Can’t login after update

[Xunzi_23Member]

Hi lgj100,
removing a kernel freed up space, if your system is running out you will get
weird symptoms like you describe, and no graphical login.

how much free do you have in root and EFI plus home if separate.

If you use a chrome based browser how big is the cache, it can reach monster
size and prevent login due lack of free space.

Also after removing the Kernel grub configuration was updated, had you also tried using grub rescue

• This reply was modified 11 months ago by Xunzi_23.
• This reply was modified 11 months ago by Xunzi_23.

[2n-1Member]

Hm, another thing to check:

I experienced lighdm would not let me log in sometimes (black screen or loop) which seemed to be a timing issue. And I found it happend when the hostname changed after it started up, e.g. when your dhcp server takes a bit to answer and supplies a specific hostname.

So maybe something to check with the antix display manager.

[2n-1Member]

For example, log in on a console and restart the display manager.

[2n-1Member]

There is some SysReq code to take back control of the keyboard from X11. Think it’s Alt-SysRq-r https://en.wikipedia.org/wiki/Magic_SysRq_key

[andfreeMember]

lgj100 wrote:

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.

I don’t claim that they are similar cases but, if installing a standard antiX kernel 6.6 can affect booting into a standard antix kernel 5.10 or 6.1, why is it strange if installing a debian kernel 6.12 can also do the same?

antiX utilities for processing PDF files (posted by ChPol)

[Xunzi_23Member]

hi andfree, different kernels bring different drivers, means some run better than others on certain hardware,
I have also experienced complete boot failure, presumably due missing drivers.

debian 6.12 did not break my setup, neither did any other up to the 6.16.12+deb14-amd64 which I am using right now.

i have seen two main problems, grub update failed and lack of free disk space.
There are almost certainly other possible causes that can only be solved on a case to case basis.

[lgj100Member]

Hi Everyone,

@xunzi_23
There is plenty of disk space on both of my systems. There is no space issue. This is a software issue. Probably similar to what you describe in your last post. I have two machines with different hardware and identical software, and they behave in the same way, and the boot problem went away on both systems when I removed the debian 6.12 kernel.

The lesson for me is to be careful and expect issues when installing a non-antix kernel. Perhaps non-antix kernels should not be included in the CLI APT Based Package Manager? Installing a on-antix kermel could mean problems….

Sincerely,
Lars.

[Xunzi_23Member]

lgj100 wrote

The lesson for me is to be careful and expect issues when installing a non-antix kernel.
Perhaps non-antix kernels should not be included in the CLI APT Based Package Manager? Installing a on-antix kermel could mean problems….

The Linux Kernel is a central system component, drivers and features vary.. The replaced version is kept, rollback in case of problems.

Maybe you misunderstood, I had most problems with antiX kernel 5.10, that does not boot on my hardware, I get improved results with a very recent Debian Kernel
rather than with the latest antiX Kernel. That is a feature not a worry.

When you change a Kernel you will get either newer or older drivers and features, those must fit to the hardware you have, if they do not use your boot
option for older or newer kernels in the advanced menu of Grub.

The Debian Kernel is a requested feature in the software list, it is needed especiually for devices a couple of years old.
For very new devices the Xanmod or Zen Kernel is a better choice. Inclusion of such a very modern kernel in the available software list would be useful.

Personal advice, if your system runs ok and you want to experiment or exchange core components make a backup then do so in a live clone of your installed system.
Positive results, make exactly the same changes to the installed system.

Please do not expect perfect results on every possible hardware configuration
or with every BIOS Why I mention BIOS, mine is buggy, will never be fixed, = I have to live with that or scrap my hardware.

See fails as a learning experience, over time we all become more advanced in system administration.

[andfreeMember]

lgj100 wrote:

I have two machines with different hardware and identical software, and they behave in the same way

This is my case, too: a Toshiba Satellite A300 & an Asus VivoBook X540MA, both behaving in the same way.

lgj100 wrote:

(…) the boot problem went away on both systems when I removed the debian 6.12 kernel.

lgj100 wrote:

Before removing 6.12.12, the system wouldn’t boot into neither 6.1.42, nor 5.10 (…) Now it boots into both 5.10 and 6.1.42.

