Pico-series microcontroller Command Line Setup

This script gives you an easy way to setup your Raspberry Pi to be able to build and run programs on your Pico-series microcontroller from the command line. Compatibility with any systems not running Raspberry Pi OS or Raspberry Pi OS Lite is not guaranteed or maintained.

To download & run this script, you can use the following commands:

wget https://raw.githubusercontent.com/raspberrypi/pico-setup/master/pico_setup.sh
chmod +x pico_setup.sh
./pico_setup.sh

For manual command line setup instructions for other operating systems, see Basic Setup on Other Operating Systems

If you want to use a GUI instead of the command line, then see the pico-vscode extension instead of this script - this supports 64-bit Windows, MacOS, Raspberry Pi OS, and most common Linux distros. This is documented in the Getting Started Guide.

Compiling and running an example

After running the setup script, you'll want to run an example to check everything's working. First go into pico-examples:

cd pico/pico-examples

Depending on the board you're using (eg pico2), replace build_pico with the relevant build directory (eg build_pico2) in the following commands.

If you're not using one of the default boards (pico, pico_w, pico2, or pico2_w), you'll need to create a new build directory for your board - you can do this with this command (replace both instances of $board with the board you are using):

cmake -S . -B build_$board -GNinja -DPICO_BOARD=$board -DCMAKE_BUILD_TYPE=Debug

To build the blink example, run the following command:

cmake --build build_pico --target blink

This builds the specified target blink in the build folder build_pico - it will probably display no work to do because blink was built earlier by pico_setup.sh

Then to run it, attach a Pico-series microcontroller in BOOTSEL mode, and run:

picotool load build_pico/blink/blink.uf2 -vx

This loads the file into Flash on the board, then verifies it was loaded correctly and reboots the board to execute it

You should now have a blinking LED on your board! For more info on the picotool command which is used to load and query binaries on the device, see its README

Console Input/Output

To view console output, you can either connect the UART output to a Debug Probe (or similar) and use stdio_uart (see the hello_serial example), or you can use stdio_usb (see the hello_usb example).

First, build & run the example for your stdio choice on your Pico-series microcontroller with the same commands as before:

cmake --build build_pico --target hello_serial
picotool load build_pico/hello_world/serial/hello_serial.uf2 -vx

Then attach minicom to view the output:

minicom -b 115200 -D /dev/ttyACM0

The port number may be different, so also try /dev/ttyACM1 etc - and on other OSes may be entirely different (eg /dev/tty.usbmodem0001 on MacOS)

To exit minicom, type Ctrl+A then X

Debugging with OpenOCD and GDB

To debug programs on the Pico-series microcontroller, you first need to attach a debugger such as the Debug Probe. Once that's done, you can attach OpenOCD to your Pico-series microcontroller with this command (replace rp2040.cfg with rp2350.cfg, if using an RP2350-based board like a Pico 2):

openocd -f interface/cmsis-dap.cfg -f target/rp2040.cfg -c "adapter speed 5000"

In a separate window, you can run GDB and connect to that OpenOCD server:

gdb -ex "target extended-remote localhost:3333"

Then in GDB, run the following:

file build_pico/blink/blink.elf
monitor reset init
load
continue

To exit GDB, use Ctrl+D twice. This will leave your Pico-series microcontroller in the halted state, so you will need to unplug and replug it to get it running again. To leave the device running, you can use:

monitor reset run

before exiting to leave the Pico running the code.

Useful GDB Commands

To configure the GDB layout, the following commands can be useful:

• layout src - displays the source code
• layout split - displays the source code and the assembly instructions
• layout regs - displays the current register contents

To step through code, you can use:

• step or s - step to the next line of code, and into any functions
• next or n - step to the next line of code, without stepping into functions
• finish or fin - step out of the current function

To step through assembly instructions, you can use:

• stepi or si - step to the next assembly instruction, and into any functions
• nexti or ni - step to the next assembly instruction, without stepping into functions

While stepping, you can just press enter again to repeat the previous command

To set breakpoints, use the break or b command plus the location of the breakpoint. The location can be:

• A function name - break main
• A line of code - break 12
• Either of those in a specific file - break blink.c:48, break blink.c:main
• A specific memory address - break *0x10000ff2

For more details on debugging with GDB, see the GDB docs

Multiple Terminals

When debugging or viewing serial output, you might want multiple programs open in different terminals, as they all need to run at the same time.

