We are currently migrating Bugzilla to GitHub issues.
Any changes made to the bug tracker now will be lost, so please do not post new bugs or make changes to them.
When we're done, all bug URLs will redirect to their equivalent location on the new bug tracker.

In an attempt to strip the SDL library to the minimal functionality as it's currently only being used for input handling on our side, I disabled as many subsystems and other build flags as seemed reasonable to me, and this included the threads flag. SDL does check for certain build flags dependencies during compilation, so I thought it would not cause any issues to disable as much as I can.

Anyhow, everything seemed fine at first: the Game Controller API was working fine and dandy for all controllers I've tested: Xbox 360, Logitech Dual Action, PS4 Dual Shock - however it breaks Switch Pro controllers.

The core issue is that the HIDAPI Switch code calls SDL_HIDAPI_LockRumble, which in turn relies on threading support. (Grepping for this function yields results for PS5 and Xbox One as well, so I assume those have similar issues but I wasn't able to test that.) This makes SDL report the error "Couldn't load stick calibration".

SDL_GameControllerOpen will end up returning null on SDL_CONTROLLERDEVICEADDED caused by the joystick driver failing to be opened.

Now, personally, I just enabled threading support to fix this issue. However, it would be nice if the issue had been less obscure. I think requiring threading to be enabled when using HIDAPI during compilation is a pretty simple solution - requiring threading to handle modern gamepads on desktop platforms doesn't seem like a lot to ask even if you're not going to end up ever using any of these few specially handled gamepads.