![]() In many cases the user just wants to make a kernel configuration change. The majority of users that are interested in building their own kernel are doing so because they have installed Ubuntu on their system and they wish to make a small change to the kernel for that system. This page describes how to build the kernel. Obtaining the source for an Ubuntu release.Of course, if you want to, you can use your own toolchain. Vagrant gives you a simple starting point without worrying too much about the whole toolchain situation. MacOS on the other hand can make trouble in my experience. Windows with WSLv2 also works very well, since it give you a Linux kernel. The toolchain for Linux is very simple to install and to use. Is this just to simplify the compilation of the source code, or is vagrant/Docker a typical part of the RIOT developer's dev/test flow? In your introductory examples (wiki) I see vagrant being used. You are free in what structure you want to use. For your own application: Just take a look into tests and copy the way of doing it to your own application. But for the start, I would say just executing them manually is good enough. As I mentioned, there are a couple of tools for that. You can, of course, automate this as well. How are RIOT apps typically tested? Do I just have to deploy it to the MCU and take the whole system up for a test spin, or are there ways of simulating/virtualizing the test phase?įor the kernel: Write some tests in the test directory. Sometimes not well documented, but an amazing set of tools.ģ. For commercial applications this looks probably different and you can make use of the CD/CI pipeline of RIOT OS itself. There is no big pipeline involved, since I only need one, maybe two devices with the application. I just have my MCUs here, write some apps, compile them and flash them. My hobbyist development pipeline: I barely think about a pipeline. So, you can ship your app logic as binary, while the RIOT OS core has to be open under LGPLv2.1 license. The LGPLv2.1 license allows you to do that. ![]() So, you are also able to ship your application as binary. You will see that there are a lot of different binaries in there. But just take a look into the bin directory after compiling your source code. It just doesn't make sense, if you work directly on the metal. I mean you maybe could implement dynamic linking on Microcontroller, but the overhead is probably not resonable. Yep, this is the normal process for firmware development. My understanding is that my compiler/linker will compile RIOT and my app TOGETHER into the same executable that can be deployed to my MCU - is this true? What are typical hobbyist deployment pipelines/pathways to get the final executable deployed/installed to the target MCU? I never did that and just looked into the API doc.Ģ. The configuration in CLion is a bit more complext though. ![]() It generates the necessary C/C++ Plugin configuration, so that autocomplete works out of the box. Unfortantly my CLion subscription expired and I had to switch to visual studio code ^^Įven though I really like CLion, the RIOT OS extension ( ) for Visual Studio Code helps in development. I just did exactly that until some weeks ago. You can use it with you CLion/GCC/CMake stack. Can I develop RIOT-based C applications using my existing CLion/GCC/CMake/Mac "stack"? Or does RIOT have its own IDE that I should be using on Mac? Is my "stack" a typical dev setup or does it send up any red flags in anyone's mind? Thanks in advance for any-and-all help here!ġ. Is this just to simplify the compilation of the source code, or is vagrant/Docker a typical part of the RIOT developer’s dev/test flow? How are RIOT apps typically tested? Do I just have to deploy it to the MCU and take the whole system up for a test spin, or are there ways of simulating/virtualizing the test phase? My understanding is that my compiler/linker will compile RIOT and my app together into the same executable that can be deployed to my MCU – is this true? What are typical hobbyist deployment pipelines/pathways to get the final executable deployed/installed to the target MCU? I would like to make some simple hobby projects using my own tiny C application + RIOT running on some type of MCU, driving some simple electronics (LEDs, speakers, motors, etc.).Īfter looking over your wiki there’s a few pieces of the puzzle I’m still missing here, and would appreciate any insight or course correction!Ĭan I develop RIOT-based C applications using my existing CLion/GCC/CMake/Mac “stack”? Or does RIOT have its own IDE that I should be using on Mac? Is my “stack” a typical dev setup or does it send up any red flags in anyone’s mind? I have done some toy C programming before and currently have a CLion/GCC/CMake stack setup on my Mac laptop.
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |