHITELIST
variables to check compatibility of tunings against their
selection of libraries.
If TUNEABI
is undefined, then every
tuning is allowed.
See the
sanity
class to see how the variable is used.
If set, the OpenEmbedded system ignores the
TUNEABI_WHITELIST
variable.
Providers that use prebuilt libraries can use the
TUNEABI_OVERRIDE
,
TUNEABI_WHITELIST
,
and
TUNEABI
variables to check compatibility of a tuning against their
selection of libraries.
See the
sanity
class to see how the variable is used.
A whitelist of permissible
TUNEABI
values.
If TUNEABI_WHITELIST
is not set,
all tunes are allowed.
Providers that use prebuilt libraries can use the
TUNEABI_WHITELIST
,
TUNEABI_OVERRIDE
,
and TUNEABI
variables to check
compatibility of a tuning against their selection of
libraries.
See the
sanity
class to see how the variable is used.
feature
]¶
Specifies CPU or Application Binary Interface (ABI)
tuning features that conflict with feature
.
Known tuning conflicts are specified in the machine include
files in the
Source Directory.
Here is an example from the
meta/conf/machine/include/mips/arch-mips.inc
include file that lists the "o32" and "n64" features as
conflicting with the "n32" feature:
TUNECONFLICTS[n32] = "o32 n64"
feature
]¶
Specifies a valid CPU or Application Binary Interface (ABI)
tuning feature.
The specified feature is stored as a flag.
Valid features are specified in the machine include files
(e.g. meta/conf/machine/include/arm/arch-arm.inc
).
Here is an example from that file:
TUNEVALID[bigendian] = "Enable big-endian mode."
See the machine include files in the Source Directory for these features.
Configures the
UBOOT_MACHINE
and can also define
IMAGE_FSTYPES
for individual cases.
Following is an example from the
meta-fsl-arm
layer.
UBOOT_CONFIG ??= "sd" UBOOT_CONFIG[sd] = "mx6qsabreauto_config,sdcard" UBOOT_CONFIG[eimnor] = "mx6qsabreauto_eimnor_config" UBOOT_CONFIG[nand] = "mx6qsabreauto_nand_config,ubifs" UBOOT_CONFIG[spinor] = "mx6qsabreauto_spinor_config"
In this example, "sd" is selected as the configuration
of the possible four for the
UBOOT_MACHINE
.
The "sd" configuration defines "mx6qsabreauto_config"
as the value for UBOOT_MACHINE
, while
the "sdcard" specifies the
IMAGE_FSTYPES
to use for the U-boot
image.
For more information on how the
UBOOT_CONFIG
is handled, see the
uboot-config
class.
Specifies the entry point for the U-Boot image.
During U-Boot image creation, the
UBOOT_ENTRYPOINT
variable is passed
as a command-line parameter to the
uboot-mkimage
utility.
Specifies the load address for the U-Boot image.
During U-Boot image creation, the
UBOOT_LOADADDRESS
variable is passed
as a command-line parameter to the
uboot-mkimage
utility.
Appends a string to the name of the local version of the U-Boot image. For example, assuming the version of the U-Boot image built was "2013.10", the full version string reported by U-Boot would be "2013.10-yocto" given the following statement:
UBOOT_LOCALVERSION = "-yocto"
Specifies the value passed on the
make
command line when building
a U-Boot image.
The value indicates the target platform configuration.
You typically set this variable from the machine
configuration file (i.e.
conf/machine/
).
machine_name
.conf
Please see the "Selection of Processor Architecture and Board Type" section in the U-Boot README for valid values for this variable.
Specifies the target called in the
Makefile
.
The default target is "all".
Points to the generated U-Boot extension.
For example, u-boot.sb
has a
.sb
extension.
The default U-Boot extension is
.bin
Specifies the target used for building U-Boot. The target is passed directly as part of the "make" command (e.g. SPL and AIS). If you do not specifically set this variable, the OpenEmbedded build process passes and uses "all" for the target during the U-Boot building process.
Specifies a list of options that, if reported by the
configure script as being invalid, should not generate a
warning during the
do_configure
task.
Normally, invalid configure options are simply not passed
to the configure script (e.g. should be removed from
EXTRA_OECONF
or
PACKAGECONFIG_CONFARGS
).
However, common options, for example, exist that are passed
to all configure scripts at a class level that might not
be valid for some configure scripts.
It follows that no benefit exists in seeing a warning about
these options.
For these cases, the options are added to
UNKNOWN_CONFIGURE_WHITELIST
.
The configure arguments check that uses
UNKNOWN_CONFIGURE_WHITELIST
is part
of the
insane
class and is only enabled if the recipe inherits the
autotools
class.
For recipes inheriting the
update-rc.d
class, UPDATERCPN
specifies
the package that contains the initscript that is
enabled.
The default value is "${PN}". Given that almost all recipes that install initscripts package them in the main package for the recipe, you rarely need to set this variable in individual recipes.
You can perform a per-recipe check for what the latest
upstream source code version is by calling
bitbake -c checkpkg
recipe
.
If the recipe source code is provided from Git
repositories, the OpenEmbedded build system determines the
latest upstream version by picking the latest tag from the
list of all repository tags.
You can use the
UPSTREAM_CHECK_GITTAGREGEX
variable to provide a regular expression to filter only the
relevant tags should the default filter not work
correctly.
UPSTREAM_CHECK_GITTAGREGEX = "git_tag_regex"
Use the UPSTREAM_CHECK_REGEX
variable
to specify a different regular expression instead of the
default one when the package checking system is parsing
the page found using
UPSTREAM_CHECK_URI
.
UPSTREAM_CHECK_REGEX = "package_regex"
You can perform a per-recipe check for what the latest
upstream source code version is by calling
bitbake -c checkpkg
recipe
.
If the source code is provided from tarballs, the latest
version is determined by fetching the directory listing
where the tarball is and attempting to find a later tarball.
When this approach does not work, you can use
UPSTREAM_CHECK_URI
to
provide a different URI that contains the link to the
latest tarball.
UPSTREAM_CHECK_URI = "recipe_url"
Determines if devtmpfs
is used for
/dev
population.
The default value used for USE_DEVFS
is "1" when no value is specifically set.
Typically, you would set USE_DEVFS
to "0" for a statically populated /dev
directory.
See the "Selecting a Device Manager" section in the Yocto Project Development Tasks Manual for information on how to use this variable.
When using SysVinit, determines whether or not to run a getty on any virtual terminals in order to enable logging in through those terminals.
The default value used for USE_VT
is "1" when no default value is specifically set.
Typically, you would set USE_VT
to "0" in the machine configuration file for machines
that do not have a graphical display attached and
therefore do not need virtual terminal functionality.
A list of classes to globally inherit.
These classes are used by the OpenEmbedded build system
to enable extra features (e.g.
buildstats
,
image-mklibs
, and so forth).
The default list is set in your
local.conf
file:
USER_CLASSES ?= "buildstats image-mklibs image-prelink"
For more information, see
meta-poky/conf/local.conf.sample
in
the
Source Directory.
If set to error
, forces the
OpenEmbedded build system to produce an error if the user
identification (uid
) and group
identification (gid
) values are not
defined in any of the files listed
in USERADD_UID_TABLES
and USERADD_GID_TABLES
. If
set to warn
, a warning will be issued
instead.
The default behavior for the build system is to dynamically
apply uid
and
gid
values.
Consequently, the USERADD_ERROR_DYNAMIC
variable is by default not set.
If you plan on using statically assigned
gid
and uid
values, you should set
the USERADD_ERROR_DYNAMIC
variable in
your local.conf
file as
follows:
USERADD_ERROR_DYNAMIC = "error"
Overriding the default behavior implies you are going to
also take steps to set static uid
and
gid
values through use of the
USERADDEXTENSION
,
USERADD_UID_TABLES
,
and
USERADD_GID_TABLES
variables.
USERADD_ERROR_DYNAMIC
to error
and setting it
to warn
. When it is set
to warn
, the build system will report a
warning for every undefined uid
and
gid
in any recipe. But when it is set
to error
, it will only report errors
for recipes that are actually built. This saves you from
having to add static IDs for recipes that you know will
never be built.
Specifies a password file to use for obtaining static
group identification (gid
) values
when the OpenEmbedded build system adds a group to the
system during package installation.
When applying static group identification
(gid
) values, the OpenEmbedded build
system looks in
BBPATH
for a files/group
file and then applies
those uid
values.
Set the variable as follows in your
local.conf
file:
USERADD_GID_TABLES = "files/group"
USERADDEXTENSION
variable to "useradd-staticids" causes the build system
to use static gid
values.
When inheriting the
useradd
class, this variable
specifies the individual packages within the recipe that
require users and/or groups to be added.
You must set this variable if the recipe inherits the class. For example, the following enables adding a user for the main package in a recipe:
USERADD_PACKAGES = "${PN}"
USERADD_PACKAGES
variable,
you need to set one or more of the
USERADD_PARAM
,
GROUPADD_PARAM
,
or
GROUPMEMS_PARAM
variables.
When inheriting the
useradd
class, this variable
specifies for a package what parameters should pass
to the useradd
command
if you add a user to the system when the package
is installed.
Here is an example from the dbus
recipe:
USERADD_PARAM_${PN} = "--system --home ${localstatedir}/lib/dbus \ --no-create-home --shell /bin/false \ --user-group messagebus"
For information on the standard Linux shell command
useradd
, see
http://linux.die.net/man/8/useradd.
Specifies a password file to use for obtaining static
user identification (uid
) values
when the OpenEmbedded build system adds a user to the
system during package installation.
When applying static user identification
(uid
) values, the OpenEmbedded build
system looks in
BBPATH
for a files/passwd
file and then applies
those uid
values.
Set the variable as follows in your
local.conf
file:
USERADD_UID_TABLES = "files/passwd"
USERADDEXTENSION
variable to "useradd-staticids" causes the build system
to use static uid
values.
When set to "useradd-staticids", causes the
OpenEmbedded build system to base all user and group
additions on a static
passwd
and
group
files found in
BBPATH
.
To use static user identification (uid
)
and group identification (gid
)
values, set the variable
as follows in your local.conf
file:
USERADDEXTENSION = "useradd-staticids"
uid
and gid
values causes the OpenEmbedded build system to employ
the
useradd-staticids
class.
If you use static uid
and
gid
information, you must also
specify the files/passwd
and
files/group
files by setting the
USERADD_UID_TABLES
and
USERADD_GID_TABLES
variables.
Additionally, you should also set the
USERADD_ERROR_DYNAMIC
variable.
Specifies the persistence of the target's
/var/log
directory, which is used to
house postinstall target log files.
By default, VOLATILE_LOG_DIR
is set
to "yes", which means the file is not persistent.
You can override this setting by setting the
variable to "no" to make the log directory persistent.
Specifies the quality assurance checks whose failures are
reported as warnings by the OpenEmbedded build system.
You set this variable in your distribution configuration
file.
For a list of the checks you can control with this variable,
see the
"insane.bbclass
"
section.
When placed in the recipe that builds your image, this
variable lists build-time dependencies.
The WKS_FILE_DEPENDS
variable is only
applicable when Wic images are active (i.e. when
IMAGE_FSTYPES
contains entries related to Wic).
If your recipe does not create Wic images, the variable
has no effect.
The WKS_FILE_DEPENDS
variable is
similar to the
DEPENDS
variable.
When you use the variable in your recipe that builds the
Wic image, dependencies you list in the
WIC_FILE_DEPENDS
variable are added to
the DEPENDS
variable.
With the WKS_FILE_DEPENDS
variable,
you have the possibility to specify a list of additional
dependencies (e.g. native tools, bootloaders, and so forth),
that are required to build Wic images.
Following is an example:
WKS_FILE_DEPENDS = "some-native-tool
"
In the previous example,
some-native-tool
would be
replaced with an actual native tool on which the build
would depend.
Specifies the location of the Wic
kickstart file that is used by the OpenEmbedded build
system to create a partitioned image
(image
.wic
).
For information on how to create a partitioned image, see
the
"Creating Partitioned Images Using Wic"
section in the Yocto Project Development Tasks Manual.
For details on the kickstart file format, see the
"OpenEmbedded Kickstart (.wks
) Reference"
Chapter.
The pathname of the work directory in which the OpenEmbedded
build system builds a recipe.
This directory is located within the
TMPDIR
directory structure and is specific to the recipe being
built and the system for which it is being built.
The WORKDIR
directory is defined as
follows:
${TMPDIR}/work/${MULTIMACH_TARGET_SYS}/${PN}/${EXTENDPE}${PV}-${PR}
The actual directory depends on several things:
As an example, assume a Source Directory top-level folder
name poky
, a default Build Directory at
poky/build
, and a
qemux86-poky-linux
machine target
system.
Furthermore, suppose your recipe is named
foo_1.3.0-r0.bb
.
In this case, the work directory the build system uses to
build the package would be as follows:
poky/build/tmp/work/qemux86-poky-linux/foo/1.3.0-r0
Specifies the packages that should be installed to
provide an X server and drivers for the current machine,
assuming your image directly includes
packagegroup-core-x11-xserver
or,
perhaps indirectly, includes "x11-base" in
IMAGE_FEATURES
.
The default value of XSERVER
, if not
specified in the machine configuration, is
"xserver-xorg xf86-video-fbdev xf86-input-evdev".
While you can use most variables in almost any context such as
.conf
, .bbclass
,
.inc
, and .bb
files,
some variables are often associated with a particular locality or context.
This chapter describes some common associations.
The following subsections provide lists of variables whose context is configuration: distribution, machine, and local.
This section lists variables whose configuration context is the distribution, or distro.
This section lists variables whose configuration context is the
local configuration through the local.conf
file.
The following subsections provide lists of variables whose context is recipes: required, dependencies, path, and extra build information.
This section lists variables that are required for recipes.
SRC_URI
- used
in recipes that fetch local or remote files.
This section lists variables that define extra build information for recipes.
35.1. | How does Poky differ from OpenEmbedded? |
The term "Poky" refers to the specific reference build system that the Yocto Project provides. Poky is based on OE-Core and BitBake. Thus, the generic term used here for the build system is the "OpenEmbedded build system." Development in the Yocto Project using Poky is closely tied to OpenEmbedded, with changes always being merged to OE-Core or BitBake first before being pulled back into Poky. This practice benefits both projects immediately. | |
35.2. | My development system does not meet the required Git, tar, and Python versions. In particular, I do not have Python 3.5.0 or greater. Can I still use the Yocto Project? |
You can get the required tools on your host development system a couple different ways (i.e. building a tarball or downloading a tarball). See the "Required Git, tar, Python and gcc Versions" section for steps on how to update your build tools. | |
35.3. | How can you claim Poky / OpenEmbedded-Core is stable? |
There are three areas that help with stability;
| |
35.4. | How do I get support for my board added to the Yocto Project? |
Support for an additional board is added by creating a Board Support Package (BSP) layer for it. For more information on how to create a BSP layer, see the "Understanding and Creating Layers" section in the Yocto Project Development Tasks Manual and the Yocto Project Board Support Package (BSP) Developer's Guide. Usually, if the board is not completely exotic, adding support in the Yocto Project is fairly straightforward. | |
35.5. | Are there any products built using the OpenEmbedded build system? |
The software running on the Vernier LabQuest is built using the OpenEmbedded build system. See the Vernier LabQuest website for more information. There are a number of pre-production devices using the OpenEmbedded build system and the Yocto Project team announces them as soon as they are released. | |
35.6. | What does the OpenEmbedded build system produce as output? |
Because you can use the same set of recipes to create output of various formats, the output of an OpenEmbedded build depends on how you start it. Usually, the output is a flashable image ready for the target device. | |
35.7. | How do I add my package to the Yocto Project? |
To add a package, you need to create a BitBake recipe. For information on how to create a BitBake recipe, see the "Writing a New Recipe" section in the Yocto Project Development Tasks Manual. | |
35.8. | Do I have to reflash my entire board with a new Yocto Project image when recompiling a package? |
The OpenEmbedded build system can build packages in various
formats such as IPK for OPKG, Debian package
( | |
35.9. |
I see the error ' |
You are probably running the build on an NTFS filesystem.
Use | |
35.10. | I see lots of 404 responses for files when the OpenEmbedded build system is trying to download sources. Is something wrong? |
Nothing is wrong. The OpenEmbedded build system checks any configured source mirrors before downloading from the upstream sources. The build system does this searching for both source archives and pre-checked out versions of SCM-managed software. These checks help in large installations because it can reduce load on the SCM servers themselves. The address above is one of the default mirrors configured into the build system. Consequently, if an upstream source disappears, the team can place sources there so builds continue to work. | |
35.11. | I have machine-specific data in a package for one machine only but the package is being marked as machine-specific in all cases, how do I prevent this? |
Set | |
35.12. | I'm behind a firewall and need to use a proxy server. How do I do that? |
Most source fetching by the OpenEmbedded build system is done
by
Following is the applicable code for setting various proxy
types in the # You can set the default proxies for Wget to use for http, https, and ftp. # They will override the value in the environment. #https_proxy = http://proxy.yoyodyne.com:18023/ #http_proxy = http://proxy.yoyodyne.com:18023/ #ftp_proxy = http://proxy.yoyodyne.com:18023/ # If you do not want to use proxy at all, set this to off. #use_proxy = on
The Yocto Project also includes a
| |
35.13. |
What’s the difference between |
The | |
35.14. | I'm seeing random build failures. Help?! |
If the same build is failing in totally different and random ways, the most likely explanation is:
The OpenEmbedded build system processes a massive amount of data that causes lots of network, disk and CPU activity and is sensitive to even single-bit failures in any of these areas. True random failures have always been traced back to hardware or virtualization issues. | |
35.15. |
When I try to build a native recipe, the build fails with |
If you get an error message that indicates GNU
#error GNU libiconv not in use but included iconv.h is from libiconv If you find a previously installed file, you should either uninstall it or temporarily rename it and try the build again.
This issue is just a single manifestation of "system
leakage" issues caused when the OpenEmbedded build system
finds and uses previously installed files during a native
build.
This type of issue might not be limited to
| |
35.16. | What do we need to ship for license compliance? |
This is a difficult question and you need to consult your lawyer for the answer for your specific case. It is worth bearing in mind that for GPL compliance, there needs to be enough information shipped to allow someone else to rebuild and produce the same end result you are shipping. This means sharing the source code, any patches applied to it, and also any configuration information about how that package was configured and built. You can find more information on licensing in the "Licensing" section in the Yocto Project Overview and Concepts Manual and also in the "Maintaining Open Source License Compliance During Your Product's Lifecycle" section in the Yocto Project Development Tasks Manual. | |
35.17. | How do I disable the cursor on my touchscreen device? |
You need to create a form factor file as described in the
"Miscellaneous BSP-Specific Recipe Files"
section in the Yocto Project Board Support Packages (BSP)
Developer's Guide.
Set the HAVE_TOUCHSCREEN=1
| |
35.18. | How do I make sure connected network interfaces are brought up by default? |
The default interfaces file provided by the netbase recipe does not automatically bring up network interfaces. Therefore, you will need to add a BSP-specific netbase that includes an interfaces file. See the "Miscellaneous BSP-Specific Recipe Files" section in the Yocto Project Board Support Packages (BSP) Developer's Guide for information on creating these types of miscellaneous recipe files. For example, add the following files to your layer: meta-MACHINE/recipes-bsp/netbase/netbase/MACHINE/interfaces meta-MACHINE/recipes-bsp/netbase/netbase_5.0.bbappend
| |
35.19. | How do I create images with more free space? |
By default, the OpenEmbedded build system creates images that are 1.3 times the size of the populated root filesystem. To affect the image size, you need to set various configurations:
| |
35.20. | Why don't you support directories with spaces in the pathnames? |
The Yocto Project team has tried to do this before but too
many of the tools the OpenEmbedded build system depends on,
such as | |
35.21. | How do I use an external toolchain? |
The toolchain configuration is very flexible and customizable.
It is primarily controlled with the
The default value of
In addition to the toolchain configuration, you also need a
corresponding toolchain recipe file.
This recipe file needs to package up any pre-built objects in
the toolchain such as | |
35.22. | How does the OpenEmbedded build system obtain source code and will it work behind my firewall or proxy server? |
The way the build system obtains source code is highly configurable. You can setup the build system to get source code in most environments if HTTP transport is available.
When the build system searches for source code, it first
tries the local download directory.
If that location fails, Poky tries
Assuming your distribution is "poky", the OpenEmbedded build
system uses the Yocto Project source
As an example, you could add a specific server for the
build system to attempt before any others by adding something
like the following to the PREMIRRORS_prepend = "\ git://.*/.* http://www.yoctoproject.org/sources/ \n \ ftp://.*/.* http://www.yoctoproject.org/sources/ \n \ http://.*/.* http://www.yoctoproject.org/sources/ \n \ https://.*/.* http://www.yoctoproject.org/sources/ \n"
These changes cause the build system to intercept Git, FTP,
HTTP, and HTTPS requests and direct them to the
Aside from the previous technique, these options also exist: BB_NO_NETWORK = "1" This statement tells BitBake to issue an error instead of trying to access the Internet. This technique is useful if you want to ensure code builds only from local sources. Here is another technique: BB_FETCH_PREMIRRORONLY = "1"
This statement limits the build system to pulling source
from the Here is another technique: BB_GENERATE_MIRROR_TARBALLS = "1" This statement tells the build system to generate mirror tarballs. This technique is useful if you want to create a mirror server. If not, however, the technique can simply waste time during the build.
Finally, consider an example where you are behind an
HTTP-only firewall.
You could make the following changes to the
PREMIRRORS_prepend = "\ ftp://.*/.* http://www.yoctoproject.org/sources/ \n \ http://.*/.* http://www.yoctoproject.org/sources/ \n \ https://.*/.* http://www.yoctoproject.org/sources/ \n" BB_FETCH_PREMIRRORONLY = "1"
These changes would cause the build system to successfully
fetch source over HTTP and any network accesses to anything
other than the
The build system also honors the standard shell environment
variables NoteYou can find more information on the "Working Behind a Network Proxy" Wiki page. | |
35.23. | Can I get rid of build output so I can start over? |
Yes - you can easily do this.
When you use BitBake to build an image, all the build output
goes into the directory created when you run the
build environment setup script (i.e.
Within the Build Directory, is the | |
35.24. |
Why do |
Executables and libraries might need to be used from a directory other than the directory into which they were initially installed. Complicating this situation is the fact that sometimes these executables and libraries are compiled with the expectation of being run from that initial installation target directory. If this is the case, moving them causes problems.
This scenario is a fundamental problem for package maintainers
of mainstream Linux distributions as well as for the
OpenEmbedded build system.
As such, a well-established solution exists.
Makefiles, Autotools configuration scripts, and other build
systems are expected to respect environment variables such as
When the OpenEmbedded build system uses a recipe to build a
target-architecture program (i.e. one that is intended for
inclusion on the image being built), that program eventually
runs from the root file system of that image.
Thus, the build system provides a value of "/usr/bin" for
Meanwhile, NoteDue to these lengthy examples, the paths are artificially broken across lines for readability.
/home/maxtothemax/poky-bootchart2/build/tmp/work/i586-poky-linux/zlib/ 1.2.8-r0/sysroot-destdir/usr/bin /home/maxtothemax/poky-bootchart2/build/tmp/work/x86_64-linux/ zlib-native/1.2.8-r0/sysroot-destdir/home/maxtothemax/poky-bootchart2/ build/tmp/sysroots/x86_64-linux/usr/bin
Even if the paths look unusual, they both are correct -
the first for a target and the second for a native recipe.
These paths are a consequence of the
| |
35.25. |
The files provided by my |
This situation results when a build system does
not recognize the environment variables supplied to it by
BitBake.
The incident that prompted this FAQ entry involved a Makefile
that used an environment variable named
|
The Yocto Project team is happy for people to experiment with the Yocto Project. A number of places exist to find help if you run into difficulties or find bugs. This presents information about contributing and participating in the Yocto Project.
The Yocto Project gladly accepts contributions. You can submit changes to the project either by creating and sending pull requests, or by submitting patches through email. For information on how to do both as well as information on how to identify the maintainer for each area of code, see the "Submitting a Change to the Yocto Project" section in the Yocto Project Development Tasks Manual.
The Yocto Project uses its own implementation of Bugzilla to track defects (bugs). Implementations of Bugzilla work well for group development because they track bugs and code changes, can be used to communicate changes and problems with developers, can be used to submit and review patches, and can be used to manage quality assurance.
Sometimes it is helpful to submit, investigate, or track a bug against the Yocto Project itself (e.g. when discovering an issue with some component of the build system that acts contrary to the documentation or your expectations).
A general procedure and guidelines exist for when you use Bugzilla to submit a bug. For information on how to use Bugzilla to submit a bug against the Yocto Project, see the following:
The "Submitting a Defect Against the Yocto Project" section in the Yocto Project Development Tasks Manual.
The Yocto Project Bugzilla wiki page
For information on Bugzilla in general, see http://www.bugzilla.org/about/.
A number of mailing lists maintained by the Yocto Project exist as well as related OpenEmbedded mailing lists for discussion, patch submission and announcements. To subscribe to one of the following mailing lists, click on the appropriate URL in the following list and follow the instructions:
http://lists.yoctoproject.org/listinfo/yocto - General Yocto Project discussion mailing list.
http://lists.openembedded.org/mailman/listinfo/openembedded-core - Discussion mailing list about OpenEmbedded-Core (the core metadata).
http://lists.openembedded.org/mailman/listinfo/openembedded-devel - Discussion mailing list about OpenEmbedded.
http://lists.openembedded.org/mailman/listinfo/bitbake-devel - Discussion mailing list about the BitBake build tool.
http://lists.yoctoproject.org/listinfo/poky - Discussion mailing list about Poky.
http://lists.yoctoproject.org/listinfo/yocto-announce - Mailing list to receive official Yocto Project release and milestone announcements.
For more Yocto Project-related mailing lists, see the Yocto Project Website.
Two IRC channels on freenode are available for the Yocto Project and Poky discussions:
#yocto
#poky
Here is a list of resources you might find helpful:
The Yocto Project website: The home site for the Yocto Project.
The Yocto Project Main Wiki Page: The main wiki page for the Yocto Project. This page contains information about project planning, release engineering, QA & automation, a reference site map, and other resources related to the Yocto Project.
OpenEmbedded: The build system used by the Yocto Project. This project is the upstream, generic, embedded distribution from which the Yocto Project derives its build system (Poky) and to which it contributes.
BitBake: The tool used to process metadata.
BitBake User Manual: A comprehensive guide to the BitBake tool. If you want information on BitBake, see this manual.
Yocto Project Quick Build: This short document lets you experience building an image using the Yocto Project without having to understand any concepts or details.
Yocto project Overview and Concepts Manual: This manual provides overview and conceptual information about the Yocto Project.
Yocto Project Development Tasks Manual: This manual is a "how-to" guide that presents procedures useful to both application and system developers who use the Yocto Project.
Yocto Project Application Development and the Extensible Software Development Kit (eSDK) manual: This guide provides information that lets you get going with the standard or extensible SDK. An SDK, with its cross-development toolchains, allows you to develop projects inside or outside of the Yocto Project environment.
Yocto Project Board Support Package (BSP) Developer's Guide: This guide defines the structure for BSP components. Having a commonly understood structure encourages standardization.
Yocto Project Linux Kernel Development Manual: This manual describes how to work with Linux Yocto kernels as well as provides a bit of conceptual information on the construction of the Yocto Linux kernel tree.
Yocto Project Reference Manual: This manual provides reference material such as variable, task, and class descriptions.
Yocto Project Mega-Manual: This manual is simply a single HTML file comprised of the bulk of the Yocto Project manuals. The Mega-Manual primarily exists as a vehicle by which you can easily search for phrases and terms used in the Yocto Project documentation set.
Yocto Project Profiling and Tracing Manual: This manual presents a set of common and generally useful tracing and profiling schemes along with their applications (as appropriate) to each tool.
Toaster User Manual: This manual introduces and describes how to set up and use Toaster. Toaster is an Application Programming Interface (API) and web-based interface to the OpenEmbedded Build System, which uses BitBake, that reports build information.
FAQ: A list of commonly asked questions and their answers.
Release Notes: Features, updates and known issues for the current release of the Yocto Project. To access the Release Notes, go to the Downloads page on the Yocto Project website and click on the "RELEASE INFORMATION" link for the appropriate release.
Bugzilla: The bug tracking application the Yocto Project uses. If you find problems with the Yocto Project, you should report them using this application.
Bugzilla Configuration and Bug Tracking Wiki Page: Information on how to get set up and use the Yocto Project implementation of Bugzilla for logging and tracking Yocto Project defects.
Internet Relay Chat (IRC):
Two IRC channels on freenode are available
for Yocto Project and Poky discussions: #yocto
and
#poky
, respectively.
Quick EMUlator (QEMU): An open-source machine emulator and virtualizer.
Toaster is a web interface to the Yocto Project's OpenEmbedded build system. The interface enables you to configure and run your builds. Information about builds is collected and stored in a database. You can use Toaster to configure and start builds on multiple remote build servers.
Toaster allows you to configure and run builds, and it provides extensive information about the build process.
Configure and Run Builds: You can use the Toaster web interface to configure and start your builds. Builds started using the Toaster web interface are organized into projects. When you create a project, you are asked to select a release, or version of the build system you want to use for the project builds. As shipped, Toaster supports Yocto Project releases 1.8 and beyond. With the Toaster web interface, you can:
Browse layers listed in the various layer sources that are available in your project (e.g. the OpenEmbedded Layer Index at http://layers.openembedded.org/layerindex/).
Browse images, recipes, and machines provided by those layers.
Import your own layers for building.
Add and remove layers from your configuration.
Set configuration variables.
Select a target or multiple targets to build.
Start your builds.
Toaster also allows you to configure and run your builds from the command line, and switch between the command line and the web interface at any time. Builds started from the command line appear within a special Toaster project called "Command line builds".
Information About the Build Process: Toaster also records extensive information about your builds. Toaster collects data for builds you start from the web interface and from the command line as long as Toaster is running.
With Toaster you can:
See what was built (recipes and packages) and what packages were installed into your final image.
Browse the directory structure of your image.
See the value of all variables in your build configuration, and which files set each value.
Examine error, warning, and trace messages to aid in debugging.
See information about the BitBake tasks executed and reused during your build, including those that used shared state.
See dependency relationships between recipes, packages, and tasks.
See performance information such as build time, task time, CPU usage, and disk I/O.
For an overview of Toaster shipped with the Yocto Project 3.1.2 Release, see the "Toaster - Yocto Project 2.2" video.
You can set Toaster up to run as a local instance or as a shared hosted service.
When Toaster is set up as a local instance, all the components reside on a single build host. Fundamentally, a local instance of Toaster is suited for a single user developing on a single build host.
Toaster as a hosted service is suited for multiple users developing across several build hosts. When Toaster is set up as a hosted service, its components can be spread across several machines:
This chapter describes how you need to prepare your system in order to use Toaster.
Before you can use Toaster, you need to first set up your build system to run the Yocto Project. To do this, follow the instructions in the "Preparing the Build Host" section of the Yocto Project Development Tasks Manual. For Ubuntu/Debian, you might also need to do an additional install of pip3.
$ sudo apt-get install python3-pip
Toaster requires extra Python dependencies in order to run.
A Toaster requirements file named
toaster-requirements.txt
defines the
Python dependencies.
The requirements file is located in the
bitbake
directory, which is located in the
root directory of the
Source Directory
(e.g. poky/bitbake/toaster-requirements.txt
).
The dependencies appear in a pip
,
install-compatible format.
You need to install the packages that Toaster requires. Use this command:
$ pip3 install --user -r bitbake/toaster-requirements.txt
The previous command installs the necessary Toaster modules
into a local python 3 cache in your
$HOME
directory.
The caches is actually located in
$HOME/.local
.
To see what packages have been installed into your
$HOME
directory, do the following:
$ pip3 list installed --local
If you need to remove something, the following works:
$ pip3 uninstall PackageNameToUninstall
Once you have set up the Yocto Project and installed the Toaster system dependencies as described in the "Preparing to Use Toaster" chapter, you are ready to start Toaster.
Navigate to the root of your
Source Directory
(e.g. poky
):
$ cd poky
Once in that directory, source the build environment script:
$ source oe-init-build-env
Next, from the build directory (e.g.
poky/build
), start Toaster using this
command:
$ source toaster start
You can now run your builds from the command line, or with Toaster as explained in section "Using the Toaster Web Interface".
To access the Toaster web interface, open your favorite browser and enter the following:
http://127.0.0.1:8000
By default, Toaster starts on port 8000.
You can use the WEBPORT
parameter to
set a different port.
For example, the following command sets the port to "8400":
$ source toaster start webport=8400
You can start a Toaster environment without starting its web server. This is useful for the following:
Capturing a command-line build’s statistics into the Toaster database for examination later.
Capturing a command-line build’s statistics when the Toaster server is already running.
Having one instance of the Toaster web server track and capture multiple command-line builds, where each build is started in its own “noweb” Toaster environment.
The following commands show how to start a Toaster environment without starting its web server, perform BitBake operations, and then shut down the Toaster environment. Once the build is complete, you can close the Toaster environment. Before closing the environment, however, you should allow a few minutes to ensure the complete transfer of its BitBake build statistics to the Toaster database. If you have a separate Toaster web server instance running, you can watch this command-line build’s progress and examine the results as soon as they are posted:
$ source toaster start noweb
$ bitbake target
$ source toaster stop
You can start a Toaster environment with the “New Projects” feature disabled. Doing so is useful for the following:
Sharing your build results over the web server while blocking others from starting builds on your host.
Allowing only local command-line builds to be captured into the Toaster database.
Use the following command to set up Toaster without a build server:
$ source toaster start nobuild webport=port
By default, Toaster binds to the loop back address
(i.e. localhost), which does not allow access from
external hosts. To allow external access, use the
WEBPORT
parameter to open an
address that connects to the network, specifically the
IP address that your NIC uses to connect to the network.
You can also bind to all IP addresses the computer
supports by using the shortcut
"0.0.0.0:port
".
The following example binds to all IP addresses on the host:
$ source toaster start webport=0.0.0.0:8400
This example binds to a specific IP address on the host's NIC:
$ source toaster start webport=192.168.1.1:8400
Toaster creates a _toaster_clones
directory inside your Source Directory
(i.e. poky
) to clone any layers
needed for your builds.
Alternatively, if you would like all of your Toaster related
files and directories to be in a particular location other than
the default, you can set the TOASTER_DIR
environment variable, which takes precedence over your current
working directory.
Setting this environment variable causes Toaster to create and use
$TOASTER_DIR./_toaster_clones
.
Toaster creates a build directory within your Source
Directory (e.g. poky
) to execute
the builds.
Alternatively, if you would like all of your Toaster related files
and directories to be in a particular location, you can set
the TOASTER_DIR
environment variable,
which takes precedence over your current working directory.
Setting this environment variable causes Toaster to use
$TOASTER_DIR/build
as the build directory.
Toaster is built on the Django framework. Django provides an administration interface you can use to edit Toaster configuration parameters.
To access the Django administration interface, you must create a superuser by following these steps:
If you used pip3
, which is
recommended, to set up the Toaster system dependencies,
you need be sure the local user path is in your
PATH
list.
To append the pip3 local user path, use the following
command:
$ export PATH=$PATH:$HOME/.local/bin
From the directory containing the Toaster database,
which by default is the
Build Directory,
invoke the createsuperuser
command
from manage.py
:
$ cd ~/poky/build $ ../bitbake/lib/toaster/manage.py createsuperuser
Django prompts you for the username, which you need to provide.
Django prompts you for an email address, which is optional.
Django prompts you for a password, which you must provide.
Django prompts you to re-enter your password for verification.
After completing these steps, the following confirmation message appears:
Superuser created successfully.
Creating a superuser allows you to access the Django administration interface through a browser. The URL for this interface is the same as the URL used for the Toaster instance with "/admin" on the end. For example, if you are running Toaster locally, use the following URL:
http://127.0.0.1:8000/admin
You can use the Django administration interface to set Toaster configuration parameters such as the build directory, layer sources, default variable values, and BitBake versions.
You can use a production instance of Toaster to share the Toaster instance with remote users, multiple users, or both. The production instance is also the setup that can handle heavier loads on the web service. Use the instructions in the following sections to set up Toaster to run builds through the Toaster web interface.
Be sure you meet the following requirements:
mod-wsgi
, and Mysql requirements.
Have all the build requirements as described in the "Preparing to Use Toaster" chapter.
Have an Apache webserver.
Have mod-wsgi
for the Apache
webserver.
Use the Mysql database server.
If you are using Ubuntu 16.04, run the following:
$ sudo apt-get install apache2 libapache2-mod-wsgi-py3 mysql-server python3-pip libmysqlclient-dev
If you are using Fedora 24 or a RedHat distribution, run the following:
$ sudo dnf install httpd python3-mod_wsgi python3-pip mariadb-server mariadb-devel python3-devel
If you are using openSUSE Leap 42.1, run the following:
$ sudo zypper install apache2 apache2-mod_wsgi-python3 python3-pip mariadb mariadb-client python3-devel
Perform the following steps to install Toaster:
Create toaster user and set its home directory to
/var/www/toaster
:
$ sudo /usr/sbin/useradd toaster -md /var/www/toaster -s /bin/false $ sudo su - toaster -s /bin/bash
Checkout a copy of poky
into the web server directory.
You will be using /var/www/toaster
:
$ git clone git://git.yoctoproject.org/poky $ git checkout dunfell
Install Toaster dependencies using the --user flag which keeps the Python packages isolated from your system-provided packages:
$ cd /var/www/toaster/ $ pip3 install --user -r ./poky/bitbake/toaster-requirements.txt $ pip3 install --user mysqlclient
Configure Toaster by editing
/var/www/toaster/poky/bitbake/lib/toaster/toastermain/settings.py
as follows:
Edit the DATABASES settings:
DATABASES = { 'default': { 'ENGINE': 'django.db.backends.mysql', 'NAME': 'toaster_data', 'USER': 'toaster', 'PASSWORD': 'yourpasswordhere', 'HOST': 'localhost', 'PORT': '3306', } }
Edit the SECRET_KEY:
SECRET_KEY = 'your_secret_key
'
Edit the STATIC_ROOT:
STATIC_ROOT = '/var/www/toaster/static_files/'
Add the database and user to the mysql
server defined earlier:
$ mysql -u root -p mysql> CREATE DATABASE toaster_data; mysql> CREATE USER 'toaster'@'localhost' identified by 'yourpasswordhere'; mysql> GRANT all on toaster_data.* to 'toaster'@'localhost'; mysql> quit
Get Toaster to create the database schema, default data, and gather the statically-served files:
$ cd /var/www/toaster/poky/ $ ./bitbake/lib/toaster/manage.py migrate $ TOASTER_DIR=`pwd` TEMPLATECONF='poky' \ ./bitbake/lib/toaster/manage.py checksettings $ ./bitbake/lib/toaster/manage.py collectstatic
In the previous example, from the poky
directory, the migrate
command
ensures the database schema changes have propagated
correctly (i.e. migrations).
The next line sets the Toaster root directory
TOASTER_DIR
and the location
of the Toaster configuration file
TOASTER_CONF
, which is relative to
TOASTER_DIR
.
The TEMPLATECONF
value reflects the
contents of poky/.templateconf
, and
by default, should include the string "poky".
For more information on the Toaster configuration
file, see the
"Configuring Toaster"
section.
This line also runs the checksettings
command, which configures the location of the Toaster
Build Directory.
The Toaster root directory TOASTER_DIR
determines where the Toaster build directory
is created on the file system.
In the example above,
TOASTER_DIR
is set as follows:
/var/www/toaster/poky
This setting causes the Toaster build directory to be:
/var/www/toaster/poky/build
Finally, the collectstatic
command
is a Django framework command that collects all the
statically served files into a designated directory to
be served up by the Apache web server as defined by
STATIC_ROOT
.
Test and/or use the Mysql integration with Toaster’s Django web server. At this point, you can start up the normal Toaster Django web server with the Toaster database in Mysql. You can use this web server to confirm that the database migration and data population from the Layer Index is complete.
To start the default Toaster Django web server with the Toaster database now in Mysql, use the standard start commands:
$ source oe-init-build-env $ source toaster start
Additionally, if Django is sufficient for your requirements, you can use it for your release system and migrate later to Apache as your requirements change.
Add an Apache configuration file for Toaster to your Apache web server's configuration directory. If you are using Ubuntu or Debian, put the file here:
/etc/apache2/conf-available/toaster.conf
If you are using Fedora or RedHat, put it here:
/etc/httpd/conf.d/toaster.conf
If you are using OpenSUSE, put it here:
/etc/apache2/conf.d/toaster.conf
Following is a sample Apache configuration for Toaster you can follow:
Alias /static /var/www/toaster/static_files <Directory /var/www/toaster/static_files> <IfModule mod_access_compat.c> Order allow,deny Allow from all </IfModule> <IfModule !mod_access_compat.c> Require all granted </IfModule> </Directory> <Directory /var/www/toaster/poky/bitbake/lib/toaster/toastermain> <Files "wsgi.py"> Require all granted </Files> </Directory> WSGIDaemonProcess toaster_wsgi python-path=/var/www/toaster/poky/bitbake/lib/toaster:/var/www/toaster/.local/lib/python3.4/site-packages WSGIScriptAlias / "/var/www/toaster/poky/bitbake/lib/toaster/toastermain/wsgi.py" <Location /> WSGIProcessGroup toaster_wsgi </Location>
If you are using Ubuntu or Debian, you will need to enable the config and module for Apache:
$ sudo a2enmod wsgi $ sudo a2enconf toaster $ chmod +x bitbake/lib/toaster/toastermain/wsgi.py
Finally, restart Apache to make sure all new configuration is loaded. For Ubuntu, Debian, and OpenSUSE use:
$ sudo service apache2 restart
For Fedora and RedHat use:
$ sudo service httpd restart
Prepare the systemd service to run Toaster builds. Here is a sample configuration file for the service:
[Unit] Description=Toaster runbuilds [Service] Type=forking User=toaster ExecStart=/usr/bin/screen -d -m -S runbuilds /var/www/toaster/poky/bitbake/lib/toaster/runbuilds-service.sh start ExecStop=/usr/bin/screen -S runbuilds -X quit WorkingDirectory=/var/www/toaster/poky [Install] WantedBy=multi-user.target
Prepare the runbuilds-service.sh
script that you need to place in the
/var/www/toaster/poky/bitbake/lib/toaster/
directory by setting up executable permissions:
#!/bin/bash #export http_proxy=http://proxy.host.com:8080 #export https_proxy=http://proxy.host.com:8080 #export GIT_PROXY_COMMAND=$HOME/bin/gitproxy cd ~/poky/ source ./oe-init-build-env build source ../bitbake/bin/toaster $1 noweb [ "$1" == 'start' ] && /bin/bash
Run the service:
# service runbuilds start
Since the service is running in a detached screen session, you can attach to it using this command:
$ sudo su - toaster $ screen -rS runbuilds
You can detach from the service again using "Ctrl-a" followed by "d" key combination.
You can now open up a browser and start using Toaster.
The Toaster web interface allows you to do the following:
Browse published layers in the OpenEmbedded Layer Index that are available for your selected version of the build system.
Import your own layers for building.
Add and remove layers from your configuration.
Set configuration variables.
Select a target or multiple targets to build.
Start your builds.
See what was built (recipes and packages) and what packages were installed into your final image.
Browse the directory structure of your image.
See the value of all variables in your build configuration, and which files set each value.
Examine error, warning and trace messages to aid in debugging.
See information about the BitBake tasks executed and reused during your build, including those that used shared state.
See dependency relationships between recipes, packages and tasks.
See performance information such as build time, task time, CPU usage, and disk I/O.
Following are several videos that show how to use the Toaster GUI:
Build Configuration: This video overviews and demonstrates build configuration for Toaster.
Build Custom Layers: This video shows you how to build custom layers that are used with Toaster.
Toaster Homepage and Table Controls: This video goes over the Toaster entry page, and provides an overview of the data manipulation capabilities of Toaster, which include search, sorting and filtering by different criteria.
Build Dashboard: This video shows you the build dashboard, a page providing an overview of the information available for a selected build.
Image Information: This video walks through the information Toaster provides about images: packages installed and root file system.
Configuration: This video provides Toaster build configuration information.
Tasks: This video shows the information Toaster provides about the tasks run by the build system.
Recipes and Packages Built: This video shows the information Toaster provides about recipes and packages built.
Performance Data: This video shows the build performance data provided by Toaster.
This section only applies if you have set up Toaster for local development, as explained in the "Starting Toaster for Local Development" section.
When you create a project in Toaster, you will be asked to provide a name and to select a Yocto Project release. One of the release options you will find is called "Local Yocto Project".
When you select the "Local Yocto Project" release, Toaster will run your builds using the local Yocto Project clone you have in your computer: the same clone you are using to run Toaster. Unless you manually update this clone, your builds will always use the same Git revision.
If you select any of the other release options, Toaster will fetch the tip of your selected release from the upstream Yocto Project repository every time you run a build. Fetching this tip effectively means that if your selected release is updated upstream, the Git revision you are using for your builds will change. If you are doing development locally, you might not want this change to happen. In that case, the "Local Yocto Project" release might be the right choice.
However, the "Local Yocto Project" release will not provide you with any compatible layers, other than the three core layers that come with the Yocto Project:
If you want to build any other layers, you will need to manually import them into your Toaster project, using the "Import layer" page.
Occasionally, a layer might provide more than one version of
the same recipe.
For example, the openembedded-core
layer
provides two versions of the bash
recipe
(i.e. 3.2.48 and 4.3.30-r0) and two versions of the
which
recipe (i.e. 2.21 and 2.18).
The following figure shows this exact scenario:
By default, the OpenEmbedded build system builds one of the
two recipes.
For the bash
case, version 4.3.30-r0 is
built by default.
Unfortunately, Toaster as it exists, is not able to override
the default recipe version.
If you would like to build bash 3.2.48, you need to set the
PREFERRED_VERSION
variable.
You can do so from Toaster, using the "Add variable" form,
which is available in the "BitBake variables" page of the
project configuration section as shown in the following screen:
To specify bash
3.2.48 as the version to build,
enter "PREFERRED_VERSION_bash" in the "Variable" field, and "3.2.48"
in the "Value" field.
Next, click the "Add variable" button:
After clicking the "Add variable" button, the settings for
PREFERRED_VERSION
are added to the bottom
of the BitBake variables list.
With these settings, the OpenEmbedded build system builds the
desired version of the recipe rather than the default version:
In order to configure and use Toaster, you should understand some
concepts and have some basic command reference material available.
This final chapter provides conceptual information on layer sources,
releases, and JSON configuration files.
Also provided is a quick look at some useful
manage.py
commands that are Toaster-specific.
Information on manage.py
commands does exist
across the Web and the information in this manual by no means
attempts to provide a command comprehensive reference.
In general, a "layer source" is a source of information about existing layers. In particular, we are concerned with layers that you can use with the Yocto Project and Toaster. This chapter describes a particular type of layer source called a "layer index."
A layer index is a web application that contains information about a set of custom layers. A good example of an existing layer index is the OpenEmbedded Layer Index. A public instance of this layer index exists at http://layers.openembedded.org. You can find the code for this layer index's web application at http://git.yoctoproject.org/cgit/cgit.cgi/layerindex-web/.
When you tie a layer source into Toaster, it can query the layer
source through a
REST
API, store the information about the layers in the Toaster
database, and then show the information to users.
Users are then able to view that information and build layers
from Toaster itself without worrying about cloning or editing
the BitBake layers configuration file
bblayers.conf
.
Tying a layer source into Toaster is convenient when you have many custom layers that need to be built on a regular basis by a community of developers. In fact, Toaster comes pre-configured with the OpenEmbedded Metadata Index.
To use your own layer source, you need to set up the layer source and then tie it into Toaster. This section describes how to tie into a layer index in a manner similar to the way Toaster ties into the OpenEmbedded Metadata Index.
The obvious first step for using a layer index is to have several custom layers that developers build and access using the Yocto Project on a regular basis. This set of layers needs to exist and you need to be familiar with where they reside. You will need that information when you set up the code for the web application that "hooks" into your set of layers.
For general information on layers, see the "The Yocto Project Layer Model" section in the Yocto Project Overview and Concepts Manual. For information on how to create layers, see the "Understanding and Creating Layers" section in the Yocto Project Development Tasks Manual.
If you want Toaster to use your layer index, you must host the web application in a server to which Toaster can connect. You also need to give Toaster the information about your layer index. In other words, you have to configure Toaster to use your layer index. This section describes two methods by which you can configure and use your layer index.
In the previous section, the code for the OpenEmbedded Metadata Index (i.e. http://layers.openembedded.org) was referenced. You can use this code, which is at http://git.yoctoproject.org/cgit/cgit.cgi/layerindex-web/, as a base to create your own layer index.
Access the administration interface through a
browser by entering the URL of your Toaster instance and
adding "/admin
" to the end of the
URL.
As an example, if you are running Toaster locally, use
the following URL:
http://127.0.0.1:8000/admin
The administration interface has a "Layer sources" section that includes an "Add layer source" button. Click that button and provide the required information. Make sure you select "layerindex" as the layer source type.
The Django fixture feature overrides the default layer
server when you use it to specify a custom URL. To use
the fixture feature, create (or edit) the file
bitbake/lib/toaster.orm/fixtures/custom.xml
,
and then set the following Toaster setting to your
custom URL:
<?xml version="1.0" ?> <django-objects version="1.0"> <object model="orm.toastersetting" pk="100"> <field name="name" type="CharField">CUSTOM_LAYERINDEX_SERVER</field> <field name="value" type="CharField">https://layers.my_organization.org/layerindex/branch/master/layers/</field> </object> <django-objects>
When you start Toaster for the first time, or if you
delete the file toaster.sqlite
and restart,
the database will populate cleanly from this layer index server.
Once the information has been updated, verify the new layer information is available by using the Toaster web interface. To do that, visit the "All compatible layers" page inside a Toaster project. The layers from your layer source should be listed there.
If you change the information in your layer index server, refresh the Toaster database by running the following command:
$ bitbake/lib/toaster/manage.py lsupdates
If Toaster can reach the API URL, you should see a message telling you that Toaster is updating the layer source information.
When you create a Toaster project using the web interface, you are asked to choose a "Release." In the context of Toaster, the term "Release" refers to a set of layers and a BitBake version the OpenEmbedded build system uses to build something. As shipped, Toaster is pre-configured with releases that correspond to Yocto Project release branches. However, you can modify, delete, and create new releases according to your needs. This section provides some background information on releases.
As shipped, Toaster is configured to use a specific set of releases. Of course, you can always configure Toaster to use any release. For example, you might want your project to build against a specific commit of any of the "out-of-the-box" releases. Or, you might want your project to build against different revisions of OpenEmbedded and BitBake.
As shipped, Toaster is configured to work with the following releases:
Yocto Project 3.1.2 "Dunfell" or OpenEmbedded "Dunfell": This release causes your Toaster projects to build against the head of the dunfell branch at http://git.yoctoproject.org/cgit/cgit.cgi/poky/log/?h=rocko or http://git.openembedded.org/openembedded-core/commit/?h=rocko.
Yocto Project "Master" or OpenEmbedded "Master": This release causes your Toaster Projects to build against the head of the master branch, which is where active development takes place, at http://git.yoctoproject.org/cgit/cgit.cgi/poky/log/ or http://git.openembedded.org/openembedded-core/log/.
Local Yocto Project or Local OpenEmbedded:
This release causes your Toaster Projects to
build against the head of the poky
or openembedded-core
clone you
have local to the machine running Toaster.
In order to use Toaster, you must configure the database with the default content. The following subsections describe various aspects of Toaster configuration.
The
bldcontrol/management/commands/checksettings.py
file controls workflow configuration.
The following steps outline the process to initially populate
this database.
The default project settings are set from
orm/fixtures/settings.xml
.
The default project distro and layers are added
from orm/fixtures/poky.xml
if poky
is installed.
If poky is not installed, they are added
from orm/fixtures/oe-core.xml
.
If the orm/fixtures/custom.xml
file
exists, then its values are added.
The layer index is then scanned and added to the database.
Once these steps complete, Toaster is set up and ready to use.
The pre-set data for Toaster is easily customizable. You can
create the orm/fixtures/custom.xml
file
to customize the values that go into to the database.
Customization is additive,
and can either extend or completely replace the existing values.
You use the orm/fixtures/custom.xml
file
to change the default project settings for the machine, distro,
file images, and layers.
When creating a new project, you can use the file to define
the offered alternate project release selections.
For example, you can add one or more additional selections that
present custom layer sets or distros, and any other local or proprietary
content.
Additionally, you can completely disable the content from the
oe-core.xml
and poky.xml
files by defining the section shown below in the
settings.xml
file.
For example, this option is particularly useful if your custom
configuration defines fewer releases or layers than the default
fixture files.
The following example sets "name" to "CUSTOM_XML_ONLY" and its value to "True".
<object model="orm.toastersetting" pk="99"> <field type="CharField" name="name">CUSTOM_XML_ONLY</field> <field type="CharField" name="value">True</field> </object>
The following is an overview of the file format used by the
oe-core.xml
, poky.xml
,
and custom.xml
files.
The following subsections describe each of the sections in the
fixture files, and outline an example section of the XML code.
you can use to help understand this information and create a local
custom.xml
file.
This section defines the default distro value for new projects.
By default, it reserves the first Toaster Setting record "1".
The following demonstrates how to set the project default value
for
DISTRO
:
<!-- Set the project default value for DISTRO --> <object model="orm.toastersetting" pk="1"> <field type="CharField" name="name">DEFCONF_DISTRO</field> <field type="CharField" name="value">poky</field> </object>
You can override other default project values by adding
additional Toaster Setting sections such as any of the
settings coming from the settings.xml
file.
Also, you can add custom values that are included in the
BitBake environment.
The "pk" values must be unique.
By convention, values that set default project values have a
"DEFCONF" prefix.
The following defines which version of BitBake is used for the following release selection:
<!-- Bitbake versions which correspond to the metadata release --> <object model="orm.bitbakeversion" pk="1"> <field type="CharField" name="name">rocko</field> <field type="CharField" name="giturl">git://git.yoctoproject.org/poky</field> <field type="CharField" name="branch">rocko</field> <field type="CharField" name="dirpath">bitbake</field> </object>
The following defines the releases when you create a new project.
<!-- Releases available --> <object model="orm.release" pk="1"> <field type="CharField" name="name">rocko</field> <field type="CharField" name="description">Yocto Project 2.4 "Rocko"</field> <field rel="ManyToOneRel" to="orm.bitbakeversion" name="bitbake_version">1</field> <field type="CharField" name="branch_name">rocko</field> <field type="TextField" name="helptext">Toaster will run your builds using the tip of the <a href="http://git.yoctoproject.org/cgit/cgit.cgi/poky/log/?h=rocko">Yocto Project Rocko branch</a>.</field> </object>
The "pk" value must match the above respective BitBake version record.
The following defines the default layers for each release:
<!-- Default project layers for each release --> <object model="orm.releasedefaultlayer" pk="1"> <field rel="ManyToOneRel" to="orm.release" name="release">1</field> <field type="CharField" name="layer_name">openembedded-core</field> </object>
The 'pk' values in the example above should start at "1" and increment uniquely. You can use the same layer name in multiple releases.
Layer definitions are the most complex.
The following defines each of the layers, and then defines the exact layer
version of the layer used for each respective release.
You must have one orm.layer
entry for each layer.
Then, with each entry you need a set of
orm.layer_version
entries that connects
the layer with each release that includes the layer.
In general all releases include the layer.
<object model="orm.layer" pk="1"> <field type="CharField" name="name">openembedded-core</field> <field type="CharField" name="layer_index_url"></field> <field type="CharField" name="vcs_url">git://git.yoctoproject.org/poky</field> <field type="CharField" name="vcs_web_url">http://git.yoctoproject.org/cgit/cgit.cgi/poky</field> <field type="CharField" name="vcs_web_tree_base_url">http://git.yoctoproject.org/cgit/cgit.cgi/poky/tree/%path%?h=%branch%</field> <field type="CharField" name="vcs_web_file_base_url">http://git.yoctoproject.org/cgit/cgit.cgi/poky/tree/%path%?h=%branch%</field> </object> <object model="orm.layer_version" pk="1"> <field rel="ManyToOneRel" to="orm.layer" name="layer">1</field> <field type="IntegerField" name="layer_source">0</field> <field rel="ManyToOneRel" to="orm.release" name="release">1</field> <field type="CharField" name="branch">rocko</field> <field type="CharField" name="dirpath">meta</field> </object> <object model="orm.layer_version" pk="2"> <field rel="ManyToOneRel" to="orm.layer" name="layer">1</field> <field type="IntegerField" name="layer_source">0</field> <field rel="ManyToOneRel" to="orm.release" name="release">2</field> <field type="CharField" name="branch">HEAD</field> <field type="CharField" name="commit">HEAD</field> <field type="CharField" name="dirpath">meta</field> </object> <object model="orm.layer_version" pk="3"> <field rel="ManyToOneRel" to="orm.layer" name="layer">1</field> <field type="IntegerField" name="layer_source">0</field> <field rel="ManyToOneRel" to="orm.release" name="release">3</field> <field type="CharField" name="branch">master</field> <field type="CharField" name="dirpath">meta</field> </object>
The layer "pk" values above must be unique, and typically start at "1". The layer version "pk" values must also be unique across all layers, and typically start at "1".
