WLoop OpenBLT

Waterloop's port(s) of OpenBLT, to be used on our G6 pod.

Project File Structure

OpenBLT is included as a submodule in this repo. Our ports are included in the root directory of this repository (one directory per device).

Inside of each port, there will be a "Boot" and a "TestProgram" directory. The "Boot" directory contains the source code of the OpenBLT port, and the "TestProgram" directory contains a program to test the OpenBLT port on.

Building

First and foremost, the submodules must be pulled:

git submodule update --init --recursive

Each "main" subdirectory has its own makefile. For example, lets build OpenBLT for an STM32G431KBT Nucleo Board:

cd ./g431_nucleo/Boot
make

Building the test program is done similarly:

cd ./g431_nucleo/TestProgram
make

To build the host utility BootCommander, which is used to actually provide the FW update via OpenBLT, you may run the following script:

./utils/build_boot_commander.sh

BootCommander has some dependencies, which will first need to be installed:

sudo apt install cmake build-essential libusb-1.0-0 libusb-1.0-0-dev

Flashing Firmware Using BootCommander and OpenBLT

For this section, we will be using the ./g431_nucleo as an example.

# Build BootCommander (if you haven't already)
./utils/build_boot_commander.sh

# Build OpenBLT for the G431 Nucleo
cd ./g431_nucleo/Boot
make

# Flash OpenBLT onto the device...
make flash

# Build the test program...
cd ../TestProgram
make

# Now, flash the test program with BootCommander (over UART)
cd ../../
./utils/BootCommander \
    -s=xcp \
    -t=xcp_rs232 \
    -d=/dev/serial/by-id/usb-STMicroelectronics_STLINK-V3_002900233331511034333834-if02 \
    ./g431_nucleo/TestProgram/build/openblt_test_program.srec

# NOTE: the -d flag will be different for everyone, it is the path to the UART <==> USB
#       device that you have connected to the device (for example, /dev/ttyUSB0)

Developing for this Repo

Intellisense and Autocomplete

Intellisense and autocomplete are available through the clangd language server. It is supported on most modern editors.

If you use VSCode, there exists an extension for it.

clangd is already installed on your vagrant VMs. It is also easily installed through apt and brew if you like.

clangd uses the auto-generated file compile_commands.json to provide intellisense and autocomplete. This file is gitignored in this repository, so you will need to generate it yourself if you want intellisense. This can be easily done using bear, which is pre-installed on your vagrant VMs.

bear -- make