DaVinci Resolve Checker

Check your system configuration and hardware for ability to successfully run DaVinci Resolve.

This project is only targeted for Linux platform. Windows and Hackintosh users should not be interested in this project, as there is no problem for them to install native version for their platform.

It checks GPUs presented in system (shows the driver in use), checks OpenGL drivers (actual renderer string), installed OpenCL drivers. If script detects configuration problem, it suggests how to solve it.

Supported distributions:

• Arch Linux
• EndeavourOS
• Garuda Linux
• Manjaro Linux

Installation:

• Clone this repository.
• Install required dependencies:
  sudo pacman -S expac mesa-utils python-distro
• Install other required dependencies from aur (assuming you use yay):
  yay -S python-pylspci

Python modules could be alternatively installed with pip:
pip install distro pip install pylspci

Usage:

Run the script with the same parameters as you intend to launch D.R. For example, if using Nvidia Optimus laptop, you probably use prime-run, so run: prime-run ./davinci-resolve-checker.py

The output of the script should be the following:

DaVinci Resolve checker 1.4.1
...
All seems good. You should be able to run DaVinci Resolve successfully.

If you have some problem, the script will tell you what is wrong with your configuration, for example:

DaVinci Resolve checker 1.4.1
Chassis type: desktop
Installed OpenCL drivers: opencl-amd-polaris opencl-nvidia
Presented GPUs:
        UHD Graphics 630 (Desktop) (kernel driver in use: i915)
        Ellesmere [Radeon RX 470/480/570/570X/580/580X/590] (kernel driver in use: amdgpu)
OpenGL vendor string: Intel

Your primary gpu is Intel. Go to your uefi settings and set primary display to PCIE. Otherwise you could not use DaVinci Resolve (I did not tested it).

The script will use the locale set in your machine to display messages in your preferred language. In case it's not implemented, default is en_US (United States English).

Override this behavior by providing a --locale xx_YY argument to the script call. Example: --locale pt_BR for Brazilian Portuguese, or --locale zh_CN for Simplified Chinese.

For AMD GPUs you have an option to force checks with proprietary stack (amdgpu-pro). For this, add --pro parameter:

$ ROC_ENABLE_PRE_VEGA=1 davinci-resolve-checker.py --pro -l en_US
Using locale en_US
DaVinci Resolve checker 5.0.0
Installed DaVinci Resolve package: davinci-resolve-studio 18.1.4-1
Chassis type: desktop
Installed OpenCL drivers: opencl-amd
Presented GPUs:
        Ellesmere [Radeon RX 470/480/570/570X/580/580X/590] (kernel driver in use: amdgpu)
        Navi 23 WKS-XL [Radeon PRO W6600] (kernel driver in use: amdgpu)
OpenGL vendor string: AMD
OpenGL renderer string: AMD Radeon RX 580 Series (polaris10, LLVM 15.0.7, DRM 3.49, 6.2.10-arch1-1)
clinfo detected platforms and devices:
        AMD Accelerated Parallel Processing (roc) (number of devices: 2)
                AMD Radeon Pro W6600
                AMD Radeon RX 580 Series
        AMD Accelerated Parallel Processing (orca) (number of devices: 1)
                AMD Radeon RX 580 Series

You have several AMD GPUs. DR Studio can utilise several GPUs. Script will check if appropriate driver for your renderer GPU is used. But keep in mind that if you use prime offloading, than your primary gpu still need appropriate driver (script does not check it).
You are not using Pro OpenGL implementation. Install amdgpu-pro-libgl and run DaVinci Resolve with progl prefix. Otherwise it will crash.

Contribution:

If you have find some error or want to ask for a feature, open a new issue and describe the problem in detail.

Localization

Currently, this script displays messages in these locales:

• en_US - English, United States
• it_IT - Italian
• pt_BR - Portiguese, Brazil
• ru_RU - Russian, Russian Federation
• zh_CN - Chinese, Simplified

To contribute translating the script, add a locale file to the languages/ directory. Use the en_US.py file as a reference.

The name of your file should match exactly the locale identification. For example, if you were to implement translation to French, France, you'd have a languages/fr_FR.py file.

After implementing the translation, run python -m unittest to certify everything is in order.

If you see any errors, it means your translation dictionary local_str has misconfigured keys. Double check against the en_US.py file, or open an issue if you need assistance.

To translate in vs code, use Vscode Google Translate extension. Also redefine shortcuts, see #74