I’ve got a PC running Tumbleweed that has a NVIDIA GeForce GTX 670 in it. Recently, I did an upgrade on that system using the previous install .iso. When it rebooted, it went straight to CLI rather than GUI. So I tried again using the latest .iso (upgrade). Same result. I verified that all NVIDIA drivers (G05 for the 600 series card) are installed, as well as KDE, and X. If I try
init 5
nothing happens. No errors or anything, just a new prompt.
I’ve tried booting to recovery mode on the most current kernel (6.7.7.1), as well as on the older (6.7.6.1) kernel, and recovery mode on that kernel as well.
Following advice from other, similar threads, I ran the following commands:
systemctl get-default
-----Output:
graphical.target
systemctl list-unit-files --state=failed
-----Output:
UNIT FILE STATE PRESET
0 unit files listed.
systemctl isolate graphical.target
No output. Didn’t do anything.
journalctl -b | grep ailed
-----Output:
Mar 04 12:28:53 localhost (udev-worker)[638]: controlC0: Process ‘/usr/sbin/alsactl restore /dev/snd/controlC0’ failed with exit code 99.
Mar 04 12:28:57 localhost (udev-worker)[636]: event14: Failed to call EVIOCSKEYCODE with scan code 0x7c, and key code 190: Invalid argument
Mar 04 12:29:05 localhost nscd[976]: 976 stat failed for file /etc/services'; will try again later: No such file or directory Mar 04 12:29:05 localhost nscd[976]: 976 stat failed for file /etc/netgroup’; will try again later: No such file or directory
Mar 04 13:03:58 localhost.localdomain sudo[4460]: tonka : TTY=tty1 ; PWD=/var/log ; USER=root ; COMMAND=/usr/bin/systemctl list-unit-files state=failed
I backed up Xorg.0.log, and deleted the original. I’ve done all this, and rebooted and have nothing in Xorg.0.log.
Can anybody give me any suggestions on how I can get my GUI back?
I’m stuck on these comments (especially the 2nd comment), and feel it’s imperative that every TW user should have the minimum awareness of how TW should be updated.
So, you’ve ONLY updated TW by downloading a 4.x gb .ISO and doing the update offline. Why?
After TW is installed, a user should execute a
zypper dup
at the command line (root user acct). Many users run that on a daily basis, some once a week.
Personally (at this point) I’d do one thing. At the command line, execute
zypper -vv dup
and see what the results is.
If no complaints (or very minimal), proceed with the dup, then reboot when it’s done. All okay? move on.
However, If there’s a multitude of conflicts for dup, I’d simply use that recent .ISO and install a afresh (don’t forget to backup /home), then read the TW documentation about basic maintenance.
Technically there is absolutely no problem to use actual snapshot ISOs to upgrade a TW machine.
BUT
you need a stock TW installation without additional repos like packman, Nvidia or others!
the snapshot ISO needs to be newer than the system snapshot which will get upgraded
That means, if you have an additional non standard repo, the upgrade variant via an offline media might fail as there will be no updates for the other repos.
To get your system back into an consistent state, do a zypper dup and if it fails, do a fresh reinstallation and remember to use zypper dup for TW upgrade…
Using the online media for upgrade is also easier as you don’t need to dump the ISO to a pen drive …
UNIT LOAD ACTIVE SUB DESCRIPTION
0 loaded units listed.
Mar 04 13:53:12 localhost kernel: x86/cpu: VMX (outside TXT) disabled by BIOS
Mar 04 13:53:13 localhost kernel: ACPI BIOS Error (bug): Could not resolve symbol [_SB.PCI0.SAT0.SPT5._GTF.DSSP], AE_NOT_FOUND (20230628/psargs-330)
Mar 04 13:53:13 localhost kernel: ACPI Error: Aborting method _SB.PCI0.SAT0.SPT5._GTF due to previous error (AE_NOT_FOUND) (20230628/psparse-529)
Mar 04 13:53:13 localhost kernel: ACPI BIOS Error (bug): Could not resolve symbol [_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND (20230628/psargs-330)
Mar 04 13:53:13 localhost kernel: ACPI Error: Aborting method _SB.PCI0.SAT0.SPT4._GTF due to previous error (AE_NOT_FOUND) (20230628/psparse-529)
Mar 04 13:53:13 localhost kernel: ACPI BIOS Error (bug): Could not resolve symbol [_SB.PCI0.SAT0.SPT5._GTF.DSSP], AE_NOT_FOUND (20230628/psargs-330)
Mar 04 13:53:13 localhost kernel: ACPI Error: Aborting method _SB.PCI0.SAT0.SPT5._GTF due to previous error (AE_NOT_FOUND) (20230628/psparse-529)
Mar 04 13:53:13 localhost kernel: ACPI BIOS Error (bug): Could not resolve symbol [_SB.PCI0.SAT0.SPT4._GTF.DSSP], AE_NOT_FOUND (20230628/psargs-330)
Mar 04 13:53:13 localhost kernel: ACPI Error: Aborting method _SB.PCI0.SAT0.SPT4._GTF due to previous error (AE_NOT_FOUND) (20230628/psparse-529)
Mar 04 13:53:23 localhost (udev-worker)[626]: event10: Failed to call EVIOCSKEYCODE with scan code 0x7c, and key code 190: Invalid argument
That makes sense. I did, in fact, have packman and NVIDIA repos setup when I did that upgrade. However, afterwards, I added those repos and updated the packages from those two distros. So wouldn’t that fix the failed upgrade?
Looking at the results from my lsmod, it looks like my machine is trying to use the motherboard integrated video card. A while back I updated the motherboard BIOS (can we still call it that?), and in the process the onboard video must have gotten re-enabled. I’ve now disabled it.
I then, re-installed the NVIDIA drivers via yast2. However, when I lsmod | grep nvidia, it shows nothing. My lsmod still shows “asus wmi” for “video”.
It looks like either my BIOS is screwed up, or there’s a problem with my NVIDIA card. When I set the BIOS to boot to the PCIe device, it switches it back to AUTO, and boots form the onboard video (Asus wmi). I’m reverting my BIOS now. If that doesn’t help, I’ll test the video card in another PC. I’ll report the results either way.
It’s looking like I’ve got a hardware problem (motherboard most likely). After rebooting over and over trying to get the BIOS to keep my setting disabling the onboard video, and changing the BIOS version in the process, and trying multiple video cards, I ran into another problem. I booted, as I had been doing, and it only went into the GRUB prompt. ls was saying that the / partition was an unknown filesystem. I had that problem with another PC just recently, and went through the troubleshooting steps that were recommended here when I posted about it. I had the same result. I had to do another update via the install .iso. When that was done, it booted right to the GUI. WTH? This was with a different video card, and the video looked like hell. So I swapped out to my original video card, and it worked. Somehow, one of my user security settings got changed in the process, and when trying to fix that, the PC is now not booting at all. It hangs at the motherboard logo and the motherboard is showing a “boot device fault” LED. I’ve seen this issue off and on in the past, but now I can’t get past this.
I think I’m going to replace the motherboard, CPU, and RAM, and see if that helps. I’ll update when I know more, but that won’t be for at least a week or so.