no GUI after upgrade to buster

[djk44883]

If you did an upgrade and not a new installation... why "I did format"???

at first I did regular upgrade (I followed the whole procedure update, dist upgrade, change source.list update.. etc - I probably followed every step and in the right order, on my other laptop it worked but there are only 2 systems and no separate /, /home)
but it failed, no GUI, so I installed buster from CD on top of lame buster which was already there. this ofcourse did not produce expected result, then I installed from CD several times, to no avail, no GUI.

oh my! You might have considered asking for assistance... before you, uhm, well here you are now.

/etc/mtab does show /home is mounted. When your system does boot "into /" Do you have any messages? Have you tried startx ?

the last line in mtab -

Code: Select all

tmpfs /run/user/0 tmpfs rw,nosuid,nodev,relatime,size=346460k,mode=700 0 0

This is above me, but I don't think that's you logged on your system as a regular user. I think you're user 1000, maybe -
tmpfs /run/user/1000 tmpfs rw,nosuid,nodev,relatime,size=815544k,mode=700,uid=1000,gid=1000 0 0
since I'm not clear on this, I don't know where to go yet. No just constantly re-installing probably won't fix something if you don't know what you're fixing.

No busybox? Not in maintenance mode? Not using root password?

One more thing, if you re-installed, then the boot menu comes from the debian installation now? or has it always?

[serandrzej]

oh my! You might have considered asking for assistance... before you, uhm, well here you are now.

yes, but I did consulted the net, and thought that I followed properly all prescribed procedure.

/etc/mtab does show /home is mounted. When your system does boot "into /" Do you have any messages? Have you tried startx ?

just tried starx. and what struck me is that it says "operating system Linux 4.9.0-8.."
while uname -a returns 4.19.0-5
also "KERNEL command line: BOOT_IMAGE_=(..) root=UUID=30a48c54-5cd5-4607-a70d-a1b7ac761ce0 ro quiet"
something feels wrong with the UUID.
then it reports also several errors:
xinit: giving up
xinit: unable to connect to X server: connection refused
xinit: server error

and when buster boots into /, yes there are quite a number of messages.

No busybox? Not in maintenance mode? Not using root password?

I can do anything only on root password, my account not on sudoers list.
I think I tried in maintenance mode (will try again) I managed to go into recovery mode - but no success. what's busy box - I don't know.

One more thing, if you re-installed, then the boot menu comes from the debian installation now? or has it always?

boot menu is always debian's. except when I ran update-grub from ubuntu. after another reistall of buster it's again debian's.

[djk44883]

just tried starx. and what struck me is that it says "operating system Linux 4.9.0-8.."
while uname -a returns 4.19.0-5
also "KERNEL command line: BOOT_IMAGE_=(..) root=UUID=30a48c54-5cd5-4607-a70d-a1b7ac761ce0 ro quiet"
something feels wrong with the UUID.
then it reports also several errors:
xinit: giving up
xinit: unable to connect to X server: connection refused
xinit: server erro

Maybe your installation isn't so messed up as your booting wrong. That UUID doesn't match any that you posed in your first blkid... so yea, that seems messed up.

Since I'm just getting pieces, can't say for sure. Your user not being in sudo, but you can log in as a "normal" user? (Of course, as root you could add yourself to sudo, but that's a different issue) Just wonder if your account exists and is functional.

You really don't want to see something about busybox. It's what I get when a system fails to boot, part way through, and pukes into maintenance mode, but only if you've set a root password.

Although systemd has it's journalctl facility, there are still some info in /var/log files. It's a pita via a console, but if you can access the filesystem via ubuntu see if there's anything. Check /var/log/installer and see if the install left any messages about not finishing.

[serandrzej]

root=UUID=30a48c54-5cd5-4607-a70d-a1b7ac761ce0 ro quiet"
That UUID doesn't match any that you posed in your first blkid... so yea, that seems messed up.

correction, UUID=30a48c54-5cd5-4607-a70d-a1b7ac761ce0 is now assigned to dev/sda5 - root.
As the partition has been formatted several times since my first blkid, the UUID must have changed. So the UUID is ok. should have updated blkid. my mistake.

