Duo sunshine button is always greyed out

[BiassedXD]

I've never used duo before, but the sunshine button hasn't work since I got it. I've tried reinstalling, still nothing.
This message keeps showing up in the logs. I don't know if it's important or not

Event Name: WUDFUnhandledException
Response: Not available
Cab Id: 0

Problem signature:
P1: UnhandledException
P2: Host
P3: c0000005
P4: 8b20
P5: duoidd.dll
P6: 1.5.9.0
P7: duoidd.dll
P8: 1.5.9.0
P9: 2.31.0
P10: (DumpedSeparately)

Attached files:
\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WER201B.tmp.WERInternalMetadata.xml

These files may be available here:
\\?\C:\ProgramData\Microsoft\Windows\WER\ReportArchive\NonCritical_UnhandledExcepti_8b6785d9e6bbadb4fded8e4e358b1d1507f15b_00000000_866ee5fd-6054-4be5-85d0-467c00298c23

Analysis symbol: 
Rechecking for solution: 0
Report Id: 866ee5fd-6054-4be5-85d0-467c00298c23
Report Status: 268435456
Hashed bucket: 2e8fbdf8f28dfeae217104762cb18811
Cab Guid: 0

I'd really appreciate some help, let me know what info you need.
Thanks.

[BiassedXD]

WudfHost_ext__171172.dmp

[BiassedXD]

It seems to work fine on my windows home edition laptop. The computer I want to run it on is on windows 10 enterprise, is that supported?

[GorigoLewis]

I have similiar issues starting from not able to pair, starting from confusing guide where the 10 step and 11 step is literally (just go there and then you have everything) along when launching app sessions do start automatically while i disabled it, pressing stop session results in an infinity closing

[BiassedXD]

Yes still an issue 1 month later would appreciate some help. 🙏

[Gnightz]

I think I have the same issue as you.

[magalz]

Hey! I dug into the dump you attached and found something useful.

This is a real crash inside Duo’s virtual-display driver, not just a generic Sunshine failure:

Process:        WUDFHost.exe
Faulting driver: DuoIdd.dll
DuoIdd version:  1.5.9.0
Faulting offset: DuoIdd+0x8B20
Exception:       0xC0000005
Invalid read:    0x220
Loaded stack:    DuoIdd.dll + IddCx.dll

The dump only has DuoIdd.dll and IddCx.dll loaded, and Sunshine and RTSP are not involved in this particular crash. It looks like DuoIDD is dereferencing an invalid internal object during virtual-display initialization or a display-mode change.

Could you share a bit more information so we can try to narrow down what triggers it? That is a lot of possible logs, lol, so:

Environment

Get-ComputerInfo |
  Select-Object WindowsProductName,WindowsVersion,OsBuildNumber

Also:

• Duo version/build;
• GPU model;
• NVIDIA driver version;
• whether the failure occurs immediately after boot or only after starting a specific Duo instance;
• whether the Sunshine button is already grey before any Moonlight connection attempt.

Event logs

Immediately after reproducing the issue, please export:

mkdir C:\DuoDebug 2>nul
wevtutil epl System C:\DuoDebug\System.evtx
wevtutil epl Application C:\DuoDebug\Application.evtx

Please also run:

wevtutil el | findstr /i "DriverFramework WUDF Display Kernel-PnP"

and export any matching DriverFrameworks/WUDF operational channel.

Duo/Sunshine logs

Please provide the relevant .log files from Duo/Sunshine.

C:\Program Files\Duo\config\<InstanceName>.log

Procmon capture

A short process-scoped Procmon capture would be especially useful. Filter only:

PROCESS NAME is WUDFHost.exe and INCLUDE
PROCESS NAME is Duo.exe and INCLUDE
PROCESS NAME is DuoRdp.exe and INCLUDE
PROCESS NAME is sunshine.exe and INCLUDE

After setting this filter, start the capture immediately before launching the Duo instance and stop when the Sunshine button becomes grey (wait some 10 seconds just to be sure), them export:

DuoIDD-startup.PML
DuoIDD-startup.CSV

The important events are:

• Process Create/Exit;
• DuoIdd.dll and IddCx.dll loading;
• display/registry access;
• IRP_MJ_DEVICE_CONTROL;
• ACCESS DENIED;
• NAME NOT FOUND;
• PATH NOT FOUND.

Please record the exact timestamp when the instance is launched and when Sunshine becomes grey.

A full WUDFHost dump is optional.

The existing dump already identifies a null/invalid object dereference inside DuoIdd. The additional logs and a synchronized Procmon trace may reveal which display or mode-change request precedes the crash.