Tty1 not usable

tty1 has recently broken for me. KDE is running on tty2 and console is on tty3 but this has changed recently (I often switch tty and would have noticed although I did not use this device for a few months). If I try to switch to tty1 the GUI is still displayed but appears to hang, the only thing I can do from tty1 is switch to another tty.

There seems to be a few posts (Login/desktop moved from tty1 to tty2 after update - #3 by Aragorn) but nothing recently.

@Aragorn do you remember the bug you mention in the above thread and know if it was resolved or not and whether it has reoccurred recently?

Usually the display manager runs on tty1 and depending on the system the desktop runs on either tty2 or tty7.

On my workstation - using plasma-login-manager - tty is occupied with an empty screen and a blinking cursor.

3 Likes

That might have been what changed it. I’m using it now too.
I can’t remember ever having a tty that wasn’t functional per se (i.e. having an empty screen and blinking cursor). Do you think this would be a bug?

For me tty1 ends when Plymouth gets its job done.

Odd. You’d think there’s be more symmetry here.

No, THIS is a bug…
:beetle:

Actually, I have no clue why - but it has always been like this, I had no TTY1 for years already.

2 Likes

No. That’s a beetle. Not a bug (heteroptera). Or well, I guess heteroptera are TRUE bugs, but anyway. :stuck_out_tongue_winking_eye:

But yeah. I don’t remember ever seeing anything interactable on tty1. It used to be just black with an underscore-cursor. Now for a long while it’s shown the tail end of the boot-up list.

1 Like

It’s not a bug, but a design choice in systemd. tty1 is normally reserved as the console for the system boot-up messages, whether you are booting with those messages hidden — via the quiet boot parameter and/or via plymouth — or visible by default. Your Plasma desktop then runs on either tty1 or tty2 — the choice appears to be rather random.

Either way, the first free tty in a Manjaro Plasma system is tty3. Switching back to the GUI could thus be a matter of switching back to either tty1 or tty2.

If after switching to tty1 you see the desktop but it’s not responding, then this may simply be an inert ghost image in the console framebuffer, and then you need to try tty2.

I know it’s somewhat annoying, but it’s systemd doing this. :man_shrugging:


Yes, but is it John, Paul, George or Ringo? :zany_face:

5 Likes

Haha for years I’ve waited for someone else who shares my hatred for lame US English… and now it comes back to bite me in the a55 when I can’t find an emoji for an actual bug…!

Quite rightly, though - bugs are generally biters and suckers, and we shouldn’t tar beetles with the same brush.

Well done!

3 Likes

Haha, but I nearly always had the text displayed, I didn’t really make the connection… so the system boots in TTY1, then you move over to TTY2 for the desktop… makes sense.

2 Likes

I think it resembles mainly this. :face_savoring_food:

1 Like

I’m currently running Plasma on TTY1, which is confirmed by the w command:

w 
 08:29:30 up 2 days, 21:33,  1 user,  load average: 0.45, 0.51, 0.57
USER     TTY       LOGIN@   IDLE   JCPU   PCPU  WHAT
scotty   tty1      Tue10    2days  0.05s  0.05s /usr/bin/startplasma-wayland

After running the above command, I then used Ctrl+Alt+F2 to switch to TTY2, and logged into a text session:

w
 08:32:58 up 2 days, 21:36,  2 users,  load average: 0.35, 0.41, 0.52
USER     TTY       LOGIN@   IDLE   JCPU   PCPU  WHAT
scotty   tty1      Tue10    2days  0.05s  0.05s /usr/bin/startplasma-wayland
scotty   tty2      08:30    2:42   0.00s   ?    -bash

After logging out of the TTY2 session with the logout command, I’m back to just TTY1:

w 
 08:33:47 up 2 days, 21:37,  1 user,  load average: 0.30, 0.39, 0.51
USER     TTY       LOGIN@   IDLE   JCPU   PCPU  WHAT
scotty   tty1      Tue10    2days  0.05s  0.05s /usr/bin/startplasma-wayland
1 Like

Well, just to illustrate the randomness, here’s mine… :backhand_index_pointing_down:

[nx-74205:/dev/pts/5][/home/aragorn]
[aragorn] >  w
 22:43:08 up 19:43,  1 user,  load average: 0.16, 0.06, 0.08
USER     TTY       LOGIN@   IDLE   JCPU   PCPU  WHAT
aragorn  tty2      03:00   19:43m  0.03s  0.03s /usr/bin/startplasma-wayland

[nx-74205:/dev/pts/5][/home/aragorn]
[aragorn] >  who
aragorn  pts/0        2026-03-05 03:00 (:0)
aragorn  pts/1        2026-03-05 03:00 (:0)
aragorn  pts/2        2026-03-05 03:00 (:0)
aragorn  pts/3        2026-03-05 14:24 (:0)
aragorn  pts/4        2026-03-05 14:25 (:0)
aragorn  pts/5        2026-03-05 22:43 (:0)

[nx-74205:/dev/pts/5][/home/aragorn]
[aragorn] >

As you can see, Plasma logged me in at tty2, not at tty1, and I have not opened any ttys since boot-up.

This is with plasma-login-manager by the way, and running fully on Wayland.

2 Likes

@Aragorn is right. But the randomness, I think, is just the order in which they get claimed on boot. But we can easily increase this, if you want. Just edit:

/etc/systemd/logind.conf

The defaults:

#NAutoVTs=6

You can go up to:

NAutoVTs=12

So you can go up to Ctrl+Alt+F12, and it will spawn gettys if you switch to them.

But there is also this default:

ReserveVT=6

So it is kind of the old tty1; one pre-spawned and always ready on tty6. So I have just grown used to it.

It seems possible to configure login managers and your compositor/X-server to take different ones.. And always have a ready tty1 on Ctrl+Alt+F1. But I’m not sure if I want to bother with that rabbit hole. :grimacing:

Let’s go with blaming systemd.

2 Likes

I would recommend against adding tty12, because that’s where journald sends the system log to in real-time — or at least, if you have that enabled. :wink:

# /etc/systemd/jounald.conf

# [...]
ForwardToConsole=yes
TTYPath=/dev/tty12
# [...]
3 Likes

Thanks everyone, this has been a good discussion and I am now educated on bugs of the emoji kind. :joy:

5 Likes

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.