Also startx reads "Build Operating System Linux 4.9.0-8 (...)
Current Operating System Linux debian1 4.19.0-5..." - so my previous information was misleading, sorry for that

[serandrzej]

/etc/mtab does show /home is mounted. When your system does boot "into /" Do you have any messages? Have you tried startx ?

I can't upload dmesg file, and pasting the result wouldn't make sense. what I'm getting on booting and also found in dmesg is that (30 or so same lines) :

[ 14.042288] pstore: crypto_comp_decompress failed, ret = -22!
[ 14.042400] pstore: decompression failed: -22
[ 14.043868] pstore: crypto_comp_decompress failed, ret = -22!
[ 14.043976] pstore: decompression failed: -22

also something abt radeon (radeon kernel modesetting for R600 or later requires firmware installed), amdgpu (amdgpu requires firmware installed) - but these ones showed also on stretch and jessie - don't have any impact probably

[serandrzej]

Although systemd has it's journalctl facility, there are still some info in /var/log files. It's a pita via a console, but if you can access the filesystem via ubuntu see if there's anything. Check /var/log/installer and see if the install left any messages about not finishing.

nothing suspicious in /var/log/installer, in status file all packages have "Status: install ok" and then "installed" or "unpacked". neither did I spot anything in hardware-summary and Xorg.0.log (though I don't know how something wrong should look like).

Then in deamon.log I found a number of entries about lightdm failure:

Aug 11 07:30:42 debian1 systemd[1]: lightdm.service: Service RestartSec=100ms expired, scheduling restart.
Aug 11 07:30:42 debian1 systemd[1]: lightdm.service: Scheduled restart job, restart counter is at 3.
Aug 11 07:30:42 debian1 systemd[1]: Stopped Light Display Manager.
Aug 11 07:30:42 debian1 systemd[1]: Starting Light Display Manager...
Aug 11 07:30:42 debian1 systemd[1]: systemd-update-utmp-runlevel.service: Succeeded.
Aug 11 07:30:42 debian1 systemd[1]: Started Update UTMP about System Runlevel Changes.
Aug 11 07:30:42 debian1 systemd[1]: Started Light Display Manager.
Aug 11 07:30:42 debian1 systemd[1]: lightdm.service: Main process exited, code=exited, status=1/FAILURE
Aug 11 07:30:42 debian1 systemd[1]: lightdm.service: Failed with result 'exit-code'.
Aug 11 07:30:42 debian1 systemd[1]: lightdm.service: Service RestartSec=100ms expired, scheduling restart.
Aug 11 07:30:42 debian1 systemd[1]: lightdm.service: Scheduled restart job, restart counter is at 4.
Aug 11 07:30:42 debian1 systemd[1]: Stopped Light Display Manager.
Aug 11 07:30:42 debian1 systemd[1]: Starting Light Display Manager...
Aug 11 07:30:43 debian1 systemd[1]: Started Light Display Manager.
Aug 11 07:30:43 debian1 systemd[1]: lightdm.service: Main process exited, code=exited, status=1/FAILURE
Aug 11 07:30:43 debian1 systemd[1]: lightdm.service: Failed with result 'exit-code'.
Aug 11 07:30:43 debian1 systemd[1]: lightdm.service: Service RestartSec=100ms expired, scheduling restart.
Aug 11 07:30:43 debian1 systemd[1]: lightdm.service: Scheduled restart job, restart counter is at 5.
Aug 11 07:30:43 debian1 systemd[1]: Stopped Light Display Manager.
Aug 11 07:30:43 debian1 systemd[1]: lightdm.service: Start request repeated too quickly.
Aug 11 07:30:43 debian1 systemd[1]: lightdm.service: Failed with result 'exit-code'.
Aug 11 07:30:43 debian1 systemd[1]: Failed to start Light Display Manager.
Aug 11 07:30:43 debian1 systemd[1]: lightdm.service: Triggering OnFailure= dependencies.
Aug 11 07:30:43 debian1 systemd[1]: Starting Terminate Plymouth Boot Screen...
Aug 11 07:30:43 debian1 systemd[1]: plymouth-quit.service: Succeeded.
Aug 11 07:30:43 debian1 systemd[1]: Started Terminate Plymouth Boot Screen.

