One key thing I've noticed about Wayland instability on my box

Before 6.0.2 Wayland was crashing all the time. Now it seems whenever it would have crashed in 6.0.0 it now loses responsiveness for a number of seconds.

Do you have any idea what the problem is and if this is being fixed?

Quite annoying and bothersome.

With that said, I am surprised 6.0.2 fixed all the crashes.

I am using this processor:

Intel(R) Pentium(R) 3805U @ 1.90GHz

And 12 GB of RAM.

Forgot to mention I am on 6.0.3 now with all the latest updates.

Without snippets of stack traces for any of the crashes, or any discernible pattern about what causes these crashes/slowdowns, it’s not really possible to tell.

The best way to have things fixed is to make sure that a developer can reproduce these issues on their development system. The second-best way is to provide enough specific information about the error so that an interested developer can make the right guess about where the problem lies, close enough to find the error in the corresponding part of the code. Without either of those, it gets excruciatingly hard to get a bugfix.

My understanding is that any regular KDE developers aren’t seeing constant interruptions in responsiveness, so it’s important to figure out somehow why those are showing up on your system but not most other systems.

Maybe the old CPU is used together with some other very old components that don’t get much testing anymore? Perhaps you’re still using an HDD and Baloo’s “get out of the way” heuristics aren’t doing their job when the disk is orders of magnitude slower than an average SSD? Perhaps your GPU driver isn’t being actively maintained anymore and Wayland’s use of more modern GPU features causes it to fail? Or maybe just the plain old “memory is starting to fail and it surfaces in all kinds of weird ways”?

Maybe you can find some relevant information in your system logs (journalctl --boot if using systemd). Or you can try selectively changing some choice settings in the Desktop Search settings, Compositing settings (if on X11). Maybe you can find other people with similar hardware as yours reporting similar issues on modern distros. Without more actionable information, I fear that your question may be left unanswered.

1 Like

“Maybe the old CPU is used together with some other very old components that don’t get much testing anymore? Perhaps you’re still using an HDD and Baloo’s “get out of the way” heuristics aren’t doing their job when the disk is orders of magnitude slower than an average SSD?”

It is a very fast M.2 SSD. Realized I left that out of the initial post, but also not sure who would use an 3805U CPU with an HDD. These CPUs are for small form factor boxes.

The baloo file indexing constantly crashes, so whenever I check it is not running.

The GPU I used is integrated on the Intel CPU.

Mesa 24.0.3, HD Graphics, 100 MHz core clock. Driver used i915.

The memory is very fast at 1600 MT/s, DDR3.

CPU temp currently is at 57 degrees C.

I enjoy this box and hardware and find it perfectly sufficient for my office-based needs. Ran KDE Neon flawlessly til version 6.

Here is the drive: Kingston Digital 60GB SSDNow mSATA Solid State Drive for Mac and PC (SMS200S3/60G):

Form Factor - mSATA
Interface - mSATA Rev. 3.0 (6Gb/s) SATA Rev. 2.0 (3Gb/s), SATA Rev. 1.0 (1.5Gb/s)
Capacity - 60GB
Sequential reads - 550MB/s
Sequential Writes - 520MB/s

I have a big swap file at over 13 GB. About 5 GB of free space on the drive besides the large swap.

Is the loss of responsiveness tied to any particular workloads that you could observe, or does it seem completely random?

1 Like

I will have to get back to you about that because right this moment I cannot replicate it.

Will revisit this thread when it happens again to describe what I was doing. Have to step out right now.

1 Like

It is not predictable. Happened once when reading a USB drive while trying to navigate in Dolphin. Another time when opening Eddie, AirVPN’s client. Another time, the worst, was when I tried to resume from sleep. The screen completely froze, and after waiting a minute I had no choice but to do a hard reboot. Currently am back in X11 after that experience.