On Raspberry Pi OS Lite, you can switch between different terminals with Alt+F1,F2,F3,F4 etc.

Alternatively you can use something like screen or tmux to allow you to open new terminals and detach from them - for example using screen:

• screen -S minicom - open a new terminal called minicom
• Ctrl+A then D - detach from the current terminal
• screen -r minicom - re-attach to an existing terminal called minicom
• screen -ls - list existing terminals

For more details on screen, see the screen docs

Basic Setup on Other Operating Systems

Prerequisites

Windows

If you're on Windows, it is strongly recommended to use WSL2 and then follow the Linux instructions inside that. You should also install usbipd to access USB devices inside WSL2 (see the docs there for instructions).

If you're not using WSL2, then you'll need to install the following tools:

• Arm GNU Toolchain
  • Pick Windows -> AArch32 bare-metal target (arm-none-eabi) -> the .exe file
• CMake
• Microsoft Visual Studio
  • When running the installer, select Desktop Development with C++

Then follow the manual setup instructions. You will need to run all commands from the "Developer PowerShell for VS ..." terminal, not your usual terminal. For the first cmake configuration in each project (eg cmake ..), you will need to replace -GNinja with -G "NMake Makefiles", and anything with sudo needs to be run as administrator. Also, you should use as short a path as possible due to path length limits on Windows.

MacOS

Install Homebrew and run these commands:

xcode-select --install
brew install cmake ninja
brew install --cask gcc-arm-embedded

Then follow the manual setup instructions.

Linux

If you have apt, then running the pico_setup.sh script should hopefully work for you, and you won't need to do any manual setup.

If it doesn't work, or you don't have apt, then you can manually install the GIT_DEPS (git & git-lfs) and SDK_DEPS (cmake, gcc-arm-none-eabi, gcc, g++ & ninja-build) from the script. You may also need to install a cross-compiled newlib if it isn't included in gcc-arm-none-eabi, such as libnewlib-arm-none-eabi & libstdc++-arm-none-eabi-newlib. Once those are installed, follow the manual setup instructions.

Setup SDK & Picotool

SDK

Run this from the path you want to store the SDK:

git clone https://github.com/raspberrypi/pico-sdk.git
git -C pico-sdk submodule update --init

Then set PICO_SDK_PATH environment variable to that path - on MacOS or Linux, just add the following to your .zshrc (MacOS) or .bashrc file:

export PICO_SDK_PATH=/path/to/pico-sdk

then reload your terminal window.

On Windows, from an administrator PowerShell run:

[System.Environment]::SetEnvironmentVariable('PICO_SDK_PATH','/path/to/pico-sdk', 'Machine')

Picotool

Run this from the path you want to store picotool:

git clone https://github.com/raspberrypi/picotool.git

Then install libusb

• On Windows, download and extract libUSB from here https://libusb.info/ (hover over Downloads, and click Latest Windows Binaries), and set LIBUSB_ROOT environment variable to the extracted directory.
• On MacOS brew install libusb
• On Linux sudo apt install libusb-1.0-0-dev

Then build and install picotool using these commands:

cd picotool
cmake -S . -B build
cmake --build build
sudo cmake --install .

To use picotool without sudo on Linux, you'll also need to install the picotool udev rules from the picotool/udev folder.

For more details on building & installing picotool, see its README

Test it's working with pico-examples

Clone pico-examples:

git clone https://github.com/raspberrypi/pico-examples.git
cd pico-examples

Build them all, replacing all occurences of $board with the pico board you are using:

cmake -S . -B build_$board -GNinja -DPICO_BOARD=$board -DCMAKE_BUILD_TYPE=Debug
cmake --build build_$board

The cmake -S flag indicates the source directory, and the -B flag tells cmake the name of the output-directory to create, which is then supplied to the --build flag

Put your board in BOOTSEL mode and use picotool to load the blink example:

picotool load build_$board/blink/blink.uf2 -vx

You should now have a blinking LED on your board.