[BUG] IVRVirtualDisplay broken on Linux

[ckiee]

Describe the bug Usage of the IVRVirtualDisplay interface in the OpenVR Driver API on Linux causes SteamVR to crash with the usual "Shared IPC Compositor Init Failed (303)" and sometimes also error 307 which is also IPC-related.

To Reproduce Steps to reproduce the behavior:

• git clone https://github.com/maxer456/steamvr-linux-experiment steamvr-le && cd steamvr-le && mkdir build && cd build && cmake .. && make -j && cd .. && cp build/libdriver_steamvr-test.so bin/linux64/driver_steamvr-linux-experiment.so
• add path for the cloned repo to ~/.config/openvr/openvrpaths.vrpath
• start SteamVR and watch it crash

Expected behavior I expected SteamVR to not crash.

System Information:

• Distribution: Arch
• SteamVR version: 1.15.12
• Steam client version:
• Opted into Steam client beta?: No
• Graphics driver version: nVidia proprietary driver 455.45.01
• Gist for SteamVR System Information

Screenshots

Additional context This issue is not a recent regression: we've been experiencing this for a while but only I've only now identified this as a good place to ask for help in.

We wrote the example we're testing here just by reading the header files and the docs/sample for windows. It'd be really nice if we had some updated documentation for Linux for this API. (and maybe DirectMode too if it's even supported on Linux..?)

Note: Commenters who are also experiencing this issue are encouraged to include the "System Information" section in their replies.

[WilsontheWolf]

I've helped test this too. System Information:

• Distribution: Ubuntu 20.04.1 LTS
• SteamVR version:
• Steam client version:
• Opted into Steam client beta?: yes
• Graphics driver version: AMD AMD RAVEN (DRM 3.35.0, 5.4.0-54-generic, LLVM 11.0.0)
• Gist for SteamVR System Information

[WilsontheWolf]

A video of my results is here: https://www.youtube.com/watch?v=RRRufq-Qqbc

[Vixea]

+1 for directModeComponent

[FriederHannenheim]

+1

[Kierek]

+1

[ckiee]

Please use the reactions button under the comment. It's spammy otherwise.

[FriederHannenheim]

The latest SteamVR Beta 2.0.10 contains a fix for IVRVirtualDisplay. Could somebody please check if it works now?

[Vixea]

I'd prefer if kisak would add input on this development otherwise I would say based on previous information that nothing changed

[kisak-valve]

Hello @Vixea, friendly reminder that I'm a moderator for Valve's issue trackers on Github, and not a SteamVR developer myself.

I don't have any real insight into this issue, but having written that I think that change log note is in response to https://github.com/ValveSoftware/openvr/issues/1775 instead of this issue report.

[TheDeveloperGuy]

Fixed on Windows (in Beta 2.0.10). Valve QA is embarrassingly bad, so they just released SteamVR 2.0 without even testing it properly (and ignored bug reports).

[Vixea]

Also keep in mind that directModeComponent is also broken which is why I didn't the "+1 for directModeComponent" it's more of saying this is also an issue than a straight plus 1 obviously people didn't get that :|

[Vixea]

https://store.steampowered.com/news/app/250820/view/4146199229186243608

[FriederHannenheim]

poggers