Skip to main content
Topic solved
This topic has been marked as solved and requires no further attention.
Topic: Error message: Opening file /tmp/.ICE-unix failed proceeding without lock (Read 571 times) previous topic - next topic
0 Members and 3 Guests are viewing this topic.

Error message: Opening file /tmp/.ICE-unix failed proceeding without lock

Hello, guys

I've recently started to notice an error message while booting:

Opening file "/tmp/.ICE-unix/996" failed, proceeding without lock: No such device or address

"996" can be any random number lower than my user (1000).
Managed to find out that this file belongs to xfce4-session.
Everything seems to be working, but the error is annoying.
I've also tried to remove everything in /tmp before rebooting, but the message persists.

UPDATE: udev (version 256) + xfce = /tmp/.ICE-unix lockfiles errors
Downgraded udev and its parts to version 255 and the error is gone.
Guess there's some bug in the upstream systemd.

Re: Error message: Opening file /tmp/.ICE-unix failed proceeding without lock

Reply #1
/tmp/.ICE-unix is the inter-client exchange protocol directory, usually created by the X login manager and primarily used by the session manager (in my case, mate-session). The socket file inside should be the PID of the session manager, so you should check if things match.

Re: Error message: Opening file /tmp/.ICE-unix failed proceeding without lock

Reply #2
/tmp/.ICE-unix is the inter-client exchange protocol directory, usually created by the X login manager and primarily used by the session manager (in my case, mate-session). The socket file inside should be the PID of the session manager, so you should check if things match.


I guess I know the reason... It could be some bug in the new etmpfiles version or udev. I'm not 100% sure yet. This is what I see in my pacman.log:

Code: [Select]
[2024-06-20T15:56:32+0300] [ALPM] transaction started
[2024-06-20T15:56:32+0300] [ALPM] upgraded libudev (255.6-1 -> 256-1)
[2024-06-20T15:56:33+0300] [ALPM] upgraded esysusers (255.6-1 -> 256-1)
[2024-06-20T15:56:33+0300] [ALPM] upgraded etmpfiles (255.6-1 -> 256-1)
[2024-06-20T15:56:33+0300] [ALPM] upgraded lib32-udev (255.6-1 -> 256-1)
[2024-06-20T15:56:33+0300] [ALPM] upgraded udev (255.6-1 -> 256-1)

That's when this error message started to appear. And this happens on two machines.

I think I'll try to downgrade these packages today and see if it helps (will report back).

I'd also willingly dive into the code but I cannot find neither esysusers nor etmpfiles on Gitea sources to see the changes...

Or is it just udev that compiles into several packages?

P.S. udev is part of systemd as it turns out. Then I'm out of reporting them bugs. It's just useless, I guess.

Re: Error message: Opening file /tmp/.ICE-unix failed proceeding without lock

Reply #3
Reporting... Downgraded udev and its parts = the error has gone. I guess there's a bug (or feature, lol) in the new systemd.

Re: Error message: Opening file /tmp/.ICE-unix failed proceeding without lock

Reply #4
Or is it just udev that compiles into several packages?
Yes

Ways to get the pkgbase of a package. (Why this isn't part of pacman -Qi I have no idea?)
expac %e PKG
pacinfo PKG (part of pacutils)
cat /var/lib/pacman/local/PKG($version)/desc

Re: Error message: Opening file /tmp/.ICE-unix failed proceeding without lock

Reply #5
Ways to get the pkgbase of a package.
expac %e PKG
pacinfo PKG (part of pacutils)
cat /var/lib/pacman/local/PKG($version)/desc

Thank you, figured it all out already. Still learning...

Anyway, the issue is udev 256 + xfce = /tmp/.ICE-unix lockfiles errors.