In my case, as I said, it was an antiX 6.6 kernel that caused boot problems to a 5.10 or a 6.1 antiX kernel. But there are also another differences, and it’s for this reason I don’t claim that they are similar cases. “My case” is actually two different main cases (the second of which is actually a combination of many sub-cases):
– an antiX 6.6 kernel prevents an antiX 5.10 kernel from booting, when the drive is divided into separate root and home partitions.
– an installed antiX 6.6 kernel prevents a frugal antiX kernel (5.10, 6.1 or even 6.6 itself) from booting (oversimplified description).
For none of these cases I saw something related to an eudev upgrade.

lgj100 wrote:

It would give some incomprehensible error and ask me to look at /var/log/Xorg.0.log.

In the first case, it gave me this:

Cannot open log file "/home/username/.local/share/xorg/Xorg.0.log"

But I’m afraid that this comparison is leading nowhere. My only conclusion is that issues can arise even with antiX kernels and

Xunzi_23 wrote:

(…) can only be solved on a case to case basis.

• This reply was modified 11 months ago by andfree.

antiX utilities for processing PDF files (posted by ChPol)

[Brian MasinickModerator]

Xunzi_23 wrote:

lgj100 wrote

The lesson for me is to be careful …
…Installing a on-antix kernel could mean problems…

The Linux Kernel is a central system component, drivers and features vary.. The replaced version is kept, rollback in case of problems.

Maybe you misunderstood, I had most problems with antiX kernel 5.10, that does not boot on my hardware, I get improved results with a very recent Debian Kernel …
Please do not expect perfect results on every possible hardware configuration
or with every BIOS Why I mention BIOS, mine is buggy, will never be fixed, = I have to live with that or scrap my hardware.

See fails as a learning experience, over time we all become more advanced in system administration.

I chopped a lot out of the original and kept a few key thoughts, namely:

A LOT of different kernel alternatives are available for very good reasons. In some cases, only older kernels that still have hardware support for certain older devices work. Why? The total size of the linux kernel with module support is enormous. At multiple occasions there have been older devices dropped but even more devices added.

ALSO, for me personally, if my hardware was supported, for the most part, since the very earliest kernels up until the present time, the current kernel worked really well. I can thing of two cases when I encountered issues:

1) A few builds in the 5.10 series did NOT work well for me, I think, but I’m not positive, that it was 5.10.142 that was erratic with my Dell Inspiron 5558. The good news is that subsequent updates worked fine.

2) I’m not sure if it was 6.1.42, but one of them in the 6.1 series, though it worked, was also a bit “questionable” in populating /dev; MOST of the time it worked, but there were occasionally “oddities”, perhaps not identical to what’s been reported here, but with some similarities.

What is common about my few, rare “issues” and the ones cited in this thread is that there is almost always one or more other alternatives that work beautifully – several examples: a newer or older (but still supported) antiX kernel, a Debian kernel, a Liquorix kernel, or a Xanmod kernel. ALL of these are solid options; each come with their advantages and disadvantages. The newest kernels load more memory cache to provide better response, but all that comes with memory overhead, so the newest kernels are most likely to work best with the newest systems. This is NOT universally true, and I can demonstrate it because I’ve tried some of the newest kernels with some of my oldest systems and they generally work.

Again, my Dell Inspiron 5558 is an excellent test model: it can function just fine with V4, 5 or 6 kernels, or Xanmod or Liquorix, Debian, or antiX kernels. The oldest ones use the least memory. While difficult to actually measure, the newest kernels, using more caching and other features AT TIMES seem to run just a tick “faster”, though they don’t necessarily LOAD faster – they also use more memory.

The point is that a lot of these options and alternatives are available in order to match and support the broadest possible range of systems. Just don’t expect all of them to excel in 100% of the use cases; each have particular things they do well at the expense of something else. The newest ones DROP hardware support for twenty year old systems; the older ones don’t have any support for systems with new hardware or peripherals.

--
Brian Masinick
Alternate "Search B":
This search page

[sybokMember]

Well, whatever the heck happened, I face the same situation on my antiX upgraded to Trixie.

Yesterday’s shutdown after several days of uptime (with interruptions when PC put to sleep, occasional updates and switching to different local accounts) was followed by “dead” login screen not accepting any prompts, cannot even switch to a TTY|console using Ctrl+Alt+Fn.
This used to be solved by clean boot, not anymore.
During the boot, the diodes on keyboard briefly light up which I believe is a good signal.

Exploring update history log reveals that ‘eudev’ was installed on 2025-10-23 (and I believe that I kept the PC on or asleep since then up to [most of] 2025-10-28 including):
eudev:amd64 (252:3.2.14-1.0antix3, 252:3.2.14-1.0antix5)
libeudev-dev:amd64 (252:3.2.14-1.0antix3, 252:3.2.14-1.0antix5)
libeudev1:amd64 (252:3.2.14-1.0antix3, 252:3.2.14-1.0antix5)

I run an antiX amd64 kernels, 6.6.101 and one from the 5.10 series (5.10.236 or 5.10.240, not sure, both records are present in ‘/boot/’), that both (always) worked fine (after clean boot).

Do you think that (creating a live USB or finding an old one,) chrooting, downgrading the libeudev-related packages and marking them temporarily on hold, might solve the problem?

BTW, there’s 1.1 GB space free on my ‘/home’ partition.

Redundant glib remark: Hmpf, when @anticapitalista asked why so little people reported the issue, perhaps, they are locked out of their (sole) devices.

PS: Writing from Devuan (multi-boot on the same device).

[abc-nixMember]

sybok wrote:

Do you think that (creating a live USB or finding an old one,) chrooting, downgrading the libeudev-related packages and marking them temporarily on hold, might solve the problem?

This has worked for others before, as reported in this thread, so go for it. You should be able to do this also from Devuan (though it is more manual than using chroot-reescue-scan).

If you are on trixie, then you can download the (what I think was the) last working eudev from this link (packages should be eudev and libeudev1, version 3.2.9.1.0antix7).

Corruption has two hands: the Right and the Left. Don't let them squeeze you.

[anticapitalistaForum Admin]

If using sysvinit make sure there is a udev, seatd (and dbus and slimski if you use it) file in /etc/init.d.
If using runit make sure there is a udev, seatd (and dbus and slimski if you use it) run file linked to/etc/services/

Philosophers have interpreted the world in many ways; the point is to change it.

antiX with runit - leaner and meaner.

[sybokMember]

@abc-nix:

This has worked for others before, as reported in this thread

Damn, I admit of making the mistake and not reading it thoroughly as using this PC as a work-station and I (felt I) was in a hurry to get it fixed while switching among several tasks (almost never a good idea in my case).

EDIT: Oh my, using the suggested download link to grab and install the three *eudev* packages that were updated (and wreaked havoc on my antiX isntall) leads to issues with ‘libgudev’

# apt install –fix-broken
Correcting dependencies… failed.
Unsatisfied dependencies:
libgudev-1.0-0 : Depends: libeudev1 (>= 252:3.2.12) but 246:3.2.9.1.0antix7 is installed
Error: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages.
Error: Unable to correct dependencies

The old versions seem to be gone according to
apt list --all-versions eudev
after the manual install via ‘dpkg’ (did I just mess ‘apt’ by installing using ‘dpkg’ making it seem to be a non-repository package?).

@anticapitalista: Yes, using SysVinit. Thank you for the tip.

EDIT: The files are there, perhaps already were

• This reply was modified 11 months ago by sybok.

[anticapitalistaForum Admin]

What brought in libgudev?

Philosophers have interpreted the world in many ways; the point is to change it.

antiX with runit - leaner and meaner.

[Author]

Posts