Hi,
This is my first Debian bug report. I would appreciate knowing to
which package the report should reference.
The following is a list and
the results of the steps I took.
1. The system upgraded the Debian11
5.10.0-36-amd64 by creating 5.10.0-37-amd64.
2. After rebooting, the
login screen appeared on the monitor connected to the vga port (which is common).
3. When I tried to log into my user account, the screen went
black for almost a second, a screen flashed what appeared to be the screen described below, and it went back to the log in window.
4. I logged in as
root (Old School) successfully.
5. My user account's home directory was
on a NAS. I changed passwd file to use the local disk as the home
directory.
6. I then was successful in logging in as a user.
7. In
preparation of installing the NVIDIA drivers, I located the X process.
8.
After killing Xwayland, the screen was filled with a very small character
which looked like a small copyright symbols.
9. Whenever I typed
anything, it continued to appear as the copyright symbol.
10. I logged in
and out. Some of the characters turned green and the rest were either dim
or bright.
When checking the log files I found the following:
Warning:
Unsupported maximum keycode 569, clipping. This was reported in 2011 and
2012. The other related messaged were also present.
Trying to read,
write, or create a file on the NAS was DENIED. The user number of the file matched the User's number.
Thanks,
Robert Sundelin


--- PyGate Linux v1.5.2
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Wed, Jan 14, 2026 at 17:44:28 -0600, sundelin@mc.net wrote:

1. The system upgraded the Debian11

After killing Xwayland, the screen was filled with a very small character which looked like a small copyright symbols.

https://lists.debian.org/debian-kernel/2025/12/msg00210.html

https://lists.debian.org/debian-user/2025/12/msg00381.html

--- PyGate Linux v1.5.2
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Wed 14 Jan 2026 at 17:44:28 (-0600), sundelin@mc.net wrote:

1. The system upgraded the Debian11
5.10.0-36-amd64 by creating 5.10.0-37-amd64.

My experience after doing that is recounted in:
https://lists.debian.org/debian-user/2025/12/msg00188.html
As you see there, booting the backup kernel, -36-, makes the
problem go away while you sort out the NAS problem, if that's
unrelated to the upgrade.

2. After rebooting, the
login screen appeared on the monitor connected to the vga port (which is common).

If this a VC login prompt, or some Display Manager?

3. When I tried to log into my user account, the screen went
black for almost a second, a screen flashed what appeared to be the screen described below, and it went back to the log in window.

4. I logged in as
root (Old School) successfully.

I'm very Old School. Logging in as root on a VC is fine, but I
wouldn't do that in a DM.

5. My user account's home directory was
on a NAS. I changed passwd file to use the local disk as the home
directory.

6. I then was successful in logging in as a user.

Again, in a VC or with a DM?

7. In
preparation of installing the NVIDIA drivers, I located the X process.

8.
After killing Xwayland, the screen was filled with a very small character which looked like a small copyright symbols.

So it seems likely that you were logged in graphically, and killed X
to get back to a VC. The question is which VC.

If you boot up and login at a VC, my experience is that VC1 looks
normal, but none of the others does. However, in their weird state,
they still accept and execute what you type as normal.

So I would try circulating around the VCs with whichever keys do that, Alt-arrows or ?-arrows typically, and see whether you find a VC that
looks normal.

9. Whenever I typed
anything, it continued to appear as the copyright symbol.

10. I logged in
and out. Some of the characters turned green and the rest were either dim
or bright.

When checking the log files I found the following:

Warning:
Unsupported maximum keycode 569, clipping. This was reported in 2011 and 2012. The other related messaged were also present.

I assume this is something to do with X and wayland. (I've never
used the latter.)

Trying to read,
write, or create a file on the NAS was DENIED. The user number of the file matched the User's number.

Again, does this relate to the upgrade, or is it still a problem
when running the older kernel version?

This is my first Debian bug report. I would appreciate knowing to
which package the report should reference.

https://lists.debian.org/debian-kernel/2025/12/msg00290.html

is a better reference than that given, as there's been some work
on finding exactly when this bug arose. (The thread starts at
msg00209.html .) I don't know whether there will be a fix soon,
but because of that discussion, I haven't filed a bug myself.

I only have one machine still running bullseye, but all five
heterogeneous machines showed the same problem with -37-. None
has the problem when running bookworm or trixie, nor have I seen
reports of the bug. Is upgrading the distribution an option for you?

Cheers,
David.

--- PyGate Linux v1.5.6
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Tue, 27 Jan 2026 at 05:13, David Wright <deblis@lionunicorn.co.uk> wrote:

On Wed 14 Jan 2026 at 17:44:28 (-0600), sundelin@mc.net wrote:

After killing Xwayland, the screen was filled with a very small character which looked like a small copyright symbols.

[...]

Again, does this relate to the upgrade, or is it still a problem
when running the older kernel version?

This is my first Debian bug report. I would appreciate knowing to
which package the report should reference.

https://lists.debian.org/debian-kernel/2025/12/msg00290.html

is a better reference than that given, as there's been some work
on finding exactly when this bug arose. (The thread starts at
msg00209.html .) I don't know whether there will be a fix soon,
but because of that discussion, I haven't filed a bug myself.

I only have one machine still running bullseye, but all five
heterogeneous machines showed the same problem with -37-. None
has the problem when running bookworm or trixie, nor have I seen
reports of the bug. Is upgrading the distribution an option for you?

Hi, I've not read carefully your conversation, but I recall seeing this
which might be the relevant Debian bug for what you are discussing:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1123750
It looks like the cause and fix has been identified, but being Debian 11 oldoldstable I'm unsure about expectations for a fixed package being
released.

--- PyGate Linux v1.5.6
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

Thank you for your informative reply. The problems were
reported on my new computer, which is less than a year
old. I bought the new computer for redundancy to this
computer which is getting a little old. Comparison of the
two is: Both have two CPU. The new one has 16 cores per
CPU, which act like 32. This one has four cores per CPU.
The new one has solid state drives. This one has hard
disk drives. The uname -a of this computer gives: Linux
gale 5.10.0-36-amd64 #1 SMP Debian 5.10.244-1 (2025-09-29)
x86_64 GNU/Linux. For comparison, uname -a on the new
computer gives: Linux keel 5.10.0-37-amd64 #1 SMP Debian
5.10.247-1 (2025-12-11) x86_64 GNU/Linux

I have been fighting annoyances cubed. I had written most
of my replies to your email when Firefox went back to email
login and all changes were lost except for an early save I
had done. When I tried to log in today, I found this
computer was running -37- Xwayland instead of -36- Xorg.
Next I found that something had caused my NAS to shutdown.

-----------------------------------------------
On Mon, 26 Jan 2026 23:13:07 -0600, David Wright
<deblis@lionunicorn.co.uk>
wrote:

On Wed 14 Jan 2026 at 17:44:28 (-0600), sundelin@mc.net wrote:

1. The system upgraded the Debian11
5.10.0-36-amd64 by creating 5.10.0-37-amd64.

My experience after doing that is recounted in:
https://lists.debian.org/debian-user/2025/12/msg00188.html
As you see there, booting the backup kernel, -36-, makes the
problem go away while you sort out the NAS problem, if that's
unrelated to the upgrade.

I use a NAS because I have six computers which collectively run
eleven operating systems. On this computer, using a NAS has
worked properly since I loaded Debian 7. On the new computer
-36- works fine. So does Sun Solaris and Apple Darwin. I did
check and saw in -37- that the files from the NAS were mounted.
If a computer has trouble mounting the file systems from a NAS,
the default is to mount them Read Only. But in this case the
error messages included that I could not read my files.

2. After rebooting, the
login screen appeared on the monitor connected to the vga port (which

is

common).

If this a VC login prompt, or some Display Manager?

I had to track down what VC meant since I only use simple terminals.
The system went back to a Display Manager with a login window.

3. When I tried to log into my user account, the screen went
black for almost a second, a screen flashed what appeared to be the
screen
described below, and it went back to the log in window.

4. I logged in as
root (Old School) successfully.

I'm very Old School. Logging in as root on a VC is fine, but I
wouldn't do that in a DM.

5. My user account's home directory was
on a NAS. I changed passwd file to use the local disk as the home
directory.

6. I then was successful in logging in as a user.

Again, in a VC or with a DM?

As with the root login, everything used a Display Manager.

7. In
preparation of installing the NVIDIA drivers, I located the X process.

8.
After killing Xwayland, the screen was filled with a very small

character

which looked like a small copyright symbols.

So it seems likely that you were logged in graphically, and killed X
to get back to a VC. The question is which VC.

If you boot up and login at a VC, my experience is that VC1 looks
normal, but none of the others does. However, in their weird state,
they still accept and execute what you type as normal.
o
So I would try circulating around the VCs with whichever keys do that, Alt-arrows or ?-arrows typically, and see whether you find a VC that
looks normal.

Things acted like a regular terminal such as a VT100. But I was not
able to switch to anything different. Sometimes when using your
combination keys, it would go back to the login screen with Xwayland
running.

9. Whenever I typed
anything, it continued to appear as the copyright symbol.

10. I logged in
and out. Some of the characters turned green and the rest were eithero

dim

or bright.

When checking the log files I found the following:

Warning:
Unsupported maximum keycode 569, clipping. This was reported in 2011

and

2012. The other related messaged were also present.

I assume this is something to do with X and wayland. (I've never
used the latter.)

It sounds reasonable that Xwayland might be causing the problems.
When I tried to blacklist Xwayland, the system would not boot.

o

Trying to read,
write, or create a file on the NAS was DENIED. The user number of the
file
matched the User's number.

Again, does this relate to the upgrade, or is it still a problem
when running the older kernel version?

As described above, the NAS problem started with the upgrade.

This is my first Debian bug report. I would appreciate knowing to
which package the report should reference.

https://lists.debian.org/debian-kernel/2025/12/msg00290.html

is a better reference than that given, as there's been some work
on finding exactly when this bug arose. (The thread starts at
msg00209.html .) I don't know whether there will be a fix soon,
but because of that discussion, I haven't filed a bug myself.

I only have one machine still running bullseye, but all five
heterogeneous machines showed the same problem with -37-. None
has the problem when running bookworm or trixie, nor have I seen
reports of the bug. Is upgrading the distribution an option for you?

Since my new computer has four solid state disks, my plan is to
put Ubuntu, buster, bullseye, and bookworm on them. If bookworm
worked properly, I would first put in on this one (older) in place
of bullseye. Once trixie has matured a little, I would then replace
bullseye on the new computer with trixie. Buster is being kept
around until the sound works properly in bookworm or trixie. The
sound does not work properly in bullseye. For a long time bullseye
would not find my Epson scanner like buster does. Now bullseye
does find it.

Cheers,
David.

-----------------------------------------------

--- PyGate Linux v1.5.8
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Fri 30 Jan 2026 at 18:50:20 (-0600), sundelin@mc.net wrote:
[ ? ]
I'll just make the odd comment.

I have been fighting annoyances cubed. I had written most
of my replies to your email when Firefox went back to email
login and all changes were lost except for an early save I
had done.

I always prepare text in emacs, so that very little is lost even
should the machine crash when a document hasn't been explicitly
saved for a while.

When I tried to log in today, I found this
computer was running -37- Xwayland instead of -36- Xorg.

If you're prepared to run buster, then I can't see anything
wrong with sticking with -36-. (My machine has had no problem
with -37- apart from the ?/? characters all over VCs 2-6.)

I use a NAS because I have six computers which collectively run
eleven operating systems. On this computer, using a NAS has
worked properly since I loaded Debian 7. On the new computer
-36- works fine. So does Sun Solaris and Apple Darwin. I did
check and saw in -37- that the files from the NAS were mounted.
If a computer has trouble mounting the file systems from a NAS,
the default is to mount them Read Only. But in this case the
error messages included that I could not read my files.

You would have to be more specific than that. Maybe missing modules?

I had to track down what VC meant since I only use simple terminals.
The system went back to a Display Manager with a login window.

As I run X with a fvwm, I don't get into that situation at login.
But from what's below, it sounds like the video driver stuff.

Things acted like a regular terminal such as a VT100. But I was not
able to switch to anything different. Sometimes when using your
combination keys, it would go back to the login screen with Xwayland
running.

Ah. Well that might be running on the one VC that's working properly,
so with a DM, you would never see that one.

Here, with -37-, I normally run X on VC1, the one that works. VCs 2-6
show nothing but ?/?. To get both a VC and X, I just start X on VC2,
by blindly typing my loginname, password, and xxx (a script that runs
startx) into the ?/? junk, which leaves VC1 available for use.

Warning:
Unsupported maximum keycode 569, clipping. This was reported in 2011

and

2012. The other related messaged were also present.

I looked that up, and it seems its cause is understood (though not by me).

It sounds reasonable that Xwayland might be causing the problems.
When I tried to blacklist Xwayland, the system would not boot.

I don't know what blacklisting Xwayland means. I thought you had been
able to run Xorg in -36-. Nothing else changed with the kernel upgrade
here:

Start-Date: 2025-12-13 21:49:11
Commandline: apt-get dist-upgrade
Install: linux-image-5.10.0-37-amd64:amd64 (5.10.247-1, automatic)
Upgrade: linux-image-amd64:amd64 (5.10.244-1, 5.10.247-1)
End-Date: 2025-12-13 21:50:01

Start-Date: 2025-12-13 21:50:59
Commandline: apt-get --purge autoremove
Purge: linux-image-5.10.0-35-amd64:amd64 (5.10.237-1)
End-Date: 2025-12-13 21:51:27

Trying to read,
write, or create a file on the NAS was DENIED. The user number of the
file
matched the User's number.

Again, does this relate to the upgrade, or is it still a problem
when running the older kernel version?

As described above, the NAS problem started with the upgrade.

The actual error message would be more helpful. For example, there
are simple file permissions; then there are more subtle things like
apparmor restrictions; and recently someone got caught out by the
addition of ProtectHome in a systemd file.

Since my new computer has four solid state disks, my plan is to
put Ubuntu, buster, bullseye, and bookworm on them. If bookworm
worked properly, I would first put in on this one (older) in place
of bullseye. Once trixie has matured a little, I would then replace
bullseye on the new computer with trixie. Buster is being kept
around until the sound works properly in bookworm or trixie. The
sound does not work properly in bullseye. For a long time bullseye
would not find my Epson scanner like buster does. Now bullseye
does find it.

Yes, I think sound on DEs has changed quite a lot with the
introduction of things like pulseaudio and pipewire. I've always just
carried on using alsamixer, controlled by FVWM's keybindings, though
I have sometimes seen pipewire running (no idea what started it).

And I got used to scanning onto a USB stick 20 years ago at work,
and have carried on doing that. It's never felt inconvenient as
I have to stand at the scanner anyway, handling the original.
Good luck, though.

Cheers,
David.

--- PyGate Linux v1.5.10
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)