Toaster has an API that allows remote management applications to directly query the state of the Toaster server and its builds in a machine-to-machine manner. This API uses the REST interface and the transfer of JSON files. For example, you might monitor a build inside a container through well supported known HTTP ports in order to easily access a Toaster server inside the container. In this example, when you use this direct JSON API, you avoid having web page parsing against the display the user sees.
Before you use remote Toaster monitoring, you should do a health check. To do this, ping the Toaster server using the following call to see if it is still alive:
http://host
:port
/health
Be sure to provide values for host
and port
.
If the server is alive, you will get the response HTML:
<!DOCTYPE html> <html lang="en"> <head><title>Toaster Health</title></head> <body>Ok</body> </html>
Sometimes it is useful to determine the status of a build in progress. To get the status of pending builds, use the following call:
http://host
:port
/toastergui/api/building
Be sure to provide values for host
and port
.
The output is a JSON file that itemizes all builds in
progress.
This file includes the time in seconds since each
respective build started as well as the progress of the
cloning, parsing, and task execution.
The following is sample output for a build in progress:
{"count": 1, "building": [ {"machine": "beaglebone", "seconds": "463.869", "task": "927:2384", "distro": "poky", "clone": "1:1", "id": 2, "start": "2017-09-22T09:31:44.887Z", "name": "20170922093200", "parse": "818:818", "project": "my_rocko", "target": "core-image-minimal" }] }
The JSON data for this query is returned in a single line. In the previous example the line has been artificially split for readability.
Once a build is completed, you get the status when you use the following call:
http://host
:port
/toastergui/api/builds
Be sure to provide values for host
and port
.
The output is a JSON file that itemizes all complete builds,
and includes build summary information.
The following is sample output for a completed build:
{"count": 1, "builds": [ {"distro": "poky", "errors": 0, "machine": "beaglebone", "project": "my_rocko", "stop": "2017-09-22T09:26:36.017Z", "target": "quilt-native", "seconds": "78.193", "outcome": "Succeeded", "id": 1, "start": "2017-09-22T09:25:17.824Z", "warnings": 1, "name": "20170922092618" }] }
The JSON data for this query is returned in a single line. In the previous example the line has been artificially split for readability.
Sometimes it is useful to determine the status of a specific build. To get the status of a specific build, use the following call:
http://host
:port
/toastergui/api/build/ID
Be sure to provide values for host
,
port
, and ID
.
You can find the value for ID
from the
Builds Completed query. See the
"Checking Status of Builds Completed"
section for more information.
The output is a JSON file that itemizes the specific build and includes build summary information. The following is sample output for a specific build:
{"build": {"distro": "poky", "errors": 0, "machine": "beaglebone", "project": "my_rocko", "stop": "2017-09-22T09:26:36.017Z", "target": "quilt-native", "seconds": "78.193", "outcome": "Succeeded", "id": 1, "start": "2017-09-22T09:25:17.824Z", "warnings": 1, "name": "20170922092618", "cooker_log": "/opt/user/poky/build-toaster-2/tmp/log/cooker/beaglebone/build_20170922_022607.991.log" } }
The JSON data for this query is returned in a single line. In the previous example the line has been artificially split for readability.
In addition to the web user interface and the scripts that start
and stop Toaster, command-line commands exist through the
manage.py
management script.
You can find general documentation on
manage.py
at the
Django
site.
However, several manage.py
commands have been
created that are specific to Toaster and are used to control
configuration and back-end tasks.
You can locate these commands in the
Source Directory
(e.g. poky
) at
bitbake/lib/manage.py
.
This section documents those commands.
When using manage.py
commands given
a default configuration, you must be sure that your
working directory is set to the
Build Directory.
Using manage.py
commands from the
Build Directory allows Toaster to find the
toaster.sqlite
file, which is located
in the Build Directory.
For non-default database configurations, it is possible
that you can use manage.py
commands
from a directory other than the Build Directory.
To do so, the
toastermain/settings.py
file must be
configured to point to the correct database backend.
buildslist
¶
The buildslist
command lists all builds
that Toaster has recorded.
Access the command as follows:
$ bitbake/lib/toaster/manage.py buildslist
The command returns a list, which includes numeric identifications, of the builds that Toaster has recorded in the current database.
You need to run the buildslist
command
first to identify existing builds in the database before
using the
builddelete
command.
Here is an example that assumes default repository and build
directory names:
$ cd ~/poky/build $ python ../bitbake/lib/toaster/manage.py buildslist
If your Toaster database had only one build, the above
buildslist
command would return something
like the following:
1: qemux86 poky core-image-minimal
builddelete
¶
The builddelete
command deletes data
associated with a build.
Access the command as follows:
$ bitbake/lib/toaster/manage.py builddelete build_id
The command deletes all the build data for the specified
build_id
.
This command is useful for removing old and unused data from
the database.
Prior to running the builddelete
command, you need to get the ID associated with builds
by using the
buildslist
command.
perf
¶
The perf
command measures Toaster
performance.
Access the command as follows:
$ bitbake/lib/toaster/manage.py perf
The command is a sanity check that returns page loading times in order to identify performance problems.
checksettings
¶
The checksettings
command verifies
existing Toaster settings.
Access the command as follows:
$ bitbake/lib/toaster/manage.py checksettings
Toaster uses settings that are based on the
database to configure the building tasks.
The checksettings
command verifies that
the database settings are valid in the sense that they have
the minimal information needed to start a build.
In order for the checksettings
command
to work, the database must be correctly set up and not have
existing data.
To be sure the database is ready, you can run the following:
$ bitbake/lib/toaster/manage.py syncdb $ bitbake/lib/toaster/manage.py migrate orm $ bitbake/lib/toaster/manage.py migrate bldcontrol
After running these commands, you can run the
checksettings
command.
runbuilds
¶
The runbuilds
command launches
scheduled builds.
Access the command as follows:
$ bitbake/lib/toaster/manage.py runbuilds
The runbuilds
command checks if
scheduled builds exist in the database and then launches them
per schedule.
The command returns after the builds start but before they
complete.
The Toaster Logging Interface records and updates the database
when the builds complete.
Welcome to the BitBake User Manual. This manual provides information on the BitBake tool. The information attempts to be as independent as possible regarding systems that use BitBake, such as OpenEmbedded and the Yocto Project. In some cases, scenarios or examples within the context of a build system are used in the manual to help with understanding. For these cases, the manual clearly states the context.
Fundamentally, BitBake is a generic task execution engine that allows shell and Python tasks to be run efficiently and in parallel while working within complex inter-task dependency constraints. One of BitBake's main users, OpenEmbedded, takes this core and builds embedded Linux software stacks using a task-oriented approach.
Conceptually, BitBake is similar to GNU Make in some regards but has significant differences:
BitBake executes tasks according to provided
metadata that builds up the tasks.
Metadata is stored in recipe (.bb
)
and related recipe "append" (.bbappend
)
files, configuration (.conf
) and
underlying include (.inc
) files, and
in class (.bbclass
) files.
The metadata provides
BitBake with instructions on what tasks to run and
the dependencies between those tasks.
BitBake includes a fetcher library for obtaining source code from various places such as local files, source control systems, or websites.
The instructions for each unit to be built (e.g. a piece of software) are known as "recipe" files and contain all the information about the unit (dependencies, source file locations, checksums, description and so on).
BitBake includes a client/server abstraction and can be used from a command line or used as a service over XML-RPC and has several different user interfaces.
BitBake was originally a part of the OpenEmbedded project. It was inspired by the Portage package management system used by the Gentoo Linux distribution. On December 7, 2004, OpenEmbedded project team member Chris Larson split the project into two distinct pieces:
BitBake, a generic task executor
OpenEmbedded, a metadata set utilized by BitBake
Today, BitBake is the primary basis of the OpenEmbedded project, which is being used to build and maintain Linux distributions such as the Angstrom Distribution, and which is also being used as the build tool for Linux projects such as the Yocto Project.
Prior to BitBake, no other build tool adequately met the needs of an aspiring embedded Linux distribution. All of the build systems used by traditional desktop Linux distributions lacked important functionality, and none of the ad hoc Buildroot-based systems, prevalent in the embedded space, were scalable or maintainable.
Some important original goals for BitBake were:
Handle cross-compilation.
Handle inter-package dependencies (build time on target architecture, build time on native architecture, and runtime).
Support running any number of tasks within a given package, including, but not limited to, fetching upstream sources, unpacking them, patching them, configuring them, and so forth.
Be Linux distribution agnostic for both build and target systems.
Be architecture agnostic.
Support multiple build and target operating systems (e.g. Cygwin, the BSDs, and so forth).
Be self-contained, rather than tightly integrated into the build machine's root filesystem.
Handle conditional metadata on the target architecture, operating system, distribution, and machine.
Be easy to use the tools to supply local metadata and packages against which to operate.
Be easy to use BitBake to collaborate between multiple projects for their builds.
Provide an inheritance mechanism to share common metadata between many packages.
Over time it became apparent that some further requirements were necessary:
Handle variants of a base recipe (e.g. native, sdk, and multilib).
Split metadata into layers and allow layers to enhance or override other layers.
Allow representation of a given set of input variables to a task as a checksum. Based on that checksum, allow acceleration of builds with prebuilt components.
BitBake satisfies all the original requirements and many more with extensions being made to the basic functionality to reflect the additional requirements. Flexibility and power have always been the priorities. BitBake is highly extensible and supports embedded Python code and execution of any arbitrary tasks.
BitBake is a program written in the Python language. At the highest level, BitBake interprets metadata, decides what tasks are required to run, and executes those tasks. Similar to GNU Make, BitBake controls how software is built. GNU Make achieves its control through "makefiles", while BitBake uses "recipes".
BitBake extends the capabilities of a simple tool like GNU Make by allowing for the definition of much more complex tasks, such as assembling entire embedded Linux distributions.
The remainder of this section introduces several concepts that should be understood in order to better leverage the power of BitBake.
BitBake Recipes, which are denoted by the file extension
.bb
, are the most basic metadata files.
These recipe files provide BitBake with the following:
Descriptive information about the package (author, homepage, license, and so on)
The version of the recipe
Existing dependencies (both build and runtime dependencies)
Where the source code resides and how to fetch it
Whether the source code requires any patches, where to find them, and how to apply them
How to configure and compile the source code
How to assemble the generated artifacts into one or more installable packages
Where on the target machine to install the package or packages created
Within the context of BitBake, or any project utilizing BitBake
as its build system, files with the .bb
extension are referred to as recipes.
Configuration files, which are denoted by the
.conf
extension, define
various configuration variables that govern the project's build
process.
These files fall into several areas that define
machine configuration, distribution configuration,
possible compiler tuning, general common
configuration, and user configuration.
The main configuration file is the sample
bitbake.conf
file, which is
located within the BitBake source tree
conf
directory.
Class files, which are denoted by the
.bbclass
extension, contain
information that is useful to share between metadata files.
The BitBake source tree currently comes with one class metadata file
called base.bbclass
.
You can find this file in the
classes
directory.
The base.bbclass
class files is special since it
is always included automatically for all recipes
and classes.
This class contains definitions for standard basic tasks such
as fetching, unpacking, configuring (empty by default),
compiling (runs any Makefile present), installing (empty by
default) and packaging (empty by default).
These tasks are often overridden or extended by other classes
added during the project development process.
Layers allow you to isolate different types of customizations from each other. While you might find it tempting to keep everything in one layer when working on a single project, the more modular your metadata, the easier it is to cope with future changes.
To illustrate how you can use layers to keep things modular,
consider customizations you might make to support a specific target machine.
These types of customizations typically reside in a special layer,
rather than a general layer, called a Board Support Package (BSP)
layer.
Furthermore, the machine customizations should be isolated from
recipes and metadata that support a new GUI environment, for
example.
This situation gives you a couple of layers: one for the machine
configurations and one for the GUI environment.
It is important to understand, however, that the BSP layer can still
make machine-specific additions to recipes within
the GUI environment layer without polluting the GUI layer itself
with those machine-specific changes.
You can accomplish this through a recipe that is a BitBake append
(.bbappend
) file.
Append files, which are files that have the
.bbappend
file extension, extend or
override information in an existing recipe file.
BitBake expects every append file to have a corresponding recipe file.
Furthermore, the append file and corresponding recipe file
must use the same root filename.
The filenames can differ only in the file type suffix used
(e.g. formfactor_0.0.bb
and
formfactor_0.0.bbappend
).
Information in append files extends or overrides the information in the underlying, similarly-named recipe files.
When you name an append file, you can use the
"%
" wildcard character to allow for matching
recipe names.
For example, suppose you have an append file named
as follows:
busybox_1.21.%.bbappend
That append file would match any busybox_1.21.
x
.bb
version of the recipe.
So, the append file would match the following recipe names:
busybox_1.21.1.bb busybox_1.21.2.bb busybox_1.21.3.bb
%
" character
is limited in that it only works directly in front of the
.bbappend
portion of the append file's
name.
You cannot use the wildcard character in any other
location of the name.
If the busybox
recipe was updated to
busybox_1.3.0.bb
, the append name would not
match.
However, if you named the append file
busybox_1.%.bbappend
, then you would have a match.
In the most general case, you could name the append file something as
simple as busybox_%.bbappend
to be entirely
version independent.
You can obtain BitBake several different ways:
Cloning BitBake: Using Git to clone the BitBake source code repository is the recommended method for obtaining BitBake. Cloning the repository makes it easy to get bug fixes and have access to stable branches and the master branch. Once you have cloned BitBake, you should use the latest stable branch for development since the master branch is for BitBake development and might contain less stable changes.
You usually need a version of BitBake that matches the metadata you are using. The metadata is generally backwards compatible but not forward compatible.
Here is an example that clones the BitBake repository:
$ git clone git://git.openembedded.org/bitbake
This command clones the BitBake Git repository into a
directory called bitbake
.
Alternatively, you can
designate a directory after the
git clone
command
if you want to call the new directory something
other than bitbake
.
Here is an example that names the directory
bbdev
:
$ git clone git://git.openembedded.org/bitbake bbdev
Installation using your Distribution Package Management System: This method is not recommended because the BitBake version that is provided by your distribution, in most cases, is several releases behind a snapshot of the BitBake repository.
Taking a snapshot of BitBake: Downloading a snapshot of BitBake from the source code repository gives you access to a known branch or release of BitBake.
The following example downloads a snapshot of BitBake version 1.17.0:
$ wget http://git.openembedded.org/bitbake/snapshot/bitbake-1.17.0.tar.gz $ tar zxpvf bitbake-1.17.0.tar.gz
After extraction of the tarball using the tar utility,
you have a directory entitled
bitbake-1.17.0
.
Using the BitBake that Comes With Your Build Checkout: A final possibility for getting a copy of BitBake is that it already comes with your checkout of a larger BitBake-based build system, such as Poky. Rather than manually checking out individual layers and gluing them together yourself, you can check out an entire build system. The checkout will already include a version of BitBake that has been thoroughly tested for compatibility with the other components. For information on how to check out a particular BitBake-based build system, consult that build system's supporting documentation.
The bitbake
command is the primary interface
to the BitBake tool.
This section presents the BitBake command syntax and provides
several execution examples.
Following is the usage and syntax for BitBake:
$ bitbake -h Usage: bitbake [options] [recipename/target recipe:do_task ...] Executes the specified task (default is 'build') for a given set of target recipes (.bb files). It is assumed there is a conf/bblayers.conf available in cwd or in BBPATH which will provide the layer, BBFILES and other configuration information. Options: --version show program's version number and exit -h, --help show this help message and exit -b BUILDFILE, --buildfile=BUILDFILE Execute tasks from a specific .bb recipe directly. WARNING: Does not handle any dependencies from other recipes. -k, --continue Continue as much as possible after an error. While the target that failed and anything depending on it cannot be built, as much as possible will be built before stopping. -f, --force Force the specified targets/task to run (invalidating any existing stamp file). -c CMD, --cmd=CMD Specify the task to execute. The exact options available depend on the metadata. Some examples might be 'compile' or 'populate_sysroot' or 'listtasks' may give a list of the tasks available. -C INVALIDATE_STAMP, --clear-stamp=INVALIDATE_STAMP Invalidate the stamp for the specified task such as 'compile' and then run the default task for the specified target(s). -r PREFILE, --read=PREFILE Read the specified file before bitbake.conf. -R POSTFILE, --postread=POSTFILE Read the specified file after bitbake.conf. -v, --verbose Enable tracing of shell tasks (with 'set -x'). Also print bb.note(...) messages to stdout (in addition to writing them to ${T}/log.do_<task>). -D, --debug Increase the debug level. You can specify this more than once. -D sets the debug level to 1, where only bb.debug(1, ...) messages are printed to stdout; -DD sets the debug level to 2, where both bb.debug(1, ...) and bb.debug(2, ...) messages are printed; etc. Without -D, no debug messages are printed. Note that -D only affects output to stdout. All debug messages are written to ${T}/log.do_taskname, regardless of the debug level. -q, --quiet Output less log message data to the terminal. You can specify this more than once. -n, --dry-run Don't execute, just go through the motions. -S SIGNATURE_HANDLER, --dump-signatures=SIGNATURE_HANDLER Dump out the signature construction information, with no task execution. The SIGNATURE_HANDLER parameter is passed to the handler. Two common values are none and printdiff but the handler may define more/less. none means only dump the signature, printdiff means compare the dumped signature with the cached one. -p, --parse-only Quit after parsing the BB recipes. -s, --show-versions Show current and preferred versions of all recipes. -e, --environment Show the global or per-recipe environment complete with information about where variables were set/changed. -g, --graphviz Save dependency tree information for the specified targets in the dot syntax. -I EXTRA_ASSUME_PROVIDED, --ignore-deps=EXTRA_ASSUME_PROVIDED Assume these dependencies don't exist and are already provided (equivalent to ASSUME_PROVIDED). Useful to make dependency graphs more appealing -l DEBUG_DOMAINS, --log-domains=DEBUG_DOMAINS Show debug logging for the specified logging domains -P, --profile Profile the command and save reports. -u UI, --ui=UI The user interface to use (knotty, ncurses or taskexp - default knotty). --token=XMLRPCTOKEN Specify the connection token to be used when connecting to a remote server. --revisions-changed Set the exit code depending on whether upstream floating revisions have changed or not. --server-only Run bitbake without a UI, only starting a server (cooker) process. -B BIND, --bind=BIND The name/address for the bitbake xmlrpc server to bind to. -T SERVER_TIMEOUT, --idle-timeout=SERVER_TIMEOUT Set timeout to unload bitbake server due to inactivity, set to -1 means no unload, default: Environment variable BB_SERVER_TIMEOUT. --no-setscene Do not run any setscene tasks. sstate will be ignored and everything needed, built. --setscene-only Only run setscene tasks, don't run any real tasks. --remote-server=REMOTE_SERVER Connect to the specified server. -m, --kill-server Terminate any running bitbake server. --observe-only Connect to a server as an observing-only client. --status-only Check the status of the remote bitbake server. -w WRITEEVENTLOG, --write-log=WRITEEVENTLOG Writes the event log of the build to a bitbake event json file. Use '' (empty string) to assign the name automatically. --runall=RUNALL Run the specified task for any recipe in the taskgraph of the specified target (even if it wouldn't otherwise have run). --runonly=RUNONLY Run only the specified task within the taskgraph of the specified targets (and any task dependencies those tasks may have).
This section presents some examples showing how to use BitBake.
Executing tasks for a single recipe file is relatively simple. You specify the file in question, and BitBake parses it and executes the specified task. If you do not specify a task, BitBake executes the default task, which is "build”. BitBake obeys inter-task dependencies when doing so.
The following command runs the build task, which is
the default task, on the foo_1.0.bb
recipe file:
$ bitbake -b foo_1.0.bb
The following command runs the clean task on the
foo.bb
recipe file:
$ bitbake -b foo.bb -c clean
There are a number of additional complexities introduced
when one wants to manage multiple .bb
files.
Clearly there needs to be a way to tell BitBake what
files are available and, of those, which you
want to execute.
There also needs to be a way for each recipe
to express its dependencies, both for build-time and
runtime.
There must be a way for you to express recipe preferences
when multiple recipes provide the same functionality, or when
there are multiple versions of a recipe.
The bitbake
command, when not using
"--buildfile" or "-b" only accepts a "PROVIDES".
You cannot provide anything else.
By default, a recipe file generally "PROVIDES" its
"packagename" as shown in the following example:
$ bitbake foo
This next example "PROVIDES" the package name and also uses
the "-c" option to tell BitBake to just execute the
do_clean
task:
$ bitbake -c clean foo
The BitBake command line supports specifying different
tasks for individual targets when you specify multiple
targets.
For example, suppose you had two targets (or recipes)
myfirstrecipe
and
mysecondrecipe
and you needed
BitBake to run taskA
for the first
recipe and taskB
for the second
recipe:
$ bitbake myfirstrecipe:do_taskA mysecondrecipe:do_taskB
BitBake is able to generate dependency graphs using
the dot
syntax.
You can convert these graphs into images using the
dot
tool from
Graphviz.
When you generate a dependency graph, BitBake writes two files to the current working directory:
task-depends.dot
:
Shows dependencies between tasks.
These dependencies match BitBake's internal task execution list.
pn-buildlist
:
Shows a simple list of targets that are to be built.
To stop depending on common depends, use the "-I" depend
option and BitBake omits them from the graph.
Leaving this information out can produce more readable graphs.
This way, you can remove from the graph
DEPENDS
from inherited classes
such as base.bbclass
.
Here are two examples that create dependency graphs. The second example omits depends common in OpenEmbedded from the graph:
$ bitbake -g foo $ bitbake -g -I virtual/kernel -I eglibc foo
BitBake is able to build multiple images or packages using a single command where the different targets require different configurations (multiple configuration builds). Each target, in this scenario, is referred to as a "multiconfig".
To accomplish a multiple configuration build, you must
define each target's configuration separately using
a parallel configuration file in the build directory.
The location for these multiconfig configuration files
is specific.
They must reside in the current build directory in
a sub-directory of conf
named
multiconfig
.
Following is an example for two separate targets:
The reason for this required file hierarchy
is because the BBPATH
variable
is not constructed until the layers are parsed.
Consequently, using the configuration file as a
pre-configuration file is not possible unless it is
located in the current working directory.
Minimally, each configuration file must define the machine and the temporary directory BitBake uses for the build. Suggested practice dictates that you do not overlap the temporary directories used during the builds.
Aside from separate configuration files for each
target, you must also enable BitBake to perform multiple
configuration builds.
Enabling is accomplished by setting the
BBMULTICONFIG
variable in the local.conf
configuration file.
As an example, suppose you had configuration files
for target1
and
target2
defined in the build
directory.
The following statement in the
local.conf
file both enables
BitBake to perform multiple configuration builds and
specifies the two extra multiconfigs:
BBMULTICONFIG = "target1 target2"
Once the target configuration files are in place and BitBake has been enabled to perform multiple configuration builds, use the following command form to start the builds:
$ bitbake [mc:multiconfigname
:]target
[[[mc:multiconfigname
:]target
] ... ]
Here is an example for two extra multiconfigs:
target1
and
target2
:
$ bitbake mc::target
mc:target1:target
mc:target2:target
Sometimes dependencies can exist between targets (multiconfigs) in a multiple configuration build. For example, suppose that in order to build an image for a particular architecture, the root filesystem of another build for a different architecture needs to exist. In other words, the image for the first multiconfig depends on the root filesystem of the second multiconfig. This dependency is essentially that the task in the recipe that builds one multiconfig is dependent on the completion of the task in the recipe that builds another multiconfig.
To enable dependencies in a multiple configuration build, you must declare the dependencies in the recipe using the following statement form:
task_or_package
[mcdepends] = "mc:from_multiconfig
:to_multiconfig
:recipe_name
:task_on_which_to_depend
"
To better show how to use this statement, consider an
example with two multiconfigs: target1
and target2
:
image_task
[mcdepends] = "mc:target1:target2:image2
:rootfs_task
"
In this example, the
from_multiconfig
is "target1" and
the to_multiconfig
is "target2".
The task on which the image whose recipe contains
image_task
depends on the
completion of the rootfs_task
used to build out image2
, which
is associated with the "target2" multiconfig.
Once you set up this dependency, you can build the "target1" multiconfig using a BitBake command as follows:
$ bitbake mc:target1:image1
This command executes all the tasks needed to create
image1
for the "target1"
multiconfig.
Because of the dependency, BitBake also executes through
the rootfs_task
for the "target2"
multiconfig build.
Having a recipe depend on the root filesystem of another
build might not seem that useful.
Consider this change to the statement in the
image1
recipe:
image_task
[mcdepends] = "mc:target1:target2:image2
:image_task
"
In this case, BitBake must create
image2
for the "target2"
build since the "target1" build depends on it.
Because "target1" and "target2" are enabled for multiple configuration builds and have separate configuration files, BitBake places the artifacts for each build in the respective temporary build directories.
The primary purpose for running BitBake is to produce some kind
of output such as a single installable package, a kernel, a software
development kit, or even a full, board-specific bootable Linux image,
complete with bootloader, kernel, and root filesystem.
Of course, you can execute the bitbake
command with options that cause it to execute single tasks,
compile single recipe files, capture or clear data, or simply
return information about the execution environment.
This chapter describes BitBake's execution process from start to finish when you use it to create an image. The execution process is launched using the following command form:
$ bitbake target
For information on the BitBake command and its options, see "The BitBake Command" section.
Prior to executing BitBake, you should take advantage of available
parallel thread execution on your build host by setting the
BB_NUMBER_THREADS
variable in your project's local.conf
configuration file.
A common method to determine this value for your build host is to run the following:
$ grep processor /proc/cpuinfo
This command returns the number of processors, which takes into
account hyper-threading.
Thus, a quad-core build host with hyper-threading most likely
shows eight processors, which is the value you would then assign to
BB_NUMBER_THREADS
.
A possibly simpler solution is that some Linux distributions
(e.g. Debian and Ubuntu) provide the ncpus
command.
The first thing BitBake does is parse base configuration
metadata.
Base configuration metadata consists of your project's
bblayers.conf
file to determine what
layers BitBake needs to recognize, all necessary
layer.conf
files (one from each layer),
and bitbake.conf
.
The data itself is of various types:
Recipes: Details about particular pieces of software.
Class Data: An abstraction of common build information (e.g. how to build a Linux kernel).
Configuration Data: Machine-specific settings, policy decisions, and so forth. Configuration data acts as the glue to bind everything together.
The layer.conf
files are used to
construct key variables such as
BBPATH
and
BBFILES
.
BBPATH
is used to search for
configuration and class files under the
conf
and classes
directories, respectively.
BBFILES
is used to locate both recipe
and recipe append files
(.bb
and .bbappend
).
If there is no bblayers.conf
file,
it is assumed the user has set the BBPATH
and BBFILES
directly in the environment.
Next, the bitbake.conf
file is located
using the BBPATH
variable that was
just constructed.
The bitbake.conf
file may also include other
configuration files using the
include
or
require
directives.
Prior to parsing configuration files, BitBake looks at certain variables, including:
The first four variables in this list relate to how BitBake treats shell environment variables during task execution. By default, BitBake cleans the environment variables and provides tight control over the shell execution environment. However, through the use of these first four variables, you can apply your control regarding the environment variables allowed to be used by BitBake in the shell during execution of tasks. See the "Passing Information Into the Build Task Environment" section and the information about these variables in the variable glossary for more information on how they work and on how to use them.
The base configuration metadata is global and therefore affects all recipes and tasks that are executed.
BitBake first searches the current working directory for an
optional conf/bblayers.conf
configuration file.
This file is expected to contain a
BBLAYERS
variable that is a space-delimited list of 'layer' directories.
