Welcome to Sming Framework
Let’s do smart things!!!
Sming is an asynchronous embedded C/C++ framework with superb performance and multiple network features. Sming is open source, modular and supports multiple architectures including ESP8266 and ESP32.
Summary
Superb performance and memory usage (Sming compiles to native firmware!)
Fast and user friendly development
Integrated host emulator to assist with developing, testing and debugging libraries and applications on a PC before uploading them to an actual microcontroller. Try it out online here.
Built-in powerful wireless modules
Compatible with standard Libraries - use popular hardware in few lines of code
Simple yet powerful hardware API wrappers
Standard Arduino-style wrappers for simplicity and flexibility
HardwareSPI library provides advanced asynchronous SPI device master capability including use of ESP8266 ‘overlapped’ mode for reduced pin usage and dual/quad I/O support.
Sming Graphics Library implements asynchronous display control model with low memory usage, transparency (alpha-blending) and flexible scripted resource management.
Modular C++ installable file system interface
New FWFS read-only filesystem can be used as base file system, with read/write filesystems mounted in sub-directories
Integrated metadata support (file times, security descriptors, user metadata)
Access to Host (linux / Windows / MacOS) filing system via same API
Integrated streaming archival system to support backups or file system compaction operations
Powerful asynchronous (async) network stack including:
Async TCP and UDP stack based on LWIP.
With clients supporting: HTTP, MQTT, WebSockets and SMTP.
And servers for: DNS, FTP, HTTP(+ WebSockets), Telnet.
With SSL support for all network clients and servers. Based on axTLS and BearSSL.
Over-The-Air (OTA) firmware upgrades via HTTP(S) and MQTT(S).
Universal Plug and Play (UPnP) framework supports both standard and custom devices with full ControlPoint support.
Generates C++ code from standard UPnP XML schema.
Companion SSDP library independently supports discovery on local network.
GoogleCast library allows control of ChromeCast devices or smart TVs supporting the GoogleCast protocol.
Hue Emulator provides simple way to implement devices compatible with Amazon Alexa.
ESP8266 features
Integrated boot loader rBoot with support for 1MB ROMs, OTA firmware updating and ROM switching.
Crash handlers for analyzing/handling system restarts due to fatal errors or WDT resets.
PWM support based on Stefan Bruens PWM.
Optional custom heap allocation based on Umm Malloc.
Based on Espressif NONOS SDK Version 3.
ESP32 features
Based on ESP IDF SDK
Getting Started
Development System Installation
Choose your preferred development environment for how to install the needed development software and toolchain(s):
Linux Installation
Quick Install
Debian (Ubuntu) and Fedora systems can use the scripted installer.
Prepare installation directory
Let’s use /opt as the main directory for tools and Sming.
Regular users may not have access to /opt, so do this:
sudo chown $USER:$USER /opt
(alternatively, use a different directory).
Install GIT
Debian:
sudo apt-get install -y git
Fedora:
dnf install -y git
Fetch the Sming repository
git clone https://github.com/SmingHub/Sming /opt/sming
Run the installer
source /opt/sming/Tools/install.sh all
If you want to save disk space then you can select which tools to install. Get a list of available options like this:
/opt/sming/Tools/install.sh
Install locations can be customised by setting environment variables before running the install. Certain variables should also be set globally. See Configuration for details.
If you want to use the stable (release) branch:
cd /opt/sming
git checkout master
git pull
Build a ‘Basic Blink’ example
Change to the application directory and build:
cd $SMING_HOME/../samples/Basic_Blink
make
Flash to your device (using default serial port):
cd $SMING_HOME/../samples/Basic_Blink
make flash
Next steps
Proceed to Configuration.
Mac-OS Installation
Pre-requisites
(You might already have it)
Xcode command line tools
xcode-select --install
Homebrew
ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
Eclipse
(Optional):
brew install Caskroom/cask/eclipse-cpp
Build tools
Required for the makefile build system:
brew install binutils coreutils automake wget gawk libtool gettext gperf grep
export PATH="/usr/local/opt/gnu-sed/libexec/gnubin:$PATH"
Install gnu-sed and make sure that it is the default sed command line application:
brew install gnu-sed --with-default-names
If you have already installed gnu-sed but it is not the default one,
then make sure to uninstall it and install it again with the correct options:
brew uninstall gnu-sed
brew install gnu-sed --with-default-names
ESP8266 Toolchain
We pull this in from the SmingTools repository:
cd ~/
export ESP_HOME=/opt/esp-quick-toolchain
curl -LO https://github.com/SmingHub/SmingTools/releases/download/1.0/x86_64-apple-darwin14.xtensa-lx106-elf-e6a192b.201211.tar.gz
sudo mkdir -p $ESP_HOME
sudo tar -zxf x86_64-apple-darwin14.xtensa-lx106-elf-e6a192b.201211.tar.gz -C $ESP_HOME
sudo chmod -R 775 $ESP_HOME
You can also build it yourself with Homebrew or with MacPorts.
Get Sming Core
Clone from the Sming repository:
cd <your-favourite-development-folder>/
git clone https://github.com/SmingHub/Sming.git
cd Sming
Warning
Do NOT use the –recursive option for the command above. Our build mechanism will take care to get the third-party sources and patch them, if needed.
You will get a copy of our develop branch which is intended for developers. It is the one where all new cool (unstable) features are landing.
If you want to use our stable branch then use the master branch:
git checkout origin/master
Finally, set the SMING_HOME environment variable to point to <your-favourite-development-folder>/Sming/Sming:
export SMING_HOME=`pwd`
Environment Variables
Open with a text editor the .profile file in your home directory, and add these lines:
export ESP_HOME=/opt/esp-quick-toolchain
export SMING_HOME=<your-favourite-development-folder>/Sming/Sming
Make sure to replace <your-favourite-development-folder> in the
command above with the actual directory on your local disk.
(Optional step)
(used by Make and Eclipse - make sure to quit Eclipse first)
If you installed Eclipse manually, substitute
/opt/homebrew-cask/Caskroom/eclipse-cpp/4.5.1/Eclipse.app to
/Applications/Eclipse/eclipse.app:
sudo /usr/libexec/PlistBuddy -c "Add :LSEnvironment:ESP_HOME string '/opt/esp-open-sdk'" /opt/homebrew-cask/Caskroom/eclipse-cpp/4.5.1/Eclipse.app/Contents/Info.plist
sudo /usr/libexec/PlistBuddy -c "Add :LSEnvironment:SMING_HOME string '/opt/sming/Sming'" /opt/homebrew-cask/Caskroom/eclipse-cpp/4.5.1/Eclipse.app/Contents/Info.plist
sudo /System/Library/Frameworks/CoreServices.framework/Frameworks/LaunchServices.framework/Support/lsregister -v -f /opt/homebrew-cask/Caskroom/eclipse-cpp/4.5.1/Eclipse.app
Using Eclipse on Mac
If you are using Eclipse to build the samples you need to make sure the path is set correctly for the make process.
Your project must also be configured to use the correct serial port for your ESP8266. You can change it like this:
make COM_PORT=/dev/tty.usbserial
Next, ensure that you can build the Basic Blink from a terminal window:
cd $SMING_HOME/../samples/Basic_Blink
make
This will also build the required framework components, so may take a few minutes.
If this works without errors then type echo $PATH and copy the resulting path
to the clipboard.
Now fire up Eclipse and go to
Eclipse ==> Preferences ==> C/C++ ==> Build ==> Environment
and add a new variable PATH. Paste in the path saved from the terminal
session above. You can also add SMING_HOME and ESP_HOME variables here
the same way as you set in the export commands above which will then be
set for all the projects.
The standard make files use miniterm.py to provide a serial Terminal for
debugging the ESP8266. Miniterm does not work inside Eclipse so you
should disable it like this:
make KILL_TERM= TERMINAL=
This will prevent Eclipse from trying to launch miniterm and throwing an error about Inappropriate ioctl for device.
You can use the built in terminal in Eclipse Oxygen by adding it using
Window ==> Show View ==> Terminal
then setting terminal type to Serial and setting the port to the port
the ESP8266 is connected to. Remember to disconnect before tying to
re-flash the device though.
Compile Sming Examples
See Sample Projects for a list of all examples provided with Sming:
cd $SMING_HOME/../samples/
If you want to test some of the examples, try this:
cd $SMING_HOME/../samples
cd Basic_Blink
make
# The command below will upload the sample code to your ESP8266 device
make flash
Next steps
Proceed to Configuration.
Windows Installation
This page describes how to install the required tools and obtain the current release version of Sming using the Chocolatey package manager.
See also:
Manual Windows Installation
MinGW
Code built for the ESP8266 uses a separate Espressif compiler, but Components such as SPIFFS require additional tools which are built as Windows executable applications.
MinGW provides a (mostly) POSIX-compliant development environment for Windows, including GNU Make and various other command-line tools.
Note
There are two versions of MinGW.
MinGW is the original, and the version recommended for use with Sming.
MinGW-w64 was forked from MinGW in 2007 in order to provide support for 64 bits and new APIs. (Which we don’t need!)
To find out if you already have GCC on your system, type:
where gcc
If it shows C:\MinGW\bin\gcc.exe then you have a standard MinGW installation. Check the gcc version:
gcc --version
The current version is 8.2.0. You can upgrade by renaming or removing your existing installation then following these instructions.
Fast install
Download MinGW.7z from the SmingTools repository.
Unzip to default location:
7z -oC:\ x MinGW.7z
Note
You can obtain 7-zip from https://www.7-zip.org/.
Update
PATHenvironment variable:SETX PATH "C:\mingw\bin;C:\mingw\msys\1.0\bin;%PATH%"
Make sure it is in this exact order. If you have Cygwin installed make sure the above entries appear first.
You will need to restart your command prompt (and Eclipse, if already running) for these changes to take effect.
Alternative install
To install from the original MinGW source repository:
Get the MingGW Setup and run it. This will create the
C:\MinGWdirectory with minimal content.Set PATH environment variable as above.
Install required MinGW packages:
mingw-get install mingw32-base-bin mingw-developer-toolkit-bin mingw32-gcc-g++-bin mingw32-pthreads-w32-dev mingw32-libmingwex
Note that you can upgrade packages using:
mingw-get update mingw-get upgrade
However this will not upgrade a 6.3.0 installation to 8.2.0.
Install ESP8266 Toolchain
Download toolchain ESP Quick Toolchain.
Unzip to default location:
7z -oC:\tools\esp-quick-toolchain x x86_64-w64-mingw32.xtensa-lx106-elf-e6a192b.201211.zip
Set
ESP_HOMEenvironment variable:SETX ESP_HOME C:\tools\esp-quick-toolchain
Note
There is NO trailing slash on the path!
If you want to set environment variables system-wide, append /M to the command. You’ll need to do this from an administrative command prompt.
Install Python
Get the current version from https://www.python.org/.
By default, python gets installed here:
C:\Users\xxx\AppData\Local\Programs\Python\Python38\python.exe
Wherever it ends up will either need to be in the path:
setx PATH "C:\Users\xxx\AppData\Local\Programs\Python\Python38;%PATH%"
or located using the PYTHON environment variable:
setx PYTHON "C:\Users\xxx\AppData\Local\Programs\Python\Python38"
Important
The PYTHON variable may not contain spaces. This is a MinGW restriction.
Install GIT
This is required to fetch and update Sming code from its repository.
Install command-line GIT client.
These steps are optional, but highly recommended:
Install Graphical client Git Extensions.
Create an account at https://github.com. This will allow you to participate in discussions, raise issues and, if you like, Contributing to the framework!
Download Sming
You can put Sming anywhere convenient, provided there are no spaces in the path! For example, C:\tools\sming:
mkdir C:\tools cd /d C:\tools
To obtain the latest develop code with all the latest features and fixes:
git clone https://github.com/SmingHub/Sming
To obtain the latest release:
git clone https://github.com/SmingHub/Sming --branch master
Set
SMING_HOMEenvironment variable:SETX SMING_HOME C:\tools\Sming\Sming
Note: there is NO trailing slash on the path!
Note
Whilst Windows filenames are not (by default) case-sensitive, the compiler tools are.
Please take care to type paths exactly as shown.
At this stage you should be able to build a sample:
cd samples\Basic_Blink
make -j
If you want to try out the Host emulator, do this:
make -j SMING_ARCH=Host
For build options:
make help
Install Eclipse IDE
Whilst building and configuring your application is generally easier and faster using the command prompt, developing and debugging code is greatly simplified using an Integrated Development Environment (IDE).
Install Java Runtime Environment.
Install Eclipse IDE for C++ Developers.
Start Eclipse IDE. When prompted, enter
C:\tools\smingas the workspace path.Select File -> Import -> General -> Existing Project into Workspace. In the line Select root directory, select the directory
C:\tools\sming\Smingand import everything.Go have a cup of coffee while Eclipse scans all the source code. It can take a while!
To build a project, right-click and select Build project. Alternatively, select the project and press F9.
Eclipse IDE variables
The only variable you should need to set within Eclipse is SMING_HOME.
You can set this within the Eclipse IDE via Window > Preferences -> C/C++ > Build > Environment.
If you set this via global environment variable before starting Eclipse then this step is not necessary.
Note
Variables set within the IDE won’t be accessible in other Eclipse sessions or the command prompt.
All other configuration should be done either in your project’s component.mk file or via command line.
For example, to switch to a Host emulator build, do this:
make SMING_ARCH=Host list-config
This also displays the current configuration settings. Whether you build from command line or Eclipse, the same settings will be used.
Next steps
Proceed to Configuration.
WSL
Building under Windows is generally slower than in Linux. This is because the current build system requires a Posix emulation layer (MinGW). However, it does offer the simplest way to use Sming on a Windows PC and does not affect the quality or functionality of your applications.
There are situations where it is highly desirable to build Sming in a Linux environment:
Making use of linux-only development tools, e.g. valgrind (dynamic bug detection system)
Integration building/testing prior to submitting a PR to the Sming repository
Need/want faster builds
Whilst a Windows computer can be configured to dual-boot with Linux, this is generally inconvenient for day-to-day use. A better solution is to run Linux inside a virtual machine environment such as VirtualBox, VmWare or Hyper-V.
Note that Docker is not a virtual environment but can in fact be run inside a virtual machine to take advantage of the process isolation and security benefits.
Windows Subsystem for Linux
https://docs.microsoft.com/en-us/windows/wsl/
“The Windows Subsystem for Linux lets developers run a GNU/Linux environment – including most command-line tools, utilities, and applications – directly on Windows, unmodified, without the overhead of a traditional virtual machine or dual-boot setup.”
There are currently two versions of WSL: this documentation relates to WSL2.
Note
WSL2 uses Hyper-V so may conflict with other virtual machines you may be using.
Installing WSL
See instructions here https://docs.microsoft.com/en-us/windows/wsl/install.
Install an up-to-date Linux distribution from the Microsoft store, currently Ubuntu-20.04.
Note
You may encounter an error message similar to this during installation:
WslRegisterDistribution failed with error: 0x80370102
Error: 0x80370102 The virtual machine could not be started because a required feature is not installed.
One thing not mentioned in the instructions is to check that the hypervisor is set to auto-start at system boot. This is the default but for various reasons it can get disabled.
To check, type:
bcdedit
At an administrative command prompt. Under the Windows Boot Loader entry you should see an entry like this:
hypervisorlaunchtype Auto
If it’s missing or set to another value (e.g. off) then change it as follows:
bcdedit /set {current} hypervisorlaunchtype auto
After a system reboot you should be able to continue with the installation.
Installing Sming
Open a WSL command prompt and follow the instructions in Linux Installation.
Please note:
A native Windows python3 distribution is required to enable access to serial ports.
Ensure that python is available in the system path for both WSL2 and Windows.
Do not set the
PYTHONenvironment variable.
This will ensure that the build system can run python scripts either in WSL2 or in Windows as necessary.
Flashing devices
WSL2 does not currently support access to USB serial devices, so the Sming build system runs
the appropriate application directly under Windows using powershell.exe.
Therefore, use the normal Windows COM port name rather than the linux ones (such as /dev/ttyUSB0):
make flash COM_PORT=COM4
Serial debugging
Again, as we have no direct access to USB COM ports a workaround is required.
A small python application can be run on Windows to act as a simple bridge between the serial port and a TCP port.
See Tools/tcp_serial_redirect.py - run without arguments to see available options.
You can start the server like this:
make tcp-serial-redirect
A new console will be created (minimised) showing something like this:
--- TCP/IP to Serial redirect on COM4 115200,8,N,1 ---
--- type Ctrl-C / BREAK to quit
Waiting for connection on 192.168.1.101:7780...
This uses the current COM_PORT and COM_SPEED_SERIAL settings.
Now we can start the debugger:
make gdb COM_PORT_GDB=192.168.1.101:7780
Valgrind
You may get an error running make valgrind advising that libc6-dbg:i386 be installed. Here’s how:
sudo dpkg --add-architecture i386
sudo apt-get update
sudo apt-get install libc6-dbg:i386
Quick Install
Open an administrative cmd.exe command prompt and paste the text from the box below and press enter:
curl -LO https://raw.githubusercontent.com/SmingHub/Sming/develop/Tools/choco-install.cmd && choco-install.cmd
At the moment the Esp32 toolchain is not installed by default. If you want to install it run the following command:
choco install -y sming.esp32
Important
After installation, please close the administrative command prompt and open a new, regular command shell.
This ensures that environment variables are set correctly.
It is also inadvisable to continue running with elevated privileges.
If you followed and executed carefully the steps above Sming should be installed and configured. You can scroll down to Build Basic_Blink to check the installation.
Optional step: Switch to stable version
The installer uses the latest develop branch. This one contains great new features and bugfixes but can be unstable at times.
Switching to our stable release will guarantee you that Sming’s code will not change so often.
On the downside you will have to wait for all new shiny features and bugfixes.
If you really want to use the latest stable release you can type the command below:
cd %SMING_HOME%
git checkout master
git pull
Optional step: Re-installation
In case something is broken, this will perform a forced re-install of all packages:
rmdir /s /q c:\tools\sming
choco install sming -y -f -x
Packages
You can find the installer sources at https://github.com/slaff/chocolatey-packages. Packages are as follows:
- git
GIT CLI client.
Please configure after installation to leave line-endings intact or else patching will fail:
git config --global core.autocrlf input
- python
Python version 3.
- cmake
-
Required to build some Components, also for Host mode.
- mingw
MinGW 32-bit.
The installer updates the system
PATHbut please check by running:where make.exe
The output should show only one result:
"C:\MinGW\msys\1.0\bin\make.exe"- esp8266-eqt
-
Sets a system-wide
ESP_HOMEvariable. - sming.esp32
Sming-compatible version of ESP-IDF and tools.
- sming.core
Latest stable version of Sming.
Sets a system-wide
SMING_HOMEenvironment variable.
Note that setting SMING_HOME and ESP_HOME as system-wide variables means they do not need to be set every time a command prompt is opened, and will be seen by eclipse without any further configuration.
Build Basic_Blink
To check the installation, open a command prompt and type these commands:
cd %SMING_HOME%\..\samples\Basic_Blink
make
The project should build without error.
Next steps
Proceed to Configuration.
Docker
Docker is a useful tool if you want to experiment with Sming in an isolated environment.
If you’re unfamiliar with Docker, you can find a good overview in the article, What is a Container?.
This page shows how to create docker containers with all the necessary tools to build Sming applications.
Install Docker
Visit the official Docker Installation Page and follow the instructions tailored for your operating system.
Install docker-compose
Docker Compose makes dealing with the orchestration processes of Docker containers (such as starting up, shutting down, and setting up intra-container linking and volumes) really easy.
With docker compose we can define the entire multi-container application in a single file and spin it up using one command.
Visit the official Docker Compose Installation Page and follow the instructions tailored for your operating system.
Building images
You can find the related Docker scripts in $SMING_HOME/Tools/Docker.
To build your own images, do this:
cd $SMING_HOME/Tools/Docker/cli
docker-compose build
cd $SMING_HOME/Tools/Docker/ide
docker-compose build
Adjust your settings
sming-ide:
build: .
volumes:
- ../../Sming/:/opt/sming/
ports:
#choose a free port to connect to the web C9 editor
- "10080:80"
devices:
# uncomment to map your serial USB adaptor
#- "/dev/ttyUSB0:/dev/ttyUSB0"
privileged: true
Start your container
cd /opt/sming/Tools/Docker/ide && docker-compose up -d
Open your browser
http://localhost:10080
cd /opt/sming/samples/Basic_Blink
make
make flash
You can also try Sming without installing anything locally. We have an interactive tutorial that can be run directly from your browser.
Configuration
Environment variables
Certain environment variables should be set globally so that command prompts and integrated development environments (IDEs) work correctly.
You can find a list of these in Tools/export.sh.
For Linux and WSL2, append values to your ~/.bashrc file:
# All architectures
export SMING_HOME=/opt/sming/Sming
# Esp8266
export ESP_HOME=/opt/esp-quick-toolchain
# Esp32
export IDF_PATH=/opt/esp-idf
export IDF_TOOLS_PATH=/opt/esp32
export ESP32_PYTHON_PATH=/usr/bin
# Rp2040
export PICO_TOOLCHAIN_PATH=/opt/rp2040
Another approach is to place these in a separate file, then add source ~/.smingrc.
For Windows, you can either edit the global variables graphically via control panel,
or use the setx command:
REM All architectures
setx SMING_HOME "c:\tools\sming\Sming"
REM Esp8266
setx ESP_HOME "c:\tools\esp-quick-toolchain"
REM Esp32
setx IDF_PATH "c:\tools\esp-idf"
setx IDF_TOOLS_PATH "c:\tools\esp32"
setx ESP32_PYTHON_PATH "c:\Python39"
REM Rp2040
setx PICO_TOOLCHAIN_PATH "c:\tools\rp2040"
In both cases values will only take effect in new sessions so close/reopen command prompts or IDEs.
Note that project-specific settings should NOT be configured globally.
Please use the project’s component.mk file for that purpose.
Configuring your device
You may need to configure your project to support the specific device being programmed:
COM_PORTIf you haven’t set this already, it will need to match the port you’re using to talk to the target device.COM_SPEED_ESPTOOLThe default value should work fine but you can usually set a much faster speed.
You can change these on the command line:
make SMING_ARCH=Esp8266 COM_PORT=/dev/ttyUSB3 COM_SPEED_ESPTOOL=921600
For Windows or WSL expect to use COM2, COM3, etc.
Once you’re happy with the settings, you can add them to your project’s component.mk file.
You may need to do this to reset the cached values:
make config-clean
The current set of configuration variables can be seen thus:
make list-config
Other hardware-specific settings are stored in the hardware configuration file. You can examine the current configuration like this:
make hwconfig
The standard config should work with all ESP8266 variants. If you want to use SPIFFS then you should add this line to your component.mk file:
HWCONFIG = spiffs
This expects your device to have at least 4MBytes of flash.
See Sming build system for further details about configuring your project.
See Features for configuring Sming options.
Documentation
In addition to our online documentation, you can also generate a complete documentation locally by following these instructions.
Examples
The examples are a great way to learn the API and brush up your C/C++ knowledge.
Further documentation about the Sample Projects is available too.
Basic Blink
Blinking is something like the “Hello World” example for the embedded world. You can check it using the commands below:
cd Sming/samples
cd Basic_Blink
make # -- compiles the application
make flash # -- tries to upload the application to your ESP8266 device.
See Basic Blink for more information.
Simple GPIO input/output
#define LED_PIN 2 // GPIO2
...
pinMode(LED_PIN, OUTPUT);
digitalWrite(LED_PIN, HIGH);
For a complete example take a look at the Basic Blink sample.
Start Serial communication
Serial.begin(9600);
Serial.println("Hello Sming! Let's do smart things.");
Connect to WiFi
WifiStation.enable(true);
WifiStation.config("LOCAL-NETWORK", "123456789087"); // Put your SSID and password here
Read DHT22 sensor
#include <Libraries/DHTesp/DHTesp.h> // This is just a popular Arduino library!
#define DHT_PIN 0 // GPIO0
DHTesp dht;
void init()
{
dht.setup(DHT_PIN, DHTesp::DHT22);
float h = dht.getHumidity();
float t = dht.getTemperature();
}
Take a look at the code of the DHT22 Humidity Sensor sample.
HTTP client
HttpClient thingSpeak;
...
thingSpeak.downloadString("http://api.thingspeak.com/update?key=XXXXXXX&field1=" + String(sensorValue), onDataSent);
void onDataSent(HttpClient& client, bool successful)
{
if (successful) {
Serial.println("Successful!");
}
else {
Serial.println("Failed");
}
}
For more examples take a look at the HTTP Client, HttpClient Instapush and ThingSpeak Http Client samples.
OTA application update
void doUpgrade()
{
// need a clean object, otherwise if run before and failed will not run again
if(otaUpdater) {
delete otaUpdater;
}
otaUpdater = new Ota::Network::HttpUpgrader();
// select rom partition to flash
auto part = ota.getNextBootPartition();
// The content located on ROM_0_URL will be stored to the new partition
otaUpdater->addItem(ROM_0_URL, part);
// and/or set a callback (called on failure or success without switching requested)
otaUpdater->setCallback(upgradeCallback);
// start update
otaUpdater->start();
}
For a complete example take a look at the Basic Ota sample.
Embedded HTTP Web Server
server.listen(80);
server.paths.set("/", onIndex);
server.paths.set("/hello", onHello);
server.paths.setDefault(onFile);
Serial.println("=== WEB SERVER STARTED ===");
Serial.println(WifiStation.getIP());
...
void onIndex(HttpRequest &request, HttpResponse &response)
{
TemplateFileStream *tmpl = new TemplateFileStream("index.html");
auto &vars = tmpl->variables();
vars["counter"] = String(counter);
vars["IP"] = WifiStation.getIP().toString();
vars["MAC"] = WifiStation.getMacAddress().toString();
response.sendTemplate(tmpl);
}
void onFile(HttpRequest &request, HttpResponse &response)
{
String file = request.getPath();
if (file[0] == '/')
file = file.substring(1);
response.setCache(86400, true);
response.sendFile(file);
}
For more examples take a look at the HttpServer Config Network, Bootstrap Http Server, HttpServer Websockets and AJAX Http Server samples.
Email client
SmtpClient emailClient;
emailClient.connect(Url("smtp://user:password@domain.com"));
MailMessage* mail = new MailMessage();
mail->from = "developers@sming";
mail->to = "iot-developers@world";
mail->subject = "Greetings from Sming";
mail->setBody("Hello");
FileStream* file= new FileStream("image.png");
mail->addAttachment(file);
emailClient.onMessageSent(onMailSent);
emailClient.send(mail);
...
int onMailSent(SmtpClient& client, int code, char* status)
{
MailMessage* mail = client.getCurrentMessage();
...
if(client.countPending() == 0) {
client.quit();
}
return 0;
}
See the SMTP Client sample for details.
Live Debugging
Applications based on Sming Framework that are flashed and running on an ESP8266 device can be debugged using interactive debuggers. In order to debug an application it has to be re-compiled with the ENABLE_GDB=1 directive. And then flashed on the device. As shown below:
cd $SMING_HOME/../samples/LiveDebug
make clean
make ENABLE_GDB=1
make flashapp # <-- this will update only the application firmware.
Once the debuggable application is flashed on the device the developers have to run GDB. The easiest way to run the command-line GDB is to execute the following command:
make gdb
Developers using Eclipse CDT can have debugging sessions like the one below:
See Live Debug sample for details.
Features
There are multiple custom features that can be enabled by default. For example: SSL support, custom LWIP, open PWM, custom heap allocation, more verbose debugging, etc.
The features available are dependent upon the architecture for which Sming is being built.
Sming (main)
This is the main Sming Component containing all architecture-independent code. All architecture-specific stuff is in either Sming (Esp8266) or Sming (Host).
Configuration variables
Serial Communications
- COM_SPEED
Default baud rate for serial port.
This will recompile your application to use the revised baud rate. Note that this will change the default speed used for both flashing and serial comms. See also Esptool and Terminal for further details.
The default rate for serial ports is 115200 baud. You can change it like this:
make COM_SPEED=921600
Debug information log level and format
- DEBUG_VERBOSE_LEVEL
When compiled in debug mode (:envvar:SMING_RELEASE undefined) there are four debug levels in increasing level of verbosity:
0: errors
1: warnings
2: information (default)
3: debug
Change it like this:
make DEBUG_VERBOSE_LEVEL=3
- DEBUG_PRINT_FILENAME_AND_LINE
Set this to 1 to include the filename and line number in every line of debug output. This will require extra space on flash.
Note
If you change these settings and want them applied to Sming, not just your project, then you’ll need to recompile all components like this:
make components-clean
make DEBUG_VERBOSE_LEVEL=3
Release builds
- SMING_RELEASE
By default, this value is undefined to produce a build with debug output. To build for release, do this:
make SMING_RELEASE=1
This remains in force until you change it back:
make SMING_RELEASE=
Localisation
- LOCALE
Sming can format dates/time values based on a country code identified by this value. This is provided as a #define symbol for your application to use. See Sming/Core/SmingLocale.h for further details.
Networking
- DISABLE_NETWORK
0 (Default) 1 - Remove core networking support
Applications which do not require networking can set this flag to avoid building or linking the core Networking Support library.
This will reduce build times, application size and RAM usage. Builds will not succeeded if network code has been inadvertently included.
- DISABLE_WIFI
Note
EXPERIMENTAL
0 (Default) 1 - Exclude WiFi initialisation code
Keeps the core Networking Support library but excludes WiFi code. Applications using ethernet can use this to reduce code size. See Basic Ethernet.
Components
FlashString
Introduction
This is a C++ library to simplify the definition and use of data structures stored in program (flash) memory in an embedded microcontroller.
It was developed for use with Sming and the ESP8266 but could be ported to other platforms relatively easily.
Perhaps the most common use for PROGMEM data is passing strings around, usually as a pointer to a ‘C’ NUL-terminated char array. There are a couple of problems with this:
If we need to know how long the string is, we need to call
strlen_P(), which is really expensive computationally.We can’t include NUL characters in the string.
Both of these are easily solved by passing the length along with the string, like this:
char myString[] PROGMEM = "hello, this is a string"; Serial.println(FPSTR(myString), sizeof(myString) - 1);Of course, passing two parameters instead of one gets tiresome and is not very C++, is it?
This library implements C++ objects stored in flash memory, using macros to create the data structures. The object interfaces are implemented using class templates for performance and flexibility.
The classes are all in the
FSTRnamespace.Objects
Introduction
An
FSTR::Objectis a class template with array-like behaviour, though it is not used directly.Instead, use one of the four classes in the library:
Each type has its own set of macros for easy data construction, and creation of the appropriate Object class which may then be used directly.
Macros follow the same pattern:
DEFINE_FSTR_*Creates a static data structure with an associated Object reference. The _LOCAL variant makes the reference
static constexpr.DECLARE_FSTR_*Use this in a header to declare an Object reference so it can be used across translation units.
Created symbols are C++ and adopt any enclosing namespaced.
Reading Object content
To read parts of an Object, use the
FSTR::Object::read()method.If the data isn’t used very often, use the
FSTR::Object::readFlash()method instead as it avoids disrupting the cache. TheFSTR::Streamclass (aliasFlashMemoryStream) does this by default.Object Internals
This section provides some examples of how structures are created, but in normal use you should use the provided macros as they simplify the task and include structure validity checks.
FSTR::ObjectBaseis a non-template POD base class, and looks like this (methods omitted):class ObjectBase { uint32_t flashLength_; // uint8_t data[]; };Attention
flashLength_must not be accessed directly; use thelength()method instead.Data structures are created like this:
constexpr const struct { ObjectBase object; char data[8]; } flashHelloData PROGMEM = { {5}, "hello" };The
objectfield may then be cast to a reference of the required type, like this:auto& str = flashHelloData.object.as<FSTR::String>();If you want to access it as an array, do this:
auto& arr = str.as<FSTR::Array<char>>();References are an efficient and convenient way to access an Object, and should not consume any memory themselves as the compiler/linker resolve them to the actual object.
However, in practice the Espressif compiler stores a full pointer to most things to support relative addressing, and if the references aren’t declared PROGMEM they’ll consume RAM.
Copy behaviour
Whilst references are the preferred way to access flash Objects, they can also be created dynamically:
FSTR::String emptyString; FSTR::String stringCopy(FS("Inline string"));Such instances are stored in RAM but only consume 4 bytes as they simply keep a pointer to the real flash Object.
Note
Don’t try to copy ObjectBase!
Here’s a somewhat contrived example to demonstrate:
DEFINE_FSTR_DATA_LOCAL(flashHelloData, "Hello"); auto myCopy = flashHelloData.object; Serial.print("myCopy.length() = "); Serial.println(myCopy.length());In debug builds, this will throw an assertion. In release builds, you’ll get a zero-length object.
Aggregate initialization
We use aggregate initialization to set up the structures so the data is fixed at link time without any constructor or initialiser functions.
This means classes cannot have:
user-provided constructors
brace-or-equal-initializers for non-static data members
private or protected non-static data members
virtual functions
base classes (until C++17)
This is why
FSTR::ObjectBaseis used to define data structures.Classes created using the
FSTR::Objecttemplate ensures the necessary constructors are available to do this:auto myCopy = flashHelloData.object.as<FSTR::String>(); Serial.print("myCopy.length() = "); Serial.println(myCopy.length());The macros create an appropriate Object& reference for you.
Structure checks
The construction macros include a sanity check to ensure the initialization is truly just Plain Old Data, without any hidden initialisers.
You may encounter one of the following errors during compilation:
The value of ‘X’ is not usable in a constant expression
FSTR structure not POD
This generally means one or more of the arguments in the initialisation data is not
constexpr. Most compilers are quite relaxed about this butGCC 4.8.5is particularly thick.In testing, this happens with references for global Objects, which of course cannot be constexpr. To fix it, the offending Object either needs to be redefined LOCAL, or if the Object data is in scope (i.e. defined in the same source file) then you can get a direct pointer to it using the
FSTR_PTR()macro.Macros
- DECLARE_FSTR_OBJECT(name, ObjectType)
Declare a global Object reference.
- Parameters:
name –
ObjectType –
- DEFINE_FSTR_REF(name, ObjectType, object)
Define a reference to an object.
- Parameters:
name – Name for reference
ObjectType – Fully qualified typename of object required, e.g. FSTR::String, FlashString, FSTR::Vector<int>, etc.
object – Object instance to cast
- DEFINE_FSTR_REF_NAMED(name, ObjectType)
- FSTR_DATA_NAME(name)
Provide internal name for generated flash string structures.
- FSTR_PTR(objref)
Given an Object& reference, return a pointer to the actual object.
However, some older compilers such as GCC 4.8.5 requires such references to be declared constexpr. For example, this fails with
FSTR structure not POD:Global references cannot be declared constexpr, so changing DEFINE_FSTR to DEFINE_FSTR_LOCAL will fix the problem.DEFINE_FSTR(globalStringRef, "This creates a global reference"); DEFINE_VECTOR(myVector, FSTR::String, &globalStringRef); ^^^Another solution is to get a direct pointer to the actual data structure:
We can only do this of course if the data structure is in scope.DEFINE_VECTOR(myVector, FSTR::String, FSTR_PTR(globalStringRef));
- Parameters:
objref – When an Object pointer is required, such when defining entries for a Vector or Map, it is usually sufficient to use &objref.
- FSTR_CHECK_STRUCT(name)
Check structure is POD-compliant and correctly aligned.
- IMPORT_FSTR_OBJECT(name, ObjectType, file)
Import an object from an external file with reference.
See also
See also
IMPORT_FSTR_DATANote
Can only be used at file scope
- Parameters:
name – Name for the object
ObjectType – Object type for reference
file – Absolute path to the file containing the content
- IMPORT_FSTR_OBJECT_LOCAL(name, ObjectType, file)
Like IMPORT_FSTR_OBJECT except reference is declared static constexpr.
Class Template
- template<class ObjectType, typename ElementType>
class Object : public FSTR::ObjectBaseBase class template for all types.
- Template Parameters:
ObjectType – The object type actually being instantiated
ElementType –
Public Functions
- inline Object()
Creates a null object.
- inline Object(const Object &obj)
Copy constructor.
Note
Objects are usually passed around by reference or as a pointer, but for ease of use we need a working copy constructor. A copy contains a pointer to the real object.
- inline constexpr const size_t length() const
Get the length of the content in elements.
- inline ElementType operator[](unsigned index) const
Array operator[].
- inline size_t read(size_t index, ElementType *buffer, size_t count) const
Read content into RAM.
- Parameters:
index – First element to read
buffer – Where to store data
count – How many elements to read
- Return values:
size_t – Number of elements actually read
- inline size_t readFlash(size_t index, ElementType *buffer, size_t count) const
Read content into RAM,using
flashmem_read()
- Parameters:
index – First element to read
buffer – Where to store data
count – How many elements to read
- Return values:
size_t – Number of elements actually read
Public Static Functions
- static inline constexpr const ObjectType &empty()
Return an empty object which evaluates to null.
Flash Strings
Introduction
Strings are basically just arrays of char, but have additional methods to allow them to be used more easily. These methods are consistent with Wiring
String, so should be reasonably familiar.
length()returns the number of characters in the String, excluding the NUL terminator
size()returns the number of bytes of storage usedFor example, “123” is actually stored as
{ '1', '2', '3', '\0' }so the length is 3 and the size is 4. However, “1234” is stored as{ '1', '2', '3', '4', '\0' }so the length is 4 and the size is 8.Using Strings
Note
You can use
FSTR::Stringor the Sming-providedFlashStringalias to work with Strings.Within a function:
DEFINE_FSTR_LOCAL(myFlashString, "This is my flash string"); Serial.println(myFlashString); Serial.printf("myFlashString has %u chars and occupies %u bytes\n", myFlashString.length(), myFlashString.size());To use Strings across translation units, we do this in the header:
DECLARE_FSTR(myFlashString);And in a source file:
DEFINE_FSTR(myFlashString, "I am a flash string\0I've got a Naughty NUL.");You can generally use a Flash String anywhere you can use a regular Wiring String as it has an implicit ::String() operator. Note that
WStringis used within the library for disambiguation.Inline Strings
Use the
FS()macro to create Flash Strings inline:Serial.println(FS("A Flash String"));Note
The macro makes use of
FS_PTR()which creates the structure and returns a pointer to it. It behaves like a function call, although the compiler inlines the code.Therefore FS() may only be used within functions. At file scope you’ll get this error:
statement-expressions are not allowed outside functions nor in template-argument lists
The example above doesn’t provide any improvement over
Fas there are no Flash String overloads available, so is equivalent to this:String s = FS("A Flash String"); Serial.println(s);However, it’s rather different if you pass it to a function which recognises Flash Strings, like this:
FSTR::println(Serial, FS("A Flash String"));This is equivalent to:
FS("A Flash String").printTo(Serial); Serial.println();
FSTR::String::printTo()uses no heap and imposes no restriction on the string length.Nested Inline Strings
It would be really useful to be able to use inline Strings this within nested structures, and this can be done provided those structures are in RAM.
Important
Inline Strings cannot be used when defining Vectors or Maps.
Here’s is a simplified structure we will attempt to initialize:
static const struct { FlashString* string; } flashData PROGMEM = { FS_PTR("Inline Flash String") }; Serial.println(*flashData.string);The static flashData structure gets initialised at runtime on first use, as per C++ rules. This attempts to copy our pointer into the flashData structure which clearly it cannot do as it’s in
PROGMEM, so we get a LOAD/STORE error. We must remove PROGMEM.Avoiding the heap
Instead of using a temporary Wiring String, you can use
LOAD_FSTR()to load the content into a temporary stack buffer:DEFINE_FSTR(globalTest, "This is a testing string"); void func() { LOAD_FSTR(local, globalTest); printf("%s, %u characters, buffer is %u bytes\n", local, globalTest.length(), sizeof(local)); }You can do this with inline Flash Strings using
FSTR_ARRAY():FSTR_ARRAY(buffer, "text");Is roughly equivalent to:
char name[] = "text";Except the buffer is word aligned, so sizeof(name) may differ.
Macros
- FS_PTR(str)
Define an inline String and return a pointer to it.
Note
The rather obscure
asmstatement is required to prevent the compiler from discarding the symbol at link time, which leads to an ‘undefined reference’ error
- FS(str)
Define an inline FSTR::String and return it as a copy.
Example:
Serial.println(FS("This is a Flash String"));
- DECLARE_FSTR(name)
Declare a global FSTR::String& reference.
Note
Define the FSTR::String object using DEFINE_STR()
- Parameters:
name –
- DEFINE_FSTR(name, str)
Define a FSTR::String object with global reference.
Example:
DEFINE_FSTR(test, “This is a test\0Another test\0hello”)
The data includes the nul terminator but the length does not.
- Parameters:
name – Name of FSTR::String& reference to define
str – Content of the FSTR::String
- DEFINE_FSTR_LOCAL(name, str)
Like DEFINE_FSTR except reference is declared static constexpr.
- DEFINE_FSTR_DATA(name, str)
Define a FSTR::String data structure.
- Parameters:
name – Name of data structure
str – Quoted string content
- LOAD_FSTR(name, fstr)
Load a FSTR::String object into a named local (stack) buffer.
Example:
DEFINE_FSTR(globalTest, "This is a testing string") ... LOAD_FSTR(local, globalTest) printf("%s, %u characters, buffer is %u bytes\n", local, globalTest.length(), sizeof(local));
- FSTR_ARRAY(name, str)
Define a flash FSTR::String and load it into a named char[] buffer on the stack.
Note
Equivalent to
char name[] = "text"except the buffer is word aligned. Faster than using a temporary Wiring String and avoids using the heap.
- Parameters:
name – Name of char[] buffer
str – Content of FSTR::String
- IMPORT_FSTR(name, file)
Define a FSTR::String containing data from an external file.
See also
See also
IMPORT_FSTR_DATA
- Parameters:
name – Name for the FSTR::String object
file – Absolute path to the file containing the content
- IMPORT_FSTR_LOCAL(name, file)
Like IMPORT_FSTR except reference is declared static constexpr.
- FSTR_TABLE(name)
declare a table of FlashStrings
- Deprecated:
Use a Vector or Map
Declares a simple table. Example:
Table entries may be accessed directly as they are word-aligned. Examples:DEFINE_FSTR(fstr1, "Test string #1"); DEFINE_FSTR(fstr2, "Test string #2"); FSTR_TABLE(table) = { &fstr1, &fstr2, };debugf("fstr1 = '%s'", FSTR::String(*table[0]).c_str()); debugf("fstr2.length() = %u", table[1]->length());
- Parameters:
name – name of the table
String Class
- class String : public FSTR::Object<String, char>
describes a counted string stored in flash memory
Public Functions
- inline size_t size() const
Get the number of bytes used to store the String.
Note
Always an integer multiple of 4 bytes
- inline flash_string_t data() const
Get a WString-compatible pointer to the flash data.
- bool equals(const char *cstr, size_t len, bool ignoreCase = false) const
Check for equality with a C-string.
Note
loads string into a stack buffer for the comparison, no heap required
- Parameters:
cstr –
len – Length of cstr (optional)
- Return values:
bool – true if strings are identical
Arrays
Introduction
Supports arrays of simple types, such as char, int, double, or POD structures (i.e. basic C structures).
FSTR::Arrayis a class template, so requires an additionalElementTypeparameter:#include <FlashString/Array.hpp> DEFINE_FSTR_ARRAY(myDoubleArray, double, PI, 53.0, 100, 1e8, 47 ); Serial.print("My double array: "); myDoubleArray.printTo(Serial); Serial.println();Note
Objects do not inherit from Printable because it is a virtual base class.
Therefore, statements like
Serial.println(myDoubleArray)are not supported.This also avoids ambiguity between implicit WString conversions.
There are some Print helper functions in the library you can use:
FSTR::println(Serial, myDoubleArray);These are templated so will handle both simple data types and Objects.
You can share Arrays between translation units by declaring it in a header:
DECLARE_FSTR_ARRAY(table);Macros
- DECLARE_FSTR_ARRAY(name, ElementType)
Declare a global Array& reference.
Note
Use
DEFINE_FSTR_ARRAYto instantiate the global Object
- Parameters:
name –
ElementType –
- DEFINE_FSTR_ARRAY(name, ElementType, ...)
Define an Array Object with global reference.
Note
Unlike String, array is not NUL-terminated
- Parameters:
name – Name of Array& reference to define
ElementType –
... – List of ElementType items
- DEFINE_FSTR_ARRAY_LOCAL(name, ElementType, ...)
Like DEFINE_FSTR_ARRAY except reference is declared static constexpr.
- DEFINE_FSTR_ARRAY_DATA(name, ElementType, ...)
Define an Array data structure.
- Parameters:
name – Name of data structure
ElementType –
... – List of ElementType items
- LOAD_FSTR_ARRAY(name, array)
Load an Array object into a named local (stack) buffer.
Note
Example:
DEFINE_FSTR_ARRAY(fsArray, double, 5.33, PI) ... LOAD_FSTR_ARRAY(arr, fsArray) printf("arr[0] = %f, %u elements, buffer is %u bytes\n", arr[0], fsArray.length(), sizeof(arr));
- FSTR_ARRAY_ARRAY(name, ElementType, ...)
Define an Array and load it into a named buffer on the stack.
Note
Equivalent to
ElementType name[] = {a, b, c}except the buffer is word-aligned
- IMPORT_FSTR_ARRAY(name, ElementType, file)
Define an Array containing data from an external file.
See also
See also
IMPORT_FSTR_DATA
- Parameters:
name – Name for the Array object
ElementType – Array element type
file – Absolute path to the file containing the content
- IMPORT_FSTR_ARRAY_LOCAL(name, ElementType, file)
Like IMPORT_FSTR_ARRAY except reference is declared static constexpr.
Classes
- template<typename ElementType>
class Array : public FSTR::Object<Array<ElementType>, ElementType>Class to access an array of integral values stored in flash.
- Template Parameters:
ElementType –
Tables
Introduction
Simple tables can be implemented using Arrays, like this:
struct TableRow { float columns[3]; int operator[](size_t index) const { return columns[index]; } }; DEFINE_FSTR_ARRAY(table, TableRow, {0.1, 0.2, 0.3}, {0.6, 0.7, 0.8} ); for(auto row: table) { Serial.printf("%f, %f, %f\n", row[0], row[1], row[2]); }Each row is a fixed size. The
FSTR::TableRowclass template is provided to simplify this:#include <FlashString/Table.hpp> using FloatRow = FSTR::TableRow<float, 3>; DEFINE_FSTR_ARRAY(table, FloatRow, {0.1, 0.2, 0.3}, {0.6, 0.7, 0.8} ); table.printTo(Serial); table.println();If you want to create a table with rows of different sizes or types, use a Vector.
Class Template
- template<typename ElementType, size_t Columns>
class TableRowClass template to define the row of a table.
Use with an Array Object to construct simple tables. Methods provide Object-like access.
- Template Parameters:
ElementType –
Columns – Number of columns in the table
Public Functions
- inline ElementType operator[](size_t index) const
Array operator.
- Parameters:
index –
- Return values:
ElementType –
- inline size_t length() const
Get number of columns.
- Return values:
size_t –
Vectors
Introduction
A
FSTR::Vectoris an array of Object pointers:struct Vector<ObjectType> { FSTR::Object object; ObjectType* entries[]; };A key use for this is the construction of string tables.
Defining Vectors
Inline Strings are not supported, so the content has to be defined first:
DEFINE_FSTR_LOCAL(str1, "Test string #1"); DEFINE_FSTR_LOCAL(str2, "Test string #2"); IMPORT_FSTR_LOCAL(str3, PROJECT_DIR "/files/somedata.json");Now we can define the Vector:
#include <FlashString/Vector.hpp> DEFINE_FSTR_VECTOR(myTable, FlashString, &str1, &str2, nullptr, &str3 );Note the use of
nullptrto indicate an invalid vector entry, as distinct from an empty String.Using Vectors
Now we can access the data using Vector methods:
debugf("table.length() = %u", table.length()); debugf("fstr1 = '%s'", String(table[0]).c_str()); debugf("fstr2.length() = %u", table[1].length()); debugf("fstr3.length() = %u", table[2].length());You can share Vectors between translation units by declaring it in a header:
DECLARE_FSTR_VECTOR(table);To search a Vector:
int i = table.indexOf("TEST STRING #1");Note
By default, searches in Vector<String> are not case-sensitive.
The
indexOfmethod has an extraignoreCaseparameter, which defaults totrue.Structure
The above example generates a structure like this:
const struct { ObjectBase object; String* entries[4]; } __fstr__myTable PROGMEM = { {16}, &str1, &str2, nullptr, &str3, }; const Vector<String>& myTable PROGMEM = __fstr__myTable.as<Vector<String>>();Note:
FSTR::namespace qualifier omitted for clarity.Macros
- DECLARE_FSTR_VECTOR(name, ObjectType)
Declare a global Vector& reference.
Note
Use
DEFINE_VECTORto instantiate the global Object
- Parameters:
name –
ObjectType –
- DEFINE_FSTR_VECTOR(name, ObjectType, ...)
Define a Vector Object with global reference.
Note
Size will be calculated
- Parameters:
name – Name of Vector& reference to define
ObjectType –
... – List of ObjectType* pointers
- DEFINE_FSTR_VECTOR_LOCAL(name, ObjectType, ...)
Like DEFINE_FSTR_VECTOR except reference is declared static constexpr.
- DEFINE_FSTR_VECTOR_SIZED(name, ObjectType, size, ...)
Define a Vector Object with global reference, specifying the number of elements.
Note
Use in situations where the array size cannot be automatically calculated
- Parameters:
name – Name of Vector& reference to define
ObjectType –
size – Number of elements
... – List of ObjectType* pointers
- DEFINE_FSTR_VECTOR_SIZED_LOCAL(name, ObjectType, size, ...)
Like DEFINE_FSTR_VECTOR_SIZED except reference is declared static constexpr.
- DEFINE_FSTR_VECTOR_DATA(name, ObjectType, ...)
Define a Vector data structure.
Note
Size will be calculated
- Parameters:
name – Name of data structure
ObjectType –
... – List of ObjectType* pointers
- DEFINE_FSTR_VECTOR_DATA_SIZED(name, ObjectType, size, ...)
Define a Vector data structure and specify the number of elements.
Note
Use in situations where the array size cannot be automatically calculated
- Parameters:
name – Name of data structure
ObjectType –
size – Number of elements
... – List of ObjectType* pointers
Class Template
- template<class ObjectType>
class Vector : public FSTR::Object<Vector<ObjectType>, const ObjectType*>Class to access a Vector of objects stored in flash.
- Template Parameters:
ObjectType –
Maps
Introduction
A
FSTR::Mapis analogous to the WiringHashMapclass, allowing content to be indexed using a key value.The Map contains an array of
FSTR::MapPairstructures:struct MapPair<KeyType, ContentType> { KeyType key_; ContentType* content_; };
KeyTypecan be any simple type such aschar,int,float,enumetc. It may also be aStringObject (or, more precisely,String*).
ContentTypecan be any Object type (String, Array, Vector or Map). This allows hierarchical structures to be created.Example: int ⇒ String
Here’s a basic example using integer keys:
#include <FlashString/Map.hpp> IMPORT_FSTR_LOCAL(content1, PROJECT_DIR "/files/index.html"); IMPORT_FSTR_LOCAL(content2, PROJECT_DIR "/files/favicon.html"); DEFINE_FSTR_MAP(intmap, int, FSTR::String, {35, &content1}, {180, &content2} );You should generally use
IMPORT_FSTR_LOCAL()when referencing imported objects. If you need global access to imported data as well, then useIMPORT_FSTR().Note
Older toolchains (generally GCC earlier than version 6) will fail to compile with error: the value of ‘FS_content1’ is not usable in a constant expression.
You can work around this as follows:
IMPORT_FSTR(content1, PROJECT_DIR "/files/index.html"); // Make this a global reference IMPORT_FSTR_LOCAL(content2, PROJECT_DIR "/files/favicon.html"); DEFINE_FSTR_MAP(intmap, int, FSTR::String, {35, &FSTR_DATA_NAME(FS_content1).as<FSTR::String>()}, // Cast the actual content {180, &content2} );We can now do this:
void printValue(int key) { auto value = intmap[key]; if(value) { Serial.printf("Found '%u' in map, containing %u chars\n", value.key(), value.content().length()); Serial.println(value.printer()); } else { Serial.printf("Couldn't find '%u' in map\n", key); } }Example: String ⇒ String
Both the key and the content are stored as Strings:
#include <FlashString/Map.hpp> DEFINE_FSTR_LOCAL(key1, "index.html"); DEFINE_FSTR_LOCAL(key2, "favicon.ico"); IMPORT_FSTR_LOCAL(content1, PROJECT_DIR "/files/index.html"); IMPORT_FSTR_LOCAL(content2, PROJECT_DIR "/files/favicon.html"); DEFINE_FSTR_MAP(fileMap, FlashString, FlashString, {&key1, &content1}, {&key2, &content2}, );We can now do this:
void onFile(HttpRequest& request, HttpResponse& response) { String fileName = request.uri.getRelativePath(); auto& value = fileMap[fileName]; if(value) { // Found Serial.printf("Found '%s' in fileMap\n", String(value.key()).c_str()); auto stream = new FlashMemoryStream(value); response.sendDataStream(stream, ContentType::fromFullFileName(fileName)); } else { Serial.printf("File '%s' not found\n", fileName.c_str()); } }Note
As with
Vector<String>,Map<String, ...>lookups are by default case-insensitive.If you require a case-sensitive lookup, use the
indexOfmethod withignoreCase = false.Structure
The macro in the first example above produces a structure like this:
constexpr const struct { ObjectBase object; MapPair<int, String> data[2]; } __fstr__intmap PROGMEM = { {16}, {35, &content1}, {180, &content2}, }; const Map<int, String>& intmap = __fstr__intmap.object.as<Map<int, String>>();Note:
FSTR::namespace qualifier omitted for clarity.Usually, each MapPair is 8 bytes, but if the key is a double or int64 it would be 12 bytes.
Macros
- DECLARE_FSTR_MAP(name, KeyType, ContentType)
Declare a global Map& reference.
Note
Use DEFINE_FSTR_MAP to instantiate the global object
- Parameters:
name – Name of the Map& reference to define
KeyType – Integral type to use for key
ContentType – Object type to declare for content
- DEFINE_FSTR_MAP(name, KeyType, ContentType, ...)
Define a Map Object with global reference.
Note
Size will be calculated
- Parameters:
name – Name of the Map& reference to define
KeyType – Integral type to use for key
ContentType – Object type to declare for content
... – List of MapPair definitions { key, &content }
- DEFINE_FSTR_MAP_LOCAL(name, KeyType, ContentType, ...)
Like DEFINE_FSTR_MAP except reference is declared static constexpr.
- DEFINE_FSTR_MAP_SIZED(name, KeyType, ContentType, size, ...)
Define a Map Object with global reference, specifying the number of elements.
- Parameters:
name – Name of the Map& reference to define
KeyType – Integral type to use for key
ContentType – Object type to declare for content
size – Number of elements
... – List of MapPair definitions { key, &content }
- DEFINE_FSTR_MAP_SIZED_LOCAL(name, KeyType, ContentType, size, ...)
Like DEFINE_FSTR_MAP_SIZED except reference is declared static.
- DEFINE_FSTR_MAP_DATA(name, KeyType, ContentType, ...)
Define a Map data structure.
Note
Size will be calculated
- Parameters:
name – Name of data structure
KeyType – Integral type to use for key
ContentType – Object type to declare for content
... – List of MapPair definitions { key, &content }
- DEFINE_FSTR_MAP_DATA_SIZED(name, KeyType, ContentType, size, ...)
Define a Map data structure, specifying the number of elements.
- Parameters:
name – Name of data structure
KeyType – Integral type to use for key
ContentType – Object type to declare for content
size – Number of elements
... – List of MapPair definitions { key, &content }
Class Templates
- template<typename KeyType, class ContentType, class Pair = MapPair<KeyType, ContentType>>
class Map : public FSTR::Object<Map<KeyType, ContentType>, MapPair<KeyType, ContentType>>Class template to access an associative map.
- Template Parameters:
KeyType –
ContentType –
Public Functions
- inline const Pair valueAt(unsigned index) const
Get a map entry by index, if it exists.
Note
Result validity can be checked using if()
- template<typename TRefKey, typename T = KeyType>
inline std::enable_if<!std::is_class<T>::value, int>::type indexOf(const TRefKey &key) constLookup an integral key and return the index.
- Parameters:
key – Key to locate, must be compatible with KeyType for equality comparison
- Return values:
int – If key isn’t found, return -1
- template<typename TRefKey, typename T = KeyType>
inline std::enable_if<std::is_same<T, String>::value, int>::type indexOf(const TRefKey &key, bool ignoreCase = true) constLookup a String key and return the index.
- Parameters:
key –
ignoreCase – Whether search is case-sensitive (default: true)
- Return values:
int – If key isn’t found, return -1
- template<typename KeyType, class ContentType>
class MapPairdescribes a pair mapping key => data for a specified key type
- Template Parameters:
Public Functions
- inline operator IfHelperType() const
Provides bool() operator to determine if Pair is valid.
- template<typename T = KeyType>
inline std::enable_if<!std::is_class<T>::value, KeyType>::type key() constGet the key (non-class key types)
- template<typename T = KeyType>
inline std::enable_if<std::is_same<T, String>::value, const KeyType&>::type key() constGet the key (String key type)
- inline const ContentType &content() const
Accessor to get a reference to the content.
Streams
- class FSTR::Stream : public IDataSourceStream
Alias:
FlashMemoryStreamThis is a Sming
IDataSourceStreamdescendant which you can use to stream the contents of any FlashString object. It’s especially useful when used withIMPORT_FSTR:IMPORT_FSTR(myLargeFile, PROJECT_DIR "/files/lots-of-stuff.txt"); FSTR::Stream fs(myLargeFile); Serial.println(myLargefile);Because the data is read in sections, it’s not limited by available RAM.
Note
Unless you need
myLargeFileto be a global symbol, you’ll generally want to useIMPORT_FSTR_LOCAL().Like a
FileStream, you can also seek randomly within aFSTR::Stream, so you can use it as the basis for an elementary read-only filesystem.See Maps for a more useful example.
- class FSTR::TemplateStream : public TemplateStream
Alias:
TemplateFlashMemoryStreamStandard templating stream for tag replacement.
Utilities
Importing files
For String and Array objects you can import data directly from a file using
IMPORT_FSTR()orIMPORT_FSTR_ARRAY(). For example:IMPORT_FSTR(myData, PROJECT_DIR "/files/myData.bin");This defines a C++ reference to the data called
myDataso it can be referred to usingDECLARE_FSTR()if required.Attention
File paths must be absolute or the compiler won’t be able to locate it reliably.
Sming provides
PROJECT_DIRandCOMPONENT_PATHto help with this.Note
A corresponding C symbol will also be defined, based on the provided name, to provide linkage with the imported data.
You generally shouldn’t have an issue with this as the symbols are restricted to file scope, but it is something to be aware of.
One use for imported files is to serve content via HTTP, like this:
void onFile(HttpRequest& request, HttpResponse& response) { Serial.printf("myData is %u bytes long\n", myData.length()); auto fs = new FSTR::Stream(myData); response.sendDataStream(fs); }Therefore files can be bound into the firmware and accessed without requiring a filing system. This idea is extended further using Maps.
Custom Imports
Use
IMPORT_FSTR_DATA()to import the contents of a file without defining any C/C++ variable:IMPORT_FSTR_DATA(myCustomData, PROJECT_DIR "/files/data.bin");You’ll need to define an appropriate symbol:
struct MyCustomStruct { uint32_t length; char name[12]; char description[20]; uint8_t data[1024]; }; extern "C" const MyCustomStruct myCustomData;You’ll still have to consider how the data is accessed. If it’s small and un-complicated you can just copy it into RAM:
MyCustomStruct buf; memcpy_P(&buf, &myCustomData, sizeof(buf));Custom Objects
A better way to handle large, complex structures is to define a custom Object to handle it. You can find an example of how to do this in
test/app/custom.cpp, which does this:
Define
MyCustomStruct:struct MyCustomStruct { FSTR::ObjectBase object; char name[12]; char description[20]; FSTR::ObjectBase dataArray; };Define a base object type (
CustomObject) using theFSTR::Objectclass template. This determines the underlying element type, generallycharoruint8_tare most useful.Derive an Object class (
MyCustomObject) to encapsulate access toMyCustomStruct.Use the
IMPORT_FSTR_OBJECT()macro to import the custom data and define a global reference (customObject) of typeMyCustomObject&.Use
DECLARE_FSTR_OBJECT()macro to declare the reference in a header.More complex examples may involve multiple custom Object types.
API Reference
- DECL(t)
Wrap a type declaration so it can be passed with commas in it.
Example:
These fail:template <typename ElementType, size_t Columns> struct MultiRow { ElementType values[Columns]; }Use DECL like this:DECLARE_FSTR_ARRAY(myArray, MultiRow<double, 3>); DECLARE_FSTR_ARRAY(myArray, (MultiRow<double, 3>));Although for this example we should probably do this:DECLARE_FSTR_ARRAY(myArray, DECL((MultiRow<double, 3>)) );using MultiRow_double_3 = MultiRow<double, 3>; DECLARE_FSTR_ARRAY(myArray, MultiRow_double_3);
- IMPORT_FSTR_DATA(name, file)
Link the contents of a file.
This provides a more efficient way to read constant (read-only) file data. The file content is bound into firmware image at link time.
We need inline assembler’s
.incbininstruction to actually import the data. We use a macro STR() so that if required the name can be resolved from a#definedvalue.Use PROJECT_DIR to locate files in your project’s source tree:
Use COMPONENT_PATH within a component.IMPORT_FSTR_DATA(myFlashData, PROJECT_DIR "/files/my_flash_file.txt");No C/C++ symbol is declared, this is type-dependent and must be done separately:
If the symbol is not referenced the content will be discarded by the linker.extern "C" FSTR::String myFlashData;
- STR(x)
- XSTR(x)
- template<typename T>
struct argument_typeSming Integration
Sming provides several aliases to provide compatibility with existing code:
Other pages
Upgrade notes
Version 1.0
FlashString was first introduced in Sming 3.7.0 on 17 November 2018, as a single file in the Wiring directory. Other than bugfixes and Host Emulator support it hasn’t had any significant updates.
Version 2.0
This library was introduced to Sming in version 4.0.1. If you are migrating from Sming 3.7.0 or later and have used FlashString in your projects, some minor changes may be necessary.
FlashString
This has been moved inside the
FSTRnamespace and renamed toString.FlashString has been retained as an alias for convenience to avoid ambiguity when used with Wiring String.
FlashString::isEqual()has been renamed toequal()for consistency with Wiring String.Tables
Table support has been improved and formalised using the
Vectorclass. Previously:DEFINE_FSTR_LOCAL(fstr1, "Test string #1"); DEFINE_FSTR_LOCAL(fstr2, "Test string #2"); static FSTR_TABLE(table) = { FSTR_PTR(fstr1), FSTR_PTR(fstr2), }; Serial.println("FSTR tables[%u]\n", ARRAY_SIZE(table)); Serial.printf(" fstr1 = '%s'\n", String(*table[0]).c_str()); Serial.printf(" fstr1.length() = %u\n", table[0]->length());Now you can do this:
DEFINE_FSTR_LOCAL(str1, "Test string #1"); DEFINE_FSTR_LOCAL(str2, "Test string #2"); DEFINE_FSTR_VECTOR(table, FlashString, &str1, &str2); Serial.printf("FSTR table[%u]\n", table.length()); Serial.printf(" fstr1 = '%s'\n", String(table[0]).c_str()); Serial.printf(" fstr1.length() = %u\n", table[0].length());And perform lookups:
Serial.print(” indexOf(‘Test STRING #1’) = “); Serial.println(table.indexOf(“Test STRING #1”));
Maps and other features
Associative maps have been added to support keys using a Flash String or an integral type. Content is typically a String, but can also be another Table or Map for building hierarchical structures.
Moving to class templates has added a lot of possibilities so I hope you have fun finding out what can be done with this library!
Suggestions for improvements and fixes always welcome :-)
Change Log
Sming 3.7.0
Here’s the PR summary from the original repository. The ‘pull-request’ links are rendered during Sming documentation build.
- Thu Sep 20 18:00:58 2018 Pull Request #1438
Initial commit, the FlashString class plus macros
- Fri Oct 5 14:52:24 2018 Pull Request #1459
Add FSTR_TABLE() macro to define a FlashString* pointer. Very basic table support so we can use it for looking up HTTP status strings, etc.
- Wed Oct 24 09:13:50 2018 Pull Request #1502
Add streaming support - Add FlashMemoryStream class - Add FlashTemplateStream class - Add IMPORT_FSTR macro to bind a file into firmware image as FlashString object - Demonstrate usage using HttpServer_ConfigNetwork sample application. Note behaviour is unchanged, but settings.html is statically linked rather than using SPIFFS.
- Sat May 11 14:25:05 2019 Pull Request #1690
Mark const FlashString& references as PROGMEM. The compiler still stores a pointer to long address so it can use relative addressing modes, and it puts these in RAM unless you tell it not to.
Also added DEFINE_FSTR_LOCAL macro to allow correct use inside function definitions.
- Thu May 30 14:49:06 2019 Pull Request #1692
Sming Host Emulator! The IMPORT_FSTR feature requires architecture-dependent compiler support, so a separate version was added to support WIN32. The standard version works as-is for Linux.
- Tue Jun 4 07:43:29 2019 Pull Request #1709
Fix array bounds issue detected via GCC-8 on Host build (#1709)
- Mon Jul 8 07:21:58 2019 Pull Request #1757
Bugfix: LOAD_FSTR() doesn’t NUL-terminate buffered string (#1757)
- Tue Jul 30 19:55:50 2019 Pull Request #1786
Revise IMPORT_FSTR macro so the name can itself be #defined (#1786)
- Mon Oct 21 08:00:48 2019 Pull Request #1899
Add optional len parameter to Flashstring::isEqual(const char*)
Sming 4.0.0
Sat Nov 2 13:01:20 2019
Sming version 4.0.0 is due for release next week, and FlashString has proven to be a useful addition to the toolkit. Time for an upgrade.
Sming 4.0.1
FlashString has been expanded from a single file into its own Component library to:
Improve support for tables and add associative mapping
Ensure each class has its own header file
Add methods so users don’t need to muck about with memcpy_P, etc.
Move some code out of the header into a separate .cpp source file
Add proper documentation
Make it easier to port to other frameworks (e.g. Arduino)
Get it out of Wiring - doesn’t really belong there and certainly don’t want to clutter up that directory
Improve maintainability
Fix compatibility with ArduinoJson 6.13.0
Change summary:
Move FlashString into a separate Component
Add FlashString::read() and readFlash() methods
Revise internals to be more consistent with naming
Improve table handling using new Vector class
Add associative mapping support with Map class
Revise structures so it contains only the length, obtaining data position using pointer arithmetic. This fixes an error with GCC 8.2.0 which didn’t like accessing zero-length arrays.
Use ObjectBase as the first element in a data structure to allow use of ‘safe’ static casting, handled using the as() method template
Documentation!
Technobabble
Why not just use a wrapper class?
A wrapper class would require this:
const struct { ... } myFlashData PROGMEM = { ... }; void func() { FlashString fs(myFlashData); Serial.println(fs); }Easy enough, just create a wrapper every time you need to use the data.
It does get a bit clunky though when you have a lot of strings, lists of strings or where strings are used in several places.
How about global/static wrappers
Like this:
FlashString fs(myFlashData); void func() { Serial.println(fs); }Each wrapper uses 4 bytes of RAM. We also need to bear in mind that at startup objects are being initialised so we cannot necessarily guarantee that our FlashString wrappers have been created by the time they’re needed in other code. Result: hang/crash.
How about a macro to create the wrapper?
Something like this:
void func() { Serial.println(FS(myFlashData)); }To be honest, I hadn’t really considered that one, but it is worth exploring.
Can’t I put the wrapper in PROGMEM?
A wrapper requires a constructor, each of which needs an initialiser which gets called at startup.
Which is why FlashString isn’t a wrapper.
FlashStrings can be used anywhere you can use PROGMEM data, including globally initialised classes.
There is a disadvantage to this approach: You can’t make copies of FlashString objects. They must always be accessed as references or pointers.
Reasons to use FlashString
Easy to use, treat like read-only
StringobjectUse methods of FlashString instead of calls to memcpy_P, strcpy_P, etc.
Blocks of data stored in flash can be passed around easily as the length is known
Can be passed to functions instead of a String parameter, but without having to load it into RAM first
Can be streamed directly using FlashMemoryStream
Can be read randomly using FlashString::read()
Aligned read and copy operations provide excellent performance
Fast equality comparisons using length field to short-circuit comparison
Data can be imported and linked directly into the program image from a local file, and accessed as a FlashString
Custom structures can be defined and accessed as a FlashString
Operates transparently with ArduinoJson (patched as from version 6.13.0)
Reasons not to use FlashString
Storing lots of small strings can be inefficient because of length and alignment, and because a NUL terminator is always appended by DEFINE_FSTR macros even though it’s not generally required. (This is because of how C requires string arrays to be declared.)
For example, “A” requires 8 bytes:
01 00 00 00 // Length 41 00 00 00 // "A\0" padded to word boundary However, this disadvantage can be overcome by storing such strings in a single block and accessing them using a :source:`Sming/Core/Data/CStringArray`.Note
Some counted string implementations store the length field at offset -4 so we always point to the actual data. In this case, that doesn’t offer any advantages as directly accessing the data is discouraged. Therefore, the length field always comes first.
Structure packing
When using Arrays with 64-bit types (including double) this is what we define:
struct { Array<int64_t> object; int64_t data[5]; } x;The object is a structure containing a single 32-bit value. Data arrays with uint8, uint16 or uint32 elements will start on the next 32-bit boundary, which is what we want.
With 64-bit values this is what the compiler does:
struct { Array<int64_t> object; uint32_t; // Packing added by compiler int64_t values[5]; } x;Which messes things up of course. Therefore Array classes and data are packed.
This is currently only an issue for Array types, but it also means that if you want to use Array with custom data structures then they should also be packed. That means you need to pay careful attention to member alignment and if packing is required then add it manually.
TODO List
- Missing methods
Behaviour of String should reflect WString. Review and add any missing methods. Note that implementations for some are likely non-trivial since we cannot assume the content will fit into RAM.
- Benchmark filemap
Compare SPIFFS vs. file map:
File access times
File transfer times
- Implement stream operator <<
For simpler printing. This is a large architectural decision as Sming doesn’t have any of this, neither it seems does Arduino although some libraries add support for it.
The advantage over Print/Printable is that support can be added using template functions without modifying the classes themselves. Formatting statements can be inserted to customise the output.
- Formatted print output
We have separate argument for Array printing, but if we want to customise the text for each item as well then we have to use a regular for-loop and handle the separators as well.
Easiest is probably to just add format arguments to printTo() method.
Simple example: print an array of integers in HEX.
Need to consider with text escaping, probably leave that for external libraries.
- Multi-dimensional arrays
This could be an array of structs, so operator[] returns an entire row. As an optimisation, could we define additional accessors for nested levels?
Alternatives:
Vector<Array>. Each array has to be defined before table.
Specialize operator for 2x2 case as it’s quite common. e.g. SubElementType operator[](unsigned row, unsigned col) const
- Type Information
The flashLength_ value can be redefined like this:
length: 20; elementSize: 3; ///< Number of bytes in each element, less one type: 5; ///< Enumeration identifying element type Char = 0, Signed, Unsigned, Float, Enum, String, Array, Vector, Map, UserDefined = 16 // Start of user-defined types- Variant Object
Interprets objects based on type information at runtime. This would allow complex data structures to be defined and used without knowing the object types. For example, translated JSON or XML documents.
- Translation tools
Say we wanted to access a read-only JSON document. How might we do that?
Easy. We write a Host application, so we can use all the existing libraries. Memory not an issue.
Load the JSON document using ArduinoJson
Serialize the document into FlashString-compatible declarations, producing a header file and a source file.
All we need to do then is write a JSON serializer to produce the appropriate structures.
About FlashString
FlashString was first introduced in Sming 3.7.0 on 17 November 2018.
I created it as a one-file solution to address these specific issues:
Regular Strings are really useful for passing around any kind of heap-allocated data, but it’s limited by available RAM.
The https://github.com/mikalhart/Flash library for Arduino provides a good approach, creating a wrapper class around the flash data. But I wanted it to be able to treat my FlashString classes as if they were just stored in flash so I wouldn’t have to wrap everything in a macro, like F() does.
Solution: The
FSTR::String(akaFlashString) class.Using PROGMEM directly for data is cumbersome, slow and fraught with danger: It’s inherently unsafe because of the alignment issues. Some smart cookie came up with a compiler patch so it could insert the correct instructions and thus avoid alignment exceptions. However, that still causes execution of inefficient code since the hardware requires we perform aligned accesses. Generating the structures correctly in the first place felt the best way forward.
Solution:
DEFINE_FSTR()Sharing flash data structures globally
Solution:
DECLARE_FSTR()How to get content into PROGMEM without having to manually convert everything into C structures. One solution to this is using external tools but that complicates the build process and felt un-necessary.
Solution:
IMPORT_FSTR()Relying on SPIFFS for serving fixed content is inefficient and problematic if/when the filesystem gets corrupted. I needed a solution which allowed large content to be served up without requiring a filesystem. The long term solution to this is, of course, a read-only filesystem but that is a complex thing indeed to do properly.
Solution:
FSTR::Stream(akaFlashMemoryStream) class.Embedded microsystems
Back in the golden era of the Commodore 64 and the ZX81, embedded really meant something. You had to get your hands dirty and poke about in the hardware.
With modern hardware it’s cheating. I mean, how on earth can a basic Android ‘App’ weigh in at 20MB? I cannot get my head around that.
Of course, it’s just economics - far cheaper to throw lots of hardware resources at a problem than actually build it right in the first place. With all that memory kicking about it’s just pure laziness. IMHO. And of course that trend will continue because it feeds the hardware manufacturers in this disposable world.
Not a fan.
Anyway, before I lose the plot my point was that as the term ‘embedded’ has been hijacked it needs at least some qualification.
So I’ll say that this library, and my other work, is focused on embedded microsystems (EMS).
Why bother?
The main limitation of the ESP8266 is probably its RAM. So why not just use an ESP32? It’s got lots more RAM.
But it’s still a finite resource, and I can think of far better ways to use it than filling it up with constant data.
QED. RAM will always be a scarce commodity in an EMS.
Thanks
This library may have never been written if not for ArduinoJson. Of specific interest is the way class templates are used, the high quality of documentation, ease of use and test-driven approach. I have attempted to emulate those particular attributes here.
Special thanks to Slavey Karadzhov for his support and assistance over the past year mentoring me through the process of upgrading the Sming embedded framework. I’ve learned a phenomenal amount about modern development practices and it’s really given my coding practices a kick up the arse. Awesome :-)
References
Used by
Sming (main) Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Hosted-Lib
This library provides replacement implementations for low-level operations which may then be accessed via the selected RPC interface.
See HostEd for further details.
References
Environment Variables
ENABLE_HOSTED_DIGITAL
ENABLE_HOSTED_WIRESoC support
host
HostEd
The hosted component allows Sming’s host emulator to run parts of the commands on an actual microcontroller. The communication is done via simplePRC and the microcontroller has to be flashed with a special application.
Overview
Sming’s host emulator allows easier debugging and development of embedded applications. This component named “Hosted” extends the host emulator and facilitates testing functionality that only a real microcontroller can provide as for example digital I/O operations or SPI operations.
For example in order to run the Basic_Blink application under the host emulator and run the actual blinking of a LED on a microcontroller we can compile the application using the following directives:
make SMING_ARCH=Host ENABLE_HOSTED=tcp HOSTED_SERVER_IP=192.168.4.1
SMING_ARCH=Hostinstructs the build system to build the application for the Host architecture.
ENABLE_HOSTED=tcpinstructs the host emulator to communicate with the real microcontroller using TCP.
HOSTED_SERVER_IP=192.168.4.1tells the host emulator where the server is running.We need to compile and flash also a special application on the desired microcontroller. This application will act as an RPC Server and will execute the commands from the host emulator on the microcontroller.
In the
samplesdirectory you will find the sample applications that will turn your microcontroller into an RPC server.The compilation and flashing for ESP32, for example, can be done using the following commands:
cd samples/tcp make SMING_ARCH=Esp32 WIFI_SSID=YourSSID WIFI_PWD=YourPassword make flashIf you replace
SMING_ARCH=Esp32withSMING_ARCH=Esp8266then the hosted application will be compiled and flashed on a ESP8266 microcontroller. Make sure to replace the values of WIFI_SSID and WIFI_PWD with the actual name and password for the Access Point (AP).Communication
At the moment the communication between an application running on the Host and the RCP server running on a microcontroller can be done using TCP or serial interface.
The
transportclasses are located underinclude/Hosted/Transport.Configuration
- ENABLE_HOSTED
Default: empty (disabled)
Enables the hosted component. Valid values for the moment are: -
tcpfor communication over TCP network. -serialfor communication over serial interface
- HOSTED_SERVER_IP
Default: 192.168.13.1
Used only when ENABLE_HOSTED=tcp is specified. Specifies the IP address of the remote RPC server.
- HOSTED_COM_PORT
Default:
COM_PORTUsed only when ENABLE_HOSTED=serial is specified. Specifies which local communication port should be used to connect to the remote RPC server.
- HOSTED_COM_SPEED
Default: 115200
Used only when ENABLE_HOSTED=serial is specified. Specifies the communication baud rate.
API Documentation
- namespace Hosted
Functions
- char convertType(const String &type)
Convert C type to format character See: https://docs.python.org/3.5/library/struct.html#format-strings
- String convertFQN(const String &name)
Converts a name as given from PRETTY_FUNCTION to internal format. Examples: void a::sub(int) -> a::sub(: i) void TwoWire::pins(uint8_t,uint8_t) -> TwoWire::pins(: B B) uint8_t digitalRead(uint16_t) -> digitalRead(B: H)
- Parameters:
name – source name
- Return values:
converted – name
Variables
- constexpr int COMMAND_NOT_FOUND = -1
- class Client : private simpleRPC::ParserCallbacks
Public Functions
- template<typename ...Args>
inline bool send(const String &functionName, Args... args)Method to send commands to the remote server.
If the command is overloaded, one command name with two or more different signatures then the name has to be containing the full function signature. Example: “void digitalWrite(uint16_t, uint8_t)”. The name with the signature MUST be the same as the one produced from PRETTY_FUNCTION -> https://gcc.gnu.org/onlinedocs/gcc/Function-Names.html
- Parameters:
functionName – Either the name or the name with the signature. Example: “digitalWrite” - will try to call the default digitalWrite function on the server
variable – arguments
- Return values:
true – on success, false if the command is not available
- template<typename R>
inline R wait()This method will block the execution until a message is detected.
- Return values:
HostedCommand –
- inline int getFunctionId(String name)
Fetches a list of commands supported on the RPC server and gives back the id of the desired command.
- Parameters:
name – command name to query
- Return values:
-1 – if not found. Otherwise the id of the function
- inline bool getRemoteCommands()
Gets list of remote command names and their ids.
- Return values:
true – on success, false otherwise
- class Serial : public Stream
Public Functions
- inline bool begin(uint32_t baud = 9600)
Initialise the serial port.
- Parameters:
baud – BAUD rate of the serial port (Default: 9600)
- inline virtual size_t write(uint8_t c) override
Writes a single character to output stream.
- Parameters:
c – Character to write to output stream
- Return values:
size_t – Quantity of characters written to output stream
- inline virtual size_t readBytes(char *buffer, size_t length) override
Read chars from stream into buffer.
Terminates if length characters have been read or timeout (see setTimeout). Returns the number of characters placed in the buffer (0 means no valid data found).
Note
Inherited classes may provide more efficient implementations without timeout.
- inline virtual size_t write(const uint8_t *buffer, size_t size)
Writes characters from a buffer to output stream.
- Parameters:
buffer – Pointer to character buffer
size – Quantity of characters to write
- Return values:
size_t – Quantity of characters written to stream
- namespace Transport
- class BaseTransport
Subclassed by Hosted::Transport::SerialTransport, Hosted::Transport::TcpTransport
- class SerialTransport : public Hosted::Transport::BaseTransport
- class TcpClientStream : public Stream
Public Functions
- inline virtual size_t readBytes(char *buffer, size_t length) override
Read chars from stream into buffer.
Terminates if length characters have been read or timeout (see setTimeout). Returns the number of characters placed in the buffer (0 means no valid data found).
Note
Inherited classes may provide more efficient implementations without timeout.
- inline virtual size_t write(const uint8_t *buffer, size_t size) override
Writes characters from a buffer to output stream.
- Parameters:
buffer – Pointer to character buffer
size – Quantity of characters to write
- Return values:
size_t – Quantity of characters written to stream
- inline virtual size_t write(uint8_t c) override
Writes a single character to output stream.
- Parameters:
c – Character to write to output stream
- Return values:
size_t – Quantity of characters written to output stream
- class TcpClientTransport : public Hosted::Transport::TcpTransport
- class TcpServerTransport : public Hosted::Transport::TcpTransport
- class TcpTransport : public Hosted::Transport::BaseTransport
Subclassed by Hosted::Transport::TcpClientTransport, Hosted::Transport::TcpServerTransport
References
Used by
Hosted-Lib Component
Hosted RCP Server over TCP Sample
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Installable File System
Created for Sming Framework Project August 2018 by mikee47
I struggled to find anything like this for embedded systems. Probably didn’t look hard enough, but it seemed like a fun thing to do so here we are.
The term ‘IFS’ came about because of the ‘I’ naming convention for virtual ‘Interface’ classes, hence IFileSystem. The term ‘installable’ is entirely appropriate because IFS allows file systems to be loaded and unloaded dynamically. Or maybe I just nicked the term from Microsoft :-)
Overview
IFS is written in C++ and has these core components:
- FileSystem API
File systems are implemented using the
IFS::IFileSystemvirtual class. This is, in essence, a single large ‘function table’ you will see in regular filesystem implementations.Class methods are similar to SPIFFS (which is POSIX-like).
Note
A single
Statstructure is used both for reading directory entries and for regularfileStat()operations.This differs from regular file APIs but is intended to simplify operation.
Applications will typically use
IFS::FileSysteminstead, which adds additional methods and overloads such as String parameter support. This used to implement the standard ‘flat’ Sming filesystem API, with a few minor changes and a number of additions.Two wrapper clases (
IFS::FileandIFS::Directory) are provided for applications to manage access to files and folders.- Firmware FileSystem (FWFS)
Files, directories and metadata are all stored as objects in read-only image. FWFS images are compact, fast to access and use very little RAM (approx. 240 bytes for file descriptors, etc.)
To support read/write data a writeable filesystem can be mounted in a sub-directory.
A python tool
fsbuildis used to build an FWFS image from user files. See Filesystem builder.This is integrated into the build system using the
fwfs-buildtarget for the partition. Example Hardware configuration fragment:"partitions": { "fwfs1": { "address": "0x280000", "size": "0x60000", "type": "data", "subtype": "fwfs", "filename": "out/fwfs1.bin", "build": { "target": "fwfs-build", // To build a FWFS image "config": "fsimage.fwfs" // Configuration for the image } } }Sming provides the Basic IFS sample application which gives a worked example of this.
Simple filesystem definitions can be defined in-situ, rather than using external file:
"build": { "target": "fwfs-build", // To build a FWFS image "config": { "name": "Simple filesystem", "source": { "/": "files" } } }The following basic IFS implementations are provided in this library:
IFS::FWFS::FileSystemFirmware Filesystem. It is designed to support all features of IFS, whereas other filesystems may only use a subset.
IFS::HYFS::FileSystemHybrid filesystem. Uses FWFS as the read-only root filesystem, with a writeable filesystem ‘layered’ on top.
When a file is opened for writing it is transparently copied to the SPIFFS partition so it can be updated. Wiping the SPIFFS partition reverts the filesystem to its original state.
Note that files marked as ‘read-only’ on the FWFS system are blocked from this behaviour.
IFS::Host::FileSystemFor Host architecture this allows access to the Linux/Windows host filesystem.
IFS::Gdb::FileSystemWhen running under a debugger this allows access to the Host filesystem. (Currently only works for ESP8266.)
IFS (and FWFS) has the following features:
- Attributes
Files have a standard set of attribute flags plus modification time and simple role-based access control list (ACL).
- Directories
Fully supported, and can be enumerated with associated file information using a standard opendir/readdir/closedir function set.
- User metadata
Supported for application use. The API for this is loosely based on Linux extended attributes (non-POSIX). Attributes are small chunks of data attached to files and directories, each identified by a numeric
IFS::AttributeTag.- Filesystem API
The Sming FileSystem functions are now wrappers around a single IFileSystem instance, which is provided by the application.
- Streaming classes
Sming provides IFS implementations for these so they can be constructed on any filesystem, not just the main (global) one.
- Dynamic loading
File systems may be loaded/created and unloaded/destroyed at runtime
- Multiple filesystems
Applications may use any supported filesystem, or write their own, or use any combination of existing filesystems to meet requirements. The API is the same.
- Mount points
FWFS is designed for use as a read-only root filing system, and supports mounting other filesystems in special directories.
FWFS
Many applications require a default, often fixed set of files. The easiest way is just to use SPIFFS. The problem is that power outages can corrupt a filesystem. For an embedded device that’s bad news. SPIFFS is also a bit overkill if you’re just storing configuration data, or it’s just for read-only use.
So what do you do if your filesystem gets wiped? Resetting a system back to a functional, default state can be tricky if the core user interface web files are gone. You could reformat and pull a standard set of files off a server somewhere. If your storage requirements are minimal, you could link the file data into your firmware as constant data blocks.
That’s kind of what FWFS does, but in a more structured and user-friendly way.
FWFS offers a more convenient solution by providing all your default files in a compact, fast, read-only format. Images can be mounted in separate partitions, linked into the program image itself or stored as files within another filesystem.
Note
This behaviour is supported by partitions (see Storage Management) using custom
Storage::Deviceobjects.Redirection
FWFS incorporates a redirector. This works by creating a mount point (a named object), which looks like an empty directory. When accessed, this get redirected to the root of another filesystem. The maximum number of mount points is fixed at compile time, but file systems can be mounted and dismounted at any time.
Mount points are identified explicitly in the build configuration file:
"mountpoints": { "path/to/use/spiffs": 0, "path/to/use/littlefs": 1 }The filesystem builder creates the MountPoint objects and tags them with the given volume indices. For example, the directory “path/to/use/littlefs” is attached to volume index #0.
Note
Unlike other filesystems you cannot use a regular directory as a mountpoint. To change the name of a mountpoint requires the filesystem image to be re-built and re-flashed.
Applications use the
IFileSystem::setVolume()method to install the actual filesystem.Streaming backup/archive support
The
IFS::FWFS::ArchiveStreamclass can be used to generate streaming filesystem backups from any supported filesystem. The archive files are in FWFS format.Here are some examples of how it can be used:
Stream filesystem (or directory) images directly to remote servers
Make local filesystem backups
Compact log files which don’t change much (think of ZIP files - just needs a compression plugin)
Backup entire filesystem a local file, an empty partition, etc.
Defragment/compact or repair a damaged filesystem by re-formatting then restoring from backup
The archiver has some additional features:
Specify whether to archive an entire filesystem or start from a specific directory
Specify whether to follow links (e.g. other filesystems in mountpoints) or not
Exclude any file or directory via custom callback (or by overriding methods)
Perform custom file data encoding such as compression or encryption via callbacks
Add additional metadata to files (comments, encryption codes, etc.)
See the Basic IFS sample for
Access Control
This came about because I wanted to secure down my ESP8266 web server applications so that only the basic index.html, stylesheets and accompanying javascript would be publicly accessibly. Everything else would require user authentication.
I also wanted to prevent certain users from accessing restricted files. Other users would also be able to edit files. So a simple role-based access control mechanism seemed appropriate.
Access control typically encapsulates two areas:
- Authentication
Is the user who they say they are? Usually performed by validating a username/password combination.
- Authorisation
What is the user permitted to do?
I’ll step aside for a brief word on security. Authentication is the weakest link because it’s exposed to public scrutiny. To avoid compromise authentication must only be done over a secured link. That means SSL.
If you have the option it’s usually best to put all your smart devices behind a secure proxy. The raspberry Pi is great for stuff like this. The Pi deals with keeping the public connection secure, and translates it into a regular HTTP connection for the ESP8266.
If you don’t have this option, but you need to connect your ESP8266 to the internet, use the SSL build for Sming.
Having done this, we don’t need to worry about encrypting passwords as the SSL layer will do that. We just need to make sure they’re good passwords.
In my applications authentication is done by matching username/password against the user database, stored in a JSON file. If successful, the session gets a token which appears in every subsequent request. The user database indicates a User Role, one of public, guest, user, manager or admin. IFS keeps an ‘Access Control List’ (ACL) for each file containing two entries (ACE), one for read access and another for write access. The ACE specifies the minimum assigned
IFS::UserRolerequired for access.This is probably as much as the filesystem needs to do. I can’t see that file ownership, inherited permissions or more finely-grained access permissions would be required, but having said that extending this system would probably be fairly straightforward.
Configuration filesystem
If an application only requires write access for configuration files, SPIFFS is overkill. These files would be updated very infrequently, so wear-levelling would be un-necessary. The names and number of files would probably also be known at build time, and an individual file could be limited to a fixed size, for example one or two flash sectors. A ConfigFileSystem implementation would not need to support file creation or deletion.
Such a system would require almost no static RAM allocation and code size would be tiny.
However, the LittleFS has excellent metadata support and is ideal for storing configuration information. This can be done using :IFS::FileSystem::setUserAttribute and read using :IFS::FileSystem::getUserAttribute or :IFS::FileSystem::enumAttributes.
Note
The ESP-IDF has a mechanism for flash-based configuration space via the
NVScomponent. It is robust and flexible but uses a significant amount of RAM for buffering which may preclude its use with the ESP8266.FWFS Objects
All files, directories and associated information elements are stored as ‘objects’. Files and directories are ‘named’ objects, which may contain other objects either directly or as references. Small objects (255 bytes or less) are stored directly, larger ones get their own file. Maximum object size is 16Mbytes.
File content is stored in un-named data objects. A named object can have any number of these and will be treated as a single entity for read/write operations. File ‘fragments’ do not need to be contiguous, and are reassembled during read operations.
Named objects can be enumerated using
IFS::IFileSystem::readdir(). Internally, FWFS uses handles to access any named object. Handles are allocated from a static pool to avoid excessive dynamic (heap) allocation. Users can attach their own data to any named object using custom object types.The filesystem layout is displayed during initial mount if this library is built with
DEBUG_VERBOSE_LEVEL= 3.Why FWFS?
There are many existing candidates for a read-only system, so why do we need another one? Here are some reasons:
SPIFFS and LittleFS could be used in read-only mode but they are not designed for space-efficiency. Images are therefore larger than necessary, sometimes considerably larger. This is also true of other such filesystems designed for Linux, etc.
FWFS is designed to produce the smallest possible images to conserve limited flash storage. It therefore has a high effective capacity, i.e. you can put a lot more in there than with other filesystems.
With ROMFS, for example, information is laid out with headers first, followed by data. The root directory and volume information are at the front.
FWFS works in reverse by writing out file contents first, then file headers and then directory records. The root directory comes at the end, followed by the volume information record. This allows images to be created as a stream because directory records can be efficiently constructed in RAM as each file or subdirectory record is written out. This keeps memory usage low.
In addition, checksums and additional metadata can be created while file data is written out. This could be required for compressing or encrypting the contents, or for error tolerance. For example, if corruption is encountered whilst reading file contents this can be noted in the metadata which is written out afterwards.
Filesystem images are therefore generated in a single pass, with each file or directory only read once.
Standard attribute support not well-suited to embedded microsystems.
The small set of standard metadata defined by IFS is designed to solve specific problems with typical IOT applications.
Code dependencies
Written initially for Sming, the library should be fairly portable to other systems.
No definitions from SPIFFS or other modules should be used in the public interface; such dependencies should be managed internally.
Applications should avoid using filesystem-dependent calls, structures or error codes. Such code, if necessary, should be placed into a separate module.
Implementation details
The traditional way to implement installable filing systems is using function tables, such as you’ll see in Linux. One reason is because the Linux kernel is written in C, not C++. For Sming, a virtual class seems the obvious choice, however there are some pros and cons.
- VMT
- Advantages
Compiler ensures correct ordering of methods, parameter type checking
Simpler coding
Extending and overriding is natural
- Function table
- Advantages
Portable to C applications (although with some fudging so are VMTs).
- Disadvantages
Care required to keep function order and parameters correct. Very likely we’d use a bunch of macros to deal with this.
Macros
We could #define the active filing system name which the FileSystem functions would map to the appropriate call. For example, fileOpen would get mapped to SPIFlashFileSystem_open(). We need to provide macros for defining file system functions.
- Advantages
Fast
- Disadvantages
Complicated
Prone to bugs
Not C++
Configuration variables
- FWFS_DEBUG
default: 0
Set to 1 to enable more detailed debugging information.
- ENABLE_FILE_SIZE64
default: disabled
Set to 1 to enable support for 64-bit files. This requires
ENABLE_STORAGE_SIZE64to be set.API
- namespace IFS
Return compression corresponding to given string
- param str:
- retval Compression::Type:
- Compression::Type getCompressionType(const char *str, Compression::Type defaultValue = Compression::Type::None)
- inline Compression::Type getCompressionType(const String &str, Compression::Type defaultValue = Compression::Type::None)
Return the access type value for the given string.
- param str:
- param defaultRole:
Returned if string isn’t recognsed
- retval UserRole:
Typedefs
- using AttributeEnumCallback = Delegate<bool(AttributeEnum &e)>
Return true to continue enumeration, false to stop.
- using ErrorCode = int
- using FileAttributes = BitSet<uint8_t, FileAttribute, size_t(FileAttribute::MAX)>
File attributes are stored as a bitmask.
- using DirHandle = struct ImplFileDir*
Enums
- enum class AttributeTag : uint16_t
Identifies a specific attribute.
Values:
- enumerator XX
- enumerator User
First user attribute.
- enum ControlCode
See
IFS::IFileSystem::fcontrolThese are weakly typed as values may be defined elsewhere.
Values:
- enumerator FCNTL_GET_MD5_HASH
Get stored MD5 hash for file.
FWFS calculates this when the filesystem image is built and can be used by applications to verify file integrity.
On success, returns size of the hash itself (16) If the file is zero-length then Error::NotFound will be returned as the hash is not stored for empty files.
- enumerator FCNTL_SET_VOLUME_LABEL
Set volume label.
- enumerator FCNTL_USER_BASE
Start of user-defined codes.
Codes before this are reserved for system use
Functions
- String getAclString(const IFS::ACL &acl)
Return a brief textual representation for an ACL Suitable for inclusion in a file listing.
- Parameters:
acl –
- Return values:
String –
- inline AttributeTag getUserAttributeTag(uint8_t value)
- size_t getAttributeSize(AttributeTag tag)
- String getFileAttributeString(FileAttributes attr)
Get the string representation for the given set of file attributes suitable for inclusion in a file listing.
- Parameters:
attr –
- Return values:
String –
- FileSystem *getDefaultFileSystem()
Framework should implement this method.
- FileSystem *createFirmwareFilesystem(Storage::Partition partition)
Create a firmware filesystem.
- Parameters:
partition –
- Return values:
FileSystem* – constructed filesystem object
- FileSystem *createHybridFilesystem(Storage::Partition fwfsPartition, IFileSystem *flashFileSystem)
Create a hybrid filesystem.
- Parameters:
fwfsPartition – Base read-only filesystem partition
flashFileSystem – The filesystem to use for writing
- Return values:
FileSystem* – constructed filesystem object
- FileSystem *mountArchive(FileSystem &fs, const String &filename)
Mount an FWFS archive.
- Parameters:
fs – Filesystem where file is located
filename – Name of archive file
- Return values:
FileSystem* – constructed filesystem object
- time_t fsGetTimeUTC()
Get current timestamp in UTC.
Note
Filing systems must store timestamps in UTC Use this function; makes porting easier.
- Return values:
time_t –
- bool isRootPath(const char *&path)
Check if path is root directory.
Paths equal to “/” or “” are empty and considered equivalent to nullptr. Methods or functions can use this macro to resolve these for simpler parsing.
- Parameters:
Path – to check, set to nullptr if it’s the root directory
- Return values:
bool – true if path is root directory
- FileSystem *createFatFilesystem(Storage::Partition partition)
Create a FAT filesystem.
- Parameters:
partition –
- Return values:
FileSystem* – constructed filesystem object
- class DirectoryTemplate : public SectionTemplate
Directory stream class.
Subclassed by IFS::HtmlDirectoryTemplate, IFS::JsonDirectoryTemplate
Public Functions
- inline virtual bool nextRecord() override
Move to next record.
- Return values:
bool – true to emit section, false to skip
- class FileStream : public IFS::FsBase, public ReadWriteStream
File stream class.
Subclassed by FileStream, GdbFileStream, HostFileStream
Public Functions
- void attach(FileHandle file, size_t size)
Attach this stream object to an open file handle.
- Parameters:
file –
size –
- bool open(const String &fileName, IFS::OpenFlags openFlags = OpenFlag::Read)
Open a file by path, and attach this stream object to it.
Note
call getLastError() to determine cause of failure
- Parameters:
fileName – Full path to file
openFlags –
- Return values:
bool – true on success, false on error
- void close()
Close file.
- inline virtual StreamType getStreamType() const override
Get the stream type.
- Return values:
StreamType – The stream type.
- virtual size_t write(const uint8_t *buffer, size_t size) override
Write chars to stream.
Note
Although this is defined in the Print class, ReadWriteStream uses this as the core output method so descendants are required to implement it
- Parameters:
buffer – Pointer to buffer to write to the stream
size – Quantity of chars to write
- Return values:
size_t – Quantity of chars written to stream
- inline virtual int read() override
Read one character and moves the stream pointer.
- Return values:
The – character that was read or -1 if none is available
- virtual size_t readBytes(char *buffer, size_t length) override
Read chars from stream into buffer.
Terminates if length characters have been read or timeout (see setTimeout). Returns the number of characters placed in the buffer (0 means no valid data found).
Note
Inherited classes may provide more efficient implementations without timeout.
- virtual uint16_t readMemoryBlock(char *data, int bufSize) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
- virtual int seekFrom(int offset, SeekOrigin origin) override
Change position in stream.
Note
This method is implemented by streams which support random seeking, such as files and memory streams.
- Parameters:
offset –
origin –
- Return values:
New – position, < 0 on error
- inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
- String fileName() const
Filename of file stream is attached to.
- Return values:
String – invalid if stream isn’t open
- inline bool fileExist() const
Determine if file exists.
- Return values:
bool – true if stream contains valid file
- inline virtual String getName() const override
Returns name of the resource.
Note
Commonly used to obtain name of file
- Return values:
String –
- virtual MimeType getMimeType() const override
Get MIME type for stream content.
- Return values:
MimeType –
- inline virtual bool isValid() const override
Determine if the stream object contains valid data.
Note
Where inherited classes are initialised by constructor this method indicates whether that was successful or not (e.g. FileStream)
- Return values:
bool – true if valid, false if invalid
- inline size_t getPos() const
Get the offset of cursor from beginning of data.
- Return values:
size_t – Cursor offset
- inline virtual int available() override
Return the maximum bytes available to read, from current position.
- Return values:
int – -1 is returned when the size cannot be determined
- virtual String id() const override
Returns unique id of the resource.
- Return values:
String – the unique id of the stream.
- bool truncate(size_t newSize)
Reduce the file size.
- Parameters:
newSize –
- Return values:
bool – true on success
- inline bool truncate()
Truncate file at current position.
- Return values:
bool – true on success
- class HtmlDirectoryTemplate : public IFS::DirectoryTemplate
Read-only stream access to directory listing with HTML output.
- class JsonDirectoryTemplate : public IFS::DirectoryTemplate
Read-only stream providing directory listing in JSON format.
- struct ACL
- struct AttributeEnum
Attribute information passed to enumeration callback.
- struct Compression
A compression descriptor.
- class Directory : public IFS::FsBase
Wrapper class for enumerating a directory.
Public Functions
- bool open(const String &dirName = nullptr)
Open a directory and attach this stream object to it.
Note
call getLastError() to determine cause of failure
- Parameters:
dirName – Default is root directory
- Return values:
bool – true on success, false on error
- void close()
Close directory.
- bool rewind()
Rewind directory stream to start so it can be re-enumerated.
Note
call getLastError() to determine cause of failure
- Return values:
bool – true on success, false on error
- inline const String &getDirName() const
Name of directory stream is attached to.
- Return values:
String – invalid if stream isn’t open
- inline bool dirExist() const
Determine if directory exists.
- Return values:
bool – true if stream is attached to a directory
- struct Extent
Defines the location of a contiguous run of file data.
e.g. offset: 0 length: 251 skip: 5 repeat: 30 Describes 30 blocks of data each of 251 bytes with a 5-byte gap between them. Thus run contains 7530 bytes of file data, consuming 7680 bytes of storage.
skip/repeat is necessary for SPIFFS to keep RAM usage sensible. Thus, a sample 267kByte application image has 1091 extents, but requires perhaps only 60 entries.
- class File : public IFS::FsBase
Wraps up all file access methods.
Common flag combinations
Public Functions
- inline bool stat(Stat &stat)
get file information
- Parameters:
stat – structure to return information in, may be null
- Return values:
bool – true on success
- inline int control(ControlCode code, void *buffer, size_t bufSize)
Low-level and non-standard file control operations.
To simplify usage the same buffer is used for both input and output. Only the size of the buffer is provided. If a specific FCNTL code requires more information then it will be contained within the provided data.
- Parameters:
code – FCNTL_XXX code
buffer – Input/Output buffer
bufSize – Size of buffer
- Return values:
int – error code or, on success, data size
- template<typename T>
inline bool open(const T &path, OpenFlags flags = OpenFlag::Read)open a file by name/path
- Parameters:
path – full path to file
flags – opens for opening file
- Return values:
bool – true on success
- inline bool close()
close an open file
- Return values:
bool – true on success
- inline int read(void *data, size_t size)
read content from a file and advance cursor
- Parameters:
data – buffer to write into
size – size of file buffer, maximum number of bytes to read
- Return values:
int – number of bytes read or error code
- inline int write(const void *data, size_t size)
write content to a file at current position and advance cursor
- Parameters:
data – buffer to read from
size – number of bytes to write
- Return values:
int – number of bytes written or error code
- inline file_offset_t seek(file_offset_t offset, SeekOrigin origin)
change file read/write position
- Parameters:
offset – position relative to origin
origin – where to seek from (start/end or current position)
- Return values:
int – current position or error code
- inline bool eof()
determine if current file position is at end of file
- Return values:
bool – true if at EOF or file is invalid
- inline file_offset_t tell()
get current file position
- Return values:
int32_t – current position relative to start of file, or error code
- inline bool truncate(file_size_t new_size)
Truncate (reduce) the size of an open file.
- Parameters:
newSize –
- Return values:
bool – true on success
- inline bool truncate()
Truncate an open file at the current cursor position.
- Return values:
bool – true on success
- inline bool flush()
flush any buffered data to physical media
- Return values:
bool – true on success
- inline bool setacl(const ACL &acl)
Set access control information for file.
- Parameters:
acl –
- Return values:
bool – true on success
- inline bool settime(time_t mtime)
Set modification time for file.
Note
any subsequent writes to file will reset this to current time
- Return values:
bool – true on success
- inline bool setcompression(const Compression &compression)
Set file compression information.
- Parameters:
compression –
- Return values:
bool – true on success
- inline bool remove()
remove (delete) an open file (and close it)
- Return values:
bool – true on success
- inline file_size_t getSize()
Get size of file.
- Return values:
uint32_t – Size of file in bytes, 0 on error
- inline file_offset_t readContent(size_t size, const ReadContentCallback &callback)
Read from current file position and invoke callback for each block read.
- Parameters:
size – Maximum number of bytes to read
callback –
- Return values:
int – Number of bytes processed, or error code
- inline file_offset_t readContent(const ReadContentCallback &callback)
Read from current file position to end of file and invoke callback for each block read.
- Parameters:
callback –
- Return values:
file_offset_t – Number of bytes processed, or error code
- inline String getContent()
Read content of the file, from current position.
Note
After calling this function the content of the file is placed in to a string. The result will be an invalid String (equates to
false) if the file could not be read. If the file exists, but is empty, the result will be an empty string “”.
- Return values:
String – String variable in to which to read the file content
- inline FileHandle release()
Return current file handle and release ownership.
- class FileCopier
Class to manage copying of files and directories including attributes.
Public Types
- class FileSystem : public IFS::IFileSystem
Installable File System base class.
Adds additional methods to ease use over base IFileSystem.
rename a file
- param oldpath:
- param newpath:
- retval int:
error code
- using ReadContentCallback = Delegate<int(const char *buffer, size_t size)>
Callback for readContent method.
- Param buffer:
- Param size:
- Retval int:
Return number of bytes consumed, < size to stop If < 0 then this is returned as error code to
readContentcall.
- inline int remove(const String &path)
remove (delete) a file by path
- Parameters:
path –
- Return values:
int – error code
- template<typename T>
inline int setacl(const T &file, const ACL &acl)Set access control information for file.
- Parameters:
file – handle or path to file
acl –
- Return values:
int – error code
- template<typename T>
inline int setattr(const T &file, FileAttributes attr)Set file attributes.
- Parameters:
file – handle or path to file
attr –
- Return values:
int – error code
- template<typename T>
inline int settime(const T &file, time_t mtime)Set modification time for file.
Note
any subsequent writes to file will reset this to current time
- Parameters:
file – handle or path to file
- Return values:
int – error code
- template<typename T>
inline int setcompression(const T &file, const Compression &compression)Set file compression information.
- Parameters:
file –
compression –
- Return values:
int – error code
- file_size_t getSize(FileHandle file)
Get size of file.
- Parameters:
file – File handle
- Return values:
file_size_t – Size of file in bytes, 0 on error
- file_size_t getSize(const char *fileName)
Get size of file.
- Parameters:
fileName – Name of file
- Return values:
file_size_t – Size of file in bytes, 0 on error
- file_offset_t readContent(FileHandle file, size_t size, const ReadContentCallback &callback)
Read from current file position and invoke callback for each block read.
- Parameters:
file –
size – Maximum number of bytes to read
callback –
- Return values:
file_offset_t – Number of bytes processed, or error code
- file_offset_t readContent(FileHandle file, const ReadContentCallback &callback)
Read from current file position to end of file and invoke callback for each block read.
- Parameters:
file –
callback –
- Return values:
file_offset_t – Number of bytes processed, or error code
- file_offset_t readContent(const String &filename, const ReadContentCallback &callback)
Read entire file content in blocks, invoking callback after every read.
- Parameters:
filename –
callback –
- Return values:
file_offset_t – Number of bytes processed, or error code
- int remove(const char *path) = 0
remove (delete) a file by path
- Parameters:
path –
- Return values:
int – error code
Read content of a file
After calling this function the content of the file is placed in to a c-string Ensure there is sufficient space in the buffer for file content plus extra trailing null, i.e. at least bufSize + 1 Always check the return value!
- param fileName:
Name of file to read from
- param buffer:
Pointer to a character buffer in to which to read the file content
- param bufSize:
Quantity of bytes to read from file
- retval size_t:
Quantity of bytes read from file or zero on failure
Returns 0 if the file could not be read
Create or replace file with defined content
This function creates a new file or replaces an existing file and populates the file with the content of a c-string buffer.
- param fileName:
Name of file to create or replace
- param content:
Pointer to c-string containing content to populate file with
- retval int:
Number of bytes transferred or error code
- param length:
(optional) number of characters to write
Public Functions
- int makedirs(const char *path)
Create a directory and any intermediate directories if they do not already exist.
- Parameters:
path – Path to directory. If no trailing ‘/’ is present the final element is considered a filename.
- Return values:
int – error code
- int truncate(const char *fileName, file_size_t newSize)
Truncate a file to a specific size.
- Parameters:
fileName – File to truncate
- Return values:
int – new file size, or error code
- String getContent(const String &fileName)
Read content of a file.
Note
After calling this function the content of the file is placed in to a string. The result will be an invalid String (equates to
false) if the file could not be read. If the file exists, but is empty, the result will be an empty string “”.
- Parameters:
fileName – Name of file to read from
- Return values:
String – String variable in to which to read the file content
- int opendir(const char *path, DirHandle &dir) = 0
open a directory for reading
- Parameters:
path – path to directory. nullptr is interpreted as root directory
dir – returns a pointer to the directory object
- Return values:
int – error code
- int mkdir(const char *path) = 0
Create a directory.
Only the final directory in the path is guaranteed to be created. Usually, this call will fail if intermediate directories are not present. Use
IFS::FileSystem::makedirs()for this purpose.
- Parameters:
path – Path to directory
- Return values:
int – error code
- int stat(const char *path, Stat *stat) = 0
get file information
Returned stat will indicate whether the path is a mountpoint or directory. For a mount point, stats for the root directory of the mounted filesystem must be obtained by opening a handle then using
fstat:int handle = fs.open("path-to-mountpoint"); Stat stat; fs.fstat(handle, &stat); fs.close(handle);
- Parameters:
path – name or path of file/directory/mountpoint
s – structure to return information in, may be null to do a simple file existence check
- Return values:
int – error code
- int fstat(FileHandle file, Stat *stat) = 0
get file information
- Parameters:
file – handle to open file
stat – structure to return information in, may be null
- Return values:
int – error code
- FileHandle open(const char *path, OpenFlags flags) = 0
open a file (or directory) by path
This function may also be used to obtain a directory handle to perform various operations such as enumerating attributes. Calls to read or write on such handles will typically fail.
- Parameters:
path – full path to file
flags – Desired access and other options
- Return values:
FileHandle – file handle or error code
- int ftruncate(FileHandle file, file_size_t new_size) = 0
Truncate (reduce) the size of an open file.
Note
In POSIX
ftruncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize
- Parameters:
file – Open file handle, must have Write access
newSize –
- Return values:
int – Error code
- int rename(const char *oldpath, const char *newpath) = 0
rename a file
- Parameters:
oldpath –
newpath –
- Return values:
int – error code
- class FsBase
Subclassed by IFS::Directory, IFS::FWFS::ArchiveStream, IFS::File, IFS::FileStream
Public Functions
- inline int getLastError()
determine if an error occurred during operation
- Return values:
int – filesystem error code
- class IFileSystem
Installable File System base class.
Construction and initialisation of a filing system is implementation-dependent so there are no methods here for that.
Methods are defined as virtual abstract unless we actually have a default base implementation. Whilst some methods could just return Error::NotImplemented by default, keeping them abstract forces all file system implementations to consider them so provides an extra check for completeness.
Note
The ‘I’ implies Installable but could be for Interface :-)
Subclassed by IFS::FAT::FileSystem, IFS::FWFS::FileSystem, IFS::FileSystem, IFS::Gdb::FileSystem, IFS::HYFS::FileSystem, IFS::Host::FileSystem
Public Functions
- inline virtual ~IFileSystem()
Filing system implementations should dismount and cleanup here.
- virtual int mount() = 0
Mount file system, performing any required initialisation.
- Return values:
error – code
- virtual int getinfo(Info &info) = 0
get filing system information
- Parameters:
info – structure to read information into
- Return values:
int – error code
- inline virtual int setProfiler(IProfiler *profiler)
Set profiler instance to enable debugging and performance assessment.
- Parameters:
profiler –
- Return values:
int – error code - profiling may not be supported on all filesystems
- inline virtual String getErrorString(int err)
get the text for a returned error code
- Return values:
String –
- inline virtual int setVolume(uint8_t index, IFileSystem *fileSystem)
Set volume for mountpoint.
- Parameters:
index – Volume index
fileSystem – The filesystem to root at this mountpoint
- Return values:
int – error code
- virtual int opendir(const char *path, DirHandle &dir) = 0
open a directory for reading
- Parameters:
path – path to directory. nullptr is interpreted as root directory
dir – returns a pointer to the directory object
- Return values:
int – error code
- virtual int readdir(DirHandle dir, Stat &stat) = 0
read a directory entry
Note
File system allocates entries structure as it usually needs to track other information. It releases memory when closedir() is called.
- Parameters:
dir –
stat –
- Return values:
int – error code
- virtual int rewinddir(DirHandle dir) = 0
Reset directory read position to start.
- Parameters:
dir –
- Return values:
int – error code
- virtual int closedir(DirHandle dir) = 0
close a directory object
- Parameters:
dir – directory to close
- Return values:
int – error code
- virtual int mkdir(const char *path) = 0
Create a directory.
Only the final directory in the path is guaranteed to be created. Usually, this call will fail if intermediate directories are not present. Use
IFS::FileSystem::makedirs()for this purpose.
- Parameters:
path – Path to directory
- Return values:
int – error code
- virtual int stat(const char *path, Stat *stat) = 0
get file information
Returned stat will indicate whether the path is a mountpoint or directory. For a mount point, stats for the root directory of the mounted filesystem must be obtained by opening a handle then using
fstat:int handle = fs.open("path-to-mountpoint"); Stat stat; fs.fstat(handle, &stat); fs.close(handle);
- Parameters:
path – name or path of file/directory/mountpoint
s – structure to return information in, may be null to do a simple file existence check
- Return values:
int – error code
- virtual int fstat(FileHandle file, Stat *stat) = 0
get file information
- Parameters:
file – handle to open file
stat – structure to return information in, may be null
- Return values:
int – error code
- inline virtual int fcontrol(FileHandle file, ControlCode code, void *buffer, size_t bufSize)
Low-level and non-standard file control operations.
To simplify usage the same buffer is used for both input and output. Only the size of the buffer is provided. If a specific FCNTL code requires more information then it will be contained within the provided data.
- Parameters:
file –
code – FCNTL_XXX code
buffer – Input/Output buffer
bufSize – Size of buffer
- Return values:
int – error code or, on success, data size
- virtual FileHandle open(const char *path, OpenFlags flags) = 0
open a file (or directory) by path
This function may also be used to obtain a directory handle to perform various operations such as enumerating attributes. Calls to read or write on such handles will typically fail.
- Parameters:
path – full path to file
flags – Desired access and other options
- Return values:
FileHandle – file handle or error code
- virtual int close(FileHandle file) = 0
close an open file
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int read(FileHandle file, void *data, size_t size) = 0
read content from a file and advance cursor
- Parameters:
file – handle to open file
data – buffer to write into
size – size of file buffer, maximum number of bytes to read
- Return values:
int – number of bytes read or error code
- virtual int write(FileHandle file, const void *data, size_t size) = 0
write content to a file at current position and advance cursor
- Parameters:
file – handle to open file
data – buffer to read from
size – number of bytes to write
- Return values:
int – number of bytes written or error code
- virtual file_offset_t lseek(FileHandle file, file_offset_t offset, SeekOrigin origin) = 0
change file read/write position
- Parameters:
file – handle to open file
offset – position relative to origin
origin – where to seek from (start/end or current position)
- Return values:
file_offset_t – current position or error code
- virtual int eof(FileHandle file) = 0
determine if current file position is at end of file
- Parameters:
file – handle to open file
- Return values:
int – 0 - not EOF, > 0 - at EOF, < 0 - error
- virtual file_offset_t tell(FileHandle file) = 0
get current file position
- Parameters:
file – handle to open file
- Return values:
file_offset_t – current position relative to start of file, or error code
- virtual int ftruncate(FileHandle file, file_size_t new_size) = 0
Truncate (reduce) the size of an open file.
Note
In POSIX
ftruncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize
- Parameters:
file – Open file handle, must have Write access
newSize –
- Return values:
int – Error code
- virtual int flush(FileHandle file) = 0
flush any buffered data to physical media
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int fsetxattr(FileHandle file, AttributeTag tag, const void *data, size_t size) = 0
Set an extended attribute on an open file.
- Parameters:
file – handle to open file
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int fgetxattr(FileHandle file, AttributeTag tag, void *buffer, size_t size) = 0
Get an extended attribute from an open file.
- Parameters:
file – handle to open file
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual int fenumxattr(FileHandle file, AttributeEnumCallback callback, void *buffer, size_t bufsize) = 0
Enumerate attributes.
- Parameters:
file – handle to open file
callback – Callback function to invoke for each attribute found
buffer – Buffer to use for reading attribute data. Use nullptr if only tags are required
bufsize – Size of buffer
- Return values:
int – error code, on success returns number of attributes read
- virtual int setxattr(const char *path, AttributeTag tag, const void *data, size_t size) = 0
Set an extended attribute for a file given its path.
- Parameters:
path – Full path to file (or directory)
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int getxattr(const char *path, AttributeTag tag, void *buffer, size_t size) = 0
Get an attribute from a file given its path.
- Parameters:
file – Full path to file (or directory)
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- inline virtual int fgetextents(FileHandle file, Storage::Partition *part, Extent *list, uint16_t extcount)
Get extents for a file.
- Parameters:
file – Handle to open file
part – Partition where the file lives (OUT, OPTIONAL)
list – Buffer for extents (OPTIONAL)
extcount – Maximum number of extents to return in
list- Return values:
int – Total number of extents for file (may be larger than ‘extcount’), or error code
- virtual int rename(const char *oldpath, const char *newpath) = 0
rename a file
- Parameters:
oldpath –
newpath –
- Return values:
int – error code
- virtual int remove(const char *path) = 0
remove (delete) a file by path
- Parameters:
path –
- Return values:
int – error code
- virtual int fremove(FileHandle file) = 0
remove (delete) a file by handle
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int format() = 0
format the filing system
Note
this does a default format, returning file system to a fresh state The filing system implementation may define more specialised methods which can be called directly.
- Return values:
int – error code
- inline virtual int check()
Perform a file system consistency check.
Note
if possible, issues should be resolved. Returns 0 if file system checked out OK. Otherwise there were issues: < 0 for unrecoverable errors,
0 for recoverable errors.
- Return values:
int – error code
- struct Info
Basic information about filing system.
Subclassed by IFS::IFileSystem::NameInfo
Public Members
- Type type = {}
The filing system type identifier.
- Attributes attr = {}
Attribute flags.
- size_t maxNameLength = {255}
Maximum length of a single file name.
- size_t maxPathLength = {255}
Maximum length of a full file path.
- uint32_t volumeID = {0}
Unique identifier for volume.
- NameBuffer name
Buffer for name.
- volume_size_t volumeSize = {0}
Size of volume, in bytes.
- volume_size_t freeSpace = {0}
Available space, in bytes.
- struct NameInfo : public IFS::IFileSystem::Info
Filing system information with buffer for name.
- struct NameBuffer
defines a ‘safe’ name buffer
Note
Instead of including a fixed name array in Stat (and IFileSystem::Info) structures, we use a NameBuffer to identify a separate buffer. This has several advantages:
There are fancier ways to do this but a structure is transparent and requires no heap allocation.- Maximum size is not fixed - Finding and copying the name is optional - Actual name length is returned in the 'length' field, regardless of size - A NameBuffer structure (or one containing it) only requires initialising once before a loop operation as buffer/size are preserved.Note
lengthalways reflects the required name/path length, and may be longer than size.Subclassed by IFS::FileNameBuffer
Public Functions
- inline NameBuffer(String &s)
Make a NameBuffer point to contents of a String.
- inline int copy(const char *src, uint16_t srclen)
copies text from a source buffer into a name buffer
Note
length field is always set to srclen, regardless of number of characters copied.
- Parameters:
src – source name
srclen – number of characters in name
- inline int addSep()
When building file paths this method simplified appending separators.
Note
if the path is not empty, a separator character is appended
- Return values:
int – error code
- inline char *endptr()
get a pointer to the next write position
Note
use space() to ensure buffer doesn’t overrun When writing text be sure to call terminate() when complete
- Return values:
char* –
- inline uint16_t space()
get the number of free characters available
Note
returns 0 if buffer has overrun
- Return values:
uint16_t –
- inline void terminate()
ensure the buffer has a nul terminator, even if it means overwriting content
- inline bool overflow() const
determine if name buffer overflowed
Note
Compares returned length with buffer size; A nul terminator is always appended, so size should be >= (length + 1)
- struct FileNameBuffer : public IFS::NameBuffer
a quick’n’dirty name buffer with maximum path allocation
- class IProfiler
Filesystems may optionally provide performance statistics.
Subclassed by IFS::Profiler
Public Functions
- virtual void erase(storage_size_t address, size_t size) = 0
Called BEFORE an erase operation.
- class Profiler : public IFS::IProfiler
Public Functions
- inline virtual void erase(storage_size_t address, size_t size) override
Called BEFORE an erase operation.
- struct Stat
- struct Stat
File Status structure.
Subclassed by IFS::NameStat
Public Functions
- inline Stat &operator=(const Stat &rhs)
assign content from another Stat structure
Note
All fields are copied as for a normal assignment, except for ‘name’, where rhs.name contents are copied into our name buffer.
- inline bool isDir() const
Is this a directory (or mountpoint) ?
Public Members
- IFileSystem *fs = {nullptr}
The filing system owning this file.
- NameBuffer name
Name of file.
- file_size_t size = {0}
Size of file in bytes.
- struct NameStat : public IFS::Stat
version of Stat with integrated name buffer
Note
provide for convenience
- struct TimeStamp
Manage IFS timestamps stored as an unsigned 32-bit value.
An unsigned 32-bit value containing seconds will overflow in about 136 years. time_t starts 1 Jan 1970 (Unix epoch).
Times are stored in UTC (GMT).
- namespace Debug
Functions
- void printFsInfo(Print &out, FileSystem &fs)
- void printAttrInfo(Print &out, FileSystem &fs, const String &filename)
- int listDirectory(Print &out, FileSystem &fs, const String &path, Options options = 0)
- namespace Error
Functions
- namespace FAT
Functions
- int translateFatfsResult(uint8_t result, bool diskio_write)
- ErrorCode calculateFatParam(Partition partition, const FormatOptions &opt, FatParam ¶m)
Deduce FAT volume parameters for given space.
- Parameters:
partition – The partition to format
opt – Formatting options
param – On success, contains calculated parameters for FAT volume
- Return values:
ErrorCode – When partitioning using MBR format, this method can be used to determine the
Sys indicatorvalue setting.
- ErrorCode formatVolume(Partition partition, const FatParam ¶m)
Format partition using pre-calculated FAT parameters.
- Parameters:
partition – The partition to format
param – Detailed FAT parameters (returned from
calculateFatParam)- Return values:
ErrorCode – This function allows fine control over exactly how a FAT partition is constructed. Generally the
calculateFatParamfunction should be used to populate theparamstructure, then any modifications can be made as required before actually formatting the volume.
- inline ErrorCode formatVolume(Partition partition, const FormatOptions &opt = {})
Format partition with a blank FAT volume.
- Parameters:
partition – The partition to format
opt – Formatting options
- Return values:
ErrorCode –
- class FileSystem : public IFS::IFileSystem
Wraps fatfs
Public Functions
- virtual int mount() override
Mount file system, performing any required initialisation.
- Return values:
error – code
- virtual int getinfo(Info &info) override
get filing system information
- Parameters:
info – structure to read information into
- Return values:
int – error code
- virtual int setProfiler(IProfiler *profiler) override
Set profiler instance to enable debugging and performance assessment.
- Parameters:
profiler –
- Return values:
int – error code - profiling may not be supported on all filesystems
- virtual String getErrorString(int err) override
get the text for a returned error code
- Return values:
String –
- virtual int opendir(const char *path, DirHandle &dir) override
open a directory for reading
- Parameters:
path – path to directory. nullptr is interpreted as root directory
dir – returns a pointer to the directory object
- Return values:
int – error code
- virtual int readdir(DirHandle dir, Stat &stat) override
read a directory entry
Note
File system allocates entries structure as it usually needs to track other information. It releases memory when closedir() is called.
- Parameters:
dir –
stat –
- Return values:
int – error code
- virtual int rewinddir(DirHandle dir) override
Reset directory read position to start.
- Parameters:
dir –
- Return values:
int – error code
- virtual int closedir(DirHandle dir) override
close a directory object
- Parameters:
dir – directory to close
- Return values:
int – error code
- virtual int mkdir(const char *path) override
Create a directory.
Only the final directory in the path is guaranteed to be created. Usually, this call will fail if intermediate directories are not present. Use
IFS::FileSystem::makedirs()for this purpose.
- Parameters:
path – Path to directory
- Return values:
int – error code
- virtual int stat(const char *path, Stat *stat) override
get file information
Returned stat will indicate whether the path is a mountpoint or directory. For a mount point, stats for the root directory of the mounted filesystem must be obtained by opening a handle then using
fstat:int handle = fs.open("path-to-mountpoint"); Stat stat; fs.fstat(handle, &stat); fs.close(handle);
- Parameters:
path – name or path of file/directory/mountpoint
s – structure to return information in, may be null to do a simple file existence check
- Return values:
int – error code
- virtual int fstat(FileHandle file, Stat *stat) override
get file information
- Parameters:
file – handle to open file
stat – structure to return information in, may be null
- Return values:
int – error code
- virtual int fcontrol(FileHandle file, ControlCode code, void *buffer, size_t bufSize) override
Low-level and non-standard file control operations.
To simplify usage the same buffer is used for both input and output. Only the size of the buffer is provided. If a specific FCNTL code requires more information then it will be contained within the provided data.
- Parameters:
file –
code – FCNTL_XXX code
buffer – Input/Output buffer
bufSize – Size of buffer
- Return values:
int – error code or, on success, data size
- virtual int fsetxattr(FileHandle file, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute on an open file.
- Parameters:
file – handle to open file
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int fgetxattr(FileHandle file, AttributeTag tag, void *buffer, size_t size) override
Get an extended attribute from an open file.
- Parameters:
file – handle to open file
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual int fenumxattr(FileHandle file, AttributeEnumCallback callback, void *buffer, size_t bufsize) override
Enumerate attributes.
- Parameters:
file – handle to open file
callback – Callback function to invoke for each attribute found
buffer – Buffer to use for reading attribute data. Use nullptr if only tags are required
bufsize – Size of buffer
- Return values:
int – error code, on success returns number of attributes read
- virtual int setxattr(const char *path, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute for a file given its path.
- Parameters:
path – Full path to file (or directory)
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int getxattr(const char *path, AttributeTag tag, void *buffer, size_t size) override
Get an attribute from a file given its path.
- Parameters:
file – Full path to file (or directory)
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual FileHandle open(const char *path, OpenFlags flags) override
open a file (or directory) by path
This function may also be used to obtain a directory handle to perform various operations such as enumerating attributes. Calls to read or write on such handles will typically fail.
- Parameters:
path – full path to file
flags – Desired access and other options
- Return values:
FileHandle – file handle or error code
- virtual int close(FileHandle file) override
close an open file
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int read(FileHandle file, void *data, size_t size) override
read content from a file and advance cursor
- Parameters:
file – handle to open file
data – buffer to write into
size – size of file buffer, maximum number of bytes to read
- Return values:
int – number of bytes read or error code
- virtual int write(FileHandle file, const void *data, size_t size) override
write content to a file at current position and advance cursor
- Parameters:
file – handle to open file
data – buffer to read from
size – number of bytes to write
- Return values:
int – number of bytes written or error code
- virtual file_offset_t lseek(FileHandle file, file_offset_t offset, SeekOrigin origin) override
change file read/write position
- Parameters:
file – handle to open file
offset – position relative to origin
origin – where to seek from (start/end or current position)
- Return values:
file_offset_t – current position or error code
- virtual int eof(FileHandle file) override
determine if current file position is at end of file
- Parameters:
file – handle to open file
- Return values:
int – 0 - not EOF, > 0 - at EOF, < 0 - error
- virtual file_offset_t tell(FileHandle file) override
get current file position
- Parameters:
file – handle to open file
- Return values:
file_offset_t – current position relative to start of file, or error code
- virtual int ftruncate(FileHandle file, file_size_t new_size) override
Truncate (reduce) the size of an open file.
Note
In POSIX
ftruncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize
- Parameters:
file – Open file handle, must have Write access
newSize –
- Return values:
int – Error code
- virtual int flush(FileHandle file) override
flush any buffered data to physical media
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int rename(const char *oldpath, const char *newpath) override
rename a file
- Parameters:
oldpath –
newpath –
- Return values:
int – error code
- virtual int remove(const char *path) override
remove (delete) a file by path
- Parameters:
path –
- Return values:
int – error code
- virtual int fremove(FileHandle file) override
remove (delete) a file by handle
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int format() override
format the filing system
Note
this does a default format, returning file system to a fresh state The filing system implementation may define more specialised methods which can be called directly.
- Return values:
int – error code
- virtual int check() override
Perform a file system consistency check.
Note
if possible, issues should be resolved. Returns 0 if file system checked out OK. Otherwise there were issues: < 0 for unrecoverable errors,
0 for recoverable errors.
- Return values:
int – error code
- struct FormatOptions
Formatting options.
- struct FatParam
- namespace FWFS
Functions
- FileAttributes getFileAttributes(Object::Attributes objattr)
- Object::Attributes getObjectAttributes(FileAttributes fileAttr)
Variables
- constexpr size_t FWFS_BASE_OFFSET = {sizeof(uint32_t)}
- constexpr uint32_t FWFILESYS_START_MARKER = {0x53465746}
- constexpr uint32_t FWFILESYS_END_MARKER = {0x46574653}
- class ArchiveStream : public IFS::FsBase, public IDataSourceStream
Supports direct streaming into FWFS archive format.
Data needs to be enumerated so that all child files and directories are written first. As this happens, the parent directory is built as a list of object references.
The size of all child objects must be known before the containing object is written out. It should be sufficient to buffer all of these into an internal stream. This would only be a problem if large child objects are used. This is possible, but the builder avoids doing it and we should too.
Top-level object is a Volume, below that is the root Directory. That means objects are written in the order: child files/directories Root directory Volume End marker
Subclassed by ArchiveStream
Public Functions
- inline ArchiveStream(FileSystem *fileSystem, const VolumeInfo &volumeInfo, const String &rootPath = nullptr, Flags flags = 0)
Construct an archive stream.
- Parameters:
fileSystem – The filesystem to read
volumeInfo – Information about volume to store in the stream
rootPath – Where to root the generated filesystem
flags –
- inline virtual bool filterStat(const Stat &stat)
Override this method to filter items.
Use to omit temporary files or directories.
- Parameters:
stat – The current item
- Return values:
bool – Return true to process the item, false to skip it
- inline virtual IBlockEncoder *createEncoder(FileInfo &file)
Override this method to implement custom encoding such as compression or encryption.
To support compression or encryption, this method can create the appropriate stream type and set the appropriate attributes using methods of
FileInfo.
- Parameters:
file – Details of the file being archived
- Return values:
IBlockEncoder* – Stream to use for file content. Return nullptr for default behaviour.
- virtual uint16_t readMemoryBlock(char *data, int bufSize) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
- virtual int seekFrom(int offset, SeekOrigin origin) override
Change position in stream.
Note
This method is implemented by streams which support random seeking, such as files and memory streams.
- Parameters:
offset –
origin –
- Return values:
New – position, < 0 on error
- inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
- inline virtual MimeType getMimeType() const override
Get MIME type for stream content.
- Return values:
MimeType –
- void reset()
Reset stream to beginning.
- class FileInfo
Passed to callbacks to allow modification of output data.
Public Functions
- int setAttribute(AttributeTag tag, const void *data, size_t size)
Set an additional attribute on the file.
These are written out before existing file metadata is copied, so will take priority. For example, if we set the compression attribute here then that is the one which the filesystem will use when mounted. However, the original compression attribute will still be present which may be helpful.
- template<typename ...ParamTypes>
inline int setUserAttribute(uint8_t tagValue, ParamTypes... params)Set an additional user attribute.
- struct VolumeInfo
- class IBlockEncoder
Virtual base class to support (file) data encryption and compression.
Encryption and compression are typically done in blocks of a fixed size. To support these operations an instance of this class is created which encodes data one block at a time. Each block is stored separately and the resulting file consists of a chain of these blocks. This is natively supported by FWFS.
If the final data size is known in advance then the implementation will return just a single data stream.
Subclassed by IFS::FWFS::BasicEncoder
Public Functions
- virtual IDataSourceStream *getNextStream() = 0
@Implement this method and return nullptr when all blocks have been encoded.
The stream returned must know it’s size (i.e. available() must not return -1). The encoder owns any stream objects created so is responsible for destroying them when finished. This allows them to be re-used if appropriate.
- class BasicEncoder : public IFS::FWFS::IBlockEncoder
Public Functions
- inline virtual IDataSourceStream *getNextStream() override
@Implement this method and return nullptr when all blocks have been encoded.
The stream returned must know it’s size (i.e. available() must not return -1). The encoder owns any stream objects created so is responsible for destroying them when finished. This allows them to be re-used if appropriate.
- class FileSystem : public IFS::IFileSystem
Implementation of firmware filing system using IFS.
Public Functions
- virtual int mount() override
Mount file system, performing any required initialisation.
- Return values:
error – code
- virtual int getinfo(Info &info) override
get filing system information
- Parameters:
info – structure to read information into
- Return values:
int – error code
- virtual String getErrorString(int err) override
get the text for a returned error code
- Return values:
String –
- virtual int setVolume(uint8_t index, IFileSystem *fileSystem) override
Set volume for mountpoint.
- Parameters:
index – Volume index
fileSystem – The filesystem to root at this mountpoint
- Return values:
int – error code
- virtual int opendir(const char *path, DirHandle &dir) override
open a directory for reading
- Parameters:
path – path to directory. nullptr is interpreted as root directory
dir – returns a pointer to the directory object
- Return values:
int – error code
- virtual int readdir(DirHandle dir, Stat &stat) override
read a directory entry
Note
File system allocates entries structure as it usually needs to track other information. It releases memory when closedir() is called.
- Parameters:
dir –
stat –
- Return values:
int – error code
- virtual int rewinddir(DirHandle dir) override
Reset directory read position to start.
- Parameters:
dir –
- Return values:
int – error code
- virtual int closedir(DirHandle dir) override
close a directory object
- Parameters:
dir – directory to close
- Return values:
int – error code
- virtual int mkdir(const char *path) override
Create a directory.
Only the final directory in the path is guaranteed to be created. Usually, this call will fail if intermediate directories are not present. Use
IFS::FileSystem::makedirs()for this purpose.
- Parameters:
path – Path to directory
- Return values:
int – error code
- virtual int stat(const char *path, Stat *stat) override
get file information
Returned stat will indicate whether the path is a mountpoint or directory. For a mount point, stats for the root directory of the mounted filesystem must be obtained by opening a handle then using
fstat:int handle = fs.open("path-to-mountpoint"); Stat stat; fs.fstat(handle, &stat); fs.close(handle);
- Parameters:
path – name or path of file/directory/mountpoint
s – structure to return information in, may be null to do a simple file existence check
- Return values:
int – error code
- virtual int fstat(FileHandle file, Stat *stat) override
get file information
- Parameters:
file – handle to open file
stat – structure to return information in, may be null
- Return values:
int – error code
- virtual int fcontrol(FileHandle file, ControlCode code, void *buffer, size_t bufSize) override
Low-level and non-standard file control operations.
To simplify usage the same buffer is used for both input and output. Only the size of the buffer is provided. If a specific FCNTL code requires more information then it will be contained within the provided data.
- Parameters:
file –
code – FCNTL_XXX code
buffer – Input/Output buffer
bufSize – Size of buffer
- Return values:
int – error code or, on success, data size
- virtual int fsetxattr(FileHandle file, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute on an open file.
- Parameters:
file – handle to open file
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int fgetxattr(FileHandle file, AttributeTag tag, void *buffer, size_t size) override
Get an extended attribute from an open file.
- Parameters:
file – handle to open file
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual int fenumxattr(FileHandle file, AttributeEnumCallback callback, void *buffer, size_t bufsize) override
Enumerate attributes.
- Parameters:
file – handle to open file
callback – Callback function to invoke for each attribute found
buffer – Buffer to use for reading attribute data. Use nullptr if only tags are required
bufsize – Size of buffer
- Return values:
int – error code, on success returns number of attributes read
- virtual int setxattr(const char *path, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute for a file given its path.
- Parameters:
path – Full path to file (or directory)
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int getxattr(const char *path, AttributeTag tag, void *buffer, size_t size) override
Get an attribute from a file given its path.
- Parameters:
file – Full path to file (or directory)
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual FileHandle open(const char *path, OpenFlags flags) override
open a file (or directory) by path
This function may also be used to obtain a directory handle to perform various operations such as enumerating attributes. Calls to read or write on such handles will typically fail.
- Parameters:
path – full path to file
flags – Desired access and other options
- Return values:
FileHandle – file handle or error code
- virtual int close(FileHandle file) override
close an open file
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int read(FileHandle file, void *data, size_t size) override
read content from a file and advance cursor
- Parameters:
file – handle to open file
data – buffer to write into
size – size of file buffer, maximum number of bytes to read
- Return values:
int – number of bytes read or error code
- virtual int write(FileHandle file, const void *data, size_t size) override
write content to a file at current position and advance cursor
- Parameters:
file – handle to open file
data – buffer to read from
size – number of bytes to write
- Return values:
int – number of bytes written or error code
- virtual file_offset_t lseek(FileHandle file, file_offset_t offset, SeekOrigin origin) override
change file read/write position
- Parameters:
file – handle to open file
offset – position relative to origin
origin – where to seek from (start/end or current position)
- Return values:
file_offset_t – current position or error code
- virtual int eof(FileHandle file) override
determine if current file position is at end of file
- Parameters:
file – handle to open file
- Return values:
int – 0 - not EOF, > 0 - at EOF, < 0 - error
- virtual file_offset_t tell(FileHandle file) override
get current file position
- Parameters:
file – handle to open file
- Return values:
file_offset_t – current position relative to start of file, or error code
- virtual int ftruncate(FileHandle file, file_size_t new_size) override
Truncate (reduce) the size of an open file.
Note
In POSIX
ftruncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize
- Parameters:
file – Open file handle, must have Write access
newSize –
- Return values:
int – Error code
- virtual int flush(FileHandle file) override
flush any buffered data to physical media
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int fgetextents(FileHandle file, Storage::Partition *part, Extent *list, uint16_t extcount) override
Get extents for a file.
- Parameters:
file – Handle to open file
part – Partition where the file lives (OUT, OPTIONAL)
list – Buffer for extents (OPTIONAL)
extcount – Maximum number of extents to return in
list- Return values:
int – Total number of extents for file (may be larger than ‘extcount’), or error code
- virtual int rename(const char *oldpath, const char *newpath) override
rename a file
- Parameters:
oldpath –
newpath –
- Return values:
int – error code
- virtual int remove(const char *path) override
remove (delete) a file by path
- Parameters:
path –
- Return values:
int – error code
- virtual int fremove(FileHandle file) override
remove (delete) a file by handle
- Parameters:
file – handle to open file
- Return values:
int – error code
- inline virtual int format() override
format the filing system
Note
this does a default format, returning file system to a fresh state The filing system implementation may define more specialised methods which can be called directly.
- Return values:
int – error code
- inline virtual int check() override
Perform a file system consistency check.
Note
if possible, issues should be resolved. Returns 0 if file system checked out OK. Otherwise there were issues: < 0 for unrecoverable errors,
0 for recoverable errors.
- Return values:
int – error code
- struct Object
Object structure.
Note
all objects conform to this structure. Only the first word (4 bytes) are required to nagivate the file system. All objects have an 8, 16 or 24-bit size field. Content is always immediately after this field. Reference objects are always 8-bit sized.
Public Types
- enum class Attribute
Object attributes.
Note
these are bit values
Values:
- enumerator Archive
Object modified flag.
Note
Object has been changed on disk. Typically used by backup applications
- enumerator Encrypted
Object data is encrypted.
This is just a hint. Applications will typically provide additional user metadata to provide any additional information required for decryption.
- enumerator MAX
Public Functions
- inline size_t contentOffset() const
return offset to start of object content
- inline uint32_t contentSize() const
return size of object content, excluding header and size fields
Note
must check return error code
- Return values:
size – or error code
- inline uint32_t size() const
total size this object occupies in the image
- Return values:
size – or error code
- class ObjectBuffer
Class to manage writing object data into a stream.
- namespace Gdb
- class FileSystem : public IFS::IFileSystem
IFS implementation of Host filing system.
Public Functions
- inline virtual int mount() override
Mount file system, performing any required initialisation.
- Return values:
error – code
- virtual int getinfo(Info &info) override
get filing system information
- Parameters:
info – structure to read information into
- Return values:
int – error code
- virtual String getErrorString(int err) override
get the text for a returned error code
- Return values:
String –
- inline virtual int opendir(const char *path, DirHandle &dir) override
open a directory for reading
- Parameters:
path – path to directory. nullptr is interpreted as root directory
dir – returns a pointer to the directory object
- Return values:
int – error code
- inline virtual int rewinddir(DirHandle dir) override
Reset directory read position to start.
- Parameters:
dir –
- Return values:
int – error code
- inline virtual int readdir(DirHandle dir, Stat &stat) override
read a directory entry
Note
File system allocates entries structure as it usually needs to track other information. It releases memory when closedir() is called.
- Parameters:
dir –
stat –
- Return values:
int – error code
- inline virtual int closedir(DirHandle dir) override
close a directory object
- Parameters:
dir – directory to close
- Return values:
int – error code
- inline virtual int mkdir(const char *path) override
Create a directory.
Only the final directory in the path is guaranteed to be created. Usually, this call will fail if intermediate directories are not present. Use
IFS::FileSystem::makedirs()for this purpose.
- Parameters:
path – Path to directory
- Return values:
int – error code
- virtual int stat(const char *path, Stat *stat) override
get file information
Returned stat will indicate whether the path is a mountpoint or directory. For a mount point, stats for the root directory of the mounted filesystem must be obtained by opening a handle then using
fstat:int handle = fs.open("path-to-mountpoint"); Stat stat; fs.fstat(handle, &stat); fs.close(handle);
- Parameters:
path – name or path of file/directory/mountpoint
s – structure to return information in, may be null to do a simple file existence check
- Return values:
int – error code
- virtual int fstat(FileHandle file, Stat *stat) override
get file information
- Parameters:
file – handle to open file
stat – structure to return information in, may be null
- Return values:
int – error code
- inline virtual int fsetxattr(FileHandle file, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute on an open file.
- Parameters:
file – handle to open file
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- inline virtual int fgetxattr(FileHandle file, AttributeTag tag, void *buffer, size_t size) override
Get an extended attribute from an open file.
- Parameters:
file – handle to open file
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- inline virtual int fenumxattr(FileHandle file, AttributeEnumCallback callback, void *buffer, size_t bufsize) override
Enumerate attributes.
- Parameters:
file – handle to open file
callback – Callback function to invoke for each attribute found
buffer – Buffer to use for reading attribute data. Use nullptr if only tags are required
bufsize – Size of buffer
- Return values:
int – error code, on success returns number of attributes read
- inline virtual int setxattr(const char *path, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute for a file given its path.
- Parameters:
path – Full path to file (or directory)
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- inline virtual int getxattr(const char *path, AttributeTag tag, void *buffer, size_t size) override
Get an attribute from a file given its path.
- Parameters:
file – Full path to file (or directory)
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual FileHandle open(const char *path, OpenFlags flags) override
open a file (or directory) by path
This function may also be used to obtain a directory handle to perform various operations such as enumerating attributes. Calls to read or write on such handles will typically fail.
- Parameters:
path – full path to file
flags – Desired access and other options
- Return values:
FileHandle – file handle or error code
- virtual int close(FileHandle file) override
close an open file
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int read(FileHandle file, void *data, size_t size) override
read content from a file and advance cursor
- Parameters:
file – handle to open file
data – buffer to write into
size – size of file buffer, maximum number of bytes to read
- Return values:
int – number of bytes read or error code
- virtual int write(FileHandle file, const void *data, size_t size) override
write content to a file at current position and advance cursor
- Parameters:
file – handle to open file
data – buffer to read from
size – number of bytes to write
- Return values:
int – number of bytes written or error code
- virtual file_offset_t lseek(FileHandle file, file_offset_t offset, SeekOrigin origin) override
change file read/write position
- Parameters:
file – handle to open file
offset – position relative to origin
origin – where to seek from (start/end or current position)
- Return values:
file_offset_t – current position or error code
- virtual int eof(FileHandle file) override
determine if current file position is at end of file
- Parameters:
file – handle to open file
- Return values:
int – 0 - not EOF, > 0 - at EOF, < 0 - error
- virtual file_offset_t tell(FileHandle file) override
get current file position
- Parameters:
file – handle to open file
- Return values:
file_offset_t – current position relative to start of file, or error code
- inline virtual int ftruncate(FileHandle file, file_size_t new_size) override
Truncate (reduce) the size of an open file.
Note
In POSIX
ftruncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize
- Parameters:
file – Open file handle, must have Write access
newSize –
- Return values:
int – Error code
- inline virtual int flush(FileHandle file) override
flush any buffered data to physical media
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int rename(const char *oldpath, const char *newpath) override
rename a file
- Parameters:
oldpath –
newpath –
- Return values:
int – error code
- virtual int remove(const char *path) override
remove (delete) a file by path
- Parameters:
path –
- Return values:
int – error code
- inline virtual int fremove(FileHandle file) override
remove (delete) a file by handle
- Parameters:
file – handle to open file
- Return values:
int – error code
- inline virtual int format() override
format the filing system
Note
this does a default format, returning file system to a fresh state The filing system implementation may define more specialised methods which can be called directly.
- Return values:
int – error code
- inline virtual int check() override
Perform a file system consistency check.
Note
if possible, issues should be resolved. Returns 0 if file system checked out OK. Otherwise there were issues: < 0 for unrecoverable errors,
0 for recoverable errors.
- Return values:
int – error code
- namespace Host
Functions
- FileSystem &getFileSystem()
- class FileSystem : public IFS::IFileSystem
IFS implementation of Host filing system.
Public Functions
- virtual int mount() override
Mount file system, performing any required initialisation.
- Return values:
error – code
- virtual int getinfo(Info &info) override
get filing system information
- Parameters:
info – structure to read information into
- Return values:
int – error code
- virtual String getErrorString(int err) override
get the text for a returned error code
- Return values:
String –
- virtual int opendir(const char *path, DirHandle &dir) override
open a directory for reading
- Parameters:
path – path to directory. nullptr is interpreted as root directory
dir – returns a pointer to the directory object
- Return values:
int – error code
- virtual int rewinddir(DirHandle dir) override
Reset directory read position to start.
- Parameters:
dir –
- Return values:
int – error code
- virtual int readdir(DirHandle dir, Stat &stat) override
read a directory entry
Note
File system allocates entries structure as it usually needs to track other information. It releases memory when closedir() is called.
- Parameters:
dir –
stat –
- Return values:
int – error code
- virtual int closedir(DirHandle dir) override
close a directory object
- Parameters:
dir – directory to close
- Return values:
int – error code
- virtual int mkdir(const char *path) override
Create a directory.
Only the final directory in the path is guaranteed to be created. Usually, this call will fail if intermediate directories are not present. Use
IFS::FileSystem::makedirs()for this purpose.
- Parameters:
path – Path to directory
- Return values:
int – error code
- virtual int stat(const char *path, Stat *stat) override
get file information
Returned stat will indicate whether the path is a mountpoint or directory. For a mount point, stats for the root directory of the mounted filesystem must be obtained by opening a handle then using
fstat:int handle = fs.open("path-to-mountpoint"); Stat stat; fs.fstat(handle, &stat); fs.close(handle);
- Parameters:
path – name or path of file/directory/mountpoint
s – structure to return information in, may be null to do a simple file existence check
- Return values:
int – error code
- virtual int fstat(FileHandle file, Stat *stat) override
get file information
- Parameters:
file – handle to open file
stat – structure to return information in, may be null
- Return values:
int – error code
- virtual int fsetxattr(FileHandle file, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute on an open file.
- Parameters:
file – handle to open file
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int fgetxattr(FileHandle file, AttributeTag tag, void *buffer, size_t size) override
Get an extended attribute from an open file.
- Parameters:
file – handle to open file
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual int fenumxattr(FileHandle file, AttributeEnumCallback callback, void *buffer, size_t bufsize) override
Enumerate attributes.
- Parameters:
file – handle to open file
callback – Callback function to invoke for each attribute found
buffer – Buffer to use for reading attribute data. Use nullptr if only tags are required
bufsize – Size of buffer
- Return values:
int – error code, on success returns number of attributes read
- virtual int setxattr(const char *path, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute for a file given its path.
- Parameters:
path – Full path to file (or directory)
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int getxattr(const char *path, AttributeTag tag, void *buffer, size_t size) override
Get an attribute from a file given its path.
- Parameters:
file – Full path to file (or directory)
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual FileHandle open(const char *path, OpenFlags flags) override
open a file (or directory) by path
This function may also be used to obtain a directory handle to perform various operations such as enumerating attributes. Calls to read or write on such handles will typically fail.
- Parameters:
path – full path to file
flags – Desired access and other options
- Return values:
FileHandle – file handle or error code
- virtual int close(FileHandle file) override
close an open file
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int read(FileHandle file, void *data, size_t size) override
read content from a file and advance cursor
- Parameters:
file – handle to open file
data – buffer to write into
size – size of file buffer, maximum number of bytes to read
- Return values:
int – number of bytes read or error code
- virtual int write(FileHandle file, const void *data, size_t size) override
write content to a file at current position and advance cursor
- Parameters:
file – handle to open file
data – buffer to read from
size – number of bytes to write
- Return values:
int – number of bytes written or error code
- virtual file_offset_t lseek(FileHandle file, file_offset_t offset, SeekOrigin origin) override
change file read/write position
- Parameters:
file – handle to open file
offset – position relative to origin
origin – where to seek from (start/end or current position)
- Return values:
file_offset_t – current position or error code
- virtual int eof(FileHandle file) override
determine if current file position is at end of file
- Parameters:
file – handle to open file
- Return values:
int – 0 - not EOF, > 0 - at EOF, < 0 - error
- virtual file_offset_t tell(FileHandle file) override
get current file position
- Parameters:
file – handle to open file
- Return values:
file_offset_t – current position relative to start of file, or error code
- virtual int ftruncate(FileHandle file, file_size_t new_size) override
Truncate (reduce) the size of an open file.
Note
In POSIX
ftruncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize
- Parameters:
file – Open file handle, must have Write access
newSize –
- Return values:
int – Error code
- virtual int flush(FileHandle file) override
flush any buffered data to physical media
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int rename(const char *oldpath, const char *newpath) override
rename a file
- Parameters:
oldpath –
newpath –
- Return values:
int – error code
- virtual int remove(const char *path) override
remove (delete) a file by path
- Parameters:
path –
- Return values:
int – error code
- inline virtual int fremove(FileHandle file) override
remove (delete) a file by handle
- Parameters:
file – handle to open file
- Return values:
int – error code
- inline virtual int format() override
format the filing system
Note
this does a default format, returning file system to a fresh state The filing system implementation may define more specialised methods which can be called directly.
- Return values:
int – error code
- inline virtual int check() override
Perform a file system consistency check.
Note
if possible, issues should be resolved. Returns 0 if file system checked out OK. Otherwise there were issues: < 0 for unrecoverable errors,
0 for recoverable errors.
- Return values:
int – error code
- namespace HYFS
- class FileSystem : public IFS::IFileSystem
Public Functions
- virtual int mount() override
Mount file system, performing any required initialisation.
- Return values:
error – code
- virtual int getinfo(Info &info) override
get filing system information
- Parameters:
info – structure to read information into
- Return values:
int – error code
- virtual String getErrorString(int err) override
get the text for a returned error code
- Return values:
String –
- virtual int setVolume(uint8_t index, IFileSystem *fileSystem) override
Set volume for mountpoint.
- Parameters:
index – Volume index
fileSystem – The filesystem to root at this mountpoint
- Return values:
int – error code
- virtual int opendir(const char *path, DirHandle &dir) override
open a directory for reading
- Parameters:
path – path to directory. nullptr is interpreted as root directory
dir – returns a pointer to the directory object
- Return values:
int – error code
- virtual int readdir(DirHandle dir, Stat &stat) override
read a directory entry
Note
File system allocates entries structure as it usually needs to track other information. It releases memory when closedir() is called.
- Parameters:
dir –
stat –
- Return values:
int – error code
- virtual int rewinddir(DirHandle dir) override
Reset directory read position to start.
- Parameters:
dir –
- Return values:
int – error code
- virtual int closedir(DirHandle dir) override
close a directory object
- Parameters:
dir – directory to close
- Return values:
int – error code
- virtual int mkdir(const char *path) override
Create a directory.
Only the final directory in the path is guaranteed to be created. Usually, this call will fail if intermediate directories are not present. Use
IFS::FileSystem::makedirs()for this purpose.
- Parameters:
path – Path to directory
- Return values:
int – error code
- virtual int stat(const char *path, Stat *stat) override
get file information
Returned stat will indicate whether the path is a mountpoint or directory. For a mount point, stats for the root directory of the mounted filesystem must be obtained by opening a handle then using
fstat:int handle = fs.open("path-to-mountpoint"); Stat stat; fs.fstat(handle, &stat); fs.close(handle);
- Parameters:
path – name or path of file/directory/mountpoint
s – structure to return information in, may be null to do a simple file existence check
- Return values:
int – error code
- virtual int fstat(FileHandle file, Stat *stat) override
get file information
- Parameters:
file – handle to open file
stat – structure to return information in, may be null
- Return values:
int – error code
- virtual int fcontrol(FileHandle file, ControlCode code, void *buffer, size_t bufSize) override
Low-level and non-standard file control operations.
To simplify usage the same buffer is used for both input and output. Only the size of the buffer is provided. If a specific FCNTL code requires more information then it will be contained within the provided data.
- Parameters:
file –
code – FCNTL_XXX code
buffer – Input/Output buffer
bufSize – Size of buffer
- Return values:
int – error code or, on success, data size
- virtual int fsetxattr(FileHandle file, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute on an open file.
- Parameters:
file – handle to open file
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int fgetxattr(FileHandle file, AttributeTag tag, void *buffer, size_t size) override
Get an extended attribute from an open file.
- Parameters:
file – handle to open file
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual int fenumxattr(FileHandle file, AttributeEnumCallback callback, void *buffer, size_t bufsize) override
Enumerate attributes.
- Parameters:
file – handle to open file
callback – Callback function to invoke for each attribute found
buffer – Buffer to use for reading attribute data. Use nullptr if only tags are required
bufsize – Size of buffer
- Return values:
int – error code, on success returns number of attributes read
- virtual int setxattr(const char *path, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute for a file given its path.
- Parameters:
path – Full path to file (or directory)
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
- virtual int getxattr(const char *path, AttributeTag tag, void *buffer, size_t size) override
Get an attribute from a file given its path.
- Parameters:
file – Full path to file (or directory)
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
- virtual FileHandle open(const char *path, OpenFlags flags) override
open a file (or directory) by path
This function may also be used to obtain a directory handle to perform various operations such as enumerating attributes. Calls to read or write on such handles will typically fail.
- Parameters:
path – full path to file
flags – Desired access and other options
- Return values:
FileHandle – file handle or error code
- virtual int close(FileHandle file) override
close an open file
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int read(FileHandle file, void *data, size_t size) override
read content from a file and advance cursor
- Parameters:
file – handle to open file
data – buffer to write into
size – size of file buffer, maximum number of bytes to read
- Return values:
int – number of bytes read or error code
- virtual int write(FileHandle file, const void *data, size_t size) override
write content to a file at current position and advance cursor
- Parameters:
file – handle to open file
data – buffer to read from
size – number of bytes to write
- Return values:
int – number of bytes written or error code
- virtual file_offset_t lseek(FileHandle file, file_offset_t offset, SeekOrigin origin) override
change file read/write position
- Parameters:
file – handle to open file
offset – position relative to origin
origin – where to seek from (start/end or current position)
- Return values:
file_offset_t – current position or error code
- virtual int eof(FileHandle file) override
determine if current file position is at end of file
- Parameters:
file – handle to open file
- Return values:
int – 0 - not EOF, > 0 - at EOF, < 0 - error
- virtual file_offset_t tell(FileHandle file) override
get current file position
- Parameters:
file – handle to open file
- Return values:
file_offset_t – current position relative to start of file, or error code
- virtual int ftruncate(FileHandle file, file_size_t new_size) override
Truncate (reduce) the size of an open file.
Note
In POSIX
ftruncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize
- Parameters:
file – Open file handle, must have Write access
newSize –
- Return values:
int – Error code
- virtual int flush(FileHandle file) override
flush any buffered data to physical media
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int fgetextents(FileHandle file, Storage::Partition *part, Extent *list, uint16_t extcount) override
Get extents for a file.
- Parameters:
file – Handle to open file
part – Partition where the file lives (OUT, OPTIONAL)
list – Buffer for extents (OPTIONAL)
extcount – Maximum number of extents to return in
list- Return values:
int – Total number of extents for file (may be larger than ‘extcount’), or error code
- virtual int rename(const char *oldpath, const char *newpath) override
rename a file
- Parameters:
oldpath –
newpath –
- Return values:
int – error code
- virtual int remove(const char *path) override
remove (delete) a file by path
- Parameters:
path –
- Return values:
int – error code
- virtual int fremove(FileHandle file) override
remove (delete) a file by handle
- Parameters:
file – handle to open file
- Return values:
int – error code
- virtual int format() override
format the filing system
Note
this does a default format, returning file system to a fresh state The filing system implementation may define more specialised methods which can be called directly.
- Return values:
int – error code
- virtual int check() override
Perform a file system consistency check.
Note
if possible, issues should be resolved. Returns 0 if file system checked out OK. Otherwise there were issues: < 0 for unrecoverable errors,
0 for recoverable errors.
- Return values:
int – error code
References
Used by
Sming (main) Component
SPI Flash Component
FatIFS Library
LittleFS Library
SPIFFS IFS Library Library
Environment Variables
FSBUILD
FSBUILD_OPTIONSSoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Networking Support
Contains core networking protocol support classes.
DNS: Domain Name System
https://en.m.wikipedia.org/wiki/Domain_Name_System
Server API
- enum class DnsReplyCode
Values:
- enumerator NoError
- enumerator FormError
- enumerator ServerFailure
- enumerator NonExistentDomain
- enumerator NotImplemented
- enumerator Refused
- enumerator YXDomain
- enumerator YXRRSet
- enumerator NXRRSet
- DNS_QR_QUERY
- DNS_QR_RESPONSE
- DNS_OPCODE_QUERY
- struct DnsHeader
- class DnsServer : public UdpConnection
DNS server class.
Public Functions
- inline void setErrorReplyCode(DnsReplyCode replyCode)
Set error reply code.
- inline void setTTL(uint32_t ttl)
Set message Time-To-Live in seconds.
- bool start(uint16_t port, const String &domainName, const IpAddress &resolvedIP)
Start the DNS server.
- Parameters:
port –
domainName –
resolvedIP –
- Return values:
bool – true if successful, false if there are no sockets available.
- void stop()
Stop the DNS server.
FTP: File Transfer Protocol
https://en.m.wikipedia.org/wiki/File_Transfer_Protocol
Server API
- class FtpServer : public CustomFtpServer
Provides FTP server.
HTTP: HyperText Transfer Protocol
https://en.m.wikipedia.org/wiki/Hypertext_Transfer_Protocol
Build Variables
- HTTP_SERVER_EXPOSE_NAME
Default: 1 (enabled)
Adds “HttpServer/Sming” to the SERVER field in response headers. If disabled, the SERVER field is omitted from all responses.
- HTTP_SERVER_EXPOSE_VERSION
Default: 0 (disabled)
Adds the current Sming build version to the SERVER field in response headers. For example, “Sming/4.0.0-rc2”.
Requires HTTP_SERVER_EXPOSE_NAME to be enabled.
- HTTP_SERVER_EXPOSE_DATE
Default: 0 (disabled)
Sets the DATE field in response headers.
API Documentation
Client API
- using RequestQueue = ObjectQueue<HttpRequest, HTTP_REQUEST_POOL_SIZE>
Provides http client connection.
- class HttpClient
Subclassed by Ota::Network::HttpUpgrader
Public Functions
- inline virtual ~HttpClient()
HttpClient destructor.
Note
DON’T call cleanup. If you want to free all resources from HttpClients the correct sequence will be to
Delete all instances of HttpClient
Call the static method HttpClient::cleanup();
- inline bool downloadString(const Url &url, RequestCompletedDelegate requestComplete, size_t maxLength = NETWORK_SEND_BUFFER_SIZE)
Queue request to download content as string (in memory)
- Parameters:
url – URL from which the content will be fetched
requestComplete – Completion callback
maxLength – maximum bytes to store in memory. If the response is bigger than
maxLengththen the rest bytes will be discarded. Use this parameter wisely as setting the value too high may consume all available RAM resulting in device restart and Denial-Of-Service
- bool downloadFile(const Url &url, const String &saveFileName, RequestCompletedDelegate requestComplete = nullptr)
Queue request to download a file.
- Parameters:
url – Source of file data
saveFileName – Path to save file to. Optional: specify nullptr to use name from url
requestComplete – Completion callback
- inline HttpRequest *createRequest(const Url &url)
Helper function to create a new request on a URL.
- Parameters:
url –
- Return values:
HttpRequest* –
Public Static Functions
- static inline void cleanup()
Use this method to clean all object pools.
Server API
- class HttpResource
Instances of this class are registered with an HttpServer for a specific URL.
Subclassed by HttpMultipartResource, WebsocketResource
Public Functions
- inline virtual void shutdown(HttpServerConnection &connection)
Takes care to cleanup the connection.
Public Members
- HttpResourceDelegate onUrlComplete = nullptr
URL is ready. Path and status code are available.
- HttpServerConnectionBodyDelegate onBody = nullptr
resource wants to process the raw body data
- HttpResourceDelegate onHeadersComplete = nullptr
headers are ready
- HttpResourceDelegate onRequestComplete = nullptr
request is complete OR upgraded
- HttpServerConnectionUpgradeDelegate onUpgrade = nullptr
request is upgraded and raw data is passed to it
- class PluginRef : public LinkedObjectTemplate<PluginRef>
- class HttpResourceTree : public ObjectMap<String, HttpResource>
Class to map URL paths to classes which handle them.
Public Functions
- inline HttpResource *setDefault(HttpResource *resource)
Set the default resource handler.
- Parameters:
resource – The default resource handler
- inline HttpResource *setDefault(const HttpResourceDelegate &onRequestComplete)
Set the default resource handler, identified by “*” wildcard.
- Parameters:
onRequestComplete – The default resource handler
- inline HttpResource *setDefault(const HttpPathDelegate &callback)
Set the default resource handler, identified by “*” wildcard.
- inline HttpResource *getDefault()
Get the current default resource handler, if any.
- Return values:
HttpResource* –
- HttpResource *set(const String &path, const HttpResourceDelegate &onRequestComplete)
Set a callback to handle the given path.
Note
Path should start with slash. Trailing slashes will be removed.
Note
Any existing handler for this path is replaced
- Parameters:
path – URL path
onRequestComplete – Delegate to handle this path
- Return values:
HttpResource* – The created resource object
- template<class ...Tail>
inline HttpResource *set(const String &path, const HttpResourceDelegate &onRequestComplete, HttpResourcePlugin *plugin, Tail... plugins)Set a callback to handle the given path, with one or more plugins.
Note
Path should start with slash. Trailing slashes will be removed.
Note
Any existing handler for this path is replaced
- Parameters:
path – URL path
onRequestComplete – Delegate to handle this path
plugin – Plugins to register for the resource
- Return values:
HttpResource* – The created resource object
- HttpResource *set(String path, const HttpPathDelegate &callback)
Add a new path resource with a callback.
Note
Path should start with slash. Trailing slashes will be removed
Note
Any existing handler for this path is replaced
- Parameters:
path – URL path
callback – The callback that will handle this path
- template<class ...Tail>
inline HttpResource *set(const String &path, const HttpPathDelegate &callback, HttpResourcePlugin *plugin, Tail... plugins)Add a new path resource with callback and one or more plugins.
Note
Path should start with slash. Trailing slashes will be removed
Note
Any existing handler for this path is replaced
- Parameters:
path – URL path
callback – The callback that will handle this path
plugin – - optional resource plugin
- Return values:
HttpResource* – The created resource object
- inline void set(const K &key, V *value)
Set a key value.
- Parameters:
key –
value –
- struct HttpServerSettings
Public Members
- uint16_t maxActiveConnections = 10
maximum number of concurrent requests..
- uint16_t keepAliveSeconds = 0
default seconds to keep the connection alive before closing it
- int minHeapSize = -1
min heap size that is required to accept connection, -1 means use server default
- bool useDefaultBodyParsers = 1
if the default body parsers, as form-url-encoded, should be used
- bool closeOnContentError = true
close the connection if a body parser or resource fails to parse the body content.
- class HttpServer : public TcpServer
Public Functions
- void configure(const HttpServerSettings &settings)
Allows changing the server configuration.
- inline void setBodyParser(const String &contentType, HttpBodyParserDelegate parser)
Allows content-type specific parsing of the body based on content-type.
- Parameters:
contentType – Can be full content-type like ‘application/json’, or ‘application/*’ or ‘*’. If there is exact match for the content-type wildcard content-types will not be used. There can be only one catch-all ‘*’ body parser and that will be the last registered
parser –
- inline void setBodyParser(MimeType mimeType, HttpBodyParserDelegate parser)
Allows content-type specific parsing of the body based on content-type.
- Parameters:
mimeType –
parser –
Public Members
- HttpResourceTree paths
Maps paths to resources which deal with incoming requests.
Support API
- namespace ContentType
Obtain MIME type value from file name or path, with extension
- MimeType fromFullFileName(const char *fileName, MimeType unknown)
- Parameters:
fileName – As NUL-terminated string
unknown – Value to return if type cannot be determined
- Return values:
MimeType –
Obtain content type string from file name or path, with extension
Functions
- MimeType fromFileExtension(const char *extension, MimeType unknown)
Obtain MIME type value from file extension.
- Parameters:
extension – excluding ‘.’ separator (e.g. “htm”, “json”)
unknown – Value to return if type cannot be determined
- Return values:
MimeType –
- String fromFileExtension(const char *extension)
Obtain content type string from file extension.
- Parameters:
extension – excluding ‘.’ separator (e.g. “htm”, “json”)
- Return values:
String –
- inline String fromFileExtension(const String &extension)
Obtain content type string from file extension.
- Parameters:
extension –
- Return values:
String –
- MimeType fromString(const char *str)
Get enumerated value for a MIME type string.
- Parameters:
str –
- Return values:
MimeType – If empty, null or unrecognised returns MIME_UNKNOWN
Defines
- ENABLE_HTTP_REQUEST_AUTH
- HTTP_MAX_HEADER_SIZE
- HTTP_REQUEST_POOL_SIZE
- XX(num, name, string)
- XX(num, name, string)
- XX(num, name, string)
- XX(num, name, string)
- XX(n, s)
- XX(n, s)
- HTTP_PARSER_ERRNO(p)
Typedefs
- using HttpFiles = ObjectMap<String, ReadWriteStream>
Enums
Functions
- String toString(HttpStatus code)
Return a descriptive string for an HTTP status code.
- inline String httpGetStatusText(unsigned code)
Return a descriptive string for an HTTP status code.
- inline String toString(HttpMethod method)
Return text for an HTTP method.
- class HttpHeaders : public HttpHeaderFields, private HashMap<HttpHeaderFieldName, String>
Encapsulates a set of HTTP header information.
- Todo:
add name and/or value escaping
Note
fields are stored as a map of field names vs. values. Standard fields may be accessed using enumeration tags. Behaviour is as for HashMap, with the addition of methods to support enumerated field names.
Subclassed by SSDP::BaseMessage< HttpHeaders >
Public Functions
- const String &operator[](const String &name) const
Fetch a reference to the header field value by name.
Note
if the field doesn’t exist a null String reference is returned
- Parameters:
name –
- Return values:
const – String& Reference to value
- inline String &operator[](const String &name)
Fetch a reference to the header field value by name.
Note
if the field doesn’t exist it is created with the default null value
- Parameters:
name –
- Return values:
String& – Reference to value
- class HeaderConst : public BaseElement<true>
- class HttpRequest
Encapsulates an incoming or outgoing request.
Set request body content
- inline HttpRequest *setBody(const String &body)
Set body from String object.
- HttpRequest *setBody(String &&body) noexcept
Set body from String object using move semantics: body will be invalid on return.
- HttpRequest *setBody(IDataSourceStream *stream)
Set body using given stream object, and retain ownership.
- HttpRequest *setBody(const uint8_t *rawData, size_t length)
Set body content by copying binary data.
- Parameters:
rawData – Data to copy
length – Number of bytes to copy
Public Types
- using SslInitDelegate = Delegate<void(Ssl::Session &session, HttpRequest &request)>
Callback delegate type used to initialise an SSL session for a given request.
Public Functions
- inline HttpRequest(const HttpRequest &value)
Copy constructor.
Note
Internal streams are not copied so these must be dealt with afterwards
- inline HttpRequest *clone() const
Clone this request into a new object using the copy constructor.
- Return values:
HttpRequest* – The new request object
- inline HttpRequest *setFile(const String &formElementName, ReadWriteStream *stream)
Sets a file to be sent.
- Parameters:
formElementName – The name of the element in the form
stream – Pointer to the stream (doesn’t have to be a FileStream)
- Return values:
HttpRequest* –
- inline const String &getHeader(const String &name)
Get header field value.
- Parameters:
name – Name of field
- Return values:
const – String& Value, will be invalid (i.e. if() == false) if field not present
- inline const String &getPostParameter(const String &name)
Get POST parameter value.
- Parameters:
name – Name of parameter
- Return values:
const – String& Value, will be invalid (i.e. if() == false) if field not present
- inline String getQueryParameter(const String &name, const String &defaultValue = nullptr) const
Get parameter from query fields.
- Parameters:
name – Name of parameter
defaultValue – Optional default value to use if requested parameter not present
- inline String getBody()
Moves content from the body stream into a String.
Note
Move semantics are used to ensure that no/minimal additional memory is required. If your application has set a non-memory stream type then the method will fail and return an invalid String. The stream content will be left unchanged.
- Return values:
String –
- inline IDataSourceStream *getBodyStream()
Return the current body stream.
Note
may return null
- Return values:
IDataSourceStream* –
- HttpRequest *setResponseStream(ReadWriteStream *stream)
Instead of storing the response body we can set a stream that will take care to process it.
Note
The response to this request will be stored in the user-provided stream.
- Parameters:
stream –
- Return values:
HttpRequest* –
- inline ReadWriteStream *getResponseStream()
Get the response stream (if any)
- void reset()
Clear buffers and reset to default state in preparation for another request.
- inline HttpRequest *onSslInit(SslInitDelegate delegate)
To customise SSL session options, provide a callback.
- Parameters:
delegate – Invoked before creating SSL connection
Public Members
- HttpMethod method = HTTP_GET
Request method.
- HttpHeaders headers
Request headers.
- HttpParams postParams
POST parameters.
- int retries = 0
how many times the request should be send again…
- void *args = nullptr
Used to store data that should be valid during a single request.
Public Static Functions
- static inline String toString(const HttpRequest &req)
Tries to present a readable version of the request.
- Parameters:
req –
- Return values:
String –
- class HttpResponse
Represents either an incoming or outgoing response to a HTTP request.
Public Functions
- bool sendFile(const String &fileName, bool allowGzipFileCheck = true)
Send file by name.
- Parameters:
fileName –
allowGzipFileCheck – If true, check file extension to see if content compressed
- Return values:
bool –
- bool sendNamedStream(IDataSourceStream *newDataStream)
Parse and send stream, using the name to determine the content type.
- Parameters:
newDataStream – If not set already, the contentType will be obtained from the name of this stream
- Return values:
bool –
- inline bool sendDataStream(IDataSourceStream *newDataStream, enum MimeType type)
Send data from the given stream object.
- Parameters:
newDataStream –
type –
- Return values:
false – on error
- bool sendDataStream(IDataSourceStream *newDataStream, const String &reqContentType = nullptr)
Send data from the given stream object.
Note
all data is submitted via stream so called by internal routines
- Parameters:
newDataStream –
reqContentType –
- Return values:
on – error returns false and stream will have been destroyed so any external references to it must be invalidated.
- inline String getBody()
Moves content from the body stream into a String.
Note
Move semantics are used to ensure that no/minimal additional memory is required. If your application has set a non-memory stream type then the method will fail and return an invalid String. The stream content will be left unchanged.
- Return values:
String –
- void reset()
reset response so it can be re-used
- void setBuffer(ReadWriteStream *buffer)
Called by connection to specify where incoming response data is written.
- Parameters:
buffer –
- void freeStreams()
release allocated stream memory
- inline bool isSuccess()
Determine if the response status indicates success.
Public Members
- HttpStatus code = HTTP_STATUS_OK
The HTTP status response code.
- HttpHeaders headers
Response headers.
- ReadWriteStream *buffer = nullptr
Internal stream for storing strings and receiving responses.
- IDataSourceStream *stream = nullptr
The body stream.
MQTT: MQ Telemetry Transport
https://en.m.wikipedia.org/wiki/MQTT
Client API
- using MqttDelegate = Delegate<int(MqttClient &client, mqtt_message_t *message)>
- using MqttRequestQueue = ObjectQueue<mqtt_message_t, MQTT_REQUEST_POOL_SIZE>
- MQTT_REQUEST_POOL_SIZE
- MQTT_CLIENT_CONNECTED
- MQTT_FLAG_RETAINED
- class MqttClient : protected TcpClient
Public Functions
- inline void setKeepAlive(uint16_t seconds)
Sets keep-alive time. That information is sent during connection to the server.
- Parameters:
seconds –
- inline void setPingRepeatTime(uint16_t seconds)
Sets the interval in which to ping the remote server if there was no activity
- Parameters:
seconds –
- bool setWill(const String &topic, const String &message, uint8_t flags = 0)
Sets last will and testament
- Parameters:
topic –
message –
flags – QoS, retain, etc flags
- Return values:
bool –
- bool connect(const Url &url, const String &uniqueClientName)
Connect to a MQTT server.
- Parameters:
url – URL in the form “mqtt://user:password@server:port” or “mqtts://user:password@server:port”
uniqueClientName –
- Return values:
bool –
- bool publish(const String &topic, const String &message, uint8_t flags = 0)
Publish a message.
- Parameters:
topic –
message – Message content as String
flags – Optional flags
- Return values:
bool –
- bool publish(const String &topic, IDataSourceStream *stream, uint8_t flags = 0)
Publish a message.
- Parameters:
topic –
message – Message content as read-only stream
flags – Optional flags
- Return values:
bool –
- bool subscribe(const String &topic)
Subscribe to a topic.
- Parameters:
topic –
- Return values:
bool –
- bool unsubscribe(const String &topic)
Unsubscribe from a topic.
- Parameters:
topic –
- Return values:
bool –
- inline void setEventHandler(mqtt_type_t type, MqttDelegate handler)
Register a callback function to be invoked on incoming event notification.
- Parameters:
type – Type of event to be notified of
handler – The callback. Pass nullptr to cancel callback.
- inline void setPayloadParser(MqttPayloadParser payloadParser = nullptr)
Sets or clears a payload parser (for PUBLISH messages from the server to us)
Note
We no longer have size limitation for incoming or outgoing messages but in order to prevent running out of memory we have a “sane” payload parser that will read up to 1K of payload
- inline void setConnectedHandler(MqttDelegate handler)
Sets a handler to be called after successful MQTT connection.
- Parameters:
handler –
- inline void setPublishedHandler(MqttDelegate handler)
Sets a handler to be called after receiving confirmation from the server for a published message from the client.
- Parameters:
handler –
- inline void setMessageHandler(MqttDelegate handler)
Sets a handler to be called after receiving a PUBLISH message from the server.
- Parameters:
handler –
- inline void setDisconnectHandler(TcpClientCompleteDelegate handler)
Sets a handler to be called on disconnect from the server.
- Parameters:
handler –
- inline void setCompleteDelegate(TcpClientCompleteDelegate completeCb = nullptr)
Set or clear the callback for connection close.
- Parameters:
completeCb – callback delegate or nullptr
Public Static Functions
- static inline uint8_t getFlags(mqtt_qos_t QoS, mqtt_retain_t retain = MQTT_RETAIN_FALSE, mqtt_dup_t dup = MQTT_DUP_FALSE)
Compute the flags value.
- Parameters:
QoS – - Quality of Service
retain – - Retain flag
dup – - Duplicate delivery
- Return values:
uint8_t – calculated flags value
NTP: Network Time Protocol
https://en.m.wikipedia.org/wiki/Network_Time_Protocol
Client API
- NTP_PORT
- NTP_PACKET_SIZE
- NTP_VERSION
- NTP_MODE_CLIENT
- NTP_MODE_SERVER
- NTP_DEFAULT_SERVER
- NTP_DEFAULT_AUTOQUERY_SECONDS
- NTP_MIN_AUTOQUERY_SECONDS
Minimum autoquery interval.
- NTP_CONNECTION_TIMEOUT_MS
Time to retry query when network connection unavailable.
- NTP_RESPONSE_TIMEOUT_MS
Time to wait before retrying NTP query.
- class NtpClient : protected UdpConnection
NTP client class.
Public Functions
- inline NtpClient()
Instantiates NTP client object.
- inline NtpClient(NtpTimeResultDelegate onTimeReceivedCb)
Instantiates NTP client object.
- Parameters:
onTimeReceivedCb – Callback delegate to be called when NTP time result is received
- NtpClient(const String &reqServer, unsigned reqIntervalSeconds, NtpTimeResultDelegate onTimeReceivedCb = nullptr)
Instantiates NTP client object.
- Parameters:
reqServer – IP address or hostname of NTP server; nullptr to use default server
reqIntervalSeconds – Quantity of seconds between NTP requests
onTimeReceivedCb – Callback delegate to be called when NTP time result is received (Default: None)
- void requestTime()
Request time from NTP server.
Note
Instigates request. Result is handled by NTP result handler function if defined
- inline void setNtpServer(const String &server)
Set the NTP server.
- Parameters:
server – IP address or hostname of NTP server
- void setAutoQuery(bool autoQuery)
Enable / disable periodic query.
- Parameters:
autoQuery – True to enable periodic query of NTP server
- void setAutoQueryInterval(unsigned seconds)
Set query period.
- Parameters:
seconds – Period in seconds between periodic queries
- inline void setAutoUpdateSystemClock(bool autoUpdateClock)
Enable / disable update of system clock.
- Parameters:
autoUpdateClock – True to update system clock with NTP result.
SMTP: Simple Mail Transfer Protocol
https://en.m.wikipedia.org/wiki/Simple_Mail_Transfer_Protocol
Client API
- enum SmtpState
Values:
- enumerator eSMTP_Banner
- enumerator eSMTP_Hello
- enumerator eSMTP_StartTLS
- enumerator eSMTP_SendAuth
- enumerator eSMTP_SendingAuthLogin
- enumerator eSMTP_RequestingAuthChallenge
- enumerator eSMTP_SendAuthResponse
- enumerator eSMTP_SendingAuth
- enumerator eSMTP_Ready
- enumerator eSMTP_SendMail
- enumerator eSMTP_SendingMail
- enumerator eSMTP_SendRcpt
- enumerator eSMTP_SendingRcpt
- enumerator eSMTP_SendData
- enumerator eSMTP_SendingData
- enumerator eSMTP_SendHeader
- enumerator eSMTP_SendingHeaders
- enumerator eSMTP_StartBody
- enumerator eSMTP_SendingBody
- enumerator eSMTP_Sent
- enumerator eSMTP_Quitting
- enumerator eSMTP_Disconnect
- using SmtpClientCallback = Delegate<int(SmtpClient &client, int code, char *status)>
- SMTP_QUEUE_SIZE
- SMTP_ERROR_LENGTH
- SMTP_CODE_SERVICE_READY
SMTP response codes
- SMTP_CODE_BYE
- SMTP_CODE_AUTH_OK
- SMTP_CODE_REQUEST_OK
- SMTP_CODE_AUTH_CHALLENGE
- SMTP_CODE_START_DATA
- SMTP_OPT_PIPELINE
- SMTP_OPT_STARTTLS
- SMTP_OPT_AUTH_PLAIN
- SMTP_OPT_AUTH_LOGIN
- SMTP_OPT_AUTH_CRAM_MD5
- class MailMessage
Public Functions
- inline MailMessage &setHeader(const String &name, const String &value)
Set a header value.
- Parameters:
name –
value –
- Return values:
MailMessage& –
- HttpHeaders &getHeaders()
Get a reference to the current set of headers.
- Return values:
HttpHeaders& –
- MailMessage &setBody(const String &body, MimeType mime = MIME_TEXT)
Sets the body of the email.
- Parameters:
body –
mime –
- Return values:
MailMessage& –
- MailMessage &setBody(String &&body, MimeType mime = MIME_TEXT) noexcept
Sets the body of the email using move semantics.
- Parameters:
body – Will be moved into message then invalidated
mime –
- Return values:
MailMessage& –
- MailMessage &setBody(IDataSourceStream *stream, MimeType mime = MIME_TEXT)
Sets the body of the email.
- Parameters:
stream –
mime –
- Return values:
MailMessage& –
- MailMessage &addAttachment(FileStream *stream)
Adds attachment to the email.
- Parameters:
stream –
- Return values:
MailMessage& –
- MailMessage &addAttachment(IDataSourceStream *stream, MimeType mime, const String &filename = "")
Adds attachment to the email.
- Parameters:
stream –
mime –
filename –
- Return values:
MailMessage& –
- MailMessage &addAttachment(IDataSourceStream *stream, const String &mime, const String &filename = "")
Adds attachment to the email.
- Parameters:
stream –
mime –
filename –
- Return values:
MailMessage& –
- class SmtpClient : protected TcpClient
Unnamed Group
- bool send(const String &from, const String &to, const String &subject, const String &body)
Queues a single message before it is sent later to the SMTP server.
- Parameters:
from –
to –
subject –
body – The body in plain text format
- Return values:
bool – true when the message was queued successfully, false otherwise
Public Functions
- bool connect(const Url &url)
Connects to remote URL.
- Parameters:
url – Provides the protocol, remote server, port and user credentials allowed protocols:
smtp - clear text SMTP
smtps - SMTP over SSL connection
- bool send(MailMessage *message)
Powerful method to queues a single message before it is sent later to the SMTP server.
- Parameters:
message –
- Return values:
bool – true when the message was queued successfully, false otherwise
- MailMessage *getCurrentMessage()
Gets the current message.
- Return values:
MailMessage* – The message, or NULL if none is scheduled
- void quit()
Sends a quit command to the server and closes the TCP connection.
- inline SmtpState getState()
Returns the current state of the SmtpClient.
- inline void onMessageSent(SmtpClientCallback callback)
Callback that will be called every time a message is sent successfully.
- Parameters:
callback –
- inline void onServerError(SmtpClientCallback callback)
Callback that will be called every an error occurs.
- Parameters:
callback –
TCP: Transmission Control Protocol
https://en.m.wikipedia.org/wiki/Transmission_Control_Protocol
Connection API
- enum TcpConnectionEvent
Values:
- enumerator eTCE_Connected
Occurs after connection establishment.
- enumerator eTCE_Received
Occurs on data receive.
- enumerator eTCE_Sent
- enumerator eTCE_Poll
- using TcpConnectionDestroyedDelegate = Delegate<void(TcpConnection&)>
- NETWORK_DEBUG
- NETWORK_SEND_BUFFER_SIZE
- class TcpConnection : public IpConnection
Subclassed by FtpDataStream, FtpServerConnection, TcpClient, TcpServer
Public Functions
- inline int writeString(const char *data, uint8_t apiflags = TCP_WRITE_FLAG_COPY)
Writes string data directly to the TCP buffer.
- Parameters:
data – null terminated string
apiflags – TCP_WRITE_FLAG_COPY, TCP_WRITE_FLAG_MORE
- Return values:
int – negative on error, 0 when retry is needed or positive on success
- inline int writeString(const String &data, uint8_t apiflags = TCP_WRITE_FLAG_COPY)
Writes string data directly to the TCP buffer.
- Parameters:
data –
apiflags – TCP_WRITE_FLAG_COPY, TCP_WRITE_FLAG_MORE
- Return values:
int – negative on error, 0 when retry is needed or positive on success
- virtual int write(const char *data, int len, uint8_t apiflags = TCP_WRITE_FLAG_COPY)
Base write operation.
- Parameters:
data –
len –
apiflags – TCP_WRITE_FLAG_COPY, TCP_WRITE_FLAG_MORE
- Return values:
int – negative on error, 0 when retry is needed or positive on success
- int write(IDataSourceStream *stream)
Writes stream data directly to the TCP buffer.
- Parameters:
stream –
apiflags – TCP_WRITE_FLAG_COPY, TCP_WRITE_FLAG_MORE
- Return values:
int – negative on error, 0 when retry is needed or positive on success
- inline void setDestroyedDelegate(TcpConnectionDestroyedDelegate destroyedDelegate)
Sets a callback to be called when the object instance is destroyed.
- Parameters:
destroyedDelegate –
Client API
- enum TcpClientState
Values:
- enumerator eTCS_Ready
- enumerator eTCS_Connecting
- enumerator eTCS_Connected
- enumerator eTCS_Successful
- enumerator eTCS_Failed
- enum TcpClientCloseAfterSentState
Values:
- enumerator eTCCASS_None
- enumerator eTCCASS_AfterSent
- enumerator eTCCASS_AfterSent_Ignore_Received
- using TcpClientEventDelegate = Delegate<void(TcpClient &client, TcpConnectionEvent sourceEvent)>
- TCP_CLIENT_TIMEOUT
- class TcpClient : public TcpConnection
Subclassed by HttpConnection, MqttClient, SmtpClient
Public Functions
- inline void setReceiveDelegate(TcpClientDataDelegate receiveCb = nullptr)
Set or clear the callback for received data.
- Parameters:
receiveCb – callback delegate or nullptr
- inline void setCompleteDelegate(TcpClientCompleteDelegate completeCb = nullptr)
Set or clear the callback for connection close.
- Parameters:
completeCb – callback delegate or nullptr
- bool send(IDataSourceStream *source, bool forceCloseAfterSent = false)
Sends data stream.
Note
This function takes ownership of the stream pointer!
- Parameters:
source – stream pointer
forceCloseAfterSent –
- Return values:
bool – true when the stream can be used. When false the stream will be deleted.
- inline void setCloseAfterSent(bool ignoreIncomingData = false)
Schedules the connection to get closed after the data is sent
- Parameters:
ignoreIncomingData – when that flag is set the connection will start ignoring incoming data.
- inline void commit()
Tries to send the pending data immediately.
Note
Call this method to decrease latency. Use it carefully.
Server API
- TCP_SERVER_TIMEOUT
- class TcpServer : public TcpConnection
Subclassed by CustomFtpServer, HttpServer
Telnet
https://en.m.wikipedia.org/wiki/Telnet
This is a very simple protocol which can be implemented using a
TcpServerclass. See TelnetServer for an example application.UDP: User Datagram Protocol
https://en.m.wikipedia.org/wiki/User_Datagram_Protocol
Connection API
- using UdpConnectionDataDelegate = Delegate<void(UdpConnection &connection, char *data, int size, IpAddress remoteIP, uint16_t remotePort)>
- class UdpConnection : public IpConnection
Subclassed by DnsServer, NtpClient, SSDP::Server, mDNS::Server
Public Functions
- bool setMulticast(IpAddress ip)
Sets the UDP multicast IP.
Note
This method works only when LWIP is compiled with LWIP_MULTICAST_TX_OPTIONS
- Parameters:
ip –
- Return values:
true – when LWIP supports this operation, false otherwise
- bool setMulticastTtl(size_t ttl)
Sets the UDP multicast Time-To-Live(TTL).
Note
This method works only when LWIP is compiled with LWIP_MULTICAST_TX_OPTIONS
- Parameters:
ttl – - time to live in hops. For example if a milticast UDP packet needs to pass through two routes to reach the receiver then the TTL should be set to 2
- Return values:
true – when LWIP supports this operation, false otherwise
URL: Uniform Resource Locator
https://en.m.wikipedia.org/wiki/URL
- URI_SCHEME_MAP(XX)
- XX(name, str, port)
Common URI scheme strings.
- class Url
Class to manage URL instance.
Note
The various URL components are stored in un-escaped format for ease of editing. Unless otherwise indicated, methods act on un-escaped text. Methods used to obtain escaped versions are clearly marked. Any attached fragment (marked bv ‘#’) in the URL is discarded
Public Functions
- inline Url(const String &urlString)
Construct a URL object from a regular escaped string.
- Parameters:
urlString – Escaped URL
- inline Url(const char *urlString)
Construct a URL object from a regular null-terminated escaped string.
- Parameters:
urlString – Escaped URL
- Url &operator=(String urlString)
Copy assignment operator.
Note
urlString is modified by so no point in passing const reference
- Parameters:
urlString – Escaped URL
- inline Url &operator=(const char *urlString)
Copy assignment operator, for C-style strings.
- Parameters:
urlString – Escaped URL
- inline int getPort() const
Obtain the actual port number to be used.
Note
If not specified, the default scheme port is returned
- Return values:
int –
- String getHostWithPort() const
Get hostname+port part of URL string.
Note
Neither of these is subject to escaping
- Return values:
String –
- String getPathWithQuery() const
Get path with any query parameters attached.
Note
Both path and query values are escaped
- Return values:
String –
- void debugPrintTo(Print &p) const
Printable output for debugging.
- Parameters:
p –
Public Members
- int Port = 0
Undefined by default.
WebSocket Protocol
https://en.m.wikipedia.org/wiki/WebSocket
Connection API
- enum WsConnectionState
Current state of Websocket connection.
Values:
- enumerator eWSCS_Ready
- enumerator eWSCS_Open
- enumerator eWSCS_Closed
- using WebsocketList = Vector<WebsocketConnection*>
- using WebsocketDelegate = Delegate<void(WebsocketConnection&)>
- using WebsocketMessageDelegate = Delegate<void(WebsocketConnection&, const String&)>
- using WebsocketBinaryDelegate = Delegate<void(WebsocketConnection&, uint8_t *data, size_t size)>
- WEBSOCKET_VERSION
- struct WsFrameInfo
- class WebsocketConnection
Subclassed by WebsocketClient
Public Functions
- WebsocketConnection(HttpConnection *connection = nullptr, bool isClientConnection = true)
Constructs a websocket connection on top of http client or server connection.
- Parameters:
connection – the transport connection
isClientConnection – true when the passed connection is an http client connection
- bool bind(HttpRequest &request, HttpResponse &response)
Binds websocket connection to an http server connection.
- Parameters:
request –
response –
- Return values:
bool – true on success, false otherwise
- bool send(const char *message, size_t length, ws_frame_type_t type = WS_FRAME_TEXT)
Sends a websocket message from a buffer.
- Parameters:
message –
length – Quantity of data in message
type –
- inline bool send(const String &message, ws_frame_type_t type = WS_FRAME_TEXT)
Sends websocket message from a String.
Note
A String may contain arbitrary data, not just text, so can use this for any frame type
- Parameters:
message – String
type –
- bool send(IDataSourceStream *source, ws_frame_type_t type = WS_FRAME_TEXT, bool useMask = false, bool isFin = true)
Sends websocket message from a stream.
- Parameters:
source – The stream to send - we get ownership of the stream
type –
useMask – MUST be true for client connections
isFin – true if this is the final frame
- Return values:
bool – true on success
- inline bool sendString(const String &message)
Sends a string websocket message.
- Parameters:
message –
- inline bool sendBinary(const uint8_t *data, size_t length)
Sends a binary websocket message.
- Parameters:
data –
length –
- void close()
Closes a websocket connection (without closing the underlying http connection)
- void reset()
Resets a websocket connection.
- inline void setUserData(void *userData)
Attaches a user data to a websocket connection.
- Parameters:
userData –
- inline void *getUserData()
Retrieves user data attached.
- Return values:
void* – The user data previously set by
setUserData()
- inline bool operator==(const WebsocketConnection &rhs) const
Test if another connection refers to the same object.
- Parameters:
rhs – The other WebsocketConnection to compare with
- Return values:
bool –
- inline void setConnectionHandler(WebsocketDelegate handler)
Sets the callback handler to be called after successful websocket connection.
- Parameters:
handler –
- inline void setMessageHandler(WebsocketMessageDelegate handler)
Sets the callback handler to be called after a websocket message is received.
- Parameters:
handler –
- inline void setBinaryHandler(WebsocketBinaryDelegate handler)
Sets the callback handler to be called after a binary websocket message is received.
- Parameters:
handler –
- inline void setPongHandler(WebsocketDelegate handler)
Sets the callback handler to be called when pong reply received.
- Parameters:
handler –
- inline void setDisconnectionHandler(WebsocketDelegate handler)
Sets the callback handler to be called before closing a websocket connection.
- Parameters:
handler –
- void activate()
Should be called after a websocket connection is established to activate the websocket parser and allow sending of websocket data.
- bool onConnected()
Call this method when the websocket connection was (re)activated.
- Return values:
bool – true on success
- inline HttpConnection *getConnection()
Gets the underlying HTTP connection.
- Return values:
HttpConnection* –
- void setConnection(HttpConnection *connection, bool isClientConnection = true)
Sets the underlying (transport ) HTTP connection.
- Parameters:
connection – the transport connection
isClientConnection – true when the passed connection is an http client connection
- inline WsConnectionState getState()
Gets the state of the websocket connection.
- Return values:
WsConnectionState –
Public Static Functions
- static void broadcast(const char *message, size_t length, ws_frame_type_t type = WS_FRAME_TEXT)
Broadcasts a message to all active websocket connections.
- Parameters:
message –
length –
type –
- static inline void broadcast(const String &message, ws_frame_type_t type = WS_FRAME_TEXT)
Broadcasts a message to all active websocket connections.
- Parameters:
message –
type –
- static inline const WebsocketList &getActiveWebsockets()
Obtain the list of active websockets.
Note
Return value is const as only restricted operations should be carried out on the list.
- Return values:
const – WebsocketList&
Client API
- class WebsocketClient : protected WebsocketConnection
Websocket Client.
Public Functions
- bool connect(const Url &url)
Connects websocket client to server.
- Parameters:
url – Url address of websocket server
- inline void sendPing(const String &payload = nullptr)
Send websocket ping to server.
- Parameters:
payload – Maximum 255 bytes
- Return values:
bool – true if the data can be send, false otherwise
- inline void sendPong(const String &payload = nullptr)
Send websocket ping to server.
- Parameters:
payload – Maximum 255 bytes
- Return values:
bool – true if the data can be send, false otherwise
- inline void setSslInitHandler(Ssl::Session::InitDelegate handler)
Set the SSL session initialisation callback.
- Parameters:
handler –
- inline void setBinaryHandler(WebsocketBinaryDelegate handler)
Sets the callback handler to be called after a binary websocket message is received.
- Parameters:
handler –
- inline void setConnectionHandler(WebsocketDelegate handler)
Sets the callback handler to be called after successful websocket connection.
- Parameters:
handler –
- inline void setDisconnectionHandler(WebsocketDelegate handler)
Sets the callback handler to be called before closing a websocket connection.
- Parameters:
handler –
- inline void setMessageHandler(WebsocketMessageDelegate handler)
Sets the callback handler to be called after a websocket message is received.
- Parameters:
handler –
- bool send(const char *message, size_t length, ws_frame_type_t type = WS_FRAME_TEXT)
Sends a websocket message from a buffer.
- Parameters:
message –
length – Quantity of data in message
type –
- inline bool send(const String &message, ws_frame_type_t type = WS_FRAME_TEXT)
Sends websocket message from a String.
Note
A String may contain arbitrary data, not just text, so can use this for any frame type
- Parameters:
message – String
type –
- bool send(IDataSourceStream *source, ws_frame_type_t type = WS_FRAME_TEXT, bool useMask = false, bool isFin = true)
Sends websocket message from a stream.
- Parameters:
source – The stream to send - we get ownership of the stream
type –
useMask – MUST be true for client connections
isFin – true if this is the final frame
- Return values:
bool – true on success
- inline bool sendBinary(const uint8_t *data, size_t length)
Sends a binary websocket message.
- Parameters:
data –
length –
- inline bool sendString(const String &message)
Sends a string websocket message.
- Parameters:
message –
- void close()
Closes a websocket connection (without closing the underlying http connection)
- inline WsConnectionState getState()
Gets the state of the websocket connection.
- Return values:
WsConnectionState –
Other networking libraries:
Configuration variables
- ENABLE_WIFI_DEBUG
default: 0 (disabled)
RP2040 only
Set to 1 to enable additional debugging output for processing WiFi events.
References
Used by
Sming (main) Component
TelnetServer Sample
Multipart Parser Library
Network Ping Sample
Environment Variables
SoC support
Storage Management
This Component provides support for using storage devices in a structured way by partitioning them into areas for specific uses. Partitions may can contain information such as:
Application (firmware) images
Filesystem(s)
Configuration/calibration/parameter data
Custom flash storage areas
A single partition table is located on the main flash device,
Storage::spiFlash, and defines all partitions with a unique name and associatedStorage::Partition::Type/Storage::Partition::SubType.Hardware configuration
Each project has an associated
Hardware configuration, specified by theHWCONFIGsetting: this is a JSON file with a.hwextension.For user convenience, the configuration file may contain comments however these are stripped before processing.
The build system locates the file by searching, in order:
{PROJECT_DIR}the root project directory
{SMING_HOME}/Arch/{SMING_ARCH}
{SMING_HOME}Each architecture provides a
standardconfiguration which defines such things as the partition table location and standard system partitions. Other configurations inherit from this by providing abase_configvalue.You can list the available configs like this:
make hwconfig-listThis also shows the file path should you wish to view or edit it.
To select and view the resulting configuration, do this:
make hwconfig HWCONFIG=spiffsor, to show the partition map:
make map HWCONFIG=spiffsNote
You can set
HWCONFIGin your project’scomponent.mkfile, however as with other configuration variables it will be overridden by the cached value set on the command line.For example, if you want to change from
standardtostandard-4mfor your project, first add this line to your component.mk file:HWCONFIG := standard-4mThen either run
make HWCONFIG=standard-4mormake config-clean.Hardware configuration options
Commonly used settings can be stored in an option library for easier use. The library files are named
options.jsonand located in the place as .hw files.For example, we can do this:
make HWCONFIG=standard HWCONFIG_OPTS=4m,spiffsThis loads the ‘standard’ profile then merges the fragments found in the option library with the given names. This is how the
standard-4mprofile is constructed.If using this approach, remember to updated your project’s
component.mkwith the desired settings, and verify the layout is correct usingmake map.OTA updates
When planning OTA updates please check that the displayed partition map corresponds to your project. For example, the partition table requires a free sector so must not overlap other partitions.
Your OTA update process must include a step to write the partition table to the correct location. See Partition table migration.
It is not necessary to update the bootloader. See rBoot for further information.
Custom configurations
To customise the hardware configuration for a project, for example ‘my_project’:
Create a new configuration file in your project root, such as
my_project.hw:{ "name": "My project config", "base_config": "spiffs", "options": ["vdd"] }You can use any available configuration as the base_config. Option fragments can be pulled in as shown. See Hardware configuration options.
If required, modify any inherited settings:
{ "name": "My config", "base_config": "standard", "devices": { "spiFlash": { "speed": 80, "mode": "qio", "size": "2M" } }, "partitions": { "rom0": { "address": "0x10000", "size": "0x80000" } } }This will adjust flash parameters (previously via SPI_SPEED, SPI_MODE and SPI_SIZE), and the location/size of the primary application partition.
Add any additional partitions:
{ "name": "My config", "base_config": "standard-4m", "partitions": { "rom0": { "address": "0x10000", "size": "0x80000" }, "spiffs1": { "address": "0x00280000", "size": "256K", "type": "data", "subtype": "spiffs", "filename": "$(FW_BASE)/spiffs1_rom.bin", "build": { "target": "spiffsgen", "files": "files/spiffs1" } } } }This adds a second SPIFFS partition, and instructs the build system to generate an image file for it using the files in the project’s
files/spiffs1directory.Select the new configuration and re-build the project:
make HWCONFIG=my_projectYou should also add this to your project’s
component.mkfile:HWCONFIG := my_projectProgram your device:
make flashThis will flash everything: bootloader, partition table and all defined partitions (those with a
filenameentry).Note
The build system isn’t smart enough to track dependencies for partition build targets.
To rebuild these manually type:
make buildpartThese will be removed when
make cleanis run, but you can also clean them separately thus:make part-cleanPartition maps
This is a concise view of your flash partitions. Display it like this:
make mapFor the Basic Storage sample application, we get this:
Basic_Storage: Invoking 'map' for Esp8266 (debug) architecture Partition map: Device Start End Size Type SubType Name Filename ---------------- ---------- ---------- ---------- -------- -------- ---------------- ------------ spiFlash 0x00000000 0x00001fff 8K Boot Sector spiFlash 0x00002000 0x00002fff 4K Partition Table spiFlash 0x00003000 0x00003fff 4K data phy phy_init $(FLASH_INIT_DATA) spiFlash 0x00004000 0x00007fff 16K data sysparam sys_param spiFlash 0x00008000 0x000fffff 992K app factory rom0 $(RBOOT_ROM_0_BIN) spiFlash 0x00100000 0x001effff 960K (unused) spiFlash 0x001f0000 0x001f3fff 16K user 0 user0 user0.bin spiFlash 0x001f4000 0x001f7fff 16K user 1 user1 spiFlash 0x001f8000 0x001fffff 32K (unused) spiFlash 0x00200000 0x0027ffff 512K data spiffs spiffs0 $(SPIFF_BIN_OUT) spiFlash 0x00280000 0x002bffff 256K data spiffs spiffs1 $(FW_BASE)/spiffs1_rom.bin spiFlash 0x002c0000 0x002fffff 256K data spiffs spiffs2 $(FW_BASE)/spiffs2_rom.bin spiFlash 0x00300000 0x003fffff 1M (unused)For comparison, here’s the output for Esp32:
Basic_Storage: Invoking 'map' for Esp32 (debug) architecture Partition map: Device Start End Size Type SubType Name Filename ---------------- ---------- ---------- ---------- -------- -------- ---------------- ------------ spiFlash 0x00000000 0x00007fff 32K Boot Sector spiFlash 0x00008000 0x00008fff 4K Partition Table spiFlash 0x00009000 0x0000efff 24K data nvs nvs spiFlash 0x0000f000 0x0000ffff 4K data phy phy_init spiFlash 0x00010000 0x001fffff 1984K app factory factory $(TARGET_BIN) spiFlash 0x001f0000 0x001f3fff 16K user 0 user0 user0.bin spiFlash 0x001f4000 0x001f7fff 16K user 1 user1 spiFlash 0x001f8000 0x001fffff 32K (unused) spiFlash 0x00200000 0x0027ffff 512K data spiffs spiffs0 $(SPIFF_BIN_OUT) spiFlash 0x00280000 0x002bffff 256K data spiffs spiffs1 $(FW_BASE)/spiffs1_rom.bin spiFlash 0x002c0000 0x002fffff 256K data spiffs spiffs2 $(FW_BASE)/spiffs2_rom.bin spiFlash 0x00300000 0x003fffff 1M (unused)To compare this with the partition map programmed into a device, do this:
make readmap mapJSON validation
When the binary partition table is built or updated, the configuration is first validated against a schema Sming/Components/Storage/schema.json.
This complements the checks performed by the
hwconfigtool.You can run the validation manually like this:
make hwconfig-validateSee JSON Schema for details about JSON schemas.
Configuration
- HWCONFIG
default: standard
Set this to the hardware configuration to use for your project.
Default configurations:
- standard
Base profile with 1MB flash size which should work on all device variants. Located in the
Sming/Arch/{SMING_ARCH}directory.
- standard-4m
Overrides
standardto set 4Mbyte flash size
- spiffs
Adds a single SPIFFS partition. See SPIFFS IFS Library.
Other configurations may be available, depending on architecture. You can see these by running
make hwconfig-list.For example, to select
spiffsadd the following line to your project:HWCONFIG := spiffsYou will also need to run
make HWCONFIG=spiffsto change the cached value (ormake config-cleanto reset everything).
- HWCONFIG_OPTS
Set this to adjust the hardware profile using option fragments. See Hardware configuration options.
- ENABLE_STORAGE_SIZE64
Build with
ENABLE_STORAGE_SIZE64=1to enable support for storage devices of more than 4GB capacity.Device and partition addresses and sizes use the
storage_size_ttype, which by default isuint32_t. Setting this value changes it touint64_t.When enabling this setting, care must be taken in code especially with
printfstyle format strings such as in debug statements. The safest way to handle both cases is like this:debug_i("Partition size: %llu", uint64_t(part.size()));Binary partition table
Sming uses the same binary partition table structure as ESP-IDF, located immediately after the boot sector. However, it is organised slightly differently to allow partitions to be registered for multiple storage devices.
Entries are fixed 32-byte structures,
Storage::esp_partition_info_t, organised as follows:
The first entry is always a
storagetype defining the mainspiFlashdevice.This is followed by regular partition entries sorted in ascending address order. There may be gaps between the partitions.
The partition table md5sum entry is inserted as normal
If any external devices are defined: - A SMING_EXTENSION entry, which the esp32 bootloader interprets as the end of the partition table. - The next entry is a
storagetype for theexternaldevice. - This is followed by regular partition entries as before. - A second md5sum entry is inserted for the entire partition table thus farThe end of the partition table is identified by an empty sector (i.e. all bytes 0xFF).
Partition API
This is a C++ interface. Some examples:
Storage::Partition part = Storage::findPartition("spiffs0"); // Find by name if(part) { Serial << part << endl; } else { Serial << "spiffs0 partition NOT Found" << endl; } // Enumerate all partitions for(auto part: Storage::findPartition()) { Serial << part << endl; } // Enumerate all SPIFFS partitions for(auto part: Storage::findPartition(Storage::Partition::SubType::Data::spiffs)) { Serial << part << endl; }A
Storage::Partitionobject is just a wrapper and can be freely copied around. It defines methods which should be used to read/write/erase the partition contents.Each partition has an associated
Storage::Device. This is usuallyStorage::spiFlashfor the main flash device.Other devices must be registered via
Storage::PartitionTable::registerStorageDevice().You can query partition entries from a Storage object directly, for example:
#include <Storage/SpiFlash.h> for(auto part: Storage::spiFlash->partitions()) { Serial << part << endl; }External Storage
If your design has additional fixed storage devices, such as SPI RAM, flash or EEPROM, you can take advantage of the partition API to manage them as follows:
Implement a class to manage the storage, inheriting from
Storage::Device.Create a custom hardware configuration for your project and add a
devicesentry describing your storage device, plus partition entries: thedevicefield identifies which device these entries relate to.Create an instance of your custom device and make a call to
Storage::registerDevice()in yourinit()function (or elsewhere if more appropriate).See Disk Storage for how devices such as SD flash cards are managed.
API
- namespace Storage
Get power of 2 for given value
See also
Use
isLog2()to confirm value is power of 2
- param value:
Must be an exact power of 2
- retval uint8_t:
Result n such that
value == 1 << nEnums
Functions
- inline bool isSize64(uint64_t value)
Determine if a value requires 64-bits to store.
- inline bool isSize64(int64_t value)
Determine if a value requires 64-bits to store.
- void initialize()
Called early in the startup phase.
- bool registerDevice(Device *device)
Register a storage device.
- Return values:
bool – true on success, false if another device already registered with same name
- bool unRegisterDevice(Device *device)
Unregister a storage device.
Use extreme care: behaviour is unpredictable if partitions are in use
- inline Iterator findPartition(Partition::Type type = Partition::Type::any, uint8_t subType = Partition::SubType::any)
Find partitions of the given type.
Variables
- constexpr uint16_t ESP_PARTITION_MAGIC = {0x50AA}
Identifies a valid partition.
- constexpr uint16_t ESP_PARTITION_MAGIC_MD5 = {0xEBEB}
Identifies an MD5 hash block.
- constexpr size_t ESP_PARTITION_TABLE_MAX_LEN = {0xC00}
- class FileDevice : public Storage::Device
Create custom storage device using backing file.
Public Functions
- inline FileDevice(const String &name, IFS::IFileSystem &fileSys, IFS::FileHandle file, storage_size_t size)
Construct a file device with custom size.
- Parameters:
name – Name of device
fileSys – File system where file is located
file – Handle to open file
size – Size of device in bytes
- inline FileDevice(const String &name, IFS::IFileSystem &fileSys, IFS::FileHandle file)
Construct a device using existing file.
Device will match size of existing file
- Parameters:
name – Name of device
fileSys – File system where file is located
file – Handle to open file
- inline virtual Type getType() const override
Obtain device type.
- inline virtual storage_size_t getSize() const override
Obtain addressable size of this device.
- Return values:
storage_size_t – Must be at least as large as the value declared in the hardware configuration
- inline virtual size_t getBlockSize() const override
Obtain smallest allocation unit for erase operations.
- virtual bool read(storage_size_t address, void *buffer, size_t len) override
Read data from the storage device.
- Parameters:
address – Where to start reading
dst – Buffer to store data
size – Size of data to be read, in bytes.
- Return values:
bool – true on success, false on error
- virtual bool write(storage_size_t address, const void *data, size_t len) override
Write data to the storage device.
- Parameters:
address – Where to start writing
src – Data to write
size – Size of data to be written, in bytes.
- Return values:
bool – true on success, false on error
- virtual bool erase_range(storage_size_t address, storage_size_t len) override
Erase a region of storage in preparation for writing.
- Parameters:
address – Where to start erasing
size – Size of region to erase, in bytes
- Return values:
bool – true on success, false on error
- class Device : public LinkedObjectTemplate<Device>
Represents a storage device (e.g. flash memory)
Subclassed by Storage::Disk::BlockDevice, Storage::FileDevice, Storage::ProgMem, Storage::SpiFlash, Storage::StreamDevice, Storage::SysMem
Public Types
Public Functions
- inline const PartitionTable &partitions() const
Provide read-only access to partition table.
- inline PartitionTable &editablePartitions()
Provide full access to partition table.
- inline bool loadPartitions(uint32_t tableOffset)
Load partition table entries @tableOffset Location of partition table to read.
- Return values:
bool – true on success, false on failure
- bool loadPartitions(Device &source, uint32_t tableOffset)
Load partition table entries from another table.
- Parameters:
source – Device to load entries from @tableOffset Location of partition table to read
- Return values:
bool – true on success, false on failure
- inline virtual uint32_t getId() const
Obtain device ID.
- Return values:
uint32_t – typically flash chip ID
- virtual size_t getBlockSize() const = 0
Obtain smallest allocation unit for erase operations.
- virtual storage_size_t getSize() const = 0
Obtain addressable size of this device.
- Return values:
storage_size_t – Must be at least as large as the value declared in the hardware configuration
- virtual bool read(storage_size_t address, void *dst, size_t size) = 0
Read data from the storage device.
- Parameters:
address – Where to start reading
dst – Buffer to store data
size – Size of data to be read, in bytes.
- Return values:
bool – true on success, false on error
- virtual bool write(storage_size_t address, const void *src, size_t size) = 0
Write data to the storage device.
- Parameters:
address – Where to start writing
src – Data to write
size – Size of data to be written, in bytes.
- Return values:
bool – true on success, false on error
- virtual bool erase_range(storage_size_t address, storage_size_t size) = 0
Erase a region of storage in preparation for writing.
- Parameters:
address – Where to start erasing
size – Size of region to erase, in bytes
- Return values:
bool – true on success, false on error
- inline virtual uint16_t getSectorSize() const
Get sector size, the unit of allocation for block-access devices.
Override this method only if the device does not support standard 512-byte sector access. For example, ‘Advanced-Format’ drives use 4096-byte sectors.
- inline virtual storage_size_t getSectorCount() const
Obtain total number of sectors on this device.
- inline virtual bool sync()
Flush any pending writes to the physical media.
Devices with intermediate buffering should implement this method.
- Return values:
bool – Return false if sync operation failed.
- class Iterator
- class Partition
Represents a flash partition.
Confirm partition is of the expected type
- bool verify(Type type, uint8_t subtype) const
Strong C++ type value.
- Parameters:
type – Expected partition type
subtype – Expected partition sub-type
- Return values:
bool – true if type is OK, false if not. Logs debug messages on failure.
- inline bool verify(uint8_t type, uint8_t subtype) const
Weak ‘type’ value.
Public Functions
- bool read(storage_size_t offset, void *dst, size_t size)
Read data from the partition.
- Parameters:
offset – Where to start reading, relative to start of partition
dst – Buffer to store data
size – Size of data to be read, in bytes.
- Return values:
bool – true on success, false on error
- bool write(storage_size_t offset, const void *src, size_t size)
Write data to the partition.
Note
Flash region must be erased first
- Parameters:
offset – Where to start writing, relative to start of partition
src – Data to write
size – Size of data to be written, in bytes.
- Return values:
bool – true on success, false on error
- bool erase_range(storage_size_t offset, storage_size_t size)
Erase part of the partition.
Note
Both offset and size must be aligned to flash sector size (4Kbytes)
- Parameters:
offset – Where to start erasing, relative to start of partition
size – Size of region to erase, in bytes
- Return values:
bool – true on success, false on error
- inline uint8_t subType() const
Obtain partition sub-type.
- inline storage_size_t address() const
Obtain partition starting address.
- Return values:
storage_size_t – Device address
- inline storage_size_t lastAddress() const
Obtain address of last byte in this this partition.
- Return values:
storage_size_t – Device address
- inline storage_size_t size() const
Obtain partition size.
- Return values:
storage_size_t – Size in bytes
- inline Flags flags() const
Get partition flags.
- inline bool isEncrypted() const
Check state of partition
encryptedflag.
- inline bool isReadOnly() const
Check state of partition
readOnlyflag.
- bool getDeviceAddress(storage_size_t &address, storage_size_t size) const
Get corresponding storage device address for a given partition offset.
- Parameters:
address – IN: Zero-based offset within partition, OUT: Device address
size – Size of data to be accessed
- Return values:
bool – true on success, false on failure Fails if the given offset/size combination is out of range, or the partition is undefined.
- String getDeviceName() const
Get name of storage device for this partition.
- Return values:
String –
- inline bool contains(storage_size_t addr) const
Determine if given address contained within this partition.
- size_t getBlockSize() const
Obtain smallest allocation unit for erase operations.
- uint16_t getSectorSize() const
Get sector size for block-addressable devices.
See also
- inline storage_size_t getSectorCount() const
Obtain total number of sectors in this partition.
- bool sync()
Flush any pending writes to the physical media.
See also
Public Static Functions
- struct FullType
Express both partition type and subtype together.
- struct Info : public LinkedObjectTemplate<Info>, public Printable
Partition information.
Subclassed by Storage::Disk::PartInfo
- struct SubType
- struct esp_partition_info_t
Internal structure describing the binary layout of a partition table entry.
- class PartitionStream : public ReadWriteStream
Stream operating directory on a Storage partition.
To support write operations, the target region must be erased first.
Public Functions
- inline PartitionStream(Partition partition, storage_size_t offset, size_t size, bool blockErase)
Access part of a partition using a stream.
If blockErase is false then region must be pre-erased before writing.
- Deprecated:
Use
modeparameter instead ofblockErase
- Parameters:
partition –
offset – Limit access to this starting offset
size – Limit access to this number of bytes from starting offset
blockErase – Set to true to erase blocks before writing
- inline PartitionStream(Partition partition, bool blockErase)
Access entire partition using stream.
If blockErase is false then partition must be pre-erased before writing.
- Deprecated:
Use
modeparameter instead ofblockErase
- Parameters:
partition –
blockErase – Set to true to erase blocks before writing
- inline PartitionStream(Partition partition, storage_size_t offset, size_t size, Mode mode = Mode::ReadOnly)
Access part of a partition using a stream.
Note
When writing in Mode::BlockErase, block erasure is only performed at the start of each block. Therefore if
offsetis not a block boundary then the corresponding block will not be erased first.
- Parameters:
partition –
offset – Limit access to this starting offset
size – Limit access to this number of bytes from starting offset
mode –
- inline PartitionStream(Partition partition, Mode mode = Mode::ReadOnly)
Access entire partition using stream.
- Parameters:
partition –
mode – If blockErase is false then partition must be pre-erased before writing.
- inline virtual int available() override
Return the total length of the stream.
- Return values:
int – -1 is returned when the size cannot be determined
- virtual uint16_t readMemoryBlock(char *data, int bufSize) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
- virtual int seekFrom(int offset, SeekOrigin origin) override
Change position in stream.
Note
This method is implemented by streams which support random seeking, such as files and memory streams.
- Parameters:
offset –
origin –
- Return values:
New – position, < 0 on error
- virtual size_t write(const uint8_t *buffer, size_t size) override
Write chars to stream.
Note
Although this is defined in the Print class, ReadWriteStream uses this as the core output method so descendants are required to implement it
- Parameters:
buffer – Pointer to buffer to write to the stream
size – Quantity of chars to write
- Return values:
size_t – Quantity of chars written to stream
- inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
- class PartitionTable
Subclassed by Storage::ProgMem::ProgMemPartitionTable, Storage::SysMem::SysMemPartitionTable
Partition search
Public Functions
- inline Partition find(const String &name) const
Find partition by name.
- Parameters:
Name – Name to search for, case-sensitive
- Return values:
Partition – Names are unique so at most only one match
- class ProgMem : public Storage::Device
Storage device to access PROGMEM using flash API.
Public Functions
- inline virtual size_t getBlockSize() const override
Obtain smallest allocation unit for erase operations.
- inline virtual storage_size_t getSize() const override
Obtain addressable size of this device.
- Return values:
storage_size_t – Must be at least as large as the value declared in the hardware configuration
- inline virtual Type getType() const override
Obtain device type.
- virtual bool read(storage_size_t address, void *dst, size_t size) override
Read data from the storage device.
- Parameters:
address – Where to start reading
dst – Buffer to store data
size – Size of data to be read, in bytes.
- Return values:
bool – true on success, false on error
- inline virtual bool write(storage_size_t address, const void *src, size_t size) override
Write data to the storage device.
- Parameters:
address – Where to start writing
src – Data to write
size – Size of data to be written, in bytes.
- Return values:
bool – true on success, false on error
- inline virtual bool erase_range(storage_size_t address, storage_size_t size) override
Erase a region of storage in preparation for writing.
- Parameters:
address – Where to start erasing
size – Size of region to erase, in bytes
- Return values:
bool – true on success, false on error
- class ProgMemPartitionTable : public Storage::PartitionTable
Public Functions
- Partition add(const String &name, const void *flashPtr, size_t size, Partition::FullType type)
Add partition entry for PROGMEM data access.
- Parameters:
name – Name for partition
flashPtr – PROGMEM pointer
size – Size of PROGMEM data
type – Partition type and subtype
- Return values:
Partition – Invalid if data is not progmem
- class SpiFlash : public Storage::Device
Main flash storage device.
Public Functions
- virtual size_t getBlockSize() const override
Obtain smallest allocation unit for erase operations.
- virtual storage_size_t getSize() const override
Obtain addressable size of this device.
- Return values:
storage_size_t – Must be at least as large as the value declared in the hardware configuration
- inline virtual Type getType() const override
Obtain device type.
- virtual uint32_t getId() const override
Obtain device ID.
- Return values:
uint32_t – typically flash chip ID
- virtual bool read(storage_size_t address, void *dst, size_t size) override
Read data from the storage device.
- Parameters:
address – Where to start reading
dst – Buffer to store data
size – Size of data to be read, in bytes.
- Return values:
bool – true on success, false on error
- virtual bool write(storage_size_t address, const void *src, size_t size) override
Write data to the storage device.
- Parameters:
address – Where to start writing
src – Data to write
size – Size of data to be written, in bytes.
- Return values:
bool – true on success, false on error
- virtual bool erase_range(storage_size_t address, storage_size_t size) override
Erase a region of storage in preparation for writing.
- Parameters:
address – Where to start erasing
size – Size of region to erase, in bytes
- Return values:
bool – true on success, false on error
- class StreamDevice : public Storage::Device
Read-only partition on a stream object.
Note
Writes not possible as streams always append data, cannot do random writes
Public Functions
- inline virtual Type getType() const override
Obtain device type.
- inline virtual bool read(storage_size_t address, void *buffer, size_t len) override
Read data from the storage device.
- Parameters:
address – Where to start reading
dst – Buffer to store data
size – Size of data to be read, in bytes.
- Return values:
bool – true on success, false on error
- inline virtual bool write(storage_size_t address, const void *data, size_t len) override
Write data to the storage device.
- Parameters:
address – Where to start writing
src – Data to write
size – Size of data to be written, in bytes.
- Return values:
bool – true on success, false on error
- inline virtual bool erase_range(storage_size_t address, storage_size_t len) override
Erase a region of storage in preparation for writing.
- Parameters:
address – Where to start erasing
size – Size of region to erase, in bytes
- Return values:
bool – true on success, false on error
- class SysMem : public Storage::Device
Storage device to access system memory, e.g. RAM.
Public Functions
- inline virtual size_t getBlockSize() const override
Obtain smallest allocation unit for erase operations.
- inline virtual storage_size_t getSize() const override
Obtain addressable size of this device.
- Return values:
storage_size_t – Must be at least as large as the value declared in the hardware configuration
- inline virtual Type getType() const override
Obtain device type.
- inline virtual bool read(storage_size_t address, void *buffer, size_t len) override
Read data from the storage device.
- Parameters:
address – Where to start reading
dst – Buffer to store data
size – Size of data to be read, in bytes.
- Return values:
bool – true on success, false on error
- inline virtual bool write(storage_size_t address, const void *data, size_t len) override
Write data to the storage device.
- Parameters:
address – Where to start writing
src – Data to write
size – Size of data to be written, in bytes.
- Return values:
bool – true on success, false on error
- inline virtual bool erase_range(storage_size_t address, storage_size_t len) override
Erase a region of storage in preparation for writing.
- Parameters:
address – Where to start erasing
size – Size of region to erase, in bytes
- Return values:
bool – true on success, false on error
- class SysMemPartitionTable : public Storage::PartitionTable
- namespace Debug
- namespace Disk
Enums
Functions
- uint32_t crc32_byte(uint32_t crc, uint8_t d)
- uint32_t crc32(uint32_t bcc, const void *data, size_t length)
- inline uint32_t crc32(const void *data, size_t length)
- Error formatDisk(BlockDevice &device, GPT::PartitionTable &table, const Uuid &diskGuid = {})
Partition a device using the GPT scheme.
- Parameters:
device –
table – Partitions to create
- Return values:
Error –
- Error formatDisk(BlockDevice &device, MBR::PartitionTable &table)
Partition a device using the MBR scheme.
- Parameters:
device –
table – Partitions to create
- Return values:
Error –
- inline SysType getSysTypeFromIndicator(SysIndicator si)
- Error validate(BasePartitionTable &table, storage_size_t firstAvailableBlock, storage_size_t totalAvailableBlocks, uint32_t blockSize)
Validate partition table entries.
If size <= 100 then the actual size is calculated as a percentage and updated
If offset = 0 then a suitable location is found and the offset updated
On success, partition entries are ordered by position.
- Parameters:
firstAvailableBlock – First block number which may be allocated to a partition
totalAvailableBlocks – Number of blocks available for partition allocation
blockSize – Size of a block
- Return values:
Error – For each partition:
Variables
- constexpr uint32_t PARTITION_ALIGN = {0x100000U}
- class BlockDevice : public Storage::Device
Base class for sector-addressable (block) devices.
Inherited classes must set the
sectorCountvalue, and implement the fourraw_xxxmethods.This class supports byte-level access using internal buffering, which if required must be enabled by the application via the
allocateBuffersmethod. Without buffering, read/writes must always be sector-aligned. Rrase must always be sector-aligned.For power-loss resiliency it is important to call
sync()at appropriate times. Filing system implementations should do this after closing a file, for example. Applications should consider this if leaving files open for extended periods, and explicitly call ‘flush’ on the filing system or ‘sync’ on the partition or device if necessary.Subclassed by Storage::Disk::HostFileDevice, Storage::SD::Card, USB::MSC::LogicalUnit
Public Functions
- virtual bool read(storage_size_t address, void *dst, size_t size) override
Read data from the storage device.
- Parameters:
address – Where to start reading
dst – Buffer to store data
size – Size of data to be read, in bytes.
- Return values:
bool – true on success, false on error
- virtual bool write(storage_size_t address, const void *src, size_t size) override
Write data to the storage device.
- Parameters:
address – Where to start writing
src – Data to write
size – Size of data to be written, in bytes.
- Return values:
bool – true on success, false on error
- virtual bool erase_range(storage_size_t address, storage_size_t size) override
Erase a region of storage in preparation for writing.
- Parameters:
address – Where to start erasing
size – Size of region to erase, in bytes
- Return values:
bool – true on success, false on error
- inline virtual size_t getBlockSize() const override
Obtain smallest allocation unit for erase operations.
- inline virtual storage_size_t getSize() const override
Obtain addressable size of this device.
- Return values:
storage_size_t – Must be at least as large as the value declared in the hardware configuration
- inline virtual storage_size_t getSectorCount() const override
Obtain total number of sectors on this device.
- virtual bool sync() override
Flush any pending writes to the physical media.
Devices with intermediate buffering should implement this method.
- Return values:
bool – Return false if sync operation failed.
- bool allocateBuffers(unsigned numBuffers)
Set number of sector buffers to use.
Required to support byte-level read/write operations on block devices. Buffering can improve performance, with diminishing returns above around 4 sectors.
- Parameters:
numBuffers – Number of buffers to allocate 1,2,4,8,etc. Pass 0 to deallocate/disable buffering.
- Return values:
bool – false on memory allocation error, or if failed to flush existing buffers to disk
- struct Buffer : public std::unique_ptr<uint8_t[]>
- class BufferList
- class HostFileDevice : public Storage::Disk::BlockDevice
Create custom storage device using backing file.
Public Functions
- HostFileDevice(const String &name, const String &filename, storage_size_t size)
Construct a file device with custom size.
- Parameters:
name – Name of device
filename – Path to file
size – Size of device in bytes
- HostFileDevice(const String &name, const String &filename)
Construct a device using existing file.
Device will match size of existing file
- Parameters:
name – Name of device
filename – Path to file
- inline virtual Type getType() const override
Obtain device type.
- struct DiskPart
Adds information specific to MBR/GPT disk partitions.
Subclassed by Storage::Disk::PartInfo
Public Functions
- size_t printTo(Print &p) const
Print full contents of this structure.
- struct PartInfo : public Storage::Partition::Info, public Storage::Disk::DiskPart
In-memory partition information.
A disk Storage::Partition refers to this instance.
Public Functions
- inline virtual const Disk::DiskPart *diskpart() const override
Obtain additional disk information.
Accessed via
Partition::diskpart()method
- virtual size_t printTo(Print &p) const override
Print important fields only.
- class BasePartitionTable : public OwnedLinkedObjectListTemplate<PartInfo>
Common type for MBR/GPT partition table.
Subclassed by Storage::Disk::GPT::PartitionTable, Storage::Disk::MBR::PartitionTable
- class Scanner
Class to iterate through disk partition tables.
- class SectorBuffer : public std::unique_ptr<uint8_t[]>
Buffer for working with disk sectors.
- namespace EXFAT
- namespace FAT
- namespace GPT
- struct SmingTypeGuid : public Uuid
- class PartitionTable : public Storage::Disk::BasePartitionTable
Public Functions
- inline bool add(const String &name, SysType sysType, storage_size_t offset, storage_size_t size, const Uuid &uniqueGuid = {}, const Uuid &typeGuid = {}, Partition::Flags flags = 0)
Add a new standard GPT partition definition.
- Parameters:
name – Partition name
sysType – Intended content for this partition
offset – Start offset, or 0 to have position calculated
size – Size of partition (in bytes), or percentage (0-100) of total partitionable disk space
uniqueGuid – Unique partition identifier (optional: will be generated if not provided)
typeGuid – Partition type GUID (default is BASIC_DATA)
flags –
- Return values:
bool – true on success
- inline bool add(const String &name, Partition::FullType type, storage_size_t offset, storage_size_t size, const Uuid &uniqueGuid = {}, Partition::Flags flags = 0)
Add a new GPT partition for a regular Sming filing system.
Note
These partitions use a custom type GUID and won’t be recognised by external software.
- Parameters:
name – Partition name
type – Sming partition type/subtype
offset – Start offset, or 0 to have position calculated
size – Size of partition (in bytes), or percentage (0-100) of total partitionable disk space
uniqueGuid – Unique partition identifier (optional: will be generated if not provided)
flags –
- Return values:
bool – true on success
- namespace MBR
- class PartitionTable : public Storage::Disk::BasePartitionTable
Public Functions
- inline bool add(SysType sysType, SysIndicator sysIndicator, storage_size_t offset, storage_size_t size, Partition::Flags flags = 0)
Add a new MBR partition definition.
Note
MBR does not have partition name field; this will appear as ‘mbr1’, ‘mbr2’, etc.
- Parameters:
sysType – Intended content for this partition (or ‘unknown’)
SysIndicator – Appropriate system code SI_xxx
offset – Start offset, or 0 to have position calculated
size – Size of partition (in bytes), or percentage (0-100) of total partitionable disk space
flags –
- Return values:
bool – true on success
- namespace SD
- class Card : public Storage::Disk::BlockDevice
Public Functions
- bool begin(uint8_t chipSelect, uint32_t freq = 0)
Initialise the card.
- Parameters:
chipSelect –
freq – SPI frequency in Hz, use 0 for maximum supported frequency
- inline virtual uint32_t getId() const
Obtain device ID.
- Return values:
uint32_t – typically flash chip ID
- inline virtual Type getType() const
Obtain device type.
- inline virtual size_t getBlockSize() const override
Obtain smallest allocation unit for erase operations.
- struct CID
- struct CSD
Subclassed by Storage::SD::CSD1, Storage::SD::CSD2, Storage::SD::CSD3
References
Used by
Sming (main) Component
Installable File System Component
Disk Storage Library
Environment Variables
PARTITION_TABLE_OFFSETSoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Arch Driver
This is an internal Component to pull together common headers and code used at a low-level by more than once Arch.
This is to:
Ease maintenance
Avoid code duplication
Provide a consistent API for the framework to use
References
Used by
Esp32 Drivers Component
Esp8266 Drivers Component
Host Drivers Component
RP2040 Drivers Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
AXTLS 8266
SSL support using the AXTLS library
References
Used by
SSL: Secure Sockets Layer Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: axtls-8266
This is an ESP8266 port of axTLS library, currently based on axTLS 2.1.4 (SVN version 277).
This library supports TLS 1.2, and the following cipher suites:
Cipher suite name (RFC)
OpenSSL name
Key exchange
Encryption
Hash
TLS_RSA_WITH_AES_128_CBC_SHA
AES128-SHA
RSA
AES-128
SHA-1
TLS_RSA_WITH_AES_256_CBC_SHA
AES256-SHA
RSA
AES-256
SHA-1
TLS_RSA_WITH_AES_128_CBC_SHA256
AES128-SHA256
RSA
AES-128
SHA-256
TLS_RSA_WITH_AES_256_CBC_SHA256
AES256-SHA256
RSA
AES-256
SHA-256
Using the library
This is not a self-sufficient library. In addition to the standard C library functions, application has to provide the following functions:
ax_port_read ax_port_write ax_port_open ax_port_close ax_get_file phy_get_rand (provided by the IoT SDK) ets_printf (in ESP8266 ROM) ets_putc (in ESP8266 ROM)For use with LwIP raw TCP API, see compat/README.md
Building .. image:: https://travis-ci.org/igrr/axtls-8266.svg
- target:
- alt:
Build status
To build, add xtensa toolchain to your path, and run
make. The library will be built inbin/directory.Credits and license
axTLS is written and maintained by Cameron Rich.
Other people have contributed to this port; see git logs for a full list.
See LICENSE file for axTLS license.
Bear SSL
SSL support using Bear SSL for ESP8266.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: bearssl
Cryptographic Support
Introduction
Contains basic cryptographic support classes for Sming.
This provides a strongly typed and flexible C++ library for commonly used routines, such as MD5, SHA hashes and HMAC.
Architecture-specific ROM or SDK routines are used where appropriate to reduce code size and improve performance.
The intention is that this library will provide the optimal implementations for any given architecture but maintain a consistent interface and allow it to be easily extended.
Hashes
You must #include the appropriate header for the hash family, for example:
#include <Crypto/Sha1.h>All hash operations are then performed via the
Crypto::Sha1class.Here’s a basic example showing how to calculate a SHA1 hash on a C string:
#include <Crypto/Sha1.h> void sha1Test(const char* buffer) { // Returns a Crypto::Sha1::Hash object auto hash = Crypto::Sha1().calculate(buffer, strlen(buffer)); Serial.print("SHA1: "); Serial.println(Crypto::toString(hash)); }If your data has multiple chunks, use the longer form:
#include <Crypto/Sha2.h> void sha256Test(const String& s1, const String& s2) { Crypto::Sha256 ctx; ctx.update(s1); ctx.update(s2); Serial.print("SHA256: "); Serial.println(Crypto::toString(ctx.getHash())); } sha256Test(F("This is some text to be hashed"), F("Hello"));Some hashes have additional optional parameters, for example:
#include <Crypto/Blake2s.h> void blake2sTest(const String& key, const String& content) { Crypto::Blake2s256 ctx(key); ctx.update(content); Serial.print("BLAKE2S-256: "); Serial.println(Crypto::toString(ctx.getHash())); }HMAC
The HMAC algorithm is commonly used for verifying both the integrity and authenticity of a message. It can be used with any defined hash, commonly MD5 or SHA1.
For example, an MD5 HMAC (as used with CRAM-MD5) may be done like this:
#include <Crypto/Md5.h> void printHmacMd5(const String& key, const String& data) { auto hash = Crypto::HmacMd5(key).calculate(data); Serial.print("HMAC.MD5 = "); Serial.println(Crypto::toString(hash)); }‘C’ API
The library also defines a standard ‘C’ api so it can be used from within existing code, such as AXTLS 8266 and Bear SSL. These definitions may be found in
Crypto/HashApi.API Documentation
- namespace Crypto
Typedefs
- template<size_t hashsize>
using Blake2s = HashContext<Blake2sEngine<hashsize>>
- template<size_t hashsize>
using HmacBlake2s = HmacContext<Blake2s<hashsize>>
- using HmacBlake2s256 = HmacBlake2s<32>
- using HmacBlake2s128 = HmacBlake2s<16>
- template<size_t size_>
using ByteArray = std::array<uint8_t, size_>Class template for fixed byte array.
Note
Until C++17 (and GCC > 5.5) inheriting from std::array<> breaks aggregate initialization.
- using Md5 = HashContext<Md5Engine>
- using HmacMd5 = HmacContext<Md5>
- using Sha1 = HashContext<Sha1Engine>
- using HmacSha1 = HmacContext<Sha1>
- using Sha224 = HashContext<Sha224Engine>
- using Sha256 = HashContext<Sha256Engine>
- using Sha384 = HashContext<Sha384Engine>
- using Sha512 = HashContext<Sha512Engine>
- using HmacSha224 = HmacContext<Sha224>
- using HmacSha256 = HmacContext<Sha256>
- using HmacSha384 = HmacContext<Sha384>
- using HmacSha512 = HmacContext<Sha512>
Functions
- class Blob
Wraps a pointer to some data with size.
- template<class Engine_>
class HashContextClass template for a Hash implementation ‘Context’.
- Template Parameters:
Engine – The HashEngine implementation
Subclassed by OtaUpgrade::ChecksumVerifier
Update hash over a given block of data
- inline HashContext &update(const Blob &blob)
Data from Blob.
- inline HashContext &update(const FSTR::ObjectBase &obj)
Data from flash object.
- inline HashContext &update(const void *data, size_t size)
Pointer to data + size.
- Parameters:
data – Data block
size – Length of data in bytes
- template<size_t size_>
inline HashContext &update(const ByteArray<size_> &array)Data in ByteArray.
Public Functions
- template<typename ...EngineArgs>
inline HashContext &reset(EngineArgs&&... engineArgs)Reset the context for a new calculation.
- template<typename ...Ts>
inline Hash calculate(Ts&&... args)Calculate hash on some data.
- Parameters:
args – See update() methods
- Return values:
Hash –
- inline Hash getHash()
Finalise and return the final hash value.
- Return values:
Hash –
- inline State getState()
Get intermediate hash state.
Note
This method is only required for core hashes, used by Bear SSL
- Parameters:
state – OUT: current state
- Return values:
uint64_t – Number of bytes processed so far
- inline void setState(const State &state)
Restore intermediate hash state.
Parameter values obtained via previous getState() call
Note
This method is only required for core hashes, used by Bear SSL
- Parameters:
state –
count –
- struct State
- template<class HashContext>
class HmacContextHMAC class template.
Implements the HMAC algorithm using any defined hash context
Public Functions
- inline HmacContext &init(const Secret &key)
Initialise HMAC with key.
- Return values:
Reference – to enable method chaining
- template<typename ...Ts>
inline HmacContext &update(Ts&&... args)Update HMAC with some message content.
- Parameters:
args – See HashContext update() methods
- Return values:
Reference – to enable method chaining
- template<typename ...Ts>
inline Hash calculate(Ts&&... args)Calculate hash for some data.
Use like this:
auto hash = Crypto::HmacMd5(mySecret).calculate(myData);
- Parameters:
args – See HashContext update() methods
- Return values:
Hash –
References
Used by
AXTLS 8266 Component
SSL: Secure Sockets Layer Component
libsodium Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Esptool
This Component provides Espressif’s tool for reading and writing firmware and other data to hardware.
Options
- SPI_SPEED
[read-only] Set by Hardware configuration.
Clock speed for flash memory (20, 26, 40 or 80). Default is 40.
- SPI_MODE
[read-only] Set by Hardware configuration.
Flash memory operating mode (quot, dio, dout, qio). Default is qio.
- SPI_SIZE
[read-only] Set by Hardware configuration.
Size of flash memory chip (256KB, 512KB, 1MB, 2MB, 4MB). Default is 512K bytes.
The default hardware profile
standardsets this to 1MB. You can setHWCONFIG=standard-4mto increase it or create a custom Hardware configuration for your project.
- ESPTOOL
Full path of esptool.py
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
Submodule: esptool
esptool.py
A Python-based, open-source, platform-independent utility to communicate with the ROM bootloader in Espressif chips.
Documentation
Visit the documentation or run
esptool.py -h.Contribute
If you’re interested in contributing to esptool.py, please check the contributions guide.
About
esptool.py was initially created by Fredrik Ahlberg (@themadinventor), and later maintained by Angus Gratton (@projectgus). It is now supported by Espressif Systems. It has also received improvements from many members of the community.
License
This document and the attached source code are released as Free Software under GNU General Public License Version 2 or later. See the accompanying LICENSE file for a copy.
HTTP Parser
http-parser is **not** actively maintained. New projects and projects looking to migrate should consider llhttp.
This is a parser for HTTP messages written in C. It parses both requests and responses. The parser is designed to be used in performance HTTP applications. It does not make any syscalls nor allocations, it does not buffer data, it can be interrupted at anytime. Depending on your architecture, it only requires about 40 bytes of data per message stream (in a web server that is per connection).
Features:
No dependencies
Handles persistent streams (keep-alive).
Decodes chunked encoding.
Upgrade support
Defends against buffer overflow attacks.
The parser extracts the following information from HTTP messages:
Header fields and values
Content-Length
Request method
Response status code
Transfer-Encoding
HTTP version
Request URL
Message body
Usage
One
http_parserobject is used per TCP connection. Initialize the struct usinghttp_parser_init()and set the callbacks. That might look something like this for a request parser:http_parser_settings settings; settings.on_url = my_url_callback; settings.on_header_field = my_header_field_callback; /* ... */ http_parser *parser = malloc(sizeof(http_parser)); http_parser_init(parser, HTTP_REQUEST); parser->data = my_socket;When data is received on the socket execute the parser and check for errors.
size_t len = 80*1024, nparsed; char buf[len]; ssize_t recved; recved = recv(fd, buf, len, 0); if (recved < 0) { /* Handle error. */ } /* Start up / continue the parser. * Note we pass recved==0 to signal that EOF has been received. */ nparsed = http_parser_execute(parser, &settings, buf, recved); if (parser->upgrade) { /* handle new protocol */ } else if (nparsed != recved) { /* Handle error. Usually just close the connection. */ }
http_parserneeds to know where the end of the stream is. For example, sometimes servers send responses without Content-Length and expect the client to consume input (for the body) until EOF. To tellhttp_parserabout EOF, give0as the fourth parameter tohttp_parser_execute(). Callbacks and errors can still be encountered during an EOF, so one must still be prepared to receive them.Scalar valued message information such as
status_code,method, and the HTTP version are stored in the parser structure. This data is only temporally stored inhttp_parserand gets reset on each new message. If this information is needed later, copy it out of the structure during theheaders_completecallback.The parser decodes the transfer-encoding for both requests and responses transparently. That is, a chunked encoding is decoded before being sent to the on_body callback.
The Special Problem of Upgrade
http_parsersupports upgrading the connection to a different protocol. An increasingly common example of this is the WebSocket protocol which sends a request likeGET /demo HTTP/1.1 Upgrade: WebSocket Connection: Upgrade Host: example.com Origin: http://example.com WebSocket-Protocol: samplefollowed by non-HTTP data.
(See RFC6455 for more information the WebSocket protocol.)
To support this, the parser will treat this as a normal HTTP message without a body, issuing both on_headers_complete and on_message_complete callbacks. However http_parser_execute() will stop parsing at the end of the headers and return.
The user is expected to check if
parser->upgradehas been set to 1 afterhttp_parser_execute()returns. Non-HTTP data begins at the buffer supplied offset by the return value ofhttp_parser_execute().Callbacks
During the
http_parser_execute()call, the callbacks set inhttp_parser_settingswill be executed. The parser maintains state and never looks behind, so buffering the data is not necessary. If you need to save certain data for later usage, you can do that from the callbacks.There are two types of callbacks:
- notification
typedef int (*http_cb) (http_parser*);Callbacks: on_message_begin, on_headers_complete, on_message_complete.
- data
typedef int (*http_data_cb) (http_parser*, const char *at, size_t length);Callbacks: (requests only) on_url,
(common) on_header_field, on_header_value, on_body;Callbacks must return 0 on success. Returning a non-zero value indicates error to the parser, making it exit immediately.
For cases where it is necessary to pass local information to/from a callback, the
http_parserobject’sdatafield can be used. An example of such a case is when using threads to handle a socket connection, parse a request, and then give a response over that socket. By instantiation of a thread-local struct containing relevant data (e.g. accepted socket, allocated memory for callbacks to write into, etc), a parser’s callbacks are able to communicate data between the scope of the thread and the scope of the callback in a threadsafe manner. This allowshttp_parserto be used in multi-threaded contexts.Example:
typedef struct { socket_t sock; void* buffer; int buf_len; } custom_data_t; int my_url_callback(http_parser* parser, const char *at, size_t length) { /* access to thread local custom_data_t struct. Use this access save parsed data for later use into thread local buffer, or communicate over socket */ parser->data; ... return 0; } ... void http_parser_thread(socket_t sock) { int nparsed = 0; /* allocate memory for user data */ custom_data_t *my_data = malloc(sizeof(custom_data_t)); /* some information for use by callbacks. * achieves thread -> callback information flow */ my_data->sock = sock; /* instantiate a thread-local parser */ http_parser *parser = malloc(sizeof(http_parser)); http_parser_init(parser, HTTP_REQUEST); /* initialise parser */ /* this custom data reference is accessible through the reference to the parser supplied to callback functions */ parser->data = my_data; http_parser_settings settings; /* set up callbacks */ settings.on_url = my_url_callback; /* execute parser */ nparsed = http_parser_execute(parser, &settings, buf, recved); ... /* parsed information copied from callback. can now perform action on data copied into thread-local memory from callbacks. achieves callback -> thread information flow */ my_data->buffer; ... }In case you parse HTTP message in chunks (i.e.
read()request line from socket, parse, read half headers, parse, etc) your data callbacks may be called more than once.http_parserguarantees that data pointer is only valid for the lifetime of callback. You can alsoread()into a heap allocated buffer to avoid copying memory around if this fits your application.Reading headers may be a tricky task if you read/parse headers partially. Basically, you need to remember whether last header callback was field or value and apply the following logic:
(on_header_field and on_header_value shortened to on_h_*) ------------------------ ------------ -------------------------------------------- | State (prev. callback) | Callback | Description/action | ------------------------ ------------ -------------------------------------------- | nothing (first call) | on_h_field | Allocate new buffer and copy callback data | | | | into it | ------------------------ ------------ -------------------------------------------- | value | on_h_field | New header started. | | | | Copy current name,value buffers to headers | | | | list and allocate new buffer for new name | ------------------------ ------------ -------------------------------------------- | field | on_h_field | Previous name continues. Reallocate name | | | | buffer and append callback data to it | ------------------------ ------------ -------------------------------------------- | field | on_h_value | Value for current header started. Allocate | | | | new buffer and copy callback data to it | ------------------------ ------------ -------------------------------------------- | value | on_h_value | Value continues. Reallocate value buffer | | | | and append callback data to it | ------------------------ ------------ --------------------------------------------Parsing URLs
A simplistic zero-copy URL parser is provided as
http_parser_parse_url(). Users of this library may wish to use it to parse URLs constructed from consecutiveon_urlcallbacks.See examples of reading in headers:
partial example in C
from Node library in Javascript
References
Used by
Networking Support Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
b64: Base64 Encoding/Decoding Routines
Overview:
libb64 is a library of ANSI C routines for fast encoding/decoding data into and from a base64-encoded format. C++ wrappers are included, as well as the source code for standalone encoding and decoding executables.
base64 consists of ASCII text, and is therefore a useful encoding for storing binary data in a text file, such as xml, or sending binary data over text-only email.
References:
- Wikipedia article:
- base64, another implementation of a commandline en/decoder:
Why?
I did this because I need an implementation of base64 encoding and decoding, without any licensing problems. Most OS implementations are released under either the GNU/GPL, or a BSD-variant, which is not what I require.
Also, the chance to actually use the co-routine implementation in code is rare, and its use here is fitting. I couldn’t pass up the chance. For more information on this technique, see “Coroutines in C”, by Simon Tatham, which can be found online here: http://www.chiark.greenend.org.uk/~sgtatham/coroutines.html
So then, under which license do I release this code? On to the next section…
License:
This work is released under into the Public Domain. It basically boils down to this: I put this work in the public domain, and you can take it and do whatever you want with it.
An example of this “license” is the Creative Commons Public Domain License, a copy of which can be found in the LICENSE file, and also online at http://creativecommons.org/licenses/publicdomain/
Commandline Use:
There is a new executable available, it is simply called base64. It can encode and decode files, as instructed by the user.
To encode a file: $ ./base64 -e filea fileb fileb will now be the base64-encoded version of filea.
To decode a file: $ ./base64 -d fileb filec filec will now be identical to filea.
Programming:
Some C++ wrappers are provided as well, so you don’t have to get your hands dirty. Encoding from standard input to standard output is as simple as
#include <b64/encode.h> #include <iostream> int main() { base64::encoder E; E.encode(std::cin, std::cout); return 0; }Both standalone executables and a static library is provided in the package,
Example code:
The ‘examples’ directory contains some simple example C code, that demonstrates how to use the C interface of the library.
Implementation:
It is DAMN fast, if I may say so myself. The C code uses a little trick which has been used to implement coroutines, of which one can say that this implementation is an example.
(To see how the libb64 codebase compares with some other BASE64 implementations available, see the BENCHMARKS file)
The trick involves the fact that a switch-statement may legally cross into sub-blocks. A very thorough and enlightening essay on co-routines in C, using this method, can be found in the above mentioned “Coroutines in C”, by Simon Tatham: http://www.chiark.greenend.org.uk/~sgtatham/coroutines.html
For example, an RLE decompressing routine, adapted from the article: 1 static int STATE = 0; 2 static int len, c; 3 switch (STATE) 4 { 5 while (1) 6 { 7 c = getchar(); 8 if (c == EOF) return EOF; 9 if (c == 0xFF) { 10 len = getchar(); 11 c = getchar(); 12 while (len–) 13 { 14 STATE = 0; 15 return c; 16 case 0: 17 } 18 } else 19 STATE = 1; 20 return c; 21 case 1: 22 } 23 } 24 }
As can be seen from this example, a coroutine depends on a state variable, which it sets directly before exiting (lines 14 and 119). The next time the routine is entered, the switch moves control to the specific point directly after the previous exit (lines 16 and 21).hands
(As an aside, in the mentioned article the combination of the top-level switch, the various setting of the state, the return of a value, and the labelling of the exit point is wrapped in #define macros, making the structure of the routine even clearer.)
The obvious problem with any such routine is the static keyword. Any static variables in a function spell doom for multithreaded applications. Also, in situations where this coroutine is used by more than one other coroutines, the consistency is disturbed.
What is needed is a structure for storing these variabled, which is passed to the routine separately. This obviously breaks the modularity of the function, since now the caller has to worry about and care for the internal state of the routine (the callee). This allows for a fast, multithreading-enabled implementation, which may (obviously) be wrapped in a C++ object for ease of use.
The base64 encoding and decoding functionality in this package is implemented in exactly this way, providing both a high-speed high-maintanence C interface, and a wrapped C++ which is low-maintanence and only slightly less performant.
References
Used by
Networking Support Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
libyuarel
Very simple and well tested C library for parsing URLs with zero-copy and no mallocs. The library does not do any validation on the URL, neither before nor after it is parsed. The different parts are parsed by searching for special characters like
:and/. For a URL should be able to be parsed by yuarel, it has to be constructed in one of the following formats:Absolute URL: scheme “:” [ “//” ] [ username “:” password “@” ] host [ “:” port ] [ “/” ] [ path ] [ “?” query ] [ “#” fragment ]
Relative URL: path [ “?” query ] [ “#” fragment ]
Parts within
[and]are optional. A minimal URL could look like this:
a:bor/Due to the fact that the library isn’t copying any strings and instead points to the parts in the URL string, the first
/in the path will be replaced with a null terminator. Therefore, the first slash will be missing in the path.To build
$ make && make check && sudo make installTry it
Compile the example in
examples/:$ make examplesRun the example program:
$ ./simpleThe structs
The struct that holds the parsed URL looks like this:
struct yuarel { char *scheme; /* scheme, without ":" and "//" */ char *username; /* username, default: NULL */ char *password; /* password, default: NULL */ char *host; /* hostname or IP address */ int port; /* port, default: 0 */ char *path; /* path, without leading "/", default: NULL */ char *query; /* query, default: NULL */ char *fragment; /* fragment, default: NULL */ };The struct that holds a parsed query string parameter looks like this:
struct yuarel_param { char *key; char *val; };Library functions
Parse a URL to a struct
int yuarel_parse(struct yuarel *url, char *url_str)
struct yuarel *url: a pointer to the struct where to store the parsed values.char *url_str: a pointer to the url to be parsed (null terminated).Note that the url string will be modified by the function.
Returns 0 on success, otherwise -1.
Split a path into several strings
int yuarel_split_path(char *path, char **parts, int max_parts)No data is copied, the slashed are used as null terminators and then pointers to each path part will be stored in
parts.
char *path: the path to split. The string will be modified.char **parts: a pointer to an array of(char *)where to store the result.int max_parts: max number of parts to parse.Note that the path string will be modified by the function.
Returns the number of parsed items. -1 on error.
Parse a query string
int yuarel_parse_query(char *query, char delimiter, struct yuarel_param *params, int max_params)
char *query: the query string to parse. The string will be modified.char delimiter: the character that separates the key/value pairs from eachother.struct yuarel_param *params: an array of(struct yuarel_param)where to store the result.int max_values: max number of parameters to parse.The query string should be a null terminated string of parameters separated by a delimiter. Each parameter are checked for the equal sign character. If it appears in the parameter, it will be used as a null terminator and the part that comes after it will be the value of the parameter.
No data are copied, the equal sign and delimiters are used as null terminators and then pointers to each parameter key and value will be stored in the yuarel_param struct.
Note that the query string will be modified by the function.
Returns the number of parsed items. -1 on error.
How to use it:
Compile with
-lyuarel.#include <stdlib.h> #include <stdio.h> #include <yuarel.h> int main(void) { int p; struct yuarel url; char *parts[3]; char url_string[] = "http://localhost:8989/path/to/test?query=yes#frag=1"; if (-1 == yuarel_parse(&url, url_string)) { fprintf(stderr, "Could not parse url!\n"); return 1; } printf("scheme:\t%s\n", url.scheme); printf("host:\t%s\n", url.host); printf("port:\t%d\n", url.port); printf("path:\t%s\n", url.path); printf("query:\t%s\n", url.query); printf("fragment:\t%s\n", url.fragment); if (3 != yuarel_split_path(url.path, parts, 3)) { fprintf(stderr, "Could not split path!\n"); return 1; } printf("path parts: %s, %s, %s\n", parts[0], parts[1], parts[2]); printf("Query string parameters:\n"); p = yuarel_parse_query(url.query, '&', params, 3); while (p-- > 0) { printf("\t%s: %s\n", params[p].key, params[p].val); } }References
Used by
Networking Support Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
LWIP
Introduction
Uses LWIP version 2 to enable networking for Sming running on a Windows or Linux Host system.
Note
Network support is enabled by default. If you don’t need it, use the
--nonetcommand-line option.Build Variables
- ENABLE_LWIPDEBUG
- 0 (default)
Standard build
- 1
Enable debugging output
- ENABLE_CUSTOM_LWIP
2 (default)
Setting this to any other value will cause a build error.
Linux
Support is provided via TAP network interface (a virtual network layer operating at the ethernet frame level). A TAP interface must be created first, and requires root privilege:
sudo ip tuntap add dev tap0 mode tap user `whoami` sudo ip a a dev tap0 192.168.13.1/24 sudo ifconfig tap0 upThis creates the
tap0interface. The emulator will automatically select the firsttapinterface found. To override this, use the--ifnameoption. An IP address will be assigned, but can be changed using the--ipaddroption.If your application needs to access the internet, additional setup is required:
sudo sysctl net.ipv4.ip_forward=1 sudo sysctl net.ipv6.conf.default.forwarding=1 sudo sysctl net.ipv6.conf.all.forwarding=1 export INTERNET_IF=wlan0 # <!--- Make sure to replace wlan0 with the network interface connected to Internet sudo iptables -t nat -A POSTROUTING -o $INTERNET_IF -j MASQUERADE sudo iptables -A FORWARD -m conntrack --ctstate RELATED,ESTABLISHED -j ACCEPT sudo iptables -A FORWARD -i tap0 -o $INTERNET_IF -j ACCEPTWindows
Requires NPCAP library to be installed. Provided with current (3.0.2) version of Wireshark.
By default, the first valid network adapter will be used, with address assigned via DHCP.
If the adapter is wrong, get a list thus:
out\Host\Windows\debug\firmware\app --ifname=?or
make run HOST_NETWORK_OPTIONS=--ifname=?produces a listing:
Available adapters: - 0: {ACC6BFB2-A15B-4CF8-B93A-8D97644D0AAC} - Oracle 192.168.56.1 / 255.255.255.0 - 1: {A12D4DD0-0EA8-435D-985E-A1F96F781EF0} - NdisWan Adapter - 2: {3D66A354-39DD-4C6A-B9C4-14EE223FC3D1} - MS NDIS 6.0 LoopBack Driver 0.0.0.0 / 255.0.0.0 - 3: {BC53D919-339E-4D70-8573-9D7A8AE303C7} - NdisWan Adapter - 4: {3CFD43EA-9CC7-44A7-83D4-EB04DD029FE7} - NdisWan Adapter - 5: {530640FF-A9C3-436B-9EA2-65102C788119} - Realtek PCIe GBE Family Controller 192.168.1.70 / 255.255.255.0 - 6: {0F649280-BAC2-4515-9CE3-F7DFBB6A1BF8} - Kaspersky Security Data Escort Adapter 10.102.37.150 / 255.255.255.252Then use the appropriate number (or GUID), with the gateway IP address - an address will be assigned via DHCP:
make run HOST_NETWORK_OPTIONS="--ifname=5 --gateway=192.168.1.254"You can find gateway addresses using the
ipconfigcommand.If you want to use a specific IP address, the appropriate adapter will be selected but you still need to specify the gateway address:
make run HOST_NETWORK_OPTIONS="--ipaddr=192.168.1.10 --gateway=192.168.1.254"References
Used by
Host WiFi Component
Host Library Component
Networking Support Component
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: lwip
INTRODUCTION
lwIP is a small independent implementation of the TCP/IP protocol suite.
The focus of the lwIP TCP/IP implementation is to reduce the RAM usage while still having a full scale TCP. This making lwIP suitable for use in embedded systems with tens of kilobytes of free RAM and room for around 40 kilobytes of code ROM.
lwIP was originally developed by Adam Dunkels at the Computer and Networks Architectures (CNA) lab at the Swedish Institute of Computer Science (SICS) and is now developed and maintained by a worldwide network of developers.
FEATURES
IP (Internet Protocol, IPv4 and IPv6) including packet forwarding over multiple network interfaces
ICMP (Internet Control Message Protocol) for network maintenance and debugging
IGMP (Internet Group Management Protocol) for multicast traffic management
MLD (Multicast listener discovery for IPv6). Aims to be compliant with RFC 2710. No support for MLDv2
ND (Neighbor discovery and stateless address autoconfiguration for IPv6). Aims to be compliant with RFC 4861 (Neighbor discovery) and RFC 4862 (Address autoconfiguration)
DHCP, AutoIP/APIPA (Zeroconf), ACD (Address Conflict Detection) and (stateless) DHCPv6
UDP (User Datagram Protocol) including experimental UDP-lite extensions
TCP (Transmission Control Protocol) with congestion control, RTT estimation fast recovery/fast retransmit and sending SACKs
raw/native API for enhanced performance
Optional Berkeley-like socket API
TLS: optional layered TCP (“altcp”) for nearly transparent TLS for any TCP-based protocol (ported to mbedTLS) (see changelog for more info)
PPPoS and PPPoE (Point-to-point protocol over Serial/Ethernet)
DNS (Domain name resolver incl. mDNS)
6LoWPAN (via IEEE 802.15.4, BLE or ZEP)
APPLICATIONS
HTTP server with SSI and CGI (HTTPS via altcp)
SNMPv2c agent with MIB compiler (Simple Network Management Protocol), v3 via altcp
SNTP (Simple network time protocol)
NetBIOS name service responder
MDNS (Multicast DNS) responder
iPerf server implementation
MQTT client (TLS support via altcp)
LICENSE
lwIP is freely available under a BSD license.
DEVELOPMENT
lwIP has grown into an excellent TCP/IP stack for embedded devices, and developers using the stack often submit bug fixes, improvements, and additions to the stack to further increase its usefulness.
Development of lwIP is hosted on Savannah, a central point for software development, maintenance and distribution. Everyone can help improve lwIP by use of Savannah’s interface, Git and the mailing list. A core team of developers will commit changes to the Git source tree.
The lwIP TCP/IP stack is maintained in the ‘lwip’ Git module and contributions (such as platform ports) are in the ‘contrib’ Git module.
See doc/savannah.txt for details on Git server access for users and developers.
The current Git trees are web-browsable: http://git.savannah.gnu.org/cgit/lwip.git http://git.savannah.gnu.org/cgit/lwip/lwip-contrib.git
Submit patches and bugs via the lwIP project page: http://savannah.nongnu.org/projects/lwip/
Continuous integration builds (GCC, clang): https://travis-ci.org/lwip-tcpip/lwip
DOCUMENTATION
Self documentation of the source code is regularly extracted from the current Git sources and is available from this web page: http://www.nongnu.org/lwip/
Also, there are mailing lists you can subscribe at http://savannah.nongnu.org/mail/?group=lwip plus searchable archives: http://lists.nongnu.org/archive/html/lwip-users/ http://lists.nongnu.org/archive/html/lwip-devel/
There is a wiki about lwIP at http://lwip.wikia.com/wiki/LwIP_Wiki You might get questions answered there, but unfortunately, it is not as well maintained as it should be.
lwIP was originally written by Adam Dunkels: http://dunkels.com/adam/
Reading Adam’s papers, the files in docs/, browsing the source code documentation and browsing the mailing list archives is a good way to become familiar with the design of lwIP.
Adam Dunkels adam@sics.se Leon Woestenberg leon.woestenberg@gmx.net
malloc_count
This Component is a modified version of the original code, intended to provide basic heap monitoring for the Sming Host Emulator.
The following is the original README.
Introduction
malloc_countprovides a set of source code tools to measure the amount of allocated memory of a program at run-time. The code library provides facilities to
measure the current and peak heap memory allocation, and
write a memory profile for plotting, see the figure on the right.
Furthermore, separate
stack_countfunction can measure stack usage.The code tool works by intercepting the standard
malloc(),free(), etc functions. Thus no changes are necessary to the inspected source code.See http://panthema.net/2013/malloc_count for the current version.
Intercepting Heap Allocation Functions
The source code of
malloc_count.[ch]intercepts the standard heap allocation functionsmalloc(),free(),realloc()andcalloc()and adds simple counting statistics to each call. Thus the program must be relinked formalloc_countto work. Each call tomalloc()and others is passed on to lower levels, and the regularmalloc()is used for heap allocation.Of course,
malloc_countcan also be used with C++ programs and maybe even script languages like Python and Perl, because thenewoperator and most script interpreters allocations all are based onmalloc.The tools are usable under Linux and probably also with Cygwin and MinGW, as they too support the standard Linux dynamic link loading mechanisms.
Memory Profile and
stack_countThe
malloc_countsource is accompanied by two further memory analysis tools:stack_countand a C++ header calledmemprofile.h.In
stack_count.[ch]two simple functions are provided that can measure the maximum stack usage between two points in a program.Maybe the most useful application of
malloc_countis to create a memory/heap profile of a program (while it is running). This profile can also be created using the well-known valgrind tool “massif”, however, massif is really slow. The overhead ofmalloc_countis much smaller, and usingmemprofile.ha statistic file can be produced, which is directly usable with Gnuplot.The source code archive contains two example applications: one which queries
malloc_countfor current heap usage, and a second which creates the memory profile in the figure on the right. See the STX B+ Tree library for another, more complex example of a memory profile.Downloads
See http://panthema.net/2013/malloc_count for the current version.
The source code is published under the MIT License (MIT), which is also found in the header of all source files.
Short Usage Guide
Compile
malloc_count.cand link it with your program. The source filemalloc_count.oshould be located towards the end of the.ofile sequence. You must also add “-ldl“ to the list of libraries.Run your program and observe that when terminating, it outputs a line like
malloc_count ### exiting, total: 12,582,912, peak: 4,194,304, current: 0If desired, increase verbosity
by setting
log_operations = 1at the top ofmalloc_count.cand adaptinglog_operations_thresholdto output only large allocations, orby including
malloc_count.hin your program and using the user-functions define therein to output memory usage at specific checkpoints. See the directorytest-malloc_count/in the source code for an example.Tip: Set the locale environment variable
LC_NUMERIC=en_GBor similar to get comma-separation of thousands in the printed numbers.The directory
test-memprofile/contains a simple program, which fills astd::vectorandstd::setwith integers. The memory usage of these containers is profiled using the facilities ofmemprofile.h, which are described verbosely in the source.Thread Safety
The current statistic methods in
malloc_count.care not thread-safe. However, the general mechanism (as described below) is per-se thread-safe. The only non-safe parts are adding and subtracting from the counters ininc_count()anddec_count().The
malloc_count.ccode contains a#define THREAD_SAFE_GCC_INTRINSICS, which enables use of gcc’s intrinsics for atomic counting operations. If you use gcc, enable this option to make themalloc_counttool thread-safe.The functions in
memprofile.hare not thread-safe.stack_countcan also be used on local thread stacks.Technicalities of Intercepting
libcFunction CallsThe method used in
malloc_countto hook the standard heap allocation calls is to provide a source file exporting the symbols “malloc“, “free“, etc. These override the libc symbols and thus the functions inmalloc_countare used instead.However,
malloc_countdoes not implement a heap allocator. It loads the symbols “malloc“, “free“, etc. directly using the dynamic link loader “dl“ from the chain of shared libraries. Calls to the overriding “malloc“ functions are forwarded to the usual libc allocator.To keep track of the size of each allocated memory area,
malloc_countuses a trick: it prepends each allocation pointer with two additional bookkeeping variables: the allocation size and a sentinel value. Thus when allocating n bytes, in truth n + c bytes are requested from the libcmalloc()to save the size (c is by default 16, but can be adapted to fix alignment problems). The sentinel only serves as a check that your program has not overwritten the size information.Closing Credits
The idea for this augmenting interception method is not my own, it was borrowed from Jeremy Kerr http://ozlabs.org/~jk/code/.
Written 2013-01-21, 2013-03-16, and 2014-09-10 by Timo Bingmann tb@panthema.net
API Documentation
- namespace MallocCount
Typedefs
- using Callback = std::function<void(size_t current)>
Callback function type.
- Param current:
Current allocated bytes
Functions
- size_t getCurrent(void)
Get the currently allocated amount of memory.
- size_t getPeak(void)
Get the peak memory allocation.
- void resetPeak(void)
Reset the peak memory allocation to current.
- size_t getTotal(void)
Get the total cumulative memory allocation.
- void resetTotal(void)
Reset the total cumulative memory allocation to zero.
- size_t getAllocCount(void)
Get the total number of allocations.
- void setAllocLimit(size_t maxBytes)
Set an allocation limit.
- Parameters:
maxBytes – Specify 0 for no limit
- void setCallback(Callback callback)
Set a callback function that is invoked on each change of the current allocation.
Note
The callback function must not use malloc()/realloc()/free() or it will go into an endless recursive loop!
- void enableLogging(bool enable)
Enable/disable logging.
- void setLogThreshold(size_t threshold)
Set minimum allocation size for log output (when enabled)
References
Used by
Heap Component
Basic Graphics Sample
Basic Alexa Sample
RingTone Player Sample
SmingTest Library
Basic SSL Sample
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
mqtt-protocol-c
Zero-copy, interruptible MQTT protocol
parserandserialiserwritten in C. Based initially on Deoxxa’s code and extended to support full client and server parsing and serialization.Overview
mqtt-protocol-c is designed for use in resource-constrained environments. To that end, it avoids making any memory allocations or assumptions about the underlying system. It handles only the binary parsing/serialising part of the MQTT protocol, leaving all the logic and networking up to the user.
Examples
Take a look at
test_serialiserandtest_parserin the bin/test.c file.License
BSD-3 Clause license. You can read the full license from here.
Copyright
2018 - present Slavey Karadzhov slav@attachix.com
2014 - Deoxxa Development
References
Used by
Networking Support Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
rBoot
Introduction
rBoot is a second-stage bootloader that allows booting application images from several pre-configured flash memory addresses, called “slots”. Sming supports up to three slots.
Note
With Sming 4.3 partitions are used to manage flash storage. A “slot” refers to a specific application partition, typically
rom0,rom1orrom2.The location or size of these partitions is determined by the Hardware configuration.
The bootloader has been modified to use the partition table as reference, identifying slots by the partition sub-type.
Where systems are to be updated Over the Air (OTA) at least two application partitions are required. The bootloader identifies these by their partition subtype:
slot #0->App/Ota_0,slot #1->App/Ota_1, etc.Fixed applications without OTA capability use a single application image. This must be the
App/Factorypartition type, and corresponds toslot #0.At startup, the bootloader will use the partition table to locate the application image. It will also ensure that the ROM slot information in the boot configuration is consistent, and update it if necessary.
Attention
Make sure that your slots do not extend beyond a 1MB boundary and do not overlap with each other, the file system (if enabled) or the RFcal/system parameters area! Sming currently has no means of detecting a misconfigured memory layout.
Slot 0
This is the default slot (
rom0, the primary application partition) which is always used.Slot 1
- RBOOT_ROM1_ADDR
[read-only] default: disabled
The start address of slot 1.
If your application includes any kind of Over-the-Air (OTA) firmware update functionality, you will need a second memory slot to store the received update image while the update routines execute from the first slot.
Note
The
spiffs-two-romsconfiguration can be used for this purpose.Upon successful completion of the update, the second slot is activated, such that on next reset rBoot boots into the uploaded application. While now running from slot 1, the next update will be stored to slot 0 again, i.e. the roles of slot 0 and slot 1 are flipped with every update.
For devices with more than 1MB of flash memory, it is advisable to choose an address for
rom1with the same offset within its 1MB block asrom0.Slot 2 (GPIO slot)
rBoot supports booting into a third slot upon explicit user request, e.g. by pressing a button during reset/power up. This is especially useful for implementing some sort of recovery mechanism.
To enable slot 2, set these values:
- RBOOT_GPIO_ENABLED
Disabled by default. Set to 1 to enable slot 2.
- RBOOT_ROM2_ADDR
[read-only]
Address for slot 2. You must create a custom Hardware configuration for your project with a definition for
rom2.{ ... "partitions": { "rom2": { "address": "0x100000", "size": "512K", "type": "app", "subtype": "ota_2" } } }Note
At present, this will only configure rBoot. Sming will not create an application image for slot 2.
You can, however, use a second Sming project to build a recovery application image as follows:
Create a new Sming project for your recovery application. This will be a simple single-slot project. Create a new Hardware configuration and configure the
rom0start address and size to the same as therom2partition of the main project.option (a)
Build and flash the recovery project as usual by typing
make flash. This will install the recovery ROM (into slot 2 of the main project) and a temporary bootloader, which will be overwritten in the next step.Go back to your main project. Build and flash it with
make flash. This will install the main application (into slot 0) and the final bootloader. You are now all set for booting into the recovery image if the need arises.option (b)
Run a normal
makefor your recovery projectLocate the firmware image file, typically
out/Esp8266/release/firmware/rom0.bin(adjust accordingly if using a debug build). Copy this image file asrom2.bininto your main project directory.Add an additional property to the
rom2partition entry in your main project:"filename": "rom2.bin"When you run
make flashin this will get written along with the other partitions.Automatically derived settings
The
RBOOT_BIG_FLASHandRBOOT_TWO_ROMSsettings are now read-only as their values are derived automatically.In earlier versions of Sming these had to be set manually.
Big Flash Mode
The ESP8266 can map only 1MB of flash memory into its internal address space at a time. As you might expect, the first megabyte is mapped by default. This is fine if your image(s) reside(s) in this range. Otherwise, rBoot has to inject a piece of code into the application startup routine to set up the flash controller with the correct mapping.
- RBOOT_BIG_FLASH
READONLY Set when
RBOOT_ROM0_ADDRorRBOOT_ROM1_ADDR>= 0x100000.See also Big flash support.
Single vs. Dual ROMs
Since every 1MB range of flash memory is mapped to an identical internal address range, the same ROM image can be used for slots 0 and 1 if (and only if!) both slots have the same address offsets within their 1MB blocks, i.e.
(RBOOT_ROM0_ADDR & 0xFFFFF) == (RBOOT_ROM1_ADDR & 0xFFFFF).Consequently, for such configurations, the Sming build system generates only one ROM image.
In all other cases, two distinct application images must be linked with different addresses for the ‘irom0_0_seg’ memory region. You should use the
two-rom-modeHardware configuration for this. The Sming build system will handle everything else, including linker script generation.
- RBOOT_TWO_ROMS
READONLY Determines if rBoot needs to generate distinct firmware images.
Further Configuration Settings
- RBOOT_SILENT
Default: 0 (verbose)
At system restart rBoot outputs debug information to the serial port. Set to 1 to disable.
- RBOOT_LD_TEMPLATE
Path to the linker script template. The actual script is output to the application build directory (e.g.
rom0.ld), replacing the irom0_0_seg entry according to the configured build settings.
- RBOOT_ROM_0
Base name for firmware image #0. Default is
rom0.
- RBOOT_ROM_1
Base name for firmware image #1. Default is
rom1.
- ESPTOOL2
READONLY Defines the path to the esptool2 tool which rBoot uses to manipulate ROM images. Use
$(ESPTOOL2)if you need it within your own projects.API Documentation
API Documentation
rboot-api.h
Functions
- rboot_config rboot_get_config(void)
Read rBoot configuration from flash.
Note
Returns rboot_config (defined in rboot.h) allowing you to modify any values in it, including the ROM layout.
- Return values:
rboot_config – Copy of the rBoot configuration
- bool rboot_set_config(rboot_config *conf)
Write rBoot configuration to flash memory.
Note
Saves the rboot_config structure back to configuration sector (BOOT_CONFIG_SECTOR) of the flash, while maintaining the contents of the rest of the sector. You can use the rest of this sector for your app settings, as long as you protect this structure when you do so.
- Parameters:
conf – pointer to a rboot_config structure containing configuration to save
- Return values:
bool – True on success
- uint8_t rboot_get_current_rom(void)
Get index of current ROM.
Note
Get the currently selected boot ROM (this will be the currently running ROM, as long as you haven’t changed it since boot or rBoot booted the rom in temporary boot mode, see rboot_get_last_boot_rom).
- Return values:
uint8_t – Index of the current ROM
- bool rboot_set_current_rom(uint8_t rom)
Set the index of current ROM.
Note
Set the current boot ROM, which will be used when next restarted.
Note
This function re-writes the whole configuration to flash memory (not just the current ROM index)
- Parameters:
rom – The index of the ROM to use on next boot
- Return values:
bool – True on success
- rboot_write_status rboot_write_init(uint32_t start_addr)
Initialise flash write process.
Note
Call once before starting to pass data to write to flash memory with rboot_write_flash function. start_addr is the address on the SPI flash to write from. Returns a status structure which must be passed back on each write. The contents of the structure should not be modified by the calling code.
- Parameters:
start_addr – Address on the SPI flash to begin write to
- bool rboot_write_end(rboot_write_status *status)
Complete flash write process.
Note
Call at the completion of flash writing. This ensures any outstanding bytes are written (if data so far hasn’t been a multiple of 4 bytes there will be a few bytes unwritten, until you call this function).
- Parameters:
status – Pointer to rboot_write_status structure defining the write status
- bool rboot_write_flash(rboot_write_status *status, const uint8_t *data, uint16_t len)
Write data to flash memory.
Note
Call repeatedly to write data to the flash, starting at the address specified on the prior call to rboot_write_init. Current write position is tracked automatically. This method is likely to be called each time a packet of OTA data is received over the network.
Note
Call rboot_write_init before calling this function to get the rboot_write_status structure
- Parameters:
status – Pointer to rboot_write_status structure defining the write status
data – Pointer to a block of uint8_t data elements to be written to flash
len – Quantity of uint8_t data elements to write to flash
- struct rboot_write_status
Structure defining flash write status.
See also
Note
The user application should not modify the contents of this structure.
rboot.h
Defines
- CHKSUM_INIT
- SECTOR_SIZE
- BOOT_CONFIG_SECTOR
- BOOT_CONFIG_MAGIC
- BOOT_CONFIG_VERSION
- MODE_STANDARD
- MODE_GPIO_ROM
- MODE_TEMP_ROM
- MODE_GPIO_ERASES_SDKCONFIG
- MODE_GPIO_SKIP
- RBOOT_RTC_MAGIC
- RBOOT_RTC_READ
- RBOOT_RTC_WRITE
- RBOOT_RTC_ADDR
- BOOT_GPIO_NUM
- MAX_ROMS
- struct rboot_config
Structure containing rBoot configuration.
Note
ROM addresses must be multiples of 0x1000 (flash sector aligned). Without BOOT_BIG_FLASH only the first 8Mbit (1MB) of the chip will be memory mapped so ROM slots containing .irom0.text sections must remain below 0x100000. Slots beyond this will only be accessible via spi read calls, so use these for stored resources, not code. With BOOT_BIG_FLASH the flash will be mapped in chunks of 8MBit (1MB), so ROMs can be anywhere, but must not straddle two 8MBit (1MB) blocks.
Public Members
- uint8_t magic
Our magic, identifies rBoot configuration - should be BOOT_CONFIG_MAGIC.
- uint8_t version
Version of configuration structure - should be BOOT_CONFIG_VERSION.
- uint8_t mode
Boot loader mode (MODE_STANDARD | MODE_GPIO_ROM | MODE_GPIO_SKIP)
- uint8_t current_rom
Currently selected ROM (will be used for next standard boot)
- uint8_t gpio_rom
ROM to use for GPIO boot (hardware switch) with mode set to MODE_GPIO_ROM.
- uint8_t count
Quantity of ROMs available to boot.
- uint8_t unused[2]
Padding (not used)
- uint32_t roms[MAX_ROMS]
Flash addresses of each ROM.
ESP8266 Cache_Read_Enable
12th June 2015 Richard
Since I haven’t seen this documented anywhere, here is my attempt to explain the Cache_Read_Enable function. Valid values and what they do (at a register level) are from decompiling the code. The outcome of those values is based on my own experimentation so my descriptions and explanations may be silly but they currently fit the observed results.
void Cache_Read_Enable(uint8 odd_even, uint8 mb_count, unt8 no_idea);
- Valid values for odd_even
0 – clears bits 24 & 25 of control register 0x3FF0000C 1 – clears bit 24, sets bit 25 other – clears bit 25, sets bit 24
- Function of odd_even
0 – allows access to even numbered mb 1 – allow access to odd numbered mb other – appears to do the same as 1, there must be a difference but I haven’t worked out what it is
- Valid values for mb_count
0-7 – set bits 16, 17 & 18 of control register 0x3FF0000C
- Function of mb_count
- Which odd or even bank to map (according to odd_even option)
e.g. mb_count = 0, odd_even = 0 -> map first 8Mbit of flash e.g. mb_count = 0, odd_even = 1 -> map second 8Mbit of flash e.g. mb_count = 1, odd_even = 0 -> map third 8Mbit of flash e.g. mb_count = 1, odd_even = 1 -> map fourth 8Mbit of flash
- Valid values for no_idea
0 – sets bit 3 of 0x3FF00024 1 – sets bit 26 of 0x3FF0000C and sets bits 3 & 4 of 0x3FF00024
- Function of no_idea
The clue is in the name, I can’t work out what this does from my experiments, but the SDK always sets this to 1.
Source: https://richard.burtons.org/2015/06/12/esp8266-cache_read_enable/
06/04/2021 @author mikee47 UPDATE
RTOS SDK code has changed, now see usage in
esp_fast_boot.c. Call looks like this:Cache_Read_Enable(sub_region, region, SOC_CACHE_SIZE);See
esp_fast_boot_restart(). Code (rearranged) looks like this:extern void pm_goto_rf_on(void); extern void clockgate_watchdog(int on); int esp_fast_boot_restart(void) { const esp_partition_t* to_boot = esp_ota_get_boot_partition(); if (!to_boot) { ESP_LOGI(TAG, "no OTA boot partition"); to_boot = esp_ota_get_running_partition(); if (!to_boot) { ESP_LOGE(TAG, "ERROR: Fail to get running partition"); return -EINVAL; } } uint32_t image_start = to_boot->address; uint32_t image_size = to_boot->size - 4; esp_image_header_t image; int ret = spi_flash_read(image_start, &image, sizeof(esp_image_header_t)); if (ret != ESP_OK) { ESP_LOGE(TAG, "ERROR: Fail to read image head from spi flash error=%d", ret); return -EIO; } uint32_t image_entry = image.entry_addr; uint8_t region; if (image_start < 0x200000) { region = 0; } else if (image_start < 0x400000) { region = 1; } else if (image_start < 0x600000) { region = 2; } else if (image_start < 0x800000) { region = 3; } else { ESP_LOGE(TAG, "ERROR: App bin error, start_addr 0x%08x image_len %d\n", image_start, image_size); return -EINVAL; } uint8_t sub_region; uint32_t image_mask = image_start & 0x1fffff; if (image_mask < 0x100000) { sub_region = 0; } else { sub_region = 1; } pm_goto_rf_on(); clockgate_watchdog(0); REG_WRITE(0x3ff00018, 0xffff00ff); SET_PERI_REG_MASK(0x60000D48, BIT1); CLEAR_PERI_REG_MASK(0x60000D48, BIT1); REG_WRITE(INT_ENA_WDEV, 0); _xt_isr_mask(UINT32_MAX); const uint32_t sp = DRAM_BASE + DRAM_SIZE - 16; Cache_Read_Disable(); Cache_Read_Enable(sub_region, region, SOC_CACHE_SIZE); __asm__ __volatile__( "mov a1, %0\n" : : "a"(sp) : "memory" ); void (*user_start)(size_t start_addr); user_start = (void *)entry_addr; user_start(image_start); }References
Used by
Sming (Esp8266) Component
Sming (Host) Component
Over-The-Air(OTA) Upgrader Library
Environment Variables
PARTITION_TABLE_OFFSET
RBOOT_DIR
RBOOT_ROM_0_BIN
RBOOT_ROM_1_BINSoC support
esp8266
host
Submodule: rboot
rBoot - An open source boot loader for the ESP8266
by Richard A Burton, richardaburton@gmail.com http://richard.burtons.org/
March 2021: Forked for use with Sming and partition tables.
rBoot is designed to be a flexible open source boot loader, a replacement for the binary blob supplied with the SDK. It has the following advantages over the Espressif loader:
Open source (written in C).
Supports up to 256 roms.
Roms can be variable size.
Able to test multiple roms to find a valid backup (without resetting).
Flash layout can be changed on the fly (with care and appropriately linked rom images).
GPIO support for rom selection.
Wastes no stack space (SDK boot loader uses 144 bytes).
Documented config structure to allow easy editing from user code.
Can validate .irom0.text section with checksum.
Temporary next-boot rom selection.
Limitations
The ESP8266 can only map 8Mbits (1MB) of flash to memory, but which 8Mbits to map is selectable. This allows individual roms to be up to 1MB in size, so long as they do not straddle an 8Mbit boundary on the flash. This means you could have four 1MB roms or 8 512KB roms on a 32Mbit flash (such as on the ESP-12), or a combination. Note, however, that you could not have, for example, a 512KB rom followed immediately by a 1MB rom because the 2nd rom would then straddle an 8MBit boundary. By default support for using more than the first 8Mbit of the flash is disabled, because it requires several steps to get it working. See below for instructions.
Building
A Makefile is included, which should work with the gcc xtensa cross compiler. There are two source files, the first is compiled and included as data in the second. When run this code is copied to memory and executed (there is a good reason for this, see my blog for an explanation). The make file will handle this for you, but you’ll need my esptool2 (see github).
To use the Makefile set
SDK_BASEto point to the root of the Espressif SDK and either setXTENSA_BINDIRto the gcc xtensa bin directory or include it in yourPATH. These can be set as environment variables or by editing the Makefile.Two small assembler stub functions allow the bootloader to launch the user code without reserving any space on the stack (while the SDK boot loader uses 144 bytes). This compiles fine with GCC, but if you use another compiler and it will not compile/work for you then uncomment the
#define BOOT_NO_ASMinrboot.hto use a C version of these functions (this uses 32 bytes).Tested with SDK v2.2 and GCC v4.8.5.
Installation
Simply write rboot.bin to the first sector of the flash. Remember to set your flash size correctly with your chosen flash tool (e.g. for esptool.py use the
-fsoption). When run rBoot will create it’s own config at the start of sector two for a simple two rom system. You can can then write your two roms to flash addresses0x2000and (half chip size +0x2000). E.g. for 8Mbit flash:esptool.py write_flash -fs 8m 0x0000 rboot.bin 0x2000 user1.bin 0x82000 user2.binNote: your device may need other options specified. E.g. The nodemcu devkit v1.0 (commonly, but incorrectly, sold as v2) also needs the
-fm diooption.For more interesting rom layouts you’ll need to write an rBoot config sector manually, see next step.
The two testload bin files can be flashed in place of normal user roms for testing rBoot. You do not need these for normal use.
rBoot Config
typedef struct { uint8_t magic; // our magic uint8_t version; // config struct version uint8_t mode; // boot loader mode uint8_t current_rom; // currently selected rom uint8_t gpio_rom; // rom to use for gpio boot uint8_t count; // number of roms in use uint8_t unused[2]; // padding uint32_t roms[MAX_ROMS]; // flash addresses of the roms #ifdef BOOT_CONFIG_CHKSUM uint8_t chksum; // boot config chksum #endif } rboot_config;Write a config structure as above to address
0x1000on the flash. If you want more than 4 roms (default) just increase MAX_ROMS when you compile rBoot. Think about how you intend to layout your flash before you start! Rom addresses must be sector aligned i.e start on a multiple of 4096.
magicshould have value0xe1(defined asBOOT_CONFIG_MAGIC).
versionis used in case the config structure changes after deployment. It is defined as0x01(BOOT_CONFIG_VERSION). I don’t intend to increase this, but you should if you choose to reflash the bootloader after deployment and the config structure has changed.
modecan be0x00(MODE_STANDARD) or0x01(MODE_GPIO_ROM). See below for an explanation ofMODE_GPIO_ROM. There is also an optional extra mode flag0x04(MODE_GPIO_ERASES_SDKCONFIG), see below for details.
current_romis the rom to boot, numbered0tocount-1.
gpio_romis the rom to boot when the GPIO is triggered at boot.
countis the number of roms available (may be less thanMAX_ROMS, but not more).
unused[2]is padding so theuint32_trom addresses are 4 bytes aligned.
romsis the array of flash address for the roms. The default generated config will contain two entries:0x00002000and0x00082000.
chksum(if enabled, not by default) should be the xor of0xeffollowed by each of the bytes of the config structure up to (but obviously not including) the chksum byte itself.Default config
A default config sector will be created on boot if one does not exists, or if an existing config is corrupted, and the default rom will be set to rom 0. If you want to have a very customised config for which the default would not be suitable, you can override the implementation in the
rboot.hheader file. See the comments and example code inrboot.hfor more information.GPIO boot mode
- RBOOT_GPIO_ENABLED
If rBoot is compiled with
BOOT_GPIO_ENABLEDset inrboot.h(orRBOOT_GPIO_ENABLEDset in the Makefile), then GPIO boot functionality will be included in the rBoot binary. The feature can then be enabled by setting the rboot_configmodefield toMODE_GPIO_ROM. You must also setgpio_romin the config to indicate which rom to boot when the GPIO is activated at boot.If the GPIO input pin reads high at boot then rBoot will start the currently selected normal or temp rom (as appropriate). However if the GPIO is pulled low then the rom indicated in config option
gpio_romis started instead.The default GPIO is 16, but this can be overridden in the Makefile (
RBOOT_GPIO_NUMBER) orrboot.h(BOOT_GPIO_NUM). If GPIOs other than 16 are used, the internal pullup resistor is enabled before the pin is read and disabled immediately afterwards. For pins that default on reset to configuration other than GPIO input, the pin mode is changed to input when reading but changed back before rboot continues.After a GPIO boot the
current_romfield will be updated in the config, so the GPIO booted rom should change this again if required.GPIO boot skip mode
- RBOOT_GPIO_SKIP_ENABLED
If rBoot is compiled with
BOOT_GPIO_SKIP_ENABLEDset inrboot.h(orRBOOT_GPIO_SKIP_ENABLEDset in the Makefile), then a GPIO can be used to skip to the next rom at boot. The feature must then be enabled by setting the rboot_config ‘mode’ field toMODE_GPIO_SKIP. This means you do not need to have a dedicated GPIO boot rom. If you have a rom that is technically good (valid checksum, etc.) but has operational problems, e.g. wifi doesn’t work or it crashes on boot, rBoot will not be able to detect that and switch rom automatically. In this scenario rebooting the device while pulling the GPIO low will force rBoot to skip this rom and try the next one instead. In a simple two rom setup this simply toggles booting of the other rom.
RBOOT_GPIO_SKIP_ENABLEDandRBOOT_GPIO_ENABLEDcannot be used at the same time.BOOT_GPIO_NUMis used to select the GPIO pin, as withRBOOT_GPIO_ENABLED.Erasing SDK configuration on GPIO boot (rom or skip mode)
If you set the
MODE_GPIO_ERASES_SDKCONFIGflag in the configuration like this:conf.mode = MODE_GPIO_ROM|MODE_GPIO_ERASES_SDKCONFIG; then a GPIO boot will also the erase the Espressif SDK persistent settings store in the final 16KB of flash. This includes removing calibration constants, saved SSIDs, etc.Note that
MODE_GPIO_ERASES_SDKCONFIGis a flag, so it has to be set as well asMODE_GPIO_ROMto take effect.Linking user code
Each rom will need to be linked with an appropriate linker file, specifying where it will reside on the flash. If you are only flashing one rom to multiple places on the flash it must be linked multiple times to produce the set of rom images. This is the same as with the SDK loader.
Because there are endless possibilities for layout with this loader I don’t supply sample linker files. Instead I’ll tell you how to make them.
For each rom slot on the flash take a copy of the
eagle.app.v6.ldlinker script from the sdk. You then need to modify just one line in it for each rom:irom0_0_seg : org = 0x40240000, len = 0x3C000Change the org address to be
0x40200000(base memory mapped location of the flash) + flash address +0x10(offset of data after the header). The logical place for your first rom is the third sector, address0x2000.0x40200000 + 0x2000 + 0x10 = 0x40202010If you use the default generated config the loader will expect to find the second rom at flash address half-chip-size +0x2000(e.g.0x82000on an 8MBit flash) so theirom0_0_segshould be:0x40200000 + 0x82000 + 0x10 = 0x40282010Due to the limitation of mapped flash (max 8MBit) if you use a larger chip and do not have big flash support enabled the second rom in the default config will still be placed at0x082000, not truly half-chip-size +0x2000. Ideally you should also adjust the len to help detect over sized sections at link time, but more important is the overall size of the rom which you need to ensure fits in the space you have allocated for it in your flash layout plan.Then simply compile and link as you would normally for OTA updates with the SDK boot loader, except using the linker scripts you’ve just prepared rather than the ones supplied with the SDK. Remember when building roms to create them as ‘new’ type roms (for use with SDK boot loader v1.2+). Or if using my esptool2 use the
-boot2option. Note: the test loads included with rBoot are built with-boot0because they do not contain a.irom0.textsection (and so the value ofirom0_0_segin the linker file is irrelevant to them) but ‘normal’ user apps always do.irom checksum
The SDK boot loader checksum only covers sections loaded into ram (data and some code). Most of the SDK and user code remains on the flash and that is not included in the checksum. This means you could attempt to boot a corrupt rom and, because it looks ok to the boot loader, there will be no attempt to switch to a backup rom. rBoot improves on this by allowing the
.irom0.textsection to be included in the checksum. To enable this uncomment#define BOOT_IROM_CHKSUMinrboot.hand build your roms with esptool2 using the-iromchksumoption.Big flash support
This only needs to be enabled if you wish to be able to memory map more than the first 8MBit of the flash. Note you can still only map 8Mbit at a time. Use this if you want to have multiple 1MB roms, or more smaller roms than will fit in 8Mbits. If you have a large flash but only need, for example, two 512KB roms you do not need to enable this mode.
Support in rBoot is enabled by uncommenting the
#define BOOT_BIG_FLASHinrboot.h.Thinking about your linker files is either simpler or more complicated, depending on your usage of the flash. If you intend to use multiple 1MB roms you will only need one linker file and you only need to link once for OTA updates. Although when you perform an OTA update the rom will be written to a different position on the flash, each 8Mbit of flash is mapped (separately) to
0x40200000. So when any given rom is run the code will appear at the same place in memory regardless of where it is on the flash. Your base address for the linker would be0x40202010. (Actually all but the first rom could base at0x40200010(because they don’t need to leave space for rBoot and config) but then you’re just making it more complicated again!)If you wanted eight 512KB roms you would need two linker files - one for the first half of any given 8Mbits of flash and another for the second half. Just remember you are really laying out within a single 8MBit area, which can then be replicated multiple times on the flash.
Now the clever bit - rBoot needs to hijack the memory mapping code to select which 8Mbits gets mapped. There is no API for this, but we can override the SDK function. First we need to slightly modify the SDK library
libmain.a, like so:xtensa-lx106-elf-objcopy -W Cache_Read_Enable_New libmain.a libmain2.aThis produces a version of libmain with a ‘weakened’
Cache_Read_Enable_Newfunction, which we can then override with our own. Modify your Makefile to link against the librarymain2instead ofmain.Next add
rboot-bigflash.c(from theappcodedirectory) &rboot.hto your project - this adds the replacementCache_Read_Enable_Newto your code.Getting gcc to apply the override correctly can be slightly tricky (I’m not sure why, it shouldn’t be). One option is to add
-u Cache_Read_Enable_Newto yourLD_FLAGSand change the order of objects on the LD command so yourobjects/.afile is before the libraries. Another way that seems easier was to#include rboot-bigflash.cinto the main .c file, rather than compiling it to a separate object file. I can’t make any sense of that, but I suggest you uncomment the message in theCache_Read_Enable_Newfunction when you first build with it, to make sure you are getting your version into the rom.Now when rBoot starts your rom, the SDK code linked in it that normally performs the memory mapping will delegate part of that task to rBoot code (linked in your rom, not in rBoot itself) to choose which part of the flash to map.
Temporary boot option and rBoot<–>app communication
- RBOOT_RTC_ENABLED
To enable communication between rBoot and your app you should enable the
BOOT_RTC_ENABLEDoption inrboot.h. rBoot will then use the RTC data area to pass a structure with boot information which can be read by the app. This will allow the app to determine the boot mode (normal, temporary or GPIO) and the booted rom (even if it is a tempoary boot). Your app can also update this structure to communicate with rBoot when the device is next rebooted, e.g. to instruct it to temporarily boot a different rom to the one saved in the config. See the api documentation and/or the rBoot sample project for more details. Note: the message “don’t use rtc mem data”, commonly seen on startup, comes from the sdk and is not related to this rBoot feature.Simple RPC
Sming wrapper for the Arduino simpleRPC library. This provides a way to export functions as remote procedure calls.
References
Used by
HostEd Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: simpleRPC
Simple RPC implementation for Arduino.
This library provides a simple way to export Arduino functions as remote procedure calls. The exported method definitions are communicated to the host, which is then able to generate an API interface.
Features:
For each method, only one line of code is needed for exporting.
Automatic parameter- and return type inference.
Support for all native C types and strings.
Support for arbitrary functions and class methods.
Optional function and parameter naming and documentation.
Support for PROGMEM’s
F()macro to reduce memory footprint.Support for compound data structures like Tuples, Objects (Tuples with internal structure), Vectors and arbitrary combinations of these.
Support for reading multidimensional C arrays (e.g.,
int**).Support for different types of I/O interfaces via plugins, e.g.,
Bluetooth.
Ethernet (untested).
Hardware serial.
RS485 serial.
Software serial (untested).
USB serial.
WiFi.
Wire (untested).
Support for using multiple interfaces at the same time.
The Arduino library is independent of any host implementation, a Python API client library is provided as a reference implementation.
Please see ReadTheDocs for the latest documentation.
Quick start
Export any function e.g.,
digitalRead()anddigitalWrite()using theinterface()function.#include <simpleRPC.h> void setup() { Serial.begin(9600); } void loop() { interface(Serial, digitalRead, "", digitalWrite, ""); }These functions are now available on the host under names
method0()andmethod1().The documentation string can be used to name and describe the method.
interface( Serial, digitalRead, "digital_read: Read digital pin. @pin: Pin number. @return: Pin value.", digitalWrite, "digital_write: Write to a digital pin. @pin: Pin number. @value: Pin value.");This is reflected on the host, where the methods are now named
digital_read()anddigital_write()and where the provided API documentation is also available. In the client reference implementation documentation, contains an example on how this works.Further reading
Please read section usage for more information about exporting normal functions, class member functions and documentation conventions.
If you want to create your own host library implementation for other programming languages, the section protocol should help you on your way.
SSL: Secure Sockets Layer
https://en.m.wikipedia.org/wiki/Transport_Layer_Security
Sming supports multiple SSL implementations, currently with adapters for:
If you want to use SSL then take a look at the Basic SSL sample for creating SSL clients, and HttpServer Config Network for SSL servers.
Configuration Variables
- ENABLE_SSL
0 (default): SSL requires lots of RAM and some intensive processing, so to conserve resources it is disabled by default.
1: to enable the default SSL adapter. At the moment that is Axtls.
Axtls: to enable SSL support using the AXTLS 8266 component.
Bearssl: to enable SSL support using the Bear SSL component.
API Documentation
SSL: Upgrading
Introduction
Sming v4.1 introduced some major changes in the SSL architecture to support multiple adapters.
The default adapter is still based on axTLS, and it can be enabled in your application by providing the
ENABLE_SSLdirective either in your component.mk file or during compilation.Migration
The access to the SSL connection from the TcpConnection is simplified and fetching information about SSL certificate and session id is easier than before.
The old code was looking like this:
SSL* ssl = connection.getSsl(); if(ssl) { const char* common_name = ssl_get_cert_dn(ssl, SSL_X509_CERT_COMMON_NAME); if(common_name) { debugf("Common Name:\t\t\t%s\n", common_name); } displayCipher(ssl); displaySessionId(ssl); }Now it should be migrated to the following shorter version:
auto ssl = connection.getSsl(); if(ssl != nullptr) { ssl->printTo(Serial); }SSL initialisation in TCP clients or servers is done using an
Ssl::Session::InitDelegatecallback.Old code looking like this:
MqttClient* getMqttClient() { if(mqtt == nullptr) { mqtt = new MqttClient(); mqtt->addSslOptions(SSL_SERVER_VERIFY_LATER); // << this is where we were setting SSL options Url url;Has to be migrated to the following code:
void sslInit(Ssl::Session& session) { session.options.verifyLater = true; } MqttClient* getMqttClient() { if(mqtt == nullptr) { mqtt = new MqttClient(); mqtt->setSslInitHandler(sslInit); // << this is where the sslInit callback is set Url url;It is possible to create an SSL enabled server. The excerpt below demonstrates this and it is part of the HttpServer Config Network sample. Pay attention to the security considerations and limitations using this on a microcontroller with limited RAM:
void startWebServer() { #ifdef ENABLE_SSL server.setSslInitHandler([](Ssl::Session& session) { debug_i("SSL Init handler: setting server keyCert"); session.keyCert.assign(serverKey, serverCert); }); server.listen(443, true); #else server.listen(80); #endif server.paths.set("/", onIndex); server.paths.set("/ipconfig", onIpConfig); server.paths.set("/ajax/get-networks", onAjaxNetworkList); server.paths.set("/ajax/connect", onAjaxConnect); server.paths.setDefault(onFile); }Setting client certificates, ssl options and pinning for a HttpRequest is done using onSslInit callback. If you look at the Basic SSL sample you will see that the old way of setting them was as shown below:
HttpRequest* request = new HttpRequest(F("https://www.grc.com/fingerprints.htm")); request->setSslOptions(SSL_SERVER_VERIFY_LATER); request->pinCertificate(fingerprints);The new one is using the following sequence of commands:
auto request = new HttpRequest(F("https://www.grc.com/fingerprints.htm")); request->onSslInit(grcSslInit);A sample callback is given below. In the callback the developer has access to the current SSL session and HTTP request and can modify them accordingly:
void grcSslInit(Ssl::Session& session, HttpRequest& request) { static const Ssl::Fingerprint::Cert::Sha1 fingerprint PROGMEM = { ... }; session.validators.pin(fingerprint); // We're using validators, so don't attempt to validate full certificate session.options.verifyLater = true; // Go with maximum buffer sizes session.maxBufferSize = Ssl::MaxBufferSize::K16; }Note also that the
Fingerprintsclass has been removed. Instead, we use methods ofsession.validatorsto:
Pin fingerprints;
Add one or more custom callback validators;
Implement custom validators by inheriting from
Ssl::Validator.Cryptographic support
Some basic class-based cryptographic support is provided via the Cryptographic Support library, organised within the
Cryptonamespace.This is primarily for use with the SSL interface but does not require SSL to be enabled.
Alternatively, the cryptographic ‘C’ libraries themselves may be used directly by your application, regardless of which SSL adapter is in use, or even if SSL is disabled.
For example the following old code is using axTLS cryptographic functions:
char* loadPsk(int* keylen) { SHA1_CTX sha_ctx; // ... SHA1_Init(&sha_ctx); SHA1_Update(&sha_ctx, (uint8_t*)buffer, strlen(buffer)); SHA1_Final(digest, &sha_ctx);For this code to work you should include the following header:
#include <axtls-8266/crypto/crypto.h>And also make sure that your application component.mk file has the following line:
COMPONENT_DEPENDS += axtls-8266SSL namespace
All SSL related classes and types are organized in a separate namespace called
Ssl. For example you should useSsl::KeyCertPairinstead ofSslKeyCertPairandSsl::Fingerprintsinstead ofSslFingerprints.Comparison of SSL implementations
Memory usage
axTLS uses dynamic (heap) allocations which causes issues with large fragment sizes. If you run into memory problems, try setting
ENABLE_CUSTOM_HEAP.Bear SSL uses fixed buffers and does not suffer from this limitation.
{ todo }
Session
- class Session
Handles all SSL activity for a TCP connection.
A session is created for every TCP connection where
useSslis specified. It is then passed to any registered session initialisation callbacks for customisation.Public Functions
- inline const SessionId *getSessionId() const
If available, return the current SSL Session ID.
- Return values:
SessionId* – If connection hasn’t been established, may return Null
- bool onAccept(TcpConnection *client, tcp_pcb *tcp)
Called when a client connection is made via server TCP socket.
- Parameters:
client – The client TCP socket
tcp – The low-level TCP connection to use for reading and writing
- Return values:
bool – true if the connection may proceed, false to abort
- inline void setConnection(Connection *connection)
Called by TcpConnection to set the established SSL connection.
- Parameters:
connection – The server connection
- inline Connection *getConnection()
Get the currently active SSL connection object.
- Return values:
Connection* –
- bool onConnect(tcp_pcb *tcp)
Handle connection event.
- Parameters:
tcp –
- Return values:
bool – true on success, false to abort the connection
- inline bool isConnected() const
Determine if an SSL connection has been fully established.
- Return values:
bool – Connection state
- void close()
End the session.
SSL typically sends a closing handshake at this point
- int read(InputBuffer &input, uint8_t *&output)
Read data from SSL connection.
- Parameters:
input – Source encrypted data
output – Points to decrypted content
- Return values:
int – Size of decrypted data returned, or negative on error
- int write(const uint8_t *data, size_t length)
Write data to SSL connection.
- Parameters:
data –
length –
- Return values:
int – Quantity of bytes actually written, or tcp error code
- bool validateCertificate()
Called by SSL adapter when certificate validation is required.
Note
SSL Internal method
- Return values:
bool – true if validation is success, false to abort connection
- void handshakeComplete(bool success)
Called by SSL adapter when handshake has been completed.
Note
SSL Internal method
- Parameters:
success – Indicates if handshake was successful
- size_t printTo(Print &p) const
For debugging.
Public Members
- String hostName
Used for SNI https://en.wikipedia.org/wiki/Server_Name_Indication.
- KeyCertPair keyCert
Required for server, optional for client.
- MaxBufferSize maxBufferSize = MaxBufferSize::Default
Controls SSL RAM usage.
- const CipherSuites::Array *cipherSuites = &CipherSuites::basic
Configure supported cipher suites. Default is basic.
- int cacheSize = 10
Set session caching.
Server: Number of cached client sessions. Suggested value: 10.
Client: Number of cached session ids. Suggested value: 1.
- ValidatorList validators
List of certificate validators used by Client.
- struct Options
Configurable options.
- enum class Ssl::MaxBufferSize
Indicate to SSL how much memory (approximately) to commit for buffers.
A remote SSL server may require data transfers in large (16K) fragments, so restricting buffer sizes may cause connections to such servers to fail.
This must be balanced against other requirements for RAM by the application, therefore this setting can be used to restrict RAM usage.
Note
The ordinal value of this enumeration corresponds to SSL fragment size as defined in Maximum Fragment Length Negotiation https://tools.ietf.org/html/rfc6066
Values:
- enumerator Default
Let SSL implementation decide.
- enumerator B512
512 bytes
- enumerator K1
1024 bytes
- enumerator K2
- enumerator K4
- enumerator K8
- enumerator K16
Cipher Suites
- enum class Ssl::CipherSuite : uint16_t
Cipher suite identifier.
The TLS standard specifies codes using two 8-bit values. We combine these into a single 16-bit value in MSB-LSB order.
For example:
TLS_RSA_WITH_AES_128_CBC_SHA = { 0x00, 0x2F } = 0x002F
See also
Refer to
CipherSuite.hfor defined values.Values:
- enumerator XX
Certificates
- class Certificate
Implemented by SSL adapter to handle certificate operations.
Public Types
Public Functions
- virtual bool getFingerprint(Fingerprint::Type type, Fingerprint &fingerprint) const = 0
Obtain certificate fingerprint.
- Parameters:
type – Which type of fingerprint to return
fingerprint – On success, returned fingerprint
- Return values:
bool – true on success, false if fingerprint not available
- virtual String getName(DN dn, RDN rdn) const = 0
Retrieve an X.509 distinguished name component.
- Parameters:
dn – The desired Distinguished Name
rdn – The component to return
- Return values:
String – The requested Distinguished Name component
- size_t printTo(Print &p) const
Debugging print support.
- class ValidatorList : public Vector<Validator>
Performs certificate validation.
Validators are created in the application’s session initialisation callback. When the certificate has been received, it is checked against each registered validator in turn until successful. All validators are destroyed during this process.
If there are no validators in the list then the certificate will not be checked and the connection accepted.
Public Functions
- inline bool add(Validator *validator)
Add a validator to the list.
- Parameters:
validator – Must be allocated on the heap
- template<class T>
inline bool pin(const T &fingerprint)Pin a fingerprint.
Creates and adds a fingerprint validator to the list
- inline bool add(ValidatorCallback callback, void *data = nullptr)
Register a custom validator callback.
- Parameters:
callback –
data – User-provided data (optional)
- bool validate(const Certificate *certificate)
Validate certificate via registered validators.
We only need one match for a successful result, but we free all the validators. This method must be called no more than ONCE.
Note
Called by SSL framework.
- Parameters:
certificate – When called with nullptr will free all validators, then fail
- Return values:
bool – true on success, false on failure
Public Members
- Fingerprint::Types fingerprintTypes
Contains a list of registered fingerprint types.
Allows implementations to avoid calculating fingerprint values which are not required, as this is computationally expensive.
- class Validator
Base validator class.
Validation is performed by invoking each validator in turn until a successful result is obtained.
Custom validators may either override this class, or use a callback.
Subclassed by Ssl::CallbackValidator, Ssl::FingerprintValidator< FP >
- union Fingerprint
Various types of fingerprint.
Applications should use the appropriate type to define a fingerprint, for example:
static const Fingerprint::Cert::Sha1 fingerprint PROGMEM = { ... };Public Types
- enum class Type
SSL Certificate fingerprint type.
Values:
- enumerator CertSha1
SHA1 Fingerprint of entire certificate.
- enumerator CertSha256
SHA256 Fingerprint of entire certificate.
- enumerator PkiSha256
SHA256 Fingerprint of Public Key Information.
- union Cert
Fingerprints for the entire Certificate.
- struct Sha1
Fingerprint based on the SHA1 value of the certificate.
The SHA1 hash of the entire certificate. This changes on each certificate renewal so needs to be updated every time the remote server updates its certificate.
Advantages: Takes less time to verify than SHA256
Disadvantages: Likely to change periodically
- struct Sha256
Fingerprint based on the SHA256 value of the certificate.
Typically displayed in browser certificate information
Public Static Attributes
- static constexpr Type type = Type::CertSha256
- union Pki
@Fingerprints for the Public Key only
- struct Sha256
Fingerprint based on the SHA256 value of the Public Key Subject in the certificate.
For HTTP public key pinning (RFC7469), the SHA-256 hash of the Subject Public Key Info (which usually only changes when the public key changes) is used.
Advantages: Doesn’t change frequently
Disadvantages: Takes more time (in ms) to verify.
- class KeyCertPair
Class to manage an SSL key certificate with optional password.
Unnamed Group
- bool assign(const uint8_t *newKey, unsigned newKeyLength, const uint8_t *newCertificate, unsigned newCertificateLength, const char *newKeyPassword = nullptr)
Create certificate using provided values.
Note
We take a new copy of the certificate
- Parameters:
newKey –
newKeyLength –
newCertificate –
newCertificateLength –
newKeyPassword –
- Return values:
bool – false on memory allocation failure
Public Functions
- inline bool assign(const KeyCertPair &keyCert)
Assign another certificate to this structure.
Note
We take a new copy of the certificate
- Parameters:
keyCert –
- Return values:
bool – false on memory allocation failure
- using Ssl::ValidatorCallback = Delegate<bool(const Certificate *certificate, void *data)>
Validator callback function.
Note
Callback must ALWAYS release any allocated memory before returning. If called with certificate = NULL then just release memory and return false.
- Param ssl:
Contains certificate to validate (may be NULL)
- Param data:
Data for the callback to use
- Retval bool:
true if validation succeeded
SSL Adapter API
These classes provide the interface between a
Ssl::Sessionand an appropriate adapter.Error codes
Error codes are implementation specific, however 0 always indicates success and < 0 for error.
To obtain a description for an error code, use
Ssl::Connection::getErrorString().SSL Alerts are reported via error codes. To obtain the alert code call
Ssl::Connection::getAlert()which returns anSsl::Alertcode. If the error code is not an alert thenAlert::INVALIDis returned.Classes
- class Factory
Implemented by SSL adapter.
- class Context
Implemented by SSL adapter to create and manage SSL connections.
Public Functions
- virtual bool init() = 0
Initializer method that must be called after object creation and before the creation of server or client connections.
- Return values:
bool – true on success
- virtual Connection *createClient(tcp_pcb *tcp) = 0
Creates client SSL connection. Your SSL client use this call to create a client connection to remote server.
- Return values:
Connection* –
- virtual Connection *createServer(tcp_pcb *tcp) = 0
Creates server SSL connection. Your SSL servers use this call to allow remote clients to connect to them and use SSL.
- Return values:
Connection* –
- class Connection : public Printable
Implemented by SSL adapter to handle a connection.
Returned
interror codes are 0 for success, or < 0 for error.The error codes themselves are implementation-specific. Use
getErrorString()to obtain the message. SSL Alerts are also reported via error codes and can be obtained using a call togetAlert().Public Functions
- virtual bool isHandshakeDone() const = 0
Checks if the handshake has finished.
- Return values:
bool – true on success
- virtual int read(InputBuffer &input, uint8_t *&output) = 0
Reads encrypted information and decrypts it.
- Parameters:
input – Source encrypted data
output – Pointer to decrypted plaintext buffer
- Return values:
0 – : handshake is still in progress > 0 : there is decrypted data < 0 : error
- virtual int write(const uint8_t *data, size_t length) = 0
Converts and sends plaintext data.
- Parameters:
data –
length –
- Return values:
int – length of the data that was actually written < 0 on error
- virtual CipherSuite getCipherSuite() const = 0
Gets the cipher suite that was used.
- Return values:
CipherSuite – IDs as defined by SSL/TLS standard
- virtual SessionId getSessionId() const = 0
Gets the current session id object. Should be called after handshake.
- Return values:
SessionId –
- virtual const Certificate *getCertificate() const = 0
Gets the certificate object. That object MUST be owned by the Connection implementation and should not be freed outside of it.
- Return values:
Certificate* – Returns NULL if there is no certificate available
- virtual size_t printTo(Print &p) const override
For debugging.
References
Used by
Networking Support Component
Environment Variables
SSL_DEBUGSoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Terminal
Introduction
This Component provides make targets to support a serial terminal for communicating with devices.
The default serial terminal is miniterm. If you don’t have it installed already, you can install it via pip using the following command:
make python-requirements(You’ll need python, of course.)
Options
- COM_PORT
Default port for device communications. Default value depends on the development platform being used.
- COM_OPTS
Additional options to pass to the terminal.
- TERMINAL
Command line to use when running
make terminal. Redefine if you want to use a different terminal application.
- KILL_TERM
Command line to use to kill the running terminal process If the terminal never runs in the background and the warnings annoy you, just clear it.
References
Used by
Sming (main) Component
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
uzlib
Deflate/Zlib-compatible LZ77 compression/decompression library
References
Used by
RP2040 Core Component Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: uzlib
uzlib - Deflate/Zlib-compatible LZ77 compression/decompression library
uzlib is a library which can decompress any valid Deflate, Zlib, and Gzip (further called just “Deflate”) bitstream, and compress data to Deflate- compatible bitstream, albeit with lower compression ratio than Zlib Deflate algorithm (very basic LZ77 compression algorithm is used instead, static Deflate Huffman tree encoding is used for bitstream).
uzlib aims for minimal code size and runtime memory requirements, and thus suitable for (deeply) embedded systems.
uzlib is based on:
tinf library by Joergen Ibsen (Deflate decompression)
Deflate Static Huffman tree routines by Simon Tatham
LZ77 compressor by Paul Sokolovsky
Library integrated and maintained by Paul Sokolovsky.
2014-2020 Paul Sokolovsky
uzlib library is licensed under Zlib license.
Decompressor features
Handling of input (compressed) stream:
Can reside (fully) in memory.
Can be received, byte by byte, from an application-defined callback function (which e.g. can read it from file or another I/O device).
Combination of the above: a chunk of input is buffered in memory, when buffer is exhausted, the application callback is called to refill it.
Handling of output (decompressed) stream:
In-memory decompression, where output stream fully resides in memory.
Streaming decompression, which allows to process arbitrary-sized streams (longer than available memory), but requires in-memory buffer for Deflate dictionary window.
Application specifies number of output bytes it wants to decompress, which can be as high as UINT_MAX to decompress everything into memory at once, or as low as 1 to decompress byte by byte, or any other value to decompress a chunk of that size.
Note that in regard to input stream handling, uzlib employs callback-based, “pull-style” design. The control flow looks as follows:
Application requests uzlib to decompress given number of bytes.
uzlib performs decompression.
If more input is needed to decompress given number of bytes, uzlib calls back into application to provide more input bytes. (An implication of this is that uzlib will always return given number of output bytes, unless end of stream (or error) happens).
The original Zlib library instead features “push-style” design:
An application prepares arbitrary number of input bytes in a buffer, and free space in output buffer, and calls Zlib with these buffers.
Zlib tries to decode as much as possible input and produce as much as possible output. It returns back to the application if input buffer is exhausted, or output buffer is full, whatever happens first.
Currently, uzlib doesn’t support push-style operation a-la Zlib.
Compressor features
Compressor uses very basic implementation of LZ77 algorithm using hash table to find repeating substrings. The size of the hash table (on which compression efficiency depends), pointer to the hashtable memory, and the size of LZ77 dictionary should be configured in
struct uzlib_comp.Currently, compressor doesn’t support streaming operation, both input and output must reside in memory. Neither it supports incremental operation, entire input buffer is compressed at once with a single call to uzlib.
API and configuration
The API is defined in the file uzlib.h and should be largely self-describing. There are also examples implementing gzip-compatible compression and decompression applications in examples/ for further reference. (You may also refer to the original
tinfREADME below for additional information, but it’s mostly provided for historical reference, anduzliblargely evolved beyond it).There are some compile-time options for the library, defined in the file uzlib_conf.h. They can be altered directly in the file, or passed as the compiler options (
-DXXX=YYY) when building library.Binary sizes
To give an impression of code/data sizes of uzlib, the following figures are provided. Numbers for *.o files are code sizes of individual components (tinflate.o is decompressor, genlz77.o and defl_static.o - compressor), and TINF_DATA is the size of the corresponding data structure. These numbers are provided for different architectures, with default uzlib configuration, and with compilers/their options as specified.
gcc -m32 -Os gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0 2891 src/tinflate.o 381 src/genlz77.o 1685 src/defl_static.o 1284 TINF_DATA arm-none-eabi-gcc -mthumb -mcpu=cortex-m4 -Os arm-none-eabi-gcc (GNU Tools for Arm Embedded Processors 9-2019-q4-major) 9.2.1 20191025 (release) [ARM/arm-9-branch revision 277599] 1620 src/tinflate.o 180 src/genlz77.o 1131 src/defl_static.o 1284 TINF_DATA
Original tinf library README
For historical reference and to provide proper credit, the original ``tinf`` library README follows. NOTE: Many parts no longer apply to ``uzlib``. In particular, API is different, features supported are largely extended, and during decompression, there are checks to avoid erroneous/undefined behavior on incorrect Deflate bitstreams.
tinf - tiny inflate library
Version 1.00
Copyright (c) 2003 Joergen Ibsen
About
tinf is a small library implementing the decompression algorithm for the deflate compressed data format (called ‘inflate’). Deflate compression is used in e.g. zlib, gzip, zip and png.
I wrote it because I needed a small in-memory zlib decompressor for a self- extracting archive, and the zlib library added 15k to my program. The tinf code added only 2k.
Naturally the size difference is insignificant in most cases. Also, the zlib library has many more features, is more secure, and mostly faster. But if you have a project that calls for a small and simple deflate decompressor, give it a try :-)
While the implementation should be fairly compliant, it does assume it is given valid compressed data, and that there is sufficient space for the decompressed data.
Simple wrappers for decompressing zlib streams and gzip’ed data in memory are supplied.
tgunzip, an example command-line gzip decompressor in C, is included.
The inflate algorithm and data format are from ‘DEFLATE Compressed Data Format Specification version 1.3’ (RFC 1951).
The zlib data format is from ‘ZLIB Compressed Data Format Specification version 3.3’ (RFC 1950).
The gzip data format is from ‘GZIP file format specification version 4.3’ (RFC 1952).
Ideas for future versions:
the fixed Huffman trees could be built by
tinf_decode_trees()using a small tablememory for the
TINF_DATAobject should be passed, to avoid using more than 1k of stack spacewrappers for unpacking zip archives and png images
implement more in x86 assembler
more sanity checks
in
tinf_uncompress, the (entry value of)destLenandsourceLenare not usedblocking of some sort, so everything does not have to be in memory
optional table-based huffman decoder
Functionality
void tinf_init();Initialise the global uninitialised data used by the decompression code. This function must be called once before any calls to the decompression functions.
int tinf_uncompress(void *dest, unsigned int *destLen, const void *source, unsigned int sourceLen);Decompress data in deflate compressed format from
source[]todest[].destLenis set to the length of the decompressed data. ReturnsTINF_OKon success, andTINF_DATA_ERRORon error.int tinf_gzip_uncompress(void *dest, unsigned int *destLen, const void *source, unsigned int sourceLen);Decompress data in gzip compressed format from
source[]todest[].destLenis set to the length of the decompressed data. ReturnsTINF_OKon success, andTINF_DATA_ERRORon error.int tinf_zlib_uncompress(void *dest, unsigned int *destLen, const void *source, unsigned int sourceLen);Decompress data in zlib compressed format from
source[]todest[].destLenis set to the length of the decompressed data. ReturnsTINF_OKon success, andTINF_DATA_ERRORon error.unsigned int tinf_adler32(const void *data, unsigned int length);Computes the Adler-32 checksum of
lengthbytes starting atdata. Used bytinf_zlib_uncompress().unsigned int tinf_crc32(const void *data, unsigned int length);Computes the CRC32 checksum of
lengthbytes starting atdata. Used bytinf_gzip_uncompress().Source Code
The source code is ANSI C, and assumes that int is 32-bit. It has been tested on the x86 platform under Windows and Linux.
The decompression functions should be endian-neutral, and also reentrant and thread-safe (not tested).
In src/nasm there are 32-bit x86 assembler (386+) versions of some of the files.
Makefiles (GNU Make style) for a number of compilers are included.
Frequently Asked Questions
Q: Is it really free? Can I use it in my commercial ExpenZip software?
- A: It’s open-source software, available under the zlib license (see
later), which means you can use it for free – even in commercial products. If you do, please be kind enough to add an acknowledgement.
Q: Did you just strip stuff from the zlib source to make it smaller?
- A: No, tinf was written from scratch, using the RFC documentation of
the formats it supports.
Q: What do you mean by: ‘the zlib library .. is more secure’?
- A: The zlib decompression code checks the compressed data for validity
while decompressing, so even on corrupted data it will not crash. The tinf code assumes it is given valid compressed data.
Q: I’m a Delphi programmer, can I use tinf?
- A: Sure, the object files produced by both Borland C and Watcom C should
be linkable with Delphi.
Q: Will tinf work on UltraSTRANGE machines running WhackOS?
A: I have no idea .. please try it out and let me know!
Q: Why are all the makefiles in GNU Make style?
- A: I’m used to GNU Make, and it has a number of features that are missing
in some of the other Make utilities.
Q: This is the first release, how can there be frequently asked questions?
A: Ok, ok .. I made the questions up ;-)
License
tinf - tiny inflate library
Copyright (c) 2003 Joergen Ibsen
This software is provided ‘as-is’, without any express or implied warranty. In no event will the authors be held liable for any damages arising from the use of this software.
Permission is granted to anyone to use this software for any purpose, including commercial applications, and to alter it and redistribute it freely, subject to the following restrictions:
The origin of this software must not be misrepresented; you must not claim that you wrote the original software. If you use this software in a product, an acknowledgment in the product documentation would be appreciated but is not required.
Altered source versions must be plainly marked as such, and must not be misrepresented as being the original software.
This notice may not be removed or altered from any source distribution.
ws_parser
ws_parser is a streaming parser for the WebSocket protocol (RFC 6455).
This library is loosely inspired by joyent/http_parser and shares many of the same attributes: it has no dependencies, makes no allocations or syscalls, and only requires 16 bytes of memory to maintain its parse state.
References
Used by
Networking Support Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
References
Environment Variables
SoC support
Sming Esp32 Architecture
Support building Sming for the Esp32 architecture.
Build variables
- IDF_PATH
This contains the base directory for the ESP-IDF toolchain used to build the framework. This variable is required and must be set accordingly.
- SDK_CUSTOM_CONFIG
Custom SDK settings for a project can be defined in a separate file and setting this value to the location, relative to the project source root directory.
These will be added to the default SDK settings.
To make the settings current, you must run
make sdk-config-clean. This will discard any changes made viamake sdk-menuconfig.
Requirements
Sming requires a slightly modified version of the Espressif SDK. You can find the SDK here https://github.com/mikee47/esp-idf/tree/sming/release/v5.2. See idf_versions below for further details.
Using the Sming installation scripts are the recommended way to install these SDK versions. See Getting Started.
You can find further details in the ESP-IDF documentation.
Building
Make sure that the IDF_PATH is set.
Also make sure that the other ESP-IDF environmental variables are set.
In Linux this can be done using the following command:
source $SMING_HOME/Tools/export.sh
Build the framework and application as usual, specifying SMING_ARCH =Esp32. For example:
cd $SMING_HOME/../samples/Basic_Serial
make SMING_ARCH=Esp32
This builds the application. Once built the application needs to be flashed on a real Esp32 microcontroller to run. Flashing can be done using the following command:
make flash
SDK
Sming comes with pre-compiled libraries and configuration files. If needed you can re-configure ESP-IDF using the command below:
make SMING_ARCH=Esp32 sdk-menuconfig
A re-compilation is required after the change of the configuration. This can be done with the following command:
make SMING_ARCH=Esp32 Sming-build all
If you want to revert to using the default SDK settings then issue the following command:
make SMING_ARCH=Esp32 sdk-config-clean
You can also configure per-project custom settings via SDK_CUSTOM_CONFIG.
SoC variants
Sming leverages the ESP IDF HAL to support multiple processor variants.
This is still at an early stage of development however basic applications should build for the following variants:
esp32 (default)
esp32s2
esp32c3
esp32s3
esp32c2
You can change variants like this:
`
make SMING_SOC=esp32c3
`
Each variant uses a different build directory, e.g. out/Esp32/esp32c3/... to avoid conflicts.
See Esp32 Core Component for further details.
IDF versions
Sming currently supports IDF versions 4.3, 4.4, 5.0 and 5.2.
The default installed IDF version is 4.4. This can be changed as follows:
INSTALL_IDF_VER=5.2 $SMING_HOME/../Tools/install.sh esp32
The installation script creates a soft-link in /opt/esp-idf pointing to the last version installed.
Use the IDF_PATH environment variable or change the soft-link to select which one to use.
After switching versions, run make clean components-clean before re-compiling.
Note
Currently, switching from version 4.x to 5.0 or vice-versa requires an additional step as they use different versions of the ‘pyparsing’ Python library.
If moving from IDF 4.x to 5.0: python -m pip install --upgrade pyparsing
Moving from IDF 5.0 to 4.x: python -m pip install 'pyparsing<2.4'
Components
Sming (Esp32)
This Component builds a library containing architecture-specific code, and defines dependencies for Sming to build for the Esp32.
Interactive debugging on the device
- ENABLE_GDB
In order to be able to debug live directly on the ESP8266 microcontroller you should re-compile your application with
ENABLE_GDB=1directive.- undefined (default)
Compile normally
- 1
Compile with debugging support provided by Esp8266 GDBSTUB for Sming. See also the Live Debug sample.
References
Used by
Sming (main) Component
Environment Variables
TARGET_BINTARGET_OUT
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
Esp32 Drivers
Provides low-level peripheral drivers.
GPIO: General-Purpose I/O
SDK definitions for GPIO.
-
enum GPIO_INT_TYPE
Values:
-
enumerator GPIO_PIN_INTR_DISABLE
Interrupt disabled for this pin
-
enumerator GPIO_PIN_INTR_POSEDGE
Interrupt occurs on positive edge
-
enumerator GPIO_PIN_INTR_NEGEDGE
Interrupt occurs on negative edge
-
enumerator GPIO_PIN_INTR_ANYEDGE
Interrupt occurs on both positive and negative edge
-
enumerator GPIO_PIN_INTR_LOLEVEL
Interrupt occurs when GPIO low
-
enumerator GPIO_PIN_INTR_HILEVEL
Interrupt occurs when GPIO high
-
enumerator GPIO_PIN_INTR_DISABLE
-
enumerator GPIO_PIN_INTR_POSEDGE
-
enumerator GPIO_PIN_INTR_NEGEDGE
-
enumerator GPIO_PIN_INTR_ANYEDGE
-
enumerator GPIO_PIN_INTR_LOLEVEL
-
enumerator GPIO_PIN_INTR_HILEVEL
-
enumerator GPIO_PIN_INTR_DISABLE
hw_timer: Hardware Timers
Driver for hardware timers.
Variables
- USE_US_TIMER
0 (default): Use default /256 prescale for Timer2 1: Use /16 prescale
The following functions depend on Timer2: - NOW() return value, the Timer2 tick count - Software timers - System time
Software timers are driven by Timer2, which by default uses a /256 prescale providing a resolution of 3.2us and a range of 1’ 54”.
Enabling this setting increases the resolution to 200ns but reduces the maximum software timer to 7” 9.5s.
API Documentation
-
enum hw_timer_clkdiv_t
Values:
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enum hw_timer_intr_type_t
Values:
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enum hw_timer_source_type_t
Values:
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enum hw_timer_clkdiv_t
Values:
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enum hw_timer_intr_type_t
Values:
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enum hw_timer_source_type_t
Values:
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enum hw_timer_clkdiv_t
Values:
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enum hw_timer_intr_type_t
Values:
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enum hw_timer_source_type_t
Values:
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
typedef void (*hw_timer_callback_t)(void *arg)
-
typedef void (*hw_timer_callback_t)(void *arg)
-
typedef void (*hw_timer_callback_t)(void *arg)
-
struct hw_timer_private_t hw_timer_private
-
void hw_timer1_attach_interrupt(hw_timer_source_type_t source_type, hw_timer_callback_t callback, void *arg)
Attach an interrupt for the timer.
- Parameters:
source_type – Ignored, uses APB clock source
callback – Callback function invoked via timer interrupt
arg – Passed to callback function
source_type –
callback – Callback function invoked via timer interrupt
arg – Passed to callback function
-
inline void hw_timer1_enable(hw_timer_clkdiv_t div, hw_timer_intr_type_t intr_type, bool auto_load)
Enable the timer.
Note: With one-shot timer application callback must stop the timer when it is no longer required. This is to reduce callback latency. If this is not done, timer will trigger again when timer counter wraps around to 0. For /16 divisor this is only 1.7s.
- Parameters:
div –
intr_type – Ignored, always level-triggered
auto_load –
div –
intr_type –
auto_load – true for repeating timer, false for one-shot
div –
intr_type –
auto_load –
-
void hw_timer1_write(uint32_t ticks)
Set the timer interval.
Set the timer interval and arm.
- Parameters:
ticks –
-
void hw_timer1_disable(void)
Disable the timer.
-
void hw_timer1_detach_interrupt(void)
Detach interrupt from the timer.
-
uint32_t hw_timer1_read(void)
Get timer1 count.
- Return values:
uint32_t – Current count value, counts from initial value down to 0
-
static uint32_t hw_timer2_read(void)
Read current timer2 value.
- Return values:
uint32_t –
-
void hw_timer_init(void)
Initialise hardware timers.
Note
Called by startup code
-
void hw_timer2_set_alarm(uint32_t ticks)
Set timer2 alarm count value.
Note
For internal use ONLY; used by software timers
- Parameters:
ticks –
-
uint32_t hw_timer_ticks()
Fetch 32-bit microsecond count.
All timers reference from a single 64-bit counter. We use only the lower 32 bits here as it provides lowest latency and compatibility with existing API.
-
HW_TIMER1_GROUP
-
HW_TIMER1_INDEX
-
MAX_HW_TIMER1_INTERVAL
Maximum timer interval in ticks.
Note
The corresponding time interval depends on the prescaler in use:
ESP32 supports a wide range of prescalers and uses 54-bit counter value. Limiting the range 31 bits avoids issues with overflows and moving to 64-bit calculations./1 - 26.84s /16 - 429.50s /256 - 6871.95s
-
MIN_HW_TIMER1_INTERVAL_US
Minimum hardware interval in microseconds.
Note
Attempting to use repetitive interrupts below this level can lead to system instabliity and lockups, due to the software overhead in servicing the interrupts.
-
NOW()
-
MAX_HW_TIMER1_INTERVAL
Maximum timer interval in ticks.
Note
The corresponding time interval depends on the prescaler in use:
/1 - 0.1048s /16 - 1.677s /256 - 26.84s
-
MIN_HW_TIMER1_INTERVAL_US
Minimum hardware interval in microseconds.
Note
Attempting to use repetitive interrupts below this level can lead to system instabliity and lockups, due to the software overhead in servicing the interrupts.
-
HW_TIMER2_CLKDIV
-
HW_TIMER2_CLK
-
MAX_HW_TIMER1_INTERVAL
Maximum timer interval in ticks.
-
MIN_HW_TIMER1_INTERVAL_US
Minimum hardware interval in microseconds.
Note
Attempting to use repetitive interrupts below this level can lead to system instabliity and lockups, due to the software overhead in servicing the interrupts.
-
HW_TIMER2_CLK
-
struct hw_timer_private_t
- #include <hw_timer.h>
I2S: Inter-IC Serial communications
Introduction
I2S was designed for transfer of digital audio data.
The ESP8266 has two I2S modules (one transmitter and one receiver), both with hardware DMA support, which means transfers from RAM to the hardware SPI FIFO can be handled directly in hardware without any CPU involvement.
Sming I2S support
The Sming driver deals with the complicated of setting up the hardware, using an API similar to that in the Espressif RTOS SDK. In addition, DMA buffers may be accessed directly to avoid double-buffering and the associated RAM and copy overhead.
Applications
Audio
Playing MIDI files, MP3 files, speech synthesis, etc. is all possible using the ESP8266, though many audio applications require considerable processing power. That means you may need to disable WiFi and set the processor to run at full 160MHz speed.
High-quality multi-channel audio requires an external I2S DAC, which is what the protocol was designed for in the first place. You may find problems with insufficient RAM, but you can always add external SPI RAM.
More realistic uses include generating simple tones, beeps, playing pre-recorded WAV audio, etc. to supplement existing projects. This can all be done in the background without disrupting the system’s main purpose, whatever that may be.
For such applications you can generate single-channel audio via the I2S OUT pin, using Pulse-density modulation.
See the Tone Generator library for a demonstration of this.
GPIO Expansion
Expand GPIO using low-cost shift registers. https://github.com/lhartmann/esp8266_reprap.
Pixel-strip control
Devices such as WS2812-based NeoPixels use a simple, single-wire protocol. I2S is ideal for this as it can be used to generate a precisely-timed bitstream with very low CPU loading.
API Documentation
Warning
doxygengroup: Cannot find group “i2s_driver” in doxygen xml output for project “api” from directory: ../api/xml/
OS Timer
-
using smg_timer_func_t = void (*)(void *arg)
-
void smg_timer_arm_ticks(os_timer_t *ptimer, uint32_t ticks, bool repeat_flag)
Set a software timer using the Timer2 tick value.
This function has been added to Sming for more efficient and flexible use of software timers. It can be used alongside the SDK
os_timer_arm_new()function.- Parameters:
ptimer – Timer structure
ticks – Tick count duration for the timer
repeat_flag – true if timer will automatically repeat
-
void smg_timer_setfn(os_timer_t *ptimer, os_timer_func_t pfunction, void *parg)
-
void smg_timer_arm_us(os_timer_t *ptimer, uint32_t time_us, bool repeat_flag)
-
void smg_timer_arm(os_timer_t *ptimer, uint32_t time_ms, bool repeat_flag)
-
void smg_timer_disarm(os_timer_t *ptimer)
-
void smg_timer_done(os_timer_t *ptimer)
-
static inline uint64_t smg_timer_expire(const os_timer_t *ptimer)
-
void os_timer_arm_ticks(os_timer_t *ptimer, uint32_t ticks, bool repeat_flag)
Set a software timer using the Timer2 tick value.
This function has been added to Sming for more efficient and flexible use of software timers. It can be used alongside the SDK
os_timer_arm_new()function.- Parameters:
ptimer – Timer structure
ticks – Tick count duration for the timer
repeat_flag – true if timer will automatically repeat
-
static inline bool os_timer_is_armed(const os_timer_t *ptimer)
-
static inline uint64_t os_timer_expire(const os_timer_t *ptimer)
-
static inline void os_timer_done(os_timer_t *ptimer)
-
os_timer_func_t
-
os_timer_t
-
os_timer_arm
-
os_timer_arm_us
-
os_timer_disarm
-
os_timer_setfn
-
os_timer_arm_ticks
-
os_timer_expire
-
os_timer_done
-
struct smg_timer_t
- #include <os_timer.h>
UART: Universal Asynchronous Receive/Transmit
Custom asynchronous driver.
Warning
doxygengroup: Cannot find group “uart_driver” in doxygen xml output for project “api” from directory: ../api/xml/
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Host Library Component
Sming (Host) Component
Sming (Rp2040) Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
Esp32 Core Component
Introduction
Contains startup code, crash handling and additional Esp32-specific support code, including the ESP-IDF SDK.
If you want to tune ESP-IDF to your needs you should run:
make SMING_ARCH=Esp32 sdk-menuconfig
Followed by:
make
Configuration variables
The following variables may need to be changed if tools are installed in a different location, or if multiple versions are installed. By default, the most current version will be used.
- ESP32_COMPILER_PATH
Location of xtensa compiler toolchain
- ESP32_ULP_PATH
Location of ULP compiler toolchain
- ESP32_OPENOCD_PATH
Location of ESP32 version of Open OCD JTAG debugging tools.
- ESP32_PYTHON_PATH
Location of ESP-IDF python.
- CREATE_EVENT_TASK
default: disabled
Warning
This setting is provided for debugging purposes ONLY.
Sming uses a custom event loop to ensure that timer and task callbacks are all executed in the same thread context.
Sometimes this behaviour can cause issues with IDF code. Setting this to 1 will create the event loop in a separate thread, which is standard IDF behaviour.
Background
An empty ESP IDF project is built which generates a set of libraries and headers which the framework can then be built against.
The project is located in project/{SMING_SOC}.
The code for this project is copied from sdk/project.
The default configuration settings are obtained from sdk/config and written
to project/{SMING_SOC}/sdkconfig.defaults.
When sdk-menuconfig is run, the project/{SMING_SOC}/sdkconfig is modified.
This can be reset using make sdk-config-clean.
If custom settings are required for a project then place these in a separate file
and set SDK_CUSTOM_CONFIG to the location, relative to the project source root directory.
References
Used by
Esp32 Drivers Component
Esp32 WiFi Component
Sming (Esp32) Component
Esp8266 SPI Flash Support Component
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
Esp32 WiFi
All related libraries for WiFi support.
References
Used by
Sming (Esp8266) Component
Host Library Component
Networking Support Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
GDB Stub for Esp32
This defines the command line to use when make gdb is run.
Esp32 debugging is handled via JTAG interface.
No additional code is required as serial debugging is not (currently) implemented.
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Sming (Rp2040) Component
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
Heap
This Component implements heap-related housekeeping functions. Heap usage is tracked using malloc_count. This also provides some validation (using sentinels to detect if memory blocks are overwritten).
- ENABLE_MALLOC_COUNT
We require malloc_count to keep track of heap usage for system_get_free_heap_size(). It does this by hooking the memory allocation routines (malloc, free, etc.). If you wish to disable this behaviour, set ENABLE_MALLOC_COUNT=0. If using tools such as Valgrind, this will provide a cleaner trace.
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
Esp32 LIBC Component
This Component accommodates the differences in runtime libraries for the various supported toolchains.
References
Used by
Esp32 Core Component Component
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Sming (Rp2040) Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
Esp8266 SPI Flash Support
Provides functions for access to flash memory.
API Documentation
-
enum SPIFlashMode
Values:
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_FAST_READ
-
enumerator MODE_SLOW_READ
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enum SPIFlashSpeed
Values:
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enum SPIFlashSize
Values:
-
enumerator SIZE_1MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_4MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
Not listed.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_1MBIT
-
enum SPIFlashMode
Values:
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_FAST_READ
-
enumerator MODE_SLOW_READ
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enum SPIFlashSpeed
Values:
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enum SPIFlashSize
Values:
-
enumerator SIZE_1MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_4MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
Not listed.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_1MBIT
-
enum SPIFlashMode
Values:
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_FAST_READ
-
enumerator MODE_SLOW_READ
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enum SPIFlashSpeed
Values:
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enum SPIFlashSize
Values:
-
enumerator SIZE_1MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_4MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
Not listed.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_1MBIT
-
static inline uint32_t flashmem_get_address(const void *memptr)
Obtain the flash memory address for a memory pointer.
Note
If memptr is not in valid flash memory it will return an offset which exceeds the internal flash memory size.
Note
The flash location is dependent on where rBoot has mapped the firmware.
- Parameters:
memptr –
- Return values:
uint32_t – Offset from start of flash memory
-
uint32_t flashmem_write(const void *from, uint32_t toaddr, uint32_t size)
Write a block of data to flash.
Note
None of the parameters need to be aligned
- Parameters:
from – Buffer to obtain data from
toaddr – Flash location to start writing
size – Number of bytes to write
- Return values:
uint32_t – Number of bytes written
-
uint32_t flashmem_read(void *to, uint32_t fromaddr, uint32_t size)
Read a block of data from flash.
Note
none of the parameters need to be aligned
- Parameters:
to – Buffer to store data
fromaddr – Flash location to start reading
size – Number of bytes to read
- Return values:
uint32_t – Number of bytes written
-
bool flashmem_erase_sector(uint32_t sector_id)
Erase a single flash sector.
- Parameters:
sector_id – the sector to erase
- Return values:
true – on success
-
SPIFlashInfo flashmem_get_info()
Get flash memory information block.
- Return values:
SPIFlashInfo – Information block
-
uint8_t flashmem_get_size_type()
Returns a number indicating the size of flash memory chip.
- Return values:
uint8_t – See SpiFlashInfo.size field for possible values
-
uint32_t flashmem_get_size_bytes()
get the total flash memory size
- Return values:
uint32_t – Size in bytes
-
uint16_t flashmem_get_size_sectors()
Get the total number of flash sectors.
- Return values:
uint16_t – Sector count
-
uint32_t flashmem_find_sector(uint32_t address, uint32_t *pstart, uint32_t *pend)
Helper function: find the flash sector in which an address resides.
Note
Optional parameters may be null
- Parameters:
address –
pstart – OUT/OPTIONAL: Start of sector containing the given address
pend – OUT/OPTIONAL: Last address in sector
- Return values:
uint32_t – Sector number for the given address
-
uint32_t flashmem_get_sector_of_address(uint32_t addr)
Get sector number containing the given address.
- Parameters:
addr –
- Return values:
uint32_t – sector number
-
uint32_t spi_flash_get_id(void)
-
uint32_t flashmem_write_internal(const void *from, uint32_t toaddr, uint32_t size)
write to flash memory
Note
All parameters MUST be aligned to 4-byte word boundaries, including the RAM pointer
- Parameters:
from – Buffer to read data from - MUST be word-aligned
toaddr – Flash address (offset) to write to - MUST be word-aligned
size – Number of bytes to write - MUST be word-aligned
- Return values:
uint32_t – Number of bytes actually written
-
uint32_t flashmem_read_internal(void *to, uint32_t fromaddr, uint32_t size)
Read from flash memory.
Note
All parameters MUST be aligned to 4-byte word boundaries, including the RAM pointer
- Parameters:
to – Buffer to store data - MUST be word-aligned
fromaddr – Flash address (offset) to read from - MUST be word-aligned
size – Number of bytes to read - MUST be word-aligned
- Return values:
uint32_t – Number of bytes actually read
-
uint32_t flashmem_get_first_free_block_address()
-
void flashmem_sfdp_read(uint32_t addr, void *buffer, size_t count)
-
INTERNAL_FLASH_WRITE_UNIT_SIZE
Flash memory access must be aligned and in multiples of 4-byte words.
-
INTERNAL_FLASH_READ_UNIT_SIZE
-
FLASH_TOTAL_SEC_COUNT
-
SYS_PARAM_SEC_COUNT
Number of flash sectors reserved for system parameters at start.
-
FLASH_WORK_SEC_COUNT
-
INTERNAL_FLASH_SECTOR_SIZE
-
INTERNAL_FLASH_SIZE
-
INTERNAL_FLASH_WRITE_UNIT_SIZE
Flash memory access must be aligned and in multiples of 4-byte words.
-
INTERNAL_FLASH_READ_UNIT_SIZE
-
FLASH_TOTAL_SEC_COUNT
-
SYS_PARAM_SEC_COUNT
Number of flash sectors reserved for system parameters at start.
-
FLASH_WORK_SEC_COUNT
-
INTERNAL_FLASH_SECTOR_SIZE
-
INTERNAL_FLASH_SIZE
-
INTERNAL_FLASH_START_ADDRESS
-
FLASH_TOTAL_SEC_COUNT
-
SYS_PARAM_SEC_COUNT
Number of flash sectors reserved for system parameters at start.
-
FLASH_WORK_SEC_COUNT
-
INTERNAL_FLASH_SECTOR_SIZE
-
INTERNAL_FLASH_SIZE
-
INTERNAL_FLASH_START_ADDRESS
-
struct SPIFlashInfo
- #include <esp_spi_flash.h>
SPI Flash memory information block. Copied from bootloader header. See
esp_image_header_t.SPI Flash memory information block. Stored at the beginning of flash memory.
-
struct STORE_TYPEDEF_ATTR
- #include <esp_spi_flash.h>
SPI Flash memory information block. Stored at the beginning of flash memory.
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Host Library Component
Sming (Host) Component
Sming (Rp2040) Component
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
Sming Esp8266 Architecture
Support building Sming for the Esp8266 architecture.
See also ESP Quick Toolchain.
Build variables
- ESP_HOME
This contains the base directory for the toolchain used to build the framework.
Components
Sming (Esp8266)
This Component builds a library containing architecture-specific code, and defines dependencies for Sming to build for the Esp8266.
SDK 3.0+
Default: OFF. In order to use SDK 3.0.0 or newer please follow the instructions here Esp8266 Core Component.
No-WiFi build
Note
This is an EXPERIMENTAL feature. Not all hardware functions may be available.
If a project does not require WiFi (or networking) then setting the DISABLE_WIFI variable
will reduce code size and RAM usage significantly.
It does this using an un-official Esp8266 No WiFi Component.
Custom LWIP
LWIP (LightWeight IP) is a small independent implementation
of the TCP/IP protocol suite used widely in embedded systems. Sming supports several versions of this,
controlled by the ENABLE_CUSTOM_LWIP setting.
- ENABLE_CUSTOM_LWIP
- 0
Use binary Esp8266 LWIP (Espressif) stack.
- 1 (default)
Use custom compiled Esp8266 Open LWIP (version 1) stack. Compared with the Espressif stack, this uses less RAM but consumes FLASH (program) memory. All espconn_* functions are turned off by default, so if you require these add the
ENABLE_ESPCONN=1 directive. The Basic Smart Config example sets this in itscomponent.mkfile.- 2
Use Esp8266 LWIP Version 2 stack. This does not have support for espconn_* functions.
- ENABLE_LWIP_DEBUG
By default, some debug information will be printed for critical errors and situations. Set this to 1 to enable printing of all debug information.
Interactive debugging on the device
- ENABLE_GDB
In order to be able to debug live directly on the ESP8266 microcontroller you should re-compile your application with
ENABLE_GDB=1directive.- undefined (default)
Compile normally
- 1
Compile with debugging support provided by Esp8266 GDBSTUB for Sming. See also the Live Debug sample.
References
Used by
Sming (main) Component
Environment Variables
SoC support
esp8266
Esp8266 Drivers
Provides low-level peripheral drivers.
GPIO: General-Purpose I/O
SDK definitions for GPIO.
-
enum GPIO_INT_TYPE
Values:
-
enumerator GPIO_PIN_INTR_DISABLE
Interrupt disabled for this pin
-
enumerator GPIO_PIN_INTR_POSEDGE
Interrupt occurs on positive edge
-
enumerator GPIO_PIN_INTR_NEGEDGE
Interrupt occurs on negative edge
-
enumerator GPIO_PIN_INTR_ANYEDGE
Interrupt occurs on both positive and negative edge
-
enumerator GPIO_PIN_INTR_LOLEVEL
Interrupt occurs when GPIO low
-
enumerator GPIO_PIN_INTR_HILEVEL
Interrupt occurs when GPIO high
-
enumerator GPIO_PIN_INTR_DISABLE
-
enumerator GPIO_PIN_INTR_POSEDGE
-
enumerator GPIO_PIN_INTR_NEGEDGE
-
enumerator GPIO_PIN_INTR_ANYEDGE
-
enumerator GPIO_PIN_INTR_LOLEVEL
-
enumerator GPIO_PIN_INTR_HILEVEL
-
enumerator GPIO_PIN_INTR_DISABLE
hw_timer: Hardware Timers
Driver for hardware timers.
Variables
- USE_US_TIMER
0 (default): Use default /256 prescale for Timer2 1: Use /16 prescale
The following functions depend on Timer2: - NOW() return value, the Timer2 tick count - Software timers - System time
Software timers are driven by Timer2, which by default uses a /256 prescale providing a resolution of 3.2us and a range of 1’ 54”.
Enabling this setting increases the resolution to 200ns but reduces the maximum software timer to 7” 9.5s.
API Documentation
-
enum hw_timer_clkdiv_t
Values:
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enum hw_timer_intr_type_t
Values:
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enum hw_timer_source_type_t
Values:
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enum hw_timer_clkdiv_t
Values:
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enum hw_timer_intr_type_t
Values:
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enum hw_timer_source_type_t
Values:
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enum hw_timer_clkdiv_t
Values:
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enum hw_timer_intr_type_t
Values:
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enum hw_timer_source_type_t
Values:
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
typedef void (*hw_timer_callback_t)(void *arg)
-
typedef void (*hw_timer_callback_t)(void *arg)
-
typedef void (*hw_timer_callback_t)(void *arg)
-
struct hw_timer_private_t hw_timer_private
-
void hw_timer1_attach_interrupt(hw_timer_source_type_t source_type, hw_timer_callback_t callback, void *arg)
Attach an interrupt for the timer.
- Parameters:
source_type – Ignored, uses APB clock source
callback – Callback function invoked via timer interrupt
arg – Passed to callback function
source_type –
callback – Callback function invoked via timer interrupt
arg – Passed to callback function
-
inline void hw_timer1_enable(hw_timer_clkdiv_t div, hw_timer_intr_type_t intr_type, bool auto_load)
Enable the timer.
Note: With one-shot timer application callback must stop the timer when it is no longer required. This is to reduce callback latency. If this is not done, timer will trigger again when timer counter wraps around to 0. For /16 divisor this is only 1.7s.
- Parameters:
div –
intr_type – Ignored, always level-triggered
auto_load –
div –
intr_type –
auto_load – true for repeating timer, false for one-shot
div –
intr_type –
auto_load –
-
void hw_timer1_write(uint32_t ticks)
Set the timer interval.
Set the timer interval and arm.
- Parameters:
ticks –
-
void hw_timer1_disable(void)
Disable the timer.
-
void hw_timer1_detach_interrupt(void)
Detach interrupt from the timer.
-
uint32_t hw_timer1_read(void)
Get timer1 count.
- Return values:
uint32_t – Current count value, counts from initial value down to 0
-
static uint32_t hw_timer2_read(void)
Read current timer2 value.
- Return values:
uint32_t –
-
void hw_timer_init(void)
Initialise hardware timers.
Note
Called by startup code
-
void hw_timer2_set_alarm(uint32_t ticks)
Set timer2 alarm count value.
Note
For internal use ONLY; used by software timers
- Parameters:
ticks –
-
uint32_t hw_timer_ticks()
Fetch 32-bit microsecond count.
All timers reference from a single 64-bit counter. We use only the lower 32 bits here as it provides lowest latency and compatibility with existing API.
-
HW_TIMER1_GROUP
-
HW_TIMER1_INDEX
-
MAX_HW_TIMER1_INTERVAL
Maximum timer interval in ticks.
Note
The corresponding time interval depends on the prescaler in use:
ESP32 supports a wide range of prescalers and uses 54-bit counter value. Limiting the range 31 bits avoids issues with overflows and moving to 64-bit calculations./1 - 26.84s /16 - 429.50s /256 - 6871.95s
-
MIN_HW_TIMER1_INTERVAL_US
Minimum hardware interval in microseconds.
Note
Attempting to use repetitive interrupts below this level can lead to system instabliity and lockups, due to the software overhead in servicing the interrupts.
-
NOW()
-
MAX_HW_TIMER1_INTERVAL
Maximum timer interval in ticks.
Note
The corresponding time interval depends on the prescaler in use:
/1 - 0.1048s /16 - 1.677s /256 - 26.84s
-
MIN_HW_TIMER1_INTERVAL_US
Minimum hardware interval in microseconds.
Note
Attempting to use repetitive interrupts below this level can lead to system instabliity and lockups, due to the software overhead in servicing the interrupts.
-
HW_TIMER2_CLKDIV
-
HW_TIMER2_CLK
-
MAX_HW_TIMER1_INTERVAL
Maximum timer interval in ticks.
-
MIN_HW_TIMER1_INTERVAL_US
Minimum hardware interval in microseconds.
Note
Attempting to use repetitive interrupts below this level can lead to system instabliity and lockups, due to the software overhead in servicing the interrupts.
-
HW_TIMER2_CLK
-
struct hw_timer_private_t
- #include <hw_timer.h>
I2S: Inter-IC Serial communications
Introduction
I2S was designed for transfer of digital audio data.
The ESP8266 has two I2S modules (one transmitter and one receiver), both with hardware DMA support, which means transfers from RAM to the hardware SPI FIFO can be handled directly in hardware without any CPU involvement.
Sming I2S support
The Sming driver deals with the complicated of setting up the hardware, using an API similar to that in the Espressif RTOS SDK. In addition, DMA buffers may be accessed directly to avoid double-buffering and the associated RAM and copy overhead.
Applications
Audio
Playing MIDI files, MP3 files, speech synthesis, etc. is all possible using the ESP8266, though many audio applications require considerable processing power. That means you may need to disable WiFi and set the processor to run at full 160MHz speed.
High-quality multi-channel audio requires an external I2S DAC, which is what the protocol was designed for in the first place. You may find problems with insufficient RAM, but you can always add external SPI RAM.
More realistic uses include generating simple tones, beeps, playing pre-recorded WAV audio, etc. to supplement existing projects. This can all be done in the background without disrupting the system’s main purpose, whatever that may be.
For such applications you can generate single-channel audio via the I2S OUT pin, using Pulse-density modulation.
See the Tone Generator library for a demonstration of this.
GPIO Expansion
Expand GPIO using low-cost shift registers. https://github.com/lhartmann/esp8266_reprap.
Pixel-strip control
Devices such as WS2812-based NeoPixels use a simple, single-wire protocol. I2S is ideal for this as it can be used to generate a precisely-timed bitstream with very low CPU loading.
API Documentation
Warning
doxygengroup: Cannot find group “i2s_driver” in doxygen xml output for project “api” from directory: ../api/xml/
OS Timer
-
using smg_timer_func_t = void (*)(void *arg)
-
void smg_timer_arm_ticks(os_timer_t *ptimer, uint32_t ticks, bool repeat_flag)
Set a software timer using the Timer2 tick value.
This function has been added to Sming for more efficient and flexible use of software timers. It can be used alongside the SDK
os_timer_arm_new()function.- Parameters:
ptimer – Timer structure
ticks – Tick count duration for the timer
repeat_flag – true if timer will automatically repeat
-
void smg_timer_setfn(os_timer_t *ptimer, os_timer_func_t pfunction, void *parg)
-
void smg_timer_arm_us(os_timer_t *ptimer, uint32_t time_us, bool repeat_flag)
-
void smg_timer_arm(os_timer_t *ptimer, uint32_t time_ms, bool repeat_flag)
-
void smg_timer_disarm(os_timer_t *ptimer)
-
void smg_timer_done(os_timer_t *ptimer)
-
static inline uint64_t smg_timer_expire(const os_timer_t *ptimer)
-
void os_timer_arm_ticks(os_timer_t *ptimer, uint32_t ticks, bool repeat_flag)
Set a software timer using the Timer2 tick value.
This function has been added to Sming for more efficient and flexible use of software timers. It can be used alongside the SDK
os_timer_arm_new()function.- Parameters:
ptimer – Timer structure
ticks – Tick count duration for the timer
repeat_flag – true if timer will automatically repeat
-
static inline bool os_timer_is_armed(const os_timer_t *ptimer)
-
static inline uint64_t os_timer_expire(const os_timer_t *ptimer)
-
static inline void os_timer_done(os_timer_t *ptimer)
-
os_timer_func_t
-
os_timer_t
-
os_timer_arm
-
os_timer_arm_us
-
os_timer_disarm
-
os_timer_setfn
-
os_timer_arm_ticks
-
os_timer_expire
-
os_timer_done
-
struct smg_timer_t
- #include <os_timer.h>
PWM: Pulse-Width Modulation
The driver interface is defined in the ESP8266 SDK.
Build variables
- ENABLE_CUSTOM_PWM
- undefined
use the Espressif PWM driver
- 1 (default)
Use the New PWM driver, a drop-in replacement for the version provided in the Espressif SDK.
API Documentation
-
void pwm_init(uint32_t period, uint32_t *duty, uint32_t pwm_channel_num, uint32_t (*pin_info_list)[3])
Initialize PWM function, including GPIO selection, period and duty cycle.
Example:
uint32 ioInfo[][3] = { {PWM_0_OUT_IO_MUX, PWM_0_OUT_IO_FUNC, PWM_0_OUT_IO_NUM}, {PWM_1_OUT_IO_MUX, PWM_1_OUT_IO_FUNC, PWM_1_OUT_IO_NUM}, {PWM_2_OUT_IO_MUX, PWM_2_OUT_IO_FUNC, PWM_2_OUT_IO_NUM} }; pwm_init(light_param.pwm_period, light_param.pwm_duty, 3, ioInfo);
Example:
uint32 io_info[][3] = { {PWM_0_OUT_IO_MUX, PWM_0_OUT_IO_FUNC, PWM_0_OUT_IO_NUM}, {PWM_1_OUT_IO_MUX, PWM_1_OUT_IO_FUNC, PWM_1_OUT_IO_NUM}, {PWM_2_OUT_IO_MUX, PWM_2_OUT_IO_FUNC, PWM_2_OUT_IO_NUM} }; pwm_init(light_param.pwm_period, light_param.pwm_duty, 3, io_info);
Note
This API can be called only once.
Note
This API can be called only once.
- Parameters:
period – PWM period
duty – duty cycle of each output
pwm_channel_num – PWM channel number
pin_info_list – Array containing an entry for each channel giving
period – PWM period
duty – duty cycle of each output
pwm_channel_num – PWM channel number
pin_info_list – Array containing an entry for each channel giving
-
void pwm_start(void)
Starts PWM.
This function needs to be called after PWM configuration is changed.
-
void pwm_set_duty(uint32_t duty, uint8_t channel)
Sets duty cycle of a PWM output.
Set the time that high-level signal will last. The range of duty depends on PWM period. Its maximum value of which can be Period * 1000 / 45.
For example, for 1-KHz PWM, the duty range is 0 ~ 22222.
- Parameters:
duty – The time that high-level single will last, duty cycle will be (duty*45)/(period*1000)
channel – PWM channel, which depends on how many PWM channels are used
-
uint32_t pwm_get_duty(uint8_t channel)
Get duty cycle of PWM output.
Duty cycle will be (duty*45) / (period*1000).
- Parameters:
channel – PWM channel, which depends on how many PWM channels are used
- Return values:
uint32 – Duty cycle of PWM output
-
void pwm_set_period(uint32_t period)
Set PWM period.
- Parameters:
period – PWM period in us. For example, 1-KHz PWM period = 1000us.
-
uint32_t pwm_get_period(void)
Get PWM period.
- Return values:
uint32 – Return PWM period in us.
-
uint32_t get_pwm_version(void)
Get version information of PWM.
- Return values:
uint32 – PWM version
UART: Universal Asynchronous Receive/Transmit
Custom asynchronous driver.
Warning
doxygengroup: Cannot find group “uart_driver” in doxygen xml output for project “api” from directory: ../api/xml/
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Host Library Component
Sming (Host) Component
Sming (Rp2040) Component
Environment Variables
SoC support
esp8266
Esp8266 LWIP (Espressif)
The Espressif SDK provides a default version of LWIP, enabled by using this Component.
See ENABLE_CUSTOM_LWIP.
References
SoC support
esp8266
Esp8266 Open LWIP (version 1)
This Component provides the default Sming LWIP stack, which uses less RAM than the Espressif version.
See ENABLE_CUSTOM_LWIP.
- ENABLE_LWIPDEBUG
- 0 (default)
Standard build
- 1
Enable debugging output
You can increase debugging for certain areas by modifying debug options in
esp-open-lwip/include/lwipopts.h.
- ENABLE_ESPCONN
The Espressif SDK defines a network API consisting of functions which start with
espconn_.- 0 (default)
Disabled
- 1
Enable espconn_ functions
References
Environment Variables
SoC support
esp8266
Submodule: esp-open-lwip
Esp8266 Core Component
Contains startup code, crash handling and additional Esp8266-specific support code.
Sming uses libraries from the ESP8266 NON-OS SDK version 3, imported as a submodule. The header and linker files are provided by this Component.
References
Used by
Esp8266 Drivers Component
Esp8266 LWIP (Espressif) Component
Esp8266 Open LWIP (version 1) Component
Esp8266 WiFi Component
Esp8266 GDBSTUB for Sming Component
Esp8266 LWIP Version 2 Component
Sming (Esp8266) Component
Esp8266 SPI Flash Support Component
Environment Variables
FLASH_INIT_DATAFLASH_INIT_DATA_VCC
SoC support
esp8266
Submodule: sdk
Support Policy for ESP8266 NonOS
Starting from December 2019,
We will not add any new features to the ESP8266 NonOS SDK.
We will only fix critical bugs in the ESP8266 NonOS SDK.
We will only maintain the master branch of ESP8266 NonOS SDK, which is a continuously bug-fix version based on v3.0. This means:
All other released branches will not be updated.
All the future versions will be released from only the master branch mentioned above.
It is suggested that the ESP8266_RTOS_SDK, instead of ESP8266 NonOS SDK, be used for your projects.
The latest ESP8266_RTOS_SDK allows users to develop applications using an architecture that are compatible with the SDKs of all Espressif chips, including ESP8266 series, ESP32 series, and the upcoming new series of chips. Switching to ESP8266_RTOS_SDK will helps users to:
Eliminate the necessity to maintain more than one applications (for different chips), thus greatly reducing maintenance costs.
Easily switch to other Espressif chips in the future for enhanced flexibility, less dependency, and reduced time-to-market.
Thank you for your interest in Espressif products.
ESP8266 NonOS 支持政策
自 2019 年 12 月起，我们将：
停止为 ESP8266 NonOS 新增任何功能。
仅修复 ESP8266 NonOS 的关键 bug。
所有更新仅在 master 分支进行，即基于 v3.0.0 的持续 bug 修复版本。这意味着：
其他任何 release 分支均不再提供维护；
所有更新均将通过上述 master 分支发布。
建议客户使用新版 ESP8266_RTOS_SDK。
简单来说，新版 ESP8266_RTOS_SDK 可帮助客户避免对单一 SDK 的依赖，允许客户应用程序同时兼容多款乐鑫芯片，包括 ESP8266 系列、ESP32 系列以及未来发布的新产品。使用 ESP8266_RTOS_SDK 允许客户：
避免同时维护针对不同芯片的多套应用程序，从而降低维护成本。
未来可轻松切换至其他乐鑫芯片，从而提高灵活性、降低对单一芯片的依赖，并缩短上市时间。
感谢大家对乐鑫的支持与关注。
ESP8266_NONOS_SDK
All documentations @ http://espressif.com/en/support/download/documents?keys=&field_type_tid%5B%5D=14
Notes
Please add user_pre_init() in your project, which will be called before user_init(). And you MUST call system_partition_table_regist() in user_pre_init to register your project partition table.
The following partition address CAN NOT be modified, and you MUST give the correct address. They are retated to the flash map, please refer to ESP8266 SDK Getting Started Guide or ESP8266 SDK 入门指南.
SYSTEM_PARTITION_BOOTLOADER
SYSTEM_PARTITION_OTA_1
SYSTEM_PARTITION_OTA_2
SYSTEM_PARTITION_SYSTEM_PARAMETER
If you donot use Non-FOTA bin, eagle.irom0.text.bin and irom0.text MUST be downloaded the fixed address, which also can be found in ESP8266 SDK Getting Started Guide or ESP8266 SDK 入门指南, and you can define their partition type after SYSTEM_PARTITION_CUSTOMER_BEGIN.
Esp8266 No WiFi
Background
Based on SDK 3.0.1 disassemblies with reference and thanks to:
https://github.com/pvvx/esp8266web Further development of
The problem we have is to disentangle the wifi functions from the SDK within the
app_main.o and user_interface.o modules from libmain.a.
In particular, the user_start() code which mixes up
wifi stuff with system code so it cannot be easily separated.
The SDKnoWiFi implements the startup code and some system functions but contains a lot of stuff which is provided by the SDK and in other parts of the Sming framework. We need to provide replacement functions to interoperate correctly with the remaining SDK code.
Advantages
Reduces code image size when networking/WiFi is not required
Eliminates need for SDK-specific partitions (rf_cal, phy_init, sys_param)
Process
This is the approach used to create this Component.
Remove user_interface from libmain:
ar d libmain.a user_interface.o
Implement
call_user_start_local. The actual entry point iscall_user_start, in vector.o, which is required. (Note: This symbol was -u in the linker script but this is un-necessary and has been removed.)Build and implement missing functions.
Where functions can be passed directly to the ROM code (e.g. system_os_task -> ets_task) these are defined in the no.wifi.ld linker script.
Other code is located in a .c file named according to the module it replaces in libmain.a.
There are various options for integrating this into Sming, but I’ve decided the most useful
way is to have it as a link-time option. We either link esp_wifi or esp_no_wifi
as required, no rebuild to the framework is necessary. This is controlled by the
DISABLE_WIFI setting.
This means that if there are any WiFi routines used in the project the link step will fail.
An alternate approach is to separate out the Network/WiFi stuff in the framework into separate Components which are themselves only included. However, some of the network code could be useful in a non-networked application. There may even be an alternate physical network layer implemented (e.g. Ethernet over SPI) so it’s simplest to just leave the framework intact.
Library Disassembly
You can create a disassembly of the relevant SDK libraries for inspection by running this command from a project directory:
make sdk-disassemble
If you want to disassemble other SDK libraries, do this:
make sdk-disassemble SDK_LIBLIST="crypto net80211"
Known issues
Further work is required to implement the following (list incomplete):
Sleep/power saving modes
Partition tables
Analogue reading
References
SoC support
esp8266
Esp8266 WiFi
All related libraries for WiFi support. Definitions are provided in the Espressif Non-OS SDK.
References
Used by
Sming (Esp8266) Component
Host Library Component
Networking Support Component
SoC support
esp8266
Esp8266 GDBSTUB for Sming
Background
This is a rewrite of gdbstub based on the esp8266 Arduino project.
To use the GNU Debugger (GDB) with Sming requires your application to
include some code (gdbstub) which communicates via the serial port.
On the ESP8266 only UART0 may be used for this as UART1 is
transmit-only.
The gdbstub code will only be built if you specify ENABLE_GDB
=1 when compiling your application. At startup, before your init()
function is called, it will claim UART0 so your application will be
unable to use it directly. Therefore, the default port for Serial
is changed to UART2.
UART2 is a ‘virtual’ serial port to enable serial communications to work correctly when GDB-enabled. Read/write calls and serial callbacks are handled via gdbstub. Baud rate changes affect UART0 directly.
Note that target refers to the application being debugged, and
host the development system running the GDB program.
Refer to the official GDB documentation for further details.
GDB
This is the application which runs on your development system and talks
to gdbstub.
Linux: A version of this should be available in
$ESP_HOME/xtensa-lx106-elf/bin/xtensa-lx106-elf-gdbWindows: At time of writing, UDK doesn’t provide a GDB application - Download and run the executable installer at SysProgs
Copy the
C:\SysGCC\esp8266\opt\xtensa-lx106-elf\bin\xtensa-lx106-elf-gdb.exeto a suitable location.
Mac: ?
Usage
Configure gdbstub by editing
gdbstub-cfg.has required. You can also configure the options by setting :USER_CFLAGSin your project’scomponent.mkfile. e.gUSER_CFLAGS=-DGDBSTUB_BREAK_ON_INIT=0.Optional: Add
gdb_do_break()statements to your application.Run
make clean, thenmake ENABLE_GDB=1 flashto build and flash the application with debugging enabledRun gdb, depending on your configuration immediately after resetting the board or after it has run into an exception. The easiest way to do it is to use the provided script:
make gdb.
To run manually, see the following variables which you can inspect using make list-config.
- GDB_CMDLINE
Command line used to run GDB.
- GDBSTUB_DIR
Location of the GDB stub component, and the
gdbcmdsfile.
- COM_SPEED_GDB
Same as
COM_SPEED_SERIAL, which is the value compiled into the gdbstub code.
- GDB_UART_SWAP
If you need to use alternate serial pins, enable this option
GDB_UART_SWAP=1
- GDB
Path to the GDB executable being used.
Useful GDB commands
c Continue execution
q Quit and detach
where Display current stopped location
bt Show stack backtrace
disass Disassemble, disass/m to mix with source code
print expr Display a variable or other value
print func() Call a function, display result, or call func() to
discard result
tui enable Provides a windowed interface within the console (only
seems to work in Linux)
x/32xw $sp Display contents of stack
info reg Display register values
info break Display details of currently set breakpoints
delete Delete all breakpoints
br Set a breakpoint at the given address or function name
hbr Set a hardware breakpoint
watch Set a hardware watchpoint to detect when the value of a
variable changes
These commands require GDBSTUB_ENABLE_HOSTIO to be enabled:
remote get targetfile hostfile Read a file from SPIFFS (on the
target)
remote put hostfile targetfile Write a file to SPIFFS
remote delete targetfile Delete a file from SPIFFS
Eclipse
Windows:
Ensure
Use external console for inferioris checked.In connection settings, specify COM port like with leading /, e.g.
/COM4
Problems connecting?
Switch to the debug perspective before connecting
Ensure serial baud rate matches your application
Remove or disable all breakpoints before attaching. Eclipse will attempt to set these on connection, and if any are invalid it will hang and timeout.
Check connectivity using command-line GDB
GDB System Calls
Applications may interact with GDB directly using system calls, for example reading input from the GDB command prompt. See the Live Debug sample for a demonstration.
Note that system calls are disabled in the default configuration, so set
GDBSTUB_ENABLE_SYSCALL =1 to use this feature with your
application.
Known Issues and Limitations
- Unable to set requested break/watch points
Cause: Due to hardware limitations, only one hardware breakpount and one hardware watchpoint are available
Solution: None (hardware limitation)
- System crashes if debugger is paused for too long
Cause: The WiFi hardware is designed to be serviced by software periodically. It has some buffers so it will behave OK when some data comes in while the processor is busy, but these buffers are not infinite. If the WiFi hardware receives lots of data while the debugger has stopped the CPU, it is bound to crash. This will happen mostly when working with UDP and/or ICMP; TCP-connections in general will not send much more data when the other side doesn’t send any ACKs.
Solution: In such situations avoid pausing the debugger for extended periods
- Software breakpoints/watchpoints (‘break’ and ‘watch’) don’t work on flash code
Cause: GDB handles these by replacing code with a debugging instruction, therefore the code must be in RAM.
Solution: Use hardware breakpoint (‘hbreak’) or use
GDB_IRAM_ATTRfor code which requires testing
- If hardware breakpoint is set, single-stepping won’t work unless code is in RAM.
Cause: GDB reverts to software breakpoints if no hardware breakpoints are available
Solution: Delete hardware breakpoint before single-stepping
- Crash occurs when setting breakpoint in HardwareTimer callback routine
Cause: By default, HardwareTimer uses Non-maskable Interrupts (NMI) which keep running when the debugger is paused
Solution: Use the timer in non-maskable mode, or enable
GDBSTUB_PAUSE_HARDWARE_TIMERoption
- If gdbstub isn’t initialised then UART2 won’t work, though initialisation will succeed
Cause: By design, uart callbacks can be registered for UART2 at any time, before or after initialisation
Solution: Not really an issue, just something to be aware of
- Error reported, “packet reply is too long”
Cause: Mismatch between GDB version and stub code
Solution: Set
GDBSTUB_GDB_PATCHED=1 or use an unpatched version of GDB
- Whilst GDB is attached, input cannot be passed to application
Cause: GDB buffers keystrokes and replays them only when the target is interrupted (e.g. via ctrl+C), rather than passing them via serial connection.
Solution: Application may use gdb_syscall interface to communicate with debugger. See
$(SMING_HOME)/system/gdb_syscall.h, and Live Debug sample.
- No apparent way to have second ‘console’ (windows terminology) separate from GDB interface
Cause: Unknown
Solution: Is this possible with remote targets?
- GDB (in Windows) doesn’t respond at all to Ctrl+C
Cause: Unknown
Solution: Press Ctrl+Break to ‘hard kill’ GDB. You’ll probably need to do the next step as well to get it back
- When GDB is running under windows, appears to hang when target reset or restarted
Cause: Unknown, may not happen on all devboards but presents with NodeMCU
- Solution
quit GDB
quitStart terminal
make terminalreset board
quit terminal
run GDB again
make gdb
- Debug messages don’t appear in Eclipse
Cause: Unknown
Solution: Use command-line GDB, or a better visual debugger
Configuration
Defines
-
ENABLE_EXCEPTION_DUMP
When enabled, an exception or crash dumps a stack trace to debug output Default is ON for debug builds, OFF for release builds
Note: Not dependent upon ENABLE_GDB
-
ENABLE_CRASH_DUMP
When enabled, an unexpected reset (i.e. system crash) dumps a stack trace to debug output Default is ON for debug builds, OFF for release builds
Note: Not dependent upon ENABLE_GDB
-
GDBSTUB_ENABLE_DEBUG
When defined, GDB communications are echoed to UART1 for testing GDB stub operation.
0: No debug output 1: Show decoded commands and responses 2: Show packet content 3: Show debug output for internal routines
-
GDBSTUB_GDB_PATCHED
Espressif provide a patched version of GDB which emits only those registered present in the lx106. Set to 0 if an unpatched version of GDB is used.
-
GDBSTUB_USE_OWN_STACK
Enable this to make the exception and debugging handlers switch to a private stack. This will use up 1K of RAM, but may be useful if you’re debugging stack or stack pointer corruption problems. It’s normally disabled because not many situations need it. If for some reason the GDB communication stops when you run into an error in your code, try enabling this.
-
GDBSTUB_STACK_SIZE
-
GDBSTUB_BREAK_ON_EXCEPTION
Enable this to cause the program to pause and wait for gdb to be connected when an exception is encountered.
-
GDBSTUB_BREAK_ON_RESTART
Enable this to cause the program to pause and wait for gdb to be connected when an unexpected system restart occurs.
-
GDBSTUB_CTRLC_BREAK
If this is defined, gdbstub will break the program when you press Ctrl-C in gdb. It does this by monitoring for the ‘x03’ character in the serial receive routine. Any preceding characters are passed through to the application via UART2. If your application uses the serial port for terminal (text) communications you should be OK, but binary transfers are likely to cause problems and this option should probably be disabled. Instead, use GDBSTUB_BREAK_ON_INIT, or call gdb_do_break() in your application.
Specify: 0 to disable Ctrl+C break checking completely 1 to allow Ctrl+C break only when debugger is attached 2 to allow Ctrl+C break at any time. Ensure you have set remote interrupt-on-connect on in GDB command file, so it will send a Ctrl+C sequence when attempting to connect
-
GDBSTUB_ENABLE_UART2
The GDB stub has exclusive access to UART0, so applications cannot use it directly and attempts to open it will fail.
If this option is enabled, the default serial port will be changed to UART2 to allow debug output (from m_printf, debug_*, os_printf, etc.) to show up in your GDB session.
Outside of GDB terminal applications should work as normal, with the following caveats:
See GDB remote serial protocol for further details.If GDBSTUB_BREAK_ON_INIT is defined, then at startup your application will display `$T05#b9` and stop. A similar thing will happen if GDBSTUB_CTRLC_BREAK=2 and you type Ctrl+C. Continue by typing `$D#44` (without the quotes), or exit the terminal and start GDB.Disabling this option releases some IRAM. You may be instead able to use UART1 for debug output, adding
Serial.setPort(UART_ID_1);in your application’sinit()function.
-
GDBSTUB_ENABLE_SYSCALL
Enable gdb_syscall_* functions for use by application. If undefined, calls will do nothing and return -1.
-
GDBSTUB_ENABLE_HOSTIO
Enable Host I/O capability, where files may be accessed via GDB command prompt using
remote get,remote putandremote deletecommands.
-
GDBSTUB_BREAK_ON_INIT
Enable this if you want the GDB stub to wait for you to attach GDB before running. It does this by breaking in the init routine; use the gdb ‘c’ command (continue) to start the program.
-
GDBSTUB_CMDENABLE_P
Some commands are not required by GDB, so if necessary can be disabled to save memory.
-
GDBSTUB_CMDENABLE_X
-
GDBSTUB_UART_READ_TIMEOUT
Specify a timeout (in milliseconds) when stub is reading from serial port. Set to 0 to wait indefinitely.
-
GDBSTUB_FORCE_IRAM
Wherever possible gdbstub code is placed in flash memory. This is fine for most cases, but if debugging whilst flash is disabled or busy (eg during SPI operations or flash write/erase) then you will need to enable this option to move stub code into IRAM.
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Sming (Rp2040) Component
Environment Variables
SoC support
esp8266
Heap
Configuration variables
- ENABLE_CUSTOM_HEAP
If your application is experiencing heap fragmentation then you can try the alternative heap.
- undefined (default)
OFF, use standard heap
- 1
Use UMM Malloc.
Warning
Do not enable custom heap allocation and -mforce-l32 compiler flag at the same time.
- UMM_FUNC_IRAM
Default: 1 (enabled)
Custom heap functions are stored in IRAM by default for performance reasons.
If you need the IRAM (about 1.5K bytes) then disable this option:
make ENABLE_CUSTOM_HEAP=1 UMM_FUNC_IRAM=0
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Environment Variables
UMM_POISON_CHECK
SoC support
esp8266
Submodule: umm_malloc
umm_malloc - Memory Manager For Small(ish) Microprocessors
This is a memory management library specifically designed to work with the ARM7 embedded processor, but it should work on many other 32 bit processors, as well as 16 and 8 bit devices.
You can even use it on a bigger project where a single process might want to manage a large number of smaller objects, and using the system heap might get expensive.
Acknowledgements
Joerg Wunsch and the avr-libc provided the first malloc() implementation
that I examined in detail.
http://www.nongnu.org/avr-libc
Doug Lea’s paper on malloc() was another excellent reference and provides a lot of detail on advanced memory management techniques such as binning.
http://g.oswego.edu/dl/html/malloc.html
Bill Dittman provided excellent suggestions, including macros to support
using these functions in critical sections, and for optimizing realloc()
further by checking to see if the previous block was free and could be
used for the new block size. This can help to reduce heap fragmentation
significantly.
Yaniv Ankin suggested that a way to dump the current heap condition might be useful. I combined this with an idea from plarroy to also allow checking a free pointer to make sure it’s valid.
Dimitry Frank contributed many helpful additions to make things more robust including a user specified config file and a method of testing the integrity of the data structures.
Usage
Copy the umm_malloc_cfg_example.h file to umm_malloc_cfg.h and
make the changes required to support your application.
The following #defines must be set to something useful for the
library to work at all
UMM_MALLOC_CFG_HEAP_ADDRmust be set to the symbol representing the starting address of the heap. The heap must be aligned on the natural boundary size of the processor.UMM_MALLOC_CFG_HEAP_SIZEmust be set to the size of the heap. The heap size must be a multiple of the natural boundary size of the processor.
The fit algorithm is defined as either:
UMM_BEST_FITwhich scans the entire free list and looksfor either an exact fit or the smallest block that will satisfy the request. This is the default fit method.
UMM_FIRST_FITwhich scans the entire free list and looksfor the first block that satisfies the request.
The following #defines are disabled by default and should
remain disabled for production use. They are helpful when
testing allocation errors (which are normally due to bugs in
the application code) or for running the test suite when
making changes to the code.
You can define them in your compiler command line or uncomment
the corresponding entries is umm_malloc_cfg.h:
UMM_INFOis used to include code that allows dumping the entire heap structure (helpful when there’s a problem).UMM_INTEGRITY_CHECKis used to include code that performs an integrity check on the heap structure. It’s up to you to call theumm_integrity_check()function.UMM_POISON_CHECKis used to include code that adds some bytes around the memory being allocated that are filled with known data. If the data is not intact when the block is checked, then somone has written outside of the memory block they have been allocated. It is up to you to call theumm_poison_check()function.
API
The following functions are available for your application:
void *umm_malloc( size_t size );void *umm_calloc( size_t num, size_t size );void *umm_realloc( void *ptr, size_t size );void umm_free( void *ptr );
They have exactly the same semantics as the corresponding standard library functions.
Background
The memory manager assumes the following things:
The standard POSIX compliant malloc/calloc/realloc/free semantics are used
All memory used by the manager is allocated at link time, it is aligned on a 32 bit boundary, it is contiguous, and its extent (start and end address) is filled in by the linker.
All memory used by the manager is initialized to 0 as part of the runtime startup routine. No other initialization is required.
The fastest linked list implementations use doubly linked lists so that its possible to insert and delete blocks in constant time. This memory manager keeps track of both free and used blocks in a doubly linked list.
Most memory managers use a list structure made up of pointers to keep track of used - and sometimes free - blocks of memory. In an embedded system, this can get pretty expensive as each pointer can use up to 32 bits.
In most embedded systems there is no need for managing a large quantity of memory block dynamically, so a full 32 bit pointer based data structure for the free and used block lists is wasteful. A block of memory on the free list would use 16 bytes just for the pointers!
This memory management library sees the heap as an array of blocks, and uses block numbers to keep track of locations. The block numbers are 15 bits - which allows for up to 32767 blocks of memory. The high order bit marks a block as being either free or in use, which will be explained later.
The result is that a block of memory on the free list uses just 8 bytes instead of 16.
In fact, we go even one step futher when we realize that the free block index values are available to store data when the block is allocated.
The overhead of an allocated block is therefore just 4 bytes.
Each memory block holds 8 bytes, and there are up to 32767 blocks available, for about 256K of heap space. If that’s not enough, you can always add more data bytes to the body of the memory block at the expense of free block size overhead.
There are a lot of little features and optimizations in this memory management system that makes it especially suited to small systems, and the best way to appreciate them is to review the data structures and algorithms used, so let’s get started.
Detailed Description
We have a general notation for a block that we’ll use to describe the different scenarios that our memory allocation algorithm must deal with:
+----+----+----+----+
c |* n | p | nf | pf |
+----+----+----+----+
Where:
c is the index of this block
is the indicator for a free block
n is the index of the next block in the heap
p is the index of the previous block in the heap
nf is the index of the next block in the free list
pf is the index of the previous block in the free list
The fact that we have forward and backward links in the block descriptors means that malloc() and free() operations can be very fast. It’s easy to either allocate the whole free item to a new block or to allocate part of the free item and leave the rest on the free list without traversing the list from front to back first.
The entire block of memory used by the heap is assumed to be initialized to 0. The very first block in the heap is special - it’t the head of the free block list. It is never assimilated with a free block (more on this later).
Once a block has been allocated to the application, it looks like this:
+----+----+----+----+
c | n | p | ... |
+----+----+----+----+
Where:
c is the index of this block
n is the index of the next block in the heap
p is the index of the previous block in the heap
Note that the free list information is gone because it’s now being used to store actual data for the application. If we had even 500 items in use, that would be 2,000 bytes for free list information. We simply can’t afford to waste that much.
The address of the ... area is what is returned to the application
for data storage.
The following sections describe the scenarios encountered during the operation of the library. There are two additional notation conventions:
?? inside a pointer block means that the data is irrelevant. We don’t care
about it because we don’t read or modify it in the scenario being
described.
... between memory blocks indicates zero or more additional blocks are
allocated for use by the upper block.
While we’re talking about “upper” and “lower” blocks, we should make a comment about adresses. In the diagrams, a block higher up in the picture is at a lower address. And the blocks grow downwards their block index increases as does their physical address.
Finally, there’s one very important characteristic of the individual blocks that make up the heap - there can never be two consecutive free memory blocks, but there can be consecutive used memory blocks.
The reason is that we always want to have a short free list of the largest possible block sizes. By always assimilating a newly freed block with adjacent free blocks, we maximize the size of each free memory area.
Operation of malloc right after system startup
As part of the system startup code, all of the heap has been cleared.
During the very first malloc operation, we start traversing the free list starting at index 0. The index of the next free block is 0, which means we’re at the end of the list!
At this point, the malloc has a special test that checks if the current block index is 0, which it is. This special case initializes the free list to point at block index 1 and then points block 1 to the last block (lf) on the heap.
BEFORE AFTER
+----+----+----+----+ +----+----+----+----+
0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 1 |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+
1 |*lf | 0 | 0 | 0 |
+----+----+----+----+
...
+----+----+----+----+
lf | 0 | 1 | 0 | 0 |
+----+----+----+----+
The heap is now ready to complete the first malloc operation.
Operation of malloc when we have reached the end of the free list and there is no block large enough to accommodate the request.
This happens at the very first malloc operation, or any time the free list is traversed and no free block large enough for the request is found.
The current block pointer will be at the end of the free list, and we know we’re at the end of the list because the nf index is 0, like this:
BEFORE AFTER
+----+----+----+----+ +----+----+----+----+
pf |*?? | ?? | cf | ?? | pf |*?? | ?? | lf | ?? |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
p | cf | ?? | ... | p | cf | ?? | ... |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ +----+----+----+----+
cf | 0 | p | 0 | pf | c | lf | p | ... |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+
lf | 0 | cf | 0 | pf |
+----+----+----+----+
As we walk the free list looking for a block of size b or larger, we get to cf, which is the last item in the free list. We know this because the next index is 0.
So we’re going to turn cf into the new block of memory, and then create a new block that represents the last free entry (lf) and adjust the prev index of lf to point at the block we just created. We also need to adjust the next index of the new block (c) to point to the last free block.
Note that the next free index of the pf block must point to the new lf because cf is no longer a free block!
Operation of malloc when we have found a block (cf) that will fit the current request of b units exactly
This one is pretty easy, just clear the free list bit in the current block and unhook it from the free list.
BEFORE AFTER
+----+----+----+----+ +----+----+----+----+
pf |*?? | ?? | cf | ?? | pf |*?? | ?? | nf | ?? |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
p | cf | ?? | ... | p | cf | ?? | ... |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ +----+----+----+----+ Clear the free
cf |* n | p | nf | pf | cf | n | p | .. | list bit here
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ +----+----+----+----+
n | ?? | cf | ... | n | ?? | cf | ... |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
nf |*?? | ?? | ?? | cf | nf | ?? | ?? | ?? | pf |
+----+----+----+----+ +----+----+----+----+
Unhooking from the free list is accomplished by adjusting the next and prev free list index values in the pf and nf blocks.
Operation of malloc when we have found a block that will fit the current request of b units with some left over
We’ll allocate the new block at the END of the current free block so we don’t have to change ANY free list pointers.
BEFORE AFTER
+----+----+----+----+ +----+----+----+----+
pf |*?? | ?? | cf | ?? | pf |*?? | ?? | cf | ?? |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
p | cf | ?? | ... | p | cf | ?? | ... |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ +----+----+----+----+
cf |* n | p | nf | pf | cf |* c | p | nf | pf |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ This is the new
c | n | cf | .. | block at cf+b
+----+----+----+----+
+----+----+----+----+ +----+----+----+----+
n | ?? | cf | ... | n | ?? | c | ... |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
nf |*?? | ?? | ?? | cf | nf | ?? | ?? | ?? | pf |
+----+----+----+----+ +----+----+----+----+
This one is prety easy too, except we don’t need to mess with the free list indexes at all becasue we’ll allocate the new block at the end of the current free block. We do, however have to adjust the indexes in cf, c, and n.
That covers the initialization and all possible malloc scenarios, so now we need to cover the free operation possibilities…
Free Scenarios
The operation of free depends on the position of the current block being freed relative to free list items immediately above or below it. The code works like this:
if next block is free
assimilate with next block already on free list
if prev block is free
assimilate with prev block already on free list
else
put current block at head of free list
Step 1 of the free operation checks if the next block is free, and if it is assimilate the next block with this one.
Note that c is the block we are freeing up, cf is the free block that follows it.
BEFORE AFTER
+----+----+----+----+ +----+----+----+----+
pf |*?? | ?? | cf | ?? | pf |*?? | ?? | nf | ?? |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
p | c | ?? | ... | p | c | ?? | ... |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ +----+----+----+----+ This block is
c | cf | p | ... | c | nn | p | ... | disconnected
+----+----+----+----+ +----+----+----+----+ from free list,
+----+----+----+----+ assimilated with
cf |*nn | c | nf | pf | the next, and
+----+----+----+----+ ready for step 2
+----+----+----+----+ +----+----+----+----+
nn | ?? | cf | ?? | ?? | nn | ?? | c | ... |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
nf |*?? | ?? | ?? | cf | nf |*?? | ?? | ?? | pf |
+----+----+----+----+ +----+----+----+----+
Take special note that the newly assimilated block (c) is completely disconnected from the free list, and it does not have its free list bit set. This is important as we move on to step 2 of the procedure…
Step 2 of the free operation checks if the prev block is free, and if it is then assimilate it with this block.
Note that c is the block we are freeing up, pf is the free block that precedes it.
BEFORE AFTER
+----+----+----+----+ +----+----+----+----+ This block has
pf |* c | ?? | nf | ?? | pf |* n | ?? | nf | ?? | assimilated the
+----+----+----+----+ +----+----+----+----+ current block
+----+----+----+----+
c | n | pf | ... |
+----+----+----+----+
+----+----+----+----+ +----+----+----+----+
n | ?? | c | ... | n | ?? | pf | ?? | ?? |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
nf |*?? | ?? | ?? | pf | nf |*?? | ?? | ?? | pf |
+----+----+----+----+ +----+----+----+----+
Nothing magic here, except that when we’re done, the current block (c) is gone since it’s been absorbed into the previous free block. Note that the previous step guarantees that the next block (n) is not free.
Step 3 of the free operation only runs if the previous block is not free. it just inserts the current block to the head of the free list.
Remember, 0 is always the first block in the memory heap, and it’s always head of the free list!
BEFORE AFTER
+----+----+----+----+ +----+----+----+----+
0 | ?? | ?? | nf | 0 | 0 | ?? | ?? | c | 0 |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
p | c | ?? | ... | p | c | ?? | ... |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ +----+----+----+----+
c | n | p | .. | c |* n | p | nf | 0 |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ +----+----+----+----+
n | ?? | c | ... | n | ?? | c | ... |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
nf |*?? | ?? | ?? | 0 | nf |*?? | ?? | ?? | c |
+----+----+----+----+ +----+----+----+----+
Again, nothing spectacular here, we’re simply adjusting a few pointers to make the most recently freed block the first item in the free list.
That’s because finding the previous free block would mean a reverse traversal of blocks until we found a free one, and it’s just easier to put it at the head of the list. No traversal is needed.
Realloc Scenarios
Finally, we can cover realloc, which has the following basic operation.
The first thing we do is assimilate up with the next free block of memory if possible. This step might help if we’re resizing to a bigger block of memory. It also helps if we’re downsizing and creating a new free block with the leftover memory.
First we check to see if the next block is free, and we assimilate it to this block if it is. If the previous block is also free, and if combining it with the current block would satisfy the request, then we assimilate with that block and move the current data down to the new location.
Assimilating with the previous free block and moving the data works like this:
BEFORE AFTER
+----+----+----+----+ +----+----+----+----+
pf |*?? | ?? | cf | ?? | pf |*?? | ?? | nf | ?? |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
cf |* c | ?? | nf | pf | c | n | ?? | ... | The data gets
+----+----+----+----+ +----+----+----+----+ moved from c to
+----+----+----+----+ the new data area
c | n | cf | ... | in cf, then c is
+----+----+----+----+ adjusted to cf
+----+----+----+----+ +----+----+----+----+
n | ?? | c | ... | n | ?? | c | ?? | ?? |
+----+----+----+----+ +----+----+----+----+
... ...
+----+----+----+----+ +----+----+----+----+
nf |*?? | ?? | ?? | cf | nf |*?? | ?? | ?? | pf |
+----+----+----+----+ +----+----+----+----+
Once we’re done that, there are three scenarios to consider:
The current block size is exactly the right size, so no more work is needed.
The current block is bigger than the new required size, so carve off the excess and add it to the free list.
The current block is still smaller than the required size, so malloc a new block of the correct size and copy the current data into the new block before freeing the current block.
The only one of these scenarios that involves an operation that has not yet been described is the second one, and it’s shown below:
BEFORE AFTER
+----+----+----+----+ +----+----+----+----+
p | c | ?? | ... | p | c | ?? | ... |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ +----+----+----+----+
c | n | p | ... | c | s | p | ... |
+----+----+----+----+ +----+----+----+----+
+----+----+----+----+ This is the
s | n | c | .. | new block at
+----+----+----+----+ c+blocks
+----+----+----+----+ +----+----+----+----+
n | ?? | c | ... | n | ?? | s | ... |
+----+----+----+----+ +----+----+----+----+
Then we call free() with the adress of the data portion of the new block (s) which adds it to the free list.
Esp8266 LIBC Component
This Component accommodates the differences in runtime libraries for the various supported toolchains.
See also ESP Quick Toolchain.
References
Used by
Esp32 Core Component Component
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Sming (Rp2040) Component
SoC support
esp8266
Esp8266 LWIP Version 2
This Component implements the current Version 2 LWIP stack. Note that at present espconn_* functions are not supported.
References
Environment Variables
SoC support
esp8266
Submodule: lwip2
lwIP-2 from original source
This repo offers a link layer for esp82xx-nonos-sdk-2. The original goal is to try and use a recent lwIP version for stability reasons. Currently lwIP-v2 is implemented, other IP stacks could be tried.
Notes
Tested
arduino NTPClient
arduino WiFiAccessPoint
arduino OTA
Sming Telnet sample
Build
makefiles are working with linux/osx, and maybe with windows (using ‘make’ included in msys from mingw…)
get lwIP sources
git submodule update --init --recursive
optionnally tune lwIP configuration in glue-lwip/lwipopts.h
build & install
make -f Makefile.<arch> install
MSS
Remember the MSS footprint: 4*MSS bytes in RAM per tcp connection. The lowest recommanded value is 536 which is the default here.
How it works
Espressif binary libraries rely on their lwip implementation. The idea, as described in this comment is to wrap espressif calls, and rewrite them for a new tcp implementation.
Example with lwip1.4’s ethernet_input() called by espressif binary blobs finally reaching lwip2’s:
-- LWIP2-----------------------------------
#define ethernet_input ethernet_input_LWIP2
- lwip2's ethernet_input_LWIP2_is called
(/ \)
| |
-- glue (new side)-----------^-v-----------
| |
glue_ethernet_input | |
- maps structures glue->new |
- calls ethernet_input_LWIP2(^ v)
- maps structures new->glue |
| |
-- glue (old side)-----------^-v-----------
| |
ethernet_input(): | |
- maps structures old->glue |
- calls glue_ethernet_input (^ v)
- maps structures glue->old |
| |
- espressif blobs -----------^-v-----------
XXXXXXXXXXXXXXXXXXXXXXXXXXXX | | XXXXXXXXXX
wifi calls ethernet_input(/ \) XXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
-------------------------------------------
History
Original cleaning and port of espressif’s patch set from lwIP-v1.4 to lwIP-v2 with references to lwIP-git-sha1
Discussion on how further work could done
First version of this implementation
Second version for arduino only
Esp8266 SPI Flash Support
Provides functions for access to flash memory.
API Documentation
-
enum SPIFlashMode
Values:
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_FAST_READ
-
enumerator MODE_SLOW_READ
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enum SPIFlashSpeed
Values:
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enum SPIFlashSize
Values:
-
enumerator SIZE_1MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_4MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
Not listed.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_1MBIT
-
enum SPIFlashMode
Values:
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_FAST_READ
-
enumerator MODE_SLOW_READ
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enum SPIFlashSpeed
Values:
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enum SPIFlashSize
Values:
-
enumerator SIZE_1MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_4MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
Not listed.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_1MBIT
-
enum SPIFlashMode
Values:
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_FAST_READ
-
enumerator MODE_SLOW_READ
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enum SPIFlashSpeed
Values:
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enum SPIFlashSize
Values:
-
enumerator SIZE_1MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_4MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
Not listed.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_1MBIT
-
static inline uint32_t flashmem_get_address(const void *memptr)
Obtain the flash memory address for a memory pointer.
Note
If memptr is not in valid flash memory it will return an offset which exceeds the internal flash memory size.
Note
The flash location is dependent on where rBoot has mapped the firmware.
- Parameters:
memptr –
- Return values:
uint32_t – Offset from start of flash memory
-
uint32_t flashmem_write(const void *from, uint32_t toaddr, uint32_t size)
Write a block of data to flash.
Note
None of the parameters need to be aligned
- Parameters:
from – Buffer to obtain data from
toaddr – Flash location to start writing
size – Number of bytes to write
- Return values:
uint32_t – Number of bytes written
-
uint32_t flashmem_read(void *to, uint32_t fromaddr, uint32_t size)
Read a block of data from flash.
Note
none of the parameters need to be aligned
- Parameters:
to – Buffer to store data
fromaddr – Flash location to start reading
size – Number of bytes to read
- Return values:
uint32_t – Number of bytes written
-
bool flashmem_erase_sector(uint32_t sector_id)
Erase a single flash sector.
- Parameters:
sector_id – the sector to erase
- Return values:
true – on success
-
SPIFlashInfo flashmem_get_info()
Get flash memory information block.
- Return values:
SPIFlashInfo – Information block
-
uint8_t flashmem_get_size_type()
Returns a number indicating the size of flash memory chip.
- Return values:
uint8_t – See SpiFlashInfo.size field for possible values
-
uint32_t flashmem_get_size_bytes()
get the total flash memory size
- Return values:
uint32_t – Size in bytes
-
uint16_t flashmem_get_size_sectors()
Get the total number of flash sectors.
- Return values:
uint16_t – Sector count
-
uint32_t flashmem_find_sector(uint32_t address, uint32_t *pstart, uint32_t *pend)
Helper function: find the flash sector in which an address resides.
Note
Optional parameters may be null
- Parameters:
address –
pstart – OUT/OPTIONAL: Start of sector containing the given address
pend – OUT/OPTIONAL: Last address in sector
- Return values:
uint32_t – Sector number for the given address
-
uint32_t flashmem_get_sector_of_address(uint32_t addr)
Get sector number containing the given address.
- Parameters:
addr –
- Return values:
uint32_t – sector number
-
uint32_t spi_flash_get_id(void)
-
uint32_t flashmem_write_internal(const void *from, uint32_t toaddr, uint32_t size)
write to flash memory
Note
All parameters MUST be aligned to 4-byte word boundaries, including the RAM pointer
- Parameters:
from – Buffer to read data from - MUST be word-aligned
toaddr – Flash address (offset) to write to - MUST be word-aligned
size – Number of bytes to write - MUST be word-aligned
- Return values:
uint32_t – Number of bytes actually written
-
uint32_t flashmem_read_internal(void *to, uint32_t fromaddr, uint32_t size)
Read from flash memory.
Note
All parameters MUST be aligned to 4-byte word boundaries, including the RAM pointer
- Parameters:
to – Buffer to store data - MUST be word-aligned
fromaddr – Flash address (offset) to read from - MUST be word-aligned
size – Number of bytes to read - MUST be word-aligned
- Return values:
uint32_t – Number of bytes actually read
-
uint32_t flashmem_get_first_free_block_address()
-
void flashmem_sfdp_read(uint32_t addr, void *buffer, size_t count)
-
INTERNAL_FLASH_WRITE_UNIT_SIZE
Flash memory access must be aligned and in multiples of 4-byte words.
-
INTERNAL_FLASH_READ_UNIT_SIZE
-
FLASH_TOTAL_SEC_COUNT
-
SYS_PARAM_SEC_COUNT
Number of flash sectors reserved for system parameters at start.
-
FLASH_WORK_SEC_COUNT
-
INTERNAL_FLASH_SECTOR_SIZE
-
INTERNAL_FLASH_SIZE
-
INTERNAL_FLASH_WRITE_UNIT_SIZE
Flash memory access must be aligned and in multiples of 4-byte words.
-
INTERNAL_FLASH_READ_UNIT_SIZE
-
FLASH_TOTAL_SEC_COUNT
-
SYS_PARAM_SEC_COUNT
Number of flash sectors reserved for system parameters at start.
-
FLASH_WORK_SEC_COUNT
-
INTERNAL_FLASH_SECTOR_SIZE
-
INTERNAL_FLASH_SIZE
-
INTERNAL_FLASH_START_ADDRESS
-
FLASH_TOTAL_SEC_COUNT
-
SYS_PARAM_SEC_COUNT
Number of flash sectors reserved for system parameters at start.
-
FLASH_WORK_SEC_COUNT
-
INTERNAL_FLASH_SECTOR_SIZE
-
INTERNAL_FLASH_SIZE
-
INTERNAL_FLASH_START_ADDRESS
-
struct SPIFlashInfo
- #include <esp_spi_flash.h>
SPI Flash memory information block. Copied from bootloader header. See
esp_image_header_t.SPI Flash memory information block. Stored at the beginning of flash memory.
-
struct STORE_TYPEDEF_ATTR
- #include <esp_spi_flash.h>
SPI Flash memory information block. Stored at the beginning of flash memory.
IRAM pre-caching
Introduction
This is taken from the esp8266 arduino core. For details see:
See Flash memory for some background.
Code is always executed from IRAM, however it must be read in from flash memory on first use which can cause issues within timing-critical code.
This can be avoided by placing the code in IRAM. However, IRAM is a very limited resource and should generally be reserved for interrupt service routines.
An alternative solution is to arrange for critical code to be pre-loaded into IRAM before it starts execution.
Note
This cannot be used for interrupt service routines.
By their nature, interrupts are essentially random events and therefore code must be available in IRAM at any time.
Usage
The steps required are:
Mark the function containing code using
IRAM_PRECACHE_ATTR()Place a call to
IRAM_PRECACHE_START()before the first line of critical codePlace a call to
IRAM_PRECACHE_END()after the last line of critical code
You must always declare a tag to avoid the risk of section name conflicts.
You can find an example of how precaching is used here:
https://github.com/esp8266/Arduino/blob/master/cores/esp8266/core_esp8266_spi_utils.cpp
API Documentation
-
void iram_precache(void *addr, uint32_t bytes)
Pre-load flash data into the flash instruction cache.
Note
All pages containing the requested region will be read to pull them into cache RAM.
- Parameters:
addr – First location to cache, specify NULL to use current location.
bytes – Number of bytes to cache
-
IRAM_PRECACHE_ATTR
-
IRAM_PRECACHE_START(tag)
-
IRAM_PRECACHE_END(tag)
-
IRAM_PRECACHE_ATTR
Tools for pre-loading code into the flash cache.
It can be useful for code that accesses/uses SPI0 which is connected to the flash chip.
Non interrupt handler code that is infrequently called but might otherwise require being in valuable IRAM - such as bit-banging I/O code or some code run at bootup can avoid being permanently in IRAM.
Mark functions containing critical code using this attribute
-
IRAM_PRECACHE_START(tag)
Place this macro before the first line of the critical code.
Note
Do not omit the tag, and be careful with naming to avoid conflicts
- Parameters:
tag – Used to create the precached section name, must be globally unique
-
IRAM_PRECACHE_END(tag)
Place this macro after the last line of critical code.
- Parameters:
tag – Must be the same tag used in IRAM_PRECACHE_START()
-
IRAM_PRECACHE_ATTR
-
IRAM_PRECACHE_START(tag)
-
IRAM_PRECACHE_END(tag)
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Host Library Component
Sming (Host) Component
Sming (Rp2040) Component
SoC support
esp8266
Sming Host Emulator
Summary
Allows Sming applications to be built as an executable to run on the development host (Windows or Linux).
This is a source-level emulator for developing and testing new framework code prior to flashing a real device.
This is not a machine emulator; if you need something operating at a lower level take a look at QEMU.
Requirements
CMake is required to build LWIP
Ensure you are using relatively recent compilers, with 32-bit libraries available.
For Linux, you may require the gcc-multilib and g++-multilib
packages to build 32-bit executables on a 64-bit OS.
For Windows, make sure your MinGW distro is up to date.
See Windows Installation for further details.
Building
Build the framework and application as usual, specifying SMING_ARCH =Host. For example:
cd $SMING_HOME/../samples/Basic_Serial
make SMING_ARCH=Host
This builds the application as an executable in, for example,
out/Host/firmware/app.exe. Various command-line options are
supported, use --help for details.
The easiest way to run the emulator is via make run. Variables are
used to pass the appropriate options to the command line.
To find out what options are in force, use make list-config.
Configuration
- CLI_TARGET_OPTIONS
Use this to add any custom options to the emulator command line. e.g.:
make run CLI_TARGET_OPTIONS=–help make run CLI_TARGET_OPTIONS=”–debug=0 –cpulimit=2”
Note: These settings are not ‘sticky’
Components
Sming (Host)
This Component builds a library containing architecture-specific code, and defines dependencies for Sming to build for the Host Emulator.
References
Used by
Sming (main) Component
SoC support
host
Serialib
Serialib is a simple C++ library for serial communication. The library has been designed to work under Linux and Windows.
More details on [Lulu’s blog](https://lucidar.me/en/serialib/cross-plateform-rs232-serial-library/)
References
Used by
Host Drivers Component
SoC support
host
Submodule: seriallib
Serialib
Serialib is a simple C++ library for serial communication. The library has been designed to work under Linux and Windows.
More details on Lulu’s blog
Host Drivers
Provides low-level peripheral support classes.
Host ADC
Stub functions are provided which insert an appropriate ‘sampling’ delay and return 0.
Host GPIO
The emulator does not provide any low-level GPIO support, this is handled at a higher level
using DigitalHooks.
API Documentation
-
class DigitalHooks
Class to customise behaviour for digital functions.
Note
By default, actions get output to console but this can get very busy. The easiest way to change the behaviour is by sub-classing DigitalHooks and passing the new class instance to
setDigitalHooks().Public Functions
-
virtual void badPin(const char *function, uint16_t pin)
Report invalid pin number.
- Parameters:
function – Name of function which was called
pin – The pin number
-
virtual bool pinMode(uint16_t pin, uint8_t mode)
Set pin mode.
- Parameters:
pin – Has already been range checked
mode –
- Return values:
true – if mode can be set for this pin, will be stored
-
virtual void digitalWrite(uint16_t pin, uint8_t val)
Change pin output.
- Parameters:
pin – Has already been range checked
val – New pin value
-
virtual uint8_t digitalRead(uint16_t pin, uint8_t mode)
Read pin state.
- Parameters:
pin – Has already been range checked
mode – The currently set mode for this pin
val – State for pin
-
virtual void pullup(uint16_t pin, bool enable)
Set or clear pullup state for a pin.
- Parameters:
pin – Has already been range checked
enable – true for pullup, false for no pullup
-
virtual unsigned long pulseIn(uint16_t pin, uint8_t state, unsigned long timeout)
Measure duration of pulse on GPIO.
- Parameters:
pin – GPIO to measure
state – State of pulse to measure [HIGH | LOW]
timeout – Maximum duration of pulse
- Return values:
unsigned – long Pulse duration in microseconds
-
virtual uint16_t analogRead(uint16_t pin)
Sample analogue input.
- Parameters:
pin – GPIO to read
- Return values:
uint16_t – Sample value
-
virtual void badPin(const char *function, uint16_t pin)
Host hw_timer driver
Hardware timers are emulated using a background thread. The O/S scheduler timeslice granularity is quite coarse and for time-critical threads is typically 1ms. Therefore, when a timer becomes close to expiry the thread sits (‘spins’) in a loop to get better resolution. For the rest of the time the thread is suspended or in a wait state.
Host PWM
A header is provided to allow code to compile but the emulator does not provide any further PWM support.
Host UART driver
Introduction
Implements a UART driver to connect via TCP socket, allowing terminal emulation using telnet, or directly to local host serial device (e.g. /dev/ttyUSB0, COM4, etc.)
If not otherwise reassigned, UART0 output is sent to the console and keyboard input is written to the UART0 receive queue.
Build variables
- ENABLE_HOST_UARTID
To enable emulation for a UART, set this value to the numbers required. You would normally add this to a project’s component.mk file.
For example:
ENABLE_HOST_UARTID = 0 1
If setting it on the command line, remember to use quotes:
make ENABLE_HOST_UARTID="0 1"
See Basic Serial which uses both serial ports like this.
- HOST_UART_PORTBASE
The base port number to use for telnet. Default is 10000, which corresponds to UART0.
This is passed to the command line
--uartportoption.
- HOST_UART_OPTIONS
By default, this value combines the above options.
Allows full customisation of UART command-line options for
make run.
TCP port emulation
Set required ports for emulation using the ENABLE_HOST_UARTID, then execute make run.
Note
As an alternative to make run, you can run the compiled application manually like this:
out/Host/debug/firmware/app --pause --uart=0 --uart=1
Now start a telnet session for each serial port, in separate command windows:
telnet localhost 10000
telnet localhost 10001
In the application window, press Enter. This behaviour is enabled by the
pause option, which stops the emulator after initialisation so
telnet can connect to it. Without pause you’ll lose any serial
output at startup.)
Note
For Windows users, putty is a good alternative to telnet. It also
has options for things like carriage-return/linefeed translation
(“\n” -> “\r\n`”). Run using:
putty telnet://localhost:10000
Port numbers are allocated sequentially from 10000. If you want to use
different port numbers, set HOST_UART_PORTBASE.
Physical device connection
Override HOST_UART_OPTIONS adding the –device option. For example:
make run HOST_UART_OPTIONS="--uart=0 --device=/dev/ttyUSB0"
The --device option must follow the --uart option. Another example:
make run HOST_UART_OPTIONS="--uart=0 --device=/dev/ttyUSB0 --uart=1 --device=/dev/ttyUSB1"
The port is opened when uart_init() gets called.
The default baud rate is whatever the application has requested. This can be overridden as follows:
make run HOST_UART_OPTIONS="--uart=0 --device=/dev/ttyUSB0 --baud=921600"
For Windows, substitute the appropriate device name, e.g. COM4 instead of /dev/ttyUSB0.
Note
If necessary, add ENABLE_HOST_UARTID= to prevent telnet windows from being created.
Console I/O may be assigned to a different port like this:
make run HOST_UART_OPTIONS="--uart=1 --device=console"
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Host Library Component
Sming (Host) Component
Sming (Rp2040) Component
Environment Variables
SoC support
host
Host ESP HAL
Provides implementations for various low-level functions similar to the Esp8266.
Task queues
Timer queues
System functions
References
Used by
Sming (Host) Component
SoC support
host
Host WiFi
Provides WiFi / network functions. The actual implementations are provided in LWIP
References
Used by
Sming (Esp8266) Component
Host Library Component
Networking Support Component
Environment Variables
HOST_NETWORK_OPTIONS
SoC support
host
GDB Stub for Host
This defines the command line to use when make gdb is run. No additional code is required to debug for the Host.
If you want to debug your application while having a separate UART then make sure to send the following commands to your debugger:
handle SIGUSR1 nostop noprint
This component provides also gdbinit file containing the optimal settings needed for debugging.
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Sming (Rp2040) Component
Environment Variables
SoC support
host
Heap
This Component implements heap-related housekeeping functions. Heap usage is tracked using malloc_count. This also provides some validation (using sentinels to detect if memory blocks are overwritten).
- ENABLE_MALLOC_COUNT
We require malloc_count to keep track of heap usage for system_get_free_heap_size(). It does this by hooking the memory allocation routines (malloc, free, etc.). If you wish to disable this behaviour, set ENABLE_MALLOC_COUNT=0. If using tools such as Valgrind, this will provide a cleaner trace.
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Environment Variables
SoC support
host
Host Library
This Components provides the core functionality for the Host Emulator:
Sockets
Classes to provide simple Berkeley socket support for both Linux and Windows
Options
Command line argument parsing. Applications may pass parameters and access them using
the CommandLine.
Startup
Initialises SPI Flash, Uart server (in Host Drivers) and LWIP
networking, then enters the main task loop. This loop services LWIP plus the task and timer queues
(implemented in Host ESP HAL).
The Ctrl+C keypress is trapped to provide an orderly exit. If the system has become stuck in a loop or is otherwise
unresponsive, subsequent Ctrl+C presses will force a process termination.
Threads and Interrupts
The emulator uses Posix threads (pthread library) to perform background processing which would probably be done in hardware or which is outside of the framework.
Ideally we’d use SCHED_FIFO to disable time-slicing and more closely resemble how interrupts work on a single-core CPU. However, this mode isn’t supported in Windows, and Linux requires privileged access and can potentially crash the system. Best avoided, I think.
All ESP code runs at a specific interrupt level, where 0 represents regular code. Interrupts are triggered from a separate thread (a CThread instance) which calls :cpp:func:` When an interrupt occurs, the level is raised according to the priority of that interrupt. Until that code has finished, only interrupts of a higher priority will preempt it.
Thread ‘interrupt’ code is sandwiched between calls to interrupt_begin() and interrupt_end(), which blocks interrupts from other threads at the same or lower level. The threads aren’t suspended but will block if they call interrupt_begin(). However, the main thread (level 0) is halted to reflect normal interrupt behaviour.
- HOST_PARAMETERS
Set this value to pass additional parameters to a Host application. For example:
make run HOST_PARAMETERS='param1=12 param2="parameter with spaces"' make run HOST_PARAMETERS="param1=12 param2=\"parameter with spaces\""
See Basic Utility for a worked example.
API
-
class CommandLine
Provides access to the command line.
Anything which doesn’t start with
-is interpreted as an application parameter. For example:Parameters which start withapp param1=value
-are handled by the Host Emulator. Anything after--is passed directly to the application:app -- --debug --verbose
Public Functions
-
inline const Parameters &getParameters()
Fetch a reference to the list of command-line parameters.
-
inline const Parameters &getParameters()
References
Used by
Host ESP HAL Component
Sming (Host) Component
Environment Variables
SoC support
host
libc
Contains implementations of any non-standard C library functions.
References
Used by
Esp32 Core Component Component
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Sming (Rp2040) Component
SoC support
host
SPI Flash
This Component emulates an embedded flash memory device using a backing file. It includes additional checks on addresses, sizes and alignments to detect common issues which can be more difficult to find when running on target hardware.
The default backing file is called flash.bin, located in the same directory as the host executable.
See Virtual Flasher for configuration details.
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Host Library Component
Sming (Host) Component
Sming (Rp2040) Component
SoC support
host
Virtual Flasher
Flash memory access is emulated using SPI Flash.
This Component implements make targets to operate on the flash backing file in a similar way to flashing a real device.
Build Variables
The following options are interpreted and used to provide command-line parameters to the emulator executable:
- FLASH_BIN
Full path to the flash backing file
- HOST_FLASH_OPTIONS
This defaults to a combination of the above variables, but you can override if necessary.
The size of the flash memory is set via Hardware configuration.
See Esptool for details and other applicable variables.
Build targets
make flashinitto clear and reset the file.make flashfsto copy the generated SPIFFS image into the backing file.make flashwrites out all required images to the backing file. For now, this is the same asmake flashfsbut that will change when support is added for custom user images.
References
Used by
Sming (Host) Component
Environment Variables
SoC support
host
Sming RP2040 Architecture
Support building Sming for the Raspberry Pi RP2040 SOC.
Testing so far has been limited to the Rasperry Pi Pico, but there are lots of other boards available.
Configure this using the PICO_BOARD setting. The default is pico.
You can find the full list here.
Special mention to the arduino-pico project https://github.com/earlephilhower/arduino-pico. Lots of helpful stuff in there!
Note
This architecture should be considered experimental at present.
Tested and working:
CPU frequency adjustment
system_get_cpu_freq(),system_update_cpu_freq()Timers working: hardware, software and CPU cycle counter
Hardware serial ports (UART driver)
Task queue (also supports queuing tasks from code running on core #1)
Flash memory routines
os_random()andos_get_random()implemented using ring oscillator. This is the best the hardware is capable of, but not crypto grade.Heap is standard newlib implementation,
system_get_free_heap_size()provided.Software watchdog implemented, timeout is 8 seconds
A disassembly and symbol file are generated for this architecture.
SDK declares flash memory size in the board header files. This is checked at compile time against the value declared in the partition table.
Reset information
system_get_rst_info()indicates watchdog or manual resets. Exception information not yet implemented.System functions
system_get_chip_id(),system_get_sdk_version().Partitions and file systems (except SD cards and FAT)
SPIClass tested with Radio_nRF24L01 sample only
WiFi networking support for the Pico-W
Standard analogue I/O via analogRead. More advanced hardware capabilities require use of the SDK directly.
Dual-core support. See below for details.
Yet to be implemented:
- USB
Best to write a separate
Sming-USBlibrary (based on TinyUSB) to support the RP2040, ESP32-S2 & ESP32-S3 variants. Needs some thought about good integration into the framework. Arduino-Pico overridesHardwareSerialto support serial devices, we can do something similar.- HardwareSPI
To support DMA, etc.
- PWM
Hardware can drive up to 16 outputs and measure input frequency/duty cycle.
- I2C
Has hardware support
- RTC
Can wake from deep sleep but requires an external clock (e.g. 32kHz crystal) and appropriate API. (Setting and reading the time is implemented.)
- Low-power modes
Deep sleep / suspend / power-saving
- PIO (Programmable I/O)
A killer feature for the RP2040. Uses range from simple glue logic to I2S, etc.
- Crash/exception handling & serial debugging
RP2040 supports JTAG debugging but requires extra hardware. Serial debugging is often enough and easier to set up. Requires GDB stub plus implementing crash handler callbacks, etc.
- Multi-boot / OTA updates.
If you run
make mapyou’ll see there is no bootloader! It’s part of the firmware image at present. Adding RP2040 support to rBoot may work, however the Pico typically has only 2MByte flash which is quite restrictive. It is also necessary to compile images at different addresses as there is no windowed XIP (eXecute In Place) capability. See Flash In-Place library for a basic method of OTA.
Requirements
These requirements are in addition to the standard Sming setup.
The easiest way to get started is with the Sming installer - see Getting Started.
Note: Windows is not currently included in the chocolatey repository. The following instructions should help.
- Compiler/linker
The RP2040 contains two ARM Cortex-M0+ cores. Tools for all platforms can be downloaded from the ARM developer website.
Unzip the archive to a suitable location (e.g.
/opt/rp2040orc:\tools\rp2040) and setPICO_TOOLCHAIN_PATHaccordingly.Note
At time of writing the Ubuntu repositories contain an older version of this toolchain. It also does not contain GDB, but can be installed separately:
sudo apt install gcc-arm-none-eabi gdb-multiarch
To use gdb-multiarch you’ll need to do this:
make gdb GDB=gdb-multiarch
- Ninja
This is used to build the RP2040 SDK code:
sudo apt install ninja-build
It is available for other platforms at https://ninja-build.org/ and consists of a single executable file. The application should either be in the system path, or set
NINJAto the full path of the executable file.If you have Sming working with the ESP32 then you’ll already have it installed.
Setup and programming
Serial support requires a 3.3v USB adapter connected to the appropriate pins:
UART0: TX = 0, RX = 1
UART1: TX = 4, RX = 5
To program your device, unplug the USB cable (i.e. remove power from the device)
then hold down the BOOTSEL button whilst plugging it back in again.
You can then run:
make flash
as usual and the device will be programmed.
Once Sming is running on the device, reprogramming is simpler and only requires pressing
the BOOTSEL button (no power cycle).
If the firmware has crashed or stalled the watchdog timer should reboot the system after 8 seconds, at which point BOOTSEL should be detected. So just hold the button down until this happens.
If all else fails, go through the initial power-cycle process.
This behaviour can be disabled using the ENABLE_BOOTSEL setting.
Boot process
Unlike the Espressif parts, the RP2040 is not programmed via the serial port, but written to the device when configured as a Mass Storage device (removable flash drive).
Data to be flashed must be in UF2 format and sent as a single file. See UF2 Support.
Once the file has finished sending the RP2040 reboots itself into normal operating mode (assuming BOOTSEL has been released).
The RP2040 can also be programmed via JTAG debugging but this requires additional hardware and setup.
Note
The RP2040 bootloader does not include support for reading flash memory via mass storage,
so commands such as make verifyflash won’t work at present.
Dual-core support
Sming is a strictly non-threaded framework, and all code runs on core #0. The SDK multicore API may still be used to run code on core #1, but this requires some care to ensure smooth operation.
The task queue (System::queueTask(), etc.) may be used to send messages to Sming from Core #1 code.
Passing messages the other way, from Sming code to core #1, could be done using a separate SDK task queue.
Flash access
Core 1 code may run directly from flash memory (via XIP) without any special considerations. However, during flash erase/write operations (e.g. file writes) XIP is disabled. If core 1 code attempts to access flash during these periods the system will hard fault.
Note
Floating-point support requires use of routines in flash memory. Integer operations should all be safe to use.
If unexplained crashes are occurring then check the build output files (in out/Rp2040/debug/build) or use a debugger to identify any errant code running from flash.
A typical use for core #1 might be to perform processing of some kind, such as processing data sampled via analogue inputs. If all code is run from RAM then it can continue uninterrupted even during filing system operations.
Alternatively some kind of synchronisation mechanism may be used to ensure that core 1 is suspended or running from RAM during any flash erase/write operations.
Networking
The Pico-W variant includes an Infineon CYW43439 bluetooth/WiFi SoC.
Raspberry Pi use the … driver. The SDK also includes an LWIP implementation.
The physical interface is SPI using a custom (PIO) implementation. This requires the use of GPIOxx which can no longer be accessed directly, but instead via xxxxx.
The CYW43 chip is initialised (via cyw43_ensure_up) when application code makes the first call into the networking API, for example by enabling station or AP access. Part of the hardware configuration here is to download firmware to the CYW43 chip (about 240KB) plus the CLM BLOB (< 1KB).
Sming contains patches which compresses this data (based on https://github.com/raspberrypi/pico-sdk/issues/909) to about 145KB. By default, it is linked into the application image, but can also be read from a separate partition.
Source code
The RP2040 is a very capable SOC, albeit without WiFi. A massive advantage is that the platform is fully open-source. Even the bootrom is published!
Here’s a summary of the various Github repositories the Raspberry Pi Foundation have made available:
- https://github.com/raspberrypi/pico-sdk
The core SDK for the RP2040 SOC. Sming includes this as a submodule.
- https://github.com/raspberrypi/picotool
This is a tool for inspecting RP2040 binaries, and interacting with RP2040 devices when they are in BOOTSEL mode. It does this by talking to a custom USB device implemented in the RP2040 bootrom.
Getting this to build is a bit fiddly. So far I’ve managed without it, but there is a
picotoolcomponent in the framework which can be used to try building it.- https://github.com/raspberrypi/pico-bootrom
Contents of the RP2040 boot rom. Very handy to be able to see this.
- https://github.com/raspberrypi/pico-examples
Examples using the pico SDK directly.
- https://github.com/raspberrypi/picoprobe
An RP2040 board can be used as a low-cost JTAG adapter using this firmware. Takes some setting up to use though. See [Getting Started PDF](https://datasheets.raspberrypi.org/pico/getting-started-with-pico.pdf) for details.
- https://github.com/raspberrypi/pico-extras
Some additional libraries which may or may not end up in the SDK.
- https://github.com/raspberrypi/pico-playground
Further examples using the pico-extras libraries.
Build variables
- PICO_TOOLCHAIN_PATH
This contains the base directory for the toolchain used to build the framework. Pre-compiled toolchains can be downloaded from the ARM Developer website.
Components
Sming (Rp2040)
This Component builds a library containing architecture-specific code, and defines dependencies for Sming to build for Raspberry Pi 2040 microcontroller-based boards.
References
Used by
Sming (main) Component
Environment Variables
TARGET_BIN
SoC support
rp2040
RP2040 Drivers
Provides low-level peripheral drivers.
GPIO: General-Purpose I/O
SDK definitions for GPIO.
-
enum GPIO_INT_TYPE
Values:
-
enumerator GPIO_PIN_INTR_DISABLE
Interrupt disabled for this pin
-
enumerator GPIO_PIN_INTR_POSEDGE
Interrupt occurs on positive edge
-
enumerator GPIO_PIN_INTR_NEGEDGE
Interrupt occurs on negative edge
-
enumerator GPIO_PIN_INTR_ANYEDGE
Interrupt occurs on both positive and negative edge
-
enumerator GPIO_PIN_INTR_LOLEVEL
Interrupt occurs when GPIO low
-
enumerator GPIO_PIN_INTR_HILEVEL
Interrupt occurs when GPIO high
-
enumerator GPIO_PIN_INTR_DISABLE
-
enumerator GPIO_PIN_INTR_POSEDGE
-
enumerator GPIO_PIN_INTR_NEGEDGE
-
enumerator GPIO_PIN_INTR_ANYEDGE
-
enumerator GPIO_PIN_INTR_LOLEVEL
-
enumerator GPIO_PIN_INTR_HILEVEL
-
enumerator GPIO_PIN_INTR_DISABLE
hw_timer: Hardware Timers
Driver for hardware timers.
API Documentation
-
enum hw_timer_clkdiv_t
Values:
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enum hw_timer_intr_type_t
Values:
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enum hw_timer_source_type_t
Values:
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enum hw_timer_clkdiv_t
Values:
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enum hw_timer_intr_type_t
Values:
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enum hw_timer_source_type_t
Values:
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enum hw_timer_clkdiv_t
Values:
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enumerator TIMER_CLKDIV_16
-
enumerator TIMER_CLKDIV_256
-
enumerator TIMER_CLKDIV_1
-
enum hw_timer_intr_type_t
Values:
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enumerator TIMER_LEVEL_INT
-
enumerator TIMER_EDGE_INT
-
enum hw_timer_source_type_t
Values:
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
enumerator TIMER_NMI_SOURCE
-
enumerator TIMER_FRC1_SOURCE
-
typedef void (*hw_timer_callback_t)(void *arg)
-
typedef void (*hw_timer_callback_t)(void *arg)
-
typedef void (*hw_timer_callback_t)(void *arg)
-
struct hw_timer_private_t hw_timer_private
-
void hw_timer1_attach_interrupt(hw_timer_source_type_t source_type, hw_timer_callback_t callback, void *arg)
Attach an interrupt for the timer.
- Parameters:
source_type – Ignored, uses APB clock source
callback – Callback function invoked via timer interrupt
arg – Passed to callback function
source_type –
callback – Callback function invoked via timer interrupt
arg – Passed to callback function
-
inline void hw_timer1_enable(hw_timer_clkdiv_t div, hw_timer_intr_type_t intr_type, bool auto_load)
Enable the timer.
Note: With one-shot timer application callback must stop the timer when it is no longer required. This is to reduce callback latency. If this is not done, timer will trigger again when timer counter wraps around to 0. For /16 divisor this is only 1.7s.
- Parameters:
div –
intr_type – Ignored, always level-triggered
auto_load –
div –
intr_type –
auto_load – true for repeating timer, false for one-shot
div –
intr_type –
auto_load –
-
void hw_timer1_write(uint32_t ticks)
Set the timer interval.
Set the timer interval and arm.
- Parameters:
ticks –
-
void hw_timer1_disable(void)
Disable the timer.
-
void hw_timer1_detach_interrupt(void)
Detach interrupt from the timer.
-
uint32_t hw_timer1_read(void)
Get timer1 count.
- Return values:
uint32_t – Current count value, counts from initial value down to 0
-
static uint32_t hw_timer2_read(void)
Read current timer2 value.
- Return values:
uint32_t –
-
void hw_timer_init(void)
Initialise hardware timers.
Note
Called by startup code
-
void hw_timer2_set_alarm(uint32_t ticks)
Set timer2 alarm count value.
Note
For internal use ONLY; used by software timers
- Parameters:
ticks –
-
uint32_t hw_timer_ticks()
Fetch 32-bit microsecond count.
All timers reference from a single 64-bit counter. We use only the lower 32 bits here as it provides lowest latency and compatibility with existing API.
-
HW_TIMER1_GROUP
-
HW_TIMER1_INDEX
-
MAX_HW_TIMER1_INTERVAL
Maximum timer interval in ticks.
Note
The corresponding time interval depends on the prescaler in use:
ESP32 supports a wide range of prescalers and uses 54-bit counter value. Limiting the range 31 bits avoids issues with overflows and moving to 64-bit calculations./1 - 26.84s /16 - 429.50s /256 - 6871.95s
-
MIN_HW_TIMER1_INTERVAL_US
Minimum hardware interval in microseconds.
Note
Attempting to use repetitive interrupts below this level can lead to system instabliity and lockups, due to the software overhead in servicing the interrupts.
-
NOW()
-
MAX_HW_TIMER1_INTERVAL
Maximum timer interval in ticks.
Note
The corresponding time interval depends on the prescaler in use:
/1 - 0.1048s /16 - 1.677s /256 - 26.84s
-
MIN_HW_TIMER1_INTERVAL_US
Minimum hardware interval in microseconds.
Note
Attempting to use repetitive interrupts below this level can lead to system instabliity and lockups, due to the software overhead in servicing the interrupts.
-
HW_TIMER2_CLKDIV
-
HW_TIMER2_CLK
-
MAX_HW_TIMER1_INTERVAL
Maximum timer interval in ticks.
-
MIN_HW_TIMER1_INTERVAL_US
Minimum hardware interval in microseconds.
Note
Attempting to use repetitive interrupts below this level can lead to system instabliity and lockups, due to the software overhead in servicing the interrupts.
-
HW_TIMER2_CLK
-
struct hw_timer_private_t
- #include <hw_timer.h>
OS Timer
Driver for software timer queues
-
using smg_timer_func_t = void (*)(void *arg)
-
void smg_timer_arm_ticks(os_timer_t *ptimer, uint32_t ticks, bool repeat_flag)
Set a software timer using the Timer2 tick value.
This function has been added to Sming for more efficient and flexible use of software timers. It can be used alongside the SDK
os_timer_arm_new()function.- Parameters:
ptimer – Timer structure
ticks – Tick count duration for the timer
repeat_flag – true if timer will automatically repeat
-
void smg_timer_setfn(os_timer_t *ptimer, os_timer_func_t pfunction, void *parg)
-
void smg_timer_arm_us(os_timer_t *ptimer, uint32_t time_us, bool repeat_flag)
-
void smg_timer_arm(os_timer_t *ptimer, uint32_t time_ms, bool repeat_flag)
-
void smg_timer_disarm(os_timer_t *ptimer)
-
void smg_timer_done(os_timer_t *ptimer)
-
static inline uint64_t smg_timer_expire(const os_timer_t *ptimer)
-
void os_timer_arm_ticks(os_timer_t *ptimer, uint32_t ticks, bool repeat_flag)
Set a software timer using the Timer2 tick value.
This function has been added to Sming for more efficient and flexible use of software timers. It can be used alongside the SDK
os_timer_arm_new()function.- Parameters:
ptimer – Timer structure
ticks – Tick count duration for the timer
repeat_flag – true if timer will automatically repeat
-
static inline bool os_timer_is_armed(const os_timer_t *ptimer)
-
static inline uint64_t os_timer_expire(const os_timer_t *ptimer)
-
static inline void os_timer_done(os_timer_t *ptimer)
-
os_timer_func_t
-
os_timer_t
-
os_timer_arm
-
os_timer_arm_us
-
os_timer_disarm
-
os_timer_setfn
-
os_timer_arm_ticks
-
os_timer_expire
-
os_timer_done
-
struct smg_timer_t
- #include <os_timer.h>
PWM: Pulse-Width Modulation
API Documentation
-
void pwm_init(uint32_t period, uint32_t *duty, uint32_t pwm_channel_num, uint32_t (*pin_info_list)[3])
Initialize PWM function, including GPIO selection, period and duty cycle.
Example:
uint32 ioInfo[][3] = { {PWM_0_OUT_IO_MUX, PWM_0_OUT_IO_FUNC, PWM_0_OUT_IO_NUM}, {PWM_1_OUT_IO_MUX, PWM_1_OUT_IO_FUNC, PWM_1_OUT_IO_NUM}, {PWM_2_OUT_IO_MUX, PWM_2_OUT_IO_FUNC, PWM_2_OUT_IO_NUM} }; pwm_init(light_param.pwm_period, light_param.pwm_duty, 3, ioInfo);
Example:
uint32 io_info[][3] = { {PWM_0_OUT_IO_MUX, PWM_0_OUT_IO_FUNC, PWM_0_OUT_IO_NUM}, {PWM_1_OUT_IO_MUX, PWM_1_OUT_IO_FUNC, PWM_1_OUT_IO_NUM}, {PWM_2_OUT_IO_MUX, PWM_2_OUT_IO_FUNC, PWM_2_OUT_IO_NUM} }; pwm_init(light_param.pwm_period, light_param.pwm_duty, 3, io_info);
Note
This API can be called only once.
Note
This API can be called only once.
- Parameters:
period – PWM period
duty – duty cycle of each output
pwm_channel_num – PWM channel number
pin_info_list – Array containing an entry for each channel giving
period – PWM period
duty – duty cycle of each output
pwm_channel_num – PWM channel number
pin_info_list – Array containing an entry for each channel giving
-
void pwm_start(void)
Starts PWM.
This function needs to be called after PWM configuration is changed.
-
void pwm_set_duty(uint32_t duty, uint8_t channel)
Sets duty cycle of a PWM output.
Set the time that high-level signal will last. The range of duty depends on PWM period. Its maximum value of which can be Period * 1000 / 45.
For example, for 1-KHz PWM, the duty range is 0 ~ 22222.
- Parameters:
duty – The time that high-level single will last, duty cycle will be (duty*45)/(period*1000)
channel – PWM channel, which depends on how many PWM channels are used
-
uint32_t pwm_get_duty(uint8_t channel)
Get duty cycle of PWM output.
Duty cycle will be (duty*45) / (period*1000).
- Parameters:
channel – PWM channel, which depends on how many PWM channels are used
- Return values:
uint32 – Duty cycle of PWM output
-
void pwm_set_period(uint32_t period)
Set PWM period.
- Parameters:
period – PWM period in us. For example, 1-KHz PWM period = 1000us.
-
uint32_t pwm_get_period(void)
Get PWM period.
- Return values:
uint32 – Return PWM period in us.
-
uint32_t get_pwm_version(void)
Get version information of PWM.
- Return values:
uint32 – PWM version
UART: Universal Asynchronous Receive/Transmit
Custom asynchronous driver.
Warning
doxygengroup: Cannot find group “uart_driver” in doxygen xml output for project “api” from directory: ../api/xml/
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Host Library Component
Sming (Host) Component
Sming (Rp2040) Component
Environment Variables
SoC support
rp2040
GDB Stub for RP2040
This defines the command line to use when make gdb is run.
TODO: Implement stub code to allow serial debugging.
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Sming (Rp2040) Component
Environment Variables
SoC support
rp2040
RP2040 LIBC Component
This Component accommodates the differences in runtime libraries for the various supported toolchains.
References
Used by
Esp32 Core Component Component
Sming (Esp32) Component
Sming (Esp8266) Component
Sming (Host) Component
Sming (Rp2040) Component
SoC support
rp2040
Picotool
Picotool is a tool for inspecting RP2040 binaries, and interacting with RP2040 devices when they are in BOOTSEL mode.
Note for full documentation see https://rptl.io/pico-get-started Appendix B.
References
Environment Variables
PICOTOOL
SoC support
rp2040
Submodule: picotool
Building
You need to set PICO_SDK_PATH in the environment, or pass it to cmake You need to install libusb-1.0.
Linux/Mac: use your favorite package tool
Windows: download from here https://libusb.info/
If you are on Windows, set LIBUSB_ROOT environment variable to the install directory
mkdir build
cd build
cmake ..
make
for Windows non MinGW/WSL:
mkdir build
cd build
cmake -G "NMake Makefiles" ..
nmake
Overview
Picotool is a tool for inspecting RP2040 binaries, and interacting with RP2040 devices when they are in BOOTSEL mode.
Note for full documentation see https://rptl.io/pico-get-started
$ picotool help
PICOTOOL:
Tool for interacting with a RP2040 device in BOOTSEL mode, or with a RP2040 binary
SYNOPSYS:
picotool info [-b] [-p] [-d] [-l] [-a] [--bus <bus>] [--address <addr>] [-f] [-F]
picotool info [-b] [-p] [-d] [-l] [-a] <filename> [-t <type>]
picotool load [-v] [-x] <filename> [-t <type>] [-o <offset>] [--bus <bus>] [--address <addr>] [-f] [-F]
picotool save [-p] [--bus <bus>] [--address <addr>] [-f] [-F] <filename> [-t <type>]
picotool save -a [--bus <bus>] [--address <addr>] [-f] [-F] <filename> [-t <type>]
picotool save -r <from> <to> [--bus <bus>] [--address <addr>] [-f] [-F] <filename> [-t <type>]
picotool verify [--bus <bus>] [--address <addr>] [-f] [-F] <filename> [-t <type>] [-r <from> <to>] [-o <offset>]
picotool reboot [-a] [-u] [--bus <bus>] [--address <addr>] [-f] [-F]
picotool version [-s]
picotool help [<cmd>]
COMMANDS:
info Display information from the target device(s) or file.
Without any arguments, this will display basic information for all connected RP2040 devices in BOOTSEL mode
load Load the program / memory range stored in a file onto the device.
save Save the program / memory stored in flash on the device to a file.
verify Check that the device contents match those in the file.
reboot Reboot the device
version Display picotool version
help Show general help or help for a specific command
Use "picotool help <cmd>" for more info
Note commands that aren’t acting on files require an RP2040 device in BOOTSEL mode to be connected.
info
So there is now Binary Information support in the SDK which allows for easily storing compact information that picotool
can find (See Binary Info section below). The info command is for reading this information.
The information can be either read from one or more connected RP2040 devices in BOOTSEL mode, or from a file. This file can be an ELF, a UF2 or a BIN file.
$ picotool help info
INFO:
Display information from the target device(s) or file.
Without any arguments, this will display basic information for all connected RP2040 devices in BOOTSEL mode
SYNOPSYS:
picotool info [-b] [-p] [-d] [-l] [-a] [--bus <bus>] [--address <addr>] [-f] [-F]
picotool info [-b] [-p] [-d] [-l] [-a] <filename> [-t <type>]
OPTIONS:
Information to display
-b, --basic
Include basic information. This is the default
-p, --pins
Include pin information
-d, --device
Include device information
-l, --build
Include build attributes
-a, --all
Include all information
TARGET SELECTION:
To target one or more connected RP2040 device(s) in BOOTSEL mode (the default)
--bus <bus>
Filter devices by USB bus number
--address <addr>
Filter devices by USB device address
To target a file
<filename>
The file name
-t <type>
Specify file type (uf2 | elf | bin) explicitly, ignoring file extension
e.g.
$ picotool info
Program Information
name: hello_world
features: stdout to UART
$ picotool info -a
Program Information
name: hello_world
features: stdout to UART
binary start: 0x10000000
binary end: 0x1000606c
Fixed Pin Information
20: UART1 TX
21: UART1 RX
Build Information
build date: Dec 31 2020
build attributes: Debug build
Device Information
flash size: 2048K
ROM version: 2
$ picotool info -bp
Program Information
name: hello_world
features: stdout to UART
Fixed Pin Information
20: UART1 TX
21: UART1 RX
$ picotool info -a lcd_1602_i2c.uf2
File lcd_1602_i2c.uf2:
Program Information
name: lcd_1602_i2c
web site: https://github.com/raspberrypi/pico-examples/tree/HEAD/i2c/lcd_1602_i2c
binary start: 0x10000000
binary end: 0x10003c1c
Fixed Pin Information
4: I2C0 SDA
5: I2C0 SCL
Build Information
build date: Dec 31 2020
save
Save allows you to save a range of memory or a program or the whole of flash from the device to a BIN file or a UF2 file
$ picotool help save
SAVE:
Save the program / memory stored in flash on the device to a file.
SYNOPSYS:
picotool save [-p] [--bus <bus>] [--address <addr>] <filename> [-t <type>]
picotool save -a [--bus <bus>] [--address <addr>] <filename> [-t <type>]
picotool save -r <from> <to> [--bus <bus>] [--address <addr>] <filename> [-t <type>]
OPTIONS:
Selection of data to save
-p, --program
Save the installed program only. This is the default
-a, --all
Save all of flash memory
-r, --range
Save a range of memory; note that the range is expanded to 256 byte boundaries
<from>
The lower address bound in hex
<to>
The upper address bound in hex
Source device selection
--bus <bus>
Filter devices by USB bus number
--address <addr>
Filter devices by USB device address
File to save to
<filename>
The file name
-t <type>
Specify file type (uf2 | elf | bin) explicitly, ignoring file extension
e.g.
$ picotool info
Program Information
name: lcd_1602_i2c
web site: https://github.com/raspberrypi/pico-examples/tree/HEAD/i2c/lcd_1602_i2c
$ picotool save spoon.uf2
Saving file: [==============================] 100%
Wrote 51200 bytes to spoon.uf2
$ picotool info spoon.uf2
File spoon.uf2:
Program Information
name: lcd_1602_i2c
web site: https://github.com/raspberrypi/pico-examples/tree/HEAD/i2c/lcd_1602_i2c
Binary Information
Binary information is machine locatable and generally machine consumable. I say generally because anyone can include any information, and we can tell it from ours, but it is up to them whether they make their data self describing.
Note that we will certainly add more binary info over time, but I’d like to get a minimum core set included in most binaries from launch!!
Basic Information
This information is really handy when you pick up a Pico and don’t know what is on it!
Basic information includes
program name
program description
program version string
program build date
program url
program end address
program features - this is a list built from individual strings in the binary, that can be displayed (e.g. we will have one for UART stdio and one for USB stdio) in the SDK
build attributes - this is a similar list of strings, for things pertaining to the binary itself (e.g. Debug Build)
The binary information is self-describing/extensible, so programs can include information picotool is not aware of (e.g. MicroPython includes a list of in-built libraries)
Pins
This is certainly handy when you have an executable called ‘hello_world.elf’ but you forgot what board it is built for…
Static (fixed) pin assignments can be recorded in the binary in very compact form:
$ picotool info --pins sprite_demo.elf
File sprite_demo.elf:
Fixed Pin Information
0-4: Red 0-4
6-10: Green 0-4
11-15: Blue 0-4
16: HSync
17: VSync
18: Display Enable
19: Pixel Clock
20: UART1 TX
21: UART1 RX
Including Binary information
Binary information is declared in the program by macros (vile warped macros); for the previous example:
$ picotool info --pins sprite_demo.elf
File sprite_demo.elf:
Fixed Pin Information
0-4: Red 0-4
6-10: Green 0-4
11-15: Blue 0-4
16: HSync
17: VSync
18: Display Enable
19: Pixel Clock
20: UART1 TX
21: UART1 RX
… there is one line in the setup_default_uart function:
bi_decl_if_func_used(bi_2pins_with_func(PICO_DEFAULT_UART_RX_PIN, PICO_DEFAULT_UART_TX_PIN, GPIO_FUNC_UART));
The two pin numbers, and the function UART are stored, then decoded to their actual function names (UART1 TX etc) by picotool.
The bi_decl_if_func_used makes sure the binary information is only included if the containing function is called.
Equally, the video code contains a few lines like this:
bi_decl_if_func_used(bi_pin_mask_with_name(0x1f << (PICO_SCANVIDEO_COLOR_PIN_BASE + PICO_SCANVIDEO_DPI_PIXEL_RSHIFT), "Red 0-4"));
Details
Things are designed to waste as little space as possible, but you can turn everything off with preprocessor var PICO_NO_BINARY_INFO=1. Additionally
any SDK code that inserts binary info can be separately excluded by its own preprocesor var.
You need
#include "pico/binary_info.h"
Basically you either use bi_decl(bi_blah(...)) for unconditional inclusion of the binary info blah, or
bi_decl_if_func_used(bi_blah(...)) for binary information that may be stripped if the enclosing function
is not included in the binary by the linker (think --gc-sections)
There are a bunch of bi_ macros in the headers
#define bi_binary_end(end) ...
#define bi_program_name(name) ...
#define bi_program_description(description) ...
#define bi_program_version_string(version_string) ...
#define bi_program_build_date_string(date_string) ...
#define bi_program_url(url) ...
#define bi_program_feature(feature) ...
#define bi_program_build_attribute(attr) ...
#define bi_1pin_with_func(p0, func) ...
#define bi_2pins_with_func(p0, p1, func) ...
#define bi_3pins_with_func(p0, p1, p2, func) ...
#define bi_4pins_with_func(p0, p1, p2, p3, func) ...
#define bi_5pins_with_func(p0, p1, p2, p3, p4, func) ...
#define bi_pin_range_with_func(plo, phi, func) ...
#define bi_pin_mask_with_name(pmask, label) ...
#define bi_pin_mask_with_names(pmask, label) ...
#define bi_1pin_with_name(p0, name) ...
#define bi_2pins_with_names(p0, name0, p1, name1) ...
#define bi_3pins_with_names(p0, name0, p1, name1, p2, name2) ...
#define bi_4pins_with_names(p0, name0, p1, name1, p2, name2, p3, name3) ...
which make use of underlying macros, e.g.
#define bi_program_url(url) bi_string(BINARY_INFO_TAG_RASPBERRY_PI, BINARY_INFO_ID_RP_PROGRAM_URL, url)
NOTE: It is easy to forget to enclose these in bi_decl etc., so an effort has been made (at the expense of a lot of kittens)
to make the build fail with a somewhat helpful error message if you do so.
For example, trying to compile
bi_1pin_with_name(0, "Toaster activator");
gives
/home/graham/dev/mu/pico_sdk/src/common/pico_binary_info/include/pico/binary_info/code.h:17:55: error: '_error_bi_is_missing_enclosing_decl_261' undeclared here (not in a function)
17 | #define __bi_enclosure_check_lineno_var_name __CONCAT(_error_bi_is_missing_enclosing_decl_,__LINE__)
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
... more macro call stack of doom
Setting common fields from CMake
You can use
pico_set_program_name(foo "not foo") # as "foo" would be the default
pico_set_program_description(foo "this is a foo")
pico_set_program_version_string(foo "0.00001a")
pico_set_program_url(foo "www.plinth.com/foo")
Note all of these are passed as command line arguments to the compilation, so if you plan to use quotes, newlines etc you may have better luck defining via bi_decl in the code.
Additional binary information/picotool features
Block devices
MicroPython and CircuitPython, eventually the SDK and others may support one or more storage devices in flash. We already have macros to define these although picotool doesn’t do anything with them yet… but backup/restore/file copy and even fuse mount in the future might be interesting.
I suggest we tag these now…
This is what I have right now off the top of my head (at the time)
#define bi_block_device(_tag, _name, _offset, _size, _extra, _flags)
with the data going into
typedef struct __packed _binary_info_block_device {
struct _binary_info_core core;
bi_ptr_of(const char) name; // optional static name (independent of what is formatted)
uint32_t offset;
uint32_t size;
bi_ptr_of(binary_info_t) extra; // additional info
uint16_t flags;
} binary_info_block_device_t;
and
enum {
BINARY_INFO_BLOCK_DEV_FLAG_READ = 1 << 0, // if not readable, then it is basically hidden, but tools may choose to avoid overwriting it
BINARY_INFO_BLOCK_DEV_FLAG_WRITE = 1 << 1,
BINARY_INFO_BLOCK_DEV_FLAG_REFORMAT = 1 << 2, // may be reformatted..
BINARY_INFO_BLOCK_DEV_FLAG_PT_UNKNOWN = 0 << 4, // unknown free to look
BINARY_INFO_BLOCK_DEV_FLAG_PT_MBR = 1 << 4, // expect MBR
BINARY_INFO_BLOCK_DEV_FLAG_PT_GPT = 2 << 4, // expect GPT
BINARY_INFO_BLOCK_DEV_FLAG_PT_NONE = 3 << 4, // no partition table
};
USB device descriptors
Seems like tagging these might be nice (we just need to store the pointer to it assuming - as is often the case - the descriptor is just a linear chunk of memory) … I assume there is a tool out there to prettify such a thing if picotool dumps the descriptor in binary.
Issues
If you ctrl+c out of the middle of a long operation, then libusb seems to get a bit confused, which means we aren’t able
to unlock our lockout of USB MSD writes (we have turned them off so the user doesn’t step on their own toes). Simply running
picotool info again will unlock it properly the next time (or you can reboot the device).
RP2040 Core Component
Contains startup code, crash handling and additional RP2040-specific support code.
Configuration variables
- PICO_BOARD
default: pico
Select development board in use. List available boards with
make list-boards.The SDK defines various useful bits of information in a board header file, such as the default LED pin, how much flash memory it has, etc. Use
make board-infoto list these values.If using custom hardware, select
noneand provide definitions as required.
- ENABLE_BOOTSEL
default: 1 for debug, 0 for release builds
This setting is provided to make it easy to re-program RP2040 boards during development. When enabled, Sming monitors the BOOTSEL button and restartS in boot mode if pressed.
- LINK_CYW43_FIRMWARE
default: 1
The Pico-W board requires a ~140K (compressed) firmware BLOB which by default is linked into the application image.
This can be wasteful when using OTA as the firmware must be contained in all application images.
Setting this value to ‘0’ will omit firmware from the image, and instead load it from a partition called ‘cyw43_fw’. This partition can be added to the standard map using the ‘cyw43_fw’
HWCONFIG_OPTsetting:make LINK_CYW43_FIRMWARE=0 HWCONFIG_OPT=cyw43_fw
This is not the default setting since the additional partition must be managed by the end application.
- PICO_DEBUG
default: 0
Set to 1 to enable additional debug output from compiled Pico SDK library. Shows things like WiFi firmware version.
References
Used by
RP2040 Drivers Component
Sming (Rp2040) Component
Rp2040 SPI Flash Support Component
Environment Variables
CYW43_FIRMWARECYW43_PIN_WL_HOST_WAKECYW43_PIN_WL_REG_ONCYW43_USES_VSYS_PINCYW43_WL_GPIO_COUNTCYW43_WL_GPIO_LED_PINCYW43_WL_GPIO_VBUS_PINPICO_BOOT_STAGE2_CHOOSE_W25Q080PICO_DEFAULT_I2CPICO_DEFAULT_I2C_SCL_PINPICO_DEFAULT_I2C_SDA_PINPICO_DEFAULT_SPIPICO_DEFAULT_SPI_CSN_PINPICO_DEFAULT_SPI_RX_PINPICO_DEFAULT_SPI_SCK_PINPICO_DEFAULT_SPI_TX_PINPICO_DEFAULT_UARTPICO_DEFAULT_UART_RX_PINPICO_DEFAULT_UART_TX_PINPICO_FLASH_SIZE_BYTESPICO_FLASH_SPI_CLKDIVPICO_RP2040_B0_SUPPORTEDPICO_RP2040_B1_SUPPORTEDPICO_VSYS_PINRASPBERRYPI_PICO_W
SoC support
rp2040
Submodule: pico-sdk
Raspberry Pi Pico SDK
The Raspberry Pi Pico SDK (henceforth the SDK) provides the headers, libraries and build system necessary to write programs for the RP2040-based devices such as the Raspberry Pi Pico 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 RP2040’s hardware include PIO (Programmable IO).
Additionally the SDK provides higher level libraries for dealing with timers, synchronization, USB (TinyUSB) and multi-core programming along with various utilities.
The SDK can be used to build anything from simple applications, to fully fledged runtime environments such as MicroPython, to low level software such as RP2040’s on-chip bootrom itself.
Additional libraries/APIs that are not yet ready for inclusion in the SDK can be found in pico-extras.
Documentation
See Getting Started with the Raspberry Pi Pico for information on how to setup your hardware, IDE/environment and for how to build and debug software for the Raspberry Pi Pico and other RP2040-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 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.
Example code
See pico-examples for example code you can build.
Getting the latest SDK code
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.
Quick-start your own project
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 C/C++ SDK
Install CMake (at least version 3.13), and GCC cross compiler
sudo apt install cmake 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 clonethis Raspberry Pi Pico SDK repositoryCopy pico_sdk_import.cmake from the SDK into your project directory
#. Set
PICO_SDK_PATHto the SDK location in your environment, or pass it (-DPICO_SDK_PATH=) to cmake later. #.Setup a
CMakeLists.txtlike:cmake_minimum_required(VERSION 3.13) # 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.txtlike:cmake_minimum_required(VERSION 3.13) # 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.txtlike: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 clonethis Raspberry Pi Pico SDK repository #.Setup a
CMakeLists.txtlike: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 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() { setup_default_uart(); 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:
$ mkdir build $ cd build $ cmake ..
When building for a board other than the Raspberry Pi Pico, you should pass
-DPICO_BOARD=board_nameto thecmakecommand above, e.g.cmake -DPICO_BOARD=pico_w ..to configure the SDK and build options accordingly for that particular board.Doing so 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 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.
$ make hello_world
You now have
hello_world.elfto load via a debugger, orhello_world.uf2that can be installed and run on your Raspberry Pi Pico via drag and drop.
Rp2040 SPI Flash Support
Provides functions for access to flash memory.
API Documentation
-
enum SPIFlashMode
Values:
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_FAST_READ
-
enumerator MODE_SLOW_READ
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enum SPIFlashSpeed
Values:
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enum SPIFlashSize
Values:
-
enumerator SIZE_1MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_4MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
Not listed.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_1MBIT
-
enum SPIFlashMode
Values:
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_FAST_READ
-
enumerator MODE_SLOW_READ
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enum SPIFlashSpeed
Values:
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enum SPIFlashSize
Values:
-
enumerator SIZE_1MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_4MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
Not listed.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_1MBIT
-
enum SPIFlashMode
Values:
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_FAST_READ
-
enumerator MODE_SLOW_READ
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enumerator MODE_QOUT
-
enumerator MODE_DIO
-
enumerator MODE_DOUT
-
enumerator MODE_SLOW_READ
Not supported.
-
enumerator MODE_FAST_READ
Not supported.
-
enumerator MODE_QIO
-
enum SPIFlashSpeed
Values:
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enumerator SPEED_26MHZ
-
enumerator SPEED_20MHZ
-
enumerator SPEED_80MHZ
-
enumerator SPEED_40MHZ
-
enum SPIFlashSize
Values:
-
enumerator SIZE_1MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_4MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
Not listed.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_4MBIT
-
enumerator SIZE_2MBIT
-
enumerator SIZE_8MBIT
-
enumerator SIZE_16MBIT
-
enumerator SIZE_32MBIT
-
enumerator SIZE_1MBIT
Not supported.
-
enumerator SIZE_1MBIT
-
static inline uint32_t flashmem_get_address(const void *memptr)
Obtain the flash memory address for a memory pointer.
Note
If memptr is not in valid flash memory it will return an offset which exceeds the internal flash memory size.
Note
The flash location is dependent on where rBoot has mapped the firmware.
- Parameters:
memptr –
- Return values:
uint32_t – Offset from start of flash memory
-
uint32_t flashmem_write(const void *from, uint32_t toaddr, uint32_t size)
Write a block of data to flash.
Note
None of the parameters need to be aligned
- Parameters:
from – Buffer to obtain data from
toaddr – Flash location to start writing
size – Number of bytes to write
- Return values:
uint32_t – Number of bytes written
-
uint32_t flashmem_read(void *to, uint32_t fromaddr, uint32_t size)
Read a block of data from flash.
Note
none of the parameters need to be aligned
- Parameters:
to – Buffer to store data
fromaddr – Flash location to start reading
size – Number of bytes to read
- Return values:
uint32_t – Number of bytes written
-
bool flashmem_erase_sector(uint32_t sector_id)
Erase a single flash sector.
- Parameters:
sector_id – the sector to erase
- Return values:
true – on success
-
SPIFlashInfo flashmem_get_info()
Get flash memory information block.
- Return values:
SPIFlashInfo – Information block
-
uint8_t flashmem_get_size_type()
Returns a number indicating the size of flash memory chip.
- Return values:
uint8_t – See SpiFlashInfo.size field for possible values
-
uint32_t flashmem_get_size_bytes()
get the total flash memory size
- Return values:
uint32_t – Size in bytes
-
uint16_t flashmem_get_size_sectors()
Get the total number of flash sectors.
- Return values:
uint16_t – Sector count
-
uint32_t flashmem_find_sector(uint32_t address, uint32_t *pstart, uint32_t *pend)
Helper function: find the flash sector in which an address resides.
Note
Optional parameters may be null
- Parameters:
address –
pstart – OUT/OPTIONAL: Start of sector containing the given address
pend – OUT/OPTIONAL: Last address in sector
- Return values:
uint32_t – Sector number for the given address
-
uint32_t flashmem_get_sector_of_address(uint32_t addr)
Get sector number containing the given address.
- Parameters:
addr –
- Return values:
uint32_t – sector number
-
uint32_t spi_flash_get_id(void)
-
uint32_t flashmem_write_internal(const void *from, uint32_t toaddr, uint32_t size)
write to flash memory
Note
All parameters MUST be aligned to 4-byte word boundaries, including the RAM pointer
- Parameters:
from – Buffer to read data from - MUST be word-aligned
toaddr – Flash address (offset) to write to - MUST be word-aligned
size – Number of bytes to write - MUST be word-aligned
- Return values:
uint32_t – Number of bytes actually written
-
uint32_t flashmem_read_internal(void *to, uint32_t fromaddr, uint32_t size)
Read from flash memory.
Note
All parameters MUST be aligned to 4-byte word boundaries, including the RAM pointer
- Parameters:
to – Buffer to store data - MUST be word-aligned
fromaddr – Flash address (offset) to read from - MUST be word-aligned
size – Number of bytes to read - MUST be word-aligned
- Return values:
uint32_t – Number of bytes actually read
-
uint32_t flashmem_get_first_free_block_address()
-
void flashmem_sfdp_read(uint32_t addr, void *buffer, size_t count)
-
INTERNAL_FLASH_WRITE_UNIT_SIZE
Flash memory access must be aligned and in multiples of 4-byte words.
-
INTERNAL_FLASH_READ_UNIT_SIZE
-
FLASH_TOTAL_SEC_COUNT
-
SYS_PARAM_SEC_COUNT
Number of flash sectors reserved for system parameters at start.
-
FLASH_WORK_SEC_COUNT
-
INTERNAL_FLASH_SECTOR_SIZE
-
INTERNAL_FLASH_SIZE
-
INTERNAL_FLASH_WRITE_UNIT_SIZE
Flash memory access must be aligned and in multiples of 4-byte words.
-
INTERNAL_FLASH_READ_UNIT_SIZE
-
FLASH_TOTAL_SEC_COUNT
-
SYS_PARAM_SEC_COUNT
Number of flash sectors reserved for system parameters at start.
-
FLASH_WORK_SEC_COUNT
-
INTERNAL_FLASH_SECTOR_SIZE
-
INTERNAL_FLASH_SIZE
-
INTERNAL_FLASH_START_ADDRESS
-
FLASH_TOTAL_SEC_COUNT
-
SYS_PARAM_SEC_COUNT
Number of flash sectors reserved for system parameters at start.
-
FLASH_WORK_SEC_COUNT
-
INTERNAL_FLASH_SECTOR_SIZE
-
INTERNAL_FLASH_SIZE
-
INTERNAL_FLASH_START_ADDRESS
-
struct SPIFlashInfo
- #include <esp_spi_flash.h>
SPI Flash memory information block. Copied from bootloader header. See
esp_image_header_t.SPI Flash memory information block. Stored at the beginning of flash memory.
Public Members
-
uint8_t mode
SPIFlashMode.
-
uint8_t speed
SPIFlashSpeed.
-
uint8_t size
SPIFlashSize.
-
uint8_t mode
-
struct STORE_TYPEDEF_ATTR
- #include <esp_spi_flash.h>
SPI Flash memory information block. Stored at the beginning of flash memory.
Public Members
-
uint8_t mode
SPIFlashMode.
-
uint8_t speed
SPIFlashSpeed.
-
uint8_t size
SPIFlashSize.
-
uint8_t mode
References
Used by
Sming (Esp32) Component
Sming (Esp8266) Component
Host Library Component
Sming (Host) Component
Sming (Rp2040) Component
SoC support
rp2040
UF2 Support
Provides support for converting binary code into UF2 format for uploading to RP2040 devices.
USB Flashing Format (UF2)
UF2 is a file format, developed by Microsoft for PXT (also known as Microsoft MakeCode), that is particularly suitable for flashing microcontrollers over MSC (Mass Storage Class; aka removable flash drive).
For a more friendly explanation, check out this blog post. Also, take a look at the list of implementations at the bottom of this document.
Overview
The UF2 file consists of 512 byte blocks, each of which is self-contained and independent of others. Each 512 byte block consists of (see below for details):
magic numbers at the beginning and at the end
address where the data should be flashed
up to 476 bytes of data
The data transfers over MSC always arrive in multiples of 512 bytes. Together with the FAT file system structure, this means that blocks of the UF2 file are always aligned with the MSC writes - the microcontroller never gets a partial file.
The magic numbers let the microcontroller distinguish an UF2 file block from other data (eg., FAT table entry, or various book-keeping files stored by some operating systems). When a UF2 block is recognized, it can be flashed immediately (unless flash page size is more than 256 bytes; in that case a buffer is needed). The actual handling of file format during writing is very simple (~10 lines of C code in simplest version).
File format
A UF2 file consists of 512 byte blocks. Each block starts with a 32 byte header, followed by data, and a final magic number. All fields, except for data, are 32 bit unsigned little endian integers.
Offset |
Size |
Value |
|---|---|---|
0 |
4 |
First magic number, |
4 |
4 |
Second magic number, |
8 |
4 |
Flags |
12 |
4 |
Address in flash where the data should be written |
16 |
4 |
Number of bytes used in data (often 256) |
20 |
4 |
Sequential block number; starts at 0 |
24 |
4 |
Total number of blocks in file |
28 |
4 |
File size or board family ID or zero |
32 |
476 |
Data, padded with zeros |
508 |
4 |
Final magic number, |
The following C struct can be used:
struct UF2_Block {
// 32 byte header
uint32_t magicStart0;
uint32_t magicStart1;
uint32_t flags;
uint32_t targetAddr;
uint32_t payloadSize;
uint32_t blockNo;
uint32_t numBlocks;
uint32_t fileSize; // or familyID;
uint8_t data[476];
uint32_t magicEnd;
} UF2_Block;
Flags
Currently, there are five flags defined:
0x00000001- not main flash - this block should be skipped when writing the device flash; it can be used to store “comments” in the file, typically embedded source code or debug info that does not fit on the device flash0x00001000- file container - see below0x00002000- familyID present - when set, thefileSize/familyIDholds a value identifying the board family (usually corresponds to an MCU)0x00004000- MD5 checksum present - see below0x00008000- extension tags present - see below
Family ID
This field is optional, and should be set only when the corresponding flag is set. It is recommended that new bootloaders require the field to be set appropriately, and refuse to flash UF2 files without it. If you’re developing your own bootloader, and your board family isn’t listed here, pick a new family ID at random. It’s good to also send a PR here, so your family can be listed.
If the familyID doesn’t match, the bootloader should disregard the
entire block, including blockNo and numBlocks fields.
In particular, writing a full UF2 file with non-matching familyID
should not reset the board.
This also allows for several files with different familyID to be
simply concatenated together, and the whole resulting file to be copied
to the device with only one actually being written to flash.
Picking numbers at random
The reason to pick numbers at random is to minimize risk of collisions
in the wild. Do not pick random numbers by banging on keyboard, or by using
0xdeadf00d, 0x42424242 etc. A good way is to use the following
shell command: printf "0x%04x%04x\n" $RANDOM $RANDOM
Another good way is the link at the bottom of https://microsoft.github.io/uf2/patcher/
This procedure was unfortunately not used for the SAMD51 and NRF52840 below.
Family list
The current master list of family IDs is maintained in a JSON file.
Rationale
The magic number at the end is meant to mitigate partial block writes.
Second and final magic numbers were randomly selected, except for the last byte
of final magic number, which was forced to be '\n' (0xA). Together with the
first magic number being "UF2\n" this makes it easy to identify UF2 blocks in
a text editor.
The header is padded to 32 bytes, as hex editors commonly use 16 or 32 bytes as line length. This way, the data payload is aligned to line start.
32 bit integers are used for all fields so that large flash sizes can be supported in future, as well as for simplicity. Little endian is used, as most microcontrollers are little endian. 8 bit microcontrollers can choose to just use the first 16 bits of various header fields.
The total number of blocks in the file and the sequential block number make it easy for the bootloader to detect that all blocks have been transferred. It requires one bit of memory per block (eg., on SAMD21G18A it’s 128 bytes). Alternatively, the bootloader might ignore that and just implement a reset after say 1 second break in incoming UF2 blocks.
Payload sizes
The number of data bytes is configurable and depends on the size of the flash page (that is the smallest size that can be erased) on the microcontroller.
if the page size is more than
476bytes, the bootloader should support any payload size, as it needs to buffer the entire page in memory anywayif the page size is less than
476bytes, the payload should be a multiple of page size, so it can be written without buffering; the target address should also be a multiple of page size
In any event, payload size and target address should always be 4-byte aligned.
Note that payload size of 256 is always correct, and makes it easy to convert
between flash addresses and UF2 file offsets.
For example, on Atmel’s SAMD21 chips the page size is 256 bytes, and this
also is the payload size. If the page size was 128 bytes, one could use
payload of 128*3. Nordic nRF51 has page size of 1024 bytes, and thus
any payload size should be allowed.
Embedding sources
Some IDEs will embed program sources in the UF2 file. This allows a UF2 files to be loaded by the IDE and serve as a natural backup and transfer format. This can be done in two ways:
using the “not main flash” flag
using normal blocks that are flashed to the device
If the bootloader can expose CURRENT.UF2 file (see below) and there is enough
flash available, than the second option is more desirable, as it allows sharing
programs directly from the board.
See https://makecode.com/source-embedding for more info.
Robustness
The file format is designed specifically to deal with the following problems:
operating system (OS) writing blocks in different order than occurs in a file
OS writing blocks multiple times
OS writing data that is not UF2 blocks
OS writing first/final part of a block, possibly for metadata detection or search indexing
The only file system assumption we make is that blocks of file are aligned with blocks on the hard drive. It’s likely true of many file systems besides FAT.
We also assume that USB MSC device reports its block size to be a multiple of 512
bytes. In the wild these devices always almost report exactly 512, and some
operating systems do not support other values.
Files exposed by bootloaders
Bootloaders may expose virtual files in their MSC devices. These are standardized here, so that flashing tools can automatically detect the bootloaders.
INFO_UF2.TXT- contains information about the bootloader build and the board on which it is runningINDEX.HTM- redirects to a page that contains an IDE or other informationCURRENT.UF2- the contents of the entire flash of the device, starting at0x00000000, with256payload size; thus, the size of this file will report as twice the size of flash
Flashing tools can use the presence of INFO_UF2.TXT (in upper or lower case,
as FAT is case-insensitive) file as an indication that a given directory is
actually a connected UF2 board. The other files should not be used for
detection.
Typical INFO_UF2.TXT file looks like this:
UF2 Bootloader v1.1.3 SFA
Model: Arduino Zero
Board-ID: SAMD21G18A-Zero-v0
The Board-ID field is machine-readable and consists of a number of dash-separated tokens.
The first token is the CPU type, second is the board type, and third is the board revision.
More tokens can be also added.
The bootloader should contain its info file as a static string somewhere in its code.
If possible, the last word of the bootloader code should point to this string.
This way, the info file can be found in the initial section of the CURRENT.UF2
file as well. Thus, a board type can be determined from the contents of CURRENT.UF2.
This is particularly useful with the source embedding (see above).
File containers
It is also possible to use the UF2 format as a container for one or more regular files (akin to a TAR file, or ZIP archive without compression). This is useful when the embedded device being flashed sports a file system.
The program to run may reside in one of the files, or in the main flash memory.
In such a usage the file container flag is set on blocks, the field fileSize
holds the file size of the current file, and the field targetAddr holds the
offset in current file.
The not main flash flag on blocks should be ignored when the file container is set.
The file name is stored at &data[payloadSize] (ie., right after the actual payload) and
terminated with a 0x00 byte. The format of filename is dependent on the
bootloader (usually implemented as some sort of file system daemon).
The bootloader will usually allow any size of the payload.
The current files on device might be exposed as multiple UF2 files, instead of
a single CURRENT.UF2. They may reside in directories, however, due to UF2 general
design, it doesn’t matter which directory the UF2 file is written to.
Typical writing procedure is as follows:
validate UF2 magic numbers
make sure that
targetAddr < fileSizeand thatfileSizeisn’t out of reasonable rangewrite
0x00atdata[475]to ensure NUL termination of file nameread file name from
&data[payloadSize]; perform any mapping on the file namecreate a directory where the file is to be written if it doesn’t exist
open the file for writing
truncate the file to
fileSizeseek
targetAddrwrite the payload (ie.,
data[0 ... payloadSize - 1])close the file
The fields blockNo and numBlocks refer to the entire UF2 file, not the current
file.
MD5 checksum
When the 0x4000 flag is set, the last 24 bytes of data[] hold the following structure:
Offset |
Size |
Value |
|---|---|---|
0 |
4 |
Start address of region |
4 |
4 |
Length of region in bytes |
8 |
16 |
MD5 checksum in binary format |
The flashing program should compute the MD5 sum of the specified region. If the region checksum matches, flashing of the current block can be skipped. Typically, many blocks in sequence will have the same region specified, and can all be skipped, if the matching succeeded. The position of the current block will typically be inside of the region. The position and size of the region should be multiple of page erase size (4k or 64k on typical SPI flash).
This is currently only used on ESP32, which is also why MD5 checksum is used.
Implementations
Bootloaders
Tiny UF2 - Support ESP32-S2, iMXRT10xx, STM32F4
RP2040 chip - native support in silicon
UF2-ChibiOS - Supports STM32H7
There’s an ongoing effort to implement UF2 in Codal.
Editors
Libraries
License
MIT
Code of Conduct
This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.
uf2conv – Packing and unpacking UF2 files
SYNOPSIS
This tool is based on the code at https://github.com/Microsoft/uf2 but has been modified for use with Sming and the RP2040:
- Handles multiple binary inputs using addr=content format
- Locations specified as offsets, relative to base address
- Defaults are set for RP2040, but can be changed (family ID and base address)
- Content is padded to flash sector size (4096 bytes) - required by RP2040 bootrom
- Content is padded with 0xff (not 0) as this represents erased (unprogrammed) flash location
- Renamed 'deploy' as 'upload' and added progress indicator
- Support for hex files and 'C' arrays removed
Run uf2conv.py -h for full list of options.
EXAMPLES
Pack binary file(s) to .uf2
Specify source using offset=content values:
uf2conv.py --convert 0x2000=cpx/firmware.bin 0x4000=cpx/partitions.bin --output cpx/upload.uf2
Offsets are relative to the start of flash memory.
Unpack a .uf2 file
uf2conv.py current.uf2 --output current.bin
Produces a list of current-{addr}.bin files for each chunk found in the .uf2 source file.
Without the --output option just displays a chunk summary.
Use the --verbose option to display details for every block in the file.
Configuration variables
- UF2CONV_PY
Path to the
uf2convutility should you wish to run it manually.
References
Used by
Sming (Rp2040) Component
Environment Variables
SoC support
rp2040
Framework
Timers and Clocks
Hardware timers
The ESP8266 has one of these, HardwareTimer, which can be programmed to trigger
when a pre-programmed interval has elapsed. The callback handler is executed in an
interrupt context so is restricted on what it can do.
The timer is appropriate where accurate and short timing is required.
The API for hardware (and Software timer queues timers) is identical, implemented using a
CallbackTimer class template for best performance.
Note
Applies to Esp8266 architecture only.
As with all Callback timers, the HardwareTimer can be one-shot or repeating.
With the Esp8266 a one-shot timer will repeat after it has been triggered, but only after the timer counter wraps around. The Esp8266 only has 23 bits of resolution so with a clock divider of 16 this will happen after about 1.7s.
Because the Esp8266 lacks any proper PWM hardware the timer latency is critical. Adding any additional code to the callback is therefore left to the programmer.
- group hardware_timer
Hardware callback timer.
Typedefs
-
template<hw_timer_clkdiv_t clkdiv = TIMER_CLKDIV_16, HardwareTimerMode mode = eHWT_NonMaskable>
using HardwareTimer1 = CallbackTimer<Timer1Api<clkdiv, mode>> Hardware Timer class template with selectable divider and interrupt mode.
-
using HardwareTimer = HardwareTimer1<>
Default hardware Timer class.
Enums
-
template<hw_timer_clkdiv_t clkdiv, HardwareTimerMode mode>
class Timer1Api : public CallbackTimerApi<Timer1Api<clkdiv, mode>> - #include <HardwareTimer.h>
Class template for Timer1 API.
Note
Provides low-level interface for timer access
-
template<hw_timer_clkdiv_t clkdiv = TIMER_CLKDIV_16, HardwareTimerMode mode = eHWT_NonMaskable>
Software timer queues
Sming has one timer queue which is managed via the Timer and SimpleTimer classes.
Like hardware timers, these are callback timers, but the callback handlers execute in the task context so there are no special restrictions on what you can do there.
Because of this, however, these timers are less precise hardware timers, and the accuracy generally depends on how busy the CPU is.
Timers can be ‘one-shot’, for timing single events, or ‘auto-reset’ repetitive timers. A repetitive timer will ensure that time intervals between successive callbacks are consistent.
Timer
The Timer class is the most flexible way to use software timers, supporting extended time intervals
and delegate callback functions so you can use it with class methods, capturing lambdas, etc.
- group timer
Extended timer queue class.
-
template<class TimerClass>
class OsTimer64Api : public CallbackTimerApi<OsTimer64Api<TimerClass>> - #include <Timer.h>
Class template implementing an extended OS Timer with 64-bit microsecond times and delegate callback support.
-
template<typename TimerApi>
class DelegateCallbackTimer : public CallbackTimer<TimerApi> - #include <Timer.h>
Class template adding delegate callback method support to the basic CallbackTimer template.
Public Functions
-
template<TimeType microseconds>
inline DelegateCallbackTimer &initializeUs(TimerDelegate delegateFunction) Initialise timer in microseconds, with static check.
- Template Parameters:
microseconds – Timer interval in microseconds
- Parameters:
delegateFunction – Function to call when timer triggers
- Return values:
ExtendedCallbackTimer& – Reference to timer
-
template<uint32_t milliseconds>
inline DelegateCallbackTimer &initializeMs(TimerDelegate delegateFunction) Initialise hardware timer in milliseconds, with static check.
- Template Parameters:
milliseconds – Timer interval in milliseconds
- Parameters:
delegateFunction – Function to call when timer triggers
- Return values:
ExtendedCallbackTimer& – Reference to timer
-
inline DelegateCallbackTimer &initializeMs(uint32_t milliseconds, TimerDelegate delegateFunction)
Initialise millisecond timer.
Note
Delegate callback method
- Parameters:
milliseconds – Duration of timer in milliseconds
delegateFunction – Function to call when timer triggers
-
inline DelegateCallbackTimer &initializeUs(uint32_t microseconds, TimerDelegate delegateFunction)
Initialise microsecond timer.
Note
Delegate callback method
- Parameters:
microseconds – Duration of timer in milliseconds
delegateFunction – Function to call when timer triggers
-
inline void setCallback(TimerDelegate delegateFunction)
Set timer trigger function using Delegate callback method.
Note
Don’t use this for interrupt timers
- Parameters:
delegateFunction – Function to be called on timer trigger
-
template<TimeType microseconds>
-
class Timer : public DelegateCallbackTimer<OsTimer64Api<Timer>>
- #include <Timer.h>
Callback timer class.
Subclassed by Profiling::MinMaxTimes< Timer >
-
class AutoDeleteTimer : public DelegateCallbackTimer<OsTimer64Api<AutoDeleteTimer>>
- #include <Timer.h>
Auto-delete callback timer class.
-
template<class TimerClass>
Simple Timer
The SimpleTimer class only supports regular function callbacks, but if you don’t need
the additional functionality that a regular Timer offers then it will save some RAM.
- group simple_timer
Basic timer queue class.
Typedefs
-
using SimpleTimer = CallbackTimer<OsTimerApi>
Basic callback timer.
Note
For delegate callback support and other features see
Timerclass.
-
class OsTimerApi : public CallbackTimerApi<OsTimerApi>
- #include <SimpleTimer.h>
Implements common system callback timer API.
Subclassed by CallbackTimer< OsTimerApi >
-
using SimpleTimer = CallbackTimer<OsTimerApi>
Callback Timer API
-
template<typename TimerApi>
class CallbackTimer : protected TimerApi Callback timer class template.
Note
Methods return object reference for Method Chaining http://en.wikipedia.org/wiki/Method_chaining This class template provides basic C-style callbacks for best performance
- Template Parameters:
TimerApi – The physical timer implementation
Subclassed by DelegateCallbackTimer< OsTimer64Api< AutoDeleteTimer > >, DelegateCallbackTimer< OsTimer64Api< Timer > >, DelegateCallbackTimer< TimerApi >
Public Functions
-
template<NanoTime::Unit unit, TimeType time>
inline CallbackTimer &initialize(TimerCallback callback, void *arg = nullptr) Initialise timer with an interval (static check) and callback.
Note
If interval out of range compilation will fail with error
- Template Parameters:
unit – Time unit for interval
time – Timer interval
- Parameters:
callback – Callback function to call when timer triggers
arg – Optional argument passed to callback
- Return values:
CallbackTimer& – Reference to timer
-
template<NanoTime::Unit unit>
inline CallbackTimer &initialize(TimeType time, TimerCallback callback, void *arg = nullptr) Initialise timer with an interval and callback.
- Template Parameters:
unit – Time unit for interval
- Parameters:
time – Timer interval
callback – Callback function to call when timer triggers
arg – Optional argument passed to callback
- Return values:
CallbackTimer& – Reference to timer
-
template<TimeType microseconds>
inline CallbackTimer &initializeUs(TimerCallback callback, void *arg = nullptr) Initialise timer in microseconds (static check) with Timer Callback and optional argument.
-
template<TimeType microseconds>
inline CallbackTimer &initializeUs(InterruptCallback callback = nullptr) Initialise timer in microseconds (static check) with optional Interrupt Callback (no argument)
-
inline CallbackTimer &initializeUs(TimeType microseconds, TimerCallback callback, void *arg = nullptr)
Initialise timer in microseconds with Timer Callback and optional argument.
-
inline CallbackTimer &initializeUs(TimeType microseconds, InterruptCallback callback = nullptr)
Initialise timer in microseconds with optional Interrupt Callback (no arg)
-
template<uint32_t milliseconds>
inline CallbackTimer &initializeMs(TimerCallback callback, void *arg = nullptr) Initialise hardware timer in milliseconds (static check) with Timer Callback and optional argument.
-
template<uint32_t milliseconds>
inline CallbackTimer &initializeMs(InterruptCallback callback = nullptr) Initialise hardware timer in milliseconds (static check) and optional Interrupt Callback (no arg)
-
inline CallbackTimer &initializeMs(uint32_t milliseconds, TimerCallback callback, void *arg = nullptr)
Initialise hardware timer in milliseconds with Timer Callback and optional argument.
-
inline CallbackTimer &initializeMs(uint32_t milliseconds, InterruptCallback callback = nullptr)
Initialise hardware timer in milliseconds with optional Interrupt Callback (no arg)
-
inline bool start(bool repeating = true)
Start timer running.
- Parameters:
repeating – True to restart timer when it triggers, false for one-shot (Default: true)
- Return values:
bool – True if timer started
-
inline bool startOnce()
Start one-shot timer.
Note
Timer starts and will run for configured period then stop
- Return values:
bool – True if timer started
-
inline void stop()
Stops timer.
-
inline bool restart()
Restart timer.
Note
Timer is stopped then started with current configuration
- Return values:
bool – True if timer started
-
inline bool isStarted() const
Check if timer is started.
- Return values:
bool – True if started
-
inline TickType getInterval() const
Get timer interval in clock ticks.
-
inline bool checkInterval(TickType ticks) const
Check timer interval is valid.
- Parameters:
ticks – Interval to check
- Return values:
bool – true if interval is within acceptable range for this timer
-
inline bool setInterval(TickType ticks)
Set timer interval in timer ticks.
- Parameters:
ticks – Interval in timer ticks
-
template<TimeType ticks>
inline void setInterval() Set timer interval in timer ticks (static check)
Note
On error, compilation fails with error message
- Template Parameters:
ticks – Interval in ticks
-
template<NanoTime::Unit unit, TimeType time>
inline void setInterval() Set timer interval in specific time unit (static check)
Note
On error, compilation fails with error message
- Template Parameters:
unit –
time – Interval to set
-
template<NanoTime::Unit unit>
inline bool setInterval(TimeType time) Set timer interval in timer ticks.
- Template Parameters:
unit –
- Parameters:
time – Interval in given units
-
inline bool setIntervalUs(TimeType microseconds)
Set timer interval in microseconds.
-
template<TimeType microseconds>
inline void setIntervalUs() Set timer interval in microseconds (static check)
-
inline bool setIntervalMs(uint32_t milliseconds)
Set timer interval in milliseconds.
-
template<uint32_t milliseconds>
inline void setIntervalMs() Set timer interval in milliseconds (static check)
-
inline void setCallback(TimerCallback callback, void *arg = nullptr)
Set timer trigger callback.
- Parameters:
callback – Function to call when timer triggers
arg – Optional argument passed to callback
-
inline void setCallback(InterruptCallback callback)
Set timer trigger callback.
Note
Provided for convenience where callback argument not required
- Parameters:
callback – Function to call when timer triggers
Public Static Functions
-
static inline constexpr Millis millis()
Get a millisecond time source.
-
static inline constexpr Micros micros()
Get a microsecond time source.
-
template<uint64_t us>
static inline constexpr uint64_t usToTicks() Convert microsecond count into timer ticks.
-
static inline TickType usToTicks(TimeType time)
Convert microsecond count into timer ticks.
-
template<uint64_t ticks>
static inline constexpr uint64_t ticksToUs() Convert timer ticks into microseconds.
-
static inline TimeType ticksToUs(TickType ticks)
Convert timer ticks into microseconds.
-
template<uint64_t ticks>
static inline constexpr void checkInterval() Check timer interval in ticks is valid (static check)
Note
On error, compilation fails with error message
- Template Parameters:
ticks – Timer interval to check
-
template<NanoTime::Unit unit, uint64_t time>
static inline constexpr void checkInterval() Check timer interval in specific time unit is valid (static check)
Note
On error, compilation fails with error message
- Template Parameters:
unit – Time unit for interval
time – Interval to check
-
template<uint64_t milliseconds>
static inline constexpr void checkIntervalMs() Check timer interval in milliseconds is valid (static check)
-
template<uint64_t microseconds>
static inline constexpr void checkIntervalUs() Check timer interval in microseconds is valid (static check)
Polled timers
Polled timers can be used to measure elapsed time intervals or to check for timeouts within code loops. Traditionally one might do this as follows:
const unsigned TIMEOUT_MS = 100;
unsigned start = millis();
while(millis() - start < TIMEOUT_MS) {
// do some stuff
}
unsigned elapsed = millis() - start;
Serial.print("Elapsed time: ");
Serial.print(elapsed);
Serial.println("ms");
Note
A common source of bugs when hand-coding such loops can be illustrated by this alternative:
unsigned timeout = millis() + TIMEOUT_MS;
while(millis() < timeout) {
// do some stuff
}
At first glance this looks better than the first approach as we don’t have a subtraction within the loop. However, when millis() exceeds MAXINT - TIMEOUT_MS the timeout calculation will wrap and the loop will never get executed. This takes a little under 25 days, but with microseconds it’ll happen in less than an hour. This may not be the desired behaviour.
It’s generally safer and easier to use a PolledTimer:
OneShotFastMs timer(TIMEOUT_MS);
while(!timer.expired()) {
// do some stuff
}
auto elapsed = timer.elapsedTime(); // Returns a `NanoTime::Time` object, value with time units
Serial.print("Elapsed time: ");
Serial.println(elapsed.toString()); // Includes units in the elapsed time interval
// Show time rounded to nearest whole seconds
Serial.println(elapsed.as<NanoTime::Seconds>().toString());
If you prefer to use microseconds, use OneShotFastUs instead or specify the units directly:
OneShotElapseTimer<NanoTime::Microseconds> timer;
Another advantage of polled timers is speed. Every call to millis (or micros) requires a calculation from clock ticks into milliseconds (or microseconds). It’s also a function call.
Polled timers measure time using hardware clock ticks, and query the hardware timer register directly without any function calls or calculations. This makes them a much better choice for tight timing loops.
Here’s the output from the BenchmarkPolledTimer module:
How many loop iterations can we achieve in 100 ms ?
Using millis(), managed 55984 iterations, average loop time = 1786ns (143 CPU cycles)
Using micros(), managed 145441 iterations, average loop time = 688ns (55 CPU cycles)
Using PolledTimer, managed 266653 iterations, average loop time = 375ns (30 CPU cycles)
API Documentation
- group polled_timer
Polled interval timers.
Unnamed Group
-
template<NanoTime::Unit unit>
using OneShotElapseTimer = PolledTimer::OneShot<PolledTimerClock, unit>
-
template<NanoTime::Unit unit>
using PeriodicElapseTimer = PolledTimer::Periodic<PolledTimerClock, unit>
-
using OneShotFastMs = OneShotElapseTimer<NanoTime::Milliseconds>
-
using PeriodicFastMs = PeriodicElapseTimer<NanoTime::Milliseconds>
-
using OneShotFastUs = OneShotElapseTimer<NanoTime::Microseconds>
-
using PeriodicFastUs = PeriodicElapseTimer<NanoTime::Microseconds>
-
using ElapseTimer = OneShotFastUs
-
template<NanoTime::Unit units>
using OneShotCpuCycleTimer = PolledTimer::OneShot<CpuCycleClockNormal, units>
-
template<NanoTime::Unit units>
using PeriodicCpuCycleTimer = PolledTimer::Periodic<CpuCycleClockNormal, units>
-
template<NanoTime::Unit units>
using OneShotCpuCycleTimerFast = PolledTimer::OneShot<CpuCycleClockFast, units>
-
template<NanoTime::Unit units>
using PeriodicCpuCycleTimerFast = PolledTimer::Periodic<CpuCycleClockFast, units>
-
using CpuCycleTimer = OneShotCpuCycleTimer<NanoTime::Nanoseconds>
-
using CpuCycleTimerFast = OneShotCpuCycleTimerFast<NanoTime::Nanoseconds>
-
typedef OneShotFastMs oneShotFastMs
-
typedef PeriodicFastMs periodicFastMs
-
typedef OneShotFastUs oneShotFastUs
-
typedef PeriodicFastUs periodicFastUs
-
template<NanoTime::Unit unit>
Timer range
The maximum interval for a timer varies depending on the clock source and the selected prescaler. The count may be further restricted by hardware. For example, Timer1 only provides a 23-bit count which means with a /16 prescaler (the default for HardwareTimer) it overflows after only 1.67 seconds.
It’s therefore important to check that timers are being used within their valid range. There are generally two ways to do this:
- Runtime checks
This means checking function/method return values and acting accordingly. This often gets omitted because it can lead to cluttered code, which then leads to undiagnosed bugs creeping in which can be very difficult to track down later on.
With a polled timer, you’d use one of these methods to set the time interval:
bool reset(const TimeType& timeInterval); bool resetTicks(const TimeType& interval);
They both return true on success.
- Static checks
These checks are performed during code compilation, so if a check fails the code won’t compile. In regular ‘C’ code you’d do this using #if statements, but C++ offers a much better way using static_assert.
To reset a polled timer and incorporate a static check, use this method:
template <uint64_t timeInterval> void reset();
Note that timeInterval cannot be a variable (even if it’s const) as the compiler must be able to determine its value. It must therefore be constexpr compatible.
You can use static checking to pre-validate the range for a timer before using it:
timer.checkTime<10>();
timer.checkTime<10000>();
This will throw a compile-time error if the timer is not capable of using intervals in the range 10 - 10000 microseconds (or whichever time unit you’ve selected). It doesn’t add any code to the application. If the code compiles, then you can be confident that the timer will function as expected and you don’t need to check return values.
You can see these checks in action in the Live Debug sample, which uses the HardwareTimer
with a /16 prescaler. If you change BLINK_INTERVAL_MS to 2000 then the code will not compile.
Clocks
Timers and their capabilities can vary considerably. For example, Timer1 can be configured with a prescaler of 1, 16 or 256 which affects both the resolution and range of the timer. One might also consider the CPU cycle counter to have a selectable prescaler of 1 or 2, depending on whether it’s running at 80MHz or 160MHz.
A Clock definition is a class template which allows us to query timer properties and perform time
conversions for a specific timer configuration. These definitions can be found in Sming/Platform/Clocks.h.
Note
A Clock is a purely virtual construct and does not provide any means to configure the hardware,
although it does provide the ticks() method to obtain the current timer value.
Clocks are made more useful by TimeSource, a generic class template defined in Sming/Core/NanoTime.h.
This provides methods to convert between time values and tick values for a specific time unit.
Let’s say we want a microsecond source using Timer2:
TimeSource<Timer2Clock, NanoTime::Microseconds, uint32_t> t2source;
We can now call methods of t2source like this:
// What's the maximum Timer2 value in microseconds?
Serial.println(t2source.maxClockTime());
// How many clock ticks per microsecond ?
Serial.println(t2source.ticksPerUnit()); // 5/1
// How many clock ticks for 100us ?
Serial.println(t2source.template timeConst<100>().ticks());
Note that all of these values are computed at compile time. Some runtime conversions:
Serial.println(t2source.timeToTicks(100));
Serial.println(t2source.ticksToTime(10000));
The results of conversions are rounded rather than truncated, which provides more accurate results and reduces timing jitter.
For debugging purposes you can print a description:
Serial.println(t2source.toString()); // "Timer2Clock/5MHz/32-bit/microseconds"
See Sming/Core/NanoTime.h for further details.
System Clock API
-
template<hw_timer_clkdiv_t clkdiv>
struct Timer1Clock : public NanoTime::Clock<Timer1Clock<clkdiv>, HW_TIMER_BASE_CLK / (1 << clkdiv), uint32_t, MAX_HW_TIMER1_INTERVAL> - #include <Clocks.h>
Clock implementation for Hardware Timer 1.
- Template Parameters:
clkdiv – Prescaler in use
NanoTime
Utilities for handling time periods at nanosecond resolution.
-
namespace NanoTime
Typedefs
Enums
Functions
-
constexpr uint64_t round_ce(double v)
-
const char *unitToLongString(Unit unit)
Get a long string identifying the given time units, e.g. “seconds”.
-
template<uint64_t time, Unit unitsFrom, Unit unitsTo, typename R = std::ratio_divide<UnitTickRatio<unitsTo>, UnitTickRatio<unitsFrom>>>
constexpr uint64_t convert() Function template to convert a constant time quantity from one unit to another.
Note
example:
uint32_t micros = convert<50, Milliseconds, Microseconds>();
- Template Parameters:
time – The time to convert
unitsFrom – Units for
timeparameterunitsTo – Units for return value
- Return values:
TimeType – Converted time
-
template<typename TimeType>
TimeType convert(const TimeType &time, Unit unitsFrom, Unit unitsTo) Function template to convert a time quantity from one unit to another.
- Template Parameters:
TimeType – Variable type to use for calculation
- Parameters:
time – The time to convert
unitsFrom – Units for
timeparameterunitsTo – Units for return value
- Return values:
TimeType – Converted time, returns TimeType(-1) if calculation overflowed
-
template<typename T>
Time<T> time(Unit unit, T value) Helper function to create a Time and deduce the type.
-
template<Unit unitsFrom, Unit unitsTo, typename TimeType>
TimeType convert(const TimeType &time) Function template to convert a time quantity from one unit to another.
- Template Parameters:
unitsFrom – Units for
timeparameterunitsTo – Units for return value
TimeType – Variable type to use for calculation
- Parameters:
time – The time to convert
- Return values:
TimeType – Converted time, returns TimeType(-1) if calculation overflowed
Variables
-
constexpr BasicRatio32 unitTicks[UnitMax + 1] = {{1000000000, 1}, {1000000, 1}, {1000, 1}, {1, 1}, {1, 60}, {1, 60 * 60}, {1, 24 * 60 * 60},}
List of clock ticks for each supported unit of time.
-
struct Frequency
- #include <NanoTime.h>
Class to represent a frequency.
-
template<class Clock_, Unit unit_, uint64_t time_>
struct TimeConst - #include <NanoTime.h>
Class template to represent a fixed time value for a specific Clock.
Note
Includes compile-time range checking. Time is taken as reference for conversions.
- Template Parameters:
Clock_ –
unit_ –
time_ –
-
template<class Clock_, uint64_t ticks_>
struct TicksConst - #include <NanoTime.h>
Class template representing a fixed clock tick count.
Note
Includes compile-time range checking
- Template Parameters:
Source_ –
ticks_ –
-
template<class Clock_, Unit unit_, typename TimeType_>
struct TimeSource : public Clock_ - #include <NanoTime.h>
Class template for accessing a Clock in specific time units.
Note
Includes compile-time range checking. Time is taken as reference for conversions.
- Template Parameters:
Clock_ –
units_ –
TimeType_ – Limits range of calculations
Subclassed by PolledTimer::Timer< NanoTime::Microseconds >, PolledTimer::Timer< NanoTime::Milliseconds >, PolledTimer::Timer< NanoTime::Seconds >, PolledTimer::Timer< NanoTime::Nanoseconds >
-
template<typename T>
struct Time - #include <NanoTime.h>
Class to handle a simple time value with associated unit.
-
template<typename Clock_, typename T>
struct Ticks - #include <NanoTime.h>
Class to handle a tick value associated with a clock.
-
template<typename ClockDef, uint32_t frequency_, typename TickType_, TickType_ maxTicks_>
struct Clock - #include <NanoTime.h>
Class template representing a physical Clock with fixed timing characteristics.
See also
Use
TimeSourceto work with a Clock in specific time unitsNote
Physical clocks are implemented using this as a base. The provided frequency accounts for any prescaler setting in force. Fixing this at compile time helps to avoid expensive runtime calculations and permits static range checks.
- Template Parameters:
Subclassed by NanoTime::TimeSource< Clock, unit_, TimeType >
-
struct TimeValue
- #include <NanoTime.h>
A time time broken into its constituent elements.
Note
Useful for analysing and printing time values
-
constexpr uint64_t round_ce(double v)
Core Framework
Program Space
Support for storing and accessing data from Program Space (flash memory).
A string literal (e.g. “string”) used in code gets emitted to the .rodata segment by the compiler. That means it gets read into RAM at startup and remains there.
To avoid this, and reclaim the RAM, the data must be stored in a different segment. This is done
using the PROGMEM macro.
Such strings are usually defined using the PSTR() macro, which also ensures that any duplicate
strings are merged and thus avoids storing them more than once. This is particularly beneficial
for debugging strings.
Templated code
Attention
PROGMEM may not work when used in templated code.
GCC silently ignores ‘section attributes’ in templated code, which means const variables will remain
in the default .rodata section.
Strings defined using PSTR() (and related macros) or FlashString definitions
are handled correctly because internally they use special names (__pstr__ and __fstr__)
which the linker picks up on.
memcpy_aligned
Once in flash memory, string data must be read into RAM before it can be used. Accessing the flash memory directly is awkward. If locations are not strictly accessed as 4-byte words the system will probably crash; I say ‘probably’ because sometimes it just behaves weirdly if the RAM address isn’t aligned.
So, the location being accessed, the RAM buffer it’s being copied to and the length all have to be
word-aligned, i.e. integer multiples of 4 bytes.
If these conditions are satisfied, then it’s safe to use a regular memcpy() call.
However, you are strongly discouraged from doing this.
Instead, use memcpy_aligned(), which will check the parameters and raise an assertion in debug mode
if they are incorrect.
FakePgmSpace
Standard string functions such as memcpy_P(), strcpy_P(), etc. are provided to enable
working with P-strings. With the new arduino-provided toolchains these are now part of the standard
library, however Sming has some additions and differences.
F()Loads a String object with the given text, which is allocated to flash:
String s = F("test");
Note
The
F()macro differs from the Arduino/Esp8266 implementation in that it instantiates aStringobject.Since the length of the string is known at compile-time, it can be passed to the String constructor which avoids an additional call to
strlen_P()._F()Like F() except buffer is allocated on stack. Most useful where nul-terminated data is required:
m_printf(_F("C-style string\n"));
This macro is faster than
F(), but you need to be careful as the temporary stack buffer becomes invalid as soon as the containing block goes out of scope. Used as a function parameter, that means the end of the function call.Examples:
println(_F("Debug started")); commandOutput->print(_F("Welcome to the Tcp Command executor\r\n"));
Bad:
char* s = _F("string")
An assignment such as this will not work because the temporary will be out of scope after the statement, hence s will point to garbage. In this instance
PSTR_ARRAY(s, "string")can be used.DEFINE_PSTR()Declares a PSTR stored in flash. The variable (name) points to flash memory so must be accessed using the appropriate xxx_P function.
LOAD_PSTR()Loads pre-defined PSTR into buffer on stack:
// The ``_LOCAL`` macro variants include static allocation DEFINE_PSTR_LOCAL(testFlash, "This is a test string\n"); LOAD_PSTR(test, testFlash) m_printf(test);
PSTR_ARRAY()Create and load a string into the named stack buffer. Unlike
_F(), this ensures a loaded string stays in scope:String testfunc() { //char * test = "This is a string"; <<- BAD PSTR_ARRAY(test, "This is a string"); m_printf(test); ... return test; // Implicit conversion to String }
Both DEFINE_PSTR() and PSTR_ARRAY() load a PSTR into a stack buffer, but using
sizeof on that buffer will return a larger value than the string itself because it’s aligned.
Calling sizeof on the original flash data will get the right value.
If it’s a regular nul-terminated string then strlen_P() will get the length, although it’s
time-consuming.
FlashString
For efficient, fast and flexible use of PROGMEM data see FlashString.
API Documentation
-
FLASH_MEMORY_START_ADDR
-
isFlashPtr(ptr)
Simple check to determine if a pointer refers to flash memory.
-
PROGMEM
Place entity into flash memory.
Attach to const variable declaration to have it stored in flash memory
Such variables should not be accessed like regular pointers as aligned instructions are required. Use the provided library functions, such as
memcpy_P, instead.
-
PROGMEM_PSTR
Place NUL-terminated string data into flash memory.
Duplicate string data will be merged according to the rules laid out in https://sourceware.org/binutils/docs/as/Section.html
-
PSTR(str)
Define and use a NUL-terminated ‘C’ flash string inline.
Note
Uses string section merging so must not contain embedded NULs
- Parameters:
str –
- Return values:
char[] – In flash memory, access using flash functions
-
PGM_P
Identifies a char pointer as living in flash memory Use to clarify code.
-
PRIPSTR
Remove ?
-
void *memcpy_aligned(void *dst, const void *src, unsigned len)
copy memory aligned to word boundaries
dst and src must be aligned to word (4-byte) boundaries
lenwill be rounded up to the nearest word boundary, so the dst buffer MUST be large enough for this.- Parameters:
dst –
src –
len – Size of the source data
-
int memcmp_aligned(const void *ptr1, const void *ptr2, unsigned len)
compare memory aligned to word boundaries
ptr1 and ptr2 must all be aligned to word (4-byte) boundaries. len is rounded up to the nearest word boundary
- Parameters:
ptr1 –
ptr2 –
len –
- Return values:
int – 0 if all bytes match
-
IS_ALIGNED(_x)
determines if the given value is aligned to a word (4-byte) boundary
-
ALIGNUP4(n)
Align a size up to the nearest word boundary.
-
ALIGNDOWN4(n)
Align a size down to the nearest word boundary.
-
printf_P_heap(f_P, ...)
-
printf_P_stack(f_P, ...)
-
printf_P(fmt, ...)
-
PSTR_COUNTED(str)
Define and use a counted flash string inline.
Note
Strings are treated as binary data so may contain embedded NULs, but duplicate strings are not merged.
- Parameters:
str –
- Return values:
char[] – In flash memory, access using flash functions
-
_F(str)
-
DEFINE_PSTR(name, str)
define a PSTR
- Parameters:
name – name of string
str – the string data
-
DEFINE_PSTR_LOCAL(name, str)
define a PSTR for local (static) use
- Parameters:
name – name of string
str – the string data
-
DECLARE_PSTR(name)
Declare a global reference to a PSTR instance.
- Parameters:
name –
-
LOAD_PSTR(name, flash_str)
Create a local (stack) buffer called
nameand load it with flash data.If defining a string within a function or other local context, must declare static.
Example:
void testfunc() { static DEFINE_PSTR(test, "This is a test string\n"); m_printf(LOAD_PSTR(test)); }
- Parameters:
name –
flash_str – Content stored in flash. The compiler knows its size (length + nul), which is rounded up to multiple of 4 bytes for fast copy.
-
_FLOAD(pstr)
-
PSTR_ARRAY(name, str)
Define a flash string and load it into a named array buffer on the stack.
For example, this:
is roughly equivalent to this:PSTR_ARRAY(myText, "some text");
where ALIGNED_SIZE is the length of the text (including NUL terminator) rounded up to the next word boundary. To get the length of the text, excluding NUL terminator, use:char myText[ALIGNED_SIZE] = "some text";
sizeof(PSTR_myText) - 1
Note
Must not contain embedded NUL characters
Core Data Classes
BitSet
Introduction
The C++ STL provides std::bitset
which emulates an array of bool elements.
The BitSet class template provides similar support for sets of strongly-typed elements.
For example:
enum class Fruit {
apple,
banana,
kiwi,
orange,
passion,
pear,
tomato,
};
using FruitBasket = BitSet<uint8_t, Fruit, unsigned(Fruit::tomato) + 1>;
static constexpr FruitBasket fixedBasket = Fruit::orange | Fruit::banana | Fruit::tomato;
A FruitBasket uses one byte of storage, with each bit representing an item of Fruit.
If the basket contains a piece of fruit, the corresponding bit is set.
If it does not, the bit is clear.
Without BitSet you implement this as follows:
using FruitBasket = uint8_t;
static constexpr FruitBasket fixedBasket = _BV(Fruit::orange) | _BV(Fruit::banana) | _BV(Fruit::tomato);
To test whether the set contains a value you’d do this:
if(fixedBasket & _BV(Fruit::orange)) {
Serial.println("I have an orange");
}
With a BitSet, you do this:
if(fixedBasket[Fruit::orange]) {
Serial.println("I have an orange");
}
And you can add an element like this:
basket[Fruit::kiwi] = true;
Bit manipulation operators are provided so you can do logical stuff like this:
FruitBasket basket1; // Create an empty basket
// Add a kiwi fruit
basket1 = fixedBasket + Fruit::kiwi;
// Create a second basket containing all fruit not in our first basket
FruitBasket basket2 = ~basket1;
// Remove some fruit
basket2 -= Fruit::orange | Fruit::tomato;
And so on.
To display the contents of a BitSet, do this:
Serial.print(_F("My basket contains: "));
Serial.println(basket1);
You will also need to provide an implementation of toString(Fruit)
or whatever type you are using for the set elements.
API
-
template<typename S, typename E, size_t size_ = sizeof(S) * 8>
class BitSet Manage a set of bit values using enumeration.
API is similar to a simplified std::bitset, but with added +/- operators.
Note
It is important to specify size correctly when using enumerated values. In the
FruitBasketexample, we use auint8_tstorage type so can have up to 8 possible values. However, theFruitenum contains only 7 values. The set operations will therefore be restricted to ensure that the unused bit is never set.- Template Parameters:
S – Storage type (e.g. uint32_t). This determines how much space to use, and must be an unsigned integer. It is safe to use this class in structures, where it will occupy exactly the required space.
E – Element type e.g. enum class. You can use any enum or unsigned integer for the elements. These must have ordinal sequence starting at 0.
size_ – Number of possible values in the set. Defaults to maximum for given storage type. Number of possible values in the set. This must be at least 1, and cannot be more than the given Storage type may contain. For example, a :cpp:type:
uint8_tmay contain up to 8 values.
Public Functions
-
constexpr BitSet() = default
Construct empty set.
-
template<typename S2>
inline constexpr BitSet(const BitSet<S2, E> &bitset) Copy constructor.
- Parameters:
bitset – The set to copy
-
inline constexpr BitSet(S value)
Construct from a raw set of bits.
- Parameters:
value – Integral type whose bits will be interpreted as set{E}
-
inline constexpr BitSet(E e)
Construct set containing a single value.
- Parameters:
e – Value to place in our new BitSet object
-
inline size_t count() const
Get the number of elements in the set, i.e. bits set to 1.
-
inline BitSet &operator&=(const BitSet &rhs)
Intersection: Leave only elements common to both sets.
-
inline bool operator[](E e) const
Read-only [] operator.
- Parameters:
e – Element to test for
- Return values:
bool – true if given element is in the set
-
inline BitRef operator[](E e)
Read/write [] operator.
This returns a temporary BitRef object to support assignment operations such as
set[x] = value- Parameters:
e – Element to read or write
- Return values:
BitRef – Temporary object used to do the read or write
-
inline bool any() const
Determine if set contains any values.
-
inline bool any(const BitSet &other) const
Determine if set contains any values from another set i.e. intersection != [].
-
inline bool all() const
Test if set contains all possible values.
-
inline bool none() const
Test if set is empty.
-
inline BitSet &set(E e, bool state = true)
Set the state of the given bit (i.e. add to or remove from the set)
- Parameters:
e – Element to change
state – true to add the element, false to remove it
Public Static Functions
-
static inline constexpr size_t size()
Get the number of possible elements in the set.
-
class BitRef
CString
Introduction
Whilst use of char* pointers is very common in Sming code, it is generally advisable to avoid pointers in C++ where possible.
The STL provides class templates such as unique_ptr which deals with memory allocation and de-allocation to avoid issues with memory leaks.
The CString class implements this on a char[] and adds some additional methods which are similar to the String class.
String vs. CString
String objects each require a minimum of 24 bytes of RAM, and always contain a length field.
A CString is much simpler and contains only a char* pointer, so a NULL string is only 4 bytes.
When storing arrays or lists of strings (or objects containing those strings) which change infrequently, such as fixed configuration data, use a CString for memory efficiency.
API Documentation
-
class CString : public std::unique_ptr<char[]>
Class to manage a NUL-terminated C-style string When storing persistent strings in RAM the regular String class can become inefficient, so using a regular
char*can be preferable. This class provides that with additional methods to simplify lifetime management and provide some interoperability with Wiring String objects.
CStringArray
Introduction
This is a class to manage a double NUL-terminated list of strings, such as "one\0two\0three\0".
It’s similar in operation to Vector<String>, but more memory-efficient as all the data is
stored in a single String object. (CStringArray is a subclass of String.)
You can see some examples in Sming/Core/DateTime.cpp and Sming/Core/Data/WebConstants.cpp.
Background
Each value in the sequence is terminated by a NUL character \0.
For clarity, placing one string per line is suggested:
// ["one", "two", "three"]
CStringArray csa = F(
"one\0"
"two\0"
"three\0"
);
Note use of the F() macro.
Assignments require a length because of the NUL characters, so this won’t work as expected:
// ["one"]
CStringArray csa =
"one\0"
"two\0"
"three\0";
When assigning sequences, the final NUL separator may be omitted (it will be added automatically):
// ["one", "two", "three"]
CStringArray csa = F(
"one\0"
"two\0"
"three"
);
Sequences may contain empty values, so this example contains four values:
// ["one", "two", "three", ""]
CStringArray csa = F(
"one\0"
"two\0"
"three\0"
"\0"
);
Adding strings
Elements can be added using standard concatenation operators:
CStringArray arr;
arr += "string1";
arr += 12;
arr += 5.4;
arr += F("data");
Be mindful that each call may require a heap re-allocation, so consider
estimating or calculating the required space and using String::reserve():
CStringArray arr;
arr.reserve(250);
// now add content
Use with FlashString
You can use a single FlashString containing these values and load them all
at the same time into a CStringArray:
DEFINE_FSTR_LOCAL(fstr_list,
"a\0"
"b\0"
"c\0"
"d\0"
"e\0"
);
CStringArray list(fstr_list);
for(unsigned i = 0; i < list.count(); ++i) {
debug_i("list[%u] = '%s'", i, list[i]);
}
Note
The entire FlashString is loaded into RAM so better suited for occasional lookups or if instantiated outside of a loop.
You may find FSTR::Array, FSTR::Vector or FSTR::Map more appropriate.
See FlashString for details.
Iterator support
Looping by index is slow because the array must be scanned from the start for each access. Iterators are simpler to use and much more efficient:
for(auto s: list) {
debug_i("'%s'", s);
}
For more complex operations:
CStringArray::Iterator pos;
for(auto it = list.begin(); it != list.end(); ++it) {
debug_i("list[%u] = '%s' @ %u", it.index(), *it, it.offset());
// Can use direct comparison with const char* or String
if(it == "c") {
pos = it; // Note position
}
}
if(pos) {
debug_i("Item '%s' found at index %u, offset %u", pos.str(), pos.index(), pos.offset());
} else {
debug_i("Item not found");
}
Pushing and popping
CStringArray can be used as a simple FIFO or stack using push/pop methods. Behaviour is similar to STL deque, except pop methods also return a value.
STACK:
CStringArray csa;
csa.pushBack("first value");
csa.pushBack("second value");
String popStack = csa.popBack(); // "second value"
FIFO:
CStringArray csa;
csa.pushBack("first value");
csa.pushBack("second value");
String deque = csa.popFront(); // "first value"
Note that popping values does not perform any memory de-allocation.
Comparison with Vector<String>
- Advantages
More memory efficient Uses only a single heap allocation (assuming content is passed to constructor) Useful for simple lookups, e.g. mapping enumerated values to strings
- Disadvantages
Slower. Items must be iterated using multiple strlen() calls Ordering and insertions / deletions not supported
API Documentation
-
class CStringArray : private String
Class to manage a double null-terminated list of strings, such as “one\0two\0three\0”.
Concatenation operators
-
template<typename T>
inline CStringArray &operator+=(T value) Append numbers, etc. to the array.
- Parameters:
value – char, int, float, etc. as supported by String
Public Functions
-
bool add(const char *str, int length = -1)
Append a new string (or array of strings) to the end of the array.
Note
If str contains any NUL characters it will be handled as an array
- Parameters:
str –
length – Length of new string in array (default is length of str)
- Return values:
bool – false on memory allocation error
-
inline bool add(const String &str)
Append a new string (or array of strings) to the end of the array.
Note
If str contains any NUL characters it will be handled as an array
- Parameters:
str –
- Return values:
bool – false on memory allocation error
-
int indexOf(const char *str, bool ignoreCase = true) const
Find the given string and return its index.
Note
Comparison is not case-sensitive
- Parameters:
str – String to find
ignoreCase – Whether search is case-sensitive or not
- Return values:
int – index of given string, -1 if not found
-
inline int indexOf(const String &str, bool ignoreCase = true) const
Find the given string and return its index.
Note
Comparison is not case-sensitive
- Parameters:
str – String to find
ignoreCase – Whether search is case-sensitive or not
- Return values:
int – index of given string, -1 if not found
-
inline bool contains(const char *str, bool ignoreCase = true) const
Check if array contains a string.
Note
Search is not case-sensitive
- Parameters:
str – String to search for
ignoreCase – Whether search is case-sensitive or not
- Return values:
bool – True if string exists in array
-
inline bool contains(const String &str, bool ignoreCase = true) const
Check if array contains a string.
Note
Search is not case-sensitive
- Parameters:
str – String to search for
ignoreCase – Whether search is case-sensitive or not
- Return values:
bool – True if string exists in array
-
const char *getValue(unsigned index) const
Get string at the given position.
- Parameters:
index – 0-based index of string to obtain
- Return values:
const – char* nullptr if index is not valid
-
inline const char *operator[](unsigned index) const
Get string at the given position.
- Parameters:
index – 0-based index of string to obtain
- Return values:
const – char* nullptr if index is not valid
-
inline const char *front() const
Get first value in array, null if empty.
-
bool pushFront(const char *str)
Insert item at start of array.
- Parameters:
str – Item to insert
- Return values:
bool – false on memory error
-
String popFront()
Pop first item from array (at index 0)
- Return values:
String – null if array is empty
-
const char *back() const
Get last item in array, null if empty.
-
inline bool pushBack(const char *str)
Add item to end of array.
- Parameters:
str – Item to add
- Return values:
bool – false on memory error
-
inline void clear()
Empty the array.
-
unsigned count() const
Get quantity of strings in array.
- Return values:
unsigned – Quantity of strings
-
String join(const String &separator = ",") const
Get contents of array as delimited string.
- Parameters:
separator – What to join elements with
- Return values:
String – e.g. CStringArray(F(“a\0b\0c”)).join() returns “a,b,c”
-
bool reserve(size_t size)
Pre-allocate String memory.
On failure, the String is left unchanged. reserve(0), if successful, will validate an invalid string (i.e., “if (s)” will be true afterwards)
- Parameters:
size –
- Return values:
bool – true on success, false on failure
-
inline const char *c_str() const
Get a constant (un-modifiable) pointer to String content.
- Return values:
const – char* Always valid, even for a null string
-
bool endsWith(char suffix) const
Compare the end of a String.
- Parameters:
suffix –
- Return values:
bool – true on match
-
bool endsWith(const String &suffix) const
Compare the end of a String.
- Parameters:
suffix –
- Return values:
bool – true on match
-
size_t getBytes(unsigned char *buf, size_t bufsize, size_t index = 0) const
Read contents of a String into a buffer.
Note
Returned data always nul terminated so buffer size needs to take this into account
- Parameters:
buf – buffer to write data
bufsize – size of buffer in bytes
index – offset to start
- Return values:
unsigned – number of bytes copied, excluding nul terminator
-
inline bool startsWith(const String &prefix) const
Compare the start of a String Comparison is case-sensitive, must match exactly.
- Parameters:
prefix –
- Return values:
bool – true on match
-
bool startsWith(const String &prefix, size_t offset) const
Compare a string portion.
mis-named as does not necessarily compare from start
Note
Comparison is case-sensitive, must match exactly
- Parameters:
prefix –
offset – Index to start comparison at
- Return values:
bool – true on match
-
inline void toCharArray(char *buf, size_t bufsize, size_t index = 0) const
Read contents of String into a buffer.
See also
See
getBytes()
-
class Iterator
-
template<typename T>
CSV Reader
-
class CsvReader
Class to parse a CSV file.
Spec: https://www.ietf.org/rfc/rfc4180.txt
Each record is located on a separate line
Line ending for last record in the file is optional
Field headings are provided either in the source data or in constructor (but not both)
Fields separated with ‘,’ and whitespace considered part of field content
Fields may or may not be quoted - if present, will be removed during parsing
Fields may contain line breaks, quotes or commas
Quotes may be escaped thus “” if field itself is quoted
Additional features:
Line breaks can be
or \r
Escapes codes within fields will be converted:
\r \t “, \
Field separator can be changed in constructor
Public Functions
-
inline CsvReader(IDataSourceStream *source, char fieldSeparator = ',', const CStringArray &headings = nullptr, size_t maxLineLength = 2048)
Construct a CSV reader.
- Parameters:
source – Stream to read CSV text from
fieldSeparator –
headings – Required if source data does not contain field headings as first row
maxLineLength – Limit size of buffer to guard against malformed data
-
void reset()
Reset reader to start of CSV file.
Cursor is set to ‘before start’. Call ‘next()’ to fetch first record.
-
inline bool next()
Seek to next record.
-
inline unsigned count() const
Get number of columns.
-
inline const char *getValue(unsigned index)
Get a value from the current row.
- Parameters:
index – Column index, starts at 0
- Return values:
const – char* nullptr if index is not valid
-
inline const char *getValue(const char *name)
Get a value from the current row.
- Parameters:
index – Column name
- Return values:
const – char* nullptr if name is not found
-
inline int getColumn(const char *name)
Get index of column given its name.
- Parameters:
name – Column name to find
- Return values:
int – -1 if name is not found
-
inline explicit operator bool() const
Determine if row is valid.
-
inline const CStringArray &getHeadings() const
Get headings.
-
inline const CStringArray &getRow() const
Get current row.
Data formatting
A standard mechanism is provided for encoding and decoding text in commonly-used text formats
using a Format::Formatter class implementation.
-
namespace Format
-
-
class Formatter
- #include <Formatter.h>
Virtual class to perform format-specific String adjustments.
Subclassed by Format::Standard
-
class Standard : public Format::Formatter
- #include <Standard.h>
Subclassed by Format::Html, Format::Json, Format::Xml
-
class Formatter
Linked Object Lists
-
class LinkedObject
- #include <LinkedObject.h>
Base virtual class to allow objects to be linked together.
This can be more efficient than defining a separate list, as each object requires only an additional pointer field for ‘next’.
Subclassed by LinkedObjectTemplate< Request >, LinkedObjectTemplate< Item >, LinkedObjectTemplate< HttpResourcePlugin >, LinkedObjectTemplate< StationInfo >, LinkedObjectTemplate< Element >, LinkedObjectTemplate< PluginRef >, LinkedObjectTemplate< Object >, LinkedObjectTemplate< Asset >, LinkedObjectTemplate< Answer >, LinkedObjectTemplate< Controller >, LinkedObjectTemplate< Handler >, LinkedObjectTemplate< Question >, LinkedObjectTemplate< Service >, LinkedObjectTemplate< Device >, LinkedObjectTemplate< PriorityNode< ObjectType > >, LinkedObjectTemplate< Renderer >, LinkedObjectTemplate< ClassType >, LinkedObjectTemplate< Info >, LinkedObjectTemplate< ObjectType >
Public Functions
-
inline virtual ~LinkedObject()
-
inline virtual LinkedObject *next() const
-
inline bool insertAfter(LinkedObject *object)
-
inline bool operator==(const LinkedObject &other) const
-
inline bool operator!=(const LinkedObject &other) const
Private Members
-
LinkedObject *mNext = {nullptr}
Friends
- friend class LinkedObjectList
-
inline virtual ~LinkedObject()
-
template<typename ObjectType>
class LinkedObjectTemplate : public LinkedObject - #include <LinkedObject.h>
Base class template for linked items with type casting.
Public Types
-
using Iterator = IteratorTemplate<ObjectType, ObjectType*, ObjectType&>
-
using ConstIterator = IteratorTemplate<const ObjectType, const ObjectType*, const ObjectType&>
-
template<typename T, typename TPtr, typename TRef>
class IteratorTemplate - #include <LinkedObject.h>
Public Types
-
using iterator_category = std::forward_iterator_tag
-
using difference_type = std::ptrdiff_t
Public Functions
-
inline IteratorTemplate(const IteratorTemplate &other)
-
inline IteratorTemplate &operator++()
-
inline IteratorTemplate operator++(int)
-
inline bool operator==(const IteratorTemplate &rhs) const
-
inline bool operator!=(const IteratorTemplate &rhs) const
-
using iterator_category = std::forward_iterator_tag
-
using Iterator = IteratorTemplate<ObjectType, ObjectType*, ObjectType&>
-
class LinkedObjectList
- #include <LinkedObjectList.h>
Singly-linked list of objects.
Note
We don’t own the items, just keep references to them
Subclassed by LinkedObjectListTemplate< Controller >, LinkedObjectListTemplate< Handler >, LinkedObjectListTemplate< Service >, LinkedObjectListTemplate< Control >, LinkedObjectListTemplate< ObjectType >
Public Functions
-
inline LinkedObjectList()
-
inline LinkedObjectList(LinkedObject *object)
-
bool add(LinkedObject *object)
-
inline bool add(const LinkedObject *object)
-
inline bool insert(LinkedObject *object)
-
inline bool insert(const LinkedObject *object)
-
bool remove(LinkedObject *object)
-
inline LinkedObject *pop()
-
inline void clear()
-
inline LinkedObject *head()
-
inline const LinkedObject *head() const
-
inline bool isEmpty() const
Protected Attributes
-
LinkedObject *mHead = {nullptr}
-
inline LinkedObjectList()
-
template<typename ObjectType>
class LinkedObjectListTemplate : public LinkedObjectList - #include <LinkedObjectList.h>
Subclassed by OwnedLinkedObjectListTemplate< Request >, OwnedLinkedObjectListTemplate< Item >, OwnedLinkedObjectListTemplate< HttpResourcePlugin >, OwnedLinkedObjectListTemplate< Surface >, OwnedLinkedObjectListTemplate< Element >, OwnedLinkedObjectListTemplate< StationInfo >, OwnedLinkedObjectListTemplate< PluginRef >, OwnedLinkedObjectListTemplate< Asset >, OwnedLinkedObjectListTemplate< Answer >, OwnedLinkedObjectListTemplate< Question >, OwnedLinkedObjectListTemplate< Device >, OwnedLinkedObjectListTemplate< PriorityNode< ObjectType > >, OwnedLinkedObjectListTemplate< ClassType >, OwnedLinkedObjectListTemplate< Info >, OwnedLinkedObjectListTemplate< ObjectType >
Public Types
-
using Iterator = typename LinkedObjectTemplate<ObjectType>::template IteratorTemplate<ObjectType, ObjectType*, ObjectType&>
-
using ConstIterator = typename LinkedObjectTemplate<ObjectType>::template IteratorTemplate<const ObjectType, const ObjectType*, const ObjectType&>
Public Functions
-
LinkedObjectListTemplate() = default
-
inline LinkedObjectListTemplate(ObjectType *object)
-
inline ObjectType *head()
-
inline const ObjectType *head() const
-
inline ConstIterator begin() const
-
inline ConstIterator end() const
-
inline bool add(ObjectType *object)
-
inline bool add(const ObjectType *object)
-
inline bool insert(ObjectType *object)
-
inline bool insert(const ObjectType *object)
-
inline ObjectType *pop()
-
inline size_t count() const
-
inline bool contains(const ObjectType &object) const
-
using Iterator = typename LinkedObjectTemplate<ObjectType>::template IteratorTemplate<ObjectType, ObjectType*, ObjectType&>
-
template<typename ObjectType>
class OwnedLinkedObjectListTemplate : public LinkedObjectListTemplate<ObjectType> - #include <LinkedObjectList.h>
Class template for singly-linked list of objects.
Note
We own the objects so are responsible for destroying them when removed
Subclassed by Storage::Disk::BasePartitionTable
Public Functions
-
OwnedLinkedObjectListTemplate() = default
-
OwnedLinkedObjectListTemplate(const OwnedLinkedObjectListTemplate &other) = delete
-
OwnedLinkedObjectListTemplate &operator=(const OwnedLinkedObjectListTemplate &other) = delete
-
inline ~OwnedLinkedObjectListTemplate()
-
inline bool remove(ObjectType *object)
-
inline void clear()
-
OwnedLinkedObjectListTemplate() = default
Object Map
-
template<typename K, typename V>
class ObjectMap - #include <ObjectMap.h>
Implementation of a HashMap for owned objects, i.e. anything created with new().
Example:
void test() { ObjectMap<String, MyType> map; MyType* object1 = new MyType(); if (map["key1"] == nullptr) { // Does NOT create entry in map map["key1"] = object1; // Entry now created, "key1" -> object1 } MyType* object2 = new MyType(); map["key1"] = object2; // object1 is destroyed, "key1" -> object2 // Demonstrate use of value reference auto value = map["key1"]; // Returns ObjectMap<String, MyType>::Value object value = new MyType(); // "key1" -> new object value = nullptr; // Free object, "key1" -> nullptr (but still in map) value.remove(); // Free object1 and remove from map // As soon as `map` goes out of scope, all contained objects are destroyed map["key1"] = new MyType(); map["key2"] = new MyType(); }Note
Once added to the map the object is destroyed when no longer required.
Public Functions
-
inline ObjectMap()
-
inline ~ObjectMap()
-
inline unsigned count() const
Get the number of entries in this map.
- Return values:
int – Entry count
-
inline const V *operator[](const K &key) const
Get value for given key, if it exists.
Note
The caller must not use
deleteon the returned value- Parameters:
key –
- Return values:
const – V* Will be null if not found in the map
-
inline Value operator[](const K &key)
Access map entry by reference.
See also
Note
If the given key does not exist in the map it will NOT be created
- Parameters:
key –
- Return values:
Value – Guarded access to mapped value corresponding to given key
-
inline Value get(const K &key)
Get map entry value.
See also
operator[]- Parameters:
key –
- Return values:
Value –
-
inline V *find(const K &key) const
Find the value for a given key, if it exists.
Note
If you need to modify the existing map entry, use
operator[]orvalueAt()- Parameters:
key –
- Return values:
V* – Points to the object if it exists, otherwise nullptr
-
inline int indexOf(const K &key) const
Get the index of a key.
- Parameters:
key –
- Return values:
int – The index of the key, or -1 if key does not exist
-
inline bool contains(const K &key) const
Check if a key is contained within this map.
- Parameters:
key – the key to check
- Return values:
bool – true if key exists
-
inline void removeAt(unsigned index)
Remove entry at given index.
- Parameters:
index – location to remove from this map
-
inline bool remove(const K &key)
Remove a key from this map.
- Parameters:
key – The key identifying the entry to remove
- Return values:
bool – true if the value was found and removed
-
inline V *extract(const K &key)
Get the value for a given key and remove it from the map, without destroying it.
Note
The returned object must be freed by the caller when no longer required
- Parameters:
key –
- Return values:
V* –
-
inline V *extractAt(unsigned index)
Get the value at a given index and remove it from the map, without destroying it.
Note
The returned object must be freed by the caller when no longer required
- Parameters:
index –
- Return values:
V* –
-
inline void clear()
Clear the map of all entries.
-
class Value
- #include <ObjectMap.h>
Class to provide safe access to mapped value.
Note
ObjectMap
operator[]returns one of these, which provides behaviour consistent with V*Public Functions
-
inline bool remove()
Remove this value from the map.
- Return values:
bool – true if the value was found and removed
-
inline bool remove()
-
inline ObjectMap()
Object Queue
Packet writing
-
struct Packet
- #include <Packet.h>
Helper class for reading/writing packet content.
Subclassed by HostPacket, NetworkPacket
Public Functions
-
inline Packet(void *data, uint16_t pos = 0)
-
inline const uint8_t *ptr() const
-
inline uint8_t *ptr()
-
inline void skip(uint16_t len) const
-
inline uint8_t peek8() const
-
inline uint8_t read8() const
-
inline void read(void *buffer, uint16_t len) const
-
inline void write8(uint8_t value)
-
inline void write(const void *s, uint16_t len)
-
inline Packet(void *data, uint16_t pos = 0)
Range
-
template<typename T>
struct TRange - #include <Range.h>
Manage a range of numbers between specified limits.
Values in the range meet the criteria (min <= value <= max)
Public Functions
-
inline constexpr TRange()
-
inline constexpr TRange()
Streams
Sming provides a set of Stream class which extend Stream methods.
IDataSourceStream is used where read-only access is required.
It introduces the IDataSourceStream::readMemoryBlock() method which
performs a regular read without updating the stream position.
This allows optimistic reading and re-sending, but cannot be handled by some stream
types and should be used with care.
ReadWriteStream is used where read/write operation is required.
Printing
The arduino Print class provides the basic output streaming mechanism.
Sming has some enhancements:
- C++ streaming operation <<
Building output is commonly done like this:
Serial.print("Temperature: "); Serial.print(temperature); Serial.print(" °C, humidity: "); Serial.print(humidity); Serial.println("%");
In Sming, this will produce exactly the same result:
Serial << "Temperature" << temperature << " °C, humidity: " << humidity << "%" << endl;
Note
Sming does NOT support the C++ STL streaming classes, such as
iostream, etc.- Number Printing
Examples:
Serial.print(12, HEX); // "c" Serial.print(12, HEX, 4); // "000c" Serial.print(12, HEX, 4, '.'); // "...c" Serial.print(12); // "12" Serial.print(12, DEC, 4); // "0012"
Similar extensions are provided for
Stringconstruction:Serial << "0x" << String(12, HEX, 8); // "0x0000000c" Serial << String(12, DEC, 4, '.'); // "..12"
- Field-width control
Supported via String methods:
Serial << String(12).padLeft(4); // " 12" Serial << String(12).padLeft(4, '0'); // "0012" Serial << String(12).padRight(4); // "12 " Serial << String(12).pad(-4, '0'); // "0012" Serial << String(12).pad(4); // "12 "
- Strongly-typed enumerations
Use of
enum classis good practice as it produces strongly-typed and scoped values. Most of these are also provided with a standardtoString(E)function overload.This allows string equivalents to be printed very easily:
auto status = HTTP_STATUS_HTTP_VERSION_NOT_SUPPORTED; auto type = MIME_HTML; Serial.print(type); // "text/html" // "Status: HTTP Version Not Supported, type: text/html" Serial << "Status: " << status << ", type: " << type << endl; // Status: 505, type: 0 Serial << "Status: " << int(status) << ", type: " << int(type) << endl;
API Documentation
IDataSourceStream
-
class IDataSourceStream : public Stream
Base class for read-only stream.
Subclassed by FSTR::Stream, Graphics::SubStream, IFS::FWFS::ArchiveStream, MultiStream, ReadWriteStream, RtttlJsonListStream, SectionStream, SharedMemoryStream< T >, StreamTransformer, TemplateStream, UPnP::DescriptionStream, UrlencodedOutputStream, XorOutputStream
Public Functions
-
inline virtual StreamType getStreamType() const
Get the stream type.
- Return values:
StreamType – The stream type.
-
inline virtual bool isValid() const
Determine if the stream object contains valid data.
Note
Where inherited classes are initialised by constructor this method indicates whether that was successful or not (e.g. FileStream)
- Return values:
bool – true if valid, false if invalid
-
virtual size_t readBytes(char *buffer, size_t length) override
Read chars from stream into buffer.
Terminates if length characters have been read or timeout (see setTimeout). Returns the number of characters placed in the buffer (0 means no valid data found).
Note
Inherited classes may provide more efficient implementations without timeout.
-
virtual uint16_t readMemoryBlock(char *data, int bufSize) = 0
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
-
virtual int read() override
Read one character and moves the stream pointer.
- Return values:
The – character that was read or -1 if none is available
-
virtual int peek() override
Read a character without advancing the stream pointer.
- Return values:
int – The character that was read or -1 if none is available
-
inline virtual int seekFrom(int offset, SeekOrigin origin)
Change position in stream.
Note
This method is implemented by streams which support random seeking, such as files and memory streams.
- Parameters:
offset –
origin –
- Return values:
New – position, < 0 on error
-
inline virtual bool seek(int len)
Move read cursor.
- Parameters:
len – Relative cursor adjustment
- Return values:
bool – True on success.
-
virtual bool isFinished() = 0
Check if all data has been read.
- Return values:
bool – True on success.
-
inline virtual int available() override
Return the total length of the stream.
- Return values:
int – -1 is returned when the size cannot be determined
-
inline virtual size_t write(uint8_t charToWrite) override
From Stream class: We don’t write using this stream.
- Parameters:
charToWrite –
-
inline virtual String id() const
Returns unique id of the resource.
- Return values:
String – the unique id of the stream.
-
inline virtual String getName() const
Returns name of the resource.
Note
Commonly used to obtain name of file
- Return values:
String –
-
inline virtual MimeType getMimeType() const
Get MIME type for stream content.
- Return values:
MimeType –
-
virtual String readString(size_t maxLen) override
Overrides Stream method for more efficient reading.
Note
Stream position is updated by this call
-
inline virtual bool moveString(String &s)
Memory-based streams may be able to move content into a String.
If the operation is not supported by the stream,
swill be invalidated and false returned.Because a String object must have a NUL terminator, this will be appended if there is sufficient capacity. In this case, the method returns true.
If there is no capacity to add a NUL terminator, then the final character of stream data will be replaced with a NUL. The method returns false to indicate this.
- Parameters:
s – String object to move data into
- Return values:
bool – true on success, false if there’s a problem.
-
size_t readBytes(char *buffer, size_t length)
Read chars from stream into buffer.
Terminates if length characters have been read or timeout (see setTimeout). Returns the number of characters placed in the buffer (0 means no valid data found).
Note
Inherited classes may provide more efficient implementations without timeout.
-
inline virtual StreamType getStreamType() const
ReadWriteStream
-
class ReadWriteStream : public IDataSourceStream
Base class for read/write stream.
Subclassed by CircularBuffer, EndlessMemoryStream, HardwareSerial, IFS::FileStream, LimitedMemoryStream, MemoryDataStream, Ota::UpgradeOutputStream, OtaUpgrade::BasicStream, Storage::PartitionStream, StreamWrapper, USB::CDC::UsbSerial
Public Functions
-
inline virtual size_t write(uint8_t charToWrite) override
From Stream class: We don’t write using this stream.
- Parameters:
charToWrite –
-
virtual size_t write(const uint8_t *buffer, size_t size) override = 0
Write chars to stream.
Note
Although this is defined in the Print class, ReadWriteStream uses this as the core output method so descendants are required to implement it
- Parameters:
buffer – Pointer to buffer to write to the stream
size – Quantity of chars to write
- Return values:
size_t – Quantity of chars written to stream
-
virtual size_t copyFrom(IDataSourceStream *source, size_t size = SIZE_MAX)
Copy data from a source stream.
- Parameters:
source – Stream to read data from
size – Quantity of chars to write, determines size of intermediate buffer to use
- Return values:
size_t – Quantity of chars actually written, may be less than requested
-
size_t write(uint8_t c) = 0
Writes a single character to output stream.
- Parameters:
c – Character to write to output stream
- Return values:
size_t – Quantity of characters written to output stream
-
inline size_t write(const char *str)
Writes a c-string to output stream.
- Parameters:
str – Pointer to c-string
- Return values:
size_t – Quantity of characters written to stream
-
size_t write(const uint8_t *buffer, size_t size)
Writes characters from a buffer to output stream.
- Parameters:
buffer – Pointer to character buffer
size – Quantity of characters to write
- Return values:
size_t – Quantity of characters written to stream
-
inline size_t write(const char *buffer, size_t size)
Writes characters from a buffer to output stream.
