Matter Open IoT SDK Example Application

These examples are built using Open IoT SDK and runs inside an emulated target through the Arm FVP model for the Corstone-300 MPS3.

The list of currently supported Matter examples:

shell
lock-app

You can use these examples as a reference for creating your own applications.

Environment setup

The VSCode devcontainer has all the dependencies pre-installed. It is the recommended way to build, run and develop with the Open IoT SDK port of the Matter Project. Please read this VSCode development guide for more information.

Before building the examples, check out the Matter repository and sync Open IoT SDK submodules using the following command:

scripts/checkout_submodules.py --shallow --recursive --platform openiotsdk

Next, bootstrap the source tree to install Pigweed (CIPD and Python packages) components inside your environment (only once).

To bootstrap:

using CLI

$ bash scripts/bootstrap.sh

using VSCode tasks

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Bootstrap

Networking setup

Running ARM Fast Model with the TAP/TUN device networking mode requires the setting up of proper network interfaces. Special scripts were designed to make the setup easy. In the scripts/setup/openiotsdk directory you can find:

  • network_setup.sh - script to create the specific network namespace and Virtual Ethernet interface to connect with the host network. Both host and namespace sides have linked IP addresses. Inside the network namespace the TAP device interface is created and bridged with a Virtual Ethernet peer. There is also an option to enable an Internet connection in the namespace by forwarding traffic to the host default interface.

    To enable the Open IoT SDK networking environment:

    ${MATTER_ROOT}/scripts/setup/openiotsdk/network_setup.sh up

    To disable the Open IoT SDK networking environment:

    ${MATTER_ROOT}/scripts/setup/openiotsdk/network_setup.sh down

    To restart the Open IoT SDK networking environment:

    ${MATTER_ROOT}/scripts/setup/openiotsdk/network_setup.sh restart

    The default scripts settings are:

    • ARM - network base name
    • current session user - network namespace user
    • fe00::1 - host side IPv6 address
    • fe00::2 - namespace side IPv6 address
    • 10.200.1.1 - host side IPv4 address
    • 10.200.1.2 - namespace side IPv4 address
    • no Internet connection support to network namespace

    Example of the OIS network environment settings:

    ARMns namespace configuration
ARMbr: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 10.200.1.2  netmask 255.255.255.0  broadcast 0.0.0.0
        inet6 fe00::2  prefixlen 64  scopeid 0x0<global>
        inet6 fe80::1809:17ff:fe6c:f566  prefixlen 64  scopeid 0x20<link>
        ether 1a:09:17:6c:f5:66  txqueuelen 1000  (Ethernet)
        RX packets 1  bytes 72 (72.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 0  bytes 0 (0.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

ARMnveth: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        ether 46:66:29:a6:91:4b  txqueuelen 1000  (Ethernet)
        RX packets 2  bytes 216 (216.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 3  bytes 270 (270.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

ARMtap: flags=4419<UP,BROADCAST,RUNNING,PROMISC,MULTICAST>  mtu 1500
        ether 1a:09:17:6c:f5:66  txqueuelen 1000  (Ethernet)
        RX packets 0  bytes 0 (0.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 0  bytes 0 (0.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536
        inet 127.0.0.1  netmask 255.0.0.0
        inet6 ::1  prefixlen 128  scopeid 0x10<host>
        loop  txqueuelen 1000  (Local Loopback)
        RX packets 0  bytes 0 (0.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 0  bytes 0 (0.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

Host configuration
ARMhveth: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 10.200.1.1  netmask 255.255.255.0  broadcast 0.0.0.0
        inet6 fe80::147c:c9ff:fe4a:c6d2  prefixlen 64  scopeid 0x20<link>
        inet6 fe00::1  prefixlen 64  scopeid 0x0<global>
        ether 16:7c:c9:4a:c6:d2  txqueuelen 1000  (Ethernet)
        RX packets 3  bytes 270 (270.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 2  bytes 216 (216.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

    Use --help to get more information about the script options.