Recall that if BitBake cannot find a bblayers.conf
file, then it is assumed the user has set the BBPATH
and BBFILES
variables directly in the environment.
For each directory (layer) in this list, a conf/layer.conf
file is located and parsed with the
LAYERDIR
variable being set to the directory where the layer was found.
The idea is these files automatically set up
BBPATH
and other variables correctly for a given build directory.
BitBake then expects to find the conf/bitbake.conf
file somewhere in the user-specified BBPATH
.
That configuration file generally has include directives to pull
in any other metadata such as files specific to the architecture,
the machine, the local environment, and so forth.
Only variable definitions and include directives are allowed
in BitBake .conf
files.
Some variables directly influence BitBake's behavior.
These variables might have been set from the environment
depending on the environment variables previously
mentioned or set in the configuration files.
The
"Variables Glossary"
chapter presents a full list of variables.
After parsing configuration files, BitBake uses its rudimentary inheritance mechanism, which is through class files, to inherit some standard classes. BitBake parses a class when the inherit directive responsible for getting that class is encountered.
The base.bbclass
file is always included.
Other classes that are specified in the configuration using the
INHERIT
variable are also included.
BitBake searches for class files in a
classes
subdirectory under
the paths in BBPATH
in the same way as
configuration files.
A good way to get an idea of the configuration files and the class files used in your execution environment is to run the following BitBake command:
$ bitbake -e > mybb.log
Examining the top of the mybb.log
shows you the many configuration files and class files
used in your execution environment.
You need to be aware of how BitBake parses curly braces. If a recipe uses a closing curly brace within the function and the character has no leading spaces, BitBake produces a parsing error. If you use a pair of curly braces in a shell function, the closing curly brace must not be located at the start of the line without leading spaces.
Here is an example that causes BitBake to produce a parsing error:
fakeroot create_shar() { cat << "EOF" > ${SDK_DEPLOY}/${TOOLCHAIN_OUTPUTNAME}.sh usage() { echo "test" ###### The following "}" at the start of the line causes a parsing error ###### } EOF }
Writing the recipe this way avoids the error:
fakeroot create_shar() { cat << "EOF" > ${SDK_DEPLOY}/${TOOLCHAIN_OUTPUTNAME}.sh usage() { echo "test" ######The following "}" with a leading space at the start of the line avoids the error ###### } EOF }
During the configuration phase, BitBake will have set
BBFILES
.
BitBake now uses it to construct a list of recipes to parse,
along with any append files (.bbappend
)
to apply.
BBFILES
is a space-separated list of
available files and supports wildcards.
An example would be:
BBFILES = "/path/to/bbfiles/*.bb /path/to/appends/*.bbappend"
BitBake parses each recipe and append file located
with BBFILES
and stores the values of
various variables into the datastore.
BBFILES
.
For each file, a fresh copy of the base configuration is
made, then the recipe is parsed line by line.
Any inherit statements cause BitBake to find and
then parse class files (.bbclass
)
using
BBPATH
as the search path.
Finally, BitBake parses in order any append files found in
BBFILES
.
One common convention is to use the recipe filename to define
pieces of metadata.
For example, in bitbake.conf
the recipe
name and version are used to set the variables
PN
and
PV
:
PN = "${@bb.parse.BBHandler.vars_from_file(d.getVar('FILE', False),d)[0] or 'defaultpkgname'}" PV = "${@bb.parse.BBHandler.vars_from_file(d.getVar('FILE', False),d)[1] or '1.0'}"
In this example, a recipe called "something_1.2.3.bb" would set
PN
to "something" and
PV
to "1.2.3".
By the time parsing is complete for a recipe, BitBake has a list of tasks that the recipe defines and a set of data consisting of keys and values as well as dependency information about the tasks.
BitBake does not need all of this information. It only needs a small subset of the information to make decisions about the recipe. Consequently, BitBake caches the values in which it is interested and does not store the rest of the information. Experience has shown it is faster to re-parse the metadata than to try and write it out to the disk and then reload it.
Where possible, subsequent BitBake commands reuse this cache of
recipe information.
The validity of this cache is determined by first computing a
checksum of the base configuration data (see
BB_HASHCONFIG_WHITELIST
)
and then checking if the checksum matches.
If that checksum matches what is in the cache and the recipe
and class files have not changed, BitBake is able to use
the cache.
BitBake then reloads the cached information about the recipe
instead of reparsing it from scratch.
Recipe file collections exist to allow the user to
have multiple repositories of
.bb
files that contain the same
exact package.
For example, one could easily use them to make one's
own local copy of an upstream repository, but with
custom modifications that one does not want upstream.
Here is an example:
BBFILES = "/stuff/openembedded/*/*.bb /stuff/openembedded.modified/*/*.bb" BBFILE_COLLECTIONS = "upstream local" BBFILE_PATTERN_upstream = "^/stuff/openembedded/" BBFILE_PATTERN_local = "^/stuff/openembedded.modified/" BBFILE_PRIORITY_upstream = "5" BBFILE_PRIORITY_local = "10"
Assuming BitBake has been instructed to execute a target
and that all the recipe files have been parsed, BitBake
starts to figure out how to build the target.
BitBake looks through the PROVIDES
list
for each of the recipes.
A PROVIDES
list is the list of names by which
the recipe can be known.
Each recipe's PROVIDES
list is created
implicitly through the recipe's
PN
variable
and explicitly through the recipe's
PROVIDES
variable, which is optional.
When a recipe uses PROVIDES
, that recipe's
functionality can be found under an alternative name or names other
than the implicit PN
name.
As an example, suppose a recipe named keyboard_1.0.bb
contained the following:
PROVIDES += "fullkeyboard"
The PROVIDES
list for this recipe becomes
"keyboard", which is implicit, and "fullkeyboard", which is explicit.
Consequently, the functionality found in
keyboard_1.0.bb
can be found under two
different names.
The PROVIDES
list is only part of the solution
for figuring out a target's recipes.
Because targets might have multiple providers, BitBake needs
to prioritize providers by determining provider preferences.
A common example in which a target has multiple providers
is "virtual/kernel", which is on the
PROVIDES
list for each kernel recipe.
Each machine often selects the best kernel provider by using a
line similar to the following in the machine configuration file:
PREFERRED_PROVIDER_virtual/kernel = "linux-yocto"
The default
PREFERRED_PROVIDER
is the provider with the same name as the target.
BitBake iterates through each target it needs to build and
resolves them and their dependencies using this process.
Understanding how providers are chosen is made complicated by the fact
that multiple versions might exist for a given provider.
BitBake defaults to the highest version of a provider.
Version comparisons are made using the same method as Debian.
You can use the
PREFERRED_VERSION
variable to specify a particular version.
You can influence the order by using the
DEFAULT_PREFERENCE
variable.
By default, files have a preference of "0".
Setting DEFAULT_PREFERENCE
to "-1" makes the
recipe unlikely to be used unless it is explicitly referenced.
Setting DEFAULT_PREFERENCE
to "1" makes it
likely the recipe is used.
PREFERRED_VERSION
overrides any
DEFAULT_PREFERENCE
setting.
DEFAULT_PREFERENCE
is often used to mark newer
and more experimental recipe versions until they have undergone
sufficient testing to be considered stable.
When there are multiple “versions” of a given recipe,
BitBake defaults to selecting the most recent
version, unless otherwise specified.
If the recipe in question has a
DEFAULT_PREFERENCE
set lower than the other recipes (default is 0), then
it will not be selected.
This allows the person or persons maintaining
the repository of recipe files to specify
their preference for the default selected version.
Additionally, the user can specify their preferred version.
If the first recipe is named a_1.1.bb
, then the
PN
variable
will be set to “a”, and the
PV
variable will be set to 1.1.
Thus, if a recipe named a_1.2.bb
exists, BitBake
will choose 1.2 by default.
However, if you define the following variable in a
.conf
file that BitBake parses, you
can change that preference:
PREFERRED_VERSION_a = "1.1"
It is common for a recipe to provide two versions -- a stable, numbered (and preferred) version, and a version that is automatically checked out from a source code repository that is considered more "bleeding edge" but can be selected only explicitly.
For example, in the OpenEmbedded codebase, there is a standard,
versioned recipe file for BusyBox,
busybox_1.22.1.bb
,
but there is also a Git-based version,
busybox_git.bb
, which explicitly contains the line
DEFAULT_PREFERENCE = "-1"
to ensure that the numbered, stable version is always preferred unless the developer selects otherwise.
Each target BitBake builds consists of multiple tasks such as
fetch
, unpack
,
patch
, configure
,
and compile
.
For best performance on multi-core systems, BitBake considers each
task as an independent
entity with its own set of dependencies.
Dependencies are defined through several variables.
You can find information about variables BitBake uses in
the Variables Glossary
near the end of this manual.
At a basic level, it is sufficient to know that BitBake uses the
DEPENDS
and
RDEPENDS
variables when
calculating dependencies.
For more information on how BitBake handles dependencies, see the "Dependencies" section.
Based on the generated list of providers and the dependency information, BitBake can now calculate exactly what tasks it needs to run and in what order it needs to run them. The "Executing Tasks" section has more information on how BitBake chooses which task to execute next.
The build now starts with BitBake forking off threads up to the limit set in the
BB_NUMBER_THREADS
variable.
BitBake continues to fork threads as long as there are tasks ready to run,
those tasks have all their dependencies met, and the thread threshold has not been
exceeded.
It is worth noting that you can greatly speed up the build time by properly setting
the BB_NUMBER_THREADS
variable.
As each task completes, a timestamp is written to the directory specified by the
STAMP
variable.
On subsequent runs, BitBake looks in the build directory within
tmp/stamps
and does not rerun
tasks that are already completed unless a timestamp is found to be invalid.
Currently, invalid timestamps are only considered on a per
recipe file basis.
So, for example, if the configure stamp has a timestamp greater than the
compile timestamp for a given target, then the compile task would rerun.
Running the compile task again, however, has no effect on other providers
that depend on that target.
The exact format of the stamps is partly configurable.
In modern versions of BitBake, a hash is appended to the
stamp so that if the configuration changes, the stamp becomes
invalid and the task is automatically rerun.
This hash, or signature used, is governed by the signature policy
that is configured (see the
"Checksums (Signatures)"
section for information).
It is also possible to append extra metadata to the stamp using
the [stamp-extra-info]
task flag.
For example, OpenEmbedded uses this flag to make some tasks machine-specific.
For more information on tasks, see the "Tasks" section.
Tasks can be either a shell task or a Python task.
For shell tasks, BitBake writes a shell script to
${
T
}/run.do_taskname.
and then executes the script.
The generated shell script contains all the exported variables,
and the shell functions with all variables expanded.
Output from the shell script goes to the file
pid
${T}/log.do_taskname.
.
Looking at the expanded shell functions in the run file and
the output in the log files is a useful debugging technique.
pid
For Python tasks, BitBake executes the task internally and logs information to the controlling terminal. Future versions of BitBake will write the functions to files similar to the way shell tasks are handled. Logging will be handled in a way similar to shell tasks as well.
The order in which BitBake runs the tasks is controlled by its task scheduler. It is possible to configure the scheduler and define custom implementations for specific use cases. For more information, see these variables that control the behavior:
It is possible to have functions run before and after a task's main
function.
This is done using the [prefuncs]
and [postfuncs]
flags of the task
that lists the functions to run.
A checksum is a unique signature of a task's inputs. The signature of a task can be used to determine if a task needs to be run. Because it is a change in a task's inputs that triggers running the task, BitBake needs to detect all the inputs to a given task. For shell tasks, this turns out to be fairly easy because BitBake generates a "run" shell script for each task and it is possible to create a checksum that gives you a good idea of when the task's data changes.
To complicate the problem, some things should not be included in
the checksum.
First, there is the actual specific build path of a given task -
the working directory.
It does not matter if the working directory changes because it should not
affect the output for target packages.
The simplistic approach for excluding the working directory is to set
it to some fixed value and create the checksum for the "run" script.
BitBake goes one step better and uses the
BB_HASHBASE_WHITELIST
variable to define a list of variables that should never be included
when generating the signatures.
Another problem results from the "run" scripts containing functions that might or might not get called. The incremental build solution contains code that figures out dependencies between shell functions. This code is used to prune the "run" scripts down to the minimum set, thereby alleviating this problem and making the "run" scripts much more readable as a bonus.
So far we have solutions for shell scripts. What about Python tasks? The same approach applies even though these tasks are more difficult. The process needs to figure out what variables a Python function accesses and what functions it calls. Again, the incremental build solution contains code that first figures out the variable and function dependencies, and then creates a checksum for the data used as the input to the task.
Like the working directory case, situations exist where dependencies should be ignored. For these cases, you can instruct the build process to ignore a dependency by using a line like the following:
PACKAGE_ARCHS[vardepsexclude] = "MACHINE"
This example ensures that the PACKAGE_ARCHS
variable does not
depend on the value of MACHINE
, even if it does reference it.
Equally, there are cases where we need to add dependencies BitBake is not able to find. You can accomplish this by using a line like the following:
PACKAGE_ARCHS[vardeps] = "MACHINE"
This example explicitly adds the MACHINE
variable as a
dependency for PACKAGE_ARCHS
.
Consider a case with in-line Python, for example, where BitBake is not
able to figure out dependencies.
When running in debug mode (i.e. using -DDD
), BitBake
produces output when it discovers something for which it cannot figure out
dependencies.
Thus far, this section has limited discussion to the direct inputs into a task. Information based on direct inputs is referred to as the "basehash" in the code. However, there is still the question of a task's indirect inputs - the things that were already built and present in the build directory. The checksum (or signature) for a particular task needs to add the hashes of all the tasks on which the particular task depends. Choosing which dependencies to add is a policy decision. However, the effect is to generate a master checksum that combines the basehash and the hashes of the task's dependencies.
At the code level, there are a variety of ways both the basehash and the dependent task hashes can be influenced. Within the BitBake configuration file, we can give BitBake some extra information to help it construct the basehash. The following statement effectively results in a list of global variable dependency excludes - variables never included in any checksum. This example uses variables from OpenEmbedded to help illustrate the concept:
BB_HASHBASE_WHITELIST ?= "TMPDIR FILE PATH PWD BB_TASKHASH BBPATH DL_DIR \ SSTATE_DIR THISDIR FILESEXTRAPATHS FILE_DIRNAME HOME LOGNAME SHELL TERM \ USER FILESPATH STAGING_DIR_HOST STAGING_DIR_TARGET COREBASE PRSERV_HOST \ PRSERV_DUMPDIR PRSERV_DUMPFILE PRSERV_LOCKDOWN PARALLEL_MAKE \ CCACHE_DIR EXTERNAL_TOOLCHAIN CCACHE CCACHE_DISABLE LICENSE_PATH SDKPKGSUFFIX"
The previous example excludes the work directory, which is part of
TMPDIR
.
The rules for deciding which hashes of dependent tasks to include through
dependency chains are more complex and are generally accomplished with a
Python function.
The code in meta/lib/oe/sstatesig.py
shows two examples
of this and also illustrates how you can insert your own policy into the system
if so desired.
This file defines the two basic signature generators OpenEmbedded-Core
uses: "OEBasic" and "OEBasicHash".
By default, there is a dummy "noop" signature handler enabled in BitBake.
This means that behavior is unchanged from previous versions.
OE-Core
uses the "OEBasicHash" signature handler by default
through this setting in the bitbake.conf
file:
BB_SIGNATURE_HANDLER ?= "OEBasicHash"
The "OEBasicHash" BB_SIGNATURE_HANDLER
is the same as the
"OEBasic" version but adds the task hash to the stamp files.
This results in any metadata change that changes the task hash, automatically
causing the task to be run again.
This removes the need to bump
PR
values, and changes to metadata automatically ripple across the build.
It is also worth noting that the end result of these signature generators is to make some dependency and hash information available to the build. This information includes:
BB_BASEHASH_task-
taskname
:
The base hashes for each task in the recipe.
BB_BASEHASH_
filename
:
taskname
:
The base hashes for each dependent task.
BBHASHDEPS_
filename
:
taskname
:
The task dependencies for each task.
BB_TASKHASH
:
The hash of the currently running task.
It is worth noting that BitBake's "-S" option lets you
debug BitBake's processing of signatures.
The options passed to -S allow different debugging modes
to be used, either using BitBake's own debug functions
or possibly those defined in the metadata/signature handler
itself.
The simplest parameter to pass is "none", which causes a
set of signature information to be written out into
STAMPS_DIR
corresponding to the targets specified.
The other currently available parameter is "printdiff",
which causes BitBake to try to establish the closest
signature match it can (e.g. in the sstate cache) and then
run bitbake-diffsigs
over the matches
to determine the stamps and delta where these two
stamp trees diverge.
You can find more information on checksum metadata in the "Task Checksums and Setscene" section.
The setscene process enables BitBake to handle "pre-built" artifacts. The ability to handle and reuse these artifacts allows BitBake the luxury of not having to build something from scratch every time. Instead, BitBake can use, when possible, existing build artifacts.
BitBake needs to have reliable data indicating whether or not an artifact is compatible. Signatures, described in the previous section, provide an ideal way of representing whether an artifact is compatible. If a signature is the same, an object can be reused.
If an object can be reused, the problem then becomes how to replace a given task or set of tasks with the pre-built artifact. BitBake solves the problem with the "setscene" process.
When BitBake is asked to build a given target, before building anything, it first asks whether cached information is available for any of the targets it's building, or any of the intermediate targets. If cached information is available, BitBake uses this information instead of running the main tasks.
BitBake first calls the function defined by the
BB_HASHCHECK_FUNCTION
variable with a list of tasks and corresponding
hashes it wants to build.
This function is designed to be fast and returns a list
of the tasks for which it believes in can obtain artifacts.
Next, for each of the tasks that were returned as possibilities,
BitBake executes a setscene version of the task that the possible
artifact covers.
Setscene versions of a task have the string "_setscene" appended to the
task name.
So, for example, the task with the name xxx
has
a setscene task named xxx_setscene
.
The setscene version of the task executes and provides the necessary
artifacts returning either success or failure.
As previously mentioned, an artifact can cover more than one task.
For example, it is pointless to obtain a compiler if you
already have the compiled binary.
To handle this, BitBake calls the
BB_SETSCENE_DEPVALID
function for each successful setscene task to know whether or not it needs
to obtain the dependencies of that task.
Finally, after all the setscene tasks have executed, BitBake calls the
function listed in
BB_SETSCENE_VERIFY_FUNCTION2
with the list of tasks BitBake thinks has been "covered".
The metadata can then ensure that this list is correct and can
inform BitBake that it wants specific tasks to be run regardless
of the setscene result.
You can find more information on setscene metadata in the "Task Checksums and Setscene" section.
In addition to the standard command line option to control how
verbose builds are when execute, bitbake also supports user defined
configuration of the
Python logging
facilities through the
BB_LOGCONFIG
variable. This variable defines a json or yaml
logging configuration
that will be intelligently merged into the default configuration.
The logging configuration is merged using the following rules:
The user defined configuration will completely replace the default
configuration if top level key
bitbake_merge
is set to the value
False
. In this case, all other rules
are ignored.
The user configuration must have a top level
version
which must match the value of
the default configuration.
Any keys defined in the handlers
,
formatters
, or filters
,
will be merged into the same section in the default
configuration, with the user specified keys taking
replacing a default one if there is a conflict. In
practice, this means that if both the default configuration
and user configuration specify a handler named
myhandler
, the user defined one will
replace the default. To prevent the user from inadvertently
replacing a default handler, formatter, or filter, all of
the default ones are named with a prefix of
"BitBake.
"
If a logger is defined by the user with the key
bitbake_merge
set to
False
, that logger will be completely
replaced by user configuration. In this case, no other
rules will apply to that logger.
All user defined filter
and
handlers
properties for a given logger
will be merged with corresponding properties from the
default logger. For example, if the user configuration adds
a filter called myFilter
to the
BitBake.SigGen
, and the default
configuration adds a filter called
BitBake.defaultFilter
, both filters
will be applied to the logger
As an example, consider the following user logging configuration
file which logs all Hash Equivalence related messages of VERBOSE or
higher to a file called hashequiv.log
{ "version": 1, "handlers": { "autobuilderlog": { "class": "logging.FileHandler", "formatter": "logfileFormatter", "level": "DEBUG", "filename": "hashequiv.log", "mode": "w" } }, "formatters": { "logfileFormatter": { "format": "%(name)s: %(levelname)s: %(message)s" } }, "loggers": { "BitBake.SigGen.HashEquiv": { "level": "VERBOSE", "handlers": ["autobuilderlog"] }, "BitBake.RunQueue.HashEquiv": { "level": "VERBOSE", "handlers": ["autobuilderlog"] } } }
BitBake files have their own syntax. The syntax has similarities to several other languages but also has some unique features. This section describes the available syntax and operators as well as provides examples.
This section provides some basic syntax examples.
The following example sets VARIABLE
to
"value".
This assignment occurs immediately as the statement is parsed.
It is a "hard" assignment.
VARIABLE = "value"
As expected, if you include leading or trailing spaces as part of an assignment, the spaces are retained:
VARIABLE = " value" VARIABLE = "value "
Setting VARIABLE
to "" sets it to an empty string,
while setting the variable to " " sets it to a blank space
(i.e. these are not the same values).
VARIABLE = "" VARIABLE = " "
You can use single quotes instead of double quotes when setting a variable's value. Doing so allows you to use values that contain the double quote character:
VARIABLE = 'I have a " in my value'
Sometimes you need to modify existing variables. Following are some cases where you might find you want to modify an existing variable:
Customize a recipe that uses the variable.
Change a variable's default value used in a
*.bbclass
file.
Change the variable in a *.bbappend
file to override the variable in the original recipe.
Change the variable in a configuration file so that the value overrides an existing configuration.
Changing a variable value can sometimes depend on how the value was originally assigned and also on the desired intent of the change. In particular, when you append a value to a variable that has a default value, the resulting value might not be what you expect. In this case, the value you provide might replace the value rather than append to the default value.
If after you have changed a variable's value and something unexplained occurs, you can use BitBake to check the actual value of the suspect variable. You can make these checks for both configuration and recipe level changes:
For configuration changes, use the following:
$ bitbake -e
This command displays variable values after the
configuration files (i.e. local.conf
,
bblayers.conf
,
bitbake.conf
and so forth) have
been parsed.
For recipe changes, use the following:
$ bitbake recipe
-e | grep VARIABLE="
This command checks to see if the variable actually makes it into a specific recipe.
Outside of functions, BitBake joins any line ending in a backslash character ("\") with the following line before parsing statements. The most common use for the "\" character is to split variable assignments over multiple lines, as in the following example:
FOO = "bar \ baz \ qaz"
Both the "\" character and the newline character
that follow it are removed when joining lines.
Thus, no newline characters end up in the value of
FOO
.
Consider this additional example where the two
assignments both assign "barbaz" to
FOO
:
FOO = "barbaz" FOO = "bar\ baz"
printf
or
echo -n
.
Variables can reference the contents of other variables using a syntax that is similar to variable expansion in Bourne shells. The following assignments result in A containing "aval" and B evaluating to "preavalpost".
A = "aval" B = "pre${A}post"
${FOO}
and not
$FOO
is recognized as an expansion of
FOO
.
The "=" operator does not immediately expand variable references in the right-hand side. Instead, expansion is deferred until the variable assigned to is actually used. The result depends on the current values of the referenced variables. The following example should clarify this behavior:
A = "${B} baz" B = "${C} bar" C = "foo" *At this point, ${A} equals "foo bar baz"* C = "qux" *At this point, ${A} equals "qux bar baz"* B = "norf" *At this point, ${A} equals "norf baz"*
Contrast this behavior with the immediate variable expansion operator (i.e. ":=").
If the variable expansion syntax is used on a variable that
does not exist, the string is kept as is.
For example, given the following assignment,
BAR
expands to the literal string
"${FOO}" as long as FOO
does not exist.
BAR = "${FOO}"
You can use the "?=" operator to achieve a "softer" assignment for a variable. This type of assignment allows you to define a variable if it is undefined when the statement is parsed, but to leave the value alone if the variable has a value. Here is an example:
A ?= "aval"
If A
is set at the time this statement is parsed,
the variable retains its value.
However, if A
is not set,
the variable is set to "aval".
It is possible to use a "weaker" assignment than in the previous section by using the "??=" operator. This assignment behaves identical to "?=" except that the assignment is made at the end of the parsing process rather than immediately. Consequently, when multiple "??=" assignments exist, the last one is used. Also, any "=" or "?=" assignment will override the value set with "??=". Here is an example:
A ??= "somevalue" A ??= "someothervalue"
If A
is set before the above statements are parsed,
the variable retains its value.
If A
is not set,
the variable is set to "someothervalue".
Again, this assignment is a "lazy" or "weak" assignment because it does not occur until the end of the parsing process.
The ":=" operator results in a variable's contents being expanded immediately, rather than when the variable is actually used:
T = "123" A := "test ${T}" T = "456" B := "${T} ${C}" C = "cval" C := "${C}append"
In this example, A
contains
"test 123", even though the final value of T
is "456".
The variable B
will end up containing "456 cvalappend".
This is because references to undefined variables are preserved as is
during (immediate)expansion. This is in contrast to GNU Make, where undefined
variables expand to nothing.
The variable C
contains "cvalappend" since ${C}
immediately
expands to "cval".
Appending and prepending values is common and can be accomplished using the "+=" and "=+" operators. These operators insert a space between the current value and prepended or appended value.
These operators take immediate effect during parsing. Here are some examples:
B = "bval" B += "additionaldata" C = "cval" C =+ "test"
The variable B
contains
"bval additionaldata" and C
contains "test cval".
If you want to append or prepend values without an inserted space, use the ".=" and "=." operators.
These operators take immediate effect during parsing. Here are some examples:
B = "bval" B .= "additionaldata" C = "cval" C =. "test"
The variable B
contains
"bvaladditionaldata" and
C
contains "testcval".
You can also append and prepend a variable's value using an override style syntax. When you use this syntax, no spaces are inserted.
These operators differ from the ":=", ".=", "=.", "+=", and "=+" operators in that their effects are applied at variable expansion time rather than being immediately applied. Here are some examples:
B = "bval" B_append = " additional data" C = "cval" C_prepend = "additional data " D = "dval" D_append = "additional data"
The variable B
becomes
"bval additional data" and C
becomes
"additional data cval".
The variable D
becomes
"dvaladditional data".
It is also possible to append and prepend to shell functions and BitBake-style Python functions. See the "Shell Functions" and "BitBake-Style Python Functions sections for examples.
You can remove values from lists using the removal override style syntax. Specifying a value for removal causes all occurrences of that value to be removed from the variable.
When you use this syntax, BitBake expects one or more strings. Surrounding spaces and spacing are preserved. Here is an example:
FOO = "123 456 789 123456 123 456 123 456" FOO_remove = "123" FOO_remove = "456" FOO2 = " abc def ghi abcdef abc def abc def def" FOO2_remove = " \ def \ abc \ ghi \ "
The variable FOO
becomes
" 789 123456 "
and FOO2
becomes
" abcdef ".
Like "_append" and "_prepend", "_remove" is applied at variable expansion time.
An advantage of the override style operations
"_append", "_prepend", and "_remove" as compared to the
"+=" and "=+" operators is that the override style
operators provide guaranteed operations.
For example, consider a class foo.bbclass
that needs to add the value "val" to the variable
FOO
, and a recipe that uses
foo.bbclass
as follows:
inherit foo FOO = "initial"
If foo.bbclass
uses the "+=" operator,
as follows, then the final value of FOO
will be "initial", which is not what is desired:
FOO += "val"
If, on the other hand, foo.bbclass
uses the "_append" operator, then the final value of
FOO
will be "initial val", as intended:
FOO_append = " val"
FOO
:
FOO_append = "bar" FOO_append = "baz"The only effect of changing the second assignment in the previous example to use "+=" would be to add a space before "baz" in the appended value (due to how the "+=" operator works).
Another advantage of the override style operations is that you can combine them with other overrides as described in the "Conditional Syntax (Overrides)" section.
Variable flags are BitBake's implementation of variable properties or attributes. It is a way of tagging extra information onto a variable. You can find more out about variable flags in general in the "Variable Flags" section.
You can define, append, and prepend values to variable flags. All the standard syntax operations previously mentioned work for variable flags except for override style syntax (i.e. "_prepend", "_append", and "_remove").
Here are some examples showing how to set variable flags:
FOO[a] = "abc" FOO[b] = "123" FOO[a] += "456"
The variable FOO
has two flags:
[a]
and [b]
.
The flags are immediately set to "abc" and "123", respectively.
The [a]
flag becomes "abc 456".
No need exists to pre-define variable flags. You can simply start using them. One extremely common application is to attach some brief documentation to a BitBake variable as follows:
CACHE[doc] = "The directory holding the cache of the metadata."
You can use inline Python variable expansion to set variables. Here is an example:
DATE = "${@time.strftime('%Y%m%d',time.gmtime())}"
This example results in the DATE
variable being set to the current date.
Probably the most common use of this feature is to extract
the value of variables from BitBake's internal data dictionary,
d
.
The following lines select the values of a package name
and its version number, respectively:
PN = "${@bb.parse.BBHandler.vars_from_file(d.getVar('FILE', False),d)[0] or 'defaultpkgname'}" PV = "${@bb.parse.BBHandler.vars_from_file(d.getVar('FILE', False),d)[1] or '1.0'}"
foo()
is called each time FOO
is expanded:
FOO = "${@foo()}"Contrast this with the following immediate assignment, where
foo()
is only called once, while the
assignment is parsed:
FOO := "${@foo()}"
For a different way to set variables with Python code during parsing, see the "Anonymous Python Functions" section.
It is possible to completely remove a variable or a variable flag from BitBake's internal data dictionary by using the "unset" keyword. Here is an example:
unset DATE unset do_fetch[noexec]
These two statements remove the DATE
and the
do_fetch[noexec]
flag.
When specifying pathnames for use with BitBake, do not use the tilde ("~") character as a shortcut for your home directory. Doing so might cause BitBake to not recognize the path since BitBake does not expand this character in the same way a shell would.
Instead, provide a fuller path as the following example illustrates:
BBLAYERS ?= " \ /home/scott-lenovo/LayerA \ "
You can export variables to the environment of running
tasks by using the export
keyword.
For example, in the following example, the
do_foo
task prints "value from
the environment" when run:
export ENV_VARIABLE ENV_VARIABLE = "value from the environment" do_foo() { bbplain "$ENV_VARIABLE" }
$ENV_VARIABLE
in this case because it lacks the obligatory
{}
.
Rather, $ENV_VARIABLE
is expanded
by the shell.
It does not matter whether
export ENV_VARIABLE
appears before or
after assignments to ENV_VARIABLE
.
It is also possible to combine export
with setting a value for the variable.
Here is an example:
export ENV_VARIABLE = "variable-value
"
In the output of bitbake -e
, variables
that are exported to the environment are preceded by "export".
Among the variables commonly exported to the environment
are CC
and CFLAGS
,
which are picked up by many build systems.
BitBake uses
OVERRIDES
to control what variables are overridden after BitBake
parses recipes and configuration files.
This section describes how you can use
OVERRIDES
as conditional metadata,
talks about key expansion in relationship to
OVERRIDES
, and provides some examples
to help with understanding.
You can use OVERRIDES
to conditionally select
a specific version of a variable and to conditionally
append or prepend the value of a variable.
Selecting a Variable:
The OVERRIDES
variable is
a colon-character-separated list that contains items
for which you want to satisfy conditions.
Thus, if you have a variable that is conditional on “arm”, and “arm”
is in OVERRIDES
, then the “arm”-specific
version of the variable is used rather than the non-conditional
version.
Here is an example:
OVERRIDES = "architecture:os:machine" TEST = "default" TEST_os = "osspecific" TEST_nooverride = "othercondvalue"
In this example, the OVERRIDES
variable lists three overrides:
"architecture", "os", and "machine".
The variable TEST
by itself has a default
value of "default".
You select the os-specific version of the TEST
variable by appending the "os" override to the variable
(i.e.TEST_os
).
To better understand this, consider a practical example
that assumes an OpenEmbedded metadata-based Linux
kernel recipe file.
The following lines from the recipe file first set
the kernel branch variable KBRANCH
to a default value, then conditionally override that
value based on the architecture of the build:
KBRANCH = "standard/base" KBRANCH_qemuarm = "standard/arm-versatile-926ejs" KBRANCH_qemumips = "standard/mti-malta32" KBRANCH_qemuppc = "standard/qemuppc" KBRANCH_qemux86 = "standard/common-pc/base" KBRANCH_qemux86-64 = "standard/common-pc-64/base" KBRANCH_qemumips64 = "standard/mti-malta64"
Appending and Prepending:
BitBake also supports append and prepend operations to
variable values based on whether a specific item is
listed in OVERRIDES
.
Here is an example:
DEPENDS = "glibc ncurses" OVERRIDES = "machine:local" DEPENDS_append_machine = " libmad"
In this example, DEPENDS
becomes
"glibc ncurses libmad".
Again, using an OpenEmbedded metadata-based
kernel recipe file as an example, the
following lines will conditionally append to the
KERNEL_FEATURES
variable based
on the architecture:
KERNEL_FEATURES_append = " ${KERNEL_EXTRA_FEATURES}" KERNEL_FEATURES_append_qemux86=" cfg/sound.scc cfg/paravirt_kvm.scc" KERNEL_FEATURES_append_qemux86-64=" cfg/sound.scc cfg/paravirt_kvm.scc"
Setting a Variable for a Single Task: BitBake supports setting a variable just for the duration of a single task. Here is an example:
FOO_task-configure = "val 1" FOO_task-compile = "val 2"
In the previous example, FOO
has the value "val 1" while the
do_configure
task is executed,
and the value "val 2" while the
do_compile
task is executed.
Internally, this is implemented by prepending
the task (e.g. "task-compile:") to the value of
OVERRIDES
for the local datastore of the do_compile
task.
You can also use this syntax with other combinations
(e.g. "_prepend
") as shown in the
following example:
EXTRA_OEMAKE_prepend_task-compile = "${PARALLEL_MAKE} "
Key expansion happens when the BitBake datastore is finalized. To better understand this, consider the following example:
A${B} = "X" B = "2" A2 = "Y"
In this case, after all the parsing is complete,
BitBake expands ${B}
into "2".
This expansion causes A2
, which was
set to "Y" before the expansion, to become "X".
Despite the previous explanations that show the different forms of variable definitions, it can be hard to work out exactly what happens when variable operators, conditional overrides, and unconditional overrides are combined. This section presents some common scenarios along with explanations for variable interactions that typically confuse users.
There is often confusion concerning the order in which overrides and various "append" operators take effect. Recall that an append or prepend operation using "_append" and "_prepend" does not result in an immediate assignment as would "+=", ".=", "=+", or "=.". Consider the following example:
OVERRIDES = "foo" A = "Z" A_foo_append = "X"
For this case, A
is
unconditionally set to "Z" and "X" is
unconditionally and immediately appended to the variable
A_foo
.
Because overrides have not been applied yet,
A_foo
is set to "X" due to the append
and A
simply equals "Z".
Applying overrides, however, changes things.
Since "foo" is listed in OVERRIDES
,
the conditional variable A
is replaced
with the "foo" version, which is equal to "X".
So effectively, A_foo
replaces A
.
This next example changes the order of the override and the append:
OVERRIDES = "foo" A = "Z" A_append_foo = "X"
For this case, before overrides are handled,
A
is set to "Z" and A_append_foo
is set to "X".
Once the override for "foo" is applied, however,
A
gets appended with "X".
Consequently, A
becomes "ZX".
Notice that spaces are not appended.
This next example has the order of the appends and overrides reversed back as in the first example:
OVERRIDES = "foo" A = "Y" A_foo_append = "Z" A_foo_append = "X"
For this case, before any overrides are resolved,
A
is set to "Y" using an immediate assignment.
After this immediate assignment, A_foo
is set
to "Z", and then further appended with
"X" leaving the variable set to "ZX".
Finally, applying the override for "foo" results in the conditional
variable A
becoming "ZX" (i.e.
A
is replaced with A_foo
).
This final example mixes in some varying operators:
A = "1" A_append = "2" A_append = "3" A += "4" A .= "5"
For this case, the type of append operators are affecting the
order of assignments as BitBake passes through the code
multiple times.
Initially, A
is set to "1 45" because
of the three statements that use immediate operators.
After these assignments are made, BitBake applies the
"_append" operations.
Those operations result in A
becoming "1 4523".
BitBake allows for metadata sharing through include files
(.inc
) and class files
(.bbclass
).
For example, suppose you have a piece of common functionality
such as a task definition that you want to share between
more than one recipe.
In this case, creating a .bbclass
file that contains the common functionality and then using
the inherit
directive in your recipes to
inherit the class would be a common way to share the task.
This section presents the mechanisms BitBake provides to
allow you to share functionality between recipes.
Specifically, the mechanisms include include
,
inherit
, INHERIT
, and
require
directives.
BitBake uses the
BBPATH
variable to locate needed include and class files.
Additionally, BitBake searches the current directory for
include
and require
directives.
BBPATH
variable is analogous to
the environment variable PATH
.
In order for include and class files to be found by BitBake,
they need to be located in a "classes" subdirectory that can
be found in BBPATH
.
inherit
Directive¶
When writing a recipe or class file, you can use the
inherit
directive to inherit the
functionality of a class (.bbclass
).
BitBake only supports this directive when used within recipe
and class files (i.e. .bb
and
.bbclass
).
The inherit
directive is a rudimentary
means of specifying functionality contained in class files
that your recipes require.
For example, you can easily abstract out the tasks involved in
building a package that uses Autoconf and Automake and put
those tasks into a class file and then have your recipe
inherit that class file.
As an example, your recipes could use the following directive
to inherit an autotools.bbclass
file.
The class file would contain common functionality for using
Autotools that could be shared across recipes:
inherit autotools
In this case, BitBake would search for the directory
classes/autotools.bbclass
in BBPATH
.
If you want to use the directive to inherit
multiple classes, separate them with spaces.
The following example shows how to inherit both the
buildhistory
and rm_work
classes:
inherit buildhistory rm_work
An advantage with the inherit directive as compared to both
the
include and
require directives
is that you can inherit class files conditionally.
You can accomplish this by using a variable expression
after the inherit
statement.
Here is an example:
inherit ${VARNAME}
If VARNAME
is going to be set, it needs
to be set before the inherit
statement
is parsed.
One way to achieve a conditional inherit in this case is to use
overrides:
VARIABLE = "" VARIABLE_someoverride = "myclass"
Another method is by using anonymous Python. Here is an example:
python () { if condition == value: d.setVar('VARIABLE', 'myclass') else: d.setVar('VARIABLE', '') }
Alternatively, you could use an in-line Python expression in the following form:
inherit ${@'classname' if condition else ''} inherit ${@functionname(params)}
In all cases, if the expression evaluates to an empty string, the statement does not trigger a syntax error because it becomes a no-op.
include
Directive¶
BitBake understands the include
directive.
This directive causes BitBake to parse whatever file you specify,
and to insert that file at that location.
The directive is much like its equivalent in Make except
that if the path specified on the include line is a relative
path, BitBake locates the first file it can find
within BBPATH
.
The include directive is a more generic method of including
functionality as compared to the
inherit directive,
which is restricted to class (i.e. .bbclass
)
files.
The include directive is applicable for any other kind of
shared or encapsulated functionality or configuration that
does not suit a .bbclass
file.
As an example, suppose you needed a recipe to include some self-test definitions:
include test_defs.inc
include
directive does not
produce an error when the file cannot be found.
Consequently, it is recommended that if the file you
are including is expected to exist, you should use
require
instead of include
.
Doing so makes sure that an error is produced if the
file cannot be found.
require
Directive¶
BitBake understands the require
directive.
This directive behaves just like the
include
directive with the exception that
BitBake raises a parsing error if the file to be included cannot
be found.
Thus, any file you require is inserted into the file that is
being parsed at the location of the directive.
The require directive, like the include directive previously
described, is a more generic method of including
functionality as compared to the
inherit directive,
which is restricted to class (i.e. .bbclass
)
files.
The require directive is applicable for any other kind of
shared or encapsulated functionality or configuration that
does not suit a .bbclass
file.
Similar to how BitBake handles
include
,
if the path specified
on the require line is a relative path, BitBake locates
the first file it can find within BBPATH
.
As an example, suppose you have two versions of a recipe
(e.g. foo_1.2.2.bb
and
foo_2.0.0.bb
) where
each version contains some identical functionality that could be
shared.
You could create an include file named foo.inc
that contains the common definitions needed to build "foo".
You need to be sure foo.inc
is located in the
same directory as your two recipe files as well.
Once these conditions are set up, you can share the functionality
using a require
directive from within each
recipe:
require foo.inc
INHERIT
Configuration Directive¶
When creating a configuration file (.conf
),
you can use the
INHERIT
configuration directive to inherit a class.
BitBake only supports this directive when used within
a configuration file.
As an example, suppose you needed to inherit a class
file called abc.bbclass
from a
configuration file as follows:
INHERIT += "abc"
This configuration directive causes the named
class to be inherited at the point of the directive
during parsing.
As with the inherit
directive, the
.bbclass
file must be located in a
"classes" subdirectory in one of the directories specified
in BBPATH
.
.conf
files are parsed
first during BitBake's execution, using
INHERIT
to inherit a class effectively
inherits the class globally (i.e. for all recipes).
If you want to use the directive to inherit
multiple classes, you can provide them on the same line in the
local.conf
file.
Use spaces to separate the classes.
The following example shows how to inherit both the
autotools
and pkgconfig
classes:
INHERIT += "autotools pkgconfig"
As with most languages, functions are the building blocks that are used to build up operations into tasks. BitBake supports these types of functions:
Shell Functions: Functions written in shell script and executed either directly as functions, tasks, or both. They can also be called by other shell functions.
BitBake-Style Python Functions:
Functions written in Python and executed by BitBake or other
Python functions using bb.build.exec_func()
.
Python Functions: Functions written in Python and executed by Python.
Anonymous Python Functions: Python functions executed automatically during parsing.
Regardless of the type of function, you can only
define them in class (.bbclass
)
and recipe (.bb
or .inc
)
files.
Functions written in shell script and executed either directly as functions, tasks, or both. They can also be called by other shell functions. Here is an example shell function definition:
some_function () { echo "Hello World" }
When you create these types of functions in your recipe
or class files, you need to follow the shell programming
rules.
The scripts are executed by /bin/sh
,
which may not be a bash shell but might be something
such as dash
.
You should not use Bash-specific script (bashisms).
Overrides and override-style operators like
_append
and
_prepend
can also be applied to
shell functions.
Most commonly, this application would be used in a
.bbappend
file to modify functions in
the main recipe.
It can also be used to modify functions inherited from
classes.
As an example, consider the following:
do_foo() { bbplain first fn } fn_prepend() { bbplain second } fn() { bbplain third } do_foo_append() { bbplain fourth }
Running do_foo
prints the following:
recipename do_foo: first recipename do_foo: second recipename do_foo: third recipename do_foo: fourth
You can use the bitbake -e
recipename
command to view the final assembled function
after all overrides have been applied.
These functions are written in Python and executed by
BitBake or other Python functions using
bb.build.exec_func()
.
An example BitBake function is:
python some_python_function () { d.setVar("TEXT", "Hello World") print d.getVar("TEXT") }
Because the Python "bb" and "os" modules are already imported, you do not need to import these modules. Also in these types of functions, the datastore ("d") is a global variable and is always automatically available.
${X}
)
are no longer expanded within Python functions.
This behavior is intentional in order to allow you
to freely set variable values to expandable expressions
without having them expanded prematurely.
If you do wish to expand a variable within a Python
function, use d.getVar("X")
.
Or, for more complicated expressions, use
d.expand()
.
Similar to shell functions, you can also apply overrides and override-style operators to BitBake-style Python functions.
As an example, consider the following:
python do_foo_prepend() { bb.plain("first") } python do_foo() { bb.plain("second") } python do_foo_append() { bb.plain("third") }
Running do_foo
prints
the following:
recipename do_foo: first recipename do_foo: second recipename do_foo: third
You can use the bitbake -e
recipename
command to view the final assembled function
after all overrides have been applied.
These functions are written in Python and are executed by other Python code. Examples of Python functions are utility functions that you intend to call from in-line Python or from within other Python functions. Here is an example:
def get_depends(d): if d.getVar('SOMECONDITION'): return "dependencywithcond" else: return "dependency" SOMECONDITION = "1" DEPENDS = "${@get_depends(d)}"
This would result in DEPENDS
containing dependencywithcond
.
Here are some things to know about Python functions:
Python functions can take parameters.
The BitBake datastore is not automatically available. Consequently, you must pass it in as a parameter to the function.
The "bb" and "os" Python modules are automatically available. You do not need to import them.
Following are some important differences between BitBake-style Python functions and regular Python functions defined with "def":
Only BitBake-style Python functions can be tasks.
Overrides and override-style operators can only be applied to BitBake-style Python functions.
Only regular Python functions can take arguments and return values.
Variable flags
such as [dirs]
,
[cleandirs]
, and
[lockfiles]
can be used
on BitBake-style Python functions, but not on
regular Python functions.
BitBake-style Python functions generate a separate
${
T
}/run.
function-name
.
pid
script that is executed to run the function, and also
generate a log file in
${T}/log.
function-name
.
pid
if they are executed as tasks.
Regular Python functions execute "inline" and do not
generate any files in ${T}
.
Regular Python functions are called with the usual
Python syntax.
BitBake-style Python functions are usually tasks and
are called directly by BitBake, but can also be called
manually from Python code by using the
bb.build.exec_func()
function.
Here is an example:
bb.build.exec_func("my_bitbake_style_function", d)
bb.build.exec_func()
can also
be used to run shell functions from Python code.
If you want to run a shell function before a Python
function within the same task, then you can use a
parent helper Python function that starts by running
the shell function with
bb.build.exec_func()
and then
runs the Python code.
To detect errors from functions executed with
bb.build.exec_func()
, you
can catch the bb.build.FuncFailed
exception.
bb.build.FuncFailed
.
Rather, bb.build.FuncFailed
should be viewed as a general indicator that the
called function failed by raising an exception.
For example, an exception raised by
bb.fatal()
will be caught inside
bb.build.exec_func()
, and a
bb.build.FuncFailed
will be raised
in response.
Due to their simplicity, you should prefer regular Python functions
over BitBake-style Python functions unless you need a feature specific
to BitBake-style Python functions.
Regular Python functions in metadata are a more recent invention than
BitBake-style Python functions, and older code tends to use
bb.build.exec_func()
more often.
Sometimes it is useful to set variables or perform other operations programmatically during parsing. To do this, you can define special Python functions, called anonymous Python functions, that run at the end of parsing. For example, the following conditionally sets a variable based on the value of another variable:
python () { if d.getVar('SOMEVAR') == 'value': d.setVar('ANOTHERVAR', 'value2') }
An equivalent way to mark a function as an anonymous function is to give it the name "__anonymous", rather than no name.
Anonymous Python functions always run at the end of parsing, regardless of where they are defined. If a recipe contains many anonymous functions, they run in the same order as they are defined within the recipe. As an example, consider the following snippet:
python () { d.setVar('FOO', 'foo 2') } FOO = "foo 1" python () { d.appendVar('BAR', ' bar 2') } BAR = "bar 1"
The previous example is conceptually equivalent to the following snippet:
FOO = "foo 1" BAR = "bar 1" FOO = "foo 2" BAR += "bar 2"
FOO
ends up with the value "foo 2",
and BAR
with the value "bar 1 bar 2".
Just as in the second snippet, the values set for the
variables within the anonymous functions become available
to tasks, which always run after parsing.
Overrides and override-style operators such as
"_append
" are applied before
anonymous functions run.
In the following example, FOO
ends
up with the value "foo from anonymous":
FOO = "foo" FOO_append = " from outside" python () { d.setVar("FOO", "foo from anonymous") }
For methods you can use with anonymous Python functions, see the "Functions You Can Call From Within Python" section. For a different method to run Python code during parsing, see the "Inline Python Variable Expansion" section.
Through coding techniques and the use of
EXPORT_FUNCTIONS
, BitBake supports
exporting a function from a class such that the
class function appears as the default implementation
of the function, but can still be called if a recipe
inheriting the class needs to define its own version of
the function.
To understand the benefits of this feature, consider
the basic scenario where a class defines a task function
and your recipe inherits the class.
In this basic scenario, your recipe inherits the task
function as defined in the class.
If desired, your recipe can add to the start and end of the
function by using the "_prepend" or "_append" operations
respectively, or it can redefine the function completely.
However, if it redefines the function, there is
no means for it to call the class version of the function.
EXPORT_FUNCTIONS
provides a mechanism
that enables the recipe's version of the function to call
the original version of the function.
To make use of this technique, you need the following things in place:
The class needs to define the function as follows:
classname
_
functionname
For example, if you have a class file
bar.bbclass
and a function named
do_foo
, the class must define the function
as follows:
bar_do_foo
The class needs to contain the EXPORT_FUNCTIONS
statement as follows:
EXPORT_FUNCTIONS functionname
For example, continuing with the same example, the
statement in the bar.bbclass
would be
as follows:
EXPORT_FUNCTIONS do_foo
You need to call the function appropriately from within your
recipe.
Continuing with the same example, if your recipe
needs to call the class version of the function,
it should call bar_do_foo
.
Assuming do_foo
was a shell function
and EXPORT_FUNCTIONS
was used as above,
the recipe's function could conditionally call the
class version of the function as follows:
do_foo() { if [ somecondition ] ; then bar_do_foo else # Do something else fi }
To call your modified version of the function as defined
in your recipe, call it as do_foo
.
With these conditions met, your single recipe can freely choose between the original function as defined in the class file and the modified function in your recipe. If you do not set up these conditions, you are limited to using one function or the other.
Tasks are BitBake execution units that make up the
steps that BitBake can run for a given recipe.
Tasks are only supported in recipes and classes
(i.e. in .bb
files and files
included or inherited from .bb
files).
By convention, tasks have names that start with "do_".
Tasks are either
shell functions or
BitBake-style Python functions
that have been promoted to tasks by using the
addtask
command.
The addtask
command can also
optionally describe dependencies between the
task and other tasks.
Here is an example that shows how to define a task
and declare some dependencies:
python do_printdate () { import time print time.strftime('%Y%m%d', time.gmtime()) } addtask printdate after do_fetch before do_build
The first argument to addtask
is the name of the function to promote to
a task.
If the name does not start with "do_", "do_" is
implicitly added, which enforces the convention that
all task names start with "do_".
In the previous example, the
do_printdate
task becomes a
dependency of the do_build
task, which is the default task (i.e. the task run by
the bitbake
command unless
another task is specified explicitly).
Additionally, the do_printdate
task becomes dependent upon the
do_fetch
task.
Running the do_build
task
results in the do_printdate
task running first.
do_printdate
task is only run
the first time you build the recipe with
the bitbake
command.
This is because BitBake considers the task "up-to-date"
after that initial run.
If you want to force the task to always be rerun for
experimentation purposes, you can make BitBake always
consider the task "out-of-date" by using the
[
nostamp
]
variable flag, as follows:
do_printdate[nostamp] = "1"You can also explicitly run the task and provide the
-f
option as follows:
$ bitbake recipe
-c printdate -f
When manually selecting a task to run with the
bitbake
recipe
-c
task
command, you can omit the "do_" prefix as part of the
task name.
You might wonder about the practical effects of using
addtask
without specifying any
dependencies as is done in the following example:
addtask printdate
In this example, assuming dependencies have not been
added through some other means, the only way to run
the task is by explicitly selecting it with
bitbake
recipe
-c printdate
.
You can use the
do_listtasks
task to list all tasks
defined in a recipe as shown in the following example:
$ bitbake recipe
-c listtasks
For more information on task dependencies, see the "Dependencies" section.
See the "Variable Flags" section for information on variable flags you can use with tasks.
As well as being able to add tasks, you can delete them.
Simply use the deltask
command to
delete a task.
For example, to delete the example task used in the previous
sections, you would use:
deltask printdate
If you delete a task using the deltask
command and the task has dependencies, the dependencies are
not reconnected.
For example, suppose you have three tasks named
do_a
, do_b
, and
do_c
.
Furthermore, do_c
is dependent on
do_b
, which in turn is dependent on
do_a
.
Given this scenario, if you use deltask
to delete do_b
, the implicit dependency
relationship between do_c
and
do_a
through do_b
no longer exists, and do_c
dependencies
are not updated to include do_a
.
Thus, do_c
is free to run before
do_a
.
If you want dependencies such as these to remain intact, use
the [noexec]
varflag to disable the task
instead of using the deltask
command to
delete it:
do_b[noexec] = "1"
When running a task, BitBake tightly controls the shell execution environment of the build tasks to make sure unwanted contamination from the build machine cannot influence the build.
BB_PRESERVE_ENV
variable.
Consequently, if you do want something to get passed into the build task environment, you must take these two steps:
Tell BitBake to load what you want from the environment
into the datastore.
You can do so through the
BB_ENV_WHITELIST
and
BB_ENV_EXTRAWHITE
variables.
For example, assume you want to prevent the build system from
accessing your $HOME/.ccache
directory.
The following command "whitelists" the environment variable
CCACHE_DIR
causing BitBake to allow that
variable into the datastore:
export BB_ENV_EXTRAWHITE="$BB_ENV_EXTRAWHITE CCACHE_DIR"
Tell BitBake to export what you have loaded into the
datastore to the task environment of every running task.
Loading something from the environment into the datastore
(previous step) only makes it available in the datastore.
To export it to the task environment of every running task,
use a command similar to the following in your local configuration
file local.conf
or your
distribution configuration file:
export CCACHE_DIR
Sometimes, it is useful to be able to obtain information
from the original execution environment.
BitBake saves a copy of the original environment into
a special variable named
BB_ORIGENV
.
The BB_ORIGENV
variable returns a datastore
object that can be queried using the standard datastore operators
such as getVar(, False)
.
The datastore object is useful, for example, to find the original
DISPLAY
variable.
Here is an example:
origenv = d.getVar("BB_ORIGENV", False) bar = origenv.getVar("BAR", False)
The previous example returns BAR
from the original
execution environment.
Variable flags (varflags) help control a task's functionality and dependencies. BitBake reads and writes varflags to the datastore using the following command forms:
variable
= d.getVarFlags("variable
") self.d.setVarFlags("FOO", {"func": True})
When working with varflags, the same syntax, with the exception of overrides, applies. In other words, you can set, append, and prepend varflags just like variables. See the "Variable Flag Syntax" section for details.
BitBake has a defined set of varflags available for recipes and classes. Tasks support a number of these flags which control various functionality of the task:
[cleandirs]
:
Empty directories that should be created before the
task runs.
Directories that already exist are removed and recreated
to empty them.
[depends]
:
Controls inter-task dependencies.
See the
DEPENDS
variable and the
"Inter-Task Dependencies"
section for more information.
[deptask]
:
Controls task build-time dependencies.
See the
DEPENDS
variable and the
"Build Dependencies"
section for more information.
[dirs]
:
Directories that should be created before the task runs.
Directories that already exist are left as is.
The last directory listed is used as the
current working directory for the task.
[lockfiles]
:
Specifies one or more lockfiles to lock while the task
executes.
Only one task may hold a lockfile, and any task that
attempts to lock an already locked file will block until
the lock is released.
You can use this variable flag to accomplish mutual
exclusion.
[noexec]
:
When set to "1", marks the task as being empty, with
no execution required.
You can use the [noexec]
flag to set up
tasks as dependency placeholders, or to disable tasks defined
elsewhere that are not needed in a particular recipe.
[nostamp]
:
When set to "1", tells BitBake to not generate a stamp
file for a task, which implies the task should always
be executed.
[nostamp]
task will always be
executed as well.
This can cause unnecessary rebuilding if you are
not careful.
[number_threads]
:
Limits tasks to a specific number of simultaneous threads
during execution.
This varflag is useful when your build host has a large number
of cores but certain tasks need to be rate-limited due to various
kinds of resource constraints (e.g. to avoid network throttling).
number_threads
works similarly to the
BB_NUMBER_THREADS
variable but is task-specific.
Set the value globally.
For example, the following makes sure the
do_fetch
task uses no more than two
simultaneous execution threads:
do_fetch[number_threads] = "2"
Setting the varflag in individual recipes rather than globally can result in unpredictable behavior.
Setting the varflag to a value greater than the
value used in the BB_NUMBER_THREADS
variable causes number_threads
to have no effect.
[postfuncs]
:
List of functions to call after the completion of the task.
[prefuncs]
:
List of functions to call before the task executes.
[rdepends]
:
Controls inter-task runtime dependencies.
See the
RDEPENDS
variable, the
RRECOMMENDS
variable, and the
"Inter-Task Dependencies"
section for more information.
[rdeptask]
:
Controls task runtime dependencies.
See the
RDEPENDS
variable, the
RRECOMMENDS
variable, and the
"Runtime Dependencies"
section for more information.
[recideptask]
:
When set in conjunction with
recrdeptask
, specifies a task that
should be inspected for additional dependencies.
[recrdeptask]
:
Controls task recursive runtime dependencies.
See the
RDEPENDS
variable, the
RRECOMMENDS
variable, and the
"Recursive Dependencies"
section for more information.
[stamp-extra-info]
:
Extra stamp information to append to the task's stamp.
As an example, OpenEmbedded uses this flag to allow
machine-specific tasks.
[umask]
:
The umask to run the task under.
Several varflags are useful for controlling how signatures are calculated for variables. For more information on this process, see the "Checksums (Signatures)" section.
[vardeps]
:
Specifies a space-separated list of additional
variables to add to a variable's dependencies
for the purposes of calculating its signature.
Adding variables to this list is useful, for example, when
a function refers to a variable in a manner that
does not allow BitBake to automatically determine
that the variable is referred to.
[vardepsexclude]
:
Specifies a space-separated list of variables
that should be excluded from a variable's dependencies
for the purposes of calculating its signature.
[vardepvalue]
:
If set, instructs BitBake to ignore the actual
value of the variable and instead use the specified
value when calculating the variable's signature.
[vardepvalueexclude]
:
Specifies a pipe-separated list of strings to exclude
from the variable's value when calculating the
variable's signature.
BitBake allows installation of event handlers within recipe
and class files.
Events are triggered at certain points during operation, such
as the beginning of operation against a given recipe
(i.e. *.bb
), the start of a given task,
a task failure, a task success, and so forth.
The intent is to make it easy to do things like email
notification on build failures.
Following is an example event handler that prints the name
of the event and the content of the
FILE
variable:
addhandler myclass_eventhandler python myclass_eventhandler() { from bb.event import getName print("The name of the Event is %s" % getName(e)) print("The file we run for is %s" % d.getVar('FILE')) } myclass_eventhandler[eventmask] = "bb.event.BuildStarted bb.event.BuildCompleted"
In the previous example, an eventmask has been set so that
the handler only sees the "BuildStarted" and "BuildCompleted"
events.
This event handler gets called every time an event matching
the eventmask is triggered.
A global variable "e" is defined, which represents the current
event.
With the getName(e)
method, you can get
the name of the triggered event.
The global datastore is available as "d".
In legacy code, you might see "e.data" used to get the datastore.
However, realize that "e.data" is deprecated and you should use
"d" going forward.
The context of the datastore is appropriate to the event in question. For example, "BuildStarted" and "BuildCompleted" events run before any tasks are executed so would be in the global configuration datastore namespace. No recipe-specific metadata exists in that namespace. The "BuildStarted" and "BuildCompleted" events also run in the main cooker/server process rather than any worker context. Thus, any changes made to the datastore would be seen by other cooker/server events within the current build but not seen outside of that build or in any worker context. Task events run in the actual tasks in question consequently have recipe-specific and task-specific contents. These events run in the worker context and are discarded at the end of task execution.
During a standard build, the following common events might occur. The following events are the most common kinds of events that most metadata might have an interest in viewing:
bb.event.ConfigParsed()
:
Fired when the base configuration; which consists of
bitbake.conf
,
base.bbclass
and any global
INHERIT
statements; has been parsed.
You can see multiple such events when each of the
workers parse the base configuration or if the server
changes configuration and reparses.
Any given datastore only has one such event executed
against it, however.
If
BB_INVALIDCONF
is set in the datastore by the event handler, the
configuration is reparsed and a new event triggered,
allowing the metadata to update configuration.
bb.event.HeartbeatEvent()
:
Fires at regular time intervals of one second.
You can configure the interval time using the
BB_HEARTBEAT_EVENT
variable.
The event's "time" attribute is the
time.time()
value when the
event is triggered.
This event is useful for activities such as
system state monitoring.
bb.event.ParseStarted()
:
Fired when BitBake is about to start parsing recipes.
This event's "total" attribute represents the number of
recipes BitBake plans to parse.
bb.event.ParseProgress()
:
Fired as parsing progresses.
This event's "current" attribute is the number of
recipes parsed as well as the "total" attribute.
bb.event.ParseCompleted()
:
Fired when parsing is complete.
This event's "cached", "parsed", "skipped", "virtuals",
"masked", and "errors" attributes provide statistics
for the parsing results.
bb.event.BuildStarted()
:
Fired when a new build starts.
BitBake fires multiple "BuildStarted" events (one per configuration)
when multiple configuration (multiconfig) is enabled.
bb.build.TaskStarted()
:
Fired when a task starts.
This event's "taskfile" attribute points to the recipe
from which the task originates.
The "taskname" attribute, which is the task's name,
includes the do_
prefix, and the
"logfile" attribute point to where the task's output is
stored.
Finally, the "time" attribute is the task's execution start
time.
bb.build.TaskInvalid()
:
Fired if BitBake tries to execute a task that does not exist.
bb.build.TaskFailedSilent()
:
Fired for setscene tasks that fail and should not be
presented to the user verbosely.
bb.build.TaskFailed()
:
Fired for normal tasks that fail.
bb.build.TaskSucceeded()
:
Fired when a task successfully completes.
bb.event.BuildCompleted()
:
Fired when a build finishes.
bb.cooker.CookerExit()
:
Fired when the BitBake server/cooker shuts down.
This event is usually only seen by the UIs as a
sign they should also shutdown.
This next list of example events occur based on specific requests to the server. These events are often used to communicate larger pieces of information from the BitBake server to other parts of BitBake such as user interfaces:
bb.event.TreeDataPreparationStarted()
bb.event.TreeDataPreparationProgress()
bb.event.TreeDataPreparationCompleted()
bb.event.DepTreeGenerated()
bb.event.CoreBaseFilesFound()
bb.event.ConfigFilePathFound()
bb.event.FilesMatchingFound()
bb.event.ConfigFilesFound()
bb.event.TargetsTreeGenerated()
BitBake supports two features that facilitate creating
from a single recipe file multiple incarnations of that
recipe file where all incarnations are buildable.
These features are enabled through the
BBCLASSEXTEND
and
BBVERSIONS
variables.
PROVIDES
,
PN
, and
DEPENDS
variables would need to be modified by the extension class.
For specific examples, see the OE-Core
native
, nativesdk
,
and multilib
classes.
BBCLASSEXTEND
:
This variable is a space separated list of classes used to "extend" the
recipe for each variant.
Here is an example that results in a second incarnation of the current
recipe being available.
This second incarnation will have the "native" class inherited.
BBCLASSEXTEND = "native"
BBVERSIONS
:
This variable allows a single recipe to build multiple versions of a
project from a single recipe file.
You can also specify conditional metadata
(using the
OVERRIDES
mechanism) for a single version, or an optionally named range of versions.
Here is an example:
BBVERSIONS = "1.0 2.0 git" SRC_URI_git = "git://someurl/somepath.git" BBVERSIONS = "1.0.[0-6]:1.0.0+ \ 1.0.[7-9]:1.0.7+" SRC_URI_append_1.0.7+ = "file://some_patch_which_the_new_versions_need.patch;patch=1"
The name of the range defaults to the original version of the
recipe.
For example, in OpenEmbedded, the recipe file
foo_1.0.0+.bb
creates a default name range
of 1.0.0+
.
This is useful because the range name is not only placed
into overrides, but it is also made available for the metadata to use
in the variable that defines the base recipe versions for use in
file://
search paths
(FILESPATH
).
To allow for efficient parallel processing, BitBake handles dependencies at the task level. Dependencies can exist both between tasks within a single recipe and between tasks in different recipes. Following are examples of each:
For tasks within a single recipe, a
recipe's do_configure
task might need to complete before its
do_compile
task can run.
For tasks in different recipes, one
recipe's do_configure
task might require another recipe's
do_populate_sysroot
task to finish first such that the libraries and headers
provided by the other recipe are available.
This section describes several ways to declare dependencies. Remember, even though dependencies are declared in different ways, they are all simply dependencies between tasks.
.bb
File¶
BitBake uses the addtask
directive
to manage dependencies that are internal to a given recipe
file.
You can use the addtask
directive to
indicate when a task is dependent on other tasks or when
other tasks depend on that recipe.
Here is an example:
addtask printdate after do_fetch before do_build
In this example, the do_printdate
task depends on the completion of the
do_fetch
task, and the
do_build
task depends on the
completion of the do_printdate
task.
For a task to run, it must be a direct or indirect dependency of some other task that is scheduled to run.
For illustration, here are some examples:
The directive
addtask mytask before do_configure
causes do_mytask
to run before
do_configure
runs.
Be aware that do_mytask
still only
runs if its input checksum
has changed since the last time it was run.
Changes to the input checksum of
do_mytask
also indirectly cause
do_configure
to run.
The directive
addtask mytask after do_configure
by itself never causes do_mytask
to run.
do_mytask
can still be run manually
as follows:
$ bitbake recipe
-c mytask
Declaring do_mytask
as a dependency
of some other task that is scheduled to run also causes
it to run.
Regardless, the task runs after
do_configure
.
BitBake uses the
DEPENDS
variable to manage build time dependencies.
The [deptask]
varflag for tasks
signifies the task of each
item listed in DEPENDS
that must
complete before that task can be executed.
Here is an example:
do_configure[deptask] = "do_populate_sysroot"
In this example, the do_populate_sysroot
task of each item in DEPENDS
must complete before
do_configure
can execute.
BitBake uses the
PACKAGES
,
RDEPENDS
, and
RRECOMMENDS
variables to manage runtime dependencies.
The PACKAGES
variable lists runtime
packages.
Each of those packages can have RDEPENDS
and
RRECOMMENDS
runtime dependencies.
The [rdeptask]
flag for tasks is used to
signify the task of each
item runtime dependency which must have completed before that
task can be executed.
do_package_qa[rdeptask] = "do_packagedata"
In the previous example, the do_packagedata
task of each item in RDEPENDS
must have
completed before do_package_qa
can execute.
Although RDEPENDS
contains entries from the
runtime dependency namespace, BitBake knows how to map them back
to the build-time dependency namespace, in which the tasks are defined.
BitBake uses the [recrdeptask]
flag to manage
recursive task dependencies.
BitBake looks through the build-time and runtime
dependencies of the current recipe, looks through
the task's inter-task
dependencies, and then adds dependencies for the
listed task.
Once BitBake has accomplished this, it recursively works through
the dependencies of those tasks.
Iterative passes continue until all dependencies are discovered
and added.
The [recrdeptask]
flag is most commonly
used in high-level
recipes that need to wait for some task to finish "globally".
For example, image.bbclass
has the following:
do_rootfs[recrdeptask] += "do_packagedata"
This statement says that the do_packagedata
task of the current recipe and all recipes reachable
(by way of dependencies) from the
image recipe must run before the do_rootfs
task can run.
BitBake allows a task to recursively depend on itself by referencing itself in the task list:
do_a[recrdeptask] = "do_a do_b"
In the same way as before, this means that the do_a
and do_b
tasks of the current recipe and all
recipes reachable (by way of dependencies) from the recipe
must run before the do_a
task can run. In this
case BitBake will ignore the current recipe's do_a
task circular dependency on itself.
BitBake uses the [depends]
flag in a more generic form
to manage inter-task dependencies.
This more generic form allows for inter-dependency
checks for specific tasks rather than checks for
the data in DEPENDS
.
Here is an example:
do_patch[depends] = "quilt-native:do_populate_sysroot"
In this example, the do_populate_sysroot
task of the target quilt-native
must have completed before the
do_patch
task can execute.
The [rdepends]
flag works in a similar
way but takes targets
in the runtime namespace instead of the build-time dependency
namespace.
BitBake provides many functions you can call from within Python functions. This section lists the most commonly used functions, and mentions where to find others.
It is often necessary to access variables in the BitBake datastore using Python functions. The BitBake datastore has an API that allows you this access. Here is a list of available operations:
Operation | Description |
---|---|
d.getVar("X", expand) | Returns the value of variable "X". Using "expand=True" expands the value. Returns "None" if the variable "X" does not exist. |
d.setVar("X", "value") | Sets the variable "X" to "value". |
d.appendVar("X", "value") | Adds "value" to the end of the variable "X".
Acts like d.setVar("X", "value")
if the variable "X" does not exist. |
d.prependVar("X", "value") | Adds "value" to the start of the variable "X".
Acts like d.setVar("X", "value")
if the variable "X" does not exist. |
d.delVar("X") | Deletes the variable "X" from the datastore. Does nothing if the variable "X" does not exist. |
d.renameVar("X", "Y") | Renames the variable "X" to "Y". Does nothing if the variable "X" does not exist. |
d.getVarFlag("X", flag, expand) | Returns the value of variable "X". Using "expand=True" expands the value. Returns "None" if either the variable "X" or the named flag does not exist. |
d.setVarFlag("X", flag, "value") | Sets the named flag for variable "X" to "value". |
d.appendVarFlag("X", flag, "value") | Appends "value" to the named flag on the
variable "X".
Acts like d.setVarFlag("X", flag, "value")
if the named flag does not exist. |
d.prependVarFlag("X", flag, "value") | Prepends "value" to the named flag on
the variable "X".
Acts like d.setVarFlag("X", flag, "value")
if the named flag does not exist. |
d.delVarFlag("X", flag) | Deletes the named flag on the variable "X" from the datastore. |
d.setVarFlags("X", flagsdict) | Sets the flags specified in
the flagsdict() parameter.
setVarFlags does not clear previous flags.
Think of this operation as addVarFlags . |
d.getVarFlags("X") | Returns a flagsdict
of the flags for the variable "X".
Returns "None" if the variable "X" does not exist. |
d.delVarFlags("X") | Deletes all the flags for the variable "X". Does nothing if the variable "X" does not exist. |
d.expand(expression) | Expands variable references in the specified
string expression.
References to variables that do not exist are left as is.
For example, d.expand("foo ${X}")
expands to the literal string "foo ${X}" if the
variable "X" does not exist. |
You can find many other functions that can be called
from Python by looking at the source code of the
bb
module, which is in
bitbake/lib/bb
.
For example,
bitbake/lib/bb/utils.py
includes
the commonly used functions
bb.utils.contains()
and
bb.utils.mkdirhier()
, which come
with docstrings.
BitBake uses checksums (or signatures) along with the setscene to determine if a task needs to be run. This section describes the process. To help understand how BitBake does this, the section assumes an OpenEmbedded metadata-based example.
These checksums are stored in
STAMP
.
You can examine the checksums using the following BitBake command:
$ bitbake-dumpsigs
This command returns the signature data in a readable format
that allows you to examine the inputs used when the
OpenEmbedded build system generates signatures.
For example, using bitbake-dumpsigs
allows you to examine the do_compile
task's “sigdata” for a C application (e.g.
bash
).
Running the command also reveals that the “CC” variable is part of
the inputs that are hashed.
Any changes to this variable would invalidate the stamp and
cause the do_compile
task to run.
The following list describes related variables:
BB_HASHCHECK_FUNCTION
:
Specifies the name of the function to call during
the "setscene" part of the task's execution in order
to validate the list of task hashes.
BB_SETSCENE_DEPVALID
:
Specifies a function BitBake calls that determines
whether BitBake requires a setscene dependency to
be met.
BB_SETSCENE_VERIFY_FUNCTION2
:
Specifies a function to call that verifies the list of
planned task execution before the main task execution
happens.
BB_STAMP_POLICY
:
Defines the mode for comparing timestamps of stamp files.
BB_STAMP_WHITELIST
:
Lists stamp files that are looked at when the stamp policy
is "whitelist".
BB_TASKHASH
:
Within an executing task, this variable holds the hash
of the task as returned by the currently enabled
signature generator.
STAMP
:
The base path to create stamp files.
STAMPCLEAN
:
Again, the base path to create stamp files but can use wildcards
for matching a range of files for clean operations.
Support for wildcard use in variables varies depending on the
context in which it is used.
For example, some variables and file names allow limited use of
wildcards through the "%
" and
"*
" characters.
Other variables or names support Python's
glob
syntax,
fnmatch
syntax, or
Regular Expression (re)
syntax.
For variables that have wildcard suport, the documentation describes which form of wildcard, its use, and its limitations.
file://
)http://
, ftp://
, https://
)(cvs://
)svn://
)git://
)gitsm://
)ccrc://
)p4://
)repo://
)BitBake's fetch module is a standalone piece of library code that deals with the intricacies of downloading source code and files from remote systems. Fetching source code is one of the cornerstones of building software. As such, this module forms an important part of BitBake.
The current fetch module is called "fetch2" and refers to the fact that it is the second major version of the API. The original version is obsolete and has been removed from the codebase. Thus, in all cases, "fetch" refers to "fetch2" in this manual.
BitBake takes several steps when fetching source code or files. The fetcher codebase deals with two distinct processes in order: obtaining the files from somewhere (cached or otherwise) and then unpacking those files into a specific location and perhaps in a specific way. Getting and unpacking the files is often optionally followed by patching. Patching, however, is not covered by this module.
The code to execute the first part of this process, a fetch, looks something like the following:
src_uri = (d.getVar('SRC_URI') or "").split() fetcher = bb.fetch2.Fetch(src_uri, d) fetcher.download()
This code sets up an instance of the fetch class.
The instance uses a space-separated list of URLs from the
SRC_URI
variable and then calls the download
method to download the files.
The instantiation of the fetch class is usually followed by:
rootdir = l.getVar('WORKDIR') fetcher.unpack(rootdir)
This code unpacks the downloaded files to the
specified by WORKDIR
.
base.bbclass
.
The SRC_URI
and WORKDIR
variables are not hardcoded into the fetcher, since those fetcher
methods can be (and are) called with different variable names.
In OpenEmbedded for example, the shared state (sstate) code uses
the fetch module to fetch the sstate files.
When the download()
method is called,
BitBake tries to resolve the URLs by looking for source files
in a specific search order:
Pre-mirror Sites:
BitBake first uses pre-mirrors to try and find source files.
These locations are defined using the
PREMIRRORS
variable.
Source URI:
If pre-mirrors fail, BitBake uses the original URL (e.g from
SRC_URI
).
Mirror Sites:
If fetch failures occur, BitBake next uses mirror locations as
defined by the
MIRRORS
variable.
For each URL passed to the fetcher, the fetcher
calls the submodule that handles that particular URL type.
This behavior can be the source of some confusion when you
are providing URLs for the SRC_URI
variable.
Consider the following two URLs:
http://git.yoctoproject.org/git/poky;protocol=git git://git.yoctoproject.org/git/poky;protocol=http
In the former case, the URL is passed to the
wget
fetcher, which does not
understand "git".
Therefore, the latter case is the correct form since the
Git fetcher does know how to use HTTP as a transport.
Here are some examples that show commonly used mirror definitions:
PREMIRRORS ?= "\ bzr://.*/.* http://somemirror.org/sources/ \n \ cvs://.*/.* http://somemirror.org/sources/ \n \ git://.*/.* http://somemirror.org/sources/ \n \ hg://.*/.* http://somemirror.org/sources/ \n \ osc://.*/.* http://somemirror.org/sources/ \n \ p4://.*/.* http://somemirror.org/sources/ \n \ svn://.*/.* http://somemirror.org/sources/ \n" MIRRORS =+ "\ ftp://.*/.* http://somemirror.org/sources/ \n \ http://.*/.* http://somemirror.org/sources/ \n \ https://.*/.* http://somemirror.org/sources/ \n"
It is useful to note that BitBake supports
cross-URLs.
It is possible to mirror a Git repository on an HTTP
server as a tarball.
This is what the git://
mapping in
the previous example does.
Since network accesses are slow, BitBake maintains a
cache of files downloaded from the network.
Any source files that are not local (i.e.
downloaded from the Internet) are placed into the download
directory, which is specified by the
DL_DIR
variable.
File integrity is of key importance for reproducing builds.
For non-local archive downloads, the fetcher code can verify
SHA-256 and MD5 checksums to ensure the archives have been
downloaded correctly.
You can specify these checksums by using the
SRC_URI
variable with the appropriate
varflags as follows:
SRC_URI[md5sum] = "value
" SRC_URI[sha256sum] = "value
"
You can also specify the checksums as parameters on the
SRC_URI
as shown below:
SRC_URI = "http://example.com/foobar.tar.bz2;md5sum=4a8e0f237e961fd7785d19d07fdb994d"
If multiple URIs exist, you can specify the checksums either directly as in the previous example, or you can name the URLs. The following syntax shows how you name the URIs:
SRC_URI = "http://example.com/foobar.tar.bz2;name=foo" SRC_URI[foo.md5sum] = 4a8e0f237e961fd7785d19d07fdb994d
After a file has been downloaded and has had its checksum checked,
a ".done" stamp is placed in DL_DIR
.
BitBake uses this stamp during subsequent builds to avoid
downloading or comparing a checksum for the file again.
If
BB_STRICT_CHECKSUM
is set, any download without a checksum triggers an
error message.
The
BB_NO_NETWORK
variable can be used to make any attempted network access a fatal
error, which is useful for checking that mirrors are complete
as well as other things.
The unpack process usually immediately follows the download.
For all URLs except Git URLs, BitBake uses the common
unpack
method.
A number of parameters exist that you can specify within the URL to govern the behavior of the unpack stage:
unpack: Controls whether the URL components are unpacked. If set to "1", which is the default, the components are unpacked. If set to "0", the unpack stage leaves the file alone. This parameter is useful when you want an archive to be copied in and not be unpacked.
dos:
Applies to .zip
and
.jar
files and specifies whether to
use DOS line ending conversion on text files.
basepath: Instructs the unpack stage to strip the specified directories from the source path when unpacking.
subdir: Unpacks the specific URL to the specified subdirectory within the root directory.
The unpack call automatically decompresses and extracts files with ".Z", ".z", ".gz", ".xz", ".zip", ".jar", ".ipk", ".rpm". ".srpm", ".deb" and ".bz2" extensions as well as various combinations of tarball extensions.
As mentioned, the Git fetcher has its own unpack method that is optimized to work with Git trees. Basically, this method works by cloning the tree into the final directory. The process is completed using references so that there is only one central copy of the Git metadata needed.
As mentioned earlier, the URL prefix determines which fetcher submodule BitBake uses. Each submodule can support different URL parameters, which are described in the following sections.
file://
)¶
This submodule handles URLs that begin with
file://
.
The filename you specify within the URL can be
either an absolute or relative path to a file.
If the filename is relative, the contents of the
FILESPATH
variable is used in the same way
PATH
is used to find executables.
If the file cannot be found, it is assumed that it is available in
DL_DIR
by the time the download()
method is called.
If you specify a directory, the entire directory is unpacked.
Here are a couple of example URLs, the first relative and the second absolute:
SRC_URI = "file://relativefile.patch" SRC_URI = "file:///Users/ich/very_important_software"
http://
, ftp://
, https://
)¶This fetcher obtains files from web and FTP servers. Internally, the fetcher uses the wget utility.
The executable and parameters used are specified by the
FETCHCMD_wget
variable, which defaults
to sensible values.
The fetcher supports a parameter "downloadfilename" that
allows the name of the downloaded file to be specified.
Specifying the name of the downloaded file is useful
for avoiding collisions in
DL_DIR
when dealing with multiple files that have the same name.
Some example URLs are as follows:
SRC_URI = "http://oe.handhelds.org/not_there.aac" SRC_URI = "ftp://oe.handhelds.org/not_there_as_well.aac" SRC_URI = "ftp://you@oe.handhelds.org/home/you/secret.plan"
SRC_URI = "http://abc123.org/git/?p=gcc/gcc.git;a=snapshot;h=a5dd47"Such URLs should should be modified by replacing semi-colons with '&' characters:
SRC_URI = "http://abc123.org/git/?p=gcc/gcc.git&a=snapshot&h=a5dd47"In most cases this should work. Treating semi-colons and '&' in queries identically is recommended by the World Wide Web Consortium (W3C). Note that due to the nature of the URL, you may have to specify the name of the downloaded file as well:
SRC_URI = "http://abc123.org/git/?p=gcc/gcc.git&a=snapshot&h=a5dd47;downloadfilename=myfile.bz2"
(cvs://
)¶This submodule handles checking out files from the CVS version control system. You can configure it using a number of different variables:
FETCHCMD_cvs
:
The name of the executable to use when running
the cvs
command.
This name is usually "cvs".
SRCDATE
:
The date to use when fetching the CVS source code.
A special value of "now" causes the checkout to
be updated on every build.
CVSDIR
:
Specifies where a temporary checkout is saved.
The location is often DL_DIR/cvs
.
CVS_PROXY_HOST
:
The name to use as a "proxy=" parameter to the
cvs
command.
CVS_PROXY_PORT
:
The port number to use as a "proxyport=" parameter to
the cvs
command.
As well as the standard username and password URL syntax, you can also configure the fetcher with various URL parameters:
The supported parameters are as follows:
"method":
The protocol over which to communicate with the CVS
server.
By default, this protocol is "pserver".
If "method" is set to "ext", BitBake examines the
"rsh" parameter and sets CVS_RSH
.
You can use "dir" for local directories.
"module": Specifies the module to check out. You must supply this parameter.
"tag": Describes which CVS TAG should be used for the checkout. By default, the TAG is empty.
"date":
Specifies a date.
If no "date" is specified, the
SRCDATE
of the configuration is used to checkout a specific date.
The special value of "now" causes the checkout to be
updated on every build.
"localdir":
Used to rename the module.
Effectively, you are renaming the output directory
to which the module is unpacked.
You are forcing the module into a special
directory relative to
CVSDIR
.
"rsh" Used in conjunction with the "method" parameter.
"scmdata": Causes the CVS metadata to be maintained in the tarball the fetcher creates when set to "keep". The tarball is expanded into the work directory. By default, the CVS metadata is removed.
"fullpath": Controls whether the resulting checkout is at the module level, which is the default, or is at deeper paths.
"norecurse": Causes the fetcher to only checkout the specified directory with no recurse into any subdirectories.
"port": The port to which the CVS server connects.
Some example URLs are as follows:
SRC_URI = "cvs://CVSROOT;module=mymodule;tag=some-version;method=ext" SRC_URI = "cvs://CVSROOT;module=mymodule;date=20060126;localdir=usethat"
svn://
)¶
This fetcher submodule fetches code from the
Subversion source control system.
The executable used is specified by
FETCHCMD_svn
, which defaults
to "svn".
The fetcher's temporary working directory is set by
SVNDIR
,
which is usually DL_DIR/svn
.
The supported parameters are as follows:
"module": The name of the svn module to checkout. You must provide this parameter. You can think of this parameter as the top-level directory of the repository data you want.
"path_spec": A specific directory in which to checkout the specified svn module.
"protocol": The protocol to use, which defaults to "svn". If "protocol" is set to "svn+ssh", the "ssh" parameter is also used.
"rev": The revision of the source code to checkout.
"scmdata": Causes the “.svn” directories to be available during compile-time when set to "keep". By default, these directories are removed.
"ssh": An optional parameter used when "protocol" is set to "svn+ssh". You can use this parameter to specify the ssh program used by svn.
"transportuser": When required, sets the username for the transport. By default, this parameter is empty. The transport username is different than the username used in the main URL, which is passed to the subversion command.
Following are three examples using svn:
SRC_URI = "svn://myrepos/proj1;module=vip;protocol=http;rev=667" SRC_URI = "svn://myrepos/proj1;module=opie;protocol=svn+ssh" SRC_URI = "svn://myrepos/proj1;module=trunk;protocol=http;path_spec=${MY_DIR}/proj1"
git://
)¶
This fetcher submodule fetches code from the Git
source control system.
The fetcher works by creating a bare clone of the
remote into
GITDIR
,
which is usually DL_DIR/git2
.
This bare clone is then cloned into the work directory during the
unpack stage when a specific tree is checked out.
This is done using alternates and by reference to
minimize the amount of duplicate data on the disk and
make the unpack process fast.
The executable used can be set with
FETCHCMD_git
.
This fetcher supports the following parameters:
"protocol": The protocol used to fetch the files. The default is "git" when a hostname is set. If a hostname is not set, the Git protocol is "file". You can also use "http", "https", "ssh" and "rsync".
"nocheckout": Tells the fetcher to not checkout source code when unpacking when set to "1". Set this option for the URL where there is a custom routine to checkout code. The default is "0".