[djk44883]

correction, UUID=30a48c54-5cd5-4607-a70d-a1b7ac761ce0 is now assigned to dev/sda5 - root.
As the partition has been formatted several times since my first blkid, the UUID must have changed. So the UUID is ok. should have updated blkid. my mistake.

alright... stop re-re-re-installing, you've proved it's not the answer.

Also startx reads "Build Operating System Linux 4.9.0-8 (...)
Current Operating System Linux debian1 4.19.0-5..." - so my previous information was misleading, sorry for that

That's not as alarming as you'd think. I checked my /var/log/Xorg.0.log -

Code: Select all

[     6.776] Build Operating System: Linux 4.9.0-8-amd64 x86_64 Debian
[     6.776] Current Operating System: Linux htpc 4.19.0-5-amd64 #1 SMP Debian 4.19.37-6 (2019-07-18) x86_64


Thinking the Build, is the system used to compile xorg-server - not your's! Misleading!? Your previous post truncated Build - "operating system Linux 4.9.0-8.." while negating the Current

also something abt radeon (radeon kernel modesetting for R600 or later requires firmware installed), amdgpu (amdgpu requires firmware installed) - but these ones showed also on stretch and jessie - don't have any impact probably

"requires" - probably no impact? https://wiki.debian.org/AtiHowTo But if you needed proprietary firmware, installer should have made it known.

Use your ubuntu system, mount the debian partition and find what errors are in /var/log/Xorg.0.log maybe, you have a video driver issue. Still, you never answer that you can log in as a regular user and access your home directory - your system is otherwise working.

Trying to understand, starting an X session and loading the desktop environment and/or window manager - is booting into "/home" and "into / root partition" is how you term a console session or the terminal?

[serandrzej]

Trying to understand, starting an X session and loading the desktop environment and/or window manager - is booting into "/home" and "into / root partition" is how you term a console session or the terminal?

yes, that's correct, I mean that's how I term those two processes, ie loading desktop env and opening console session.

[serandrzej]

Use your ubuntu system, mount the debian partition and find what errors are in /var/log/Xorg.0.log maybe, you have a video driver issue.

having run

Code: Select all

cat /var/log/Xorg.0.log | grep EE

in debian console I get:

Code: Select all

(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[    27.282] (EE) open /dev/dri/card0: No such file or directory
[    27.282] (EE) open /dev/dri/card0: No such file or directory
[    27.283] (EE) Unable to find a valid framebuffer device
[    27.284] (EE) Screen 0 deleted because of no matching config section.
[    27.284] (EE) Screen 0 deleted because of no matching config section.
[    27.284] (EE) Screen 0 deleted because of no matching config section.
[    27.284] (EE)
[    27.284] (EE) Cannot run in framebuffer mode. Please specify busIDs        for all framebuffer devices
[    27.284] (EE)
[    27.285] (EE)
[    27.285] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[    27.285] (EE)
[    27.295] (EE) Server terminated with error (1). Closing log file.

(the same entries I found viewing the file from ubuntu GUI.)
also, yes I may have issues with video drivers, but what puzzles me is why linux mint fails to load, even if for some time it did load into GUI (after upgrade to buster, that is).
I found this https://bbs.archlinux.org/viewtopic.php?id=238302 - but this is about nvidia. now, sorry if it's a silly question - if I have radeon this means that I don't use nvidia, right?
Moreover, I recall having issues with lightdm on upgrade in the past but I don't remember how I solved them, probably by clean install. maybe it's an issue related to xfce flavour?

Still, you never answer that you can log in as a regular user and access your home directory - your system is otherwise working.

what do you mean exactly? I can log as a non-root user, just don't have sudo rights, at the moment I don't need them. do you need a feedback for this?