- Parameters:
buffer – Pointer to character buffer
size – Quantity of characters to write
- Return values:
size_t – Quantity of characters written to stream
-
inline virtual size_t write(uint8_t charToWrite) override
Stream Classes
- group stream
Data stream classes.
Typedefs
-
using TemplateFlashMemoryStream = FSTR::TemplateStream
Template stream using content stored in flash.
-
class CircularBuffer : public ReadWriteStream
- #include <CircularBuffer.h>
Circular stream class.
Base class for data source stream
-
class IDataSourceStream : public Stream
- #include <DataSourceStream.h>
Base class for read-only stream.
Subclassed by FSTR::Stream, Graphics::SubStream, IFS::FWFS::ArchiveStream, MultiStream, ReadWriteStream, RtttlJsonListStream, SectionStream, SharedMemoryStream< T >, StreamTransformer, TemplateStream, UPnP::DescriptionStream, UrlencodedOutputStream, XorOutputStream
-
class EndlessMemoryStream : public ReadWriteStream
- #include <EndlessMemoryStream.h>
Memory stream that stores unlimited number of bytes.
Memory is allocated on write and released when all written bytes have been read out. This behaviour differs from a circular buffer as the size is not fixed.
-
class FileStream : public IFS::FileStream
- #include <FileStream.h>
File stream class.
-
class GdbFileStream : public IFS::FileStream
- #include <GdbFileStream.h>
GDB File stream class to provide access to host files whilst running under debugger.
-
class DirectoryTemplate : public SectionTemplate
- #include <DirectoryTemplate.h>
Directory stream class.
Subclassed by IFS::HtmlDirectoryTemplate, IFS::JsonDirectoryTemplate
-
class FileStream : public IFS::FsBase, public ReadWriteStream
- #include <FileStream.h>
File stream class.
Subclassed by FileStream, GdbFileStream, HostFileStream
-
class HtmlDirectoryTemplate : public IFS::DirectoryTemplate
- #include <HtmlDirectoryTemplate.h>
Read-only stream access to directory listing with HTML output.
-
class LimitedMemoryStream : public ReadWriteStream
- #include <LimitedMemoryStream.h>
Memory stream operating on fixed-size buffer Once the limit is reached the stream will discard incoming bytes on write.
-
class LimitedWriteStream : public StreamWrapper
- #include <LimitedWriteStream.h>
A stream wrapper class that limits the number of bytes that can be written. Helpful when writing on a file system or memory should be limited to the available size of the media.
-
class MemoryDataStream : public ReadWriteStream
- #include <MemoryDataStream.h>
Read/write stream using expandable memory buffer.
This is intended to allow data to be streamed into it, then streamed back out at a later date.
It is not intended to have data continuously written in and read out; memory is not reclaimed as it is read.
Subclassed by JsonObjectStream, UPnP::ActionResponse::Stream
-
class MultiStream : public IDataSourceStream
- #include <MultiStream.h>
Base class for read-only stream which generates output from multiple source streams.
Subclassed by MultipartStream, StreamChain
-
class ReadWriteStream : public IDataSourceStream
- #include <ReadWriteStream.h>
Base class for read/write stream.
Subclassed by CircularBuffer, EndlessMemoryStream, HardwareSerial, IFS::FileStream, LimitedMemoryStream, MemoryDataStream, Ota::UpgradeOutputStream, OtaUpgrade::BasicStream, Storage::PartitionStream, StreamWrapper, USB::CDC::UsbSerial
- #include <SharedMemoryStream.h>
Memory stream operating on fixed shared buffer.
One reason for templating this class is for distinction between
charorconst chartypes, to avoid dangerous casts. Elements may be structures or other types.
-
class StreamWrapper : public ReadWriteStream
- #include <StreamWrapper.h>
An abstract class that provides a wrapper around a stream.
Subclassed by LimitedWriteStream, PartCheckerStream
-
class TemplateFileStream : public TemplateStream
- #include <TemplateFileStream.h>
Template stream using content from the filesystem.
-
class TemplateStream : public IDataSourceStream
- #include <TemplateStream.h>
Stream which performs variable-value substitution on-the-fly.
Template uses {varname} style markers which are replaced as the stream is read.
Note: There must be no whitespace after the opening brace. For example,
{ varname }will be emitted as-is without modification.This allows inclusion of CSS fragments such as
td { padding: 0 10px; }in HTML.If necessary, use double-braces
{{varname}}in templates and enable by callingsetDoubleBraces(true).Invalid tags, such as
{"abc"}will be ignored, so JSON templates do not require special treatment.Subclassed by FSTR::TemplateStream, SectionTemplate, TemplateFileStream
-
class XorOutputStream : public IDataSourceStream
- #include <XorOutputStream.h>
Xors original stream content with the specified mask.
-
class HostFileStream : public IFS::FileStream
- #include <HostFileStream.h>
Host File stream class.
-
class TemplateStream : public TemplateStream
- #include <TemplateStream.hpp>
Template Flash memory stream class.
-
class Base64OutputStream : public StreamTransformer
- #include <Base64OutputStream.h>
Read-only stream to emit base64-encoded content from source stream.
-
class ChunkedStream : public StreamTransformer
- #include <ChunkedStream.h>
Read-only stream to obtain data using HTTP chunked encoding.
Used where total length of stream is not known in advance
-
class MultipartStream : public MultiStream
- #include <MultipartStream.h>
Read-only stream for creating HTTP multi-part content.
-
class QuotedPrintableOutputStream : public StreamTransformer
- #include <QuotedPrintableOutputStream.h>
Read-only stream that transforms bytes of data into quoted printable data stream.
-
class UrlencodedOutputStream : public IDataSourceStream
- #include <UrlencodedOutputStream.h>
Represents key-value pairs as urlencoded string content.
-
class PartitionStream : public ReadWriteStream
- #include <PartitionStream.h>
Stream operating directory on a Storage partition.
To support write operations, the target region must be erased first.
-
class JsonObjectStream : public MemoryDataStream
- #include <JsonObjectStream.h>
JsonObject stream class.
-
using TemplateFlashMemoryStream = FSTR::TemplateStream
-
class StreamTransformer : public IDataSourceStream
Class that can be used to transform streams of data on the fly.
Subclassed by Base64OutputStream, ChunkedStream, QuotedPrintableOutputStream
Public Functions
-
inline virtual void saveState()
A method that backs up the current state.
-
inline virtual void restoreState()
A method that restores the last backed up state.
Protected Functions
-
virtual size_t transform(const uint8_t *in, size_t inLength, uint8_t *out, size_t outLength) = 0
Inherited class implements this method to transform a block of data.
Note
Called with
in = nullptrandinLength = 0at end of input stream- Parameters:
in – source data
inLength – source data length
out – output buffer
outLength – size of output buffer
- Return values:
size_t – number of output bytes written
-
inline virtual void saveState()
Template Streams
Sming provides several classes to assist with serving dynamic content:
Basic Templating
The TemplateStream class is a stream which performs variable-value substitution using
{varname} style markers, which are replaced as the stream is read.
You can find a simple demonstration of how this class is used in the Bootstrap Http Server sample application.
Note
There must be no whitespace after the opening brace.
For example, { varname } will be emitted as-is without modification.
This allows inclusion of CSS fragments such as td { padding: 0 10px; } in HTML
without resorting to double-braces.
If necessary, use double-braces {{varname}} in your template and
call TemplateStream::setDoubleBraces() (true).
Invalid tags, such as {"abc"} will be ignored, so JSON templates do not require
special treatment.
Variable values can be set using TemplateStream::setVar() or TemplateStream::setVars().
These are stored in a HashMap which can be accessing directly via TemplateStream::variables().
To support calculated values and external lookups, an optional callback may
be provided via TemplateStream::onGetValue().
This is invoked only if a variable is not found in the map.
Another option is to use TemplateStream as a base class and override the TemplateStream::getValue() method.
Important
If required, text must be escaped appropriately for the output format.
For example, encoding reserved HTML characters can be handled using Format::Html::escape().
Advanced Templating
Introduction
The SectionTemplate class extends TemplateStream to provide more advanced dataset processing capabilities.
It is intended to be used as the base class for a data provider.
One such implementation is the IFS::DirectoryTemplate class.
The Basic IFS sample demonstrates how it can be used to provide a formatted directory
listing in multiple formats, using a different template for each format.
The Basic Templates sample illustrates a similar approach using data from CSV data files.
If the output format requires escaping, create an instance of the appropriate Format::Formatter
and call SectionTemplate::setFormatter().
If providing custom values via callback, obtain the current formatter via SectionTemplate::formatter()
class and call the escape method.
Note that for performance reasons this is not done automatically as often variable values
do not require escaping. User-provided values or filenames must always be properly escaped.
Sections
Templates typically contain multiple sections.
The IFS::DirectoryTemplate, for example, uses 3 sections for header, content and footer.
The header and footer are emitted exactly once, but the content section is repeated for each available data record.
The SectionStream class is used internally so that all sections can be provided within a single file.
Sections are (by default) marked {SECTION} … {/SECTION}.
Everything outside of these markers is ignored, so can contain comments.
Using SectionTemplate
Implementations should provide the following methods:
- nextRecord
This method is called before a new content record is about to be output. Here’s the annotated
IFS::DirectoryTemplateimplementation:// Return true if we have a new valid record, false if not bool nextRecord() override { // Content section we fetch the next directory record, if there is one if(sectionIndex() == 1) { return directory->next(); } // This code emits the header and footer sections exactly once // Returning false suppresses their output completely return recordIndex() < 0; }
This sets up the ‘current’ directory information record.
- getValue
Lookup values for a given field:
String getValue(const char* name) override { // return ... }
Important
If required, text must be escaped appropriately for the output format. Use
SectionTemplate::formatter()to obtain the current For example, encoding reserved HTML characters can be handled usingFormat::Html::escape().
Control language
A basic control language is implemented using ! escaped tags.
Commands may have zero or more arguments, separated by :.
Numbers must be decimal and start with a digit, e.g.
11or5.6Strings must be quoted “…”
Sub-expressions must be contained in braces {…}
Anything else is treated as a variable name. Variable names beginning with $ are reserved for internal use. The following values are currently defined:
$section The current section index
$record The current record index
Conditional if/else/endif statements may be nested.
This is the current command list:
{!int:A}Output A as integer{!float:A}Output A as float{!string:A}Output A as quoted string{!mime_type:A}Get MIME type string for a filename{!replace:A:B:C}Copy of A with all occurrences of B replaced with C{!length:A}Number of characters in A{!pad:A:B:C}Copy of A padded to at least B characters with C (default is space). Use -ve B to left-pad. C{!repeat:A:B}Repeat A, number of iterations is B{!kb:A}Convert A to KB{!ifdef:A}emit block if A is not zero-length{!ifdef:A}emit block if A is zero-length{!ifeq:A:B}emit block if A == B{!ifneq:A:B}emit block if A != B{!ifgt:A:B}emit block if A > B{!iflt:A:B}emit block if A < B{!ifge:A:B}emit block if A >= B{!ifle:A:B}emit block if A <= B{!ifbtw:A:B:C}emit block if B <= A <= C{!ifin:A:B}emit block if A contains B{!ifin:A:B}emit block if A does not contain B{!else}{!endif}{!add:A:B}A + B{!sub:A:B}A - B{!goto:A}move to section A{!count:A}emit number of records in section A{!index:A}emit current record index for section A
Note
See Sming/Core/Data/Streams/SectionTemplate.h for an up-to-date list of commands and internal variables.
Here’s an excerpt from the Basic_IFS sample, displaying information for a single file:
{!iflt:$record:100} <!-- If $record < 100 -->
<tr>
<td>{$record}</td>
<td>{file_id}</td>
<td><a href="{path}{name}"><span style='font-size:20px'>{icon}</span> {name}</a></td>
<td>{!mime_type:name}</td>
<td>{modified}</td>
{!ifin:attr:"D"} <!-- Value of 'attr' variable contains "D" ->
<td></td><td></td>
{!else}
<td>{size}<br>{!kb:size} KB</td>
<td>{original_size}<br>{!kb:original_size} KB</td>
{!endif}
<td>{!replace:attr_long:", ":"<br>"}</td>
<td>{compression}</td>
<td>{access_long}</td>
</tr>
{!else} <!-- $record >= 100 -->
Too many records {$record}
{!endif}
API Reference
-
class TemplateStream : public IDataSourceStream
Stream which performs variable-value substitution on-the-fly.
Template uses {varname} style markers which are replaced as the stream is read.
Note: There must be no whitespace after the opening brace. For example,
{ varname }will be emitted as-is without modification.This allows inclusion of CSS fragments such as
td { padding: 0 10px; }in HTML.If necessary, use double-braces
{{varname}}in templates and enable by callingsetDoubleBraces(true).Invalid tags, such as
{"abc"}will be ignored, so JSON templates do not require special treatment.Subclassed by FSTR::TemplateStream, SectionTemplate, TemplateFileStream
Public Types
Public Functions
-
inline TemplateStream(IDataSourceStream *stream, bool owned = true)
Create a template stream.
- Parameters:
stream – source of template data
owned – If true (default) then stream will be destroyed when complete
-
inline virtual StreamType getStreamType() const override
Get the stream type.
- Return values:
StreamType – The stream type.
-
virtual uint16_t readMemoryBlock(char *data, int bufSize) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
-
virtual int seekFrom(int offset, SeekOrigin origin) override
Change position in stream.
Note
This method is implemented by streams which support random seeking, such as files and memory streams.
- Parameters:
offset –
origin –
- Return values:
New – position, < 0 on error
-
inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
-
inline void setVar(const String &name, const String &value)
Set value of a variable in the template file.
Note
Sets and existing variable or adds a new variable if variable does not already exist
- Parameters:
name – Name of variable
value – Value to assign to the variable
-
inline void setVars(const Variables &vars)
Set multiple variables in the template file.
- Parameters:
vars – Template Variables
-
inline Variables &variables()
Get the template variables.
- Return values:
TemplateVariables – Reference to the template variables
-
inline virtual String getName() const override
Returns name of the resource.
Note
Commonly used to obtain name of file
- Return values:
String –
-
inline void onGetValue(GetValueDelegate callback)
Set a callback to obtain variable values.
- Parameters:
callback – Invoked only if variable name not found in map
-
inline void enableOutput(bool enable)
During processing applications may suppress output of certain sections by calling this method from within the getValue callback.
-
inline bool isOutputEnabled() const
Determine if stream output is active.
Used by SectionTemplate class when processing conditional tags.
-
inline void setDoubleBraces(bool enable)
Use two braces {{X}} to mark tags.
- Parameters:
enable – true: use two braces, false (default): single brace only
-
virtual String evaluate(char *&expr)
Evaluate a template expression.
This method is overridden by SectionTemplate to support more complex expressions.
- Parameters:
expr – IN: First character after the opening brace(s) OUT: First character after the closing brace(s)
- Return values:
String – Called internally and an opening brace (“{” or “{{”) has been found. Default behaviour is to locate the closing brace(s) and interpret the bounded text as a variable name, which is passed to
getValue.
-
inline TemplateStream(IDataSourceStream *stream, bool owned = true)
-
class SectionTemplate : public TemplateStream
Provides enhanced template tag processing for use with a SectionStream.
Subclassed by IFS::DirectoryTemplate
Public Types
Public Functions
-
inline void onGetValue(GetValue callback)
Set a callback to be invoked.
Alternative to subclassing.
-
inline void setFormatter(Formatter &formatter)
Associate a text format with this template stream.
- Parameters:
formatter – Provide formatter so we can call escape(), etc. as required
-
inline Formatter &formatter() const
Get the stream format.
- Return values:
Formatter& – The formatter in effect. Default is :cpp:class:Format::Standard.
-
inline virtual MimeType getMimeType() const override
Get the MIME type associated with this template stream.
- Return values:
MimeType – As defined by the formatter. Default is MIME_TEXT.
-
inline const SectionStream &stream() const
Access the underlying section stream.
Provided for debugging and other purposes. Applications should not use this method.
- Return values:
SectionStream& – Wraps source stream provided in constructor
-
inline int sectionIndex() const
Get the index for the current section.
- Return values:
int – Indices are 0-based, returns -1 if ‘Before Start’
-
inline uint8_t sectionCount() const
Get number of sections in source stream.
- Return values:
uint8_t – Source is scanned in constructor so this is always valid
-
inline int recordIndex() const
Get current record index.
- Return values:
int – Indices are 0-based, returns -1 if ‘Before Start’
-
bool gotoSection(uint8_t index)
Discard current output and change current section.
- Parameters:
uint8_t – Index of section to move to
- Return values:
bool – true on success, false if section index invalid
-
inline void onNextRecord(NextRecord callback)
Set a callback to be invoked when a new record is required.
Can be used as alternative to subclassing.
-
virtual String evaluate(char *&expr) override
Evaluate a template expression.
This method is overridden by SectionTemplate to support more complex expressions.
- Parameters:
expr – IN: First character after the opening brace(s) OUT: First character after the closing brace(s)
- Return values:
String – Called internally and an opening brace (“{” or “{{”) has been found. Default behaviour is to locate the closing brace(s) and interpret the bounded text as a variable name, which is passed to
getValue.
-
inline void onGetValue(GetValue callback)
-
class SectionStream : public IDataSourceStream
Presents each section within a source stream as a separate stream.
Sections are (by default) marked {!SECTION} … {/SECTION} This is typically used with templating but can be used with any stream type provided the tags do not conflict with content.
Public Types
-
using NextSection = Delegate<void()>
Application notification callback when section changes.
-
using NextRecord = Delegate<bool()>
Application callback to move to next record.
- Retval bool:
true to emit section, false to skip
Public Functions
-
inline SectionStream(IDataSourceStream *source, uint8_t maxSections = 5)
Construct a section stream with default options.
-
inline SectionStream(IDataSourceStream *source, uint8_t maxSections, const String &startTag, const String &endTag)
Construct a section stream.
- Parameters:
source – Contains all section data, must support random seeking
startTag – Unique text used to mark start of a section
endTag – Marks end of a section
-
inline virtual int available() override
Return the total length of the stream.
- Return values:
int – -1 is returned when the size cannot be determined
-
virtual uint16_t readMemoryBlock(char *data, int bufSize) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
-
virtual int seekFrom(int offset, SeekOrigin origin) override
Change position in stream.
Note
This method is implemented by streams which support random seeking, such as files and memory streams.
- Parameters:
offset –
origin –
- Return values:
New – position, < 0 on error
-
inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
-
inline size_t count() const
Get number of sections in this stream.
-
inline const Section *getSection() const
Get description of the current section.
- Return values:
Section* – The section information, or nullptr if there is no current section
-
inline const Section *getSection(unsigned index) const
Get description for any section given its index.
- Return values:
Section* – The section information, or nullptr if section was not found
-
inline void onNextSection(NextSection callback)
Register a callback to be invoked when moving to a new section.
-
inline void onNextRecord(NextRecord callback)
Register a callback to be invoked when moving to a new record.
-
bool gotoSection(uint8_t index)
Goto a new section immediately.
-
inline bool setNewSection(int8_t index)
Goto a new section after current tag has been processed.
-
struct Section
-
using NextSection = Delegate<void()>
Date and Time
DateTime class
-
class DateTime
Date and time class.
Date and time functions mostly work with Unix time, the quantity of seconds since 00:00:00 1970-01-01. There is no support for leap seconds which are added (and in theory, removed) occasionally to compensate for earth rotation variation. This means that timespan calculation and free-running clocks may be inaccurate if they span leap seconds. To facilitate leap seconds, reference must be made to leap second table. This will not be done within the Sming framework and must be handled by application code if required.
32-bit signed values support a range of +/-68 years; the Unix epoch is midnight 1 Jan 1970, so overflows at about 3am on 19 Jan 2038. The value is signed which allows dates prior to 1970 to be represented.
An unsigned 32-bit value can be used to store dates provided they are after 1970. These are good until February 2106.
Note
time_t is a signed 64-bit value for all architectures except Windows Host and esp32 ESP-IDF 4.x.
Public Functions
-
inline DateTime()
Instantiate an uninitialised date and time object.
-
inline DateTime(time_t time)
Instantiate a date and time object from a Unix timestamp.
- Parameters:
time – Unix time to assign to object
-
inline operator time_t()
Get current Unix time.
- Return values:
time_t – Quantity of seconds since 00:00:00 1970-01-01
-
void setTime(time_t time)
Set time using Unix timestamp.
- Parameters:
time – Unix time to set object time to
-
inline void setTime(uint8_t sec, uint8_t min, uint8_t hour, uint8_t day, uint8_t month, uint16_t year)
Set time using time and date component values.
- Parameters:
sec – Seconds
min – Minutes
hour – Hour
day – Day of month
month – Month (0=Jan, 11=Dec)
year – Year
-
bool fromHttpDate(const String &httpDate)
Parse a HTTP full date and set time and date.
Note
Also supports obsolete RFC 850 date format, e.g. Sunday, 06-Nov-94 08:49:37 GMT where 2 digit year represents range 1970-2069
Note
GMT suffix is optional and is always assumed / ignored
- Parameters:
httpDate – HTTP full date in RFC 1123 format, e.g. Sun, 06 Nov 1994 08:49:37 GMT
- Return values:
bool – True on success
-
bool fromISO8601(const String &datetime)
Parse an ISO8601 date/time string.
Basic formatdoesn’t include separators, whereasExtended formatdoes.See also
Acceptable date formats:
Acceptable time formats:YYYY-MM-DD or YYYYMMDD YYYY-MM (but not YYYYMM)
Thh:mm:ss.sss or Thhmmss.sss Thh:mm:ss or Thhmmss Thh:mm.mmm or Thhmm.mmm Thh:mm or Thhmm Thh.hhh Thh
- Parameters:
datetime – Date and optional time in ISO8601 format, e.g. “1994-11-06”, “1994-11-06T08:49:37”. Separators are optional.
- Return values:
bool – True on success
-
bool isNull() const
Check if time date object is initialised.
- Return values:
True – if object has no value. False if initialised.
-
time_t toUnixTime() const
Get Unix time.
Note
Unix time does not account for leap seconds.
- Return values:
time_t – Unix time, quantity of seconds since 00:00:00 1970-01-01
-
String toShortDateString() const
Get human readable date.
- Return values:
String – Date in requested format, e.g. DD.MM.YYYY
-
String toShortTimeString(bool includeSeconds = false) const
Get human readable time.
- Parameters:
includeSeconds – True to include seconds (Default: false)
- Return values:
String – Time in format hh:mm or hh:mm:ss
-
String toFullDateTimeString() const
Get human readable date and time.
- Return values:
String – Date and time in format DD.MM.YYYY hh:mm:ss
-
String toISO8601() const
Get human readable date and time.
- Return values:
String – Date and time in format YYYY-MM-DDThh:mm:ssZ
-
String toHTTPDate() const
Get human readable date and time.
- Return values:
String – Date and time in format DDD, DD MMM YYYY hh:mm:ss GMT
-
void addMilliseconds(long add)
Add time to date time object.
- Parameters:
add – Quantity of milliseconds to add to object
-
String format(const char *formatString) const
Create string formatted with time and date placeholders.
Note
Uses strftime style formatting, e.g. format(“Today is %a, %d %b %Y”) returns “Today is Mon, 10 Dec 2018”
Note
Localisation may be implemented in libsming at compile time by setting LOCALE, e.g. LOCALE=LOCALE_DE_DE
Note
Default localisation is EN_GB
Note
Formatting parameters
Param
Description
Locale
%a
Abbreviated weekday name
*
%A
Full weekday name
*
%b
Abbreviated month name
*
%B
Full month name
*
%c
Locale preferred date and time format
*
%C
Century number (2 digits)
%d
Day of month as decimal number with leading zero (2 digits)
%D
US date format (mm/dd/yyyy)
%e
Day of month as decimal number with leading space (2 digits)
%F
ISO 8601 date format (YYYY-mm-dd)
%h
Equivalent to %b
*
%H
Hour as a decimal number using a 24-hour clock (range 00 to 23)
%I
Hour as a decimal number using a 12-hour clock (range 00 to 12)
%j
Day of the year as a decimal number (range 001 to 366)
%m
Month as a decimal number (range 01 to 12)
%M
Minute as a decimal number (range 00 to 59)
%n
Newline
%p
Meridiem indicator: AM or PM. Midnight is AM and noon is PM
%r
Locale 12-hour clock time notation. This is equivalent to %I:%M:%S %p
*
%R
Time in 24-hour notation (HH:MM)
%S
Second as a decimal number (range 00 to 60)
%t
Horizontal tab
%T
Time in 24-hour notation (HH:MM:SS)
%u
Day of the week as a decimal (range 1 to 7, Monday is 1)
%U
Week number as a decimal number (range 00 to 53, first Sunday as the first day of week 01)
%V
ISO 8601 week number as a decimal number (range 01 to 53, where week 1 is the first week including a Thursday)
%w
Day of the week as a decimal (range 0 to 6, Sunday is 0)
%W
Week number as a decimal number (range 00 to 53, first Monday as the first day of week 01)
%x
Locale preferred date representation
*
%X
Locale preferred time representation
*
%y
Year as a decimal number without a century (range 00 to 99)
%Y
Year as a decimal number (range 1970 to …)
%%
Percent sign
- Parameters:
formatString – String including date and time formatting
- Return values:
String – Formatted string
Public Members
-
uint16_t Year = 0
Full Year number.
-
uint16_t DayofYear = 0
Day of year (0-365)
-
uint8_t DayofWeek = 0
Day of week (0-6 Sunday is day 0)
-
uint8_t Month = 0
Month (0-11 Jan is month 0)
-
uint8_t Day = 0
Day of month (1-31)
-
uint8_t Hour = 0
Hour (0-23)
-
uint8_t Minute = 0
Minute (0-59)
-
uint8_t Second = 0
Second (0-59)
-
uint16_t Milliseconds = 0
Milliseconds (0-999)
Public Static Functions
-
static void fromUnixTime(time_t timep, uint8_t *psec, uint8_t *pmin, uint8_t *phour, uint8_t *pday, uint8_t *pwday, uint8_t *pmonth, uint16_t *pyear)
Convert from Unix time to individual time components.
Note
This is a more compact version of the C library localtime function
Note
Pass the Unix timedate value and pointers to existing integers. The integers are updated with the converted values
Note
This static function may be used without instantiating a DateTime object, e.g. DateTime::convertFromUnixTime(…);
Note
32-bit Unix time has year 2036 issue.
Note
Unix time does not account for leap seconds.
Note
All of the return values are optional, specify nullptr if not required
- Parameters:
timep – Unix time date value to convert
psec – Pointer to integer to hold resulting seconds
pmin – Pointer to integer to hold resulting minutes
phour – Pointer to integer to hold resulting hour
pday – Pointer to integer to hold resulting day of month
pwday – Pointer to integer to hold resulting day of week
pmonth – Pointer to integer to hold resulting month
pyear – Pointer to integer to hold resulting year
-
static time_t toUnixTime(int sec, int min, int hour, int day, uint8_t month, uint16_t year)
Convert from individual time components to Unix time.
Note
Seconds, minutes, hours and days may be any value, e.g. to calculate the value for 300 days since 1970 (epoch), set day=300
Note
This static function may be used without instantiating a DateTime object, e.g.
time_t unixTime = DateTime::toUnixTime(...);Note
Unix time does not account for leap seconds.
- Parameters:
sec – Seconds
min – Minutes
hour – Hours
day – Days
month – Month (0-11, Jan=0, Feb=1, …Dec=11), or enum (dtJanuary, …)
year – Year, either full 4 digit year or 2 digits for 2000-2068
- Return values:
time_t – Number of seconds since unix epoch (Midnight, Jan 1 1970)
-
inline DateTime()
System Clock
Enums
Variables
-
SystemClockClass SystemClock
Global instance of system clock object.
Note
Use SystemClock.function to access system clock functions
Note
Example:
SystemClock.setTimeZone(+9.5); //Darwin, Australia
-
class SystemClockClass
- #include <SystemClock.h>
Public Functions
-
time_t now(TimeZone timeType = eTZ_Local) const
Get the current date and time.
- Parameters:
timeType – Time zone to use (UTC / local)
- Return values:
DateTime – Current date and time
-
bool setTime(time_t time, TimeZone timeType)
Set the system clock’s time.
Note
System time is always stored as UTC time. If setting using the value retrieved from a time server using NTP, specify eTZ_UTC. If passing a local value using eTZ_Local, ensure that the time zone has been set correctly as it will be converted to UTC before storing.
- Parameters:
time – Unix time to set clock to (quantity of seconds since 00:00:00 1970-01-01)
timeType – Time zone of Unix time, i.e. is time provided as local or UTC?
-
String getSystemTimeString(TimeZone timeType = eTZ_Local) const
Get current time as a string.
Note
Date separator may be changed by adding
#define DT_DATE_SEPARATOR "/"to source code- Parameters:
timeType – Time zone to present time as, i.e. return local or UTC time
- Return values:
String – Current time in format:
dd.mm.yy hh:mm:ss
-
bool setTimeZoneOffset(int seconds)
Sets the local time zone offset.
- Parameters:
seconds – Offset from UTC of local time zone in seconds (-720 < offset < +720)
- Return values:
bool – true on success, false if value out of range
-
inline bool setTimeZone(float localTimezoneOffset)
Set the local time zone offset in hours.
- Parameters:
localTimezoneOffset – Offset from UTC of local time zone in hours (-12.0 < offset < +12.0)
- Return values:
bool – true on success, false if value out of range
-
inline int getTimeZoneOffset() const
Get the current time zone offset.
- Return values:
int – Offset in seconds from UTC
-
time_t now(TimeZone timeType = eTZ_Local) const
File System
Defines
-
CHECK_FS(_method)
Typedefs
-
using file_t = IFS::FileHandle
-
using FileHandle = IFS::FileHandle
-
using FileAttribute = IFS::FileAttribute
-
using FileAttributes = IFS::FileAttributes
Functions
-
inline IFS::FileSystem *getFileSystem()
Get the currently active file system, if any.
- Return values:
IFS::FileSystem* –
-
void fileSetFileSystem(IFS::IFileSystem *fileSystem)
Sets the currently active file system.
Note
Any existing filing system is destroyed. Typically the filing system implementation has helper functions which create and initialise the file system to a valid state. The last step is to call this function to make it active. Call this function with nullptr to deactivate the filing system.
- Parameters:
fileSystem –
-
inline void fileFreeFileSystem()
-
bool fileMountFileSystem(IFS::IFileSystem *fs)
Mount a constructed filesystem with debug messages.
-
bool fwfs_mount()
Mount the first available FWFS volume.
- Return values:
bool – true on success
-
bool fwfs_mount(Storage::Partition partition)
Mount SPIFFS volume from a specific partition.
- Return values:
bool – true on success
-
inline IFS::FileSystem *fileMountArchive(const String &filename)
Mount a backup archive.
- Parameters:
filename – Path to archive file
- Return values:
IFS::FileSystem* – Mounted filesystem. Caller must delete when finished.
-
template<typename T>
inline FileHandle fileOpen(const T &path, FileOpenFlags flags = File::ReadOnly) Open file by path.
- Parameters:
path – Full path to file
flags – Mode to open file
- Return values:
file – File handle or negative error code
-
inline int fileClose(FileHandle file)
Clode file.
Note
File Handle is returned from fileOpen function
- Parameters:
file – Handle of file to close
-
inline int fileWrite(FileHandle file, const void *data, size_t size)
Write to file.
- Parameters:
file – File handle
data – Pointer to data to write to file
size – Quantity of data elements to write to file
- Return values:
int – Quantity of data elements actually written to file or negative error code
-
inline int fileTouch(FileHandle file)
Update file modification time.
- Parameters:
file – File handle
- Return values:
int – Error code
-
inline int fileRead(FileHandle file, void *data, size_t size)
Read from file.
- Parameters:
file – File handle
data – Pointer to data buffer in to which to read data
size – Quantity of data elements to read from file
- Return values:
int – Quantity of data elements actually read from file or negative error code
-
inline file_offset_t fileSeek(FileHandle file, file_offset_t offset, SeekOrigin origin)
Position file cursor.
- Parameters:
file – File handle
offset – Quantity of bytes to move cursor
origin – Position from where to move cursor
- Return values:
file_offset_t – Offset within file or negative error code
-
inline bool fileIsEOF(FileHandle file)
Check if at end of file.
- Parameters:
file – File handle
- Return values:
bool – true if at end of file
-
inline file_offset_t fileTell(FileHandle file)
Get position in file.
- Parameters:
file – File handle
- Return values:
file_offset_t – Read / write cursor position or error code
-
inline int fileFlush(FileHandle file)
Flush pending writes.
- Parameters:
file – File handle
- Return values:
int – Size of last file written or error code
-
inline String fileGetErrorString(int err)
get the text for a returned error code
- Parameters:
err –
- Return values:
String –
-
template<typename TFileName>
inline int fileSetContent(const TFileName &fileName, const char *content, size_t length) Create or replace file with defined content.
Note
This function creates a new file or replaces an existing file and populates the file with the content of a c-string buffer.
- Parameters:
fileName – Name of file to create or replace
content – Pointer to c-string containing content to populate file with
length – Number of characters to write
- Return values:
int – Number of bytes transferred or error code
-
template<typename TFileName, typename TContent>
inline int fileSetContent(const TFileName &fileName, TContent content)
-
template<typename TFileName>
inline file_size_t fileGetSize(const TFileName &fileName) Get size of file.
- Parameters:
fileName – Name of file
- Return values:
file_size_t – Size of file in bytes, 0 on error
-
inline int fileTruncate(FileHandle file, file_size_t newSize)
Truncate (reduce) the size of an open file.
Note
In POSIX
ftruncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize- Parameters:
file – Open file handle, must have Write access
newSize –
- Return values:
int – Error code
-
inline int fileTruncate(FileHandle file)
Truncate an open file at the current cursor position.
- Parameters:
file – Open file handle, must have Write access
- Return values:
int – Error code
-
template<typename TFileName>
int fileTruncate(const TFileName &fileName, file_size_t newSize) Truncate (reduce) the size of a file.
Note
In POSIX
truncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize- Parameters:
fileName –
newSize –
- Return values:
int – Error code
-
inline int fileRename(const char *oldName, const char *newName)
Rename file.
- Parameters:
oldName – Original name of file to rename
newName – New name for file
- Return values:
int – Error code
-
template<typename TFileName>
String fileGetContent(const TFileName &fileName) Read content of a file.
Note
After calling this function the content of the file is placed in to a string. The result will be an invalid String (equates to
false) if the file could not be read. If the file exists, but is empty, the result will be an empty string “”.- Parameters:
fileName – Name of file to read from
- Return values:
String – String variable in to which to read the file content
-
template<typename TFileName>
inline size_t fileGetContent(const TFileName &fileName, char *buffer, size_t bufSize) Read content of a file.
Note
After calling this function the content of the file is placed in to a c-string Ensure there is sufficient space in the buffer for file content plus extra trailing null, i.e. at least bufSize + 1 Always check the return value!
Note
Returns 0 if the file could not be read
- Parameters:
fileName – Name of file to read from
buffer – Pointer to a character buffer in to which to read the file content
bufSize – Quantity of bytes to read from file
- Return values:
size_t – Quantity of bytes read from file or zero on failure
-
inline int fileStats(const char *fileName, FileStat &stat)
Get file statistics.
- Parameters:
name – File name
stat – Pointer to SPIFFS statistic structure to populate
- Return values:
int – Error code
-
inline int fileStats(FileHandle file, FileStat &stat)
brief Get file statistics
- Parameters:
file – File handle
stat – Pointer to SPIFFS statistic structure to populate
- Return values:
int – Error code
-
inline int fileDelete(const char *fileName)
Delete file.
- Parameters:
name – Name of file to delete
- Return values:
int – Error code
-
inline int fileDelete(FileHandle file)
Delete file.
- Parameters:
file – handle of file to delete
- Return values:
int – Error code
-
inline bool fileExist(const char *fileName)
Check if a file exists on file system.
- Parameters:
fileName – Full path to file to check for
- Return values:
bool – true if file exists
-
inline bool dirExist(const char *dirName)
Check if a directory exists on file system.
- Parameters:
dirName – Full path to directory to check for
- Return values:
bool – true if directory exists
-
inline int fileOpenDir(const char *dirName, DirHandle &dir)
Open a named directory for reading.
- Parameters:
name – Name of directory to open, empty or “/” for root
dir – Directory object
- Return values:
int – Error code
-
inline int fileCloseDir(DirHandle dir)
close a directory object
- Parameters:
dir – directory to close
- Return values:
int – Error code
-
inline int fileReadDir(DirHandle dir, FileStat &stat)
Read a directory entry.
- Parameters:
dir – The directory object returned by fileOpenDir()
stat – The returned information, owned by DirHandle object
- Return values:
int – Error code
-
inline int fileRewindDir(DirHandle dir)
Rewind to start of directory entries.
- Parameters:
dir – The directory object returned by fileOpenDir()
- Return values:
int – Error code
-
inline int fileGetSystemInfo(IFS::FileSystem::Info &info)
Get basic file system information.
- Return values:
int – Error code
-
IFS::FileSystem::Type fileSystemType()
Get the type of file system currently mounted (if any)
- Return values:
FileSystemType – the file system type
-
inline int fileSystemCheck()
Perform a consistency check/repair on the active file system.
- Return values:
int – 0 if OK, < 0 unrecoverable errors, > 0 repairs required
-
inline int fileSetACL(FileHandle file, const IFS::ACL &acl)
Set access control information.
- Parameters:
file – File handle
acl –
- Return values:
int – Error code
-
template<typename T>
int fileSetAttr(const T &filename, FileAttributes attr) Set file attributes.
- Parameters:
filename –
attr –
- Return values:
int – Error code
-
inline int fileSetTime(FileHandle file, time_t mtime)
Set access control information for file.
Note
any writes to file will reset this to current time
- Parameters:
file – handle to open file, must have write access
- Return values:
int – Error code
-
namespace SmingInternal
Variables
-
IFS::FileSystem *activeFileSystem
Global file system instance.
Filing system implementations should use helper functions to setup and initialise a filing system. If successful, call fileSetFileSystem() to make it active. This ensures that any active filing system is dismounted destroyed first.
-
IFS::FileSystem *activeFileSystem
Platform Support
WiFi Access Point
- group wifi_ap
Control and monitoring of WiFi access point interface.
- Todo:
How is wifi access point dhcp controlled?
See also
See also
Note
The WiFi access point interface provides a WiFi network access point. Control of WiFi AP including WiFi SSID and password and IP address.
Variables
-
AccessPointClass &WifiAccessPoint
Global instance of WiFi access point object.
Note
Use WifiAccessPoint.function to access WiFi access point functions
Note
Example:
if(WifiAccessPoint.config("ESP_AP", AUTH_OPEN)) WifiAccessPoint.enable(true);
-
class AccessPointClass
- #include <AccessPoint.h>
Access point class.
Public Functions
-
virtual void enable(bool enabled, bool save = false) = 0
Enable or disable WiFi AP.
- Parameters:
enabled – True to enable AP. False to disable.
save – True to save operational mode to flash, False to set current operational mode only
-
virtual bool isEnabled() const = 0
Get WiFi AP enable status.
- Return values:
bool – True if WiFi AP enabled.
-
virtual bool config(const String &ssid, String password, WifiAuthMode mode, bool hidden = false, int channel = 7, int beaconInterval = 200) = 0
Configure WiFi AP.
- Parameters:
ssid – WiFi AP SSID
password – WiFi AP password
mode – WiFi AP mode
hidden – True to hide WiFi AP (Default: Visible)
channel – WiFi AP channel (Default: 7)
beaconInterval – WiFi AP beacon interval in milliseconds (Default: 200ms)
- Return values:
bool – True on success
-
virtual IpAddress getIP() const = 0
Get WiFi AP IP address.
- Return values:
IpAddress – WiFi AP IP address
-
virtual bool setIP(IpAddress address) = 0
Set WiFi AP IP address.
- Parameters:
address – New IP address for WiFi AP
- Return values:
bool – True on success
-
virtual MacAddress getMacAddress() const = 0
Get WiFi AP MAC address.
- Return values:
MacAddress –
-
String getMAC(char sep = '\0') const
Get WiFi AP MAC address.
- Parameters:
sep – separator between bytes (e.g. ‘:’)
- Return values:
String – WiFi AP MAC address
-
virtual bool setMacAddress(const MacAddress &addr) const = 0
Set Access Point MAC address.
Must be called from
init()before activating Access Point. Espressif place certain limitations on MAC addresses:Bit 0 of the first byte of the MAC address can not be 1. For example:
OK: “1a:XX:XX:XX:XX:XX” NOT OK: “15:XX:XX:XX:XX:XX”
- Parameters:
addr – The new MAC address
- Return values:
bool – true on success
-
virtual IpAddress getNetworkMask() const = 0
Get WiFi AP network mask.
- Return values:
IpAddress – WiFi AP network mask
-
virtual IpAddress getNetworkGateway() const = 0
Get WiFi AP default gateway.
- Return values:
IpAddress – WiFi AP default gateway
-
virtual IpAddress getNetworkBroadcast() const = 0
Get WiFi AP broadcast address.
- Return values:
IpAddress – WiFi AP broadcast address
-
inline bool isLocal(IpAddress address)
Determine if the given address is on the same subnet.
Note
Use to prevent external access to services
- Parameters:
address –
- Return values:
bool – true if address is local
-
virtual String getSSID() const = 0
Get WiFi access point SSID.
- Return values:
String – WiFi access point SSID
-
virtual String getPassword() const = 0
Get WiFi access point password.
- Return values:
String – WiFi access point password
-
virtual std::unique_ptr<StationList> getStations() const = 0
Gets a list of stations connected to the access point.
- Return values:
StationList –
-
virtual void enable(bool enabled, bool save = false) = 0
Ethernet
Currently only supported on ESP32 using embedded MAC.
-
class EmbeddedEthernet : public Ethernet::IdfService
Ethernet provider using ESP32 embedded MAC. Requires an external PHY.
See https://docs.espressif.com/projects/esp-idf/en/v4.3/esp32/api-reference/network/esp_eth.html.
These are the reduced (RMII) PHY connections. Note that the high-speed signals may not be re-allocated.
Note: Configuring clock options must be done via SDK (make sdk-menuconfig). ESP-IDF v4.4 will add the ability to override these in software.Signal GPIO Direction ------ ---- --------- TXD0 19 OUT TXD1 22 OUT TX_EN 21 OUT RXD0 25 IN RXD1 26 IN CRS_DV 27 IN Receive Data Valid MDC 23 OUT: configurable MDIO 18 OUT: configurable CLK_MII 0 IN: No other pins supported OUT: alternate pins: 16, 17 50MHz clock provided either by the PHY or the MAC. Default is PHY.
The following connections are optional:
PHY_RESET - OUT (set via PhyConfig)
Public Functions
-
struct Config
-
struct Config
-
namespace Ethernet
Typedefs
-
using EventDelegate = Delegate<void(Ethernet::Event event)>
Delegate type for Ethernet events.
- Param event:
Which event occurred
-
using GotIpDelegate = Delegate<void(IpAddress ip, IpAddress netmask, IpAddress gateway)>
Delegate type for ‘got IP address’ event.
Enums
Variables
-
constexpr int8_t PIN_DEFAULT = {-2}
Use default pin for platform.
-
constexpr int8_t PIN_UNUSED = {-1}
Do not configure this pin.
Only applies if pin is optional, otherwise it will be interpreted as ‘auto detect’.
-
constexpr int8_t PHY_ADDR_AUTO = {-1}
Automatically detect PHY address during initialization.
-
class Dp83848 : public Ethernet::PhyFactory
- #include <Dp83848.h>
DP 83848 PHY interface.
-
class Ip101 : public Ethernet::PhyFactory
- #include <Ip101.h>
IP101 PHY interface.
-
class Ksz8041 : public Ethernet::PhyFactory
- #include <Ksz8041.h>
KSZ 8041 PHY interface.
-
class Lan8720 : public Ethernet::PhyFactory
- #include <Lan8720.h>
LAN 8720 PHY interface.
-
class Rtl8201 : public Ethernet::PhyFactory
- #include <Rtl8201.h>
RTL 8201 PHY interface.
-
struct PhyConfig
- #include <Ethernet.h>
PHY configuration.
Public Members
-
int8_t phyAddr = PHY_ADDR_AUTO
PHY address.
-
int8_t resetPin = PIN_UNUSED
Reset GPIO number */.
-
uint16_t resetTimeout = 100
Reset timeout value in milliseconds.
-
uint16_t autoNegTimeout = 4000
Auto-negotiation timeout in milliseconds.
-
int8_t phyAddr = PHY_ADDR_AUTO
-
class PhyFactory
- #include <Ethernet.h>
Virtual class used to construct a specific PHY instance.
Applications provide an instance of this factory class so that the Service can create and configure it at the correct point in initialisation or teardown.
Subclassed by Ethernet::DM9051PhyFactory, Ethernet::Dp83848, Ethernet::Ip101, Ethernet::Ksz8041, Ethernet::Lan8720, Ethernet::Rtl8201, Ethernet::W5500PhyFactory
-
class Service
- #include <Ethernet.h>
Abstract Service class.
Provides a common implementation for TCP/IP ethernet support.
An Ethernet interface requires a MAC layer plus PHY.
The ESP32, for example, contains a MAC but requires an external PHY. Other solutions, such as the W5500, contain MAC+PHY and require the correct PhyFactory to work.
Ethernet implementations should provide appropriate setup methods which are called by the application before invoking
begin().Subclassed by Ethernet::IdfService
Public Functions
-
virtual void end() = 0
Tear down the ethernet connection.
-
virtual MacAddress getMacAddress() const = 0
Get MAC address.
-
virtual bool setMacAddress(const MacAddress &addr) = 0
Set MAC address.
-
virtual bool setFullDuplex(bool enable) = 0
Set duplex mode of MAC.
-
virtual bool setLinkState(bool up) = 0
Set link status of MAC.
-
virtual bool setPromiscuous(bool enable) = 0
Set MAC promiscuous mode.
-
virtual IpAddress getIP() const = 0
Get current IP address.
-
virtual bool setIP(IpAddress address, IpAddress netmask, IpAddress gateway) = 0
Set static IP address.
-
virtual bool isEnabledDHCP() const = 0
Determine if DHCP is active for this interface.
-
virtual bool enableDHCP(bool enable) = 0
Enable/disable DHCP on this interface.
-
inline void onEvent(EventDelegate callback)
Set callback for ethernet events.
-
inline void onGotIp(GotIpDelegate callback)
Set callback for ‘station connected with IP address’ event.
-
virtual void end() = 0
-
class DM9051PhyFactory : public Ethernet::PhyFactory
- #include <DM9051.h>
-
class DM9051Service : public Ethernet::SpiService
- #include <DM9051.h>
Ethernet provider using W5500 SPI.
-
class SpiService : public Ethernet::IdfService
- #include <SpiService.h>
SPI ethernet provider.
Subclassed by Ethernet::DM9051Service, Ethernet::W5500Service
-
struct Config
- #include <SpiService.h>
-
struct Config
-
class W5500PhyFactory : public Ethernet::PhyFactory
- #include <W5500.h>
-
class W5500Service : public Ethernet::SpiService
- #include <W5500.h>
Ethernet provider using W5500 SPI.
-
class IdfService : public Ethernet::Service
- #include <IdfService.h>
Base Ethernet service for IDF SDK.
Subclassed by EmbeddedEthernet, Ethernet::SpiService
Public Functions
-
virtual void end() override
Tear down the ethernet connection.
-
virtual MacAddress getMacAddress() const override
Get MAC address.
-
virtual bool setMacAddress(const MacAddress &addr) override
Set MAC address.
-
virtual bool setFullDuplex(bool enable) override
Set duplex mode of MAC.
-
virtual bool setLinkState(bool up) override
Set link status of MAC.
-
virtual bool setPromiscuous(bool enable) override
Set MAC promiscuous mode.
-
virtual IpAddress getIP() const override
Get current IP address.
-
virtual bool setIP(IpAddress address, IpAddress netmask, IpAddress gateway) override
Set static IP address.
-
virtual bool isEnabledDHCP() const override
Determine if DHCP is active for this interface.
-
virtual bool enableDHCP(bool enable) override
Enable/disable DHCP on this interface.
-
virtual void end() override
-
using EventDelegate = Delegate<void(Ethernet::Event event)>
WiFi Station
- group wifi_sta
Control and monitoring of WiFi station interface.
See also
See also
Note
The WiFi station interface provides client access to a WiFi network. Control of WiFi connection including WiFi SSID and password and IP address, DHCP, etc.
Typedefs
-
using ScanCompletedDelegate = Delegate<void(bool success, BssList &list)>
Scan complete handler function.
-
using SmartConfigDelegate = Delegate<bool(SmartConfigEvent event, const SmartConfigEventInfo &info)>
Smart configuration handler function.
- Param event:
- Param info:
- Retval bool:
return true to perform default configuration
Enums
-
enum StationConnectionStatus
WiFi station connection states.
Values:
-
enumerator eSCS_Idle
Connection idle.
-
enumerator eSCS_Connecting
Connecting.
-
enumerator eSCS_WrongPassword
Wrong password.
-
enumerator eSCS_AccessPointNotFound
AP not found.
-
enumerator eSCS_ConnectionFailed
Connection failed.
-
enumerator eSCS_GotIP
Got IP address.
-
enumerator eSCS_Idle
-
enum SmartConfigType
Smart configuration type.
Values:
-
enumerator SCT_None
-
enumerator SCT_EspTouch
ESP Touch.
-
enumerator SCT_AirKiss
Air Kiss.
-
enumerator SCT_EspTouch_AirKiss
ESP Touch and Air Kiss.
-
enumerator SCT_EspTouch_V2
ESP Touch version 2.
-
enumerator SCT_None
Variables
-
StationClass &WifiStation
Global instance of WiFi station object.
Note
Use WifiStation.function to access WiFi station functions
Note
Example:
if(WifiStation.config("My_WiFi", "My_Password")) WifiStation.enable(true);
-
struct SmartConfigEventInfo
- #include <Station.h>
Smart Config callback information.
-
class StationClass
- #include <Station.h>
WiFi station class.
Public Functions
-
virtual void enable(bool enabled, bool save = false) = 0
Enable / disable WiFi station.
Note
Disabling WiFi station will also disable and clear the handler set with waitConnection.
- Parameters:
enabled – True to enable station. False to disable.
save – True to save operational mode to flash, False to set current operational mode only
-
virtual bool isEnabled() const = 0
Get WiFi station enable status.
- Return values:
bool – True if WiFi station enabled
-
virtual bool config(const Config &config) = 0
Configure WiFi station.
- Parameters:
ssid – WiFi SSID
password – WiFi password
autoConnectOnStartup – True to auto connect. False for manual. (Default: True)
save – True to save the SSID and password in Flash. False otherwise. (Default: True)
-
inline bool config(const String &ssid, const String &password, bool autoConnectOnStartup = true, bool save = true)
Configure WiFi station.
- Parameters:
ssid – WiFi SSID
password – WiFi password
autoConnectOnStartup – True to auto connect. False for manual. (Default: True)
save – True to save the SSID and password in Flash. False otherwise. (Default: True)
-
virtual bool connect() = 0
Connect WiFi station to network.
-
virtual bool disconnect() = 0
Disconnect WiFi station from network.
-
bool isConnected() const
Get WiFi station connectoin status.
- Return values:
bool – True if connected.
-
bool isConnectionFailed() const
Get WiFi station connection failure status.
- Return values:
bool – True if connection failed
-
virtual StationConnectionStatus getConnectionStatus() const = 0
Get WiFi station connection status.
- Return values:
StationConnectionStatus – Connection status structure
-
String getConnectionStatusName() const
Get WiFi station connection status name.
- Return values:
String – String representing connection status
-
virtual bool isEnabledDHCP() const = 0
Get WiFi station DHCP enabled status.
- Return values:
bool – True if DHCP enabled
-
virtual void enableDHCP(bool enable) = 0
Enable or disable WiFi station DHCP.
- Parameters:
enable – True to enable WiFi station DHCP
-
virtual void setHostname(const String &hostname) = 0
Set WiFi station DHCP hostname.
- Parameters:
hostname – - WiFi station DHCP hostname
-
virtual String getHostname() const = 0
Set WiFi station DHCP hostname.
- Return values:
WiFi – station DHCP hostname
-
virtual IpAddress getIP() const = 0
Get WiFi station IP address.
- Return values:
IpAddress – IP address of WiFi station
-
virtual MacAddress getMacAddress() const = 0
Get WiFi station MAC address.
- Return values:
MacAddress –
-
String getMAC(char sep = '\0') const
Get WiFi station MAC address.
- Parameters:
sep – Optional separator between bytes (e.g. ‘:’)
- Return values:
String – WiFi station MAC address
-
virtual bool setMacAddress(const MacAddress &addr) const = 0
Set WiFi station MAC address.
Must be called from
init()before activating station. Espressif place certain limitations on MAC addresses:Bit 0 of the first byte of the MAC address can not be 1. For example:
OK: “1a:XX:XX:XX:XX:XX” NOT OK: “15:XX:XX:XX:XX:XX”
- Parameters:
addr – The new MAC address
- Return values:
bool – true on success
-
virtual IpAddress getNetworkMask() const = 0
Get WiFi station network mask.
- Return values:
IpAddress – WiFi station network mask
-
virtual IpAddress getNetworkGateway() const = 0
Get WiFi station default gateway.
- Return values:
IpAddress – WiFi station default gateway
-
virtual IpAddress getNetworkBroadcast() const = 0
GetWiFi station broadcast address.
- Return values:
IpAddress – WiFi station broadcast address
-
inline bool isLocal(IpAddress address)
Determine if the given address is on the same subnet.
Note
Use to prevent external access to services
- Parameters:
address –
- Return values:
bool – true if address is local
-
bool setIP(IpAddress address)
Set WiFi station IP address.
- Parameters:
address – IP address
- Return values:
bool – True on success
-
virtual bool setIP(IpAddress address, IpAddress netmask, IpAddress gateway) = 0
Set WiFi station IP parameters.
- Parameters:
address – IP address
netmask – Network mask
gateway – Default gateway
- Return values:
bool – True on success
-
virtual String getSSID() const = 0
Get WiFi station SSID (Service Set Identifier)
- Return values:
String – WiFi station SSID
-
virtual MacAddress getBSSID() const = 0
Get BSSID (Basic Service Set Identifier) for connected AP.
- Return values:
MacAddress – Identifier of connected Access Point
-
virtual String getPassword() const = 0
Get WiFi station password.
- Return values:
String – WiFi station password
-
virtual int8_t getRssi() const = 0
Get WiFi signal strength.
- Return values:
int8_t – Value in dBm
-
virtual uint8_t getChannel() const = 0
Get active WiFi channel.
- Return values:
uint8_t – channel number
-
virtual bool startScan(ScanCompletedDelegate scanCompleted) = 0
Start WiFi station network scan.
- Parameters:
scanCompleted – Function to call when scan completes
- Return values:
bool – True on success
-
struct Config
- #include <Station.h>
Station configuration passed to config method.
-
virtual void enable(bool enabled, bool save = false) = 0
-
using ScanCompletedDelegate = Delegate<void(bool success, BssList &list)>
System
Task Queue
Sming has a task queue which allows execution of a function to be deferred until
the system is less busy. This is done by calling SystemClass::queueCallback().
Callbacks are executed as soon as possible, and allow other higher priority tasks (such as servicing the WiFi stack) to be handled in a timely manner.
A common use for the queue is to initiate processing from an interrupt service routine.
You must not spend too much time in the callback. How much time depends on the nature of your application, but tasks consuming more than 100ms will probably affect responsiveness and should be broken into smaller chunks. You might do this by wrapping such tasks in a class together with some state information. At the end of the initial callback if there is further work to be done simply make another call to queueCallback().
The task queue size is fixed, so the call to queueCallback() will fail if there is no room.
- TASK_QUEUE_LENGTH
Maximum number of entries in the task queue (default 10).
- ENABLE_TASK_COUNT
If problems are suspected with task queuing, it may be getting flooded. For this reason you should check the return value from queueCallback().
You can enable this option to keep track of the number of active tasks,
SystemClass::getTaskCount(), and the maximum,SystemClass::getMaxTaskCount().By default this is disabled and both methods will return 255. This is because interrupts must be disabled to ensure an accurate count, which may not be desirable.
API Documentation
- group system
Access to the ESP8266 system Provides system control and monitoring of the ESP8266.
Typedefs
-
using TaskCallback32 = void (*)(uint32_t param)
Task callback function type, uint32_t parameter.
Note
Callback code does not need to be in IRAM
-
using TaskCallback = void (*)(void *param)
Task callback function type, void* parameter.
Note
Callback code does not need to be in IRAM
-
using TaskDelegate = Delegate<void()>
Task Delegate callback type.
-
using SystemReadyDelegate = TaskDelegate
Handler function for system ready.
Enums
-
enum CpuFrequency
Common CPU frequencies.
Values:
-
enumerator eCF_80MHz
-
enumerator eCF_125MHz
-
enumerator eCF_133MHz
-
enumerator eCF_160MHz
-
enumerator eCF_240MHz
-
enumerator eCF_80MHz
-
enum DeepSleepOptions
Deep sleep options.
Values:
-
enumerator eDSO_RF_CAL_BY_INIT_DATA
RF_CAL or not after deep-sleep wake up, depends on init data byte 108.
-
enumerator eDSO_RF_CAL_ALWAYS
RF_CAL after deep-sleep wake up, there will be large current.
-
enumerator eDSO_RF_CAL_NEVER
no RF_CAL after deep-sleep wake up, there will only be small current.
-
enumerator eDSO_DISABLE_RF
disable RF after deep-sleep wake up, just like modem sleep, there will be the smallest current.
-
enumerator eDSO_RF_CAL_BY_INIT_DATA
Variables
-
SystemClass System
Global instance of system object.
Note
Use system.function to access system functions
Note
Example:
system.reset();
-
class ISystemReadyHandler
- #include <System.h>
Interface class implemented by classes to support on-ready callback.
Subclassed by WDTClass, WifiSniffer
Public Functions
-
virtual void onSystemReady() = 0
Handle system ready events.
-
virtual void onSystemReady() = 0
-
class SystemClass
- #include <System.h>
System class.
Public Functions
-
inline bool isReady()
Check if system ready.
- Return values:
bool – True if system initialisation is complete and system is now ready
-
void restart(unsigned deferMillis = 0)
Request a restart of the system.
Note
A delay is often required to allow network callback code to complete correctly. The restart is always deferred, either using the task queue (if deferMillis == 0) or using a timer. This method always returns immediately.
- Parameters:
deferMillis – defer restart request by a number of milliseconds
-
inline bool setCpuFrequency(CpuFrequency freq)
Set the CPU frequency.
- Parameters:
freq – Frequency to set CPU
- Return values:
bool – true on success
-
inline CpuFrequency getCpuFrequency()
Get the CPU frequency.
- Return values:
CpuFrequency – The frequency of the CPU
-
bool deepSleep(uint32_t timeMilliseconds, DeepSleepOptions options = eDSO_RF_CAL_BY_INIT_DATA)
Enter deep sleep mode. Deep sleep turns off processor and keeps only the RTC memory active.
Note
Determine reset cause like this:
auto info = system_get_rst_info(); if(info->reason == REASON_DEEP_SLEEP_AWAKE) { // ... }
Note
ESP8266: Ensure GPIO 16 (XPD_DCDC) is connected to RST (EXT_RSTB). and call pinMode(16, WAKEUP_PULLUP) to enable wakeup from deep sleep.
- Parameters:
timeMilliseconds – Quantity of milliseconds to remain in deep sleep mode
options – Deep sleep options
-
inline void onReady(SystemReadyDelegate readyHandler)
Set handler for system ready event.
Note
if system is ready, callback is executed immediately without deferral
- Parameters:
readyHandler – Function to handle event
-
inline void onReady(ISystemReadyHandler *readyHandler)
Set handler for system ready event.
Note
if system is ready, callback is executed immediately without deferral
- Parameters:
readyHandler – Function to handle event
Public Static Functions
-
static bool initialize()
System initialisation.
Note
Called by user_main: applications should not call this function or the task queue will be re-initialised and any currently queued tasks won’t be called.
- Return values:
bool – true on success
-
static bool queueCallback(TaskCallback32 callback, uint32_t param = 0)
Queue a deferred callback.
Note
It is important to check the return value to avoid memory leaks and other issues, for example if memory is allocated and relies on the callback to free it again. Note also that this method is typically called from interrupt context so must avoid things like heap allocation, etc.
- Parameters:
callback – The function to be called
param – Parameter passed to the callback (optional)
- Return values:
bool – false if callback could not be queued
-
static inline bool queueCallback(TaskCallback callback, void *param = nullptr)
Queue a deferred callback, with optional void* parameter.
-
static inline bool queueCallback(InterruptCallback callback)
Queue a deferred callback with no callback parameter.
-
static bool queueCallback(TaskDelegate callback)
Queue a deferred Delegate callback.
Note
Provides flexibility and ease of use for using capturing lambdas, etc. but requires heap allocation and not as fast as a function callback. DO NOT use from interrupt context, use a Task/Interrupt callback.
- Parameters:
callback – The Delegate to be called
- Return values:
bool – false if callback could not be queued
-
static inline unsigned getTaskCount()
Get number of tasks currently on queue.
- Return values:
unsigned –
-
static inline unsigned getMaxTaskCount()
Get maximum number of tasks seen on queue at any one time.
Note
If return value is higher than maximum task queue TASK_QUEUE_LENGTH then the queue has overflowed at some point and tasks have been left un-executed.
- Return values:
unsigned –
-
inline bool isReady()
-
using TaskCallback32 = void (*)(uint32_t param)
WiFi
Build variables
- ENABLE_WPS
Set to 1 to enable WiFi Protected Setup (WPS).
WPS is not enabled by default to preserve resources, and because there may be security implications which you should consider carefully.
- ENABLE_SMART_CONFIG
Set to 1 to enable WiFi Smart Configuration API.
SmartConfig requires extra libraries and
ENABLE_ESPCONN.See Basic Smart Config sample application.
If you want to provide a default SSID and Password for connection to your default Access Point, you can do this:
make WIFI_SSID=MyAccessPoint WIFI_PWD=secret
These are provided as #defined symbols for your application to use. See Basic WiFi for a simple example, or MeteoControl for a more flexible solution using configuration files.
- WIFI_SSID
SSID identifying default Access Point to connect to. By default, this is undefined.
API Documentation
WiFi Events
- group wifi_ev
Event callback interface for WiFi events.
See also
See also
Defines
-
WIFI_DISCONNECT_REASON_CODES_MAP(XX)
Common set of reason codes to IEEE 802.11-2007.
Some acronymns used here - see the full standard for more precise definitions.
SSID: Service Set Identifier (the visible name given to an Access Point)
BSSID: Basic Service Set Identifier (a MAC address physically identifying the AP)
IE: Information Element (standard piece of information carried within WiFi packets)
STA: Station (any device which supports WiFi, including APs though the term commonly refers to a client)
AP: Access Point (device to which other stations may be associated)
RSN: Robust Security Network
AUTH: Authentication (how a station proves its identity to another)
Note
Codes at 200+ are non-standard, defined by Espressif.
Typedefs
-
using StationConnectDelegate = Delegate<void(const String &ssid, MacAddress bssid, uint8_t channel)>
Delegate type for ‘station connected’ event.
Note
This event occurs when the station successfully connects to the target AP. Upon receiving this event, the DHCP client begins the process of getting an IP address.
- Param ssid:
- Param bssid:
- Param channel:
-
using StationDisconnectDelegate = Delegate<void(const String &ssid, MacAddress bssid, WifiDisconnectReason reason)>
Delegate type for ‘station disconnected’ event.
Note
This event can be generated in the following scenarios:
When the station is already connected to the AP, and a manual disconnect or re-configuration is taking place. e.g.
WifiStation.disconnect()When
WifiStation.connect()is called, but the Wi-Fi driver fails to set up a connection with the AP due to certain reasons, e.g. the scan fails to find the target AP, authentication times out, etc. If there are more than one AP with the same SSID, the disconnected event is raised after the station fails to connect all of the found APs.When the Wi-Fi connection is disrupted because of specific reasons, e.g., the station continuously loses N beacons, the AP kicks off the station, the AP’s authentication mode is changed, etc.
- Param ssid:
SSID from which we’ve disconnected
- Param bssid:
- Param reason:
Why the connection was dropped
-
using StationAuthModeChangeDelegate = Delegate<void(WifiAuthMode oldMode, WifiAuthMode newMode)>
Delegate type for ‘station authorisation mode changed’ event.
Note
This event arises when the AP to which the station is connected changes its authentication mode, e.g., from ‘no auth’ to WPA. Generally, the application event callback does not need to handle this.
- Param oldMode:
- Param newMode:
-
using StationGotIPDelegate = Delegate<void(IpAddress ip, IpAddress netmask, IpAddress gateway)>
Delegate type for ‘station got IP address’ event.
Note
This event arises when the DHCP client successfully gets the IPV4 address from the DHCP server, or when the IPV4 address is changed. The IPV4 may be changed because of the following reasons:
The DHCP client fails to renew/rebind the IPV4 address, and the station’s IPV4 is reset to 0.
The DHCP client rebinds to a different address.
The static-configured IPV4 address is changed.
- Param ip:
- Param netmask:
- Param gateway:
-
using AccessPointConnectDelegate = Delegate<void(MacAddress mac, uint16_t aid)>
Delegate type for ‘Access Point Connect’ event.
Note
This event occurs every time a station is connected to our Access Point.
- Param mac:
MAC address of the station
- Param aid:
Association ID representing the connected station
-
using AccessPointDisconnectDelegate = Delegate<void(MacAddress mac, uint16_t aid)>
Delegate type for ‘Access Point Disconnect’ event.
Note
This event occurs every time a station is disconnected from our Access Point.
- Param mac:
MAC address of the station
- Param aid:
Association ID assigned to the station
-
using AccessPointProbeReqRecvedDelegate = Delegate<void(int rssi, MacAddress mac)>
Delegate type for ‘Access Point Probe Request Received’ event.
Note
Probe Requests are a low-level management frame which are used to determine information about our Access Point, such as which authentication modes are supported.
- Param rssi:
Signal strength
- Param mac:
Enums
-
enum WifiDisconnectReason
Reason codes for WiFi station disconnection.
See also
Values:
Variables
-
WifiEventsClass &WifiEvents
Global reference to architecture-specific implementation.
-
class WifiEventsClass
- #include <WifiEvents.h>
WiFi events class.
Public Functions
-
inline void onStationConnect(StationConnectDelegate delegateFunction)
Set callback for ‘station connected’ event.
-
inline void onStationDisconnect(StationDisconnectDelegate delegateFunction)
Set callback for ‘station disconnected’ event.
-
inline void onStationAuthModeChange(StationAuthModeChangeDelegate delegateFunction)
Set callback for ‘station authorisation mode change’ event.
-
inline void onStationGotIP(StationGotIPDelegate delegateFunction)
Set callback for ‘station connected with IP address’ event.
-
inline void onAccessPointConnect(AccessPointConnectDelegate delegateFunction)
Set callback for ‘access point client connected’ event.
-
inline void onAccessPointDisconnect(AccessPointDisconnectDelegate delegateFunction)
Set callback for ‘access point client disconnected’ event.
-
inline void onAccessPointProbeReqRecved(AccessPointProbeReqRecvedDelegate delegateFunction)
Set callback for ‘access point probe request received’ event.
Public Static Functions
-
static String getDisconnectReasonName(WifiDisconnectReason reason)
Get short name for disconnection reason.
-
static String getDisconnectReasonDesc(WifiDisconnectReason reason)
Get descriptive explanation for disconnect reason.
-
inline void onStationConnect(StationConnectDelegate delegateFunction)
-
WIFI_DISCONNECT_REASON_CODES_MAP(XX)
WiFi Sniffer
See WiFi Sniffer for an example of how to use this.
API Documentation
- group wifi_sniffer
WiFi promiscuous mode sniffer support.
Defines
-
ETH_MAC_LEN
Typedefs
-
using BeaconInfoList = BeaconOrClientListTemplate<BeaconInfo>
For applications to use to manage list of unique beacons.
-
using ClientInfoList = BeaconOrClientListTemplate<ClientInfo>
For applications to use to manage list of unique clients.
-
using WifiSnifferCallback = Delegate<void(uint8_t *data, uint16_t length)>
-
using WifiBeaconCallback = Delegate<void(const BeaconInfo &beacon)>
-
using WifiClientCallback = Delegate<void(const ClientInfo &client)>
-
struct BeaconInfo
- #include <WifiSniffer.h>
Decoded Wifi beacon (Access Point) information.
-
struct ClientInfo
- #include <WifiSniffer.h>
Decoded Wifi client information.
-
class WifiSniffer : public ISystemReadyHandler
- #include <WifiSniffer.h>
Public Functions
-
void begin()
Initialise the sniffer.
-
void end()
Stop the sniffer.
-
inline void onBeacon(WifiBeaconCallback callback)
Register notification for beacon (AP) info.
-
inline void onClient(WifiClientCallback callback)
Register notification for client info.
-
inline void onSniff(WifiSnifferCallback callback)
Register notification for all incoming data.
Note
Callback invoked for all packet types, including beacon/client
-
inline void setChannel(unsigned channel)
Set the channel to listen on.
- Parameters:
channel –
-
inline unsigned getChannel()
Get the current channel being listened on.
-
void begin()
-
ETH_MAC_LEN
Services
Profiling
CPU Usage
Class to provide a CPU usage indication based on task callback availability. To produce useful results it requires a couple of seconds at startup to calibrate.
Use like this:
#include <Services/Profiling/CpuUsage.h>
// instantiate a single instance of :cpp:class:`Profiling::CpuUsage`
Profiling::CpuUsage cpuUsage;
// Will be called when CpuUsage calibration has completed
void onReady()
{
// Continue with application initialisation
}
void init()
{
// Begin clock calibration
cpuUsage.begin(onReady);
}
See Basic Tasks for a more detailed example.
CPU usage is calculated over an update period which begins with a call to Profiling::CpuUsage::reset().
The actual update period must be managed elsewhere, using a callback timer, web request
or other mechanism. It doesn’t need to be exact as the actual elapsed time in CPU
cycles is used for the calculation.
After the update period has elapsed, call Profiling::CpuUsage::getUtilisation() to obtain a CPU usage figure.
This figure is obtained using the number of task callbacks made within the update period.
- loop cycles
Set up repeating task callback and measure invocations between successive calls
- total cycles
The total number of CPU cycles between calls to
Profiling::CpuUsage::update().- used
total - loop
- utilisation
used / total
-
class CpuUsage
Class to provide a CPU usage indication based on task callback availability.
Public Functions
-
inline void begin(InterruptCallback ready)
Calibrate the baseline figure for minimum CPU usage.
Note
Typically call this in
init()- Parameters:
ready – Function to call when calibration is complete
-
inline void reset()
Reset counters to start a new update period.
-
inline unsigned getLoopIterations()
Get the number of task callbacks made so far.
-
inline uint32_t getElapsedCycles()
Get the total number of CPU cycles since the last call to reset()
-
inline uint32_t getMinLoopCycles()
Get the figure used as the baseline cycle count.
-
inline unsigned getUtilisation()
Get the CPU utilisation figure in 1/100ths of a percent.
-
inline void begin(InterruptCallback ready)
Min-Max
-
template<typename T>
class MinMax Class to track minimum and maximum values of a set of data, with average, total and count.
Subclassed by Profiling::MinMaxTimes< Timer >
Min-Max Times
TaskStat
Example of use:
#include <Services/Profiling/TaskStat.h>
Profiling::TaskStat taskStat(Serial);
Timer statTimer;
void init()
{
...
// Set up timer to print task information to Serial terminal every 2 seconds
statTimer.initializeMs<2000>(InterruptCallback([]() { taskStat.update(); }));
statTimer.start();
}
-
class TaskStat
Helper class to support printing of real-time task information.
Code is refactored from the FreeRTOS Real Time Stats Example.
Requires these SDK configuration settings to be set:
FREERTOS_USE_TRACE_FACILITY
FREERTOS_GENERATE_RUN_TIME_STATS
FREERTOS_VTASKLIST_INCLUDE_COREID (optional)
Wiring
The Wiring Framework is common to all Arduino platforms.
Some adaptations and improvements have been made for Sming which are documented here.
Vector
-
template<typename Element>
class Vector : public Countable<Element> Vector class template.
Public Functions
-
bool setSize(unsigned int newSize)
Reduce or increase number of items.
If increasing number of items, new items will be set to current
nilvalue. If reducing number of items, old items will be deleted.- Return values:
true – on success, false on memory reallocation failure
-
inline void trimToSize()
Reduce capacity to match current size.
-
template<bool is_const>
class Iterator
-
bool setSize(unsigned int newSize)
HashMap
-
template<typename K, typename V>
class HashMap HashMap class template.
Public Types
-
using SortCompare = bool (*)(const ElementConst &e1, const ElementConst &e2)
Return true if key1 < key2.
Public Functions
-
void sort(SortCompare compare)
Sort map entries.
-
template<bool is_const>
struct BaseElement
-
template<bool is_const>
class Iterator
-
using SortCompare = bool (*)(const ElementConst &e1, const ElementConst &e2)
Wiring String
Small String Optimisation
The String class is probably the most used class in Arduino and Sming.
Unfortunately it gets the blame for one of the most indidious problems in the embedded world, heap fragmentation.
To alleviate this problem, Sming uses a technique known as Small String Optimisation, which uses the available space inside the String object itself to avoid using the heap for small allocations of 10 characters or fewer.
This was lifted from the Arduino Esp8266 core. Superb work - thank you!
Configuration Variables
- STRING_OBJECT_SIZE
minimum: 12 bytes (default) maximum: 128 bytes
Must be an integer multiple of 4 bytes.
This is an experimental feature which lets you increase the size of a String object to reduce heap allocations further. The effect of this will vary depending on your application, but you can see some example figures in Pull Request #1951.
Benefits of increasing
STRING_OBJECT_SIZE:Increase code speed
Fewer heap allocations
Drawbacks:
Increased static memory usage for global/static String objects or embedded within global/static class instances.
A String can use SSO or the heap, but not both together, so in heap mode any additional SSO space will remain unused.
Allows the size of a String object to be changed to increase the string length available before the heap is used.
Note
The current implementation uses one byte for a NUL terminator, and another to store the length, so the maximum SSO string length is (STRING_OBJECT_SIZE - 2) characters.
However, the implementation may change so if you need to check the maximum SSO string size in your code, please use
String::SSO_CAPACITY.
API Documentation
-
class String
The String class.
Note that a string object’s default constructor creates an empty string. This is not the same as a null string. A null string evaluates to false, but an empty string evaluates to true.
Small String Optimisation means that heap is only used for strings longer than 10 characters, not including the NUL terminator. This is simply making use of existing storage within the String object.
This length can be increased using STRING_OBJECT_SIZE, but note the additional space remains unused when switching to heap storage for longer Strings.
Subclassed by CStringArray, StringSumHelper
Copy constructors
If the initial value is null or invalid, or if memory allocation fails, the string will be marked as invalid (i.e. “if (s)” will be false).
Copy operators
If the value is null or invalid, or if the memory allocation fails, the String will be marked as invalid (“if (s)” will be false).
Move operators
Move content from one String to another without any heap allocation.
Move operators are automatically selected by the compiler when it is able, such as when returning temporary String objects from functions.
In other situations, use
std::move:String original("A String"); String copy("This is the content for the copy"); copy = std::move(myString);
copywill now contain “A String”, whilstoriginalwill be invalidated.Concatenation methods
Works with built-in types. On failure, the string is left unchanged. If the argument is null or invalid, the concatenation is considered unsuccessful.
- retval bool:
true on success, false on failure
Concatenation operators
If there’s not enough memory for the concatenated value, the string will be left unchanged (but this isn’t signalled in any way)
Comparison methods
Works with String
and ‘c’ string
Comparisons are case-sensitive, binary comparison null strings (including cstr == nullptr) are treated as empty.
- retval int:
Returns < 0 if String is lexically before the argument, > 0 if after or 0 if the same
Test for equality
Compares content byte-for-byte using binary comparison
null strings (including cstr == nullptr) are treated as empty.
- retval bool:
Returns true if strings are identical
Equality operator ==
- retval bool:
true if Strings are identical
In-equality operator !=
- retval bool:
Returns true if strings are not identical
Test for equality, without case-sensitivity
null strings are treated as empty.
- retval bool:
true if strings are considered the same
Array operators
If index is invalid, returns NUL \0
int indexOf(…)
Locate a character or String within another String
.
By default, searches from the beginning of the
String, but can also start from a given index, allowing for the locating of all instances of the character or String.- retval int:
Index if found, -1 if not found
int lastIndexOf(…)
Locate a character or String within another String By default, searches from the end of the String, but can also work backwards from a given index, allowing for the locating of all instances of the character or String.
- retval int:
Index if found, -1 if not found
String substring(…)
Get a substring of a String
.
The starting index is inclusive (the corresponding character is included in the substring), but the optional ending index is exclusive (the corresponding character is not included in the substring).
- param from:
Index of first character to retrieve
- param to:
(optional) One-past the ending character to retrieve
If the ending index is omitted, the substring continues to the end of the String.
If you don’t need the original String, consider using remove() instead:
This produces the same result:String original("This is the original string."); String sub = original.substring(0, 13);
original.remove(13);
replace(…)
Replace all instances of a given character or substring with another character or substring.
Replacing a single character always succeeds as this is handled in-place.
- retval bool:
true on success, false on allocation failure
Where
replaceis longer thanfindthe String may need to be re-allocated, which could fail. If this happens the method returns false and the String is left unchanged.remove()
Remove characters from a String
.
If no count is provided then all characters from the given index to the end of the
String are removed.Note
The String is modified in-situ without any reallocation
- param index:
Index of the first character to remove
- param count:
Number of characters to remove
Pad string to a minimum length
This is used, for example, when outputting tabular data. The string is modified in-situ to minimise memory reallocations.
Methods may be chained like this::
Serial << String(value).padLeft(10, '.') << endl;
-
inline String &padLeft(uint16_t minWidth, char c = ' ')
Insert padding at start of string if length is less than given width.
Public Functions
-
bool reserve(size_t size)
Pre-allocate String memory.
On failure, the String is left unchanged. reserve(0), if successful, will validate an invalid string (i.e., “if (s)” will be true afterwards)
- Parameters:
size –
- Return values:
bool – true on success, false on failure
-
bool setLength(size_t length)
set the string length accordingly, expanding if necessary
Note
extra characters are undefined
- Parameters:
length – required for string (nul terminator additional)
- Return values:
true – on success, false on failure
-
bool setBuffer(const Buffer &buffer)
Set String content using move semantics from external memory buffer.
Note
length MUST be <
size- A NUL character is written at this location- Parameters:
buffer – We’ll take ownership of this buffer
- Return values:
bool – true on success; on failure, ownership of buffer is not transferred
-
Buffer getBuffer()
Get String content using move semantics.
Note
String is invalidated by this call. Caller is responsible for buffer memory.
- Return values:
Buffer –
-
inline operator StringIfHelperType() const
Provides safe bool() operator.
Evaluates as false if String is null, otherwise evaluates as true
-
inline bool startsWith(const String &prefix) const
Compare the start of a String Comparison is case-sensitive, must match exactly.
- Parameters:
prefix –
- Return values:
bool – true on match
-
bool startsWith(const String &prefix, size_t offset) const
Compare a string portion.
mis-named as does not necessarily compare from start
Note
Comparison is case-sensitive, must match exactly
- Parameters:
prefix –
offset – Index to start comparison at
- Return values:
bool – true on match
-
bool endsWith(char suffix) const
Compare the end of a String.
- Parameters:
suffix –
- Return values:
bool – true on match
-
bool endsWith(const String &suffix) const
Compare the end of a String.
- Parameters:
suffix –
- Return values:
bool – true on match
-
inline char charAt(size_t index) const
Obtain the character at the given index.
Note
If index is invalid, returns NUL \0
- Parameters:
index –
- Return values:
char –
-
void setCharAt(size_t index, char c)
Sets the character at a given index.
Note
If index is invalid, does nothing
- Parameters:
index –
c –
-
size_t getBytes(unsigned char *buf, size_t bufsize, size_t index = 0) const
Read contents of a String into a buffer.
Note
Returned data always nul terminated so buffer size needs to take this into account
- Parameters:
buf – buffer to write data
bufsize – size of buffer in bytes
index – offset to start
- Return values:
unsigned – number of bytes copied, excluding nul terminator
-
inline void toCharArray(char *buf, size_t bufsize, size_t index = 0) const
Read contents of String into a buffer.
See also
See
getBytes()
-
inline const char *c_str() const
Get a constant (un-modifiable) pointer to String content.
- Return values:
const – char* Always valid, even for a null string
-
inline char *begin()
Get a modifiable pointer to String content.
Note
If String is NUL, returns nullptr.
Public Static Attributes
-
static constexpr size_t SSO_CAPACITY = STRING_OBJECT_SIZE - 2
Max chars. (excluding NUL terminator) we can store in SSO mode.
-
struct Buffer
Used with setBuffer and getBuffer methods.
Sample Projects
Sming comes with a set of samples to illustrate and test various aspects of the framework. Some are related to specific Libraries.
MPU6050 Six-Axis (Gyro + Accelerometer)
MPU6050 sensor reader.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MMA7455 Accelerometer
Simple demonstration which reads and displays output from an I2C accelerometer.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic APA102
Demonstrates use of APA102 library for controlling smart LEDs via SPI.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic AWS
How to set up Amazon Web Services to use as an IOT gateway.
openssl rsa -in APP.private.key -out APP.private.key.der -outform DER openssl x509 -in APP.cert.pem -out APP.cert.der -outform DER
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Audio
Demonstrates how to set up I2S driver to produce audio output.
References
SoC support
esp8266
Basic Blink
Simple blink example to confirm that the basic build system is working with your system.
We use Timer instead of a loop because we want to allow WiFi communications to work in the background. See Multitasking.
The LED on many development boards is connected to GPIO2, so this is the default.
If you get no response then check the documentation or schematic as your system may differ and change the LED_PIN definition accordingly.
For example, the NodeMCU ESP-C3 kits have an RGB LED connected to GPIO 3, 4 & 5.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Capsense
Simple demonstration showing raw output from a capacitive sensor.
References
SoC support
esp8266
Basic DateTime
A basic demonstration of Sming’s date and time handling functions.
Connect a serial terminal and enter a unix timestamp as a numeric value, then press ENTER to display it in various formats.
You can delete characters, or press ESC to clear the entry.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Delegates
Demonstrates the various ways you can use callbacks within Sming.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Ethernet
Demonstrates creating an ethernet connection.
Currently only supported for ESP32 using C++ wrappers around the ESP-IDF implementation.
See https://docs.espressif.com/projects/esp-idf/en/v4.3/esp32/api-reference/network/esp_eth.html.
Embedded MAC
The standard ESP32 contains an embedded ethernet MAC but requires an external PHY. This demonstration uses a commonly available LAN8270 module.
See Components/Network/Arch/Esp32/include/Platform/EmbeddedEthernet.h for connections.
SPI ethernet
SPI connections depend on the device variant being used. See Sming/Libraries/HardwareSPI/src/Arch/Esp32/Controller.cpp for details.
See Sming/Components/Network/Arch/Esp32/Network/W5500.cpp for additional pin connections for W5500-based devices.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
Basic Hardware PWM
Demonstrates how to generate PWM signals on multiple pins using Sming’s HardwarePWM class.
References
SoC support
esp8266
Basic IFS
Simple Webserver demonstration using IFS.
View the filesystem using a web browser.
To see directory content in a different format, append ?format=XX, one of:
jsontexthtml
Use the format archive to retrieve an archive/backup of the directory tree as an FWFS image.
Building
By default, data is stored in a read-only FWFS (Firmware Filesystem) partition.
This sample also demonstrates how to store the data in a FlashString object:
make config-clean
make ENABLE_FLASHSTRING_IMAGE=1
Because the data is linked into the program image this is only suitable for small filesystem images. This could be used to store default recovery data, especially with OTA updates because each program image is self-contained.
To add support for SD Cards to this sample:
make ENABLE_SDCARD=1
See FatIFS for further details of SD Card and FAT filing system support.
To add support for a USB storage device:
make ENABLE_USB_STORAGE=1
References
Environment Variables
ENABLE_FLASHSTRING_IMAGEENABLE_SDCARDENABLE_USB_STORAGE
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Interrupts
Simple example of how interrupts can be used within Sming.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic NFC
Sample application to read an RFID tag using a suitable reader connected via SPI interface.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Neopixel
Demonstration of the Adafruit Neopixel library for controlling single-wire-based LED pixels and strip.
References
SoC support
esp32
esp32c3
esp32s2
esp32s3
esp8266
rp2040
Basic Ota
Introduction
This sample integrates Over-The-Air(OTA) Upgrader, Over-The-Air(OTA) Network Upgrader and Sming. It demonstrates dual rom booting, big flash support, OTA updates and dual spiffs filesystems. This sample should work on all supported architectures.
Esp8266
On Esp8266 we use rBoot as bootloader. When using rBoot big flash support with multiple 1MB slots only one rom image needs to be created. If you don’t want to use big flash support (e.g. for a device with smaller flash) see the separate instructions below. You can easily take the ota files and add them to your own project to add OTA support.
Building
Set
WIFI_SSID&WIFI_PWDenvironment variables with your wifi details.Edit the OTA server details defined in the application
component.mkfile.make && make flashPut rom0.bin and spiff_rom.bin in the root of your webserver for OTA.
Interact with the sample using a terminal (
make terminal). Sorry - no web-gui (yet).
Testing
For testing purposes we provide an Ota server that can be started on your desktop machine:
make otaserver
The server listens on port 9999 and all network interfaces. If your desktop has the following IP address 192.168.1.30
after connecting to your WIFI router then you can compile the sample to use this IP address and the testing OTA server:
make ROM_0_URL=http://192.168.1.30:9999/rom0.bin SPIFFS_URL=http://192.168.1.30:9999/spiff_rom.bin
make flash
Make sure to replace 192.168.1.30 with your WIFI IP address.
Flash layout considerations
Esp8266
If you want to use, for example, two 512k roms in the first 1MB block of flash (old style) then Sming will automatically create two separately linked roms. If you are flashing a single rom to multiple 1MB flash blocks, all using the same offset inside their 1MB blocks, only a single rom is created. See rBoot for further details.
If using a very small flash (e.g. 512k) there may be no room for a spiffs filesystem, so use HWCONFIG = standard
After building copy all the rom*.bin files to the root of your web server.
If you want more than two roms you must be an advanced user and should be able to work out what to copy and edit to achieve this!
Configuration
- ROM_0_URL
Default: http://192.168.7.5:80/rom0.bin
The URL where the firmware for the first application partition can be downloaded.
- ROM_1_URL
Default: http://192.168.7.5:80/rom1.bin
Used when
RBOOT_TWO_ROMSis set. The URL where the firmware for the second application partition can be downloaded.
- SPIFFS_URL
Default: http://192.168.7.5:80/spiff_rom.bin
The URL where the spiffs partition attached can be downloaded.
Credits
The initial sample was made possible with the assistance of piperpilot, gschmott and robotiko on the esp8266.com forum.
References
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
Basic ProgMem
This sample application demonstrates the various ways to access data stored in Flash memory.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic I2C Scanner
Classic Arduino I2C scanner with adaptations for Sming framework.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Serial
Demonstrates Sming’s asynchronous capabilities using the UART to send and receive simultaneously using two serial ports.
Use the primary serial port to enter commands:
- cat
Send the contents of the Readme.md file to the second serial port.
- text
Echo a block of text
Note that you can continue to type commands while serial data is being transmitted as all operations are fully asynchronous. This becomes more obvious if you try using a low baud rate, for example:
make COM_SPEED_SERIAL=9600 COM_SPEED_ESPTOOL=921600
We’d still like a decent speed for flashing though!
If you want to set the two ports to different speeds the code will need to be modified.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Servo
Demonstrates use of the Servo library for controlling multiple servo actuators.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Smart Config
Introduction
SmartConfig is a mechanism to more easily configure an ESP device using a smart phone.
Calling smartConfigStart() starts a search for an Access Point (AP) with a special signature. It then tries to extract data like SSID and password from it. The App on your smart phone sends out that information.
The example here shows how to use ESP_TOUCH method to do smart configuration on the device. It is ported from the C code that Espressif provides in their SDK examples.
You will need an App for your Smart Phone, such as:
ESP8266 SmartConfig (search on Android Play)
See also https://www.espressif.com/en/products/software/esp-touch/overview.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
Basic SSL
Compilation
In Sming the SSL support is not enabled by default.
In order to enable it you should compile your project with the
ENABLE_SSL =1 directive. This can be done using the following command:
make ENABLE_SSL=1
Now you can flash your application to your ESP8266 device.
Debug Information
If you want to see more debug information during compile type you should
add the directive SSL_DEBUG =1, like this:
make ENABLE_SSL=1 SSL_DEBUG=1
Slow SSL negotiation
The initial SSL negotiation is CPU intensive. By default SmingFramework switches the CPU frequency from 80 to 160 MHz. After the negotiation the CPU is switched back to 80 MHz.
If your device is running on battery this can drain the battery much
faster. If you do not want the switch from 80 to 160 MHz to happen then
make sure to recompile SmingFramework with SSL_SLOW_CONNECT directive:
make USER_CFLAGS="SSL_SLOW_CONNECT=1"
Memory usage
SSL uses a significant amount of RAM. You can build this sample to track heap usage and output statistics every 5 seconds:
make ENABLE_MALLOC_COUNT=1
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Storage
This sample application demonstrates various ways to manage and access flash memory using the Storage Management system.
It also shows how to create and partition custom storage devices.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Tasks
This sample demonstrates using the Task class to efficiently perform intensive processing.
Let’s suppose we need to continuously sample an analogue input as fast as possible, do some processing and output the results.
This sample attempts to do some (very rough) fourier analysis on sampled ADC data, displaying the average amplitudes in a set of frequency ‘buckets’ (ranges).
The timing jitter using this approach is quite bad, so it attempts to correct but generally this form of processing is best done with more capable hardware (e.g. ESP32, FPGA).
In this sample application, we just write the results to the debug terminal, but a real application might stream this to waiting network clients.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Templates
Simple demonstration of how to use Template streams.
The program uses two sets of data:
cars.csv from https://corgis-edu.github.io/corgis/csv/cars/
classics.csv from https://corgis-edu.github.io/corgis/csv/classics/
Templates are used to display filtered data from these in text, JSON and XML formats.
Although the data is just printed to the serial terminal, it can just as easily be sent over a network link as in the Basic IFS sample.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Utility
Demonstrates how to pass parameters to Host applications.
This allows creation of Host-only utility applications.
After building, try this:
make run HOST_PARAMETERS='command=testWebConstants'
To run the application directly, you can do this:
out/Host/debug/firmware/app --nonet command=testWebConstants
--nonet is optional.
You may find --debug=0 useful to suppress all but host error messages.
Specify 1 to include warnings.
Use --help to see available emulator options.
See HOST_PARAMETERS for further details.
References
SoC support
host
Basic Web Skeleton
Introduction
Basic application that can be used as a start point for some useful App.
Features:
can setup wifi ssid and wifi password for STA (wifi client) mode either from own AP or as connected to some wifi network
demonstrate new way of catching wif-events with WifiEvents class
if preconfigured wifi network is unreachable start AP named TyTherm with hardcoded password (see source)
can enable/disable STA (wifi client) mode
own AP autodisable after successful connection to preconfigured wifi network
form population and sending is done with json+ajax
demonstrate usage of getting raw http request body to be processed as json
demonstrate how to fill html template on client side with more flexible than Smings Templating - JavaScript
App called TyTherm because it is base for TinY TermOmeter :)
FlashString
This sample also demonstrates how to use FlashString maps as an alternative to using SPIFFS for serving files.
To test this out, build the application without a filesystem image:
make HWCONFIG=standard ENABLE_FLASHSTRING_MAP=1
See webserver.cpp for the details.
References
Environment Variables
ENABLE_FLASHSTRING_MAP
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Web Skeleton (LTS)
Introduction
This is a copy of the Basic_WebSkeletonApp from the Sming LTS branch, modified to build for Sming 4.0. Sming uses ArduinoJson version 6 by default, so this sample demonstrates how to continue using version 5.
This is how it was done:
When including SmingCore.h, just use #include <SmingCore.h>.
Remove the file include/user_config.h: Sming does not require it. If you have made changes for your project then move the changes into your project’s
component.mkfile.Remove any #include <user_config.h> statements from your files.
Rename Makefile-user.mk file to component.mk.
Replace the file Makefile with the one from the Basic_Blink sample project. If you’ve ignored the instructions and modified the file (!) then you’ll need to move those changes into your new component.mk file instead.
Sming uses #pragma once style for header guards, so consider updating your own files if you’re not already doing this.
We use the OneWire Arduino Library, so add ARDUINO_LIBRARIES := OneWire to our component.mk file. Sming now only builds libraries which are specifically requested for a project.
We’re using ArduinoJson, version 6. Sming supports version 5 or 6, but you must specify which one:
Add ARDUINO_LIBRARIES=ArduinoJson6 to the project’s component.mk file.
Add #include <JsonObjectStream.h>. If you’re not using the stream class, add #include <ArduinoJson6.h> to code.
Update callback function parameter lists for
STADisconnectandSTAGotIP. We can also get a description for disconnection reasons, so display that instead of just a number. See Sming/Platform/WifiEvents.h.In webserver.cpp, the onConfiguration() function uses request.getBody(). It is more efficient to use streams, so just replace with request.getBodyStream().
That’s it.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
Basic WiFi
Demonstrates WiFi network connection and scanning functionality.
Enables an unsecured Software Access Point called Sming InternetOfThings which you can connect to from your smart phone (or other WiFi device).
Prints details of any probe requests, so you can see who’s scanning your device.
Scans list of available Access Points and displays them.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
CANBUS
Demonstrates communication with MCP2515-compatible devices via SPI interface.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HMC5883L Compass
HMC5883L sensor reader.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
DFPlayer Mini
Introduction
The DFPlayer Mini MP3 Player is a small and low-cost MP3 module with a simplified output directly to the speaker. The module can be used as a stand alone module with attached battery, speaker and push buttons or used in combination with an ESP8266/ESP32 via RX/TX.
For more details see this link: https://www.dfrobot.com/wiki/index.php/DFPlayer_Mini_SKU:DFR0299
Building
Steps to test this sample: * Connect DFPlayer Mini using RX/TX pins to TX/RX pins respectively of ESP8266. * Copy one or more mp3 files to your SD card. The example playes one file for some time and moves to the next one. * Insert your SD card in the slot of the DF Player module. * Flash and run the sample code.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
DS3232RTC NTP Setter
Connecting a real-time clock chip allows time to be maintained when OFF or in deep sleep modes without requiring an active network connection.
This example demonstrates how to do this using the popular DS3232 I2C device, and also how to set it up using a network time server.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
TM1637 Display
Demonstrates use of 7-segment LED displays using TM1637 controller. This only requires 2 GPIO connections.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Distance measurement with Vl530xl sensor
This sample demonstrates how to use an inexpensive laser sensor VL53L0X to measure distance via I2C interface.
In addition to the original library a “long range” mode is added to measure distances up to 2 meter.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
DNS Captive Portal
Demonstrates how to use Sming to create a default landing page when configured as an Access Point.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Echo SSL
This example talks to an SSL-enabled TCP server. If you do not have such you can use ncat (from the nmap package) and run something like this:
ncat -vvv -l 0.0.0.0 4444 --ssl --ssl-key ~/Private/x.key --ssl-cert ~/Private/x.cert
See Basic SSL for information on compiling and configuring SSL.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
FTP Server Files
This example sets up an FTP server with a couple of files stored in SPIFFS. It mounts this on top of an FWFS volume (a hybrid filesystem).
The sample creates three users with different roles (guest, user and administrator).
User |
Password |
Role |
guest |
(none) |
Guest |
me |
“123” |
User |
admin |
“1234” |
Admin |
You’ll need to have WiFi configured. You can set this information when building like this:
make WIFI_SSID=ssid WIFI_PWD=password
substituting your actual Access Point details for ssid and password.
Flash to your device:
make flash
You should be able to connect using an FTP client:
ftp ipaddress
and when prompted log in with one of the above usernames.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
APDS-9660 Gesture
Gesture sensors allow ambient light and colour measuring, proximity detection and touchless gesture sensing.
This sample demonstrates how to use the APDS-9660 which is connected via I2C interface.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HTTP Client
See Basic SSL for information on compiling and configuring SSL.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HttpClient Instapush
Instapush is an event notification service created in 2014.
For details, see https://pushbots.com/.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
ThingSpeak Http Client
Example of a HttpClient pushing data to ThingSpeak .
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
AJAX Http Server
Demonstration of a simple web server with page updates using AJAX.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Bootstrap Http Server
Embedded web server.
This sample will download all required files from a remote server.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HttpServer Config Network
Introduction
The HTTP server coming with Sming is quite powerful but it is limited from the available resources of the underlining hardware (your favorite ESP8266 microcontroller). Serving multiple files at once can be problematic. It is not the size of the files that can cause problems, but the number of simultaneous files that need to be delivered. Therefore if you serve multiple CSS or JS files you can optimize your web application before uploading it into your ESP8266 using the advice below.
Optimizing File Delivery
In this example you will see how to combine CSS and JS files, compress them and deliver the optimized content via the HTTP server.
Installation
The file optimization uses gulp. To install it and the needed gulp
packages you need to install first npm. Npm
is the Node.JS package manager. Once you are done with the installation
you can run from the command line the following:
npm install
The command above will install gulp and its dependencies.
Usage
During the development of your web application you should work only in
the web/dev/ folder. Once you are ready with the application you can
pack the resources and upload them to your device. The commands
are
make web-pack make web-upload
That should be it.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HttpServer Firmware Upload
This example combines the libraries Multipart Parser for file uploads and Over-the-Air Firmware Upgrade, to create a browser based firmware upgrade solution akin to what’s found in many consumer devices. The example is kept as minimal as possible to serve as a starting point for your own applications.
About HTTP server file upload
The HTTP server coming with Sming is quite powerful but it is limited by the available resources of the underlining hardware (your favorite ESP8266 microcontroller).
This sample demonstrates how to use the Multipart Parser library to enable file upload of the HTTP server. On a normal computer the file uploads are usually using temporary space on the hard disk or in memory to store the incoming data.
On an embedded device that is a luxury that we can hardly afford. In this sample we demonstrate how to define which file upload fields should be recognized and what (file) streams are responsible for processing and storing the data. If a field is not specified then its content will be discarded.
About OtaUpgrade
The OtaUpgrade component provides the OtaUpgradeStream class which
is hooked up to the web server to process a firmware upgrade file uploaded to
the device.
The component is also responsible for creating the upgrade files during the build
process. A single upgrade file conveniently encapsulates all ROM images, thereby
relieving the user from having to know the slot that is updated and manually
selecting the corresponding ROM image in a Two-ROM configuration.
The file format also supports state-of-the-art security features like a digital
signature, encryption and downgrade protection. You are invited to play with
them and observe their impact on code size. See also the Over-the-Air Firmware Upgrade
documentation for further advice on how to use the security features properly.
Usage instructions
Configure your flash memory layout. See Hardware configuration.
Build the example by running:
makeConnect your device via USB/Serial cable and run:
make flashconfig
to clear any remains of the previous flash layout configuration, followed by:
make flash
to install the example firmware. You need to do this only once. Subsequent updates can be performed wirelessly using the web interface.
Point the browser to your ESP’s IP address to open the firmware upgrade page.
Select the upgrade file, which has been automatically created alongside step 2 from
out/Esp8266/<build-type>/firmware/firmware.otaand hit the “Update” button.After a few seconds, you should see a confirmation the the upgrade was successful. The device now reboots into the upgraded firmware.
If the upgrade is not successful, rebuild with debug output enabled and check the serial console for error messages.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HttpServer Plugins
Simple example that demonstrate extending the HttpServer functionality with additional plugins. Using such examples one can add:
URL Rewriting
Authentication. Sming has already existing plugins for Basic Authentication, IP Limiting
Content extracting/processing. For example a gzip decoder plugin can be implemented.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HttpServer Websockets
Simple example of how to support persistent websocket connections in a server.
References
Environment Variables
ENABLE_CMD_HANDLER
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
AM2321 Humidity Sensor
Example application demonstrating the use of the AM2321 Temperature/Humidity Sensor library.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
BME280 Humidity Sensor
Example application demonstrating the use of the Adafruit BME280 Library library.
This is a port of the bme280test example from the library.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
DHT22 Humidity Sensor
Example application demonstrating the use of the DHT ESP Temperature/Humidity Sensors library with a DHT22 humidity sensor.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SI7021 Humidity Sensor
Introduction
Example code for I2C SI7021 sensor board
This code uses the following GPIO:
#define I2C_SCL 5 // SCL
#define I2C_SDA 4 // SCA
Build instructions
Use make and make flash to build and flash the firmware to the
ESP8266 board.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
IR Library
Example application demonstrating how to receive IR remote signals using the ESP8266.
References
SoC support
esp8266
WS2812 LEDs
Example application to demonstrate control of WS2812 RGB pixel-strip LEDs.
References
SoC support
esp8266
Yeelight LED bulbs
Demonstrates how to control Yeelight WiFi devices.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
BH1750 Light Sensor
ESP8266 BH1750 sensor reader.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Liquid Crystal 44780
Introduction
Example code for I2C LiquidCrystal LCDs.
This code uses the following GPIO:
GPIO0 SCL
GPIO2 SDA
Build instructions
Use make and make flash to build and flash the firmware to the
ESP8266 board.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Live Debug
This project is an example of how to integrate GDB debugging into your project. It provides a basic command interface which you can use via regular serial terminal or with the GDB application.
To use this sample application with the command-line GDB application, simply build and flash the project as usual:
make clean
make flash
You should be presented with the GDB command prompt. Enter ‘c’ to continue running the application:
(gdb) c
Continuing.
(attached)
The (attached) prompt is displayed by the LiveDebug application.
Type help to get a list of available commands.
Note that if you run this application via serial terminal
(make terminal) you’ll get the (Detached) prompt instead.
Debugging under eclipse
Interacting with the GDB console causes problems for eclipse, so compile
with the ENABLE_GDB_CONSOLE=0 option:
make clean
make flash ENABLE_CONSOLE=0
Alternatively, use the consoleOff command from the GDB command
prompt, then quit the debugger and connect via eclipse.
Exception Handling
Sming comes with a built-in exception handling that takes care to display the stack trace leading to the issue. Usually it looks like this:
***** Fatal exception 28 (LOAD_PROHIBITED)
pc=0x40100e96 sp=0x3ffff640 excvaddr=0x000015b8
ps=0x00000033 sar=0x00000018 vpri=0x000000f0
r00: 0x40100d69=1074793833 r01: 0x3ffff640=1073739328 r02: 0x3fff3900=1073690880
r03: 0x2b265ed4= 723934932 r04: 0x3fffbff0=1073725424 r05: 0x000015b8= 5560
r06: 0x000015b8= 5560 r07: 0x14a8433b= 346571579 r08: 0x00000008= 8
r09: 0x14a842f3= 346571507 r10: 0x3fff22d0=1073685200 r11: 0x00000003= 3
r12: 0x00000048= 72 r13: 0x3fff38c0=1073690816 r14: 0x3ffe9da0=1073651104
r15: 0x3fff1138=1073680696
Stack dump:
To decode the stack dump call from command line:
make decode-stacktrace
and copy & paste the text enclosed in '===='.
================================================================
3ffff640: 40100e96 00000033 00000018 000000f0
3ffff650: 40100d69 3fff3900 2b265ed4 3fffbff0
3ffff660: 000015b8 000015b8 14a8433b 00000008
3ffff670: 14a842f3 3fff22d0 00000003 00000048
3ffff680: 3fff38c0 3ffe9da0 3fff1138 0000001c
3ffff690: 002222fb c0fb5c01 0bc10000 facfd1fb
3ffff6a0: 620020c0 6162802d 0020c004 59062c52
3ffff6b0: 0020c051 61492c48 210020c0 7c38fb50
...
================================================================
Enter make decode-stacktrace then copy & paste the text
to produce something readable like this:
0x40100e96: pvPortRealloc at ??: ?
0x40100d69: pvPortMalloc at ??:?
0x402455f0: ax_port_malloc at C:\tools\Sming-3.1.2\Sming/third-party/axtls-8266/ replacements/mem.c:51
0x4024561a: ax_port_calloc at C:\tools\Sming-3.1.2\Sming/third-party/axtls-8266/ replacements/mem.c:63
0x40230acc: x509_new at c:\tools\Sming-3.1.2\Sming\third-party\axtls-8266/ssl/x5 09.c:81
0x4023d3e4: m_vsnprintf at C:\tools\Sming-3.1.2\Sming/system/m_printf.cpp:69
0x4023d4a6: m_vprintf at C:\tools\Sming-3.1.2\Sming/system/m_printf.cpp:83
0x40000a39: ?? ??:0
0x4021418a: pp_attach at ??:?
0x40221d60: pbuf_alloc at ??:?
0x40221f0a: pbuf_copy at ??:?
0x4023d3e4: m_vsnprintf at C:\tools\Sming-3.1.2\Sming/system/m_printf.cpp:69
Note that you can also put the stack trace into a file, for example dump.txt then enter:
make decode-stacktrace TRACE=dump.txt
Using the information about the type of the exception (ex:
***** Fatal exception 28) and the sequence of commands might help us
figure out the issue.
But that information might not be enough. And finding the root cause may take quite some time.
GDB Debugging
Debugging is a powerful technique allowing you to interactively run your code and be able to see much more information about the things that went wrong.
To use, (Re)compile your project with the ENABLE_GDB option and
flash it to the board:
make clean
make ENABLE_GDB=1
make flash
Instead of a terminal, the GDB console will be opened automatically.
If you need to run GDB manually after resetting the board or after it has run into an exception, use the provided script:
make gdb
Note that software breakpoints (‘br’) only work on code that is in RAM. During development you can use the GDB_IRAM_ATTR attribute in your function declarations. Code in flash can only have a hardware breakpoint (‘hbr’).
Read the GDB stub Notes for more information.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MeteoControl
More complex example of Internet of Things device. Can read humidity and temperature data from sensor and output it to screen with actual time. Time loaded directly from Google. Also with device can automatically control any external load.
Features:
temperature
humidity
actual time from internet
built-in display
web control interface
automatically control external load
HTTP API for data reading & writing
initialization from internet at first run
How web interface looks like:
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MeteoControl MQTT
Similar to MeteoControl sample without LCD interface and instead controlled via MQTT.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MQTT Client Hello
Simple MQTT client example.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Network Ping
Sample demonstrating the usage of a network ping to check for connectivity issues.
References
SoC support
esp8266
MCP23017 I2C Port Expander
You can easily add more GPIO connections for an ESP8266 using a port expander. This example demonstrates how to do that with a common I2C port expander.
If you need a faster connection, see MCP23S17 SPI Port Expander.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MCP23S17 SPI Port Expander
You can easily add more GPIO connections for an ESP8266 using a port expander. This example demonstrates how to do that with a common SPI port expander.
These devices are perhaps a little less common than their I2C equivalents, see MCP23017 I2C Port Expander.
References
SoC support
esp8266
BMP180 Pressure Sensor
Introduction
Example code for I2C BMP180 sensor board
This code uses the following GPIO:
GPIO0 SCL
GPIO2 SDA
Build instructions
Use make and make flash to build and flash the firmware to the
ESP8266 board.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
RCSwitch Library
Demonstration of how to control wireless (radio-controlled) devices such as switched mains sockets, TVs, etc.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
nRF24L01 Radio
Example application showing how to interface with the popular nRF24L01 2.4GHz radio transceivers.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SI443 Radio
Example application for radio module Si4432 aka RF22 driver. Link: http://www.electrodragon.com/w/SI4432_433M-Wireless_Transceiver_Module_%281.5KM_Range,_Shield-Protected%29
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SD Card
SDCard/FAT file usage and write benchmark.
References
Used by
SD Card Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
5110 LCD Screen
Demonstration of how to interface to Monochrome Nokia 5110 LCD Displays.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
rp2040
SSD1306 OLED Screen
Example of direct work with SSD1306 OLED screen on ESP8266.
Minimal requirements: ESP-03
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
ILI9163C TFT Screen
Example of direct work with ILI9163C 1.44” TFT screen on ESP8266.
Minimal requirements: ESP-03
References
SoC support
esp8266
ILI9340 and ILI9341 TFT Screens
Demonstration of how to interface with displays such as the Adafruit 2.8” Touch Shield V2 (SPI).
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
ST7735 TFT Screen
Demonstration of how to interface with various SPI displays using the Adafruit ST7735 library.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SMTP Client
Purpose
To show the basics of sending an email via SMTP
SMTP
When an IOT device detects a malfunction, it is good to be able to notify the user. We can do that either by sending them an email or via a service like MQTT.
This sample shows how to send an email via SMTP directly from the ESP8266.
smtp2go conveniently provides a free account.
Create an account here: https://www.smtp2go.com/setupguide/arduino/ .
It needs some configuration: * the name of the SMTP server, eg “mail.smtp2go.com” * a username and password * the name and email of the person from whom the email will be sent * the name and email of the person to send the email to
Edit the sample to replace these values and the SSID etc.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SystemClock NTP
Demonstrates various methods for keeping the system clock updated from a Network Time Server using the NtpClient class.
The Timezone class is used to convert between UTC and local times, accounting for daylight savings.
The SolarCalculator class is used to calculate times of next sunrise and sunset, which you might want to automate activation of lights, for example.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
TCP Client NarodMon
An example of sending data to narodmon.ru using a TCP client.
https://narodmon.ru is a geo-information project to display on the world map and control (on PCs, smartphones and other gadgets) the sensor readings of its members (temperature, humidity, pressure, wind speed and direction, radiation, energy consumption and any other values), as well as private and urban webcams for public or private viewing.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
DS1820 Temperature Sensor
DS1820/DS18B20 sensor reader.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
UDP Server Echo
Sets up a UDP server which echoes back incoming packets. By default listens on port 1234.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
UDP Server mDNS
Instructions
The multicast Domain Name System (mDNS) resolves host names to IP addresses within small networks that do not include a local name server.
More info on mDNS can be found at https://en.wikipedia.org/wiki/Multicast_DNS
mDNS has two parts:
Advertise
Discover
This example just does Advertise. See Basic MDNS for discovery example.
In short this code will advertise other machines about its ipaddress.
How to use mDNS:
ADD your WIFI_SSID / WIFI_PWD
Flash the Complied code to your ESP8266/ESP32 device
Open a web browser and go to “http://sming.local/” to open a sample webpage.
You should also be able to ping using the advertised name:
ping UDP_Server.local
Linux
You need to install Avahi mDNS/DNS-SD daemon.
In your browser type “http://sming.local/” to open a sample webpage.
Android
You need to install ZeroConf Browser or Bonjour Browser.
In those app you would be able to see IP address of your ESP module.
In android Chrome “http://sming.local/” does not translate to IP address, so android Chrome is not supporting mDNS.
But you can make your own app using Network Service Discovery. See http://developer.android.com/training/connect-devices-wirelessly/nsd.html for details.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HCSR04 Ultrasonic Transducer
Shows how to use an HCSR04 module to measure distances.
Warning
Ultrasonic_HCSR04 modules usually work with 5v power and TTL levels, so you NEED voltage divider or level shifter for ECHO pin.
Trigger pin is tolerant to 3.3v and should work without problems.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
WebcamServer sample
A sample providing access to webcams via HTTP server.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Websocket Client
This is a simple demo of the WebsocketClient class. It shows connection, closing and reconnection methods of WebsocketClient.
The client tries to connect to a websocket echo server. It sents 10 messages then client connection is closed. This sequence repeats after 20 seconds.
The sample was originally written to communicate with echo.websocket.org
but that service no longer exists.
Instead, run make wsserver to run a local test server.
This has the advantage of showing detailed diagnostic information which may be helpful.
Build with WS_URL set to the server address. For example:
make WS_URL=ws://192.168.1.10:8000
References
Environment Variables
WS_URL
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
WiFi Sniffer
Introduction
This is an adaptation of the WiFi promiscuous mode demo code from Arduino
See https://www.hackster.io/rayburne/esp8266-mini-sniff-f6b93a
// Notes.h tab in Arduino IDE is only for comments and references!
// based on RandDruid/esp8266-deauth (MIT) https://github.com/RandDruid/esp8266-deauth
// inspired by kripthor/WiFiBeaconJam (no license) https://github.com/kripthor/WiFiBeaconJam
// https://git.schneefux.xyz/schneefux/jimmiejammer/src/master/jimmiejammer.ino
// requires SDK v1.3: install esp8266/Arduino from git and checkout commit 1c5751460b7988041fdc80e0f28a31464cdf97a3
// Modified by M. Ray Burnette for publication as WiFi Sniffer 20161013
/*
Arduino 1.6.12 on Linux Mint 17.3
Sketch uses 227,309 bytes (21%) of program storage space. Maximum is 1,044,464 bytes.
Global variables use 45,196 bytes (55%) of dynamic memory, leaving 36,724 bytes for local variables. Maximum is 81,920 bytes.
*/
/*
// beacon template
uint8_t template_beacon[128] = { 0x80, 0x00, 0x00, 0x00,
/*4*/
0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
/*10*/ 0x01, 0x02, 0x03, 0x04, 0x05, 0x06,
/*16*/ 0x01, 0x02, 0x03, 0x04, 0x05, 0x06,
/*22*/ 0xc0, 0x6c,
/*24*/ 0x83, 0x51, 0xf7, 0x8f, 0x0f, 0x00, 0x00, 0x00,
/*32*/ 0x64, 0x00,
/*34*/ 0x01, 0x04,
/* SSID */
/*36*/ 0x00, 0x06, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x01, 0x08, 0x82, 0x84, 0x8b, 0x96, 0x24, 0x30, 0x48, 0x6c,
0x03, 0x01,
/*56*/ 0x04
}
;
* /
/* Notes:
Ref: http://www.esp8266.com/viewtopic.php?f=32&t=7025
In the ESP8266WiFi.h, there is the function getNetworkInfo() which I presume allows you to get
info for hidden AP.
bool getNetworkInfo(uint8_t networkItem, String &ssid, uint8_t &encryptionType, int32_t &RSSI, uint8_t* &BSSID, int32_t &channel, bool &isHidden);
CODE: SELECT ALL
/**
loads all infos from a scanned wifi in to the ptr parameters
@param networkItem uint8_t
@param ssid const char*
@param encryptionType uint8_t
@param RSSI int32_t
@param BSSID uint8_t *
@param channel int32_t
@param isHidden bool
@return (true if ok)
*/
/* Serial Console Sample Output:
ESP8266 mini-sniff by Ray Burnette http://www.hackster.io/rayburne/projects
Type: /-------MAC------/-----WiFi Access Point SSID-----/ /----MAC---/ Chnl RSSI
BEACON: <=============== [ TardisTime] 1afe34a08bc9 8 -76
BEACON: <=============== [ xfinitywifi] 56571a0730c0 11 -90
BEACON: <=============== [ ] 52571a0730c0 11 -91
BEACON: <=============== [ ATTGH6Gs22] 1005b1d6ff90 11 -95
BEACON: <=============== [ ATT4P3G9f8] 1c1448777420 11 -92
BEACON: <=============== [ HOME-30C2] 5c571a0730c0 11 -91
BEACON: <=============== [ ATT8Q4z656] b077acc4dfd0 11 -92
BEACON: <=============== [ HOME-B1C2] 94877c55b1c0 11 -94
BEACON: <=============== [ HUXU2012] 0c54a5d6e480 6 -94
BEACON: <=============== [ xfinitywifi] 0c54a5d6e482 6 -97
BEACON: <=============== [ ] 0c54a5d6e481 6 -96
DEVICE: 18fe34fdc2b8 ==> [ TardisTime] 1afe34a08bc9 8 -79
DEVICE: 18fe34f977a0 ==> [ TardisTime] 1afe34a08bc9 8 -94
DEVICE: 6002b4484f2d ==> [ ATTGH6Gs22] 0180c2000000 11 -98
BEACON: <=============== [ HOME-01FC-2.4] 84002da251d8 6 -100
DEVICE: 503955d34834 ==> [ ATT8Q4z656] 01005e7ffffa 11 -87
BEACON: <=============== [ ] 84002da251d9 6 -98
BEACON: <=============== [ xfinitywifi] 84002da251da 6 -95
BEACON: <=============== [ ] fa8fca34e26c 11 -94
DEVICE: cc0dec048363 ==> [ ATT8Q4z656] 01005e7ffffa 11 -88
BEACON: <=============== [ ] fa8fca95bad3 11 -92
BEACON: <=============== [ HOME-5475] 58238c3b5475 1 -96
BEACON: <=============== [ xfinitywifi] 5a238c3b5477 1 -94
BEACON: <=============== [ ] 5a238c3b5476 1 -96
DEVICE: 1859330bf08e ==> [ ATT8Q4z656] 01005e7ffffa 11 -92
BEACON: <=============== [ ] 92877c55b1c0 11 -92
DEVICE: f45fd47bd5e0 ==> [ ATTGH6Gs22] ffffffffffff 11 -93
BEACON: <=============== [ Lynch] 744401480a27 11 -96
BEACON: <=============== [ xfinitywifi] 96877c55b1c0 11 -93
DEVICE: f43e9d006c10 ==> [ xfinitywifi] 8485066ff726 6 -96
DEVICE: 285aeb4f16bf ==> [ ATTGH6Gs22] 3333ffb3c678 11 -94
DEVICE: 006b9e7fab90 ==> [ ATTGH6Gs22] 01005e7ffffa 11 -91
DEVICE: 78456155b9f0 ==> [ Lynch] 01005e7ffffa 11 -95
DEVICE: 6cadf84a419d ==> [ HOME-30C2] 88cb8787697a 11 -89
BEACON: <=============== [ Verizon-SM-G935V-6526] a608ea306526 11 -92
*/
References
SoC support
esp8266
Hosted RPC Server over Serial
Overview
This application creates a RPC server that will communicate over serial interface. To compile it for an Esp8266 microcontroller you can use the following command:
make SMING_ARCH=Esp8266
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Hosted RCP Server over TCP
Overview
This application creates a RPC server that will communicate over TCP. You can either start an Access Point from the controller or connect the application to an existing WIFI Access point. The latter can be compiled using the following command:
make SMING_ARCH=Esp8266 CONNECT_TO_WIFI=1 WIFI_SSID="MySSID" WIFI_PWD="Secr3tP4Ssw0rd"
Configuration
References
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic AnimatedGIF
Sample demonstating the usage of the optimized AnimatedGIF library together with Sming Graphics Library library.
You should be able to see the following animated image:
Image source: Animated Images Dot Org.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
TensorFlow_HelloWorld
The power of Machine Learning (ML) inside you microcontroller! This sample is a remake of the standard HelloWorld example from TensorFlowLite for Arduino. It demonstrated how a ML “model” can be used inside an ESP8266 mircocontroller with the help of Sming.
If you are not familiar with TensorFlow take a look at their documentation.
Compilation
This sample can be compiled and run on a ESP8266 microcontroller or on the Host architecture. In order to compile it for ESP8266 you must use ESP Quick Toolchain. The ESP Quick Toolchain is more feature-complete and allows the compilation to run without many modifications to the original library.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Webcam
Sample demonstrating communication with a web camera via AT commands using At Client library. The complete set of commands for the sample web camera are in the docs/ directory.
More about the web camera module and commands can be found here.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Bluetooth Gamepad
Introduction
This sample turns the ESP32 into a Bluetooth LE gamepad that presses buttons and moves axis
Possible buttons are: BUTTON_1 through to BUTTON_16 (16 buttons supported by default. Library can be configured to support up to 128)
Possible DPAD/HAT switch position values are: DPAD_CENTERED, DPAD_UP, DPAD_UP_RIGHT, DPAD_RIGHT, DPAD_DOWN_RIGHT, DPAD_DOWN, DPAD_DOWN_LEFT, DPAD_LEFT, DPAD_UP_LEFT (or HAT_CENTERED, HAT_UP etc)
bleGamepad.setAxes takes the following int16_t parameters for the Left/Right Thumb X/Y, Left/Right Triggers plus slider1 and slider2, and hat switch position as above: (Left Thumb X, Left Thumb Y, Right Thumb X, Right Thumb Y, Left Trigger, Right Trigger, Hat switch position ^ (1 hat switch (dpad) supported by default. Library can be configured to support up to 4)
Library can also be configured to support up to 5 simulation controls (can be set with setSimulationControls) (rudder, throttle, accelerator, brake, steering), but they are not enabled by default.
Testing
You can use one of the following applications on your PC to test and see all buttons that were clicked.
On Linux install jstest-gtk to test the ESP32 gamepad. Under Ubuntu this can be done by typing the following command:
sudo apt install jstest-gtk
On Windows use this Windows test application.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Bluetooth Keyboard
This sample demonstrates how to turn an Esp32 device into external keyboard. The “keyboard” and your PC will be communicating using Bluetooth Low Energy (BLE). The “keyboard” will write words, press Enter, press a media key and, if enabled in the sample code, Ctrl+Alt+Delete.
Usage
Once this sample is flashed and running on your ESP32 you can test it. Open a new text editor on your PC. Then search from your PC for new bluetooth devices. A device named “Sming BLE Keyboard” should show up. Connect to it and focus/open you text editor window. Be fast. Soon enough a “Hello World” text will start to be “magically” typed inside your text editor.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
CS5460 generic sample
The sample prints the measured RMS voltage each second (with CS5460 voltage and current filters enabled).
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Arducam
A demonstration application for controlling ArduCAM camera modules via web or telnet interface.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
CommandLine
Demonstrates Sming’s CommandProcessing capability via serial interface.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
TelnetServer
A demonstration of a telnet server built using CommandProcessing.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Control your DIAL-enabled smart monitor/TV using Sming
Demonstrates starting of YouTube, playing a video and closing YouTube after a small period of time.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic FlashIP
Introduction
This is a simplified version of the Basic_OTA sample using the FlashIP approach instead.
There is one application partition and one LittleFS filing system partition. NB. You can also use SPIFFS and FWFS.
Building
Testing
For testing purposes Sming offers a simple python webserver that can be run on your development machine:
make otaserver
The server listens on port 9999 and all network interfaces.
The current directory is changed to the firmware directory for the current project, so you could,
for example, open a separate terminal window in the Basic_Serial directory and do this:
make SMING_SOC=rp2040
make otaserver
This will then be ready to serve up the app.bin image file for the Basic_Serial sample.
This sample should be compiled and flashed as follows, replacing 192.168.1.30 with your development machine’s IP address
and inserting details for your local WiFi:
make flash WIFI_SSID=... WIFI_PWD=... FIRMWARE_URL=http://192.168.1.30:9999/app.bin
Once connected to WiFi, you can enter ota to download and re-program the new firmware.
Configuration
- FIRMWARE_URL
The URL for the application image to be downloaded
References
Environment Variables
SoC support
host
rp2040
Google Cast Client
Sample application that demonstrates how to send information directly from your application to your Smart TV using the Google Cast protocol.
A media file is started playing. From the terminal, you can control as follows:
‘f’: Skip forward 10 seconds ‘r’: Skip back 10 seconds space: Toggle between pause and play ‘q’: Quit playback
The chromecast device must be online at startup, or a connection error will result. Re-start the application to retry.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Advanced Animation
Lots of rectangles flying around the screen. Again.
Does the same thing as the Basic Animation sample but uses the display driver directly which eliminates the Scene construction step.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Animation
Lots of rectangles flying around the screen.
Port of Bob’s animation tutorial code. Thanks Bob!
https://www.youtube.com/watch?v=q9xWvLZg7Lc
https://bytesnbits.co.uk/basic-animation-spi-touchscreen/
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Graphics
Demonstration of how to use the Sming Graphics Library library with the Host Virtual Screen or displays such as the Adafruit 2.8” Touch Shield V2 (SPI).
To try this out from the Host Emulator you can use the virtual screen:
make virtual-screen
The IP address:port will be shown at the command prompt and in the title bar, for example 192.168.1.105:7780. Build the application:
make SMING_ARCH=Host VSADDR=192.168.1.105 VSPORT=7780
Note: The screen server address and port are passed by command line so the application doesn’t require rebuilding. For example:
out/Host/debug/firmware/app vsaddr=192.168.1.105 vsport=7780
References
Environment Variables
GUITIMER_INTERVALMAX_LOOP_COUNTSHOWFONT_INTERVAL
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Touch
Demonstration of how to use the Sming Graphics Library library with touch screens such as the Adafruit 2.8” Touch Shield V2 (SPI).
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Bresenham
Exploration of advanced anti-aliased line and curve rendering algorithms.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Color Test
Simple test that should draw from left to right rectangles with red, green, blue and white color.
Useful for display driver diagnostics. The TFT_eSPI project describes quite nicely the different switches that can affect the colors in MIPI displays:
Different hardware manufacturers use different colour order
configurations at the hardware level. This may result in
incorrect colours being displayed.
Incorrectly displayed colours could also be the result of
using the wrong display driver...
Typically displays have a control register (MADCTL) that can
be used to set the Red Green Blue (RGB) colour order to RGB
or BRG so that red and blue are swapped on the display.
This control register is also used to manage the display
rotation and coordinate mirroring. The control register
typically has 8 bits, for the ILI9341 these are:
Bit Function
7 Mirror Y coordinate (row address order)
6 Mirror X coordinate (column address order)
5 Row/column exchange (for rotation)
4 Refresh direction (top to bottom or bottom to top in portrait orientation)
3 RGB order (swaps red and blue)
2 Refresh direction (top to bottom or bottom to top in landscape orientation)
1 Not used
0 Not used
The control register bits can be written with this example command sequence:
tft.writecommand(TFT_MADCTL);
tft.writedata(0x48); // Bits 6 and 3 set
0x48 is the default value for ILI9341 (0xA8 for ESP32 M5STACK)
in rotation 0 orientation.
Another control register can be used to "invert" colours,
this swaps black and white as well as other colours (e.g.
green to magenta, red to cyan, blue to yellow).
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Custom Object
Demonstates how to create a custom object and render it.
This sample draws a Mandelbrot set which is particular CPU intensive with lots of iteration and floating point calculations.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Alexa
If you have an Echo Dot or other Amazon Alexa device, this shows how to provide simple support by emulating a Philips Hue lighting bridge.
Pairing is permanently enabled in this sample application so all you need to do is ask Alexa to discover devices and the lights should appear.
You can verify this is working by requesting a list of registered lights via HTTP:
http://IP_ADDRESS/api/user/lights
Turning the ESP LED on and off can be done by sending a POST request to light 101.
The RESTED plugin for firefox is very useful for this sort of thing. The endpoint URL is:
http://IP_ADDRESS/api/user/lights/101/state
To turn the LED ON, the body of the request would contain:
{"on":true}
And to turn it off again:
{"on":false}
Remember to set the Content-Type header to application/json.
Here’s how to do it with CURL:
curl -X POST -H "Content-Type: application/json" -d "{on: true}" http://IP_ADDRESS/api/user/lights/101/state
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic RS485
Demonstrates basic use of IO Control library to set up one MODBUS device and one DMX device on serial port 0.
Debug output is moved to serial port 1.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Fan Controller
Using RPi Pico to control three 4-pin PWM fans for inverter.
Output 20kHz PWM to each fan independently
Measure sync pulses from each fan to confirm rotation rate and detect failure
Provide modbus slave control
References
SoC support
rp2040
LittleFS inspector
Builds a basic filesystem and includes simple code to display commits by reading partition directly. Helps with understanding structure and evaluating performance.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic MDNS
Demonstrates use of the MDNS library to discover local devices.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MHZ19 CO2 sensor reading
Example application demonstrating the use of the MHZ19 CO2 Sensor library.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MHZ19 Span-calibration
Example demonstrating span-calibration procedure.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MHZ19 Zero-Calibration
Example demonstrating Zero-calibration procedure.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
ModbusMaster RTU generic sample
The generic sample calls mbLoop() each second. In mbLoop() the globalSeconds variable is incremented and sent to a slave device. Then the same register address of the same slave device is read and the result is output via debugf() using UART1.
The modbus response timeout can be changed using MB_RESPONSE_TIMEOUT.
Check out my video.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HttpServer Upload
This example demonstrates how to upload multiple files on the server.
About HTTP server file upload
The HTTP server coming with Sming is quite powerful but it is limited by available hardware resources.
This sample demonstrates how to use the Multipart Parser library to enable file upload of the HTTP server. On a normal computer the file uploads are usually using temporary space on the hard disk or in memory to store the incoming data.
On an embedded device that is a luxury that we can hardly afford. In this sample we demonstrate how to define which file upload fields should be recognized and what (file) streams are responsible for processing and storing the data. If a field is not specified then its content will be discarded.
Usage instructions
Configure your flash memory layout. See Hardware configuration.
Build the example by running:
makeConnect your device via USB/Serial cable and run:
make flash
to install the example firmware.
Point the browser to your device’s IP address to open the firmware upgrade page.
Select the upload file and hit the “Update” button.
If the file size isn’t bigger than the MAX_FILE_SIZE defined in the application code you should see a confirmation that the upload was successful.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
OTA over MQTT
Introduction
This example demonstrates how you can create an application that updates its firmware Over The Air (OTA) using the MQTT protocol. This application uses OTA Firmware Upgrade via MQTT and follows the recommended versioning principles.
Based on ENABLE_OTA_ADVANCED the firmware data can be either without any encoding or be signed and encrypted.
Tools
There are two tools that facilitate the packiging and deployment of a new firmware.
For more information read Firmware packaging in the documentation of the OTA Firmware Upgrade via MQTT component.
Security
Depending on ENABLE_SSL a standard SSL/TLS can be enabled:
The communication between the application and the server will be encrypted using standard SSL.
To prove that the server is the correct one: The MQTT clients should pin the public key fingerprint on the server. OR have a list of public key fingerprints that are allowed.
Depending on
ENABLE_CLIENT_CERTIFICATEthe application can send a client certificate that is signed by the server.
Configuration
If ENABLE_SSL is enabled (highly recommended), OTA upgrade files will be transferred securely over TLS/SSL.
- APP_ID
Default: “test”
This variable contains the unique application name.
- APP_VERSION
Default: not set
Contains the application major and minor versions separated by comma. Example “4.2”. If not set will use the current major and minor version from Sming.
- APP_VERSION_PATCH
Default: not set
Contains the application patch version as integer. For stable versions you can use 0 until 255. For unstable versions the current timestamp can be used as a patch version.
- ENABLE_CLIENT_CERTIFICATE
Default: 0 (disabled)
Used in combination with
ENABLE_SSL. Set to 1 if the remote server requires the application to authenticate via client certificate.
- MQTT_URL
Default: depends on
ENABLE_SSLandENABLE_CLIENT_CERTIFICATEvaluesUrl containing the location of the firmware update MQTT server.
References
Environment Variables
MQTT_FINGERPRINT_SHA1
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
RingTone Player
Introduction
You may find the need to include basic audio features such as haptic feedback or tune generation to support an existing application. The advantage of using I2S for this is the very low CPU overhead.
This sample demonstrates how simple audio can be added to an application using the I2S ToneGenerator and RingTone libraries. The sample contains a selection of tunes.
The sample may be controlled via serial terminal, or by web interface.
Serial connection
Because I2S uses the serial RX pin, the serial port uses the alternate pin mappings. See Tone Generator for connection details.
You’ll still need the regular serial port for programming. Here’s a sample setup for Windows:
Using NodeMCU with standard serial port connected to COM4
Second USB/UART converter connected to alternate serial pins as COM8
The command line to build, flash and open a terminal would be:
make -j COM_PORT=COM8 COM_SPEED=115200 COM_PORT_ESPTOOL=COM4 COM_SPEED_ESPTOOL=921600
make flashinit
make flash
Web interface
This is built using Bootstrap 4.3.1 and jQuery 3.4.1.
RebootUn-installs the I2S driver then restarts the system after 5-seconds. The memory consumption drops as the DMA buffers are de-allocated.
Play controls- ▢ Stop
Tune playback stops at the current position, and the I2S driver is stopped so I2S interrupts are disabled. DMA buffers remain allocated.
- || Pause
Tune playback stops at the current position, but I2S interrupts remain active.
- ▷ Play
Resumes tune playback from stop/pause states.
Voice and ModeSelects available voice for the tone generator (Sine, Triangular, Sawtooth or Square wave).
Playback mode
⇢ Sequential
🔀 Random
Playback speed
Tune selection
- ↤ Previous
Play previous tune
- ↲ Rewind
Rewind to start of current tune
- ↦ Next
Play next tune
- → Skip
Skip to a random tune
Shows the current tune and elapsed play time. A drop-down list shows all available tunes.
Graphs
In addition to play controls, there are three graphs showing some statistics over a two-minute period. The data is sent from the ESP8266 via websocket once per second.
- CPU Usage
Gives an indication of processor loading. Try connecting a second web client to see what happens.
- Fill Time
Shows the time taken by the Tone Generator to create and buffer tone waveforms.
This graph has three traces, indicating time in milliseconds over the 1-second update period:
Red: Longest time taken
Green: Shortest time
Black: Average time
You’ll see more activity during complex, fast tunes but also for lower notes which require more samples (and larger buffers).
The progress bar indicates the total time taken over the update period.
- Memory Usage
The graph has three traces:
Black: Available memory
Red: Maximum memory used during the update period
Green: Memory in use at the time of update
The red/green traces generally follow each other and represent the allocation/deallocation of tone buffers.
References
SoC support
esp8266
host
Switch Joycon
Introduction
This sample turns the ESP32 into a Switch Joycon (Bluetooth LE gamepad) that presses buttons and moves axis
Possible buttons are 0 through to 15.
Possible HAT switch position values are: Centered, Up, UpRight, Right, DownRight, Down, DownLeft, Left, UpLeft.
Testing
You can use one of the following applications on your PC to test and see all buttons that were clicked.
On Linux install jstest-gtk to test the ESP32 gamepad. Under Ubuntu this can be done by typing the following command:
sudo apt install jstest-gtk
On Windows use this Windows test application.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Screen TFT_S1D13781
This is a port of the demo from the Epson S1D13781 display controller. See TFT_S1D13781.
Evaluation boards are inexpensive and is a useful way to evaluate display modules with TFT interfaces.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic ControlPoint
Demonstrates use of UPnP library to create and use control points.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic UPnP
Demonstrates use of UPnP library.
The sample devices here can be fully enumerated over the network.
Services
UPnP devices may also provide services which can be enumerated and used to control it.
The Wemo device Wemo::Controllee has two services for events and metadata.
The device and service descriptions are stored in the schema directory.
This is used by UPnP to generate class code, so all we need to do is implement the action methods.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Device
Application demonstrating how to set up USB device with multiple interfaces.
The USB configuration is described in basic_device.usbcfg, and the corresponding
TinyUSB configuration files generated in, for example, out/Rp2040/debug/USB.
C++ classes are provided for some of the standard interfaces
The MIDI device can be tested using the linux aplaymidi utility.
Type aplaymidi -l to show available devices.
Play a MIDI file with aplaymidi -p 24:0 test.mid.
Note
This sample works as-is for the rp2040.
The esp32s2 has limited endpoints so some of the interfaces need to be commented-out in basic_device.usbcfg.
Perhaps try starting with just midi interface.
References
SoC support
esp32s2
esp32s3
host
rp2040
Basic Host
Application demonstrating how to set up USB host with multiple interfaces.
References
SoC support
host
rp2040
CDC MSC HID Host sample
This is a port of the hid_composite example from TinyUSB (https://github.com/hathach/tinyusb/tree/master/examples/host/cdc_msc_hid).
References
SoC support
rp2040
HID Composite Device sample
This is a port of the hid_composite example from TinyUSB (https://github.com/hathach/tinyusb/tree/master/examples/device/hid_composite).
Once connected, the PICO LED will respond to the state of keyboard CAPS LOCK.
References
SoC support
esp32s2
esp32s3
host
rp2040
Basic FlatBuffers Sample
Basic sample demonstrating encoding and decoding of data using flatbuffers. For more complicated examples take a look at the official flatbuffers tutorial.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Advanced_Jsvm
To save space and be able to run JerryScript on an embedded device, this library builds without parser support.
This means that the JavaScript files have to be compiled into ‘snap’ files before landing on the device.
This can be done during normal compilation and flashing, or indirectly via web application.
Manual compilation method
You should have your device connected physically to your computer via USB cable.
In this case the compilation will be done on your computer and the compiled files will be saved to your device.
In order to configure the JavaScript file compilation one can use APP_JS_SOURCE_DIR and APP_JS_SOURCE_DIR variables.
All .js files in the APP_JS_SOURCE_DIR directory will be compiled into .snap files and written to APP_JS_SNAPDIR directory.
Example:
make APP_JS_SOURCE_DIR=files/js APP_JS_SNAP_DIR=out/web
In this advanced sample the variables are directly stored in component.mk file.
The generated .snap files may be flashed or uploaded to your device for execution via any appropriate mechanism.
Web Browser compilation method
Quite often we don’t have physical access to our devices. And here comes the true beauty of this sample. With only network access to the device and a modern browser one can still compile JavaScript files and upload them to the device. No external tools are required.
The advanced sample starts a web server with a simple page.
In it you will see the initial source code of the main.js file.
There are two buttons in the HTML page. The first one starts and stops the JavaScript VM.
If you press Run you will be able to see the result from the execution in the device terminal.
If you have your device connected to your computer then type the following to see the output:
make terminal
If you want to change the JavaScript code and test it again on your device you have to press Stop.
Do directly in the browser the desired changes and press Compile. If the modified JavaScript is valid
then it will be compiled directly in your browser and sent to your device.
Again no external tools are required - just a modern browser.
After that you can press Run in the browser and enjoy the modified result in the terminal.
Credits
The initial work on the JerryScript library for Sming was done as part of the U:Kit project.
References
Environment Variables
GZIP
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic Context
Jerryscript provides the optional External context feature to allow applications to provide an external buffer or isolated engine contexts. The heap size is configured at runtime and enables loading of multiple javascript applications.
This sample demonstrates how to use the Jerryscript JS::Context class to implement containerised applications.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic_Jsvm
This sample demonstrates running JavaScript applications in a sandbox inside a microcontroller. The sample uses the jerryscript library for Sming.
To save space and be able to run JerryScript on an embedded device Sming compiles this library without a parser. This means that the JavaScript files have to be compiled before landing on the device. This sample demonstrates how this is done automatically for you.
Snapshots
The initial JavaScript code used in this sample is located in files/main.js.
The code has one global variable x and two functions setup and loop.
As shown below:
var x;
/**
* This is a function that will be called once.
*/
function setup() {
x = 0;
print ('Setup: X='+x);
}
/**
* This is a function that will be called every half a second
*/
function loop() {
x++;
print('Loop: X='+x);
}
Once compiled the source code will be converted into a byte-code that will be executed inside of your microcontroller.
In JerryScript this final byte-code is called snapshot or just snap.
Calls
This sample will run first the setup JavaScript function and then execute in an endless loop the loop function.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Event_Jsvm
This sample demonstrates how to run your JavaScript application as performing some action in response to an event, input or stimulus. Since Sming framework itself has an event-driven architecture such JavaScript applications will be able to use all the advantages that Sming has in terms of CPU, RAM and power usage.
JavaScript Code
Similar to Sming this sample calls the init JavaScript function. And in this function
we register the events that we are interested in. As shown below.
/**
* This is a function that will be called once.
*/
function init() {
x = 0;
print('Init: X=' + x);
// This is how we register an event listener in JavaScript
addEventListener("EVENT_TEMP", function(event) {
// An event has a name and multiple params
// the params have string keys and string values
print('Event name: ' + event.name + ', value: ' + event.params['temp']);
});
// ...
}
The addEventListener mimics the browser addEventListener function.
It accepts two parameters - a unique event name and a callable function. Similar to the browser function
you can add multiple callable functions to one event.
C/C++ Code
The addEventListener function is not present in the standard JerryScript VM. In this sample we register it as a built-in JavaScript function.
Excerpt from the application.cpp file:
void startJsvm()
{
/*
* This is how we register a new function in JavaScript
* that will communicate directly with our C/C++ code.
*/
jsVm.registerFunction("addEventListener", addEventListener);
// ...
The actual implementation of the addEventListener C++ function is in the vm_functions.cpp file.
There you will find also the triggerEvent function which is available only in the C/C++ code.
The latter is used to trigger events from the C/C++ code inside the JavaScript application:
JsEventData params;
params["temp"]="20";
triggerEvent("EVENT_TEMP", params);
The code above simulates a temperature sensor sending the temperature in Celsius (C).
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Modbusino RTU generic sample
The sample code provides three element uint16_t array used for the modbus slave registers. These values are printed each second through UART1. mbPrint() is triggered on valid write multiple registers (0x10) command.
Several environment variables (envvar:RS485_RE_PIN RS485_TX_LEVEL) can be used for configuration.
The slave address is defined with envvar:MB_SLAVE_ADDR. For example:
make MB_SLAVE_ADDR=2
These variables can be listed with:
make list-config
References
Environment Variables
MB_SLAVE_ADDR
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Basic_VM
This sample is a starting point for working with the rBPF system on Sming.
See rBPF Femto-Container support for toolchain setup.
The sample contains three functions:
increment()simply adds 1 to the parameter value and returns it.multiply()instead stores the output value in the context parameters.store()demonstrates parameter passing using the stores.
The source code for these can be found in the container subdirectory.
Each file contains a single function which is compiled into eRBF code for execution by the virtual machine.
Build the sample like any regular Sming application by running make. You can try it out without any hardware using the Host Emulator:
make SMING_ARCH=Host
make run
You should see this:
All up, running the Femto-Container application now
Calling 'increment()' in VM
input 0, result 1, expected 1
Calling 'multiply()' in VM
input (120000005, 120000023), output 14400003360000115, expected 14400003360000115, result 0
Calling 'store()' in VM
output (1001234, 2005678), result (1234, 5678)
Note that if a runtime error occurs then an appropriate error message is displayed.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
UserCalls
This sample demonstrates how to provide application functions for use by VM containers.
- Declare calls
This sample provides three application calls:
int bpf_user_get_magic(); void bpf_user_set_magic(int value); size_t bpf_user_send_packet(char* data, size_t len) \ size_t bpf_user_read_packet(char* buffer, size_t len)
These are declared using the
BPF_SYSCALL_APPmacro.The
include/bpf_appcalls.hmust be provided in the container directory.
- Implement calls
The application must provide implementations for the declared functions in the
rBPF::VMnamespace. Implementations are identical to the VM calls but with an additional pointer to the VM :cpp:typedef:`bpf_t` instance:int bpf_user_get_magic(bpf_t* bpf); void bpf_user_set_magic(bpf_t* bpf, int value); size_t bpf_user_send_packet(bpf_t* bpf, char* data, size_t len) \ size_t bpf_user_read_packet(bpf_t* bpf, char* buffer, size_t len)
Note that when passing pointers these must be checked using either
bpf_store_allowed()orbpf_load_allowed().See
appcalls.cppfor the implementations.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Libraries
Sming comes with a number of ported libraries that cover many areas of embedded programming but for sure not all.
To reduce the size of the framework, some libraries are not included directly and must be retrieved if you wish to access any sample applications there.
For example, the Basic UPnP sample application is contained in the UPnP library, which can be retrieved like this:
cd $SMING_HOME
make fetch UPnP
You should then get UPnP: found in ‘Libraries/UPnP’, so now we can build the sample:
cd Libraries/UPnP/samples/Basic_UPnP
make
Note
If your project references any libraries they will automatically be pulled in during the build.
These are all the libraries included with Sming:
AM2321 Temperature/Humidity Sensor
AM2321 sensor support for Arduino.
See Arduino AM2321
Example:
#include <AM2321.h>
void readByAM2321()
{
AM2321 am2321;
am2321.read();
Serial.print("(");
Serial.print(am2321.temperature/10.0);
Serial.print(", ");
Serial.print(am2321.humidity/10.0);
Serial.println(')');
}
References
Used by
AM2321 Humidity Sensor Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
APA102 LED
APA102 library by HappyCodingRobot@github.com
APA102 class enhanced to work with hardware SPI and software SPI. Changed to work with the new SPI implementation.
See Pull Request #787.
References
Used by
Basic APA102 Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Adafruit BME280 Library
This is a library for the Adafruit BME280 Humidity, Barometric Pressure + Temp sensor
Designed specifically to work with the Adafruit BME280 Breakout.
Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
About this Driver
Written by Ladyada for Adafruit Industries.
BSD license, check license.txt for more information
All text above must be included in any redistribution
References
Used by
BME280 Humidity Sensor Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Adafruit Bus IO Library
This is a helper library to abstract away I2C & SPI transactions and registers
Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
MIT license, all text above must be included in any redistribution
References
Used by
Adafruit BME280 Library Library
Adafruit GFX Library Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Adafruit GFX Library
This is the core graphics library for all our displays, providing a common set of graphics primitives (points, lines, circles, etc.). It needs to be paired with a hardware-specific library for each display device we carry (to handle the lower-level functions).
Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
Written by Limor Fried/Ladyada for Adafruit Industries. BSD license, check license.txt for more information. All text above must be included in any redistribution.
Recent Arduino IDE releases include the Library Manager for easy installation. Otherwise, to download, click the DOWNLOAD ZIP button, uncompress and rename the uncompressed folder Adafruit_GFX. Confirm that the Adafruit_GFX folder contains Adafruit_GFX.cpp and Adafruit_GFX.h. Place the Adafruit_GFX library folder your ArduinoSketchFolder/Libraries/ folder. You may need to create the Libraries subfolder if its your first library. Restart the IDE.
You will also need to install the latest Adafruit BusIO library. Search for “Adafruit BusIO” in the library manager, or install by hand from https://github.com/adafruit/Adafruit_BusIO
Useful Resources
Image2Code: This is a handy Java GUI utility to convert a BMP file into the array code necessary to display the image with the drawBitmap function. Check out the code at ehubin’s GitHub repository: https://github.com/ehubin/Adafruit-GFX-Library/tree/master/Img2Code
drawXBitmap function: You can use the GIMP photo editor to save a .xbm file and use the array saved in the file to draw a bitmap with the drawXBitmap function. See the pull request here for more details: https://github.com/adafruit/Adafruit-GFX-Library/pull/31
‘Fonts’ folder contains bitmap fonts for use with recent (1.1 and later) Adafruit_GFX. To use a font in your Arduino sketch, #include the corresponding .h file and pass address of GFXfont struct to setFont(). Pass NULL to revert to ‘classic’ fixed-space bitmap font.
‘fontconvert’ folder contains a command-line tool for converting TTF fonts to Adafruit_GFX header format.
You can also use this GFX Font Customiser tool (*web version here*) to customize or correct the output from fontconvert, and create fonts with only a subset of characters to optimize size.
Roadmap
The PRIME DIRECTIVE is to maintain backward compatibility with existing Arduino sketches – many are hosted elsewhere and don’t track changes here, some are in print and can never be changed! This “little” library has grown organically over time and sometimes we paint ourselves into a design corner and just have to live with it or add progressively more ungainly workarounds.
We are grateful for everyone’s contributions, but pull requests for the following will NOT be merged:
Additional or incompatible font formats (see Prime Directive above). There are already two formats and the code is quite bloaty there as it is. This also creates liabilities for tools and documentation. What’s there isn’t perfect but it does the job.
Additional or incompatible bitmap formats, for similar reasons. It’s getting messy.
Adding background color to custom fonts to erase prior screen contents. The ONLY acceptable methods are to clear the area with a filled rect, or (to avoid flicker) draw text into a GFXcanvas1 and copy to screen with drawBitmap() w/background color. This is on purpose and by design. We’ve discussed this. Glyphs can overlap.
Scrolling, whether hardware- or software-based. Such implementations tend to rely on hardware-specific features (not universally available), read access to the screen’s framebuffer (ditto) and/or the addition of virtual functions in GFX which them must be added in every subclass, of which there are many. The GFX API is largely “set” at this point and this is just a limitation we live with now.
Please don’t reformat code for the sake of reformatting code. The resulting large “visual diff” makes it impossible to untangle actual bug fixes from merely rearranged lines. clang-format will be the final arbiter.
Please no more pentagram-drawing PRs. Any oddly-specific drawing functions can go in your own code and aren’t helpful in a library context.
If you must have one of these features, consider creating a fork with the features required for your project…it’s easy to keep synced with the upstream code.
References
Used by
Adafruit ILI9341 Arduino Library Library
Adafruit_SSD1306 Library
Adafruit ST7735 Display Library
ILI9163C TFT Display Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Adafruit ILI9341 Arduino Library
This is a library for the Adafruit ILI9341 display products
This library works with the Adafruit 2.8” Touch Shield V2 (SPI)
Adafruit 2.4” TFT LCD with Touchscreen Breakout w/MicroSD Socket - ILI9341
2.8” TFT LCD with Touchscreen Breakout Board w/MicroSD Socket - ILI9341
2.2” 18-bit color TFT LCD display with microSD card breakout - ILI9340
TFT FeatherWing - 2.4” 320x240 Touchscreen For All Feathers
Check out the links above for our tutorials and wiring diagrams. These displays use SPI to communicate, 4 or 5 pins are required to interface (RST is optional).
BMP image-loading examples have been moved to the Adafruit_ImageReader library: https://github.com/adafruit/Adafruit_ImageReader
Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
Written by Limor Fried/Ladyada for Adafruit Industries. MIT license, all text above must be included in any redistribution
To download. click the DOWNLOADS button in the top right corner, rename the uncompressed folder Adafruit_ILI9341. Check that the Adafruit_ILI9341 folder contains Adafruit_ILI9341.cpp and Adafruit_ILI9341.
Place the Adafruit_ILI9341 library folder your arduinosketchfolder/libraries/ folder. You may need to create the libraries subfolder if its your first library. Restart the IDE
Also requires the Adafruit_GFX library for Arduino.
References
Used by
Environment Variables
TFT_CS_PINTFT_DC_PINTFT_RESET_PIN
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Adafruit NeoPixel Library
Arduino library for controlling single-wire-based LED pixels and strip such as the Adafruit 60 LED/meter Digital LED strip, the Adafruit FLORA RGB Smart Pixel, the Adafruit Breadboard-friendly RGB Smart Pixel, the Adafruit NeoPixel Stick, and the Adafruit NeoPixel Shield.
After downloading, rename folder to ‘Adafruit_NeoPixel’ and install in Arduino Libraries folder. Restart Arduino IDE, then open File->Sketchbook->Library->Adafruit_NeoPixel->strandtest sketch.
Compatibility notes: Port A is not supported on any AVR processors at this time
Features
### Simple to use
Controlling NeoPixels “from scratch” is quite a challenge, so we provide a library letting you focus on the fun and interesting bits.
### Give back
The library is free; you don’t have to pay for anything. Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
### Supported Chipsets
We have included code for the following chips - sometimes these break for exciting reasons that we can’t control in which case please open an issue!
AVR ATmega and ATtiny (any 8-bit) - 8 MHz, 12 MHz and 16 MHz
Teensy 3.x and LC
Arduino Due
Arduino 101
ATSAMD21 (Arduino Zero/M0 and other SAMD21 boards) @ 48 MHz
ATSAMD51 @ 120 MHz
Adafruit STM32 Feather @ 120 MHz
ESP8266 any speed
ESP32 any speed
Nordic nRF52 (Adafruit Feather nRF52), nRF51 (micro:bit)
Infineon XMC1100 BootKit @ 32 MHz
Infineon XMC1100 2Go @ 32 MHz
Infineon XMC1300 BootKit @ 32 MHz
Infineon XMC4700 RelaxKit, XMC4800 RelaxKit, XMC4800 IoT Amazon FreeRTOS Kit @ 144 MHz
Check forks for other architectures not listed here!
### GNU Lesser General Public License
Adafruit_NeoPixel is free software: you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
Functions
begin()
updateLength()
updateType()
show()
delay_ns()
setPin()
setPixelColor()
fill()
ColorHSV()
getPixelColor()
setBrightness()
getBrightness()
clear()
gamma32()
Examples
There are many examples implemented in this library. One of the examples is below. You can find other examples here
Simple
#include <Adafruit_NeoPixel.h>
#ifdef __AVR__
#include <avr/power.h>
#endif
#define PIN 6
#define NUMPIXELS 16
Adafruit_NeoPixel pixels(NUMPIXELS, PIN, NEO_GRB + NEO_KHZ800);
#define DELAYVAL 500
void setup() {
#if defined(__AVR_ATtiny85__) && (F_CPU == 16000000)
clock_prescale_set(clock_div_1);
#endif
pixels.begin();
}
void loop() {
pixels.clear();
for(int i=0; i<NUMPIXELS; i++) {
pixels.setPixelColor(i, pixels.Color(0, 150, 0));
pixels.show();
delay(DELAYVAL);
}
}
Contributing
If you want to contribute to this project:
Report bugs and errors
Ask for enhancements
Create issues and pull requests
Tell others about this library
Contribute new protocols
Please read CONTRIBUTING.md for details on our code of conduct, and the process for submitting pull requests to us.
Roadmap
The PRIME DIRECTIVE is to maintain backward compatibility with existing Arduino sketches – many are hosted elsewhere and don’t track changes here, some are in print and can never be changed!
Please don’t reformat code for the sake of reformatting code. The resulting large “visual diff” makes it impossible to untangle actual bug fixes from merely rearranged lines. (Exception for first item in wishlist below.)
Things I’d Like To Do But There’s No Official Timeline So Please Don’t Count On Any Of This Ever Being Canonical:
For the show() function (with all the delicate pixel timing stuff), break out each architecture into separate source files rather than the current unmaintainable tangle of #ifdef statements!
Please don’t use updateLength() or updateType() in new code. They should not have been implemented this way (use the C++ ‘new’ operator with the regular constructor instead) and are only sticking around because of the Prime Directive. setPin() is OK for now though, it’s a trick we can use to ‘recycle’ pixel memory across multiple strips.
In the M0 and M4 code, use the hardware systick counter for bit timing rather than hand-tweaked NOPs (a temporary kludge at the time because I wasn’t reading systick correctly). (As of 1.4.2, systick is used on M4 devices and it appears to be overclock-compatible. Not for M0 yet, which is why this item is still here.)
As currently written, brightness scaling is still a “destructive” operation – pixel values are altered in RAM and the original value as set can’t be accurately read back, only approximated, which has been confusing and frustrating to users. It was done this way at the time because NeoPixel timing is strict, AVR microcontrollers (all we had at the time) are limited, and assembly language is hard. All the 32-bit architectures should have no problem handling nondestructive brightness scaling – calculating each byte immediately before it’s sent out the wire, maintaining the original set value in RAM – the work just hasn’t been done. There’s a fair chance even the AVR code could manage it with some intense focus. (The DotStar library achieves nondestructive brightness scaling because it doesn’t have to manage data timing so carefully…every architecture, even ATtiny, just takes whatever cycles it needs for the multiply/shift operations.)
Credits
This library is written by Phil “Paint Your Dragon” Burgess for Adafruit Industries, with contributions by PJRC, Michael Miller and other members of the open source community.
License
Adafruit_NeoPixel is free software: you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. Adafruit_NeoPixel is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with NeoPixel. If not, see this
References
Used by
Basic Neopixel Sample
SoC support
esp32
esp32c3
esp32s2
esp32s3
esp8266
rp2040
Adafruit PCD8544 Nokia 5110 LCD Library
This is a library for our Monochrome Nokia 5110 LCD Displays
Pick one up today in the Adafruit shop! ——> https://www.adafruit.com/product/338
These displays use SPI to communicate, 4 or 5 pins are required to
interface.
Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
Written by Limor Fried/Ladyada for Adafruit Industries.
BSD license, check license.txt for more information.
All text above must be included in any redistribution.
To install, use the Arduino Library Manager to search for ‘Adafruit VCNL4010’ and install the library.
You will also need the Adafruit GFX Graphics core which does all the circles, text, rectangles, etc. You can get it by searching for ‘Adafruit GFX’ in the Arduino Library Manager or check it out here: https://github.com/adafruit/Adafruit-GFX-Library
References
Used by
5110 LCD Screen Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
rp2040
Adafruit_SSD1306
This is a library for our Monochrome OLEDs based on SSD1306 drivers
Pick one up today in the adafruit shop! ——> http://www.adafruit.com/category/63_98
These displays use I2C or SPI to communicate, 2 to 5 pins are required to interface.
Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
Written by Limor Fried/Ladyada for Adafruit Industries, with contributions from the open source community. Scrolling code contributed by Michael Gregg. Dynamic buffer allocation based on work by Andrew Canaday. BSD license, check license.txt for more information. All text above must be included in any redistribution
Preferred installation method is to use the Arduino IDE Library Manager. To download the source from Github instead, click “Clone or download” above, then “Download ZIP.” After uncompressing, rename the resulting folder Adafruit_SSD1306. Check that the Adafruit_SSD1306 folder contains Adafruit_SSD1306.cpp and Adafruit_SSD1306.h.
You will also have to install the Adafruit GFX library which provides graphics primitves such as lines, circles, text, etc. This also can be found in the Arduino Library Manager, or you can get the source from https://github.com/adafruit/Adafruit-GFX-Library
Changes
- Pull Request:
(November 2021)
Added define
SSD1306_NO_SPLASHto opt-out of including splash images inPROGMEMand drawing to display duringbegin.
- Pull Request:
(September 2019)
new #defines for SSD1306_BLACK, SSD1306_WHITE and SSD1306_INVERSE that match existing #define naming scheme and won’t conflict with common color names
old #defines for BLACK, WHITE and INVERSE kept for backwards compat (opt-out with #define NO_ADAFRUIT_SSD1306_COLOR_COMPATIBILITY)
Version 1.2 (November 2018) introduces some significant changes:
Display dimensions are now specified in the constructor…you no longer need to edit the .h file for different screens (though old sketches can continue to work that way).
SPI transactions are used and SPI bitrate can be specified (both require Arduino 1.6 or later).
SPI and Wire (I2C) interfaces other than the defaults are supported.
Compatibility
MCU |
Tested Works |
Doesn’t Work |
Not Tested |
Notes |
|---|---|---|---|---|
Atmega328 |
X |
|||
Atmega32u4 |
X |
|||
Atmega2560 |
X |
|||
ESP8266 |
X |
Change OLED_RESET to different pin if using default I2C pins D4/D5. |
||
ESP32 |
X |
|||
ATSAM3X8E |
X |
|||
ATSAMD21 |
X |
|||
Intel Curie |
X |
|||
WICED |
X |
No hardware SPI - bitbang only |
||
ATtiny85 |
X |
|||
Particle |
X |
ATmega328 : Arduino UNO, Adafruit Pro Trinket, Adafruit Metro 328, Adafruit Metro Mini
ATmega32u4 : Arduino Leonardo, Arduino Micro, Arduino Yun, Teensy 2.0, Adafruit Flora, Bluefruit Micro
ATmega2560 : Arduino Mega
ESP8266 : Adafruit Huzzah
ATSAM3X8E : Arduino Due
ATSAMD21 : Arduino Zero, M0 Pro, Adafruit Metro Express, Feather M0
ATtiny85 : Adafruit Gemma, Arduino Gemma, Adafruit Trinket
Particle: Particle Argon
References
Used by
SSD1306 OLED Screen Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Adafruit ST7735 Display
This is a library for several Adafruit displays based on ST77* drivers.
- Works with the Adafruit 1.8” TFT Breakout w/SD card
- The 1.8” TFT shield
- The 1.44” TFT breakout
- as well as Adafruit raw 1.8” TFT display
Check out the links above for our tutorials and wiring diagrams. These displays use SPI to communicate, 4 or 5 pins are required to interface (RST is optional).
Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
Written by Limor Fried/Ladyada for Adafruit Industries. MIT license, all text above must be included in any redistribution.
Recent Arduino IDE releases include the Library Manager for easy installation. Otherwise, to download, click the DOWNLOAD ZIP button, uncompress and rename the uncompressed folder Adafruit_ST7735. Confirm that the Adafruit_ST7735 folder contains Adafruit_ST7735.cpp, Adafruit_ST7735.h and related source files. Place the Adafruit_ST7735 library folder your ArduinoSketchFolder/Libraries/ folder. You may need to create the Libraries subfolder if its your first library. Restart the IDE.
Also requires the Adafruit_GFX library for Arduino.
References
Used by
ST7735 TFT Screen Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Adafruit Unified Sensor Driver
Many small embedded systems exist to collect data from sensors, analyse the data, and either take an appropriate action or send that sensor data to another system for processing.
One of the many challenges of embedded systems design is the fact that parts you used today may be out of production tomorrow, or system requirements may change and you may need to choose a different sensor down the road.
Creating new drivers is a relatively easy task, but integrating them into existing systems is both error prone and time consuming since sensors rarely use the exact same units of measurement.
By reducing all data to a single sensors_event_t ‘type’ and settling on specific, standardised SI units for each sensor family the same sensor types return values that are comparable with any other similar sensor. This enables you to switch sensor models with very little impact on the rest of the system, which can help mitigate some of the risks and problems of sensor availability and code reuse.
The unified sensor abstraction layer is also useful for data-logging and data-transmission since you only have one well-known type to log or transmit over the air or wire.
Unified Sensor Drivers
The following drivers are based on the Adafruit Unified Sensor Driver:
Accelerometers
Gyroscope
Light
Magnetometers
Barometric Pressure
Humidity & Temperature
Humidity, Temperature, & Barometric Pressure
Orientation
All in one device
Adafruit_LSM9DS0 (accelerometer, gyroscope, magnetometer)
Adafruit_LSM9DS1 (accelerometer, gyroscope, magnetometer)
How Does it Work?
Any driver that supports the Adafruit unified sensor abstraction layer will implement the Adafruit_Sensor base class. There are two main typedefs and one enum defined in Adafruit_Sensor.h that are used to ‘abstract’ away the sensor details and values:
Sensor Types (sensors_type_t)
These pre-defined sensor types are used to properly handle the two related typedefs below, and allows us determine what types of units the sensor uses, etc.
/** Sensor types */
typedef enum
{
SENSOR_TYPE_ACCELEROMETER = (1),
SENSOR_TYPE_MAGNETIC_FIELD = (2),
SENSOR_TYPE_ORIENTATION = (3),
SENSOR_TYPE_GYROSCOPE = (4),
SENSOR_TYPE_LIGHT = (5),
SENSOR_TYPE_PRESSURE = (6),
SENSOR_TYPE_PROXIMITY = (8),
SENSOR_TYPE_GRAVITY = (9),
SENSOR_TYPE_LINEAR_ACCELERATION = (10),
SENSOR_TYPE_ROTATION_VECTOR = (11),
SENSOR_TYPE_RELATIVE_HUMIDITY = (12),
SENSOR_TYPE_AMBIENT_TEMPERATURE = (13),
SENSOR_TYPE_VOLTAGE = (15),
SENSOR_TYPE_CURRENT = (16),
SENSOR_TYPE_COLOR = (17)
} sensors_type_t;
Sensor Details (sensor_t)
This typedef describes the specific capabilities of this sensor, and allows us to know what sensor we are using beneath the abstraction layer.
/* Sensor details (40 bytes) */
/** struct sensor_s is used to describe basic information about a specific sensor. */
typedef struct
{
char name[12];
int32_t version;
int32_t sensor_id;
int32_t type;
float max_value;
float min_value;
float resolution;
int32_t min_delay;
} sensor_t;
The individual fields are intended to be used as follows:
name: The sensor name or ID, up to a maximum of twelve characters (ex. “MPL115A2”)
version: The version of the sensor HW and the driver to allow us to differentiate versions of the board or driver
sensor_id: A unique sensor identifier that is used to differentiate this specific sensor instance from any others that are present on the system or in the sensor network
type: The sensor type, based on sensors_type_t in sensors.h
max_value: The maximum value that this sensor can return (in the appropriate SI unit)
min_value: The minimum value that this sensor can return (in the appropriate SI unit)
resolution: The smallest difference between two values that this sensor can report (in the appropriate SI unit)
min_delay: The minimum delay in microseconds between two sensor events, or ‘0’ if there is no constant sensor rate
Sensor Data/Events (sensors_event_t)
This typedef is used to return sensor data from any sensor supported by the abstraction layer, using standard SI units and scales.
/* Sensor event (36 bytes) */
/** struct sensor_event_s is used to provide a single sensor event in a common format. */
typedef struct
{
int32_t version;
int32_t sensor_id;
int32_t type;
int32_t reserved0;
int32_t timestamp;
union
{
float data[4];
sensors_vec_t acceleration;
sensors_vec_t magnetic;
sensors_vec_t orientation;
sensors_vec_t gyro;
float temperature;
float distance;
float light;
float pressure;
float relative_humidity;
float current;
float voltage;
sensors_color_t color;
};
} sensors_event_t;
It includes the following fields:
version: Contain ‘sizeof(sensors_event_t)’ to identify which version of the API we’re using in case this changes in the future
sensor_id: A unique sensor identifier that is used to differentiate this specific sensor instance from any others that are present on the system or in the sensor network (must match the sensor_id value in the corresponding sensor_t enum above!)
type: the sensor type, based on sensors_type_t in sensors.h
timestamp: time in milliseconds when the sensor value was read
data[4]: An array of four 32-bit values that allows us to encapsulate any type of sensor data via a simple union (further described below)
Required Functions
In addition to the two standard types and the sensor type enum, all drivers based on Adafruit_Sensor must also implement the following two functions:
bool getEvent(sensors_event_t*);
Calling this function will populate the supplied sensors_event_t reference with the latest available sensor data. You should call this function as often as you want to update your data.
void getSensor(sensor_t*);
Calling this function will provide some basic information about the sensor (the sensor name, driver version, min and max values, etc.
Standardised SI values for sensors_event_t
A key part of the abstraction layer is the standardisation of values on SI units of a particular scale, which is accomplished via the data[4] union in sensors_event_t above. This 16 byte union includes fields for each main sensor type, and uses the following SI units and scales:
acceleration: values are in meter per second per second (m/s^2)
magnetic: values are in micro-Tesla (uT)
orientation: values are in degrees
gyro: values are in rad/s
temperature: values in degrees centigrade (Celsius)
distance: values are in centimeters
light: values are in SI lux units
pressure: values are in hectopascal (hPa)
relative_humidity: values are in percent
current: values are in milliamps (mA)
voltage: values are in volts (V)
color: values are in 0..1.0 RGB channel luminosity and 32-bit RGBA format
The Unified Driver Abstraction Layer in Practice
Using the unified sensor abstraction layer is relatively easy once a compliant driver has been created.
Every compliant sensor can now be read using a single, well-known ‘type’ (sensors_event_t), and there is a standardised way of interrogating a sensor about its specific capabilities (via sensor_t).
An example of reading the TSL2561 light sensor can be seen below:
Adafruit_TSL2561 tsl = Adafruit_TSL2561(TSL2561_ADDR_FLOAT, 12345);
...
/* Get a new sensor event */
sensors_event_t event;
tsl.getEvent(&event);
/* Display the results (light is measured in lux) */
if (event.light)
{
Serial.print(event.light); Serial.println(" lux");
}
else
{
/* If event.light = 0 lux the sensor is probably saturated
and no reliable data could be generated! */
Serial.println("Sensor overload");
}
Similarly, we can get the basic technical capabilities of this sensor with the following code:
sensor_t sensor;
sensor_t sensor;
tsl.getSensor(&sensor);
/* Display the sensor details */
Serial.println("------------------------------------");
Serial.print ("Sensor: "); Serial.println(sensor.name);
Serial.print ("Driver Ver: "); Serial.println(sensor.version);
Serial.print ("Unique ID: "); Serial.println(sensor.sensor_id);
Serial.print ("Max Value: "); Serial.print(sensor.max_value); Serial.println(" lux");
Serial.print ("Min Value: "); Serial.print(sensor.min_value); Serial.println(" lux");
Serial.print ("Resolution: "); Serial.print(sensor.resolution); Serial.println(" lux");
Serial.println("------------------------------------");
Serial.println("");
References
Used by
Adafruit BME280 Library Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Adafruit VL53L0X Library
This is a library for the Adafruit VL53L0X time-of-flight breakout:
Check out the links above for our tutorials and wiring diagrams. This chip uses I2C to communicate
Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
Written by Limor Fried/Ladyada for Adafruit Industries. MIT license, all text above must be included in any redistribution
References
Used by
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
AnimatedGIF
Introduction
AnimatedGIF is an optimized library for playing animated GIFs on embedded devices.
Features
Supports any MCU with at least 24K of RAM (Cortex-M0+ is the simplest I’ve tested).
Optimized for speed; the main limitation will be how fast you can copy the pixels to the display. You can use SPI+DMA to help.
GIF image data can come from memory (FLASH/RAM), SDCard or any media you provide.
GIF files can be any length, (e.g. hundreds of megabytes)
Simple C++ class and callback design allows you to easily add GIF support to any application.
The C code doing the heavy lifting is completely portable and has no external dependencies.
Does not use dynamic memory (malloc/free/new/delete), so it’s easy to build it for a minimal bare metal system.
Using
Add
COMPONENT_DEPENDS += AnimatedGIFto your application componenent.mk file.Add these lines to your application:
#include <AnimatedGifTask.h> namespace { AnimatedGifTask* task; // ... } // namespace void init() { // ... initDisplay(); tft.setOrientation(Graphics::Orientation::deg270); auto surface = tft.createSurface(); assert(surface != nullptr); task = new AnimatedGifTask(*surface, gifData); task->resume(); }
References
Used by
Basic AnimatedGIF Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: AnimatedGIF
AnimatedGIF
Copyright (c) 2020 BitBank Software, Inc.
Written by Larry Bank
larry@bitbanksoftware.com
I optimize other people’s code for a living. This library is a good example of the kind of work I do for my commercial clients; it contains many unique and clever optimizations that allows it to perform better than anything else available. I’m happy to contribute optimized libraries to the open source community in addition to working on commercial projects. Whatever platform you’re using, I can make significant improvements to your native code. Please contact me so that I can show you how.
I started working with image and video files around 1989 and soon turned my interest into a Document Imaging product based on my own imaging library. Over the years I added support for more and more formats until I had supported all of the standard ones, including DICOM. I sold my Document Imaging business in 1997, but still found uses for my imaging code in other projects such as retro gaming. I recently went looking to see if there was a good GIF player for Arduino and only found Adafruit’s library. Unfortunately it only runs on their ATSAMD51 boards. I thought it would be possible to use my old code to create a universal GIF player that could run on any MCU/SoC with at least 24K of RAM, so I started modifying my old GIF code for this purpose. The focus of this project is speed and to use as little RAM as possible, so there are some limitations.
Limitations
To save memory, the code limits the maximum image width to 320 pixels. This is just a constant defined in AnimatedGIF.h, so you can set it larger if you like. Animated GIF images have a lot of options to save space when encoding the changes from frame to frame. Many of the tools which generate animated GIFs don’t make use of every option and that’s helpful because the frame disposal options are not supported in the library code. They can be implemented in your drawing callback if you want. The reason they’re not provided with this code is because one of the image disposal options requires you to keep an entire copy of the previous frame and this would prevent it from working on low memory MCUs. If would also require dynamic memory allocation which is another area I avoided with this code.
Designed for Speed
My work is always focused on code optimization, so this project is no different. I’ve profiled this code and optimized it for 32-bit CPUs. I discovered while testing it that seeking on micro SD cards (on Arduino) is very very slow. I had to use a little extra RAM to buffer incoming data to avoid seeking. The code does need to seek, but it’s done very rarely. I also wrote code to buffer and de-chunk some of the incoming LZW data to avoid checking for chunk boundaries in the inner decode loop. There are a number of clever optimizations that should allow this to run faster than any existing GIF solutions on Arduino boards. The speed gained from my decoder will be lost if it takes too long to display the pixels. For this reason, the GIFDRAW callback passes an RGB565 palette in the byte order of your choosing and the example sketches uses functions to write entire lines of pixels to the SPI TFT displays in a single shot. If you implement your own GIFDRAW callback and have to pass 1 pixel at a time to whatever display device you’re using, this will cause a major slowdown in the display of the frames.
A note about performance
The chart above shows the total time to decode an 8-frame sequence included in the test_images folder. The decoding speed of your particular image depends on the complexity (how much compressed data) and if/how transparent pixels are used. Small runs of transparent pixels will interfere with the performance of displaying the image on an SPI LCD. This particular image doesn’t use transparency, so the time is purely for decoding the data.
Features:
Supports any MCU with at least 24K of RAM (Cortex-M0+ is the simplest I’ve tested).
Optimized for speed; the main limitation will be how fast you can copy the pixels to the display. You can use SPI+DMA to help.
GIF image data can come from memory (FLASH/RAM), SDCard or any media you provide.
GIF files can be any length, (e.g. hundreds of megabytes)
Simple C++ class and callback design allows you to easily add GIF support to any application.
The C code doing the heavy lifting is completely portable and has no external dependencies.
Does not use dynamic memory (malloc/free/new/delete), so it’s easy to build it for a minimal bare metal system.
Acquiring GIF files to play:
You’ll notice that the images provided in the test_images folder have been turned into C code. Each byte is now in the form 0xAB so that it can be compiled into your program and stored in FLASH memory alongside your other code. You can use a command I wrote called image_to_c (https://github.com/bitbank2/image_to_c) to convert a binary file into this type of text. If you use another tool, make sure to add a const modifier in front of the GIF data array to ensure that it gets written to FLASH and not RAM by your build environment.
The Callback functions:
One of the reasons that this is apparently the first universal GIF library for Arduino is because the lack of available RAM and myriad display options would make it difficult to support all MCUs and displays properly. I decided that to solve this issue, I would isolate the GIF decoding from the display and file I/O with callback functions. This allows the core code to run on any system, but you need to help it a little. At a minimum, your code must provide a function to draw (or store) each scan line of image. If you’re playing a GIF file from memory, this is the only function you need to provide. In the examples folder there are multiple sketches to show how this is done on various display libraries. For reading from SD cards, 4 other functions must be provided: open, close, read, seek. There is an example for implementing these in the examples folder as well.
Note:
If you’re using the ESP32 or ESP8266 and playing GIF images stored in RAM, you’ll need to provide the 4 file callback functions or modify the existing ones because RAM and FLASH are in different adddress spaces (Harvard architecture). The code assumes the source of the GIF data is in FLASH and uses memcpy_P() instead of memcpy() to access it.
The API:
Please consult the Wiki for detailed info about each method exposed by the AnimatedGIF class.
If you find this code useful, please consider sending a donation or becoming a Github sponsor.
ArduCAM Library
Introduction
This is a opensource library for taking high resolution still images and short video clip on Arduino based platforms using ArduCAM’s camera modules.
The camera breakout boards should work with ArduCAM shield before connecting to the Arduino boards.
ArduCAM mini series camera modules like Mini-2MP, Mini-5MP(Plus) can be connected to Arduino boards directly.
In addition to Arduino, the library can be ported to any hardware platforms as long as they have I2C and SPI interface based on this ArduCAM library.
Now Supported Cameras
OV7660 0.3MP
OV7670 0.3MP
OV7675 0.3MP
OV7725 0.3MP
MT9V111 0.3MP
MT9M112 1.3MP
MT9M001 1.3MP
MT9D111 2MP
OV2640 2MP JPEG
MT9T112 3MP
OV3640 3MP
OV5642 5MP JPEG
OV5640 5MP JPEG
Supported MCU Platform
Theoretically support all Arduino families
Arduino UNO R3 (Tested)
Arduino MEGA2560 R3 (Tested)
Arduino Leonardo R3 (Tested)
Arduino Nano (Tested)
Arduino DUE (Tested)
Arduino Genuion 101 (Tested)
Raspberry Pi (Tested)
ESP8266-12 (Tested) (http://www.arducam.com/downloads/ESP8266_UNO/package_ArduCAM_index.json)
Feather M0 (Tested with OV5642)
Note: ArduCAM library for ESP8266 is maintained in another repository ESP8266 using a json board manager script.
Libraries Structure
The basic libraries are composed by two sub-libraries one is ArduCAM and the other is UTFT4ArduCAM_SPI. These two libraries should be copied right under the libraries of Arduino directory in order to be recognized by the Arduino IDE.
The ArduCAM library is the core library for ArduCAM shields. It contains supported image sensor drivers and user land API functions which issue capture or image data read commands .There is also an example directory inside the ArduCAM library which illustrates most
function of the ArduCAM shields. The existing examples are plug and play without need to write a single line of code.
The UTFT4ArduCAM_SPI library is modified version of UTFT which is written by Henning Karlsen. We ported it to support ArduCAM
shield with LCD screen. So the UTFT4ArduCAM_SPI library is only needed when using the ArduCAM-LF model.
How to use
The libraries should be configured before running any examples, or else you will get a compilation error message.
1. Edit memorysaver.h file
Open the memorysaver.h file in the ArduCAM folder and enable the hardware platform and camera module which matches to your hardware by comment or
uncomment the macro definition in the file. For example, if you got a ArduCAM-Mini-2MP you
should uncomment the line #define OV2640_MINI_2MP and comment all the other lines. And
if you got a ArduCAM-Shield-V2 and a OV5642 camera module, you should uncomment the line #define ARDUCAM_SHIELD_V2
and the line #define OV5642_CAM then comment other lines.
2. Choose correct CS pin for your camera
Open one of the examples, wiring SPI and I2C interface especially CS pins to ArduCAM shield according to the examples. Hardware and software should be consistent to run the examples correctly.