    Open IoT SDK network setup is also supported via VScode tasks:

    • Open the Command Palette: F1
    • Select Tasks: Run Task
    • Select Setup Open IoT SDK network
    • Enter the network namespace name
    • Choose command

    The VSCode task invokes network_setup.sh with the selected parameters.

  • connect_if.sh - script that connects specified network interfaces with the default route interface. It creates a bridge and links all interfaces to it. The bridge becomes the default interface.

    Example:

    ${MATTER_ROOT}/scripts/setup/openiotsdk/connect_if.sh ARMhveth

    Use --help to get more information about the script options.

Open IoT SDK network setup scripts contain commands that require root permissions. Use sudo to run the scripts in a user account with root privileges.

After setting up the Open IoT SDK network environment the user will be able to run Matter examples on FVP in an isolated network namespace in TAP device mode.

To execute a command in a specific network namespace use the helper script scripts/run_in_ns.sh.

Example:

${MATTER_ROOT}/scripts/run_in_ns.sh ARMns <command to run>

Use --help to get more information about the script options.

💡 Notes:

For Docker environment users it's recommended to use the default bridge network for a running container. This guarantees full isolation of the Open IoT SDK network from host settings.

Debugging setup

Debugging the Matter application running on FVP model requires GDB Remote Connection Plugin for Fast Model. More details GDBRemoteConnection.

The Fast Models FVP add-on package can be downloaded from the ARM developer website Fast models. After login in to the ARM developer platform search for Fast Models, choose Fast Models (FM000A) on the list of results, then choose the revision r11p16-16rel0 and download the Third Party Add-ons for Fast Models 11.16 (Linux) package. Then unpack the package in the selected location on the host machine.

Now you should add the GDB Remote Connection Plugin to your development environment:

  • Linux host environment:

    • install Fast Model Extension package by executing the command ./setup.bin, and follow the installation instructions. After installation, the GDB Remote Connection Plugin should be visible in <installation directory>/FastModelsPortfolio_11.16/plugins/Linux64_GCC-9.3 directory.

    • add GDB plugin path to environment variable as FAST_MODEL_PLUGINS_PATH.

      Example:

      export FAST_MODEL_PLUGINS_PATH=<installation directory>/FastModelsPortfolio_11.16/plugins/Linux64_GCC-9.3

  • Docker container environment:

    • pass the Fast Model Extension package to Docker container development environment by mounting it into the /opt/FastModels_ThirdParty_IP_11-16_b16_Linux64 directory in the container. Add a volume bound to this directory Add local file mount.

      You can edit the .devcontainer/devcontainer.json file, for example:

      ...
"mounts": [ ...
"source=/opt/FastModels_ThirdParty_IP_11-16_b16_Linux64,target=/opt/FastModels_ThirdParty_IP_11-16_b16_Linux64,type=bind,consistency=cached"
... ],
...

      Or if you launch the Docker container directly from CLI, use the above arguments with docker run command:

      docker run ... --mount type=bind,source=/opt/FastModels_ThirdParty_IP_11-16_b16_Linux64,target=/opt/FastModels_ThirdParty_IP_11-16_b16_Linux64 ...

    • install the Fast Model Extension package via setup script inside Docker container:

      ${MATTER_ROOT}/scripts/setup/openiotsdk/debugging_setup.sh

    • the GDB Remote Connection Plugin should be visible in /opt/FastModelsPortfolio_11.16/plugins/Linux64_GCC-9.3 directory.

      • For VScode devcontainer use the environment variable FAST_MODEL_PLUGINS_PATH to point to the correct directory.
      • If the Docker container is directly launched remember to add the GDB Remote Connection Plugin path to the environment variable FAST_MODEL_PLUGINS_PATH inside the container:
        export FAST_MODEL_PLUGINS_PATH=/opt/FastModelsPortfolio_11.16/plugins/Linux64_GCC-9.3

Testing setup

The Matter Python packages are required for the integration test suite. They are not provided as part of the VSCode devcontainer. To install these run the following command from the CLI:

${MATTER_ROOT}/scripts/run_in_build_env.sh \
    './scripts/build_python.sh --install_virtual_env out/venv'

source out/venv/bin/activate

More information about the Python tools you can find here.

