Linux board Hello World

Prerequisites​

• A Raspberry Pi 4B / 5 (or any Linux board, e.g. Taishan Pi 3 or Dongshan Pi A1; this guide uses the Raspberry Pi), with an OS flashed, internet access, and SSH enabled.
• Your computer and the board are on the same LAN and you can ping its IP.
• The extension is installed and the SDK is initialized — see Install TuyaOpen IDE.

How a Linux board differs from an MCU board​

The T5AI_Board from Practice 1 is flashed over serial. A Linux board has no serial-flashing step — once the IDE detects that the target is a Linux board, it automatically repurposes the same set of toolbar buttons:

The Linux board's Project Details page also gains a Board connection section (for SSH info).

Step 1: Create a Linux board project​

1. Click the Board Catalogue icon in the left activity bar to browse the supported boards, chips, and SoC platforms with their full specs.

   

2. Find Raspberry Pi (or your actual Linux board) in the list and click Create project from this board.

   

3. Open the Project Details page and confirm that the top Flash button now reads Deploy and Monitor reads Run, with a Board connection form below — that is the sign the IDE has recognized it as a Linux board.

   

Step 2: Configure the SSH connection​

Before filling in the form, make sure the board is powered on, on the same LAN as you, has SSH running, and that you have a working username and password.

How to get the username / IP / password: the username and password are whatever you set when you flashed the OS; for the IP, run hostname -I on the board.

Fill in the Board connection form:

How to choose a login method: the "Login method" field decides which SSH authentication the IDE uses to reach the board.

• Password (default): the password is stored in the editor's secure credential store (the OS keychain — never written to disk, never placed in the project files). If sshpass is installed on your machine, the IDE reuses the stored password automatically and you never have to type it; multiple ssh/scp calls within the same operation even reuse a single connection, so you are not prompted repeatedly.
• Private key (advanced): authenticates with an SSH key instead of a password. The IDE defaults to ~/.ssh/id_rsa and lets you pick another key file in the config; the corresponding public key must already be registered in the board's ~/.ssh/authorized_keys.

Fill it in, click Save connection, then click Test SSH. On success it prints:

SSH OK
Linux raspberrypi 6.x.x ... aarch64 GNU/Linux
pi

Seeing SSH OK plus the kernel info and the login user means you are connected. If it fails, the IDE gives a categorized hint — for an auth failure, double-check the username and password; for a network problem, ping <IP> first; for a first-connection host-key issue, run ssh <username>@<IP> once on your computer to confirm it.

Step 3: Build​

Click Build at the top. On success it produces a Linux executable (.elf), which the Project Details firmware area lists as "Application (ELF)".

Expected output (build success):

[NOTE]: 
====================[ BUILD SUCCESS ]===================
 Target    : <project name>_QIO_1.0.0.bin
 Output    : /home/<your-username>/TuyaOpenIDE/projects/<project name>/source/embedded/dist/<project name>_1.0.0
 Platform  : LINUX
 Chip      : Raspberry_Pi
 Board     : Raspberry_Pi
 Framework : base
========================================================

Hello World depends on no model or cloud credentials, so the build output is a clean Linux executable — no auth code or extra assets required.

Step 4: Deploy​

Click Deploy at the top. The IDE deploys to the remote target machine over an SSH channel:

Expected output (deploy success):

>>> Uploading ELF…
>>> Deploy complete.
>>> Remote path: /home/pi/tuyaopen-apps/<project name>/<project name>_1.0.0.elf
>>> Local path: /home/<your-username>/TuyaOpenIDE/projects/<project name>/source/embedded/dist/<project name>_1.0.0/<project name>_1.0.0.elf (<size> bytes)

Step 5: Run​

Click Run at the top (the same spot as the MCU board's Monitor button). The IDE opens a dedicated terminal, logs into the board over ssh -t, and runs the executable in the foreground; its output streams here in real time.

Expected output (app startup):

[01-01 00:00:00 ty N][sample_project.c:38] Application information:
...
[01-01 00:00:00 ty D][sample_project.c:48] hello world

When you see hello world at the end, you are done. After printing its info, the app enters a resident loop and keeps running; to stop it, press Ctrl + C or close the terminal.

Bonus: push resource files (images, videos, …) to the board​

Beyond the program itself, the IDE can push your project's resource files — images, videos, 3D models, and so on — to the board in one click, saving you a manual scp.

1. Put the resource files in the project's source/embedded/resources/ directory (you can create subdirectories; the IDE uploads recursively and preserves the structure).

2. Run TuyaOpen: Upload Resources to Board from the command palette (Ctrl+Shift+P), or click the Upload Resources button.

   

The resources land under <remote directory>/resources/ on the board (default ~/tuyaopen-apps/<project name>/resources/); for example, a local resources/hero.glb becomes ~/tuyaopen-apps/<project name>/resources/hero.glb on the board.

Debug with Ubuntu (X86)​

Besides the Raspberry Pi (ARM/aarch64), the IDE also supports Ubuntu as a Linux deploy target — it is an X86_64 "Linux board". The flow is identical to the Raspberry Pi (both are cross-deploy: build the ELF → deploy over SSH → run); only the build target's architecture differs. When creating the project, pick Ubuntu as the board type, then follow from Step 2 onward.

Why use it for debugging:

• A Linux stand-in when you have no real ARM board — if you don't have a Raspberry Pi or Taishan Pi at hand, use an Ubuntu (X86) machine, VM, or cloud server as your "Linux board" to get the app logic and the "build → deploy → run" loop working first, then port to the real board without waiting for hardware.
• See the UI right on the desktop — on an Ubuntu desktop, the LVGL UI renders in an SDL2 window, so you can tune the UI without a real display attached.
• X86 is easy to set up and reproduce — VMs, containers, and cloud hosts all work; they snapshot and reproduce well, which suits application-layer debugging and automation.

FAQ​

• You want password-free login / it says sshpass is missing — for non-interactive use under the password method, install sshpass on your machine (macOS: Homebrew; Linux: package manager; Windows: WSL); once detected, the IDE reuses the stored password automatically.
• Deploy says ".elf file not found" — go back to Step 3 and build first, and confirm the artifact was produced.
• You changed code and need to redeploy — order: Build → Deploy → Run.

Next step​

You have deployed a program to a Linux board in one click and run through the Linux board's "build → deploy → run" workflow. Go deeper: