nixos/hardware.opentabletdriver: Stuck on "Connecting to OpenTabletDriver Daemon" after upgrade to 25.11

[arthurrump]

Nixpkgs version

• Stable (25.11)

Describe the bug

After upgrading from 25.05 to 25.11, the OpenTabletDriver daemon no longer starts correctly. This results in the application being stuck on "Connecting to OpenTabletDriver Daemon..."

The user service seems to fail with an error that the "Daemon is already running":

❯ systemctl status --user opentabletdriver.service
○ opentabletdriver.service - Open source, cross-platform, user-mode tablet driver
     Loaded: loaded (/etc/systemd/user/opentabletdriver.service; enabled; preset: ignored)
     Active: inactive (dead) since Tue 2025-12-09 16:58:52 CET; 9min ago
   Duration: 1.097s
 Invocation: 837d09fb9143416dac594bfbf1e70b94
    Process: 4741 ExecStart=/nix/store/gzx7pb3kax86pnk34f2i9q2ahj4fjzhx-OpenTabletDriver-0.6.6.2/bin/otd-daemon (code=exited, >
   Main PID: 4741 (code=exited, status=0/SUCCESS)
   Mem peak: 8.9M
        CPU: 97ms

dec 09 16:58:51 host systemd[4567]: Started Open source, cross-platform, user-mode tablet driver.
dec 09 16:58:51 host otd-daemon[4741]: OpenTabletDriver Daemon is already running.

A grep on the boot logs shows that there is indeed already another otd-daemon running, which does not show up in the --user version of those logs:

❯ journalctl -b 0 | grep otd-daemon
dec 09 16:58:44 host otd-daemon[3551]: [Detect:Info]        Searching for tablets...
dec 09 16:58:44 host otd-daemon[3551]: [Detect:Debug]        Searching for tablet 'Acepen AP1060'
...
dec 09 16:58:44 host otd-daemon[3551]: [Detect:Debug]        Searching for tablet 'Wacom CTL-672'
dec 09 16:58:44 host otd-daemon[3551]: [Device:Debug]        Initializing device 'CTL-672' /sys/devices/pci0000:00/.../hidraw/hidraw4
dec 09 16:58:44 host otd-daemon[3551]: [Device:Debug]        Using report parser type 'OpenTabletDriver.Configurations.Parsers.Wacom.Intuos.IntuosReportParser'
dec 09 16:58:44 host otd-daemon[3551]: [Device:Debug]        Set device feature: 02-02
dec 09 16:58:44 host otd-daemon[3551]: [Detect:Info]        Found tablet 'Wacom CTL-672'
dec 09 16:58:44 host otd-daemon[3551]: [Detect:Debug]        Searching for tablet 'Wacom CTL-680'
...
dec 09 16:58:44 host otd-daemon[3551]: [Detect:Debug]        Searching for tablet 'XP-Pen Star G960S'
dec 09 16:58:44 host otd-daemon[3551]: [Display:Warning]        Neither Wayland nor X11 were detected, defaulting to X11.
dec 09 16:58:46 host otd-daemon[3892]: [Detect:Info]        Searching for tablets...
dec 09 16:58:46 host otd-daemon[3892]: [Detect:Debug]        Searching for tablet 'Acepen AP1060'
...
dec 09 16:58:46 host otd-daemon[3892]: [Detect:Debug]        Searching for tablet 'Wacom CTL-672'
dec 09 16:58:46 host otd-daemon[3892]: [Device:Debug]        Initializing device 'CTL-672' /sys/devices/pci0000:00/.../hidraw/hidraw4
dec 09 16:58:46 host otd-daemon[3892]: [Device:Debug]        Using report parser type 'OpenTabletDriver.Configurations.Parsers.Wacom.Intuos.IntuosReportParser'
dec 09 16:58:46 host otd-daemon[3892]: [Device:Debug]        Set device feature: 02-02
dec 09 16:58:46 host otd-daemon[3892]: [Detect:Info]        Found tablet 'Wacom CTL-672'
dec 09 16:58:46 host otd-daemon[3892]: [Detect:Debug]        Searching for tablet 'Wacom CTL-680'
...
dec 09 16:58:46 host otd-daemon[3892]: [Detect:Debug]        Searching for tablet 'XP-Pen Star G960S'
dec 09 16:58:46 host otd-daemon[3892]: [Evdev:Debug]        Successfully initialized virtual pressure sensitive tablet. (code NONE)
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Output mode: Artist Mode
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Display area: [5120x1600@<2560, 800>:0°],
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Tablet area: [216x135@<108, 67.5>:0°],
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Clipping: Enabled
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Ignoring reports outside area: Disabled
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Pressure: Enabled
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Tilt: Enabled
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Tip Binding: [Adaptive Binding: Tip]@1%
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Eraser Binding: [Adaptive Binding: Eraser]@1%
dec 09 16:58:46 host otd-daemon[3892]: [Wacom CTL-672:Info]        Pen Bindings: Adaptive Binding: Button 1, Adaptive Binding: Button 2
dec 09 16:58:46 host otd-daemon[3892]: [Settings:Info]        Driver is enabled.
dec 09 16:58:51 host otd-daemon[4741]: OpenTabletDriver Daemon is already running.

❯ journalctl -b 0 --user | grep otd-daemon
dec 09 16:58:51 host otd-daemon[4741]: OpenTabletDriver Daemon is already running.

This did not seem to happen on 25.05.

I was able to reproduce this with both niri and GNOME, started from GDM. These logs complain not being able to detect X11 or wayland, but I have the exact same issue on another device (with no tablet connected), where this line does not show up in the log.

Steps to reproduce

hardware.opentabletdriver.enable = true;

in 25.11, reproduced at least with GNOME and GDM

Expected behaviour

The otd-daemon should only be started from the user service

Screenshots

No response

Relevant log output

Additional context

No response

System metadata

 - system: `"x86_64-linux"`
 - host os: `Linux 6.12.59-xanmod1, NixOS, 25.11 (Xantusia), 25.11pre-git`
 - multi-user?: `yes`
 - sandbox: `yes`
 - version: `nix-env (Nix) 2.31.2`
 - nixpkgs: `/nix/store/mmbrs76ksccdn0bvvqs4plivwh6xlcs5-source`

Notify maintainers

Maintainers for the opentabletdriver package: @gepbird @thiagokokada

---

Note for maintainers: Please tag this issue in your pull request description. (i.e. Resolves #ISSUE.)

I assert that this issue is relevant for Nixpkgs

• I assert that this is a bug and not a support request.
• I assert that this is not a duplicate of an existing issue.
• I assert that I have read the NixOS Code of Conduct and agree to abide by it.

Is this issue important to you?

Add a 👍 reaction to issues you find important.

[gepbird]

That's weird that you have multiple processes running for otd-daemon, when I do journalctl -b 0 | grep otd-daemon I only get a single process ID (I'm on unstable which currently should be very close to 25.11).
Does this happen right after just upgrading from 25.05 to 25.11 WITHOUT restarting the machine?
If it's reproducible after a reboot, can you share your whole config pinned to a commit where you experience this?

[arthurrump]

This is reproducible after a reboot; I did not try the upgrade without restarting. I'm not comfortable sharing my full configuration publicly, but I'll have a go at finding a more minimal config that gives the same issue. The fact that things are working fine for you suggests there is something in my configuration that causes the issue, so that gives me some motivation to find where that is.

