Wayland session crashes and falls back to X11

Another experiment that would be useful: Can you create another user account and see if the same app startup issue exists on the new account?

Another check:
Open a terminal, become root with sudo -i then run:
ostree fsck
This will check system files for any possible corruption

Well, I have already created a new user on my PC, I have tested opening those applications that have some error and open without any problem, I am going to attach the diagnosis in case it is useful with that new user.
eos-diagnostic-240506_143218_UTC-0300.txt (344,8 KB)

This is what appears after executing the command:

james@janad:~$ sudo -i then run:
ostree fsck
[sudo] contraseña para james:
-bash: -c: línea 1: error sintáctico cerca del elemento inesperado then' -bash: -c: línea 1: then run:’
Validating refs…
Validating refs in collections…
Enumerating commits…
Verifying content integrity of 2 commit objects…
fsck objects (736/109197) [ ] 0%
error: In commits 8fe6cde4febf42b7103c9620a1d0e58f43a3d865a27007f05769405709f82a38: fsck content object 5f9b7e8a1e7951d6eda73f10a2010a2f634674feedbd4d346016fe565e732457: Opening content object 5f9b7e8a1e7951d6eda73f10a2010a2f634674feedbd4d346016fe565e732457: openat: Permiso denegado

Sorry I was not clear with those instructions. Please run it as:

sudo ostree fsck

This is what appears after executing the command:

james@janad:~$ coredumpctl
No coredumps found.

This is what appears after executing the command:

james@janad:~$ sudo ostree fsck
Validating refs…
Validating refs in collections…
Enumerating commits…
Verifying content integrity of 2 commit objects…
fsck objects (109197/109197) [=============] 100%
object fsck of 2 commits completed successfully - no errors found.

If i run coredumpctl info /usr/bin/gnome-shell > crash.txt, thats hapend:

james@janad:~$ coredumpctl info /usr/bin/gnome-shell > crash.txt
No coredumps found.

And appear and crash.txt empty that i can´t upload here.

Hello, is there still something that can be done? The same application execution problems continue and have begun to greatly hinder my academic work. Could you tell me if it is possible to return to the previous version? or if the problem is with my user, how can I format it without losing my information? I will be attentive to any response.

I think the simplest thing for now is to force Xorg to be used. In a terminal, run sudoedit /etc/gdm3/daemon.conf. Make 2 changes to the file:

  1. Uncomment the line #WaylandEnable=false by removing the # character at the beginning of the line.
  2. Uncomment the line #Enable=true at the bottom of the file by removing the # character at the beginning of the line.

Save the file after making the changes. You’ll have to reboot for the changes to take effect.

If things are working, then opening a terminal after rebooting and running echo $XDG_SESSION_TYPE should show x11. If it still says wayland, then the changes haven’t taken effect.

I have done each of the steps, and this is what resulted in the terminal;

Captura desde 2024-05-20 17-04-41

Don’t spawn anywhere “wyland”, now what should I do, or what should I expect to happen? I will be attentive to send any information.

Try running any apps you were having problems with before like LibreOffice and see if they display correctly.

I have carried out new tests with different applications, but the problem persists randomly; Sometimes, applications open without any problem, but more often they do not open, or they only run, after several attempts, in safe mode.

That is very strange. Can you upload a diagnostics file after that happens?

Well, I’m going to add the diagnosis, but I’m going to add other information.

Diagnosis: eos-diagnostic-240521_195844_UTC-0300.txt (1,8 MB)

Additional Information
After the tests carried out yesterday, the following message has appeared in the application center:

The thing is that I don’t have Rstudio installed, but I tried to install it some time ago, and I had used it through RCRAN, however, I don’t understand why this error appears.

In addition, the icons of some installed applications, and others not installed, do not appear or appear with an error, as shown in the following images:

Finally, the operation of the applications continues to be erratic, some work, others do not, they take a long time to open, they open in safe mode, or they simply do not open.

Additionally, I performed an update through the sudo flatpak repair command, waiting for the solution on the applications that are presenting errors, and this is the message that was returned:

Working on the system installation at /var/lib/flatpak
[8/211] Verificando flathub:runtime/org.kde.WaylandDecoration.QGnomePlatform-dec

[211/211] Verificando flathub:runtime/org.kde.Platform.Locale/x86_64/5.15-22.08…
Comprobando remotos…
Podando objetos
Eliminando .removed
Reinstalling removed refs
Instalando app/com.google.Chrome/x86_64/eos3

Note that ‘/var/lib/flatpak/exports/share’ is not in the search path
set by the XDG_DATA_HOME and XDG_DATA_DIRS
environment variables, so applications may not
be able to find it until you set them. The
directories currently searched are:

  • /root/.local/share
  • /usr/local/share/
  • /usr/share/

james@janad:~$

I will be available to continue sending information. And thanks for help

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.

After some problems in one of the previous posts, where I reported that my applications are not working correctly, or that they never open, I have seen that the Wayland system on my laptop is not working; I have reached this conclusion after reviewing the suggestion about removing the # symbol from the #WaylandEnable=false line and the same symbol from the #Enable=true line, with the aim of forcing the system. Once that symbol was deactivated, after executing the command echo $XDG_SESSION_TYPE , wayland was not active, all this did not solve the problem and my applications continued to crash, but then I thought of putting the # symbol back on the lines where I had removed it and run the command echo $XDG_SESSION_TYPE , to see if wayland would be activated, but to my surprise, it is not active, only the message x11 appears; Isn’t it supposed to be active and working correctly? Is it possible that this error is causing my applications to not work or open?

Yes this roughly is in agreement with our investigation so far. The system tries to launch Wayland, but it crashes in a very unusual place, and so it falls back on the X11 display server. But the desktop continues to launch in confused state, where it expects to talk to Wayland, and then all the apps get confused.

The question is indeed: why is Wayland crashing? Sorry that we haven’t had time to further investigate this very strange case. We have not had any other reports of this.

Something in your user’s home directory is causing this crash - we don’t know what. Practically speaking it probably makes most sense for you to “start again” with a new user account.

I understand, so I have a question regarding the use of a new user; Can I somehow bring all my information, data, documents, etc. to the new user? or all the information would be lost. I will be attentive to any response.