A chip-shell project on the Texas Instruments CC13XX_26XX family of Wireless MCUs.
Some initial setup is necessary for preparing the build environment. This section will need to be done when migrating to new versions of the SDK. This guide assumes that the environment is linux based, and recommends Ubuntu 20.04.
Download and install SysConfig. This can be done simply with the following commands.
$ cd ~ $ wget https://dr-download.ti.com/software-development/ide-configuration-compiler-or-debugger/MD-nsUM6f7Vvb/1.18.1.3343/sysconfig-1.18.1_3343-setup.run $ chmod +x sysconfig-1.18.1_3343-setup.run $ ./sysconfig-1.18.1_3343-setup.run
Run the bootstrap script to setup the build environment.
Note, a recursive submodule checkout is required to utilize TI's Openthread reference commit.
$ cd ~/connectedhomeip $ source ./scripts/bootstrap.sh $ ./scripts/checkout_submodules.py --shallow --platform cc13xx_26xx --recursive
It is necessary to activate the environment in every new shell. Then run GN and Ninja to build the executable.
Activate the build environment with the repository activate script.
$ cd ~/connectedhomeip $ source ./scripts/activate.sh
Run the build to produce a default executable. By default on Linux both the TI SimpleLink SDK and Sysconfig are located in a ti folder in the user's home directory, and you must provide the absolute path to them. For example /home/username/ti/sysconfig_1.18.1. On Windows the default directory is C:\ti. Take note of this install path, as it will be used in the next step.
$ cd ~/connectedhomeip/examples/shell/cc13x4_26x4 $ gn gen out/debug --args="ti_sysconfig_root=\"$HOME/ti/sysconfig_1.18.1" $ ninja -C out/debug
If you would like to define arguments on the command line you may add them to the GN call.
gn gen out/debug --args="ti_sysconfig_root=\"$HOME/ti/sysconfig_1.18.1\" target_defines=[\"CC13X4_26X4_ATTESTATION_CREDENTIALS=1\"] chip_generate_link_map_file=true"
Loading the built image onto a LaunchPad is supported through two methods; Uniflash and Code Composer Studio (CCS). UniFlash can be used to load the image. Code Composer Studio can be used to load the image and debug the source code.
Programming with CCS will allow for a full debug environment within the IDE. This is accomplished by creating a target connection to the XDS110 debugger and starting a project-less debug session. The CCS IDE will attempt to find the source files on the local machine based on the debug information embedded within the ELF. CCS may prompt you to find the source code if the image was built on another machine or the source code is located in a different location than is recorded within the ELF.
Download and install Code Composer Studio.
First open CCS and create a new workspace.
Create a target connection (sometimes called the CCXML) for your target SoC and debugger as described in the Manual Method section of the CCS User's Guide.
Next initiate a project-less debug session as described in the Manual Launch section of the CCS User's Guide.
CCS should switch to the debug view described in the After Launch section of the User‘s Guide. The SoC core will likely be disconnected and symbols will not be loaded. Connect to the core as described in the Debug View section of the User’s Guide. Once the core is connected, use the Load button on the toolbar to load the ELF image.
Note that the default configuration of the CCXML uses 2-wire cJTAG instead of the full 4-wire JTAG connection to match the default jumper configuration of the LaunchPad.
Uniflash is Texas Instrument's uniform programming tool for embedded processors. This will allow you to erase, flash, and inspect the SoC without setting up a debugging environment.
Download and install UniFlash.
First open UniFlash. Debug probes connected to the computer will usually be displayed under the Detected Devices due to the automatic device detection feature. If your device does not show up in this view it my be disconnected, or you may have to create a New Configuration. If you already have a CCXML for your SoC and debug connection you can use that in the section at the bottom. Once your device is selected, click the Start button within the section to launch the session.
Select the ELF image to load on the device with the Browse button. This file is placed in the out/debug folder by this guide and ends with the *.out file extension. For OTA enabled applications, the standalone image will instead end with the *-mcuboot.hex file extension. This this is a combined image with application and and MCUBoot included.
Finally click the Load Image button to load the executable image onto the device. You should be able to see the log output over the XDS110 User UART.
Note that programming the device through JTAG sets the Halt-in-Boot flag and may cause issues when performing a software reset. This flag can be reset by power-cycling the LaunchPad.
By default the log output will be sent to the Application/User UART. Open a terminal emulator to that port to see the output with the following options:
| Parameter | Value |
|---|---|
| Speed (baud) | 115200 |
| Data bits | 8 |
| Stop bits | 1 |
| Parity | None |
| Flow control | None |
Once a device has been flashed with this example, it can now join and operate in an existing Matter network. The following sections assume that a Matter network is already active, and has at least one OpenThread Border Router.
For insight into what other components are needed to run this example, please refer to our Matter Getting Started Guide.
For help with the shell itself, refer to the shell example README.
For technical support, please consider creating a post on TI's E2E forum. Additionally, we welcome any feedback.