[arthurrump]

I created a minimal configuration with basically just Gnome and the opentabletdriver enabled (alongside secure boot and root on zfs, so I could just reboot into it). The issue remains there, with the same symptoms. Then I tried downgrading that config to 25.05 to test what an upgrade would look like, but now it is also failing on 25.05, though with slightly different symptoms:

❯ journalctl -b 0 --user | grep otd-daemon
dec 09 21:43:47 blipa otd-daemon[2023]: [Plugin:Debug]        Loading plugin 'DualActionBinds'
dec 09 21:43:47 blipa otd-daemon[2023]: [Plugin:Debug]        Loading plugin 'Slimy Scylla'
dec 09 21:43:47 blipa otd-daemon[2023]: [Detect:Info]        Searching for tablets...
dec 09 21:43:47 blipa otd-daemon[2023]: [Detect:Debug]        Searching for tablet 'Acepen AP1060'
...
dec 09 21:43:47 blipa otd-daemon[2023]: [Detect:Debug]        Searching for tablet 'XP-Pen Star G960S'
dec 09 21:43:47 blipa otd-daemon[2023]: [Detect:Info]        No tablets were detected.
dec 09 21:43:47 blipa otd-daemon[2023]: [SocketException:Fatal]        Address already in use
dec 09 21:43:47 blipa otd-daemon[2023]:    at System.Net.Sockets.Socket.UpdateStatusAfterSocketErrorAndThrowException(SocketError error, Boolean disconnectOnFailure, String callerName)
dec 09 21:43:47 blipa otd-daemon[2023]:    at System.Net.Sockets.Socket.DoBind(EndPoint endPointSnapshot, SocketAddress socketAddress)
dec 09 21:43:47 blipa otd-daemon[2023]:    at System.Net.Sockets.Socket.Bind(EndPoint localEP)
dec 09 21:43:47 blipa otd-daemon[2023]:    at System.IO.Pipes.NamedPipeServerStream.SharedServer..ctor(String path, Int32 maxCount, Boolean isFirstPipeInstance)
dec 09 21:43:47 blipa otd-daemon[2023]:    at System.IO.Pipes.NamedPipeServerStream.SharedServer.Get(String path, Int32 maxCount, PipeOptions pipeOptions)
dec 09 21:43:47 blipa otd-daemon[2023]:    at System.IO.Pipes.NamedPipeServerStream.Create(String pipeName, PipeDirection direction, Int32 maxNumberOfServerInstances, PipeTransmissionMode transmissionMode, PipeOptions options, Int32 inBufferSize, Int32 outBufferSize, HandleInheritability inheritability)
dec 09 21:43:47 blipa otd-daemon[2023]:    at System.IO.Pipes.NamedPipeServerStream..ctor(String pipeName, PipeDirection direction, Int32 maxNumberOfServerInstances, PipeTransmissionMode transmissionMode, PipeOptions options)
dec 09 21:43:47 blipa otd-daemon[2023]:    at OpenTabletDriver.Desktop.RPC.RpcHost`1.CreateStream() in /build/source/OpenTabletDriver.Desktop/RPC/RpcHost.cs:line 55
dec 09 21:43:47 blipa otd-daemon[2023]:    at OpenTabletDriver.Desktop.RPC.RpcHost`1.Run(T host) in /build/source/OpenTabletDriver.Desktop/RPC/RpcHost.cs:line 26
