Configure VSCode for Embedded project (STM32) with Make | VIDEO 43

[@klebolek]

I've been looking for tutorial to finally switch from CubeIDE and try to understand how whole build process works and i found this video. Nice one, thank you!

[@ZiggyPebble]

These are great tutorials. Exactly what I was looking for. Keep the great content coming!

[@hongeric3520]

hi，some question occuerd when i use stm32H755ZI , there is no makefile Options in cubemx ，please give me some advices ,thx

[@maciej_sliv]

Do you might know any way of tracing FreeRTOS tasks in VSC?

[@Netryon]

That's why should use already configured payed solution. Thought you be db coding, but that turned out to be just configuring. My only warning here is Littlefuse. It's already a technology you may use 0:56 very easy and interesting. Problem with these have to write your make file and do more configuring, when only thing should see is some basic running project not diving deep into just meaningless code lines. Just Copy it - may not know security vulnerabilities or if it really deletes a folder you need. Wish to do it in the description area this , but that is not how it runs escaping all the < >.

[@jr8699]

Hi, can not see the first option at bulb for Add to "includePath" - what is wrong?

[@leondong285]

It's a fantastic video.

[@MrSerlui17]

Great tutorial just what I was looking for, I still don't understand a lot of thing of how does VScode work but this was all well explained for me

[@中国动漫-o3g]

I like your videos. Thank you very much, it helped me a lot

[@smaleky09]

Thank you very much! Very helpful. God Bless.

[@jrioublanc]

Thanks, very valuable video

[@shahidkhan-md3dw]

Tell me about stm32 based products. because I want to work with stm32

[@abhinavlal3252]

So, for VS code we have to do those settings for every project?

That's right. It have a template .vscode folder that I copy to a new project, change name of the microcontroller and then open vscode. It doesn't take that much time and works great.

[@ericscoerg3763]

Where's that template? @

[@qualia_]

Great channel, keep going!

[@samc4499]

I have a question about includepath. Isn't the /** in the ${workspaceFolder}/** recursively search for all the sub-folders inside the workspace directory? Why do wee still need to add those sub folder one by one?

You are correct. But for some reason this can fault and not work. I found a couple articles and posts about this and the solution is to add all paths. If ${workspaceFolder}/** works for your project then no problem. You can use it for your compiler paths as well.

[@BenBilesBB-box]

whatever I do I get, cannot open source file "stdio.h"C/C++(1696) ..for instance cannot open source file "stdlib.h" (dependency of "stm32f769i_discovery_audio.h")C/C++(1696) when I hover over the red underlined includes

It looks like the ide cannot find your system libraries. Does your project compile or it's just a intellisense problem? Have you tried adding the include path to arm-gcc headers? On ubuntu and fedora it is installed in `/usr/arm-none-eabi/include`.

[@BenBilesBB-box]

@ Hi Matej , thanks for the reply , the project built all of the files using the make file but fails on the linking part. its a project off of github and really quite complex. Its got FPU flags enabled etc for some DSP stuff. I tried stm32CubeIDE and think its more useful for what I'm doing since it comes with a debugger and GCC all in one and I can just select the MC i'm working with. I'll keep watching your videos as it helps me understand make file basics etc

[@kcmorenachos]

nice tutorial. what is the name of this theme? looks like stm32cubIDE.

It is the default dark theme. I found that the default syntax highlighting works the best for C and C++ as well after trying loads of other popular themes.

[@bobby9568]

You are a genius!

[@THeMin1000]

Hello, can you also make a video of how to make a project for writing ARM assembly in a similar way.

I don't write much assemly at all. I can read it when I need to debug something. Also I don't see a reason to use it, unless you're making a small task scheduler or custom adder/multiplier or core control (those are already made by mcu manufacturers).

[@s94012629]

Hi, I'm a novice to learn stm32. I have a question about using printf and scanf on board(STM32F303). It cannot execute with Make method. I used MicroLIB in Keil to solve this problem before. Can you give me some suggestion to this trouble ? Thanks!

The thing is that printf and scanf are just functions that write to and read from a data stream. In case of a C application running on your computer, the stream is standard output/input -> that's why you can see printf output in console where you can also input characters and read them with scanf. On microcontroller you don't have that. It's a independent computer let's say and you have to provide a connection or a stream to and from your console. The most obvious one is UART. You can use uart as a stream pipe and route printf and scanf function over there. Your library probably did something like that, but you have to remember to enable USART peripheral and wire it up. What that library did was probably something similar. I found a great article on that note: programming.vip/docs/stm32-printf-and-scanf-function-redirection.html. You can also create your own and not use printf and scanf commands. Hope this helps.

[@SuperBiggestking]

The make file option for Toolchain/IDE is not available for the board I am working with -- the stm32h747 disco board.

I see ... Maybe it has to do somethig with the IC having two separate cpu cores and funcions, but I wouldn't know why would that affect it. I can suggest you to create CubeIde project and then run and configure build (Debug/Run) from the ide and after one build you will have Makefile inside that Debug/Release folder and you can run that. But I don't know much about the development process for these kind of beast ICs. Best or luck.

I just took a look into a such project in CubeIde. It looks like a package with two "separate" projects inside. You could apply same logic to that and just treat them separate. Use their created mekafiles and use one global makefile to run those. Therefore CubeIde takes care of configuration of it when it needs changing, but your makefile runs those and other things I discuss in the series plus many more to come :)

[@SuperBiggestking]

Thank you, Matej! I just ordered the F746 board to not have to deal with the hassle of two chips. With the F746 I give the make file option. Hopefully that means I can follow along with your video. I will get back to you about that.

[@kakmal5]

Hai.. i just want to know.. why would we use VStudio compare to STM32IDE?

It's just preference and this video shows how you can use it and retain similar functionality. My opinion is that the IDE should not matter, if we ignore advance functionalities. All you need is a build system and a compiler and learning how an IDE works limits your knowledge of the core technology. If you work with people you cannot expect them to use the same tools as yourself, so you try limiting amount of dependencies your project has. Text editor or IDE should be your choice. Right now I use neovim for all my embedded programming as I spend more time writing code than debugging it.

[@kakmal5]

@ Tq so much, you help me grow.

[@shahidkhan-md3dw]

Goodmorning my friends

[@IPridek]

Excellent video, but im stopped , Why i cant see the .vscode below my TEST project.

You will have to make some changes in order for vscode to create the folder with your configuration files. As for cpp settings, you can copy mine from the github and go from there.

[@maciej_sliv]

much appreciate!

[@sstijn577]

Thx so much for the very good video, i dont know why but only the uint32_t datatype is not recognized as a datatype, do you have any idea why this could be like this, any help is welcome, i tried a lot of things already but to no avail, still a great video, because i much prefer code formatting and also my shortcuts etc are configured in vscode etc it would be nice to have it work like this

Well vscode in my experience with the c/c++ plugin does work on and off. I found that providing it with the "compile_commands.json" file from meta build systems like cmake, meson, works the best. Give it a try.

[@cozycactus]

i prefer to use LSP with vscode and clangd for intellisense

[@edgull_tlt]

Спасибо

[@nguyenduygiang1162]

how to add library to project ?

In case of a Makefile, you need to add the path to the library headers to the list of directories, and sources to the list of sources to compile during the build. That's the same with CMake, but with different syntax. Check my other videos.

[@nguyenduygiang1162]

@ tks u guy, i did it

[@szymonpiotr5898]

8:23 Why u declare c++20 ,but path is c++/10.2.0 ? BTW: Great tutorial :)

You comment was held for review, I don't know why. Sorry for beeing late. This is arm gcc version 10.2 and does not have anything to do with c++ or c version name wise. I think that c++20 is enabled on at least 10.x.y versions. I tried 9.3.y and it doesn't have c++20 yet.

[@szymonpiotr5898]

@ ty

[@enochchan9710]

love u

[@apurbakimarkar172]

i unable to understand.......

[@sabahattinerdogan6077]

Hi sir, your tutorials are great. I have a problem. I want to use vs code at stm projects but I can't. Can you help me?

Can you describe the problem you are having? Have you followed the video?