Combined sleep fixes (#3127) Multiple pico_sync/pico_time related fixes, comments and new tests for both bare-metal SDK, and RTOS integration * Add PICO_SYNC_RP2350_SPIN_LOCK_WORKAROUND define, and fix lock_internal_ functions to workaround issues with RP2350 and s/w spin locks * Add PICO_TIME_USE_SLEEP_NOTIFIER which, if unset, will use bare __sev() and __wfe() instead of lock_internal_ functions on a dummy spin_lock to handle notifications. This is simpler, avoids the RP2350 s/w spinlock issue, but would be inappropriate for use with an RTOS since __wfe() would not yield. This defaults to 1 if any of the lock_internal_ functions are overridden (which they aren't in the bare SDK) * Add LOCK_INTERNAL_SPIN_UNLOCK_WITH_NOTIFY_WAKES_ALL which defaults to 1 unless any of the lock_internal_ functions are overridden. The original contract in lock_core.h expected that a lock_internal_spin_unlock_with_notify() would wake all waiters, but the FreeRTOS ports already violate this constraint under certain race conditions. When the value is 0, the primitives such as a semaphore where multiple waiters might be awoken use the new lock_internal_spin_unlock_maybe_notify() macro to potentially wake up more waiters. * re-arm the ta_timer before processing cancellations (fix invariant violation and #2706) * Assert alarm_pool_destroy() is called on the pool's own core, and say so * fix wrong cleanup order in ta_disable_irq_handler * Distinguish pico_time_test/pico_time_test_sw, and rename pico_sync_test_hw to pico_sync_test * pico_sync_test: run every section on both cores * Add short_sleep_test * Add sync_interop_test for sdk and sdk/rtos interaction tests of pico_sync and pico_time * Add freertos_sync_alias_test. Note this looks for an issue with event bit aliasing which has not yet been reproduced - possibly due to luck in the FreeRTOS implementation, but kept in case the problem does arise due to future changes * Fix host implementation/tests Unrelated changes, included here: * let PICOTEST_CHECK take printf arguments * add PICO_PLATFORM_STRING define set by CMake build from PICO_PLATFORM Co-authored-by: Luke Wren <luke@raspberrypi.com>
The Raspberry Pi Pico SDK (henceforth the SDK) provides the headers, libraries and build system necessary to write programs for the RP-series microcontroller-based devices such as the Raspberry Pi Pico or Raspberry Pi Pico 2 in C, C++ or assembly language.
The SDK is designed to provide an API and programming environment that is familiar both to non-embedded C developers and embedded C developers alike. A single program runs on the device at a time and starts with a conventional main() method. Standard C/C++ libraries are supported along with C-level libraries/APIs for accessing all of the RP-series microcontroller's hardware including PIO (Programmable IO).
Additionally, the SDK provides higher level libraries for dealing with timers, synchronization, Wi-Fi and Bluetooth networking, USB and multicore programming. These libraries should be comprehensive enough that your application code rarely, if at all, needs to access hardware registers directly. However, if you do need or prefer to access the raw hardware registers, you will also find complete and fully-commented register definition headers in the SDK. There's no need to look up addresses in the datasheet.
The SDK can be used to build anything from simple applications, fully-fledged runtime environments such as MicroPython, to low level software such as the RP-series microcontroller's on-chip bootrom itself.
The design goal for entire SDK is to be simple but powerful.
Additional libraries/APIs that are not yet ready for inclusion in the SDK can be found in pico-extras.
See Getting Started with the Raspberry Pi Pico-Series for information on how to setup your hardware, IDE/environment and how to build and debug software for the Raspberry Pi Pico and other RP-series microcontroller based devices.
See Connecting to the Internet with Raspberry Pi Pico W to learn more about writing applications for your Raspberry Pi Pico W that connect to the internet.
See Raspberry Pi Pico-Series C/C++ SDK to learn more about programming using the SDK, to explore more advanced features, and for complete PDF-based API documentation.
See Online Raspberry Pi Pico SDK API docs for HTML-based API documentation.
See pico-examples for example code you can build.
The master branch of pico-sdk on GitHub contains the latest stable release of the SDK. If you need or want to test upcoming features, you can try the develop branch instead.
You can install the Raspberry Pi Pico Visual Studio Code extension in VS Code.
These instructions are extremely terse, and Linux-based only. For detailed steps, instructions for other platforms, and just in general, we recommend you see Raspberry Pi Pico-Series C/C++ SDK
Install CMake (at least version 3.13), python 3, a native compiler, and a GCC cross compiler
sudo apt install cmake python3 build-essential gcc-arm-none-eabi libnewlib-arm-none-eabi libstdc++-arm-none-eabi-newlib
Set up your project to point to use the Raspberry Pi Pico SDK
Either by cloning the SDK locally (most common) :
git clone this Raspberry Pi Pico SDK repository
Copy pico_sdk_import.cmake from the SDK into your project directory
Set PICO_SDK_PATH to the SDK location in your environment, or pass it (-DPICO_SDK_PATH=) to cmake later.