3. Upload the examples
In the example folder there are seven sub directories for different ArduCAM models and the host application.
The Mini folder is for ArduCAM-Mini-2MP and ArduCAM-Mini-5MP modules.
The Mini_5MP_Plus folder is for ArduCAM-Mini-5MP-Plus (OV5640/OV5642) modules.
The RevC folder is for ArduCAM-Shield-RevC or ArduCAM-Shield-RevC+ shields.
The Shield_V2 folder is for ArduCAM-Shield-V2 shield.
The host_app folder is host capture and display application for all of ArduCAM modules.
The RaspberryPi folder is examples used for Raspberry Pi platform, see more instruction.
The ESP8266 folder is for ArduCAM-ESP8266-UNO board examples for library compatibility. Please try repository ESP8266 using josn board manager script instead.
Selecting correct COM port and Arduino boards then upload the sketches.
Arducam MINI Camera Demo Tutorial for Arduino
Arducam Camera Shield V2 Demo Tutorial for Arduino
4. How To Connect Bluetooth Module
Using this demo
https://github.com/ArduCAM/Arduino/blob/master/ArduCAM/examples/mini/ArduCAM_Mini_Video_Streaming_Bluetooth/ArduCAM_Mini_Video_Streaming_Bluetooth.ino
5. How to download the Host V2 ?
For ArduCAM_Host_V2.0_Mac.app, please refer to this link: https://www.arducam.com/downloads/app/ArduCAM_Host_V2.0_Mac.app.zip
For ArduCAM_Mini_V2.0_Linux_x86_64bit, Please refer to this link: https://www.arducam.com/downloads/app/ArduCAM_Mini_V2.0_Linux_x86_64bit.zip
References
Used by
Arducam Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
arduinoFFT
Fast Fourier Transform for Arduino
This is a fork from https://code.google.com/p/makefurt/ which has been abandoned since 2011.
This is a C++ library for Arduino for computing FFT. Now it works both on Arduino and C projects.
Tested on Arduino 1.6.11
Installation on Arduino
Use the Arduino Library Manager to install and keep it updated. Just look for arduinoFFT. Only for Arduino 1.5+
Manual installation on Arduino
To install this library, just place this entire folder as a subfolder in your Arduino installation
When installed, this library should look like:
ArduinolibrariesarduinoFTT (this library’s folder) ArduinolibrariesarduinoFTTarduinoFTT.cpp (the library implementation file, uses 32 bits floats vectors) ArduinolibrariesarduinoFTTarduinoFTT.h (the library header file, uses 32 bits floats vectors) ArduinolibrariesarduinoFTTkeywords.txt (the syntax coloring file) ArduinolibrariesarduinoFTTexamples (the examples in the “open” menu) ArduinolibrariesarduinoFTTreadme.md (this file)
Building on Arduino
After this library is installed, you just have to start the Arduino application. You may see a few warning messages as it’s built.
To use this library in a sketch, go to the Sketch | Import Library menu and select arduinoFTT. This will add a corresponding line to the top of your sketch:
#include <arduinoFTT.h>
TODO
Ratio table for windowing function.
Document windowing functions advantages and disadvantages.
Optimize usage and arguments.
Add new windowing functions.
* Spectrum table?
API
arduinoFFT(void);
arduinoFFT(double *vReal, double *vImag, uint16_t samples, double samplingFrequency); Constructor
~arduinoFFT(void); Destructor
ComplexToMagnitude(double *vReal, double *vImag, uint16_t samples);
ComplexToMagnitude();
Compute(double *vReal, double *vImag, uint16_t samples, uint8_t dir);
Compute(double *vReal, double *vImag, uint16_t samples, uint8_t power, uint8_t dir);
Compute(uint8_t dir); Calcuates the Fast Fourier Transform.
DCRemoval(double *vData, uint16_t samples);
DCRemoval(); Removes the DC component from the sample data.
MajorPeak(double *vD, uint16_t samples, double samplingFrequency);
MajorPeak();
MajorPeakParabola(); Looks for and returns the frequency of the biggest spike in the analyzed signal.
Revision(void); Returns the library revision.
Windowing(double *vData, uint16_t samples, uint8_t windowType, uint8_t dir);
Windowing(uint8_t windowType, uint8_t dir); Performs a windowing function on the values array. The possible windowing options are:
FFT_WIN_TYP_RECTANGLE
FFT_WIN_TYP_HAMMING
FFT_WIN_TYP_HANN
FFT_WIN_TYP_TRIANGLE
FFT_WIN_TYP_NUTTALL
FFT_WIN_TYP_BLACKMAN
FFT_WIN_TYP_BLACKMAN_NUTTALL
FFT_WIN_TYP_BLACKMAN_HARRIS
FFT_WIN_TYP_FLT_TOP
FFT_WIN_TYP_WELCH
Exponent(uint16_t value); Calculates and returns the base 2 logarithm of the given value.
References
Used by
Basic Tasks Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
ArduinoJson Version 5
Provided to support existing applications. New projects should use ArduinoJson Version 6.
Attention
Issue with JSON keys (applies to version 5 only)
According to the ArduinoJson docs it should take an internal copy of
char* strings, but it doesn’t! This can occur using the _F() macro:
root[_F("offset")] = something;
This won’t work. Instead, use the F() macro:
root[F("offset")] = something;
References
Used by
Basic Web Skeleton (LTS) Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: ArduinoJson
ArduinoJson is a C++ JSON library for Arduino and IoT (Internet Of Things).
Features
JSON decoding (comments are supported)
JSON encoding (with optional indentation)
Elegant API, easy to use
Fixed memory allocation (zero malloc)
No data duplication (zero copy)
Portable (written in C++98, can be used in any C++ project)
Self-contained (no external dependency)
Small footprint
Input and output streams
Compatibility
ArduinoJson works on the following hardware:
RedBearLab boards: BLE Nano, BLE Mini, WiFi Micro, LOLIN32…
Teensy boards
Intel boards: Edison, Galileo…
Texas Instruments boards: MSP430…
ArduinoJson compiles with zero warning on the following compilers, IDEs, and platforms:
Quickstart
Deserialization
Here is a program that parses a JSON document with ArduinoJson.
char json[] = "{\"sensor\":\"gps\",\"time\":1351824120,\"data\":[48.756080,2.302038]}";
StaticJsonBuffer<200> jsonBuffer;
JsonObject& root = jsonBuffer.parseObject(json);
const char* sensor = root["sensor"];
long time = root["time"];
double latitude = root["data"][0];
double longitude = root["data"][1];
See the tutorial on arduinojson.org
Serialization
Here is a program that generates a JSON document with ArduinoJson:
StaticJsonBuffer<200> jsonBuffer;
JsonObject& root = jsonBuffer.createObject();
root["sensor"] = "gps";
root["time"] = 1351824120;
JsonArray& data = root.createNestedArray("data");
data.add(48.756080);
data.add(2.302038);
root.printTo(Serial);
// This prints:
// {"sensor":"gps","time":1351824120,"data":[48.756080,2.302038]}
See the tutorial on arduinojson.org
Documentation
The documentation is available on arduinojson.org, here are some shortcuts:
The Examples show how to use the library in various situations.
The API Reference contains the description of each class and function.
The FAQ has the answer to virtually every question.
The ArduinoJson Assistant writes programs for you!
Do you like this library? Please star this project on GitHub!
What? You don’t like it but you love it? We don’t take donations anymore, but we sell a book, so you can help and learn at the same time!
ArduinoJson Version 6
Current version is 6.12.0.
If you’re upgrading from version 5, some changes will be required to your code. See the Version 6 Migration Guide for details.
Some methods of JsonVariant have been removed, replacements are:
asString() -> as<char*>() or as<const char*>. Note that as<String> produces a
serialized version, so you’ll get “null” instead of an empty/invalid result String.
asArray() -> as<JsonArray>()
asObject() -> as<JsonObject>()
There are also some useful helper functions available in the Json namespace. See Libraries/ArduinoJson6/include/ArduinoJson.h.
Sming definitions
References
Used by
GoogleCast Library
Hue Emulator Library
IO Control Library
Basic RS485 Sample
Fan Controller Sample
RingTone Player Sample
Yeelight Library
Basic Web Skeleton Sample
HttpClient Instapush Sample
AJAX Http Server Sample
HttpServer Config Network Sample
MeteoControl Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: ArduinoJson
ArduinoJson is a C++ JSON library for Arduino and IoT (Internet Of Things).
Features
-
Supports single quotes as a string delimiter
Compatible with NDJSON and JSON Lines
Efficient
Versatile
Supports custom allocators (to use external RAM chip, for example)
Supports Arduino’s ``String` <https://arduinojson.org/v6/api/config/enable_arduino_string/>`_ and STL’s ``std::string` <https://arduinojson.org/v6/api/config/enable_std_string/?utm_source=github&utm_medium=readme>`_
Supports Arduino’s
Streamand STL’s ``std::istream`/std::ostream<https://arduinojson.org/v6/api/config/enable_std_stream/?utm_source=github&utm_medium=readme>`_Supports custom readers and custom writers
Portable
Usable on any C++ project (not limited to Arduino)
Compatible with C++98
Zero warnings with
-Wall -Wextra -pedanticand/W4Works with virtually any board
Tested on all major development environments
Well designed
Self-contained (no external dependency)
constfriendly``for` friendly <https://arduinojson.org/v6/api/jsonobject/begin_end/?utm_source=github&utm_medium=readme>`_
Handles integer overflows
Well tested
Well documented
Vibrant user community
Most popular of all Arduino libraries on GitHub and PlatformIO
Quickstart
Deserialization
Here is a program that parses a JSON document with ArduinoJson.
char json[] = "{\"sensor\":\"gps\",\"time\":1351824120,\"data\":[48.756080,2.302038]}";
DynamicJsonDocument doc(1024);
deserializeJson(doc, json);
const char* sensor = doc["sensor"];
long time = doc["time"];
double latitude = doc["data"][0];
double longitude = doc["data"][1];
See the tutorial on arduinojson.org
Serialization
Here is a program that generates a JSON document with ArduinoJson:
DynamicJsonDocument doc(1024);
doc["sensor"] = "gps";
doc["time"] = 1351824120;
doc["data"][0] = 48.756080;
doc["data"][1] = 2.302038;
serializeJson(doc, Serial);
// This prints:
// {"sensor":"gps","time":1351824120,"data":[48.756080,2.302038]}
See the tutorial on arduinojson.org
Support the project
Do you like this library? Please star this project on GitHub!
What? You don’t like it but you love it?
We don’t take donations anymore, but we sell a book, so you can help and learn at the same time.
Arduino TensorFlow Lite
This library runs TensorFlow machine learning models on microcontrollers, allowing you to build AI/ML applications powered by deep learning and neural networks.
With the included examples, you can recognize speech, detect people using a camera, and recognise “magic wand” gestures using an accelerometer.
The examples work best with the Arduino Nano 33 BLE Sense board, which has a microphone and accelerometer.
References
Used by
TensorFlow_HelloWorld Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
At Client
Introduction
AT commands were initially instructions used to control a modem. AT is the abbreviation of ATtention. Every command line starts with “AT” or “at”. If interested a good background article is Hayes Command Set.
Nowadays also other devices allow communication via AT commands as for example GSM/GPRS modems, GPS trackers, web cameras and more.
This library simplifies the communication with such devices.
Usage
Add
COMPONENT_DEPENDS += AtClientto your application component.mk file.Add these lines to your application:
#include <AtClient.h> namespace { AtClient* atClient; // ... } // namespace void init() { Serial.begin(SERIAL_BAUD_RATE); Serial.systemDebugOutput(true); atClient = new AtClient(Serial); atClient->send("ATE0\r"); // ... }
References
Used by
Webcam Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
BH1750FVI Light Sensor
https://github.com/Genotronex/BH1750FVI_Master.git
Digital Light Sensor BH1750
- /*
This is a simple code to test BH1750FVI Light senosr
communicate using I2C Protocol
this library enable 2 slave device address
Main address 0x23
secondary address 0x5C
connect this sensor as following :
VCC >>> 3.3V
SDA >>> A4
SCL >>> A5
addr >> A3 “Optional and use address 0x23 “
Gnd >>>Gnd
please after download this library unzip and rename it to BH1750FVI and put it in the libraries folder in the arduino path , then restart Arduino IDE
Written By : Mohannad Rawashdeh
26/9/2013
for more information : http://www.instructables.com/id/BH1750-Digital-Light-Sensor/
contact me on my email
*/
References
Used by
BH1750 Light Sensor Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
ESP32 BLE Gamepad
Introduction
This library allows you to make the ESP32 act as a Bluetooth gamepad and control what it does. The library uses ESP32 NimBLE for faster and lighter communication.
Features
Using this library you can do the following:
Button press (128 buttons)
Button release (128 buttons)
Axes movement (6 axes (16 bit) (x, y, z, rZ, rX, rY) –> (Left Thumb X, Left Thumb Y, Right Thumb X, Right Thumb Y, Left Trigger, Right Trigger))
2 Sliders (16 bit) (Slider 1 and Slider 2)
4 point of view hats (ie. d-pad plus 3 other hat switches)
Simulation controls (rudder, throttle, accelerator, brake, steering)
Configurable HID descriptor
Report optional battery level to host (basically works, but it doesn’t show up in Android’s status bar)
Customize Bluetooth device name/manufacturer
Uses efficient NimBLE bluetooth library
Compatible with Windows
Compatible with Android (Android OS maps default buttons / axes / hats slightly differently than Windows)
Compatible with Linux (limited testing)
Compatible with MacOS X (limited testing)
Using
Add
COMPONENT_DEPENDS += BLEGamepadto your application componenent.mk file.Add these lines to your application:
#include <BleGamepad.h> namespace { BleGamepad bleGamepad; // ... } // namespace void init() { // ... bleGamepad.begin(); }
Notes
By default, reports are sent on every button press/release or axis/slider/hat/simulation movement, however this can be disabled, and then you manually call sendReport on the gamepad instance as shown in the IndividualAxes.ino example.
There is also Bluetooth specific information that you can use (optional):
Instead of BleGamepad bleGamepad; you can do BleGamepad bleGamepad("Bluetooth Device Name", "Bluetooth Device Manufacturer", 100);.
The third parameter is the initial battery level of your device.
Adjusting the battery level later on doesn’t work.
By default the battery level will be set to 100%, the device name will be ESP32 BLE Gamepad and the manufacturer will be Espressif.
References
Used by
Bluetooth Gamepad Sample
SoC support
esp32
esp32c3
esp32s3
Submodule: ESP32-BLE-Gamepad
POSSIBLE BREAKING CHANGES - PLEASE READ
A large code rebase (configuration class) along with some extra features (start, select, menu, home, back, volume up, volume down and volume mute buttons) has been committed thanks to @dexterdy
Since version 5 of this library, the axes and simulation controls have configurable min and max values The decision was made to set defaults to 0 for minimum and 32767 for maximum (previously -32767 to 32767) This was due to the fact that non-Windows operating systems and some online web-based game controller testers didn’t play well with negative numbers. Existing sketches should take note, and see the DrivingControllerTest example for how to set back to -32767 if wanted
This version of the library has been tested against NimBLE-Arduino version 1.4.1; the latest released version –> https://github.com/h2zero/NimBLE-Arduino/releases/tag/1.4.1
Please see updated examples
NimBLE
Since version 3 of this library, the more efficient NimBLE library is used instead of the default BLE implementation Please use the library manager to install it, or get it from here: https://github.com/h2zero/NimBLE-Arduino Since version 3, this library also supports a configurable HID desciptor, which allows users to customise how the device presents itself to the OS (number of buttons, hats, axes, sliders, simulation controls etc). See the examples for guidance.
ESP32-BLE-Gamepad
License
Published under the MIT license. Please see license.txt.
It would be great however if any improvements are fed back into this version.
Features
[x] Button press (128 buttons)
[x] Button release (128 buttons)
[x] Axes movement (6 axes (configurable resolution up to 16 bit) (x, y, z, rZ, rX, rY) –> (Left Thumb X, Left Thumb Y, Right Thumb X, Right Thumb Y, Left Trigger, Right Trigger))
[x] 2 Sliders (configurable resolution up to 16 bit) (Slider 1 and Slider 2)
[x] 4 point of view hats (ie. d-pad plus 3 other hat switches)
[x] Simulation controls (rudder, throttle, accelerator, brake, steering)
[x] Special buttons (start, select, menu, home, back, volume up, volume down, volume mute) all disabled by default
[x] Configurable HID descriptor
[x] Configurable VID and PID values
[x] Configurable BLE characteristics (name, manufacturer, model number, software revision, serial number, firmware revision, hardware revision)
[x] Report optional battery level to host
[x] Uses efficient NimBLE bluetooth library
[x] Compatible with Windows
[x] Compatible with Android (Android OS maps default buttons / axes / hats slightly differently than Windows)
[x] Compatible with Linux (limited testing)
[x] Compatible with MacOS X (limited testing)
[ ] Compatible with iOS (No - not even for accessibility switch - This is not a “Made for iPhone” (MFI) compatible device)
Installation
(Make sure you can use the ESP32 with the Arduino IDE. Instructions can be found here.)
Download the latest release of this library from the release page.
In the Arduino IDE go to “Sketch” -> “Include Library” -> “Add .ZIP Library…” and select the file you just downloaded.
In the Arduino IDE go to “Tools” -> “Manage Libraries…” -> Filter for “NimBLE-Arduino” by h2zero and install.
You can now go to “File” -> “Examples” -> “ESP32 BLE Gamepad” and select an example to get started.
Example
/*
* This example turns the ESP32 into a Bluetooth LE gamepad that presses buttons and moves axis
*
* At the moment we are using the default settings, but they can be canged using a BleGamepadConfig instance as parameter for the begin function.
*
* Possible buttons are:
* BUTTON_1 through to BUTTON_16
* (16 buttons by default. Library can be configured to use up to 128)
*
* Possible DPAD/HAT switch position values are:
* DPAD_CENTERED, DPAD_UP, DPAD_UP_RIGHT, DPAD_RIGHT, DPAD_DOWN_RIGHT, DPAD_DOWN, DPAD_DOWN_LEFT, DPAD_LEFT, DPAD_UP_LEFT
* (or HAT_CENTERED, HAT_UP etc)
*
* bleGamepad.setAxes sets all axes at once. There are a few:
* (x axis, y axis, z axis, rx axis, ry axis, rz axis, slider 1, slider 2)
*
* Library can also be configured to support up to 5 simulation controls
* (rudder, throttle, accelerator, brake, steering), but they are not enabled by default.
*
* Library can also be configured to support different function buttons
* (start, select, menu, home, back, volume increase, volume decrease, volume mute)
* start and select are enabled by default
*/
#include <Arduino.h>
#include <BleGamepad.h>
BleGamepad bleGamepad;
void setup()
{
Serial.begin(115200);
Serial.println("Starting BLE work!");
bleGamepad.begin();
// The default bleGamepad.begin() above enables 16 buttons, all axes, one hat, and no simulation controls or special buttons
}
void loop()
{
if (bleGamepad.isConnected())
{
Serial.println("Press buttons 5, 16 and start. Move all enabled axes to max. Set DPAD (hat 1) to down right.");
bleGamepad.press(BUTTON_5);
bleGamepad.press(BUTTON_16);
bleGamepad.pressStart();
bleGamepad.setAxes(32767, 32767, 32767, 32767, 32767, 32767, 32767, 32767);
bleGamepad.setHat1(HAT_DOWN_RIGHT);
// All axes, sliders, hats etc can also be set independently. See the IndividualAxes.ino example
delay(500);
Serial.println("Release button 5 and start. Move all axes to min. Set DPAD (hat 1) to centred.");
bleGamepad.release(BUTTON_5);
bleGamepad.releaseStart();
bleGamepad.setHat1(HAT_CENTERED);
bleGamepad.setAxes(0, 0, 0, 0, 0, 0, 0, 0);
delay(500);
}
}
By default, reports are sent on every button press/release or axis/slider/hat/simulation movement, however this can be disabled, and then you manually call sendReport on the gamepad instance as shown in the IndividualAxes.ino example.
VID and PID values can be set. See TestAll.ino for example.
There is also Bluetooth specific information that you can use (optional):
Instead of BleGamepad bleGamepad; you can do BleGamepad bleGamepad("Bluetooth Device Name", "Bluetooth Device Manufacturer", 100);.
The third parameter is the initial battery level of your device.
By default the battery level will be set to 100%, the device name will be ESP32 BLE Gamepad and the manufacturer will be Espressif.
Battery level can be set during operation by calling, for example, bleGamepad.setBatteryLevel(80); Update sent on next gamepad update if auto reporting is not enabled
Credits
Credits to T-vK as this library is based on his ESP32-BLE-Mouse library (https://github.com/T-vK/ESP32-BLE-Mouse) that he provided.
Credits to chegewara as the ESP32-BLE-Mouse library is based on this piece of code that he provided.
Credits to wakwak-koba for the NimBLE code that he provided.
Notes
This library allows you to make the ESP32 act as a Bluetooth Gamepad and control what it does.
Relies on NimBLE-Arduino
Use this Windows test app to test/see all of the buttons
You might also be interested in:
or the NimBLE versions at
ESP32 BLE Keyboard
Introduction
This library allows you to make the ESP32 act as a Bluetooth keyboard and control what it does. The library uses ESP32 NimBLE for faster and lighter communication.
Features
Using this library you can do the following:
Send key strokes
Send text
Press/release individual keys
Media keys are supported
Set battery level (basically works, but doesn’t show up in Android’s status bar)
Compatible with Android
Compatible with Windows
Compatible with Linux
Compatible with MacOS X (not stable, some people have issues, doesn’t work with old devices)
Compatible with iOS (not stable, some people have issues, doesn’t work with old devices)
Using
Add
COMPONENT_DEPENDS += BLEKeyboardto your application componenent.mk file.Add these lines to your application:
#include <BleKeyboard.h> namespace { BleKeyboard bleKeyboard; // ... } // namespace void init() { // ... bleKeyboard.begin(); }
API documentation
The BleKeyboard interface is almost identical to the Keyboard Interface, so you can use documentation right here: https://www.arduino.cc/reference/en/language/functions/usb/keyboard/
In addition to that you can send media keys (which is not possible with the USB keyboard library). Supported are the following:
KEY_MEDIA_NEXT_TRACK
KEY_MEDIA_PREVIOUS_TRACK
KEY_MEDIA_STOP
KEY_MEDIA_PLAY_PAUSE
KEY_MEDIA_MUTE
KEY_MEDIA_VOLUME_UP
KEY_MEDIA_VOLUME_DOWN
KEY_MEDIA_WWW_HOME
KEY_MEDIA_LOCAL_MACHINE_BROWSER // Opens “My Computer” on Windows
KEY_MEDIA_CALCULATOR
KEY_MEDIA_WWW_BOOKMARKS
KEY_MEDIA_WWW_SEARCH
KEY_MEDIA_WWW_STOP
KEY_MEDIA_WWW_BACK
KEY_MEDIA_CONSUMER_CONTROL_CONFIGURATION // Media Selection
KEY_MEDIA_EMAIL_READER
There is also Bluetooth specific information that you can set (optional):
Instead of BleKeyboard bleKeyboard; you can do BleKeyboard bleKeyboard("Bluetooth Device Name", "Bluetooth Device Manufacturer", 100);. (Max length is 15 characters, anything beyond that will be truncated.)
The third parameter is the initial battery level of your device. To adjust the battery level later on you can simply call e.g. bleKeyboard.setBatteryLevel(50) (set battery level to 50%).
By default the battery level will be set to 100%, the device name will be ESP32 Bluetooth Keyboard and the manufacturer will be Espressif.
There is also a setDelay method to set a delay between each key event. E.g. bleKeyboard.setDelay(10) (10 milliseconds). The default is 8.
This feature is meant to compensate for some applications and devices that can’t handle fast input and will skip letters if too many keys are sent in a small time frame.
References
Used by
Bluetooth Keyboard Sample
SoC support
esp32
esp32c3
esp32s3
Submodule: ESP32-BLE-Keyboard
ESP32 BLE Keyboard library
This library allows you to make the ESP32 act as a Bluetooth Keyboard and control what it does.
You might also be interested in:
Features
[x] Send key strokes
[x] Send text
[x] Press/release individual keys
[x] Media keys are supported
[ ] Read Numlock/Capslock/Scrolllock state
[x] Set battery level (basically works, but doesn’t show up in Android’s status bar)
[x] Compatible with Android
[x] Compatible with Windows
[x] Compatible with Linux
[x] Compatible with MacOS X (not stable, some people have issues, doesn’t work with old devices)
[x] Compatible with iOS (not stable, some people have issues, doesn’t work with old devices)
Installation
(Make sure you can use the ESP32 with the Arduino IDE. Instructions can be found here.)
Download the latest release of this library from the release page.
In the Arduino IDE go to “Sketch” -> “Include Library” -> “Add .ZIP Library…” and select the file you just downloaded.
You can now go to “File” -> “Examples” -> “ESP32 BLE Keyboard” and select any of the examples to get started.
Example
/**
* This example turns the ESP32 into a Bluetooth LE keyboard that writes the words, presses Enter, presses a media key and then Ctrl+Alt+Delete
*/
#include <BleKeyboard.h>
BleKeyboard bleKeyboard;
void setup() {
Serial.begin(115200);
Serial.println("Starting BLE work!");
bleKeyboard.begin();
}
void loop() {
if(bleKeyboard.isConnected()) {
Serial.println("Sending 'Hello world'...");
bleKeyboard.print("Hello world");
delay(1000);
Serial.println("Sending Enter key...");
bleKeyboard.write(KEY_RETURN);
delay(1000);
Serial.println("Sending Play/Pause media key...");
bleKeyboard.write(KEY_MEDIA_PLAY_PAUSE);
delay(1000);
//
// Below is an example of pressing multiple keyboard modifiers
// which by default is commented out.
//
/* Serial.println("Sending Ctrl+Alt+Delete...");
bleKeyboard.press(KEY_LEFT_CTRL);
bleKeyboard.press(KEY_LEFT_ALT);
bleKeyboard.press(KEY_DELETE);
delay(100);
bleKeyboard.releaseAll();
*/
}
Serial.println("Waiting 5 seconds...");
delay(5000);
}
API docs
The BleKeyboard interface is almost identical to the Keyboard Interface, so you can use documentation right here: https://www.arduino.cc/reference/en/language/functions/usb/keyboard/
Just remember that you have to use bleKeyboard instead of just Keyboard and you need these two lines at the top of your script:
#include <BleKeyboard.h>
BleKeyboard bleKeyboard;
In addition to that you can send media keys (which is not possible with the USB keyboard library). Supported are the following:
KEY_MEDIA_NEXT_TRACK
KEY_MEDIA_PREVIOUS_TRACK
KEY_MEDIA_STOP
KEY_MEDIA_PLAY_PAUSE
KEY_MEDIA_MUTE
KEY_MEDIA_VOLUME_UP
KEY_MEDIA_VOLUME_DOWN
KEY_MEDIA_WWW_HOME
KEY_MEDIA_LOCAL_MACHINE_BROWSER // Opens “My Computer” on Windows
KEY_MEDIA_CALCULATOR
KEY_MEDIA_WWW_BOOKMARKS
KEY_MEDIA_WWW_SEARCH
KEY_MEDIA_WWW_STOP
KEY_MEDIA_WWW_BACK
KEY_MEDIA_CONSUMER_CONTROL_CONFIGURATION // Media Selection
KEY_MEDIA_EMAIL_READER
There is also Bluetooth specific information that you can set (optional):
Instead of BleKeyboard bleKeyboard; you can do BleKeyboard bleKeyboard("Bluetooth Device Name", "Bluetooth Device Manufacturer", 100);. (Max lenght is 15 characters, anything beyond that will be truncated.)
The third parameter is the initial battery level of your device. To adjust the battery level later on you can simply call e.g. bleKeyboard.setBatteryLevel(50) (set battery level to 50%).
By default the battery level will be set to 100%, the device name will be ESP32 Bluetooth Keyboard and the manufacturer will be Espressif.
There is also a setDelay method to set a delay between each key event. E.g. bleKeyboard.setDelay(10) (10 milliseconds). The default is 8.
This feature is meant to compensate for some applications and devices that can’t handle fast input and will skip letters if too many keys are sent in a small time frame.
NimBLE-Mode
The NimBLE mode enables a significant saving of RAM and FLASH memory.
Comparison (SendKeyStrokes.ino at compile-time)
Standard
RAM: [= ] 9.3% (used 30548 bytes from 327680 bytes)
Flash: [======== ] 75.8% (used 994120 bytes from 1310720 bytes)
NimBLE mode
RAM: [= ] 8.3% (used 27180 bytes from 327680 bytes)
Flash: [==== ] 44.2% (used 579158 bytes from 1310720 bytes)
Comparison (SendKeyStrokes.ino at run-time)
Standard |
NimBLE mode |
difference |
|
|---|---|---|---|
|
296.804 |
321.252 |
+ 24.448 |
|
143.572 |
260.764 |
+ 117.192 |
|
994.224 |
579.264 |
- 414.960 |
How to activate NimBLE mode?
ArduinoIDE:
Uncomment the first line in BleKeyboard.h
#define USE_NIMBLE
PlatformIO:
Change your platformio.ini to the following settings
lib_deps =
NimBLE-Arduino
build_flags =
-D USE_NIMBLE
Credits
Credits to chegewara and the authors of the USB keyboard library as this project is heavily based on their work!
Also, credits to duke2421 who helped a lot with testing, debugging and fixing the device descriptor!
And credits to sivar2311 for adding NimBLE support, greatly reducing the memory footprint, fixing advertising issues and for adding the setDelay method.
BME280 Barometric Pressure Sensor
BME280 is Barometric Pressure Sensor.
- Datasheet for BME280:
http://www.bosch-sensortec.com/content/language1/downloads/BST-BMP180-DS000-07.pdf
Copyright (C) 2012 Love Electronics Ltd (loveelectronics.com).
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
BMP180 Pressure/Temperature Sensor
Arduino library for BMP180 Bosch pressure/temperature sensor
Copyright (C) 2012 Love Electronics Ltd (loveelectronics.com)
References
Used by
MeteoControl MQTT Sample
BMP180 Pressure Sensor Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Bounce library for Arduino
Version 1.6
by Thomas Ouellet Fredericks with contributions from: Eric Lowry Jim Schimpf Tom Harkaway
contact: mrtoftrash@gmail.com
See the online documentation here: http://www.arduino.cc/playground/Code/Bounce
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
CS5460 energy meter IC
Product information: https://www.cirrus.com/products/cs5460a/
The CS5460(A) is a highly integrated power measurement solution which combines:
Two Analog-to-digital Converters (ADCs)
High-speed power calculation functions
A serial interface
on a single chip. It is designed to accurately measure and calculate
Real (True) Energy
Instantaneous Power
I RMS
V RMS
for single phase 2- or 3-wire power metering applications.
References
Used by
CS5460 generic sample Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: CS5460
CS5460
Arduino library for CS5460
Using SPI interface to connect, tested on Arduino UNO.
Current functions include set config register arguments manually, read measurements in raw binary format and float number and calibrate gain and offset.
See detail information of CS5460: datasheet
Todo: Make full use of EOUT, EDIR and INT ports.
Capacitive Sensor Library
CapacitiveSensor lets you create sensors that can detect touch or proximity.
http://www.pjrc.com/teensy/td_libs_CapacitiveSensor.html
http://playground.arduino.cc/Main/CapacitiveSensor
http://www.youtube.com/watch?v=BHQPqQ_5ulc
CapacitiveSensor was originally written by Paul Badger and is now maintained by Paul Stoffregen.
References
Used by
Basic Capsense Sample
SoC support
esp8266
Command Processing
Introduction
Command handler provides a common command line interface (CLI). Command line must be text. Commands should be separated with a single character. CLI can be used with:
Serial
Network (Websockets, Telnet)
and all communication protocols that are exchanging text data.
Commands can be added to and removed from the command handler. Each command will trigger a defined Delegate.
A welcome message may be shown when a user connects and end-of-line (EOL) character may be defined. An automatic “help” display is available.
For more examples take a look at the CommandLine, TelnetServer and HttpServer Websockets samples.
Using
Add these lines to your application componenent.mk file:
COMPONENT_DEPENDS += CommandProcessing
Add these lines to your application:
#include <CommandProcessing/Utils.h>
Basic example:
#include <CommandProcessing/Utils.h> CommandProcessing::Handler commandHandler; void processShutdownCommand(String commandLine, ReadWriteStream& commandOutput) { // ... } void init() { commandHandler.registerSystemCommands(); commandHandler.registerCommand({CMDP_STRINGS("shutdown", "Shutdown Server Command", "Application"), processShutdownCommand}); }
- CMDPROC_FLASHSTRINGS
default: undefined (RAM strings) 1: Store strings in flash memory
Command
name,helpandgroupstrings default to RAM. This maintains backward compatibility with existing applications which may define commands like this:commandHandler.registerCommand({"shutdown", "Shutdown Server Command", "Application", processShutdownCommand});
To save RAM, update your code to use the
CMDP_STRINGSmacro and build withCMDPROC_FLASHSTRINGS=1.
API Documentation
-
CMDP_STRINGS(name, help, group)
References
Used by
Arducam Sample
CommandLine Sample
TelnetServer Sample
HttpServer Websockets Sample
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
DFPlayer - A Mini MP3 Player For Arduino
DFPlayer - A Mini MP3 Player For Arduino https://www.dfrobot.com/index.php?route=product/product&product_id=1121
This example shows the all the function of library for DFPlayer.
Created 2016-12-07 By Angelo qiao
GNU Lesser General Public License. See http://www.gnu.org/licenses/ for details. All above must be included in any redistribution
Notice and Trouble shooting
1.Connection and Diagram can be found here https://www.dfrobot.com/wiki/index.php/DFPlayer_Mini_SKU:DFR0299#Connection_Diagram 2.This code is tested on Arduino Uno, Leonardo, Mega boards.
References
Used by
DFPlayer Mini Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
DHT ESP Temperature/Humidity Sensors
An Arduino library for reading the DHT family of temperature and humidity sensors.
Forked from arduino-DHT
Original written by Mark Ruys, mark@paracas.nl.
Why did I clone this library instead of forking the original repo and push the changes?
When I searched through Github for DHT libraries, I found a lot of them, some of them offers additional functions, some of them only basic temperature and humidity values. I wanted to combine all interesting functions into one library. In addition, none of the DHT libraries I found were written to work without errors on the ESP32. For ESP32 (a multi core/ multi processing SOC) task switching must be disabled while reading data from the sensor.
Another problem I found is that many of the available libraries use the same naming (dht.h, dht.cpp), which easily leads to conflicts if different libraries are used for different platforms.
*According to users, the library works as well with DHT33 and DHT44 sensors. But as I do not own these sensors, I cannot test and confirm it. However, if you want to use this sensors, you can do so by using ``setup(pin, DHTesp::DHT22)`` and it should work. Please give me feedback in the issues if you successfull use these sensors. Thank you_
The library is tested as well on ESP8266 and should work on AVR boards as well.
Changes to the original library:
2017-12-12: Renamed DHT class to DHTesp and filenames from dht.* to DHTesp.* to avoid conflicts with other libraries - beegee-tokyo, beegee@giesecke.tk.
2017-12-12: Updated to work with ESP32 - beegee-tokyo, beegee@giesecke.tk.
2017-12-12: Added function computeHeatIndex. Reference: Adafruit DHT library.
2017-12-14: Added function computeDewPoint. Reference: idDHTLib.
2017-12-14: Added function getComfortRatio. Reference: libDHT. (References about Human Comfort invalid)
2017-12-15: Added function computePerception. Reference: WikiPedia Dew point==> Relationship to human comfort - beegee-tokyo, beegee@giesecke.tk.
2018-01-02: Added example for multiple sensors usage.
2018-01-03: Added function getTempAndHumidity which returns temperature and humidity in one call.
2018-01-03: Added retry in case the reading from the sensor fails with a timeout.
2018-01-08: Added ESP8266 (and probably AVR) compatibility.
2018-03-11: Updated DHT example
2018-06-19: Updated DHT example to distinguish between ESP8266 examples and ESP32 examples
2018-07-06: Fixed bug in ESP32 example
2018-07-17: Use correct field separator in keywords.txt
2019-03-07: Added computeAbsoluteHumidity which returns the absolute humidity in g/m³. Reference: How to convert relative humidity to absolute humidity kudos to Wurstnase
2019-03-22: Fixed auto detection problem
2019-07-31: Make getPin() public, Updated ESP8266 example
2019-10-01: Using noInterrupts() & interrupts() instead of cli and sei
2019-10-05: Reduce CPU usage and add decimal part for DHT11 (thanks to Swiftyhu)
2019-10-06: Back to working version by removing the last commit
Features
Support for DHT11 and DHT22, AM2302, RHT03
Auto detect sensor model
Determine heat index
Determine dewpoint
Determine thermal comfort:
Empiric comfort function based on comfort profiles(parametric lines)
Multiple comfort profiles possible. Default based on http://epb.apogee.net/res/refcomf.asp (References invalid)
Determine if it’s too cold, hot, humid, dry, based on current comfort profile
More info at Determining Thermal Comfort Using a Humidity and Temperature Sensor
Determine human perception based on humidity, temperature and dew point according to Horstmeyer, Steve (2006-08-15). Relative Humidity….Relative to What? The Dew Point Temperature…a better approach
Functions
**``void setup(uint8_t pin, DHT_MODEL_t model=AUTO_DETECT);``*_
Call to initialize the interface, define the GPIO pin to which the sensor is connected and define the sensor type. Valid sensor types are:
AUTO_DETECT Try to detect which sensor is connected (default if 2nd parameter is not used)
DHT11
DHT22
AM2302 Packaged DHT22
RHT03 Equivalent to DHT22
**``void resetTimer();``*_
Reset last time the sensor was read
**``float getTemperature();``*_
Get the temperature in degree Centigrade from the sensor
Either one of *getTemperature()* or *getHumidity()* or *getTempAndHumidity()* initiates reading a value from the sensor if the last reading was older than the minimal refresh time of the sensor.
See example _`DHT_ESP32.ino`_ or _`DHT_Test.ino`_
**``float getHumidity();``*_
Get the humidity from the sensor
Either one of *getTemperature()* or *getHumidity()* or *getTempAndHumidity()* initiates reading a value from the sensor if the last reading was older than the minimal refresh time of the sensor.
See example _`DHT_ESP32.ino`_ or _`DHT_Test.ino`_
**``TempAndHumidity getTempAndHumidity();``*_
Get the temperature and humidity from the sensor
Either one of *getTemperature()* or *getHumidity()* or *getTempAndHumidity()* initiates reading a value from the sensor if the last reading was older than the minimal refresh time of the sensor.
Return value is a struct of type *TempAndHumidity* with temperature and humidity as float values. See example _`DHT_Multi.ino`_
**``DHT_ERROR_t getStatus();``*_
Get last error if reading from the sensor failed. Possible values are:
ERROR_NONE no error occured
ERROR_TIMEOUT timeout reading from the sensor
ERROR_CHECKSUM checksum of received values doesn’t match
*`const char* getStatusString();`_
Get last error as a char *
**``DHT_MODEL_t getModel()``*_
Get detected (or defined) sensor type
**``int getMinimumSamplingPeriod();``*_
Get minimmum possible sampling period. For DHT11 this is 1000ms, for other sensors it is 2000ms
**``int8_t getNumberOfDecimalsTemperature();``*_
Get number of decimals in the temperature value. For DHT11 this is 0, for other sensors it is 1
**``int8_t getLowerBoundTemperature();``*_
Get lower temperature range of the sensor. For DHT11 this is 0 degree Centigrade, for other sensors this is -40 degree Centrigrade
**``int8_t getUpperBoundTemperature();``*_
Get upper temperature range of the sensor. For DHT11 this is 50 degree Centigrade, for other sensors this is 125 degree Centrigrade
**``int8_t getNumberOfDecimalsHumidity();``*_
Get number of decimals in the humidity value. This is always 0.
**``int8_t getLowerBoundHumidity();``*_
Get lower humidity range of the sensor. For DHT11 this is 20 percent, for other sensors this is 0 percent
**``int8_t getUpperBoundHumidity();``*_
Get upper temperature range of the sensor. For DHT11 this is 90 percent, for other sensors this is 100 percent
**``static float toFahrenheit(float fromCelcius);``*_
Convert Centrigrade value to Fahrenheit value
**``static float toCelsius(float fromFahrenheit) { return (fromFahrenheit - 32.0) / 1.8; };``*_
Convert Fahrenheit value to Centigrade value
**``float computeHeatIndex(float temperature, float percentHumidity, bool isFahrenheit=false);``*_
Compute the heat index. Default temperature is in Centrigrade.
**``float computeDewPoint(float temperature, float percentHumidity, bool isFahrenheit=false);``*_
Compute the dew point. Default temperature is in Centrigrade.
**``float computeAbsoluteHumidity(float temperature, float percentHumidity, bool isFahrenheit=false);``*_
Compute the absolute humidity in g/m³. Default temperature is in Centrigrade.
**``float getComfortRatio(ComfortState& destComfStatus, float temperature, float percentHumidity, bool isFahrenheit=false);``*_
Compute the comfort ratio. Default temperature is in Centrigrade. Return values:
0 -> OK
1 -> Too Hot
2 -> Too cold
4 -> Too dry
8 -> Too humid
9 -> Hot and humid
5 -> Hot and dry
10 -> Cold and humid
6 -> Cold and dry
**``byte computePerception(float temperature, float percentHumidity, bool isFahrenheit=false);``*_
Compute the human perception. Default temperature is in Centrigrade. Return values:
0 -> Dry
1 -> Very comfortable
2 -> Comfortable
3 -> Ok
4 -> Uncomfortable
5 -> Quite uncomfortable
6 -> Very uncomfortable
7 -> Severe uncomfortable
**``uint8_t getPin(void);``*_
Returns the assigned GPIO for this instance. Usefull when connecting multiple sensors
Usage
Installation
In Arduino IDE open Sketch->Include Library->Manage Libraries then search for *DHT ESP:raw-html-m2r:`<br>`
In PlatformIO open PlatformIO Home, switch to libraries and search for ***DHT ESP32*. Or install the library in the terminal with **platformio lib install 2029**_
For manual installation download the archive, unzip it and place the DHTesp folder into the library directory.
In Arduino IDE this is usually **``<arduinosketchfolder>/libraries/``*:raw-html-m2r:`<br>`
In PlatformIO this is usually **<user/.platformio/lib>**_
References
Used by
DHT22 Humidity Sensor Sample
MeteoControl Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
DIscovery And Launch (DIAL)
Introduction
DIAL—for DIscovery And Launch—is a simple protocol that second-screen devices can use to discover and launch apps on first-screen devices. For example, your can stream a video from your embedded device on your connected TV.
Using
Add
COMPONENT_DEPENDS += DIALto your application componenent.mk file.Add these lines to your application:
#include <Dial/Client.h> static UPnP::ControlPoint controlPoint; static Dial::Client* myClient; // Call when IP address has been obtained void onIp(IpAddress ip, IpAddress mask, IpAddress gateway) { // ... /* Use UPnP to auto-discover all DIAL-enabled servers */ Dial::discover(controlPoint, [](Dial::Client& client) { // Are we looking for a specific device? Can match on friendlyName, UDN, etc. if(client.friendlyName() == F("FriendlyNameToFind")) { // Take a reference to the device myClient = &client; // Get an app and do something... auto& app = myClient->getApp("YouTube"); // Keep this device return true; } // Don't want this device, destroy it return false; }); // ... }
See the Control your DIAL-enabled smart monitor/TV using Sming sample application.
API Documentation
-
namespace Dial
Functions
-
bool discover(UPnP::ControlPoint &controlPoint, Client::Discovered callback)
-
class App
- #include <App.h>
-
class Client : public UPnP::DeviceControl
- #include <Client.h>
Public Functions
-
App &getApp(const String &applicationId)
Get application object by name.
- Parameters:
applicationId – the unique application. A list of registered ids can be found here: http://www.dial-multiscreen.org/dial-registry/namespace-database#TOC-Registered-Names
- Return values:
App& – Application object reference
-
virtual void onConnected(HttpConnection &connection) override
Inherited classes may override this to pull out any additional information from received response headers, etc. Invoked after description has been processed.
-
App &getApp(const String &applicationId)
-
bool discover(UPnP::ControlPoint &controlPoint, Client::Discovered callback)
References
Used by
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
DS18S20 Temperature Sensor
This library provides access to DS18S20 temperature sensors connected via 1-Wire bus to a single GPIO The DS18S20 can run in several modes, with varying degrees of resolution. The highest resolution is 12-bit which provides 0.0625C resolution. 12-bit measurement takes 750ms. With 4 sensors connected, measurement will take 3s.
Created on: 01-09-2015 Author: flexiti and Anakod
References
Used by
DS1820 Temperature Sensor Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Arduino DS3232RTC Library v1.0
https://github.com/JChristensen/DS3232RTC
ReadMe file
Jack Christensen Mar 2013
Introduction
DS3232RTC is an Arduino library that supports the Maxim Integrated DS3232 and DS3231 Real-Time Clocks. This library is intended to be used with the Arduino Time library.
The DS3232RTC library is a drop-in replacement for the DS1307RTC.h library by Michael Margolis that is supplied with the Arduino Time library above. To change from using a DS1307 RTC to an DS323x RTC, it is only necessary to use #include <DS3232RTC.h> instead of #include <DS1307RTC.h>.
This library is **not** a drop-in replacement for PJRC’s newer version of the DS1307RTC library.
DS3232RTC also implements functions to support the additional features of the DS3232 and DS3231. The DS3231 has the same features as the DS3232 except: (1) Battery-backed SRAM, (2) Battery-backed 32kHz output (BB32kHz bit in Control/Status register 0x0F), and (3) Adjustable temperature sensor sample rate (CRATE1:0 bits in the Control/Status register).
“Arduino DS3232RTC Library” by Jack Christensen is licensed under CC BY-SA 4.0.
Installation
To use the DS3232RTC library:
Go to https://github.com/JChristensen/DS3232RTC, click the Download ZIP button and save the ZIP file to a convenient location on your PC.
Uncompress the downloaded file. This will result in a folder containing all the files for the library, that has a name that includes the branch name, usually DS3232RTC-master.
Rename the folder to just DS3232RTC.
Copy the renamed folder to the Arduino sketchbooklibraries folder.
Examples
The following example sketches are included with the DS3232RTC library:
SetSerial: Set the RTC’s date and time from the Arduino serial monitor. Displays date, time and temperature.
TimeRTC: Same as the example of the same name provided with the Time library, demonstrating the interchangeability of the DS3232RTC library with the DS1307RTC library.
tiny3232_KnockBang: Demonstrates interfacing an ATtiny45/85 to a DS3231 or DS3232 RTC.
Usage notes
When using the DS3232RTC library, the user is responsible for ensuring that reads and writes do not exceed the device’s address space (0x00-0x12 for DS3231, 0x00-0xFF for DS3232); no bounds checking is done by the library.
Similar to the DS1307RTC library, the DS3232RTC library instantiates an RTC object; the user does not need to do this.
To use the DS3232RTC library, the Time and Wire libraries must also be included. For brevity, these includes are not repeated in the examples below:
#include <DS3232RTC.h> //http://github.com/JChristensen/DS3232RTC
#include <Time.h> //http://www.arduino.cc/playground/Code/Time
#include <Wire.h> //http://arduino.cc/en/Reference/Wire (included with Arduino IDE)
Enumerations
SQWAVE_FREQS_t
Description
Symbolic names used with the squareWave() method (described below).
Values
SQWAVE_NONE
SQWAVE_1_HZ
SQWAVE_1024_HZ
SQWAVE_4096_HZ
SQWAVE_8192_HZ
ALARM_TYPES_t
Description
Symbolic names used with the setAlarm() method (described below).
Values for Alarm 1
ALM1_EVERY_SECOND – causes an alarm once per second.
ALM1_MATCH_SECONDS – causes an alarm when the seconds match (i.e. once per minute).
ALM1_MATCH_MINUTES – causes an alarm when the minutes and seconds match.
ALM1_MATCH_HOURS – causes an alarm when the hours and minutes and seconds match.
ALM1_MATCH_DATE – causes an alarm when the date of the month and hours and minutes and seconds match.
ALM1_MATCH_DAY – causes an alarm when the day of the week and hours and minutes and seconds match.
Values for Alarm 2
ALM2_EVERY_MINUTE – causes an alarm once per minute.
ALM2_MATCH_MINUTES – causes an alarm when the minutes match (i.e. once per hour).
ALM2_MATCH_HOURS – causes an alarm when the hours and minutes match.
ALM2_MATCH_DATE – causes an alarm when the date of the month and hours and minutes match.
ALM2_MATCH_DAY – causes an alarm when the day of the week and hours and minutes match.
Methods for setting and reading the time
get(void)
Description
Reads the current date and time from the RTC and returns it as a time_t value. Returns zero if an I2C error occurs (RTC not present, etc.).
Syntax
RTC.get();
Parameters
None.
Returns
Current date and time (time_t)
Example
time_t myTime;
myTime = RTC.get();
set(time_t t)
Description
Sets the RTC date and time to the given time_t value. Clears the oscillator stop flag (OSF) bit in the control/status register. See the oscStopped() function and also the DS323x datasheet for more information on the OSF bit.
Syntax
RTC.set(t);
Parameters
t: The date and time to set the RTC to (time_t)
Returns
I2C status (byte). Returns zero if successful.
Example
//this example first sets the system time (maintained by the Time library) to
//a hard-coded date and time, and then sets the RTC from the system time.
//the setTime() function is part of the Time library.
setTime(23, 31, 30, 13, 2, 2009); //set the system time to 23h31m30s on 13Feb2009
RTC.set(now()); //set the RTC from the system time
read(tmElements_t &tm)
Description
Reads the current date and time from the RTC and returns it as a tmElements_t structure. See the Arduino Time library for details on the tmElements_t structure.
Syntax
RTC.read(tm);
Parameters
tm: Address of a tmElements_t structure to which the date and time are returned.
Returns
I2C status (byte). Returns zero if successful. The date and time read from the RTC are returned to the tm parameter.
Example
tmElements_t tm;
RTC.read(tm);
Serial.print(tm.Hour, DEC);
Serial.print(':');
Serial.print(tm.Minute,DEC);
Serial.print(':');
Serial.println(tm.Second,DEC);
write(tmElements_t &tm)
Description
Sets the RTC to the date and time given by a tmElements_t structure. Clears the oscillator stop flag (OSF) bit in the control/status register. See the oscStopped() function and also the DS323x datasheet for more information on the OSF bit.
Syntax
RTC.write(tm);
Parameters
tm: Address of a tmElements_t structure used to set the date and time.
Returns
I2C status (byte). Returns zero if successful.
Example
tmElements_t tm;
tm.Hour = 23; //set the tm structure to 23h31m30s on 13Feb2009
tm.Minute = 31;
tm.Minute = 30;
tm.Day = 13;
tm.Month = 2;
tm.Year = 2009 - 1970; //tmElements_t.Year is the offset from 1970
RTC.write(tm); //set the RTC from the tm structure
Methods for reading and writing RTC registers or static RAM (SRAM) for the DS3232
The DS3232RTC.h file defines symbolic names for the timekeeping, alarm, status and control registers. These can be used for the addr argument in the functions below.
writeRTC(byte addr, byte *values, byte nBytes)
Description
Write one or more bytes to RTC memory.
Syntax
RTC.writeRTC(addr, values, nbytes);
Parameters
addr: First SRAM address to write (byte). The valid address range is 0x00-0x12 for DS3231, 0x00-0xFF for DS3232. The general-purpose SRAM for the DS3232 begins at address 0x14. Address is not checked for validity by the library.
values: An array of values to write (*byte)
nBytes: Number of bytes to write (byte). Must be between 1 and 31 (Wire library limitation) but is not checked by the library.
Returns
I2C status (byte). Returns zero if successful.
Example
//write 1, 2, ..., 8 to the first eight DS3232 SRAM locations
byte buf[8] = {1, 2, 3, 4, 5, 6, 7, 8};
RTC.sramWrite(0x14, buf, 8);
writeRTC(byte addr, byte value)
Description
Write a single byte to RTC memory.
Syntax
RTC.writeRTC(addr, value);
Parameters
addr: SRAM address to write (byte). The valid address range is 0x00-0x12 for DS3231, 0x00-0xFF for DS3232. The general-purpose SRAM for the DS3232 begins at address 0x14. Address is not checked for validity by the library. value: Value to write (byte)
Returns
I2C status (byte). Returns zero if successful.
Example
RTC.writeRTC(3, 14); //write the value 14 to SRAM address 3
readRTC(byte addr, byte *values, byte nBytes)
Description
Read one or more bytes from RTC RAM.
Syntax
RTC.readRTC(addr, values, nbytes);
Parameters
addr: First SRAM address to read (byte). The valid address range is 0x00-0x12 for DS3231, 0x00-0xFF for DS3232. The general-purpose SRAM for the DS3232 begins at address 0x14. Address is not checked for validity by the library.
values: An array to receive the values read (*byte)
nBytes: Number of bytes to read (byte). Must be between 1 and 32 (Wire library limitation) but is not checked by the library.
Returns
I2C status (byte). Returns zero if successful.
Example
//read the last eight locations of SRAM into buf
byte buf[8];
RTC.sramRead(248, buf, 8);
readRTC(byte addr)
Description
Reads a single byte from RTC RAM.
Syntax
RTC.readRTC(addr);
Parameters
addr: SRAM address to read (byte). The valid address range is 0x00-0x12 for DS3231, 0x00-0xFF for DS3232. The general-purpose SRAM for the DS3232 begins at address 0x14. Address is not checked for validity by the library.
Returns
Value read from the RTC (byte)
Example
byte val;
val = RTC.readRTC(3); //read the value from SRAM location 3
Alarm methods
The DS3232 and DS3231 have two alarms. Alarm1 can be set to seconds precision; Alarm2 can only be set to minutes precision.
setAlarm(ALARM_TYPES_t alarmType, byte seconds, byte minutes, byte hours, byte daydate)
Description
Set an alarm time. Sets the alarm registers only. To cause the INT pin to be asserted on alarm match, use alarmInterrupt(). This method can set either Alarm 1 or Alarm 2, depending on the value of alarmType (use the ALARM_TYPES_t enumeration above). When setting Alarm 2, the seconds value must be supplied but is ignored, recommend using zero. (Alarm 2 has no seconds register.)
Syntax
RTC.setAlarm(alarmType, seconds, minutes, hours, dayOrDate);
Parameters
alarmType: A value from the ALARM_TYPES_t enumeration, above. (ALARM_TYPES_t)
seconds: The seconds value to set the alarm to. (byte)
minutes: The minutes value to set the alarm to. (byte)
hours: The hours value to set the alarm to. (byte)
dayOrDate: The day of the week or the date of the month. For day of the week, use a value from the Time library timeDayOfWeek_t enumeration, i.e. dowSunday, dowMonday, dowTuesday, dowWednesday, dowThursday, dowFriday, dowSaturday. (byte)
Returns
None.
Example
//Set Alarm1 for 12:34:56 on Sunday
RTC.setAlarm(ALM1_MATCH_DAY, 56, 34, 12, dowSunday);
setAlarm(ALARM_TYPES_t alarmType, byte minutes, byte hours, byte daydate)
Description
Set an alarm time. Sets the alarm registers only. To cause the INT pin to be asserted on alarm match, use alarmInterrupt(). This method can set either Alarm 1 or Alarm 2, depending on the value of alarmType (use the ALARM_TYPES_t enumeration above). However, when using this method to set Alarm 1, the seconds value is set to zero. (Alarm 2 has no seconds register.)
Syntax
RTC.setAlarm(alarmType, minutes, hours, dayOrDate);
Parameters
alarmType: A value from the ALARM_TYPES_t enumeration, above. (ALARM_TYPES_t)
minutes: The minutes value to set the alarm to. (byte)
hours: The hours value to set the alarm to. (byte)
dayOrDate: The day of the week or the date of the month. For day of the week, use a value from the Time library timeDayOfWeek_t enumeration, i.e. dowSunday, dowMonday, dowTuesday, dowWednesday, dowThursday, dowFriday, dowSaturday. (byte)
Returns
None.
Example
//Set Alarm2 for 12:34 on the 4th day of the month
RTC.setAlarm(ALM1_MATCH_DATE, 34, 12, 4);
alarmInterrupt(byte alarmNumber, boolean alarmEnabled)
Description
Enable or disable an alarm “interrupt”. Note that this “interrupt” causes the RTC’s INT pin to be asserted. To use this signal as an actual interrupt to a microcontroller, it will need to be connected properly and programmed in the application firmware. on the RTC.
Syntax
RTC.alarmInterrupt(alarmNumber, enable);
Parameters
alarmNumber: The number of the alarm to enable or disable, ALARM_1 or ALARM_2 (byte)
alarmEnabled: true or false (boolean)
Returns
None.
Example
RTC.alarmInterrupt(ALARM_1, true); //assert the INT pin when Alarm1 occurs.
RTC.alarmInterrupt(ALARM_2, false); //disable Alarm2
alarm(byte alarmNumber)
Description
Tests whether an alarm has been triggered. If the alarm was triggered, returns true and resets the alarm flag in the RTC, else returns false.
Syntax
RTC.alarm(alarmNumber);
Parameters
alarmNumber: The number of the alarm to test, ALARM_1 or ALARM_2 (byte)
Returns
Description (type)
Example
if ( RTC.alarm(ALARM_1) ) { //has Alarm1 triggered?
//yes, act on the alarm
}
else {
//no alarm
}
Other methods
temperature(void)
Description
Returns the RTC temperature.
Syntax
RTC.temperature();
Parameters
None.
Returns
RTC temperature as degrees Celsius times four. (int)
Example
int t = RTC.temperature();
float celsius = t / 4.0;
float fahrenheit = celsius * 9.0 / 5.0 + 32.0;
squareWave(SQWAVE_FREQS_t freq)
Description
Enables or disables the square wave output.
Syntax
RTC.squareWave(freq);
Parameters
freq: a value from the SQWAVE_FREQS_t enumeration above. (SQWAVE_FREQS_t)
Returns
None.
Example
RTC.squareWave(SQWAVE_1_HZ); //1 Hz square wave
RTC.squareWave(SQWAVE_NONE); //no square wave
oscStopped(bool clearOSF)
Description
Returns the value of the oscillator stop flag (OSF) bit in the control/status register which indicates that the oscillator is or was stopped, and that the timekeeping data may be invalid. Optionally clears the OSF bit depending on the argument passed. If the clearOSF argument is omitted, the OSF bit is cleared by default. Calls to set() and write() also clear the OSF bit.
Syntax
RTC.oscStopped(clearOSF);
Parameters
clearOSF: an optional true or false value to indicate whether the OSF bit should be cleared (reset). If not supplied, a default value of true is used, resetting the OSF bit. (bool)
Returns
True or false (bool)
Example
if ( RTC.oscStopped(false) ) { //check the oscillator
//may be trouble
}
else {
//all is well
}
References
Used by
DS3232RTC NTP Setter Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Disk Storage
Provides support for using disk storage devices.
Block Devices
Internal devices such as SPI flash may be read and written in bytes, however disk devices must be accessed in blocks. Historically these blocks are referred to as ‘sectors’, which are 512 bytes for SD cards.
Sector-addressable (block) devices should be based on Disk::BlockDevice.
An example is the Storage::SD::Card provided by the SD Storage library.
Note that devices larger than 4GB will require the ENABLE_STORAGE_SIZE64 project setting.
Partitioning
Partitions can be enumerated and created using this library. Both MBR and GPT partitioning schemes are supported.
Call Storage::Disk::scanPartitions() to populate the device partition table.
For each partition, additional information can be obtained via Storage::Partition::diskpart() call.
Partition information is written using Storage::Disk::formatDisk().
Buffering
Block devices are typically used with a sector-based filing system such as FAT (see FatIFS).
Note that these filing systems are generally unsuitable for use with SPI flash without some kind
of intermediate wear-levelling management layer as provided by SD card controllers.
By default, block devices must be strictly accessed only by sector, i.e. for SD cards aligned chunks of 512 bytes.
This is rather inflexible so BlockDevice supports byte-level access using internal buffering,
which applications may enable using the allocateBuffers method.
This allows other filing systems to be used. LittleFS seems to work OK, although SPIFFS IFS Library does not. Partitions may also be used directly without any filing system.
Important
The Device::sync() method must be called at appropriate times to ensure data is actually written
to disk. Filing systems should take care of this internally when files are closed, for example.
Testing
Linux provides excellent support for testing generated image files.
Windows users may find this tool useful: https://www.diskinternals.com/linux-reader/.
Configuration
- DISK_MAX_SECTOR_SIZE
default: 512
Determines the minimum supported read/write block size for storage devices.
For example, AF disks use 4096-byte sectors so internal reads and writes must be a multiple of this value. This will increase the internal buffer sizes and so consume more RAM.
Acknowledgements
Code in this library is based on http://elm-chan.org/fsw/ff/00index_e.html. It has been heavily reworked using structures (from the Linux kernel) and separated from the actual FAT filing system implementation (FatIFS).
API
-
namespace Disk
-
Enums
-
enum class Error
Values:
-
enumerator XX
-
enumerator XX
-
enum class SysType : uint8_t
Identifies exact disk volume type.
Values:
-
enumerator unknown
Partition type not recognised.
-
enumerator fat12
-
enumerator fat16
-
enumerator fat32
-
enumerator exfat
-
enumerator unknown
-
enum SysIndicator
MBR partition system type indicator values.
Values:
-
enumerator SI_FAT12
-
enumerator SI_FAT16
FAT16 with fewer than 65536 sectors.
-
enumerator SI_FAT16B
FAT16B with 65536 or more sectors.
-
enumerator SI_IFS
-
enumerator SI_EXFAT
-
enumerator SI_FAT32X
FAT32 with LBA.
-
enumerator SI_FAT12
Functions
-
template<typename T>
auto getBlockCount(T byteCount, uint32_t blockSize)
-
uint32_t crc32_byte(uint32_t crc, uint8_t d)
-
uint32_t crc32(uint32_t bcc, const void *data, size_t length)
-
inline uint32_t crc32(const void *data, size_t length)
-
inline bool operator!(Error err)
-
Error formatDisk(BlockDevice &device, GPT::PartitionTable &table, const Uuid &diskGuid = {})
Partition a device using the GPT scheme.
- Parameters:
device –
table – Partitions to create
- Return values:
Error –
-
Error formatDisk(BlockDevice &device, MBR::PartitionTable &table)
Partition a device using the MBR scheme.
- Parameters:
device –
table – Partitions to create
- Return values:
Error –
-
inline SysType getSysTypeFromIndicator(SysIndicator si)
-
Error validate(BasePartitionTable &table, storage_size_t firstAvailableBlock, storage_size_t totalAvailableBlocks, uint32_t blockSize)
Validate partition table entries.
If size <= 100 then the actual size is calculated as a percentage and updated
If offset = 0 then a suitable location is found and the offset updated
On success, partition entries are ordered by position.
- Parameters:
firstAvailableBlock – First block number which may be allocated to a partition
totalAvailableBlocks – Number of blocks available for partition allocation
blockSize – Size of a block
- Return values:
Error – For each partition:
-
bool scanPartitions(Device &device)
Variables
-
constexpr uint32_t PARTITION_ALIGN = {0x100000U}
-
class BlockDevice : public Storage::Device
- #include <BlockDevice.h>
Base class for sector-addressable (block) devices.
Inherited classes must set the
sectorCountvalue, and implement the fourraw_xxxmethods.This class supports byte-level access using internal buffering, which if required must be enabled by the application via the
allocateBuffersmethod. Without buffering, read/writes must always be sector-aligned. Rrase must always be sector-aligned.For power-loss resiliency it is important to call
sync()at appropriate times. Filing system implementations should do this after closing a file, for example. Applications should consider this if leaving files open for extended periods, and explicitly call ‘flush’ on the filing system or ‘sync’ on the partition or device if necessary.Subclassed by Storage::Disk::HostFileDevice, Storage::SD::Card, USB::MSC::LogicalUnit
Public Functions
-
virtual bool read(storage_size_t address, void *dst, size_t size) override
Read data from the storage device.
- Parameters:
address – Where to start reading
dst – Buffer to store data
size – Size of data to be read, in bytes.
- Return values:
bool – true on success, false on error
-
virtual bool write(storage_size_t address, const void *src, size_t size) override
Write data to the storage device.
- Parameters:
address – Where to start writing
src – Data to write
size – Size of data to be written, in bytes.
- Return values:
bool – true on success, false on error
-
virtual bool erase_range(storage_size_t address, storage_size_t size) override
Erase a region of storage in preparation for writing.
- Parameters:
address – Where to start erasing
size – Size of region to erase, in bytes
- Return values:
bool – true on success, false on error
-
inline virtual size_t getBlockSize() const override
Obtain smallest allocation unit for erase operations.
-
inline virtual storage_size_t getSize() const override
Obtain addressable size of this device.
- Return values:
storage_size_t – Must be at least as large as the value declared in the hardware configuration
-
inline virtual storage_size_t getSectorCount() const override
Obtain total number of sectors on this device.
-
virtual bool sync() override
Flush any pending writes to the physical media.
Devices with intermediate buffering should implement this method.
- Return values:
bool – Return false if sync operation failed.
-
bool allocateBuffers(unsigned numBuffers)
Set number of sector buffers to use.
Required to support byte-level read/write operations on block devices. Buffering can improve performance, with diminishing returns above around 4 sectors.
- Parameters:
numBuffers – Number of buffers to allocate 1,2,4,8,etc. Pass 0 to deallocate/disable buffering.
- Return values:
bool – false on memory allocation error, or if failed to flush existing buffers to disk
-
struct Stat
- #include <BlockDevice.h>
-
struct Func
- #include <BlockDevice.h>
-
struct Func
-
virtual bool read(storage_size_t address, void *dst, size_t size) override
-
struct Buffer : public std::unique_ptr<uint8_t[]>
- #include <Buffer.h>
-
class BufferList
- #include <Buffer.h>
-
class HostFileDevice : public Storage::Disk::BlockDevice
- #include <HostFileDevice.h>
Create custom storage device using backing file.
Public Functions
-
HostFileDevice(const String &name, const String &filename, storage_size_t size)
Construct a file device with custom size.
- Parameters:
name – Name of device
filename – Path to file
size – Size of device in bytes
-
HostFileDevice(const String &name, const String &filename)
Construct a device using existing file.
Device will match size of existing file
- Parameters:
name – Name of device
filename – Path to file
-
inline virtual String getName() const override
Obtain unique device name.
-
inline virtual Type getType() const override
Obtain device type.
-
HostFileDevice(const String &name, const String &filename, storage_size_t size)
-
struct DiskPart
- #include <PartInfo.h>
Adds information specific to MBR/GPT disk partitions.
Subclassed by Storage::Disk::PartInfo
Public Functions
-
size_t printTo(Print &p) const
Print full contents of this structure.
Public Members
-
Uuid typeGuid
GPT type GUID.
-
Uuid uniqueGuid
GPT partition unique GUID.
-
SysType systype = {}
Identifies volume filing system type.
-
SysIndicator sysind = {}
Partition sys value.
-
size_t printTo(Print &p) const
-
struct PartInfo : public Storage::Partition::Info, public Storage::Disk::DiskPart
- #include <PartInfo.h>
In-memory partition information.
A disk Storage::Partition refers to this instance.
Public Functions
-
inline virtual const Disk::DiskPart *diskpart() const override
Obtain additional disk information.
Accessed via
Partition::diskpart()method
-
virtual size_t printTo(Print &p) const override
Print important fields only.
-
inline virtual const Disk::DiskPart *diskpart() const override
-
class BasePartitionTable : public OwnedLinkedObjectListTemplate<PartInfo>
- #include <PartInfo.h>
Common type for MBR/GPT partition table.
Subclassed by Storage::Disk::GPT::PartitionTable, Storage::Disk::MBR::PartitionTable
-
class Scanner
- #include <Scanner.h>
Class to iterate through disk partition tables.
Supports MBR and GPT partitioning schemes.
Public Functions
-
std::unique_ptr<PartInfo> next()
Obtains the next partition entry (if any)
-
std::unique_ptr<PartInfo> next()
-
class SectorBuffer : public std::unique_ptr<uint8_t[]>
- #include <SectorBuffer.h>
Buffer for working with disk sectors.
-
namespace EXFAT
-
namespace FAT
-
namespace GPT
-
-
struct SmingTypeGuid : public Uuid
- #include <GPT.h>
-
class PartitionTable : public Storage::Disk::BasePartitionTable
- #include <GPT.h>
Public Functions
-
inline bool add(const String &name, SysType sysType, storage_size_t offset, storage_size_t size, const Uuid &uniqueGuid = {}, const Uuid &typeGuid = {}, Partition::Flags flags = 0)
Add a new standard GPT partition definition.
- Parameters:
name – Partition name
sysType – Intended content for this partition
offset – Start offset, or 0 to have position calculated
size – Size of partition (in bytes), or percentage (0-100) of total partitionable disk space
uniqueGuid – Unique partition identifier (optional: will be generated if not provided)
typeGuid – Partition type GUID (default is BASIC_DATA)
flags –
- Return values:
bool – true on success
-
inline bool add(const String &name, Partition::FullType type, storage_size_t offset, storage_size_t size, const Uuid &uniqueGuid = {}, Partition::Flags flags = 0)
Add a new GPT partition for a regular Sming filing system.
Note
These partitions use a custom type GUID and won’t be recognised by external software.
- Parameters:
name – Partition name
type – Sming partition type/subtype
offset – Start offset, or 0 to have position calculated
size – Size of partition (in bytes), or percentage (0-100) of total partitionable disk space
uniqueGuid – Unique partition identifier (optional: will be generated if not provided)
flags –
- Return values:
bool – true on success
-
inline bool add(const String &name, SysType sysType, storage_size_t offset, storage_size_t size, const Uuid &uniqueGuid = {}, const Uuid &typeGuid = {}, Partition::Flags flags = 0)
-
struct SmingTypeGuid : public Uuid
-
namespace MBR
-
class PartitionTable : public Storage::Disk::BasePartitionTable
- #include <MBR.h>
Public Functions
-
inline bool add(SysType sysType, SysIndicator sysIndicator, storage_size_t offset, storage_size_t size, Partition::Flags flags = 0)
Add a new MBR partition definition.
Note
MBR does not have partition name field; this will appear as ‘mbr1’, ‘mbr2’, etc.
- Parameters:
sysType – Intended content for this partition (or ‘unknown’)
SysIndicator – Appropriate system code SI_xxx
offset – Start offset, or 0 to have position calculated
size – Size of partition (in bytes), or percentage (0-100) of total partitionable disk space
flags –
- Return values:
bool – true on success
-
inline bool add(SysType sysType, SysIndicator sysIndicator, storage_size_t offset, storage_size_t size, Partition::Flags flags = 0)
-
class PartitionTable : public Storage::Disk::BasePartitionTable
-
enum class Error
References
Used by
FatIFS Library
SD Storage Library
USB Library
Environment Variables
ENABLE_BLOCK_DEVICE_STATS
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
FatIFS
IFS library for Sming supporting FAT filesystems.
Formatting
Use IFS::FAT::formatVolume() to format a partition:
part = ... // Storage::Partition
int err = IFS::FAT::formatVolume(part);
Serial << "formatVolume: " << IFS::Error::toString(err) << endl;
An optional IFS::FAT::FormatOptions parameter can be used to provide additional settings:
IFS::FAT::FormatOptions opt{
.volumeLabel = "My FAT volume", // Volume label (distinct from partition name)
.types = Storage::Disk::SysType::fat32, // Only use FAT32 for this volume
};
int err = IFS::FAT::formatVolume(part, opt);
For more detailed low-control over the formatting:
IFS::FAT::FatParam param;
int err = IFS::FAT::calculateFatParam(part, opt, param);
if(err == FS_OK) {
// .. adjust settings in `param` before formatting
int err = IFS::FAT::formatVolume(part, param);
}
Mounting FAT volumes
Unlike SPIFFS or LittleFS, failure to mount a filing system does not auto-format the volume. Instead, this is left to the application to handle. For example:
auto fs = IFS::createFatFilesystem(part);
if(fs != nullptr) {
int err = fs->mount();
if(err == Error::BadFileSystem) {
// Filesystem layout is invalid...
fs->format();
err = fs->mount();
}
if(err != FS_OK) {
// Handle error
}
}
Configuration options
- ENABLE_EXFAT
default: 0 (disabled)
Set to 1 to enable support for EXFAT volumes. Requires
ENABLE_STORAGE_SIZE64option.
- ENABLE_FAT_TRIM
default: disabled
- When using devices with TRIM support this setting should be enabled.
- FAT_CODEPAGE
default: 437 (US English)
Acknowledgements
ChanN’s fatfs library <http://elm-chan.org/fsw/ff/00index_e.html>`
Linux Kernel <https://github.com/torvalds/linux/>
Linux utilities https://github.com/util-linux/util-linux
Linux FUSE exFAT filing system implementation https://github.com/relan/exfat
API Documentation
-
namespace FAT
Functions
-
int translateFatfsResult(uint8_t result, bool diskio_write)
-
String fatfsErrorToStr(uint8_t err)
-
ErrorCode calculateFatParam(Partition partition, const FormatOptions &opt, FatParam ¶m)
Deduce FAT volume parameters for given space.
- Parameters:
partition – The partition to format
opt – Formatting options
param – On success, contains calculated parameters for FAT volume
- Return values:
ErrorCode – When partitioning using MBR format, this method can be used to determine the
Sys indicatorvalue setting.
-
ErrorCode formatVolume(Partition partition, const FatParam ¶m)
Format partition using pre-calculated FAT parameters.
- Parameters:
partition – The partition to format
param – Detailed FAT parameters (returned from
calculateFatParam)
- Return values:
ErrorCode – This function allows fine control over exactly how a FAT partition is constructed. Generally the
calculateFatParamfunction should be used to populate theparamstructure, then any modifications can be made as required before actually formatting the volume.
-
inline ErrorCode formatVolume(Partition partition, const FormatOptions &opt = {})
Format partition with a blank FAT volume.
- Parameters:
partition – The partition to format
opt – Formatting options
- Return values:
ErrorCode –
-
class FileSystem : public IFS::IFileSystem
- #include <FileSystem.h>
Wraps fatfs
Public Functions
-
virtual int mount() override
Mount file system, performing any required initialisation.
- Return values:
error – code
-
virtual int getinfo(Info &info) override
get filing system information
- Parameters:
info – structure to read information into
- Return values:
int – error code
-
virtual int setProfiler(IProfiler *profiler) override
Set profiler instance to enable debugging and performance assessment.
- Parameters:
profiler –
- Return values:
int – error code - profiling may not be supported on all filesystems
-
virtual String getErrorString(int err) override
get the text for a returned error code
- Return values:
String –
-
virtual int opendir(const char *path, DirHandle &dir) override
open a directory for reading
- Parameters:
path – path to directory. nullptr is interpreted as root directory
dir – returns a pointer to the directory object
- Return values:
int – error code
-
virtual int readdir(DirHandle dir, Stat &stat) override
read a directory entry
Note
File system allocates entries structure as it usually needs to track other information. It releases memory when closedir() is called.
- Parameters:
dir –
stat –
- Return values:
int – error code
-
virtual int rewinddir(DirHandle dir) override
Reset directory read position to start.
- Parameters:
dir –
- Return values:
int – error code
-
virtual int closedir(DirHandle dir) override
close a directory object
- Parameters:
dir – directory to close
- Return values:
int – error code
-
virtual int mkdir(const char *path) override
Create a directory.
Only the final directory in the path is guaranteed to be created. Usually, this call will fail if intermediate directories are not present. Use
IFS::FileSystem::makedirs()for this purpose.- Parameters:
path – Path to directory
- Return values:
int – error code
-
virtual int stat(const char *path, Stat *stat) override
get file information
Returned stat will indicate whether the path is a mountpoint or directory. For a mount point, stats for the root directory of the mounted filesystem must be obtained by opening a handle then using
fstat:int handle = fs.open("path-to-mountpoint"); Stat stat; fs.fstat(handle, &stat); fs.close(handle);
- Parameters:
path – name or path of file/directory/mountpoint
s – structure to return information in, may be null to do a simple file existence check
- Return values:
int – error code
-
virtual int fstat(FileHandle file, Stat *stat) override
get file information
- Parameters:
file – handle to open file
stat – structure to return information in, may be null
- Return values:
int – error code
-
virtual int fcontrol(FileHandle file, ControlCode code, void *buffer, size_t bufSize) override
Low-level and non-standard file control operations.
To simplify usage the same buffer is used for both input and output. Only the size of the buffer is provided. If a specific FCNTL code requires more information then it will be contained within the provided data.
- Parameters:
file –
code – FCNTL_XXX code
buffer – Input/Output buffer
bufSize – Size of buffer
- Return values:
int – error code or, on success, data size
-
virtual int fsetxattr(FileHandle file, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute on an open file.
- Parameters:
file – handle to open file
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
-
virtual int fgetxattr(FileHandle file, AttributeTag tag, void *buffer, size_t size) override
Get an extended attribute from an open file.
- Parameters:
file – handle to open file
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
-
virtual int fenumxattr(FileHandle file, AttributeEnumCallback callback, void *buffer, size_t bufsize) override
Enumerate attributes.
- Parameters:
file – handle to open file
callback – Callback function to invoke for each attribute found
buffer – Buffer to use for reading attribute data. Use nullptr if only tags are required
bufsize – Size of buffer
- Return values:
int – error code, on success returns number of attributes read
-
virtual int setxattr(const char *path, AttributeTag tag, const void *data, size_t size) override
Set an extended attribute for a file given its path.
- Parameters:
path – Full path to file (or directory)
tag – The attribute to write
data – Content of the attribute. Pass nullptr to remove the attribute (if possible).
size – Size of the attribute in bytes
- Return values:
int – error code
-
virtual int getxattr(const char *path, AttributeTag tag, void *buffer, size_t size) override
Get an attribute from a file given its path.
- Parameters:
file – Full path to file (or directory)
tag – The attribute to read
buffer – Buffer to receive attribute content
size – Size of the buffer
- Return values:
int – error code, on success returns size of attribute (which may be larger than size)
-
virtual FileHandle open(const char *path, OpenFlags flags) override
open a file (or directory) by path
This function may also be used to obtain a directory handle to perform various operations such as enumerating attributes. Calls to read or write on such handles will typically fail.
- Parameters:
path – full path to file
flags – Desired access and other options
- Return values:
FileHandle – file handle or error code
-
virtual int close(FileHandle file) override
close an open file
- Parameters:
file – handle to open file
- Return values:
int – error code
-
virtual int read(FileHandle file, void *data, size_t size) override
read content from a file and advance cursor
- Parameters:
file – handle to open file
data – buffer to write into
size – size of file buffer, maximum number of bytes to read
- Return values:
int – number of bytes read or error code
-
virtual int write(FileHandle file, const void *data, size_t size) override
write content to a file at current position and advance cursor
- Parameters:
file – handle to open file
data – buffer to read from
size – number of bytes to write
- Return values:
int – number of bytes written or error code
-
virtual file_offset_t lseek(FileHandle file, file_offset_t offset, SeekOrigin origin) override
change file read/write position
- Parameters:
file – handle to open file
offset – position relative to origin
origin – where to seek from (start/end or current position)
- Return values:
file_offset_t – current position or error code
-
virtual int eof(FileHandle file) override
determine if current file position is at end of file
- Parameters:
file – handle to open file
- Return values:
int – 0 - not EOF, > 0 - at EOF, < 0 - error
-
virtual file_offset_t tell(FileHandle file) override
get current file position
- Parameters:
file – handle to open file
- Return values:
file_offset_t – current position relative to start of file, or error code
-
virtual int ftruncate(FileHandle file, file_size_t new_size) override
Truncate (reduce) the size of an open file.
Note
In POSIX
ftruncate()can also make the file bigger, however SPIFFS can only reduce the file size and will return an error if newSize > fileSize- Parameters:
file – Open file handle, must have Write access
newSize –
- Return values:
int – Error code
-
virtual int flush(FileHandle file) override
flush any buffered data to physical media
- Parameters:
file – handle to open file
- Return values:
int – error code
-
virtual int rename(const char *oldpath, const char *newpath) override
rename a file
- Parameters:
oldpath –
newpath –
- Return values:
int – error code
-
virtual int remove(const char *path) override
remove (delete) a file by path
- Parameters:
path –
- Return values:
int – error code
-
virtual int fremove(FileHandle file) override
remove (delete) a file by handle
- Parameters:
file – handle to open file
- Return values:
int – error code
-
virtual int format() override
format the filing system
Note
this does a default format, returning file system to a fresh state The filing system implementation may define more specialised methods which can be called directly.
- Return values:
int – error code
-
virtual int check() override
Perform a file system consistency check.
Note
if possible, issues should be resolved. Returns 0 if file system checked out OK. Otherwise there were issues: < 0 for unrecoverable errors,
0 for recoverable errors.
- Return values:
int – error code
-
virtual int mount() override
-
struct FormatOptions
- #include <Format.h>
Formatting options.
Public Members
-
SysTypes types = {0}
Valid partition format types.
-
uint8_t numFats = {1}
Number of FATs (1 or 2)
-
unsigned align = {0}
Data area alignment (sector)
-
unsigned numRootEntries = {512}
Number of root directory entries.
-
uint32_t clusterSize = {4096}
Cluster size (byte)
-
SysTypes types = {0}
-
struct FatParam
- #include <Format.h>
Public Members
-
uint32_t sectorsPerBlock
Flash erase block size.
-
uint16_t sectorsPerCluster
Set to 0 for auto-calculation.
-
uint32_t sectorsPerBlock
-
int translateFatfsResult(uint8_t result, bool diskio_write)
References
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Flash In-Place
This library is intended to support in-place application updates where there is insufficient available flash storage for more than one application partition.
This is aimed at the (Rp2040) Pico (W) which has only 2MBytes of flash.
This library takes a firmware image stored elsewhere in flash, then runs a simple piece of code running from RAM. This copies the new firmware into the application partition then reboots.
This is similar to the Arduino Pico approach, but it doesn’t require any bootloader changes and the critical copier code (in RAM) is much simpler as it doesn’t need to know anything about filing systems; all it does is copy blocks from one region of flash to another.
Operation
- Verify the source image
The application handles this e.g. CRC32 checksum.
Build a block list
Most file systems will store larger files (especially of 2-300KBytes or more) in multiple, non-contiguous regions of flash storage. These regions are known as ‘extents’, and this information is obtained via the
IFS::IFilelSystem::fgetextents()call.The
FlashIPclass uses this to build a list, which the copier code then executes from RAM. This keeps the copier code very simple.Example:
// Get the target partition auto part = *Storage::findPartition(Storage::Partition::Type::app); // Build the block list FlashIP fip; if(fip.addFile(part, 0, "firmware.bin")) { // Copy the image and reboot fip.execute(); }
- Write the block list to flash
This code runs exclusively in RAM. All interrupts and DMA are disabled first. The block list is then iterated with a read/erase/program procedure.
Reboot
Issues
- Recovery
There is no recovery available if the update operation is interrupted part-way. If physical device access is available then firmware can be re-flashed via USB.
Writing a second-stage bootloader and storing the block list in FLASH would mitigate this issue. For example, we’d reserve the first two flash sectors for this, with the application itself relocated. Each block-list entry will be invalidated once complete so the bootloader need only write incomplete blocks.
- Compression
Using a compressed firmware image would require e.g. uzlib to be located in RAM. We can add a build variable to make that happen.
A quick test on the release
HttpServer_ConfigNetworkimage using gzip reduces the size from 341136 to 280084 bytes (82% of the original size).- Encryption
Note that where updating via local network, encryption may not be necessary.
If encryption is used, the application would need to decrypt that first. Typically this is done as the file is received, before writing the data to the local file.
References
Used by
Basic FlashIP Sample
SoC support
host
rp2040
GoogleCast
This component allows you to communicate with Chrome Cast dongles or smart TVs that support the Google Cast Protocol.
Using
Add these lines to your application componenent.mk file:
COMPONENT_DEPENDS += GoogleCast ENABLE_SSL := Bearssl
Add these lines to your application:
#include <Network/GoogleCast/Client.h>Basic example:
#include <Network/GoogleCast/Client.h> GoogleCast::Client castClient; void gotIp(IpAddress ip, IpAddress mask, IpAddress gateway) { // connect directly to the IP of the devise castClient.connect(IpAddress("192.168.10.15")); castClient.onConnect([](bool success) { Serial.print(F("Client connect: ")); Serial.println(success ? "OK" : "FAIL"); if(success) { // Starting YouTube on the device castClient.receiver.launch(APPID_YOUTUBE); } }); }
Re-Generating C files from proto file
You can re-generate the C files from the proto file. This can be useful if you want to use a newer version of the Google Cast proto file. Make sure to compile it using nanopb. The re-generation can be done with the commands below:
cd $SMING_HOME/Libraries/GoogleCast/samples/basic
make python-requirements # should be executed once to download and install the required python packages
make rebuild-cast-proto
Protocol
There are multiple documents in internet that describe the Google Cast protocol and its inner workings. As a start one can take a look at the presentation and documents below.
App-Ids
List of current APP-IDS https://clients3.google.com/cast/chromecast/device/baseconfig
Individual app configurations are obtained from this url: https://clients3.google.com/cast/chromecast/device/app?a={appid}
Where {appid} is the id of the app, and the output of this is JSON in the format of a receiver app definition.
References
Used by
Google Cast Client Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Sming Graphics Library
An asynchronous graphics management library intended for SPI-based displays. See the Basic Graphics sample application for demonstration of how to use it.
For a discussion about multitasking and how Sming works, see Multitasking.
Getting started
On first use several required python modules must be installed. Try building a sample application:
cd $SMING_HOME
make fetch Graphics
cd Libraries/Graphics/samples/Basic_Graphics
make python-requirements
make
Virtual Screen
This is a display device for use by the Host Emulator. The ‘display’ is a python application which communicates
with the Sming Graphics::Display::Virtual display driver using a TCP socket.
Graphics processing is handled using SDL2.
To start the virtual screen server type make virtual-screen from your project directory.
The default TCP port is 7780. If you need to change this, use:
make virtual-screen VSPORT=7780
The screen server’s IP address is shown in the caption bar. Build and run your project in host mode as follows:
make SMING_ARCH=Host
make run VSADDR=192.1.2.3
Some definitions
Graphics::ObjectDefines a specific graphical item, such as a line, circle or image.
Graphics::SceneObjectList of objects describing the components of a displayed image (line, rectangle, circle, image, text, etc.) Also contains a list of assets which allows objects to be re-used within a scene.
Applications use methods such as drawCircle to construct a scene.
Graphics::TextBuilderHelper object to simplify text layout, handling wrapping, alignment etc.
Graphics::SurfaceRectangular array of pixels representing a display or offscreen bitmap
Graphics::DeviceA display device. This provides a single
Surfacefor access to the screen.Graphics::RendererClass which performs background rendering tasks such as drawing a sequence of lines, filling a circle, etc. Each renderer performs a single, specific-task. The renderers are created by the Surface which allows display devices to provide customised implementations.
Graphics::PixelFormatDescribes the format of pixel data required by a Surface.
For example, the ILI9341 display uses RGB565 colours (2 bytes) with the MSB first. It can use other pixel formats (RGB 6:6:6), but this is how the driver sets the display up during initialisation.
Applications do not need to be aware of this, however, and simply express colours in a 32-bit
Graphics::Colortype. This includes an alpha channel, where 255 is fully opaque and 0 is fully transparent. Conversion is handled in a single step (viaGraphics::pack()) so further manipulation (even byte-swapping) is not required by the display driver.Note that the ILI9341 display has another quirk in that display data is always read back in 24-bit format (18 significant bits). This detail is managed by the display driver which does the conversion during read operations. See
Graphics::Surface::readDataBuffer().Graphics::BlendThis is a virtual base class used to implement custom colour blending operations. Currently only supported for memory images using
Graphics::ImageSurface.
Resources
Projects describe which fonts and image resources they require in a JSON resource script.
- This is compiled to produce two files:
- $(OUT_BASE)/resource.h
A header file describing the resources and containing indices and other lookup tables. Include this ONCE from a source file.
- $(OUT_BASE)/resource.bin
Contains image data and font glyph bitmaps. This file can be linked into the application or stored in a file. A better approach is to store it in a dedicated partition as this avoids any filing system overhead. The library accesses this as a stream - applications must call
Graphics::Resource::init().
The goal is to enable an optimal set of resources to be produced for the target display from original high-quality assets.
Usage
Create a directory to contain project resources, e.g.
resource.Place any custom images, fonts, etc. into this directory
Create a resource script file, e.g.
resource/graphics.rc. See below for details on editing this file.Add an entry to the project
component.mkfile:RESOURCE_SCRIPT := resource/graphics.rcOptionally, change the directory where compiled resources are written:
RESOURCE_OUTPUT := out
The general structure of a resource script is:
{
"resources": {
"<type>": {
"<name>": {
...
}
}
}
}
Fonts
A Graphics::Font contains up to four typefaces which correspond
to the selected style (normal, italic, bold or bold+italic).
The default font is Graphics::LcdFont, a simple Arduino fixed Adafruit GFX font.
Only the ‘normal’ typeface is defined.
This is linked into the program image.
All other fonts are loaded from resources. These are parsed from their original format and converted to an efficient internal format based on the Arduino GFX font library.
Resource script entries look like this:
Styles are optional but a font must have at least one typeface.
By default, all ASCII characters from 0x20 (space) to 0x7e (~).
The codepoints parameter is a comma-separated list of ranges:
a-z,A-Z,0-9,0x4000-0x4050
This overrides the default and includes only characters and digits, plus unicode characters in the given range.
The chars parameter is a simple list of characters, e.g. “Include these chars”.
Both lists are combined, de-duplicated and sorted in ascending order.
The following font classes are currently supported:
- GFX
Adafruit GFX library font files. These are identified using a “gfx/” prefix for the font files. See
resource/fonts/GFX.- Linux
Linux console bitmap fonts. These are in .c format, identified by “linux/” path prefix. See
resource/fonts/Linux.- PFI
The “.pfi” file extension is a Portable Font Index. Font glyph bitmap information is contained in an associated .pbm file. See
resource/fonts/PFI.- VLW
Smoothed fonts produced by the https://processing.org/ library with a “.vlw” file extension. These are from the TFT_eSPI library. See
resource/fonts/VLW.Note that TTF/OTF scalable vector fonts are supported directly by this library so is the preferred format for new fonts.
- freetype
- Font type is identified from the file extension:
- .ttf
TrueType
- .otf
OpenType
- .pcf, .pcf.gz
X11 bitmap font
The
freetypelibrary supports other types so if required these are easily added.
- These fonts have some additional parameters:
- “mono”: <True/False>
Whether to produce monochrome (1-bit) or grayscale (8-bit alpha) glyphs. If not specified, defaults to grayscale.
- “size”: <Point size of font>
e.g. 16, 14.5
Images
Resource script entries look like this:
"image": {
"<name>": {
"format": "<target format>", // RGB565, RGB24 or BMP
"source": "filename" | "url",
// Optional list of transformations to apply
"transform": {
"width": target width, // Height will be auto-scaled
"height": target height, // Width will be auto-scaled
"crop": "width,height", // Crop image to given size from (0, 0)
"crop": "x,y,width,height", // Crop to selected area of image
"resize": "width,height", // Resize image
"flip": "left_right" | "top_bottom", // Flip the image horizontally or vertically
"rotate": angle in degrees, // Rotate the image
"color": { // Convert image to grayscale then colorise it
"black-color": blackpoint, // Color at blackpoint, e.g. "darkblue": 0
"mid-color": midpoint, // OPTIONAL
"white-color": whitepoint // Color at whitepoint, e.g. "lightred": 255
}
}
}
}
The Python pillow library is used to read and process images.
The format parameter is intended to produce a raw, un-compressed image compatible with the
target display device.
Alternatively, specify “BMP” to output a standared .bmp file or omit to store the original
image contents un-processed.
Scene construction
Instead of writing directly to the display, applications create a Graphics::SceneObject
which will contain a description of the scene to be drawn.
The scene can then be constructed using methods such as drawLine, drawRect, etc.
These create an object describing the operation and adds it to the scene.
When the scene is fully described, the application calls the display device’s render method
and the rendering begins in the background.
A callback function may be provided which will be invoked when the scene has been drawn.
The application is free to start creating further scenes, etc.
Note: Scenes are never modified during rendering so can be drawn multiple times if required.
Rendering
This is the process of converting the scene objects into pixels which are then drawn into a Graphics::Surface.
Computation and primitive rendering is done in small, manageable chunks via the task queue. Data is sent to the screen using the appropriate display driver. The ILI9341 driver uses interrupts and hardware (SPI) transfers to do its work.
If the primitive object is simple (e.g. block fill, horizontal/vertical line) then it can be written to the display buffer immediately.
More complex shapes (images, text, diagonal lines, rectangles, triangles, circles, etc.) create a
specific Graphics::Renderer instance to do the work.
Transparency (alpha-blending) involves a read-modify-write cycle. The ILI9341 driver can also handle small transparent rectangles (including individual pixels) to assist with line drawing. Larger areas are handled by the appropriate renderer.
Shapes such as rectangles, triangles, circles, etc. are described as a set of connected lines and points
which is then iterated through and rendered using fill operations.
The standard Graphics::Surface implementation uses a standard set of renderers to do this work,
however these can be overridden by display devices to make use of available hardware features.
Transparency
With sufficient RAM this is a trivial exercise as we just render everything to a memory surface then copy it to the display. For updating small areas of the screen this may be the best approach. However, even a small 240x320 pixel display would require 150kBytes with RGB565.
The more general approach adopted by this library is to read a small block of pixels from the display,
combine them with new data as required then write the block back to the display.
The Graphics::TextRenderer does this by default. The algorithm also ensures that text can be rendered
correctly with filled backgrounds (solid, transparent or custom brushes).
Note
It may not be possible to read from some displays.
For example, a small 1-inch display is commonly available using the Graphics::Display::ST7789V
controller connected in 4-wire mode, but with the SDO line unconnected internally.
In such cases transparency must be handled by the application using memory surfaces.
Display driver
The ILI9341 display driver buffers requests into a set of hardware-specific commands with associated data. When the request buffer is full, it is sent it to the hardware using the HardwareSPI library. This particular display is typically configured in 4-wire SPI mode which requires an additional GPIO to select between command and data transfers. The driver manages this within its interrupt service routines.
Two request buffers are used so that when one buffer has completed transfer the next can begin immediately. This switch is handled within the interrupt service routine, which also schedules a task callback to the renderer so it may re-fill the first request buffer.
Configuration variables
- RESOURCE_SCRIPT
Location of .rc resource script for your project, relative to project directory. Define this in your project’s
component.mkfile.
- RESOURCE_OUTPUT
Path to directory where compiled resources will be written, relative to project directory. Default is project build directory, e.g.
out/Rp2040/debug/build.
- VSADDR
e.g. 192.168.1.105:7780
TCP address of the virtual screen server application, as shown in the title bar.
- VSPORT
default: 7780
Port number to use for virtual screen.
- ENABLE_GRAPHICS_DEBUG
default 0 (off)
Set to ‘1’ to enable additional debug output from renderers
- ENABLE_GRAPHICS_RAM_TRACKING
default 0 (off)
Set to ‘1’ to enable additional diagnistics to assist with tracking RAM usage
Further work
The list of possible updates to the library is endless. Some thoughts:
- Text wrapping
Currently wraps at common punctuation characters. Add options to specify how to break text (character boundaries, word boundaries, break characters, etc.)
- Rich Text
Add support for HTML, markdown and other suitable formats. This could be handled by the resource compiler by converting to Graphics::Drawing resources.
- Transparent bitmaps
Simple transparency can be handled using
BlendImages with per-pixel alpha are not currently implemented. The resource compiler would be extended to support creation of images in ARGB format, with appropriate renderer updates.- Text glyph alignment
Text is drawn using the upper-left corner as reference. It may be desirable to change this, using the baseline or bottom corner instead. This can be added as an option.
- Metafiles
Rather than describing scenes in code, we should be able to draw them using other tools and export the layouts as metafiles. Similarly, we might like to export our scenes as metafiles for later re-use.
An example use case is for a ‘scribble’ application, recording pen strokes a touch-enabled screen.
The
DrawingAPI is intended for this use, providing a simple Graphics Description Language (GDL). The providedGDRAW_xxxmacros allow embedding of these scripts within programs. This feature is restricted so appropriate only for very simple constructions.Drawings could be described in textual resource files and compiled during build.
A simple utility program can be written to convert between text and binary GDL formats. Other conversion programs then only need to generate the text equivalent. This will allow the binary format to change as required.
It would be highly desirable to give subroutines textual identifiers. These would be resolved into numeric identifiers for the binary format.
- Interactive objects
Drawing buttons, for example, might be done by defining the basic button in GDL.
The button routine could use the following:
pen #0: Inactive surround pen #1: Active surround brush #0: Inactive fill brush #1: Active fill
These slots can be set depending on the desired state by program code. It would be more efficient if GDL can do this based on other parameters, or perhaps by registering callbacks or control objects to make these decisions.
For example, a
ButtonObjectcould use GDL to draw itself and provide callbacks to handle state control, etc. This would also be required for control purposes.To support use of touch screens we need the ability to query scenes and identify objects at a specific position. Those objects may then be adjusted and redrawn to reflect ‘push’ operations, etc.
For example, we might create a button control by defining a scene describing how the button looks, then create an InteractiveObject containing a reference to that scene plus some kind of user identifier and location of the control ‘hot spot’.
- Brushes, Gradient fills, etc.
Rectangular gradient fills can be done by creating a custom ImageObject and drawing it. The object just generates the pixel data as requested according to configured parameters.
This is the basis for Brush objects which would be used to render circles, rounded rectangles and glyphs. That would allow for all sorts of useful effects, including clipping images to other shapes.
Tiling an image can be done by filling a rectangle with an ImageBrush, which deals with coordinate wrapping and returns the corresponding region from a source ImageObject.
- Cursors, sprites and priority drawing
Response to user input should be prioritised, such as movement of a cursor or mouse pointer.
Cursors are drawn over everything else. Although some hardware may support cursors directly, for now we’ll assume not. Similarly text carets (e.g. ‘_’, ‘>’ prompts) can be drawn with priority so that changes in their position are updated ahead of other items.
A read-xor-write pattern is simple enough to implement within a Command List, so that the XOR operation is done in interrupt context. This should provide sufficient priority.
Sprites can be managed using a Surface filter layer which tracks Address Window operations. If data is written to an area occupied by a sprite, then that data can be modified (using XOR) before passing through to provide flicker-free drawing.
Repeat fills make this a little tricky though. An alternative is to identify when a sprite will be affected and, if so, remove it before drawing the new information then re-draw it afterwards. This could be implemented using Regions so that only the affect portion is updated.
Each sprite must store both its own contents (foreground), plus the original display contents (background) excluding any other sprites. Multiple overlapping are combined at point of rendering.
References
- File formats
BMP File Format (Wikipedia)
- Fonts
Bitmap fonts Collection of monospaced bitmap fonts for X11
- Graphics libraries
- Metafiles
WebCGM 2.1 Computer Graphics Metafile standard
- Papers
The Beauty of Bresenham’s Algorithm Discuss anti-aliasing techniques
API
-
namespace Graphics
Typedefs
-
using GlyphOptions = TextOptions
-
using CustomObject = ObjectTemplate<Object::Kind::Custom>
Base class for a custom object.
-
using AssetID = uint16_t
Numeric identifiers for re-useable objects.
Enums
-
enum class Color : uint32_t
Standard colour definitions.
Values:
-
enumerator None
-
enumerator XX
-
enumerator XX
-
enumerator None
-
enum ColorOrder
Order refers to colour order within bitstream.
Values:
-
enumerator orderRGB
-
enumerator orderBGR
-
enumerator orderRGB
-
enum class Orientation
Defines orientation of display.
Values:
-
enumerator normal
-
enumerator deg0
-
enumerator deg90
-
enumerator deg180
-
enumerator deg270
-
enumerator normal
-
enum class Align
Values:
-
enumerator Near
-
enumerator Centre
-
enumerator Far
-
enumerator Left
-
enumerator Top
-
enumerator Center
-
enumerator Right
-
enumerator Bottom
-
enumerator Near
-
enum class Origin
Points on a compass.
These are ordered by angles in increments of 45 degrees.
Values:
-
enumerator E
-
enumerator NE
-
enumerator N
-
enumerator NW
-
enumerator W
-
enumerator SW
-
enumerator S
-
enumerator SE
-
enumerator Centre
-
enumerator TopLeft
-
enumerator Top
-
enumerator TopRight
-
enumerator Left
-
enumerator Center
-
enumerator Right
-
enumerator BottomLeft
-
enumerator Bottom
-
enumerator BottomRight
-
enumerator E
Functions
-
inline constexpr uint32_t getColorValue(uint8_t r, uint8_t g, uint8_t b, uint8_t a = 255)
Obtain 24-bit colour value from red, green and blue components.
-
inline constexpr Color makeColor(uint8_t r, uint8_t g, uint8_t b, uint8_t a = 255)
Function to create a custom colour.
-
inline uint8_t getBytesPerPixel(PixelFormat format)
Get number of bytes required to store a pixel in the given format.
-
PixelBuffer pack(PixelBuffer src, PixelFormat format)
Convert RGB colour into packed format.
- Parameters:
color – The RGB colour to convert
format – The desired pixel format
- Return values:
PackedColor –
-
inline PackedColor pack(Color color, PixelFormat format)
-
PixelBuffer unpack(PixelBuffer src, PixelFormat format)
Convert packed colour into RGB.
- Parameters:
src – PixelBuffer loaded with packed colour
format – The exact format of the source colour
- Return values:
Color – The corresponding RGB colour
-
inline Color unpack(PackedColor packed, PixelFormat format)
-
inline Color unpack(uint32_t packed, PixelFormat format)
-
size_t writeColor(void *buffer, PackedColor color, PixelFormat format)
Store a packed colour value into memory.
- Parameters:
buffer – Where to write the colour value
color – The value to write
- Return values:
size_t – The number of bytes written
-
inline size_t writeColor(void *buffer, Color color, PixelFormat format)
-
size_t writeColor(void *buffer, PackedColor color, PixelFormat format, size_t count)
Store a block of packed colours into memory.
- Parameters:
buffer – Where to write the colour value
color – The value to write
count – The number of times to repeat the colour
- Return values:
size_t – The number of bytes written
-
inline size_t writeColor(void *buffer, Color color, PixelFormat format, size_t count)
-
size_t convert(const void *srcData, PixelFormat srcFormat, void *dstBuffer, PixelFormat dstFormat, size_t numPixels)
Convert block of data from one pixel format to another.
-
void highlightText(SceneObject &scene)
Highlight text segments and print details.
-
void highlightText(TextObject &object)
-
uint16_t swapBytes(uint16_t w)
-
uint32_t makeWord(uint16_t w1, uint16_t w2)
-
inline Origin opposite(Origin o)
Get the origin for the opposite side of the rectangle.
e.g. E -> W, NE -> SW
-
inline constexpr Size rotate(Size size, Orientation orientation)
-
uint16_t normaliseAngle(int angle)
Make 0 <= angle < 360.
-
class AbstractDisplay : public Graphics::Device, public Graphics::RenderTarget
- #include <AbstractDisplay.h>
Use this class as a base for all types of display drivers.
Subclassed by Graphics::Display::NullDevice, Graphics::Display::Virtual, Graphics::SpiDisplay
-
struct AddressWindow
- #include <AddressWindow.h>
Manages a rectangular area of display memory with position information.
Accesses to display memory is controlled by first setting an active Address Window. This is a rectangular area into which following writes (or reads) will store data.
Although the display hardware usually manages this some operations require tracking the position within the driver.
Public Functions
-
inline size_t getPixelCount() const
Get remaining pixels in window from current position.
-
inline size_t getPixelCount() const
-
class Asset : public LinkedObjectTemplate<Asset>, public Graphics::Meta
- #include <Asset.h>
An asset is used to render an Object, but is not itself drawable.
Subclassed by Graphics::AssetTemplate< AssetType::Object >, Graphics::AssetTemplate< AssetType::Text >, Graphics::AssetTemplate< AssetType::Font >, Graphics::AssetTemplate< AssetType::Surface >, Graphics::AssetTemplate< AssetType::TextureBrush >, Graphics::AssetTemplate< AssetType::Pen >, Graphics::AssetTemplate< AssetType::Typeface >, Graphics::AssetTemplate< AssetType::SolidBrush >, Graphics::AssetTemplate< AssetType::Blend >, Graphics::AssetTemplate< asset_type >
-
class SolidBrush : public Graphics::AssetTemplate<AssetType::SolidBrush>
- #include <Asset.h>
-
class TextureBrush : public Graphics::AssetTemplate<AssetType::TextureBrush>
- #include <Asset.h>
Subclassed by Graphics::GradientBrush, Graphics::ImageBrush
-
class GradientBrush : public Graphics::TextureBrush
- #include <Asset.h>
-
class ImageBrush : public Graphics::TextureBrush
- #include <Asset.h>
Brush using pixels from image.
-
class Brush : public Graphics::Meta
- #include <Asset.h>
The source of colour for drawing.
Subclassed by Graphics::Pen
-
class Pen : public Graphics::Brush
- #include <Asset.h>
Subclassed by Graphics::PenAsset
-
class PenAsset : public Graphics::AssetTemplate<AssetType::Pen>, public Graphics::Pen
- #include <Asset.h>
-
class TypeFace : public Graphics::AssetTemplate<AssetType::Typeface>
- #include <Asset.h>
Base class for a loaded typeface, e.g. Sans 16pt bold.
Subclassed by Graphics::LcdTypeFace, Graphics::ResourceTypeface
Public Functions
-
virtual FontStyles getStyle() const = 0
Style of this typeface (bold, italic, etc.)
-
virtual uint8_t height() const = 0
Get height of typeface, same for all characters.
-
virtual uint8_t descent() const = 0
How many pixels from bottom of em-square to baseline
-
virtual GlyphMetrics getMetrics(char ch) const = 0
Get metrics for a character.
-
virtual std::unique_ptr<GlyphObject> getGlyph(char ch, const GlyphOptions &options) const = 0
Get the glyph for a character.
Caller is responsible for destroying the glyph when no longer required.
- Parameters:
ch –
options – Options to control how the glyph is drawn (colour, shading, etc)
- Return values:
GlyphObject* – The glyph, nullptr if no glyph exists in the typeface for this character
-
inline uint8_t baseline() const
Get baseline relative to top of mbox.
-
uint16_t getTextWidth(const char *text, uint16_t length) const
Compute displayed width for a text string.
-
virtual FontStyles getStyle() const = 0
-
class Font : public Graphics::AssetTemplate<AssetType::Font>
- #include <Asset.h>
Base class for a loaded font.
As there are various formats for defining fonts we need a consistent interface for accessing and rendering them, which this class defines.
Fonts are specified by family and size, and can be registered with various typefaces.
Subclassed by Graphics::LcdFont, Graphics::ResourceFont
-
class ResourceTypeface : public Graphics::TypeFace
- #include <Asset.h>
Public Functions
-
inline virtual FontStyles getStyle() const override
Style of this typeface (bold, italic, etc.)
-
inline virtual uint8_t height() const override
Get height of typeface, same for all characters.
-
inline virtual uint8_t descent() const override
How many pixels from bottom of em-square to baseline
-
virtual GlyphMetrics getMetrics(char ch) const override
Get metrics for a character.
-
virtual std::unique_ptr<GlyphObject> getGlyph(char ch, const GlyphOptions &options) const override
Get the glyph for a character.
Caller is responsible for destroying the glyph when no longer required.
- Parameters:
ch –
options – Options to control how the glyph is drawn (colour, shading, etc)
- Return values:
GlyphObject* – The glyph, nullptr if no glyph exists in the typeface for this character
-
inline virtual FontStyles getStyle() const override
-
class TextAsset : public Graphics::AssetTemplate<AssetType::Text>
- #include <Asset.h>
-
class ObjectAsset : public Graphics::AssetTemplate<AssetType::Object>
- #include <Asset.h>
-
class AssetList : public OwnedLinkedObjectListTemplate<Asset>
- #include <Asset.h>
-
class Blend : public Graphics::AssetTemplate<AssetType::Blend>
- #include <Blend.h>
Blend operations.
See MemoryImageSurface::write
Subclassed by Graphics::BlendTemplate< BlendXor, BlendMode::Xor >, Graphics::BlendTemplate< BlendWrite, BlendMode::Write >, Graphics::BlendTemplate< BlendMask, BlendMode::Mask >, Graphics::BlendTemplate< BlendXNor, BlendMode::XNor >, Graphics::BlendTemplate< BlendAlpha, BlendMode::Alpha >, Graphics::BlendTemplate< BlendTransparent, BlendMode::Transparent >, Graphics::BlendTemplate< Class, blendMode >
-
template<class Class, BlendMode blendMode>
class BlendTemplate : public Graphics::Blend - #include <Blend.h>
-
class BlendWrite : public Graphics::BlendTemplate<BlendWrite, BlendMode::Write>
- #include <Blend.h>
-
class BlendXor : public Graphics::BlendTemplate<BlendXor, BlendMode::Xor>
- #include <Blend.h>
-
class BlendXNor : public Graphics::BlendTemplate<BlendXNor, BlendMode::XNor>
- #include <Blend.h>
-
class BlendMask : public Graphics::BlendTemplate<BlendMask, BlendMode::Mask>
- #include <Blend.h>
-
class BlendTransparent : public Graphics::BlendTemplate<BlendTransparent, BlendMode::Transparent>
- #include <Blend.h>
-
class BlendAlpha : public Graphics::BlendTemplate<BlendAlpha, BlendMode::Alpha>
- #include <Blend.h>
- #include <Buffer.h>
Shared heap-allocated data buffer.
Used for write operations with data outside Command List.
- #include <Buffer.h>
-
struct ReadBuffer
- #include <Buffer.h>
Buffer used for reading pixel data from device.
Subclassed by Graphics::ReadStatusBuffer
Public Members
-
SharedBuffer data
Buffer to read pixel data.
-
uint16 offset = {0}
Offset from start of buffer to start writing.
-
PixelFormat format = {}
Input: Requested pixel format, specify ‘None’ to get native format.
-
SharedBuffer data
-
struct ReadStatus
- #include <Buffer.h>
Stores result of read operation.
Public Members
-
size_t bytesRead = {0}
On completion, set to actual length of data read.
-
PixelFormat format = {}
Format of data.
-
size_t bytesRead = {0}
-
struct ReadStatusBuffer : public Graphics::ReadBuffer
- #include <Buffer.h>
Composite ReadBuffer with status.
-
union PixelFormatStruct
- #include <Colors.h>
Public Functions
-
inline uint8_t byteCount() const
-
inline uint8_t bitsPerPixel() const
-
inline ColorOrder colorOrder() const
Public Members
-
PixelFormat format
-
uint8_t mByteCount
-
uint8_t mBitsPerPixel
-
uint8_t mColorOrder
-
struct Graphics::PixelFormatStruct::[anonymous] [anonymous]
-
inline uint8_t byteCount() const
-
struct PackedColor
- #include <Colors.h>
Colour in device pixel format.
-
union PixelBuffer
- #include <Colors.h>
Structure used to perform pixel format conversions.
-
struct BGR24
- #include <Colors.h>
-
struct BGRA32
- #include <Colors.h>
-
struct RGB24
- #include <Colors.h>
-
struct RGB565
- #include <Colors.h>
-
struct BGR24
-
class Console : public Print
- #include <Console.h>
Public Functions
-
inline Console(AbstractDisplay &display, RenderQueue &renderQueue)
Console constructor.
- Parameters:
display – Output device
renderQueue – Allows shared access to display
-
void systemDebugOutput(bool enable)
Use console for debug output.
- Parameters:
enable – true to divert m_puts to console, false to disable debug output
-
void pause(bool state)
Suspend/resume output to display.
While paused, content will be buffered in RAM.
- Parameters:
state – true to stop output, false to resume
-
inline bool isPaused() const
Determine if output is paused.
-
virtual size_t write(const uint8_t *data, size_t size) override
Writes characters from a buffer to output stream.
- Parameters:
buffer – Pointer to character buffer
size – Quantity of characters to write
- Return values:
size_t – Quantity of characters written to stream
-
inline virtual size_t write(uint8_t c) override
Writes a single character to output stream.
- Parameters:
c – Character to write to output stream
- Return values:
size_t – Quantity of characters written to output stream
-
inline Console(AbstractDisplay &display, RenderQueue &renderQueue)
-
class Button : public Graphics::Control
- #include <Button.h>
Basic interactive button on screen.
Responsible for drawing area within designated rectangle
-
class Control : public Graphics::ObjectTemplate<object_kind>
- #include <Control.h>
Basic interactive button on screen.
Responsible for drawing area within designated rectangle
Subclassed by Graphics::Button, Graphics::Label, Graphics::TimeClock
Public Functions
-
class List : public LinkedObjectListTemplate<Control>
- #include <Control.h>
-
class List : public LinkedObjectListTemplate<Control>
-
class Screen
- #include <Screen.h>
Public Types
-
using DrawMethod = Delegate<bool(SceneObject &scene)>
Invoked when screen is drawn.
- Param scene:
Where to compose screen
- Retval bool:
Return true to continue default processing
-
using ControlMethod = Delegate<bool(ControlEvent event, Control &control)>
Invoked in response to user input.
- Param event:
- Param control:
- Retval bool:
Return true to continue default processing
-
using DrawMethod = Delegate<bool(SceneObject &scene)>
-
class TimeClock : public Graphics::Control
- #include <TimeClock.h>
-
struct HMS
- #include <TimeClock.h>
-
struct HMS
-
class Device
- #include <Device.h>
A physical display device.
Subclassed by Graphics::AbstractDisplay
Public Functions
-
virtual bool setOrientation(Orientation orientation) = 0
Set display orientation.
-
virtual Size getNativeSize() const = 0
Get physical size of display.
- Return values:
Size – Dimensions for NORMAL orientation
-
inline Orientation getOrientation()
Get current display orientation.
-
virtual bool setScrollMargins(uint16_t top, uint16_t bottom) = 0
Set margins for hardware scrolling.
Area between top/bottom can be scrolled using
scroll()method.- Parameters:
top – Number of fixed pixels at top of screen
bottom – Number of fixed pixels at bottom of screen
-
virtual bool scroll(int16_t y) = 0
Scroll region of display up or down using hardware scrolling.
- Parameters:
y – Number of lines to scroll. Positive values scroll content down, negative values scroll up.
-
virtual bool setOrientation(Orientation orientation) = 0
-
struct FillInfo
- #include <DisplayList.h>
Supports DisplayList blend operations.
See also
-
class DisplayList
- #include <DisplayList.h>
Stores list of low-level display commands.
Used by hardware surfaces to efficiently buffer commands which are then executed in interrupt context.
The ILI9341 4-wire SPI mode is awkard to use so to allow more efficient access the command and data information is buffered as various types of ‘chunk’ using this class. A single HSPI request packet is used for all requests and is re-filled in interrupt context from this list.
Count values (data length) are stored as either 1 or 2 bytes. Values less than 0x80 bytes are stored in 1 byte, values from 0x0080 to 0x7fff are stored MSB first with the top bit set, followed by the LSB. For example, 0x1234 would be stored as 0x92 0x34.
Commands are stored in 1 byte. To allow use of this class with other displays may require adjusting this, perhaps converting this into a class template to accommodate these differences more efficiently.
Standard chunk. Contains a display-specific command byte followed by optional data.
command (1 byte)
data length (1 or 2 bytes)
data (variable length)
Data chunk (no command).
COMMAND_DATA (1 byte)
data length (1 or 2 bytes)
data (variable length)
Repeated data chunk (no command). Used to perform colour fills.
COMMAND_DATA_REPEAT (1 byte)
data length (1 or 2 bytes)
repeat count (1 or 2 bytes)
data (variable length)
Read command. Defines a single read command packet. Reading is particularly awkward on the ILI9341 as it ‘forgets’ the read position after each read packet. Reading a block of display memory therefore requires the address window to be set immediately before the RAMRD instruction. The final block in the sequence is a CALLBACK command.
COMMAND_READ (1 byte)
data length (1 or 2 bytes)
command (1 byte) The display-specific command to issue
buffer address (4 bytes)
Callback. Invokes a callback function (see Callback type). Note that the single parameter points to a copy of the original data which is stored in the list.
COMMAND_CALLBACK (1 byte)
parameter data length (1 or 2 bytes)
callback function pointer (4 bytes)
callback parameters (variable)
Subclassed by Graphics::SpiDisplayList
Public Types
-
enum CodeArgLengths
Obtain maximum size for command, not including variable data which may be added.
Used to check space before starting a command sequence. Actual space used may be less than this.
Values:
-
enumerator XX
-
enumerator XX
-
using Callback = void (*)(void *parameterData)
Queued callback.
The parameter data is copied so the caller does not need to allocate memory dynamically. The data must start on a 32-bit word boundary.
- Param parameterData:
A copy of the original parameter data
Public Functions
-
inline DisplayList(AddressWindow &addrWindow, size_t bufferSize)
- Parameters:
commands – Codes corresponding to specific display commands
addrWindow – Reference to current device address window, synced between lists
bufferSize –
-
inline DisplayList(AddressWindow &addrWindow, const FSTR::ObjectBase &data)
Create pre-defined display list from flash data.
Used for initialisation data
-
inline DisplayList(AddressWindow &addrWindow, const void *data, size_t length)
Create initialised display list from RAM data.
Used for initialisation data
-
void reset()
Reset the display list ready for re-use List MUST NOT be in use!
-
inline bool isEmpty() const
Determine if any commands have been stored for execution.
-
inline uint16_t freeSpace() const
Get number of bytes remaining in buffer.
-
inline uint16_t readOffset() const
Get current read position.
-
inline uint16_t used() const
Get number of bytes stored in buffer.
-
inline const uint8_t *getContent() const
Get read-only pointer to start of buffer.
-
uint8_t *getBuffer(uint16_t &available)
Get some space in the list to write pixel data.
Call
commitafter the data has been written- Parameters:
available – (OUT) How much space is available
- Return values:
uint8_t* – Where to write data
-
inline uint8_t *getBuffer(uint16_t minBytes, uint16_t &available)
Get some space in the list to write pixel data.
Call
commitafter the data has been written- Parameters:
minBytes – Minimum bytes required
available – (OUT) How much space is available
- Return values:
uint8_t* – Where to write data, nullptr if required space not available
-
void commit(uint16_t length)
Commit block of data to the list.
Note
MUST NOT call with more data than returned from
availablein getBuffer call.- Parameters:
length – How many bytes have been written
-
inline bool writeCommand(uint8_t command, uint32_t data, uint8_t length)
Write command with 1-4 bytes of parameter data.
-
bool writeCommand(uint8_t command, const void *data, uint16_t length)
Write command with variable amount of parameter data.
-
bool writeData(const void *data, uint16_t length)
Add WRITE command plus data.
Add WRITE command plus external data.
- Parameters:
data – Will be locked until display list has been executed
-
bool blockFill(const void *data, uint16_t length, uint32_t repeat)
Perform a block fill operation with repeat, e.g. multiple pixel fill or repeated pattern.
-
bool setPixel(PackedColor color, uint8_t bytesPerPixel, Point pt)
Set a single pixel.
-
bool readMem(void *buffer, uint16_t length)
Read a block of display memory.
-
bool writeCallback(Callback callback, void *params, uint16_t paramLength)
Request a callback.
The callback will be invoked (in interrupt context) when read from the display list
- Parameters:
callback – Callback function to invoke
params – Parameter data, will be stored in list
paramLength – Size of parameter data
-
bool fill(const Rect &rect, PackedColor color, uint8_t bytesPerPixel, FillInfo::Callback callback)
Perform a block fill operation with blending.
Performs a read/blend/write operation. Use to perform transparent fills or other blend operations on small regions of display memory.
- Parameters:
rect – Area to fill
color –
bytesPerPixel –
callback – Invoked in interrupt context to perform blend operation
-
inline bool canLockBuffer()
Enforce maximum number of locked buffers to conserve memory.
- Return values:
bool – true if call to
lockBuffer()will succeed
Lock a shared buffer by storing a reference to it. This will be released when
reset()is called.
-
inline bool require(uint16_t length)
Check if list has space for the given number of bytes.
- Return values:
bool – true if there’s room
-
bool readEntry(Entry &info)
Read next command list entry.
Used by virtual display and for debugging.
SpiDisplayListuses a different mechanism for reading.- Parameters:
info –
- Return values:
bool – false if there are no more commands
-
struct Entry
- #include <DisplayList.h>
Values returned from
readEntry
-
union Header
- #include <DisplayList.h>
Each list entry starts with a header.
This information is interpreted by the display driver but should translate to a single display command.
Each command has fixed content with optional variable data. Length can be packed with command (0-31), or an additional uint8_t or uint16_t.
setPixel(uint8_t x, uint8_t y, void* data, uint8_t len) COL(uint8_t x, uint8_t len) ROW(uint8_t y, 1) DAT data, len
setPixel(uint16_t x, uint8_t y, void* data, uint16_t len) COL(uint16_t x, uint16_t len) ROW(uint8_t y, 1) DAT data, len
fillRect(uint16_t x, uint16_t y, uint16_t w, uint16_t h, data len) COL16 x, w ROW16 y, h
Public Members
-
Code code
-
uint8_t len
-
struct Graphics::DisplayList::Header::[anonymous] [anonymous]
-
uint8_t u8
Public Static Attributes
-
static constexpr uint8_t lenMax = {15}
-
Code code
-
class ImageSurface : public Graphics::Surface
- #include <ImageSurface.h>
Virtual class to access an image as a Surface.
Use to create off-screen bitmaps by drawing or copying areas from display memory.
Subclassed by Graphics::FileImageSurface, Graphics::MemoryImageSurface
Public Functions
-
virtual int readDataBuffer(ReadBuffer &buffer, ReadStatus *status, ReadCallback callback, void *param) override
Read some pixels.
Call
setAddrWindowto set up region to be read. Returns true when all pixels have been queued for reading.- Parameters:
buffer – Details requested format and buffer to read
status – Optional. Stores result of read operation.
callback – Optional. Invoked when read has completed
param – Parameters passed to callback
- Return values:
int – Number of pixels queued for reading (or read); 0 if no further pixels to read, < 0 to try again later
-
inline virtual void reset() override
Reset surface ready for more commands.
-
virtual bool present(PresentCallback callback, void *param) override
Present surface to display device.
Hardware devices will queue buffered commands to the display device then return. The surface will be marked as BUSY. Attempting to call present() on a BUSY surface must return true. If surface is EMPTY (no buffered commands), must return false.
- Parameters:
callback – Invoked when surface is available for more commands
param – Passed to callback
- Return values:
bool – true If callback has been queued, false if surface is empty
-
virtual int readDataBuffer(ReadBuffer &buffer, ReadStatus *status, ReadCallback callback, void *param) override
-
class MemoryImageSurface : public Graphics::ImageSurface
- #include <ImageSurface.h>
Image surface using RAM as backing store.
Useful for sprites, etc.
-
class FileImageSurface : public Graphics::ImageSurface
- #include <ImageSurface.h>
Image surface using filing system as backing store.
Slower than RAM but size is unrestricted. Use for constructing complex scenes for faster drawing.
-
class LcdGlyph : public Graphics::GlyphObject
- #include <LcdFont.h>
Public Functions
-
inline virtual bool init() override
Initialise the object, e.g. parse header content and obtain dimensions.
-
virtual void readAlpha(void *buffer, Point origin, size_t stride) const override
Obtain glyph information as block of 8-bit alpha values.
This method is called with a positive origin to accommodate negative x/y glyph offsets. Italic and script typefaces do this a lot!
- Parameters:
buffer –
origin – Location of cursor within buffer
stride – Number of bytes per row in buffer
-
inline virtual bool init() override
-
class LcdTypeFace : public Graphics::TypeFace
- #include <LcdFont.h>
Public Functions
-
inline virtual FontStyles getStyle() const override
Style of this typeface (bold, italic, etc.)
-
inline virtual uint8_t height() const override
Get height of typeface, same for all characters.
-
inline virtual uint8_t descent() const override
How many pixels from bottom of em-square to baseline
-
inline virtual GlyphObject::Metrics getMetrics(char ch) const override
Get metrics for a character.
-
virtual std::unique_ptr<GlyphObject> getGlyph(char ch, const GlyphObject::Options &options) const override
Get the glyph for a character.
Caller is responsible for destroying the glyph when no longer required.
- Parameters:
ch –
options – Options to control how the glyph is drawn (colour, shading, etc)
- Return values:
GlyphObject* – The glyph, nullptr if no glyph exists in the typeface for this character
-
inline virtual FontStyles getStyle() const override
-
class Meta
- #include <Meta.h>
Empty base class to support object enumeration Non-virtual to avoid bloat.
Subclassed by Graphics::Asset, Graphics::Brush, Graphics::Object, Graphics::TextObject::Element, Graphics::TextOptions
-
class MetaWriter
- #include <Meta.h>
Writes object content in readable format for debugging.
-
class MipiDisplay : public Graphics::SpiDisplay
- #include <MipiDisplay.h>
Subclassed by Graphics::Display::ILI9341, Graphics::Display::ST7789V
Public Functions
-
inline void setNativeSize(Size screenSize)
Sets the screen size. Must be called before calling begin()
-
inline virtual Size getNativeSize() const override
Get physical size of display.
- Return values:
Size – Dimensions for NORMAL orientation
-
virtual bool setScrollMargins(uint16_t top, uint16_t bottom) override
Set margins for hardware scrolling.
Area between top/bottom can be scrolled using
scroll()method.- Parameters:
top – Number of fixed pixels at top of screen
bottom – Number of fixed pixels at bottom of screen
-
virtual bool scroll(int16_t y) override
Scroll region of display up or down using hardware scrolling.
- Parameters:
y – Number of lines to scroll. Positive values scroll content down, negative values scroll up.
-
virtual bool setOrientation(Orientation orientation) override
Set display orientation.
-
inline virtual PixelFormat getPixelFormat() const override
All surfaces support the same pixel format.
-
inline void setNativeSize(Size screenSize)
-
class MipiSurface : public Graphics::Surface
- #include <MipiDisplay.h>
Public Functions
-
inline virtual void reset() override
Reset surface ready for more commands.
-
virtual int readDataBuffer(ReadBuffer &buffer, ReadStatus *status, ReadCallback callback, void *param) override
Read some pixels.
Call
setAddrWindowto set up region to be read. Returns true when all pixels have been queued for reading.- Parameters:
buffer – Details requested format and buffer to read
status – Optional. Stores result of read operation.
callback – Optional. Invoked when read has completed
param – Parameters passed to callback
- Return values:
int – Number of pixels queued for reading (or read); 0 if no further pixels to read, < 0 to try again later
-
virtual bool render(const Object &object, const Rect &location, std::unique_ptr<Renderer> &renderer) override
Start rendering an object.
Surfaces may override this method to implement alternative rendering using specific hardware features of the display device.
Software renderers should be run by calling
surface::execute.- Parameters:
object – What to render
location – Placement information
renderer – If operation cannot be completed in hardware, create a renderer instance to manage the process
- Return values:
Return – true on success, false to retry later
-
virtual bool present(PresentCallback callback, void *param) override
Present surface to display device.
Hardware devices will queue buffered commands to the display device then return. The surface will be marked as BUSY. Attempting to call present() on a BUSY surface must return true. If surface is EMPTY (no buffered commands), must return false.
- Parameters:
callback – Invoked when surface is available for more commands
param – Passed to callback
- Return values:
bool – true If callback has been queued, false if surface is empty
-
inline virtual void reset() override
-
class Renderer : public LinkedObjectTemplate<Renderer>
- #include <Object.h>
Virtual base class to manage rendering of various types of information to a surface.
Subclassed by Graphics::BlendRenderer, Graphics::CircleRenderer, Graphics::CopyRenderer, Graphics::EllipseRenderer, Graphics::FilledCircleRenderer, Graphics::FilledEllipseRenderer, Graphics::FilledRectRenderer, Graphics::FilledRoundedRectRenderer, Graphics::GfxLineRenderer, Graphics::ImageRenderer, Graphics::LineRenderer, Graphics::MultiRenderer, Graphics::PolylineRenderer, Graphics::RectRenderer, Graphics::RoundedRectRenderer, Graphics::ScrollRenderer, Graphics::SurfaceRenderer, Graphics::TextRenderer
-
class Object : public LinkedObjectTemplate<Object>, public Graphics::Meta
- #include <Object.h>
A drawable object inherits from this virtual base class.
Subclassed by Graphics::ObjectTemplate< Object::Kind::FilledCircle >, Graphics::ObjectTemplate< Object::Kind::Reference >, Graphics::ObjectTemplate< Object::Kind::Line >, Graphics::ObjectTemplate< Object::Kind::Rect >, Graphics::ObjectTemplate< Object::Kind::FilledEllipse >, Graphics::ObjectTemplate< Object::Kind::Ellipse >, Graphics::ObjectTemplate< Object::Kind::FilledArc >, Graphics::ObjectTemplate< Object::Kind::Arc >, Graphics::ObjectTemplate< Object::Kind::Surface >, Graphics::ObjectTemplate< Object::Kind::FilledRect >, Graphics::ObjectTemplate< Object::Kind::Drawing >, Graphics::ObjectTemplate< Object::Kind::Polyline >, Graphics::ObjectTemplate< Object::Kind::Point >, Graphics::ObjectTemplate< Object::Kind::Copy >, Graphics::ObjectTemplate< Object::Kind::Circle >, Graphics::ObjectTemplate< Object::Kind::Text >, Graphics::ObjectTemplate< Object::Kind::Image >, Graphics::ObjectTemplate< Object::Kind::Scroll >, Graphics::ObjectTemplate< object_kind >, Graphics::SceneObject
-
template<Object::Kind object_kind>
class ObjectTemplate : public Graphics::Object - #include <Object.h>
Subclassed by Graphics::Control
-
class ReferenceObject : public Graphics::ObjectTemplate<Object::Kind::Reference>
- #include <Object.h>
Reference to another object.
Objects are owned by a Scene, so this allows objects to be re-used in multiple scenes, or to other classes which expose an Object interface.
-
class PointObject : public Graphics::ObjectTemplate<Object::Kind::Point>
- #include <Object.h>
A single pixel == 1x1 rectangle.
-
class RectObject : public Graphics::ObjectTemplate<Object::Kind::Rect>
- #include <Object.h>
A rectangular outline.
-
class FilledRectObject : public Graphics::ObjectTemplate<Object::Kind::FilledRect>
- #include <Object.h>
A filled rectangle.
-
class LineObject : public Graphics::ObjectTemplate<Object::Kind::Line>
- #include <Object.h>
A drawn line.
-
class PolylineObject : public Graphics::ObjectTemplate<Object::Kind::Polyline>
- #include <Object.h>
A sequence of lines.
By default, lines are connected. This is used to draw rectangles, triangles or any other type of polygon. A line is drawn between points 0-1, 1-2, 2-3, etc.
Setting the
connectedproperty to false allows the lines to be discontinuous, so a line is drawn between points 0-1, 2-3, 3-4, etc.
-
class CircleObject : public Graphics::ObjectTemplate<Object::Kind::Circle>
- #include <Object.h>
A circle outline.
Public Functions
-
class FilledCircleObject : public Graphics::ObjectTemplate<Object::Kind::FilledCircle>
- #include <Object.h>
A filled circle.
Public Functions
-
class EllipseObject : public Graphics::ObjectTemplate<Object::Kind::Ellipse>
- #include <Object.h>
An ellipse outline.
-
class FilledEllipseObject : public Graphics::ObjectTemplate<Object::Kind::FilledEllipse>
- #include <Object.h>
A filled ellipse.
-
class ArcObject : public Graphics::ObjectTemplate<Object::Kind::Arc>
- #include <Object.h>
An arc outline.
-
class FilledArcObject : public Graphics::ObjectTemplate<Object::Kind::FilledArc>
- #include <Object.h>
A filled arc.
-
class ImageObject : public Graphics::ObjectTemplate<Object::Kind::Image>
- #include <Object.h>
Virtual base class for an image.
Subclassed by Graphics::GlyphObject, Graphics::StreamImageObject
Public Functions
-
virtual Renderer *createRenderer(const Location &location) const override
Create a software renderer for this object.
Return nullptr if object cannot/should not be rendered
- Parameters:
location –
- Return values:
renderer – Returned renderer object
-
virtual bool init() = 0
Initialise the object, e.g. parse header content and obtain dimensions.
-
virtual PixelFormat getPixelFormat() const = 0
Get native pixel format.
- Return values:
PixelFormat – Return None if ambivalent about format (e.g. calculated pixel data)
-
virtual size_t readPixels(const Location &loc, PixelFormat format, void *buffer, uint16_t width) const = 0
Read pixels in requested format.
- Parameters:
loc – Start position
format – Required pixel format
buffer – Buffer for pixels
width – Number of pixels to read
- Return values:
size_t – Number of bytes written
-
virtual Renderer *createRenderer(const Location &location) const override
-
class StreamImageObject : public Graphics::ImageObject
- #include <Object.h>
Image whose contents are stored in a stream, typically in a file or flash memory.
Subclassed by Graphics::BitmapObject, Graphics::RawImageObject
-
class BitmapObject : public Graphics::StreamImageObject
- #include <Object.h>
A BMP format image.
Code based on https://github.com/adafruit/Adafruit-GFX-Library
Public Functions
-
virtual bool init() override
Initialise the object, e.g. parse header content and obtain dimensions.
-
inline virtual PixelFormat getPixelFormat() const override
Get native pixel format.
- Return values:
PixelFormat – Return None if ambivalent about format (e.g. calculated pixel data)
-
virtual size_t readPixels(const Location &loc, PixelFormat format, void *buffer, uint16_t width) const override
Read pixels in requested format.
- Parameters:
loc – Start position
format – Required pixel format
buffer – Buffer for pixels
width – Number of pixels to read
- Return values:
size_t – Number of bytes written
-
virtual bool init() override
-
class RawImageObject : public Graphics::StreamImageObject
- #include <Object.h>
Image stored as raw pixels in a specific format.
Use images stored in native display format for best performance as no conversion is required.
Subclassed by Graphics::FileImageObject, Graphics::MemoryImageObject
Public Functions
-
inline virtual bool init() override
Initialise the object, e.g. parse header content and obtain dimensions.
-
inline virtual PixelFormat getPixelFormat() const override
Get native pixel format.
- Return values:
PixelFormat – Return None if ambivalent about format (e.g. calculated pixel data)
-
virtual size_t readPixels(const Location &loc, PixelFormat format, void *buffer, uint16_t width) const override
Read pixels in requested format.
- Parameters:
loc – Start position
format – Required pixel format
buffer – Buffer for pixels
width – Number of pixels to read
- Return values:
size_t – Number of bytes written
-
inline virtual bool init() override
-
class RenderTarget
- #include <Object.h>
Interface for objects which support writing via surfaces.
Subclassed by Graphics::AbstractDisplay, Graphics::FileImageObject, Graphics::MemoryImageObject
Public Functions
-
virtual PixelFormat getPixelFormat() const = 0
All surfaces support the same pixel format.
-
virtual PixelFormat getPixelFormat() const = 0
-
class MemoryImageObject : public Graphics::RawImageObject, public Graphics::RenderTarget
- #include <Object.h>
Public Functions
-
inline virtual PixelFormat getPixelFormat() const override
Get native pixel format.
- Return values:
PixelFormat – Return None if ambivalent about format (e.g. calculated pixel data)
-
inline virtual Surface *createSurface(size_t bufferSize = 0) override
Create a surface for use with this render target.
Caller is responsible for destroying the surface when no longer required.
- Parameters:
bufferSize – Size of internal command/data buffer
- Return values:
Surface* – The surface to use
-
inline virtual PixelFormat getPixelFormat() const override
-
class FileImageObject : public Graphics::RawImageObject, public Graphics::RenderTarget
- #include <Object.h>
Public Functions
-
inline virtual PixelFormat getPixelFormat() const override
Get native pixel format.
- Return values:
PixelFormat – Return None if ambivalent about format (e.g. calculated pixel data)
-
inline virtual PixelFormat getPixelFormat() const override
-
class GlyphObject : public Graphics::ImageObject
- #include <Object.h>
A character glyph image.
Characters are accessed like regular images but there are some specialisations which may be necessary which this class exposes:
To support devices with custom renderers we require access to the raw monochrome bits
Images currently don’t support transparency, although this could be enabled using a flag bits in the PackedColor format. The display driver or renderer would then need to interpret this flag and render accordingly. An alternative method is to nominate a transparent colour.
Subclassed by Graphics::LcdGlyph
Public Functions
-
inline virtual PixelFormat getPixelFormat() const override
Get native pixel format.
- Return values:
PixelFormat – Return None if ambivalent about format (e.g. calculated pixel data)
-
virtual size_t readPixels(const Location &loc, PixelFormat format, void *buffer, uint16_t width) const override
Read pixels in requested format.
- Parameters:
loc – Start position
format – Required pixel format
buffer – Buffer for pixels
width – Number of pixels to read
- Return values:
size_t – Number of bytes written
-
virtual void readAlpha(void *buffer, Point origin, size_t stride) const = 0
Obtain glyph information as block of 8-bit alpha values.
This method is called with a positive origin to accommodate negative x/y glyph offsets. Italic and script typefaces do this a lot!
- Parameters:
buffer –
origin – Location of cursor within buffer
stride – Number of bytes per row in buffer
-
class TextObject : public Graphics::ObjectTemplate<Object::Kind::Text>
- #include <Object.h>
A block of text consisting of zero or more segments.
A segment is a straight run of text using a specific typeface and style.
Public Functions
-
class ColorElement : public Graphics::TextObject::Element
- #include <Object.h>
-
class Element : public LinkedObjectTemplate<Element>, public Graphics::Meta
- #include <Object.h>
Subclassed by Graphics::TextObject::ColorElement, Graphics::TextObject::FontElement, Graphics::TextObject::RunElement, Graphics::TextObject::TextElement
-
class FontElement : public Graphics::TextObject::Element
- #include <Object.h>
-
class RunElement : public Graphics::TextObject::Element
- #include <Object.h>
-
class TextElement : public Graphics::TextObject::Element
- #include <Object.h>
-
class ColorElement : public Graphics::TextObject::Element
-
class SurfaceObject : public Graphics::ObjectTemplate<Object::Kind::Surface>
- #include <Object.h>
Describes a target surface and corresponding source location.
Used in surface copy operations to read display memory contents.
-
class CopyObject : public Graphics::ObjectTemplate<Object::Kind::Copy>
- #include <Object.h>
Describes a copy operation within the same surface.
-
class ScrollObject : public Graphics::ObjectTemplate<Object::Kind::Scroll>
- #include <Object.h>
Describes a scrolling operation.
-
class DrawingObject : public Graphics::ObjectTemplate<Object::Kind::Drawing>
- #include <Object.h>
A collection of line and curve drawing operations.
Stored in a compact format which can be streamed.
-
template<typename T>
class ItemList - #include <Renderer.h>
Fixed list of types.
Rendering algorithms create small sets of points, lines or rectangles. Buffering these in a small list simplifies logic considerably.
-
class PointList : public Graphics::ItemList<Point>
- #include <Renderer.h>
Small list of points for drawing.
Algorithms generate multiple points within a loop so buffering them in a list simplifies logic considerably.
-
class RectList : public Graphics::ItemList<Rect>
- #include <Renderer.h>
Small list of rectangles, similar to PointList.
Subclassed by Graphics::ArcRectList
-
class MultiRenderer : public Graphics::Renderer
- #include <Renderer.h>
Base class to render multiple objects.
Subclassed by Graphics::Drawing::Renderer, Graphics::RenderQueue, Graphics::SceneRenderer
-
class SceneRenderer : public Graphics::MultiRenderer
- #include <Renderer.h>
A scene is a list of other objects, so we just iterate through the list and draw each in turn.
Rendering is performed by calling
Surface::render(). Surfaces are provided by devices so may be able to provide optimised renderers for their hardware.
-
class GfxLineRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws 1-pixel lines.
Code based on https://github.com/adafruit/Adafruit-GFX-Library
-
class PolylineRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws series of lines defined by a
PolylineObject
-
class RectRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws a rectangle as a polyline.
-
class FilledRectRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws a filled rectangle.
To accommodate transparency, etc.
-
class RoundedRectRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws a rectangle outline with rounded corners.
Code based on https://github.com/adafruit/Adafruit-GFX-Library
-
class FilledRoundedRectRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws a filled rectangle with rounded corners.
Code based on https://github.com/adafruit/Adafruit-GFX-Library
-
class CircleRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws a circle outline.
Code based on https://github.com/adafruit/Adafruit-GFX-Library
-
class FilledCircleRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws a filled circle.
Code based on https://github.com/adafruit/Adafruit-GFX-Library
-
struct Ellipse
- #include <Renderer.h>
State information for tracing an ellipse outline.
-
class EllipseRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws an ellipse outline.
Uses McIlroy’s Ellipse Algorithm
See http://enchantia.com/graphapp/doc/tech/ellipses.html
Subclassed by Graphics::ArcRenderer
-
class FilledEllipseRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draws a filled ellipse.
See http://enchantia.com/graphapp/doc/tech/ellipses.html
Subclassed by Graphics::FilledArcRenderer
-
class ArcRenderer : public Graphics::EllipseRenderer
- #include <Renderer.h>
Render arc outline with adjustable line width.
-
class FilledArcRenderer : public Graphics::FilledEllipseRenderer
- #include <Renderer.h>
Render arc outline with adjustable line width.
-
class SurfaceRenderer : public Graphics::Renderer
- #include <Renderer.h>
Copy an area to another surface.
Typically used to copy display memory into RAM
-
class CopyRenderer : public Graphics::Renderer
- #include <Renderer.h>
Copy an area within the same surface.
Subclassed by Graphics::ImageCopyRenderer
-
class ImageCopyRenderer : public Graphics::CopyRenderer
- #include <Renderer.h>
-
class TextRenderer : public Graphics::Renderer
- #include <Renderer.h>
Draw a line of text.
If foreground and background colours are the same then the text is renderered transparently.
-
class RenderQueue : private Graphics::MultiRenderer
- #include <RenderQueue.h>
Top-level manager to queue objects for rendering to a specific target.
Use this to render single objects, typically Scenes or Drawings
Public Functions
-
inline RenderQueue(RenderTarget &target, uint8_t surfaceCount = 2, size_t bufferSize = 0)
Constructor.
Surfaces are created by the target display device.
For minimum RAM usage use a single surface.
For best performance use two, so one can be prepared whilst the other is being written to the screen.
The RenderQueue owns these surfaces.
- Parameters:
target – Where to render scenes
bufferSize – Size of each allocated surface buffer. Specify 0 to use default.
surfaceCount – Number of surfaces to allocate
-
template<typename T>
inline void render(T *object, const Location &location, typename T::Callback callback = nullptr, uint16_t delayMs = 0) Add object to the render queue and start rendering if it isn’t already.
- Parameters:
object – Scene, Drawing, etc. to render
location – Where to draw the object
callback – Optional callback to invoke when render is complete
delayMs – Delay between render completion and callback
-
inline RenderQueue(RenderTarget &target, uint8_t surfaceCount = 2, size_t bufferSize = 0)
-
class SceneObject : public Graphics::Object
- #include <Scene.h>
A Scene containing multiple objects.
Public Functions
-
virtual Renderer *createRenderer(const Location &location) const override
Create a software renderer for this object.
Return nullptr if object cannot/should not be rendered
- Parameters:
location –
- Return values:
renderer – Returned renderer object
-
template<typename T>
inline T *addObject(T *obj) Add a new object to the scene.
Use this method to add custom objects. To draw an object multiple times use
drawObjectwhich will add a reference instead.- Parameters:
obj – This will be owned by the scene
-
inline void clear(const Brush &brush = Color::Black)
Clear the scene and fill with a chosen colour.
-
inline CopyObject *copy(const Rect &source, Point dest)
Copy region of display to another.
- Parameters:
source – Area to copy
dest – Top-left corner to copy to
-
inline ScrollObject *scroll(const Rect &area, int16_t cx, int16_t cy, bool wrapx = false, bool wrapy = false, Color fill = Color::None)
Scroll display memory.
- Parameters:
area – Region to scroll
cx – Distance to scroll horizontally
cy – Distance to scroll vertically
wrapx – true to scroll, false to clip in X direction
wrapx – Y scroll/clip
fill – Optional color to fill in clip mode
-
virtual Renderer *createRenderer(const Location &location) const override
-
class SpiDisplay : protected HSPI::Device, public Graphics::AbstractDisplay
- #include <SpiDisplay.h>
Subclassed by Graphics::MipiDisplay
Public Functions
-
inline virtual HSPI::IoModes getSupportedIoModes() const override
Return set of IO modes supported by a device implementation.
-
inline virtual HSPI::IoModes getSupportedIoModes() const override
-
class SpiDisplayList : public Graphics::DisplayList
- #include <SpiDisplayList.h>
Display list for hardware SPI devices.
A single HSPI request packet is used for all requests and is re-filled in interrupt context from this list.
Public Functions
-
bool fillRequest()
Called from interrupt context to re-fill the SPI request packet.
- Return values:
bool – true on success, false if there are no further requests
-
struct Commands
- #include <SpiDisplayList.h>
Commonly-used display-specific command codes.
Short codes are used to represent these commands. Other commands are stored directly in the list.
-
bool fillRequest()
-
class WriteStream
- #include <Stream.h>
-
class ReadStream
- #include <Stream.h>
-
class SubStream : public IDataSourceStream
- #include <Stream.h>
Public Functions
-
inline virtual int available() override
Return the total length of the stream.
- Return values:
int – -1 is returned when the size cannot be determined
-
inline virtual uint16_t readMemoryBlock(char *data, int bufSize) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
-
inline virtual int seekFrom(int offset, SeekOrigin origin) override
Change position in stream.
Note
This method is implemented by streams which support random seeking, such as files and memory streams.
- Parameters:
offset –
origin –
- Return values:
New – position, < 0 on error
-
inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
-
inline virtual int available() override
-
class Surface : public Graphics::AssetTemplate<AssetType::Surface>
- #include <Surface.h>
Interface for a drawing surface.
Represents a rectangular area of pixels which can be read or written.
A display device has at least one of these, representing the primary display area. More complex devices with large amounts of display memory may allow additional surfaces to be used to perform screen updates by ‘flipping’ (switching active surface) or fast copies using display hardware.
Subclassed by Graphics::ImageSurface, Graphics::MipiSurface
Public Types
-
using ReadCallback = void (*)(ReadBuffer &data, size_t length, void *param)
Callback for readPixel() operations.
- Param buffer:
Buffer passed to readPixel() call
Public Functions
-
virtual int readDataBuffer(ReadBuffer &buffer, ReadStatus *status = nullptr, ReadCallback callback = nullptr, void *param = nullptr) = 0
Read some pixels.
Call
setAddrWindowto set up region to be read. Returns true when all pixels have been queued for reading.- Parameters:
buffer – Details requested format and buffer to read
status – Optional. Stores result of read operation.
callback – Optional. Invoked when read has completed
param – Parameters passed to callback
- Return values:
int – Number of pixels queued for reading (or read); 0 if no further pixels to read, < 0 to try again later
-
virtual bool render(const Object &object, const Rect &location, std::unique_ptr<Renderer> &renderer)
Start rendering an object.
Surfaces may override this method to implement alternative rendering using specific hardware features of the display device.
Software renderers should be run by calling
surface::execute.- Parameters:
object – What to render
location – Placement information
renderer – If operation cannot be completed in hardware, create a renderer instance to manage the process
- Return values:
Return – true on success, false to retry later
-
bool render(const Object &object, const Rect &location)
Render an object in one cycle.
Use this method for simple renders which should complete in one cycle. Typically used for memory surface renders or where an object is expected to render within a single surface.
- Parameters:
surface – Where to render the object to
location –
- Return values:
bool – true on success
-
inline bool execute(std::unique_ptr<Renderer> &renderer)
Execute a renderer.
- Parameters:
renderer – Will be released when render has completed
- Return values:
bool – true if render is complete
-
virtual void reset() = 0
Reset surface ready for more commands.
-
virtual bool present(PresentCallback callback = nullptr, void *param = nullptr) = 0
Present surface to display device.
Hardware devices will queue buffered commands to the display device then return. The surface will be marked as BUSY. Attempting to call present() on a BUSY surface must return true. If surface is EMPTY (no buffered commands), must return false.
- Parameters:
callback – Invoked when surface is available for more commands
param – Passed to callback
- Return values:
bool – true If callback has been queued, false if surface is empty
-
bool fillSmallRect(const Brush &brush, const Rect &location, const Rect &rect)
Fill a small rectangle using a non-transparent brush.
-
bool drawHLine(PackedColor color, uint16_t x0, uint16_t x1, uint16_t y, uint16_t w)
Draw a simple horizontal line using a filled rectangle.
-
bool drawVLine(PackedColor color, uint16_t x, uint16_t y0, uint16_t y1, uint16_t w)
Draw a simple vertical line using a filled rectangle.
-
struct Stat
- #include <Surface.h>
-
using ReadCallback = void (*)(ReadBuffer &data, size_t length, void *param)
-
class TextParser
- #include <TextBuilder.h>
Simplifies construction of TextObject instances.
Subclassed by Graphics::TextBuilder
-
class TextBuilder : public Graphics::TextParser, public Print
- #include <TextBuilder.h>
Simplifies construction of TextObject instances.
Public Functions
-
inline virtual size_t write(uint8_t c) override
Writes a single character to output stream.
- Parameters:
c – Character to write to output stream
- Return values:
size_t – Quantity of characters written to output stream
-
inline virtual size_t write(const uint8_t *buffer, size_t size) override
Writes characters from a buffer to output stream.
- Parameters:
buffer – Pointer to character buffer
size – Quantity of characters to write
- Return values:
size_t – Quantity of characters written to stream
-
size_t write(uint8_t c) = 0
Writes a single character to output stream.
- Parameters:
c – Character to write to output stream
- Return values:
size_t – Quantity of characters written to output stream
-
inline size_t write(const char *str)
Writes a c-string to output stream.
- Parameters:
str – Pointer to c-string
- Return values:
size_t – Quantity of characters written to stream
-
size_t write(const uint8_t *buffer, size_t size)
Writes characters from a buffer to output stream.
- Parameters:
buffer – Pointer to character buffer
size – Quantity of characters to write
- Return values:
size_t – Quantity of characters written to stream
-
inline size_t write(const char *buffer, size_t size)
Writes characters from a buffer to output stream.
- Parameters:
buffer – Pointer to character buffer
size – Quantity of characters to write
- Return values:
size_t – Quantity of characters written to stream
-
inline virtual size_t write(uint8_t c) override
-
class Touch
- #include <Touch.h>
Subclassed by Graphics::VirtualTouch, Graphics::XPT2046
Public Types
-
using Callback = Delegate<void()>
Callback function.
Public Functions
-
virtual bool setOrientation(Orientation orientation)
Set display orientation.
-
virtual Size getNativeSize() const = 0
Get physical size of display.
- Return values:
Size – Dimensions for NORMAL orientation
-
inline virtual void setCallback(Callback callback)
Register callback to be invoked when state changes.
-
inline Orientation getOrientation()
Get current display orientation.
-
struct Calibration
- #include <Touch.h>
-
struct State
- #include <Touch.h>
-
using Callback = Delegate<void()>
-
template<typename T>
struct TPoint - #include <Types.h>
An (x, y) display coordinate.
-
class Region
- #include <Types.h>
Represents the intersection of two rectangles.
This produces up to 4 separate, non-overlapping rectangles.
-
struct Location
- #include <Types.h>
Identifies position within bounding rectangle.
Public Members
-
Rect source
Reference source area.
This is generally used by brushes to locate colour information. For example, with an ImageBrush objects can be filled to either re-use the same portion of the reference image, or to reveal a particular part of it.
-
Rect source
-
class ColorRange
- #include <Types.h>
-
class Scale
- #include <Types.h>
-
struct GlyphMetrics
- #include <Types.h>
Glyph metrics.
-
namespace Display
-
class Virtual : public Graphics::AbstractDisplay
- #include <Virtual.h>
Virtual display device for Host.
Talks to python virtual screen application via TCP
Public Functions
-
inline virtual Size getNativeSize() const override
Get physical size of display.
- Return values:
Size – Dimensions for NORMAL orientation
-
virtual bool setOrientation(Orientation orientation) override
Set display orientation.
-
inline virtual PixelFormat getPixelFormat() const override
All surfaces support the same pixel format.
-
virtual bool setScrollMargins(uint16_t top, uint16_t bottom) override
Set margins for hardware scrolling.
Area between top/bottom can be scrolled using
scroll()method.- Parameters:
top – Number of fixed pixels at top of screen
bottom – Number of fixed pixels at bottom of screen
-
virtual bool scroll(int16_t y) override
Scroll region of display up or down using hardware scrolling.
- Parameters:
y – Number of lines to scroll. Positive values scroll content down, negative values scroll up.
-
inline virtual Size getNativeSize() const override
-
class ILI9341 : public Graphics::MipiDisplay
- #include <ILI9341.h>
-
class NullDevice : public Graphics::AbstractDisplay
- #include <Null.h>
Null display device, discards data.
Used for testing performance and algorithms.
Public Functions
-
inline virtual Size getNativeSize() const override
Get physical size of display.
- Return values:
Size – Dimensions for NORMAL orientation
-
inline virtual bool setOrientation(Orientation orientation) override
Set display orientation.
-
inline virtual bool setScrollMargins(uint16_t top, uint16_t bottom) override
Set margins for hardware scrolling.
Area between top/bottom can be scrolled using
scroll()method.- Parameters:
top – Number of fixed pixels at top of screen
bottom – Number of fixed pixels at bottom of screen
-
inline virtual bool scroll(int16_t y) override
Scroll region of display up or down using hardware scrolling.
- Parameters:
y – Number of lines to scroll. Positive values scroll content down, negative values scroll up.
-
inline virtual PixelFormat getPixelFormat() const override
All surfaces support the same pixel format.
-
inline virtual Size getNativeSize() const override
-
class ST7789V : public Graphics::MipiDisplay
- #include <ST7789V.h>
-
class Virtual : public Graphics::AbstractDisplay
-
namespace Drawing
Enums
-
union Header
- #include <Header.h>
Command header structure.
Public Types
Public Members
-
uint8_t index
Register index.
-
uint32_t param
-
ResourceKind kind
-
LengthSize lengthSize
-
uint8_t index
-
class Reader
- #include <Reader.h>
-
struct Registers
- #include <Registers.h>
-
class Renderer : public Graphics::MultiRenderer
- #include <Renderer.h>
A drawing contains a compact list of drawing commands, like a virtual plotter.
-
class Target
- #include <Target.h>
-
class Writer
- #include <Writer.h>
-
union Header
-
namespace Mipi
Enums
-
enum SerialInterfaceCommand
Values:
-
enumerator DSI_V_SYNC_START
-
enumerator DSI_V_SYNC_END
-
enumerator DSI_H_SYNC_START
-
enumerator DSI_H_SYNC_END
-
enumerator DSI_COMPRESSION_MODE
-
enumerator DSI_END_OF_TRANSMISSION
-
enumerator DSI_COLOR_MODE_OFF
-
enumerator DSI_COLOR_MODE_ON
-
enumerator DSI_SHUTDOWN_PERIPHERAL
-
enumerator DSI_TURN_ON_PERIPHERAL
-
enumerator DSI_GENERIC_SHORT_WRITE_0_PARAM
-
enumerator DSI_GENERIC_SHORT_WRITE_1_PARAM
-
enumerator DSI_GENERIC_SHORT_WRITE_2_PARAM
-
enumerator DSI_GENERIC_READ_REQUEST_0_PARAM
-
enumerator DSI_GENERIC_READ_REQUEST_1_PARAM
-
enumerator DSI_GENERIC_READ_REQUEST_2_PARAM
-
enumerator DSI_DCS_SHORT_WRITE
-
enumerator DSI_DCS_SHORT_WRITE_PARAM
-
enumerator DSI_DCS_READ
-
enumerator DSI_EXECUTE_QUEUE
-
enumerator DSI_SET_MAXIMUM_RETURN_PACKET_SIZE
-
enumerator DSI_NULL_PACKET
-
enumerator DSI_BLANKING_PACKET
-
enumerator DSI_GENERIC_LONG_WRITE
-
enumerator DSI_DCS_LONG_WRITE
-
enumerator DSI_PICTURE_PARAMETER_SET
-
enumerator DSI_COMPRESSED_PIXEL_STREAM
-
enumerator DSI_LOOSELY_PACKED_PIXEL_STREAM_YCBCR20
-
enumerator DSI_PACKED_PIXEL_STREAM_YCBCR24
-
enumerator DSI_PACKED_PIXEL_STREAM_YCBCR16
-
enumerator DSI_PACKED_PIXEL_STREAM_30
-
enumerator DSI_PACKED_PIXEL_STREAM_36
-
enumerator DSI_PACKED_PIXEL_STREAM_YCBCR12
-
enumerator DSI_PACKED_PIXEL_STREAM_16
-
enumerator DSI_PACKED_PIXEL_STREAM_18
-
enumerator DSI_PIXEL_STREAM_3BYTE_18
-
enumerator DSI_PACKED_PIXEL_STREAM_24
-
enumerator DSI_V_SYNC_START
-
enum SerialTransactionType
Values:
-
enumerator DSI_RX_ACKNOWLEDGE_AND_ERROR_REPORT
-
enumerator DSI_RX_END_OF_TRANSMISSION
-
enumerator DSI_RX_GENERIC_SHORT_READ_RESPONSE_1BYTE
-
enumerator DSI_RX_GENERIC_SHORT_READ_RESPONSE_2BYTE
-
enumerator DSI_RX_GENERIC_LONG_READ_RESPONSE
-
enumerator DSI_RX_DCS_LONG_READ_RESPONSE
-
enumerator DSI_RX_DCS_SHORT_READ_RESPONSE_1BYTE
-
enumerator DSI_RX_DCS_SHORT_READ_RESPONSE_2BYTE
-
enumerator DSI_RX_ACKNOWLEDGE_AND_ERROR_REPORT
-
enum DisplayCommandSet
Values:
-
enumerator DCS_NOP
-
enumerator DCS_SOFT_RESET
-
enumerator DCS_GET_COMPRESSION_MODE
-
enumerator DCS_GET_DISPLAY_ID
-
enumerator DCS_GET_ERROR_COUNT_ON_DSI
-
enumerator DCS_GET_RED_CHANNEL
-
enumerator DCS_GET_GREEN_CHANNEL
-
enumerator DCS_GET_BLUE_CHANNEL
-
enumerator DCS_GET_DISPLAY_STATUS
-
enumerator DCS_GET_POWER_MODE
-
enumerator DCS_GET_ADDRESS_MODE
-
enumerator DCS_GET_PIXEL_FORMAT
-
enumerator DCS_GET_DISPLAY_MODE
-
enumerator DCS_GET_SIGNAL_MODE
-
enumerator DCS_GET_DIAGNOSTIC_RESULT
-
enumerator DCS_ENTER_SLEEP_MODE
-
enumerator DCS_EXIT_SLEEP_MODE
-
enumerator DCS_ENTER_PARTIAL_MODE
-
enumerator DCS_ENTER_NORMAL_MODE
-
enumerator DCS_GET_IMAGE_CHECKSUM_RGB
-
enumerator DCS_GET_IMAGE_CHECKSUM_CT
-
enumerator DCS_EXIT_INVERT_MODE
-
enumerator DCS_ENTER_INVERT_MODE
-
enumerator DCS_SET_GAMMA_CURVE
-
enumerator DCS_SET_DISPLAY_OFF
-
enumerator DCS_SET_DISPLAY_ON
-
enumerator DCS_SET_COLUMN_ADDRESS
-
enumerator DCS_SET_PAGE_ADDRESS
-
enumerator DCS_WRITE_MEMORY_START
-
enumerator DCS_WRITE_LUT
-
enumerator DCS_READ_MEMORY_START
-
enumerator DCS_SET_PARTIAL_ROWS
-
enumerator DCS_SET_PARTIAL_COLUMNS
-
enumerator DCS_SET_SCROLL_AREA
-
enumerator DCS_SET_TEAR_OFF
-
enumerator DCS_SET_TEAR_ON
-
enumerator DCS_SET_ADDRESS_MODE
-
enumerator DCS_SET_SCROLL_START
-
enumerator DCS_EXIT_IDLE_MODE
-
enumerator DCS_ENTER_IDLE_MODE
-
enumerator DCS_SET_PIXEL_FORMAT
-
enumerator DCS_WRITE_MEMORY_CONTINUE
-
enumerator DCS_SET_3D_CONTROL
-
enumerator DCS_READ_MEMORY_CONTINUE
-
enumerator DCS_GET_3D_CONTROL
-
enumerator DCS_SET_VSYNC_TIMING
-
enumerator DCS_SET_TEAR_SCANLINE
-
enumerator DCS_GET_SCANLINE
-
enumerator DCS_SET_DISPLAY_BRIGHTNESS
-
enumerator DCS_GET_DISPLAY_BRIGHTNESS
-
enumerator DCS_WRITE_CONTROL_DISPLAY
-
enumerator DCS_GET_CONTROL_DISPLAY
-
enumerator DCS_WRITE_POWER_SAVE
-
enumerator DCS_GET_POWER_SAVE
-
enumerator DCS_SET_CABC_MIN_BRIGHTNESS
-
enumerator DCS_GET_CABC_MIN_BRIGHTNESS
-
enumerator DCS_READ_DDB_START
-
enumerator DCS_READ_PPS_START
-
enumerator DCS_READ_DDB_CONTINUE
-
enumerator DCS_READ_PPS_CONTINUE
-
enumerator DCS_NOP
-
enum DcsAddressMode
Values:
-
enumerator DCS_ADDRESS_MODE_MIRROR_Y
-
enumerator DCS_ADDRESS_MODE_MIRROR_X
-
enumerator DCS_ADDRESS_MODE_SWAP_XY
-
enumerator DCS_ADDRESS_MODE_REFRESH_BT
-
enumerator DCS_ADDRESS_MODE_BGR
-
enumerator DCS_ADDRESS_MODE_RGB
-
enumerator DCS_ADDRESS_MODE_LATCH_RL
-
enumerator DCS_ADDRESS_MODE_FLIP_X
-
enumerator DCS_ADDRESS_MODE_FLIP_Y
-
enumerator DCS_ADDRESS_MODE_MIRROR_Y
-
enum SerialInterfaceCommand
-
namespace Resource
Functions
-
void init(IDataSourceStream *stream)
Application calls this method to set source for graphics resourcess.
Graphics resource data such is compiled into a single binary file which the application must mount in some way (typically as a dedicated partition), then create a stream so the graphics library can access it.
- Parameters:
stream – Where to obtain resource data from
-
IDataSourceStream *createSubStream(uint32_t offset, size_t size)
Graphics objects call this method to obtain access to resource data.
The resource compiler generates a header file containing resource descriptions. The application passes this to the appropriate object constructor, which in turn calls this function to access the related binary data (e.g. font or image bitmaps).
-
struct GlyphResource
- #include <resource.h>
Describes glyph bitmap and position.
Public Members
-
uint16_t bmOffset
Offset relative to TypefaceResource::bmpOffset.
-
uint8_t width
Bitmap dimensions in pixels.
-
uint8_t height
Bitmap dimensions in pixels.
-
int8_t xOffset
X dist from cursor pos to UL corner.
-
int8_t yOffset
Y dist from cursor pos to UL corner.
-
uint8_t xAdvance
Distance to advance cursor (x axis)
-
uint16_t bmOffset
-
struct GlyphBlock
- #include <resource.h>
Identifies a run of unicode characters.
-
struct TypefaceResource
- #include <resource.h>
Public Members
-
uint32_t bmOffset
Start of bitmap data in resource stream.
-
uint32_t bmOffset
-
struct FontResource
- #include <resource.h>
-
struct ImageResource
- #include <resource.h>
-
void init(IDataSourceStream *stream)
-
using GlyphOptions = TextOptions
References
Used by
AnimatedGIF Library
Advanced Animation Sample
Basic Animation Sample
Basic Graphics Sample
Basic Touch Sample
Bresenham Sample
Color Test Sample
Custom Object Sample
Environment Variables
ENABLE_VIRTUAL_SCREEN
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HMC5883L Compass
https://github.com/jrowberg/i2cdevlib.git
References
Used by
HMC5883L Compass Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
HardwareSPI
Asynchronous SPI device stack for Sming. Initially written for ESP8266 but supports other architectures.
Problem statement
The ESP8266 has limited I/O and the most useful way to expand it is using serial devices. There are several types available:
- I2C
Can be fast-ish but more limited than SPI, however slave devices generally cheaper, more widely available and simpler to interface as they only require 2 pins. The ESP8266 does have hardware support however, so requires a bit-banging solution. Inefficient.
- I2S
Designed for streaming audio devices but has other uses. See Esp8266 Drivers.
- RS232
Tied in with RS485, Modbus, DMX512, etc. Well-supported with asynchronous driver.
- SPI
Speed generally limited by slave device capability, can be multiplexed (‘overlapped’) onto SPI0 (flash memory) pins. The two SPI modules are identical, but the additional pins for quad/dual modes are only brought out for SPI0. In addition, three user chip selects are available in this mode; there is only one in normal mode although this can be handled using a regular GPIO.
The purpose of this driver is to provide a consistent interface which makes better use of the hardware capabilities.
Classes may inherit from
HSPI::Deviceto provide support for specific devices such as Flash memory, SPI RAM, LCD controllers, etc.Device objects are attached to the stack via specified PinSet (overlapped or normal) and chip select.
Multiple concurrent devices are supported, limited only by available chip selects and physical constraints such as wire length, bus speeds.
2 and 4-bit modes are supported via overlap pins.
A
HSPI::Requestobject supports transfers of up to 64K. The controller splits these into smaller transactions as required.Asynchronous execution so application does not block during SPI transfer. Application may provide a per-request callback to be notified when request has completed.
Blocking requests are also supported.
Support for moving data between Sming Streams and SPI memory devices using the
HSPI::StreamAdapter.
SPI system expansion
A primary use-case for this driver is to provide additional resources for the ESP8266 using SPI devices such as:
MC23S017 SPI bus expander. Operates at 10MHz (regular SPI) and provides 16 additional I/O with interrupt capability.
High-speed shift registers. These can be wired directly to SPI buses to expand GPIO capability.
Epson S1D13781 display controller. See TFT_S1D13781. Evaluation boards are inexpensive and is a useful way to evaluate display modules with TFT interfaces. The Newhaven NHD-5.0-800480TF-ATXL#-CTP was used during development of this driver.
Bridgetek FT813 EVE TFT display controller. This supports dual/quad modes and clocks up to 30MHz.
NRF24L01 RF transceiver. Rated bus speed is 8MHz but it seems to work fine at 40MHz.
Serial memory devices. The library contains a driver for the IS62/65WVS2568GALL fast serial RAM, which clocks up to 45MHz and supports SDI/SQI modes.
Software operation
Overview
We have:
Controller: SPI hardware
Device: Slave device on the SPI bus
Request: Details of a transaction for a specific device
The Controller maintains an active list of requests (as a linked-list), but does not own the request objects. These will be allocated by the application, configured then submitted to the Controller for execution.
The Device specifies how a slave is connected to the bus, and that may change dynamically. For example, at reset an SPI RAM may be in SPI mode but can be switched into SDI/SQI modes.
This is device-specific so would be implemented by superclassing HSPI::Device.
Requests
Each HSPI::Request is split into transactions.
A transaction has four phases: command - address - MOSI - dummy - MISO.
All phases are optional.
The dummy bits are typically used in read modes and specified by the device datasheet.
No data is transferred during this phase.
The ESP8266 hardware FIFO is used for MOSI/MISO phases and is limited to 64 bytes, so larger transfers must be broken into chunks. The driver handles this automatically.
Requests may be executed asynchronously so the call will not block and the CPU can continue with normal operations. An optional callback is invoked when the request has completed. As an example, consider moving a 128KByte file from flash storage into FT813 display memory:
Read the first file chunk into a RAM buffer, submit an SPI request1 to transfer it asynchronously
Read the second file chunk into another RAM buffer, and prepare request2 for that (but do not submit it yet)
When request1 has completed, submit request2 (from the interrupt callback). Schedule a task to read the next chunk and prepare request1.
When request2 has completed, continue from step (2) to submit request1, etc.
Timing
A 64-byte data transfer (full hardware FIFO with 1 command byte and 3-byte address) at 26MHz would take 21us (5.25us in QIO mode) or 1680 (420) CPU cycles. To transfer 128Kbytes would take 2048 such transactions, 43ms (11ms for QIO), not including memory copy overheads.
In practice request sizes will be much smaller due to RAM constraints. Nevertheless, at high clock speeds the interrupt rate increases to the point where it consumes more CPU cycles than the actual transfer. The driver therefore disables interrupts in these situations and executes the request in task mode.
Bear in mind that issuing a blocking request will also require all queued requests to complete.
The driver does not currently support out-of-order execution, which might prioritise faster devices.
Pin Set
To avoid confusion, we’ll refer to the flash memory SPI bus as SPI0, and the user bus as SPI1. This driver doesn’t support direct use of SPI0 as on most devices it is reserved for flash memory. However, an overlap mode is supported which makes use of hardware arbitration to perform SPI1 transactions using SPI0 pins. This has several advantages:
Liberates three GPIO which would normally be required for MOSI, MISO and SCLK.
Only one additional pin is required for chip select.
Additional 2/4 bits-per-clock modes are available for supported devices.
For the ESP8266, these are the HSPI::PinSet assignments:
- PinSet::normal
- MISO=GPIO12, MOSI=GPIO13, SCLK=GPIO14. One chip select:
GPIO15 (HSPI CS)
- PinSet::overlap
- MISO=SD0, MOSI=SD1, IO2=SD3, IO3=SD2, SCLK = CLK. Three chip selects:
GPIO15 (HSPI_CS)
GPIO1 (SPI_CS1 / UART0_TXD). This conflicts with the normal serial TX pin which should be swapped to GPIO2 if required.
GPIO0 (SPI_CS2)
- PinSet::manual
Typically a GPIO will be assigned to perform chip select (CS). The application should register a callback function via
HSPI::onSelectDevice()which performs the actual switching. This MUST be in IRAM.
Note
The connections for IO2/3 look wrong above, but on two different models of SPI RAM chip these have been verified as correct by writing in SPIHD mode and reading in quad mode.
Multiplexed CS
Multiple devices can be supported on a single CS using, for example using a HC138 3:8 decoder. The CS line is connected to an enable input, with three GPIO outputs setting A0-2.
A custom controller should be created like this:
class CustomController: public HSPI::Controller
{
public:
bool startDevice(Device& dev, PinSet pinSet, uint8_t chipSelect) override
{
/*
* You should perform any custom validation here and return false on failure.
* For example, if we're only using 3 of the 8 available outputs.
*/
auto addr = chipSelect & 0x07;
if(addr > 3) {
debug_e("Invalid CS addr: %u", addr);
return false;
}
/*
* Provide a callback to route chip select signal as required.
* Note this only needs to be done once, so could be called externally without
* subclassing HSPI::Controller if the other mechanisms aren't required.
*/
onSelectDevice(selectDevice);
/*
* Initialise hardware Controller
*/
auto cs = chipSelect >> 3;
return HSPI::Controller::startDevice(dev, pinSet, cs);
}
private:
static void IRAM_ATTR selectDevice(uint8_t chipSelect, bool active);
uint8_t activeChipSelect{0};
};
Now in the .cpp file:
CustomController spi;
void IRAM_ATTR CustomController::selectDevice(uint8_t chipSelect, bool active)
{
// Only perform GPIO if CS changes as GPIO is expensive
if(active && chipSelect != activeChipSelect) {
auto addr = chipSelect & 0x07;
digitalWrite(PIN_MUXADDR0, addr & 0x01);
digitalWrite(PIN_MUXADDR1, addr & 0x02);
// As we only need 2 address lines, can leave this one
// digitalWrite(PIN_MUXADDR2, addr & 0x03);
activeChipSelect = chipSelect;
}
// If using a hardware CS output then we're done.
if(spi.getActivePinSet() == HSPI::PinSet::manual) {
// For manual chip select operation, set the state directly here
}
}
IO Modes
Not to be confused with HSPI::ClockMode, the HSPI::IoMode determines how
the command, address and data phases are transferred:
.
Bits per clock
.
IO Mode
Command
Address
Data
Duplex
SPI
1
1
1
Full
SPIHD
1
1
1
Half
SPI3WIRE
1
1
1
Half
DUAL
1
1
2
Half
DIO
1
2
2
Half
SDI
2
2
2
Half
QUAD
1
1
4
Half
QIO
1
4
4
Half
SQI
4
4
4
Half
Note
SDI and SQI are not supported directly by hardware, but is implemented within the driver using the address phase. In these modes, commands are limited to 8 bits.
This seems to be consistent with the ESP32 IDF driver, as in spi_ll.h:
/** IO modes supported by the master. */
typedef enum {
SPI_LL_IO_MODE_NORMAL = 0, ///< 1-bit mode for all phases
SPI_LL_IO_MODE_DIO, ///< 2-bit mode for address and data phases, 1-bit mode for command phase
SPI_LL_IO_MODE_DUAL, ///< 2-bit mode for data phases only, 1-bit mode for command and address phases
SPI_LL_IO_MODE_QIO, ///< 4-bit mode for address and data phases, 1-bit mode for command phase
SPI_LL_IO_MODE_QUAD, ///< 4-bit mode for data phases only, 1-bit mode for command and address phases
} spi_ll_io_mode_t;
Somne devices (e.g. W25Q32 flash) have specific commands to support these modes, but others (e.g. IS62/65WVS2568GALL fast serial RAM) do not, and the SDI/SQI mode setting applies to all phases. This needs to be implemented in the driver as otherwise the user code is more complex than necesssary and performance suffers considerably.
Streaming
The HSPI::StreamAdapter provides support for streaming of data to/from memory devices.
This would be used, for example, to transfer content to or from a FileStream
or FlashMemoryStream to SPI RAM asynchronously.
Supported devices must inherit from HSPI::MemoryDevice.
API
-
enum class HSPI::ClockMode : uint8_t
SPI clock polarity (CPOL) and phase (CPHA)
Values:
-
enumerator mode0
CPOL: 0 CPHA: 0.
-
enumerator mode1
CPOL: 0 CPHA: 1.
-
enumerator mode2
CPOL: 1 CPHA: 0.
-
enumerator mode3
CPOL: 1 CPHA: 1.
-
enumerator mode0
-
enum class HSPI::IoMode : uint8_t
Mode of data transfer.
Values:
-
enumerator SPI
One bit per clock, MISO stage concurrent with MISO (full-duplex)
-
enumerator SPIHD
One bit per clock, MISO stage follows MOSI (half-duplex)
-
enumerator SPI3WIRE
Half-duplex using MOSI for both sending and receiving data.
-
enumerator MAX
-
enumerator SPI
-
enum class HSPI::PinSet
How SPI hardware pins are connected.
Values:
-
enumerator none
Disabled.
-
enumerator normal
Standard HSPI pins.
-
enumerator manual
HSPI pins with manual chip select.
-
enumerator overlap
Overlapped with SPI 0.
-
enumerator none
-
struct Request
Defines an SPI Request Packet.
Request fields may be accessed directly or by use of helper methods.
Application is responsible for managing Request object construction/destruction. Queuing is managed as a linked list so the objects aren’t copied.
Applications will typically only require a couple of Request objects, so one can be prepared whilst the other is in flight. This helps to minimises the setup latency between SPI transactions.
Set value for command phase
-
inline void setCommand(uint16_t command, uint8_t bitCount)
- Parameters:
command –
bitCount – Length of command in bits
-
inline void setCommand8(uint8_t command)
Set 8-bit command.
- Parameters:
command –
-
inline void setCommand16(uint16_t command)
Set 16-bit command.
- Parameters:
command –
Set value for address phase
-
inline void setAddress(uint32_t address, uint8_t bitCount)
- Parameters:
address –
bitCount – Length of address in bits
-
inline void setAddress24(uint32_t address)
Set 24-bit address.
- Parameters:
address –
Public Functions
Public Members
-
Request *next = {nullptr}
Controller uses this to queue requests.
-
uint16_t cmd = {0}
Command value.
-
uint8_t cmdLen = {0}
Command bits, 0 - 16.
-
uint8_t async
Set for asynchronous operation.
-
uint8_t task
Controller will execute this request in task mode.
-
uint32_t addr = {0}
Address value.
-
uint8_t addrLen = {0}
Address bits, 0 - 32.
-
uint8_t dummyLen = {0}
Dummy read bits between address and read data, 0 - 255.
-
uint8_t sizeAlign = {0}
Required size alignment of each transaction (if split up)
-
void *param = {nullptr}
User parameter.
-
inline void setCommand(uint16_t command, uint8_t bitCount)
-
struct Data
Specifies a block incoming or outgoing data.
Data can be specified directly within
Data, or as a buffer reference.Command or address are stored in native byte order and rearranged according to the requested byteOrder setting. Data is always sent and received LSB first (as stored in memory) so any re-ordering must be done by the device or application.
Set internal data value of 1-4 bytes
Note
Data is sent LSB, MSB (native byte order)
-
inline void set8(uint8_t data)
Set to single 8-bit value.
