Anybody else getting constant micro freezees (Like your mouse freezes for couple of seconds) and then randomly the whole GNOME just freezes, I can hear the music still working in background but only hard reset makes the system work again.
It was fine until I decided to update to Fedora 44.
Tried downgrading NVIDIA drivers to 580 branch but it did not help.
Yes, this is the third or fourth thread about it.
Others have reported that KDE is not exhibiting the behaviour.
I think bug reports are in order.
gsettings set org.gnome.desktop.interface enable-animations true
Im getting micro freezes in KDE also (not getting a whole freeze), from a clean boot and with no applications running. This is also on F43 and AMD. Only started a couple of days ago for me
Have a look at your
dnf history
to see what has recently been upgraded, and try
dnf downgrade packagename
with that.
good point, unfortunately I upgraded with discover so I cant just dnf downgrade, and
matthew@fedora ~> rpm -qa --last | grep “04 Apr 2026”
(plus the one for april 03) gives a total of 175 packages 
looks like I’ve got a bit of work to do
I have the same freezes as you describe here, but since a slightly longer time if I’m not wrong – perhaps since last week of March. At the beginning I thought it was related to Steam, but today I got it again even though I didn’t start this application.
I have the impression it always happen after resuming from sleep mode (but not always the first one). Only my mouse cursor does not seem to be affected.
I’m on Fedora 43, GNOME Shell and I have an AMD Radeon™ RX 5600 XT GPU.
Maybe it would be a good idea to watch what the system is doing to catch where these issues are originating from? You can open a terminal tab and just keep it running. When the crash/ freezes happens you should be able to catch it live.
journalctl -f -p err..emerg
If you need to do this after a restart :
journalctl -b -1 -p err..emerg
You could review the logs and we could help from there. Also review a previous boot logs and see if anything comes up.
journalctl -b -1 | grep -E "error|fail|timeout|drm|nvidia|amdgpu|gnome-shell|wayland"
I was having the freezes just right now, so I ran the first command and got this output:
avril 07 08:30:07 fedora kernel: kvm_intel: VMX not supported by CPU 5
avril 07 08:30:07 fedora kernel: kvm_intel: VMX not supported by CPU 1
avril 07 08:30:45 filius gdm-password][3264]: gkr-pam: unable to locate daemon control file
avril 07 08:30:48 filius systemd[3414]: Failed to start app-gnome-vmware\x2duser-4001.scope - Application launched by gnome-session-service.
avril 07 11:03:03 fedora bluetoothd[998]: Failed to set default system config for hci0
avril 07 11:03:07 fedora kernel: ata2: softreset failed (device not ready)
avril 07 14:53:09 fedora bluetoothd[998]: Failed to set default system config for hci0
avril 07 14:53:13 fedora kernel: ata2: softreset failed (device not ready)
avril 07 15:34:08 fedora bluetoothd[998]: Failed to set default system config for hci0
avril 07 15:34:12 fedora kernel: ata2: softreset failed (device not ready)
avril 07 16:16:44 fedora bluetoothd[998]: Failed to set default system config for hci0
avril 07 16:16:48 fedora kernel: ata2: softreset failed (device not ready)
avril 07 16:19:29 fedora bluetoothd[998]: Failed to set default system config for hci0
avril 07 16:19:33 fedora kernel: ata2: softreset failed (device not ready)
I can’t tell about the first lines but I can say that the four last ones have appeared just after resuming from sleep.
did you run journalctl -f -p or was this a log from previous boot.
Also check your BIOS/UEFI settings:
Then watch journalctl the next time you log back in.
Sorry, I ran journalctl -f -p err..emerg. It’s my current boot.
Also check your BIOS/UEFI settings:
- Make sure SATA mode is set to AHCI (not RAID or IDE).
Then watch journalctl the next time you log back in.
