05-06-2020 03:41 PM
Solved! Go to Solution.
05-19-2020 01:09 PM
kojack said:
I gave Laminar a detailed bug report. Hopefully they'll take it seriously.
05-19-2020 03:16 PM
05-19-2020 03:38 PM
1960Steve said:
What's Vulkan mode?
05-19-2020 05:50 PM
kojack said:
kojack said:
I gave Laminar a detailed bug report. Hopefully they'll take it seriously.
Well, looks like they didn't. Their response was that it's Oculus' fault. They've reported it and are waiting for Oculus to fix it (the thing Oculus has been saying for 1.5 years to not do, but X-Plane keeps doing).Someone on the X-Plane forum said they reported it to Oculus and was told it's Laminar's fault for not updating their code.So both are blaming each other.
05-19-2020 07:57 PM
Well the plot thickens. Today I tried xp11.41, 11.50b9 Vulkan, and 11.50b9 OpenGl with my Oculus Quest wireless with Virtual Desktop. All these ran fine, and surprisingly very well with 11.50b9 Vulkan btw. They also all shutdown properly.
In order to run xp11 with the Quest wireless w/VD you need to startup SteamVR and then open xp11. I have added xp11 manually to my Steam library so it shows up on my SteamVR dash menu. Alternatively, using a desktop shortcut works as well.
I'm really not sure why/how this works but it does. It may be due to some of the secret sauce in Virtual Desktop. Who knows? My Quest is currently on its v16 but I don't think that this correlates much to Rift v17. The Quest v17 is rolling out over the next week so I will try this experiment again later when I receive it. I have not tried all this with my Quest w/Link cable but I doubt that it will work with either xp11.41 or xp11.50b9 OpenGL since it then just becomes essentially a Rift with Link. I'll leave that for someone else to try.
Aside; My thoughts regarding who's to blame regarding the current Rift shutdown bug;
The xp11 shutdown bug was fixed +1.5 years ago with the Oculus update 1.35 (may have been 1.34?). At that time there had been no changes to xp11. It remained fixed until earlier this year when the Oculus v13PTC update rolled out. Also, there were no xp11 updates at that time. I did report this to Oculus (along with a few other v13PTC complaints) but this never went anywhere (= usual, not our problem) and when v13 became final the problem(s) was still there. So, it seems to me that the shutdown bug was caused by something that Oculus introduced with their v13 update. I don't have the expertise to look into this in any detail, just cause and effect analysis on my part.
05-19-2020 08:29 PM
05-19-2020 09:46 PM
moody11hamed said:
From people past experience, how long this will last the before its fixed!!!
05-19-2020 10:51 PM
05-21-2020 04:40 AM
I found out a tiny bit more about the hang when shutting down X-Plane. I got the new preview version of WinDbg (official windows debugger) and it actually found debug symbols for X-Plane! (They list things like function names inside of X-Plane).
Anyway, using WinDbg, the hang happens due to an access violation exception inside of NVAPI (nvidia's library for doing driver level stuff). Nvidia try to hide stuff in NVAPI behind a query system and use NDAs to block the public from knowing what's in there, so I can tell it's calling there but not which hidden part.
Basically at the hang X-Plane is shutting down a boost serialization singleton, That calls Kernel32's ExitProcessImplementation, which calls the NT kernel's RtlExitUserProcess, which calls the kernel's LdrShutdownProcess, which calls LdrpCallInitRoutine, which calls Nvidia's QueryInterface which recursively calls itself 2 more times, then it throws the exception.
Ok, ignore all that technical crap. The important bit is LdrpCallInitRoutine. That is how libraries are loaded and unloaded.
So what seems to be happening is X-Plane is doing a shutdown, part of that tries to unload what I'm guessing is a Nvidia library, which freaks out.
05-21-2020 06:01 AM