Configuration

Trusted Firmware-M

To add TF-M support to Matter example you need to set TFM_SUPPORT variable inside main application CMakeLists.txt file.

set(TFM_SUPPORT YES)

This causes the Matter example to be built as non-secure application in Non-secure Processing Environment (NSPE). The bootloader and the secure part are also built from TF-M sources. All components are merged into a single executable file at the end of the building process.

You can also provide the own version of Matter example by setting TFM_NS_APP_VERSION variable.

set(TFM_NS_APP_VERSION "0.0.1")

Trusted Firmware-M Protected Storage

By default, the Block Device storage is used for storing Matter key-value data.

There is an option to add TF-M Protected Storage Service support for key-value storage component in the Matter examples. Set the variable CONFIG_CHIP_OPEN_IOT_SDK_USE_PSA_PS to YES to add TF-M Protected Storage support to your application. You can put it inside the main application CMakeLists.txt file:

set(CONFIG_CHIP_OPEN_IOT_SDK_USE_PSA_PS YES)

or add as a Cmake command-line parameter:

cmake -G <...> -DCONFIG_CHIP_OPEN_IOT_SDK_USE_PSA_PS=YES <...>

This option causes key-value objects will be stored in a secure part of flash memory and the Protected Storage Service takes care of their encryption and authentication.

💡 Notes:

The TF-M Protected Storage option requires enabling TF-M support.

The -k/--kvsstore option in Open IoT SDK build script selects key-value storage implementation for the Matter's examples. It demonstrates how to use the CONFIG_CHIP_OPEN_IOT_SDK_USE_PSA_PS variable.

Storing persistent memory block in external files

The persistent storage is required to store key-value data of the Matter examples.

Two storage types are supported:

  • Block device storage: The memory partition is located in non-secure SRAM
  • TF-M protected storage: The memory partition is located in secure QSPI_RAM

Fast models offers option to load and dump memory content. More details are available here. Depending on the storage implementation, different flags are used in the FVP options.

For block device storage use:

--dump mps3_board.sram=<file-path>@0:0x0,0x100000
--data mps3_board.sram=<file-path>@0:0x0

For TF-M protected storage use:

--dump mps3_board.qspi_sram=<file-path>@0:0x660000,0x12000
--data mps3_board.qspi_sram=<file-path>@0:0x660000

💡 Notes:

The file-path must exist to use the --data option.

Open IoT SDK build script provides the -K,--kvsfile option to use the persistence options listed above.

Crypto backend

Open IoT SDK port supports two crypto backend implementations:

The CMake variable CONFIG_CHIP_CRYPTO controls how cryptographic operations are implemented in Matter. It accepts two values:

This variable can be set in the main application CMakeLists.txt:

set(CONFIG_CHIP_CRYPTO <mbedtls | psa>)

The variable can also be defined with CMake CLI:

cmake -G <...> -DCONFIG_CHIP_CRYPTO=<mbedtls | psa> <...>

💡 Notes:

The TF-M PSA crypto option requires enabling TF-M support.

Building

You can build examples using the dedicated VSCode task or by calling directly the build script from the command line.

Building using the VSCode task

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Build Open IoT SDK example
  • Decide on debug mode support
  • Decide on LwIP debug logs support
  • Choose crypto algorithm
  • Choose example name

This will call the script with the selected parameters.

Building using CLI

You can call the script directly yourself.

${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh <example name>

Use --help to get more information about the script options.

Running

The application runs in the background and opens a telnet session. The telnet client connects to the port used by the FVP. When the telnet process is terminated it also terminates the FVP instance.