- Parameters:
data –
-
inline void set16(uint16_t data)
Set to single 16-bit value.
- Parameters:
data –
-
inline void set32(uint32_t data, uint8_t len = 4)
Set to 32-bit data.
- Parameters:
data –
len – Length in bytes (1 - 4)
-
inline void set8(uint8_t data)
-
class Device
Manages a specific SPI device instance attached to a controller.
Subclassed by Graphics::SpiDisplay, Graphics::XPT2046, HSPI::MemoryDevice
Public Functions
-
inline bool begin(PinSet pinSet, uint8_t chipSelect, uint32_t clockSpeed)
Register device with controller and prepare for action.
- Parameters:
pinSet – Use PinSet::normal for Esp32, other values for Esp8266
chipSelect – Identifies the CS number for ESP8266, or the GPIO pin for ESP32
clockSpeed – Bus speed
-
inline bool isReady() const
Determine if the device is initialised.
- Return values:
bool –
-
virtual IoModes getSupportedIoModes() const = 0
Return set of IO modes supported by a device implementation.
-
inline bool begin(PinSet pinSet, uint8_t chipSelect, uint32_t clockSpeed)
-
class MemoryDevice : public HSPI::Device
Base class for read/write addressable devices.
Subclassed by HSPI::RAM::IS62_65, HSPI::RAM::PSRAM64
Prepare a write request
Prepare a read request
Public Functions
-
inline void write(uint32_t address, const void *data, size_t len)
Write a block of data.
Note
Limited by current operating mode
- Parameters:
address –
data –
len –
-
inline void read(uint32_t address, void *buffer, size_t len)
Read a block of data.
Note
Limited by current operating mode
- Parameters:
address –
data –
len –
-
inline void write(uint32_t address, const void *data, size_t len)
-
class PSRAM64 : public HSPI::MemoryDevice
PSRAM64(H) pseudo-SRAM.
-
class IS62_65 : public HSPI::MemoryDevice
IS62/65WVS2568GALL fast serial RAM.
-
class Controller : public ControllerBase
Manages access to SPI hardware.
Public Types
-
using SelectDevice = void (*)(uint8_t chipSelect, bool active)
Interrupt callback for custom Controllers.
For manual CS (PinSet::manual) the actual CS GPIO must be asserted/de-asserted.
Expanding the SPI bus using a HC138 3:8 multiplexer, for example, can also be handled here, setting the GPIO address lines appropriately.
- Param chipSelect:
The value passed to
startDevice()- Param active:
true when transaction is about to start, false when completed
Public Functions
-
void end()
Disable HSPI controller.
Note
Reverts HSPI pins to GPIO and disables the controller
-
IoModes getSupportedIoModes(const Device &dev) const
Determine which IO modes are supported for the given device.
May be restricted by both controller and device capabilities.
-
inline void onSelectDevice(SelectDevice callback)
Set interrupt callback to use for manual CS control (PinSet::manual) or if CS pin is multiplexed.
Note
Callback MUST be marked IRAM_ATTR
-
virtual bool startDevice(Device &dev, PinSet pinSet, uint8_t chipSelect, uint32_t clockSpeed)
Assign a device to a CS# using a specific pin set. Only one device may be assigned to any CS.
Custom controllers should override this method to verify/configure chip selects, and also provide a callback (via
onSelectDevice()).
-
void configChanged(Device &dev)
Devices call this method to tell the Controller about configuration changes. Internally, we just set a flag and update the register values when required.
-
inline SpiBus getBusId() const
Get the active bus identifier.
On successful call to begin() returns actual bus in use.
-
bool loopback(bool enable)
For testing, tie MISO <-> MOSI internally.
enable true to enable loopback, false for normal receive operation
- Return values:
true – on success, false if loopback not supported
-
using SelectDevice = void (*)(uint8_t chipSelect, bool active)
-
class StreamAdapter
Helper class for streaming data to/from SPI devices.
References
Used by
Sming Graphics Library Library
TFT_S1D13781 Library
Basic Ethernet Sample
Environment Variables
HSPI_ENABLE_STATSHSPI_ENABLE_TESTPINSHSPI_TESTPIN1HSPI_TESTPIN2
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Hue Emulator
A framework for emulating Hue smart light devices via the Hue::Bridge class.
A real bridge talks to Hue devices via ZigBee, however with Sming you can control anything you want using the published API. Refer to specifications available at https://developers.meethue.com (free account login required).
Setup
Refer to the Basic Alexa sample for details of how to use this library. Here are a few key notes.
A HttpServer object is required to allow the framework to respond to requests.
Note that Gen 3+ Hue Bridges listen on port 80 (standard HTTP), however older versions use port 1901.
This library has only been tested on port 80.
In your application, remember to add bodyparsers for JSON and XML:
server.setBodyParser(MIME_JSON, bodyToStringParser);
server.setBodyParser(MIME_XML, bodyToStringParser);
Without these, you’ll get empty bodies for incoming requests.
The sample demonstrates use of provided On/Off, Dimmable and Colour device types with a global callback function.
Ideally you should provide your own custom Hue devices by inheriting from Hue::Device.
This is demonstrated using MyHueDevice. The device ID is 666.
API
-
class Bridge : public UPnP::schemas_upnp_org::device::Basic1Template<Bridge>
Public Types
-
using ConfigDelegate = Delegate<void(const Config &config)>
Called when a new user key is created.
The application should use this to store new users in persistent memory. At startup, these should be passed back via the
configure()method.
-
using StateChangeDelegate = Delegate<void(const Hue::Device &device, Hue::Device::Attributes attr)>
A global callback may be provided to perform actions when device states change.
The callback is invoked only when all request actions have been completed. The current state may be quereied using
device::getAttribute.- Param device:
The device which has been updated
- Param attr:
A set of flags indicating which attributes were changed
Public Functions
-
inline Bridge(Hue::Device::Enumerator &devices)
Constructor.
- Parameters:
devices – List of devices to present
-
void configure(const Config &config)
Perform a configuration action.
- Parameters:
config – The action to perform
-
inline void enablePairing(bool enable)
Enable creation of new users.
This could be enabled via web page on local Access Point, or physical push-button. It should also be time limited, so exits pairing mode after maybe 30 seconds. If a user creation request is received then this is disabled automatically.
Note
DO NOT leave this permanently enabled!
-
inline const Stats &getStats()
Get bridge statistics.
- Return values:
const – Stats&
-
inline void resetStats()
Clear the bridge statistics.
-
inline const UserMap &getUsers() const
Access the list of users.
- Return values:
const – UserMap&
-
void getStatusInfo(JsonObject json)
Get bridge status information in JSON format.
- Parameters:
json – Where to write information
-
struct Config
-
using ConfigDelegate = Delegate<void(const Config &config)>
-
class Device : public UPnP::Item
Subclassed by Hue::OnOffDevice
Public Types
-
using Callback = Delegate<void(Status status, int errorCode)>
Callback invoked when setAttribute() has completed.
Note
Any status other than
successis considered a failure- Param status:
Result of the operation
- Param errorCode:
Application-specific error code
Public Functions
-
virtual Status setAttribute(Attribute attr, unsigned value, Callback callback) = 0
Set a device attribute.
Note
DO NOT invoke the callback directly: only use it if pended.
- Parameters:
attr – The attribute to change
value – Value for the attribute (exact type is attribute-specific)
callback – If you return Status::pending, invoke this callback when completed
- Return values:
Status –
-
virtual bool getAttribute(Attribute attr, unsigned &value) const = 0
Get the (cached) device attribute value.
- Parameters:
attr –
value –
- Return values:
bool – true on success, false if attribute not supported or value unknown
-
virtual String getUniqueId() const
Returns the unique device ID string.
Note
Other forms of ID string may be used, however for maximum compatibility the standard format should be used. By default, this method uses the WiFi station MAC address, with 00:11 appended plus the 8-bit device ID.
- Return values:
String – Unique ID of the form AA:BB:CC:DD:EE:FF:00:11-XX, consisting of a 64-bit Zigbee MAC address plus unique endpoint ID.
-
class Enumerator : public UPnP::Enumerator<Device, Enumerator>
Abstract class to manage a list of devices.
Note
Applications must provide an implementation of this for the bridge. Returned device objects may only be considered valid for the duration of the current task call as they may be destroyed at any time.
Subclassed by Hue::DeviceListEnumerator
-
using Callback = Delegate<void(Status status, int errorCode)>
-
class OnOffDevice : public Hue::Device
Subclassed by Hue::DimmableDevice
-
class DimmableDevice : public Hue::OnOffDevice
Subclassed by Hue::ColourDevice
-
class ColourDevice : public Hue::DimmableDevice
-
enum class Hue::Status
Status of a
setAttributerequest.Values:
-
enumerator success
The action was performed immediately without error.
-
enumerator pending
The action was accepted but requires further processing.
Use this to perform requests asynchronously. You MUST invoked the provided
Callbackfunction to complete the request.When controlling remote devices, for example connected via serial link, you might issue the command immediately and then return
pending. When the serial response is received, or a timeout occurs, then the request can be completed. Note that the error code passed to the callback is optional and will be specific to your application: it will be output in verbose debug mode so may be useful.
-
enumerator error
Action could not be completed.
If the Attribute not supported by your device, or an internal I/O error occurred then return this value.
-
enumerator success
References
Used by
Basic Alexa Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
I2C Device Class
This is a generic class which abstracts bit and byte I2C R/W functions. It is part of The I2C Device Library.
There are examples in many of the classes that demonstrate basic usage patterns. The I2Cdev class is built to be used statically, reducing the memory requirement if you have multiple I2C devices in your project. Only one instance of the I2Cdev class is required.
https://github.com/jrowberg/i2cdevlib.git
References
Used by
HMC5883L Compass Library
MPU6050 Gyro / Accelerometer Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
IO Control
Introduction
This library is an asynchronous Input/Output device stack for embedded devices such as the ESP8266. The main reason for building this was to provide a mechanism for serialising modbus requests. We also needed to serialise commands to a 433MHz RF switch.
The goal is to abstract away differences between different communication protocols and provide a uniform API for controlling a network of device nodes.
To keep objects generic (i.e. not too modbus-centric) we define the following objects and features:
- Controller
The hardware which talks over RS485 (or whatever). There is one controller instance for each physical device available. The controller serialises requests.
- Device
An instance of a modbus slave or RF switch transmitter. These sit on top of a controller to define characteristics specific to the device (or slave). Multiple devices may be attached to a controller.
- Request
Encapsulates a command request for a specific device class. Requests are created, queued, executed then destroyed. Callbacks are issued to notify the start of execution and completion.
- Device Manager
Keeps track of all registered controllers and provides methods for creating and issuing requests.
Configuration is managed using JSON. Controller instances are provided by the application. Devices are defined using a JSON configuration file as devices may be added or removed from a deployed network.
The system for which this library was developed uses a basic HTML/javascript front-end which deals with configuration and generating command requests (in JSON). This keeps embedded memory requirements to a minimum.
Controllers are given an ID which indicates their class, plus index within that class:
rs485#0RS485 serial controller #0rfswitch#0RF Controller #0 (i-Lumos RF Light switch)
Device identifiers can be anything but must be globally unique, e.g. mb1, rf0, dmx1, etc. See Basic RS485 sample, config/devices.json.
We can command a device thus:
{ "device": "mb1", "node": 1, "command": "toggle" }
{ "device": "rf0", "code": 0x123456 }
Commands are defined using a command set of codes via IO::Command.
Not all device types support all commands.
API Documentation
DMX512
https://en.wikipedia.org/wiki/DMX512
Written to support DMX lighting controllers for dimmable lights. The model shown is for controlling RGB LED strip lighting.
DMX512 uses the RS485 physical interface, but a different protocol and usually higher baud rates than MODBUS. Nevertheless, it is possible to mix difference device types on the same network.
Requests to other devices will generally appear as garbage so shouldn’t have any bad side-effects.
- namespace DMX512
DMX512/Device.h
Created on: 5 November 2018
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
DMX512/Request.h
Created on: 5 November 2018
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
- struct NodeData
- class Device : public IO::RS485::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- inline virtual uint16_t maxNodes() const override
Determine maximum number of nodes supported by the device.
- Return values:
uint16_t – 0 if device doesn’t support nodes
- virtual void handleEvent(IO::Request *request, Event event) override
Implementations may override this method to customise event handling.
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- class Request : public IO::Request
Public Functions
- virtual ErrorCode parseJson(JsonObjectConst json) override
Fill this request from a JSON description.
- virtual void getJson(JsonObject json) const override
Get result of a completed request in JSON format.
- virtual bool setNode(DevNode node) override
If nodes are supported, implement this method.
/
/**
- inline virtual bool setValue(int value) override
If nodes support values, implement this method.
- virtual void submit() override
Submit a request.
The request is added to the controller’s queue. If the queue is empty, it starts execution immediately. The result of the request is posted to the callback routine.
MODBUS
https://en.wikipedia.org/wiki/Modbus
Implements MODBUS-RTU protocols.
Can use this directly, but generally preferable to create custom Device and Request classes. See R421A Relay Boards for an example.
Device properties include:
- Address
The address of a modbus slave. Modbus docs. call this the slave ID.
- Node
Represents something a slave device does. Modbus relay boards have one node for each output it controls.
- Node ID
The channel a node lives on. For the R421Axx relay boards this is the address or channel number. In a modbus transaction this is the address field.
For efficiency, nodes aren’t implemented as objects, just identifiers used by a Device / Controller. Nodes can be controlled using generic commands such as ‘open’, ‘close’, ‘toggle’, etc.
- namespace Modbus
Modbus/ADU.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Modbus/Debug.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Modbus/Device.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Modbus/Exception.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Modbus/Function.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Modbus/GenericRequest.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Modbus/PDU.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Modbus/Request.h
Created on: 1 May 2018
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Modbus/Slave.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Debug print support, outputs contents of packet
Enums
- enum class Exception
Modbus exception codes returned in response packets.
Values:
- enumerator Success
No exception, transaction completed normally.
- enumerator IllegalFunction
Function not allowed/supported/implemented, or device in wrong state to process request.
For example, an unconfigured device may be unable to return register values.
- enumerator IllegalDataAddress
Data address not allowed.
More specifically, the combination of reference number and transfer length is invalid. For a controller with 100 registers, the ADU addresses the first register as 0, and the last one as 99.
If a request is submitted with a starting register address of 96 and a quantity of registers of 4, then this request will successfully operate (address-wise at least) on registers 96, 97, 98, 99.
If a request is submitted with a starting register address of 96 and a quantity of registers of 5, then this request will fail with Exception Code 0x02 “Illegal Data Address” since it attempts to operate on registers 96, 97, 98, 99 and 100, and there is no register with address 100.
- enumerator IllegalDataValue
Data value not allowed.
This indicates a fault in the structure of the remainder of a complex request, such as that the implied length is incorrect. It specifically does NOT mean that a data item submitted for storage in a register has a value outside the expectation of the application program, since the MODBUS protocol is unaware of the significance of any particular value of any particular register.
- enumerator SlaveDeviceFailure
Protocol slave device failure exception.
An unrecoverable error occurred while the server (or slave) was attempting to perform the requested action.
Functions
- ErrorCode readRequest(RS485::Controller &controller, ADU &adu)
- void sendResponse(RS485::Controller &controller, ADU &adu)
- struct ADU
Prepare outgoing packet
The
slaveAddressandpdufields must be correctly initialised.
- retval size_t:
Size of ADU, 0 on error
Parse a received packet
- param receivedSize:
How much data was received
- retval ErrorCode:
Result of parsing
- class Device : public IO::RS485::Device
A virtual device, represents a modbus slave device.
Actual devices must implement:
callback()
fillRequestData()
Subclassed by IO::Modbus::NT18B07::Device, IO::Modbus::R421A::Device, IO::Modbus::RID35::Device, IO::Modbus::STM8Relay::Device, IO::Modbus::STS::Fan::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- virtual void handleEvent(IO::Request *request, Event event) override
Implementations may override this method to customise event handling.
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- struct PDU
Protocol Data Unit.
Content is independent of the communication layer. Structure does NOT represent over-the-wire format as packing issues make handling PDU::Data awkward. Therefore, the other fields are unaligned and adjusted as required when sent over the wire.
byteCountfieldSome functions with a
byteCountfield are marked ascalculated. This means it doesn’t need to be set when constructing requests or responses, and will be filled in later based on the values of other fields.In other cases, a
setCountmethod is provided to help ensure the byteCount field is set correctly.Get PDU size based on content
Calculation uses byte count so doesn’t access any 16-bit fields
- union Data
Public Members
- uint8_t exceptionCode
- ReadDiscreteInputs readDiscreteInputs
- ReadRegisters readHoldingRegisters
- ReadRegisters readInputRegisters
- WriteSingleCoil writeSingleCoil
- WriteSingleRegister writeSingleRegister
- ReadExceptionStatus readExceptionStatus
- GetComEventCounter getComEventCounter
- GetComEventLog getComEventLog
- WriteMultipleCoils writeMultipleCoils
- WriteMultipleRegisters writeMultipleRegisters
- ReportServerId reportServerId
- MaskWriteRegister maskWriteRegister
- ReadWriteMultipleRegisters readWriteMultipleRegisters
- union GetComEventLog
- union ReadCoils
- struct Request
- struct Response
- union ReportServerId
- class Request : public IO::RS485::Request
Subclassed by IO::Modbus::GenericRequest, IO::Modbus::NT18B07::Request, IO::Modbus::R421A::Request, IO::Modbus::RID35::Request, IO::Modbus::STM8Relay::Request, IO::Modbus::STS::Fan::Request
- namespace NT18B07
Modbus/NT18B07/Device.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Rayleigh Instruments RI-D35 energy meter
Device registers are read-only. The Node ID corresponds to the register address.
Modbus/NT18B07/Request.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Typedefs
- using TempData = int16_t[channelCount]
Variables
- const size_t channelCount = {7}
- class Device : public IO::Modbus::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- int16_t getIntValue(uint8_t channel) const
Get temperature value in 0.1C increments (avoids floating point)
- inline virtual uint16_t maxNodes() const override
Determine maximum number of nodes supported by the device.
- Return values:
uint16_t – 0 if device doesn’t support nodes
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- namespace R421A
Modbus/R421A/Device.h
Created on: 1 May 2018
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
R421A08 modbus 8-channel relay board
Channels are numbered 1 - 8. This is the 16-bit address field in a request.
To simplify operation we use a bitmask to specify relay states. We use the channel number directly so values can range between 0x0001 and 0x01FE
The query command returns a range of states, but other commands work with only a single channel. We therefore implement a mechanism to iterate through all requested channels using the same request.
There is a similar 4-channel board but with no markings and (as yet) no documentation. However all commands appear to work so a designation of R421A04 seems appropriate. Some typos in the 8-channel datasheet indicate that it was adapted from that of a 4-channel version.
R421A04 - 32 addresses set with DIP1-5, DIP6 ON for RTU mode R421A08 - 64 addresses set with DIP1-6, RTU mode only
Modbus/R421A/Request.h
Created on: 1 May 2018
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
- struct StateMask
Tracks state of multiple relays.
Public Members
- BitSet32 channelMask
Identifies valid channels.
- class Device : public IO::Modbus::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- inline virtual DevNode::ID nodeIdMin() const override
Get minimum valid Node ID for this device.
Typically devices have a contiguous valid range of node IDs
- inline virtual uint16_t maxNodes() const override
Determine maximum number of nodes supported by the device.
- Return values:
uint16_t – 0 if device doesn’t support nodes
- virtual DevNode::States getNodeStates(DevNode node) const override
Return the current set of states for all nodes controlled by this device.
Used to determine if, say, all nodes are ON, OFF or a combination.
- virtual void handleEvent(IO::Request *request, Event event) override
Implementations may override this method to customise event handling.
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- class Request : public IO::Modbus::Request
Public Functions
- virtual ErrorCode parseJson(JsonObjectConst json) override
Fill this request from a JSON description.
- virtual void getJson(JsonObject json) const override
Get result of a completed request in JSON format.
- virtual bool setNode(DevNode node) override
If nodes are supported, implement this method.
/
/**
- virtual DevNode::States getNodeStates(DevNode node) override
Query node status from response.
- namespace RID35
Modbus/RID35/Request.h
Created on: 24 March 2022
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Enums
Variables
- constexpr uint16_t stdRegBase = 0x01
- constexpr uint16_t ovfRegBase = 0x96
- constexpr size_t registerCount = stdRegCount + ovfRegCount
- class Device : public IO::Modbus::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- namespace STM8Relay
Modbus/STM8Relay/Device.h
Created on: 24 June 2022
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
STM8S103 modbus relay board
4-channel relay boards with optocouplers and digital inputs. Advertised as using STM8S103 microcontroller, but no model number. Relays controlled using Modbus ‘coil’ commands. Digital inputs read as ‘discrete inputs’. Connected directly to 3.3v microcontroller. Grounding pin sets bit.
Modbus/STM8Relay/Request.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
- struct StateMask
Tracks state of multiple relays.
Public Members
- BitSet32 channelMask
Identifies valid channels.
- class Device : public IO::Modbus::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- inline virtual DevNode::ID nodeIdMin() const override
Get minimum valid Node ID for this device.
Typically devices have a contiguous valid range of node IDs
- inline virtual uint16_t maxNodes() const override
Determine maximum number of nodes supported by the device.
- Return values:
uint16_t – 0 if device doesn’t support nodes
- virtual DevNode::States getNodeStates(DevNode node) const override
Return the current set of states for all nodes controlled by this device.
Used to determine if, say, all nodes are ON, OFF or a combination.
- virtual void handleEvent(IO::Request *request, Event event) override
Implementations may override this method to customise event handling.
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- class Request : public IO::Modbus::Request
Public Functions
- virtual ErrorCode parseJson(JsonObjectConst json) override
Fill this request from a JSON description.
- virtual void getJson(JsonObject json) const override
Get result of a completed request in JSON format.
- virtual bool setNode(DevNode node) override
If nodes are supported, implement this method.
/
/**
- virtual DevNode::States getNodeStates(DevNode node) override
Query node status from response.
- namespace STS
- namespace Fan
Modbus/STS/Fan/Device.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
STS Fan Controller - see samples/Fan
Modbus/STS/Fan/Request.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Variables
- const size_t channelCount = {3}
- struct FanData
- class Device : public IO::Modbus::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- inline virtual uint16_t maxNodes() const override
Determine maximum number of nodes supported by the device.
- Return values:
uint16_t – 0 if device doesn’t support nodes
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- class Request : public IO::Modbus::Request
Public Functions
- inline virtual bool setNode(DevNode node) override
If nodes are supported, implement this method.
/
/**
- inline virtual bool setValue(int value) override
If nodes support values, implement this method.
- virtual ErrorCode callback(PDU &pdu) override
Process a received PDU.
- Parameters:
pdu –
- Return values:
ErrorCode – If request is re-submitted, return Error::pending, otherwise request will be completed with given error.
- virtual void getJson(JsonObject json) const override
Get result of a completed request in JSON format.
NT18B07 7-channel NTC temperature interface board
Registers 0-6 contain temperature readings in 0.1 degree celcius increments
Register 254 contains slave ID, R/W, default 1
- namespace NT18B07
Modbus/NT18B07/Device.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Rayleigh Instruments RI-D35 energy meter
Device registers are read-only. The Node ID corresponds to the register address.
Modbus/NT18B07/Request.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Typedefs
- using TempData = int16_t[channelCount]
Variables
- const size_t channelCount = {7}
- class Device : public IO::Modbus::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- int16_t getIntValue(uint8_t channel) const
Get temperature value in 0.1C increments (avoids floating point)
- inline virtual uint16_t maxNodes() const override
Determine maximum number of nodes supported by the device.
- Return values:
uint16_t – 0 if device doesn’t support nodes
- struct Config
NT18B07 device configuration.
- struct Comp
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
R421A Relay Boards
These are inexpensive and readily available MODBUS multi-channel relay boards. Both 8 and 4-channel versions have been tested.
Device properties include:
- Address
The address of a modbus slave. Modbus docs. call this the slave ID.
- Node
Represents something a slave device does. Modbus relay boards have one node for each output it controls.
- Node ID
The channel a node lives on. For the R421Axx relay boards this is the address or channel number. In a modbus transaction this is the address field.
- 8-channel
DIP switches set the slave address, from 0x00 to 0x3f. e.g. A0 ON=#1, A1 ON=#2, etc.
- 4-channel
Set A5 ON for MODBUS RTU mode (OFF is for AT mode).
Set A0-A4 to slave address, from 0x00 to 0x1f.
API
- namespace R421A
Modbus/R421A/Device.h
Created on: 1 May 2018
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
R421A08 modbus 8-channel relay board
Channels are numbered 1 - 8. This is the 16-bit address field in a request.
To simplify operation we use a bitmask to specify relay states. We use the channel number directly so values can range between 0x0001 and 0x01FE
The query command returns a range of states, but other commands work with only a single channel. We therefore implement a mechanism to iterate through all requested channels using the same request.
There is a similar 4-channel board but with no markings and (as yet) no documentation. However all commands appear to work so a designation of R421A04 seems appropriate. Some typos in the 8-channel datasheet indicate that it was adapted from that of a 4-channel version.
R421A04 - 32 addresses set with DIP1-5, DIP6 ON for RTU mode R421A08 - 64 addresses set with DIP1-6, RTU mode only
Modbus/R421A/Request.h
Created on: 1 May 2018
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Variables
- constexpr uint8_t R421_CHANNEL_MIN = 1
- constexpr uint8_t R421A_MAX_CHANNELS = 16
- struct StateMask
Tracks state of multiple relays.
Public Members
- BitSet32 channelMask
Identifies valid channels.
- class Device : public IO::Modbus::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- inline virtual DevNode::ID nodeIdMin() const override
Get minimum valid Node ID for this device.
Typically devices have a contiguous valid range of node IDs
- inline virtual uint16_t maxNodes() const override
Determine maximum number of nodes supported by the device.
- Return values:
uint16_t – 0 if device doesn’t support nodes
- virtual DevNode::States getNodeStates(DevNode node) const override
Return the current set of states for all nodes controlled by this device.
Used to determine if, say, all nodes are ON, OFF or a combination.
- virtual void handleEvent(IO::Request *request, Event event) override
Implementations may override this method to customise event handling.
- struct Config
R421A device configuration.
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- class Request : public IO::Modbus::Request
Public Functions
- virtual ErrorCode parseJson(JsonObjectConst json) override
Fill this request from a JSON description.
- virtual void getJson(JsonObject json) const override
Get result of a completed request in JSON format.
- virtual bool setNode(DevNode node) override
If nodes are supported, implement this method.
/
/**
- virtual DevNode::States getNodeStates(DevNode node) override
Query node status from response.
RF Switch
Supports basic 433MHz AM transmitter attached to a GPIO pin via 5v buffer. You may be able to get away with connecting the GPIO directly and running from 3.3v, but timing and performance won’t be as good.
Uses hardware timer to generate PWM output using interrupts.
Developed for use with i-Lumos lightswitches which use a 24-bit code. Timing parameters are programmable though so may work with other devices.
- namespace RFSwitch
RFSwitch/Controller.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
RFSwitch/ControlleDevice.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
RFSwitch/Request.h
Created on: 1 May 2018
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
- class Controller : public IO::Controller
Controller for 433MHz transmitter.
Protocol is flexible but tested only with i-Lumos light switches. Written specifically for ESP8266 and uses the hardware timer to generate PWM signal via interrupts.
Public Functions
- inline virtual const FlashString &classname() const override
Get the class name for this Controller.
- virtual void handleEvent(IO::Request *request, Event event) override
Implementations override this method to process events as they pass through the stack.
- struct Timing
Protocol timings in microseconds.
- class Device : public IO::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- struct Config
- class Factory : public IO::Device::Factory
Public Functions
- inline virtual IO::Device *createDevice(IO::Controller &controller, const char *id) const override
Create a new device instance.
Called by
DeviceManager::createDevice()
- Parameters:
controller – The owning controller
id – Unique identifier for the device
- Return values:
Device* – The constructed instance
- inline virtual const FlashString &controllerClass() const override
Return the expected controller type for this device class, e.g. ‘rs485’.
The Device Manager uses this value to verify that devices are constructed using the correct controller.
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- class Request : public IO::Request
Public Functions
- virtual ErrorCode parseJson(JsonObjectConst json) override
Fill this request from a JSON description.
- virtual void getJson(JsonObject json) const override
Get result of a completed request in JSON format.
- inline virtual bool setNode(DevNode node) override
If nodes are supported, implement this method.
/
/**
RI-D35 Energy Meter
These are inexpensive and readily available MODBUS single-phase energy meters.
- namespace RID35
Modbus/RID35/Request.h
Created on: 24 March 2022
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Enums
- enum class Unit
Values:
- enumerator NONE
- enumerator KW
- enumerator KVAR
- enumerator KVA
- enumerator KWH
- enumerator KVARH
- enumerator KVAH
- enumerator VOLT
- enumerator AMP
- enumerator HERTZ
- enum class Register
Values:
- enumerator XX
- enumerator RID35_STDREG_MAP
Variables
- constexpr uint16_t stdRegBase = 0x01
- constexpr uint16_t ovfRegBase = 0x96
- constexpr size_t stdRegCount = (1 + unsigned(Register::MaxDemandApparentPower)) * 2
- constexpr size_t registerCount = stdRegCount + ovfRegCount
- class Device : public IO::Modbus::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
RS485
https://en.wikipedia.org/wiki/RS-485
RS485 is a very inexpensive way to wire devices together in a network. A single twisted pair (such as two wires from regular UTP network cable) at low speeds can operate over a kilometre.
Devices must be daisy-chained together, and long runs should be properly terminated typically with 120 ohm resistors at either end.
Connecting devices in star topology causes signal reflections and generally won’t work except for very short runs.
One solution is to use multiple MAX485 (or similar) transceivers to create multiple physical network segments. Only one UART is required unless you need concurrent requests.
This example allows for one transceiver to be in receive mode, with others in transmit. That means slave addresses must be unique across both segments. Resistors R2 and R8 are current limit resistors to guard against both transceivers being switched into receive mode at the same time.
The logic for controlling this would be in a callback registered via
IO::RS485::Controller::onSetDirection().For completeness, here’s example connections for a 3.3v transceiver:
R22 is optional, typically installed only on longer runs. D3 is additional transient protection - always a good idea as a first line of defence even if the transceiver itself has some built in.
- namespace RS485
RS485/Controller.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
RS485/Device.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
RS485/Request.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
- class Controller : public IO::Controller
Public Types
- using SetDirectionCallback = void (*)(uint8_t segment, Direction direction)
Callback to handle hardware transmit/receive selection Typically called from interrupt context so implementation MUST be marked IRAM_ATTR.
- Param segment:
Identifies physical connection for shared/multiplexed port
- Param direction:
Public Functions
- inline virtual const FlashString &classname() const override
Get the class name for this Controller.
- virtual void start() override
Start the controller.
Controllers may now initiate communications with registered devices. Typically they’ll query device status, etc.
- virtual void stop() override
Stop all controllers.
This method is called by the Device Manager, applications shouldn’t need it.
Note
MUST call canStop() first to ensure it’s safe to stop!
- inline void onSetDirection(SetDirectionCallback callback)
Set the transmit callback handler.
Typically with RS485 a GPIO is used to toggle between transmit/receive modes. Using a callback allows flexibility for your particular hardware implementation. You don’t need to set this up if your hardware handles the switch automatically.
- Parameters:
callback –
- inline void setDirection(IO::Direction direction)
Whilst a port is acquired, call this method to being or end transmission.
Note
Port should normally be left in receive mode on request completion.
- Parameters:
direction –
- class Device : public IO::Device
Base device class for communicating with an RS485 slave.
Subclassed by IO::DMX512::Device, IO::Modbus::Device
Public Functions
- inline virtual uint16_t address() const override
Devices with a numeric address should implement this method.
- virtual void handleEvent(IO::Request *request, Event event) override
Implementations may override this method to customise event handling.
- struct Config
RS485 configuration.
- struct Slave
Public Members
- uint16_t address
Network device address or ‘slave ID’
- uint8_t segment
Application-defined value for multiplexed serial ports. For example, several MAX485 transceivers can be connected to one serial port with appropriate multiplexing logic.
- unsigned baudrate
Serial link speed
- unsigned timeout
Max time between command/response in milliseconds.
- template<class DeviceClass>
class FactoryTemplate : public IO::Device::FactorySubclassed by IO::DMX512::Device::Factory, IO::Modbus::Device::Factory, IO::Modbus::NT18B07::Device::Factory, IO::Modbus::R421A::Device::Factory, IO::Modbus::RID35::Device::Factory, IO::Modbus::STM8Relay::Device::Factory, IO::Modbus::STS::Fan::Device::Factory
Public Functions
- inline virtual IO::Device *createDevice(IO::Controller &controller, const char *id) const override
Create a new device instance.
Called by
DeviceManager::createDevice()
- Parameters:
controller – The owning controller
id – Unique identifier for the device
- Return values:
Device* – The constructed instance
- inline virtual const FlashString &controllerClass() const override
Return the expected controller type for this device class, e.g. ‘rs485’.
The Device Manager uses this value to verify that devices are constructed using the correct controller.
- class Request : public IO::Request
Subclassed by IO::Modbus::Request
STM8S103 Relay Boards
General
These are MODBUS multi-channel relay boards with optocouplers and digital inputs. Advertised as using STM8S103F3 microcontroller, but no model number. 1, 2, 4 and 8-channel versions are available. Only 4-channel version has been tested.
There are a number of improvements over the cheaper boards (such as R421A):
Barrel jack power connector provided in addition to screw terminals
Relays are driven using optocouplers
Relay outputs use pluggable board connectors
PCB layout has good electrical isolation between relay outputs and coil side
Readable digital inputs - 3.3v logic-level, active low, connected directly to microcontroller.
Programming connections are made available so can be programmed with custom firmware.
Slave address default is 1, can be changed using custom commands (no DIP switches)
Device properties include:
- Address
The address of a modbus slave. Modbus docs. call this the slave ID.
- Node
Represents something a slave device does. Modbus relay boards have one node for each output it controls.
- Node ID
The channel a node lives on. In a modbus transaction this is the address field.
Commands
- ReadCoil, WriteCoil
Read or set state of Relays
- ReadDiscreteInputs
Read state of digital inputs. Bits are set to ‘1’ if input is pulled to ground.
- ReadHoldingRegister
Registers must be read individually. Attempting to read multiple registers returns invalid data. 0x0004 Software month as 2 characters, e.g. “Ap” 0x0008 Software year in BCD, e.g. 0x2018 for 2018 0x0010 Software time in BCD, e.g. 0x1459 for 14:59 hrs 0x0020 Hardware version, e.g. 0x006a 0x4000 Slave address, typically broadcast (to address #0)
- WriteHoldingRegister
0x4000 Set slave address, typically broadcast (to address #0)
- namespace STM8Relay
Modbus/STM8Relay/Device.h
Created on: 24 June 2022
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
STM8S103 modbus relay board
4-channel relay boards with optocouplers and digital inputs. Advertised as using STM8S103 microcontroller, but no model number. Relays controlled using Modbus ‘coil’ commands. Digital inputs read as ‘discrete inputs’. Connected directly to 3.3v microcontroller. Grounding pin sets bit.
Modbus/STM8Relay/Request.h
Copyright 2022 mikee47 mike@sillyhouse.net
This file is part of the IOControl Library
This library is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 or later.
This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this library. If not, see https://www.gnu.org/licenses/.
Variables
- constexpr uint8_t RELAY_CHANNEL_MIN = 1
- constexpr uint8_t RELAY_MAX_CHANNELS = 16
- struct StateMask
Tracks state of multiple relays.
Public Members
- BitSet32 channelMask
Identifies valid channels.
- class Device : public IO::Modbus::Device
Public Functions
- virtual IO::Request *createRequest() override
Create a request object for this device.
- Return values:
Request* – Caller must destroy or submit the request
- inline virtual DevNode::ID nodeIdMin() const override
Get minimum valid Node ID for this device.
Typically devices have a contiguous valid range of node IDs
- inline virtual uint16_t maxNodes() const override
Determine maximum number of nodes supported by the device.
- Return values:
uint16_t – 0 if device doesn’t support nodes
- virtual DevNode::States getNodeStates(DevNode node) const override
Return the current set of states for all nodes controlled by this device.
Used to determine if, say, all nodes are ON, OFF or a combination.
- virtual void handleEvent(IO::Request *request, Event event) override
Implementations may override this method to customise event handling.
- struct Config
R421A device configuration.
- class Factory : public IO::RS485::Device::FactoryTemplate<Device>
Public Functions
- inline virtual const FlashString &deviceClass() const override
Return the Device class name, e.g. ‘r421a’.
- class Request : public IO::Modbus::Request
Public Functions
- virtual ErrorCode parseJson(JsonObjectConst json) override
Fill this request from a JSON description.
- virtual void getJson(JsonObject json) const override
Get result of a completed request in JSON format.
- virtual bool setNode(DevNode node) override
If nodes are supported, implement this method.
/
/**
- virtual DevNode::States getNodeStates(DevNode node) override
Query node status from response.
References
Used by
Basic RS485 Sample
Fan Controller Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
IRremoteESP8266 Library
This library enables you to send *and* receive infra-red signals on an ESP8266 or an ESP32 using the Arduino framework using common 940nm IR LEDs and common IR receiver modules. e.g. TSOP{17,22,24,36,38,44,48}* demodulators etc.
v2.7.7 Now Available
Version 2.7.7 of the library is now available. You can view the Release Notes for all the significant changes.
Upgrading from pre-v2.0
Usage of the library has been slightly changed in v2.0. You will need to change your usage to work with v2.0 and beyond. You can read more about the changes required on our Upgrade to v2.0 page.
Upgrading from pre-v2.5
The library has changed from using constants declared as #define to
const with
the appropriate naming per the
C++ style guide.
This may potentially cause old programs to not compile.
The most likely externally used #defines have been aliased for limited
backward compatibility for projects using the old style. Going forward, only the
new kConstantName style will be supported for new protocol additions.
In the unlikely case, it does break your code, then you may have been referencing
something you likely should not have. You should be able to quickly determine
the new name from the old. e.g. CONSTANT_NAME to kConstantName.
Use common sense or examining the library’s code if this does affect code.
Supported Protocols
You can find the details of which protocols & devices are supported here.
Troubleshooting
Before reporting an issue or asking for help, please try to follow our Troubleshooting Guide first.
Frequently Asked Questions
Some common answers to common questions and problems are on our F.A.Q. wiki page.
Library API Documentation
This library uses Doxygen to automatically document the library’s API. You can find it here.
Installation
Click the “Sketch” -> “Include Library” -> “Manage Libraries…” Menu items.
Enter
IRremoteESP8266into the “Filter your search…” top right search box.Click on the IRremoteESP8266 result of the search.
Select the version you wish to install and click “Install”.
Click on “Clone or Download” button, then “`Download ZIP <https://github.com/crankyoldgit/IRremoteESP8266/archive->master.zip>`_“ on the page.
Extract the contents of the downloaded zip file.
Rename the extracted folder to “IRremoteESP8266”.
Move this folder to your libraries directory. (under windows:
C:\Users\YOURNAME\Documents\Arduino\libraries\)Restart your Arduino IDE.
Check out the examples.
cd ~/Arduino/libraries
git clone https://github.com/crankyoldgit/IRremoteESP8266.git
cd ~/Arduino/libraries/IRremoteESP8266 && git pull
Contributing
If you want to contribute to this project, consider:
Reporting bugs and errors
Ask for enhancements
Improve our documentation
Tell other people about this library
Contributors
Available here
Library History
This library was originally based on Ken Shirriff’s work (https://github.com/shirriff/Arduino-IRremote/)
Mark Szabo has updated the IRsend class to work on ESP8266 and Sebastien Warin the receiving & decoding part (IRrecv class).
As of v2.0, the library was almost entirely re-written with the ESP8266’s resources in mind.
References
Used by
IR Library Sample
SoC support
esp8266
Nextion Serial Displays
Introduction
Nextion Arduino library provides an easy-to-use way to manipulate Nextion serial displays. Users can use the libarry freely, either in commercial projects or open-source prjects, without any additional condiitons.
To get your Nextion display, please visit iMall.
To discuss the project? Request new features? Report a BUG? please visit the Forums
Download Source Code
Latest version is unstable and a mass of change may be applied in a short time without any notification for users. Commonly, it is for developers of this library.
Release version is recommended for you, unless you are one of developers of this library.
Release notes is at https://github.com/itead/ITEADLIB_Arduino_Nextion/blob/master/release_notes.md.
Latest(unstable)
Latest source code(master branch) can be downloaded: https://github.com/itead/ITEADLIB_Arduino_Nextion/archive/master.zip.
You can also clone it via git:
git clone https://github.com/itead/ITEADLIB_Arduino_Nextion
Releases(stable)
https://github.com/itead/ITEADLIB_Arduino_Nextion/archive/v0.7.0.zip
https://github.com/itead/ITEADLIB_Arduino_Nextion/archive/v0.7.0.tar.gz
All releases can be available from: https://github.com/itead/ITEADLIB_Arduino_Nextion/releases.
Documentation
Offline Documentation’s entry doc/Documentation/index.html shipped
with source code can be open in your browser such as Chrome, Firefox or
any one you like.
Supported Mainboards
All boards, which has one or more hardware serial, can be supported.
For example:
Iteaduino MEGA2560
Iteaduino UNO
Arduino MEGA2560
Arduino UNO
Configuration
In configuration file NexConfig.h, you can find two macros below:
dbSerial: Debug Serial (baudrate:9600), needed by beginners for debug your nextion applications or sketches. If your complete your work, it will be a wise choice to disable Debug Serial.
nexSerial: Nextion Serial, the bridge of Nextion and your mainboard.
Note: the default configuration is for MEGA2560.
Redirect dbSerial and nexSerial
If you want to change the default serial to debug or communicate with Nextion , you need to modify the line in configuration file:
#define dbSerial Serial ---> #define dbSerial Serialxxx
#define nexSerial Serial2 ---> #define nexSeria Serialxxx
Disable Debug Serial
If you want to disable the debug information,you need to modify the line in configuration file:
#define DEBUG_SERIAL_ENABLE ---> //#define DEBUG_SERIAL_ENABLE
UNO-like Mainboards
If your board has only one hardware serial, such as UNO, you should
disable dbSerial and redirect nexSerial to Serial(Refer to
section:Serial configuration).
Useful Links
http://blog.iteadstudio.com/nextion-tutorial-based-on-nextion-arduino-library/
License
The MIT License (MIT)
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
References
Used by
Nextion Button Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Liquid Crystal
Introduction
Welcome to the LCD Library for Arduino and Chipkit. It is a derivate of the original LiquidCrystal Library as sourced in the Arduino SDK. It has been developed to be compatible with the current LiquidCrystal library, its performance is almost 5 times faster and fully extendable if need be.
It supports most Hitachi HD44780 based LCDs, or compatible, connected to any project using: 4, 8
wire parallel interface, I2C IO port expander (native I2C and bit bang) and Shift Register.
It currently supports 4 types of connections:
4 bit parallel LCD interface
8 bit parallel LCD interface
I2C IO bus expansion board with the PCF8574* I2C IO expander ASIC such as I2C LCD extra IO.
ShiftRegister adaptor board as described Shift Register project home or in the HW configuration described below, 2 and 3 wire configurations supported.
ShiftRegister 3 wire latch adaptor board as described ShiftRegister 3 Wire Home
Support for 1 wire shift register ShiftRegister 1 Wire
I2C bus expansion using general purpose IO lines.
How do I get set up?
Please refer to the project’s wiki
Contributors
The library has had the invaluable contribution of:
piccaso - Florian Fida - Flo, thanks for testing and improving the SR library, initial version of the 1 wire interface and speed improvements.
Perry - bperrybap@opensource.billsworld.billandterrie.com, with his thoughtful contribution, speed improvements and development support for the SR2W library.
Adrian Piccioli, with his contribution to the i2c GPIO support.
todbot Tod E. Kurt for the softwarei2cmaster library.
felias-fogg - Bernhard for the softwarei2cmaster fast
Contribution guidelines
Writing tests
Code review
Help out with bug fixing
Setup a project logo
Write new drivers to support more LCDs.
Who do I talk to?
Repo owner or admin
For SoftI2CMaster latest versions, updates and support, please refer to SoftI2CMaster
References
Used by
Liquid Crystal 44780 Sample
MeteoControl Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
LittleFS
Sming IFS integration of the LittleFS filesystem https://github.com/littlefs-project/littlefs.
References
Used by
Basic FlashIP Sample
Fan Controller Sample
LittleFS inspector Sample
Basic IFS Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: `littlefs <>`__
littlefs
A little fail-safe filesystem designed for microcontrollers.
| | | .---._____
.-----. | |
--|o |---| littlefs |
--| |---| |
'-----' '----------'
| | |
Power-loss resilience - littlefs is designed to handle random power failures. All file operations have strong copy-on-write guarantees and if power is lost the filesystem will fall back to the last known good state.
Dynamic wear leveling - littlefs is designed with flash in mind, and provides wear leveling over dynamic blocks. Additionally, littlefs can detect bad blocks and work around them.
Bounded RAM/ROM - littlefs is designed to work with a small amount of memory. RAM usage is strictly bounded, which means RAM consumption does not change as the filesystem grows. The filesystem contains no unbounded recursion and dynamic memory is limited to configurable buffers that can be provided statically.
Example
Here’s a simple example that updates a file named boot_count every time
main runs. The program can be interrupted at any time without losing track
of how many times it has been booted and without corrupting the filesystem:
#include "lfs.h"
// variables used by the filesystem
lfs_t lfs;
lfs_file_t file;
// configuration of the filesystem is provided by this struct
const struct lfs_config cfg = {
// block device operations
.read = user_provided_block_device_read,
.prog = user_provided_block_device_prog,
.erase = user_provided_block_device_erase,
.sync = user_provided_block_device_sync,
// block device configuration
.read_size = 16,
.prog_size = 16,
.block_size = 4096,
.block_count = 128,
.cache_size = 16,
.lookahead_size = 16,
.block_cycles = 500,
};
// entry point
int main(void) {
// mount the filesystem
int err = lfs_mount(&lfs, &cfg);
// reformat if we can't mount the filesystem
// this should only happen on the first boot
if (err) {
lfs_format(&lfs, &cfg);
lfs_mount(&lfs, &cfg);
}
// read current count
uint32_t boot_count = 0;
lfs_file_open(&lfs, &file, "boot_count", LFS_O_RDWR | LFS_O_CREAT);
lfs_file_read(&lfs, &file, &boot_count, sizeof(boot_count));
// update boot count
boot_count += 1;
lfs_file_rewind(&lfs, &file);
lfs_file_write(&lfs, &file, &boot_count, sizeof(boot_count));
// remember the storage is not updated until the file is closed successfully
lfs_file_close(&lfs, &file);
// release any resources we were using
lfs_unmount(&lfs);
// print the boot count
printf("boot_count: %d\n", boot_count);
}
Usage
Detailed documentation (or at least as much detail as is currently available) can be found in the comments in lfs.h.
littlefs takes in a configuration structure that defines how the filesystem operates. The configuration struct provides the filesystem with the block device operations and dimensions, tweakable parameters that tradeoff memory usage for performance, and optional static buffers if the user wants to avoid dynamic memory.
The state of the littlefs is stored in the lfs_t type which is left up
to the user to allocate, allowing multiple filesystems to be in use
simultaneously. With the lfs_t and configuration struct, a user can
format a block device or mount the filesystem.
Once mounted, the littlefs provides a full set of POSIX-like file and directory functions, with the deviation that the allocation of filesystem structures must be provided by the user.
All POSIX operations, such as remove and rename, are atomic, even in event of power-loss. Additionally, file updates are not actually committed to the filesystem until sync or close is called on the file.
Other notes
Littlefs is written in C, and specifically should compile with any compiler
that conforms to the C99 standard.
All littlefs calls have the potential to return a negative error code. The
errors can be either one of those found in the enum lfs_error in
lfs.h, or an error returned by the user’s block device operations.
In the configuration struct, the prog and erase function provided by the
user may return a LFS_ERR_CORRUPT error if the implementation already can
detect corrupt blocks. However, the wear leveling does not depend on the return
code of these functions, instead all data is read back and checked for
integrity.
If your storage caches writes, make sure that the provided sync function
flushes all the data to memory and ensures that the next read fetches the data
from memory, otherwise data integrity can not be guaranteed. If the write
function does not perform caching, and therefore each read or write call
hits the memory, the sync function can simply return 0.
Design
At a high level, littlefs is a block based filesystem that uses small logs to store metadata and larger copy-on-write (COW) structures to store file data.
In littlefs, these ingredients form a sort of two-layered cake, with the small logs (called metadata pairs) providing fast updates to metadata anywhere on storage, while the COW structures store file data compactly and without any wear amplification cost.
Both of these data structures are built out of blocks, which are fed by a common block allocator. By limiting the number of erases allowed on a block per allocation, the allocator provides dynamic wear leveling over the entire filesystem.
root
.--------.--------.
| A'| B'| |
| | |-> |
| | | |
'--------'--------'
.----' '--------------.
A v B v
.--------.--------. .--------.--------.
| C'| D'| | | E'|new| |
| | |-> | | | E'|-> |
| | | | | | | |
'--------'--------' '--------'--------'
.-' '--. | '------------------.
v v .-' v
.--------. .--------. v .--------.
| C | | D | .--------. write | new E |
| | | | | E | ==> | |
| | | | | | | |
'--------' '--------' | | '--------'
'--------' .-' |
.-' '-. .-------------|------'
v v v v
.--------. .--------. .--------.
| F | | G | | new F |
| | | | | |
| | | | | |
'--------' '--------' '--------'
More details on how littlefs works can be found in DESIGN.md and SPEC.md.
Testing
The littlefs comes with a test suite designed to run on a PC using the emulated block device found in the emubd directory. The tests assume a Linux environment and can be started with make:
make test
License
The littlefs is provided under the BSD-3-Clause license. See LICENSE.md for more information. Contributions to this project are accepted under the same license.
Individual files contain the following tag instead of the full license text.
SPDX-License-Identifier: BSD-3-Clause
This enables machine processing of license information based on the SPDX License Identifiers that are here available: http://spdx.org/licenses/
MCP23008 Port Expander
Library for communicating with MCP23008 8-bit port expander. Created by Garrett Blanton January, 24, 2014. Released into the public domain.
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MCP23017 I2C Port Expander
https://github.com/adafruit/Adafruit-MCP23017-Arduino-Library.git
This is a library for the MCP23017 I2c Port Expander
These chips use I2C to communicate, 2 pins required to interface
Adafruit invests time and resources providing this open source code, please support Adafruit and open-source hardware by purchasing products from Adafruit!
Written by Limor Fried/Ladyada for Adafruit Industries.
BSD license, check license.txt for more information
All text above must be included in any redistribution
To download. click the DOWNLOADS button in the top right corner, rename the uncompressed folder Adafruit_MCP23017. Check that the Adafruit_MCP23017 folder contains Adafruit_MCP23017.cpp and Adafruit_MCP23017.h
Place the Adafruit_MCP23017 library folder your
References
Used by
MCP23017 I2C Port Expander Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MCP23S17 SPI Port Expander
https://github.com/n0mjs710/MCP23S17.git
Arduino Driver for Microchip MCP23S17
MCP23S17 Class for Arduino
Introduction:
This class is written to simplify using the Microchip MCP23S17 general purpose I/O expander IC in the Arduino environment. Some understanding of the MCP23S17 is required, so if you are not familiar with it, download the datasheet for it and have a look. The rest of this description will assume a basic understanding of the chip.
Implementation:
The goal of this implementation is to provide a software interface that mimics the existing Arduino I/O functions:
pinMode(pin, mode)
digitalWrite(pin, value)
digitalRead(pin)
The class does include several more methods that can be used to simplify configuration in the same “Arduino-ish” way, methods for writing/reading 8-bit registers (configuration and I/O ports) at once, as well as writing/reading consecutive registers (allowing all 16 bits to be read or written with one method call). The interrupt features of the chip are not directly supported with method for specifically configuring them, however, the byte and word read/write methods may be used to configure and use the interrupt features. These features can get somewhat complicated, and any user prepared to use them will likely prefer the more generic methods for controlling them.
Upon initialization of an MCP23S17 as an object, ALL MCP23S17s on the SPI bus (sharing the same slave select) will be placed into hardware addressing mode. This allows up to 8 MCP23S17s to be used with a single slave select.
**Methods:**
MCP()
Description
Instantiate an MCP23S17 device as an object.
Syntax
MCP object_name(address, slave_select)
Parameters
object_name: any arbitrary name given to create the object
address: address (0-7) of the device configured with address (pins A0, A1, A2)
slave_select: a valid *Arduino* pin number.
Returns
none
Example
MCP onechip(1, 10); // create an object at address 1 called "onechip", using pin 10 as slave select
MCP twochip(2, 10); // create an object at address 2 called "twochip", using pin 10 as slave select
pinMode()
Description
Configure pin(s) as either input or output on the selected object (device specified by an address)
Syntax
object.name.pinMode(pin, mode);
or
object.name.pinMode(mode);
Parameters
object_name: the name given when this object was created
pin: the pin number (1-16) on which to set as input or output
mode: if a pin is specified, either a "HIGH" (1) for input (default) or a "LOW" (0) for output. If a pin is not specified, mode should be a word indicating the mode of each of the 16 I/O pins on the chip.
Returns
none
Example
void setup() {
onechip.pinMode(4, HIGH); // sets pin 4 as an input
onechip.pinMode(16, LOW); // sets pin 16 as an output
twochip.pinMode(0B0000111100001111); // sets pins 1-4 and 9-12 as input, 5-8 and 13-16 as output
}
pullupMode()
Description
Configure the weak pull-up resistors on pins defined as inputs
This has no effect on pins that are configured as outputs.
Syntax
object_name.pullupMode(pin, mode);
or
object_name.pullupMode(mode);
Parameters
object_name: the name given when this object was created
pin: the pin number (1-16) on which to enable or disable the internal weak pull-up resistor
mode: if a pin is specified, either "HIGH" (1) to enable or "LOW" (0) to disable (default) the weak pull-up resistor on the specified pin. If a pin is not specified, mode should be a word indicating the pull-up mode of each of the 16 pins on the chip. Configuring pull-up has no effect on pins while they are configured as output.
Returns
none
Example
void setup() {
onechip.pullupMode(4, HIGH); // enable the pull-up on pin 4
twochip.pullupMode(0B0000111100000000); // enable the pull-ups on pins 9-12
}
inputInvert()
Description
Configure inversion on pins configured as inputs.
This will cause an inverted input pin to read as “LOW” (0) when it is actually in a high state, or as “HIGH” (1) when it is actually in a low state. This has no effect on pins that are configured as outputs.
Syntax
object_name.inputInvert(pin, inversion);
or
object_name.inputInvert(inversion);
Parameters
object_name: the name given when this object was created
pin: the pin number (1-16) on which to set or clear inversion
inversion: if a pin is specified, either "HIGH" (1) is specified to enable, or "LOW" (0) to disable (default) inversion on the specified pin. If a pin is not specified, mode should be a word indicating the inversion state of each of the 16 pins on the chip. Configuring inversion has no effect on pins while they are configured as output.
Returns
none
Example
void setup() {
onechip.inputInvert(4, LOW); // disable inversion on pin 4
twochip.inputInvert(0B0000000000001111); // enable inversion on pins 1-4
}
digitalWrite()
Description
Write a “HIGH” or “LOW” value to a digital I/O pin(s)
Syntax
object_name.digitalWrite(pin, value);
or
object_name.digitalWrite(value);
Parameters
object_name: the name given when this object was created
pin: the pin number (1-16) who's value will be set
value: if a pin is specified, either a "HIGH" (1) or a "LOW" (0) value may be set on the specified pin. If a pin is not specified, value should be a word indicating the output state of all 16 pins on the device. Writing pins configured as inputs has no effect.
Returns
none
Example
void loop() {
onechip.digitalWrite(16, HIGH); // set pin 16 to "HIGH"
twochip.digitalWrite(0B1100000000110000); // Set 5, 6, 15 & 16 to high, 7,8, 13 & 14 to low - inputs ignored
}
digitalRead()
Description
Reads the value of input pin(s), either “HIGH” (“1”) or “LOW” (“0)
Syntax
object_name.digitalRead(pin);
or
object_name.digitalRead();
Parameters
object_name: the name given when this object was created
pin: the pin number (1-16) who's value will be read. If no pin number is supplied, a word will be read containing the input state of all pins. The values for pins configured as output should be disregarded if the "word-mode" version is used.
Returns
"HIGH" (1) or "LOW" (0) if a pin is supplied. a word (16 bits) is returned if no pin argument is given
Example
void loop() {
int onevalue;
int twovalue;
onevalue = onechip.digitalRead(4); // assigns the value of pin4 to onevalue
twovalue = twochip.digitalRead(); // assigns the value of all 16 I/O pins to twovalue
}
wordWrite()
Description:
This is an advanced method to write a register pair in the MCP23S17. This class operates the MCP23S17 in “BANK=0” mode. The intention is that a registers for both ports may be written by supplying a single word as an argument. The low byte is written to the register address supplied, and the high byte to the next higher register address.
Syntax”
object_name.wordWrite(base register, value);
Parameters:
object_name: the name given when this object was created
base register: the beginning register address to write, for example, if 0x02 is given, the low byte of "value" will be written to 0x02 and the high byte of "value" to the register at 0x03
value: a word (unsigned int) that will be broken into two bytes and written to two consecutive registers, starting with the "base register" address
Returns:
none
Example
void loop() {
onechip.wordWrite(0x12, 0xFF00); // Set GPIOA to 0x00 and GPIOB to OxFF
}
byteWrite()
Description:
This is an advanced method to write any single register in the MCP23S17.
Syntax:
object_name.byteWrite(register, value);
Parameters:
object_name: the name given when this object was created
register: the register address to write
value: a byte (unsigned char) that will be written to the specified registers
Returns:
none
Example:
void loop() {
twochip.byteWrite(0x13, 0xF0); // Set GPIOB (portB) to 0xF0 (0B11110000)
}
byteRead()
Description:
This is an advanced method to read any single register in the MCP23S17.
Syntax:
object_name.byteRead(register);
Parameters:
object_name: the name given when this object was created
register: the register address to be read
Returns:
unsigned char (uint8_t)
Example:
void loop() {
int twovalue;
twovalue = twochip.byteRead(0x12); // Read GPIOA (portA)
}
Full Example:
MCP onechip(1, 10); // create an object at address 1 called "onechip", using pin 10 as slave select
MCP twochip(2, 10); // create an object at address 2 called "twochip", using pin 10 as slave select
void setup() {
onechip.pinMode(4, HIGH); // sets pin 4 as an input
onechip.pinMode(16, LOW); // sets pin 16 as an output
twochip.pinMode(0B0000111100001111); // sets pins 1-4 and 9-12 as input, 5-8 and 13-16 as output
onechip.pullupMode(4, HIGH); // enable the pull-up on pin 4
twochip.pullupMode(0B0000111100000000); // enable the pull-ups on pins 9-1
onechip.inputInvert(4, LOW); // disable inversion on pin 4
twochip.inputInvert(0B0000000000001111); // enable inversion on pins 1-4
}
void loop() {
int onevalue;
int twovalue;
onechip.digitalWrite(16, HIGH); // set pin 16 to "HIGH"
twochip.digitalWrite(0B1100000000110000); // Set 5, 6, 15 & 16 to high, 7,8, 13 & 14 to low - inputs ignored
onevalue = onechip.digitalRead(4); // assigns the value of pin4 to onevalue
twovalue = twochip.digitalRead(); // assigns the value of all 16 I/O pins to twovalue
/* These are for context only - use them only if you really know what you're doing
onechip.wordWrite(0x12, 0xFF00); // Set GPIOA to 0x00 and GPIOB to OxFF
twochip.byteWrite(0x13, 0xF0); // Set GPIOB (portB) to 0xF0 (0B11110000)
twovalue = twochip.byteRead(0x12); // Read GPIOA (portA)
*/
}
References
Used by
MCP23S17 SPI Port Expander Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MCP_CAN Library for Arduino
MCP_CAN library v1.5 This library is compatible with any shield or board that uses the MCP2515 or MCP25625 CAN protocol controller.
This version supports setting the ID filter mode of the protocol controller, the BAUD rate with clock speed with the begin() function. Baudrates 5k, 10k, 20k, 50k, 100k, 125k, 250k, 500k, & 1000k using 16MHz clock on the MCP2515 are confirmed to work using a Peak-System PCAN-USB dongle as a reference. Baudrates for 8MHz and 20MHz crystals are yet to be confirmed but were calculated appropriately.
The readMsgBuf() functions bring in the message ID. The getCanId() function is obsolete and no longer exists, don’t use it.
The readMsgBuf(*ID, *DLC, *DATA) function will return the ID type (extended or standard) and it will bring back the remote request status bit.
If the ID AND 0x80000000 EQUALS 0x80000000, the ID is of the Extended type, otherwise it is standard.
If the ID AND 0x40000000 EQUALS 0x40000000, the message is a remote request.
The readMsgBuf(*ID, *EXT, *DLC, *DATA) function will return the ID unaltered and doesn’t inform us of a remote request.
If EXT is true, the ID is extended.
The sendMsgBuf(ID, DLC, DATA) function can send extended or standard IDs.
To mark an ID as extended, OR the ID with 0x80000000.
To send a remote request, OR the ID with 0x40000000.
The sendMsgBuf(ID, EXT, DLC, DATA) has not changed other than fixing return values.
Using the setMode() function the sketch can now put the protocol controller into sleep, loop-back, or listen-only modes as well as normal operation. Right now the code defaults to loop-back mode after the begin() function runs. I have found this to increase the stability of filtering when the controller is initialized while connected to an active bus.
User can enable and disable (default) One-Shot transmission mode from the sketch using enOneShotTX() or disOneShotTX() respectively.
To wake up from CAN bus activity while in sleep mode enable the wake up interrupt with setSleepWakeup(1). Passing 0 will disable the wakeup interrupt (default).
Installation
Copy this into the “[…/MySketches/]libraries/” folder and restart the Arduino editor.
NOTE: If an older version of the library exists (e.g. CAN_BUS_Shield) be sure to remove it from the libraries folder or replace the files with those in this library to avoid conflicts.
Help and Support
This is primarily for non-bug related issues: Please start a new thread in an appropriate area at Seeedstudio forums or Arduino.cc forums and then send me (coryjfowler) a link through the PM system, my user name is the same as it is here. I will receive an email about the PM and generally get to it with-in a week or less. Keep in mind, I do this in my spare time.
Happy Coding!
References
Used by
CANBUS Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
mDNS: Multicast Domain Name System
https://en.wikipedia.org/wiki/Multicast_DNS
Responder
Sming provides the mDNS::Responder class to allow applications
to advertise themselves on the local network.
To use:
Add
COMPONENT_DEPENDS += MDNSto your application componenent.mk file.Add these lines to your application:
#include <Mdns/Responder.h> static mDNS::Responder responder; // Call when IP address has been obtained void startmDNS() { responder.begin(F("myhostname"); }
This will advertise the device as myhostname.local.
To provide a custom service, implement a mDNS::Service class
and call mDNS::Responder::addService().
See the UDP Server mDNS sample application.
Discovery
This library also provides support for device discovery using a separate set of classes,
based on the mDNS::Server.
See Basic MDNS for an example.
Testing
For linux, you can use avahi to perform mDNS queries and confirm output is as expected:
sudo apt install avahi
avahi-browse --all -r
References
Multicast DNS RFC6762 https://tools.ietf.org/html/rfc6762
Zero-configuration networking (DNS-SD) https://en.wikipedia.org/wiki/Zero-configuration_networking
DNS-Based Service Discovery https://tools.ietf.org/html/rfc6763
DNS record types https://en.wikipedia.org/wiki/List_of_DNS_record_types
Domain Names: Implementation and Specification https://tools.ietf.org/html/rfc1035
API Documentation
-
namespace mDNS
-
Functions
Variables
-
constexpr uint32_t MDNS_IP = {0xFB0000E0}
-
constexpr uint16_t MDNS_TARGET_PORT = {5353}
-
constexpr uint16_t MDNS_SOURCE_PORT = {5353}
-
constexpr uint16_t MDNS_TTL = {255}
-
constexpr uint16_t MAX_PACKET_SIZE = {1024}
-
class Answer : public LinkedObjectTemplate<Answer>
- #include <Answer.h>
-
class Handler : public LinkedObjectTemplate<Handler>
- #include <Handler.h>
Virtual base class used for chaining message handlers.
Subclassed by mDNS::Responder
-
class Message
- #include <Message.h>
Encapsulates a message packet for flexible introspection.
Subclassed by mDNS::Request
-
class Name
- #include <Name.h>
Encoded DNS name.
mDNS-SD names are represented as instance.service.domain. See https://tools.ietf.org/html/rfc6763#section-4.1
Instance names should be friendly and not attempt to be unique. See https://tools.ietf.org/html/rfc6763#appendix-D
Example: “UDP Server._http._tcp.local” instance: “UDP Server” service: “_http._tcp” name: “http” protocol: Protocol::Tcp domain: “local”
-
class Question : public LinkedObjectTemplate<Question>
- #include <Question.h>
-
class Request : public mDNS::Message
- #include <Request.h>
Subclassed by mDNS::Query, mDNS::Reply
-
class Responder : public mDNS::Handler
- #include <Responder.h>
Special name for querying list of services.
-
class Server : protected UdpConnection
- #include <Server.h>
Locates mDNS services by issuing queries.
-
class Service : public LinkedObjectTemplate<Service>
- #include <Service.h>
Describes a basic service.
The default methods translate to a DNS-SD name of “Sming._http._tcp.local”. See :cpp:class:
mDNS::Namefor a description of how names are defined.
-
namespace Resource
-
-
class Record
- #include <Resource.h>
Resource Record with no specific type.
Subclassed by mDNS::Resource::A, mDNS::Resource::AAAA, mDNS::Resource::HINFO, mDNS::Resource::PTR, mDNS::Resource::SRV, mDNS::Resource::TXT
-
class PTR : public mDNS::Resource::Record
- #include <Resource.h>
‘PTR’ record containing pointer to a canonical name
-
class HINFO : public mDNS::Resource::Record
- #include <Resource.h>
‘HINFO’ record containing Host information
-
class TXT : public mDNS::Resource::Record
- #include <Resource.h>
‘TXT’ record containing attribute list
Originally for arbitrary human-readable text in a DNS record. Content is a set of name=value pairs. Value can be binary.
-
class Record
-
constexpr uint32_t MDNS_IP = {0xFB0000E0}
References
Used by
GoogleCast Library
Basic MDNS Sample
UDP Server mDNS Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MFRC522 RFID Module
Note
This is a copy of the readme from 12/6/19, however the code in this library was last updated 23/3/2017.
Arduino library for MFRC522 and other RFID RC522 based modules.
Read and write different types of Radio-Frequency IDentification (RFID) cards on your Arduino using a RC522 based reader connected via the Serial Peripheral Interface (SPI) interface.
Development
The development by owner miguelbalboa has ended.
Feature status: complete freeze; no function or API change
Code status: partial freeze; just fixes/typos or documentation updates; no extensions for other boards; no new examples
Maintenance status: sporadically
Why no further development? This library has a long history and is used in many projects. This projects often do not document what version they use. Committing changes maybe brake those old project and lead to bad experience (for beginners) and support request. For those reasons the library is in freeze mode. You can still commit typo, documentation or bug fixes.
Before buy
Please notice that there are many sellers (ebay, aliexpress, ..) who sell mfrc522 boards. The quality of these boards are extremely different. Some are soldered with wrong/low quality capacitors or fake/defect mfrc522.
Please consider buying several devices from different suppliers. So the chance of getting a working device is higher.
If you got a bad board and you can tell us how to detect those boards (silk, chip description, ..), please share your knowledge.
What works and not?
Works
Communication (Crypto1) with MIFARE Classic (1k, 4k, Mini).
Communication (Crypto1) with MIFARE Classic compatible PICCs.
Firmware self check of MFRC522.
Set the UID, write to sector 0, and unbrick Chinese UID changeable MIFARE cards.
Manage the SPI chip select pin (aka SS, SDA)
Works partially
Communication with MIFARE Ultralight.
Other PICCs (Ntag216).
More than 2 modules, require a multiplexer #191.
Doesn’t work
MIFARE DESFire, MIFARE DESFire EV1/EV2, not supported by software.
Communication with 3DES or AES, not supported by software.
Peer-to-peer (ISO/IEC 18092), not supported by hardware.
Communication with smart phone, not supported by hardware.
Card emulation, not supported by hardware.
Use of IRQ pin. But there is a proof-of-concept example.
With Arduino Yun see #111, not supported by software.
With Intel Galileo (Gen2) see #310, not supported by software.
Power reduction modes #269, not supported by software.
I2C instead of SPI #240, not supported by software.
UART instead of SPI #281, not supported by software.
Need more?
If software: code it and make a pull request.
If hardware: buy a more expensive like PN532 (supports NFC and many more, but costs about $15 and not usable with this library).
Compatible IDE
This library works with Arduino IDE 1.6, older versions are not supported and will cause compiler errors. The built-in library manager is supported.
If you use your own compiler, you have to enable c++11-support.
Compatible boards
!!!Only for advanced users!!!
This library is compatible with the Teensy and ESP8266 if you use the board plugin of the Arduino IDE. Not all examples are available for every board. You also have to change pins. See pin layout.
Some user made some patches/suggestions/ports for other boards:
ESP8266 (native): https://github.com/miguelbalboa/rfid/pull/235
LPCOPen (in C): https://github.com/miguelbalboa/rfid/pull/258
Note that the main target/support of library is still Arduino.
Support/issue
First checkout what works and not and troubleshooting .
- It seems to be a hardware issue or you need support to program your project?
Please ask in the official Arduino forum, where you would get a much faster answer than on Github.
- It seems to be a software issue?
Open an issue on Github.
Code style
Please use fixed integers, see stdint.h. Why? This library is compatible with different boards which use different architectures (16bit and 32bit.) Unfixed int variables have different sizes in different environments and may cause unpredictable behaviour.
Pin Layout
The following table shows the typical pin layout used:
PCD MFRC522 |
Arduino Uno / 101 |
Teensy Mega |
Nano v3 |
Leonardo / Micro |
Pro Micro |
2.0 |
++ 2.0 |
3.1 |
|
|---|---|---|---|---|---|---|---|---|---|
Signal |
Pin |
Pin |
Pin |
Pin |
Pin |
Pin |
Pin |
Pin |
Pin |
RST/Reset |
RST |
9 [1] |
5 [1] |
D9 |
RESET / ICSP-5 |
RST |
7 |
4 |
9 |
SPI SS |
SDA [3] |
10 [2] |
53 [2] |
D10 |
10 |
10 |
0 |
20 |
10 |
SPI MOSI |
MOSI |
11 / ICSP-4 |
51 |
D11 |
ICSP-4 |
16 |
2 |
22 |
11 |
SPI MISO |
MISO |
12 / ICSP-1 |
50 |
D12 |
ICSP-1 |
14 |
3 |
23 |
12 |
SPI SCK |
SCK |
13 / ICSP-3 |
52 |
D13 |
ICSP-3 |
15 |
1 |
21 |
13 |
ESP8266 Wemos D1 mini |
|
|---|---|
Signal |
Pin |
RST/Reset |
D3 |
SPI SS |
D8 |
SPI MOSI |
D7 |
SPI MISO |
D6 |
SPI SCK |
D5 |
Hardware
There are three hardware components involved:
Micro Controller:
An Arduino or compatible executing the Sketch using this library.
Prices vary from USD 7 for clones, to USD 75 for “starter kits” (which might be a good choice if this is your first exposure to Arduino; check if such kit already includes the Arduino, Reader, and some Tags).
Proximity Coupling Device (PCD):
The PCD is the actual RFID Reader based on the NXP MFRC522 Contactless Reader Integrated Circuit.
Readers can be found on eBay for around USD 5: search for “rc522”.
You can also find them on several web stores. They are often included in “starter kits”, so check your favourite electronics provider as well.
Proximity Integrated Circuit Card (PICC):
The PICC is the RFID Card or Tag using the ISO/IEC 14443A interface, for example Mifare or NTAG203.
One or two might be included with the Reader or “starter kit” already.
Protocols
The micro controller and the reader use SPI for communication.
The protocol is described in the NXP MFRC522 datasheet.
See the Pin Layout section for details on connecting the pins.
The reader and the tags communicate using a 13.56 MHz electromagnetic field.
The protocol is defined in ISO/IEC 14443-3:2011 Part 3 Type A.
Details are found in chapter 6 “Type A – Initialization and anticollision”.
See http://wg8.de/wg8n1496_17n3613_Ballot_FCD14443-3.pdf for a free version of the final draft (which might be outdated in some areas).
The reader does not support ISO/IEC 14443-3 Type B.
Security
The UID of a card can not be used as an unique identification for security related projects. Some Chinese cards allow to change the UID which means you can easily clone a card. For projects like access control, door opener or payment systems you must implement an additional security mechanism like a password or normal key.
This library only supports crypto1-encrypted communication. Crypto1 has been known as broken for a few years, so it does NOT offer ANY security, it is virtually unencrypted communication. Do not use it for any security related applications!
This library does not offer 3DES or AES authentication used by cards like the Mifare DESFire, it may be possible to be implemented because the datasheet says there is support. We hope for pull requests :).
Troubleshooting
I don’t get input from reader or WARNING: Communication failure, is the MFRC522 properly connected?
Check your physical connection, see Pin Layout .
Check your pin settings/variables in the code, see Pin Layout .
Check your pin header soldering. Maybe you have cold solder joints.
Check voltage. Most breakouts work with 3.3V.
SPI only works with 3.3V, most breakouts seem 5V tolerant, but try a level shifter.
SPI does not like long connections. Try shorter connections.
SPI does not like prototyping boards. Try soldered connections.
According to reports #101, #126 and #131, there may be a problem with the soldering on the MFRC522 breakout. You could fix this on your own.
Firmware Version: 0x12 = (unknown) or other random values
The exact reason of this behaviour is unknown.
Some boards need more time after PCD_Init() to be ready. As workaround add a delay(4) directly after PCD_Init() to give the PCD more time.
If this sometimes appears, a bad connection or power source is the reason.
If the firmware version is reported permanent, it is very likely that the hardware is a fake or has a defect. Contact your supplier.
Sometimes I get timeouts or sometimes tag/card does not work.
Try the other side of the antenna.
Try to decrease the distance between the MFRC522 and your tag.
Increase the antenna gain per firmware:
mfrc522.PCD_SetAntennaGain(mfrc522.RxGain_max);Use better power supply.
Hardware may be corrupted, most products are from china and sometimes the quality is really poor. Contact your seller.
My tag/card doesn’t work.
Distance between antenna and token too large (>1cm).
You got the wrong type PICC. Is it really 13.56 MHz? Is it really a Mifare Type A?
NFC tokens are not supported. Some may work.
Animal RFID tags are not supported. They use a different frequency (125 kHz).
Hardware may be corrupted, most products are from china and sometimes the quality is really poor. Contact your seller.
Newer versions of Mifare cards like DESFire/Ultralight maybe not work according to missing authentication, see security or different protocol.
Some boards bought from Chinese manufactures do not use the best components and this can affect the detection of different types of tag/card. In some of these boards, the L1 and L2 inductors do not have a high enough current so the signal generated is not enough to get Ultralight C and NTAG203 tags to work, replacing those with same inductance (2.2uH) but higher operating current inductors should make things work smoothly. Also, in some of those boards the harmonic and matching circuit needs to be tuned, for this replace C4 and C5 with 33pf capacitors and you are all set. (Source: Mikro Elektronika)
My mobile phone doesn’t recognize the MFRC522 or my MFRC522 can’t read data from other MFRC522
Card simulation is not supported.
Communication with mobile phones is not supported.
Peer to peer communication is not supported.
I can only read the card UID.
Maybe the AccessBits have been accidentally set and now an unknown password is set. This can not be reverted.
Probably the card is encrypted. Especially official cards like public transport, university or library cards. There is no way to get access with this library.
I need more features.
If software: code it and make a pull request.
If hardware: buy a more expensive chip like the PN532 (supports NFC and many more, but costs about $15)
License
This is free and unencumbered software released into the public domain.
Anyone is free to copy, modify, publish, use, compile, sell, or distribute this software, either in source code form or as a compiled binary, for any purpose, commercial or non-commercial, and by any means.
In jurisdictions that recognize copyright laws, the author or authors of this software dedicate any and all copyright interest in the software to the public domain. We make this dedication for the benefit of the public at large and to the detriment of our heirs and successors. We intend this dedication to be an overt act of relinquishment in perpetuity of all present and future rights to this software under copyright law.
THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
For more information, please refer to https://unlicense.org/
Dependency
Arduino.h
From: Arduino IDE / target specific
License: (target: Arduino) GNU Lesser General Public License 2.1
SPI.h
From: Arduino IDE / target specific
License: (target: Arduino) GNU Lesser General Public License 2.1
stdint.h
From: Arduino IDE / Compiler and target specific
License: different
History
The MFRC522 library was first created in Jan 2012 by Miguel Balboa (from http://circuitito.com) based on code by Dr. Leong (from http://B2CQSHOP.com) for “Arduino RFID module Kit 13.56 Mhz with Tags SPI W and R By COOQRobot”.
It was translated into English and rewritten/refactored in the fall of 2013 by Søren Thing Andersen (from http://access.thing.dk).
It has been extended with functionality to alter sector 0 on Chinese UID changeable MIFARE card in Oct 2014 by Tom Clement (from http://tomclement.nl).
Maintained by miguelbalboa until 2016. Maintained by Rotzbua from 2016 until 2018.
References
Used by
Basic NFC Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MHZ19 CO2 Sensor
Sming library supporting the MH-Z19 and MH-Z19B CO2 sensors.
Essentially a re-write of https://github.com/crisap94/MHZ19 to use Hardware serial port with callbacks.
See https://github.com/WifWaf/MH-Z19 for some very detailed and helpful background to the device.
Functionality is split into two main classes:
MHZ19::UartAccess the sensor via serial port
MHZ19::PwmReaderAsynchronously decodes PWM signal from sensor
ESP8266 Connections
Communication requires use of UART0, as UART1 is output-only.
The samples in this library reconfigure the default Serial class to use UART1 for debug output.
UART0 is reconfigured to use the alternate pins by calling HardwareSerial::swap().
Connect as follows:
From |
To |
Notes |
|---|---|---|
MHZ19 PWM |
GPIO14 (D5) |
|
MHZ19 TX |
GPIO13 (D7) |
UART0 RX |
MHZ19 RX |
GPIO15 (D8) |
UART0 TX |
GPIO2 (D4) |
GPIO1 (TX) |
Route UART1 output to USB |
Note: A similar approach is used for I2S, see Tone Generator.
References
API Documentation
-
namespace MHZ19
Typedefs
-
using MeasurementCallback = Delegate<void(Measurement &m)>
Enums
-
enum class DetectionRange
Device may be configured to output CO2 PPM values in various ranges.
Values:
-
enumerator PPM_2000
-
enumerator PPM_5000
-
enumerator PPM_10000
-
enumerator PPM_2000
Functions
-
unsigned pwmRead(uint8_t pwmPin, DetectionRange range)
Read PWM output from sensor.
Note
This will hang CPU for 1-2 seconds. Use PwmReader instead.
- Parameters:
pwmPin – GPIO to which the sensor is connected
range – Range sensor is configured for
-
class PwmReader
- #include <PwmReader.h>
Reads input pulse width asynchronously.
The ESP8266 lacks any timing capture hardware but as the pulse output from the MHZ19 is very slow it can be decoded using interrupts with high accuracy.
Once started, the PwmReader runs continuously and the last value can be obtained by calling
getMeasurement.Public Types
-
using Callback = Delegate<void(uint16_t ppm)>
Callback for regular readings.
Public Functions
-
void begin(uint8_t pin, DetectionRange range)
Start the PWM reader.
Runs continuously in background until end() is called.
- Parameters:
pin – GPIO to read PWM signal on
range – Configured device range
-
void end()
Stop the PWM reader.
-
uint16_t getMeasurement() const
Obtain the most recent measurement.
-
inline void setCallback(Callback callback)
Set a callback to be invoked whenever a valid reading is obtained.
-
bool suspend()
Temporarily suspend readings and disable interrupts on the PWM pin.
- Return values:
bool – true on success, false if reader not initialised
-
union Pulse
- #include <PwmReader.h>
Used internally to measure a high/low pulse pair.
Public Functions
-
inline uint16_t getCycleTime() const
-
inline bool isValid() const
-
inline uint16_t calculatePpm(DetectionRange range)
-
inline uint16_t getCycleTime() const
-
using Callback = Delegate<void(uint16_t ppm)>
-
struct Request
- #include <Uart.h>
-
struct Response
- #include <Uart.h>
-
struct Measurement
- #include <Uart.h>
-
class Uart
- #include <Uart.h>
Access MHZ19 sensor via serial port.
Public Functions
-
inline Uart(HardwareSerial &serial)
Use device in UART mode.
- Parameters:
serial – Port and pins must be pre-configured
-
inline void setSensorNumber(uint8_t num)
Change sensor number field.
Original datasheet identifies first byte as sensor number, others as ‘reserved’. Provided if this value changes in future hardware versions.
-
inline uint8_t getSensorNumber() const
Get currently configured sensor number.
-
void setAutoCalibration(bool enable)
Enable/disable zero-point auto-calibration feature. On by default.
Calibrates 400ppm zero point reference automatically at power-on and every 24 hours. Suitable for home/office environment only. Refer to datasheet for details.
-
void calibrateZero()
Calibrate zero point manually.
Sensor must be in stable 400ppm CO2 environment. Refer to datasheet for detailed procedure.
-
void calibrateSpan(uint16_t ppm)
Calibrate span point.
Typical calibration value is 2000ppm, but must be above 1000ppm. Refer to datasheet for detailed procedure.
-
void setDetectionRange(DetectionRange range)
MHZ-19B has configurable detection range.
-
bool getMeasurement(MeasurementCallback callback)
Read measurement from device.
-
inline Uart(HardwareSerial &serial)
-
using MeasurementCallback = Delegate<void(Measurement &m)>
References
Used by
MHZ19 CO2 sensor reading Sample
MHZ19 Span-calibration Sample
MHZ19 Zero-Calibration Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MMA-7455 Accelerometer
Library for the MMA-7455 3-axis accelerometer
References
Used by
MMA7455 Accelerometer Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
MPU6050 Gyro / Accelerometer
MPU6050 Six-Axis (Gyro + Accelerometer) Based on code from jrowberg/i2cdevlib @ 605a740. Most of the code is the same, except:
Removed MPU6050::ReadRegister function due to incompatibility. It is also not used anywhere in the original code.
MPU6050_6Axis_MotionApps20.h and MPU6050_9Axis_MotionApps41.h are not included due to deps to freeRTOS. helper_3dmath.h is also not included since it is only used in the above mentioned files.
Removed map function in favor of the Sming built-in one.
Adapted include path, coding style and applied clangformat
Deleted Calibration and Memory Block related code for code quality reason
API Documentation
-
class MPU6050
Public Functions
-
inline MPU6050()
Default constructor, uses default I2C address.
See also
MPU6050_DEFAULT_ADDRESS
-
inline MPU6050(uint8_t address)
Specific address constructor.
See also
MPU6050_DEFAULT_ADDRESS
See also
MPU6050_ADDRESS_AD0_LOW
See also
MPU6050_ADDRESS_AD0_HIGH
- Parameters:
address – I2C address
-
void initialize()
Power on and prepare for general usage. This will activate the device and take it out of sleep mode (which must be done after start-up). This function also sets both the accelerometer and the gyroscope to their most sensitive settings, namely +/- 2g and +/- 250 degrees/sec, and sets the clock source to use the X Gyro for reference, which is slightly better than the default internal clock source.
-
inline bool testConnection()
Verify the I2C connection. Make sure the device is connected and responds as expected.
- Returns:
True if connection is valid, false otherwise
-
inline uint8_t getAuxVDDIOLevel()
Get the auxiliary I2C supply voltage level. When set to 1, the auxiliary I2C bus high logic level is VDD. When cleared to 0, the auxiliary I2C bus high logic level is VLOGIC. This does not apply to the MPU-6000, which does not have a VLOGIC pin.
- Returns:
I2C supply voltage level (0=VLOGIC, 1=VDD)
-
inline void setAuxVDDIOLevel(uint8_t level)
Set the auxiliary I2C supply voltage level. When set to 1, the auxiliary I2C bus high logic level is VDD. When cleared to 0, the auxiliary I2C bus high logic level is VLOGIC. This does not apply to the MPU-6000, which does not have a VLOGIC pin.
- Parameters:
level – I2C supply voltage level (0=VLOGIC, 1=VDD)
-
inline uint8_t getRate()
Get gyroscope output rate divider. The sensor register output, FIFO output, DMP sampling, Motion detection, Zero Motion detection, and Free Fall detection are all based on the Sample Rate. The Sample Rate is generated by dividing the gyroscope output rate by SMPLRT_DIV:
Sample Rate = Gyroscope Output Rate / (1 + SMPLRT_DIV)
where Gyroscope Output Rate = 8kHz when the DLPF is disabled (DLPF_CFG = 0 or 7), and 1kHz when the DLPF is enabled (see Register 26).
Note: The accelerometer output rate is 1kHz. This means that for a Sample Rate greater than 1kHz, the same accelerometer sample may be output to the FIFO, DMP, and sensor registers more than once.
For a diagram of the gyroscope and accelerometer signal paths, see Section 8 of the MPU-6000/MPU-6050 Product Specification document.
See also
MPU6050_RA_SMPLRT_DIV
- Returns:
Current sample rate
-
inline void setRate(uint8_t rate)
Set gyroscope sample rate divider.
See also
See also
MPU6050_RA_SMPLRT_DIV
- Parameters:
rate – New sample rate divider
-
inline uint8_t getExternalFrameSync()
Get external FSYNC configuration. Configures the external Frame Synchronization (FSYNC) pin sampling. An external signal connected to the FSYNC pin can be sampled by configuring EXT_SYNC_SET. Signal changes to the FSYNC pin are latched so that short strobes may be captured. The latched FSYNC signal will be sampled at the Sampling Rate, as defined in register 25. After sampling, the latch will reset to the current FSYNC signal state.
The sampled value will be reported in place of the least significant bit in a sensor data register determined by the value of EXT_SYNC_SET according to the following table.
- Returns:
FSYNC configuration value
-
inline void setExternalFrameSync(uint8_t sync)
Set external FSYNC configuration.
See also
See also
MPU6050_RA_CONFIG
- Parameters:
sync – New FSYNC configuration value
-
inline uint8_t getDLPFMode()
Get digital low-pass filter configuration. The DLPF_CFG parameter sets the digital low pass filter configuration. It also determines the internal sampling rate used by the device as shown in the table below.
Note: The accelerometer output rate is 1kHz. This means that for a Sample Rate greater than 1kHz, the same accelerometer sample may be output to the FIFO, DMP, and sensor registers more than once.
See also
MPU6050_RA_CONFIG
See also
MPU6050_CFG_DLPF_CFG_BIT
See also
MPU6050_CFG_DLPF_CFG_LENGTH
- Returns:
DLFP configuration
-
inline void setDLPFMode(uint8_t mode)
Set digital low-pass filter configuration.
See also
getDLPFBandwidth()
See also
MPU6050_DLPF_BW_256
See also
MPU6050_RA_CONFIG
See also
MPU6050_CFG_DLPF_CFG_BIT
See also
MPU6050_CFG_DLPF_CFG_LENGTH
- Parameters:
mode – New DLFP configuration setting
-
inline uint8_t getFullScaleGyroRange()
Get full-scale gyroscope range. The FS_SEL parameter allows setting the full-scale range of the gyro sensors, as described in the table below.
See also
MPU6050_GYRO_FS_250
See also
MPU6050_RA_GYRO_CONFIG
See also
MPU6050_GCONFIG_FS_SEL_BIT
See also
MPU6050_GCONFIG_FS_SEL_LENGTH
- Returns:
Current full-scale gyroscope range setting
-
inline void setFullScaleGyroRange(uint8_t range)
Set full-scale gyroscope range.
See also
getFullScaleRange()
See also
MPU6050_GYRO_FS_250
See also
MPU6050_RA_GYRO_CONFIG
See also
MPU6050_GCONFIG_FS_SEL_BIT
See also
MPU6050_GCONFIG_FS_SEL_LENGTH
- Parameters:
range – New full-scale gyroscope range value
-
uint8_t getAccelXSelfTestFactoryTrim()
Get self-test factory trim value for accelerometer X axis.
See also
MPU6050_RA_SELF_TEST_X
- Returns:
factory trim value
-
uint8_t getAccelYSelfTestFactoryTrim()
Get self-test factory trim value for accelerometer Y axis.
See also
MPU6050_RA_SELF_TEST_Y
- Returns:
factory trim value
-
uint8_t getAccelZSelfTestFactoryTrim()
Get self-test factory trim value for accelerometer Z axis.
See also
MPU6050_RA_SELF_TEST_Z
- Returns:
factory trim value
-
uint8_t getGyroXSelfTestFactoryTrim()
Get self-test factory trim value for gyro X axis.
See also
MPU6050_RA_SELF_TEST_X
- Returns:
factory trim value
-
uint8_t getGyroYSelfTestFactoryTrim()
Get self-test factory trim value for gyro Y axis.
See also
MPU6050_RA_SELF_TEST_Y
- Returns:
factory trim value
-
uint8_t getGyroZSelfTestFactoryTrim()
Get self-test factory trim value for gyro Z axis.
See also
MPU6050_RA_SELF_TEST_Z
- Returns:
factory trim value
-
inline bool getAccelXSelfTest()
Get self-test enabled setting for accelerometer X axis.
See also
MPU6050_RA_ACCEL_CONFIG
- Returns:
Self-test enabled value
-
inline void setAccelXSelfTest(bool enabled)
Get self-test enabled setting for accelerometer X axis.
See also
MPU6050_RA_ACCEL_CONFIG
- Parameters:
enabled – Self-test enabled value
-
inline bool getAccelYSelfTest()
Get self-test enabled value for accelerometer Y axis.
See also
MPU6050_RA_ACCEL_CONFIG
- Returns:
Self-test enabled value
-
inline void setAccelYSelfTest(bool enabled)
Get self-test enabled value for accelerometer Y axis.
See also
MPU6050_RA_ACCEL_CONFIG
- Parameters:
enabled – Self-test enabled value
-
inline bool getAccelZSelfTest()
Get self-test enabled value for accelerometer Z axis.
See also
MPU6050_RA_ACCEL_CONFIG
- Returns:
Self-test enabled value
-
inline void setAccelZSelfTest(bool enabled)
Set self-test enabled value for accelerometer Z axis.
See also
MPU6050_RA_ACCEL_CONFIG
- Parameters:
enabled – Self-test enabled value
-
inline uint8_t getFullScaleAccelRange()
Get full-scale accelerometer range. The FS_SEL parameter allows setting the full-scale range of the accelerometer sensors, as described in the table below.
See also
MPU6050_ACCEL_FS_2
See also
MPU6050_RA_ACCEL_CONFIG
See also
MPU6050_ACONFIG_AFS_SEL_BIT
See also
MPU6050_ACONFIG_AFS_SEL_LENGTH
- Returns:
Current full-scale accelerometer range setting
-
inline void setFullScaleAccelRange(uint8_t range)
Set full-scale accelerometer range.
See also
- Parameters:
range – New full-scale accelerometer range setting
-
inline uint8_t getDHPFMode()
Get the high-pass filter configuration. The DHPF is a filter module in the path leading to motion detectors (Free Fall, Motion threshold, and Zero Motion). The high pass filter output is not available to the data registers (see Figure in Section 8 of the MPU-6000/ MPU-6050 Product Specification document).
The high pass filter has three modes:
See also
MPU6050_DHPF_RESET
See also
MPU6050_RA_ACCEL_CONFIG
- Returns:
Current high-pass filter configuration
-
inline void setDHPFMode(uint8_t bandwidth)
Set the high-pass filter configuration.
See also
See also
MPU6050_DHPF_RESET
See also
MPU6050_RA_ACCEL_CONFIG
- Parameters:
bandwidth – New high-pass filter configuration
-
inline uint8_t getFreefallDetectionThreshold()
Get free-fall event acceleration threshold. This register configures the detection threshold for Free Fall event detection. The unit of FF_THR is 1LSB = 2mg. Free Fall is detected when the absolute value of the accelerometer measurements for the three axes are each less than the detection threshold. This condition increments the Free Fall duration counter (Register 30). The Free Fall interrupt is triggered when the Free Fall duration counter reaches the time specified in FF_DUR.
For more details on the Free Fall detection interrupt, see Section 8.2 of the MPU-6000/MPU-6050 Product Specification document as well as Registers 56 and 58 of this document.
See also
MPU6050_RA_FF_THR
- Returns:
Current free-fall acceleration threshold value (LSB = 2mg)
-
inline void setFreefallDetectionThreshold(uint8_t threshold)
Get free-fall event acceleration threshold.
See also
See also
MPU6050_RA_FF_THR
- Parameters:
threshold – New free-fall acceleration threshold value (LSB = 2mg)
-
inline uint8_t getFreefallDetectionDuration()
Get free-fall event duration threshold. This register configures the duration counter threshold for Free Fall event detection. The duration counter ticks at 1kHz, therefore FF_DUR has a unit of 1 LSB = 1 ms.
The Free Fall duration counter increments while the absolute value of the accelerometer measurements are each less than the detection threshold (Register 29). The Free Fall interrupt is triggered when the Free Fall duration counter reaches the time specified in this register.
For more details on the Free Fall detection interrupt, see Section 8.2 of the MPU-6000/MPU-6050 Product Specification document as well as Registers 56 and 58 of this document.
See also
MPU6050_RA_FF_DUR
- Returns:
Current free-fall duration threshold value (LSB = 1ms)
-
inline void setFreefallDetectionDuration(uint8_t duration)
Get free-fall event duration threshold.
See also
See also
MPU6050_RA_FF_DUR
- Parameters:
duration – New free-fall duration threshold value (LSB = 1ms)
-
inline uint8_t getMotionDetectionThreshold()
Get motion detection event acceleration threshold. This register configures the detection threshold for Motion interrupt generation. The unit of MOT_THR is 1LSB = 2mg. Motion is detected when the absolute value of any of the accelerometer measurements exceeds this Motion detection threshold. This condition increments the Motion detection duration counter (Register 32). The Motion detection interrupt is triggered when the Motion Detection counter reaches the time count specified in MOT_DUR (Register 32).
The Motion interrupt will indicate the axis and polarity of detected motion in MOT_DETECT_STATUS (Register 97).
For more details on the Motion detection interrupt, see Section 8.3 of the MPU-6000/MPU-6050 Product Specification document as well as Registers 56 and 58 of this document.
See also
MPU6050_RA_MOT_THR
- Returns:
Current motion detection acceleration threshold value (LSB = 2mg)
-
inline void setMotionDetectionThreshold(uint8_t threshold)
Set motion detection event acceleration threshold.
See also
See also
MPU6050_RA_MOT_THR
- Parameters:
threshold – New motion detection acceleration threshold value (LSB = 2mg)
-
inline uint8_t getMotionDetectionDuration()
Get motion detection event duration threshold. This register configures the duration counter threshold for Motion interrupt generation. The duration counter ticks at 1 kHz, therefore MOT_DUR has a unit of 1LSB = 1ms. The Motion detection duration counter increments when the absolute value of any of the accelerometer measurements exceeds the Motion detection threshold (Register 31). The Motion detection interrupt is triggered when the Motion detection counter reaches the time count specified in this register.
For more details on the Motion detection interrupt, see Section 8.3 of the MPU-6000/MPU-6050 Product Specification document.
See also
MPU6050_RA_MOT_DUR
- Returns:
Current motion detection duration threshold value (LSB = 1ms)
-
inline void setMotionDetectionDuration(uint8_t duration)
Set motion detection event duration threshold.
See also
See also
MPU6050_RA_MOT_DUR
- Parameters:
duration – New motion detection duration threshold value (LSB = 1ms)
-
inline uint8_t getZeroMotionDetectionThreshold()
Get zero motion detection event acceleration threshold. This register configures the detection threshold for Zero Motion interrupt generation. The unit of ZRMOT_THR is 1LSB = 2mg. Zero Motion is detected when the absolute value of the accelerometer measurements for the 3 axes are each less than the detection threshold. This condition increments the Zero Motion duration counter (Register 34). The Zero Motion interrupt is triggered when the Zero Motion duration counter reaches the time count specified in ZRMOT_DUR (Register 34).
Unlike Free Fall or Motion detection, Zero Motion detection triggers an interrupt both when Zero Motion is first detected and when Zero Motion is no longer detected.
When a zero motion event is detected, a Zero Motion Status will be indicated in the MOT_DETECT_STATUS register (Register 97). When a motion-to-zero-motion condition is detected, the status bit is set to 1. When a zero-motion-to- motion condition is detected, the status bit is set to 0.
For more details on the Zero Motion detection interrupt, see Section 8.4 of the MPU-6000/MPU-6050 Product Specification document as well as Registers 56 and 58 of this document.
See also
MPU6050_RA_ZRMOT_THR
- Returns:
Current zero motion detection acceleration threshold value (LSB = 2mg)
-
inline void setZeroMotionDetectionThreshold(uint8_t threshold)
Set zero motion detection event acceleration threshold.
See also
See also
MPU6050_RA_ZRMOT_THR
- Parameters:
threshold – New zero motion detection acceleration threshold value (LSB = 2mg)
-
inline uint8_t getZeroMotionDetectionDuration()
Get zero motion detection event duration threshold. This register configures the duration counter threshold for Zero Motion interrupt generation. The duration counter ticks at 16 Hz, therefore ZRMOT_DUR has a unit of 1 LSB = 64 ms. The Zero Motion duration counter increments while the absolute value of the accelerometer measurements are each less than the detection threshold (Register 33). The Zero Motion interrupt is triggered when the Zero Motion duration counter reaches the time count specified in this register.
For more details on the Zero Motion detection interrupt, see Section 8.4 of the MPU-6000/MPU-6050 Product Specification document, as well as Registers 56 and 58 of this document.
See also
MPU6050_RA_ZRMOT_DUR
- Returns:
Current zero motion detection duration threshold value (LSB = 64ms)
-
inline void setZeroMotionDetectionDuration(uint8_t duration)
Set zero motion detection event duration threshold.
See also
See also
MPU6050_RA_ZRMOT_DUR
- Parameters:
duration – New zero motion detection duration threshold value (LSB = 1ms)
-
inline bool getTempFIFOEnabled()
Get temperature FIFO enabled value. When set to 1, this bit enables TEMP_OUT_H and TEMP_OUT_L (Registers 65 and 66) to be written into the FIFO buffer.
See also
MPU6050_RA_FIFO_EN
- Returns:
Current temperature FIFO enabled value
-
inline void setTempFIFOEnabled(bool enabled)
Set temperature FIFO enabled value.
See also
See also
MPU6050_RA_FIFO_EN
- Parameters:
enabled – New temperature FIFO enabled value
-
inline bool getXGyroFIFOEnabled()
Get gyroscope X-axis FIFO enabled value. When set to 1, this bit enables GYRO_XOUT_H and GYRO_XOUT_L (Registers 67 and 68) to be written into the FIFO buffer.
See also
MPU6050_RA_FIFO_EN
- Returns:
Current gyroscope X-axis FIFO enabled value
-
inline void setXGyroFIFOEnabled(bool enabled)
Set gyroscope X-axis FIFO enabled value.
See also
See also
MPU6050_RA_FIFO_EN
- Parameters:
enabled – New gyroscope X-axis FIFO enabled value
-
inline bool getYGyroFIFOEnabled()
Get gyroscope Y-axis FIFO enabled value. When set to 1, this bit enables GYRO_YOUT_H and GYRO_YOUT_L (Registers 69 and 70) to be written into the FIFO buffer.
See also
MPU6050_RA_FIFO_EN
- Returns:
Current gyroscope Y-axis FIFO enabled value
-
inline void setYGyroFIFOEnabled(bool enabled)
Set gyroscope Y-axis FIFO enabled value.
See also
See also
MPU6050_RA_FIFO_EN
- Parameters:
enabled – New gyroscope Y-axis FIFO enabled value
-
inline bool getZGyroFIFOEnabled()
Get gyroscope Z-axis FIFO enabled value. When set to 1, this bit enables GYRO_ZOUT_H and GYRO_ZOUT_L (Registers 71 and 72) to be written into the FIFO buffer.
See also
MPU6050_RA_FIFO_EN
- Returns:
Current gyroscope Z-axis FIFO enabled value
-
inline void setZGyroFIFOEnabled(bool enabled)
Set gyroscope Z-axis FIFO enabled value.
See also
See also
MPU6050_RA_FIFO_EN
- Parameters:
enabled – New gyroscope Z-axis FIFO enabled value
-
inline bool getAccelFIFOEnabled()
Get accelerometer FIFO enabled value. When set to 1, this bit enables ACCEL_XOUT_H, ACCEL_XOUT_L, ACCEL_YOUT_H, ACCEL_YOUT_L, ACCEL_ZOUT_H, and ACCEL_ZOUT_L (Registers 59 to 64) to be written into the FIFO buffer.
See also
MPU6050_RA_FIFO_EN
- Returns:
Current accelerometer FIFO enabled value
-
inline void setAccelFIFOEnabled(bool enabled)
Set accelerometer FIFO enabled value.
See also
See also
MPU6050_RA_FIFO_EN
- Parameters:
enabled – New accelerometer FIFO enabled value
-
inline bool getSlave2FIFOEnabled()
Get Slave 2 FIFO enabled value. When set to 1, this bit enables EXT_SENS_DATA registers (Registers 73 to 96) associated with Slave 2 to be written into the FIFO buffer.
See also
MPU6050_RA_FIFO_EN
- Returns:
Current Slave 2 FIFO enabled value
-
inline void setSlave2FIFOEnabled(bool enabled)
Set Slave 2 FIFO enabled value.
See also
See also
MPU6050_RA_FIFO_EN
- Parameters:
enabled – New Slave 2 FIFO enabled value
-
inline bool getSlave1FIFOEnabled()
Get Slave 1 FIFO enabled value. When set to 1, this bit enables EXT_SENS_DATA registers (Registers 73 to 96) associated with Slave 1 to be written into the FIFO buffer.
See also
MPU6050_RA_FIFO_EN
- Returns:
Current Slave 1 FIFO enabled value
-
inline void setSlave1FIFOEnabled(bool enabled)
Set Slave 1 FIFO enabled value.
See also
See also
MPU6050_RA_FIFO_EN
- Parameters:
enabled – New Slave 1 FIFO enabled value
-
inline bool getSlave0FIFOEnabled()
Get Slave 0 FIFO enabled value. When set to 1, this bit enables EXT_SENS_DATA registers (Registers 73 to 96) associated with Slave 0 to be written into the FIFO buffer.
See also
MPU6050_RA_FIFO_EN
- Returns:
Current Slave 0 FIFO enabled value
-
inline void setSlave0FIFOEnabled(bool enabled)
Set Slave 0 FIFO enabled value.
See also
See also
MPU6050_RA_FIFO_EN
- Parameters:
enabled – New Slave 0 FIFO enabled value
-
inline bool getMultiMasterEnabled()
Get multi-master enabled value. Multi-master capability allows multiple I2C masters to operate on the same bus. In circuits where multi-master capability is required, set MULT_MST_EN to 1. This will increase current drawn by approximately 30uA.
In circuits where multi-master capability is required, the state of the I2C bus must always be monitored by each separate I2C Master. Before an I2C Master can assume arbitration of the bus, it must first confirm that no other I2C Master has arbitration of the bus. When MULT_MST_EN is set to 1, the MPU-60X0’s bus arbitration detection logic is turned on, enabling it to detect when the bus is available.
See also
MPU6050_RA_I2C_MST_CTRL
- Returns:
Current multi-master enabled value
-
inline void setMultiMasterEnabled(bool enabled)
Set multi-master enabled value.
See also
See also
MPU6050_RA_I2C_MST_CTRL
- Parameters:
enabled – New multi-master enabled value
-
inline bool getWaitForExternalSensorEnabled()
Get wait-for-external-sensor-data enabled value. When the WAIT_FOR_ES bit is set to 1, the Data Ready interrupt will be delayed until External Sensor data from the Slave Devices are loaded into the EXT_SENS_DATA registers. This is used to ensure that both the internal sensor data (i.e. from gyro and accel) and external sensor data have been loaded to their respective data registers (i.e. the data is synced) when the Data Ready interrupt is triggered.
See also
MPU6050_RA_I2C_MST_CTRL
- Returns:
Current wait-for-external-sensor-data enabled value
-
inline void setWaitForExternalSensorEnabled(bool enabled)
Set wait-for-external-sensor-data enabled value.
See also
See also
MPU6050_RA_I2C_MST_CTRL
- Parameters:
enabled – New wait-for-external-sensor-data enabled value
-
inline bool getSlave3FIFOEnabled()
Get Slave 3 FIFO enabled value. When set to 1, this bit enables EXT_SENS_DATA registers (Registers 73 to 96) associated with Slave 3 to be written into the FIFO buffer.
See also
MPU6050_RA_MST_CTRL
- Returns:
Current Slave 3 FIFO enabled value
-
inline void setSlave3FIFOEnabled(bool enabled)
Set Slave 3 FIFO enabled value.
See also
See also
MPU6050_RA_MST_CTRL
- Parameters:
enabled – New Slave 3 FIFO enabled value
-
inline bool getSlaveReadWriteTransitionEnabled()
Get slave read/write transition enabled value. The I2C_MST_P_NSR bit configures the I2C Master’s transition from one slave read to the next slave read. If the bit equals 0, there will be a restart between reads. If the bit equals 1, there will be a stop followed by a start of the following read. When a write transaction follows a read transaction, the stop followed by a start of the successive write will be always used.
See also
MPU6050_RA_I2C_MST_CTRL
- Returns:
Current slave read/write transition enabled value
-
inline void setSlaveReadWriteTransitionEnabled(bool enabled)
Set slave read/write transition enabled value.
See also
See also
MPU6050_RA_I2C_MST_CTRL
- Parameters:
enabled – New slave read/write transition enabled value
-
inline uint8_t getMasterClockSpeed()
Get I2C master clock speed. I2C_MST_CLK is a 4 bit unsigned value which configures a divider on the MPU-60X0 internal 8MHz clock. It sets the I2C master clock speed according to the following table:
See also
MPU6050_RA_I2C_MST_CTRL
- Returns:
Current I2C master clock speed
-
inline void setMasterClockSpeed(uint8_t speed)
Set I2C master clock speed. @reparam speed Current I2C master clock speed
See also
MPU6050_RA_I2C_MST_CTRL
-
uint8_t getSlaveAddress(SlaveId slaveId)
Get the I2C address of the specified slave (0-3). Note that Bit 7 (MSB) controls read/write mode. If Bit 7 is set, it’s a read operation, and if it is cleared, then it’s a write operation. The remaining bits (6-0) are the 7-bit device address of the slave device.
In read mode, the result of the read is placed in the lowest available EXT_SENS_DATA register. For further information regarding the allocation of read results, please refer to the EXT_SENS_DATA register description (Registers 73 - 96).
The MPU-6050 supports a total of five slaves, but Slave 4 has unique characteristics, and so it has its own functions (getSlave4* and setSlave4*).
I2C data transactions are performed at the Sample Rate, as defined in Register 25. The user is responsible for ensuring that I2C data transactions to and from each enabled Slave can be completed within a single period of the Sample Rate.
The I2C slave access rate can be reduced relative to the Sample Rate. This reduced access rate is determined by I2C_MST_DLY (Register 52). Whether a slave’s access rate is reduced relative to the Sample Rate is determined by I2C_MST_DELAY_CTRL (Register 103).
The processing order for the slaves is fixed. The sequence followed for processing the slaves is Slave 0, Slave 1, Slave 2, Slave 3 and Slave 4. If a particular Slave is disabled it will be skipped.
Each slave can either be accessed at the sample rate or at a reduced sample rate. In a case where some slaves are accessed at the Sample Rate and some slaves are accessed at the reduced rate, the sequence of accessing the slaves (Slave 0 to Slave 4) is still followed. However, the reduced rate slaves will be skipped if their access rate dictates that they should not be accessed during that particular cycle. For further information regarding the reduced access rate, please refer to Register 52. Whether a slave is accessed at the Sample Rate or at the reduced rate is determined by the Delay Enable bits in Register 103.
See also
MPU6050_RA_I2C_SLV0_ADDR
- Parameters:
slaveId – Slave ID (0-3)
- Returns:
Current address for specified slave
-
void setSlaveAddress(SlaveId slaveId, uint8_t address)
Set the I2C address of the specified slave (0-3).
See also
See also
MPU6050_RA_I2C_SLV0_ADDR
- Parameters:
slaveId – Slave ID (0-3)
address – New address for specified slave
-
uint8_t getSlaveRegister(SlaveId slaveId)
Get the active internal register for the specified slave (0-3). Read/write operations for this slave will be done to whatever internal register address is stored in this MPU register.
The MPU-6050 supports a total of five slaves, but Slave 4 has unique characteristics, and so it has its own functions.
See also
MPU6050_RA_I2C_SLV0_REG
- Parameters:
slaveId – Slave ID (0-3)
- Returns:
Current active register for specified slave
-
void setSlaveRegister(SlaveId slaveId, uint8_t reg)
Set the active internal register for the specified slave (0-3).
See also
See also
MPU6050_RA_I2C_SLV0_REG
- Parameters:
slaveId – Slave ID (0-3)
reg – New active register for specified slave
-
bool getSlaveEnabled(SlaveId slaveId)
Get the enabled value for the specified slave (0-3). When set to 1, this bit enables Slave 0 for data transfer operations. When cleared to 0, this bit disables Slave 0 from data transfer operations.
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
- Returns:
Current enabled value for specified slave
-
void setSlaveEnabled(SlaveId slaveId, bool enabled)
Set the enabled value for the specified slave (0-3).
See also
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
enabled – New enabled value for specified slave
-
bool getSlaveWordByteSwap(SlaveId slaveId)
Get word pair byte-swapping enabled for the specified slave (0-3). When set to 1, this bit enables byte swapping. When byte swapping is enabled, the high and low bytes of a word pair are swapped. Please refer to I2C_SLV0_GRP for the pairing convention of the word pairs. When cleared to 0, bytes transferred to and from Slave 0 will be written to EXT_SENS_DATA registers in the order they were transferred.
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
- Returns:
Current word pair byte-swapping enabled value for specified slave
-
void setSlaveWordByteSwap(SlaveId slaveId, bool enabled)
Set word pair byte-swapping enabled for the specified slave (0-3).
See also
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
enabled – New word pair byte-swapping enabled value for specified slave
-
bool getSlaveWriteMode(SlaveId slaveId)
Get write mode for the specified slave (0-3). When set to 1, the transaction will read or write data only. When cleared to 0, the transaction will write a register address prior to reading or writing data. This should equal 0 when specifying the register address within the Slave device to/from which the ensuing data transaction will take place.
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
- Returns:
Current write mode for specified slave (0 = register address + data, 1 = data only)
-
void setSlaveWriteMode(SlaveId slaveId, bool mode)
Set write mode for the specified slave (0-3).
See also
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
mode – New write mode for specified slave (0 = register address + data, 1 = data only)
-
bool getSlaveWordGroupOffset(SlaveId slaveId)
Get word pair grouping order offset for the specified slave (0-3). This sets specifies the grouping order of word pairs received from registers. When cleared to 0, bytes from register addresses 0 and 1, 2 and 3, etc (even, then odd register addresses) are paired to form a word. When set to 1, bytes from register addresses are paired 1 and 2, 3 and 4, etc. (odd, then even register addresses) are paired to form a word.
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
- Returns:
Current word pair grouping order offset for specified slave
-
void setSlaveWordGroupOffset(SlaveId slaveId, bool enabled)
Set word pair grouping order offset for the specified slave (0-3).
See also
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
enabled – New word pair grouping order offset for specified slave
-
uint8_t getSlaveDataLength(SlaveId slaveId)
Get number of bytes to read for the specified slave (0-3). Specifies the number of bytes transferred to and from Slave 0. Clearing this bit to 0 is equivalent to disabling the register by writing 0 to I2C_SLV0_EN.
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
- Returns:
Number of bytes to read for specified slave
-
void setSlaveDataLength(SlaveId slaveId, uint8_t length)
Set number of bytes to read for the specified slave (0-3).
See also
See also
MPU6050_RA_I2C_SLV0_CTRL
- Parameters:
slaveId – Slave ID (0-3)
length – Number of bytes to read for specified slave
-
inline uint8_t getSlave4Address()
Get the I2C address of Slave 4. Note that Bit 7 (MSB) controls read/write mode. If Bit 7 is set, it’s a read operation, and if it is cleared, then it’s a write operation. The remaining bits (6-0) are the 7-bit device address of the slave device.
See also
See also
MPU6050_RA_I2C_SLV4_ADDR
- Returns:
Current address for Slave 4
-
inline void setSlave4Address(uint8_t address)
Set the I2C address of Slave 4.
See also
See also
MPU6050_RA_I2C_SLV4_ADDR
- Parameters:
address – New address for Slave 4
-
inline uint8_t getSlave4Register()
Get the active internal register for the Slave 4. Read/write operations for this slave will be done to whatever internal register address is stored in this MPU register.
See also
MPU6050_RA_I2C_SLV4_REG
- Returns:
Current active register for Slave 4
-
inline void setSlave4Register(uint8_t reg)
Set the active internal register for Slave 4.
See also
See also
MPU6050_RA_I2C_SLV4_REG
- Parameters:
reg – New active register for Slave 4
-
inline void setSlave4OutputByte(uint8_t data)
Set new byte to write to Slave 4. This register stores the data to be written into the Slave 4. If I2C_SLV4_RW is set 1 (set to read), this register has no effect.
See also
MPU6050_RA_I2C_SLV4_DO
- Parameters:
data – New byte to write to Slave 4
-
inline bool getSlave4Enabled()
Get the enabled value for the Slave 4. When set to 1, this bit enables Slave 4 for data transfer operations. When cleared to 0, this bit disables Slave 4 from data transfer operations.
See also
MPU6050_RA_I2C_SLV4_CTRL
- Returns:
Current enabled value for Slave 4
-
inline void setSlave4Enabled(bool enabled)
Set the enabled value for Slave 4.
See also
See also
MPU6050_RA_I2C_SLV4_CTRL
- Parameters:
enabled – New enabled value for Slave 4
-
inline bool getSlave4InterruptEnabled()
Get the enabled value for Slave 4 transaction interrupts. When set to 1, this bit enables the generation of an interrupt signal upon completion of a Slave 4 transaction. When cleared to 0, this bit disables the generation of an interrupt signal upon completion of a Slave 4 transaction. The interrupt status can be observed in Register 54.
See also
MPU6050_RA_I2C_SLV4_CTRL
- Returns:
Current enabled value for Slave 4 transaction interrupts.
-
inline void setSlave4InterruptEnabled(bool enabled)
Set the enabled value for Slave 4 transaction interrupts.
See also
See also
MPU6050_RA_I2C_SLV4_CTRL
- Parameters:
enabled – New enabled value for Slave 4 transaction interrupts.
-
inline bool getSlave4WriteMode()
Get write mode for Slave 4. When set to 1, the transaction will read or write data only. When cleared to 0, the transaction will write a register address prior to reading or writing data. This should equal 0 when specifying the register address within the Slave device to/from which the ensuing data transaction will take place.
See also
MPU6050_RA_I2C_SLV4_CTRL
- Returns:
Current write mode for Slave 4 (0 = register address + data, 1 = data only)
-
inline void setSlave4WriteMode(bool mode)
Set write mode for the Slave 4.
See also
See also
MPU6050_RA_I2C_SLV4_CTRL
- Parameters:
mode – New write mode for Slave 4 (0 = register address + data, 1 = data only)
-
inline uint8_t getSlave4MasterDelay()
Get Slave 4 master delay value. This configures the reduced access rate of I2C slaves relative to the Sample Rate. When a slave’s access rate is decreased relative to the Sample Rate, the slave is accessed every:
This base Sample Rate in turn is determined by SMPLRT_DIV (register 25) and DLPF_CFG (register 26). Whether a slave’s access rate is reduced relative to the Sample Rate is determined by I2C_MST_DELAY_CTRL (register 103). For further information regarding the Sample Rate, please refer to register 25.1 / (1 + I2C_MST_DLY) samples
See also
MPU6050_RA_I2C_SLV4_CTRL
- Returns:
Current Slave 4 master delay value
-
inline void setSlave4MasterDelay(uint8_t delay)
Set Slave 4 master delay value.
See also
See also
MPU6050_RA_I2C_SLV4_CTRL
- Parameters:
delay – New Slave 4 master delay value
-
inline uint8_t getSlate4InputByte()
Get last available byte read from Slave 4. This register stores the data read from Slave 4. This field is populated after a read transaction.
See also
MPU6050_RA_I2C_SLV4_DI
- Returns:
Last available byte read from to Slave 4
-
inline bool getPassthroughStatus()
Get FSYNC interrupt status. This bit reflects the status of the FSYNC interrupt from an external device into the MPU-60X0. This is used as a way to pass an external interrupt through the MPU-60X0 to the host application processor. When set to 1, this bit will cause an interrupt if FSYNC_INT_EN is asserted in INT_PIN_CFG (Register 55).
See also
MPU6050_RA_I2C_MST_STATUS
- Returns:
FSYNC interrupt status
-
inline bool getSlave4IsDone()
Get Slave 4 transaction done status. Automatically sets to 1 when a Slave 4 transaction has completed. This triggers an interrupt if the I2C_MST_INT_EN bit in the INT_ENABLE register (Register 56) is asserted and if the SLV_4_DONE_INT bit is asserted in the I2C_SLV4_CTRL register (Register 52).
See also
MPU6050_RA_I2C_MST_STATUS
- Returns:
Slave 4 transaction done status
-
inline bool getLostArbitration()
Get master arbitration lost status. This bit automatically sets to 1 when the I2C Master has lost arbitration of the auxiliary I2C bus (an error condition). This triggers an interrupt if the I2C_MST_INT_EN bit in the INT_ENABLE register (Register 56) is asserted.
See also
MPU6050_RA_I2C_MST_STATUS
- Returns:
Master arbitration lost status
-
inline bool getSlave4Nack()
Get Slave 4 NACK status. This bit automatically sets to 1 when the I2C Master receives a NACK in a transaction with Slave 4. This triggers an interrupt if the I2C_MST_INT_EN bit in the INT_ENABLE register (Register 56) is asserted.
See also
MPU6050_RA_I2C_MST_STATUS
- Returns:
Slave 4 NACK interrupt status
-
inline bool getSlave3Nack()
Get Slave 3 NACK status. This bit automatically sets to 1 when the I2C Master receives a NACK in a transaction with Slave 3. This triggers an interrupt if the I2C_MST_INT_EN bit in the INT_ENABLE register (Register 56) is asserted.
See also
MPU6050_RA_I2C_MST_STATUS
- Returns:
Slave 3 NACK interrupt status
-
inline bool getSlave2Nack()
Get Slave 2 NACK status. This bit automatically sets to 1 when the I2C Master receives a NACK in a transaction with Slave 2. This triggers an interrupt if the I2C_MST_INT_EN bit in the INT_ENABLE register (Register 56) is asserted.
See also
MPU6050_RA_I2C_MST_STATUS
- Returns:
Slave 2 NACK interrupt status
-
inline bool getSlave1Nack()
Get Slave 1 NACK status. This bit automatically sets to 1 when the I2C Master receives a NACK in a transaction with Slave 1. This triggers an interrupt if the I2C_MST_INT_EN bit in the INT_ENABLE register (Register 56) is asserted.
See also
MPU6050_RA_I2C_MST_STATUS
- Returns:
Slave 1 NACK interrupt status
-
inline bool getSlave0Nack()
Get Slave 0 NACK status. This bit automatically sets to 1 when the I2C Master receives a NACK in a transaction with Slave 0. This triggers an interrupt if the I2C_MST_INT_EN bit in the INT_ENABLE register (Register 56) is asserted.
See also
MPU6050_RA_I2C_MST_STATUS
- Returns:
Slave 0 NACK interrupt status
-
inline bool getInterruptMode()
Get interrupt logic level mode. Will be set 0 for active-high, 1 for active-low.
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_INT_LEVEL_BIT
- Returns:
Current interrupt mode (0=active-high, 1=active-low)
-
inline void setInterruptMode(bool mode)
Set interrupt logic level mode.
See also
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_INT_LEVEL_BIT
- Parameters:
mode – New interrupt mode (0=active-high, 1=active-low)
-
inline bool getInterruptDrive()
Get interrupt drive mode. Will be set 0 for push-pull, 1 for open-drain.
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_INT_OPEN_BIT
- Returns:
Current interrupt drive mode (0=push-pull, 1=open-drain)
-
inline void setInterruptDrive(bool drive)
Set interrupt drive mode.
See also
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_INT_OPEN_BIT
- Parameters:
drive – New interrupt drive mode (0=push-pull, 1=open-drain)
-
inline bool getInterruptLatch()
Get interrupt latch mode. Will be set 0 for 50us-pulse, 1 for latch-until-int-cleared.
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_LATCH_INT_EN_BIT
- Returns:
Current latch mode (0=50us-pulse, 1=latch-until-int-cleared)
-
inline void setInterruptLatch(bool latch)
Set interrupt latch mode.
See also
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_LATCH_INT_EN_BIT
- Parameters:
latch – New latch mode (0=50us-pulse, 1=latch-until-int-cleared)
-
inline bool getInterruptLatchClear()
Get interrupt latch clear mode. Will be set 0 for status-read-only, 1 for any-register-read.
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_INT_RD_CLEAR_BIT
- Returns:
Current latch clear mode (0=status-read-only, 1=any-register-read)
-
inline void setInterruptLatchClear(bool clear)
Set interrupt latch clear mode.
See also
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_INT_RD_CLEAR_BIT
- Parameters:
clear – New latch clear mode (0=status-read-only, 1=any-register-read)
-
inline bool getFSyncInterruptLevel()
Get FSYNC interrupt logic level mode.
See also
getFSyncInterruptMode()
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_FSYNC_INT_LEVEL_BIT
- Returns:
Current FSYNC interrupt mode (0=active-high, 1=active-low)
-
inline void setFSyncInterruptLevel(bool level)
Set FSYNC interrupt logic level mode.
See also
getFSyncInterruptMode()
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_FSYNC_INT_LEVEL_BIT
- Parameters:
mode – New FSYNC interrupt mode (0=active-high, 1=active-low)
-
inline bool getFSyncInterruptEnabled()
Get FSYNC pin interrupt enabled setting. Will be set 0 for disabled, 1 for enabled.
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_FSYNC_INT_EN_BIT
- Returns:
Current interrupt enabled setting
-
inline void setFSyncInterruptEnabled(bool enabled)
Set FSYNC pin interrupt enabled setting.
See also
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_FSYNC_INT_EN_BIT
- Parameters:
enabled – New FSYNC pin interrupt enabled setting
-
inline bool getI2CBypassEnabled()
Get I2C bypass enabled status. When this bit is equal to 1 and I2C_MST_EN (Register 106 bit[5]) is equal to 0, the host application processor will be able to directly access the auxiliary I2C bus of the MPU-60X0. When this bit is equal to 0, the host application processor will not be able to directly access the auxiliary I2C bus of the MPU-60X0 regardless of the state of I2C_MST_EN (Register 106 bit[5]).
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_I2C_BYPASS_EN_BIT
- Returns:
Current I2C bypass enabled status
-
inline void setI2CBypassEnabled(bool enabled)
Set I2C bypass enabled status. When this bit is equal to 1 and I2C_MST_EN (Register 106 bit[5]) is equal to 0, the host application processor will be able to directly access the auxiliary I2C bus of the MPU-60X0. When this bit is equal to 0, the host application processor will not be able to directly access the auxiliary I2C bus of the MPU-60X0 regardless of the state of I2C_MST_EN (Register 106 bit[5]).
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_I2C_BYPASS_EN_BIT
- Parameters:
enabled – New I2C bypass enabled status
-
inline bool getClockOutputEnabled()
Get reference clock output enabled status. When this bit is equal to 1, a reference clock output is provided at the CLKOUT pin. When this bit is equal to 0, the clock output is disabled. For further information regarding CLKOUT, please refer to the MPU-60X0 Product Specification document.
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_CLKOUT_EN_BIT
- Returns:
Current reference clock output enabled status
-
inline void setClockOutputEnabled(bool enabled)
Set reference clock output enabled status. When this bit is equal to 1, a reference clock output is provided at the CLKOUT pin. When this bit is equal to 0, the clock output is disabled. For further information regarding CLKOUT, please refer to the MPU-60X0 Product Specification document.
See also
MPU6050_RA_INT_PIN_CFG
See also
MPU6050_INTCFG_CLKOUT_EN_BIT
- Parameters:
enabled – New reference clock output enabled status
-
inline uint8_t getIntEnabled()
Get full interrupt enabled status. Full register byte for all interrupts, for quick reading. Each bit will be set 0 for disabled, 1 for enabled.
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_FF_BIT
- Returns:
Current interrupt enabled status
-
inline void setIntEnabled(uint8_t enabled)
Set full interrupt enabled status. Full register byte for all interrupts, for quick reading. Each bit should be set 0 for disabled, 1 for enabled.
See also
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_FF_BIT
- Parameters:
enabled – New interrupt enabled status
-
inline bool getIntFreefallEnabled()
Get Free Fall interrupt enabled status. Will be set 0 for disabled, 1 for enabled.
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_FF_BIT
- Returns:
Current interrupt enabled status
-
inline void setIntFreefallEnabled(bool enabled)
Set Free Fall interrupt enabled status.
See also
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_FF_BIT
- Parameters:
enabled – New interrupt enabled status
-
inline bool getIntMotionEnabled()
Get Motion Detection interrupt enabled status. Will be set 0 for disabled, 1 for enabled.
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_MOT_BIT
- Returns:
Current interrupt enabled status
-
inline void setIntMotionEnabled(bool enabled)
Set Motion Detection interrupt enabled status.
See also
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_MOT_BIT
- Parameters:
enabled – New interrupt enabled status
-
inline bool getIntZeroMotionEnabled()
Get Zero Motion Detection interrupt enabled status. Will be set 0 for disabled, 1 for enabled.
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_ZMOT_BIT
- Returns:
Current interrupt enabled status
-
inline void setIntZeroMotionEnabled(bool enabled)
Set Zero Motion Detection interrupt enabled status.
See also
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_ZMOT_BIT
- Parameters:
enabled – New interrupt enabled status
-
inline bool getIntFIFOBufferOverflowEnabled()
Get FIFO Buffer Overflow interrupt enabled status. Will be set 0 for disabled, 1 for enabled.
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_FIFO_OFLOW_BIT
- Returns:
Current interrupt enabled status
-
inline void setIntFIFOBufferOverflowEnabled(bool enabled)
Set FIFO Buffer Overflow interrupt enabled status.
See also
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_FIFO_OFLOW_BIT
- Parameters:
enabled – New interrupt enabled status
-
inline bool getIntI2CMasterEnabled()
Get I2C Master interrupt enabled status. This enables any of the I2C Master interrupt sources to generate an interrupt. Will be set 0 for disabled, 1 for enabled.
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_I2C_MST_INT_BIT
- Returns:
Current interrupt enabled status
-
inline void setIntI2CMasterEnabled(bool enabled)
Set I2C Master interrupt enabled status.
See also
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_I2C_MST_INT_BIT
- Parameters:
enabled – New interrupt enabled status
-
inline bool getIntDataReadyEnabled()
Get Data Ready interrupt enabled setting. This event occurs each time a write operation to all of the sensor registers has been completed. Will be set 0 for disabled, 1 for enabled.
See also
MPU6050_RA_INT_ENABLE
See also
MPU6050_INTERRUPT_DATA_RDY_BIT
- Returns:
Current interrupt enabled status
-
inline void setIntDataReadyEnabled(bool enabled)
Set Data Ready interrupt enabled status.
See also
See also
MPU6050_RA_INT_CFG
See also
MPU6050_INTERRUPT_DATA_RDY_BIT
- Parameters:
enabled – New interrupt enabled status
-
inline uint8_t getIntStatus()
Get full set of interrupt status bits. These bits clear to 0 after the register has been read. Very useful for getting multiple INT statuses, since each single bit read clears all of them because it has to read the whole byte.
See also
MPU6050_RA_INT_STATUS
- Returns:
Current interrupt status
-
inline bool getIntFreefallStatus()
Get Free Fall interrupt status. This bit automatically sets to 1 when a Free Fall interrupt has been generated. The bit clears to 0 after the register has been read.
See also
MPU6050_RA_INT_STATUS
See also
MPU6050_INTERRUPT_FF_BIT
- Returns:
Current interrupt status
-
inline bool getIntMotionStatus()
Get Motion Detection interrupt status. This bit automatically sets to 1 when a Motion Detection interrupt has been generated. The bit clears to 0 after the register has been read.
See also
MPU6050_RA_INT_STATUS
See also
MPU6050_INTERRUPT_MOT_BIT
- Returns:
Current interrupt status
-
inline bool getIntZeroMotionStatus()
Get Zero Motion Detection interrupt status. This bit automatically sets to 1 when a Zero Motion Detection interrupt has been generated. The bit clears to 0 after the register has been read.
See also
MPU6050_RA_INT_STATUS
See also
MPU6050_INTERRUPT_ZMOT_BIT
- Returns:
Current interrupt status
-
inline bool getIntFIFOBufferOverflowStatus()
Get FIFO Buffer Overflow interrupt status. This bit automatically sets to 1 when a Free Fall interrupt has been generated. The bit clears to 0 after the register has been read.
See also
MPU6050_RA_INT_STATUS
See also
MPU6050_INTERRUPT_FIFO_OFLOW_BIT
- Returns:
Current interrupt status
-
inline bool getIntI2CMasterStatus()
Get I2C Master interrupt status. This bit automatically sets to 1 when an I2C Master interrupt has been generated. For a list of I2C Master interrupts, please refer to Register 54. The bit clears to 0 after the register has been read.
See also
MPU6050_RA_INT_STATUS
See also
MPU6050_INTERRUPT_I2C_MST_INT_BIT
- Returns:
Current interrupt status
-
inline bool getIntDataReadyStatus()
Get Data Ready interrupt status. This bit automatically sets to 1 when a Data Ready interrupt has been generated. The bit clears to 0 after the register has been read.
See also
MPU6050_RA_INT_STATUS
See also
MPU6050_INTERRUPT_DATA_RDY_BIT
- Returns:
Current interrupt status
-
Motion6 getMotion6()
Get raw 6-axis motion sensor readings (accel/gyro). Retrieves all currently available motion sensor values.
See also
See also
See also
MPU6050_RA_ACCEL_XOUT_H
- Returns:
container for 3-axis accelerometer and 3-axis gyroscope values
-
Motion3 getAcceleration()
Get 3-axis accelerometer readings. These registers store the most recent accelerometer measurements. Accelerometer measurements are written to these registers at the Sample Rate as defined in Register 25.
The accelerometer measurement registers, along with the temperature measurement registers, gyroscope measurement registers, and external sensor data registers, are composed of two sets of registers: an internal register set and a user-facing read register set.
The data within the accelerometer sensors’ internal register set is always updated at the Sample Rate. Meanwhile, the user-facing read register set duplicates the internal register set’s data values whenever the serial interface is idle. This guarantees that a burst read of sensor registers will read measurements from the same sampling instant. Note that if burst reads are not used, the user is responsible for ensuring a set of single byte reads correspond to a single sampling instant by checking the Data Ready interrupt.
Each 16-bit accelerometer measurement has a full scale defined in ACCEL_FS (Register 28). For each full scale setting, the accelerometers’ sensitivity per LSB in ACCEL_xOUT is shown in the table below:
See also
MPU6050_RA_GYRO_XOUT_H
- Parameters:
x – 16-bit signed integer container for X-axis acceleration
y – 16-bit signed integer container for Y-axis acceleration
z – 16-bit signed integer container for Z-axis acceleration
-
inline int16_t getAccelerationX()
Get X-axis accelerometer reading.
See also
See also
MPU6050_RA_ACCEL_XOUT_H
- Returns:
X-axis acceleration measurement in 16-bit 2’s complement format
-
inline int16_t getAccelerationY()
Get Y-axis accelerometer reading.
See also
See also
MPU6050_RA_ACCEL_YOUT_H
- Returns:
Y-axis acceleration measurement in 16-bit 2’s complement format
-
inline int16_t getAccelerationZ()
Get Z-axis accelerometer reading.
See also
See also
MPU6050_RA_ACCEL_ZOUT_H
- Returns:
Z-axis acceleration measurement in 16-bit 2’s complement format
-
inline int16_t getTemperature()
Get current internal temperature.
See also
MPU6050_RA_TEMP_OUT_H
- Returns:
Temperature reading in 16-bit 2’s complement format
-
Motion3 getAngularRate()
Get 3-axis gyroscope readings. These gyroscope measurement registers, along with the accelerometer measurement registers, temperature measurement registers, and external sensor data registers, are composed of two sets of registers: an internal register set and a user-facing read register set. The data within the gyroscope sensors’ internal register set is always updated at the Sample Rate. Meanwhile, the user-facing read register set duplicates the internal register set’s data values whenever the serial interface is idle. This guarantees that a burst read of sensor registers will read measurements from the same sampling instant. Note that if burst reads are not used, the user is responsible for ensuring a set of single byte reads correspond to a single sampling instant by checking the Data Ready interrupt.
Each 16-bit gyroscope measurement has a full scale defined in FS_SEL (Register 27). For each full scale setting, the gyroscopes’ sensitivity per LSB in GYRO_xOUT is shown in the table below:
See also
See also
MPU6050_RA_GYRO_XOUT_H
- Returns:
container for 3-axis gyro values
-
inline int16_t getAngularRateX()
Get X-axis gyroscope reading.
See also
See also
MPU6050_RA_GYRO_XOUT_H
- Returns:
X-axis rotation measurement in 16-bit 2’s complement format
-
inline int16_t getAngularRateY()
Get Y-axis gyroscope reading.
See also
See also
MPU6050_RA_GYRO_YOUT_H
- Returns:
Y-axis rotation measurement in 16-bit 2’s complement format
-
inline int16_t getAngularRateZ()
Get Z-axis gyroscope reading.
See also
See also
MPU6050_RA_GYRO_ZOUT_H
- Returns:
Z-axis rotation measurement in 16-bit 2’s complement format
-
inline uint8_t getExternalSensorByte(int position)
Read single byte from external sensor data register. These registers store data read from external sensors by the Slave 0, 1, 2, and 3 on the auxiliary I2C interface. Data read by Slave 4 is stored in I2C_SLV4_DI (Register 53).
External sensor data is written to these registers at the Sample Rate as defined in Register 25. This access rate can be reduced by using the Slave Delay Enable registers (Register 103).
External sensor data registers, along with the gyroscope measurement registers, accelerometer measurement registers, and temperature measurement registers, are composed of two sets of registers: an internal register set and a user-facing read register set.
The data within the external sensors’ internal register set is always updated at the Sample Rate (or the reduced access rate) whenever the serial interface is idle. This guarantees that a burst read of sensor registers will read measurements from the same sampling instant. Note that if burst reads are not used, the user is responsible for ensuring a set of single byte reads correspond to a single sampling instant by checking the Data Ready interrupt.
Data is placed in these external sensor data registers according to I2C_SLV0_CTRL, I2C_SLV1_CTRL, I2C_SLV2_CTRL, and I2C_SLV3_CTRL (Registers 39, 42, 45, and 48). When more than zero bytes are read (I2C_SLVx_LEN > 0) from an enabled slave (I2C_SLVx_EN = 1), the slave is read at the Sample Rate (as defined in Register 25) or delayed rate (if specified in Register 52 and 103). During each Sample cycle, slave reads are performed in order of Slave number. If all slaves are enabled with more than zero bytes to be read, the order will be Slave 0, followed by Slave 1, Slave 2, and Slave 3.
Each enabled slave will have EXT_SENS_DATA registers associated with it by number of bytes read (I2C_SLVx_LEN) in order of slave number, starting from EXT_SENS_DATA_00. Note that this means enabling or disabling a slave may change the higher numbered slaves’ associated registers. Furthermore, if fewer total bytes are being read from the external sensors as a result of such a change, then the data remaining in the registers which no longer have an associated slave device (i.e. high numbered registers) will remain in these previously allocated registers unless reset.
If the sum of the read lengths of all SLVx transactions exceed the number of available EXT_SENS_DATA registers, the excess bytes will be dropped. There are 24 EXT_SENS_DATA registers and hence the total read lengths between all the slaves cannot be greater than 24 or some bytes will be lost.
Note: Slave 4’s behavior is distinct from that of Slaves 0-3. For further information regarding the characteristics of Slave 4, please refer to Registers 49 to 53.
EXAMPLE: Suppose that Slave 0 is enabled with 4 bytes to be read (I2C_SLV0_EN = 1 and I2C_SLV0_LEN = 4) while Slave 1 is enabled with 2 bytes to be read so that I2C_SLV1_EN = 1 and I2C_SLV1_LEN = 2. In such a situation, EXT_SENS_DATA _00 through _03 will be associated with Slave 0, while EXT_SENS_DATA _04 and 05 will be associated with Slave 1. If Slave 2 is enabled as well, registers starting from EXT_SENS_DATA_06 will be allocated to Slave 2.
If Slave 2 is disabled while Slave 3 is enabled in this same situation, then registers starting from EXT_SENS_DATA_06 will be allocated to Slave 3 instead.
REGISTER ALLOCATION FOR DYNAMIC DISABLE VS. NORMAL DISABLE: If a slave is disabled at any time, the space initially allocated to the slave in the EXT_SENS_DATA register, will remain associated with that slave. This is to avoid dynamic adjustment of the register allocation.
The allocation of the EXT_SENS_DATA registers is recomputed only when (1) all slaves are disabled, or (2) the I2C_MST_RST bit is set (Register 106).
This above is also true if one of the slaves gets NACKed and stops functioning.
- Parameters:
position – Starting position (0-23)
- Returns:
Byte read from register
-
inline uint16_t getExternalSensorWord(int position)
Read word (2 bytes) from external sensor data registers.
See also
- Parameters:
position – Starting position (0-21)
- Returns:
Word read from register
-
inline uint32_t getExternalSensorDWord(int position)
Read double word (4 bytes) from external sensor data registers.
See also
- Parameters:
position – Starting position (0-20)
- Returns:
Double word read from registers
-
inline uint8_t getMotionStatus()
Get full motion detection status register content (all bits).
See also
MPU6050_RA_MOT_DETECT_STATUS
- Returns:
Motion detection status byte
-
inline bool getXNegMotionDetected()
Get X-axis negative motion detection interrupt status.
See also
MPU6050_RA_MOT_DETECT_STATUS
See also
MPU6050_MOTION_MOT_XNEG_BIT
- Returns:
Motion detection status
-
inline bool getXPosMotionDetected()
Get X-axis positive motion detection interrupt status.
See also
MPU6050_RA_MOT_DETECT_STATUS
See also
MPU6050_MOTION_MOT_XPOS_BIT
- Returns:
Motion detection status
-
inline bool getYNegMotionDetected()
Get Y-axis negative motion detection interrupt status.
See also
MPU6050_RA_MOT_DETECT_STATUS
See also
MPU6050_MOTION_MOT_YNEG_BIT
- Returns:
Motion detection status
-
inline bool getYPosMotionDetected()
Get Y-axis positive motion detection interrupt status.
See also
MPU6050_RA_MOT_DETECT_STATUS
See also
MPU6050_MOTION_MOT_YPOS_BIT
- Returns:
Motion detection status
-
inline bool getZNegMotionDetected()
Get Z-axis negative motion detection interrupt status.
See also
MPU6050_RA_MOT_DETECT_STATUS
See also
MPU6050_MOTION_MOT_ZNEG_BIT
- Returns:
Motion detection status
-
inline bool getZPosMotionDetected()
Get Z-axis positive motion detection interrupt status.
See also
MPU6050_RA_MOT_DETECT_STATUS
See also
MPU6050_MOTION_MOT_ZPOS_BIT
- Returns:
Motion detection status
-
inline bool getZeroMotionDetected()
Get zero motion detection interrupt status.
See also
MPU6050_RA_MOT_DETECT_STATUS
See also
MPU6050_MOTION_MOT_ZRMOT_BIT
- Returns:
Motion detection status
-
void setSlaveOutputByte(SlaveId slaveId, uint8_t data)
Write byte to Data Output container for specified slave. This register holds the output data written into Slave when Slave is set to write mode. For further information regarding Slave control, please refer to Registers 37 to 39 and immediately following.
See also
MPU6050_RA_I2C_SLV0_DO
- Parameters:
slaveId – Slave ID (0-3)
data – Byte to write
-
inline bool getExternalShadowDelayEnabled()
Get external data shadow delay enabled status. This register is used to specify the timing of external sensor data shadowing. When DELAY_ES_SHADOW is set to 1, shadowing of external sensor data is delayed until all data has been received.
See also
MPU6050_RA_I2C_MST_DELAY_CTRL
See also
MPU6050_DELAYCTRL_DELAY_ES_SHADOW_BIT
- Returns:
Current external data shadow delay enabled status.
-
inline void setExternalShadowDelayEnabled(bool enabled)
Set external data shadow delay enabled status.
See also
See also
MPU6050_RA_I2C_MST_DELAY_CTRL
See also
MPU6050_DELAYCTRL_DELAY_ES_SHADOW_BIT
- Parameters:
enabled – New external data shadow delay enabled status.
-
bool getSlaveDelayEnabled(SlaveId slaveId)
Get slave delay enabled status. When a particular slave delay is enabled, the rate of access for the that slave device is reduced. When a slave’s access rate is decreased relative to the Sample Rate, the slave is accessed every:
This base Sample Rate in turn is determined by SMPLRT_DIV (register * 25) and DLPF_CFG (register 26).1 / (1 + I2C_MST_DLY) Samples
For further information regarding I2C_MST_DLY, please refer to register 52. For further information regarding the Sample Rate, please refer to register 25.
See also
MPU6050_RA_I2C_MST_DELAY_CTRL
See also
MPU6050_DELAYCTRL_I2C_SLV0_DLY_EN_BIT
- Parameters:
slaveId – Slave ID (0-4)
- Returns:
Current slave delay enabled status.
-
inline void setSlaveDelayEnabled(SlaveId slaveId, bool enabled)
Set slave delay enabled status.
See also
MPU6050_RA_I2C_MST_DELAY_CTRL
See also
MPU6050_DELAYCTRL_I2C_SLV0_DLY_EN_BIT
- Parameters:
slaveId – Slave ID (0-4)
enabled – New slave delay enabled status.
-
inline void resetGyroscopePath()
Reset gyroscope signal path. The reset will revert the signal path analog to digital converters and filters to their power up configurations.
See also
MPU6050_RA_SIGNAL_PATH_RESET
See also
MPU6050_PATHRESET_GYRO_RESET_BIT
-
inline void resetAccelerometerPath()
Reset accelerometer signal path. The reset will revert the signal path analog to digital converters and filters to their power up configurations.
See also
MPU6050_RA_SIGNAL_PATH_RESET
See also
MPU6050_PATHRESET_ACCEL_RESET_BIT
-
inline void resetTemperaturePath()
Reset temperature sensor signal path. The reset will revert the signal path analog to digital converters and filters to their power up configurations.
See also
MPU6050_RA_SIGNAL_PATH_RESET
See also
MPU6050_PATHRESET_TEMP_RESET_BIT
-
inline uint8_t getAccelerometerPowerOnDelay()
Get accelerometer power-on delay. The accelerometer data path provides samples to the sensor registers, Motion detection, Zero Motion detection, and Free Fall detection modules. The signal path contains filters which must be flushed on wake-up with new samples before the detection modules begin operations. The default wake-up delay, of 4ms can be lengthened by up to 3ms. This additional delay is specified in ACCEL_ON_DELAY in units of 1 LSB = 1 ms. The user may select any value above zero unless instructed otherwise by InvenSense. Please refer to Section 8 of the MPU-6000/MPU-6050 Product Specification document for further information regarding the detection modules.
See also
MPU6050_RA_MOT_DETECT_CTRL
See also
MPU6050_DETECT_ACCEL_ON_DELAY_BIT
- Returns:
Current accelerometer power-on delay
-
inline void setAccelerometerPowerOnDelay(uint8_t delay)
Set accelerometer power-on delay.
See also
See also
MPU6050_RA_MOT_DETECT_CTRL
See also
MPU6050_DETECT_ACCEL_ON_DELAY_BIT
- Parameters:
delay – New accelerometer power-on delay (0-3)
-
inline uint8_t getFreefallDetectionCounterDecrement()
Get Free Fall detection counter decrement configuration. Detection is registered by the Free Fall detection module after accelerometer measurements meet their respective threshold conditions over a specified number of samples. When the threshold conditions are met, the corresponding detection counter increments by 1. The user may control the rate at which the detection counter decrements when the threshold condition is not met by configuring FF_COUNT. The decrement rate can be set according to the following table:
When FF_COUNT is configured to 0 (reset), any non-qualifying sample will reset the counter to 0. For further information on Free Fall detection, please refer to Registers 29 to 32.
See also
MPU6050_RA_MOT_DETECT_CTRL
See also
MPU6050_DETECT_FF_COUNT_BIT
- Returns:
Current decrement configuration
-
inline void setFreefallDetectionCounterDecrement(uint8_t decrement)
Set Free Fall detection counter decrement configuration.
See also
MPU6050_RA_MOT_DETECT_CTRL
See also
MPU6050_DETECT_FF_COUNT_BIT
- Parameters:
decrement – New decrement configuration value
-
inline uint8_t getMotionDetectionCounterDecrement()
Get Motion detection counter decrement configuration. Detection is registered by the Motion detection module after accelerometer measurements meet their respective threshold conditions over a specified number of samples. When the threshold conditions are met, the corresponding detection counter increments by 1. The user may control the rate at which the detection counter decrements when the threshold condition is not met by configuring MOT_COUNT. The decrement rate can be set according to the following table:
When MOT_COUNT is configured to 0 (reset), any non-qualifying sample will reset the counter to 0. For further information on Motion detection, please refer to Registers 29 to 32.
-
inline void setMotionDetectionCounterDecrement(uint8_t decrement)
Set Motion detection counter decrement configuration.
See also
See also
MPU6050_RA_MOT_DETECT_CTRL
See also
MPU6050_DETECT_MOT_COUNT_BIT
- Parameters:
decrement – New decrement configuration value
-
inline bool getFIFOEnabled()
Get FIFO enabled status. When this bit is set to 0, the FIFO buffer is disabled. The FIFO buffer cannot be written to or read from while disabled. The FIFO buffer’s state does not change unless the MPU-60X0 is power cycled.
See also
MPU6050_RA_USER_CTRL
See also
MPU6050_USERCTRL_FIFO_EN_BIT
- Returns:
Current FIFO enabled status
-
inline void setFIFOEnabled(bool enabled)
Set FIFO enabled status.
See also
See also
MPU6050_RA_USER_CTRL
See also
MPU6050_USERCTRL_FIFO_EN_BIT
- Parameters:
enabled – New FIFO enabled status
-
inline bool getI2CMasterModeEnabled()
Get I2C Master Mode enabled status. When this mode is enabled, the MPU-60X0 acts as the I2C Master to the external sensor slave devices on the auxiliary I2C bus. When this bit is cleared to 0, the auxiliary I2C bus lines (AUX_DA and AUX_CL) are logically driven by the primary I2C bus (SDA and SCL). This is a precondition to enabling Bypass Mode. For further information regarding Bypass Mode, please refer to Register 55.
See also
MPU6050_RA_USER_CTRL
See also
MPU6050_USERCTRL_I2C_MST_EN_BIT
- Returns:
Current I2C Master Mode enabled status
-
inline void setI2CMasterModeEnabled(bool enabled)
Set I2C Master Mode enabled status.
See also
See also
MPU6050_RA_USER_CTRL
See also
MPU6050_USERCTRL_I2C_MST_EN_BIT
- Parameters:
enabled – New I2C Master Mode enabled status
-
inline void switchSPIEnabled(bool enabled)
Switch from I2C to SPI mode (MPU-6000 only) If this is set, the primary SPI interface will be enabled in place of the disabled primary I2C interface.
-
inline void resetFIFO()
Reset the FIFO. This bit resets the FIFO buffer when set to 1 while FIFO_EN equals 0. This bit automatically clears to 0 after the reset has been triggered.
See also
MPU6050_RA_USER_CTRL
See also
MPU6050_USERCTRL_FIFO_RESET_BIT
-
inline void resetI2CMaster()
Reset the I2C Master. This bit resets the I2C Master when set to 1 while I2C_MST_EN equals 0. This bit automatically clears to 0 after the reset has been triggered.
See also
MPU6050_RA_USER_CTRL
See also
MPU6050_USERCTRL_I2C_MST_RESET_BIT
-
inline void resetSensors()
Reset all sensor registers and signal paths. When set to 1, this bit resets the signal paths for all sensors (gyroscopes, accelerometers, and temperature sensor). This operation will also clear the sensor registers. This bit automatically clears to 0 after the reset has been triggered.
When resetting only the signal path (and not the sensor registers), please use Register 104, SIGNAL_PATH_RESET.
See also
MPU6050_RA_USER_CTRL
See also
MPU6050_USERCTRL_SIG_COND_RESET_BIT
-
inline void reset()
Trigger a full device reset. A small delay of ~50ms may be desirable after triggering a reset.
See also
MPU6050_RA_PWR_MGMT_1
See also
MPU6050_PWR1_DEVICE_RESET_BIT
-
inline bool getSleepEnabled()
Get sleep mode status. Setting the SLEEP bit in the register puts the device into very low power sleep mode. In this mode, only the serial interface and internal registers remain active, allowing for a very low standby current. Clearing this bit puts the device back into normal mode. To save power, the individual standby selections for each of the gyros should be used if any gyro axis is not used by the application.
See also
MPU6050_RA_PWR_MGMT_1
See also
MPU6050_PWR1_SLEEP_BIT
- Returns:
Current sleep mode enabled status
-
inline void setSleepEnabled(bool enabled)
Set sleep mode status.
See also
See also
MPU6050_RA_PWR_MGMT_1
See also
MPU6050_PWR1_SLEEP_BIT
- Parameters:
enabled – New sleep mode enabled status
-
inline bool getWakeCycleEnabled()
Get wake cycle enabled status. When this bit is set to 1 and SLEEP is disabled, the MPU-60X0 will cycle between sleep mode and waking up to take a single sample of data from active sensors at a rate determined by LP_WAKE_CTRL (register 108).
See also
MPU6050_RA_PWR_MGMT_1
See also
MPU6050_PWR1_CYCLE_BIT
- Returns:
Current sleep mode enabled status
-
inline void setWakeCycleEnabled(bool enabled)
Set wake cycle enabled status.
See also
See also
MPU6050_RA_PWR_MGMT_1
See also
MPU6050_PWR1_CYCLE_BIT
- Parameters:
enabled – New sleep mode enabled status
-
inline bool getTempSensorEnabled()
Get temperature sensor enabled status. Control the usage of the internal temperature sensor.
Note: this register stores the disabled value, but for consistency with the rest of the code, the function is named and used with standard true/false values to indicate whether the sensor is enabled or disabled, respectively.
See also
MPU6050_RA_PWR_MGMT_1
See also
MPU6050_PWR1_TEMP_DIS_BIT
- Returns:
Current temperature sensor enabled status
-
inline void setTempSensorEnabled(bool enabled)
Set temperature sensor enabled status. Note: this register stores the disabled value, but for consistency with the rest of the code, the function is named and used with standard true/false values to indicate whether the sensor is enabled or disabled, respectively.
See also
See also
MPU6050_RA_PWR_MGMT_1
See also
MPU6050_PWR1_TEMP_DIS_BIT
- Parameters:
enabled – New temperature sensor enabled status
-
inline uint8_t getClockSource()
Get clock source setting.
See also
MPU6050_RA_PWR_MGMT_1
See also
MPU6050_PWR1_CLKSEL_BIT
See also
MPU6050_PWR1_CLKSEL_LENGTH
- Returns:
Current clock source setting
-
inline void setClockSource(uint8_t source)
Set clock source setting. An internal 8MHz oscillator, gyroscope based clock, or external sources can be selected as the MPU-60X0 clock source. When the internal 8 MHz oscillator or an external source is chosen as the clock source, the MPU-60X0 can operate in low power modes with the gyroscopes disabled.
Upon power up, the MPU-60X0 clock source defaults to the internal oscillator. However, it is highly recommended that the device be configured to use one of the gyroscopes (or an external clock source) as the clock reference for improved stability. The clock source can be selected according to the following table:
See also
See also
MPU6050_RA_PWR_MGMT_1
See also
MPU6050_PWR1_CLKSEL_BIT
See also
MPU6050_PWR1_CLKSEL_LENGTH
- Parameters:
source – New clock source setting
-
inline uint8_t getWakeFrequency()
Get wake frequency in Accel-Only Low Power Mode. The MPU-60X0 can be put into Accerlerometer Only Low Power Mode by setting PWRSEL to 1 in the Power Management 1 register (Register 107). In this mode, the device will power off all devices except for the primary I2C interface, waking only the accelerometer at fixed intervals to take a single measurement. The frequency of wake-ups can be configured with LP_WAKE_CTRL as shown below:
For further information regarding the MPU-60X0’s power modes, please refer to Register 107.
See also
MPU6050_RA_PWR_MGMT_2
- Returns:
Current wake frequency
-
inline void setWakeFrequency(uint8_t frequency)
Set wake frequency in Accel-Only Low Power Mode.
See also
MPU6050_RA_PWR_MGMT_2
- Parameters:
frequency – New wake frequency
-
inline bool getStandbyXAccelEnabled()
Get X-axis accelerometer standby enabled status. If enabled, the X-axis will not gather or report data (or use power).
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_XA_BIT
- Returns:
Current X-axis standby enabled status
-
inline void setStandbyXAccelEnabled(bool enabled)
Set X-axis accelerometer standby enabled status.
See also
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_XA_BIT
- Parameters:
New – X-axis standby enabled status
-
inline bool getStandbyYAccelEnabled()
Get Y-axis accelerometer standby enabled status. If enabled, the Y-axis will not gather or report data (or use power).
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_YA_BIT
- Returns:
Current Y-axis standby enabled status
-
inline void setStandbyYAccelEnabled(bool enabled)
Set Y-axis accelerometer standby enabled status.
See also
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_YA_BIT
- Parameters:
New – Y-axis standby enabled status
-
inline bool getStandbyZAccelEnabled()
Get Z-axis accelerometer standby enabled status. If enabled, the Z-axis will not gather or report data (or use power).
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_ZA_BIT
- Returns:
Current Z-axis standby enabled status
-
inline void setStandbyZAccelEnabled(bool enabled)
Set Z-axis accelerometer standby enabled status.
See also
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_ZA_BIT
- Parameters:
New – Z-axis standby enabled status
-
inline bool getStandbyXGyroEnabled()
Get X-axis gyroscope standby enabled status. If enabled, the X-axis will not gather or report data (or use power).
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_XG_BIT
- Returns:
Current X-axis standby enabled status
-
inline void setStandbyXGyroEnabled(bool enabled)
Set X-axis gyroscope standby enabled status.
See also
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_XG_BIT
- Parameters:
New – X-axis standby enabled status
-
inline bool getStandbyYGyroEnabled()
Get Y-axis gyroscope standby enabled status. If enabled, the Y-axis will not gather or report data (or use power).
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_YG_BIT
- Returns:
Current Y-axis standby enabled status
-
inline void setStandbyYGyroEnabled(bool enabled)
Set Y-axis gyroscope standby enabled status.
See also
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_YG_BIT
- Parameters:
New – Y-axis standby enabled status
-
inline bool getStandbyZGyroEnabled()
Get Z-axis gyroscope standby enabled status. If enabled, the Z-axis will not gather or report data (or use power).
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_ZG_BIT
- Returns:
Current Z-axis standby enabled status
-
inline void setStandbyZGyroEnabled(bool enabled)
Set Z-axis gyroscope standby enabled status.
See also
See also
MPU6050_RA_PWR_MGMT_2
See also
MPU6050_PWR2_STBY_ZG_BIT
- Parameters:
New – Z-axis standby enabled status
-
inline uint16_t getFIFOCount()
Get current FIFO buffer size. This value indicates the number of bytes stored in the FIFO buffer. This number is in turn the number of bytes that can be read from the FIFO buffer and it is directly proportional to the number of samples available given the set of sensor data bound to be stored in the FIFO (register 35 and 36).
- Returns:
Current FIFO buffer size
-
inline uint8_t getFIFOByte()
Get byte from FIFO buffer. This register is used to read and write data from the FIFO buffer. Data is written to the FIFO in order of register number (from lowest to highest). If all the FIFO enable flags (see below) are enabled and all External Sensor Data registers (Registers 73 to 96) are associated with a Slave device, the contents of registers 59 through 96 will be written in order at the Sample Rate.
The contents of the sensor data registers (Registers 59 to 96) are written into the FIFO buffer when their corresponding FIFO enable flags are set to 1 in FIFO_EN (Register 35). An additional flag for the sensor data registers associated with I2C Slave 3 can be found in I2C_MST_CTRL (Register 36).
If the FIFO buffer has overflowed, the status bit FIFO_OFLOW_INT is automatically set to 1. This bit is located in INT_STATUS (Register 58). When the FIFO buffer has overflowed, the oldest data will be lost and new data will be written to the FIFO.
If the FIFO buffer is empty, reading this register will return the last byte that was previously read from the FIFO until new data is available. The user should check FIFO_COUNT to ensure that the FIFO buffer is not read when empty.
- Returns:
Byte from FIFO buffer
-
inline void setFIFOByte(uint8_t data)
Write byte to FIFO buffer.
See also
See also
MPU6050_RA_FIFO_R_W
-
inline uint8_t getDeviceID()
Get Device ID. This register is used to verify the identity of the device (0b110100, 0x34).
See also
MPU6050_RA_WHO_AM_I
See also
MPU6050_WHO_AM_I_BIT
See also
MPU6050_WHO_AM_I_LENGTH
- Returns:
Device ID (6 bits only! should be 0x34)
-
inline void setDeviceID(uint8_t id)
Set Device ID. Write a new ID into the WHO_AM_I register (no idea why this should ever be necessary though).
See also
See also
MPU6050_RA_WHO_AM_I
See also
MPU6050_WHO_AM_I_BIT
See also
MPU6050_WHO_AM_I_LENGTH
- Parameters:
id – New device ID to set.
-
struct Motion3
-
struct Motion6
-
inline MPU6050()
References
Used by
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Mirf for NRF24L01
Arduino port of Mirf for the NRF24L01 modules
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
ModbusMaster RTU Library
This library handles Modbus master RTU communication. The library supports callbacks for pre- and post-transmission, receive and transmit logging.
- MB_RESPONSE_TIMEOUT
The patch provides changeable response timeout using
MB_RESPONSE_TIMEOUT(in milliseconds).
The original author of the library is 4-20ma: https://github.com/4-20ma/ModbusMaster/
The one included in Sming is nomis’ fork: https://github.com/nomis/ModbusMaster (see branch fixes-2.0.1) that isn’t yet merged to the original repository.
References
Used by
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: ModbusMaster
ModbusMaster
` .. image:: https://img.shields.io/github/release/4-20ma/ModbusMaster.svg?maxAge=3600
- target:
https://img.shields.io/github/release/4-20ma/ModbusMaster.svg?maxAge=3600
- alt:
GitHub release
` .. image:: https://img.shields.io/travis/4-20ma/ModbusMaster.svg?maxAge=3600
` .. image:: https://img.shields.io/github/license/4-20ma/ModbusMaster.svg?maxAge=3600
- target:
https://img.shields.io/github/license/4-20ma/ModbusMaster.svg?maxAge=3600
- alt:
license
<LICENSE>`_
` .. image:: https://img.shields.io/badge/%E2%9D%A4-code%20of%20conduct-blue.svg?maxAge=3600
- target:
https://img.shields.io/badge/%E2%9D%A4-code%20of%20conduct-blue.svg?maxAge=3600
- alt:
code of conduct
<CODE_OF_CONDUCT.md>`_
Overview
This is an Arduino library for communicating with Modbus slaves over RS232/485 (via RTU protocol).
Features
The following Modbus functions are available:
Discrete Coils/Flags
0x01 - Read Coils
0x02 - Read Discrete Inputs
0x05 - Write Single Coil
0x0F - Write Multiple Coils
Registers
0x03 - Read Holding Registers
0x04 - Read Input Registers
0x06 - Write Single Register
0x10 - Write Multiple Registers
0x16 - Mask Write Register
0x17 - Read Write Multiple Registers
Both full-duplex and half-duplex RS232/485 transceivers are supported. Callback functions are provided to toggle Data Enable (DE) and Receiver Enable (/RE) pins.
Installation
Library Manager
Install the library into your Arduino IDE using the Library Manager (available from IDE version 1.6.2). Open the IDE and click Sketch > Include Library > Manage Libraries…
Scroll or search for ModbusMaster, then select the version of the library you want to install. Quit/re-launch the IDE to refresh the list; new versions are automatically added to the list, once released on GitHub.
Refer to Arduino Tutorials > Libraries Using the Library Manager.
Zip Library
Refer to Arduino Tutorials > Libraries Importing a .zip Library.
Manual
Refer to Arduino Tutorials > Libraries Manual Installation.
Hardware
This library has been tested with an Arduino Duemilanove, PHOENIX CONTACT nanoLine controller, connected via RS485 using a Maxim MAX488EPA transceiver.
Caveats
Conforms to Arduino IDE 1.5 Library Specification v2.1 which requires Arduino IDE >= 1.5.
Arduinos prior to the Mega have one serial port which must be connected to USB (FTDI) for uploading sketches and to the RS232/485 device/network for running sketches. You will need to disconnect pin 0 (RX) while uploading sketches. After a successful upload, you can reconnect pin 0.
Support
Please submit an issue for all questions, bug reports, and feature requests. Email requests will be politely redirected to the issue tracker so others may contribute to the discussion and requestors get a more timely response.
Example
The library contains a few sketches that demonstrate use of the ModbusMaster library. You can find these in the examples folder.
/*
Basic.pde - example using ModbusMaster library
Library:: ModbusMaster
Author:: Doc Walker <4-20ma@wvfans.net>
Copyright:: 2009-2016 Doc Walker
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
#include <ModbusMaster.h>
// instantiate ModbusMaster object
ModbusMaster node;
void setup()
{
// use Serial (port 0); initialize Modbus communication baud rate
Serial.begin(19200);
// communicate with Modbus slave ID 2 over Serial (port 0)
node.begin(2, Serial);
}
void loop()
{
static uint32_t i;
uint8_t j, result;
uint16_t data[6];
i++;
// set word 0 of TX buffer to least-significant word of counter (bits 15..0)
node.setTransmitBuffer(0, lowWord(i));
// set word 1 of TX buffer to most-significant word of counter (bits 31..16)
node.setTransmitBuffer(1, highWord(i));
// slave: write TX buffer to (2) 16-bit registers starting at register 0
result = node.writeMultipleRegisters(0, 2);
// slave: read (6) 16-bit registers starting at register 2 to RX buffer
result = node.readHoldingRegisters(2, 6);
// do something with data if read is successful
if (result == node.ku8MBSuccess)
{
for (j = 0; j < 6; j++)
{
data[j] = node.getResponseBuffer(j);
}
}
}
Project inspired by `Arduino Modbus Master <http://sites.google.com/site/jpmzometa/arduino-mbrt/arduino-modbus-master>`_.
Multipart Parser
Component to manage processing of multipart form data according to RFC7578, mostly used to support file uploads via HTTP/POST.
Usage
While setting up your web server, register the body parser provided by the library:
#include <MultipartParser.h>
HttpServer server;
...
server.setBodyParser(MIME_FORM_MULTIPART, formMultipartParser);
Now add a HttpMultipartResource for the path to receive the multipart data:
void fileUploadMapper(HttpFiles& files)
{
files["file"] = <writable_stream_to_process_file>;
}
int onUpload(HttpServerConnection& connection, HttpRequest& request, HttpResponse& response)
{
// `file` points to the stream from `fileUploadMapper`.
auto* file = request.files["file"];
// TODO: use methods of `file` to check if data was processed successfully
// TODO: setup HTTP response
return 0;
}
...
server.paths.set("/upload", new HttpMultipartResource(fileUploadMapper, onUpload));
See HttpServer Firmware Upload for further details.
Upgrade Notes
The functionality provided by this lirbary was previously controlled by the config option ENABLE_HTTP_SERVER_MULTIPART.
To upgrade, you have to replace:
ENABLE_HTTP_SERVER_MULTIPART := 1
by:
ARDUINO_LIBRARIES += MultipartParser
in your component.mk. In addition, body parser registration must now be done explicitly by application (see above).
API Documentation
MultipartParser API
-
size_t formMultipartParser(HttpRequest &request, const char *at, int length)
Body parser for content-type
form-data/multipartMust be added to the web server’s list of body parsers explicitly:
#include <MultipartParser.h> ... HttpServer server; ... server.setBodyParser(MIME_FORM_MULTIPART, formMultipartParser);
-
class HttpMultipartResource : public HttpResource
HttpResource that allows handling of HTTP file upload.
Public Functions
-
inline HttpMultipartResource(const HttpFilesMapper &mapper, HttpResourceDelegate complete)
Create and configure a HttpResource for handling file upload.
On a normal computer the file uploads are usually using temporary space on the hard disk or in memory to store the incoming data. On an embedded device that is a luxury that we can hardly afford. Therefore we should define a
mapthat specifies explicitly where every form field will be stored. If a field is not specified then its content will be discarded.- Parameters:
mapper – callback that provides information where the desired upload fields will be stored.
complete – callback that will be called after the request has completed.
-
virtual int setFileMap(HttpServerConnection &connection, HttpRequest &request, HttpResponse &response)
Callback implementation for HttpResource::onHeadersComplete. Not to be used by application code.
-
virtual void shutdown(HttpServerConnection &connection) override
Takes care to cleanup the connection.
-
inline HttpMultipartResource(const HttpFilesMapper &mapper, HttpResourceDelegate complete)
References
Used by
HttpServer Upload Sample
Advanced_Jsvm Sample
HttpServer Firmware Upload Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: multipart-parser
Multipart form data parser
Features
No dependencies
Works with chunks of a data - no need to buffer the whole request
Almost no internal buffering. Buffer size doesn’t exceed the size of the boundary (~60-70 bytes)
Tested as part of Cosmonaut HTTP server.
Implementation based on node-formidable by Felix Geisendörfer.
Inspired by http-parser by Ryan Dahl.
Usage (C)
This parser library works with several callbacks, which the user may set up at application initialization time.
multipart_parser_settings callbacks;
memset(&callbacks, 0, sizeof(multipart_parser_settings));
callbacks.on_header_field = read_header_name;
callbacks.on_header_value = read_header_value;
These functions must match the signatures defined in the multipart-parser header file. For this simple example, we’ll just use two of the available callbacks to print all headers the library finds in multipart messages.
Returning a value other than 0 from the callbacks will abort message processing.
int read_header_name(multipart_parser* p, const char *at, size_t length)
{
printf("%.*s: ", length, at);
return 0;
}
int read_header_value(multipart_parser* p, const char *at, size_t length)
{
printf("%.*s\n", length, at);
return 0;
}
When a message arrives, callers must parse the multipart boundary from the Content-Type header (see the RFC for more information and examples), and then execute the parser.
multipart_parser* parser = multipart_parser_init(boundary, &callbacks);
multipart_parser_execute(parser, body, length);
multipart_parser_free(parser);
Usage (C++)
In C++, when the callbacks are static member functions it may be helpful to pass the instantiated multipart consumer along as context. The following (abbreviated) class called MultipartConsumer shows how to pass this to callback functions in order to access non-static member data.
class MultipartConsumer
{
public:
MultipartConsumer(const std::string& boundary)
{
memset(&m_callbacks, 0, sizeof(multipart_parser_settings));
m_callbacks.on_header_field = ReadHeaderName;
m_callbacks.on_header_value = ReadHeaderValue;
m_parser = multipart_parser_init(boundary.c_str(), &m_callbacks);
multipart_parser_set_data(m_parser, this);
}
~MultipartConsumer()
{
multipart_parser_free(m_parser);
}
int CountHeaders(const std::string& body)
{
multipart_parser_execute(m_parser, body.c_str(), body.size());
return m_headers;
}
private:
static int ReadHeaderName(multipart_parser* p, const char *at, size_t length)
{
MultipartConsumer* me = (MultipartConsumer*)multipart_parser_get_data(p);
me->m_headers++;
}
multipart_parser* m_parser;
multipart_parser_settings m_callbacks;
int m_headers;
};
Contributors
© 2012 Igor Afonov
ESP32 NimBLE
Introduction
NimBLE is a completely open source Bluetooth Low Energy (BLE) stack produced by Apache. It is more suited to resource constrained devices than bluedroid and has now been ported to the ESP32 by Espressif.
Using
Add
COMPONENT_DEPENDS += NimBLEto your application component.mk file.Add these lines to your application:
#include <NimBLEDevice.h> void init() { // ... BLEDevice::init(""); }
References
Used by
ESP32 BLE Gamepad Library
ESP32 BLE Keyboard Library
Switch Joycon Library
SoC support
esp32
esp32c3
esp32s3
Submodule: esp-nimble-cpp
`Latest release .. image:: https://img.shields.io/github/release/h2zero/esp-nimble-cpp.svg?style=plastic
- target:
https://img.shields.io/github/release/h2zero/esp-nimble-cpp.svg?style=plastic
- alt:
Release Version
Need help? Have questions or suggestions? Join the .. image:: https://badges.gitter.im/NimBLE-Arduino/community.svg
esp-nimble-cpp
NimBLE CPP library for use with ESP32 that attempts to maintain compatibility with the nkolban cpp_uitls BLE API.
**An Arduino version of this library, including NimBLE, can be found here.**
This library significantly reduces resource usage and improves performance for ESP32 BLE applications as compared
with the bluedroid based library. The goal is to maintain, as much as reasonable, compatibility with the original
library but refactored to use the NimBLE stack. In addition, this library will be more actively developed and maintained
to provide improved capabilites and stability over the original.
Testing shows a nearly 50% reduction in flash use and approx. 100kB less ram consumed vs the original!
Your results may vary
What is NimBLE?
NimBLE is a completely open source Bluetooth Low Energy stack produced by Apache.
It is more suited to resource constrained devices than bluedroid and has now been ported to the ESP32 by Espressif.
Installation
ESP-IDF v4.0+
Download as .zip and extract or clone into the components folder in your esp-idf project.
Run menuconfig, go to Component config->Bluetooth enable Bluetooth and in Bluetooth host NimBLE.
Configure settings in NimBLE Options.
#include "NimBLEDevice.h" in main.cpp.
Call NimBLEDevice::init(""); in app_main.
ESP-IDF v3.2 & v3.3
The NimBLE component does not come with these versions of IDF (now included in 3.3.2 and above).
A backport that works in these versions has been created and is available here.
Download or clone that repo into your project/components folder and run menuconfig.
Configure settings in main menu -> NimBLE Options.
#include "NimBLEDevice.h" in main.cpp.
Call NimBLEDevice::init(""); in app_main.
Using
This library is intended to be compatible with the original ESP32 BLE functions and types with minor changes.
If you have not used the original Bluedroid library please refer to the New user guide.
If you are familiar with the original library, see: The migration guide for details about breaking changes and migration.
Also see Improvements_and_updates for information about non-breaking changes.
Full API documentation and class list can be found here.
When using this library along with Arduino and compiling with CMake you must add add_compile_definitions(ARDUINO_ARCH_ESP32=1)
in your project/CMakeLists.txt after the line include($ENV{IDF_PATH}/tools/cmake/project.cmake) to prevent Arduino from releasing BLE memory.
Acknowledgments
nkolban and chegewara for the original esp32 BLE library this project was derived from.
beegee-tokyo for contributing your time to test/debug and contributing the beacon examples.
Jeroen88 for the amazing help debugging and improving the client code.
OneWire for Arduino
https://github.com/PaulStoffregen/OneWire.git
References
Used by
DS18S20 Temperature Sensor Library
Basic Web Skeleton (LTS) Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Over-The-Air(OTA) Upgrader
Introduction
This architecture-agnostic component adds support for Over-The-Air upgrades.
Usage
Add
COMPONENT_DEPENDS += Otato your application component.mk file.Add these lines to your application:
#include <Ota/Manager.h>
After that you will have access to a global OtaManager instance that can be used to manage your OTA upgrade process.
You can use
OtaManagerto get information about the bootable partitions and update them. The code below will display the current bootable and running partition:void init() { // ... auto partition = OtaManager.getRunningPartition(); Serial.printf("\r\nCurrently running %s @ 0x%08x.\r\n", partition.name().c_str(), partition.address()); }
If needed you can also create your own instance of the of OtaUpgrader as shown below:
// Call when IP address has been obtained void onIp(IpAddress ip, IpAddress mask, IpAddress gateway) { // ... OtaUpgrader ota; auto part = ota.getNextBootPartition(); ota.begin(part); // ... write all the data to the partition ota.end(); // ... }
See the Basic Ota sample application.
API Documentation
-
namespace Ota
-
class IdfUpgrader : public Ota::UpgraderBase
- #include <IdfUpgrader.h>
ESP32 OTA Upgrader implementation.
Public Functions
-
virtual bool begin(Partition partition, size_t size = 0) override
Prepares a partition for an upgrade. The preparation is bootloader and architecture dependent.
- Parameters:
partition –
size –
- Return values:
bool –
-
virtual size_t write(const uint8_t *buffer, size_t size) override
Writes chunk of data to the partition set in
begin().- Parameters:
buffer –
size –
- Return values:
size_t – actually written bytes
-
inline virtual bool end() override
Finalizes the partition upgrade.
-
inline virtual bool abort() override
Aborts a partition upgrade.
-
inline virtual bool setBootPartition(Partition partition, bool save = true) override
Sets the default partition from where the application will be booted on next restart.
- Parameters:
partition –
save – if true the change is persisted on the flash, otherwise it will be valid only for the next boot
- Return values:
bool –
-
inline virtual Partition getBootPartition() override
Gets information about the partition that is set as the default one to boot.
Note
The returned partition can be different than the current running partition.
- Return values:
partition –
-
inline virtual Partition getRunningPartition() override
Gets information about the partition from which the current application is running.
Note
The returned partition can be different than the default boot partition.
- Return values:
partition –
-
inline virtual Partition getNextBootPartition(Partition startFrom = {}) override
Gets the next bootable partition that can be used after successful OTA upgrade.
- Parameters:
startFrom – - optional
- Return values:
partition –
-
virtual bool begin(Partition partition, size_t size = 0) override
-
class RbootUpgrader : public Ota::UpgraderBase
- #include <RbootUpgrader.h>
ESP8266 rBoot OTA Upgrader implementation.
Public Functions
-
virtual bool begin(Partition partition, size_t size = 0) override
Prepare the partition for.
-
virtual size_t write(const uint8_t *buffer, size_t size) override
Writes chunk of data to the partition set in
begin().- Parameters:
buffer –
size –
- Return values:
size_t – actually written bytes
-
inline virtual bool end() override
Finalizes the partition upgrade.
-
virtual bool setBootPartition(Partition partition, bool save = true) override
Sets the default partition from where the application will be booted on next restart.
- Parameters:
partition –
save – if true the change is persisted on the flash, otherwise it will be valid only for the next boot
- Return values:
bool –
-
inline virtual Partition getBootPartition() override
Gets information about the partition that is set as the default one to boot.
Note
The returned partition can be different than the current running partition.
- Return values:
partition –
-
virtual Partition getRunningPartition() override
Gets information about the partition from which the current application is running.
Note
The returned partition can be different than the default boot partition.
- Return values:
partition –
-
inline virtual Partition getNextBootPartition(Partition startFrom = {}) override
Gets the next bootable partition that can be used after successful OTA upgrade.
- Parameters:
startFrom – - optional
- Return values:
partition –
-
virtual bool begin(Partition partition, size_t size = 0) override
-
class UpgradeOutputStream : public ReadWriteStream
- #include <UpgradeOutputStream.h>
Write-only stream type used during firmware upgrade.
Public Functions
-
inline UpgradeOutputStream(Partition partition, size_t maxLength = 0)
Construct a stream for the given partition.
- Parameters:
partition –
-
virtual size_t write(const uint8_t *data, size_t size) override
Write chars to stream.
Note
Although this is defined in the Print class, ReadWriteStream uses this as the core output method so descendants are required to implement it
- Parameters:
buffer – Pointer to buffer to write to the stream
size – Quantity of chars to write
- Return values:
size_t – Quantity of chars written to stream
-
inline virtual StreamType getStreamType() const override
Get the stream type.
- Return values:
StreamType – The stream type.
-
inline virtual uint16_t readMemoryBlock(char *data, int bufSize) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
-
inline virtual bool seek(int len) override
Move read cursor.
- Parameters:
len – Relative cursor adjustment
- Return values:
bool – True on success.
-
inline virtual int available() override
Return the total length of the stream.
- Return values:
int – -1 is returned when the size cannot be determined
-
inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
-
inline UpgradeOutputStream(Partition partition, size_t maxLength = 0)
-
class UpgraderBase
- #include <UpgraderBase.h>
Subclassed by Ota::IdfUpgrader, Ota::RbootUpgrader
Public Functions
-
virtual bool begin(Partition partition, size_t size = 0) = 0
Prepares a partition for an upgrade. The preparation is bootloader and architecture dependent.