SATA mode is already set to AHCI. After reboot and logging in again, I get the following logs:
$ journalctl -f -p err..emerg
avril 07 16:42:00 fedora kernel: tpm_crb MSFT0101:00: [Firmware Bug]: ACPI region does not cover the entire command/response buffer. [mem 0xda1bf000-0xda1bffff flags 0x200] vs da1bf000 4000
avril 07 16:42:00 fedora kernel: tpm_crb MSFT0101:00: [Firmware Bug]: ACPI region does not cover the entire command/response buffer. [mem 0xda1c3000-0xda1c3fff flags 0x200] vs da1c3000 4000
avril 07 16:42:00 fedora systemd-modules-load[312]: Failed to find module 'v4l2loopback'
avril 07 16:42:08 fedora kernel: [drm:drm_mode_config_cleanup] *ERROR* connector Unknown-1 leaked!
avril 07 16:42:11 fedora bluetoothd[1092]: Failed to set default system config for hci0
avril 07 16:42:12 fedora kernel: kvm_intel: VMX not supported by CPU 11
avril 07 16:42:12 fedora kernel: kvm_intel: VMX not supported by CPU 9
avril 07 16:42:12 fedora kernel: kvm_intel: VMX not supported by CPU 11
avril 07 16:42:40 filius gdm-password][3168]: gkr-pam: unable to locate daemon control file
keep that terminal open and the command going until it happens again. I’m interested to see what comes up.
I can confirm now the issue happens after resuming from sleep, and it is cumulative: each time your PC goes to sleep and you wake it up, the freezes are more and more important.
Here is my complete journal. 17:08 and at 17:19 are the timestamps at which I resumed from sleep, and I have the freezes.
$ journalctl -f -p err..emerg
avril 07 16:42:00 fedora kernel: tpm_crb MSFT0101:00: [Firmware Bug]: ACPI region does not cover the entire command/response buffer. [mem 0xda1bf000-0xda1bffff flags 0x200] vs da1bf000 4000
avril 07 16:42:00 fedora kernel: tpm_crb MSFT0101:00: [Firmware Bug]: ACPI region does not cover the entire command/response buffer. [mem 0xda1c3000-0xda1c3fff flags 0x200] vs da1c3000 4000
avril 07 16:42:00 fedora systemd-modules-load[312]: Failed to find module 'v4l2loopback'
avril 07 16:42:08 fedora kernel: [drm:drm_mode_config_cleanup] *ERROR* connector Unknown-1 leaked!
avril 07 16:42:11 fedora bluetoothd[1092]: Failed to set default system config for hci0
avril 07 16:42:12 fedora kernel: kvm_intel: VMX not supported by CPU 11
avril 07 16:42:12 fedora kernel: kvm_intel: VMX not supported by CPU 9
avril 07 16:42:12 fedora kernel: kvm_intel: VMX not supported by CPU 11
avril 07 16:42:40 filius gdm-password][3168]: gkr-pam: unable to locate daemon control file
avril 07 17:08:05 fedora bluetoothd[1092]: Failed to set default system config for hci0
avril 07 17:08:09 fedora kernel: ata2: softreset failed (device not ready)
avril 07 17:19:34 fedora bluetoothd[1092]: Failed to set default system config for hci0
avril 07 17:19:38 fedora kernel: ata2: softreset failed (device not ready)
journalctl -b | grep -E “PM: suspend|PM: resume|suspending|resuming|sleep|wake” -A 10 -B 5
Try this to see if we get more info.
The output is too large (too many lines) to publish here, is there a recommended pastebin for this forum?
(Sorry, I’m new here ^^)
Yes, we have a fpaste I forgot what it is . . . let me look in the forum.
journalctl -b | grep -E “PM: suspend|PM: resume|suspending|resuming|sleep|wake” -A 10 -B 5 | fpaste --raw-url
Thank you, here are the logs: https://paste.centos.org/view/raw/bf2404a2
I hate to say, you might have to get more info on this. So, you could try to sleep/suspend a couple of times to see what come up. I personally think this is a power issue, but the ata errors would say it’s a controller issue. Either way, if you can update the firmware on your system as soon as posisble that could provide some assistance.
You can force update the firmware or there could be some in Software Center.
AFAIK this is a nvidia specific fix, but give this a try
how to install
# dnf copr enable anotheruser/gnome50
# dnf reinstall mutter\* --repo=copr:copr.fedorainfracloud.org:anotheruser:gnome50
# dnf copr disable anotheruser/gnome50
how to revert
# dnf distro-sync mutter\*
# dnf copr remove anotheruser/gnome50
it’s vanilla mutter 50.0 with 3 merge requests merged:
Revert "kms/impl-device: Remove META_DEADLINE_TIMER_STATE_INHIBITED" (!5018) · Merge requests · GNOME / mutter · GitLab
Guard early swap bail behind render source check (!5003) · Merge requests · GNOME / mutter · GitLab
onscreen/native: Move nvidia deadlock prevention to assign_next_frame (!5021) · Merge requests · GNOME / mutter · GitLab
EDIT: looks like this thread has been taken over by a F43 issue…