I’ve been experiencing a strange issue with my Linux system recently. Whenever I move my mouse or drag a window, the display lags significantly. This lag wasn’t present before, and normally everything runs smoothly. I’m not sure what’s causing this problem.
I’m on Tumbleweed Plasma with an AMD Radeon 780m iGPU.
Is there a way to debug this issue further or any specific logs I should check?
Which TW version do you now have now? How old is this installation? How often have you been upgrading? How have you been upgrading?
With GUI problems the traditional place to start is examining Xorg.0.log, if it exists. Normal location is /var/log/Xorg.0.log, but it may be found in ~/.local/share/xorg/Xorg.0.log. (EE) errors after the first one or two are serious problems. If using Wayland to run Plasma instead of Xorg, then some messages that might have appeared in Xorg.0.log may appear in the Journal.
dmesg | less will allow you to scan for kernel error messages. System messages from current boot can be scanned with sudo journalctl -b. Specific errors can be found in journal or dmesg using grep, e.g.: sudo journalctl -b | grep aile would show lines that include string “Failed”; and dmesg | egrep ‘fault|aile’ would show lines that include failed or segfault.
I’ve got the last one (with kernel 6.10) and I’ve this one since 3 weeks now. I upgrade the system every 2 days (with discover first and after zypper dup to see if something is missing).
With sudo journalctl -b | grep aile, I’ve got something wrong with ACPI
août 01 09:09:55 localhost kernel: ACPI: _OSC evaluation for CPUs failed, trying _PDC
août 01 09:10:00 localhost systemd-vconsole-setup[831]: KD_FONT_OP_GET failed while trying to get the font metadata: Invalid argument
août 01 09:10:01 localhost (udev-worker)[1050]: controlC0: Process '/usr/sbin/alsactl restore /dev/snd/controlC0' failed with exit code 99.
août 01 09:10:02 localhost kernel: ACPI: \: failed to evaluate _DSM e8c3a8d2-694b-004f-82bd-fe8607803aa7 (0x1001)
août 01 09:10:02 localhost kernel: ACPI: \: failed to evaluate _DSM e8c3a8d2-694b-004f-82bd-fe8607803aa7 (0x1001)
août 01 09:10:01 localhost (udev-worker)[1050]: controlC1: Process '/usr/sbin/alsactl restore /dev/snd/controlC1' failed with exit code 99.
août 01 09:10:02 localhost irqbalance[1405]: thermal: failed to receive messages.
août 01 09:10:02 localhost irqbalance[1405]: Failed to initialize thermal events.
Bios version is from 1rst of July (new computer with AMD Ryzen 7 7840HS if it usefull )
Now that you have suspects, you can take a better look at each, lines before and lines following, looking for context, either by opening entire journal and searching for the found strings, with more specific searches, such as e.g. sudo journalctl -g | grep -B4 -A6 udev-worker to see 11 lines with udev-worker near the middle. Similar scan can be done with dmesg, e.g. dmesg | less, or broader look at one spot, e.g. dmesg | grep -B3 -A5 ACPI.
Note - whenever failures occur with acpi among them, acpi is less likely the main problem than most other failures. They’re often purely red herrings, journal noise.
AAhhh ! Finally, I managed to isolate the problem. It turns out that the issue is related to Bluetooth. For example, when I use the touchpad, there are no lags at all, so it doesn’t seem to be a major issue. When I search in journalctl with “sudo journalctl -b -u bluetooth” I’ve got this “src/device.c:set_wake_allowed_complete() Set device flags return status: Invalid Parameters”, so I know where I can investigate now. Thank you so much for your help!