- Parameters:
partition –
size –
- Return values:
bool –
-
virtual size_t write(const uint8_t *buffer, size_t size) = 0
Writes chunk of data to the partition set in
begin().- Parameters:
buffer –
size –
- Return values:
size_t – actually written bytes
-
virtual bool end() = 0
Finalizes the partition upgrade.
-
inline virtual bool abort()
Aborts a partition upgrade.
-
virtual bool setBootPartition(Partition partition, bool save = true) = 0
Sets the default partition from where the application will be booted on next restart.
- Parameters:
partition –
save – if true the change is persisted on the flash, otherwise it will be valid only for the next boot
- Return values:
bool –
-
virtual Partition getBootPartition() = 0
Gets information about the partition that is set as the default one to boot.
Note
The returned partition can be different than the current running partition.
- Return values:
partition –
-
virtual Partition getRunningPartition() = 0
Gets information about the partition from which the current application is running.
Note
The returned partition can be different than the default boot partition.
- Return values:
partition –
-
virtual Partition getNextBootPartition(Partition startFrom = {}) = 0
Gets the next bootable partition that can be used after successful OTA upgrade.
- Parameters:
startFrom – - optional
- Return values:
partition –
-
inline Storage::Iterator getBootPartitions()
Gets information about all bootable partitions.
- Return values:
Storage::Iterator –
-
inline uint8_t getSlot(Partition partition)
Gets slot number for a partition.
- Parameters:
partition –
- Return values:
uint8_t – slot number
-
virtual bool begin(Partition partition, size_t size = 0) = 0
-
namespace Network
Variables
-
constexpr uint8_t NO_ROM_SWITCH = {0xff}
Magic value for ROM slot indicating slot won’t change after successful OTA.
-
class HttpUpgrader : protected HttpClient
- #include <HttpUpgrader.h>
Public Functions
-
inline bool addItem(const String &firmwareFileUrl, Partition partition, ReadWriteStream *stream = nullptr)
Add an item to update.
- Parameters:
firmwareFileUrl –
partition – Target partition to write
stream –
- Return values:
bool –
-
inline void switchToRom(uint8_t romSlot)
On completion, switch to the given ROM slot.
- Parameters:
romSlot – specify NO_ROM_SWITCH (the default) to cancel any previously set switch
-
inline void setBaseRequest(HttpRequest *request)
Sets the base request that can be used to pass.
- default request parameters, like request headers... - default SSL options - default SSL fingeprints - default SSL client certificates
- Parameters:
request –
-
struct Item
- #include <HttpUpgrader.h>
-
inline bool addItem(const String &firmwareFileUrl, Partition partition, ReadWriteStream *stream = nullptr)
-
constexpr uint8_t NO_ROM_SWITCH = {0xff}
-
class IdfUpgrader : public Ota::UpgraderBase
References
Used by
Over-the-Air Firmware Upgrade Library
OTA Firmware Upgrade via MQTT Library
Environment Variables
SoC support
esp8266
host
Over-The-Air(OTA) Network Upgrader
Introduction
This architecture-agnostic component adds support for Over-The-Air upgrades.
Usage
Add
COMPONENT_DEPENDS += Otato your application componenent.mk file.Add these lines to your application:
#include <Ota/Manager.h>
After that you will have access to a global OtaManager instance that can be used to manage your OTA upgrade process.
You can use
OtaManagerto get information about the bootable partitions and update them. The code below will display the current bootable and running partition:void init() { // ... auto partition = OtaManager.getRunningPartition(); Serial.printf("\r\nCurrently running %s @ 0x%08x.\r\n", partition.name().c_str(), partition.address()); }
If needed you can also create your own instance of the of OtaUpgrader as shown below:
// Call when IP address has been obtained void onIp(IpAddress ip, IpAddress mask, IpAddress gateway) { // ... OtaUpgrader ota; auto partition = ota.getNextBootPartition(); ota.begin(partition); // ... write all the data to the partition ota.end(); // ... }
See the Basic Ota sample application.
API Documentation
-
namespace Network
Variables
-
constexpr uint8_t NO_ROM_SWITCH = {0xff}
Magic value for ROM slot indicating slot won’t change after successful OTA.
-
class HttpUpgrader : protected HttpClient
- #include <HttpUpgrader.h>
Public Functions
-
inline bool addItem(const String &firmwareFileUrl, Partition partition, ReadWriteStream *stream = nullptr)
Add an item to update.
- Parameters:
firmwareFileUrl –
partition – Target partition to write
stream –
- Return values:
bool –
-
inline void switchToRom(uint8_t romSlot)
On completion, switch to the given ROM slot.
- Parameters:
romSlot – specify NO_ROM_SWITCH (the default) to cancel any previously set switch
-
inline void setBaseRequest(HttpRequest *request)
Sets the base request that can be used to pass.
- default request parameters, like request headers... - default SSL options - default SSL fingeprints - default SSL client certificates
- Parameters:
request –
-
inline const ItemList &getItems() const
Allow read access to item list.
-
struct Item
- #include <HttpUpgrader.h>
-
inline bool addItem(const String &firmwareFileUrl, Partition partition, ReadWriteStream *stream = nullptr)
-
constexpr uint8_t NO_ROM_SWITCH = {0xff}
References
Used by
Basic Ota Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
Over-the-Air Firmware Upgrade
This library offers a writable stream that decodes and applies [Over-the-Air] firmware upgrade files, as well as a small python utility to generate those upgrade files as part of Sming’s build process. It may be combined with any transport mechanism that is compatible with Sming’s stream classes. Check out the HttpServer Firmware Upload example, which demonstrates how the integrated HTTP server can be used to provide a web form for uploading new firmware images from the browser.
Prerequisites
Every in-system firmware upgrade mechanism for ESP-based devices requires partitioning the flash into two slots: One slot holds the currently running firmware, while the other slot receives the upgrade. As a consequence only half of the available flash memory can be used for the firmware. (Actually a bit less because a few sectors are reserved for the bootloader and various parameter blobs.)
In most cases, it is sufficient to set RBOOT_ROM1_ADDR to the offset address of the second slot.
See the rBoot documentation for further options and considerations.
If your partitioning choice results in two ROM images being created, they are transparently combined such that there
is always a single OTA upgrade file. During the upgrade, the OTA code will automatically select the right image and
ignore the one for the other slot.
Attention
Make sure that the ROM slots do not overlap with each other or with areas of the flash allocated to other purposes (file system, RF calibration parameters, etc.). Sming will not detect a misconfigured flash layout.
Security features leverage libsodium, which is automatically pulled in as a dependency when signing and/or
encryption is enabled. You also have to install libsodium bindings for python on your development computer, either
using python -m pip install PyNaCl (recommended for Windows users) or, if your are on Linux, preferably via your
distribution’s package manager (search for a package named ‘python-nacl’).
Usage
The OtaUpgradeStream class
The library provides the class OtaUpgradeStream (actually, an alias for either
OtaUpgrade::BasicStream or OtaUpgrade::EncryptedStream, depending on
ENABLE_OTA_ENCRYPTION.), which derives from ReadWriteStream, but, despite its base class, is only
a writable stream.
At construction time, the address and size of the slot to receive the new firmware is automatically determined from the
rBoot configuration. No further setup is required. Just feed the OTA upgrade file into the
OtaUpgradeStream::write method in arbitrarily sized chunks. The flash
memory is updated on the fly as data arrives and upon successful validation, the updated slot is activated in the rRoot
configuration.
Once the file is complete, call OtaUpgradeStream::hasError to check for
any errors that might have occurred during the upgrade process. The actual error, if any, is stored in the public member
OtaUpgradeStream::errorCode and can be converted to an error message
using OtaUpgradeStream::errorToString.
In addition, you may also examine the return value of the
OtaUpgradeStream::write method, which will be equal to the given chunk
size, unless there is an error with the file or the upgrade process.
Building
The library is fully integrated into the Sming build process. Just run:
make
and find the OTA upgrade file in out/<arch>/<config>/firmware/firmware.ota.
If security features are enabled but no secret key file does exist yet, a new one is generated during the first build.
You may change it later by modifying OTA_KEY or using the Key/Settings rollover process.
Now install the OTA-enabled firmware once via USB/Serial cable and you are all set to do future upgrades wirelessly over your chosen communication channel.
A convenience target:
make ota-upload OTA_UPGRADE_URL=http://<your-ip>/upgrade
is provided for the not too uncommon use case of uploading the OTA file as a HTTP/POST request (but obviously is of no value for other transport mechanisms). The URL is cached and can be omitted from subsequent invocations.
Configuration and Security features
- ENABLE_OTA_SIGNING
Default: 1 (enabled)
If set to 1 (highly recommended), OTA upgrade files are protected against unauthorized modification by a digital signature. This is implemented using libsodium’s crypto_verify_… API, which encapsulates a public key algorithm: A secret (or ‘private’) signing key never leaves the development computer, while a non-secret (‘public’) verification key is embedded into the firmware. Public key algorithms cannot be broken even if an attacker gains physical access to one of your devices and extracts the verification key from flash memory, because only someone in possession of the secret signing key (see
OTA_KEY) is able to create upgrade files with a valid signature.Note
You may disable signing in order to save some program memory if your communication channel already establishes a comparable level of trust, e.g. TLS with a pinned certificate.
- OTA_ENABLE_ENCRYPTION
Default: 0 (disabled)
Set to 1 to enable encryption of the upgrade file using libsodium’s crypto_secretstream_… API, in order to protect confidential data embedded in your firmware (WiFi credentials, server certificates, etc.).
It is generally unnecessary to sign encrypted upgrade files, as encryption is also authenticating, i.e. only someone in possession of the secret encryption key can generate upgrade files that decrypt successfully. There is, however, one catch: Unlike signing, encryption can be broken if an attacker is able to extract the decryption key (which is identical to the encryption key) from flash memory, in which case all current and future files encrypted with the same key are compromised. Moreover, the attacker will be able to generate new valid upgrade files modified to his or her agenda. Hence, you should only ever rely on encryption if it is impossible for an attacker to gain physical access to your device(s). But otherwise, you shouldn’t have stored confidential data on such device(s) in the first place. Conversely, you should not encrypt upgrade files that do not contain confidential data, to avoid the risk of accidentally exposing a key you might want to reuse later. For this reason, encryption is disabled by default.
Note: To mitigate a catastrophic security breach when the encryption key is revealed involuntarily, encryption and signing can be enabled at the same time. This way, an attacker (who probably has access to your WiFi by now) will at least be unable to take over more devices wirelessly. But keep in mind: it is still not a good idea to store confidential data on an unsecured device.
Note also that the described weakness is not a property of the selected encryption algorithm, but a rather general one. It can only be overcome by encrypting the communication channel instead of the upgrade file, e.g. with TLS, which uses a key exchange protocol to negotiate a temporary encryption key that is never written to flash memory. But even then, it is still unwise to embed confidential data into the firmware of a device that is physically accessible to an attacker - now you have been warned!
- OTA_KEY
Path to the secret encryption/signing key. The default is
ota.keyin the root directory of your project. If the key file does not exist, it will be generated during the first build. It can also be (re-)generated manually using the following command (usually as part of a Key/Settings rollover process):make ota-genkeyThe key file must be kept secret for obvious reasons. In particular, set up your .gitignore (or equivalent VCS mechanism) carefully to avoid accidentally pushing the key file to a public repository.
By pointing
OTA_KEYto a shared location, the same key file can be used for multiple projects, even if their security settings differ, since the key file format is independent of the security settings. (In fact, it is just a string of random numbers, from which the actual algorithm keys are derived.)
- ENABLE_OTA_DOWNGRADE
Default: 0 (disabled)
By default,
OtaUpgradeStreamrefuses to downgrade to an older firmware version, in order to prevent an attacker from restoring already patched security vulnerabilities. This is implemented by comparing timestamps embedded in the firmware and the upgrade file. To disable downgrade protection, set ENABLE_OTA_DOWNGRADE to 1.Downgrade protection must be combined with encryption or signing to be effective. A warning is issued by the build system otherwise.
- OTA_UPLOAD_URL
URL used by the
make ota-uploadcommand.
- OTA_UPLOAD_NAME
Field name for the upgrade file in the HTTP/POST request issued by
make ota-upload, corresponding to thenameattribute of the HTML input element:<input type="file" name="firmware" />
The default is “firmware”.
Key/Settings rollover process
There might be occasions where you want to change the encryption/signing key and or other OTA security settings (e.g. switch from signing to encryption or vice versa). While you could always install the new settings via USB/serial cable, you can also follow the steps below to achieve the same goal wirelessly:
Before modifying any security-related settings, start the rollover process by issuing:
make ota-rolloverNow modify security settings as desired, e.g. generate a new key using
make ota-genkey.Run
maketo build a rollover upgrade file. The firmware image(s) contained in this file use the new security settings, while the upgrade file itself is created with the old settings (saved by the command in step 1) and thus is still compatible with the firmware currently running on your device(s).Upgrade wirelessly using the rollover file created in step 3. The new security settings are now installed.
Finalize the rollover process using the command:
make ota-rollover-doneThis will delete temporary files created by step 1.
OTA upgrade file format
Basic file format
The following layout is used for unencrypted upgrade files, as well as for the data inside the encrypted container (see next paragraph). All fields are stored in little-endian byte order.
Field size (bytes) |
Field description |
|---|---|
4 |
Magic number for file format identification:
0xf01af02a for signed images0xf01af020 for images without signature |
8 |
OTA upgrade file timestamp in milliseconds since 1900/01/01 (used for downgrade protection) |
1 |
Number of ROM images (1 or 2) |
3 |
reserved, always zero |
variable |
ROM images, see below |
64 (signed)
16 (otherwise)
|
With signature: Digital signature over the whole file up to this point.
Otherwise: MD5 HASH over the whole file up to this point. This is
not a security measure but merely protects the integrity of the file. MD5
was selected, because it already available in the ESP8266’s on-chip ROM.
|
Each ROM image has the following format:
Field size (bytes) |
Field description |
|---|---|
4 |
Start address in flash memory (i.e. |
4 |
Size of ROM in bytes |
variable (see previous field) |
ROM image content |
More content may be added in a future version (e.g. SPIFFS images, bootloader image, RF calibration data blob). The reserved bytes in the file header are intended to announce such additional content.
Encryption Container format
Encrypted files are stored in chunks suitable for consumption by libsodium’s crypto_secretstream_… API.
The first chunk is always 24 bytes and is fed into crypto_secretstream_pull_init to initialize the decryption
algorithm.
Subsequent chunks are composed of:
A 2 byte header indicating the length of the chunk minus 1. The default chunk size used by otatool.py is 2 kB.
The data of the chunk, which is fed into
crypto_secretstream_pull.
For further information on the data stored in the header and the chunks, refer to libsodium’s documentation and/or source code.
API Documentation
OTA Upgrade Stream classes
-
typedef OtaUpgrade::BasicStream OtaUpgradeStream
Alias for either
OtaUpgrade::BasicStreamorOtaUpgrade::EncryptedStream, depending on encryption settings.Application code should use this alias to avoid source code modifications when changing OTA upgrade security settings.
-
class BasicStream : public ReadWriteStream
A write-only stream to parse and apply firmware unencrypted upgrade files generated by otatool.py.
The class fully automates the firmware upgrade process without any manual configuration. At construction time, the rBoot configuration is read to determine the unused ROM slot which should receive the upgrade. Just feed the upgrade file content into the
write()method in arbitrarily sized chunks. The relevant portion(s) of the Flash memory (currently only the application rom) are updated on the fly as data arrives. When the file is complete and signature validation (if enabled) was successful, the updated slot is activated in the rBoot configuration. CallhasError()and/or check the publicerrorCodemember to determine if everything went smoothly.For further information on configuration options and the file format, refer to the library’s documentation.
See also
EncryptedStreamfor encryption support.Subclassed by OtaUpgrade::EncryptedStream
-
class EncryptedStream : public OtaUpgrade::BasicStream
Encryption wrapper for BasicStream.
The class processes encrypted firmware upgrade files created by otatool.py. A buffer is allocated dynamically to fit the largest chunk of the encryption container (2kB unless otatool.py was modified). The actual processing of the decrypted data is deferred to BasicStream.
References
Used by
HttpServer Firmware Upload Sample
Environment Variables
ENABLE_OTA_ENCRYPTION
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
OTA Firmware Upgrade via MQTT
Introduction
This library allows Sming applications to upgrade their firmware Over-The-Air (OTA) using the MQTT protocol. MTQTT has less overhead compared to HTTP and can be used for faster delivery of application updates.
Using
Add
COMPONENT_DEPENDS += OtaUpgradeMqttto your application componenent.mk file.Add these lines to your application:
#include <OtaUpgrade/Mqtt/RbootPayloadParser.h> #if ENABLE_OTA_ADVANCED #include <OtaUpgrade/Mqtt/AdvancedPayloadParser.h> #endif MqttClient mqtt; // Call when IP address has been obtained void onIp(IpAddress ip, IpAddress mask, IpAddress gateway) { // ... mqtt.connect(Url(MQTT_URL), "sming"); #if ENABLE_OTA_ADVANCED /* * The advanced parser suppors all firmware upgrades supported by the `OtaUpgrade` library. * `OtaUpgrade` library provides firmware signing, firmware encryption and so on. */ auto parser = new OtaUpgrade::Mqtt::AdvancedPayloadParser(APP_VERSION_PATCH); #else /* * The command below uses class that stores the firmware directly * using RbootOutputStream on a location provided by us */ auto parser = new OtaUpgrade::Mqtt::RbootPayloadParser(part, APP_VERSION_PATCH); #endif mqtt.setPayloadParser([parser] (MqttPayloadParserState& state, mqtt_message_t* message, const char* buffer, int length) -> int { return parser->parse(state, message, buffer, length); }); String updateTopic = "a/test/u/4.3"; mqtt.subscribe(updateTopic); // ... }
See the OTA over MQTT sample application.
Versioning Principles
To simplify the OTA process we strongly recommend the following versioning principles for your application:
Use semantic versioning. If your current application version is 4.3.1 then 4 is the major, 3 is the minor and 1 is the patch version number.
Every application firmware knows its version.
An application with the same major and minor version should be compatible for update no matter what the patch number is. If the new firmware is not compatible then a new minor or major version should be used.
Theory Of Operation
On a period of time the application connects to check if there is a new version of the firmware. In your application this period has to be carefully selected so that OTA updates occur when the device has enough resources: memory, space on flash, power and time to complete such an update. Also there should be no critical task running at the moment. Depending on the size of the new firmware and the speed of the connection an update can take 10 to 20 seconds.
The application connects via MQTT to a remote server and subscribes to a special topic. The topic is based on the application id and its current version. If the current application id is
testand version is4.3.1then the topic that will be used for OTA isa/test/u/4.3.If there is a need to support both stable and unstable/nightly builds then the topic name can have s or u suffix. For example all stable versions should be published and downloaded from the topic
a/test/u/4.3/s. For the unstable ones we can use the topica/test/u/4.3/u. If an application is interested in both stable and unstable versions then it can subscribe using the following patterna/test/u/4.3/+.The application is waiting for new firmware. When the application is on battery than it makes sense to wait for a limited time and if there is no message coming back to disconnect.
Firmware packaging
The firmware update must come as one MQTT message. The MQTT protocol allows messages with a maximum size of 268435455 bytes approx 260MB. This should be perfectly enough for a device that has maximum 1MB available for an application ROM.
One MQTT message contains:
patch version of the firmware
followed by the firmware data itself
Based on the ENABLE_OTA_VARINT_VERSION the patch version can be encoded either using one byte or a varint.
Based on ENABLE_OTA_ADVANCED the firmware data can be either without any encoding or be signed and encrypted.
To simplify the packaging this library comes with a tool called deployer. To create a package type the following from your application:
make ota-pack OTA_PATCH_VERSION=127
Replace 127 with the desired patch version.
If the option OTA_PATCH_VERSION is omitted from the command line then the patch version will be generated automatically and it will contain the current unix timestamp.
Once a package is created it can be deployed to the firmware MQTT server using the command below:
make ota-deploy MQTT_FIRMWARE_URL=mqtt://relser:relpassword@attachix.com/a/test/u/4.3
The MQTT_FIRMWARE_URL above specifies that
protocol is: mqtt without SSL. Allowed values here are
mqttandmqtts. The latter uses SSL.user is: relser
password is: relpassword
host is: attachix.com
path is: /a/test/u/4.3. The path without leading and ending slashes is used to generate the topic name
a/test/u/4.3.
Make sure to replace the MQTT_FIRMWARE_URL value with your MQTT server credentials, host and topic.
Security
For additional security a standard SSL/TLS can be used
The communication should be secured using standard SSL.
To prove that the server is the correct one: The MQTT clients should pin the public key fingerprint on the server. OR have a list of public key fingerprints that are allowed.
To prove that the clients are allowed to connect: Every MQTT client should also have a client certificate that is signed by the server.
Configuration
- ENABLE_OTA_VARINT_VERSION
Default: 1 (enabled)
If set to 1 the OTA upgrade mechanism and application will use a varint encoding for the patch version. Thus allowing unlimited number of patch versions. Useful for enumerating unstable/nightly releases. A bit more difficult to read and write but allows for unlimited versions.
If set to 0 the OTA upgrade mechanism and application will use one byte for the patch version which will limit it to 256 possible patch versions. Useful for enumerating stable releases. Easier to write and read but limited to 256 versions only.
- ENABLE_OTA_ADVANCED
Default: 0 (disabled)
If set to 1 the library will work with OtaUpgradeStream which supports signature and encryption of the firmware data itself. See Over-the-Air Firmware Upgrade for details. In the application the AdvancedPayloadParser can be used to do the MQTT message handling.
API
API Documentation
-
namespace Mqtt
Typedefs
-
using RbootPayloadParser = StandardPayloadParser
- Deprecated:
Variables
-
constexpr int8_t VERSION_NOT_READY = {-1}
-
constexpr int8_t ERROR_INVALID_MQTT_MESSAGE = {-1}
-
constexpr int8_t ERROR_INVALID_PATCH_VERSION = {-2}
-
constexpr int8_t ERROR_UNKNOWN_REASON = {-10}
-
class AdvancedPayloadParser : public OtaUpgrade::Mqtt::PayloadParser
- #include <AdvancedPayloadParser.h>
This parser allows the processing of firmware data that can be encrypted or have signature
Public Functions
-
virtual bool switchRom(const UpdateState &updateState) override
This method is responsible for switching the rom. This method is NOT restarting the system. It will happen lated in the parse method.
- Return values:
true – if the switch was successful
-
virtual ReadWriteStream *getStorageStream(size_t storageSize) override
Creates new stream to store the firmware update.
- Parameters:
size_t – storageSize the requested storage size
- Return values:
ReadWriteStream* –
-
inline PayloadParser(size_t currentPatchVersion, size_t allowedVersionBytes = 24)
- Parameters:
currentPatchVersion –
allowedVersionBytes – - maximum allowed bytes to be used for describing a patch version
-
virtual bool switchRom(const UpdateState &updateState) override
-
class PayloadParser
- #include <PayloadParser.h>
Subclassed by OtaUpgrade::Mqtt::AdvancedPayloadParser, OtaUpgrade::Mqtt::StandardPayloadParser
Public Functions
-
inline PayloadParser(size_t currentPatchVersion, size_t allowedVersionBytes = 24)
- Parameters:
currentPatchVersion –
allowedVersionBytes – - maximum allowed bytes to be used for describing a patch version
-
virtual bool switchRom(const UpdateState &updateState) = 0
This method is responsible for switching the rom. This method is NOT restarting the system. It will happen lated in the parse method.
- Return values:
true – if the switch was successful
-
virtual ReadWriteStream *getStorageStream(size_t storageSize) = 0
Creates new stream to store the firmware update.
- Parameters:
size_t – storageSize the requested storage size
- Return values:
ReadWriteStream* –
-
int parse(MqttPayloadParserState &state, mqtt_message_t *message, const char *buffer, int length)
This method takes care to read the incoming MQTT message and pass it to the stream that is responsoble for storing the data.
- Return values:
int – 0 when everything is ok <0 when an error has occurred
-
struct UpdateState
- #include <PayloadParser.h>
-
inline PayloadParser(size_t currentPatchVersion, size_t allowedVersionBytes = 24)
-
class StandardPayloadParser : public OtaUpgrade::Mqtt::PayloadParser
- #include <StandardPayloadParser.h>
This parser allows the processing of firmware data that is directly stored to the flash memory using RbootOutputStream.
Public Functions
-
virtual bool switchRom(const UpdateState &updateState) override
This method is responsible for switching the rom. This method is NOT restarting the system. It will happen lated in the parse method.
- Return values:
true – if the switch was successful
-
virtual ReadWriteStream *getStorageStream(size_t storageSize) override
Creates new stream to store the firmware update.
- Parameters:
size_t – storageSize the requested storage size
- Return values:
ReadWriteStream* –
-
virtual bool switchRom(const UpdateState &updateState) override
-
using RbootPayloadParser = StandardPayloadParser
References
Used by
OTA over MQTT Sample
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rc-switch
Use your Arduino or Raspberry Pi to operate remote radio controlled devices
Download
Wiki
Info
Send RC codes
Use your Arduino or Raspberry Pi to operate remote radio controlled devices. This will most likely work with all popular low cost power outlet sockets. If yours doesn’t work, you might need to adjust the pulse length.
All you need is a Arduino or Raspberry Pi, a 315/433MHz AM transmitter and one or more devices with one of the supported chipsets:
SC5262 / SC5272
HX2262 / HX2272
PT2262 / PT2272
EV1527 / RT1527 / FP1527 / HS1527
Intertechno outlets
Receive and decode RC codes
Find out what codes your remote is sending. Use your remote to control your Arduino.
All you need is an Arduino, a 315/433MHz AM receiver (although there is no instruction yet, yes it is possible to hack an existing device) and a remote hand set.
For the Raspberry Pi, clone the https://github.com/ninjablocks/433Utils project to compile a sniffer tool and transmission commands.
References
Used by
RCSwitch Library Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Arduino driver for nRF24L01 2.4GHz Wireless Transceiver
Design Goals: This library is designed to be…
Maximally compliant with the intended operation of the chip
Easy for beginners to use
Consumed with a public interface that’s similar to other Arduino standard libraries
Built against the standard SPI library.
Please refer to:
This chip uses the SPI bus, plus two chip control pins. Remember that pin 10 must still remain an output, or the SPI hardware will go into ‘slave’ mode.
References
Used by
nRF24L01 Radio Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
RapidXML
This is a port of https://github.com/dwd/rapidxml with minimal patch, and additional support code for Sming.
API Documentation
-
namespace XML
Unnamed Group
-
bool deserialize(Document &doc, char *content)
De-serialise XML text into a document.
Note
Document maintains references to content so MUST stay in scope for lifetime of doc, or until doc.clear() is called
- Return values:
bool – false on parse error
-
bool deserialize(Document &doc, const FlashString &content)
Unnamed Group
Unnamed Group
-
Node *appendNode(Node *parent, const char *name, const char *value = nullptr, size_t name_size = 0, size_t value_size = 0)
Append a child element with optional value.
Note
If provide, sizes do not include nul-terminator Node MUST already be added to the document tree
- Parameters:
parent – Node to append child to
name – Name of child node (copy taken)
value – Optional node value (copy taken)
- Return values:
Node* – Child node
-
Node *appendNode(Node *parent, const String &name, const FlashString &value)
Unnamed Group
Gets child node by relative path
Leading/trailing separators are not permitted.
- param doc:
- param path:
Node path using slash as separator
- param ns:
Optional namespace
- retval Node*:
nullptr if none is found
Gets node from a document by path
Leading separator is important: if present, search starts at root node. If omitted, first element must match the root node name.
- param doc:
- param path:
Node path using slash as separator
- param ns:
Optional namespace
- retval Node*:
nullptr if none is found
Trailing separator is not allowed.
Example 1: /s:Body/u:GetContentResponse Will search for s:Body node that is a child of the root node. The node u:GetContentResponse should be child of the s:Body node
Example 2: s:Envelope/s:Body/u:GetContentResponse Will search for s:Body node that is a child of the root node named s:Envelope. The node u:GetContentResponse should be child of the s:Body node
Get node value
- param node:
- param name:
- param name_size:
- retval String:
invalid if node or name not found
Get attribute value
- param node:
- param name:
- param name_size:
- retval String:
invalid if node or name not found
Typedefs
-
using Document = rapidxml::xml_document<char>
-
using Node = rapidxml::xml_node<char>
-
using NodeType = rapidxml::node_type
-
using Attribute = rapidxml::xml_attribute<char>
Functions
-
Node *insertDeclaration(Document &doc)
Add a declaration to the document if there isn’t one already.
Note
By default, declarations are included when parsed during de-serialisation. If you need one in the output serialisation, call this function. (We’re talking about the “<?xml version=”1.0” ?>” at the start)
-
bool deserialize(Document &doc, char *content)
References
Used by
Basic Graphics Sample
UPnP Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Embedded RingBufCPP
This is a simple ring (FIFO) buffer queuing library for embedded platforms, such as Arduino’s. This library is based on a previous one I wrote, only now in C++ instead of C. It uses C++’s templating, so no more weird pointer casting, and deep copies of objects are supported. It has concurrency protection built in, so feel free to do operations on the buffer inside of ISR’s. All memory is statically allocated at compile time, so no heap memory is used. It can buffer any fixed size object (ints, floats, structs, objects, etc…).
FAQ’s
- I only have a C compiler for my platform
- No worries, try the vanilla C version of the library.
Use Cases
A ring buffer is used when passing asynchronous io between two threads. In the case of the Arduino, it is very useful for buffering data in an interrupt routine that is later processed in your void loop().
Supported Platforms
The library currently supports:
AVR
ESP8266
Any other platform (just implement the
RB_ATOMIC_STARTandRB_ATOMIC_ENDmacros)
Install
This library is now available in the Arduino Library Manager, directly in the IDE. Go to Sketch > Include Library > Manage Libraries and search for RingBufCPP. Then #include <RingBufCPP.h> in your sketch.
To manually install this library, download this file as a zip, and extract the resulting folder into your Arduino Libraries folder. Installing an Arduino Library.
Examples
Look at the examples folder for several examples.
Contributing
If you find this Arduino library helpful, click the Star button, and you will make my day.
Feel free to improve this library. Fork it, make your changes, then submit a pull request!
API
Constructor
RingBufCPP<typename Type, size_t MaxElements>();
Creates a new RingBuf object that can buffer up to MaxElements of type Type.
Methods
add()
bool add(const Type &obj, bool overwrite=false)
Append an element to the buffer. If there is already MaxElements in the buffer, the oldest element will either be overwritten (when overwrite is true) or this add will have no effect (when overwrite is false). Return false on a full buffer.
peek()
Type *peek(uint16_t num);
Peek at the num’th element in the buffer. Returns a pointer to the location of the num’th element. If num is out of bounds or the num’th element is empty, a NULL pointer is returned. Note that this gives you direct memory access to the location of the num’th element in the buffer, allowing you to directly edit elements in the buffer. Note that while all of RingBuf’s public methods are atomic (including this one), directly using the pointer returned from this method is not safe. If there is a possibility an interrupt could fire and remove/modify the item pointed to by the returned pointer, disable interrupts first with noInterrupts(), do whatever you need to do with the pointer, then you can re-enable interrupts by calling interrupts().
pull()
bool pull(Type *dest);
Pull the first element out of the buffer. The first element is copied into the location pointed to by dest. Returns false if the buffer is empty, otherwise returns true on success.
numElements()
size_t numElements();
Returns number of elements in the buffer.
isFull()
bool isFull();
Returns true if buffer is full, otherwise false.
isEmpty()
bool isEmpty();
Returns true if buffer is empty, false otherwise.
License
This library is open-source, and licensed under the MIT license. Do whatever you like with it, but contributions are appreciated.
References
Used by
IRremoteESP8266 Library Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
RingTone
This library provides support for parsing and playing tunes in RTTTL format.
RTTTL conversion code based on https://github.com/end2endzone/NonBlockingRTTTL.
The parser is stream-based and allows random seeking where the stream supports it.
An RTTTL writer is also included to assist with editing/creation of RTTTL files.
API Documentation
-
namespace RingTone
Enums
-
enum class Note
Note numbers, defined here for convenience.
Values:
-
enumerator MUTE
-
enumerator C
-
enumerator C_Sharp
-
enumerator D_Flat
-
enumerator D
-
enumerator D_Sharp
-
enumerator E_Flat
-
enumerator E
-
enumerator F
-
enumerator F_Sharp
-
enumerator G
-
enumerator G_Sharp
-
enumerator A_Flat
-
enumerator A
-
enumerator A_Sharp
-
enumerator B_Flat
-
enumerator B
-
enumerator MUTE
Functions
-
template<unsigned octave, unsigned note>
static constexpr unsigned calculateFrequency()
-
unsigned charToNoteValue(char c)
Get the corresponding note number for a letter.
Note
To sharpen a note, add 1
- Parameters:
c –
- Return values:
unsigned – Notes start at 1, 0 indicates error or pause/mute
-
unsigned getNoteFrequency(unsigned octave, unsigned note)
Convert a scale/note combination into frequency.
- Parameters:
octave –
note –
- Return values:
unsigned – Frequency, 0 if out of range
-
unsigned getClosestNote(unsigned frequency, unsigned &octave)
Convert a frequency into a scale/note combination into frequency.
- Parameters:
frequency –
octave – Octave for the note
- Return values:
unsigned – The note number, 0 if out of range
-
const char *getNoteName(unsigned noteValue)
Get text for a given note number.
Variables
-
static unsigned noteFrequencyA4 = 440
Reference note frequency.
-
static constexpr float frequencyRoot = pow(2, 1.0 / 12)
-
struct NoteDef
- #include <RingTone.h>
-
class Parser
- #include <RingTone.h>
Base parser class.
Subclassed by RingTone::RtttlParser
-
class Player
- #include <RingTonePlayer.h>
Base class to support playback of tunes.
Note
This class doesn’t produce any sound. To do this provide set the
onPrepareNoteandonPlayNotecallbacks, or override theprepareNoteandplayNotemethods in an inherited class.Public Functions
-
void begin(RingTone::Parser *parser)
Initialise player.
Note
We don’t own the parser, just take a reference
- Parameters:
parser – The source of ringtone data
-
void end()
Stop the player and un-reference parser.
-
bool start(unsigned delayMs = 0)
Start or continue playing the tune.
-
void stop()
Stop/pause playing the tune.
-
inline void resetPlayTime()
Stop playback and reset play time.
-
inline bool isStarted()
Determine if a tune is being played.
-
inline unsigned getSpeed() const
Get playback speed factor.
- Return values:
unsigned – 100 = normal speed, 50 = half speed, 200 = double speed, etc.
-
inline unsigned setSpeed(unsigned speed)
Set playback speed factor.
- Return values:
unsigned – New speed factor
-
inline unsigned adjustSpeed(int adjust)
Make a relative adjustment to playback speed.
- Return values:
unsigned – New speed factor
-
void begin(RingTone::Parser *parser)
-
struct RtttlHeader
- #include <RtttlParser.h>
-
struct RtttlParserState
- #include <RtttlParser.h>
Subclassed by RingTone::RtttlParser
-
class RtttlParser : public RingTone::Parser, private RingTone::RtttlParserState
- #include <RtttlParser.h>
Class to parse RTTTL files RTTTL (RingTone Text Transfer Language) format.
Public Functions
-
bool begin(IDataSourceStream *source)
Initialise the parser with the given stream.
-
void end()
Release the source stream.
-
bool nextTune()
Locate next tune and read header.
-
bool seekTune(unsigned index)
Find a tune by index, starting at #0.
-
inline unsigned getIndex()
Get the current tune index.
-
unsigned getCount()
Get the number of tunes in this file.
-
inline bool rewind()
Rewind to start of tune.
-
bool begin(IDataSourceStream *source)
-
class RtttlWriter
- #include <RtttlWriter.h>
-
enum class Note
-
class RtttlJsonListStream : public IDataSourceStream
A forward-only stream for listing contents of a tune file.
Note
Tune files can be large so we only output one tune title at a time
Public Functions
-
inline RtttlJsonListStream(const String &name, RingTone::RtttlParser *parser)
Construct a list stream.
- Parameters:
name – Identifies this stream, will have .json appended
parser – Pre-initialised parser to obtain tunes from
-
inline virtual bool isValid() const override
Determine if the stream object contains valid data.
Note
Where inherited classes are initialised by constructor this method indicates whether that was successful or not (e.g. FileStream)
- Return values:
bool – true if valid, false if invalid
-
virtual uint16_t readMemoryBlock(char *data, int bufSize) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
-
virtual bool seek(int len) override
Move read cursor.
- Parameters:
len – Relative cursor adjustment
- Return values:
bool – True on success.
-
inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
-
inline RtttlJsonListStream(const String &name, RingTone::RtttlParser *parser)
References
Used by
RingTone Player Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SD Card
Low-level support code for accessing SD Cards using FATFS.
References
Used by
SD Card Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SI7020/SI7021 Environmental Sensors
Arduino library for SI7020 and SI7021 environmental sensors
Examples:
References
Used by
SI7021 Humidity Sensor Sample
MeteoControl MQTT Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SPI Library
Provides support for both hardware and software-base SPI master devices.
Software SPI
Software SPI is likely to be of use only for ESP8266, but is enabled for all architectures.
The ESP8266 can manage a minimum of about 70ns between bit edges, with a maximum of about 1.8 MBit/s at normal CPU clock frequency or 2.5 MHz if CPU clock set to fast.
Build variables
- SPISOFT_DELAY_VARIABLE
default: 0 (disabled)
This setting must be enabled in order to use
SPISoft::setDelay(). The base clock speed (i.e. delay=0) is reduced to about 1.4 MBit/s (2.1 MBit/s for fast CPU).The appropriate delay factor can be provided in the
SPISoftconstructor, or by callingSPISoft::setDelay().With the ESP8266, then the appropriate delay factor is calculated automatically from the speed passed in
SPISettings. This will override any previously set delay factor. Use a speed of 0 to use the manually configured delay value.
- SPISOFT_DELAY_FIXED
default: 0 (disabled) maximum: 10
Adds the requested number of ‘NOP’ CPU instructions to every clock transition.
Has less impact than enabling variable delays.
Use variable delays if 10 is insufficient.
Will be ignored if variable delays are enabled.
Performance
The following information has been obtained by measurement. The test application was built using:
make SPISOFT_CALIBRATE=1 SPISOFT_DELAY_VARIABLE=1 CPU_FAST=0
then flashed to an ESP-12F. An oscilliscope was hooked up to the SCK output and the maximum frequency at each delay setting noted. (Not too awkward, just note down the frequency each time the figure changes.)
Entering these into the spisoft.ods spreadsheet allows the values to be charted and co-efficients calculated.
Curiously values from 8 upward form a straight line so we can calculate those. Smaller values are non-linear so a small lookup table is used.
Repeat for CPU_FAST=1.
Note that these figures are concerned with maximum (i.e. burst) frequencies at each delay setting, since that is what devices are sensitive to.
80 MHz CPU
. |
Frequency (kHz) |
|
|---|---|---|
Delay |
Variable |
Fixed |
0 |
1357 |
2288 |
1 |
1315 |
1869 |
2 |
1215 |
1824 |
3 |
1129 |
1780 |
4 |
1055 |
1744 |
5 |
990 |
1709 |
6 |
932 |
1639 |
7 |
881 |
1575 |
8 |
819 |
1511 |
9 |
743 |
1462 |
10 |
680 |
1408 |
11 |
626 |
1338 |
12 |
578 |
1294 |
20 |
367 |
|
30 |
251 |
|
40 |
192 |
|
50 |
155 |
|
60 |
129 |
|
70 |
112 |
|
80 |
98 |
|
90 |
87 |
|
100 |
79 |
|
160 MHz CPU
. |
Frequency (kHz) |
|
|---|---|---|
Delay |
Variable |
Fixed |
0 |
2133 |
2857 |
1 |
2126 |
2510 |
2 |
2106 |
2500 |
3 |
1958 |
2439 |
4 |
1861 |
2429 |
5 |
1745 |
2372 |
6 |
1668 |
2353 |
7 |
1573 |
2294 |
8 |
1482 |
2283 |
9 |
1357 |
2242 |
10 |
1250 |
2222 |
11 |
1160 |
2174 |
12 |
1081 |
2165 |
20 |
702 |
|
30 |
488 |
|
40 |
374 |
|
50 |
303 |
|
60 |
255 |
|
70 |
220 |
|
80 |
193 |
|
90 |
172 |
|
100 |
156 |
|
API Documentation
-
class SPISoft : public SPIBase
- #include <SPISoft.h>
Software-based SPI master.
Intended for ESP8266 due to limited I/O but will work on any architecture.
Constructors
-
SPISoft()
Default constructor uses same pins as hardware SPI.
-
inline SPISoft(uint8_t miso, uint8_t mosi, uint8_t sck, uint8_t delay = 0)
Specify pins to use plus optional delay.
Delay is ignored if code is not compiled with SPISOFT_DELAY < 0.
-
SPISoft(uint8_t delay)
Use default pins but provide a delay.
Public Functions
-
virtual bool begin() override
Initialize the SPI bus by setting SCK and MOSI to outputs, pulling SCK and MOSI low.
-
inline virtual void end() override
Disable the SPI bus (leaving pin modes unchanged).
-
virtual void endTransaction() override
Stop using the SPI bus. Normally this is called after de-asserting the chip select, to allow other libraries to use the SPI bus.
-
virtual uint32_t transfer32(uint32_t val, uint8_t bits = 32) override
Send/receive a word of variable size.
Word is transferred either MSB first (bits-1) or LSB first (bit 0) depending on the currently applied bitOrder setting.
- Parameters:
val – Word to send
bits – Size of word
-
virtual void transfer(uint8_t *buffer, size_t size) override
Send/receive a variable-length block of data.
- Parameters:
buffer – IN: The data to send; OUT: The received data
size – Number of bytes to transfer
-
inline void setDelay(uint8_t delay)
Set delay factor for the SCK signal. Impacts SPI speed.
Requires code to be compiled with SPISOFT_DELAY < 0.
ESP8266 only: The delay will be automatically calculated for a requested clock speed when
begin()orbeginTransaction()are called. To use only the manually programmed delay, set the clock speed to zero.
-
inline virtual bool loopback(bool enable) override
For testing, tie MISO <-> MOSI internally.
Note: Not included in std Arduino lib
-
inline uint8_t transfer(uint8_t val)
Send/receive one byte of data.
- Parameters:
val – The byte to send
- Return values:
uint8_t – The received byte
-
void transfer(uint8_t *buffer, size_t size) = 0
Send/receive a variable-length block of data.
- Parameters:
buffer – IN: The data to send; OUT: The received data
size – Number of bytes to transfer
-
SPISoft()
Build variables
- ENABLE_SPI_DEBUG
default: 0 (disabled)
Enable to print additional debug messages.
API Documentation
-
enum SpiMode
Values:
-
enumerator SPI_MODE0
-
enumerator SPI_MODE1
-
enumerator SPI_MODE2
-
enumerator SPI_MODE3
-
enumerator SPI_MODE0
-
const uint32_t SPI_SPEED_DEFAULT = 4000000UL
-
static constexpr uint8_t SPI_PIN_DEFAULT = {0xff}
SPI driver uses default pin assignment.
-
struct SpiPins
- #include <SPIBase.h>
SPI pin connections.
-
class SPIBase
- #include <SPIBase.h>
Subclassed by SPIClass, SPISoft
Send/receive some data
SPI transfer is based on a simultaneous send and receive: the received data is returned in receivedVal (or receivedVal16). In case of buffer transfers the received data is stored in the buffer in-place (the old data is replaced with the data received).
receivedVal = SPI.transfer(val) receivedVal16 = SPI.transfer16(val16) SPI.transfer(buffer, size)
-
inline uint8_t transfer(uint8_t val)
Send/receive one byte of data.
- Parameters:
val – The byte to send
- Return values:
uint8_t – The received byte
-
inline uint16_t transfer16(uint16_t val)
Send/receive one 16-bit word of data.
Word is transferred either MSB first (bit 15) or LSB first (bit 0) depending on the currently applied bitOrder setting.
- Parameters:
val – The word to send
- Return values:
uint16_t – The received word
-
inline virtual uint32_t transfer32(uint32_t val, uint8_t bits = 32)
Send/receive a word of variable size.
Word is transferred either MSB first (bits-1) or LSB first (bit 0) depending on the currently applied bitOrder setting.
- Parameters:
val – Word to send
bits – Size of word
-
virtual void transfer(uint8_t *buffer, size_t size) = 0
Send/receive a variable-length block of data.
- Parameters:
buffer – IN: The data to send; OUT: The received data
size – Number of bytes to transfer
Public Functions
-
virtual bool begin() = 0
Initialize the SPI bus by setting SCK and MOSI to outputs, pulling SCK and MOSI low.
-
virtual void end() = 0
Disable the SPI bus (leaving pin modes unchanged).
-
inline void beginTransaction(SPISettings &settings)
Initialize the SPI bus using the defined SPISettings.
-
inline virtual void endTransaction()
Stop using the SPI bus. Normally this is called after de-asserting the chip select, to allow other libraries to use the SPI bus.
-
inline virtual uint8_t read8()
Read one byte from SPI without setting up registers.
Used for performance tuning when doing continuous reads this method does not reset the registers, so make sure that a regular transfer(data) call was performed
Note: this method is not found on the Arduino API
USE WITH CARE !!
- Parameters:
none –
- Return values:
byte – received
-
virtual bool loopback(bool enable) = 0
For testing, tie MISO <-> MOSI internally.
Note: Not included in std Arduino lib
Public Members
-
SPISettings SPIDefaultSettings
Default settings used by the SPI bus until reset by beginTransaction(SPISettings)
Note: Not included in std Arduino lib
-
inline uint8_t transfer(uint8_t val)
-
enum class SpiBus
Identifies bus selection.
Values:
-
enumerator INVALID
-
enumerator MIN
-
enumerator SPI1
-
enumerator FSPI
-
enumerator SPI2
-
enumerator HSPI
-
enumerator MAX
-
enumerator DEFAULT
-
enumerator INVALID
-
enumerator MIN
-
enumerator SPI1
-
enumerator MAX
-
enumerator DEFAULT
-
enumerator INVALID
-
enumerator MIN
-
enumerator SPI1
-
enumerator SPI2
-
enumerator SPI3
-
enumerator MAX
-
enumerator DEFAULT
-
enumerator INVALID
-
enumerator MIN
-
enumerator SPI1
-
enumerator SPI2
-
enumerator MAX
-
enumerator DEFAULT
-
enumerator INVALID
-
using Callback = bool (*)(Request &request)
SPI completion callback routine.
- Param request:
- Retval bool:
Return true if request is finished, false to re-queue it immediately
-
struct Data
- #include <Data.h>
Specifies a block incoming or outgoing data.
Data can be specified directly within
Data, or as a buffer reference.Command or address are stored in native byte order and rearranged according to the requested byteOrder setting. Data is always sent and received LSB first (as stored in memory) so any re-ordering must be done by the device or application.
Set internal data value of 1-4 bytes
Note
Data is sent LSB, MSB (native byte order)
-
inline void set8(uint8_t data)
Set to single 8-bit value.
- Parameters:
data –
-
inline void set16(uint16_t data)
Set to single 16-bit value.
- Parameters:
data –
-
inline void set32(uint32_t data, uint8_t len = 4)
Set to 32-bit data.
- Parameters:
data –
len – Length in bytes (1 - 4)
Public Functions
-
inline void clear()
Reset to zero-length.
-
inline void set(const void *data, uint16_t count)
Set to reference external data block.
- Parameters:
data – Location of data
count – Number of bytes
Public Members
-
void *ptr
Pointer to data.
-
uint16_t length
Number of bytes of data.
-
uint16_t isPointer
If set, data is referenced indirectly, otherwise it’s stored directly.
-
inline void set8(uint8_t data)
-
class Device
- #include <Device.h>
Manages a specific SPI device instance attached to a controller.
Subclassed by Graphics::SpiDisplay, Graphics::XPT2046, HSPI::MemoryDevice
Public Functions
-
inline bool begin(PinSet pinSet, uint8_t chipSelect, uint32_t clockSpeed)
Register device with controller and prepare for action.
- Parameters:
pinSet – Use PinSet::normal for Esp32, other values for Esp8266
chipSelect – Identifies the CS number for ESP8266, or the GPIO pin for ESP32
clockSpeed – Bus speed
-
inline bool isReady() const
Determine if the device is initialised.
- Return values:
bool –
-
virtual IoModes getSupportedIoModes() const = 0
Return set of IO modes supported by a device implementation.
-
inline bool isSupported(IoMode mode) const
Determine if the device/controller combination supports an IO mode Must be called after
begin()as other settings (e.g. pinset) can affect support.
-
inline void onTransfer(Callback callback)
Set a callback to be invoked before a request is started, and when it has finished.
- Parameters:
callback – Invoked in interrupt context, MUST be in IRAM
-
inline bool begin(PinSet pinSet, uint8_t chipSelect, uint32_t clockSpeed)
-
class MemoryDevice : public HSPI::Device
- #include <MemoryDevice.h>
Base class for read/write addressable devices.
Subclassed by HSPI::RAM::IS62_65, HSPI::RAM::PSRAM64
Prepare a write request
-
virtual void prepareWrite(HSPI::Request &req, uint32_t address) = 0
Prepare request without data.
Prepare a read request
-
virtual void prepareRead(HSPI::Request &req, uint32_t address) = 0
Prepare without buffer.
-
inline void prepareRead(HSPI::Request &req, uint32_t address, void *buffer, size_t len)
Prepare with buffer.
- Parameters:
buffer – Where to write incoming data
len – Size of buffer
Public Functions
-
inline void write(uint32_t address, const void *data, size_t len)
Write a block of data.
Note
Limited by current operating mode
- Parameters:
address –
data –
len –
-
inline void read(uint32_t address, void *buffer, size_t len)
Read a block of data.
Note
Limited by current operating mode
- Parameters:
address –
data –
len –
-
virtual void prepareWrite(HSPI::Request &req, uint32_t address) = 0
-
class IS62_65 : public HSPI::MemoryDevice
- #include <IS62-65.h>
IS62/65WVS2568GALL fast serial RAM.
Public Types
Public Functions
-
inline virtual IoModes getSupportedIoModes() const override
Return set of IO modes supported by a device implementation.
-
inline bool begin(PinSet pinSet, uint8_t chipSelect, uint32_t clockSpeed)
Configure the RAM into a known operating mode.
-
inline OpMode getOpMode() const
Get current operating mode (cached value)
- Return values:
OpMode – No device access is performed.
-
inline virtual IoModes getSupportedIoModes() const override
-
class PSRAM64 : public HSPI::MemoryDevice
- #include <PSRAM64.h>
PSRAM64(H) pseudo-SRAM.
Public Functions
-
inline virtual IoModes getSupportedIoModes() const override
Return set of IO modes supported by a device implementation.
-
inline bool begin(PinSet pinSet, uint8_t chipSelect, uint32_t clockSpeed)
Configure the RAM into a known operating mode.
-
inline virtual IoModes getSupportedIoModes() const override
-
struct Request
- #include <Request.h>
Defines an SPI Request Packet.
Request fields may be accessed directly or by use of helper methods.
Application is responsible for managing Request object construction/destruction. Queuing is managed as a linked list so the objects aren’t copied.
Applications will typically only require a couple of Request objects, so one can be prepared whilst the other is in flight. This helps to minimises the setup latency between SPI transactions.
Set value for command phase
-
inline void setCommand(uint16_t command, uint8_t bitCount)
- Parameters:
command –
bitCount – Length of command in bits
-
inline void setCommand8(uint8_t command)
Set 8-bit command.
- Parameters:
command –
-
inline void setCommand16(uint16_t command)
Set 16-bit command.
- Parameters:
command –
Set value for address phase
-
inline void setAddress(uint32_t address, uint8_t bitCount)
- Parameters:
address –
bitCount – Length of address in bits
-
inline void setAddress24(uint32_t address)
Set 24-bit address.
- Parameters:
address –
Public Functions
-
inline void setAsync(Callback callback = nullptr, void *param = nullptr)
Set request to asynchronous execution with optional callback.
Public Members
-
Device *device = {nullptr}
Target device for this request.
-
Request *next = {nullptr}
Controller uses this to queue requests.
-
uint16_t cmd = {0}
Command value.
-
uint8_t cmdLen = {0}
Command bits, 0 - 16.
-
uint8_t async
Set for asynchronous operation.
-
uint8_t task
Controller will execute this request in task mode.
-
uint8_t busy
Request in progress.
-
uint32_t addr = {0}
Address value.
-
uint8_t addrLen = {0}
Address bits, 0 - 32.
-
uint8_t dummyLen = {0}
Dummy read bits between address and read data, 0 - 255.
-
uint8_t sizeAlign = {0}
Required size alignment of each transaction (if split up)
-
Data out
Outgoing data.
-
Data in
Incoming data.
-
Callback callback = {nullptr}
Completion routine.
-
void *param = {nullptr}
User parameter.
-
inline void setCommand(uint16_t command, uint8_t bitCount)
-
class StreamAdapter
- #include <StreamAdapter.h>
Helper class for streaming data to/from SPI devices.
-
class SPIClass : public SPIBase
- #include <SPI.h>
Hardware SPI class.
Public Types
-
using IoCallback = void (*)(uint16_t c, uint8_t bits, bool read)
Used for testing purposes only.
- Param c:
Value being read/written
- Param bits:
Size of value in bits
- Param read:
true for incoming value, false for outgoing
Public Functions
-
bool setup(SpiBus id, SpiPins pins = {})
Alternative to defining bus and pin set in constructor. Use this method to change global
SPIinstance setup.IMPORTANT: Must be called before begin().
-
virtual bool begin() override
Initialize the SPI bus by setting SCK and MOSI to outputs, pulling SCK and MOSI low.
-
virtual void end() override
Disable the SPI bus (leaving pin modes unchanged).
-
virtual uint8_t read8() override
Read one byte from SPI without setting up registers.
Used for performance tuning when doing continuous reads this method does not reset the registers, so make sure that a regular transfer(data) call was performed
Note: this method is not found on the Arduino API
USE WITH CARE !!
- Parameters:
none –
- Return values:
byte – received
-
virtual uint32_t transfer32(uint32_t val, uint8_t bits = 32) override
Send/receive a word of variable size.
Word is transferred either MSB first (bits-1) or LSB first (bit 0) depending on the currently applied bitOrder setting.
- Parameters:
val – Word to send
bits – Size of word
-
virtual void transfer(uint8_t *buffer, size_t numberBytes) override
Send/receive a variable-length block of data.
- Parameters:
buffer – IN: The data to send; OUT: The received data
size – Number of bytes to transfer
-
virtual bool loopback(bool enable) override
For testing, tie MISO <-> MOSI internally.
Note: Not included in std Arduino lib
-
void setDebugIoCallback(IoCallback callback)
Used for testing purposes only Must be called after begin().
Used to verify serialisation/de-searialisation bit ordering
-
inline uint8_t transfer(uint8_t val)
Send/receive one byte of data.
- Parameters:
val – The byte to send
- Return values:
uint8_t – The received byte
-
void transfer(uint8_t *buffer, size_t size) = 0
Send/receive a variable-length block of data.
- Parameters:
buffer – IN: The data to send; OUT: The received data
size – Number of bytes to transfer
-
using IoCallback = void (*)(uint16_t c, uint8_t bits, bool read)
References
Used by
Sming (main) Component
SD Storage Library
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SSDP
https://en.wikipedia.org/wiki/Simple_Service_Discovery_Protocol
Provides an SSDP server implementation aiming to be fully standards compliant.
Configuration Variables
- SSDP_DEBUG_VERBOSE
0 (default)
1: Output details of all requests. Warning: This can produce a lot of output!
- UPNP_VERSION
UPnP standard to follow.
1.0: (default)
1.1: Incomplete
2.0: Incomplete
Versions 1.1 and 2.0 require
BOOTID.UPNP.ORGandCONFIGID.UPNP.ORGfields which are not currently implemented.
Key points from UPnP 2.0 specification
I’m choosing to avoid the issue of ‘multi-homed’ devices, where both IPv4 and IPv6 are being used, and probably also WiFi+Ethernet combinations. Basically whenever multiple IP stacks are involved.
Advertisement
Multicast using ssdp:alive messages.
3 messages for each root device:
upnp:rootdevice
uuid:device-UUID
urn:<domain>:<deviceType:ver>
2 for each embedded device:
uuid:device-UUID
urn:<domain>:<deviceType:ver>
1 for each service type in each device:
urn<domain>:<serviceType:ver>
Initial advertisements ‘should be sent as quickly as possible’. Subsequent advertisements ‘are allowed to be spread over time’.
For orderly shutdown, multicast ssdp:byebye messages as for ssdp:alive.
Note that ssdp:update messages are only mentioned in the spec. wrt. mult-homed devices.
Devices should wait a random interval (e.g. 0 - 100 ms) before sending an initial
set of advertisements in order to reduce the likelihood of network storms;
this random interval should also be applied on occasions where the device obtains
a new IP address or a new UPnP-enabled interface is installed.
Due to the unreliable nature of UDP, devices should send the entire set of discovery
messages more than once with some delay between sets e.g. a few hundred milliseconds.
To avoid network congestion discovery messages should not be sent more than three times.
The device shall re-send its advertisements periodically prior to expiration of the
duration specified in the CACHE-CONTROL header field; it is Recommended that such
refreshing of advertisements be done at a randomly-distributed interval of less than
one-half of the advertisement expiration time, so as to provide the opportunity
for recovery from lost advertisements before the advertisement expires, and to
distribute over time the advertisement refreshment of multiple devices on the network
in order to avoid spikes in network traffic.
Note that UDP packets are also bounded in length (perhaps as small as 512
Bytes in some implementations); each discovery message shall fit entirely in a single UDP
packet. There is no guarantee that the above 3+2d+k messages will arrive in a particular
order.
Search request
M-SEARCH: “Please tell me about yourselves, but don’t all shout at once.”
ssdp:all
Search for all devices and services.
upnp:rootdevice
Search for root devices only.
uuid:device-UUID
Search for a particular device.
urn:<domain>:device:deviceType:ver
Search for any device of this type.
urn:<domain>:service:serviceType:ver
Search for any service of this type.
Period characters in <domain> are always substituted with hyphens (RFC 2141).
Not clear on how to handle version numbers at present. The specs. say only minor versions
are backward compatible, which why perhaps we only see major numbers in interface
definitions. e.g. Basic:1 not Basic:1.0.
Search response
Any device responding to a unicast M-SEARCH should respond within 1 second.
In response to an M-SEARCH request, if ST header in request was:
ssdp:all
Respond 3+2d+k times for a root device with d embedded devices and s embedded services
but only k distinct service types.
Value for ST header must be the same as for the NT header in NOTIFY messages with ssdp:alive.
upnp:rootdevice
Respond once for root device.
uuid:device-UUID
Respond once for each matching device, root or embedded.
urn:<domain>:device:deviceType:v
Respond once for each matching device, root or embedded.
Should specify the version of the device type contained in the M-SEARCH request.
urn:<domain>:service:serviceType:v
Respond once for each matching service type.
Should specify the version of the service type contained in the M-SEARCH request.
Descriptions
The LOCATION field is for the device description or enclosing device in the case of a service.
This implies that we never respond with a service description, which makes sense:
The device description provides key information about its services
The service description contains action lists or state variable tables
Only the device description is required to learn about services, whilst the service description is only required if the Control Point needs to interact with that service.
Points arising
So we need a filter which then gets passed through the device stack. Each response must be sent on a schedule, not all together, so we’ll need to set up a timer. We’ll also need to track state something like the DescriptionStream. Actually, what we can do is create an enumerator which iterates through the entire device stack. That will take out the complexity from here and DescriptionStream. We’ll need an additional Item tag so we can differentiate. This can either be a virtual method or we could use a union with all the different Item types plus a separate tag field. That could also contain the search filter information as input.
Move all this stuff into an SsdpResponder class?
API Documentation
-
namespace SSDP
Typedefs
-
using MessageDelegate = Delegate<void(MessageSpec *ms)>
A callback function must be provided to do the actual sending.
- Param ms:
Message spec. to action, must delete when finished with it
-
using ReceiveDelegate = Delegate<void(BasicMessage &message)>
Callback type for handling an incoming message.
-
using SendDelegate = Delegate<void(Message &msg, MessageSpec &ms)>
Callback type for sending outgoing message.
Note
The message spec. is provided by the UPnP Device Host, which then gets called back to construct the message content. It then calls
sendMessage().- Param msg:
Message with standard fields completed
- Param ms:
Parameters for constructing message
Enums
-
enum class SearchTarget
SSDP Search target types.
Values:
-
enumerator root
Root devices only:
upnp:rootdevice
-
enumerator type
or
urn:{domain}:service:{serviceType}:{v}Search for device/service type:
urn:{domain}:device:{deviceType}:{v}
-
enumerator uuid
Search for specific device:
uuid:{device-UUID}
-
enumerator all
All devices and services:
ssdp::all
-
enumerator root
Functions
- static const IpAddress multicastIp (239, 255, 255, 250)
-
DECLARE_FSTR(SSDP_DISCOVER)
-
DECLARE_FSTR(UPNP_ROOTDEVICE)
-
DECLARE_FSTR(SSDP_ALL)
-
NotifySubtype getNotifySubtype(const char *subtype)
-
DECLARE_FSTR(defaultProductNameAndVersion)
-
template<class HeaderClass>
class BaseMessage : public HeaderClass - #include <Message.h>
class template for messages
-
class BasicMessage : public SSDP::BaseMessage<BasicHttpHeaders>
- #include <Message.h>
Handles incoming messages.
Note
Contains name/value pairs as pointers.
-
class Message : public SSDP::BaseMessage<HttpHeaders>
- #include <Message.h>
Message using regular HTTP header management class.
Note
More flexible than BasicMessage but requires additional memory allocations
-
class MessageQueue
- #include <MessageQueue.h>
Queue of objects managed by a single timer.
Public Functions
-
inline void setCallback(MessageDelegate delegate)
Set a callback to handle sending a message @Param delegate.
-
void add(MessageSpec *ms, uint32_t intervalMs)
Schedule a message to start after the given interval has elapsed.
The UPnP spec. requires that messages are sent after random delays, hence the interval. MessagesSpec objects must be created using the
newallocator and are deleted after sending.- Parameters:
ms – The template spec. for constructing the message(s)
intervalMs – How long to wait before sending
-
bool contains(const MessageSpec &ms) const
Determine if a matching message specification is already queued.
See
MessageSpecoperator== definition for how comparison is performed.- Parameters:
ms –
- Return values:
bool – true if the given spec. is already queued.
-
unsigned remove(void *object)
Remove any messages for this object.
- Return values:
unsigned – Number of messages removed
-
inline void setCallback(MessageDelegate delegate)
-
class MessageSpec
- #include <MessageSpec.h>
Defines the information used to create an outgoing message.
The message queue stores these objects as a linked list.
Public Functions
-
inline MessageSpec(const MessageSpec &ms, SearchMatch match, void *object)
Construct a new message spec for a specific match type.
- Parameters:
ms – Template message spec
match – The match type
object – Target for message
-
inline IpAddress remoteIp() const
Get the remote IP address.
-
inline uint16_t remotePort() const
Get the remote port number.
-
template<class Object>
inline Object *object() const Get the target object pointer.
This is templated to provide cleaner code. Example:
MyObject* object = ms.object<MyObject>();
-
inline MessageType type() const
Get the message type.
-
inline NotifySubtype notifySubtype() const
Get the notification sub-type.
-
inline SearchMatch match() const
Get the search match type.
-
inline SearchTarget target() const
Get the search target.
-
inline void setTarget(SearchTarget target)
Set the search target.
-
inline void setRemote(IpAddress address, uint16_t port)
Set the remote address and port.
-
inline void setRepeat(uint8_t count)
Set number of times to repeat message.
-
inline uint8_t repeat() const
Get current repeat value.
-
inline bool shouldRepeat()
Check if message should be repeated and adjust counter.
-
inline MessageSpec(const MessageSpec &ms, SearchMatch match, void *object)
-
class Server : private UdpConnection
- #include <Server.h>
Listens for incoming messages and manages queue of outgoing messages.
- Todo:
Randomise the time as required by MX and keep queue ordered by time. Each message is 12 bytes, adding time would make this 16. Need to handle alives < 1/2 expiry time as well so timer will always be active. Could also use a linked list so an additional pointer would make it 20 bytes.
Note: This is basically another timer queue, so we could use software timers directly but potentially there could be a lot of them. Better I think to use a single
Timerand drive it from that.Note
The spec. talks about random intervals, etc. but to keep things simple we just use a timer to spread all these messages out at regular intervals.
Public Functions
-
bool begin(ReceiveDelegate receiveCallback, SendDelegate sendCallback)
Called from UPnP library to start SSDP server.
Note
May only be called once
- Return values:
bool – true on success
-
inline bool isActive()
Determine if server is running.
- Return values:
bool –
-
bool buildMessage(Message &msg, MessageSpec &ms)
Construct a message from the given template spec.
- Parameters:
msg – Fields of this message will be filled out
ms – Spec to use for constructing message
- Return values:
bool – Returns false if validation failed: message should not be sent
-
using MessageDelegate = Delegate<void(MessageSpec *ms)>
References
Used by
UPnP Library
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SD Storage
Provides Storage Management support for SD cards.
This code is ported from the SD Card library.
Currently uses only SPI communication, which offers only a limited interface and is not supported by SDUC cards.
SD Card connections
MMC pin |
SD pin |
miniSD pin |
microSD pin |
Name |
I/O |
Logic |
Description |
|---|---|---|---|---|---|---|---|
1 |
1 |
1 |
2 |
nCS |
I |
PP |
SPI Card Select [CS] (Negative logic) |
2 |
2 |
2 |
3 |
DI |
I |
PP |
SPI Serial Data In [MOSI] |
3 |
3 |
3 |
VSS |
S |
S |
Ground |
|
4 |
4 |
4 |
4 |
VDD |
S |
S |
Power |
5 |
5 |
5 |
5 |
CLK |
I |
PP |
SPI Serial Clock [SCLK] |
6 |
6 |
6 |
6 |
VSS |
S |
S |
Ground |
7 |
7 |
7 |
7 |
DO |
O |
PP |
SPI Serial Data Out [MISO] |
8 |
8 |
8 |
NC nIRQ |
O |
OD |
Unused (memory cards) Interrupt (SDIO cards) (negative logic) |
|
9 |
9 |
1 |
NC |
Unused |
|||
10 |
NC |
Reserved |
|||||
11 |
NC |
Reserved |
Usage
Code should be built with
ENABLE_STORAGE_SIZE64=1 unless using a very small card (1 or 2GB):Add required libraries to your project’s component.mk file:
COMPONENT_DEPENDS := SdStorage FatIFS
Create and initialise a Card Device instance:
#include <Storage/SD/Card.h> // GPIO used for Chip Select (Esp8266) #define PIN_CARD_CS 5 // Create the SD card Device and register with the Storage library auto card = new Storage::SD::Card("card1", SPI); Storage::registerDevice(card); if(!card->begin(PIN_CARD_CS)) { // Handle error here return; } // Display some information Serial << "CSD" << endl << card->csd << endl; Serial << "CID" << endl << card->cid;
At this point, the card can be accessed directly using Storage::Device methods.
If the card has been formatted then the partitions can be accessed using the standard Storage API.
For example:
Storage::Partition part;
// Find the first FAT partition on the card
part = *card->partitions().find(Storage::Partition::SubType::Data::fat);
// Get the first partition (of any type)
part = *card->partitions().begin();
// Iterate through all partitions
for(part: card->partitions()) {
Serial << part << endl;
}
To create a partition table on the card:
#include <Storage/Disk/GPT.h> Storage::Disk::GPT::PartitionTable table; // Create one partition using 100% of all available space table.add("My FAT partition", 0, 100); auto err = Storage::Disk::formatDisk(*card, table); Serial << "formatDisk: " << err << endl; if(!!err) { // If formatting fails, the disk will left in an indeterminate state! ... return; } // Fetch the first (and only) partition auto part = card->partitions().begin();
Note
To use the entire card as a single partition, it is not necessary to pre-format the disk. Instead, construct a partition spanning the entire disk:
auto part = card->editablePartitions().add("My Partition",
Storage::Partition::SubType::Data::fat, 0, card->getSize());
This approach is sometimes referred to as an SFD (Single Filing-system Device).
The partition can now be formatted with the required filing system (FAT):
int res = IFS::FAT::formatVolume(part);
Serial << "formatVolume : " << IFS::Error::toString(res) << endl;
Once formatted, the partition can be mounted as usual:
auto fatfs = IFS::createFatFilesystem(part);
if(fatfs != nullptr) {
if(fatfs->mount() == FS_OK) {
// OK, access the filing system
Serial << fatfs->getContent("My file.txt") << endl;
} else {
// Mount failed, destroy the objects
delete fs;
delete card; // Also de-registers the card with the Storage API
}
}
API Documentation
-
namespace SD
-
class Card : public Storage::Disk::BlockDevice
- #include <Card.h>
Public Functions
-
bool begin(uint8_t chipSelect, uint32_t freq = 0)
Initialise the card.
- Parameters:
chipSelect –
freq – SPI frequency in Hz, use 0 for maximum supported frequency
-
inline virtual String getName() const override
Obtain unique device name.
-
inline virtual uint32_t getId() const
Obtain device ID.
- Return values:
uint32_t – typically flash chip ID
-
inline virtual Type getType() const
Obtain device type.
-
inline virtual size_t getBlockSize() const override
Obtain smallest allocation unit for erase operations.
-
bool begin(uint8_t chipSelect, uint32_t freq = 0)
-
struct CID
- #include <CID.h>
Public Members
-
uint8_t mid
Manufacturer ID.
-
char oid[2]
OEM / Application ID.
-
char pnm[5]
Product name.
-
uint8_t prv
Product revision.
-
uint32_t psn
Product serial number.
-
uint16_t mdt
Manufacturing date.
-
uint8_t not_used
Always 1.
-
uint8_t crc
7-bit checksum
-
uint8_t mid
-
struct CSD
- #include <CSD.h>
Subclassed by Storage::SD::CSD1, Storage::SD::CSD2, Storage::SD::CSD3
-
class Card : public Storage::Disk::BlockDevice
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Source: SERV-03-MI (Micro Servo)
Servo RC PWM Control
Library to control RC servos with PWM signals.
Change History
This library was revised for Sming version 4.0. These are the main changes:
Interrupts remain enabled during updates
ServoChannel value now specifies actual pulse duration in microseconds, such that
minValue <= value <= maxValue. Previously it was0 <= value <= (maxValue - minValue).Max channels increased to 5.
See Pull Request #1870 for further details.
Brief introduction
There are generally two types of servo actuator (digital and analogue) and this library supports only the analogue variety.
Servo actuators have a logic-level control signal to set the position using an active-high pulse of around 1 - 2 ms. This pulse must be repeated every 20ms or so to ensure the position is maintained
Servos are generally insensitive to the exact period provided it’s no more than about 25ms. It’s the duration of the pulse which is critical.
For most servos 1.5ms is the default/centre position, however the min/max values will vary between models depending on the exact type and range of motion. These values are therefore configurable for each channel.
Physical connection
Servos typically use a 5V logic-level input but are usually fine with the 3.3v output from the ESP8266.
Warning
Like relays a servo is an electro-mechanical device, but it also has integrated control circuitry so doesn’t require flyback diodes, etc. to protect from current spikes.
However, remember to always insert a protection resistor of at least 200 ohms between the GPIO and the servo. This limits the current to <10mA if 5V is present on the line. For 12V servos a 1K resistor will perform the same function.
Technical Explanation
Each servo actuator is connected to a GPIO pin, which is toggled by an ISR driven from the Hardware Timer. The ServoChannel class represents this connection, and defines the current value (in microseconds) plus the range (minimum and maximum values) to which the value is constrained.
The hardware timer interrupt is managed by a single instance of the Servo class. All channels are updated sequentially at the start of each frame period (20ms in duration):
The first channel is set ON, then after the required time it is set OFF and the next channel set ON. The final interrupt turns the active channel OFF. This requires (NumChannels + 1) interrupts per frame, and the process repeats continuously whilst there is at least one active channel.
Channel updates (via calls to ServoChannel::setValue()) are not handled immediately, but deferred using a 10ms software timer (half the frame period). This allows more efficient updating and ensures the positions of all affected servos are changed at the same time.
A double-buffering technique is used for updates to avoid disabling interrupts, which allows use of the non-maskable timer interrupts for best timing accuracy and eliminates glitches in the output.
Updates involve re-calculating the list of active pins and timer intervals, which is stored into a second frame buffer. This is made active by the ISR when the current frame has completed.
If the ISR hasn’t yet processed a previous update, it will be retried after a further 10ms.
References
Used by
Basic Servo Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Signal Generator
This Component provides the SignalGenerator class which can be used to synthesise a wide range of analogue signals.
Ported from https://www.codeproject.com/Articles/30180/Simple-Signal-Generator
Simple Signal Generator
29 Oct 2008 CPOL. Tefik Becirovic
Useful software equivalent for a real, simple signal generator device
Introduction
The Simple Signal Generator is a C# class designed to generate four simple periodic waveforms including sine, square, triangle, and sawtooth. The class is provided for testing software and hardware components during the development of measurement applications. A generated signal varies across time domains, and by default, is normalized by the amplitude A € [-1,1] and period T € [0,1]. It is possible to set an arbitrary amplitude, frequency, DC-offset, and phase shift, and to invert the signal.
Background Display
There are a couple of articles on The Code Project that describe in detail about developing different user defined components for dynamically changing single values or signal shapes such as bars, graphs, charts, gauges, and a diversity of other instruments.
For testing those components, functions like y=F(x) are very often used. These functions are typically periodical, mostly sine, and their variable x will be modified in a loop with constant steps across an interval [x0,xn].
Hardware developers use a totally different technique. For analyzing and troubleshooting electronic systems, they use an additional external device called signal or function or waveform generator. The signal generator is an important piece of electronic test equipment, and should not be missed in any electronics laboratory.
On the device, we can select a built-in periodical function like y=F(t); the variable t represents real time, and we can change some parameters like frequency, amplitude, and the DC-offset. Typically, there are four basic signal types.
After customization, the desired output signal can be wired on to the input of the component that should be tested, and we monitor their response function in the time domain using an oscilloscope.
Displayed here is a typical low-cost signal generator that has on its front panel, a radio-button with four positions to choose signal types and three potentiometers to adjust frequency, amplitude, and DC-offset. It is no big problem for hardware dudes to make something such as this, especially with tips from the book [1].
Of course, there are much better and more expensive devices on the market with more functions, more parameters to choose, better range and a finer scale for these parameters, integrated display to show selected settings, user-defined functions, on-board memory, better stability, and two or more independent channels.
Note that, there are two very significant differences between the hardware and the software approach. Hardware developers use an external device, and a test signal is in the time domain, which implies that the test signal varies totally independent from the assembly that will be tested.`
The class presented offers a software equivalent for the above described real, simple signal generator device.
Using the Code
For using the class, we have a public function in the form y=F(t). The parameter t is the real time here:
float GetValue();
After a lot of consideration, I added another overload of the function in the form y=F(x). The Parameter x is explicitly given as a time here:
float GetValue(float time);
This should be used publicly only for DEBUG purposes, eventually during tests by adding new signal types!
All signals have a normalized period T € [0,1] in cycles, and not the usual T € [0,2Pi] in radian. In this case, a value 1 corresponds to a full cycle. The basis for this is an easier deal with normalized values, for example, for scaling operations by fitting in a window.
The signal generator is given as a class, but it is quite easy to transform it to a non-visual, or yet to a visual component, or maybe to a form like a front panel.
For generating all the given signals, very simple functions are used, but they are not the simplest. Some functions can be made simpler, but a development goal of this project was to generate signals exactly like the following example on Wikipedia:
Sine, square, triangle, and sawtooth waveforms By Omegatron - Own work, CC BY-SA 3.0
References
Used by
Basic Audio Sample
Basic Tasks Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SmingTest
An extensible test framework for Sming, with integrated profiling and scheduling support.
Getting started
A Sample test skeleton project is provided which you can copy to provide a starting point.
See the Sming /tests/HostTests test application for a more comprehensive example.
A test module contains all code related to testing a specific Component, Library or framework module
If a module contains more than one file then place all files in a sub-directory
Each module must have a registration function:
Use the
REGISTER_TESTmacro to name the functionAdd an entry to the Sample/app/modules.h file
Each module contains one or more test groups:
Each group is a class inheriting from TestGroup
Add a call to
registerGroup()to the registration function for each test class
Keep each group brief. Use multiple, simpler groups if necessary.
The group
TestGroup::execute()method is called to run the tests. On return the group is considered to have been successfully completed and destroyed.Call
TEST_ASSERTat appropriate points; passingfalsewill fail the test and abort the process, displaying the source location of the failure.If a test fails then additional details may be shown before calling
TEST_ASSERT(false).For asynchronous testing calling
TestGroup::pending()before returning. When the tests have been completed callTestGroup::complete(). (See /tests/HostTests/modules/Timers.cpp for an example.)
The following macros are added for ease of importing tests from other frameworks:
What happens
The registerGroup() function creates a factory function which is added to the SmingTest::Runner::groupFactories list.
The test runner creates, executes and destroys each group in turn, and deals with scheduling.
Notes
Tests are run with DEBUG_VERBOSE_LEVEL at WARNING level, so debug_i statements will not normally be shown.
Tests can use other debug_X functions as required, or Serial print methods.
Tests should compile and run for all architectures.
API Documentation
Defines
-
REGISTER_TEST(name)
Provides consistent global name for test factory function.
Test modules should use this macro to implement factory function:
#include <SmingTest.h> class SampleTest: public TestGroup { ... }; void REGISTER_TEST(sample) { registerGroup<SampleTest>(); }
- Parameters:
name – Name of test
Functions
-
namespace SmingTest
Typedefs
-
using Callback = Delegate<void()>
-
using Callback = Delegate<void()>
Internal check macros
-
INTERNAL_CHECK(expr, verbose)
-
INTERNAL_CHECK2(res, expr, verbose)
-
INTERNAL_CHECK_EQ(a, b, verbose)
-
INTERNAL_CHECK_NEQ(a, b, verbose)
Check an expression, print message for success or failure (verbose)
-
CHECK(expr)
Check expression evaluates to true.
- Parameters:
expr –
-
CHECK2(res, expr)
Provide separate test result and expression.
- Parameters:
res – Result of test
expr – Expression to display
-
CHECK_EQ(a, b)
Check two values are the same.
- Parameters:
a –
b –
-
CHECK_NEQ(a, b)
Check two values are not the same.
- Parameters:
a –
b –
Check an expression, but only print message on failure
-
REQUIRE(expr)
Check expression evaluates to true.
- Parameters:
expr –
-
REQUIRE2(res, expr)
Provide separate test result and expression.
- Parameters:
res – Result of test
expr – Expression to display
-
REQUIRE_EQ(a, b)
Check two values are the same.
- Parameters:
a –
b –
-
REQUIRE_NEQ(a, b)
Check two values are not the same.
- Parameters:
a –
b –
Defines
-
TEST_ASSERT(result)
Check a test result.
Note
Failure generates an assertion so when run in the host emulator the process fails.
- Parameters:
result – true if test was successful, false on failure
-
class TestBase
- #include <TestBase.h>
Base class supporting verification for test assertions.
Subclassed by TestGroup
Public Functions
-
inline virtual ~TestBase()
-
virtual bool testVerify(bool res, const TestParam ¶m)
Print result of a test.
- Parameters:
res – Result of the operation
param – Details of the test for display
- Return values:
bool – Same as res
-
inline bool test_verify(bool res, const char *expr, const String &value1, const String &value2, bool verbose)
-
template<typename V>
inline std::enable_if<std::is_arithmetic<V>::value, bool>::type test_verify(bool res, const char *expr, const V &value1, const V &value2, bool verbose)
-
template<typename V>
inline std::enable_if<!std::is_same<V, String>::value && !std::is_arithmetic<V>::value, bool>::type test_verify(bool res, const char *expr, const V &value1, const V &value2, bool verbose)
-
inline virtual void fail(const char *func)
-
struct TestParam
- #include <TestBase.h>
Contains details for test verification.
-
inline virtual ~TestBase()
Defines
-
startTest(s)
-
TEST_CASE_1_ARG(name)
Start a test item.
Use like this:
TEST_CASE("My Test", "description") { ... }
Note: Description is optional.
-
TEST_CASE_2_ARGS(name, desc)
-
GET_3RD_ARG(arg1, arg2, arg3, ...)
-
TEST_CASE_ARG_CHOOSER(...)
-
TEST_CASE(...)
-
class TestGroup : public TestBase
- #include <TestGroup.h>
Class to simplify generation of begin/end messages for a test group.
Public Types
Public Functions
-
void commenceTest()
-
virtual void execute() = 0
Implement this method to define the test.
Note
If tests are asynchronous, call
pending()before returning and callcomplete()when the group has completed execution (e.g. via timer callback, etc.)
-
void startItem(const String &tag, const String &description = nullptr)
Note the start of a test item within a group.
-
virtual void fail(const char *func) override
Called when test fails to identify location.
-
void initialiseAndExecute()
Called by test runner.
Protected Functions
-
inline void pending()
Call to mark test as pending so it will be executed asynchronously Call
complete()when test is finished.
-
void complete()
Call to complete pending (asynchronous) test.
-
void commenceTest()
References
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Solar Calculator
Arduino library to support calculation of sunrise / sunset times
See SystemClock NTP for example usage.
This is straight port of the code used by https://www.esrl.noaa.gov/gmd/grad/solcalc/
Javascript reference: https://www.esrl.noaa.gov/gmd/grad/solcalc/main.js
API Documentation
-
class SolarCalculator
Calculation of apparent time of sunrise and sunset.
Note: Months are 1-based
Unnamed Group
-
int sunRiseSet(bool isRise, int y, int m, int d)
@briefCalculate a sunrise or sunset figure for a given day.
- Parameters:
isRise – true for sunrise, false for sunset
y – Absolute year
m – Month number (1 - 12)
d – Day of month (1 - 31)
- Return values:
int – Minutes since midnight, -1 if there is no sunrise/sunset
Public Functions
-
inline SolarCalculator()
Default constructor, uses Royal Observatory, Greenwich as default.
-
int sunRiseSet(bool isRise, int y, int m, int d)
-
struct SolarRef
A location is required to compute solar times.
References
Used by
SystemClock NTP Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SparkFun APDS9960 RGB and Gesture Sensor Arduino Library
*Avago APDS-9960 Breakout Board (SEN-12787)*
Getting Started
Download the Git repository as a ZIP (“Download ZIP” button)
Unzip
Copy the entire library directory (APDS-9960_RGB_and_Gesture_Sensor_Arduino_Library ) to <Arduino installation directory>/libraries
Open the Arduino program
Select File -> Examples -> SparkFun_APDS9960 -> GestureTest
Plug in your Arduino and APDS-9960 with the following connections
-OR-
Use the library manager
Arduino Pin |
APDS-9960 Board |
Function |
|---|---|---|
3.3V |
VCC |
Power |
GND |
GND |
Ground |
A4 |
SDA |
I2C Data |
A5 |
SCL |
I2C Clock |
2 |
INT |
Interrupt |
Go to Tools -> Board and select your Arduino board
Go to Tools -> Serial Port and select the COM port of your Arduino board
Click “Upload”
Go to Tools -> Serial Monitor
Ensure the baud rate is set at 9600 baud
Swipe your hand over the sensor in various directions!
Repository Contents
/examples - Example sketches for the library (.ino). Run these from the Arduino IDE.
/src - Source files for the library (.cpp, .h).
library.properties - General library properties for the Arduino package manager.
Documentation
**Installing an Arduino Library Guide** - Basic information on how to install an Arduino library.
**Product Repository** - Main repository (including hardware files) for the SparkFun_APDS9960 RGB and Gesture Sensor.
**Hookup Guide** - Basic hookup guide for the sensor.
Products that use this Library
SEN-12787- Avago APDS-9960
Version History
V_1.4.1 - Removing blank files, updating library.properties file.
V_1.4.0 - Updated to new library structure
V_1.3.0 - Implemented disableProximitySensor(). Thanks to jmg5150 for catching that!
V_1.2.0 - Added pinMode line to GestureTest demo to fix interrupt bug with some Arduinos
V_1.1.0 - Updated GestureTest demo to not freeze with fast swipes
V_1.0.0: Initial release
Ambient and RGB light sensing implemented
Ambient light interrupts working
Proximity sensing implemented
Proximity interrupts working
Gesture (UP, DOWN, LEFT, RIGHT, NEAR, FAR) sensing implemented
License Information
This product is *open source_!
The code is beerware; if you see me (or any other SparkFun employee) at the local, and you’ve found our code helpful, please buy us a round!
Please use, reuse, and modify these files as you see fit. Please maintain attribution to SparkFun Electronics and release anything derivative under the same license.
Distributed as-is; no warranty is given.
Your friends at SparkFun.
References
Used by
APDS-9660 Gesture Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
SPIFFS IFS Library
This Component provides SPIFFS filesystem support for all architectures.
A single SPIFFS partition is defined using HWCONFIG =spiffs, which supports these build variables:
- DISABLE_SPIFFS
[deprecated and removed]
This value is no longer supported. Please remove it from your project’s component.mk file.
- SPIFF_SIZE
[deprecated and removed]
Size (in bytes) of the SPIFFS area in Flash memory. To change this, edit the Hardware configuration.
- SPIFF_FILES
default:
filesThe SPIFFS image is built using files from this directory, which must exist or the build will fail.
If you set this to an empty value, then an empty filesystem will be created.
- SPIFF_BIN
Filename to use for the generated SPIFFS filesystem image. The default is
spiff_rom.
- SPIFF_BIN_OUT
[read-only] Shows the full path to the generated image file.
For more control over the SPIFFS partition you can create your own partition definition in a custom Hardware configuration.
- SPIFF_FILEDESC_COUNT
Default: 7
Number of file descriptors allocated. This sets the maximum number of files which may be opened at once.
- SPIFFS_OBJ_META_LEN
Default: 16
Maximum size of metadata which SPIFFS stores in each file index header (after the filename). If this value is changed, existing SPIFFS images will not be readable.
The default value given here is provided to support Installable File System extended file attribute information.
The first 16 bytes are used for system attributes (e.g. modified time), so setting this to, say, 64 leaves 48 bytes for user metadata. Each attribute has a 2-byte header (tag + size) so a single user attribute can be stored of up to 46 bytes, or multiple tags up to this limit.
Note: LittleFS provides better support for user metadata.
References
Used by
Sming (main) Component
Basic IFS Sample
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: spiffs
SPIFFS (SPI Flash File System)
V0.3.7
Copyright (c) 2013-2017 Peter Andersson (pelleplutt1976 at gmail.com)
For legal stuff, see LICENSE. Basically, you may do whatever you want with the source. Use, modify, sell, print it out, roll it and smoke it - as long as I won’t be held responsible.
Love to hear feedback though!
INTRODUCTION
Spiffs is a file system intended for SPI NOR flash devices on embedded targets.
Spiffs is designed with following characteristics in mind:
Small (embedded) targets, sparse RAM without heap
Only big areas of data (blocks) can be erased
An erase will reset all bits in block to ones
Writing pulls one to zeroes
Zeroes can only be pulled to ones by erase
Wear leveling
BUILDING
mkdir build; make
Otherwise, configure the builddir variable towards the top of makefile as something opposed to the default build. Sanity check on the host via make test and refer to .travis.yml for the official in-depth testing procedure. See the wiki for integrating spiffs into projects and spiffsimg from nodemcu is a good example on the subject.
FEATURES
What spiffs does:
Specifically designed for low ram usage
Uses statically sized ram buffers, independent of number of files
Posix-like api: open, close, read, write, seek, stat, etc
It can run on any NOR flash, not only SPI flash - theoretically also on embedded flash of a microprocessor
Multiple spiffs configurations can run on same target - and even on same SPI flash device
Implements static wear leveling
Built in file system consistency checks
Highly configurable
What spiffs does not:
Presently, spiffs does not support directories. It produces a flat structure. Creating a file with path tmp/myfile.txt will create a file called tmp/myfile.txt instead of a myfile.txt under directory tmp.
It is not a realtime stack. One write operation might last much longer than another.
Poor scalability. Spiffs is intended for small memory devices - the normal sizes for SPI flashes. Going beyond ~128Mbyte is probably a bad idea. This is a side effect of the design goal to use as little ram as possible.
Presently, it does not detect or handle bad blocks.
One configuration, one binary. There’s no generic spiffs binary that handles all types of configurations.
NOTICE
0.4.0 is under construction. This is a full rewrite and will change the underlying structure. Hence, it will not be compatible with earlier versions of the filesystem. The API is the same, with minor modifications. Some config flags will be removed (as they are mandatory in 0.4.0) and some features might fall away until 0.4.1. If you have any worries or questions, it can be discussed in issue #179
MORE INFO
See the wiki for configuring, integrating, using, and optimizing spiffs.
For design, see docs/TECH_SPEC.
For a generic spi flash driver, see this.
HISTORY
0.3.7
fixed prevent seeking to negative offsets #158
fixed file descriptor offsets not updated for multiple fds on same file #157
fixed cache page not closed for removed files #156
fixed a lseek bug when seeking exactly to end of a fully indexed first level LUT #148
fixed wear leveling issue #145
fixed attempt to write out of bounds in flash #130,
set file offset when seeking over end #121 (thanks @sensslen)
fixed seeking in virgin files #120 (thanks @sensslen)
Optional file metadata #128 (thanks @cesanta)
AFL testing framework #100 #143 (thanks @pjsg)
Testframe updates
New API functions:
SPIFFS_update_meta, SPIFFS_fupdate_meta- updates metadata for a file
New config defines:
SPIFFS_OBJ_META_LEN- enable possibility to add extra metadata to files
0.3.6
Fix range bug in index memory mapping #98
Add index memory mapping #97
Optimize SPIFFS_read for large files #96
Add temporal cache for opening files #95
More robust gc #93 (thanks @dismirlian)
Fixed a double write of same data in certain cache situations
Fixed an open bug in READ_ONLY builds
File not visible in SPIFFS_readdir #90 (thanks @benpicco-tmp)
Cache load code cleanup #92 (thanks @niclash)
Fixed lock/unlock asymmetry #88 #87 (thanks @JackJefferson, @dpruessner)
Testframe updates
New API functions:
SPIFFS_ix_map- map index meta data to memory for a fileSPIFFS_ix_unmap- unmaps index meta data for a fileSPIFFS_ix_remap- changes file offset for index metadata mapSPIFFS_bytes_to_ix_map_entries- utility, get length of needed vector for given amount of bytesSPIFFS_ix_map_entries_to_bytes- utility, get number of bytes a vector can represent given length
New config defines:
SPIFFS_IX_MAP- enable possibility to map index meta data to memory for reading fasterSPIFFS_TEMPORAL_FD_CACHE- enable temporal cache for opening files fasterSPIFFS_TEMPORAL_CACHE_HIT_SCORE- for tuning the temporal cache
0.3.5
Fixed a bug in fs check
API returns actual error codes #84) (thanks @Nails)
Fix compiler warnings for non-gcc #83 #81 (thanks @Nails)
Unable to recover from full fs #82 (thanks @rojer)
Define SPIFFSO* flags #80
Problem with long filenames #79 (thanks @psjg)
Duplicate file name bug fix #74 (thanks @igrr)
SPIFFS_eof and SPIFFS_tell return wrong value #72 (thanks @ArtemPisarenko)
Bunch of testframe updates #77 #78 #86 (thanks @dpreussner, @psjg a.o)
0.3.4
Added user callback file func.
Fixed a stat bug with obj id.
SPIFFS_probe_fs added
Add possibility to compile a read-only version of spiffs
Make magic dependent on fs length, if needed (see #59 & #66) (thanks @hreintke)
Exposed SPIFFS_open_by_page_function
Zero-size file cannot be seek #57 (thanks @lishen2)
Add tell and eof functions #54 (thanks @raburton)
Make api string params const #53 (thanks @raburton)
Preserve user_data during mount() #51 (thanks @rojer)
New API functions:
SPIFFS_set_file_callback_func- register a callback informing about file eventsSPIFFS_probe_fs- probe a spi flash trying to figure out size of fsSPIFFS_open_by_page- open a file by page indexSPIFFS_eof- checks if end of file is reachedSPIFFS_tell- returns current file offset
New config defines:
SPIFFS_READ_ONLYSPIFFS_USE_MAGIC_LENGTH
0.3.3
Might not be compatible with 0.3.2 structures. See issue #40
Possibility to add integer offset to file handles
Truncate function presumes too few free pages #49
Bug in truncate function #48 (thanks @PawelDefee)
Update spiffs_gc.c - remove unnecessary parameter (thanks @PawelDefee)
Update INTEGRATION docs (thanks @PawelDefee)
Fix pointer truncation in 64-bit platforms (thanks @igrr)
Zero-sized files cannot be read #44 (thanks @rojer)
(More) correct calculation of max_id in obj_lu_find #42 #41 (thanks @lishen2)
Check correct error code in obj_lu_find_free #41 (thanks @lishen2)
Moar comments for SPIFFS_lseek (thanks @igrr)
Fixed padding in spiffs_page_object_ix #40 (thanks @jmattsson @lishen2)
Fixed gc_quick test (thanks @jmattsson)
Add SPIFFS_EXCL flag #36
SPIFFS_close may fail silently if cache is enabled #37
User data in callbacks #34
Ignoring SINGLETON build in cache setup (thanks Luca)
Compilation error fixed #32 (thanks @chotasanjiv)
Align cand_scores (thanks @hefloryd)
Fix build warnings when SPIFFS_CACHE is 0 (thanks @ajaybhargav)
New config defines:
SPIFFS_FILEHDL_OFFSET
0.3.2
Limit cache size if too much cache is given (thanks pgeiem)
New feature - Controlled erase. #23
SPIFFS_rename leaks file descriptors #28 (thanks benpicco)
moved dbg print defines in test framework to params_test.h
lseek should return the resulting offset (thanks hefloryd)
fixed type on dbg ifdefs
silence warning about signed/unsigned comparison when spiffs_obj_id is 32 bit (thanks benpicco)
Possible error in test_spiffs.c #21 (thanks yihcdaso-yeskela)
Cache might writethrough too often #16
even moar testrunner updates
Test framework update and some added tests
Some thoughts for next gen
Test sigsevs when having too many sectors #13 (thanks alonewolfx2)
GC might be suboptimal #11
Fix eternal readdir when objheader at last block, last entry
New API functions:
SPIFFS_gc_quick- call a nonintrusive gcSPIFFS_gc- call a full-scale intrusive gc
0.3.1
Removed two return warnings, was too triggerhappy on release
0.3.0
Added existing namecheck when creating files
Lots of static analysis bugs #6
Added rename func
Fix SPIFFS_read length when reading beyond file size
Added reading beyond file length testcase
Made build a bit more configurable
Changed name in spiffs from “errno” to “err_code” due to conflicts compiling in mingw
Improved GC checks, fixed an append bug, more robust truncate for very special case
GC checks preempts GC, truncate even less picky
Struct alignment needed for some targets, define in spiffs config #10
Spiffs filesystem magic, definable in config
New config defines:
SPIFFS_USE_MAGIC- enable or disable magic check upon mountSPIFFS_ALIGNED_OBJECT_INDEX_TABLES- alignment for certain targets
New API functions:
SPIFFS_rename- rename filesSPIFFS_clearerr- clears last errnoSPIFFS_info- returns info on used and total bytes in fsSPIFFS_format- formats the filesystemSPIFFS_mounted- checks if filesystem is mounted
Switch Joycon
Introduction
This library allows you to make the ESP32 act as a Nintendo Switch Joycon and control what it does. The library uses ESP32 NimBLE for faster and lighter communication.
Disclaimer
We are not affiliated, associated, authorized, endorsed by, or in any way officially connected with Nintendo, or any of its subsidiaries or its affiliates. The names Nintendo, Nintendo Switch and Joycon as well as related names, marks, emblems and images are registered trademarks of their respective owners.
Features
Using this library you can do the following:
Button press and release (16 buttons)
Switch Hat (1 hat )
Rotate 4 Axis
Using
Add
COMPONENT_DEPENDS += SwitchJoyconto your application componenent.mk file.Add these lines to your application:
#include <SwitchJoycon.h> namespace { SwitchJoycon joycon; // ... } // namespace void init() { // ... joycon.begin(); }
Notes
By default, reports are sent on every button press/release or axis/hat movement, however this can be disabled:
joycon.setAutoReport(false);
and then you should manually call sendReport on the joycon instance as shown below:
joycon.sendReport();
HID Debugging
On Linux you can install hid-tools using the command below:
sudo pip3 install .
Once installed hid-recorder can be used to check the device HID report description and sniff the different reports:
sudo hid-recorder
Useful Links
References
Used by
Switch Joycon Sample
SoC support
esp32
esp32c3
esp32s3
ILI9163C TFT Display
ILI9163C- A fast SPI driver for TFT that use Ilitek ILI9163C.
Link to a video:
https://www.youtube.com/watch?v=y5f-VNBxgEk&feature=youtu.be
Features:
- Very FAST!, especially with Teensy 3.x where uses DMA SPI.
- It uses just 4 or 5 wires.
- Compatible at command level with Adafruit display series so it's easy to adapt existing code.
- It uses the standard Adafruit_GFX Library (you need to install).
Pay Attention to connections!!!!:
- This display has logic at 3V3 volt so YOU NEED A VOLTAGE CONVERTER if you plan to use with arduino.
If you try to connect directly you can burn it very fast so PAY ATTENTION!
- My display works at 3V3 volt for logic but LED background has resistor for 5V. Your can be different
so carefully check out before connect it.
- My library works only in SPI mode by using MOSI,SCLK and a CS pin plus an additional pin for DC (or RS).
I've used also the reset pin but you can save it by connect it at 3V3 volt and use the constructor without
the reset pin. The initialization routine will automatically use the software reset.
- People using Teensy3 should remember that have to choose for CS and DC a pin that satisfy:
if (pin == 2 || pin == 6 || pin == 9) return true;
if (pin == 10 || pin == 15) return true;
if (pin >= 20 && pin <= 23) return true;
Background:
I got one of those displays from a chinese ebay seller but unfortunately I cannot get
any working library so I decided to hack it. ILI9163C looks pretty similar to other
display driver but it uses it's own commands so it's tricky to work with it unlsess you
carefully fight with his gigantic and not so clever datasheet.
My display it's a 1.44"", 128x128 that suppose to substitute Nokia 5110 LCD and here's the
first confusion! Many sellers claim that it's compatible with Nokia 5110 (that use a philips
controller) but the only similarity it's the pin names since that this one it's color and
have totally different controller that's not compatible. Altrough I discovered that it's not
128x128 but 128x160 (!??)... Check links below to see if it's similar to yours.
http://www.elecrow.com/144-128x-128-tft-lcd-with-spi-interface-p-855.html
Pay attention that ILI9163C can drive different resolutions and your display can be
160*128 or whatever, also there's a strain of this display with a black PCB that a friend of mine
got some weeks ago and need some small changes in library to get working.
If you look at TFT_ILI9163C.h file you can add your modifications and let me know so I
can include for future versions.
Code Optimizations:
The purpose of this library it's SPEED. I have tried to use hardware optimized calls
where was possible and results are quite good for most applications.
Of course it can be improved so feel free to add suggestions.
Copyright (c) 2014, .S.U.M.O.T.O.Y., coded by Max MC Costa.
TFT_ILI9163C Library is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
TFT_ILI9163C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with Foobar. If not, see <http://www.gnu.org/licenses/>.
This file needs the following Libraries:
*Adafruit_GFX by Adafruit:
https://github.com/adafruit/Adafruit-GFX-Library
Remember to update GFX library often to have more features with this library!
*From this version I'm using my version of Adafruit_GFX library:
https://github.com/sumotoy/Adafruit-GFX-Library
It has faster char rendering and some small little optimizations but you can
choose one of the two freely since are both fully compatible.
Special Thanks:
Thanks Adafruit for his Adafruit_GFX!
Thanks to Paul Stoffregen for his beautiful Teensy3 and DMA SPI.
Version:
0.1a1: First release, compile correctly. Altrough not fully working!
0.1a3: Some bugfix, still some addressing problems, partial rotation solved.
0.1b1: Beta version! Fully working but still not tested with Arduino and DUE (altrough it compile)
0.2b2: Code cleaned and added support for 2.2" RED PCB displays.
0.2b4: Bug fixes and added color space support.
BugList of the current version:
- Beta version fully working!
References
Used by
ILI9163C TFT Screen Sample
SoC support
esp8266
TFT_S1D13781
Introduction
This is a port of the driver from the Epson S1D13781 display controller using the HardwareSPI library.
Evaluation boards are inexpensive and is a useful way to evaluate display modules with TFT interfaces.
The Newhaven NHD-5.0-800480TF-ATXL#-CTP was used during development.
Video line: https://youtu.be/UgLX9gEdz6A
Notes
This library was developed as a stepping stone to the Bridgetek FT81x series EVE Embedded Video Engines.
These are more advanced and work by buffering graphics command primitives which are then executed in real time to construct each displayed frame.
These devices are used in the Gameduino. The author has an open source library.
This library is not fully asynchronous, but has couple of key improvements over the original source:
- Cached register values
These do not need to be read for every operation. Considerable increase in performance.
- Asynchronous writes
The HardwareSPI library can execute requests asynchronously, and this is used to provide some improvement in performance where no response is required from the display controller.
The next stage of development would be to build a generic graphics library to support multiple display controllers. Like the EVE controllers, it would incorporate a graphics instruction pipeline so that the application can buffer drawing requests in a fully asynchronous manner, eliminating any bottlenecks or wait states in the application.
References
Used by
Screen TFT_S1D13781 Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
TM1637 LED Driver
Description
An Arduino library for 7-segment display modules based on the TM1637 chip, such as Seeed Studio’s Grove 4 digit display. The TM1637 chip also has keyboard input capability, but it’s not implemented in this library.
Hardware Connection
The display modules has two signal connection (and two power connections) which are CLK and DIO. These pins can be connected to any pair of digital pins on the Arduino. When an object is created, the pins should be configured. There is no limitation on the number of instances used concurrently (as long as each instance has a pin pair of its own)
Installation
The library is installed as any Arduino library, by copying the files into a directory on the library search path of the Arduino IDE
Usage
The library provides a single class named TM1637Display. An instance of this class provides the following functions:
setSegments- Set the raw value of the segments of each digitshowNumberDec- Display a decimal numbershowNumberDecEx- Display a decimal number with decimal points or colonsetBrightness- Sets the brightness of the display
The information given above is only a summary. Please refer to TM1637Display.h for more information. An example is included, demonstrating the operation of most of the functions.
References
Used by
TM1637 Display Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Timezone
Arduino library to support local/UTC time conversions using rules
See SystemClock NTP for example usage.
Port of https://github.com/JChristensen/Timezone for Sming.
API Documentation
-
class Timezone
Class to support local/UTC time conversions using rules.
Public Functions
-
time_t toLocal(time_t utc, const TimeChangeRule **p_tcr = nullptr)
Convert the given UTC time to local time, standard or daylight time.
- Parameters:
utc – Time in UTC
p_tcr – Optionally return the rule used to convert the time
- Return values:
time_t – The local time
-
time_t toUTC(time_t local)
Convert the given local time to UTC time.
WARNING: This function is provided for completeness, but should seldom be needed and should be used sparingly and carefully.
Ambiguous situations occur after the Standard-to-DST and the DST-to-Standard time transitions. When changing to DST, there is one hour of local time that does not exist, since the clock moves forward one hour. Similarly, when changing to standard time, there is one hour of local times that occur twice since the clock moves back one hour.
This function does not test whether it is passed an erroneous time value during the Local -> DST transition that does not exist. If passed such a time, an incorrect UTC time value will be returned.
If passed a local time value during the DST -> Local transition that occurs twice, it will be treated as the earlier time, i.e. the time that occurs before the transition.
Calling this function with local times during a transition interval should be avoided.
Note
- Parameters:
local – Local time
- Return values:
time_t – Time in UTC
-
bool utcIsDST(time_t utc)
Determine whether the UTC time is within the DST interval or the Standard time interval.
- Parameters:
utc –
- Return values:
bool – true if time is within DST
-
bool locIsDST(time_t local)
Determine whether the given local time is within the DST interval or the Standard time interval.
- Parameters:
local – Local time
- Return values:
bool – true if time is within DST
-
inline const char *timeTag(bool isDst) const
Return the appropriate dalight-savings tag to append to displayed times.
- Parameters:
isDst – true if DST tag is required, otherwise non-DST tag is returned
- Return values:
const – char* The tag
-
inline const char *utcTimeTag(time_t utc)
Return the appropriate time tag for a UTC time.
- Parameters:
utc – The time in UTC
- Return values:
const – char* Tag, such as UTC, BST, etc.
-
inline const char *localTimeTag(time_t local)
Return the appropriate time tag for a local time.
- Parameters:
local – The local time
- Return values:
const – char* Tag, such as UTC, BST, etc.
-
time_t toLocal(time_t utc, const TimeChangeRule **p_tcr = nullptr)
-
struct TimeChangeRule
-
enum week_t
Values:
-
enumerator Last
-
enumerator First
-
enumerator Second
-
enumerator Third
-
enumerator Fourth
-
enumerator Last
References
Used by
SystemClock NTP Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Tone Generator
Uses the I2S module to synthesize tones using a minimal amount of Flash memory, RAM and CPU time.
Audio is output via the I2S_TX_DATA pin using Delta-Sigma modulation. This is a high-frequency bitstream which requires low-pass filtering and suitable amplification for driving speaker/headphones.
Warning
Do not connect the output directly to a loudspeaker or headphone as you may damage the GPIO. Always buffer the signal using a transistor, op-amp, etc.
See Esp8266 Drivers for further information about I2S.
As the output pin is normally used as UART0 RX, the alternate UART pin configuration is used:
GPIO |
Alternate |
NodeMCU |
Notes |
|---|---|---|---|
3 |
I2S_TX_DATA |
D9 |
Audio output |
13 |
RXD2 |
D7 |
Debug serial input |
15 |
TXD2 |
D8 |
Debug serial output |
GPIO3 is still required for programming so you’ll need to make sure the audio buffer is disconnected, or has a reasonably high impedance load. You’ll also need an additional USB-serial converter connected to GPIO13 & 15 for terminal operation.
API Documentation
-
class ToneGenerator
Generates tones with smooth transitions using Tone Buffers.
Note
During tone playback, a single
activeToneBuffer feeds the I2S with samples. When a new tone is prepared viaqueueTone(), a ToneBuffer is added to thependingqueue. If transitioning between tone and silence then additional fade in/out buffers are queued to implement low-pass filtering on these relatively abrupt transitions and thus reduce clicking. CallingsubmitPending()appends thependingqueue onto thetransitionqueue. The transition to the new tone (or silence) is made withini2sWrite(), at which point buffers are released back to theavailqueue.Public Functions
-
bool begin(unsigned sampleRate)
Initialise the tone generator and I2S.
- Parameters:
sampleRate –
- Return values:
bool – true on success
-
void end()
Stop playback and un-initialise I2S.
Note
Releases all allocated memory
-
bool start()
Start tone playback.
-
void stop()
Stop tone playback and release any memory allocated for buffers.
Note
Leaves I2S initialised. Call
start()to resume.
-
ToneBuffer *createTone(Voice voice, unsigned frequency, ToneEffect effect, unsigned repeatCount = 0)
Create a tone and queue it.
- Parameters:
voice – Voice to use for tone
frequency – Frequency in Hz
effect –
repeatCount – Specify non-zero to repeat this tone a specific number of times
- Return values:
ToneBuffer* – The queued buffer, nullptr on error
-
void queueTone(Voice voice, unsigned frequency)
Create a tone with appropriate filtering.
- Parameters:
voice – Voice to use for tone @frequency Frequency in Hz @
-
inline void submitPending()
Submit queued tone buffers for playback.
-
bool begin(unsigned sampleRate)
-
class ToneBuffer
Contains samples for one full signal cycle at a specific frequency.
Note
Data is stored Delta-Sigma modulated to minimise I2S transfer overhead.
References
Used by
RingTone Player Sample
Basic Audio Sample
SoC support
esp8266
host
UPnP Schema
This is a separate library for UPnP schema. It contains standard and reusable schema.
Code and header files are generated directly from schema to simplify the task of implementing hosted UPnP devices and controlling them.
You can find these under
Sming/out/{SMING_ARCH}/{debug|release}/build/UPnP-Schema/src.
Controlling UPnP devices
Devices and services must be registered with the UPnP framework so that the correct objects can be constructed during discovery. This information is generated in groups, with one group for each domain. This means you only need to include one file in your project. For example:
#include <Network/UPnP/schemas-sming-org/ClassGroup.h>
void initUPnP()
{
UPnP::schemas_upnp_org::registerClasses();
}
This tells UPnP about all the standard devices and services.
Custom schema
Custom schema may be imported from the application or another Component.
Create a schema sub-directory and arrange as:
schema/
{domain}/
device/
...
service/
...
Suitable schema are generated by the UPnP scan tool. These may be edited and customised with additional detail, comments, etc. as required.
Build variables
- UPNP_SCHEMA
Read-only. Contains a list of all discovered schema directories to be built for the project.
References
Used by
Hue Emulator Library
Basic ControlPoint Sample
Basic UPnP Sample
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
UPnP
Introduction
A C++ library for managing UPnP devices and services using the Sming framework.
About UPnP
Introduction
A C++ library for managing UPnP devices and services using the Sming framework.
Universal Plug and Play (UPnP) is a set of networking protocols that permits networked devices, such as personal computers, printers, Internet gateways, Wi-Fi access points and mobile devices to seamlessly discover each other’s presence on the network and establish functional network services for data sharing, communications, and entertainment. (Wikipedia)
UPnP provides a set of standards which allows user applications (Control Points such as a mobile phone App.) to discover and control embedded devices through standardised services.
Please refer to the official UPnP Device Architecture version 2.0 for further details.
Background
There are various open source UPnP libraries available, however we have a very specific set of requirements which none of them can really meet.
If (like me) you’re only dimly aware of UPnP and how it works, researching the requirements is an important first step and in itself is time consuming.
What we’d really like is a simple API which lets us throw together a fully UPnP compliant solution in the time it takes to drink a cup of coffee, with only a basic knowledge of how it all works.
- The ESP8266 is a ‘Class II IoT Device’
This classification is from RFC7228: Terminology for Constrained-Node Networks, and considers the resources available for an embedded device.
Basically, it can’t run Linux but it can support IP networking.
Whilst the ESP32 is ‘less constrained’, it’s still class II and without care and attention to framework (and application) design all that extra RAM and horsepower will rapidly disappear.
- UPnP (and SSDP) is just UDP, TCP, XML and JSON, so what’s the problem?
True, you can just throw together a custom solution which implements the various protocols and works OK. But like everything to do it properly takes time and effort.
If you want it to be reliable and behave like a good neighbour then sticking to the standards is important:
Be nice, don’t flood the network with search requests
But at the same time UDP is ‘unreliable’ so we should repeat messages
How to listen to SSDP broadcasts without hanging the system
- What’s this about XML? Why can’t we just use JSON?
One of the advantages of JSON is it’s compact, human-friendly format. XML is less so, but it is more structured and well-suited for machine-to-machine information transfer.
It’s a core part of UPnP. See UPnP Resources.
One of the many cool things about XML is XSLT, which lets us easily convert existing XML documents into any format we like. The tools to do this are readily available; we can perform these transformations in a web browser, for example, hence the wealth of online tools to do just this.
We can also use the published Schema to ensure strict standards compliance.
- How do we deal with XML on a memory-constrained device?
For very simple devices it’s not an issue, but larger sets of description information will need to be handled in small chunks and assembled before it gets sent out. We want our classes to take care of all this transparently.
If we want to discover other networked devices and control them, we’ll also need to parse the XML data in a similar way.
Tiny XML-2 is a very popular library. It’s a DOM parser so memory consumption is likely to be an issue.
Micro XPath is tiny, perhaps a little too simple though
Expat is a SAX parser written in C, could be just the ticket.
- Most of the UPnP libraries are for Linux or Windows
The OSGi Alliance provides a dynamic module system for java which includes many good design ideas and a great deal of useful information, such as:
Architecture, see list at bottom for references to the various open-source implementations.
OSGi Compendium Release 7, see org.osgi.service.upnp package
There are a couple of C/C++ frameworks which implemenent the OSGi specifications:
Others:
gUPnP object-oriented framework in C.
Curiously, I’ve failed to turn up anything for FreeRTOS.
I should probably mention Microsoft as they’re the ones who created UPnP in the first place! See UPnP APIs.
- Good integration with the Sming framework
I like Arduino-upnp, it’s a good approach and has some really great ideas. So if any of this library seems familiar that’ll be why!
In the end I decided to just build something from scratch. These are the specific areas I wanted to address:
- Support for existing Sming applications
You already have an embedded device but want to make it accessible via UPnP services. That means we want a lightweight layer which is easily added on top.
- Be careful with RAM
We don’t want to waste RAM keeping track of state and other information which is already available elsewhere.
Instead of keeping lists of services or devices we use an abstract enumerator. For example, a lighting controller may track the state of dozens or hundreds of lights, so the overhead of creating an object instance for each one is best avoided.
Also, there is a considerably amount of UPnP configuration information but it can be stored in Flash and only read out when actually required.
Classes
An application implements their devices and services using these provided base classes.
- DeviceHost
Implements SSDP discovery and notification supporting multiple root devices. The SSDP library handles the protocol details.
- Device
All devices are implemented using this base class, including root devices. A common example of a root device is a Television, with separate (embedded) devices controlling subsystems such as sound, vision, networking, etc.
Sming in a TV, now there’s an idea…
- Service
A service implements actions and manages state to control a device. Like when a REST request asks for a light to be turned on, it’ll be a service that performs the action and tracks state. The advantage with UPnP is that services are self-documented. You can explore this using various UPnP Tools.
- Item
All UPnP classes are implemented using the Item class template, which allows them to be efficiently enumerated as a linked list. Class templates are ideal because they avoid the complication of dynamic type casting and generate efficient code.
- List
A singly-linked list of items, such as devices or services.
Features
- Discovery
UPnP requires a minimal amount of information exchange to advertise services, however device descriptions can be relatively large and therefore unsafe to manipulate in a limited RAM system.
Sming’s template streams are one possible solution to this problem. The IMPORT_FSTR feature allows applications to easily define their own descriptions (templates or otherwise). The alternative is to use SPIFFS, however when Partition Tables are supported this will provide the best of both worlds.
However, the application should not normally need to do all this as the framework will, by default, enumerate device fields and build the device description information ‘on the fly’.
- Memory efficiency
Much of the UPnP framework is concerned with discovery and notification, which requires a significant amount of configuration data. This data is obtained via callbacks as required which allows device and service implementations to fetch it from flash memory storage or create it on demand, thus saving on RAM.
Using linked lists also avoids the need for separate RAM allocation and simpler enumeration. Applications are responsible for device and service memory allocation, but unless services need to be dynamically created or destroyed it’s simplest to just create them statically.
- Enumeration
One way to manage lists of many objects is to implement an enumerator with a single Service class instance. Every call to
enumerator.next()returns the same object instance but with its internal state updated.The main caveat to this approach is that if you need to keep hold of one these objects then you must make a copy; you cannot hold onto references. For this reason enumerators have a
clone()method and objects have copy constructors.
API Documentation
-
namespace UPnP
-
Enums
-
enum ErrorValues
Values:
Variables
-
DeviceHost deviceHost
-
class ActionRequest : public UPnP::ActionResponse
- #include <ActionRequest.h>
-
class ActionRequestControl : public UPnP::ActionResponse
- #include <ActionRequest.h>
-
class ActionResponse : public UPnP::LinkedItem
- #include <ActionResponse.h>
Class to handle action requests and responses.
Subclassed by UPnP::ActionRequest, UPnP::ActionRequestControl
-
class Stream : public MemoryDataStream
- #include <ActionResponse.h>
-
class Stream : public MemoryDataStream
-
class Base64
- #include <Base64.h>
-
struct SearchFilter
- #include <BaseObject.h>
-
class BaseObject : public UPnP::LinkedItem
- #include <BaseObject.h>
Objects which hook into the SSDP message stack.
Subclassed by UPnP::ObjectTemplate< ControlPoint, BaseObject >, UPnP::Object
Public Functions
-
virtual bool formatMessage(Message &msg, MessageSpec &ms) = 0
Standard fields have been completed.
Note
Fields can be modified typically by adding any custom fields before sending response.
- Parameters:
msg – The message being constructed
ms – Template spec. for message
- Return values:
bool – Return true to send message, false to cancel
-
virtual void sendMessage(Message &msg, MessageSpec &ms)
Called by framework to construct then send a message.
- Parameters:
msg – Message to send
ms – Template spec.
-
virtual bool formatMessage(Message &msg, MessageSpec &ms) = 0
-
template<typename ObjectType, typename BaseObjectType>
class ObjectTemplate : public BaseObjectType - #include <BaseObject.h>
Base class template for linked items with type casting.
-
class Iterator
- #include <BaseObject.h>
-
class Iterator
-
struct ClassGroup
- #include <ClassGroup.h>
-
class ClassGroupList : public Vector<ClassGroup>
- #include <ClassGroup.h>
-
class ControlPoint : public UPnP::ObjectTemplate<ControlPoint, BaseObject>
- #include <ControlPoint.h>
Public Functions
-
inline ControlPoint(size_t maxResponseSize = 2048)
Constructor.
- Parameters:
maxResponseSize – Limits size of stream used to receive HTTP responses
-
inline void reset()
Cancel any outstanding search and reset the list of known unique service names.
-
inline void clear()
Perform a reset and destroy all created devices.
-
inline bool beginSearch(const Urn &urn, SsdpSearch::Callback callback)
Searches for UPnP device or service and returns SSDP response messages.
- Parameters:
urn – unique identifier of the service or device to find
callback – Invoked with SSDP response message
- Return values:
bool – true on success, false if request queue is full
-
inline bool beginSearch(const Urn &urn, DescriptionSearch::Callback callback)
Searches for UPnP device or service and fetches its description.
- Parameters:
urn – unique identifier of the service or device to find
callback – Invoked with device description document
- Return values:
bool – true on success, false if request queue is full
-
inline bool beginSearch(const ObjectClass &cls, DeviceSearch::Callback callback)
Searches for UPnP device.
- Parameters:
cls – Device class object
callback – Invoked with constructed control object
- Return values:
bool – true on success, false if request queue is full
-
inline bool beginSearch(const ObjectClass &cls, ServiceSearch::Callback callback)
Searches for UPnP service.
- Parameters:
cls – Service class object
callback – Invoked with constructed control object
- Return values:
bool – true on success, false if request queue is full
-
inline bool isSearchActive() const
Determine if there’s an active search in progress.
-
bool cancelSearch()
Cancel any active search operation.
- Todo:
Set timeout on search operation and call this automatically Need to inform application though - perhaps a generic callback on the class?
- Return values:
bool – true if a search was active, false if there was no active search
-
virtual void onNotify(BasicMessage &msg)
Called by framework to handle an incoming SSDP message.
- Parameters:
msg –
-
virtual bool formatMessage(Message &msg, MessageSpec &ms) override
Standard fields have been completed.
Note
Fields can be modified typically by adding any custom fields before sending response.
- Parameters:
msg – The message being constructed
ms – Template spec. for message
- Return values:
bool – Return true to send message, false to cancel
-
bool sendRequest(HttpRequest *request)
Send a request.
- Parameters:
request – Completed request object: leave response stream unassigned, will be set later
- Return values:
bool – true on success, false if queue is full
-
bool requestDescription(const String &url, DescriptionSearch::Callback callback)
Send a request for description document.
- Parameters:
request – Description URL
callback – To be invoked with requested document
- Return values:
bool – true on success, false if queue is full
Public Static Functions
-
static const ObjectClass *findClass(const Urn &objectType)
Find a registered class.
- Parameters:
objectType – Urn uniquely identifying the class
- Return values:
ObjectClass* – The class, or nullptr if not found
-
static inline void registerClasses(const FlashString &domain, const ObjectClass::List &classes)
Use this method to register all device and service classes you need to control.
Class information is required in order to instantiate device or service objects in response to incoming descriptions. This information must be pre-registered with the control point.
- Parameters:
domain –
classes – List of classes to register
-
inline ControlPoint(size_t maxResponseSize = 2048)
-
class DescriptionStream : public IDataSourceStream
- #include <DescriptionStream.h>
Public Functions
-
inline DescriptionStream(Object &object, const String &descriptionUrl)
Construct a description stream.
- Parameters:
object – The Object to enumerate
-
void reset()
Reset back to start.
Note
Handy if you want to re-use this stream to send it somewhere else
-
inline virtual bool isValid() const override
Determine if the stream object contains valid data.
Note
Where inherited classes are initialised by constructor this method indicates whether that was successful or not (e.g. FileStream)
- Return values:
bool – true if valid, false if invalid
-
virtual uint16_t readMemoryBlock(char *data, int bufSize) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
-
virtual bool seek(int len) override
Move read cursor.
- Parameters:
len – Relative cursor adjustment
- Return values:
bool – True on success.
-
inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
-
inline virtual String getName() const override
Returns name of the resource.
Note
Commonly used to obtain name of file
- Return values:
String –
-
inline virtual MimeType getMimeType() const override
Get MIME type for stream content.
- Return values:
MimeType –
-
inline DescriptionStream(Object &object, const String &descriptionUrl)
-
struct SpecVersion
- #include <Device.h>
-
class Device : public UPnP::ObjectTemplate<Device, Object>
- #include <Device.h>
Represents any kind of device, including a root device.
Subclassed by UPnP::DeviceControl
Public Functions
-
String resolvePath(const String &path) const
Resolve a path (relative or absolute) into an absolute path.
-
virtual void search(const SearchFilter &filter) override
Called during SSDP search operation.
-
virtual bool formatMessage(Message &msg, MessageSpec &ms) override
Standard fields have been completed.
Note
Fields can be modified typically by adding any custom fields before sending response.
- Parameters:
msg – The message being constructed
ms – Template spec. for message
- Return values:
bool – Return true to send message, false to cancel
-
virtual bool onHttpRequest(HttpServerConnection &connection) override
Called by framework to handle an incoming HTTP request.
- Parameters:
connection –
request –
response –
- Return values:
bool – true if request was handled
-
virtual IDataSourceStream *createDescription() override
Called by framework to construct a device description response stream.
By default, the framework generates a stream constructed from the device information fields, but this method may be overridden if, for example, a fixed description is stored in an .xml file.
- Return values:
IDataSourceStream* – The XML description content
-
String resolvePath(const String &path) const
-
class DeviceControl : public UPnP::Device
- #include <DeviceControl.h>
Subclassed by Dial::Client
Public Functions
-
bool configureRoot(ControlPoint &controlPoint, const String &location, XML::Node *device)
Called on root device only during discovery.
-
inline DeviceControl &root()
Get the root device.
-
inline virtual String getUrl(const String &path) const override
Get fully-qualified URL given a relative path.
-
inline ControlPoint &controlPoint() const
Get managing control point for this device.
-
template<typename T>
inline ServiceControl *getService(const T &serviceType) Find a service for this device given its class.
- Template Parameters:
serviceType – ObjectClass or Urn of service to locate
- Return values:
ServiceControl* – May be nullptr if not found
-
template<typename T>
inline DeviceControl *getDevice(const T &deviceType) Find a child device given its class.
- Template Parameters:
deviceType – ObjectClass or Urn of device to locate
- Return values:
DeviceControl* – May be nullptr if not found
-
inline virtual void onConnected(HttpConnection &connection)
Inherited classes may override this to pull out any additional information from received response headers, etc. Invoked after description has been processed.
-
inline const Description &description()
Get device description.
-
struct Description
- #include <DeviceControl.h>
-
bool configureRoot(ControlPoint &controlPoint, const String &location, XML::Node *device)
-
class DeviceHost
- #include <DeviceHost.h>
Public Functions
-
IDataSourceStream *generateDebugPage(const String &title)
Create an HTML page which applications may serve up to assist with debugging.
-
IDataSourceStream *generateDebugPage(const String &title)
-
template<typename ItemType, class EnumeratorType>
class Enumerator - #include <Enumerator.h>
Abstract class to enumerate items.
todo: We want a generic enumerator which returns Items, uses virtual methods so no upcasting required
Note
Returned items may only be considered valid for the duration of the current task call as they may be destroyed at any time.
Public Functions
-
virtual EnumeratorType *clone() = 0
Make a copy of this enumerator.
Note
Each copy maintains position independently
-
virtual void reset() = 0
Reset enumerator to start of list.
Note
Call to
next()will return first item
-
virtual EnumeratorType *clone() = 0
-
class Envelope
- #include <Envelope.h>
Class to manage a SOAP envelope for service request/response.
Unnamed Group
-
Error load(String &&content)
Load a SOAP document.
- Parameters:
content – MUST remain valid for the lifetime of this Envelope, or until initialise() is called.
Public Functions
-
void clear()
Wipe the envelope contents.
-
inline ContentType contentType() const
Get the current envelope content type.
-
inline void convertToResponse()
Set a flag that this should be converted to Response on next setArg() call.
-
inline void prepareResponse()
If Response is required but hasn’t been prepared yet, do it now. This wipes out the incoming request.
-
class Fault
- #include <Envelope.h>
-
Error load(String &&content)
-
class Item
- #include <Item.h>
Subclassed by Hue::Device, UPnP::LinkedItem
-
class ItemEnumerator : public UPnP::Enumerator<Item, ItemEnumerator>
- #include <ItemEnumerator.h>
Public Functions
-
inline virtual ItemEnumerator *clone() override
Make a copy of this enumerator.
Note
Each copy maintains position independently
-
inline virtual void reset() override
Reset enumerator to start of list.
-
inline virtual ItemEnumerator *clone() override
-
class LinkedItem : public UPnP::Item
- #include <LinkedItem.h>
Base class template for items in a list.
Subclassed by UPnP::ActionResponse, UPnP::BaseObject
-
class LinkedItemList
- #include <LinkedItemList.h>
Singly-linked list of items.
Note
We don’t own the items, just keep references to them
Subclassed by UPnP::ObjectList< Device >, UPnP::ObjectList< ObjectType >
-
class Object : public UPnP::BaseObject
- #include <Object.h>
Subclassed by UPnP::ObjectTemplate< Service, Object >, UPnP::ObjectTemplate< Device, Object >
Public Functions
-
virtual void search(const SearchFilter &filter) = 0
Called during SSDP search operation.
-
inline virtual bool onHttpRequest(HttpServerConnection &connection)
Called by framework to handle an incoming HTTP request.
- Parameters:
connection –
request –
response –
- Return values:
bool – true if request was handled
-
inline virtual IDataSourceStream *createDescription()
Called by framework to construct a device description response stream.
By default, the framework generates a stream constructed from the device information fields, but this method may be overridden if, for example, a fixed description is stored in an .xml file.
- Return values:
IDataSourceStream* – The XML description content
-
virtual void search(const SearchFilter &filter) = 0
-
struct ObjectClass
- #include <ObjectClass.h>
Describes device or service class.
Public Types
-
using Version = uint8_t
Interface version number.
-
using CreateObject = Object *(*)(DeviceControl *owner)
Object constructor function.
-
using Version = uint8_t
-
template<typename ObjectType>
class ObjectList : public UPnP::LinkedItemList - #include <ObjectList.h>
Class template for singly-linked list of objects.
Note
We don’t own the objects, just keep references to them
Subclassed by UPnP::OwnedObjectList< DeviceControl >, UPnP::OwnedObjectList< Service >, UPnP::OwnedObjectList< ObjectType >
Public Functions
-
template<typename T>
inline ObjectType *find(const T &objectType) Search list for matching entry.
- Template Parameters:
Urn – or String
- Return values:
ObjectType* – Located definition or nullptr if not found
-
inline ObjectType *find(const ObjectClass &objectClass)
Search list for matching entry given its class @objectClass Class information for object.
- Return values:
ObjectType* – Located definition or nullptr if not found
-
template<typename T>
-
template<typename ObjectType>
class OwnedObjectList : public UPnP::ObjectList<ObjectType> - #include <ObjectList.h>
Class template for singly-linked list of objects.
Note
We own the objects so are responsible for destroying them when removed
-
struct Search
- #include <Search.h>
This is a helper class used by ControlPoint to manage different search types.
Subclassed by UPnP::DescriptionSearch, UPnP::DeviceSearch, UPnP::ServiceSearch, UPnP::SsdpSearch
-
struct SsdpSearch : public UPnP::Search
- #include <Search.h>
Public Types
-
using Callback = Delegate<void(SSDP::BasicMessage &message)>
Callback invoked for every matching SSDP message.
-
using Callback = Delegate<void(SSDP::BasicMessage &message)>
-
struct DeviceSearch : public UPnP::Search
- #include <Search.h>
Public Types
-
using Callback = Delegate<bool(DeviceControl &device)>
Callback invoked when device has been located.
Note
If callback sends out action requests then must return true
- Param device:
A newly constructed instance matching the search criteria
- Retval bool:
Return true to keep the device, false to destroy it
-
using Callback = Delegate<bool(DeviceControl &device)>
-
struct ServiceSearch : public UPnP::Search
- #include <Search.h>
Public Types
-
using Callback = Delegate<bool(DeviceControl &device, ServiceControl &service)>
Callback invoked when service has been located.
Note
If callback sends out action requests then must return true
- Param device:
A newly constructed instance matching the search criteria
- Param service:
Requested service interface
- Retval bool:
Return true to keep the device, false to destroy it
-
using Callback = Delegate<bool(DeviceControl &device, ServiceControl &service)>
-
class Service : public UPnP::ObjectTemplate<Service, Object>
- #include <Service.h>
Represents any kind of device, including a root device.
Subclassed by UPnP::ServiceControl
Public Functions
-
virtual void search(const SearchFilter &filter) override
Called during SSDP search operation.
-
virtual bool formatMessage(Message &msg, MessageSpec &ms) override
Standard fields have been completed.
Note
Fields can be modified typically by adding any custom fields before sending response.
- Parameters:
msg – The message being constructed
ms – Template spec. for message
- Return values:
bool – Return true to send message, false to cancel
-
virtual bool onHttpRequest(HttpServerConnection &connection) override
Called by framework to handle an incoming HTTP request.
- Parameters:
connection –
request –
response –
- Return values:
bool – true if request was handled
-
virtual IDataSourceStream *createDescription() override
Called by framework to construct a device description response stream.
By default, the framework generates a stream constructed from the device information fields, but this method may be overridden if, for example, a fixed description is stored in an .xml file.
- Return values:
IDataSourceStream* – The XML description content
-
inline virtual bool sendRequest(HttpRequest *request) const
Implemented in ServiceControl.
-
virtual Error handleAction(ActionRequest &req) = 0
An action request has been received.
- Parameters:
env – Contains the action request
- Return values:
ErrorCode – Implementation should place response into
envafter reading parameters, e.g.:This is usually handled by generated wrapper class templates.if(env.actionName() == "MyAction") { String arg1; int arg2; env.getArg("arg1", arg1); env.getArg("arg2", arg2); auto& response = env.createResponse(); // Process command here int arg3 = processMyAction(arg1, arg2); response.setArg("arg3", arg3); return ErrorCode::Success; } return ErrorCode::InvalidAction;
-
virtual void search(const SearchFilter &filter) override
-
class ServiceControl : public UPnP::Service
- #include <ServiceControl.h>
Public Functions
-
inline DeviceControl &root()
Get the root device.
-
virtual bool sendRequest(HttpRequest *request) const override
Implemented in ServiceControl.
-
inline virtual Error handleAction(ActionRequest &req) override
An action request has been received.
- Parameters:
env – Contains the action request
- Return values:
ErrorCode – Implementation should place response into
envafter reading parameters, e.g.:This is usually handled by generated wrapper class templates.if(env.actionName() == "MyAction") { String arg1; int arg2; env.getArg("arg1", arg1); env.getArg("arg2", arg2); auto& response = env.createResponse(); // Process command here int arg3 = processMyAction(arg1, arg2); response.setArg("arg3", arg3); return ErrorCode::Success; } return ErrorCode::InvalidAction;
-
inline const Description &description()
Get service description.
-
struct Description
- #include <ServiceControl.h>
-
inline DeviceControl &root()
-
enum ErrorValues
Development Notes
These may be of some interest.
Device state management
It should be possible to add and remove devices dynamically.
For example, an application may contain a MODBUS controller which communicates with various slaves to switch lights, garden pumps, etc. We’ve also decided to create a UPnP device object for each MODBUS slave.
If we add a new slave to the network, then our application creates a new UPnP device object and passes it to e.g. ‘UPnP::addDevice()’, which does this:
Add the device to the stack
Queue a set of ‘ssdp:alive’ announcements
Notify the application via callback that the device has been added
If a slave is to be taken out of the network:
Remove the device from the stack so it won’t respond to any new requests
Queue a set of ‘ssdp:byebye’ announcements
When the final announcement has been sent, notify the application via callback that the device has been removed
When the application gets a ‘remove’ notification it can safely destroy the object, if appropriate, as UPnP has finished using it.
Device tree
Add
parentdevice to embedded devices so we can track back up the tree. Still not clear on whether a device needs to know about its direct parent, or whether a reference to the root device is adequate. A service certainly needs to know about its parent. So we need both root and parent. Root can be obtained by traversing back up the tree.
Icons
Add IconList support
Control Points
This involves sending search requests and responding to advertisements. Requesting descriptions also parsing XML. Using a SAX parser (as listed above) would allow template classes to be populated and passed to an application callback for handling. It would be the application’s responsibility to decide what information to keep and how to manage it.
UPnP version
Decide how to handle versioning. We should support previous versions but selectively enforce version 1.0, 1.1 or 2.0 as required. The version must be consistent throughout a device stack so we should add this as a configurable value for root devices. This then gets propagated into embedded object descriptions as well. We can have multiple root devices at different versions.
URL handling
Responses from VR900 are illuminating. Gateway is on port 1900, media server on 8200. There doesn’t seem to be any technical reason why multiple root devices can share the same port, but using separate ports does mean we can use separate HttpServer instances.
It would not make much sense for a URL to be somewhere completely different, but it doesn’t appear to be prohibited by the V1.0 spec. V2, however, deprecates URLBase and mandates that URLs are all relative to the SSDP response location (where the description file is served from).
Demonstration of discovery
Write a sample for Host which performs an ssdp:all discovery and writes all the description files into a directory tree structure.
Issue multicast search, register callback to receive responses For each response, store in a table (Vector)
- If not already in table, queue an HTTP request for the description file
- On receipt, write out the description file into a directory:
<ip>/<port>/<path>/
- Queue an HTTP request for the presentation page
Write to same location
Components required:
- Item table. Implement using Vector. Contains:
IP Address Port number
HTTP request queue. Only service one fetch at a time.
Schema
A separate UPnP Schema is used to manage UPnP device and service schema. It generates C++ classes and sample code directly from these, which you can then use in your application.
Generation of suitable schema can be done using the Device Scanner tool.
Controlling devices
The Basic ControlPoint sample shows how this is done.
- Registration
To control UPnP-enabled devices on your local network they must first be located.
In order for this to happen, the framework must be able to match the C++ class implementation against the schema.
The data structures are generated by the UPnP Schema, which creates a function
registerClassesfor each domain. For example,UPnP::schemas_upnp_org::registerClasses()registers all devices and services for the standard namespace.Alternatively a subset of classes may be registered by calling
UPnP::ControlPoint::registerClasses()directly.Here’s an example from the ControlPoint sample:
#include <Network/UPnP/schemas-upnp-org/ClassGroup.h> // Let's make things a little easier for ourselves using namespace UPnP::schemas_upnp_org::device; using namespace UPnP::schemas_upnp_org::service; void initUPnP() { UPnP::schemas_upnp_org::registerClasses(); }
- Discovery
This is done using a
UPnP::ControlPoint::beginSearch()method, which takes two parameters: the first identifies what you are looking for, the second is a callback which gets invoked when a match has been found. You’ll typically implement this callback using a lambda function.For example, let’s find all
MediaRenderer1devices:// Only one active search is permitted so be sure to cancel any existing ones first controlPoint.cancelSearch(); controlPoint.beginSearch(Delegate<bool(MediaRenderer1&)>([](auto& device) { // We can now do stuff with the located device // Return true to keep the device, false to destroy it return false; });
This method takes a template parameter which is the C++ class type defining the device you are searching for. The framework will fetch the description for each corresponding device and construct a
UPnP::DeviceControlobject with appropriate services and embedded devices.- Control
Your search callback function gets a reference to a located device. These devices are created on the heap and owned by the
UPnP::ControlPoint. If you want to keep the device, you should take a reference to it and returntruefrom the callback.To actually do anything useful typically requires use of a
UPnP::ServiceControlobject. You’ll usually get this by callingUPnP::DeviceControl::getService<UPnP::ObjectClass>()or one of the generated helper methods. Note that this returns a pointer, which will benullptrif the service isn’t available:auto render = device.getRenderingControl(); if(render != nullptr) { // ... }
Once you have a Service object, you can control it using action methods:
render->getVolume(0, RenderingControl1::Channel::fs_Master, [&device](auto response) { // Process response here });
Action methods take a list of zero or more input parameters, with the final argument for the response.
Note
The exact type of the response can be determined for you by the compiler. Here’s the explicit call:
render->getVolume(0, RenderingControl1::Channel::fs_Master, [&device](RenderingControl1::GetVolume::Response response response) {
// ...
});
OK, so handling the action method response. You can get the result values using methods of response,
but you must first check that the device did not return a fault:
Serial.println();
Serial.println(_F("render->getVolume(0, Master):"));
// Sample uses a `checkResponse` helper function
if(auto fault = response.fault()) {
fault.printTo(Serial);
} else {
Serial.print(device.friendlyName());
Serial.print(_F(": Current Volume = "));
Serial.println(response.getCurrentVolume());
}
Serial.println();
Implementing devices
The Basic UPnP sample contains a couple of examples of how to create your own hosted devices.
The TeaPot device is the simplest possible implementation, with no services.
The Wemo device is more elaborate and has two services.
Both of these are constructed using code generated from custom schema.
These are located in the project’s schema directory which is picked up automatically
when the UPnP Schema library is built.
The framework generates a class template for each device and service. For example, take a look in samples/Basic_UPnP/include/Wemo.h#L59-L91:
class BasicEventService : public service::basicevent1Template<BasicEventService>
{
public:
// Need access to constructors
using basicevent1Template::basicevent1Template;
// Override methods if you need to customise any fields
String getField(Field desc) const override
{
switch(desc) {
case Field::serviceId:
// You could also put this in the schema
return F("urn:Belkin:serviceId:basicevent1");
default:
return basicevent1Template::getField(desc);
}
}
// Access to our device implementation
Controllee& controllee()
{
return reinterpret_cast<Controllee&>(device());
}
/* Here are the action methods */
Error getBinaryState(GetBinaryState::Response response)
{
response.setBinaryState(controllee().getState());
return Error::Success;
}
Error setBinaryState(bool state, SetBinaryState::Response response)
{
controllee().setState(state);
return Error::Success;
}
};
This perhaps slightly strange construction uses CRTP to use static polymorphism and avoid virtual method tables. This allows the compiler to generate more efficient code.
UPnP Tools
Windows:
Linux:
Under Ubuntu Linux you can install gupnp-tools:
sudo apt install gupnp-toolsAnd then discover devices on the local network using the following command:
gupnp-universal-cp
You can also start a “software” smart bulb device and use it to test your control point application:
gupnp-network-light
References
Used by
DIscovery And Launch (DIAL) Library
Hue Emulator Library
UPnP Schema Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
USB
Sming library for TinyUSB.
Note that TinyUSB uses a large number of configuration variables and callback functions, so is compiled and linked directly with the application rather than as a separate library.
Device stack
The TinyUSB example applications generally consist of these three things:
- tusb_config.h
Contains definitions such as which classes are supported, how many endpoints (each), buffer sizes, etc.
- usb_descriptors.c
Provides descriptor/report structure definitions, interface and endpoint assignments and string definitions, with callback function implementations to provide these to the TinyUSB stack.
- main.c and other translation units
Implements class-specific callbacks.
Applications may choose to use the raw TinyUSB API; this is demonstrated in the HID Composite Device sample sample. A drawback of this approach is that the configuration files have to be composed manually, which requires some knowledge of TinyUSB and can also be error-prone.
Instead, the configuration can be described in a JSON file and the configuration generated by the library.
The format of this file is defined in schema.json.
For vscode, the .usbcfg file extension is association with this schema by make ide-vscode.
This enables use of auto-completion in the editor. See Using with MS Visual Studio Code.
This approach is demonstrated in the Basic Device sample.
In the application component.mk file, set USB_CONFIG to the name of the configuration file.
The configuration will be validated and configuration files generated in, for example, out/Rp2040/debug/USB.
Device classes
C++ Device implementations are provided for some of the standard interfaces. Use is demonstrated in the Basic Device sample.
- CDC
Communications Device Class.
USB::CDC::Device.TinyUSB implements only the ACM modes which apparently have issues with Windows. Appears in linux as
/dev/ttyACM0, etc.- DFU
Device Firmware Update.
USB::DFU::Device. Use the linuxdfu-utiltool to test.- ECM_RNDIS and NCM
https://en.wikipedia.org/wiki/Ethernet_over_USB
Not yet implemented.
USB::ECM_RNDIS::DeviceandUSB::NCM::Device.- HID
Human Interface Device.
USB::HID::Device.- MIDI
Musical Instrument Digital Interface (over USB).
USB::MIDI::Device.- MSC
Mass Storage Class.
USB::MSC::Device.- VENDOR
Devices are identifed by VID:PID and require appropriate host driver.
USB::VENDOR::Device. TinyUSB implements a simple read/write interface for this class. This is implemented like a serial port to allow asynchronous streaming, etc.
Host stack
See Basic Host for an example.
Note
At present, there is no host support for Esp32. Samples will build for Rp2040 only.
- HUB
When connected to a hub (or multiple hubs) this must be defined in the configuration.
- HID
- CDC
- MSC
Allows attachment of USB storage.
USB::MSC::HostDeviceSee Basic IFS for a real-world example.- VENDOR
Support access to custom devices.
USB::MSC::HostDevice. The sample contains a demonstration for connecting an original XBOX-360 joypad controller.
Configuration variables
- USB_DEBUG_LEVEL
default: 0 (disable)
Set to a value from 1-3 to enable debug output messages from the TinyUSB stack.
- USB_CONFIG
default: undefined
This identifies the name of the JSON USB configuration file for the application. This allows the TinyUSB configuration data to be generated rather than written manually.
API
-
namespace USB
Typedefs
-
using GetDeviceDescriptor = Delegate<const tusb_desc_device_t*(const tusb_desc_device_t &desc)>
Callback to support provision of dynamic device descriptors.
Application typically copies the source descriptor to a statically allocated buffer, then amends values as required.
- Param desc:
The statically configured device descriptor
- Param const:
tusb_desc_device_t* Application returns a pointer to a statically allocated descriptor to use
-
using GetDescriptorString = Delegate<const Descriptor*(uint8_t index)>
Application-provided callback to customise string responses.
Note
Returned descriptor MUST NOT be on the stack! Typically this is statically allocated.
- Param index:
String index to fetch (STRING_INDEX_xxxx)
- Retval const:
StringDescriptor* Pointer to persistent buffer containing descriptor. Return nullptr to use default value from string table.
Functions
-
void onGetDeviceDescriptor(GetDeviceDescriptor callback)
-
void onGetDescriptorSting(GetDescriptorString callback)
-
bool begin()
-
struct Descriptor
- #include <Descriptors.h>
Structure of a USB descriptor.
Subclassed by USB::StringDescriptor< max_chars >
Public Functions
Public Members
-
uint8_t length
Total size (in bytes) including this header.
-
uint8_t type
e.g. TUSB_DESC_STRING
-
union Type
- #include <Descriptors.h>
-
uint8_t length
-
struct DescriptorList
- #include <Descriptors.h>
Buffer containing list of descriptors.
Supports C++ iteration.
Subclassed by USB::HID::Report
-
class Iterator
- #include <Descriptors.h>
-
class Iterator
-
template<size_t max_chars>
struct StringDescriptor : public USB::Descriptor - #include <Descriptors.h>
Template for making a USB string descriptor.
-
class DeviceInterface
- #include <DeviceInterface.h>
Base class to support a USB device interface implementation.
Subclassed by USB::CDC::Device, USB::DFU::Device, USB::ECM_RNDIS::Device, USB::HID::Device, USB::MIDI::Device, USB::MSC::Device, USB::NCM::Device, USB::VENDOR::Device
Public Functions
-
inline DeviceInterface(uint8_t instance, const char *name)
Constructor.
- Parameters:
instance – TinyUSB instance or index (class-specific)
name – Declared name for this interface instance
-
inline DeviceInterface(uint8_t instance, const char *name)
-
class HostInterface
- #include <HostInterface.h>
Common base class to support Host USB access.
Subclassed by USB::CDC::HostDevice, USB::HID::HostDevice, USB::MSC::HostDevice, USB::VENDOR::HostDevice
Public Functions
-
inline void begin(const Instance &inst)
Descendant classes should override this method to peform initialisation.
-
inline virtual void end()
Called when device is disconnected. Override as required.
-
struct Instance
- #include <HostInterface.h>
Identifies a TinyUSB host interface.
-
inline void begin(const Instance &inst)
-
namespace CDC
Typedefs
-
using MountCallback = Delegate<HostDevice*(const HostInterface::Instance &inst)>
Application callback to notify connection of a new device.
- Param inst:
TinyUSB device instance
- Retval HostDevice*:
Application returns pointer to implementation, or nullptr to ignore this device
-
using UnmountCallback = Delegate<void(HostDevice &dev)>
Application callback to notify disconnection of a device.
- Param dev:
The device which has been disconnected
Functions
-
void onMount(MountCallback callback)
Application should call this method to receive device connection notifications.
- Parameters:
callback –
-
void onUnmount(UnmountCallback callback)
Application should call this method to receive device disconnection notifications.
- Parameters:
callback –
-
class Device : public USB::DeviceInterface, public USB::CDC::UsbSerial
- #include <Device.h>
Serial device implementation, in ACM mode.
Public Functions
-
inline virtual size_t setRxBufferSize(size_t size) override
Sets receiving buffer size.
- Parameters:
size – requested size
- Return values:
size_t – actual size
-
inline virtual size_t setTxBufferSize(size_t size) override
Sets transmit buffer size.
- Parameters:
size – requested size
- Return values:
size_t – actual size
-
inline virtual int available() override
Return the total length of the stream.
- Return values:
int – -1 is returned when the size cannot be determined
-
inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
-
inline virtual int read() override
Read one character and moves the stream pointer.
- Return values:
The – character that was read or -1 if none is available
-
inline virtual size_t readBytes(char *buffer, size_t length) override
Read chars from stream into buffer.
Terminates if length characters have been read or timeout (see setTimeout). Returns the number of characters placed in the buffer (0 means no valid data found).
Note
Inherited classes may provide more efficient implementations without timeout.
-
inline virtual int peek() override
Read a character without advancing the stream pointer.
- Return values:
int – The character that was read or -1 if none is available
-
inline virtual void clear(SerialMode mode = SERIAL_FULL) override
Clear the serial port transmit/receive buffers.
Note
All un-read buffered data is removed and any error condition cleared
- Parameters:
mode – Whether to flush TX, RX or both (the default)
-
virtual size_t write(const uint8_t *buffer, size_t size) override
Write chars to stream.
Note
Although this is defined in the Print class, ReadWriteStream uses this as the core output method so descendants are required to implement it
- Parameters:
buffer – Pointer to buffer to write to the stream
size – Quantity of chars to write
- Return values:
size_t – Quantity of chars written to stream
-
inline virtual size_t setRxBufferSize(size_t size) override
-
class HostDevice : public USB::HostInterface, public USB::CDC::UsbSerial
- #include <HostDevice.h>
Implements CDC interface for a connected serial device.
Public Functions
-
inline virtual size_t setRxBufferSize(size_t size) override
Sets receiving buffer size.
- Parameters:
size – requested size
- Return values:
size_t – actual size
-
inline virtual size_t setTxBufferSize(size_t size) override
Sets transmit buffer size.
- Parameters:
size – requested size
- Return values:
size_t – actual size
-
inline virtual int available() override
Return the total length of the stream.
- Return values:
int – -1 is returned when the size cannot be determined
-
inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
-
inline virtual int read() override
Read one character and moves the stream pointer.
- Return values:
The – character that was read or -1 if none is available
-
inline virtual size_t readBytes(char *buffer, size_t length) override
Read chars from stream into buffer.
Terminates if length characters have been read or timeout (see setTimeout). Returns the number of characters placed in the buffer (0 means no valid data found).
Note
Inherited classes may provide more efficient implementations without timeout.
-
inline virtual int peek() override
Read a character without advancing the stream pointer.
- Return values:
int – The character that was read or -1 if none is available
-
inline virtual void clear(SerialMode mode = SERIAL_FULL) override
Clear the serial port transmit/receive buffers.
Note
All un-read buffered data is removed and any error condition cleared
- Parameters:
mode – Whether to flush TX, RX or both (the default)
-
virtual size_t write(const uint8_t *buffer, size_t size) override
Write chars to stream.
Note
Although this is defined in the Print class, ReadWriteStream uses this as the core output method so descendants are required to implement it
- Parameters:
buffer – Pointer to buffer to write to the stream
size – Quantity of chars to write
- Return values:
size_t – Quantity of chars written to stream
-
inline virtual size_t setRxBufferSize(size_t size) override
-
class UsbSerial : public ReadWriteStream
- #include <UsbSerial.h>
Base class for both device and host serial port modes.
- Todo:
We could inherit from HardwareSerial here, or preferably provide an abstract base class for all serial devices.
Subclassed by USB::CDC::Device, USB::CDC::HostDevice, USB::VENDOR::Device
Public Functions
-
virtual size_t setRxBufferSize(size_t size) = 0
Sets receiving buffer size.
- Parameters:
size – requested size
- Return values:
size_t – actual size
-
virtual size_t setTxBufferSize(size_t size) = 0
Sets transmit buffer size.
- Parameters:
size – requested size
- Return values:
size_t – actual size
-
inline void setTxWait(bool wait)
Governs write behaviour when UART transmit buffers are full.
- Parameters:
wait – If false, writes will return short count; applications can use the txComplete callback to send more data. If true, writes will wait for more buffer space so that all requested data is written
-
inline virtual uint16_t readMemoryBlock(char *buf, int max_len) override
Read a block of memory.
- Todo:
Should IDataSourceStream::readMemoryBlock return same data type as its bufSize param?
- Parameters:
data – Pointer to the data to be read
bufSize – Quantity of chars to read
- Return values:
uint16_t – Quantity of chars read
-
inline virtual bool seek(int len) override
Move read cursor.
- Parameters:
len – Relative cursor adjustment
- Return values:
bool – True on success.
-
virtual void clear(SerialMode mode = SERIAL_FULL) = 0
Clear the serial port transmit/receive buffers.
Note
All un-read buffered data is removed and any error condition cleared
- Parameters:
mode – Whether to flush TX, RX or both (the default)
-
void systemDebugOutput(bool enabled)
Configure serial port for system debug output and redirect output from debugf.
Note
If enabled, port will issue system debug messages
- Parameters:
enabled – True to enable this port for system debug output
-
unsigned getStatus()
Get status error flags and clear them.
See also
SerialStatus
- Return values:
unsigned – Status flags, combination of SerialStatus bits
-
using MountCallback = Delegate<HostDevice*(const HostInterface::Instance &inst)>
-
namespace DFU
-
class Callbacks
- #include <Device.h>
Applications must implement this class and pass an instance to Device::begin().
Public Functions
-
virtual uint32_t getTimeout(Alternate alt, dfu_state_t state) = 0
Invoked right before tud_dfu_download_cb() (state=DFU_DNBUSY) or tud_dfu_manifest_cb() (state=DFU_MANIFEST)
Application return timeout in milliseconds (bwPollTimeout) for the next download/manifest operation. During this period, USB host won’t try to communicate with us.
-
virtual void download(Alternate alt, uint32_t offset, const void *data, uint16_t length) = 0
Invoked when received DFU_DNLOAD (wLength>0) following by DFU_GETSTATUS (state=DFU_DNBUSY) requests.
This callback could be returned before flashing op is complete (async). Once finished flashing, application must call
complete()
-
virtual void manifest(Alternate alt) = 0
Invoked when download process is complete, received DFU_DNLOAD (wLength=0) following by DFU_GETSTATUS (state=Manifest)
Application can do checksum, or actual flashing if buffered entire image previously. Once finished flashing, application must call
Device::finishFlashing()
-
virtual uint16_t upload(Alternate alt, uint32_t offset, void *data, uint16_t length) = 0
Invoked when received DFU_UPLOAD request Application must populate data with up to length bytes and return the number of written bytes.
-
virtual void abort(Alternate alt) = 0
Invoked when the Host has terminated a download or upload transfer.
-
virtual void detach() = 0
Invoked when a DFU_DETACH request is received.
-
virtual uint32_t getTimeout(Alternate alt, dfu_state_t state) = 0
-
class Device : public USB::DeviceInterface
- #include <Device.h>
Public Static Functions
-
static inline void complete(dfu_status_t status)
Applications call this method from download and manifest callbacks.
This mechanism supports use of asynchronous writes.
-
static inline void complete(dfu_status_t status)
-
class Callbacks
-
namespace ECM_RNDIS
-
class Device : public USB::DeviceInterface
- #include <Device.h>
Public Functions
-
inline DeviceInterface(uint8_t instance, const char *name)
Constructor.
- Parameters:
instance – TinyUSB instance or index (class-specific)
name – Declared name for this interface instance
-
inline DeviceInterface(uint8_t instance, const char *name)
-
class Device : public USB::DeviceInterface
-
namespace HID
Typedefs
-
using MountCallback = Delegate<HostDevice*(const HostInterface::Instance &inst, const Report &report)>
Application callback to notify connection of a new device.
- Param inst:
TinyUSB device instance
- Param report:
HID report descriptors for this interface
- Retval HostDevice*:
Application returns pointer to implementation, or nullptr to ignore this device
-
using UnmountCallback = Delegate<void(HostDevice &dev)>
Application callback to notify disconnection of a device.
- Param dev:
The device which has been disconnected
Functions
-
void onMount(MountCallback callback)
Application should call this method to receive device connection notifications.
- Parameters:
callback –
-
void onUnmount(UnmountCallback callback)
Application should call this method to receive device disconnection notifications.
- Parameters:
callback –
-
class Device : public USB::DeviceInterface
- #include <Device.h>
Public Functions
-
inline DeviceInterface(uint8_t instance, const char *name)
Constructor.
- Parameters:
instance – TinyUSB instance or index (class-specific)
name – Declared name for this interface instance
-
inline DeviceInterface(uint8_t instance, const char *name)
-
struct Report : public USB::DescriptorList
- #include <HostDevice.h>
-
class HostDevice : public USB::HostInterface
- #include <HostDevice.h>
-
using MountCallback = Delegate<HostDevice*(const HostInterface::Instance &inst, const Report &report)>
-
namespace MIDI
-
-
union Packet
- #include <Device.h>
-
class Device : public USB::DeviceInterface
- #include <Device.h>
-
union Packet
-
namespace MSC
Typedefs
-
using MountCallback = Delegate<HostDevice*(const HostInterface::Instance &inst)>
Application callback to notify connection of a new device.
- Param inst:
TinyUSB device instance
- Retval HostDevice*:
Application returns pointer to implementation, or nullptr to ignore this device
-
using UnmountCallback = Delegate<void(HostDevice &dev)>
Application callback to notify disconnection of a device.
- Param dev:
The device which has been disconnected
Functions
-
void onMount(MountCallback callback)
Application should call this method to receive device connection notifications.
- Parameters:
callback –
-
void onUnmount(UnmountCallback callback)
Application should call this method to receive device disconnection notifications.
- Parameters:
callback –
-
class LogicalUnit : public Storage::Disk::BlockDevice
- #include <Device.h>
A physical device instance managed by an MSC interface.
-
class Device : public USB::DeviceInterface
- #include <Device.h>
Public Functions
-
inline DeviceInterface(uint8_t instance, const char *name)
Constructor.
- Parameters:
instance – TinyUSB instance or index (class-specific)
name – Declared name for this interface instance
-
inline DeviceInterface(uint8_t instance, const char *name)
-
struct Inquiry
- #include <HostDevice.h>
Information provided by SCSI inquiry operation.
-
class HostDevice : public USB::HostInterface
- #include <HostDevice.h>
A USB mass storage device supports one or more logical units, each of which is a physical storage device.
Public Types
-
using EnumCallback = Delegate<bool(LogicalUnit &unit, const Inquiry &inquiry)>
Callback passed to enumerate() method.
- Param unit:
The logical unit attached to a device
- Param inquiry:
Detailed information
- Retval bool:
true to continue enumeration, false to stop
Public Functions
-
virtual void end()
Called when device is disconnected. Override as required.
-
bool enumerate(EnumCallback callback)
Enumerate all logical units managed by this device.
- Parameters:
callback – Invoked for each discovered Logical Unit
-
inline LogicalUnit *operator[](unsigned lun) const
Access a specific logical unit by number.
- Parameters:
lun – The logical Unit Number
- Return values:
LogicalUnit* – The corresponding LU, or nullptr if invalid
-
inline size_t getSectorSize(uint8_t lun) const
Get the declared block/sector size for a unit.
- Parameters:
lun – The logical Unit Number
- Return values:
size_t – Block size in bytes, or 0 if invalid
-
inline storage_size_t getSectorCount(uint8_t lun) const
Get the number of blocks/sectors for a unit.
- Parameters:
lun – The logical Unit Number
- Return values:
size_t – Number of blocks, 0 if invalid
-
bool read_sectors(uint8_t lun, uint32_t lba, void *dst, size_t size)
Read data from a unit.
- Parameters:
lun – The logical Unit Number
lba – Starting Logical Block Address
dst – Buffer to store data
size – Number of sectors to read
- Return values:
bool – true on success
-
bool write_sectors(uint8_t lun, uint32_t lba, const void *src, size_t size)
Write data to a unit.
- Parameters:
lun – The logical Unit Number
lba – Starting Logical Block Address
src – Data to write
size – Number of sectors to write
- Return values:
bool – true on success
-
bool wait()
Wait for all outstanding operations to complete.
Write operations are asynchronous so calling this method ensures that the operation has completed.
- Parameters:
lun – The logical Unit Number
- Return values:
bool – false on error (e.g. device forceably disconnected)
-
using EnumCallback = Delegate<bool(LogicalUnit &unit, const Inquiry &inquiry)>
-
using MountCallback = Delegate<HostDevice*(const HostInterface::Instance &inst)>
-
namespace NCM
-
class Device : public USB::DeviceInterface
- #include <Device.h>
Not currently implemented.
-
class Device : public USB::DeviceInterface
-
namespace VENDOR
Typedefs
-
using MountCallback = Delegate<HostDevice*(const HostInterface::Instance &inst, const HostDevice::Config &cfg)>
Application callback to notify connection of a new device.
- Param inst:
TinyUSB device instance
- Param cfg:
HostDevice configuration
- Retval HostDevice*:
Application returns pointer to implementation, or nullptr to ignore this device
-
using UnmountCallback = Delegate<void(HostDevice &dev)>
Application callback to notify disconnection of a device.
- Param dev:
The device which has been disconnected
Functions
-
void onMount(MountCallback callback)
Application should call this method to receive device connection notifications.
- Parameters:
callback –
-
void onUnmount(UnmountCallback callback)
Application should call this method to receive device disconnection notifications.
- Parameters:
callback –
-
class Device : public USB::DeviceInterface, public USB::CDC::UsbSerial
- #include <Device.h>
The TinyUSB vendor API is very much like a serial port. Each instance corresponds to a bi-directional interface.
Public Functions
-
inline virtual size_t setRxBufferSize(size_t size) override
Sets receiving buffer size.
- Parameters:
size – requested size
- Return values:
size_t – actual size
-
inline virtual size_t setTxBufferSize(size_t size) override
Sets transmit buffer size.
- Parameters:
size – requested size
- Return values:
size_t – actual size
-
inline virtual int available() override
Return the total length of the stream.
- Return values:
int – -1 is returned when the size cannot be determined
-
inline virtual bool isFinished() override
Check if all data has been read.
- Return values:
bool – True on success.
-
inline virtual int read() override
Read one character and moves the stream pointer.
- Return values:
The – character that was read or -1 if none is available
-
inline virtual size_t readBytes(char *buffer, size_t length) override
Read chars from stream into buffer.
Terminates if length characters have been read or timeout (see setTimeout). Returns the number of characters placed in the buffer (0 means no valid data found).
Note
Inherited classes may provide more efficient implementations without timeout.
-
inline virtual int peek() override
Read a character without advancing the stream pointer.
- Return values:
int – The character that was read or -1 if none is available
-
inline virtual void clear(SerialMode mode = SERIAL_FULL) override
Clear the serial port transmit/receive buffers.
Note
All un-read buffered data is removed and any error condition cleared
- Parameters:
mode – Whether to flush TX, RX or both (the default)
-
virtual size_t write(const uint8_t *buffer, size_t size) override
Write chars to stream.
Note
Although this is defined in the Print class, ReadWriteStream uses this as the core output method so descendants are required to implement it
- Parameters:
buffer – Pointer to buffer to write to the stream
size – Quantity of chars to write
- Return values:
size_t – Quantity of chars written to stream
-
inline virtual size_t setRxBufferSize(size_t size) override
-
class HostDevice : public USB::HostInterface
- #include <HostDevice.h>
Base class to use for custom devices.
Public Functions
-
inline virtual void end() override
Called when device is disconnected. Override as required.
-
virtual bool setConfig(uint8_t itf_num) = 0
Set active configuration.
Implementations typically start communicating.
-
struct Config
- #include <HostDevice.h>
Device configuration received during mount procedure.
Public Members
-
uint16_t vid
Vendor ID.
-
uint16_t pid
Product ID.
-
DescriptorList list
Interface descriptor list.
-
uint16_t vid
-
struct Transfer
- #include <HostDevice.h>
Structure passed to ‘transferComplete’ method.
-
inline virtual void end() override
-
using MountCallback = Delegate<HostDevice*(const HostInterface::Instance &inst, const HostDevice::Config &cfg)>
-
using GetDeviceDescriptor = Delegate<const tusb_desc_device_t*(const tusb_desc_device_t &desc)>
References
Used by
Basic Device Sample
Basic Host Sample
CDC MSC HID Host sample Sample
HID Composite Device sample Sample
Environment Variables
SoC support
esp32s2
esp32s3
host
rp2040
Submodule: `tinyusb <>`__
Ultrasonic
Tested on modules:
HC-SR04 - ranges: 2-400cm, power: 5v, levels: TTL, for work with 3.3v need voltage divider on ECHO pin
US-100 - power: 3.3v-5v, temp. compensation
References
Used by
HCSR04 Ultrasonic Transducer Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
VT100 Emulator
This is a vt100 emulator written for devices with under 4kb of ram (for example the ATMega microcontroller). The emulator also uses ili9340 display by default, but can be compiled to use another display with only a few changes to the source code.
Note that because of the small size of the available memory and the slow speed of the microcontroller, it is a little tricky to implement all the escape sequences that are supported in vt100.
Examples
see demo.cpp for code that tests the terminal
Note: since vt100 is 80x24 lines, and our terminal only supports 40x40 lines, you need to run “stty cols 40 rows 40” command to tell terminal programs that only 40 columns are available.
Compiling
Don’t be put off by cmake. The CMakeLists.txt file is provided for convenience only. You can basically just compile using:
avr-gcc -O3 -std=c99 -mmcu=atmega328p -DF_CPU=16000000UL -c ili9340.c uart.c vt100.c
avr-g++ -O3 -std=c++11 -mmcu=atmega328p -DF_CPU=16000000UL -o demo.elf demo.cpp ili9340.o uart.o vt100.o
Compatibility
The aim of this project is to be a vt100 compliant terminal. However, at the moment it can not really be called a vt100 terminal because it is far from complete in terms of supporting all of the vt100 terminal sequences. But I do support a few and doing normal shell stuff works fine. But some functions require some more hacking to implement with only 2kb of ram (such as cursor blink and character deletion, to name a few).
Nonetheless, I’m proud to have come this far with it.
The driver currently supports the following escape sequences:
/======================================================\
| VT100 Escape sequences supported by the this driver |
\======================================================/
- ESC is the escape character "\e" or "\x1b"
- (yes) - the driver supports the sequence
- (no) - the driver currently does not support the sequence
- (?) - do we really need this function? Or: what does this do exactly?
VT52 Compatible Mode
--------------------
- (yes) ESC A Cursor up with scroll
- (yes) ESC B Cursor down with scroll
- (yes) ESC C Cursor right
- (yes) ESC D Cursor left
- (?) ESC F Special graphics character set
- (?) ESC G Select ASCII character set
- (yes) ESC H Cursor to home (but margins currently not supported!)
- (?) ESC I Reverse line feed
- (yes) ESC J Erase to end of screen
- (yes) ESC K Erase to end of line
- (no) ESC Ylc Direct cursor address (See note 1)
- (no) ESC Z Identify (See note 2)
- (?) ESC = Enter alternate keypad mode
- (?) ESC > Exit alternate keypad mode
- (?) ESC 1 Graphics processor on (See note 3)
- (?) ESC 2 Graphics processor off (See note 3)
- (?) ESC < Enter ANSI mode
ANSI Compatible Mode
--------------------
- (yes) ESC [ Pn A Cursor up Pn lines (without scroll)
- (yes) ESC [ Pn B Cursor down Pn lines
- (yes) ESC [ Pn C Cursor forward Pn characters (right)
- (yes) ESC [ Pn D Cursor backward Pn characters (left)
- (yes) ESC [ Pl;PcH Direct cursor addressing, where Pl is line#, Pc is column#
- NOTE: proper action is to move relative to margins. We don't support margins yet so the moves are absolute screen based character coordinates.
- (yes) ESC [ Pl;Pcf Same as above
- (yes) ESC D Index - moves cursor one line down and performs scroll if at bottom
- (yes) ESC M Reverse index - moves cursor up (with eventual scroll)
- (yes) ESC 7 Save cursor and attributes
- (yes) ESC 8 Restore cursor and attributes
- (?) ESC #3 Change this line to double-height top half
- (?) ESC #4 Change this line to double-height bottom half
- (?) ESC #5 Change this line to single-width single-height
- (?) ESC #6 Change this line to double-width single-height
- (yes) ESC [ Ps..Ps m
Ps refers to selective parameter. Multiple parameters are
Text attributes
0 All attributes off
1 Bold on
4 Underscore (on monochrome display adapter only)
5 Blink on
7 Reverse video on
8 Concealed on
Foreground colors
30 Black
31 Red
32 Green
33 Yellow
34 Blue
35 Magenta
36 Cyan
37 White
Background colors
40 Black
41 Red
42 Green
43 Yellow
44 Blue
45 Magenta
46 Cyan
47 White
- (yes) ESC [ K Erase from cursor to end of line
- (yes) ESC [ 0K Same
- (yes) ESC [ 1K Erase from beginning of line to cursor
- (yes) ESC [ 2K Erase line containing cursor
- (yes) ESC [ J Erase from cursor to end of screen
- (yes) ESC [ 0J Same
- (yes) ESC [ 2J Erase entire screen
- (?) ESC [ Ps..Ps q Programmable LEDs: Ps are selective parameters separated by
semicolons (073 octal) and executed in order, as follows:
0 or None All LEDs off
1 L1 On
2 L2 On
3 L3 On
4 L4 On
G0 designator G1 designator Character set
- (?) ESC ( A ESC ) A United Kingdom (UK)
- (?) ESC ( B ESC ) B United States (USASCII)
- (?) ESC ( 0 ESC ) 0 Special graphics/line drawing set
- (?) ESC ( 1 ESC ) 1 Alternative character ROM
- (?) ESC ( 2 ESC ) 2 Alternative graphic ROM
- (?) ESC K Pt;Pb r Set top scrolling window (Pt) and bottom scrolling window
(Pb). Pb must be greater than Pb.
- (no) ESC H Set tab at current column
- (no) ESC [ g Clear tab at current column
- (no) ESC [ 0g Same
- (no) ESC [ 3g Clear all tabs
Modes
-----
The following commands are currently not supported.
Mode Name Mode To set seq Mode To reset seq
- (no) Line feed/new New line ESC [20h Line feed ESC [20l
- (no) Cursor key Application ESC [?1h Cursor ESC [?1l
- (no) ANSI/VT52 ANSI n/a VT52 ESC [?2l
- (no) Column mode 132 col ESC [?3h 80 col ESC [?3l
- (no) Scrolling Smooth ESC [?4h Jump ESC [?4l
- (no) Screen mode Reverse ESC [?5h Normal ESC [?5l
- (yes) Origin mode Relative ESC [?6h Absolute ESC [?6l
- (yes) Wraparound On ESC [?7h Off ESC [?7l
- (?) Autorepeat On ESC [?8h Off ESC [?8l
- (?) Interface On ESC [?9h Off ESC [?9l
Reports
-------
- (no) ESC [ 6n Cursor position report
- (no) ESC [ Pl;PcR (response; Pl=line#; Pc=column#)
- (no) ESC [ 5n Status report
- (no) ESC [ c (response; terminal Ok)
- (no) ESC [ 0c (response; teminal not Ok)
- (no) ESC [ c What are you?
- (no) ESC [ 0c Same
- (no) ESC [?1;Ps c response; where Ps is option present:
0 Base VT100, no options
1 Preprocessor option (STP)
2 Advanced video option (AVO)
3 AVO and STP
4 Graphics processor option (GO)
5 GO and STP
6 GO and AVO
7 GO, STP, and AVO
- (no) ESC c Causes power-up reset routine to be executed
- (no) ESC #8 Fill screen with "E"
- (no) ESC [ 2;Ps y Invoke Test(s), where Ps is a decimal computed by adding the
numbers of the desired tests to be executed:
1 Power up test
2 Data loop back
4 EIA modem control signal test
8 Repeat test(s) indefinitely
TERMINAL COMMANDS
----------------
- (yes) ESC c Reset
- (no) [ ! p Soft Reset
- (yes) # 8 Fill Screen with E's
- (no) } 1 * Fill screen with * test
- (no) } 2 Video attribute test display
- (no) } 3 Character sets display test
KEYBOARD COMMANDS
-----------------
- (?) [ 2 h Keyboard locked
- (?) [ 2 l Keyboard unlocked
- (?) [ ? 8 h Autorepeat ON
- (?) [ ? 8 l Autorepeat OFF
- (?) [ 0 q Lights all off on keyboard
- (?) [ * q Light * on
PROGRAMMABLE KEY COMMANDS
-------------------------
- (?) ! pk Program a programmable key (local)
- (?) @ pk Program a programmable key (on-line)
- (?) % pk Transmit programmable key contents
SCREEN FORMAT (not supported)
-------------
- (?) [ ? 3h 132 Characters on
- (?) [ ? 3l 80 Characters on
- (?) [ ? 4h Smooth Scroll on
- (?) [ ? 4l Jump Scroll on
- (?) [ *t ; *b r Scrolling region selected, line *t to *b
- (?) [ ? 5 h Inverse video on
- (?) [ ? 5 l Normal video off
- (?) [ ? 7 h Wraparound ON
- (?) [ ? 7 l Wraparound OFF
- (?) [ ? 75 h Screen display ON
- (?) [ ? 75 l Screen display OFF
CHARACTER SETS AND LABELS
-------------------------
- (?) ( A British
- (?) ( B North American ASCII set
- (?) ( C Finnish
- (?) ( E Danish or Norwegian
- (?) ( H Swedish
- (?) ( K German
- (?) ( Q French Canadian
- (?) ( R Flemish or French/Belgian
- (?) ( Y Italian
- (?) ( Z Spanish
- (?) ( 0 Line Drawing
- (?) ( 1 Alternative Character
- (?) ( 2 Alternative Line drawing
- (?) ( 4 Dutch
- (?) ( 5 Finnish
- (?) ( 6 Danish or Norwegian
- (?) ( 7 Swedish
- (?) ( = Swiss (French or German)
[Note all ( may be replaced with )]
ATTRIBUTES AND FIELDS
-------
- (?) [ 0 m Clear all character attributes
- (?) [ 1 m Alternate Intensity ON
- (?) [ 4 m Underline ON
- (?) [ 5 m Blink ON
- (?) [ 7 m Inverse video ON
- (?) [ 22 m Alternate Intensity OFF
- (?) [ 24 m Underline OFF
- (?) [ 25 m Blink OFF
- (?) [ 27 m Inverse Video OFF
- (?) [ 0 } Protected fields OFF
- (?) [ 1 } Protected = Alternate Intensity
- (?) [ 4 } Protected = Underline
- (?) [ 5 } Protected = Blinking
- (?) [ 7 } Protected = Inverse
- (?) [ 254 } Protected = All attributes OFF
CURSOR COMMANDS
-------
- (?) [ ? 25 l Cursor OFF
- (?) [ ? 25 h Cursor ON
- (?) [ ? 50 l Cursor OFF
- (?) [ ? 50 h Cursor ON
- (yes) 7 Save cursor position and character attributes
- (yes) 8 Restore cursor position and character attributes
- (yes) D Line feed
- (yes) E Carriage return and line feed
- (yes) M Reverse Line feed
- (yes) [ A Cursor up one line
- (yes) [ B Cursor down one line
- (yes) [ C Cursor right one column
- (yes) [ D Cursor left one column
- (yes) [ * A Cursor up * lines
- (yes) [ * B Cursor down * lines
- (yes) [ * C Cursor right * columns
- (yes) [ * D Cursor left * columns
- (yes) [ H Cursor home
- (yes) [ *l ; *c H Move cursor to line *l, column *c
- (yes) [ *l ; *c f Move curosr to line *l, column *c
- (no) Y nl nc Direct cursor addressing (line/column number)
- (no) H Tab set at present cursor position
- (no) [ 0 g Clear tab at present cursor position
- (no) [ 3 g Clear all tabs
EDIT COMMANDS
-------
- (no) [ 4 h Insert mode selected
- (no) [ 4 l Replacement mode selected
- (no) [ ? 14 h Immediate operation of ENTER key
- (no) [ ? 14 l Deferred operation of ENTER key
- (no) [ ? 16 h Edit selection immediate
- (no) [ ? 16 l Edit selection deferred
- (no) [ P Delete character from cursor position
- (no) [ * P Delete * chars from curosr right
- (no) [ M Delete 1 char from cursor position
- (no) [ * M Delete * lines from cursor line down
- (yes) [ J Erase screen from cursor to end
- (yes) [ 1 J Erase beginning of screen to cursor
- (yes) [ 2 J Erase entire screen but do not move cursor
- (yes) [ K Erase line from cursor to end
- (yes) [ 1 K Erase from beginning of line to cursor
- (yes) [ 2 K Erase entire line but do not move cursor
- (no) [ L Insert 1 line from cursor position
- (no) [ * L Insert * lines from cursor position
LICENSE
-------
- Multilicense: GNU GPL free for all, by default.
For anything else, drop me an email. :)
.............................................................................
: INTERNET: info@fortmax.se :: Tel: 0733-387694 Int: +46-733-387694 :
: AUTHOR: Martin K. Schröder COPYRIGHT: 2014 SWEDEN :
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
References
Used by
Screen TFT_S1D13781 Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
WS2812 Neopixel
http://wp.josh.com/2014/05/13/ws2812-neopixels-are-not-so-finicky-once-you-get-to-know-them/
References
Used by
WS2812 LEDs Sample
SoC support
esp8266
WebCam
Introduction
This component provides a webcamera(aka webcam) stream and camera drivers that can help you stream your webcam images as video stream through the embedded web server.
References
Used by
WebcamServer sample Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Yeelight
Library to control Yeelight devices via JSON
See https://www.yeelight.com/en_US/activity
Specification: https://www.yeelight.com/download/Yeelight_Inter-Operation_Spec.pdf
References
Used by
Yeelight LED bulbs Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
FAT Filing System
Required by the SDCard library.
http://elm-chan.org/fsw/ff/00index_e.html
References
Used by
SD Card Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Flatbuffers
FlatBuffers is an efficient cross platform serialization library for C++, C#, C, Go, Java, Kotlin, JavaScript, Lobster, Lua, TypeScript, PHP, Python, Rust and Swift. It was originally created at Google for game development and other performance-critical applications.
It is available as Open Source on GitHub under the Apache license, v2 (see LICENSE.txt).
Using
Step 1. Add these lines to your application componenent.mk file:
COMPONENT_DEPENDS += flatbuffers
Step 2. Add these lines to your application:
#include <flatbuffers/flatbuffers>
Or directly use the header file generated from the flatc compiler.
Step 3. Basic example:
#include "monster.h"
using namespace MyGame::Example;
void init()
{
Serial.begin(SERIAL_BAUD_RATE);
// demonstration how to encode data into a flatbuffer
flatbuffers::FlatBufferBuilder builder;
auto name = builder.CreateString("Sming Monster");
MonsterBuilder monster_builder(builder);
monster_builder.add_name(name);
auto orc = monster_builder.Finish();
// Call `Finish()` to instruct the builder that this monster is complete.