dec 09 21:43:47 blipa otd-daemon[2023]:    at OpenTabletDriver.Daemon.Program.Main(String[] args) in /build/source/OpenTabletDriver.Daemon/Program.cs:line 102
dec 09 21:43:49 blipa otd-daemon[2158]: [Plugin:Debug]        Loading plugin 'DualActionBinds'
dec 09 21:43:49 blipa otd-daemon[2158]: [Plugin:Debug]        Loading plugin 'Slimy Scylla'
dec 09 21:43:49 blipa otd-daemon[2158]: [Detect:Info]        Searching for tablets...
dec 09 21:43:49 blipa otd-daemon[2158]: [Detect:Debug]        Searching for tablet 'Acepen AP1060'
...
dec 09 21:43:49 blipa otd-daemon[2158]: [Detect:Debug]        Searching for tablet 'XP-Pen Star G960S'
dec 09 21:43:49 blipa otd-daemon[2158]: [Detect:Info]        No tablets were detected.
dec 09 21:43:49 blipa otd-daemon[2158]: [SocketException:Fatal]        Address already in use
dec 09 21:43:49 blipa otd-daemon[2158]:    at System.Net.Sockets.Socket.UpdateStatusAfterSocketErrorAndThrowException(SocketError error, Boolean disconnectOnFailure, String callerName)
dec 09 21:43:49 blipa otd-daemon[2158]:    at System.Net.Sockets.Socket.DoBind(EndPoint endPointSnapshot, SocketAddress socketAddress)
dec 09 21:43:49 blipa otd-daemon[2158]:    at System.Net.Sockets.Socket.Bind(EndPoint localEP)
dec 09 21:43:49 blipa otd-daemon[2158]:    at System.IO.Pipes.NamedPipeServerStream.SharedServer..ctor(String path, Int32 maxCount, Boolean isFirstPipeInstance)
dec 09 21:43:49 blipa otd-daemon[2158]:    at System.IO.Pipes.NamedPipeServerStream.SharedServer.Get(String path, Int32 maxCount, PipeOptions pipeOptions)
dec 09 21:43:49 blipa otd-daemon[2158]:    at System.IO.Pipes.NamedPipeServerStream.Create(String pipeName, PipeDirection direction, Int32 maxNumberOfServerInstances, PipeTransmissionMode transmissionMode, PipeOptions options, Int32 inBufferSize, Int32 outBufferSize, HandleInheritability inheritability)
dec 09 21:43:49 blipa otd-daemon[2158]:    at System.IO.Pipes.NamedPipeServerStream..ctor(String pipeName, PipeDirection direction, Int32 maxNumberOfServerInstances, PipeTransmissionMode transmissionMode, PipeOptions options)
dec 09 21:43:49 blipa otd-daemon[2158]:    at OpenTabletDriver.Desktop.RPC.RpcHost`1.CreateStream() in /build/source/OpenTabletDriver.Desktop/RPC/RpcHost.cs:line 55
dec 09 21:43:49 blipa otd-daemon[2158]:    at OpenTabletDriver.Desktop.RPC.RpcHost`1.Run(T host) in /build/source/OpenTabletDriver.Desktop/RPC/RpcHost.cs:line 26
dec 09 21:43:49 blipa otd-daemon[2158]:    at OpenTabletDriver.Daemon.Program.Main(String[] args) in /build/source/OpenTabletDriver.Daemon/Program.cs:line 102