Setup a CMakeLists.txt like:
cmake_minimum_required(VERSION 3.13...3.27) # initialize the SDK based on PICO_SDK_PATH # note: this must happen before project() include(pico_sdk_import.cmake) project(my_project) # initialize the Raspberry Pi Pico SDK pico_sdk_init() # rest of your project
Or with the Raspberry Pi Pico SDK as a submodule :
Clone the SDK as a submodule called pico-sdk
Setup a CMakeLists.txt like:
cmake_minimum_required(VERSION 3.13...3.27) # initialize pico-sdk from submodule # note: this must happen before project() include(pico-sdk/pico_sdk_init.cmake) project(my_project) # initialize the Raspberry Pi Pico SDK pico_sdk_init() # rest of your project
Or with automatic download from GitHub :
Copy pico_sdk_import.cmake from the SDK into your project directory
Setup a CMakeLists.txt like:
cmake_minimum_required(VERSION 3.13) # initialize pico-sdk from GIT # (note this can come from environment, CMake cache etc) set(PICO_SDK_FETCH_FROM_GIT on) # pico_sdk_import.cmake is a single file copied from this SDK # note: this must happen before project() include(pico_sdk_import.cmake) project(my_project) # initialize the Raspberry Pi Pico SDK pico_sdk_init() # rest of your project
Or by cloning the SDK locally, but without copying pico_sdk_import.cmake:
git clone this Raspberry Pi Pico SDK repository
Setup a CMakeLists.txt like:
cmake_minimum_required(VERSION 3.13) # initialize the SDK directly include(/path/to/pico-sdk/pico_sdk_init.cmake) project(my_project) # initialize the Raspberry Pi Pico SDK pico_sdk_init() # rest of your project
Write your code (see pico-examples or the Raspberry Pi Pico-Series C/C++ SDK documentation for more information)
About the simplest you can do is a single source file (e.g. hello_world.c)
#include <stdio.h> #include "pico/stdlib.h" int main() { stdio_init_all(); printf("Hello, world!\n"); return 0; }
And add the following to your CMakeLists.txt:
add_executable(hello_world hello_world.c ) # Add pico_stdlib library which aggregates commonly used features target_link_libraries(hello_world pico_stdlib) # create map/bin/hex/uf2 file in addition to ELF. pico_add_extra_outputs(hello_world)
Note this example uses the default UART for stdout; if you want to use the default USB see the hello-usb example.
Setup a CMake build directory. For example, if not using an IDE:
$ cmake -S . -B build
The cmake -S flag indicates the source directory, and the -B flag tells cmake the name of the output-directory to create. This doesn't have to be named “build”, you can call it whatever you want.
When building for a board other than the Raspberry Pi Pico, you should pass -DPICO_BOARD=board_name to the cmake command above, e.g. cmake -S . -B build -DPICO_BOARD=pico2 or cmake -S . -B build -DPICO_BOARD=pico_w to configure the SDK and build options accordingly for that particular board.
Specifying PICO_BOARD=<boardname> sets up various compiler defines (e.g. default pin numbers for UART and other hardware) and in certain cases also enables the use of additional libraries (e.g. wireless support when building for PICO_BOARD=pico_w) which cannot be built without a board which provides the requisite hardware functionality.
For a list of boards defined in the SDK itself, look in this directory which has a header for each named board.
Make your target from the build directory you created.
$ cmake --build build --target hello_world
The directory-name supplied to the
--buildflag needs to match the directory-name that was passed to the-Bflag in the earlier cmake command.
You now have hello_world.elf to load via a debugger, or hello_world.uf2 that can be installed and run on your Raspberry Pi Pico-series device via drag and drop.
See Raspberry Pi Pico-series C/C++ SDK for information on setting up a build environment for RISC-V on RP2350.
The pico-sdk-tools repository contains some prebuilt versions of the RISC-V compiler.
You can use these to get a working RISC-V compiler on Raspberry Pi OS for example.
wget https://github.com/raspberrypi/pico-sdk-tools/releases/download/v2.0.0-5/riscv-toolchain-14-aarch64-lin.tar.gz sudo mkdir -p /opt/riscv/riscv-toolchain-14 sudo chown $USER /opt/riscv/riscv-toolchain-14 tar xvf riscv-toolchain-14-aarch64-lin.tar.gz -C /opt/riscv/riscv-toolchain-14
To use the RISC-V compiler to build code you need to set a couple of environment variables and run cmake from fresh.
export PICO_TOOLCHAIN_PATH=/opt/riscv/riscv-toolchain-14/ export PICO_PLATFORM=rp2350-riscv