To exit the telnet session, type CTRL + ]. This changes the command prompt to show as:

telnet>

Back in the terminal, type in the word ‘close’ to terminate the session.

telnet> close

You can run an example by using a VSCode task or by calling the run script directly from the command line.

Running using the VSCode task

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Run Open IoT SDK example
  • Enter network namespace
  • Enter network interface
  • Choose example name

This will call the script with the selected example name.

Running using CLI

You can call the script directly yourself.

${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -C run <example name>

Run example in specific network namespace with TAP device mode:

${MATTER_ROOT}/scripts/run_in_ns.sh ARMns ${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -C run -n ARMtap <example name>

Testing

Run the Pytest integration test for the specific application.

The test result can be found in the src/test_driver/openiotsdk/integration-tests/<example name>/test_report.json file.

You can test an example by using a VSCode task or by calling the test script directly from the command line.

Testing using the VSCode task

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Test Open IoT SDK example
  • Enter network namespace
  • Enter network interface
  • Choose example name

This will call the scripts with the selected example name.

Testing using CLI

You can call the script directly yourself.

${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -C test <example name>

Testing an example in a specific network namespace with TAP device mode:

${MATTER_ROOT}/scripts/run_in_ns.sh ARMns ${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -C test -n ARMtap <example name>

Debugging

Before debugging ensure the following:

  1. The debug environment is correctly setup: debugging setup.

  2. The example is compiled with debug symbols enabled:

    For CLI:

    ${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -d true <example name>

    For the VSCode task:

    => Use debug mode (true)

  3. The test network is correctly setup (if required): see networking setup.

General instructions

  • Click Run and Debug from the primary side menu or press Ctrl+Shift+D
  • Select Debug Open IoT SDK example application from the drop down list
  • Click Start Debugging(green triangle) or press F5
  • Choose example name
  • Enter GDB target address
  • Enter network namespace
  • Enter network interface
  • Choose example name

As soon as a debugging session starts, the DEBUG CONSOLE panel is displayed and shows the debugging output. Use debug controls to debug the current application.

For debugging remote targets (i.e. run in other network namespaces) you need to pass the hostname/IP address of the external GDB target that you want to connect to (GDB target address).

In the case of using the Open IoT SDK network environment the GDB server runs inside a namespace and has the same IP address as the bridge interface.

${MATTER_ROOT}/scripts/run_in_ns.sh <namespace_name> ifconfig <bridge_name>

The network namespace name and TAP interface name are also required then.

The application with GDB Remote Connection Plugin runs in the background and opens a telnet session in terminal. The telnet client connects to the port used by the FVP. When the telnet process is terminated it will also terminate the FVP instance.

To exit the telnet session, type CTRL + ]. This changes the command prompt to show as:

telnet>

Back in the terminal, type in the word ‘close’ to terminate the session.

telnet> close

💡 Notes:

As you can see above, you will need to select the name of the example twice. This is because the debug task needs to launch the run task and currently VS code has no way of passing parameters between tasks.

There are issues with debugging examples when the Docker container use the network host and VPN connection is established. Changing routing negatively affects debugging process. It is recommended not to use VPN connections while debugging.

Specific examples

Build lock-app example with PSA crypto backend support and run it in the network namespace

Using CLI

${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -b psa lock-app

export TEST_NETWORK_NAME=OIStest

sudo ${MATTER_ROOT}/scripts/setup/openiotsdk/network_setup.sh -n $TEST_NETWORK_NAME restart

${MATTER_ROOT}/scripts/examples/scripts/run_in_ns.sh ${TEST_NETWORK_NAME}ns
${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -C run -n ${TEST_NETWORK_NAME}tap
lock-app

Using the VSCode task

Build example:

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Build Open IoT SDK example
  • Deny debug mode support false
  • Deny LwIP debug logs support false
  • Choose crypto algorithm psa
  • Choose example name lock-app

Setup network environment:

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Setup Open IoT SDK network
  • Enter the network namespace name OIStest
  • Choose command restart

Run example:

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Run Open IoT SDK example
  • Enter network namespace OIStestns
  • Enter network interface OIStesttap
  • Choose example name lock-app

The example output should be seen in the terminal window.