This continues on for a few more times, until systemd decides there have been enough attempts.

Notable here is that journalctl --user and plain journalctl show the same logs in this case, so there is no other process started before the first one that fails. Somehow, some state must have become messed up, but I'm completely lost on how that could've happened or how I can find out.

I also originally verified that the issue was introduced in the upgrade to 25.11 by booting into my last 25.05 build through the boot menu, where it worked normally. So that now makes me extra confused... Any pointers on how to debug this would be much appreciated!

I have published the minimal configuration I tested (now at 25.05) at https://github.com/arthurrump/configuration-469340.

[gepbird]

Thanks for creating a minimal repro, but I still couldn't get that error.
I only did this modification to your repro and started it in a VM: gepbird/configuration-469340@d679a49

From issues like this my first idea is the state of you machine: maybe you accidentally added otd to auto start and it's racing with the NixOS enabled service? Or some previously saved otd configuration is messing things up?

If you stop the service with sudo systemctl stop --user opentabletdriver, do you see otd running with ps x | grep OpenTabletDriver? If so, can you check which process started it with eg. pstree?

[gepbird]

From your previous error message looks like this is OpenTabletDriver/OpenTabletDriver#4349 or OpenTabletDriver/OpenTabletDriver#3804.

Can you try rm /tmp/CoreFxPipe_OpenTabletDriver.Daemon then systemctl restart --user opentabletdriver?

[arthurrump]

From your previous error message looks like this is OpenTabletDriver/OpenTabletDriver#4349 or OpenTabletDriver/OpenTabletDriver#3804.

Can you try rm /tmp/CoreFxPipe_OpenTabletDriver.Daemon then systemctl restart --user opentabletdriver?

Yes, looks like that is indeed the issue for the second error message. I added boot.tmp.cleanOnBoot = true; (which I normally have enabled) to the minimal configuration and it started working again on 25.05.

I then updated the config to 25.11 and after a live switch, OpenTabletDriver still works. After a reboot, however, I see the original problem. I copied over your VM configuration and can now reproduce it in a fresh VM (thanks for the tip btw! I never realised it was that easy).

The annoying thing is that the process that blocks the normal user service from starting is already gone by the time I can check the status, so ps x | grep OpenTabletDriver gives me nothing.

[arthurrump]

The pipe that is left behind (probably due to OpenTabletDriver/OpenTabletDriver#3804) is owned by gdm-greeter:

> l /tmp
...
srwxr-xr-x  1 gdm-greeter gdm      0 13 dec 17:04 CoreFxPipe_OpenTabletDriver.Daemon

I don't know if the service is supposed to be started for GDM, especially since the normal user service seems to want to start before the GDM-owned service is stopped:

> journalctl -b 0 | grep 'tablet driver\|otd-daemon'
dec 13 17:04:53 blipa systemd[2331]: Started Open source, cross-platform, user-mode tablet driver.
dec 13 17:04:55 blipa otd-daemon[2507]: [Detect:Info]        Searching for tablets...
dec 13 17:04:55 blipa otd-daemon[2507]: [Detect:Debug]        Searching for tablet 'Acepen AP1060'
...
dec 13 17:04:55 blipa otd-daemon[2507]: [Detect:Debug]        Searching for tablet 'XP-Pen Star G960S'
dec 13 17:04:55 blipa otd-daemon[2507]: [Detect:Info]        No tablets were detected.
dec 13 17:06:41 blipa systemd[3297]: Started Open source, cross-platform, user-mode tablet driver.
dec 13 17:06:41 blipa otd-daemon[3431]: OpenTabletDriver Daemon is already running.
dec 13 17:06:51 blipa systemd[2331]: Stopping Open source, cross-platform, user-mode tablet driver...
dec 13 17:06:51 blipa systemd[2331]: Stopped Open source, cross-platform, user-mode tablet driver.

[gepbird]

Now I can also reproduce it (I see you recently updated the repro to use 25.11, I thought it was already on that and didn't check).

I'm confused why is it also running as a system level service, I'm bisecting which change caused it between 25.05 and 25.11.
A good commit is 6ee6f60, bad is c92e612. There's no changes to opentabletdriver in between.

[xXx-zwz-xXx]

It gets stuck when I run it too. I don't know enough to help proactively but might be useful to know it's effecting others. I have it in my config like on the nix wiki:

   # Enable OpenTabletDriver
   hardware.opentabletdriver.enable = true;

   # Required by OpenTabletDriver
   hardware.uinput.enable = true;
   boot.kernelModules = [ "uinput" ];

[gepbird]

Thanks for the report! I assume you're using GNOME too, right?

I'm still trying to find out why did it break, I'm 75% confident it is caused by GNOME 49, as most of the commits between a working and a broken state include gnome package bumps.
While trying to narrow down the issue, I tried to make the repetative process of logging in to GNOME, opening a terminal and checking systemctl status --user opentabletdriver simpler by enabling automatic login.
Surprisingly auto login got rid of the open tablet driver issue, so I lost a few hours of building+testing.

As weird as it sounds, enabling auto login is a workaround for this issue.