Forum › Forums › New users › New Users and General Questions › Can’t login after update
- This topic has 52 replies, 9 voices, and was last updated Nov 3-9:30 am by sybok.
-
AuthorPosts
-
October 24, 2025 at 11:13 am #188395
lgj100
MemberHi Everyone,
This morning I did a regular apt update; apt upgrade on my system. The last one before that was perhaps a week or two back. The upgrade today was very big, 600MB or so.
After the upgrade I rebooted and my graphical login screen is suddenly unresponsive (again). I see the login field in the graphical login manager, but there is no response when I type, and the only way to shut down is to hold he power button for 5 seconds. The keyboard doesn’t work. The behavior is *exactly* as described in the thread: http://www.antixforum.com/forums/topic/suddenly-cant-login-through-graphical-display-manager/ where the problem was caused by new version of eudev. Could that be the same here?
Sincerely,
Lars.October 24, 2025 at 12:40 pm #188414
sybokMemberThis is a thing I experience occasionally – it appears, then it goes away, then it reappears… I have no idea why it is so in my case.
Did your breaking update include ‘eudev’ (see ‘/var/log/apt/history.log’ if unsure)?
October 24, 2025 at 1:26 pm #188428lgj100
Member@sybok
Yes, eudev was a part of the update. I haven’t had the time to try the same fix, but I would assume it works again. I would venture that the reason you see this problem keep coming back is that the regular eudev is part of the debian update and the antix folks have not yet had the time to make a custom antix version of eudev. I’m only speculating….but it isn’t great if this happens every time debian updates eudev. Is there a way for antix to handle this better?
Sincerely,
Lars.October 24, 2025 at 9:38 pm #188490stevesr0
MemberHi lgj100,
If the issue is the change of packages from an antiX or nosystemd version to a Debian version, this can be prevented by doing
sudo apt upgrade -s | grep antixand
sudo apt upgrade -s | grep nosystemdand holding all packages that change from an antiX or nosystemd version to a Debian one.
To check if a eudev upgrade is involved, use
sudo apt upgrade -s | grep eudevto see if eudev is changing from an antiX version to a Debian one, before allowing the upgrade.
October 25, 2025 at 10:11 pm #188595lgj100
MemberHi Everyone,
This was definitely causes by the eudev upgrade. I uninstalled the antix5 versions of eudev, libeudev1 and libeudev-dev, and installed the antix1 versions, just as Robin suggested the last time this happened, after the upgrade that installed the antix2 versions of above packages. Instructions are here, http://www.antixforum.com/forums/topic/suddenly-cant-login-through-graphical-display-manager/page/3/#post-183003 Downgrading is a cumbersome process….I couldn’t get apt to do it – I had to use dpkg with –force-all. And now I can’t upgrade my system again because it has broken packages….
My question is why this is happening again? Wasn’t whatever was wrong with the antix2 versions fixed in the antix5 versions? I hope someone can get to the bottom of this so it doesn’t happen again. A regular apt upgrade shouldn’t disable the system this way….
Sincerely,
Lars.October 26, 2025 at 3:15 am #1885982n-1
MemberMy question is why this is happening again? Wasn’t whatever was wrong with the antix2 versions fixed in the antix5 versions? I hope someone can get to the bottom of this so it doesn’t happen again. A regular apt upgrade shouldn’t disable the system this way….
Sounds like something devuan solved by basing its updates on an “amprolla”-merged repository: https://www.devuan.org/os/development
- This reply was modified 11 months ago by 2n-1.
October 26, 2025 at 3:30 am #188601
anticapitalistaForum Admin@lgj100 What I find strange is that you are the only one to post about this unfortunate issue.
The updates to eudev have been made to enable multi-init booting.Philosophers have interpreted the world in many ways; the point is to change it.
antiX with runit - leaner and meaner.
October 26, 2025 at 7:22 am #188624
sybokMemberWhat I find strange is that you are the only one to post about this unfortunate issue.
I posted about the same symptoms in the past when running testing, http://www.antixforum.com/forums/topic/testing-can-no-longer-login-with-kernel-5-10/ , but I accepted is it as a quirk and work-around about it in my multi-boot system by switching the PC off and taking a short break.
October 26, 2025 at 7:34 am #188627
anticapitalistaForum Admin@sybok – your issue (from your link) was 3 years ago so I doubt it is the same.
Philosophers have interpreted the world in many ways; the point is to change it.
antiX with runit - leaner and meaner.
October 26, 2025 at 8:34 am #1886432n-1
MemberMaybe a glitch/delay in an antix mirror syncing (falling behind debian packages) that didn’t affect all users?
October 26, 2025 at 12:10 pm #188666lgj100
Member@anticapitalista
There seem to be some strange correlation between (i) the graphical login screen resolution, (ii) whether the graphical login screen responds, i.e. whether the keyboard works on the screen or not, and (iii) the console font that comes up when I boot. If the console font comes up “small” then the graphical login screen comes up in high resolution, and it is responsive and I can login. If the console font is large upon boot, then the resolution of on the graphical login screen is low and the page is unresponsive….
Also, the only kernel that allows me to login through the graphics manager, if I’m able to login, is the debian 6.12.12-1 bpo12+1 kernel.
It is somewhat random what console font the system boots with. How do I force the small console font upon boot?
Not sure why…
Sincerely,
Lars.- This reply was modified 11 months ago by lgj100.
- This reply was modified 11 months ago by lgj100.
October 26, 2025 at 5:53 pm #188705lgj100
MemberHi Everyone,
OK, so I booted the live usb, and ran the chroot rescue and removed the debian 6.12.12 kernel, and the system seems to be booting normally again, even with the antix5 versions of eudev. I find it strange that installing debian kernel 6.12.12 can affect booting into the standard antix kernels, like 6.1.42 which I also have installed. Before removing 6.12.12, the system wouldn’t boot into neither 6.1.42, nor 5.10 (except in “single” mode). It would give some incomprehensible error and ask me to look at /var/log/Xorg.0.log. Now it boots into both 5.10 and 6.1.42. I had installed the debian 6.12 kernel half a year ago or so from the CLI Apt-Based Package Manager.
Sincerely,
Lars.- This reply was modified 11 months ago by lgj100.
- This reply was modified 11 months ago by lgj100.
- This reply was modified 11 months ago by lgj100.
October 26, 2025 at 6:15 pm #188712
anticapitalistaForum Admin@lgj100 – glad to hear it seems to be ok now.
Since the issue seems to be random, I would think the issue is hardware related – ie a dying drive.
Make sure you have any important data saved elsewhere.Philosophers have interpreted the world in many ways; the point is to change it.
antiX with runit - leaner and meaner.
October 26, 2025 at 7:37 pm #188718lgj100
Member@anticapitalista
That would not be the case, because this happens on two PCs with different hardware and identical software setup.
Any other suggestions?
Sincerely,
Lars.October 27, 2025 at 1:18 am #188728
sybokMember@anticapitalista: I know it is old and you might be true.
On the other hand, it could be something similar as @stevesr0 .
The issue was present in testing, then it appeared in stable and it comes and goes as I wrote above.
I do not reboot (into antiX) that often (I do not use the other systems that much, mostly keeping them up to date) hence I could not reliably track its appearance or disappearance, hence no solid information as @lgj100 posted tying it to a specific package. -
AuthorPosts
- You must be logged in to reply to this topic.