Build lock-app example with mbedtls crypto backend support and execute its test in the network namespace

Using CLI

${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -b mbedtls lock-app

export TEST_NETWORK_NAME=OIStest

sudo ${MATTER_ROOT}/scripts/setup/openiotsdk/network_setup.sh -n $TEST_NETWORK_NAME restart

${MATTER_ROOT}/scripts/examples/scripts/run_in_ns.sh ${TEST_NETWORK_NAME}ns
${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -C test -n ${TEST_NETWORK_NAME}tap
lock-app

Using the VSCode task

Build example:

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Build Open IoT SDK example
  • Deny debug mode support false
  • Deny LwIP debug logs support false
  • Choose crypto algorithm mbedtls
  • Choose example name lock-app

Setup network environment:

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Setup Open IoT SDK network
  • Enter the network namespace name OIStest
  • Choose command restart

Test example:

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Test Open IoT SDK example
  • Enter network namespace OIStestns
  • Enter network interface OIStesttap
  • Choose example name lock-app

Build lock-app example with mbedtls crypto backend support in debug mode and debug it in the network namespace using the VSCode task

Build example:

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Build Open IoT SDK example
  • Confirm debug mode support true
  • Deny LwIP debug logs support false
  • Choose crypto algorithm mbedtls
  • Choose example name lock-app

Setup network environment:

  • Open the Command Palette: F1
  • Select Tasks: Run Task
  • Select Setup Open IoT SDK network
  • Enter the network namespace name OIStest
  • Choose command restart

Debug example:

  • Click Run and Debug from the primary side menu or press Ctrl+Shift+D
  • Select Debug Open IoT SDK example application from the drop down list
  • Click Start Debugging(green triangle) or press F5
  • Choose example name lock-app
  • Enter GDB target address 10.200.1.2
  • Enter network namespace OIStestns
  • Enter network interface OIStesttap
  • Choose example name lock-app

Use debug controls to debug the application.

Add new example

This chapter describes how to add a new Matter example based on Open IoT SDK platform.

In the description below we use the placeholder example_name as the name of the example to create. Replace it with the name of your example.

💡 Notes:

Remember to update the list of currently supported Matter examples at the top of this document.

Files structure

A new example should be put into examples/<example_name>/openiotsdk directory. It should contain:

  • application source files and headers in the main sub-directory
  • application CMakeLists.txt file
  • .gitignore file with with all sources to skip
  • README.md file with example description
  • additional directories with required configuration for used components. Use component_name-config pattern, e.g freertos-config

Target name

A new application target name should be created with chip-openiotsdk-<example_name>-example(_ns) pattern. The _ns suffix is required for TF-M applications.

Example:

set(APP_TARGET chip-openiotsdk-new-example-example_ns)

Example tools

Add a new example name to the list in the examples/platform/openiotsdk/supported_examples.txt file. After that the new example is available in all necessary tools such as helper script scripts/examples/openiotsdk_example.sh or VSCode tasks.

Example:

...
example_name
...

CI

To add a new example to the Matter CI edit the .github/workflows/examples-openiotsdk.yaml file and add the next step for openiotsdk job step that build this example.

Example:

...
- name: Build new-example example
    id: build_new_example
    run: |
        scripts/examples/openiotsdk_example.sh -b ${{ matrix.cryptoBackend }} new-example
        .environment/pigweed-venv/bin/python3 scripts/tools/memory/gh_sizes.py \
        openiotsdk release new-example \
        examples/new-example/openiotsdk/build/chip-openiotsdk-new-example-example.elf \
        /tmp/bloat_reports/
...