"rebaseable": Indicates that the upstream Git repository can be rebased. You should set this parameter to "1" if revisions can become detached from branches. In this case, the source mirror tarball is done per revision, which has a loss of efficiency. Rebasing the upstream Git repository could cause the current revision to disappear from the upstream repository. This option reminds the fetcher to preserve the local cache carefully for future use. The default value for this parameter is "0".
"nobranch": Tells the fetcher to not check the SHA validation for the branch when set to "1". The default is "0". Set this option for the recipe that refers to the commit that is valid for a tag instead of the branch.
"bareclone": Tells the fetcher to clone a bare clone into the destination directory without checking out a working tree. Only the raw Git metadata is provided. This parameter implies the "nocheckout" parameter as well.
"branch": The branch(es) of the Git tree to clone. If unset, this is assumed to be "master". The number of branch parameters much match the number of name parameters.
"rev": The revision to use for the checkout. The default is "master".
"tag": Specifies a tag to use for the checkout. To correctly resolve tags, BitBake must access the network. For that reason, tags are often not used. As far as Git is concerned, the "tag" parameter behaves effectively the same as the "rev" parameter.
"subpath": Limits the checkout to a specific subpath of the tree. By default, the whole tree is checked out.
"destsuffix":
The name of the path in which to place the checkout.
By default, the path is git/
.
"usehead":
Enables local git://
URLs to use the
current branch HEAD as the revision for use with
AUTOREV
.
The "usehead" parameter implies no branch and only works
when the transfer protocol is
file://
.
Here are some example URLs:
SRC_URI = "git://git.oe.handhelds.org/git/vip.git;tag=version-1" SRC_URI = "git://git.oe.handhelds.org/git/vip.git;protocol=http"
gitsm://
)¶
This fetcher submodule inherits from the
Git fetcher and extends
that fetcher's behavior by fetching a repository's submodules.
SRC_URI
is passed to the Git fetcher as described in the
"Git Fetcher (git://
)"
section.
You must clean a recipe when switching between
'git://
' and
'gitsm://
' URLs.
The Git Submodules fetcher is not a complete fetcher implementation. The fetcher has known issues where it does not use the normal source mirroring infrastructure properly. Further, the submodule sources it fetches are not visible to the licensing and source archiving infrastructures.
ccrc://
)¶This fetcher submodule fetches code from a ClearCase repository.
To use this fetcher, make sure your recipe has proper
SRC_URI
,
SRCREV
, and
PV
settings.
Here is an example:
SRC_URI = "ccrc://cc.example.org/ccrc;vob=/example_vob;module=/example_module" SRCREV = "EXAMPLE_CLEARCASE_TAG" PV = "${@d.getVar("SRCREV", False).replace("/", "+")}"
The fetcher uses the rcleartool
or
cleartool
remote client, depending on
which one is available.
Following are options for the SRC_URI
statement:
vob
:
The name, which must include the
prepending "/" character, of the ClearCase VOB.
This option is required.
module
:
The module, which must include the
prepending "/" character, in the selected VOB.
module
and vob
options are combined to create the load
rule in
the view config spec.
As an example, consider the vob
and
module
values from the
SRC_URI
statement at the start of this section.
Combining those values results in the following:
load /example_vob/example_module
proto
:
The protocol, which can be either http
or
https
.
By default, the fetcher creates a configuration specification.
If you want this specification written to an area other than the default,
use the CCASE_CUSTOM_CONFIG_SPEC
variable
in your recipe to define where the specification is written.
SRCREV
loses its functionality if you
specify this variable.
However, SRCREV
is still used to label the
archive after a fetch even though it does not define what is
fetched.
Here are a couple of other behaviors worth mentioning:
When using cleartool
, the login of
cleartool
is handled by the system.
The login require no special steps.
In order to use rcleartool
with authenticated
users, an "rcleartool login" is necessary before using the fetcher.
p4://
)¶
This fetcher submodule fetches code from the
Perforce
source control system.
The executable used is specified by
FETCHCMD_p4
, which defaults
to "p4".
The fetcher's temporary working directory is set by
P4DIR
,
which defaults to "DL_DIR/p4".
To use this fetcher, make sure your recipe has proper
SRC_URI
,
SRCREV
, and
PV
values.
The p4 executable is able to use the config file defined by your
system's P4CONFIG
environment variable in
order to define the Perforce server URL and port, username, and
password if you do not wish to keep those values in a recipe
itself.
If you choose not to use P4CONFIG
,
or to explicitly set variables that P4CONFIG
can contain, you can specify the P4PORT
value,
which is the server's URL and port number, and you can
specify a username and password directly in your recipe within
SRC_URI
.
Here is an example that relies on P4CONFIG
to specify the server URL and port, username, and password, and
fetches the Head Revision:
SRC_URI = "p4://example-depot/main/source/..." SRCREV = "${AUTOREV}" PV = "p4-${SRCPV}" S = "${WORKDIR}/p4"
Here is an example that specifies the server URL and port, username, and password, and fetches a Revision based on a Label:
P4PORT = "tcp:p4server.example.net:1666" SRC_URI = "p4://user:passwd@example-depot/main/source/..." SRCREV = "release-1.0" PV = "p4-${SRCPV}" S = "${WORKDIR}/p4"
S
to "${WORKDIR}/p4"
in your recipe.
repo://
)¶
This fetcher submodule fetches code from
google-repo
source control system.
The fetcher works by initiating and syncing sources of the
repository into
REPODIR
,
which is usually
DL_DIR
/repo
.
This fetcher supports the following parameters:
"protocol": Protocol to fetch the repository manifest (default: git).
"branch": Branch or tag of repository to get (default: master).
"manifest":
Name of the manifest file (default: default.xml
).
Here are some example URLs:
SRC_URI = "repo://REPOROOT;protocol=git;branch=some_branch;manifest=my_manifest.xml" SRC_URI = "repo://REPOROOT;protocol=file;branch=some_branch;manifest=my_manifest.xml"
Fetch submodules also exist for the following:
Bazaar (bzr://
)
Mercurial (hg://
)
npm (npm://
)
OSC (osc://
)
Secure FTP (sftp://
)
Secure Shell (ssh://
)
Trees using Git Annex (gitannex://
)
No documentation currently exists for these lesser used fetcher submodules. However, you might find the code helpful and readable.
We need to document AUTOREV
and
SRCREV_FORMAT
here.
This chapter lists common variables used by BitBake and gives an overview of their function and contents.
The variables listed in this glossary are specific to BitBake. Consequently, the descriptions are limited to that context.
Also, variables exist in other systems that use BitBake (e.g. The Yocto Project and OpenEmbedded) that have names identical to those found in this glossary. For such cases, the variables in those systems extend the functionality of the variable as it is described here in this glossary.
Finally, there are variables mentioned in this glossary that do not appear in the BitBake glossary. These other variables are variables used in systems that use BitBake.
A B C D E F G H I L M O P R S T
Lists recipe names
(PN
values) BitBake does not attempt to build.
Instead, BitBake assumes these recipes have already been
built.
In OpenEmbedded-Core, ASSUME_PROVIDED
mostly specifies native tools that should not be built.
An example is git-native
, which
when specified allows for the Git binary from the host to
be used rather than building
git-native
.
The directory in which BitBake executes functions during a recipe's build process.
Specifies a space-delimited list of hosts that the fetcher is allowed to use to obtain the required source code. Following are considerations surrounding this variable:
This host list is only used if
BB_NO_NETWORK
is either not set or set to "0".
Limited support for the "*
"
wildcard character for matching against the
beginning of host names exists.
For example, the following setting matches
git.gnu.org
,
ftp.gnu.org
, and
foo.git.gnu.org
.
BB_ALLOWED_NETWORKS = "*.gnu.org"
The use of the "*
"
character only works at the beginning of
a host name and it must be isolated from
the remainder of the host name.
You cannot use the wildcard character in any
other location of the name or combined with
the front part of the name.
For example,
*.foo.bar
is supported,
while *aa.foo.bar
is not.
Mirrors not in the host list are skipped and logged in debug.
Attempts to access networks not in the host list cause a failure.
Using BB_ALLOWED_NETWORKS
in
conjunction with
PREMIRRORS
is very useful.
Adding the host you want to use to
PREMIRRORS
results in the source code
being fetched from an allowed location and avoids raising
an error when a host that is not allowed is in a
SRC_URI
statement.
This is because the fetcher does not attempt to use the
host listed in SRC_URI
after a
successful fetch from the
PREMIRRORS
occurs.
Specifies the path to a log file into which BitBake's user interface writes output during the build.
Contains the name of the currently running task.
The name does not include the
do_
prefix.
Defines how BitBake handles situations where an append
file (.bbappend
) has no
corresponding recipe file (.bb
).
This condition often occurs when layers get out of sync
(e.g. oe-core
bumps a
recipe version and the old recipe no longer exists and the
other layer has not been updated to the new version
of the recipe yet).
The default fatal behavior is safest because it is the sane reaction given something is out of sync. It is important to realize when your changes are no longer being applied.
The default task to use when none is specified (e.g.
with the -c
command line option).
The task name specified should not include the
do_
prefix.
Monitors disk space and available inodes during the build and allows you to control the build based on these parameters.
Disk space monitoring is disabled by default. When setting this variable, use the following form:
BB_DISKMON_DIRS = "<action>,<dir>,<threshold> [...]" where: <action> is: ABORT: Immediately abort the build when a threshold is broken. STOPTASKS: Stop the build after the currently executing tasks have finished when a threshold is broken. WARN: Issue a warning but continue the build when a threshold is broken. Subsequent warnings are issued as defined by the BB_DISKMON_WARNINTERVAL variable, which must be defined. <dir> is: Any directory you choose. You can specify one or more directories to monitor by separating the groupings with a space. If two directories are on the same device, only the first directory is monitored. <threshold> is: Either the minimum available disk space, the minimum number of free inodes, or both. You must specify at least one. To omit one or the other, simply omit the value. Specify the threshold using G, M, K for Gbytes, Mbytes, and Kbytes, respectively. If you do not specify G, M, or K, Kbytes is assumed by default. Do not use GB, MB, or KB.
Here are some examples:
BB_DISKMON_DIRS = "ABORT,${TMPDIR},1G,100K WARN,${SSTATE_DIR},1G,100K" BB_DISKMON_DIRS = "STOPTASKS,${TMPDIR},1G" BB_DISKMON_DIRS = "ABORT,${TMPDIR},,100K"
The first example works only if you also set
the BB_DISKMON_WARNINTERVAL
variable.
This example causes the build system to immediately
abort when either the disk space in ${TMPDIR}
drops
below 1 Gbyte or the available free inodes drops below
100 Kbytes.
Because two directories are provided with the variable, the
build system also issues a
warning when the disk space in the
${SSTATE_DIR}
directory drops
below 1 Gbyte or the number of free inodes drops
below 100 Kbytes.
Subsequent warnings are issued during intervals as
defined by the BB_DISKMON_WARNINTERVAL
variable.
The second example stops the build after all currently
executing tasks complete when the minimum disk space
in the ${TMPDIR}
directory drops below 1 Gbyte.
No disk monitoring occurs for the free inodes in this case.
The final example immediately aborts the build when the
number of free inodes in the ${TMPDIR}
directory
drops below 100 Kbytes.
No disk space monitoring for the directory itself occurs
in this case.
Defines the disk space and free inode warning intervals.
If you are going to use the
BB_DISKMON_WARNINTERVAL
variable, you must
also use the
BB_DISKMON_DIRS
variable
and define its action as "WARN".
During the build, subsequent warnings are issued each time
disk space or number of free inodes further reduces by
the respective interval.
If you do not provide a BB_DISKMON_WARNINTERVAL
variable and you do use BB_DISKMON_DIRS
with
the "WARN" action, the disk monitoring interval defaults to
the following:
BB_DISKMON_WARNINTERVAL = "50M,5K"
When specifying the variable in your configuration file, use the following form:
BB_DISKMON_WARNINTERVAL = "<disk_space_interval>,<disk_inode_interval>" where: <disk_space_interval> is: An interval of memory expressed in either G, M, or K for Gbytes, Mbytes, or Kbytes, respectively. You cannot use GB, MB, or KB. <disk_inode_interval> is: An interval of free inodes expressed in either G, M, or K for Gbytes, Mbytes, or Kbytes, respectively. You cannot use GB, MB, or KB.
Here is an example:
BB_DISKMON_DIRS = "WARN,${SSTATE_DIR},1G,100K" BB_DISKMON_WARNINTERVAL = "50M,5K"
These variables cause BitBake to
issue subsequent warnings each time the available
disk space further reduces by 50 Mbytes or the number
of free inodes further reduces by 5 Kbytes in the
${SSTATE_DIR}
directory.
Subsequent warnings based on the interval occur each time
a respective interval is reached beyond the initial warning
(i.e. 1 Gbytes and 100 Kbytes).
Specifies the internal whitelist of variables to allow
through from the external environment into BitBake's
datastore.
If the value of this variable is not specified
(which is the default), the following list is used:
BBPATH
,
BB_PRESERVE_ENV
,
BB_ENV_WHITELIST
,
and
BB_ENV_EXTRAWHITE
.
Specifies an additional set of variables to allow through
(whitelist) from the external environment into BitBake's
datastore.
This list of variables are on top of the internal list
set in
BB_ENV_WHITELIST
.
When set to "1", causes BitBake's fetcher module to only
search
PREMIRRORS
for files.
BitBake will not search the main
SRC_URI
or
MIRRORS
.
Contains the filename of the recipe that owns the currently
running task.
For example, if the do_fetch
task that
resides in the my-recipe.bb
is
executing, the BB_FILENAME
variable
contains "/foo/path/my-recipe.bb".
Causes tarballs of the Git repositories, including the
Git metadata, to be placed in the
DL_DIR
directory.
Anyone wishing to create a source mirror would want to
enable this variable.
For performance reasons, creating and placing tarballs of the Git repositories is not the default action by BitBake.
BB_GENERATE_MIRROR_TARBALLS = "1"
Lists variables that are excluded from base configuration checksum, which is used to determine if the cache can be reused.
One of the ways BitBake determines whether to re-parse the
main metadata is through checksums of the variables in the
datastore of the base configuration data.
There are variables that you typically want to exclude when
checking whether or not to re-parse and thus rebuild the
cache.
As an example, you would usually exclude
TIME
and DATE
because these variables are always changing.
If you did not exclude them, BitBake would never reuse the
cache.
Lists variables that are excluded from checksum and dependency data. Variables that are excluded can therefore change without affecting the checksum mechanism. A common example would be the variable for the path of the build. BitBake's output should not (and usually does not) depend on the directory in which it was built.
Specifies the name of the function to call during the "setscene" part of the task's execution in order to validate the list of task hashes. The function returns the list of setscene tasks that should be executed.
At this point in the execution of the code, the objective is to quickly verify if a given setscene function is likely to work or not. It's easier to check the list of setscene functions in one pass than to call many individual tasks. The returned list need not be completely accurate. A given setscene task can still later fail. However, the more accurate the data returned, the more efficient the build will be.
Used in combination with the
ConfigParsed
event to trigger
re-parsing the base metadata (i.e. all the
recipes).
The ConfigParsed
event can set the
variable to trigger the re-parse.
You must be careful to avoid recursive loops with this
functionality.
Specifies the name of a config file that contains the user logging configuration. See Logging for additional information
Specifies the name of the log files saved into
${
T
}
.
By default, the BB_LOGFMT
variable
is undefined and the log file names get created using the
following form:
log.{task}.{pid}
If you want to force log files to take a specific name, you can set this variable in a configuration file.
Allows BitBake to run at a specific priority
(i.e. nice level).
System permissions usually mean that BitBake can reduce its
priority but not raise it again.
See
BB_TASK_NICE_LEVEL
for additional information.
Disables network access in the BitBake fetcher modules. With this access disabled, any command that attempts to access the network becomes an error.
Disabling network access is useful for testing source mirrors, running builds when not connected to the Internet, and when operating in certain kinds of firewall environments.
The maximum number of tasks BitBake should run in parallel at any one time. If your host development system supports multiple cores, a good rule of thumb is to set this variable to twice the number of cores.
Sets the number of threads BitBake uses when parsing. By default, the number of threads is equal to the number of cores on the system.
Contains a copy of the original external environment in which BitBake was run. The copy is taken before any whitelisted variable values are filtered into BitBake's datastore.
Disables whitelisting and instead allows all variables through from the external environment into BitBake's datastore.
Specifies the name of the executable script files
(i.e. run files) saved into
${
T
}
.
By default, the BB_RUNFMT
variable
is undefined and the run file names get created using the
following form:
run.{task}.{pid}
If you want to force run files to take a specific name, you can set this variable in a configuration file.
Contains the name of the currently executing task.
The value includes the "do_" prefix.
For example, if the currently executing task is
do_config
, the value is
"do_config".
Selects the name of the scheduler to use for the scheduling of BitBake tasks. Three options exist:
basic - The basic framework from which everything derives. Using this option causes tasks to be ordered numerically as they are parsed.
speed - Executes tasks first that have more tasks depending on them. The "speed" option is the default.
completion - Causes the scheduler to try to complete a given recipe once its build has started.
Defines custom schedulers to import.
Custom schedulers need to be derived from the
RunQueueScheduler
class.
For information how to select a scheduler, see the
BB_SCHEDULER
variable.
Specifies a function BitBake calls that determines whether BitBake requires a setscene dependency to be met.
When running a setscene task, BitBake needs to know which dependencies of that setscene task also need to be run. Whether dependencies also need to be run is highly dependent on the metadata. The function specified by this variable returns a "True" or "False" depending on whether the dependency needs to be met.
Specifies a function to call that verifies the list of planned task execution before the main task execution happens. The function is called once BitBake has a list of setscene tasks that have run and either succeeded or failed.
The function allows for a task list check to see if they make sense. Even if BitBake was planning to skip a task, the returned value of the function can force BitBake to run the task, which is necessary under certain metadata defined circumstances.
Lists variable flags (varflags) that can be safely excluded from checksum and dependency data for keys in the datastore. When generating checksum or dependency data for keys in the datastore, the flags set against that key are normally included in the checksum.
For more information on varflags, see the "Variable Flags" section.
Defines the name of the signature handler BitBake uses. The signature handler defines the way stamp files are created and handled, if and how the signature is incorporated into the stamps, and how the signature itself is generated.
A new signature handler can be added by injecting a class
derived from the
SignatureGenerator
class into the
global namespace.
Defines the behavior of the fetcher when it interacts with
source control systems and dynamic source revisions.
The BB_SRCREV_POLICY
variable is
useful when working without a network.
The variable can be set using one of two policies:
cache - Retains the value the system obtained previously rather than querying the source control system each time.
clear - Queries the source controls system every time. With this policy, there is no cache. The "clear" policy is the default.
Defines the mode used for how timestamps of stamp files are compared. You can set the variable to one of the following modes:
perfile - Timestamp comparisons are only made between timestamps of a specific recipe. This is the default mode.
full - Timestamp comparisons are made for all dependencies.
whitelist -
Identical to "full" mode except timestamp
comparisons are made for recipes listed in the
BB_STAMP_WHITELIST
variable.
Lists files whose stamp file timestamps are compared when
the stamp policy mode is set to "whitelist".
For information on stamp policies, see the
BB_STAMP_POLICY
variable.
Sets a more strict checksum mechanism for non-local URLs. Setting this variable to a value causes BitBake to report an error if it encounters a non-local URL that does not have at least one checksum specified.
Allows adjustment of a task's Input/Output priority.
During Autobuilder testing, random failures can occur
for tasks due to I/O starvation.
These failures occur during various QEMU runtime timeouts.
You can use the BB_TASK_IONICE_LEVEL
variable to adjust the I/O priority of these tasks.
BB_TASK_NICE_LEVEL
variable except with a task's I/O priorities.
Set the variable as follows:
BB_TASK_IONICE_LEVEL = "class
.prio
"
For class
, the default value is
"2", which is a best effort.
You can use "1" for realtime and "3" for idle.
If you want to use realtime, you must have superuser
privileges.
For prio
, you can use any
value from "0", which is the highest priority, to "7",
which is the lowest.
The default value is "4".
You do not need any special privileges to use this range
of priority values.
device
is the device
(e.g. sda, sdb, and so forth):
$ sudo sh -c “echo cfq > /sys/block/device
/queu/scheduler
Allows specific tasks to change their priority (i.e. nice level).
You can use this variable in combination with task overrides to raise or lower priorities of specific tasks. For example, on the Yocto Project autobuilder, QEMU emulation in images is given a higher priority as compared to build tasks to ensure that images do not suffer timeouts on loaded systems.
Within an executing task, this variable holds the hash of the task as returned by the currently enabled signature generator.
Controls how verbose BitBake is during builds. If set, shell scripts echo commands and shell script output appears on standard out (stdout).
Specifies if the current context is executing a task. BitBake sets this variable to "1" when a task is being executed. The value is not set when the task is in server context during parsing or event handling.
Allows you to extend a recipe so that it builds variants
of the software.
Some examples of these variants for recipes from the
OpenEmbedded-Core metadata are "natives" such as
quilt-native
, which is a copy of
Quilt built to run on the build system; "crosses" such
as gcc-cross
, which is a compiler
built to run on the build machine but produces binaries
that run on the target MACHINE
;
"nativesdk", which targets the SDK machine instead of
MACHINE
; and "mulitlibs" in the form
"multilib:
multilib_name
".
To build a different variant of the recipe with a minimal amount of code, it usually is as simple as adding the variable to your recipe. Here are two examples. The "native" variants are from the OpenEmbedded-Core metadata:
BBCLASSEXTEND =+ "native nativesdk"
BBCLASSEXTEND =+ "multilib:multilib_name
"
Internally, the BBCLASSEXTEND
mechanism generates recipe variants by rewriting
variable values and applying overrides such as
_class-native
.
For example, to generate a native version of a recipe,
a
DEPENDS
on "foo" is rewritten to a DEPENDS
on "foo-native".
Even when using BBCLASSEXTEND
, the
recipe is only parsed once.
Parsing once adds some limitations.
For example, it is not possible to
include a different file depending on the variant,
since include
statements are
processed when the recipe is parsed.
Sets the BitBake debug output level to a specific value
as incremented by the -D
command line
option.
Lists the names of configured layers.
These names are used to find the other BBFILE_*
variables.
Typically, each layer appends its name to this variable in its
conf/layer.conf
file.
Variable that expands to match files from
BBFILES
in a particular layer.
This variable is used in the conf/layer.conf
file and must
be suffixed with the name of the specific layer (e.g.
BBFILE_PATTERN_emenlow
).
Assigns the priority for recipe files in each layer.
This variable is useful in situations where the same recipe appears in
more than one layer.
Setting this variable allows you to prioritize a
layer against other layers that contain the same recipe - effectively
letting you control the precedence for the multiple layers.
The precedence established through this variable stands regardless of a
recipe's version
(PV
variable).
For example, a layer that has a recipe with a higher PV
value but for
which the BBFILE_PRIORITY
is set to have a lower precedence still has a
lower precedence.
A larger value for the BBFILE_PRIORITY
variable results in a higher
precedence.
For example, the value 6 has a higher precedence than the value 5.
If not specified, the BBFILE_PRIORITY
variable is set based on layer
dependencies (see the
LAYERDEPENDS
variable for
more information.
The default priority, if unspecified
for a layer with no dependencies, is the lowest defined priority + 1
(or 1 if no priorities are defined).
bitbake-layers show-layers
to list
all configured layers along with their priorities.
A space-separated list of recipe files BitBake uses to build software.
When specifying recipe files, you can pattern match using
Python's
glob
syntax.
For details on the syntax, see the documentation by
following the previous link.
Contains a space-separated list of all of all files that BitBake's parser included during parsing of the current file.
If set to a value, enables printing the task log when reporting a failed task.
If
BBINCLUDELOGS
is set, specifies the maximum number of lines from the
task log file to print when reporting a failed task.
If you do not set BBINCLUDELOGS_LINES
,
the entire log is printed.
Lists the layers to enable during the build.
This variable is defined in the bblayers.conf
configuration
file in the build directory.
Here is an example:
BBLAYERS = " \ /home/scottrif/poky/meta \ /home/scottrif/poky/meta-yocto \ /home/scottrif/poky/meta-yocto-bsp \ /home/scottrif/poky/meta-mykernel \ "
This example enables four layers, one of which is a custom, user-defined layer
named meta-mykernel
.
Sets the base location where layers are stored.
This setting is used in conjunction with
bitbake-layers layerindex-fetch
and
tells bitbake-layers
where to place
the fetched layers.
Prevents BitBake from processing recipes and recipe append files.
You can use the BBMASK
variable
to "hide" these .bb
and
.bbappend
files.
BitBake ignores any recipe or recipe append files that
match any of the expressions.
It is as if BitBake does not see them at all.
Consequently, matching files are not parsed or otherwise
used by BitBake.
The values you provide are passed to Python's regular expression compiler. Consequently, the syntax follows Python's Regular Expression (re) syntax. The expressions are compared against the full paths to the files. For complete syntax information, see Python's documentation at http://docs.python.org/3/library/re.html#re.
The following example uses a complete regular expression
to tell BitBake to ignore all recipe and recipe append
files in the meta-ti/recipes-misc/
directory:
BBMASK = "meta-ti/recipes-misc/"
If you want to mask out multiple directories or recipes, you can specify multiple regular expression fragments. This next example masks out multiple directories and individual recipes:
BBMASK += "/meta-ti/recipes-misc/ meta-ti/recipes-ti/packagegroup/" BBMASK += "/meta-oe/recipes-support/" BBMASK += "/meta-foo/.*/openldap" BBMASK += "opencv.*\.bbappend" BBMASK += "lzma"
Enables BitBake to perform multiple configuration builds
and lists each separate configuration (multiconfig).
You can use this variable to cause BitBake to build
multiple targets where each target has a separate
configuration.
Define BBMULTICONFIG
in your
conf/local.conf
configuration file.
As an example, the following line specifies three multiconfigs, each having a separate configuration file:
BBMULTIFONFIG = "configA configB configC"
Each configuration file you use must reside in the
build directory within a directory named
conf/multiconfig
(e.g.
build_directory
/conf/multiconfig/configA.conf
).
For information on how to use
BBMULTICONFIG
in an environment that
supports building targets with multiple configurations,
see the
"Executing a Multiple Configuration Build"
section.
Used by BitBake to locate class
(.bbclass
) and configuration
(.conf
) files.
This variable is analogous to the
PATH
variable.
If you run BitBake from a directory outside of the
build directory,
you must be sure to set
BBPATH
to point to the
build directory.
Set the variable as you would any environment variable
and then run BitBake:
$ BBPATH="build_directory
" $ export BBPATH $ bitbaketarget
Points to the server that runs memory-resident BitBake. The variable is only used when you employ memory-resident BitBake.
Allows you to use a configuration file to add to the list of command-line target recipes you want to build.
Allows a single recipe to build multiple versions of a
project from a single recipe file.
You also able to specify conditional metadata
using the
OVERRIDES
mechanism for a single version or for an optionally named
range of versions.
For more information on BBVERSIONS
,
see the
"Variants - Class Extension Mechanism"
section.
Used to specify the UI module to use when running BitBake.
Using this variable is equivalent to using the
-u
command-line option.
A name assigned to the build. The name defaults to a datetime stamp of when the build was started but can be defined by the metadata.
The directory in which files checked out of a Bazaar system are stored.
Specifies a weak bias for recipe selection priority.
The most common usage of this is variable is to set
it to "-1" within a recipe for a development version of a
piece of software.
Using the variable in this way causes the stable version
of the recipe to build by default in the absence of
PREFERRED_VERSION
being used to build the development version.
DEFAULT_PREFERENCE
is weak and is overridden by
BBFILE_PRIORITY
if that variable is different between two layers
that contain different versions of the same recipe.
Lists a recipe's build-time dependencies (i.e. other recipe files).
Consider this simple example for two recipes named "a" and
"b" that produce similarly named packages.
In this example, the DEPENDS
statement appears in the "a" recipe:
DEPENDS = "b"
Here, the dependency is such that the
do_configure
task for recipe "a"
depends on the do_populate_sysroot
task of recipe "b".
This means anything that recipe "b" puts into sysroot
is available when recipe "a" is configuring itself.
For information on runtime dependencies, see the
RDEPENDS
variable.
A long description for the recipe.
The central download directory used by the build process to
store downloads.
By default, DL_DIR
gets files
suitable for mirroring for everything except Git
repositories.
If you want tarballs of Git repositories, use the
BB_GENERATE_MIRROR_TARBALLS
variable.
Directs BitBake to exclude a recipe from world builds (i.e.
bitbake world
).
During world builds, BitBake locates, parses and builds all
recipes found in every layer exposed in the
bblayers.conf
configuration file.
To exclude a recipe from a world build using this variable, set the variable to "1" in the recipe.
EXCLUDE_FROM_WORLD
may still be built during a world build in order to satisfy
dependencies of other recipes.
Adding a recipe to EXCLUDE_FROM_WORLD
only ensures that the recipe is not explicitly added
to the list of build targets in a world build.
Contains the command to use when running a shell script
in a fakeroot environment.
The FAKEROOT
variable is obsolete
and has been replaced by the other
FAKEROOT*
variables.
See these entries in the glossary for more information.
Lists environment variables to set when executing
the command defined by
FAKEROOTCMD
that starts the bitbake-worker process
in the fakeroot environment.
Contains the command that starts the bitbake-worker process in the fakeroot environment.
Lists directories to create before running a task in the fakeroot environment.
Lists environment variables to set when running a task
in the fakeroot environment.
For additional information on environment variables and
the fakeroot environment, see the
FAKEROOTBASEENV
variable.
Lists environment variables to set when running a task
that is not in the fakeroot environment.
For additional information on environment variables and
the fakeroot environment, see the
FAKEROOTENV
variable.
Defines the command the BitBake fetcher module
executes when running fetch operations.
You need to use an override suffix when you use the
variable (e.g. FETCHCMD_git
or FETCHCMD_svn
).
Points at the current file. BitBake sets this variable during the parsing process to identify the file being parsed. BitBake also sets this variable when a recipe is being executed to identify the recipe file.
Specifies directories BitBake uses when searching for
patches and files.
The "local" fetcher module uses these directories when
handling file://
URLs.
The variable behaves like a shell PATH
environment variable.
The value is a colon-separated list of directories that
are searched left-to-right in order.
The directory in which a local copy of a Git repository is stored when it is cloned.
Causes the named class or classes to be inherited globally.
Anonymous functions in the class or classes
are not executed for the
base configuration and in each individual recipe.
The OpenEmbedded build system ignores changes to
INHERIT
in individual recipes.
For more information on INHERIT
, see
the
"INHERIT
Configuration Directive"
section.
Lists the layers, separated by spaces, upon which this recipe depends.
Optionally, you can specify a specific layer version for a dependency
by adding it to the end of the layer name with a colon, (e.g. "anotherlayer:3"
to be compared against
LAYERVERSION
_anotherlayer
in this case).
BitBake produces an error if any dependency is missing or
the version numbers do not match exactly (if specified).
You use this variable in the conf/layer.conf
file.
You must also use the specific layer name as a suffix
to the variable (e.g. LAYERDEPENDS_mylayer
).
When used inside the layer.conf
configuration
file, this variable provides the path of the current layer.
This variable is not available outside of layer.conf
and references are expanded immediately when parsing of the file completes.
When used inside the layer.conf
configuration
file, this variable provides the path of the current layer,
escaped for use in a regular expression
(BBFILE_PATTERN
).
This variable is not available outside of layer.conf
and references are expanded immediately when parsing of the file completes.
Optionally specifies the version of a layer as a single number.
You can use this variable within
LAYERDEPENDS
for another layer in order to depend on a specific version
of the layer.
You use this variable in the conf/layer.conf
file.
You must also use the specific layer name as a suffix
to the variable (e.g. LAYERDEPENDS_mylayer
).
The list of source licenses for the recipe.
Specifies additional paths from which BitBake gets source code.
When the build system searches for source code, it first
tries the local download directory.
If that location fails, the build system tries locations
defined by
PREMIRRORS
,
the upstream source, and then locations specified by
MIRRORS
in that order.
Allows you to suppress BitBake warnings caused when building two separate recipes that provide the same output.
BitBake normally issues a warning when building two
different recipes where each provides the same output.
This scenario is usually something the user does not
want.
However, cases do exist where it makes sense, particularly
in the virtual/*
namespace.
You can use this variable to suppress BitBake's warnings.
To use the variable, list provider names (e.g.
recipe names, virtual/kernel
,
and so forth).
BitBake uses OVERRIDES
to control
what variables are overridden after BitBake parses
recipes and configuration files.
Following is a simple example that uses an overrides list based on machine architectures:
OVERRIDES = "arm:x86:mips:powerpc"
You can find information on how to use
OVERRIDES
in the
"Conditional Syntax (Overrides)"
section.
The directory in which a local copy of a Perforce depot is stored when it is fetched.
The list of packages the recipe creates.
A promise that your recipe satisfies runtime dependencies
for optional modules that are found in other recipes.
PACKAGES_DYNAMIC
does not actually satisfy the dependencies, it only states that
they should be satisfied.
For example, if a hard, runtime dependency
(RDEPENDS
)
of another package is satisfied during the build
through the PACKAGES_DYNAMIC
variable, but a package with the module name is never actually
produced, then the other package will be broken.
The epoch of the recipe. By default, this variable is unset. The variable is used to make upgrades possible when the versioning scheme changes in some backwards incompatible way.
Specifies the directory BitBake uses to store data that should be preserved between builds. In particular, the data stored is the data that uses BitBake's persistent data API and the data used by the PR Server and PR Service.
Specifies the recipe or package name and includes all version and revision
numbers (i.e. eglibc-2.13-r20+svnr15508/
and
bash-4.2-r1/
).
The recipe name.
The revision of the recipe.
Determines which recipe should be given preference when
multiple recipes provide the same item.
You should always suffix the variable with the name of the
provided item, and you should set it to the
PN
of the recipe to which you want to give precedence.
Some examples:
PREFERRED_PROVIDER_virtual/kernel ?= "linux-yocto" PREFERRED_PROVIDER_virtual/xserver = "xserver-xf86" PREFERRED_PROVIDER_virtual/libgl ?= "mesa"
Determines which recipe should be given preference for
cases where multiple recipes provide the same item.
Functionally,
PREFERRED_PROVIDERS
is identical to
PREFERRED_PROVIDER
.
However, the PREFERRED_PROVIDERS
variable lets you define preferences for multiple
situations using the following form:
PREFERRED_PROVIDERS = "xxx:yyy aaa:bbb ..."
This form is a convenient replacement for the following:
PREFERRED_PROVIDER_xxx = "yyy" PREFERRED_PROVIDER_aaa = "bbb"
If there are multiple versions of recipes available, this
variable determines which recipe should be given preference.
You must always suffix the variable with the
PN
you want to select, and you should set
PV
accordingly for precedence.
The PREFERRED_VERSION
variable
supports limited wildcard use through the
"%
" character.
You can use the character to match any number of
characters, which can be useful when specifying versions
that contain long revision numbers that potentially change.
Here are two examples:
PREFERRED_VERSION_python = "2.7.3" PREFERRED_VERSION_linux-yocto = "4.12%"
%
" character
is limited in that it only works at the end of the
string.
You cannot use the wildcard character in any other
location of the string.
Specifies additional paths from which BitBake gets source code.
When the build system searches for source code, it first
tries the local download directory.
If that location fails, the build system tries locations
defined by PREMIRRORS
, the upstream
source, and then locations specified by
MIRRORS
in that order.
Typically, you would add a specific server for the build system to attempt before any others by adding something like the following to your configuration:
PREMIRRORS_prepend = "\ git://.*/.* http://www.yoctoproject.org/sources/ \n \ ftp://.*/.* http://www.yoctoproject.org/sources/ \n \ http://.*/.* http://www.yoctoproject.org/sources/ \n \ https://.*/.* http://www.yoctoproject.org/sources/ \n"
These changes cause the build system to intercept
Git, FTP, HTTP, and HTTPS requests and direct them to
the http://
sources mirror.
You can use file://
URLs to point
to local directories or network shares as well.
A list of aliases by which a particular recipe can be
known.
By default, a recipe's own
PN
is implicitly already in its PROVIDES
list.
If a recipe uses PROVIDES
, the
additional aliases are synonyms for the recipe and can
be useful satisfying dependencies of other recipes during
the build as specified by
DEPENDS
.
Consider the following example
PROVIDES
statement from a recipe
file libav_0.8.11.bb
:
PROVIDES += "libpostproc"
The PROVIDES
statement results in
the "libav" recipe also being known as "libpostproc".
In addition to providing recipes under alternate names,
the PROVIDES
mechanism is also used
to implement virtual targets.
A virtual target is a name that corresponds to some
particular functionality (e.g. a Linux kernel).
Recipes that provide the functionality in question list the
virtual target in PROVIDES
.
Recipes that depend on the functionality in question can
include the virtual target in
DEPENDS
to leave the choice of provider open.
Conventionally, virtual targets have names on the form "virtual/function" (e.g. "virtual/kernel"). The slash is simply part of the name and has no syntactical significance.
The network based
PR
service host and port.
Following is an example of how the PRSERV_HOST
variable is
set:
PRSERV_HOST = "localhost:0"
You must set the variable if you want to automatically
start a local PR service.
You can set PRSERV_HOST
to other
values to use a remote PR service.
The version of the recipe.
Lists a package's runtime dependencies (i.e. other packages) that must be installed in order for the built package to run correctly. If a package in this list cannot be found during the build, you will get a build error.
Because the RDEPENDS
variable applies
to packages being built, you should always use the variable
in a form with an attached package name.
For example, suppose you are building a development package
that depends on the perl
package.
In this case, you would use the following
RDEPENDS
statement:
RDEPENDS_${PN}-dev += "perl"
In the example, the development package depends on
the perl
package.
Thus, the RDEPENDS
variable has the
${PN}-dev
package name as part of the
variable.
BitBake supports specifying versioned dependencies.
Although the syntax varies depending on the packaging
format, BitBake hides these differences from you.
Here is the general syntax to specify versions with
the RDEPENDS
variable:
RDEPENDS_${PN} = "package
(operator
version
)"
For operator
, you can specify the
following:
= < > <= >=
For example, the following sets up a dependency on version
1.2 or greater of the package foo
:
RDEPENDS_${PN} = "foo (>= 1.2)"
For information on build-time dependencies, see the
DEPENDS
variable.
The directory in which a local copy of a
google-repo
directory is stored
when it is synced.
A list of package name aliases that a package also provides.
These aliases are useful for satisfying runtime dependencies
of other packages both during the build and on the target
(as specified by
RDEPENDS
).
As with all package-controlling variables, you must always use the variable in conjunction with a package name override. Here is an example:
RPROVIDES_${PN} = "widget-abi-2"
A list of packages that extends the usability of a package
being built.
The package being built does not depend on this list of
packages in order to successfully build, but needs them for
the extended usability.
To specify runtime dependencies for packages, see the
RDEPENDS
variable.
BitBake supports specifying versioned recommends.
Although the syntax varies depending on the packaging
format, BitBake hides these differences from you.
Here is the general syntax to specify versions with
the RRECOMMENDS
variable:
RRECOMMENDS_${PN} = "package
(operator
version
)"
For operator
, you can specify the
following:
= < > <= >=
For example, the following sets up a recommend on version
1.2 or greater of the package foo
:
RRECOMMENDS_${PN} = "foo (>= 1.2)"
The section in which packages should be categorized.
The list of source files - local or remote.
This variable tells BitBake which bits
to pull for the build and how to pull them.
For example, if the recipe or append file needs to
fetch a single tarball from the Internet, the recipe or
append file uses a SRC_URI
entry that specifies that tarball.
On the other hand, if the recipe or append file needs to
fetch a tarball and include a custom file, the recipe or
append file needs an SRC_URI
variable
that specifies all those sources.
The following list explains the available URI protocols:
file://
-
Fetches files, which are usually files shipped with
the metadata,
from the local machine.
The path is relative to the
FILESPATH
variable.
bzr://
- Fetches files from a
Bazaar revision control repository.
git://
- Fetches files from a
Git revision control repository.
osc://
- Fetches files from
an OSC (OpenSUSE Build service) revision control repository.
repo://
- Fetches files from
a repo (Git) repository.
http://
- Fetches files from
the Internet using HTTP.
https://
- Fetches files
from the Internet using HTTPS.
ftp://
- Fetches files
from the Internet using FTP.
cvs://
- Fetches files from
a CVS revision control repository.
hg://
- Fetches files from
a Mercurial (hg
) revision control repository.
p4://
- Fetches files from
a Perforce (p4
) revision control repository.
ssh://
- Fetches files from
a secure shell.
svn://
- Fetches files from
a Subversion (svn
) revision control repository.
Here are some additional options worth mentioning:
unpack
- Controls
whether or not to unpack the file if it is an archive.
The default action is to unpack the file.
subdir
- Places the file
(or extracts its contents) into the specified
subdirectory.
This option is useful for unusual tarballs or other archives that
do not have their files already in a subdirectory within the archive.
name
- Specifies a
name to be used for association with SRC_URI
checksums
when you have more than one file specified in SRC_URI
.
downloadfilename
- Specifies
the filename used when storing the downloaded file.
The date of the source code used to build the package. This variable applies only if the source was fetched from a Source Code Manager (SCM).
The revision of the source code used to build the package.
This variable applies only when using Subversion, Git, Mercurial and Bazaar.
If you want to build a fixed revision and you want
to avoid performing a query on the remote repository every time
BitBake parses your recipe, you should specify a SRCREV
that is a
full revision identifier and not just a tag.
Helps construct valid
SRCREV
values when multiple source controlled URLs are used in
SRC_URI
.
The system needs help constructing these values under these
circumstances.
Each component in the SRC_URI
is assigned a name and these are referenced
in the SRCREV_FORMAT
variable.
Consider an example with URLs named "machine" and "meta".
In this case, SRCREV_FORMAT
could look
like "machine_meta" and those names would have the SCM
versions substituted into each position.
Only one AUTOINC
placeholder is added
and if needed.
And, this placeholder is placed at the start of the
returned string.
Specifies the base path used to create recipe stamp files. The path to an actual stamp file is constructed by evaluating this string and then appending additional information.
Specifies the base path used to create recipe stamp files.
Unlike the
STAMP
variable, STAMPCLEAN
can contain
wildcards to match the range of files a clean operation
should remove.
BitBake uses a clean operation to remove any other stamps
it should be removing when creating a new stamp.
A short summary for the recipe, which is 72 characters or less.
The directory in which files checked out of a Subversion system are stored.
The simplest example commonly used to demonstrate any new programming language or tool is the "Hello World" example. This appendix demonstrates, in tutorial form, Hello World within the context of BitBake. The tutorial describes how to create a new project and the applicable metadata files necessary to allow BitBake to build it.
See the "Obtaining BitBake" section for information on how to obtain BitBake. Once you have the source code on your machine, the BitBake directory appears as follows:
$ ls -al total 100 drwxrwxr-x. 9 wmat wmat 4096 Jan 31 13:44 . drwxrwxr-x. 3 wmat wmat 4096 Feb 4 10:45 .. -rw-rw-r--. 1 wmat wmat 365 Nov 26 04:55 AUTHORS drwxrwxr-x. 2 wmat wmat 4096 Nov 26 04:55 bin drwxrwxr-x. 4 wmat wmat 4096 Jan 31 13:44 build -rw-rw-r--. 1 wmat wmat 16501 Nov 26 04:55 ChangeLog drwxrwxr-x. 2 wmat wmat 4096 Nov 26 04:55 classes drwxrwxr-x. 2 wmat wmat 4096 Nov 26 04:55 conf drwxrwxr-x. 3 wmat wmat 4096 Nov 26 04:55 contrib -rw-rw-r--. 1 wmat wmat 17987 Nov 26 04:55 COPYING drwxrwxr-x. 3 wmat wmat 4096 Nov 26 04:55 doc -rw-rw-r--. 1 wmat wmat 69 Nov 26 04:55 .gitignore -rw-rw-r--. 1 wmat wmat 849 Nov 26 04:55 HEADER drwxrwxr-x. 5 wmat wmat 4096 Jan 31 13:44 lib -rw-rw-r--. 1 wmat wmat 195 Nov 26 04:55 MANIFEST.in -rw-rw-r--. 1 wmat wmat 2887 Nov 26 04:55 TODO
At this point, you should have BitBake cloned to a directory that matches the previous listing except for dates and user names.
First, you need to be sure that you can run BitBake. Set your working directory to where your local BitBake files are and run the following command:
$ ./bin/bitbake --version BitBake Build Tool Core version 1.23.0, bitbake version 1.23.0
The console output tells you what version you are running.
The recommended method to run BitBake is from a directory of your
choice.
To be able to run BitBake from any directory, you need to add the
executable binary to your binary to your shell's environment
PATH
variable.
First, look at your current PATH
variable
by entering the following:
$ echo $PATH
Next, add the directory location for the BitBake binary to the
PATH
.
Here is an example that adds the
/home/scott-lenovo/bitbake/bin
directory
to the front of the PATH
variable:
$ export PATH=/home/scott-lenovo/bitbake/bin:$PATH
You should now be able to enter the bitbake
command from the command line while working from any directory.
The overall goal of this exercise is to build a complete "Hello World" example utilizing task and layer concepts. Because this is how modern projects such as OpenEmbedded and the Yocto Project utilize BitBake, the example provides an excellent starting point for understanding BitBake.
To help you understand how to use BitBake to build targets,
the example starts with nothing but the bitbake
command, which causes BitBake to fail and report problems.
The example progresses by adding pieces to the build to
eventually conclude with a working, minimal "Hello World"
example.
While every attempt is made to explain what is happening during the example, the descriptions cannot cover everything. You can find further information throughout this manual. Also, you can actively participate in the http://lists.openembedded.org/mailman/listinfo/bitbake-devel discussion mailing list about the BitBake build tool.
As stated earlier, the goal of this example is to eventually compile "Hello World". However, it is unknown what BitBake needs and what you have to provide in order to achieve that goal. Recall that BitBake utilizes three types of metadata files: Configuration Files, Classes, and Recipes. But where do they go? How does BitBake find them? BitBake's error messaging helps you answer these types of questions and helps you better understand exactly what is going on.
Following is the complete "Hello World" example.
Create a Project Directory: First, set up a directory for the "Hello World" project. Here is how you can do so in your home directory:
$ mkdir ~/hello $ cd ~/hello
This is the directory that BitBake will use to do all of its work. You can use this directory to keep all the metafiles needed by BitBake. Having a project directory is a good way to isolate your project.
Run BitBake:
At this point, you have nothing but a project directory.
Run the bitbake
command and see what
it does:
$ bitbake The BBPATH variable is not set and bitbake did not find a conf/bblayers.conf file in the expected location. Maybe you accidentally invoked bitbake from the wrong directory? DEBUG: Removed the following variables from the environment: GNOME_DESKTOP_SESSION_ID, XDG_CURRENT_DESKTOP, GNOME_KEYRING_CONTROL, DISPLAY, SSH_AGENT_PID, LANG, no_proxy, XDG_SESSION_PATH, XAUTHORITY, SESSION_MANAGER, SHLVL, MANDATORY_PATH, COMPIZ_CONFIG_PROFILE, WINDOWID, EDITOR, GPG_AGENT_INFO, SSH_AUTH_SOCK, GDMSESSION, GNOME_KEYRING_PID, XDG_SEAT_PATH, XDG_CONFIG_DIRS, LESSOPEN, DBUS_SESSION_BUS_ADDRESS, _, XDG_SESSION_COOKIE, DESKTOP_SESSION, LESSCLOSE, DEFAULTS_PATH, UBUNTU_MENUPROXY, OLDPWD, XDG_DATA_DIRS, COLORTERM, LS_COLORS
The majority of this output is specific to environment variables
that are not directly relevant to BitBake.
However, the very first message regarding the
BBPATH
variable and the
conf/bblayers.conf
file
is relevant.
When you run BitBake, it begins looking for metadata files.
The
BBPATH
variable is what tells BitBake where to look for those files.
BBPATH
is not set and you need to set it.
Without BBPATH
, BitBake cannot
find any configuration files (.conf
)
or recipe files (.bb
) at all.
BitBake also cannot find the bitbake.conf
file.
Setting BBPATH
:
For this example, you can set BBPATH
in the same manner that you set PATH
earlier in the appendix.
You should realize, though, that it is much more flexible to set the
BBPATH
variable up in a configuration
file for each project.
From your shell, enter the following commands to set and
export the BBPATH
variable:
$ BBPATH="projectdirectory
"
$ export BBPATH
Use your actual project directory in the command. BitBake uses that directory to find the metadata it needs for your project.
Run BitBake:
Now that you have BBPATH
defined, run
the bitbake
command again:
$ bitbake ERROR: Traceback (most recent call last): File "/home/scott-lenovo/bitbake/lib/bb/cookerdata.py", line 163, in wrapped return func(fn, *args) File "/home/scott-lenovo/bitbake/lib/bb/cookerdata.py", line 173, in parse_config_file return bb.parse.handle(fn, data, include) File "/home/scott-lenovo/bitbake/lib/bb/parse/__init__.py", line 99, in handle return h['handle'](fn, data, include) File "/home/scott-lenovo/bitbake/lib/bb/parse/parse_py/ConfHandler.py", line 120, in handle abs_fn = resolve_file(fn, data) File "/home/scott-lenovo/bitbake/lib/bb/parse/__init__.py", line 117, in resolve_file raise IOError("file %s not found in %s" % (fn, bbpath)) IOError: file conf/bitbake.conf not found in /home/scott-lenovo/hello ERROR: Unable to parse conf/bitbake.conf: file conf/bitbake.conf not found in /home/scott-lenovo/hello
This sample output shows that BitBake could not find the
conf/bitbake.conf
file in the project
directory.
This file is the first thing BitBake must find in order
to build a target.
And, since the project directory for this example is
empty, you need to provide a conf/bitbake.conf
file.
Creating conf/bitbake.conf
:
The conf/bitbake.conf
includes a number of
configuration variables BitBake uses for metadata and recipe
files.
For this example, you need to create the file in your project directory
and define some key BitBake variables.
For more information on the bitbake.conf
file,
see
http://git.openembedded.org/bitbake/tree/conf/bitbake.conf.
Use the following commands to create the conf
directory in the project directory:
$ mkdir conf
From within the conf
directory, use
some editor to create the bitbake.conf
so that it contains the following:
PN = "${@bb.parse.BBHandler.vars_from_file(d.getVar('FILE', False),d)[0] or 'defaultpkgname'}"
TMPDIR = "${TOPDIR}/tmp" CACHE = "${TMPDIR}/cache" STAMP = "${TMPDIR}/${PN}/stamps" T = "${TMPDIR}/${PN}/work" B = "${TMPDIR}/${PN}"
PN
, the
variables STAMP
,
T
, and B
,
prevent more than one recipe from working. You can fix
this by either setting PN
to have
a value similar to what OpenEmbedded and BitBake use
in the default bitbake.conf
file
(see previous example). Or, by manually updating each
recipe to set PN
. You will also
need to include PN
as part of the
STAMP
, T
, and
B
variable definitions in the
local.conf
file.
The TMPDIR
variable establishes a directory
that BitBake uses for build output and intermediate files other
than the cached information used by the
Setscene process.
Here, the TMPDIR
directory is set to
hello/tmp
.
tmp
directory in order to rebuild a BitBake target.
The build process creates the directory for you
when you run BitBake.
For information about each of the other variables defined in this example, click on the links to take you to the definitions in the glossary.
Run BitBake:
After making sure that the conf/bitbake.conf
file exists, you can run the bitbake
command again:
$ bitbake ERROR: Traceback (most recent call last): File "/home/scott-lenovo/bitbake/lib/bb/cookerdata.py", line 163, in wrapped return func(fn, *args) File "/home/scott-lenovo/bitbake/lib/bb/cookerdata.py", line 177, in _inherit bb.parse.BBHandler.inherit(bbclass, "configuration INHERITs", 0, data) File "/home/scott-lenovo/bitbake/lib/bb/parse/parse_py/BBHandler.py", line 92, in inherit include(fn, file, lineno, d, "inherit") File "/home/scott-lenovo/bitbake/lib/bb/parse/parse_py/ConfHandler.py", line 100, in include raise ParseError("Could not %(error_out)s file %(fn)s" % vars(), oldfn, lineno) ParseError: ParseError in configuration INHERITs: Could not inherit file classes/base.bbclass ERROR: Unable to parse base: ParseError in configuration INHERITs: Could not inherit file classes/base.bbclass
In the sample output, BitBake could not find the
classes/base.bbclass
file.
You need to create that file next.
Creating classes/base.bbclass
:
BitBake uses class files to provide common code and functionality.
The minimally required class for BitBake is the
classes/base.bbclass
file.
The base
class is implicitly inherited by
every recipe.
BitBake looks for the class in the classes
directory of the project (i.e hello/classes
in this example).
Create the classes
directory as follows:
$ cd $HOME/hello $ mkdir classes
Move to the classes
directory and then
create the base.bbclass
file by inserting
this single line:
addtask build
The minimal task that BitBake runs is the
do_build
task.
This is all the example needs in order to build the project.
Of course, the base.bbclass
can have much
more depending on which build environments BitBake is
supporting.
Run BitBake:
After making sure that the classes/base.bbclass
file exists, you can run the bitbake
command again:
$ bitbake Nothing to do. Use 'bitbake world' to build everything, or run 'bitbake --help' for usage information.
BitBake is finally reporting no errors. However, you can see that it really does not have anything to do. You need to create a recipe that gives BitBake something to do.
Creating a Layer: While it is not really necessary for such a small example, it is good practice to create a layer in which to keep your code separate from the general metadata used by BitBake. Thus, this example creates and uses a layer called "mylayer".
Minimally, you need a recipe file and a layer configuration
file in your layer.
The configuration file needs to be in the conf
directory inside the layer.
Use these commands to set up the layer and the conf
directory:
$ cd $HOME $ mkdir mylayer $ cd mylayer $ mkdir conf
Move to the conf
directory and create a
layer.conf
file that has the following:
BBPATH .= ":${LAYERDIR}" BBFILES += "${LAYERDIR}/*.bb" BBFILE_COLLECTIONS += "mylayer" BBFILE_PATTERN_mylayer := "^${LAYERDIR_RE}/"
For information on these variables, click the links to go to the definitions in the glossary.
You need to create the recipe file next.
Inside your layer at the top-level, use an editor and create
a recipe file named printhello.bb
that
has the following:
DESCRIPTION = "Prints Hello World" PN = 'printhello' PV = '1' python do_build() { bb.plain("********************"); bb.plain("* *"); bb.plain("* Hello, World! *"); bb.plain("* *"); bb.plain("********************"); }
The recipe file simply provides a description of the
recipe, the name, version, and the do_build
task, which prints out "Hello World" to the console.
For more information on these variables, follow the links
to the glossary.
Run BitBake With a Target: Now that a BitBake target exists, run the command and provide that target:
$ cd $HOME/hello $ bitbake printhello ERROR: no recipe files to build, check your BBPATH and BBFILES? Summary: There was 1 ERROR message shown, returning a non-zero exit code.
We have created the layer with the recipe and the layer
configuration file but it still seems that BitBake cannot
find the recipe.
BitBake needs a conf/bblayers.conf
that
lists the layers for the project.
Without this file, BitBake cannot find the recipe.
Creating conf/bblayers.conf
:
BitBake uses the conf/bblayers.conf
file
to locate layers needed for the project.
This file must reside in the conf
directory
of the project (i.e. hello/conf
for this
example).
Set your working directory to the hello/conf
directory and then create the bblayers.conf
file so that it contains the following:
BBLAYERS ?= " \ /home/<you>/mylayer \ "
You need to provide your own information for
you
in the file.
Run BitBake With a Target:
Now that you have supplied the bblayers.conf
file, run the bitbake
command and provide
the target:
$ bitbake printhello Parsing recipes: 100% |##################################################################################| Time: 00:00:00 Parsing of 1 .bb files complete (0 cached, 1 parsed). 1 targets, 0 skipped, 0 masked, 0 errors. NOTE: Resolving any missing task queue dependencies NOTE: Preparing RunQueue NOTE: Executing RunQueue Tasks ******************** * * * Hello, World! * * * ******************** NOTE: Tasks Summary: Attempted 1 tasks of which 0 didn't need to be rerun and all succeeded.
BitBake finds the printhello
recipe and
successfully runs the task.
bitbake printhello
again will not
result in a BitBake run that prints the same console
output.
The reason for this is that the first time the
printhello.bb
recipe's
do_build
task executes
successfully, BitBake writes a stamp file for the task.
Thus, the next time you attempt to run the task
using that same bitbake
command,
BitBake notices the stamp and therefore determines
that the task does not need to be re-run.
If you delete the tmp
directory
or run bitbake -c clean printhello
and then re-run the build, the "Hello, World!" message will
be printed again.