[djk44883]

cat /var/log/Xorg.0.log | grep EE

Sometimes, the thing with error, warning, panic, or fatal messages - you kinda want to know what happened before situations was created.

It is a great way to extract error messages.

[djk44883]

Here are other issues similar to your, using lightdm... see if there's a clue:

viewtopic.php?f=17&t=143041#p704874
viewtopic.php?f=10&t=142383#p700796

[serandrzej]

Here are other issues similar to your, using lightdm... see if there's a clue:

viewtopic.php?f=17&t=143041#p704874
viewtopic.php?f=10&t=142383#p700796

didn't work for me.
however, I installed gdm3 - now I can see graphical greeting screen but when I try to log on as my user, nothing happens, the screen is back to user lists. gdm3 cannot start X either.
startx again "is unable to connect to X server: Connection refused'

since GUI, as it seems, is able to work, if only for gdm3 greeting screen, Im wondering maybe I should edit some boot files and indicate home partition as the one system should boot into, instead of root partition.
that is, replace the line (somewhere)
BOOT_IMAGE=/boot/vmlinuz-4.19.0-5-amd64 root=UUID=30a48c54...761ce0 ro quiet
with
BOOT_IMAGE=/boot/vmlinuz-4.19.0-5-amd64 home=UUID=9f475ba3..e4a720 ro quiet

[djk44883]

since GUI, as it seems, is able to work, if only for gdm3 greeting screen, Im wondering maybe I should edit some boot files and indicate home partition as the one system should boot into, instead of root partition.

You system should boot should boot into an environment, not a partition. Filesystems or partitions are mounted during the booting process.

Im wondering maybe I should edit some boot files and indicate home partition as the one system should boot into, instead of root partition.
that is, replace the line (somewhere)
BOOT_IMAGE=/boot/vmlinuz-4.19.0-5-amd64 root=UUID=30a48c54...761ce0 ro quiet
with
BOOT_IMAGE=/boot/vmlinuz-4.19.0-5-amd64 home=UUID=9f475ba3..e4a720 ro quiet

I'm wondering WTF? If you found some option to add to the kernel line to "boot into /home" then you can declare victory.

[serandrzej]

You system should boot should boot into an environment, not a partition.

yes, sorry for my awkward use of technical terms, I have background in linguistics and literature, so I'm rather a laymen when it comes to IT.

Im wondering maybe I should edit some boot files and indicate home partition as the one system should boot into, instead of root partition.
that is, replace the line (somewhere)
BOOT_IMAGE=/boot/vmlinuz-4.19.0-5-amd64 root=UUID=30a48c54...761ce0 ro quiet
with
BOOT_IMAGE=/boot/vmlinuz-4.19.0-5-amd64 home=UUID=9f475ba3..e4a720 ro quiet

I'm wondering WTF? If you found some option to add to the kernel line to "boot into /home" then you can declare victory.

I wouldn't like to get ahead of myself and prematurely declare victory. If I recall correctly, on stretch it was always home partition that was mounted, not root, so I was just hypothesising on a possibility of instructing buster to boot into graphical environment by placing this entry "BOOT_IMAGE=/boot/vmlinuz-4.19.0-5-amd64 home=UUID=9f475ba3..e4a720 ro quiet" in a file which controls booting. The question is whether it would be workable, and besides, I have not a foggiest idea which file(s) I should put this entry in. any tips where such instruction could be placed? grub.cfg?

[Head_on_a_Stick]

But if you needed proprietary firmware, installer should have made it known.

The installer doesn't warn about firmware requirements for video cards.

@OP: try installing the firmware-linux-nonfree metapackage.

I was just hypothesising on a possibility of instructing buster to boot into graphical environment by placing this entry "BOOT_IMAGE=/boot/vmlinuz-4.19.0-5-amd64 home=UUID=9f475ba3..e4a720 ro quiet" in a file which controls booting

The kernel doesn't understand home= as a command line parameter and anyway it would make no sense to only mount /home