[Lucid] maintenance shell and tty1 compete for/share keyboard input

[Philip Muškovac]

Binary package hint: upstart

I don't know when that happened or what exactly causes this bug, but when I boot up and need to use a tty I noticed that tty1 was unusable due to the maintenance shell and the regular tty1 shell both being active:

init: plymouth-log main process (1103) terminated with status 111
mountall:mountall.c:1296: Not reached assertion failed in mounted
General error mounting filesystems.
A maintenance shell will now be started.
CONTROL-D will terminate this shell and reboot the system.
Give root password for maintenance
(or type CONTROL-D to continue): * Starting init crypto disks... [OK]

Ubuntu lucid (development branch) yofel-eee tty1

yofel-eee login:
Login incorrect.
Give root password for maintenance
(or type CONTROL-D to continue):

if you constantly try to input the root password you'll actually manage to login to the maintenance shell, but as the tty1 shell is still active you'll get login failure messages printed all the time rendering the maintenance shell useless. You have to use tty2-6 if you want to do anything.

ProblemType: Bug
Architecture: i386
Date: Sun Jan 10 19:39:18 2010
DistroRelease: Ubuntu 10.04
Package: upstart 0.6.3-11
ProcEnviron:
 LANGUAGE=en_US.UTF-8
 PATH=(custom, user)
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.32-10.14-generic
SourcePackage: upstart
Tags: lucid
Uname: Linux 2.6.32-10-generic i686

[Jonas Bygdén (jbygden)]

I have something similar in karmic.

I have quite a lot of external USB-drives connected to my headless karmic server and every boot there's a couple of drives that hasn't finished spinning up when mountall tries to mount their filesystems. This is no problem per se for me, the system is ready for login quicker than if I had to wait for them.

I get the message:

"One or more of the mounts listed in /etc/fstab cannot yet be mounted:
(ESC for recovery shell)"

Which is quite ok.

What's not OK though is that after I've waited for the disk to spin up, mountall to either fsck and then mount or just mount it tty1 is unusable until I've gone through the entire process of entering and exiting the recovery shell once per filesystem that had to be waited on.

The one thing that annoys me the most is when I have to log in to do a quick edit of a file, and forget about the ESC + Ctrl-D that I should have done before logging in. I log in, vi [file], edit something, press ESC to exit edit mode in vi and get hit with:
"Give root password for maintenance
(or type CONTROL-D to continue):"

Isn't there a way to remove the trap for ESC once all mounts have finished successfully?

[Nikolaus Rath (nikratio)]

Setting to Confirmed according to comment #2.

[Scott James Remnant (Canonical) (canonical-scott)]

The assertion error that caused this has been fixed.

It's not supposed to be possible to ever get both a maintenance shell and the getty (since that depends on mountall exiting) - you hit a rare bug that's been fixed