nRF Connect: How to build from command line after doing automatic install

RE: nRF Connect: How to build from command line after doing automatic install

liteyear — Fri, 16 Feb 2024 00:53:17 GMT

OMG, what a mission. Here's what worked for me on macOS, based on your excellent suggestion:

I couldn't find nrfutil anywhere on my system despite much searching, so I downloaded it fresh from https://www.nordicsemi.com/Products/Development-tools/nrf-util
It just comes as a binary, so then I put it on my path with: sudo mv ~/Downloads/nrfutil /usr/local/bin/
And made it executable: sudo chmod 755 /usr/local/bin/nrfutil
And then worked around the developer signing issue by opening the folder in Finder with open /usr/local/bin , right-clicking the file, selecting "Open", and approving the unverified executable.
I still couldn't run the command until I installed toolchain manager with nrfutil install toolchain-manager
Finally, the command worked and the output looked good: nrfutil toolchain-manager env --as-script
Then to tie it all together, I exported the contents to the file that gets picked up by zephyr-env.sh with nrfutil toolchain-manager env --as-script > ~/.zephyrrc
Now "all" I need to do is open a terminal and run one command to get a build environment: source /opt/nordic/ncs/v2.3.0/zephyr/zephyr-env.sh

Phew! I don't know why these basics are so hard...

RE: nRF Connect: How to build from command line after doing automatic install

Kyrre Aalerud — Wed, 31 Jan 2024 15:59:39 GMT

When using docker with sdk 2.5.1 I found the execution of single unit tests via west build no longer worked. Changing to the right folder was no longer possible as the launching of a shell via the entry point meant I would either have a shell with environment that did not handle the commands, or a shell without the environment that would.

Using the nrfutil command Peter posted allowed preparing the shell and then west build could run the specific unit tests again. The big question is, what is the intended way of running a single test seeing how this seems like a workaround?

RE: nRF Connect: How to build from command line after doing automatic install

Kyrre Aalerud — Wed, 31 Jan 2024 15:31:01 GMT

Wow! Thank you for noting this!

Now, if only they would allow us to just have it update the environment so different builds could actually work again that would be nice. Or am I missing something?

RE: nRF Connect: How to build from command line after doing automatic install

Peter Bone — Mon, 27 Nov 2023 09:55:51 GMT

For anyone ending up here with the same need for a way to setup the environment without using the toolchain manager to launch a shell, the toolchain manager itself provides a way to export the current environment as a script:

~/ncs-lcs/nrfutil toolchain-manager env --as-script

RE: nRF Connect: How to build from command line after doing automatic install

HarjitS — Tue, 08 Nov 2022 19:55:23 GMT

This is the source for Toolchain Manager v1.2.1 that launches the terminal:

github.com/.../terminal.ts

For Windows, looks like they use: nrfutil-toolchain-manager.exe in:
c:\Users\<YOUR ALIAS>>\.nrfconnect-apps\node_modules\pc-nrfconnect-toolchain-manager\resources\nrfutil-toolchain-manager\win32

There is a similar directory for Linux, Apple, etc. The "win32" changes to the appropriate name for the OS.

I really hope Nordic will provide a "nRFCmdStart" or something like that that does the equivalent of above.

RE: nRF Connect: How to build from command line after doing automatic install

HarjitS — Tue, 08 Nov 2022 19:34:34 GMT

Thank you for sharing. Unfortunately, this will break if you update the SDK. I really would like to see Nordic provide a script file.

The Toolchain manager uses the Nordic script and also VSCode->Terminal->+->nRF Connect does.

I tried to find the script location but was not able to.

RE: nRF Connect: How to build from command line after doing automatic install

mchartier — Tue, 08 Nov 2022 18:38:16 GMT

I could not find one so I made my own.

# nRF Connect
export NCS_SDK=/opt/nordic/ncs/v2.1.0
export NCS_TOOLCHAIN=/opt/nordic/ncs/toolchains/v2.1.0
export ZEPHYR_SDK_INSTALL_DIR=$NCS_TOOLCHAIN/opt/zephyr-sdk
export PATH=$PATH:$NCS_TOOLCHAIN/bin:$NCS_TOOLCHAIN/usr/bin:$NCS_TOOLCHAIN/usr/local/bin:$NCS_TOOLCHAIN:/opt/bin:$NCS_TOOLCHAIN/opt/nanopb/generator-bin:$NCS_TOOLCHAIN/opt/zephyr-sdk/arm-zephyr-eabi/bin
# source $NCS_SDK/zephyr/zephyr-env.sh

RE: nRF Connect: How to build from command line after doing automatic install

HarjitS — Tue, 08 Nov 2022 17:59:52 GMT

Is there a script or batch file we can execute that is equivalent to starting the shell from the toolchain manager? This way, we can always just execute that script or batch file and it will work with any updates/changes from Nordic.

RE: nRF Connect: How to build from command line after doing automatic install

Michal — Mon, 07 Nov 2022 09:43:33 GMT

Hello,

Thank you for the information, I will close it then.

Best regards,

Michal

RE: nRF Connect: How to build from command line after doing automatic install

mchartier — Fri, 04 Nov 2022 14:03:42 GMT

I compared the env vars between a regular shell and the ‘ncs’ shell and manually copied over the changes to my shell. Seems crude but so far it seems to work. As for vs code, I did figure out how to build the app using vs code, but the trick was manually configuring the extra kconfig files required by the app. You can close this out.

RE: nRF Connect: How to build from command line after doing automatic install

Michal — Mon, 10 Oct 2022 13:40:25 GMT

Hello,

The thing that happens when you start the shell from toolchain manager is that it sets up the environmental variables for a given SDK version.

You can add those variables by yourself if desired in your ~/.bashrc file for example.

VS Code runs the same west commands under the hood as you would in the terminal. There should be no need to do things differently.

When you build in VS Code you should be able to see the exact command used to build the application in the compiler output.

Best regards,

Michal

RE: nRF Connect: How to build from command line after doing automatic install

mchartier — Fri, 07 Oct 2022 20:16:16 GMT

Let me add some background here. When I started using nRF Connect I chose the automatic installation method. This seems like the best choice to get started quickly. Then I enrolled in the nRF Connect SDK Fundamentals course. So far so good. But now I started using some of the sample projects and they all seem to use the command line to build & flash. This works fine if you open the terminal from the toolchain manager. What is not clear to me is how to switch between command line and VS Code on the same project. For example, some of the example projects use kconfig overlay files which are listed on the build commands. If I switch over to VS Code I guess I need to edit the application setup and add the kconfig file there? This behavior is new to me. I'm used to running 'make' from the command line with no options because all the details are stored in the makefiles. Then your IDE can just run the same make command.

In short, I was hoping to learn how to use VS Code and the command line on the same project. In the short term I want to make this work for the sample projects but later I need to get the same behavior working with my own project.