I’ve installed video-hybrid-intel-nvidia-470xx-prime, rebooted and see no changes.
The icons and fonts has recovered.
I noticed that even just checking off Enabled on “Buil-in Screen” causes system to hang out/slow down.
Should I proceed with sudo DOWNGRADE_FROM_ALA=1 downgrade kscreen libkscreen ?
yes, you have nothing to lose… but you have to install the package downgrade first, then proceed with the command, and select the previous versions, where it worked, reboot … (thats assuming kscreen is the cause)
well, I did a couple of downgrades up to 5.25 … and it did not helped.
It’s unclear how many downgrades(with reboots) should be made.
well thats unfortunate… did you downgrade both of the packages, the kscreen + libkscreen?
do you have a timeshift backup?
also provide logs, maybe there will be something:
journalctl -b0 -p4 --no-pager
yes
no
(forum throws error due to big file, here’s a link)
I have some trouble in one time
For me help Downgrade kscreen and libkscreen to 5.25.1
- Install downgrade
sudo pamac install downgrade - Enter the command
sudo DOWNGRADE_FROM_ALA=1 downgrade kscreen libkscreen
and choose verison 5.25.1 - Reboot
sudo reboot
i found this in other topic
Icons on the taskbar are not visible (drivers problem)
I too have nvidia gpu + intel gpu
I try delete cache, edit /etc/default/grub, reinstall plasma \ xorg
But now we have only 2 work way how to fix it
- Downgrade
kscreenandlibkscreento 5.25.1<= - Install Plasma LTSC ( include
kscreenandlibkscreento 5.25.1<= ) - Not per-moment - When your log on in plasma, try to upload new wallper< you screen to bee a black and all to be normal ( after logout, need do it again )
I use video-hybrid-intel-nvidia-prime
and for me second display output only normal in thunderbolt → hdmi
Second fix your can try -
Install optimus manager
and use only Nvidia gpu
optimus-manager --switch nvidia
you should use timeshift, very usefull …
in the logs there was some logs referring to bbswitch, and i think 6 coredumps of plasmashell…
post output from:
pacman -Qm
pacman -Qs 'optimus|bbswitch|bumblebee'
sorry for delay (I’ve been under power cut due to russian fascists fire attacks to our critical infrastructure)
I’ve installed timeshift (though current snapshots are “erroneous”)
$ pacman -Qm
abr2gbr 1.0.2-1
advancecomp 2.1-3
audio-recorder 3.3.4-1
crow-translate 2.9.1-1
cryptoplugin 1.2.2-2
deadbeef 1.8.8-3
djvu2pdf 0.9.2-1
electron12 12.2.3-4
etcher 1.7.9-1
exaile 4.1.1-2
extundelete 0.2.4-5
flashplugin 32.0.0.465-1
fotoxx 22.30-1
glib 1.2.10-18
google-chrome 105.0.5195.125-1
jdk 19.0.1-1
jdk8 8u251-1
jre 19.0.1-1
kde-servicemenus-rootactions 2.9.1-1
key-mapper-git 1.0.0+62+g3e20cb9-1
libcanberra-gstreamer 0.30+2+gc0620e4-3
libdbusmenu-qt4 0.9.3+16.04.20160218-2
libguess 1.2-4
libreoffice-dev-bin 7.3.0.1-1
masterpdfeditor-free 4.3.89-1
metis 5.1.0.p10-2
noto-fonts-compat 20151217-1
ocrmypdf 13.2.0-1
openssl-1.0 1.0.2.u-1
package-query 1.12-1
pdf-xchange 9.2.359.0-1
pdfsam 4.2.10-1
playonlinux 4.3.4-4
python-coloredlogs 15.0.1-2
python2 2.7.18-5
python2-wxpython3 3.0.2.0-3
qpdfview 0.4.18-2
qt4 4.8.7-34
qt5-canvas3d 5.12.4-1
qt5-webkit 5.212.0alpha4-18
sky 2.1.7801-1
skypeforlinux-preview-bin 8.80.76.112-1
sni-qt 0.2.7+17.04.20170112-1
teamviewer 15.25.5-1
trimage 1.0.6-2
ttf-wps-fonts 1.0-5
tuxguitar 1.5.4-2
tuxguitar-common 1.5.4-1
tuxguitar-fluidsynth 1.2.20-1
wkhtmltopdf 0.12.6-1
wps-office 11.1.0.10702-2
wps-office-extension-russian-dictionary 20131121-4
wps-office-extension-ukrainian-dictionary 20131121-4
wps-office-fonts 1.0-2
xdg-su 1.2.3-2
xf86-input-keyboard 1.9.0-3
xf86-input-mouse 1.9.3-1
yaourt 1.9-1
zoom 5.9.1-1
And no output for pacman -Qs 'optimus|bbswitch|bumblebee'.
Thanks, I’ve already tried downgrade kscreen and libkscreen to 5.25.1 - it did not help.
As for downgrading to plasma LTS 5.25: I’m not sure if that could effect newly installed kde applications that rely on the latest version.
not sure if optimus is better than nvidia-prime overall.
nothing suspicious in the pacman -Qm output…
you had something in the logs regarding bbswitch…
switching completely to nvidia could be worth a try… use envycontrol:
pamac build envycontrol
when done, switch to nvidia:
sudo envycontrol -s nvidia
it will prompt you for reboot, so do it, and see if that helps…
Doesn’t it eliminate the role of Intel graphics card as the one that freeing up the Nvidia card from full load?
The confrontational “activity” between current nvidia-prime and envycontrol is excluded?
well yes, you will be 100% on nvidia …
but this is only for testing purposes…
and what activity? dont understand what you mean
I mean: could nvidia-prime somehow conflict with envycontrol (when both installed) in managing hybrid configuration?
no it doesnt conflict, i have also intel+nvidia prime, and use envycontrol, and no issues …
Building envycontrol...
==> Making package: envycontrol 2.1.1-1 (вт, 27-гру-2022 23:19:43 +0200)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
-> Updating envycontrol git repo...
==> Validating source files with sha256sums...
envycontrol ... Skipped
==> Removing existing $srcdir/ directory...
==> Extracting sources...
-> Creating working copy of envycontrol git repo...
Cloning into 'envycontrol'...
done.
Switched to a new branch 'makepkg'
==> Removing existing $pkgdir/ directory...
==> Entering fakeroot environment...
==> Starting package()...
Traceback (most recent call last):
File "/var/tmp/pamac-build-roman/envycontrol/src/envycontrol/setup.py", line 2, in <module>
from setuptools import setup
File "/usr/lib/python3.10/site-packages/setuptools/__init__.py", line 247, in <module>
monkey.patch_all()
File "/usr/lib/python3.10/site-packages/setuptools/monkey.py", line 97, in patch_all
patch_for_msvc_specialized_compiler()
File "/usr/lib/python3.10/site-packages/setuptools/monkey.py", line 134, in patch_for_msvc_specialized_compiler
msvc = import_module('setuptools.msvc')
File "/usr/lib/python3.10/importlib/__init__.py", line 126, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
File "/usr/lib/python3.10/site-packages/setuptools/msvc.py", line 24, in <module>
from packaging.version import LegacyVersion
ImportError: cannot import name 'LegacyVersion' from 'packaging.version' (/usr/lib/python3.10/site-packages/packaging/version.py)
==> ERROR: A failure occurred in package().
Aborting...
(packaging and setuptools are installed in latest versions)
no errors reported on aur site … can you try build some other package from aur, for example fastfetch:
pamac build fastfetch
to see if it works
I’ve launched pamac install jdk today (as timeshift dependency) successfully.
...
usr/bin/fastfetch
usr/bin/flashfetch
==> Creating package "fastfetch"...
-> Generating .PKGINFO file...
-> Generating .BUILDINFO file...
-> Generating .MTREE file...
-> Compressing package...
==> Leaving fakeroot environment.
==> Finished making: fastfetch 1.7.5-1 (вт, 27-гру-2022 23:46:11 +0200)
==> Cleaning up...
Checking keyring... [1/1]
Checking integrity... [1/1]
Loading packages files... [1/1]
Checking file conflicts... [1/1]
Checking available disk space... [1/1]
Installing fastfetch (1.7.5-1)... [1/1]
Transaction successfully finished.
all right, so theres something wrong with envycontrol…
so install optimus manager, it resets the graphics configs, so maybe it will help:
sudo pacman -S optimus-manager
dont install any dependecies like acpi call, bbswitch, only optimus…
then do this:
edit the file kate /etc/sddm.conf and simply put a # before the line starting with DisplayCommand and the one starting with DisplayStopCommand - you may not have these lines there…
reboot… if it didnt helped, try switching to nvidia:
1 Like
Ok, that seems working to get superficially expected behavior.
But, optimus-manager --switch hybrid still extremely lags/slows down the system.
so switching to nvidia works and you have no issue with the monitor?
what are the results when you switch to integrated?