// Note: Regardless of how you created the `orc`, you still need to call
// `Finish()` on the `FlatBufferBuilder`.
builder.Finish(orc);
// and then decode it
uint8_t *buffer = builder.GetBufferPointer();
auto monster = GetMonster(buffer);
Serial.printf("Monster name: %s\n", monster->name()->c_str());
}
Further reading
Take a look at the official flatbuffers tutorial.
References
Used by
Basic FlatBuffers Sample Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: src
JerryScript
A JavaScript Virtual Machine based on JerryScript.
Introduction
This library allows running JavaScript in a sandbox on all architectures supported by Sming. The library uses JerryScript as JavaScript Virtual Machine (VM).
The diagram above shows the interactions between the major components of JerryScript: Parser and VM. Parser performs translation of input ECMAScript application into the byte-code. Prepared bytecode is executed by the Virtual Machine that performs interpretation. Source: Official JerryScript site.
To save space and be able to run JerryScript on an embedded device Sming compiles this library without a parser. This means that the JavaScript files have to be compiled before landing on the device. See the samples below to learn more.
Console messages
Two external functions are made available for scripts to use:
printfor normal console output (shown in AQUA)alertfor messages in debug mode (show in RED)
VM integrity
The Jerryscript VM may encounter situations where it cannot continue execution, such as memory allocation failure or invalid bytecode.
Fatal errors cause engine execution to halt. It must be “re-booted” (in a virtual sense). This requires application co-operation as scripts may need to be re-loaded, external functions re-registered, etc.
If intending to run unverified javascript it is especially important to guard all calls into Jerryscript using the library exception handler:
JS_TRY()
{
// Call into jerryscript engine
}
JS_CATCH()
{
// Engine has failed
}
- JS_TRY:
Place code here which sets up jerryscript values, etc. then calls into Jerryscript. Data is allocated on local stack or in jerryscript engine.
DO NOT perform any system heap allocation here, such as creating or modifying String) objects. If required, instantiate those above.
- JS_CATCH:
Jerryscript ust be re-initialised before any further calls.
De-allocate any globally held jerryscript values
Call JS::cleanup()
Call JS::init() and any other necessary initialisation
It is recommended to provide a single function for both initial jerryscript initialisation and re-initialisation.
See the Event_Jsvm sample for a demonstratation.
Note that the VM uses own static heap (sized by JERRY_GLOBAL_HEAP_SIZE) so main system heap is unaffected.
When JERRY_ENABLE_DEBUG is enabled it may not be possible to recover because VM objects may be left with
invalid reference counts, for example, which will cause Jerryscript::cleanup() to fail.
Applications should generally be built with this setting disabled (the default).
Watchdog
Sming is a single-threaded framework so requires co-operation to perform multi-tasking. This means javascript calls should be brief so as not to disrupt the system.
If a javascript call takes too long then the system watchdog will timeout and reset the system. This is generally not helpful, so a separate watchdog is implemented for the Jerryscript virtual machine.
By default, the watchdog is disabled. To enable it, add a call during initialisation:
// Enable the watchdog with a 100ms timeout (0 to disable it)
JS::Watchdog::setPeriod(100);
Now, if a call takes more than 100ms then a fatal error will be thrown as discussed above.
The setting can be changed at any time and is unaffected by jerryscript engine resets.
External Contexts
By default, a single, global VM context is created in which all code is loaded and executed. If multiple javascript applications are running then if one fails it will take out all the others.
Enabling the JERRY_EXTERNAL_CONTEXT setting allows use of the JS::Context
class to run multiple isolated instances. See the Basic Context example for details.
Configuration variables
- JERRY_ENABLE_DEBUG
default: 0 (disabled)
Enable assertions and debug messages in jerryscript library. Should be left disabled unless debugging the jerryscript engine itself.
- JERRY_MEM_STATS
default: 1 (enabled)
Enable jerryscript heap memory tracking.
- JERRY_ERROR_MESSAGES
default: 0 (disabled)
Enable text messages for common errors. Consumes ~2K RAM on ESP8266.
- JERRY_COMPACT_PROFILE
default: 1 (enabled)
Compact (minimal profile) compilation profile makes the JerryScript library smaller.
Set to 0 to use es.next profile.
- JERRY_PARSER
default: 0 (disabled)
Enable to build library with javascript parser enabled. Required for use of
Jerryscript::eval()function.
- JERRY_EXTERNAL_CONTEXT
default: 0 (disabled)
Enable this setting to make use of
JS::Contextto run multiple isolated instances.
- JERRY_GLOBAL_HEAP_SIZE
default: 1 (in KB)
Size of the jerryscript global heap in kilobytes. Increase as required.
- JERRY_SNAPSHOT_TOOL
Read-only. Path to the snapshot compiler for use by applications.
- APP_JS_SOURCE_DIR
default: undefined
Snap files can be created during the build stage by setting this variable in your project’s component.mk file:
APP_JS_SOURCE_DIR := files
All
.jsfiles will be compiled into.snapfiles and written toAPP_JS_SNAPDIR.
- APP_JS_SNAP_DIR
default:
out/jerryscriptLocation to write generated .snap files.
- APP_JS_SNAP_UPDATED
default: undefined
Set this if action required when snaps are updated, for example to rebuild SPIFFS image or update application.
- JERRY_WEB_COMPILER
(read-only)
Location of pre-compiled web compiler. See
Advanced-Jsvmsample for example usage.
Credits
The initial work on the JerryScript library for Sming was done as part of the U:Kit project.
Updating JerryScript
Currently we use Jerryscript version v2.4 from JerryScript, imported as a submodule.
The web-assembly (WASM) snapshot compiler uses the emscripten SDK.
Update
In order to update the version of JerryScript two tasks have to be executed:
In this repository update the submodule version of JerryScript to point to the new version
Rebuild the snapshot compiler
The snapshot compiler can be rebuilt manually as follows:
Install the emscripten SDK:
git clone --depth 1 https://github.com/emscripten-core/emsdk.git emsdk emsdk/emsdk install latest emsdk/emsdk activate latest
run
makefrom theemscripten-snapshot-compilerdirectory.
The following three files are generated in emscripten-snapshot-compiler/build/bin/{PROFILE}:
jsc.js
jsc.wasm
jsc.wasm.map
Copy these to the jsc sub-directory for application use as demonstrated in the Advanced_Jsvm sample.
Dockerfile
The snapshot compiler can be rebuilt using a Docker container which also includes a WebAssembly toolkit. It can be built and run with the following commands:
docker build -f Dockerfile -t jerryscript-ems .
docker run -it jerryscript-ems
Inside the container a new jerryscript snapshot compiler can be built by calling:
node /jerryscript/build/bin/minimal/jsc.js -o /tmp/test.js.snap /tmp/test.js
Where test.js.snap is the output file and /tmp/test.js is an existing JavaScript file.
To update the jsc.* files you run the following from the application directory:
docker run -v $(pwd)/files/web/:/web -it jerryscript-ems
And run the following commands inside the container:
cp -r /jerryscript/build/bin/minimal/jsc.* /web
The build system handles compilation of the main.js file,
but if you prefer to use pre-compiled snaps this can also be done inside the container:
node /jerryscript/build/bin/minimal/jsc.js -o /web/main.js.snap /web/main.js
API Documentation
Classes and types
-
using Jerryscript::HeapStats = jerry_heap_stats_t
-
size_t Jerryscript::getHeapUsed()
-
bool Jerryscript::printHeap()
-
jerry_value_t Jerryscript::create_arg_count_error(const char *functionName)
-
inline Object Jerryscript::global()
Get global context.
-
void Jerryscript::initialise(jerry_init_flag_t flags = JERRY_INIT_EMPTY)
Initializes the JavaScript VM.
-
void Jerryscript::cleanup()
Clean up the virtual machine by unloading snapshots, freeing allocated memory, etc.
Note that this does not release Jerryscript heap memory allocated to existing values. This is done via
Valuedestructor or by manually callingValue::reset().
-
inline bool Jerryscript::isFeatureEnabled(Feature feature)
Check if optional feature is available.
- Parameters:
feature –
- Return values:
bool – true if library has been compiled with requested feature
-
inline void Jerryscript::gc(bool maximumEffort = false)
Perform memory garbage collection.
-
class Context
- #include <Context.h>
Jerryscript external context.
Applications may be segregated by running in separate contexts. Each context has its own dynamically allocated heap.
Subclassed by Jerryscript::ContextTemplate< ClassType >
-
template<class ClassType>
class ContextTemplate : public LinkedObjectTemplate<ClassType>, public Jerryscript::Context - #include <Context.h>
Implement a custom Context class.
-
template<class ClassType>
class ContextList : public OwnedLinkedObjectListTemplate<ClassType> - #include <Context.h>
Manages a list of contexts.
Public Functions
-
inline void foreach(Callback callback)
Invoke callback once for each context via task queue.
When calling into contexts we can do this:
JS::ContextList<MyContext> contexts; ... String str = F("Some text"); int value = 12; for(auto& ctx: contexts) { ctx.customNotify(str, value); }
However, with many containers system responsiveness may be adversely affected. Instead, separate the calls out like this:
contexts.foreach([=](auto& ctx){ ctx.customNotify(str, value); });
Note that all parameters are captured by copy (using
[=]) as, for example,strwill be destroyed when it goes out of scope.
-
inline void foreach(Callback callback)
-
class Except
- #include <Except.h>
-
struct CallInfo
- #include <Function.h>
Maps directly onto jerry_call_info_t structure.
-
class Task : public Task
- #include <Task.h>
Task that runs the
loopJavaScript function in the background
Public Functions
-
inline virtual void loop() override
Inherited classes override this to perform actual work.
-
struct OwnedValue
- #include <Types.h>
Use to initialise Value object by taking ownership of native/raw jerryscript value.
e.g.
Value myValue = OwnedValue{raw_value};This typecast is necessary because
jerry_value_tis weakly typed (as uint32_t).
-
struct CopyValue
- #include <Types.h>
Use to initialise Value object by copying native/raw jerryscript value.
-
struct StringValue
- #include <Types.h>
Use to initialise Value object to a string given a native/raw jerryscript value.
e.g.
Value myValue = StringValue{raw_value};
-
class Value
- #include <Types.h>
Represents a Jerryscript value.
This class resembles
std::unique_ptrto manage a raw/nativejerry_value_tvalue. Theget(),reset()andrelease()methods all behave in the same manner.In addition, it provides intrinsic casts from numeric/boolean C++ types plus
String.You can therefore create a value like this::
Value value = 12; Value str = "My String"; Value str2 = F("My other String");
If you need to working with floating point values be sure to compile with
JERRY_COMPACT_PROFILE=0.IMPORTANT: When dealing with raw/native values, ALWAYS use either
OwnedValueorCopyValuecasts.This is correct: Value value = OwnedValue{jerry_create_object()};
This is wrong: Value value = jerry_create_object(); // Ends up with memory leak plus garbage
Subclassed by Jerryscript::Error, Jerryscript::ExternalFunction, Jerryscript::Object
Construct a value from a simple type
-
Value(int value)
Integer.
-
Value(unsigned value)
Unsigned integer.
-
inline Value(double value)
floating-point
-
inline Value(bool value)
Boolean.
-
inline Value(const char *s)
NUL-terminated ‘C’ string.
Get raw/native value for use with jerryscript C API
Checks on value type
-
inline bool isArray() const
Can this object be accessed as an array? If so, can cast to
Arraytype.
-
inline bool isEmpty() const
An empty Value contains nothing, i.e. no javascript type has been assigned. This gets interpreted as ‘undefined’ if any attempt is made to use it in javascript.
-
inline bool isDefined() const
Contains a javascript value, but contents undefined.
-
inline bool isBoolean() const
A true/false value type.
-
inline bool isFalse() const
Is this a Boolean type set to False?
-
inline bool isTrue() const
Is this a Boolean type set to True?
-
inline bool isNull() const
Is this a NULL value?
-
inline bool isNumber() const
Does this value contain a Number?
Public Functions
-
inline Value()
Construct An empty (unused) value.
-
inline Value(const OwnedValue &value)
Construct a Value and take ownership of the given native value.
-
inline Value(const CopyValue &value)
Construct a Value using a copy (or reference to) the given native value.
-
inline Value(const StringValue &value)
Construct a string Value from the given native value.
-
inline Value &reset(jerry_value_t value = jerry_value_t(Ecma::VALUE_EMPTY))
Reset contents of object to new value (default is unassigned)
-
inline jerry_value_t release()
Get raw/native value and release ownership.
-
size_t readString(unsigned offset, char *buffer, size_t length) const
Get content from within a string value.
- Parameters:
offset – First character position to read, starting from 0
buffer – Where to store character data
length – Number of characters to read
- Return values:
size_t – Number of characters read
-
class ExternalFunction : public Jerryscript::Value
- #include <Types.h>
Object representing an external function implementation.
-
class Object : public Jerryscript::Value
- #include <Types.h>
Objects support named properties.
Subclassed by Jerryscript::Array, Jerryscript::Callable
Access object properties by name
-
inline NamedItem operator[](const String &name)
operator[] uses
NamedItemproxy object so value can be assigned or read
-
inline Value getProperty(const Value &name) const
Get a property value.
- Parameters:
name – Property name
- Return values:
Value – The property value, or Error
Public Functions
-
inline Object()
Default constructor creates a new, empty object.
-
Value runFunction(const String &name, Value &arg)
Call a specified JavaScript function with exactly one argument.
- Parameters:
name – Name of function to run (a property name)
arg – The argument
- Return values:
Value – Return value from function, or Error
-
Value runFunction(const String &name, std::initializer_list<Value> args = {})
Call a specified JavaScript function with zero or more arguments.
- Parameters:
name – Name of function to run (a property name)
- Return values:
Value – Return value from function, or Error
-
inline bool registerFunction(const String &name, jerry_external_handler_t handler)
Register an external function so it may be called from javascript.
- Parameters:
name – Name of the function (property name)
handler – The function handler, see Function.h for details
bool – true on success
-
inline bool unregisterFunction(const String &name)
Unregister an external function.
- Parameters:
name – Name of the function
bool – true on success
-
Callable getFunction(const String &name)
Retrieve the given property as a function.
- Return values:
The – callable object, or error (property not found or isn’t callable)
-
inline Value()
Construct An empty (unused) value.
-
inline Value(const OwnedValue &value)
Construct a Value and take ownership of the given native value.
-
inline Value(const CopyValue &value)
Construct a Value using a copy (or reference to) the given native value.
-
inline Value(const StringValue &value)
Construct a string Value from the given native value.
-
Value(int value)
Integer.
-
Value(unsigned value)
Unsigned integer.
-
inline Value(double value)
floating-point
-
inline Value(bool value)
Boolean.
-
inline Value(const char *s)
NUL-terminated ‘C’ string.
-
struct NamedItem
- #include <Types.h>
Iterator and operator[] access uses this wrapper class so items may be written or read.
-
inline NamedItem operator[](const String &name)
-
class Error : public Jerryscript::Value
- #include <Types.h>
Error object class.
Subclassed by Jerryscript::ArgumentError
Create an error object
Public Functions
-
operator String() const
Formulate error message Base operator in Value class returns empty value for errors. To obtain error message, check for
isError()then cast to Error():if (value.isError()) { Serial.println(JS::Error(value)); }
-
inline Value()
Construct An empty (unused) value.
-
inline Value(const OwnedValue &value)
Construct a Value and take ownership of the given native value.
-
inline Value(const CopyValue &value)
Construct a Value using a copy (or reference to) the given native value.
-
inline Value(const StringValue &value)
Construct a string Value from the given native value.
-
Value(int value)
Integer.
-
Value(unsigned value)
Unsigned integer.
-
inline Value(double value)
floating-point
-
inline Value(bool value)
Boolean.
-
inline Value(const char *s)
NUL-terminated ‘C’ string.
-
operator String() const
-
class ArgumentError : public Jerryscript::Error
- #include <Types.h>
Provides consistent error message when checking external function arguments.
-
class Array : public Jerryscript::Object
- #include <Types.h>
Array objects have properties accessed by index.
Array element/property access by index
-
inline IndexedItem operator[](unsigned index)
operator[] uses
IndexedItemproxy object so value can be assigned or read
Public Functions
-
inline Array(size_t size)
Create a new, fixed-size array with the given number of elements.
-
inline size_t count() const
Get number of elements in the array.
-
inline Object()
Default constructor creates a new, empty object.
-
struct IndexedItem
- #include <Types.h>
Iterator and operator[] access uses this wrapper class so items may be written or read.
-
class Iterator
- #include <Types.h>
-
inline IndexedItem operator[](unsigned index)
-
class Callable : public Jerryscript::Object
- #include <Types.h>
Callable object represent functions.
Public Functions
-
Value call(const Object &thisValue, std::initializer_list<Value> args = {})
Call with zero or multiple arguments.
e.g.
call(myObject, {1, 2, 3});
-
inline FunctionType functionType() const
Get specific type of callable object.
-
inline Object()
Default constructor creates a new, empty object.
-
Value call(const Object &thisValue, std::initializer_list<Value> args = {})
Macros
- group jerryscript
Macros to implement methods callable from javascript within a Context class
Must be used inside a class constructed using
JS::ContextTemplate.As with all external functions, must be registered using
JS::Object::registerFunctionto make available to javascript.-
JS_DEFINE_METHOD(method, ...)
Argument list is fixed.
Example:
class MyContext: public JS::ContextTemplate<MyContext> { public: void init() { select(); JS::initialise(); JS::global().registerFunction(F("myMethod"), myMethod); } protected: JS_DEFINE_METHOD(myMethod, 2, JS::Value& arg1, JS::Callable& arg2) { ... return ...; } };
- Parameters:
method – Name of jerryscript wrapper function
... – Argument definitions
-
JS_DEFINE_METHOD_VAR(method)
Arguments are passed as array.
Method declaration has the following prototype:
JS::Value js_method(const JS::CallInfo& callInfo, JS::Value args[], unsigned argCount)
Example:
class MyContext: public JS::ContextTemplate<MyContext> { public: void init() { select(); JS::initialise(); JS::global().registerFunction(F("myMethod"), myMethod); } protected: JS_DEFINE_METHOD_VAR(myMethod) { for(unsigned i = 0; i < argCount; ++i) { Serial.print(args[i]); } return Value(); } };
- Parameters:
method – Name of jerryscript wrapper function
Macros to implement functions callable from javascript
As with all external functions, must be registered using
JS::Object::registerFunctionto make available to javascript.Function arguments should be validated, returning
JS::Errorobject on failure.-
JS_DEFINE_FUNCTION(func, ...)
Argument list is fixed.
Example:
JS_DEFINE_FUNCTION(myFunction, JS::Value& arg1, JS::Array& arg2) { ... return ...; }
- Parameters:
func – Name of jerryscript wrapper function
... – Argument definitions
-
JS_DEFINE_FUNCTION_VAR(func)
Arguments are passed as array.
Example:
JS_DEFINE_FUNCTION_VAR(myFunction) { for(unsigned i = 0; i < argCount; ++i) { Serial.print(args[i]); } return Value(); }
- Parameters:
func – Name of jerryscript wrapper function
-
JS_DECLARE_FUNCTION(func)
Declare a function wrapper.
Use in a header file or as forward declaration
- Parameters:
func – Name of wrapper function
-
JS_DEFINE_METHOD(method, ...)
References
Used by
Advanced_Jsvm Sample
Basic Context Sample
Basic_Jsvm Sample
Event_Jsvm Sample
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: `jerryscript <>`__
libsodium
From its documentation: Sodium is a modern, easy-to-use software library for encryption, decryption, signatures, password hashing and more.
Usage
This component integrates libsodium v1.0.18 into Sming. To use it, simply add ARDUINO_LIBRARIES += libsodium to your application’s component.mk and #include <sodium.h> (or one of the more specific headers in sodium/*).
For further information, see libsodiums documentation.
Build Details
To build the library, Sming’s standard component build process is used in favor of libsodium’s original autotools based build process, which is not compatible with the xtensa-lx106-elf architecture. The list of source files, as well as compiler definitions, are hard-coded in component.mk according to the outcomes of (a hacked version of) the configure script.
All optimizations leveraging x86/ARM architecture-specific assembly instructions are disabled and only C reference implementations are used instead. This is true even when compiling for the “Host” architecture.
As a side effect, there is no need to invoke ``sodium_init()`` on application startup (which would otherwise detect available CPU features and select optimal implementations accordingly).
Notes on Random Number Generation
By default, the randombytes_... family of functions is hooked up to the ESP8266 hardware random number generator via os_[get_]random, which is also available in the Host emulator. However, due to the lack of documentation, it is unclear if the hardware random number generator provides sufficiently high quality random numbers for cryptographic purposes. Some additional information can be found here. Also note that the host emulator may not use a high-quality random number source and therefore should not be trusted with generating private keys and other sensitive data.
Alternatively, libsodium offers the possibility to install a custom random number generator implementation via randombytes_set_implementation(), which is fully controllable by the user.
References
Used by
Over-the-Air Firmware Upgrade Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: libsodium
Sodium is a new, easy-to-use software library for encryption, decryption, signatures, password hashing and more.
It is a portable, cross-compilable, installable, packageable fork of NaCl, with a compatible API, and an extended API to improve usability even further.
Its goal is to provide all of the core operations needed to build higher-level cryptographic tools.
Sodium supports a variety of compilers and operating systems, including Windows (with MingW or Visual Studio, x86 and x64), iOS, Android, as well as Javascript and Webassembly.
Documentation
The documentation is available on Gitbook and built from the libsodium-doc repository:
libsodium documentation - online, requires Javascript.
offline documentation in PDF format.
Integrity Checking
The integrity checking instructions (including the signing key for libsodium) are available in the installation section of the documentation.
Community
A mailing-list is available to discuss libsodium.
In order to join, just send a random mail to sodium-subscribe {at}
pureftpd {dot} org.
License
modbusino RTU Library (modbus slave)
modbusino is lightweight RTU Modbus slave library that supports ‘read holding registers’ and ‘write multiple registers’ functions. Please note that prior to commit 02dff3c (branch Sming, port URL https://github.com/kmihaylov/modbusino) a delay may occur after sending a message (more information can be found in the PR thread #2043, https://github.com/SmingHub/Sming/pull/2043#issuecomment-615945823).
Configuration variables
- RS485_RE_PIN
Default: 15
GPIO pin number for RE (Receive-Enable) output.
- RS485_TX_LEVEL
Default: HIGH.
Active level for RE pin during transmission.
References
Used by
Modbusino RTU generic sample Sample
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: modbusino
Modbusino
Introduction
Modbusino is a ISC licensed library to handle Modbus requests on Arduino (slave).
Features
To keep it simple and to reduce memory consumption, only the two following Modbus functions are supported:
read holding registers (0x03)
write multiple registers (0x10)
Example
#include <Modbusino.h>
/* Initialize the slave with the ID 1 */
ModbusinoSlave modbusino_slave(1);
/* Allocate a mapping of 10 values */
uint16_t tab_reg[10];
void setup() {
/* The transfer speed is set to 115200 bauds */
modbusino_slave.setup(115200);
}
void loop() {
/* Initialize the first register to have a value to read */
tab_reg[0] = 0x1234;
/* Launch Modbus slave loop with:
- pointer to the mapping
- max values of mapping */
modbusino_slave.loop(tab_reg, 10);
}
Contribute
I want to keep this library very basic and small so if you want to contribute:
Check for open issues or open a fresh issue to start a discussion around a feature idea or a bug.
Fork the repository on Github to start making your changes on another branch.
Send a pull request (of your small and atomic changes).
Bug the maintainer if he’s too busy to answer :)
Nano Protocol-Buffer
Introduction
This component adds support for Nano Protocol-Buffer implementation.
Nanopb is a small code-size Protocol Buffers implementation in ansi C. It is especially suitable for use in microcontrollers, but fits any memory restricted system.
C file generation from Proto files
Once this component is installed you can use it to add Nano Protocol-Buffer support to your project and generate C and header files from Proto files. One possible way to call the generator is to go to the directory where the proto file is located and run the generator. As shown below:
make -C $SMING_HOME fetch nano-pb
cd <folder-with-proto-file>
python $SMING_HOME/Libraries/nanopb/nanopb/generator/nanopb_generator.py <desired-proto-file>.proto
After the generator tool is run you will have newly generated C and header files that can be used in your Sming application.
Using
Add
COMPONENT_DEPENDS += nanopbto your application componenent.mk file.Add these lines to your application:
#include <PbUtils.h> // The line below should be replaced with the generated header file #include "cast_channel.pb.h"
Example:
#include <PbUtils.h> #include "cast_channel.pb.h" void doSomething(const uint8_t* data, size_t length) { // ... extensions_api_cast_channel_CastMessage message = extensions_api_cast_channel_CastMessage_init_default; message.protocol_version = extensions_api_cast_channel_CastMessage_ProtocolVersion_CASTV2_1_0; message.source_id.funcs.encode = &pbEncodeData; message.source_id.arg = new PbData(sourceId); message.destination_id.funcs.encode = &pbEncodeData; message.destination_id.arg = new PbData(destinationId); message.nameSpace.funcs.encode = &pbEncodeData; message.nameSpace.arg = new PbData(ns); message.payload_type = extensions_api_cast_channel_CastMessage_PayloadType_STRING; message.payload_utf8.funcs.encode = &pbEncodeData; message.payload_utf8.arg = new PbData((uint8_t*)data, length); // ... }
References
Used by
GoogleCast Library
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Submodule: nanopb
Nanopb - Protocol Buffers for Embedded Systems
Nanopb is a small code-size Protocol Buffers implementation in ansi C. It is especially suitable for use in microcontrollers, but fits any memory restricted system.
Homepage: https://jpa.kapsi.fi/nanopb/
Documentation: https://jpa.kapsi.fi/nanopb/docs/
Downloads: https://jpa.kapsi.fi/nanopb/download/
Nightly builds: https://jpa.kapsi.fi/jenkins/job/nanopb/
Using the nanopb library
To use the nanopb library, you need to do two things:
Compile your .proto files for nanopb, using
protoc.Include pb_encode.c, pb_decode.c and pb_common.c in your project.
The easiest way to get started is to study the project in “examples/simple”. It contains a Makefile, which should work directly under most Linux systems. However, for any other kind of build system, see the manual steps in README.txt in that folder.
Generating the headers
Protocol Buffers messages are defined in a .proto file, which follows a standard
format that is compatible with all Protocol Buffers libraries. To use it with nanopb,
you need to generate .pb.c and .pb.h files from it:
python generator/nanopb_generator.py myprotocol.proto # For source checkout
generator-bin/nanopb_generator myprotocol.proto # For binary package
(Note: For instructions for nanopb-0.3.9.x and older, see the documentation of that particular version here)
The binary packages for Windows, Linux and Mac OS X should contain all necessary
dependencies, including Python, python-protobuf library and protoc. If you are
using a git checkout or a plain source distribution, you will need to install
Python separately. Once you have Python, you can install the other dependencies
with pip install protobuf grpcio-tools.
You can further customize the header generation by creating an .options file.
See documentation for details.
Running the tests
If you want to perform further development of the nanopb core, or to verify
its functionality using your compiler and platform, you’ll want to run the
test suite. The build rules for the test suite are implemented using Scons,
so you need to have that installed (ex: sudo apt install scons or pip install scons).
To run the tests:
cd tests
scons
This will show the progress of various test cases. If the output does not end in an error, the test cases were successful.
Note: Mac OS X by default aliases ‘clang’ as ‘gcc’, while not actually
supporting the same command line options as gcc does. To run tests on
Mac OS X, use: scons CC=clang CXX=clang. Same way can be used to run
tests with different compilers on any platform.
For embedded platforms, there is currently support for running the tests
on STM32 discovery board and simavr
AVR simulator. Use scons PLATFORM=STM32 and scons PLATFORM=AVR to run
these tests.
Build systems and integration
Nanopb C code itself is designed to be portable and easy to build
on any platform. Often the bigger hurdle is running the generator which
takes in the .proto files and outputs .pb.c definitions.
There exist build rules for several systems:
Makefiles:
extra/nanopb.mk, seeexamples/simpleCMake:
extra/FindNanopb.cmake, seeexamples/cmakeSCons:
tests/site_scons(generator only)Bazel:
BUILDin source rootConan:
conanfile.pyin source rootPlatformIO: https://platformio.org/lib/show/431/Nanopb
PyPI/pip: https://pypi.org/project/nanopb/
And also integration to platform interfaces:
Building nanopb - Using vcpkg
You can download and install nanopb using the vcpkg dependency manager:
git clone https://github.com/Microsoft/vcpkg.git
cd vcpkg
./bootstrap-vcpkg.sh
./vcpkg integrate install
./vcpkg install nanopb
The nanopb port in vcpkg is kept up to date by Microsoft team members and community contributors. If the version is out of date, please create an issue or pull request on the vcpkg repository.
rBPF Femto-Container support
rBPF: RIOT-style Berkeley Packet Filters
The rBPF subsystem provides a minimal virtual machine for microcontrollers. This allows code to be run in isolated environments to increase platform security.
The particular implementation here is referred to as a Femto-Container
as only a very limited set of functionality is supported compared to a regular Virtual Machine.
Container code is compiled into BPF object code, which is linked into the Sming application as a BLOB (Binary Large OBject). The generated code has a very simple 64-bit instruction set which is then executed using a runtime interpreter.
See rBPF: Riot-style Berkeley Packet Filters for further details.
Toolchain setup
The following applications are required:
Clang/LLVM compiler suite
From a sample or project directory, install as follows.
Ubuntu:
sudo apt-get install clang
make python-requirements
Windows:
winget install llvm
set CLANG="c:\Program Files\LLVM\bin\clang"
make python-requirements
Sming interface
All .c or .cpp files are compiled into eBPF object code.
The resulting binary data is linked into the project using FSTR::Array objects.
These can be executed using a rBPF::VirtualMachine instance.
Compiled containers are located in the :cpp:namespace:`rBPF::Container` namespace.
Note that parameters are passed to container functions as a pointer. You should always use a structured type for this as shown in the samples.
The compiler will use the first public function in each source file as the entry point.
It is recommended that all other functions be declared static or placed within an anonymous namespace.
Low-level details
Clang is used to compile container source code to an eBPF object file. This is then converted to a Femto-Container-specific format using python.
For example, the increment.c container is compiled to increment.o then converted to increment.bin.
The .o file can be inspected using standard binutils tools such as objdump.
The Femto-Container application binary can be inspected as follows:
make rbpf-dump
Output from the sample increment container looks like this:
Magic: 0x46504272
Version: 0
flags: 0x0
Data length: 0 B
RoData length: 16 B
Text length: 24 B
No. functions: 1
functions:
"increment": 0x0
data:
rodata:
0: 0x69 0x6e 0x63 0x72 0x65 0x6d 0x65 0x6e
8: 0x74 0x00 0x00 0x00 0x00 0x00 0x00 0x00
text:
<increment>
0x0: 79 10 00 00 00 00 00 00 r0 = *(uint64_t*)(r1 + 0)
0x8: 07 00 00 00 01 00 00 00 r0 += 1
0x10: 95 00 00 00 00 00 00 00 Return r0
This shows the:
application header
list of functions
read-only data containing the function name and some padding
the application code
The application code fetches the value from the pointer in r1 (the context
argument) and increments the value in the second instruction.
The return parameter is stored in register r0.
Build variables
- RBPF_CONTAINER_PATH
default:
containerLocation of Femto-Container applications. Place all .c and .cpp source modules here.
- BPF_STORE_NUM_VALUES
default: 16
Maximum number of stored values.
Space is shared between all stores (global and local).
API Documentation
-
namespace rBPF
-
-
class LocalStore : public rBPF::Store, public rBPF::Store
- #include <Store.h>
Public Functions
-
inline virtual bool update(Key key, Value value) override
Update value in store.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
inline virtual bool fetch(Key key, Value &value) override
Fetch value from store.
If key is not found in the store then its added and set to 0.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
virtual bool update(Key key, Value value) override
Update value in store.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
virtual bool fetch(Key key, Value &value) override
Fetch value from store.
If key is not found in the store then its added and set to 0.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
inline virtual bool update(Key key, Value value) override
-
class GlobalStore : public rBPF::Store, public rBPF::Store
- #include <Store.h>
Public Functions
-
inline virtual bool update(Key key, Value value) override
Update value in store.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
inline virtual bool fetch(Key key, Value &value) override
Fetch value from store.
If key is not found in the store then its added and set to 0.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
virtual bool update(Key key, Value value) override
Update value in store.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
virtual bool fetch(Key key, Value &value) override
Fetch value from store.
If key is not found in the store then its added and set to 0.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
inline virtual bool update(Key key, Value value) override
-
class Store
- #include <Store.h>
Subclassed by rBPF::GlobalStore, rBPF::GlobalStore, rBPF::LocalStore, rBPF::LocalStore
Public Functions
-
virtual bool update(Key key, Value value) = 0
Update value in store.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
virtual bool fetch(Key key, Value &value) = 0
Fetch value from store.
If key is not found in the store then its added and set to 0.
- Parameters:
key –
value –
- Return values:
bool – true on success, false if store is full
-
inline Value get(Key key)
Fetch value from store.
- Parameters:
key –
-
class Entry
- #include <Store.h>
-
virtual bool update(Key key, Value value) = 0
-
class VirtualMachine
- #include <VirtualMachine.h>
Create a VM and load a container
- param container:
Container code blob
-
bool load(const Container &container, size_t stackSize = defaultStackSize)
Load container and initialise it.
- Return values:
bool – true on success
-
void unload()
Unload container and free any allocated resources.
Run the container
- param ctx:
IN/OUT Passed to container. Must be persistent.
- retval int64_t:
Result returned from container
Public Functions
-
VirtualMachine()
Create an uninitialised VM.
-
inline int getLastError() const
Get error code from last call to execute()
- Return values:
int – 0 on success. Retrieve text for error code using
getErrorString()
-
class LocalStore : public rBPF::Store, public rBPF::Store
References
Used by
Environment Variables
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
Si4432 RF Transceiver
SI4432 library for Arduino - v0.1
Please note that Library uses standard SS pin for NSEL pin on the chip. This is 53 for Mega, 10 for Uno.
made by Ahmet (theGanymedes) Ipkin (2014)
https://github.com/ADiea/si4432.git
References
Used by
SI443 Radio Sample
SoC support
esp32
esp32c2
esp32c3
esp32s2
esp32s3
esp8266
host
rp2040
API Documentation
This is a separate section built using DOXYGEN, which will replace this file when built.
Information
Developing for Sming
Contributing to Sming Framework
All contributions (PRs) to Sming Framework have to be done to the develop branch.
master: Branch that contains latest production (stable) release. No PRs other than Final Release updates will be merged into master. develop: Main development branch: contains all current new features.
This means that all contributors will have to submit a PR to a develop , it will be tested and then merged to develop for automated integration testing (via TravisCI, Jenkins or SemaphoreCI), as well as manual testing on a real device.
Sming Contributing flow:
Fork *Sming* repo
After that clone your own fork.
git clone https://github.com/<my-fork>/Sming.gitCreate a new branch off the develop branch
cd Sming git checkout develop git branch feature/<short-explanation> git checkout feature/<short-explanation>
Make sure to replace
short-explanationwith one or two words describing the purpose of the feature. If you want to commit a fix usefix/as prefix for your branch name.Build, test your code
Make sure that your code compiles and it is tested on real device. Sometimes you will be asked for a proof (usually screenshot) that it was tested on your device(s).
Document your code
As a bare minimum, please include a
README.rstorREADME.mdfile. See Documentation System for further information.Commit changes
git add <changed-files> git commit -m "<Description of the change(s)>"
Push your changes to your fork on github
git push
Rebase if needed
If your branch cannot be merged automatically because there are new changes in the official develop branch that conflict with yours then make sure to rebase your branch. The following steps can help you do this.
First step: You will need to add the
upstreamrepository. This step should be executed ONLY once.cd Sming git remote add upstream https://github.com/SmingHub/Sming.git git fetch upstream develop git checkout develop git reset --hard upstream/develop
Second step: If you have already defined
upstreamrepository and synchronized yourdevelopbranch to fetch the updates fromupstream( the commands above do this) the next step is to get the latest changes from the officialdevelopbranch.This can be done using
cd Sming git checkout develop git pull
Final step: Now you are ready to merge the latest changes from official
developbranch into your branch and place your changes on top. The commands below help you achieve this.cd Sming git checkout develop git pull git checkout feature/<short-explanation> git merge develop # Fix any merge conflicts if needed. git rebase develop # Fix any merge conflicts if needed.
If there were merge conflicts you will have to resolve them locally. This usually involves calling the following commands:
git mergetool # After resolving conflicts you should type the command below to see what are the next steps # Newer versions of `git` are kind enough to show hints git status
After that operation you should have a branch that has all latest changes from develop with your changes on top.
Submit PR to the main Sming repo, develop branch.
Work with other contributors to test your feature and get it merged to develop
This is the most common approach for a git-flow: http://nvie.com/posts/a-successful-git-branching-model/
Sming Documentation
This directory contains the build system for Sming’s documentation. It is intended to be read online via Read The Docs.
The source directory contains some top-level files however most of the
information is obtained from the various README files associated with the source code.
Setup
Linux:
../Tools/install.sh doc
Windows:
..\Tools\install.cmd doc
Building
Build the documentation like this:
make html
This will:
Pull in and patch every submodule
Generate doxygen API information
Build documentation in HTML format
If you make changes to any source documentation files these will be
picked up automatically when make html is next run.
If you make any changes to source code comments, you’ll need to re-build the doxygen information first:
make api -B
make html
Component Building Guide
Introduction
This section will take you through creating a Sming Component from start to finish, based on a real example from Pull Request #1992.
The aim here is to make an existing library, https://github.com/nomis/ModbusMaster, available for applications to use within the framework.
Various options and settings are only mentioned here, but are linked through to Sming build system where you can find full details of what they are and how they work.
You will need to be familiar with GIT and the use of submodules.
Structure
You should carefully consider how to incorporate third-party code. The choices are:
Add an existing third-party library directly as a Component, patching if required;
Add the a submodule as a sub-directory within a Component;
Copy code from an existing library (honouring any license agreement conditions) and modify/rewrite as necessary;
Start from scratch. If drawing on ideas from other implementations those should be attributed.
We could use (a), but that is appropriate only for very simple libraries, or those written for Sming but stored in an external repository (e.g. FlashString). You can find more details on this process in Adding External Sources.
Note
Bear in mind that the asynchronous architecture of Sming can be at odds with code written for Arduino.
In this example, the main ModbusMasterTransaction code incorporates a loop which
effectively halts program execution until a timeout occurs or a response is received.
Fixing that could require extensive patching, so option (c) or (d) may be more appropriate.
For now we are using option (b), so we’ll start by creating the Component directory:
cd $SMING_HOME/Libraries
mkdir ModbusMaster
cd ModbusMaster
Note
Do not use $SMING_HOME/Components. This is reserved for framework libraries.
Now add the submodule:
git submodule add --name Libraries.ModbusMaster https://github.com/nomis/ModbusMaster
This does three things:
Clones the external repository to the
ModbusMastersub-directory;Adds an entry to the
.gitmodulesfile with the URL, local path and specified name;Stages both changes, with a link indicating which commit should be used.
Note
The dotted name format is a Sming convention.
Before committing, we need to edit the .gitmodules file. Open in a text editor and
follow the instructions therein.
You should now re-stage and commit the changes before proceeding.
Patching
Where minor changes are required to third-party code then these can be applied automatically as patches. See Adding External Sources and GIT Submodules for details.
For ModbusMaster, we want to make a slight change to how the timeout is configured,
so this is provided in a file called ModbusMaster.patch.
Supported architectures
Unless there are specific reasons not to do so, Components should work on all supported architectures. In particular, it should build and run under the Host Emulator.
In order to do this, you should remove any low-level code from the library by:
Using Sming support classes or drivers (see Esp8266 Drivers); or
Placing low-level code into separate code modules or header files.
If a Component is intended only to work with specific hardware then ensure appropriate checks are incorporated so that building fails gracefully on un-supported architectures. You can do this in the component.mk file:
ifneq ($(SMING_ARCH),Esp8266)
$(error MyComponent currently only supports the ESP8266 architecture)
endif
Component configuration
For very simple Components the default settings are adequate:
Source code must be in the base directory or a sub-directory called
srcPublic header files must be in a sub-directory called
include
The source files will be compiled into a library, in this case clib-ModbusMaster.a.
To change the defaults, provide a component.mk file and set entries as required:
- Submodules
We need to tell Sming about the submodules:
COMPONENT_SUBMODULES := ModbusMaster
- Source code
Put the source code into a separate directory (or directories) and add those to
COMPONENT_SRCDIRS.You can also use
COMPONENT_SRCFILES.For ModbusMaster, all the source code is in the submodule so we set this to
COMPONENT_SRCDIRS := ModbusMaster/src.- Header files
Keep public and private
.hor.hppfiles in separate directories.Add the public ones to
COMPONENT_INCDIRS.Any private headers can be set in
EXTRA_INCDIR.For ModbusMaster, we set this to
COMPONENT_INCDIRS := ModbusMaster/src- Configuration options
ModbusMaster provides the
MB_RESPONSE_TIMEOUTvariable.This is mapped onto a
#definevalue with the same name usingCOMPONENT_CXXFLAGS.Note: Don’t confuse this with
COMPONENT_CFLAGSwhich is only used when building.csource files.If the value changes we want Sming to rebuild both the library and any code which uses it, so we assign it to the
COMPONENT_VARSvariable list. Users can check the value by runningmake list-config.- Dependencies
If your library depends on other libraries to build, these must be declared in the
component.mkfile by settingCOMPONENT_DEPENDSvariable.ModbusMaster doesn’t have any so this entry is not required.
Documentation
A Component MUST have a README.md (markdown) or README.rst (reStructuredText) file
with a level 1 heading and brief introduction at an absolute minimum.
Note
You may not be familiar with .rst files but they are a considerably improvement on markdown
and well worth investing a little time to learn.
See Contributing for further details.
Adding External Sources
Introduction
In Sming we have source code from other repositories such as rboot, spiffs, etc.
Having local copies of those modules brings some disadvantages with it:
We cannot easily update the local copy from the original source code.
We cannot easily send improvements to those projects once we make local changes in our local copy.
Sming uses GIT submodules which allows the build system to fetch source code from an external repository on demand.
If modifications are required then the submodule can be patched.
This simplifies the process of integrating changes from those third-party projects.
Where to place external code
Submodules may be a Component by itself (such as FlashString), or contained within a Component (e.g. rBoot).
The location must be chosen carefully:
Code required within the core Sming framework
If the Component supports multiple architectures, place it in
Sming/Components. Otherwise, use the appropriateSming/Arch/*/Componentsdirectory.
Code for general use
Create a new Library in
Sming/Libraries
Please consult Sming build system for further details about how Components are constructed.
Copying Source Code
If the source code does not have a publicly accessible GIT repository then the source code needs to be copied locally.
In order to track changes more easily, the initial commit should be an exact copy of the original.
Please either comment the code or add notes to the documentation to detail any required changes for compatibility.
Add submodules
As an example, this is how the new PWM submodule was added to the Esp8266 Drivers Component:
Add the submodule using GIT:
cd $SMING_HOME git submodule add \ --name ESP8266.new-pwm \ https://github.com/StefanBruens/ESP8266_new_pwm.git \ Arch/Esp8266/Components/driver/new-pwm
This adds an entry to the end of the .gitmodules file:
[submodule "ESP8266.new-pwm"]
path = Sming/Arch/Esp8266/Components/driver/new-pwm
url = https://github.com/StefanBruens/ESP8266_new_pwm.git
For naming submodules, please follow the convention used for the other entries in
.gitmodules, which is determined by the local path:
- ``Sming/Components``: just use the name of the submodule
- ``Sming/Arch/ARCH/Components``: Use ``ARCH.name``
- ``Sming/Libraries``: Use ``Libraries.name``
Open
.gitmodulesin a text editor and:
Move the entry to a more suitable location in the file, i.e. at the end of the section listing all the ESP8266-specific submodules
Add the line
ignore = dirty
Applying Patches
If a submodule requires changes to work with Sming, this can be handled using patches.
This is generally done by pulling in the original submodule, making the required changes
and then running a diff to create a patch file, like this:
cd <Sming-root-folder>/third-party/<module-name>
git diff --no-ext-diff > <module-name>.patch
If using a GUI such as GIT Extensions then this can be done using a right-click.
See GIT Submodules for further details about how patches are used and where they should be placed.
Using submodules
If the submodule is added as a Component in its own right, no further action is required. Applications can use it by adding the name to their COMPONENT_DEPENDS or ARDUINO_LIBARIES entries in component.mk as appropriate.
Submodules contained within a Component must be declared by adding them to the
COMPONENT_SUBMODULES entry in component.mk.
Moving submodules
If you need to change the location of a submodule, here’s a suggested approach. In this example, we’re going to move the Adafruit_Sensor submodule into a Component:
# Move the submodule temporarily
Sming/Libraries$ git mv Adafruit_Sensor tmp
# Create our new Component directory
Sming/Libraries$ mkdir Adafruit_Sensor
# Move the submodule back as a sub-directory
Sming/Libraries$ git mv tmp Adafruit_Sensor/Adafruit_Sensor
Now we can add a component.mk file, README.rst, etc. as required for the component.
Sming build system
Introduction
This guide is provided to assist with understanding, developing and modifying the build system.
A Sming project is built from a set of static libraries (object archives). Typically the application code is one library, built from the user’s source code, whilst the other libraries are common to all projects and stored in a separate, shared location.
Until recently Sming itself has always been built as one large library, but this is now broken into a number of discrete Component libraries. The concept is borrowed from Espressif’s ESP-IDF build system and whilst there are some similarities the two systems are completely independent.
Building applications
Setup
These are the main variables you need to be aware of:
- SMING_HOME
must be set to the full path of the
Smingdirectory.
- SMING_ARCH
Defines the target architecture
Esp8266 The default if not specified.
ESP_HOMEmust also be provided to locate SDK & tools.Esp32 Supports ESP32 architecture.
Host builds a version of the library for native host debugging on Linux or Windows
Rp2040 Supports Raspberry Pi RP2040-based boards.
- SMING_SOC
Some architectures support families of SOCs with different capabilities. Set this value to the specific variant being targeted.
Will automatically set SMING_ARCH to the appropriate value.
- SMING_CPP_STD
The build standard applied for the framework.
This defaults to C++17 if the toolchain supports it (GCC 5+), C++11 otherwise. You can override to use other standards, such as
c++2afor experimental C++20 support.
These variables are available for application use:
- PROJECT_DIR
Path to the project’s root source directory, without trailing path separator. This variable is available within makefiles, but is also provided as a #defined C string to allow references to source files within application code, such as with the
IMPORT_FSTRmacro.
COMPONENT_PATH As for PROJECT_DIR, but provides the path to the
current component’s root source directory.
Converting existing projects
Instead of Makefile-user.mk a project should provide a component.mk. To convert to the new style:
Copy
Makefileandcomponent.mkfrom theBasic_Blinksample projectCopy any customisations from
Makefile-user.mkintocomponent.mkfile. (Or, renameMakefile-user.mktocomponent.mkthen edit it.)Delete
Makefile-user.mkIf the project uses any Arduino libraries, set the
ARDUINO_LIBRARIESvariable
Targets You can add your own targets to component.mk as usual. It’s a good idea to add a comment for the target, like this:
##@Building
.PHONY: mytarget
mytarget: ##This is my target
When you type make help it will appear in the list.
If you need a target to be added as a dependency to the main application
build, add it to CUSTOM_TARGETS - the Basic Serial sample
contains a simple example of this.
- ARDUINO_LIBRARIES
If your project uses any Arduino libraries, you must set this value appropriately.
Source files Use COMPONENT_SRCDIRS instead of MODULES. Use
COMPONENT_SRCFILES to add individual files.
Include paths Use COMPONENT_INCDIRS instead of EXTRA_INCDIR,
unless the paths are only required to build this Component.
See component.mk for a full list of variables.
Building
You should normally work from the project directory. Examples:
Type
maketo build the project and any required Components. To speed things up, use parallel building, e.g.make -j5will build using a maximum of 5 concurrent jobs. The optimum value for this is usually (CPU CORES + 1). Usingmake -jwill use unlimited jobs, but can cause problems in virtual environments.Type
make helpfrom the project directory to get a list of available build targets.
To switch to a different build architecture, for example:
Type
make SMING_ARCH=Hostto build the project for the host emulatorType
make flashto copy any SPIFFS image (if enabled) to the virtual flash, and run the application. (Note that you don’t need to set SMING_ARCH again, the value is cached.)
To inspect the current build configuration, type make list-config.
Hardware configuration
The appropriate hardware configuration should be selected in the project’s component.mk file. Use one of the standard configurations or create your own. See Hardware configuration.
Configuration variables
Configuration variables should be set in the project’s component.mk file. If appropriate, they can also be set as environment variables.
During development, the easiest way to change variables is on the
make command line. These are cached so persist between make
sessions, and will override any values set in your project’s
component.mk file. For example:
Type
make SPIFF_BIN=test-romto build the project and (if enabled) create a SPIFFS image file calledtest-rom.binType
make flash COM_PORT=COM4to flash the project andtest-romSPIFFS image using the provided flash memory settingsNext time you type
make flash, the same settings will be used, no need to type them again
A separate cache is maintained for each build type (arch + release/debug). For example:
Type
make SMING_RELEASE=1 list-configto switch to release build and display the active configuration
Type make config-clean to clear all caches and revert to defaults.
For reference, a copy of all build variables are stored in a file with each firmware image created in the ‘firmware’ directory.
Component repositories
Placing Components in a common location allows them to be used by
multiple projects. To set up your own Component repository, create a
directory in a suitable location which will contain your Components and
set COMPONENT_SEARCH_DIRS to the full path of that directory. For
example:
|_ opt/
|_ shared/
|_ Components/ The repository
|_ MyComponent/
|_ AnotherComponent/
|_ spiffs/ Will be used instead of Sming version
User repositories are searched first, which allows replacement of any
Component for a project. In this example, our spiffs component will
be selected instead of the one provided with Sming.
Directory layout
The main Sming repo. is laid out like this:
|_ sming/
|_ .appveyor.yml CI testing
|_ .travis.yml CI testing
|_ .readthedocs.yml Documentation build
|_ lgtm.yml CI Static code analysis
|_ docs/ Sming documentation
|_ samples/ Samples to demonstrate specific Sming features or libraries
|_ Sming/
| |_ Makefile Builds documentation, performs global actions on the framework
| |_ project.mk Main makefile to build a project
| |_ build.mk Defines the build environment
| |_ component.mk Sming Component definition file
| |_ component-wrapper.mk Used to build each Component using a separate make instance
| |_ Arch/ Architecture-specific makefiles and code
| | |_ Esp8266/
| | | |_ sming.mk Defines architecture-specific Components and libraries
| | | |_ app.mk Link the project, create output binaries
| | | | and perform architecture-specific actions
| | | |_ build.mk Architecture-specific build definitions, such as compiler paths
| | | |_ Compiler/
| | | |_ Components/
| | | |_ Core/
| | | |_ Platform/
| | | |_ System/
| | | |_ Tools/ Pre-compiled or scripted tools
| | |_ Esp32/
| | | |_ ...
| | |_ Host/
| | |_ ...
| |_ Components/ Framework support code, not to be used directly by applications
| |_ Core/ Main framework core
| |_ Libraries/ Arduino Libraries
| | |_ ...
| |_ out/ All generated shared files are written here
| | |_ Esp8266/ The Arch
| | | |_ debug/ The build type
| | | |_ build/ Intermediate object files
| | | | |_ Lib/ Generated libraries
| | | | |_ tools/ Generated tools
| | | |_ release/
| | | |_ ...
| | |_ Host/
| | |_ ...
| |_ Platform/ System-level classes
| | |_ ...
| |_ Services/ Modules not considered as part of Core
| | |_ ...
| |_ System/ Common framework low-level system code
| | |_ include/
| |_ Wiring/
| |_ ...
|_ tests/ Integration test applications
|_ ...
|_ Tools/
|_ ci CI testing
|_ Docker
|_ ide IDE environment support tools
|_ Python Shared python scripts
|_ travis CI testing
A typical Project looks like this:
|_ Basic_Blink/
|_ Makefile Just includes project.mk
|_ component.mk Project-specific definitions
|_ app/ Default application source directory
|_ include/ Default application include directory
|_ out/ All generated shared files are written here
|_ Esp8266/ The Architecture
| |_ debug/ The build type
| | |_ build/ Intermediate object files
| | |_ firmware/ Target output files
| | |_ lib/ Generated libraries
| | |_ tools/ Generated tools
| |_ release/
| |_ ...
|_ Host
|_ ...
Component
The purpose of a Component is to encapsulate related elements for selective inclusion in a project, for easy sharing and reuse:
Shared Library with associated header files
App Code Source files to be compiled directly into the user’s project
Header files without any associated source or library
Build targets to perform specific actions, such as flashing binary data to hardware
By default, a Component is built into a shared library using any source
files found in the base or src directories. All Arduino Libraries
are built as Components. Note that the application is also built as a
Component library, but the source directory defaults to app instead
of src.
Components are referred to simply by name, defined by the directory in
which it is stored. The Component itself is located by looking in all
the directories listed by COMPONENT_SEARCH_DIRS, which contains a
list of repositories. (Every sub-directory of a repository is considered
to be a Component.) If there are Components with the same name in
different search directories, the first one found will be used.
Components are customised by providing an optional component.mk
file.
You can see details of all Components used in a project using
make list-components. Add V=1 to get more details.
Note that the application itself is also built as a Component, and may be configured in a similar way to any other Component.
Library variants
Libraries can often be built using different option settings, so a mechanism is required to ensure that libraries (including the application) are rebuilt if those settings change. This is handled using variants, which modifies the library name using a hash of the settings values. Each variant gets its own build sub-directory so incremental building works as usual.
There are several types of config variable:
Variable type |
Cached? |
Rebuild Component? |
Rebuild application ? |
Relink application |
|---|---|---|---|---|
COMPONENT |
Y |
Y |
Y |
Y |
CONFIG |
Y |
N |
Y |
Y |
RELINK |
Y |
N |
N |
Y |
CACHE |
Y |
N |
N |
N |
DEBUG |
N |
N |
N |
N |
Variables are usually defined in the context of a Component, in the component.mk file. All Components see the full configuration during building, not just their own variables.
The type of a configuration variable is defined by adding its name to one of the following lists:
- CONFIG_VARS
The Application library derives its variant from these variables. Use this type if the Component doesn’t require a rebuild, but the application does.
- COMPONENT_VARS
A Component library derives its variant from these variables. Any variable which requires a rebuild of the Component library itself must be listed. For example, the
esp-open-lwipComponent defines this asENABLE_LWIPDEBUG ENABLE_ESPCONN. The default values for these producesENABLE_LWIPDEBUG=0 ENABLE_ESPCONN=0, which is hashed (using MD5) to producea46d8c208ee44b1ee06f8e69cfa06773, which is appended to the library name.All dependent Components (which list this one in
COMPONENT_DEPENDS) will also have a variant created.
- COMPONENT_RELINK_VARS
Behaves just like
COMPONENT_VARSexcept dependent Components are not rebuilt. This is appropriate where the public interface (header files) are not affected by the variable setting, but the library implementation still requires a variant.
- RELINK_VARS
Code isn’t re-compiled, but libraries are re-linked and firmware images re-generated if any of these variables are changed. For example,
make RBOOT_ROM_0=new-rom-filerewrites the firmware image using the given filename. (Also, as the value is cached, if you then domake flashappthat same image gets flashed.)
- CACHE_VARS
These variables have no effect on building, but are cached. Variables such as
COM_SPEED_ESPTOOLfall into this category.
- DEBUG_VARS
These are generally for information only, and are not cached (except for
SMING_ARCHandSMING_RELEASE).
Note that the lists not prefixed COMPONENT_xx are global and so should only
be appended, never assigned.
Dependencies
COMPONENT_DEPENDS identifies a list of Components upon which this
one depends. These are established as pre-requisites so will trigger a
rebuild. In addition, all dependent COMPONENT_VARS are (recursively)
used in creation of the library hash.
For example, the axtls-8266 Component declares SSL_DEBUG as a
COMPONENT_VAR. Because Sming depends on sming-arch, which in
turn depends on axtls-8266, all of these Components get rebuilt as
different variants when SSL_DEBUG changes values. The project code
(App Component) also gets rebuilt as it implicitly depends on Sming.
GIT Submodules
Sming uses source code from other repositories. Instead of including local copies, these are handled using GIT submodules. Where changes are required, patches may be provided as a diff .patch file and/or set of files to be added/replaced. Only those submodules necessary for a build are pulled in, as follows:
The submodule is fetched from its remote repository
If a .patch file exists, it is applied
Any additional files are copied into the submodule directory
An empty
.submodulefile is created to tells the build system that the submodule is present and correct.
The patch file must have the same name as the submodule, with a .patch extension. It can be located in the submodule’s parent directory:
|_ Components/
|_ heap/
|_ .component.mk Component definition
|_ umm_malloc.patch Diff patch file
|_ umm_malloc/ Submodule directory
|_ .submodule Created after successful patching
...
However, if the Component is itself a submodule, then patch files must
be placed in a ../.patches directory:
|_ Libraries/
|_ .patches/
| |_ Adafruit_SSD1306.patch Diff patch file
| |_ Adafruit_SSD1306/
| |_ component.mk This file is added to submodule
|_ Adafruit_SSD1306/ The submodule directory
|_ .submodule Created after successful patching
...
This example includes additional files for the submodule. There are some advantages to this approach:
Don’t need to modify or create .patch
Changes to the file are easier to follow than in a .patch
IMPORTANT Adding a component.mk file in this manner allows the build system to resolve dependencies before any submodules are fetched.
In the above example, the component.mk file defines a dependency on
the Adafruit_GFX library, so that will automatically get pulled in
as well.
Component configuration
The component.mk is parsed twice, first from the top-level makefile
and the second time from the sub-make which does the actual building. A
number of variables are used to define behaviour.
These values are for reference only and should not be modified.
- COMPONENT_NAME
Name of the Component
- COMPONENT_PATH
Base directory path for Component, no trailing path separator
- COMPONENT_BUILD_DIR
The current directory.
This should be used if the Component provides any application code or targets to ensure it is built in the correct directory (but not by this makefile).
This value changes depending on the build variant.
- COMPONENT_BUILD_BASE
This value does not change with build variant.
If the Component generates source code, for example, it can be placed here (in a sub-directory).
- COMPONENT_LIBDIR
Location to store created Component (shared) libraries
- COMPONENT_VARIANT
Name of the library to build
- COMPONENT_LIBPATH
Full path to the library to be built
These values may be used to customise Component behaviour and may be changed as required.
- COMPONENT_LIBNAME
By default, the library has the same name as the Component but can be changed if required. Note that this will be used as the stem for any variants.
Set
COMPONENT_LIBNAME :=if the Component doesn’t create a library. If you don’t do this, a default library will be built but will be empty if no source files are found.
- COMPONENT_TARGETS
Set this to any additional targets to be built as part of the Component, prefixed with
$(COMPONENT_RULE).If targets should be built for each application, use
CUSTOM_TARGETSinstead. See SPIFFS IFS Library for an example.
- COMPONENT_PREREQUISITES
These targets will be built before anything else. If your library generates source code, for example, then it should be done by setting this value to the appropriate targets.
- COMPONENT_RULE
This is a special value used to prefix any custom targets which are to be built as part of the Component. The target must be prefixed by
$(COMPONENT_RULE)without any space between it and the target. This ensures the rule only gets invoked during a component build, and is ignored by the top-level make.
- COMPONENT_SUBMODULES
Relative paths to dependent submodule directories for this Component. These will be fetched/patched automatically before building.
Default behaviour is to initialise submodules recursively. To prevent this behaviour and initialise only the top-level submodule, add a file to the parent directory with the same name as the submodule and a
.no-recursiveextension.
- COMPONENT_SRCDIRS
Locations for source code relative to COMPONENT_PATH (defaults to “. src”)
- COMPONENT_SRCFILES
Individual source files. Useful for conditional includes.
- COMPONENT_INCDIRS
Default: “include”.
Include directories available when building ALL Components (not just this one). Paths may be relative or absolute
- EXTRA_INCDIR
Include directories for just this Component. Paths may be relative or absolute
- INCDIR
The resultant set of include directories used to build this Component. Will contain include directories specified by all other Components in the build. May be overridden if required.
- COMPONENT_APPCODE
List of directories containing source code to be compiled directly with the application. (Ignore in the project.mk file - use
COMPONENT_SRCDIRSinstead).
- CUSTOM_BUILD
Set to 1 if providing an alternative build method. See Custom building section.
- EXTRA_OBJ
Absolute paths to any additional binary object files to be added to the Component archive library.
- COMPONENT_DEPENDS
Set to the name(s) of any dependent Components.
- EXTRA_LIBS
Set to names of any additional libraries to be linked.
- EXTRA_LDFLAGS
Set to any additional flags to be used when linking.
- COMPONENT_PYTHON_REQUIREMENTS
If the component requires uncommon Python modules (e. g. as part of a custom build step), set this variable to one or more requirements.txt files. This allows installation of all python requirements of the project by invoking:
make python-requirements [PIP_ARGS=...]
Note
A requirements.txt file in the root directory of the Component is detected automatically without setting this variable. To prevent autodetection (e.g. if the python requirements depend on another configuration variable) you must set this variable to an empty value.
- PIP_ARGS
These values are global so must only be appended to (with +=) ,
never overwritten.
- CUSTOM_TARGETS
Identifies targets to be built along with the application. These will be invoked directly by the top-level make.
- GLOBAL_CFLAGS
Use only if you need to provide additional compiler flags to be included when building all Components (including Application) and custom targets.
- APP_CFLAGS
Used when building application and custom targets.
- COMPONENT_CFLAGS
Will be visible ONLY to C code within the component.
- COMPONENT_CXXFLAGS
Will be visible ONLY to C++ code within the component.
- COMPONENT_CPPFLAGS
Will be visible to both C and C++ code within the component.
Important
During initial parsing, many of these variables (specifically, the
COMPONENT_xxx ones) do not keep their values. For this reason it
is usually best to use simple variable assignment using :=.
For example, in Esp8266/Components/gdbstub we define
GDB_CMDLINE. It may be tempting to do this:
GDB_CMDLINE = trap '' INT; $(GDB) -x $(COMPONENT_PATH)/gdbcmds -b $(COM_SPEED_GDB) -ex "target remote $(COM_PORT_GDB)"
That won’t work! By the time GDB_CMDLINE gets expanded,
COMPONENT_PATH could contain anything. We need GDB_CMDLINE to be
expanded only when used, so the solution is to take a simple copy of
COMPONENT_PATH and use it instead, like this:
GDBSTUB_DIR := $(COMPONENT_PATH)
GDB_CMDLINE = trap '' INT; $(GDB) -x $(GDBSTUB_DIR)/gdbcmds -b $(COM_SPEED_GDB) -ex "target remote $(COM_PORT_GDB)"
These values are global and should be used ONLY in the Sming/Arch/*/build.mk files to tune the architecture compilation flags.
These values must only be appended to (with +=), never overwritten.
- CPPFLAGS
Used to provide both C and C++ flags that are applied globally.
- CFLAGS
Used to provide ONLY C flags that are applied globally.
- CXXFLAGS
Used to provide ONLY C++ flags that are applied globally.
- SMING_C_STD
Used to provide the C language standard. The default is
c11.
Important
Do NOT set CPPFLAGS, CFLAGS and CXXFLAGS outside of the Sming/Arch/*/build.mk files.
Building
For faster builds use make with the -j (jobs) feature of make. It is
usually necessary to specify a limit for the number of jobs, especially
on virtual machines. There is usually no point in using a figure greater
than (CPU cores + 1). The CI builds use -j3.
Note that Makefile-app.mk enforces sequential building to ensure
submodules are fetched and patched correctly. This also ensures that
only one Component is built at a time which keeps the build logs quite
clean and easy to follow.
Components can be rebuilt and cleaned individually. For example:
make spiffs-buildruns the Component ‘make’ for spiffs, which contains the SPIFFS library.make spiffs-cleanremoves all intermediate build files for the Componentmake spiffs-rebuildcleans and then re-builds the Component
By default, a regular make performs an incremental build on the
application, which invokes a separate (recursive) make for the App
Component. All other Components only get built if any of their targets
don’t exist (e.g. variant library not yet built). This makes application
building faster and less ‘busy’, which is generally preferable for
regular application development. For Component development this
behaviour can be changed using the FULL_COMPONENT_BUILD variable
(which is cached). Examples:
make FULL_COMPONENT_BUILD=lwipwill perform an incremental build on thelwipComponentmake FULL_COMPONENT_BUILD=1will incrementally build all Components
Custom Building
To use an external makefile or other build system (such as CMake) to
create the Component library, or to add additional shared libraries or
other targets, customise the component.mk file as follows:
Set
CUSTOM_BUILD=1Define the custom rule, prefixed with
$(COMPONENT_RULE). Note that Components are built using a separate make instance with the current directory set to the build output directory, not the source directory.
It is important that the rule uses the provided values for
COMPONENT_LIBNAME, COMPONENT_LIBPATH and COMPONENT_LIBDIR so
that variant building, cleaning, etc. work correctly. See below under
‘Building’, and the Host lwip Component for an example.
Components are built using a make instance with the current directory set to the build output directory, not the source directory. If any custom building is done then these variables must be obeyed to ensure variants, etc. work as expected:
COMPONENT_LIBNAME as provided by component.mk, defaults to component
name, e.g. Sming COMPONENT_LIBHASH hash of the component
variables used to create unique library names,
e.g. 13cd2ddef79fda79dae1644a33bf48bb COMPONENT_VARIANT name
of the library to be built, including hash.
e.g. Sming-13cd2ddef79fda79dae1644a33bf48bb COMPONENT_LIBDIR
directory where any generated libraries must be output,
e.g. /home/user/sming/Sming/out/Esp8266/debug/lib/
COMPONENT_LIBPATH full path to the library to be created,
e.g. /home/user/sming/Sming/out/Esp8266/debug/lib/clib-Sming-13cd2ddef79fda79dae1644a33bf48bb.a
COMPONENT_BUILDDIR where to write intermediate object files,
e.g. /home/user/sming/Sming/out/Esp8266/debug/build/Sming/Sming-13cd2ddef79fda79dae1644a33bf48bb
Porting existing libraries
to be completed
Known Issues
Cleaning Components are not cleaned unless defined.
e.g. make axtls-8266-clean will fail unless you also specify
ENABLE_SSL=1.
Empty libraries Components without any source code produce an empty library. This is because, for simplicity, we don’t want to add a component.mk to every Arduino library.
Empty Component directories Every sub-directory in the
COMPONENT_SEARCH_DIRS is interpreted as a Component. For example,
spiffs was moved out of Arch/Esp8266/Components but if an empty
directory called ‘spiffs’ still remains then it will be picked up
instead of the main one. These sorts of issues can be checked using
make list-components to ensure the correct Component path has been
selected.
Continuous Integration Testing
It is important to ensure that any change made to framework code does not introduce additional bugs. In practice, this is impossible to guarantee but we can at least perform a full build of the entire framework with all samples (both in the framework any any associated libraries).
This is done using the integration testing framework using linux and Windows build environments.
In addition, a number of integration tests are run using Host builds which verify the logic of a large proportion of the code. Testing low-level operation requires real hardware and this must be done manually, but in general libraries and samples can be largely tested using Host builds and carefully constructed tests.
The SmingTest library should be used for such test applications to ensure they are supported on all architectures. It also provides a mechanism for logging test results.
Github Actions
We use Github Actions to manage all test builds. This service is free of charge for open-source projects.
Note
Appveyor has been removed in favour of GitHub Actions.
We used to use Travis but this is no longer free of charge.
Sming performs the build and test logic is handled using scripts, which are intended to be easily portable to other CI services if necessary. Mostly batch scripts (.cmd) are used for Windows, and bash scripts (.sh) for GNU/Linux but where practical powershell core is used as this runs on either.
Note
Sming doesn’t perform CI builds for MacOS.
Library CI support
Sming libraries may be separately built and tested whether or not they are included as part of the Sming repository (or a fork).
There are two mechanisms available.
GitHub Actions
The library.yml reusable workflow is provided, which takes care of these tasks:
Checking in the library to test
Checking in the Sming framework
Installing build tools
Builds all applications within the library’s
samplesdirectory, for all supported architecturesIf a test application is provided then that should be located in a
testdirectory. This is built for all architectures, and also executed for Host.
Builds are handled using Tools/ci/library/Makefile.
See also https://docs.github.com/en/actions/using-workflows/reusing-workflows.
To use this in a project, add a suitable workflow to the .github/workflows directory.
Templates are provided in the .github/workflows/library directory.
Here is the basic push scenario:
name: CI Push
on: [push]
jobs:
build:
uses: SmingHub/Sming/.github/workflows/library.yml@develop
# Inputs are all optional, defaults are shown
with:
# Repository to fetch Sming from
sming_repo: 'https://github.com/SmingHub/Sming'
# Sming branch to run against
sming_branch: 'develop'
# Library alias
alias: ''
The sming_repo and sming_branch inputs are provided if your library requires modifications
to Sming which are not (yet) in the main repository.
The alias input is required where the library repository name does not correspond with
the working title.
For example, the jerryscript library is in a repository called Sming-jerryscript,
so must be checked out using a different name.
If Sming contains a library (or Component) with the same name then it will be overridden,
with a warning Multiple matches found for Component 'jerryscript' in the build log.
The ci-dispatch.yml example demonstrates manual triggering, which allows these inputs to be easily changed.
See https://docs.github.com/en/actions/managing-workflow-runs/manually-running-a-workflow.
Note that the workflow must be available in the library’s default branch, or it will not appear in the github web page.
Appveyor
Appveyor may be configured to test a Sming library separately. Steps to enable:
- Add project to appveyor account
Projects->New Projectand select from list- Set
Custom Configuration to https://raw.githubusercontent.com/SmingHub/Sming/develop/Tools/ci/library/appveyor.txt.
- Set
Project URL slug If the library under test already exists in the Sming framework then the test directory MUST have the same name to ensure it gets picked up.
For example, testing the
Sming-jerryscriptlibrary requires this value to be set tojerryscriptto match the Sming library name. Build logs should then report a warningMultiple matches found for Component 'jerryscript'.- Set sming fork/branch
By default builds use the main Sming
developbranch. If testing a library which requires changes to the framework, you’ll need to use a fork and addSMING_REPOandSMING_BRANCHenvironment variables to the project settings.Note that environment variables set here will override any values set in appveyor.txt.
The provided default makefile
builds all applications within the library’s samples directory.
If a test application is provided then that should be located in a test directory.
This is built for all architectures, and also executed for Host.
Build on your own ‘cloud’
Resources are limited to one concurrent build job per appveyor account. Each build gets two virtual CPUs but they’re not particular fast. Network connectivity is, on the other hand, excellent!
One very useful feature that appveyor provides is Bring Your Own Cloud or Computer. This allows the actual builds to be run on other hardware.
Builds can have up to 5 concurrent jobs and as many CPUs as are available. In addition, build images can be pre-installed with toolchains. This can reduce total run times from 5+ hours to around 30 minutes.
Configuration
Full support requires a Windows server with Hyper-V, WSL2 and Docker installed. Hyper-V is built into Windows 10/11 professional edition. WSL2 should be available on all Windows versions.
Linux/MacOS are supported but only for GNU/Linux images.
Note that whilst Docker supports both Windows and Linux images, both cannot be used at the same time: it is necessary to manually switch between Linux/Windows containers. However, testing shows much better performance using Hyper-V for Windows builds.
Add Docker build cloud for Linux builds:
Appveyor -> BYOC -> Add Cloud
Cloud Provider: Docker
Operating system: Windows
Base Image:
Ubuntu 20.04 MinimalShell commands:
git clone https://github.com/SmingHub/Sming --branch develop --depth 1 /tmp/sming pwsh /tmp/sming/Tools/Docker/appveyor/setup.ps1
Image name:
linux
Execute commands as indicated in the resulting screen.
Wait for the image to be built.
The final stage updates the cloud information in your appveyor account. Customise as follows:
- Name
Change this so it contains only letters, numbers and dash (-). Default names contain a space, e.g.
COMPUTER Dockerso change toCOMPUTER-Docker
- Custom Docker command arguments
Customise CPU resources, RAM usage, etc. For example:
--cpus=8
See https://docs.docker.com/engine/reference/commandline/run/.
- Failover strategy
Default values will fail a job if no worker is available to service it. The following settings are suggested:
Job start timeout: 60 Provisioning attempts: 100
Add Hyper-V build cloud for Windows builds:
Same as (1) above except:
Cloud Provider: Hyper-V
Base Image:
Windows Server Core 2019 MinimalImage name:
windows
When complete, fix the build cloud name as previously, e.g. COMPUTER-HyperV.
Also check CPU cores, RAM allocation, failover strategy.
Fix authorization token
The above steps will also install the Appveyor Host Agent software on your computer. This is the software which communicates with the Appveyor server and directs the build jobs.
The authorization token used by the agent can be found in the registry:
Computer\HKEY_LOCAL_MACHINE\SOFTWARE\AppVeyor\HostAgent
Make sure that both clouds have the same token.
Configure BYOC images
Select
BYOC->Imagesand amend mappings as follows:Image Name:
Ubuntu2004OS Type:
LinuxBuild cloud: “COMPUTER-Docker” (as configured above)
Image Name:
Visual Studio 2019OS Type:
WindowsBuild cloud: “COMPUTER-HyperV” (as configured above)
Now, when a build is started it should use your own server. To revert back to normal operation change the
Image Namefields in the entries. It’s not necessary to delete them: just add, say, “X” to the name so they’re not recognised.Note
Clouds may also be configured on a per-project basis by setting the
APPVEYOR_BUILD_WORKER_CLOUDenvironment variable to the appropriate cloud name.To get both Linux and Windows builds working concurrently using this approach would require a single cloud to support dual images.
Rebuilding docker images
Appveyor images are customised by pre-installing Sming build tools. When these are updated images must be re-built.
The easiest way to do this is using the provided dockerfiles:
cd $SMING_HOME/../Tools/Docker/appveyor
docker build --no-cache -t linux -f Dockerfile-Linux .
docker build --no-cache -t windows -f Dockerfile-Windows .
Custom images
To use a Sming fork for building the image simply replace the repo URL and branch in the Shell Commands given above.
These may also be passed to docker build as follows:
docker build -t linux-test -f Dockerfile-Linux --build-arg SMING_REPO=https://github.com/myrepo/Sming --build-arg SMING_BRANCH=feature/appveyor-revisions .
Issues
If you get error image not supported by cloud this probably means an image has been mapped to the wrong clould.
Goto Appveyor -> BYOC -> Images and update/delete the offending entries.
If either cloud is shown as offline then check the authorization token (step 4 above).
It may be necessary to restart the Appveyor Host Agent service (via Windows service manager).
Clang Tools
clang-format is a tool that implements automatic source code formatting. It can be used to automatically enforce the layout rules for Sming.
clang-tidy is a C++ “linter” tool to assist with diagnosing and fixing typical programming errors including portability/readability issues, bug-prone code constructs, interface misuse, or bugs that can be deduced via static analysis.
You can find details for the current release at https://releases.llvm.org/download.html. Note that clang-format is part of the main Clang project, whilst clang-tidy can be found in clang-tools-extra.
Installation
In Ubuntu you should be able to install them using the following command:
sudo apt-get install clang-format clang-tidy
See the the download page of the Clang project for installation instructions for other operating systems.
Important
Different versions of clang-format can produce different results, despite using the same configuration file.
We are using version 8.0.1 of clang-format on our Continuous Integration (CI) System.
You should install the same version on your development computer.
clang-format
Rules
The coding rules are described in the .clang-format file, located in the root directory of the framework.
You should not edit this file unless it is a discussed and agreed coding style change.
IDE integration
There are multiple existing integrations for IDEs. You can find details in the ClangFormat documentation.
For the Eclipse IDE we recommend installing the CppStyle plugin. You can configure your IDE to auto-format the code on “Save” using the recommended coding style and/or format according to our coding style rules using Ctrl-Shift-F (for formatting of whole file or selection of lines). Read Configure CppStyle for details.
Command Line
Single File
If you want to directly apply the coding standards from the command line you can run the following command:
cd $SMING_HOME clang-format -style=file -i Core/<modified-file>Where
Core/<modified-file>should be replaced with the path to the file that you have modified.
All files
The following command will run again the coding standards formatter over all C, C++ and header files inside the
Sming/Core,samplesand other key directories:cd $SMING_HOME make csThe command needs time to finish. So be patient. It will go over all files and will try to fix any coding style issues.
If you wish to apply coding style to your own project, add an empty
.csmarker file to any directory containing source code or header files. All source/header files in that directory and any sub-directories will be formatted when you run:make csfrom your project directory.
Eclipse
If you have installed CppStyle as described above you can configure Eclipse to auto-format your files on Save.
Alternatively, you can manually apply the coding style rules by selecting the source code of a
C, C++ or header file or a selection in it and run the Format command
(usually Ctrl-Shift-F).
clang-tidy
Configuration
No specific version is required but generally you should aim to use the most recent version available in your distribution. Version 17.0.6 was used at time of writing these notes.
The default tool configuration is defined in the .clang-tidy file, located in the root directory of the framework.
Note
Unlike clang-format, clang-tidy has to be able to compile the target code in order to perform static analysis. Code must build without errors for Host architecture.
Clang has more limited constexpr support than GCC so cannot parse some of the FlashString and NanoTime code without modification. However, these changes are only made when __clang__ is defined and do not affect regular builds with GCC.
No object code is generated by clang-tidy.
Usage
Only source files which haven’t been built are inspected. So, to restrict which code gets processed built the entire application normally, then ‘clean’ the relevant modules before proceeding with clang-tidy.
For example:
cd $SMING_HOME/../samples/Basic_Servo
make -j SMING_SOC=host
make clean Servo-clean
make CLANG_TIDY=clang-tidy
If you want to fix a particular type of problem, it’s usually best to be explicit:
make CLANG_TIDY="clang-tidy --checks='-*,modernize-use-equals-default' --fix"
Remember to run make cs and check the output before committing!
If you want to provide a custom configuration file:
make CLANG_TIDY="clang-tidy --config-file=myTidyConfig"
Note
clang-tidy can take a long time to do its work, so it’s tempting to use the -j option to speed things up. You may see some corrupted output though as the output from multiple clang-tidy instances aren’t serialised correctly. It’s usually fine to get a rough ‘first-pass’ indication of any problems though.
However, if attempting to apply fixes DO NOT use the -j option as this will result in corrupted output.
Coding Style Rules
The benefits of coding standards are readability, maintainability and compatibility. Any member of the development team in Sming should be able to read the code of another developer. The developer who maintains a piece of code tomorrow may not be the coder who programmed it today.
Therefore we enforce coding standards as described in this guide. The coding style rules are mandatory for most of the framework, and Pull Request that does not adhere to the coding style rules will not be merged until those rules are applied.
The rules are optional for libraries, but recommended.
Tools will help you adhere to these coding standards without the need to know them by heart. See Clang Tools for further details.
Please also bookmark the C++ Core Guidelines. This is an invaluable reference for writing good code and making the best use of this powerful language.
Indentation
We use tabs for indentation. Configure your editor to display a tab as long as 4 spaces. For reference, the corresponding settings in clang-format are:
TabWidth: 4
UseTab: Always
IndentWidth: 4
Naming
Classes, Structures, type aliases
Must be nouns in UpperCamelCase, with the first letter of every word capitalised. Use whole words — avoid acronyms and abbreviations (unless the abbreviation is much more widely used than the long form, such as URL or HTML).
Examples:
class HttpClient {} class HttpClientConnection {} using LargeValue = uint32_t; struct MyStruct { ... }; enum MyEnum { a, ///< Comment if required b, c, };Note
The trailing , on the final item in an enumeration declaration will ensure that clang-format places each item on a separate line. This makes for easier reading and the addition of line comments if appropriate.
Methods
Must be either verbs in lowerCamelCase, or a multi-word name that begins with a verb in lowercase; that is, with the first letter lowercase and the first letters of subsequent words in uppercase.
Examples:
bind(); getStatus();
Variables
Local variables, instance variables, and class variables must also be written in lowerCamelCase. Variable names must not start with, end with or contain underscore (_) or dollar sign ($) characters. This is in contrast to some coding conventions which prefix all instance variables with underscore, however this is reserved by the C++ standard and can create problems.
Variable names should be short yet meaningful. The choice of a variable name should be mnemonic — that is, designed to indicate to the casual observer the intent of its use. One-character variable names should be avoided except for temporary “throwaway” variables. Common names for temporary variables are i, j, k, m, and n for integers; c, d, and e for characters.
Examples:
int i; char c; WebsocketClient* client;
Pre-processor definitions
#defined macros must be written in uppercase characters separated by underscores. Names may contain digits if appropriate, but not as the first character. For example:
#define MAX_PARTICIPANTS 10
Constants
Typically declared using
constorconstexprand, like variables, should be lower-camelcase. Names MUST NOT be all-uppercase as these may be confused with #defined values.See C++ Core Guidelines.
Use of typedef in C++
Use of typedef in C++ code is not recommended.
The using keyword has been available since C++11 and offers a more natural way to express type definitions. It is also necessary in certain situations such as templating.
For example:
using ValueType = uint32_t;
is more readable than:
typedef uint32_t ValueType;
Especially in multiple type declarations the subject is always immediately after the using keyword
and makes reading much easier.
enum/struct declarations
This:
typedef struct _MyStructTag {
...
} MyStruct;
typedef enum _MyEnumTag {
...
} MyEnum;
is overly verbose and un-necessary. It’s a hangover from ‘C’ code and should generally be avoided for readability and consistency. This is the preferred definition:
struct MyStruct {
...
};
enum MyEnum {
.............
};
It’s also un-necessary to qualify usage with enum. i.e. MyEnum e; is sufficient, don’t need enum MyEnum e;.
C++ Standard
For the moment we recommend the use of C++11. The corresponding settings in clang-format are:
Standard: Cpp11
Cpp11BracedListStyle: true
Starting and ending spaces
We don’t recommend the use of a starting or ending space in angles, container literals, c-style cast parentheses, parentheses and square brackets. Our settings are:
SpaceAfterCStyleCast: false
SpaceBeforeParens: Never
SpaceInEmptyParentheses: false
SpacesInAngles: false
SpacesInContainerLiterals: false
SpacesInCStyleCastParentheses: false
SpacesInParentheses: false
SpacesInSquareBrackets: false
See the meaning of those keys and their selected values in the ClangFormatStyleOptions document.
Line length
We are living in the 21st century so most of the monitors should be capable of displaying 120 characters per line. If a line is longer than those characters it will be split whenever possible:
ColumnLimit: 120
Empty Lines
Two or more empty lines will be compacted to one. Also we delete empty lines at the start of a block:
KeepEmptyLinesAtTheStartOfBlocks: false
MaxEmptyLinesToKeep: 1
Braces
See the meaning of these keys and their selected values in the ClangFormatStyleOptions document:
BraceWrapping:
AfterClass: false
AfterControlStatement: false
AfterEnum: true
AfterFunction: true
AfterObjCDeclaration: false
AfterStruct: false
BeforeElse: true
IndentBraces: false
BreakBeforeBraces: Linux
Pointer Alignment
Always on the left:
PointerAlignment: Left
Includes
We don’t re-sort includes although it is highly recommended to order the headers alphabetically whenever possible:
SortIncludes: false
Spaces
For readability put always spaces before assignment operators:
SpaceBeforeAssignmentOperators: true
Standard file headers
Please use the standard Sming header with copyright notice:
/****
* Sming Framework Project - Open Source framework for high efficiency native ESP8266 development.
* Created 2015 by Skurydin Alexey
* http://github.com/anakod/Sming
* All files of the Sming Core are provided under the LGPL v3 license.
*
* [Insert filename here] - [optional brief description of file]
*
* @author [date] [name] [email]
*
* [comments]
*
****/
Do not include details of minor changes to the file as this is handled by GIT. It may be appropriate to add notes to identify major changes or contributions. These should be marked with a new @author tag.
Deprecating code
Where a change in the Sming API may break existing users’ code, then the existing type/method/function/variable must be maintained for a time to allow time for migration to the new technique. Such changes should only be made if there is a good reason, for example improved reliability, performance, ease of use.
Deprecation requires two steps:
Step 1: Add a @deprecated tag to the method header comment so the change
is flagged in the auto-generated API documentation. Include a brief
explanation of the new method or technique to be adopted. See also
Documenting the API.
Example:
/** @deprecated Use `anotherMethod()` instead */
Step 2: Append SMING_DEPRECATED to the method declaration so the
compiler will flag a warning if that method is used during compilation.
The framework and samples must build without referencing any deprecated methods, functions or variables.
Virtual Classes
Sming makes extensive use of virtual classes. If you are modifying or adding virtual methods then please follow these guidelines:
Rule: The base class must have a virtual destructor, even if it doesn’t do anything. Example:
virtual ~Stream() {}
Rule: Inherited classes must not prepend virtual or append
override to any destructor. Example:
~IDataSourceStream();
Rationale: virtual destructors do not behave like regular virtual
methods - they are ‘chained’ rather than overridden - therefore
override is not appropriate and virtual is both un-necessary
and unhelpful
Rule: Use the override directive on inherited virtual methods:
int read() override;
Rationale: The compiler will ensure there is actually a base method to inherit from and generate a warning if one is not found, or if parameters do not correspond.
Rule: Don’t use empty destructors in inherited virtual classes
Rationale: They’re not necessary
Common issues
Some notes on commonly occurring issues:
/**
* @brief Basic example class
*/
class VirtualBuffer
{
public:
virtual ~VirtualBase
{
}
virtual unsigned getLength() const = 0;
};
/**
* @brief Descendant example class
*/
class MemoryBuffer : public VirtualBuffer
{
public:
/*
Note: Omit destructor if not required in descendant
*/
~VirtualDescendant()
{
/*
Note: delete includes null pointer check so you don't have to
*/
delete buffer;
}
/*
Use `const` qualifier for methods which don't modify object
*/
const char* getBuffer() const
{
return pos;
}
/*
Trivial code should go into the class header file where possible.
Rationale: Compiler is better able to optimise code. Easier to read.
Use `override` on virtual methods
*/
unsigned getLength() const override
{
return length;
}
/*
Use methods to access member variables rather than making them public
Rationale: Protects data, helps when tracking down bugs
*/
void setBuffer(char* newBuffer, unsigned newLength)
{
delete buffer;
buffer = newBuffer;
length = newLength;
}
private:
/*
Each class should operate on a small, well-defined item of data.
*/
/*
Class variables should be defined with initialisers, rather than using code in the constructor.
Rationale: Reduces/eliminates risk of un-initialised data causing unpredictable behaviour.
*/
char* buffer = nullptr;
/*
Remember `int` can be unsigned! If a value doesn't need to be signed, don't make it so.
Rationale: unsigned values are simpler to check, less likely to introduce bugs, compiler can better optimise computations
*/
unsigned length = 0;
};
Documentation System
Read the Docs and Sphinx
Online documentation is managed via Read the Docs, which uses Sphinx as the documentation build system.
This page describes specific details for the Sming documentation build system.
reStructuredText
These have a .rst file extension. Some references you should have a read through:
Markdown .md files are supported (except for samples) but please ensure there is only one
main heading (*HEADING) at the top containing the title of the Library or Component or it
may be incorrectly rendered in the table of contents. If you run into problems trying to get
your markdown to display correctly, then it’s probably time to move to reStructuredText!
If you need to convert existing documentation into reStructuredText take a look at pandoc. Note that their online tool doesn’t handle long files.
Source file location
Sphinx requires all document files (documents, images, etc.) to be
located in one place, namely inside the docs/source directory.
During the build process, documentation files from the Component, Library
and Sample directories are copied:
README.md
*.rst
*.svg
*.png
*.jpg
This is also performed for any submodules declared by the COMPONENT_SUBMODULES variable in a component.mk file.
This should be sufficient for most cases. However, if additional
documentation is referred to, such as in submodules belonging to a
Component, then these must be listed in the component.mk file like
this:
COMPONENT_DOCFILES := submodule/docs/api-guide.md submodule/api.txt
See Sming build system for more information about component.mk files.
README files
All Components, Libraries and Samples must include a README.rst or README.md file as follows:
Main Heading: This will be used as the title in the online documentation, so keep it brief but informative. For example,
BME280 Barometric Pressure Sensor.Purpose: What is the purpose of the code?
References: Is this based on or does it use existing code? Please include details.
Datasheets: If appropriate, please include links to manufacturer’s or external development websites. Note that any submodules or dependencies are automatically documented: see Sming build system for details, specifically
COMPONENT_SUBMODULESandCOMPONENT_DEPENDS.
You should also try to include any other information which could be useful information for a new developer. The purpose of samples projects is to demonstrate specific features or libraries, so please ensure this is adequately described.
Optionally, you may also like to include a screenshot or other diagrams or illustrations.
Please use .png, .jpg or .svg files.
Attention
The README filename is case-sensitive
Attention
Please ensure there is only one top-level heading or the documentation contents will not display correctly.
You should use the available annotations to make browsing the documentation easier. Using the Sphinx roles <https://www.sphinx-doc.org/en/master/usage/restructuredtext/roles.html> will insert hyper links to the corresponding definitions. Additional roles are provided specifically for the Sming documentation - see link_roles below.
Code blocks
There are multiple ways to show syntax-higlighted literal code blocks in Sphinx. See Showing code examples for full details.
Use the code-block directive like so:
.. code-block:: c++
for(int i = 0; i < 100; ++i) {
goto hell;
}
The language for highlighting is indicated. You can find a full list at pygments.org, however for consistency it is suggested that you use one of these:
text Doesn't highlight anything
c++ C++ code examples
bash Linux shell code
batch Windows batch file or commands
make Makefile
rst reStructuredText
You can set a default like this:
.. highlight:: c++
which will apply to any subsequent use of:
.. code:block::
or, the short-hand version:
::
API Documentation
Function, structure, class and type information is extracted from comments in the source code (see Documenting the API). This is parsed using Doxygen into XML, which is then made available using the Breathe sphinx extension. You can then pull in definitions like this:
.. doxygenclass:: String
:members:
If you wish to refer to a type within documentation, you can add a link to the definition like this:
The :cpp:class:`String` class is really useful.
This is handled using cpp inline expressions.
See Esp8266 GDBSTUB for Sming for a more complex example. At the bottom of the file we pull in the documentation for all the #defined configuration using:
.. doxygenfile:: gdbstub-cfg.h
We can then refer to a macro like this:
Don't wait on startup by setting :c:macro:`GDBSTUB_BREAK_ON_INIT` =0
In many cases including a file like this is not the best approach, perhaps using a group:
.. dogygengroup:: wstring
Or individual classes. Some experimentation may be necessary but there are plenty of examples within the main documentation to guide you.
You can use the following build variables within your Component’s component.mk file to direct doxygen parsing:
- COMPONENT_DOXYGEN_INPUT
Specifies directories or files to be parsed by Doxygen. All paths are relative to the Component directory.
If you need to specify an absolute path, append directly to DOXYGEN_INPUT instead.
- COMPONENT_DOXYGEN_INCLUDE
Add any directories or files which should be pre-processed but not included in the output.
If you need to specify an absolute path, append directly to DOXYGEN_INCLUDE_PATH instead.
- COMPONENT_DOXYGEN_PREDEFINED
Specify any necessary pre-processor definitions. An example where this is required is for function attributes #defines which would otherwise be incorrectly interpreted as variable names and cause parsing errors:
CUSTOM_ATTR void myFunc(); ^^^
So we can do this:
COMPONENT_DOXYGEN_PREDEFINED := \ CUSTOM_ATTR=
Build (environment) variables
These are defined in the README for the corresponding Component using:
:envvar::`COM_SPEED`
Determines default serial port speed
You can refer to them like this:
Change baud rate using the :envvar:`COM_SPEED` variable.
Link Roles
The documentation build system provides some custom roles for creating links.
Components
Inserting a link to a Component page, using the title of that page by default:
See :library:`Spiffs` for details of the flash filing system.
We use :component-esp8266:`axtls-8266` for SSL support.
The host has a special :component-host:`UART Driver <driver>`.
The last example shows how to change the hyperlink text. It defaults to the README description.
Libraries
As for Components, refer to libraries like this:
Use the :library:`Adafruit_ST7735` library to do some fancy display stuff.
Sample applications
To refer to a sample application README:
See the :sample:`Basic_Blink` sample for a simple introduction to Sming.
Samples may be located in the main Sming samples, but a growing number are located in a samples sub-directory of the associated Component or library.
Another example:
Where is the :sample:`generic` sample?
There are currently three libraries with a sample called ‘generic’. Documents within the appropriate library will be matched correctly.
The given target ‘generic’ is matched as follows:
Is the document contained in a Component or library? - Yes: Look in the Component’s samples sub-directory, if there is one
No match? Then check the main sming samples directory.
Still no match? Pick the first match from any other Component.
No match found? Text will appear unchanged in the output document.
To refer to a sample in a specific Component, do this:
I'm looking for the :sample:`CS5460/generic` sample.
Where is :sample:`Generic CS5460 sample <CS5460/generic>` sample?
Within main samples directory:
Is there a main :sample:`/generic` sample? Actually, no.
But there is a :sample:`/CanBus` sample!
Source code
- Use within the framework
Use the file or directory path relative to the root directory. For example:
See :source:`Sming/Core/DateTime.h`
will create a hyperlink to the source file in the Sming repository on github.
- Use within a Component or library
The source link will be interpreted relative to the Component root directory. You can find an example of this usage in the UPnP README:
:source:`samples/Basic_UPnP/include/Wemo.h#L59-L91`
If you wish to refer to a source file elsewhere in Sming, prefix with ‘/’:
See :source:`/tests/HostTests/modules/Timers.cpp` for an example.
Issues and Pull Requests
If you want to refer to discussions on github, insert links like this:
See :pull-request:`787`
See :issue:`1764`
Eclipse
You can find a good plugin editor for Eclipse by searching the
marketplace for rest editor. For example,
http://resteditor.sourceforge.net/. A useful feature is dealing with
heading underscores, just type this:
My Heading
==
Then when you save the file it gets formatted like this:
My Heading
==========
Tables, unfortunately, do take a bit of manual formatting to get right.
Sphinx Extensions
The documentation system is easily extended to support new features. This section summarises the extensions included.
- m2r2
Provides support for markdown content.
- breathe
To support Doxygen integration. See API Documentation.
- link-roles
A custom extension implemented in link-roles.py. See Link Roles.
- sphinxcontrib.wavedrom
For implementing timing and other waveform diagrams within documents. See Servo RC PWM Control for an example.
- sphinxcontrib.seqdiag
For embedding sequence diagrams, such as in the Tasks page.
Documenting the API
This guide describes how to document the Sming API. It is aimed at developers of the framework, not users of the framework.
Sming API documentation is created using javadoc style comments within the source code. Comments are added to header files and doxygen is used to create HTML documentation. API documentation is organised as a set of modules, each of which represents a logical topic, e.g. Date and Time.
The API documentation covers the core Sming framework and relevant parts of the ESP SDK (where this is required to use the framework). The API does not cover contributed libraries which should be documented by the upstream contributor. Within the source code tree, this includes:
Services
SmingCore
Wiring
rboot
The following elements of source code are documented:
Global constants
Classes
Class public members: functions, variables, operators
Macros
Structures
Enumerations
Javadoc comments use the following style:
Start with /**
End with */
Each line (between start and end) starts with a space and a *
Doxygen commands are prefixed with @
Constants documentation use the command ///<
Each API documentation comment should be part of a module. To define a module, use the @defgroup command. This should be performed only once within the Sming framework source code. The format is:
@defgroup groupname Brief description of the group
To configure a group to span code, use the @{ and @} commands. When appended to a comment block, @{ will include all subsequent javadoc comments until the next @} command. For example:
/**
* @defgroup DateTime Date and time functions
* @{
*/
...
/**
* @brief Date and time class
...
* @}
*/
To add some documentation comment to a group, use the @addtogroup command. This may be performed several times within the Sming framework source code. The format is:
@addtogroup groupname
Use the @{ @} commands to span comments. This avoids the need to add @addtogroup commands to each comment.
To create a hierarchy of groups, use the @ingroup command for child groups, for example:
/**
* @defgroup SystemTime System time functions
* @ingroup DataTime
...
It is common to be able to define a group and span a whole file using @defgroup and @{ @} commands. This is often possible when all content of a header file relates to a common topic.
Constant values should be included in the constants group, for example:
/**
* @brief ESP GPIO pin configuration
* @ingroup constants
*/
To document a function, use the following comment style:
/**
* @brief Brief description of function
* @param "Parameter name" Description of parameter (Default: value if defined)
* @param ... rest of parameters
* @retval DataType Description of return value
* @note Any extra notes that may assist application developer
*/
For example:
/** @brief Get human readable date and time
* @param format Select the date format, e.g. dtDateUS for mm.dd.yyyy (Default: dd.mm.yyyy)
* @retval String Date and time in format dd.mm.yyyy hh:mm:ss
* @note Date separator may be changed by adding #define DT_DATE_SEPARATOR "/" to source code
*/
String toFullDateTimeString(dtDateFormat_t format = dtDateFormatEU);
To document a Sming framework constant, use command ///<, for example:
int8_t DayofWeek; ///< Day of week (0-6 Sunday is day 0)
To reference another method or function, write its name with ():
Use `otherMethod()` instead
This will match any method in the current class called otherMethod,
regardless of arguments. Note the backticks ` are there just to
highlight the code. If you need to be specific:
Use `otherMethod(const char*, int)` instead
To add a See Also section, use @see. Example:
@see See `OtherClass()` for details
Multitasking
Pre-emptive Multitasking
Modern computers have evolved so that you can write a program and it will (mostly) run without interfering with anything else that is also running.
With multiple CPUs tasks can actually run at the same time (concurrently), but as there are always many more threads (tasks) these have to be shared. This is especially true on older systems which only have a single CPU.
The OS does this using a mechanism called Pre-emptive Multitasking. As far as your program is concerned, it just runs without any apparent interruptions, but in reality it only gets a little ‘slice’ of time before the operating system forcibly switches to a different program and lets it run for another ‘slice’ of time. (Hence the term, time slicing.)
With pre-emptive multitasking the operating system has to maintain state for every single task that is active. This requires additional RAM.
There is also the overhead of switching between tasks. Each task also requires its own stack space so there is usually more restriction on how it is used.
FreeRTOS is perhaps the most well-known example of a pre-emptive embedded OS.
Co-operative Multitasking
By contrast, Co-operative Multitasking, requires applications to ‘play fair’ and not hog the CPU. This means that whenever you get called to do some work, you must release control back to the system in a timely manner.
This technique is well-suited to resource-limited systems such as the Esp8266.
The Esp8266 has only a single CPU, and relatively limited RAM; about 50KBytes is available for application use. The heap grows from the bottom of RAM, and the stack starts at the top.
Sming uses the ESP8266 Non-OS SDK, which manages much of the low-level system control including the WiFi stack. It contains the main control loop, which in pseudo-code goes something like this:
for(;;) {
ServiceWifi();
ServiceTimers();
ServiceTasks();
FeedWatchdog();
}
Application code gets called from there in response to a network request, or a timer that you’ve set up, and you must deal with the request and return promptly to allow things to continue.
If you don’t do this, the system will very likely behave erratically and suffer from dropped WiFi connections and poor responsiveness. In extreme cases, the system will reset as a self-defence mechanism via Watchdog Timer; If it didn’t do this, the device would remain unresponsive until physically reset, which is generally a bad thing for an embedded device!
Note that whilst the ESP32 implementation uses the ESP IDF framework based on FreeRTOS, which is a pre-emptive OS, the programming model for Sming is the same.
Attention
Although there are functions available to manually reset watchdog timers, you should endeavour to avoid doing this from application code unless absolutely necessary.
Events
Introduction
The Sming framework has an event-driven architecture. In other words, we need to think of our application as performing some action in response to an event, input or stimulus.
Software events
Let’s look at the Basic Blink sample. We define a couple of global variables:
Timer procTimer;
bool state = true;
The one to note is procTimer, which is a software timer (see Timers and Clocks).
The startup entry point for a Sming application is the init() function:
void init()
{
pinMode(LED_PIN, OUTPUT);
procTimer.initializeMs(1000, blink).start();
}
This sets up a repeating timer, so that every 1000ms the blink function gets called:
void blink()
{
digitalWrite(LED_PIN, state);
state = !state;
}
Note that what we don’t do is sit in a loop waiting for something to happen.
Interrupt events
The Basic Interrupts sample can be summarised as follows:
We have a button connected to GPIO0 as an input
- Button pressed
- Hardware interrupt on GPIO0
Change output state of LED via GPIO2
There are two ways we can determine when the state of our GPIO0 pin changes:
- Polling
We read the input very quickly and compare the current value with the previous value, which we keep a note of. Our event occurs when the value changes, or if the input goes HIGH, or LOW, etc.
- Interrupt
If we need to catch fast pulses then polling may not work as the CPU might be doing something else and we’ll miss it. So instead, we’ll get the hardware to monitor the input for us and generate an event when it changes.
Both techniques have advantages and disadvantages, and interrupts are certainly not appropriate in all situations.
Bear in mind that every time an interrupt occurs the CPU has to stop executing your regular program, save any critical registers then jump to the interrupt service routine to run your code. All this takes time, so if the input changes very fast and very frequently then it can consume a lot of CPU and make the system very sluggish (or even crash it).
Callbacks
In Sming, we generally register a callback function to be invoked when an event occurs, such as if a timer expires.
This can be a regular ‘C’ callback function, which you should use for handling interrupts.
For regular application code, a Delegate provides more flexibility and allows you to create
simpler, cleaner code. See Delegation
for a bit of background.
The Basic Delegates sample shows how to use various types of callback.
One common requirement is to use a class method as a callback, which is easily done using Delegates.
This flexibility comes at a cost, however:
Speed. They are slower than their regular C-function counterparts
Memory. Some calls may use the heap in the background.
These are the main reasons why you should not use Delegates in an interrupt context.
See Pull Request #1734 for some further details about the relative speeds.
Memory
Map
You can find a map for the ESP8266 memory layout in the Wiki.
Memory Types
The ESP8266 has several types of memory, and it is important to have a basic apprecation of what they are and how they’re used.
DRAM
Data RAM where variables, etc. are stored. Contains the stack (which starts at the top of RAM) and the heap (which starts near the bottom, after any static data).
IRAM
Instruction RAM. All code executes from here, but there’s not much of it so so only parts of your application are loaded at any one time. This caching happens in the background by mapping certain memory address ranges into physical flash locations.
If a location is accessed which is already in the cache (a ‘hit’) then the access runs at full RAM speed. If it’s not in the cache (a ‘miss’) then an interrupt (exception) is generated internally which performs the flash read operation.
This is why interrupt service routines must not access PROGMEM directly, and must
be marked using IRAM_ATTR to ensure it’s always available.
You may get a performance improvement using IRAM_ATTR, but as frequently-used
code will be cached anyway it won’t necessarily run faster.
If the code is timing critical it may benefit from pre-caching. See Esp8266 SPI Flash Support.
Flash
Main storage for your application, libraries, the Espressif SDK code, etc. Flash memory is accessed via external serial bus and is relatively slow. For the ESP8266, it’s approximately 12x slower, though this only applies on cache misses.
See Flash memory for further details.
ROM
Fixed code stored in the Esp8266 itself containing very low-level support code which is factory-programmed and cannot be changed.
Initialisation
At startup, only the non-volatile Flash and ROM memories contain anything useful, whilst DRAM and IRAM will probably just contain garbage. The Arduino platform was initially designed to work with much simpler hardware, where the program was executed directly from Flash memory on hardware reset.
BIOS
The ESP8266 and ESP32 are far more complex, and most of the low-level initialisation happens in the ROM code. The ROM essentially contains the systems BIOS, with various low-level routines which may be used instead of accessing hardware directly. It is also responsible for setting up memory caching.
Runtime libraries
Control is passed to runtime libraries provided by Espressif, stored in Flash memory. Both ROM and runtime code are closed-source and not generally available for inspection, though disassemblies do exist.
Boot loader
The first point we really see what’s going on is in the bootloader (rBoot). The bootloader identifies the correct program image (as there can be more than one), loads the starting portion into IRAM and jumps there. It also configures the caching mechanism so that the correct program image is loaded. You can find more details about this in the rBoot documentation.
Memory initialisation
Code is copied from flash into IRAM, and const data copied into DRAM. Also static and global variable values are initialised from tables stored in flash. Static and global variables without an initialised value are initialised to 0.
Sming initialisation
{todo}.
Flash memory
Introduction
ESP8266 flash memory sizes vary from 512Kbytes on the ESP-01 up to 4Mbytes on the ESP12F. Up to 16MBytes are supported for custom designs.
Sming version 4.3 introduced partition tables to support multiple architectures, different hardware variants and custom flash layouts without restriction.
See Hardware configuration for details.
A typical layout for a 4MByte device might look like this:
Address
Config variable
Size
Source filename
Description
(hex)
(if any)
(KB)
(if applicable)
000000
1
rboot.bin
Boot loader
001000
4
rBoot configuration
002000
ROM_0_ADDR
rom0.bin
First ROM image
102000
ROM_1_ADDR
rom1.bin
Second ROM image
200000
RBOOT_SPIFFS_0
300000
RBOOT_SPIFFS_1
3FA000
4
Partition table
3FB000
4
blank.bin
RF Calibration data (Initialised to FFh)
3FC000
4
esp_init_data_default.bin
PHY configuration data
3FD000
12
blank.bin
System parameter area
Note
This was the previous layout for a 4MByte flash device:
Address |
Config variable |
Size |
Source filename |
Description |
|---|---|---|---|---|
(hex) |
(if any) |
(KB) |
(if applicable) |
|
000000 |
1 |
rboot.bin |
Boot loader |
|
001000 |
4 |
rBoot configuration |
||
002000 |
ROM_0_ADDR |
rom0.bin |
First ROM image |
|
100000 |
RBOOT_SPIFFS_0 |
|||
202000 |
ROM_1_ADDR |
rom1.bin |
Second ROM image |
|
300000 |
RBOOT_SPIFFS_1 |
|||
3FB000 |
4 |
blank.bin |
RF Calibration data (Initialised to FFh) |
|
3FC000 |
4 |
esp_init_data_default.bin |
PHY configuration data |
|
3FD000 |
12 |
blank.bin |
System parameter area |
Speed and caching
Flash memory on the ESP8266 is accessed via an external SPI bus, so reading it takes about 12x longer than reading from internal RAM. To mitigate this, some of the internal RAM is used to cache data. Part of this is managed in hardware, which means if the data required is already in the cache then there is no difference in speed. In general, then, frequently accessed data is read as if it were already in RAM.
Bear in mind that every time new data is read via the cache, something else will get thrown away
and have to be re-read. Therefore, if you have large blocks of infrequently accessed data then
it’s a good idea to read it directly using flashmem_read(). You can get the address for a
memory location using flashmem_get_address().
See Program Space for details of how to store data in flash, and access it.
Bulk Storage and Filing Systems
Sming uses a class-based layered approach to bulk storage.
The Storage Management Component defines the Storage::Device abstract class,
which devices such as SPI flash implement to provide raw read/write/erase access.
Devices are partitioned into areas for specific uses which applications
access using a Storage::Partition object.
The Disk Storage library provides support for block-access devices which use standard partitioning schemes (MBR, GPT). SD cards are supported via the SD Storage library.
Sming uses an installable (virtual) filing system mechanism based on IFS::IFileSystem.
This is managed by the Installable File System Component and contains the FWFS lightweight read-only filing system.
Additional filing system implementations are provided in separate libraries:
Note that when using bulk storage of more than about 4GB in size applications should be built with
ENABLE_STORAGE_SIZE64 =1. This changes all sizes and offsets to 64-bit.
If manipulating files greater than about 2GB (signed 32-bit value) then the ENABLE_FILE_SIZE64
setting should also be enabled.
Strings
Sming provides three classes to deal with string content:
Wiring String for flexible RAM string handling
CString for efficient storage of C-style char* RAM strings
CStringArray for handling small strings lists
Flash Strings for storing and handling strings stored in flash memory
Interrupts
Rules
Normal code runs in a Task context, that is to say there are no particular restrictions on what functions can be called. Sming is a single-threaded framework, so all tasks execute sequentially.
However, there are some rules you need to follow when writing code which runs in an interrupt context:
Use
IRAM_ATTRattribute on the interrupt handlerDon’t call any code which isn’t in IRAM, or marked with
__forceinline. Note:inlineby itself is just a compiler ‘hint’ and there is no guarantee the code will actually be inlined.Keep the code as brief as possible
Don’t use the heap (malloc, free, new, etc.). Buffers, etc. must be pre-allocated in task context.
Put processing code into a separate function and add to the task queue
SystemClass::queueCallback(), or use aSimpleTimerorTimer.Be wary if using
Delegatecallbacks as the compiler may need to use the heap, if parameter lists do not match. For this reason, avoid capturing lambdas and method callbacks. If in doubt, stick with regular ‘C’ function pointers as defined by theInterruptCallbacktype.
Contact bounce
If you use a jumper wire to ground the input pin (or even a proper switch) then it will bounce around as contact is made and then broken. This means you’ll get multiple outputs instead of a clean signal, which will cause havoc with interrupt handlers.
One solution is to use a ‘debounce circuit’, which can be as simple as this:
3v3
_|_
___ 100nF
|
|
INPUT >--------------> GPIO PIN
Provided pull-ups are enabled on the GPIO, this is sufficient to slow the input.
An alternative solution is to poll the input instead. See Tasks.
Tasks
Introduction
If you need to perform any kind of background processing task, then you will need to consider how to break it up so it executes in small slices to allow other system functions to continue normally.
You can use callback timers to schedule periodic tasks, but if you simply need to run the task as soon as possible you should use the Task Queue.
The BasicTask class
This class uses the task queue plus a timer to provide an easy way to write a background task. All you need to is define a loop() function which does the work. The task has three states:
To see this in operation, have a look at the Basic Tasks sample.
Task Schedulers
If you want to read further about schedulers, there are various libraries available for Arduino.
These are quite simple and generic:
This one is rather more complex:
Using a scheduler is a powerful technique which allows the programmer to focus on the task at hand, rather than how to get it to run.
Debugging
Debugging is a powerful technique allowing you to interactively run your code and be able to see much more information about the things that went wrong.
Architectures
Debugging on the supported architectures requires the installation and initial setup of different debugging tools. The links below will give you more information and step-by-step instructions.
Debugging on Host
Required tools and hardware
A GNU C/C++ debugger is the only requirement for the Host architecture. Make sure that you have the following executable in your PATH:
gdb
No additional hardware is required.
Recompilation is required
In order to debug applications based on Sming Framework make sure that you are using Sming version 3.8.0 or newer.
Compilation directives
If you want to debug your application and the Sming Framework code make sure to
(re)compile it with ENABLE_GDB =1 directive:
cd $SMING_HOME/../samples/LiveDebug
make dist-clean
make ENABLE_GDB=1
The commands above will re-compile Sming with debug symbols and optimizations for debugging. These commands need to be executed once.
You can recompile Sming with the following directives to debug better Sming and the LWIP TCP/IP stack
cd $SMING_HOME/../samples/LiveDebug
make Sming-build all ENABLE_GDB=1 ENABLE_LWIPDEBUG=1
Application
To use, (re)compile your application with the ENABLE_GDB option and flash it to the board. For this example we will use the Live Debug sample application:
cd $SMING_HOME/../samples/LiveDebug
make clean
make ENABLE_GDB=1 # -- recompiles your application with debugging support
The next step is to start the debugger. This can be done with the command below:
make gdb
After that a new interactive debugging session will be started:
Welcome to SMING!
Type 'r' to run application
To start the execution of the application type r or run:
(gdb) r
Starting program: /x/Sming/samples/LiveDebug/out/Host/debug/firmware/app --flashfile=out/Host/debug/firmware/flash.bin --flashsize=4M --pause
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/i386-linux-gnu/libthread_db.so.1".
[New Thread 0xf7bdcb40 (LWP 16428)]
Welcome to the Sming Host emulator
host_flashmem_init: Created blank "out/Host/debug/firmware/flash.bin", 4194304 bytes
...
main: >> Starting Sming <<
You can pause the program execution by pressing Ctrl-C. And work further using some further GDB commands. The next paragraph describes some of them.
GDB commands
There are multiple commands supported in GDB and we will mention only some of them.
List current source code
One possibility is to see the source code of the current line where the
execution has stopped. To achieve this you should type list in the gdb
console:
(gdb) list
102 }
103 }
104
105 int main(int argc, char* argv[])
106 {
107 trap_exceptions();
108
109 host_printf("\nWelcome to the Sming Host emulator\n\n");
110
111 static struct {
Break the execution
This command will pause the debugger once it reaches a specific function
or line in the code. This is called breakpoint and can be set like this:
(gdb) break blink
Breakpoint 1 at 0x40105d4c: file app/application.cpp, line 66.
Continue the execution
To continue the execution of the application we can use the continue
command:
(gdb) continue
Continuing.
Breakpoint 1, blink () at app/application.cpp:66
66 {
(gdb)
Because we have set already a breakpoint for the blink function the
execution will be paused when the blink function is reached and from
here you can go to the next line or see the current values of the
variables.
Go to the next line
This can be done using next:
(gdb) next
67 digitalWrite(LED_PIN, ledState);
See variable value
The command to see a value is print followed by the name of the
value. For example to see the value of the ledState variable inside
the blink function we could type:
(gdb) print ledState
$1 = true
You can see more useful commands here.
Or watch the following short video
Debugging with visual debuggers like Eclipse CDT
A good visualization helps us understand things faster. What we can do is use Eclipse CDT and its debugging plugins to do remote debugging as we did from the command line.
Here is how this can be done:
Start Eclipse CDT and import the Live Debug sample:
Select File -> New -> Project -> C/C++ -> Makefile Project with Existing Code
Point Eclipse to the location of the LiveDebug sample
Import the Sming Framework (if you haven’t done it yet)
Import Project
Once the two projects are in Eclipse, set the LiveDebug project to reference the Sming project.
Now create a new Debugging Configuration:
Select Run -> Debug Configurations -> C/C++ Application
Right-click and create a new C/C++ Application
In the Main tab set, set:
Project: Basic_Build
C/C++ Application: out/Host/debug/firmware/app
disable for now the auto build
C/C++ Application
Then go to the Debugger tab and point the GDB debugger to your
gdb binary. (Type make list-config and look for GDB.)
Debugger configuration
We are now ready for debugging. Press the Debug button. (In the screenshot above the Debug button is in the bottom-right corner.) After some seconds your debugging session should be up and running and you can enjoy live debugging.
Live Debugging Session
You will be able to see the current variables and their values. You should be able to go step by step, go inside of functions, add add breakpoints and watchpoints.
Debugging on ESP8266
Required tools and hardware
A debugger is the only requirement for ESP8266. It is usually part of the provided toolchain. Make sure that you have the following executable in your PATH:
xtensa-lx106-elf-gdb
No additional hardware is required (Apart from a standard USB-to-UART adapter).
Recompilation is required
In order to debug applications based on Sming Framework make sure that you are using Sming version 3.8.0 or newer.
Framework
If you want to debug inside of the Sming Framework make sure to
(re)compile it with ENABLE_GDB =1 directive:
cd $SMING_HOME
make dist-clean
make ENABLE_GDB=1
The commands above will re-compile Sming with debug symbols and optimizations for debugging. These commands need to be executed once.
Application
To use, (re)compile your application with the ENABLE_GDB option and flash it to the board. For this example we will use the Live Debug sample application:
cd $SMING_HOME/../samples/LiveDebug
make clean
make ENABLE_GDB=1 # -- recompiles your application with debugging support
make flashapp # flashes ONLY the (re)compiled application
The device will restart then wait for a debugger to be connected:
make gdb
This will start a new debugging session where you can run your code interactively:
Remote debugging using /dev/ttyUSB0
gdbstub_init () at /home/slavey/dev/esp8266.dev.box/dev/Sming/Sming//gdb/gdbstub.cpp:914
914 gdb_do_break();
(gdb)
If the debugger is exited, the application will continue execution as normal. Re-connecting the debugger will pause execution.
GDB commands
There are multiple commands supported in GDB and we will mention only some of them.
List current source code
One possibility is to see the source code of the current line where the
execution has stopped. To achieve this you should type list in the gdb
console:
(gdb) list
909 SD(GDBSTUB_ENABLE_HOSTIO);
910 #undef SD
911
912 #if GDBSTUB_BREAK_ON_INIT
913 if(gdb_state.enabled) {
914 gdb_do_break();
915 }
916 #endif
917 }
918
Break the execution
This command will pause the debugger once it reaches a specific function
or line in the code. This is called breakpoint and can be set like this:
(gdb) break blink
Breakpoint 1 at 0x40105d4c: file app/application.cpp, line 66.
Notice: break sets a software breakpoint. This means that the
blink function must be in IRAM. Otherwise the execution will fail.
If you take a look at samples/LiveDebug/app/application.cpp#L663,
you will see a in the definition of the init function the following
attribute GDB_IRAM_ATTR:
void GDB_IRAM_ATTR init()
This attribute is used to put the init function in IRAM when the
code is compiled with the ENABLE_GDB=1 directive.
Continue the execution
To continue the execution of the application we can use the continue
command:
(gdb) continue
Continuing.
LiveDebug sample
Explore some capabilities of the GDB debugger.
[OS] mode : sta..
...
[OS] cnt
Breakpoint 1, blink () at app/application.cpp:66
66 {
(gdb)
Because we have set already a breakpoint for the blink function the
execution will be paused when the blink function is reached and from
here you can go to the next line or see the current values of the
variables.
Go to the next line
This can be done using next:
(gdb) next
67 digitalWrite(LED_PIN, ledState);
See variable value
The command to see a value is print followed by the name of the
value. For example to see the value of the ledState variable inside
the blink function we could type:
(gdb) print ledState
$1 = true
You can see more useful commands here.
Or watch the following short video
Debugging with visual debuggers like Eclipse CDT
A good visualization helps us understand things faster. What we can do is use Eclipse CDT and its debugging plugins to do remote debugging as we did from the command line.
Here is how this can be done:
Start Eclipse CDT and import the Live Debug sample:
Select File -> New -> Project -> C/C++ -> Makefile Project with Existing Code
Point Eclipse to the location of the LiveDebug sample
Import the Sming Framework (if you haven’t done it yet)
Import Project
Once the two projects are in Eclipse, set the LiveDebug project to reference the Sming project.
Now create a new Remote Debugging Configuration:
Select Run -> Debug Configurations -> C/C++ Remote Application
Right-click and create a new C/C++ Remote Application
In the Main tab set, set:
Project: Basic_Build
C/C++ Application: out/build/Esp8266/Debug/app.out
disable for now the auto build
Remote Debugging Session
Then go to the Debugger tab and point the GDB debugger to your
Xtensa-gdb binary. (Type make list-config and look for GDB.)
Remote Debugging Session
Make sure to load also GDB command file. To find out its location, run make list-config
and look for GDBSTUB_DIR. The file is called gdbcmds, and you may wish to place
a copy of the file somewhere else, especially if you intend to modify it.
You can see the file here Sming/Arch/Esp8266/Components/gdbstub/gdbcmds.
Finally we should configure the remote connection. Go to the Debugger -> Connection tab and set:
type: Serial
device: /dev/ttyUSB0 (or as required for your operating system)
speed: 115200
Set remote connection
We are now ready for debugging. Press the Debug button. (In the screenshot above the Debug button is in the bottom-right corner.) After some seconds your debugging session should be up and running and you can enjoy live debugging.
Live Debugging Session
You will be able to see the current variables and their values. You should be able to go step by step, go inside of functions, add breakpoints to code in RAM or add breakpoints to code that was in FLASH, after it was executed executed at least once.
Debugging on ESP32
Serial debugging
If an exception occurs in debug builds then a prompt will be printed to the serial terminal
such as Entering gdb stub.
As with the ESP8266, if such an exception occurs you can stop the serial debug terminal and type make gdb.
More advanced debugging is available via JTAG if you have the appropriate tools.
See https://docs.espressif.com/projects/esp-idf/en/release-v4.3/esp32/api-guides/fatal-errors.html for further details.
Required tools and hardware
A debugger and a JTAG hardware are required. The debugger is part of the provided toolchain. Make sure that you have the following executable in your PATH:
xtensa-esp32-elf-gdb
Debugging with JTAG is explained in details in the ESP-IDF documentation. Make sure to read it carefully.
For the purposes of this documentation we will be using ESP-Prog JTAG adapter and ESP32-Cam microcontroller from AI-Thinker.
Configure Hardware
The JTAG adapter has to be connected to your ESP32 microcontroller. The following pins from the JTAG adapter have to be connected to ESP32 for the communication to work.
ESP32 Pin
nodeMCU Pin
USB Blaster
JTAG Signal
1
VCC
3V3
4
VCC
2
MTDO / GPIO15
D15
3
TDO
3
MTDI / GPIO12
D12
9
TDI
4
MTCK / GPIO13
D13
1
TCK
5
MTMS / GPIO14
D14
5
TMS
6
GND
GND
2
GND
Running OpenOCD
Once the JTAG adapter is connected to the microcontroller and to a computer we have to start the OpenOCD server that will communicate with the JTAG adapter. For our specific hardware the following command has to be executed:
openocd -f interface/ftdi/esp32_devkitj_v1.cfg -f target/esp32.cfg
If you have configured your JTAG adapter correctly the following messages should show up:
Open On-Chip Debugger v0.10.0-esp32-20190313 (2019-03-13-09:52)
Licensed under GNU GPL v2
For bug reports, read
http://openocd.org/doc/doxygen/bugs.html
none separate
adapter speed: 20000 kHz
Info : Configured 2 cores
esp32 interrupt mask on
Info : Listening on port 6666 for tcl connections
Info : Listening on port 4444 for telnet connections
Info : ftdi: if you experience problems at higher adapter clocks, try the command "ftdi_tdo_sample_edge falling"
Info : clock speed 20000 kHz
Info : JTAG tap: esp32.cpu0 tap/device found: 0x120034e5 (mfg: 0x272 (Tensilica), part: 0x2003, ver: 0x1)
Info : JTAG tap: esp32.cpu1 tap/device found: 0x120034e5 (mfg: 0x272 (Tensilica), part: 0x2003, ver: 0x1)
Info : esp32: Debug controller 0 was reset (pwrstat=0x5F, after clear 0x0F).
Info : esp32: Core 0 was reset (pwrstat=0x5F, after clear 0x0F).
Info : esp32: Debug controller 1 was reset (pwrstat=0x5F, after clear 0x0F).
Info : esp32: Core 1 was reset (pwrstat=0x5F, after clear 0x0F).
Info : Detected debug stubs @ 3ffb42ac on core0 of target 'esp32'
Info : Listening on port 3333 for gdb connections
Info : accepting 'gdb' connection on tcp/3333
Recompilation is required
In order to debug applications based on Sming Framework make sure that you are using Sming version 3.8.0 or newer.
Compilation directives
If you want to debug your application and the Sming Framework code make sure to
(re)compile it with ENABLE_GDB =1 directive:
cd $SMING_HOME/../samples/Basic_Blink
make clean components-clean
make ENABLE_GDB=1
The commands above will re-compile Sming with debug symbols and optimizations for debugging.
Application
To use, (re)compile your application with the ENABLE_GDB option and flash it to the board. For this example we will use the Basic Blink sample application:
cd $SMING_HOME/../samples/Basic_Blink
make clean
make ENABLE_GDB=1 # -- recompiles your application with debugging support
make flashapp # flashes ONLY the (re)compiled application
The device will restart then wait for a debugger to be connected. Before starting the debugger you must be sure that the OpenOCD server is running and listening for incoming connections on localhost port 3333.
Now start the debugger:
make gdb
This will start a new debugging session. The debugger will try to connect to OpenOCD server and in the OpenOCD logs you should see a message similar to the one below:
Info : accepting 'gdb' connection on tcp/3333
Info : Target halted. PRO_CPU: PC=0x4012F7EE (active) APP_CPU: PC=0x4012F7EE
Info : Target halted. PRO_CPU: PC=0x4009171A (active) APP_CPU: PC=0x4012F7EE
Info : Flash mapping 0: 0x10020 -> 0x3f400020, 89 KB
Info : Flash mapping 1: 0x30018 -> 0x400d0018, 388 KB
Info : Target halted. PRO_CPU: PC=0x4009171A (active) APP_CPU: PC=0x4012F7EE
Info : Auto-detected flash size 4096 KB
Info : Using flash size 4096 KB
And in the GDB console you will see a message similar to this one:
Reading symbols from out/Esp32/debug/build/app.out...done.
0x4012f7ee in is_wifi_clk_peripheral (periph=PERIPH_LEDC_MODULE)
at /x/esp-idf/components/driver/periph_ctrl.c:225
225 switch(periph) {
JTAG tap: esp32.cpu0 tap/device found: 0x120034e5 (mfg: 0x272 (Tensilica), part: 0x2003, ver: 0x1)
JTAG tap: esp32.cpu1 tap/device found: 0x120034e5 (mfg: 0x272 (Tensilica), part: 0x2003, ver: 0x1)
esp32: Debug controller 0 was reset (pwrstat=0x5F, after clear 0x0F).
esp32: Core 0 was reset (pwrstat=0x5F, after clear 0x0F).
esp32: Debug controller 1 was reset (pwrstat=0x5F, after clear 0x5F).
esp32: Core 1 was reset (pwrstat=0x5F, after clear 0x5F).
Target halted. PRO_CPU: PC=0x5000004B (active) APP_CPU: PC=0x00000000
esp32: Core 0 was reset (pwrstat=0x1F, after clear 0x0F).
Target halted. PRO_CPU: PC=0x40000400 (active) APP_CPU: PC=0x40000400
Hardware assisted breakpoint 1 at 0x400e1cd3: file /x/Sming/Sming/Arch/Esp32/Components/esp32/startup.cpp, line 21.
(gdb)
If the debugger is exited, the application will continue execution as normal. Re-connecting the debugger will pause execution.
GDB commands
There are multiple commands supported in GDB and we will mention only some of them.
List current source code
One possibility is to see the source code of the current line where the
execution has stopped. To achieve this you should type list in the gdb
console:
(gdb) list
220
221 static bool is_wifi_clk_peripheral(periph_module_t periph)
222 {
223 /* A small subset of peripherals use WIFI_CLK_EN_REG and
224 CORE_RST_EN_REG for their clock & reset registers */
225 switch(periph) {
226 case PERIPH_SDMMC_MODULE:
227 case PERIPH_SDIO_SLAVE_MODULE:
228 case PERIPH_EMAC_MODULE:
229 case PERIPH_RNG_MODULE:
(gdb)
Break the execution
This command will pause the debugger once it reaches a specific function
or line in the code. This is called breakpoint and can be set like this:
(gdb) break blink
Breakpoint 2 at 0x400e1dc4: file app/application.cpp, line 9.
Continue the execution
To continue the execution of the application we can use the continue
command:
(gdb) continue
Continuing.
Target halted. PRO_CPU: PC=0x400E1DC4 (active) APP_CPU: PC=0x4012F7EE
[New Thread 1073483724]
[New Thread 1073514968]
[New Thread 1073494600]
[New Thread 1073487892]
[Switching to Thread 1073412944]
Breakpoint 1, blink () at app/application.cpp:9
9 {
(gdb)
Because we have set already a breakpoint for the blink function the
execution will be paused when the blink function is reached and from
here you can go to the next line or see the current values of the
variables.
Go to the next line
This can be done using next:
(gdb) next
10 digitalWrite(LED_PIN, state);
See variable value
The command to see a value is print followed by the name of the
value. For example to see the value of the ledState variable inside
the blink function we could type:
(gdb) print state
$1 = true
You can see more useful commands here.
Or watch the following short video
Debugging with visual debuggers like Eclipse CDT
A good visualization helps us understand things faster. What we can do is use Eclipse CDT and its debugging plugins to do remote debugging as we did from the command line.
Here is how this can be done:
Start Eclipse CDT and import the Basic Blink sample:
Select File -> New -> Project -> C/C++ -> Makefile Project with Existing Code
Point Eclipse to the location of the Basic_Blink sample
Import the Sming Framework (if you haven’t done it yet)
Import Project
Once the two projects are in Eclipse, set the Basic_Blink project to reference the Sming project.
Now create a new Remote Debugging Configuration:
Select Run -> Debug Configurations -> C/C++ Remote Application
Right-click and create a new C/C++ Remote Application
In the Main tab set, set:
Project: Basic_Build
C/C++ Application: out/build/Esp8266/Debug/app.out
disable for now the auto build
Remote Debugging Session
Then go to the Debugger tab and point the GDB debugger to your
Xtensa-gdb binary. (Type make list-config and look for GDB.)
Remote Debugging Session
Make sure to load also GDB command file. The file is called gdbinit, and you may wish to place
a copy of the file somewhere else, especially if you intend to modify it.
You can see the file here Sming/Arch/Esp32/Tools/gdbinit.
Finally we should configure the remote connection. Go to the Debugger -> Connection tab and set:
type: TCP
host: localhost
port: 3333
Set remote connection
We are now ready for debugging. Press the Debug button. (In the screenshot above the Debug button is in the bottom-right corner.) After some seconds your debugging session should be up and running and you can enjoy live debugging.
Live Debugging Session
You will be able to see the current variables and their values. You should be able to go step by step, go inside of functions, add breakpoints to code in RAM or add breakpoints to code that was in FLASH.
Debugging on RP2040
Currently debugging support on Sming is limited to serial output and inspecting disassembly using GDB.
See https://www.raspberrypi.org/documentation/microcontrollers/raspberry-pi-pico.html
rBoot and OTA updates
Introduction
rBoot is an open source bootloader for the ESP8266, which is now fully integrated with Sming. rBoot is a more flexible alternative to the closed source bootloader supplied by Espressif with the SDK. A bootloader allows you to have more than one application on the esp8266, either completely different apps, or different version of the same app, which can be updated over the air.
The example Basic Ota demonstrates the use of rBoot, but if you want to add it to an existing project this little tutorial will guide you.
Need to know
The esp8266 can only memory map 1MB of flash at a time, your ROM must fit within the a single 1MB block of flash.
Different 1MB blocks can be mapped at different times, so your ROM does not have to be within the first 1MB block. The support for this in rBoot is called big flash mode. A single compiled and linked ROM image can be placed at the same relative position in any 1MB block of flash.
If you have two smaller ROMs in a single 1MB of flash (your only option if you flash is 1MB or smaller) this is referred to as two ROM mode in rBoot. Each ROM needs to be separately linked to have the correct memory mapped flash offset. Examples are given below for common scenarios.
Building
The default rBoot options are perfect for a 4MB flash, so if this is what you have (e.g. an esp-12) you don’t need to do anything else. Otherwise see information below about configuration.
Run
makeas normal, rBoot and your app ROM will both be built for you.Running
make flashwill flash rBoot and the first ROM slot.
Configuring for two ROM mode
If you have a 1MB flash, you will need to have two 512KB ROM slots, both in the same 1MB block of flash. You can accommodate this by setting the appropriate hardware configuration in your project’s component.mk file:
.. code-block:: make
HWCONFIG = two-rom-mode
See Sming/Arch/Esp8266/two-rom-mode.hw for details.
You can copy this and customise it in your project.
SPIFFS
To use SPIFFS think about where you want your SPIFFS to sit on the flash.
If you have a 4MB flash the default position is for the first ROM
to be placed in the first 1MB block and the second ROM to be placed in
the third 1MB block of flash. This leaves a whole 1MB spare after each
ROM in which you can put your SPIFFS. This is the behaviour when you
set HWCONFIG = spiffs in your project’s component.mk file.
If you have a smaller flash the SPIFFS will have to share the 1MB block with the ROM. For example, the first part of each 1MB block may contain the ROM, and the second part the SPIFFS (but does not have to be split equally in half). So for the 4MB example you could put the SPIFFS for your first ROM at flash address at 0x100000 and the SPIFFS for your second ROM at 0x300000; in each case that is the 1MB block after the ROM.
To mount your SPIFFS at boot time add the following code to init:
String name = F("spiffs");
name += rboot_get_current_rom();
auto part = Storage::findPartition(name);
if(part) {
//debugf("trying to mount SPIFFS at %x, length %d", part.address(), part.size());
spiffs_mount(part);
} else {
debug_e("SPIFFS partition missing for slot #%u", slot);
}
Over-the-air (OTA) updates
Instead of insisting on a “one-solution-fits-all” approach, Sming provides you with the ingredients to build an OTA upgrade mechanism tailored to your application. This involves selecting a transport protocol and a backend that interacts with the flash memory. Any protocol from Sming’s rich set of network classes can be used, ranging from raw TCP sockets to HTTP, FTP, MQTT, with or without SSL, etc. To conserve program memory, you might prefer a protocol already employed by your application.
On the backend side, there is OtaUpgradeStream from the Over-the-Air Firmware Upgrade
library, which supports multiple ROM images in one upgrade file, as well as
state-of-the-art security features like a digital signatures and encryption.
Check out the HttpServer Firmware Upload example, which demonstrates a
browser-based firmware upgrade mechanism similar to what is found in many consumer
products.
A more lightweight solution is provided by RbootOutputStream, which
is just a thin wrapper around rBoot’s flash API, in combination with RbootHttpUpdater,
which pulls individual ROM image from an HTTP server.
Tips and Tricks
Reading VCC on ESP8266
If you are running on a battery operated device then function system_get_vdd33() from the official ESP8266 NONOS SDK can help you read the power voltage. For the latter to work properly you should make a small change in your application component.mk file and add vdd to the HWCONFIG_OPTS configuration variable.
If you cannot see such a variable in your component.mk file then append the following line to it:
HWCONFIG := vdd
You can have multiple options selected. They should be separated by comma. For example the command below will add 4MB flash, spiffs and vdd support:
HWCONFIG := 4m,spiffs,vdd
You can check the final hardware configuration using the command below:
make hwconfig
If a custom hardware configuration is needed then read Storage Management.
Minimising Memory Usage
Literals use up memory, so its a good idea to move them to flash. See Program Space and FlashString.
Webpages and Spiffs
FlashString turns out to be very useful for sending web pages, javascript, CSS and so on. Many examples for the ESP8266 exist where a Spiffs file system is used for this purpose, but in fact Spiffs is not ideal. If you want to release a new version of your software, and your web pages are in spiffs, you now have two things to release, so there is double the chance of something going wrong. Plus you have the challenge of preserving any user files while refreshing just a few.
One solution is to use a FlashString hooked up to a FlashMemoryStream instead. In the example below, the CSS file is sent compressed to save time and space. The browser asks for core.js and gets a compressed version:
IMPORT_FSTR(flash_corejsgz, PROJECT_DIR "/web/build/core.js.gz")
void onSendCore_js(HttpRequest &request, HttpResponse &response)
{
response.headers[HTTP_HEADER_CONTENT_ENCODING] = _F("gzip");
auto stream = new FlashMemoryStream(flash_corejsgz);
response.sendDataStream(stream, MimeType::MIME_JS);
}
See FlashString for further details.
Webpages Performance
HTML markup can get quite large and the bigger the file the slower the page loads. One way to deal with that is to remove the white space, this process is called minifying. The downside is that the result is difficult for a human to read. I recommend against it, at least in the early stages of your project.
To support the HTML files there are CSS files and JS files, which must be kept locally on the server if one wants things to work even when the internet is absent.
I use the bootstrap library and the CSS I write goes into another special file. The file count is now three, an HTML file and two CSS files. This is already a lot of files for a microcontroller to deal with especially if it gets download requests for all three at once. A browser will start a download request for each file it sees, and for the ESP, any more than three is a problem, meaning we need to keep this under control.
One way to deal with that is to combine the CSS files together into one.
Next we have JavaScript files which includes the custom code, the bootstrap library and the jquery library. Three extra files. Once again we can deal with these by combining them into one, in which We are back to having 3, one HTML file one CSS file and one JavaScript file.
But the files are big and this is a problem not just because it is slow. The watchdog does not like things to take a long time, and you will almost certainly end up with a timeout.
When a browser asks for a file it doesn’t mind receiving a compressed version using gzip. (Note that you need to add “Content-Encoding/gzip” to the header in the response from the server). Using gzip vastly reduces the sizes of files and it’s well worth doing.
Another size optimisation for CSS files is to remove unused CSS (UNCSS) - I recommend against this as it was too aggressive at removing stuff I really needed - YMMV.
I use gulp to automate the extraction and concatenation and compression of the CSS and JS files, here is the relevant part of my gulpfile.js:
function htm() {
return gulp.src(htmConfig.src)
.pipe(useref())
.pipe(gzip()) // compresses to a gzip file
.pipe(size({ showFiles: true }))
.pipe(gulp.dest('web/build/'))
}
My webpage looks like this
<!-- build:css core.css -->
<link rel="stylesheet" type="text/css" href="bootstrap.css">
<link rel="stylesheet" type="text/css" href="style.css">
<!-- endbuild -->
After gulp runs it looks like this
<link rel="stylesheet" href="core.css">
Upgrading
If you are migrating from version older than 3.8 you should read the release log to find the needed steps to migrate. For newer versions we have dedicated pages.
From v5.1 to v5.2
PartitionStream: Breaking change
The Storage::PartitionStream constructors with blockErase parameter have been deprecated.
The intended default behaviour is read-only, however previously this also allowed writing without block erase.
This can result in corrupted flash contents where the flash has not been explicitly erased beforehand.
The new constructors instead use a Storage::Mode so behaviour is more explicit.
The default is read-only and writes will now be failed.
64-bit time_t
There is some variability in whether time_t is 32 or 64 bits. See issue #2758.
This is dependent on the toolchain and accompanying C library.
32-bits:
Esp32 IDF 4.x
Windows Host (using mingw)
Linux host builds prior to Sming v5.2
Range of time_t:
-0x80000000: 1901-12-13 20:45:52 0x00000000: 1970-01-01 00:00:00 0x7fffffff: 2038-01-19 03:14:07
All others use 64-bit values.
For reference, C library source code can be found here https://sourceware.org/git/?p=newlib-cygwin.git;a=blob;f=newlib/libc/
Rp2040 builds with standard ARM toolkit so probably accommodated by the standard repo.
Espressif toolchains use forks:
esp8266: https://github.com/earlephilhower/newlib-xtensa/blob/xtensa-4_0_0-lock-arduino/newlib/libc/ esp32: https://github.com/espressif/newlib-esp32/blob/esp-4.3.0/newlib/libc/
From v4.7 to v5.1
Command Processing
The CommandProcessing service has been refactored and moved to a component.
This means that the following classes CommandHandler, CommandExecutor and CommandOutput are no longer available.
Enabling
The command processing component used to be enabled by setting the directive ENABLE_CMD_EXECUTOR to 1 in your component.mk file or during compilation.
This has to be replaced with the directive COMPONENT_DEPENDS += CommandProcessing in your component.mk file.
Including Header Files
To include the command processing headers in your C/C++ application we used to do the following
For example:
#include <Services/CommandProcessing/CommandProcessingDependencies.h>
becomes:
#include <CommandProcessing/Utils.h>
Usage
There is no longer a global instance of commandHandler. This means that you will need to create one yourself when you need it. This can be done using the code below:
CommandProcessing::CommandHandler commandHandler;
In order to register a command the old example code:
commandHandler.registerCommand(
CommandDelegate("example", "Example Command", "Application", processExampleCommand));
becomes:
commandHandler.registerCommand(
CommandProcessing::Command("example", "Example Command", "Application", processExampleCommand));
HardwareSerial no longer is dependent on CommandProcessing classes. And the following command will no longer work:
Serial.commandProcessing(true);
The line above has to be replaced with:
CommandProcessing::enable(commandProcessing, Serial);
See the modified samples CommandLine, TelnetServer and HttpServer Websockets for details.
From v4.6 to v4.7
Storage Partition methods
The Storage::Partition::getDevice() method has been removed because
it could be used to bypass protections offered by the partitioning API.
Storage Device Partitions
The CustomDevice class has been removed as it is simpler and more flexible to instead use PartitionTable methods.
The Storage::Device::partitions() method returns a read-only (const) Storage::PartitionTable object
for general use to avoid inadvertent modification.
Use the Storage::Device::editablePartitions() method to make partition table changes.
For example:
part = device->createPartition("archive", Storage::Partition::SubType::Data::fwfs, startOffset, size);
becomes:
part = device->editablePartitions().add("archive", Storage::Partition::SubType::Data::fwfs, startOffset, size);
Custom Partition Types
Creating custom partition types now require use of Storage::Partition::FullType.
For example:
part = device->createPartition("fs_app", Storage::Partition::Type::data, 100, startOffset, size);
becomes:
part = device->editablePartitions().add("fs_app", {Storage::Partition::Type::data, 100}, startOffset, size);
Note how the type and subtype values are enclosed in braces (instantiating a FullType struct).
This avoids confusing the subtype value 100 with the start offset.
From v4.5 to v4.6
SPI Byte/Bit ordering
The SPISettings class has been revised to be consistent with Arduino ESP8266, ESP32, RP2040, etc.
which provide a bitOrder setting but not byteOrder.
Bytes are now always sent LSB first, which corresponds with the system endianness. Bit order simply indicates how the bits within each byte are sent on the wire. This is predominantly MSB first, the default.
The SPISoft class has been upgraded to support bit ordering, transactions
and includes automatic delay calculation for the ESP8266.
See Issue #1428.
From v4.4 to v4.5
Template Streams
The TemplateStream class has been updated to improve tag recognition (Pull Request #2400).
This means regular {varname} tags are sufficient for most purposes, including JSON templates.
The IFS::JsonDirectoryTemplate previously used double-brace tags such as {{varname}}.
It now uses regular tags by default, so if you use this class either:
Add a call to
TemplateStream::setDoubleBraces()in your code, orUpdate your templates to use single braces
Eclipse Project Files
Starting with version 4.5.x we don’t provide Eclipse meta files in our samples.
These can be generated using the ide-eclipse makefile target.
For more information read the updated Using with Eclipse CDT.
Esp8266 toolchain
Sming now requires the ESP Quick Toolchain for building.
Support for the old legacy toolchains (ESP open SDK, UDK) have been dropped. They may still work but are no longer tested.
user_config.h
This header file is part of the original ESP8266 SDK code and is now deprecated. Libraries should use only the necessary headers. Applications do not require it.
Breaking Changes
Undeprecated HttpRequest::getQueryParameter() and removed a lot of old deprecated code.
Removed
WebsocketClient::disconnect(). UseWebsocketClient::close()instead.Removed
TimerDelegateStdFunction. UseTimerDelegateinstead.Removed class
URL. Use classUrlinstead.Removed
TemplateVariables. UseTemplateStream::Variablesinstead.Removed
StreamTransformer::transformCallback. Create inherited class and overrideStreamTransformer::transform()method instead.Removed
StreamTransformer::StreamTransformer(IDataSourceStream* stream, const StreamTransformerCallback& callback, size_t resultSize, size_t blockSize). Instead, create inherited class, overrideStreamTransformer::transform()method and use alternative constructor.Removed
SslValidatorCallback. UseSsl::ValidatorCallbackinstead.Removed
SslSessionIdandSSLSessionId. UseSsl::SessionIdinstead.Removed
SslKeyCertPairandSSLKeyCertPair. UseSsl::KeyCertPairinstead.Removed
SslCertificate. UseSsl::Certificateinstead.Removed
SeekOriginFlags. UseSeekOrigininstead.Removed
eSO_FileStart. UseSeekOrigin::Startinstead.Removed
eSO_CurrentPos. UseSeekOrigin::Currentinstead.Removed
eSO_FileEnd. UseSeekOrigin::Endinstead.Removed
OtaUpgrade::BasicStream::errorToString(). UsetoString()instead.Removed deprecated stuff in Mqtt, including
MQTT_MAX_BUFFER_SIZEandMQTT_MSG_PUBREC.Removed
MqttClient::publishWithQoS(const String& topic, const String& message, int QoS, bool retained, MqttMessageDeliveredCallback onDelivery). Usebool MqttClient::publish(const String& topic, const String& message, uint8_t flags)instead. If you want to have a callback that should be triggered on successful delivery of messages, useMqttClient::setEventHandler().Removed
MqttClient::setCallback(MqttStringSubscriptionCallback subscriptionCallback). UseMqttClient::setEventHandler()instead.Removed
MqttClient::setWill (const String& topic, const String& message, int QoS, bool retained). UseMqttClient::setWill()instead.Removed
MqttMessageDeliveredCallbackandMqttStringSubscriptionCallback. UseMqttDelegateinstead.Removed
IDataSourceStream::length(). UseIDataSourceStream::available()instead.Removed
HttpServer::setDefaultResource(HttpResource* resource). Usepaths.setDefault()instead.Removed
HttpServer::addPath(String path, const HttpPathDelegate& callback),HttpServer::addPath (const String& path, const HttpResourceDelegate& onRequestComplete)andHttpServer::addPath (const String& path, HttpResource* resource). Use paths.set() instead.Removed
HttpResponse::toString(const HttpResponse& res). Use :HttpResponse::toString()method ortoString()global function instead.Removed
HttpResponse::sendTemplate(IDataSourceStream* newTemplateInstance). UseHttpResponse::sendNamedStream()instead.Renamed
commandFunctionDelegatetoCommandFunctionDelegate.Removed
DateTime::convertFromUnixTime(time_t timep, int8_t* psec, int8_t* pmin, int8_t* phour, int8_t* pday, int8_t* pwday, int8_t* pmonth, int16_t* pyear). UseDateTime::fromUnixTime()instead.Removed
DateTime::convertToUnixTime (uint8_t sec, uint8_t min, uint8_t hour, uint8_t day, uint8_t month, uint16_t year). UseDateTime::toUnixTime()instead.Removed
DateTime::fromUnixTime(time_t timep, int8_t* psec, int8_t* pmin, int8_t* phour, int8_t* pday, int8_t* pwday, int8_t* pmonth, int16_t* pyear). Use unsigned version insteadvoid DateTime::fromUnixTime(time_t, uint8_t*, uint8_t*, uint8_t*, uint8_t*, uint8_t*, uint8_t*, uint16_t*).Removed
DateTime::parseHttpDate(const String& httpDate). UseDateTime::fromHttpDate()instead.Renamed
DNSServerclass toDnsServer.Removed
eFO_Append. UseFile::Appendinstead.Removed
eFO_CreateIfNotExist. UseFile::Createinstead.Removed
eFO_CreateNewAlways. UseFile::CreateNewAlwaysinstead.Removed
eFO_ReadOnly. UseFile::ReadOnlyinstead.Removed
eFO_ReadWrite. UseFile::ReadWriteinstead.Removed
eFO_Truncate. UseFile::Truncateinstead.Removed
eFO_WriteOnly. UseFile::WriteOnlyinstead.Removed
eSO_FileStart. UseSeekOrigin::Startinstead.Removed
eSO_CurrentPos. UseSeekOrigin::Currentinstead.Removed
eSO_FileEnd. UseSeekOrigin::Endinstead.Removed
fileList()function. UseDirectoryobject, orfileOpenDir()/fileReadDir()/fileCloseDir()functions.Removed
FileStream::attach(const String& fileName, FileOpenFlags openFlags=File::ReadOnly). UseFileStream::open()instead.Removed
FTPServer. UseFtpServerinstead.Removed
FtpServer::checkUser(const String& login, const String& pass). UseFtpServer::validateUser()insteadRenamed
Hardware_TimertoHardwareTimer.Renamed
HardwareSerial::setCallback(StreamDataReceivedDelegate dataReceivedDelegate)toHardwareSerial::onDataReceived().Removed
HttpClient::request(const String& url). UseHttpClient::createRequest()instead.Removed
HttpConnection::getLastModifiedDate(). UsegetResponse()->headers.getLastModifiedDate()instead.Removed
HttpConnection::getResponseCode(). UsegetResponse()->codeinstead.Removed
HttpConnection::getResponseHeader(const String& headerName, const String& defaultValue). UsegetResponse()->headers[]instead.Removed
HttpConnection::getResponseHeaders(). UsegetResponse()->headersinstead.Removed
HttpConnection::getResponseString (). Use ``getResponse()->getBody()instead.Removed
HttpConnection::getServerDate (). Use ``getResponse()->headers.getServerDate()instead.Removed
httpGetErrorName (HttpError err). UsetoString()instead.Renamed
HttpPartProducerDelegatetype toMultipartStream::Producer.Renamed
HttpPartResulttype toMultipartStream::BodyPart.Removed
HttpRequest::getPath(). Userequest.uri.Pathinstead.Removed
HttpRequest::operator=(const HttpRequest& rhs). UseHttpRequest::clone()instead.Removed
HttpRequest::setPostParameters(const HttpParams& params). Userequest.postParams = paramsinstead.Removed
HttpResponse::hasHeader(const String& name). Useheaders.contains()instead.Removed
HttpResponse::forbidden(). Useresponse.code = HTTP_STATUS_FORBIDDENinstead.Removed
HttpResponse::notFound(). Useresponse.code = HTTP_STATUS_NOT_FOUNDinstead.Removed
HttpResponse::redirect(const String& location). Useheaders[HTTP_HEADER_LOCATION]instead.
From v4.3 to v4.4
Network support
If WiFi is not required then the DISABLE_WIFI setting can be used to reduce code size.
This has a more pronounced effect for the ESP8266 which uses an experimental library.
The core network code has been moved out of Sming/Core/Network and into a separate component at Components/Network.
Some support code has been moved into Core/Data/WebHelpers: applications should still build OK but you will get
a compiler warning advising of the changes.
Note that Network/WebHelpers/aw-sha1.h has been removed in favour of the Cryptographic Support library.
Ethernet support has been added, currently only for the ESP32 embedded MAC.
If WiFi is not used then the DISABLE_WIFI setting can be used to reduce code size.
The DISABLE_NETWORK setting can be used to exclude all networking support for significant code size reduction.
ESP32
If you have made use of Sming’s experimental ESP32 support then you will need to update the ESP IDF SDK and tools to version 4.3. See Sming Esp32 Architecture.
From v4.2 to v4.3
Storage Management
The layout of flash memory has been revised and is now managed via partition table. Hardware configuration is stored in a JSON file (with .hw extension).
If your project has minimal customisation then you may only need to change
the HWCONFIG setting.
You can find full details in the Storage Management library. See also background on Flash memory.
- Removed build targets
- spiffs-image-update
Use the new
buildparttarget instead- spiffs-image-clean
Use the new
part-cleantarget instead
- New and updated build targets
- hwconfig
Displays the current configuration in JSON format
- hwconfig-list
Show available hardware configs
- hwconfig-options
Show available hardware configuration options
- map
Display the current flash memory map
- readmap
Read partition table from device and display the map
- readpart
Read contents of partition into a file, e.g.
make readpart PART=spiffs0will createout/Esp8266/debug/spiffs0.read.bin- flash
Flash partition table and all partitions. Previously it was necessary to run
make flashinitto write system parameter information. Failure to do this was a common problem and should now be a thing of the past.- flashinit
This now just erases the flash memory. The ESP8266 initialisation data gets written when running
make flash.- flashmap
Flash just the partition map
- flashpart
Flash a single partition, e.g.
make flashpart PART=spiffs0- erasepart
Erase a partition, e.g.
make erasepart PART=spiffs0- buildpart
Re-builds images associated with partitions, such as SPIFFS or other filesystem images.
- part-clean
Removes any partition images with build information. This is done as part of a normal project clean.
- Configuration variables
A number of configuration variables have been removed or made read-only, as these are now generated from the Hardware configuration.
SPI_MODE,SPI_SIZE,SPI_SPEEDThe variables are still used internally but are read-only; they cannot be set at the command line. Values are read from the hardware configuration under
/devices/spiFlash.RBOOT_ROM0_ADDR,RBOOT_ROM1_ADDR,RBOOT_ROM2_ADDRUsed by rBoot, and are now read-only. Values are read from the
addressproperty ofrom0-2in the hardware configuration.RBOOT_SPIFFS_0,RBOOT_SPIFFS_1Removed.
SPIFF_SIZERemoved. Attempting to set this automatically within a hardware configuration is liable to cause more problems than it solves, so updating the hardware config is the now only way to change this setting.
SPIFF_FILES[deprecated]
You can still use this to specify the source location for the primary SPIFFS partition (spiffs0). The preferred method is to set the
filesproperty in a partitionbuildkey.The default SPIFFS partition settings can be overridden in a custom profile. For example:
{ ... "base_config": "spiffs", "partitions": { "spiffs0": { "size": "128K", "build": { "files": "some/other/folder" } } } }
Installable File System (IFS)
Sming now supports multiple filesystems via Installable File System.
See Basic IFS for a demonstration.
Core/FileSystem.h has been modified to use IFS but the API remains largely unchanged, although somewhat expanded. Functions are now mainly just wrappers around filing system calls.
A single global IFileSystem instance is used.
- SPIFFS
All access is now managed via the
IFS::SPIFFS::FileSystemimplementation.Applications should not use SPIFFS functions directly.
Important
SPIFFS is now built with
SPIFFS_OBJ_META_LEN=16to store extended attribute information. Existing volumes built with other values will not be directly compatible; the file listing may be correct but file contents will not.To accommodate use of existing pre-built SPIFFS images,
SPIFFS_OBJ_META_LENhas been added:make SPIFFS_OBJ_META_LEN=0
You will, however, lose the additional file information (such as modification time).
- File open flags
e.g. eFO_ReadOnly. These will still work but are now deprecated and should be replaced with their C++ equivalent such as
File::ReadOnly.
From v4.1 to v4.2
Summary
Stream methods
The Stream::readBytes() has been virtualised and overridden for IDataSourceStream
descendents for more efficient operation, especially with ArduinoJson.
For normal read operations where the stream position is to be updated, applications should use
this method in preference to IDataSourceStream::readMemoryBlock().
An addition method IDataSourceStream::moveString() has been added to support extracting
the content of a memory-based stream into a String object without duplicating the data.
This is supported by LimitedMemoryStream and MemoryDataStream.
Stream / file seeking
To provide a consistent interface SeekOrigin is now used with all seek methods for streams,
and also by file system functions. Mappings are:
SeekOrigin::Start instead of eSO_FileStart
SeekOrigin::Current instead of eSO_CurrentPos
SeekOrigin::End instead of eSO_FileEnd
These map to the standard C SEEK_SET, SEEK_CUR and SEEK_END but as SeekOrigin is strongly typed
it offers compile-time checking, and has a toString(SeekOrigin) overload.
getBody methods
The HttpRequest::getBody() and HttpResponse::getBody() methods have been revised to use
move semantics. Previously, the data was copied into a new String which effectively doubled memory usage.
If you have set a non-memory stream type (e.g. FileStream) which does not implement IDataSourceStream::moveString()
then an invalid String will be returned. In this situation you should use HttpResponse::getBodyStream() instead.
ContentType / MIME types
String toString(MimeType)() has been moved out of the ContentType namespace,
so no longer requires the ContentType:: qualifier.
From v4.0 to v4.1
Summary
With Sming version 4.1 there are some backwards incompatible changes. This page is provided to help with migrating your applications.
SSL
In version 4.1 one can choose between different SSL implementations. At the moment Sming supports axTLS and BearSSL for creating SSL enabled clients and servers.
In order to allow multiple SSL adapters and seamless integration the library code had to be refactored and that introduced some breaking changes.
Detailed documentation can be found in SSL: Secure Sockets Layer.
See SSL: Upgrading for a migration guide.
MultipartParser
The MultipartParser component has been decoupled from the framework and converted into a Library. In the process, the former config option ENABLE_HTTP_SERVER_MULTIPART has been removed. Therefore, in your components.mk, replace:
ENABLE_HTTP_SERVER_MULTIPART := 1
by:
ARDUINO_LIBRARIES += MultipartParser
Also, the body parser for handling HTTP requests of content-type multipart/form-data must now be registered explicitly by the application code:
#include <MultipartParser.h>
HttpServer server;
...
server.setBodyParser(MIME_FORM_MULTIPART, formMultipartParser);
From v3.8 to v4.0
Summary
With Sming version 4.0 there are some backwards incompatible changes. This page is provided to help with migrating your applications.
In particular, the build system has changed so the process is now driven from the project directory.
See Sming build system for detailed information and a list of known issues.
Header files
The folder structure has been revised to provide support for additional architectures, such as the Host Emulator, and in the future the ESP32.
The Sming/Arch directory should never be accessed directly: it
contains specific files for the target architecture, and may provide
different header file versions to the main ones. Instead, consider these
directories to be your ‘root’ include directories:
Sming
Sming/Components
Sming/Core
Sming/Wiring
Examples of #include statements:
Old-style |
Recommended |
|---|---|
|
|
|
|
|
|
|
|
|
|
Changed Headers
If you use some of the includes below directly in your application make sure to apply the following changes:
Description |
Old name |
New name |
|---|---|---|
uart driver |
|
|
flesh memory |
|
|
C compatible types |
|
|
user_include.h
This file is generally #included ahead of everything else so that common architecture-specific definitions are available. Unless you’ve made changes to the file, it is not required and you should delete it: Sming provides a default which will be used.
If you have made customisations, please amend the file as follows:
#pragma once
#include_next <user_config.h>
<< User Customisations here >>
If you’re using #include <SmingCore.h> then you don’t need
#include <user_config.h> as this is included automatically.
Application Makefile
Rename
Makefile-user.mkfile tocomponent.mk.Replace the file
Makefilewith the one from theBasic_Blinksample project. If you’ve ignored the instructions and modified the file (!) then you’ll need to move those changes into your newcomponent.mkfile instead.Sming uses the
#pragma oncestatement for header guards, so consider updating your own files if you’re not already doing this.
Arduino Libraries
Your project must specify which Arduino Libraries it uses (if any). Do
this by setting ARDUINO_LIBRARIES in your project’s
component.mk file. For example:
ARDUINO_LIBRARIES := OneWire
This change means only the libraries you require for a project need to be built.
The following libraries have changes in their API:
JSON
ArduinoJson is now an optional Component, so you need to make a couple of changes to use it:
Add
ArduinoJson6to theARDUINO_LIBRARIESvariable in your project’scomponent.mkfile. (e.g.ARDUINO_LIBRARIES = ArduinoJson6) To support migration of existing projects, you can elect to continue using version 5 by specifyingArduinoJson5instead.Add
#include <JsonObjectStream.h>to your source code. If you’re not using the stream class, add#include <ArduinoJson.h>instead.
See library documentation for further details:
WiFi Classes
Additions
New class to handle MAC addresses: Sming/Wiring/MacAddress.h.
Sming/Platform/Station.h
getConnectionStatusName() returns String instead of char*
EStationConnectionStatus renamed to StationConnectionStatus
Sming/Platform/WifiEvents.h
Callback handler parameter lists have changed
Hardware Timers
The following functions have been removed from HardwareTimer.h:
usToTimerTicks
timerTicksToUs
Their presence is confusing. The ESP8266 has two hardware timers:
- Timer1:
A 23-bit count-down timer accessed via the HardwareTimer class. Use the usToTicks and ticksToUs methods.
- Timer2
A 32-bit counter accessed bia the ElapseTimer class. The current tick value is obtained using the NOW() macro. The clock frequency is defined by HW_TIMER2_CLK (in driver/hw_timer.h) and depends on the USE_US_TIMER compiler flag.
Deprecated / Changed types
Deprecated types will generate a compiler warning. See Deprecated List.
Experimental Stuff
Enabling SSL in HttpServer
At the moment any TCP based server in Sming can use TLS/SSL.
That applies to HttpServer (also with Websockets).
But make sure to read the security considerations and limitations.
Enabling SSL in HttpServer
The listen method in the TcpServer, and the child HttpServer class, accepts a second optional
parameter. If you look at the original code:
samples/HttpServer_WebSockets/app/application.cpp#L95-L99.
That can be changed to something like:
void startWebServer()
{
// TODO: Make sure to set a server certificate and key
server.listen(443, true);
And what is left is the actual setting of the server certificate:
void startWebServer()
{
// Assign the certificate
server.setSslInitHandler([](Ssl::Session& session) {
session.keyCert.assign(serverKey, serverCert);
});
server.listen(443, true);
The final code can be something like:
void startWebServer()
{
#ifdef ENABLE_SSL
server.setSslInitHandler([](Ssl::Session& session) {
session.keyCert.assign(serverKey, serverCert);
});
server.listen(443, true);
#else
server.listen(80);
#endif
server.paths.set("/", onIndex);
//...
Security Considerations
Does it really make sense to use SSL for an HttpServer on an ESP8266 device?
The certificate/private key pair should make it impossible for an external user to decrypt your traffic so that the things that you sent are kept private, but there are some complications with this:
The private key will not stay private for long. The private key should be kept encrypted on the flash memory, to prevent casual reading. But even with decryption there is a high probability that someone will be able to disassemble your application and figure out how to decrypt the key.
Costs for certificate. Let’s imagine that you have overcome the first issue. Then comes the second issue - if you want your users to accept the certificate it has to be signed by one of the trusted certificate authorities. And that costs money. And if you want to use a unique certificate/private key pair for every device than it will make things worse, moneywise. Note: Free SSL certificates are now available, for example https://letsencrypt.org/. These will expire if not kept up to date so adds additional complexity to your application.
You can handle up to 2 or maximum 3 connections. SSL needs 16K of memory to make the initial handshake. The memory consumption after a successful handshake can decrease to 4K, just for the SSL, per request. But realistically this means that you will end up with a server that can handle maximum 2 or 3 simultaneous connections before the heap memory is consumed and has to be released.
Therefore, in our humble opinion, it would be better to rely on the WIFI security that your Access Point (AP) provides and make this AP accessible only for your IoT devices.
Signed OTA Updating
Introduction
Deploying embedded devices with (automatic) OTA functionality introduces new risks to local networks and the whole internet. If an attacker takes over the update server or runs a MITM attack, he might be able to turn the devices into zombies.
To prevent this, you can either provide a secure connection between device and update server (e. g. VPN or TLS) or add a cryptographic signature to all OTA files. Pull Request #893 provides hooks to the OTA functionality to allow checking of such signatures.
A proven method for this is, for example, ECDSA in conjunction with SHA-256. For both steps libraries are available (micro-ecc and Arduino Cryptosuite).
To use it, you can subclass RbootOutputStream like this:
#define PREFIX_MAGIC 0x54, 0x49, 0x55, 0x53
#define PREFIX_TYPE 0x00, 0x01
#define PREFIX_SIZE sizeof(_my_prefix)
#define SIGNATURE_SIZE 64
const u8 _my_prefix[6] = { PREFIX_MAGIC, PREFIX_TYPE };
struct MyHdr {
u8 prefix[PREFIX_SIZE];
u8 signature[SIGNATURE_SIZE];
};
//-----------------------------------------------------------------------------
class MyStream : public RbootOutputStream {
public:
MyStream(uint32_t startAddress, size_t maxLength = 0): RbootOutputStream(startAddress, maxLength)
{
// do some initialization if needed.
}
size_t write(const uint8_t* data, size_t size) override;
bool close() override;
virtual ~MyStream()
{
delete sha256;
}
protected:
bool init() override;
private:
Sha256 *sha256 = nullptr;
u8 hdr_len;
MyHdr hdr;
};
//-----------------------------------------------------------------------------
bool MyStream::init() {
RbootOutputStream::init();
delete sha256;
sha256 = new Sha256;
hdr_len = 0;
}
size_t MyStream::write(const uint8_t* data, size_t size) {
// store header
u8 missing = sizeof(hdr) - hdr_len;
if (missing) {
if (size < missing) missing = size;
memcpy( &hdr, data, missing );
size -= missing;
data += missing;
hdr_len += missing;
// check prefix
if ( hdr_len >= PREFIX_SIZE ) {
if ( memcmp(hdr.prefix, _my_prefix, PREFIX_SIZE) ) {
debugf("invalid prefix");
return 0;
}
}
}
// update hash
sha256->update(data, size);
// save data
return RbootOutputStream::write(data, size);
}
bool MyStream::close() {
if (!RbootOutputStream::close()) {
return false;
}
u8 hash[SHA256_BLOCK_SIZE];
sha256->final( hash );
bool sig_ok = /* add signature check here */;
if (!sig_ok) {
debugf("wrong signature");
// TODO: if needed delete the block at the startAddress
return 0;
}
return 1;
}
And then in your application you can use your MyStream with the following setup:
RbootHttpUpdater* otaUpdater = new RbootHttpUpdater();
MyStream* stream = new MyStream(1234); // Replace 1234 with the right start address
otaUpdater->addItem(ROM_0_URL, new MyStream()); // << the second parameter specifies that your stream will be used to store the data.
// and/or set a callback (called on failure or success without switching requested)
otaUpdater->setCallback(OtaUpdate_CallBack);
Simulator: Wokwi
At the moment Sming and Wokwi have experimenta support for the Esp32 architecture.
Only VS Code IDE is supported
The Esp32 simulator lacks certain features and may encounter failures; hence, it is recommended for use with simple applications.
Wokwi Support into VS Code
To integrate Wokwi support into your VS Code, run the following command from your application’s root directory:
make ide-vscode ENABLE_WOKWI=1 SMING_ARCH=Esp32
Install the recommended extensions
Ensure that you have installed the newly recommended extensions. If none appear, manually install the wokwi.wokwi-vscode extension.
Merge all partitions to a single flash file
From the root directory of your application, run the following command to consolidate all partitions for the simulator and flash them into a single file:
make mergeflash SMING_ARCH=Esp32
Usage
Basic_Blink
The Basic_Blink sample can be compiled and executed directly on a simulator before deploying it to a physical microcontroller. Follow the commands below to get started:
cd $SMING_ARCH/../samples/Basic_Blink
make ide-vscode ENABLE_WOKWI=1 SMING_ARCH=Esp32
make mergedflash
Once the compilation is complete, open the folder in VS Code, install the recommended extensions, and either open the diagram.json file or press F1 and type Wokwi.
From the options, choose to start the Wokwi simulator.
Debugging
Running the Basic_Blink sample in the simulator enables you to debug it directly in VS Code.
Set a breakpoint in the init function in the Basic_Blink app/application.cpp file.
Press F1 and select “Start Simulator and Wait for Debugger.” In the Launch configurations, choose “Wokwi GDB” and click the play button.
This initiates a new debugging session, allowing you to debug the code running in the simulator.
Diagram Editor
The diagram.json file, which includes elements and their connections, can be edited on the Wokwi official website.
You can add new elements such as extra LEDs or servos. Ensure to copy the modified contents of the diagram.json from the website to your local environment.
Contributing
How you can help
You can contribute to Sming by
Providing Pull Requests with new features, bug fixes, new ideas, etc.
Testing our latest source code and reporting issues.
Supporting us financially to acquire hardware for testing and implementing or out of gratitude.
See Developing for Sming for details.
Financial contributions
We welcome financial contributions in full transparency on our open collective page. They help us improve the project and the community around it. If you would like to support us you can become a backer or a sponsor.
In addition to that anyone who is helping this project can file an expense. If the expense makes sense for the development of the community, it will be “merged” in the ledger of our open collective by the core contributors and the person who filed the expense will be reimbursed.
Backers and sponsors
Thank you to all the people who have backed Sming or sponsored it.
Tools Integration
Details of Sming support for Integrated Development Environments (IDEs).
Using with CLion
Note
This information is no longer current.
However, if you are able to set up an external make project with CLion that should work.
To get intelliense (or whatever the equivalent for CLion is) will require setting up a list of the correct #include paths. For vscode we have a tool to handle that:
run
make ide-vscodefrom a sample project directoryexamine generated path list in
.vscode/c_cpp_properties.json
Something similar could be done for CLion perhaps?
Developing with the Sming framework can also be done in CLion.
Copy and paste Makefiles into the project
Create
appfolder and addapplication.cppthereEdit
CMakeLists.txt
## Edit CMakeLists.txt
include_directories("/opt/Sming/Sming")
include_directories("/opt/Sming/Sming/Libraries")
include_directories("/opt/Sming/Sming/system/include")
include_directories("/opt/esp-open-sdk/sdk/include")
set(SOURCE_FILES app/application.cpp)
add_executable(Pathfinder ${SOURCE_FILES})
Build the project using terminal
make && make flash
Using with Eclipse CDT
The Eclipse CDT Project provides a fully functional C and C++ Integrated Development Environment based on the Eclipse platform. Eclipse is a free (as in “free beer”) and Open Source code editor for Windows, Linux and Mac.
For easier integration make sure you have both ESP_HOME and
SMING_HOME exported in your working environment.
Software involved
Installation
Install Eclipse CDT using your operating system packaging tools.
Configuration
First you should import the Sming project.
This can be done by going to menu File -> Import and from there choosing
General -> Existing project into Workspace. When asked for the location of the project
point Eclipse to the location of your SMING_HOME folder.
Once you have imported the Sming project you can import some of the samples. You need to open in terminal the location of the sample of your interest and type:
cd /path/to/sample/
make ide-eclipse
This will create the minimum required meta files for an Eclipse CDT project which you can import into your Workspace.
/path/to/sample/ should be replaced with the actual location of the sample.
For the Basic_Blink sample we can use the following commands:
cd $SMING_HOME/../samples/Basic_Blink # Replace $SMING_HOME with %SMING_HOME% if you use Windows
make ide-eclipse
After this you go back to your Eclipse and run again File -> Import and from there choose General -> Existing project into Workspace.
Then select the Basic_Blink directory under samples. Once the project is imported and the Eclipse CDT indexes
are build you will be able to program like a pro and have code completion and debug your code.
For the latter see Live Debug.
You can import also your own Sming-based applications in Eclipse. If your application is not containing .project and .cproject files inside its root folder then you can create such files by going to the root folder of your application and then typing the command below:
make ide-eclipse
Using with MS Visual Studio Code
Microsoft Visual Studio Code is a free (as in “free beer”) and Open Source code editor for Windows, Linux and Mac.
Software involved
Installation
Install VS Code, extensions and tools
Navigate to project folder and create configuration as described below
Open workspace. If vscode is in your system path you can do this:
code .
Configuration
Please make sure you have critical environment variables set globally before starting vscode. See Configuration for details.
One of the strengths of vscode is the use of well-documented configuration files. You can find comprehensive documentation for these online.
However, setting these up is time-consuming so the build system can create them for you. The vscode workspace root directory is your project directory.
Change to your project directory (e.g. samples/Basic_Blink) and run these commands:
make ide-vscode SMING_ARCH=Esp8266
make ide-vscode SMING_ARCH=Host
Now open the workspace in vscode, and open a source file (.c, .cpp, .h). You should now be able to select the architecture from the icon in the bottom-right corner:
VS Code language selection
A selection of tasks are provided which you can view via Terminal -> Run Task.
To debug your application, follow these steps:
Select the appropriate architecture (e.g. Host, Esp8266)
Select
Terminal->Run Task->Full rebuild (with debugging)Confirm that the baud rate (
COM_SPEED_GDB) and port (COM_PORT_GDB) are set correctly:make gdb SMING_ARCH=Esp8266 COM_PORT_GDB=/dev/ttyUSB0 COM_SPEED_GDB=115200
Update the vscode configuration:
make ide-vscodeIn vscode, select the require ‘Run’ task:
VS Code debug selection
Manual configuration changes
When you run make ide-vscode the configuration files are actually generated using a python script
Tools/vscode/setup.py. Configuration variables are passed from the project makefile.
If you make any changes to the configuration files, please note the following behaviour:
The
Host,Esp32orEsp8266intellisense settings will be overwritten.The
Esp8266 GDBandHost GDBlaunch configurations will be overwrittenThe
sming.code-workspaceand.vscode/tasks.jsonfiles will be created if they do not already exist. To re-create these they must first be deleted.
Ideally the vscode configuration files should not need to be kept under configuration control, but generated when required.
Some settings are necessarily configured via the setup.py script, however most settings can
be changed by editing the files in Tools/vscode/template.
If you do this, remember to keep a copy as they’ll be overwritten otherwise.
And, please consider contributing any changes or suggestions to the community!
Known issues / features
The vscode configuration files are only updated when you manually run
make ide-vscode. If you update change critical build variables or add/remove Components to your project, you may need to run it again to update them.When running
make ide-vscode, comments in the configuration files will be discarded.make ide-vscodemay overwrite parts of your configuration: be warned!When debugging for esp8266 output in the console is not formatted correctly. Lines appear with @ in front of them.
You may find vscode uses powershell instead of cmd.exe to execute tasks. Sming should work OK with either, but you can change this in the sming.code-workspace file via the
terminal.integrated.defaultProfile.windowssetting.
Troubleshooting
Troubleshooting Windows
If something goes wrong - don’t worry, community will be able to help you out. Don’t forget to check User questions before posting a github issues. Maybe someone else had a similar issue!
If nothing found, please make sure to provide all required information
when posting issues. Here’s the minimum that you will need to get:
Start cmd.exe and provide output of the following commands:
echo %PATH%
echo %SMING_HOME%
echo %ESP_HOME%
dir %SMING_HOME%
dir %ESP_HOME%
where make
Common issues & solutions
SMING_HOMEshould be set toc:\tools\sming\Sming, with\as a path separator, and NO backslash\at the end.ESP_HOMEshould be set toc:\Espressif, using\as a path separator, and NO backslash\at the end.MinGW paths should be at the start of PATH environment variable (before other items).
If you update your sming-core source don’t forget to do
cd c:\tools\sming\Sming && make clean && make
Random Restarts
First try setting the baud rate to 74880. Example for Linux:
python -m serial.tools.miniterm /dev/ttyUSB0 74880
The random symbols should become readable messages.
If you see repeating messages containing rf_cal[0] !=0x05 then most
probably you should initialize the flash memory on your ESP8266
device.
To achieve this do the following:
Check the Hardware configuration especially
flash_sizesetting.
Note
If you’re not sure what size the flash memory is on your device, run make flashid to check.
Run
flashinit:cd $SMING_HOME/../samples/Basic_Blink make flashinit
This resets the flash memory to a default state, erasing any existing application, configuration or data stored there.
Re-program your device:
make flash
Verify flash data has been written successfully
make verifyflash
Sample Compilation
The first thing that you need to do is to make sure that you have a clean source code. And second if the sample is still not compiling you have to provide us with more information.
Let’s start with the first part: “Clean Source Code State”. If you are
familiar with git you can run git status to get more
information. Sometimes this won’t be enough therefore we recommend you
the following steps ( They are working on Linux with bash shell. If you
have another OS and shell you should adjust them accordingly).
cd /tmp
git clone https://github.com/SmingHub/Sming.git SmingForTest
cd /tmp/SmingForTest/Sming
export SMING_HOME=/tmp/SmingForTest/Sming
The commands above will fetch the latest source code in a directory that should be completely different than the Sming directory that you are using on a daily basis. Also it points the SMING_HOME variable to the new temporary directory with the clean source code.
Now let’s go to the second step: “Compile a sample and report issues, if any”. We will use the Basic_Ssl sample. Before we compile a sample we need to compile the Sming library. This can be done calling the following commands:
cd /tmp/SmingForTest/Sming
export SMING_HOME=/tmp/SmingForTest/Sming
make clean
The last makes sure to clean any intermediate files or directories. If
we run now make. It will fetch the needed submodules, compile the
code and build a library out of it. In our case we need to compile Sming
with an optional SSL support. In order to compile Sming with SSL we need
to add the ENABLE_SSL =1 directive. The command that we need to run now
will look like this:
make ENABLE_SSL=1
On error If the compilation stops with an error, please, copy the output that the command above produces and append it to your bug report. Now run the same command one more time but with the V=1 directive. This will produce a more verbose output that can help the Sming developers figure out the issue.
make ENABLE_SSL=1 V=1
Make sure to append that output too to your bug report. Tell the Sming developers also what is your SDK (esp-open-sdk, esp-alt-sdk, …) and version, operating system & version, git version, make & version, so that the developers can reproduce your problem.
On success It is time to compile the Basic_Ssl sample. Do this by executing the commands below:
cd /tmp/SmingForTest/samples/Basic_Ssl
export SMING_HOME=/tmp/SmingForTest/Sming
make clean
make
On error 2 If that compilation fails make sure to append the output to your bug report. Now compile the sample with the V=1 flags, similar to the compilation of the Sming library.
cd /tmp/SmingForTest/samples/Basic_Ssl
export SMING_HOME=/tmp/SmingForTest/Sming
make V=1
Check User questions before posting a github issues. Maybe someone else had a similar issue!
About
Sming
Sming is an asynchronous C/C++ framework with superb performance and multiple network features. Sming is open source and is tailored towards embedded devices. It supports multiple architectures.
The project was started in 2015 and is actively developed.
Arduino
Arduino is an open-source electronics platform based on easy-to-use hardware and software. Sming is compatible with (most) standard Arduino libraries, which means that you can use any popular hardware in few lines of code.
ESP8266
ESP8266 is a microcontroller with Wi-Fi, manufactured by Espressif Systems. It is the first microcontroller that was supported from Sming. Sming provides access to all ESP8266 functions such as GPIO, timers, WiFi configuration, etc.
ESP32
ESP32 is the second microcontroller by Espressif Systems. There are also a number of more recent variants such as the esp32-s2 and esp32-c3.
Sming currently provides support for these devices but is more limited and should be considered ‘experimental’.
Licenses
- Sming Core
LGPL v3
- Espressif SDK
ESPRESSIF MIT License (with some closed-source blobs)
- Libraries
See each library for details of its own open source license
Comments
We try not to split comment lines into smaller ones and also we add one space between code and trailing comment: