Yocto Project Development Manual

Scott Rifenbark

Intel Corporation

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.

Note

For the latest version of this manual associated with this Yocto Project release, see the Yocto Project Development Manual from the Yocto Project website.
Revision History
Revision 1.1 6 October 2011
The initial document released with the Yocto Project 1.1 Release.
Revision 1.2 April 2012
Released with the Yocto Project 1.2 Release.
Revision 1.3 October 2012
Released with the Yocto Project 1.3 Release.
Revision 1.4 April 2013
Released with the Yocto Project 1.4 Release.
Revision 1.5 October 2013
Released with the Yocto Project 1.5 Release.
Revision 1.5.1 January 2014
Released with the Yocto Project 1.5.1 Release.
Revision 1.6 April 2014
Released with the Yocto Project 1.6 Release.
Revision 1.7 October 2014
Released with the Yocto Project 1.7 Release.
Revision 1.8 April 2015
Released with the Yocto Project 1.8 Release.
Revision 1.9 October 2015
Released with the Yocto Project 1.9 Release.

Table of Contents

1. The Yocto Project Development Manual
1.1. Introduction
1.2. What This Manual Provides
1.3. What this Manual Does Not Provide
1.4. Other Information
2. Getting Started with the Yocto Project
2.1. Introducing the Yocto Project
2.2. Getting Set Up
2.3. Building Images
2.4. Using Pre-Built Binaries and QEMU
3. The Yocto Project Open Source Development Environment
3.1. Open Source Philosophy
3.2. Using the Yocto Project in a Team Environment
3.2.1. System Configurations
3.2.2. Source Control Management (SCM)
3.2.3. Autobuilders
3.2.4. Policies and Change Flow
3.2.5. Summary
3.3. Yocto Project Source Repositories
3.4. Yocto Project Terms
3.5. Licensing
3.6. Git
3.6.1. Repositories, Tags, and Branches
3.6.2. Basic Commands
3.7. Workflows
3.8. Tracking Bugs
3.9. How to Submit a Change
3.9.1. Using Scripts to Push a Change Upstream and Request a Pull
3.9.2. Using Email to Submit a Patch
4. Common Development Models
4.1. System Development Workflow
4.1.1. Developing a Board Support Package (BSP)
4.1.2. Modifying the Kernel
4.2. Application Development Workflow
4.2.1. Workflow Using the ADT and Eclipse
4.2.2. Working Within Eclipse
4.2.3. Workflow Using Stand-Alone Cross-Development Toolchains
4.3. Modifying Source Code
4.3.1. Using devtool in Your Workflow
4.3.2. devtool Quick Reference
4.3.3. Using Quilt in Your Workflow
4.3.4. Finding Temporary Source Code
4.4. Image Development Using Toaster
4.5. Image Development Using Hob
4.6. Using a Development Shell
5. Common Tasks
5.1. Understanding and Creating Layers
5.1.1. Layers
5.1.2. Creating Your Own Layer
5.1.3. Best Practices to Follow When Creating Layers
5.1.4. Enabling Your Layer
5.1.5. Using .bbappend Files
5.1.6. Prioritizing Your Layer
5.1.7. Managing Layers
5.1.8. Creating a General Layer Using the yocto-layer Script
5.2. Customizing Images
5.2.1. Customizing Images Using local.conf
5.2.2. Customizing Images Using Custom IMAGE_FEATURES and EXTRA_IMAGE_FEATURES
5.2.3. Customizing Images Using Custom .bb Files
5.2.4. Customizing Images Using Custom Package Groups
5.2.5. Customizing an Image Hostname
5.3. Writing a New Recipe
5.3.1. Overview
5.3.2. Locate or Automatically Create a Base Recipe
5.3.3. Storing and Naming the Recipe
5.3.4. Understanding Recipe Syntax
5.3.5. Running a Build on the Recipe
5.3.6. Fetching Code
5.3.7. Unpacking Code
5.3.8. Patching Code
5.3.9. Licensing
5.3.10. Configuring the Recipe
5.3.11. Compilation
5.3.12. Installing
5.3.13. Enabling System Services
5.3.14. Packaging
5.3.15. Properly Versioning Pre-Release Recipes
5.3.16. Post-Installation Scripts
5.3.17. Testing
5.3.18. Examples
5.4. Adding a New Machine
5.4.1. Adding the Machine Configuration File
5.4.2. Adding a Kernel for the Machine
5.4.3. Adding a Formfactor Configuration File
5.5. Working With Libraries
5.5.1. Including Static Library Files
5.5.2. Combining Multiple Versions of Library Files into One Image
5.5.3. Installing Multiple Versions of the Same Library
5.6. Creating Partitioned Images
5.6.1. Background
5.6.2. Requirements
5.6.3. Getting Help
5.6.4. Operational Modes
5.6.5. Using an Existing Kickstart File
5.6.6. Examples
5.6.7. Plugins
5.6.8. OpenEmbedded Kickstart (.wks) Reference
5.7. Configuring the Kernel
5.7.1. Using  menuconfig
5.7.2. Creating a  defconfig File
5.7.3. Creating Configuration Fragments
5.7.4. Fine-Tuning the Kernel Configuration File
5.8. Patching the Kernel
5.8.1. Create a Layer for your Changes
5.8.2. Finding the Kernel Source Code
5.8.3. Creating the Patch
5.8.4. Set Up Your Layer for the Build
5.8.5. Set Up for the Build
5.8.6. Build the Modified QEMU Kernel Image
5.8.7. Boot the Image and Verify Your Changes
5.9. Making Images More Secure
5.9.1. General Considerations
5.9.2. Security Flags
5.9.3. Considerations Specific to the OpenEmbedded Build System
5.9.4. Tools for Hardening Your Image
5.10. Creating Your Own Distribution
5.11. Creating a Custom Template Configuration Directory
5.12. Building a Tiny System
5.12.1. Overview
5.12.2. Goals and Guiding Principles
5.12.3. Understand What Contributes to Your Image Size
5.12.4. Trim the Root Filesystem
5.12.5. Trim the Kernel
5.12.6. Remove Package Management Requirements
5.12.7. Look for Other Ways to Minimize Size
5.12.8. Iterate on the Process
5.13. Building Images for More than One Machine
5.14. Working with Packages
5.14.1. Excluding Packages from an Image
5.14.2. Incrementing a Package Revision Number
5.14.3. Handling Optional Module Packaging
5.14.4. Using Runtime Package Management
5.14.5. Testing Packages With ptest
5.15. Working with Source Files
5.15.1. Setting up Effective Mirrors
5.15.2. Getting Source Files and Suppressing the Build
5.16. Building Software from an External Source
5.17. Selecting an Initialization Manager
5.17.1. Using systemd Exclusively
5.17.2. Using systemd for the Main Image and Using SysVinit for the Rescue Image
5.18. Selecting a Device Manager
5.18.1. Using Persistent and Pre-Populated/dev
5.18.2. Using devtmpfs and a Device Manager
5.19. Using an External SCM
5.20. Creating a Read-Only Root Filesystem
5.20.1. Creating the Root Filesystem
5.20.2. Post-Installation Scripts
5.20.3. Areas With Write Access
5.21. Performing Automated Runtime Testing
5.21.1. Enabling Tests
5.21.2. Running Tests
5.21.3. Exporting Tests
5.21.4. Writing New Tests
5.22. Debugging With the GNU Project Debugger (GDB) Remotely
5.22.1. Set Up the Cross-Development Debugging Environment
5.22.2. Launch Gdbserver on the Target
5.22.3. Launch GDB on the Host Computer
5.22.4. Connect to the Remote GDB Server
5.22.5. Use the Debugger
5.23. Debugging Parallel Make Races
5.23.1. The Failure
5.23.2. Reproducing the Error
5.23.3. Creating a Patch for the Fix
5.23.4. Testing the Build
5.24. Profiling with OProfile
5.24.1. Profiling on the Target
5.24.2. Using OProfileUI
5.25. Maintaining Open Source License Compliance During Your Product's Lifecycle
5.25.1. Providing the Source Code
5.25.2. Providing License Text
5.25.3. Providing Compilation Scripts and Source Code Modifications
5.26. Using the Error Reporting Tool
5.26.1. Enabling and Using the Tool
5.26.2. Disabling the Tool
5.26.3. Setting Up Your Own Error Reporting Server
6. Using the Quick EMUlator (QEMU)
6.1. Overview
6.2. Running QEMU
6.2.1. Setting Up the Environment
6.2.2. Using the runqemu Command
6.3. Running Under a Network File System (NFS) Server
6.3.1. Setting Up to Use NFS
6.3.2. Starting and Stopping NFS
6.4. Tips and Tricks