Repeating "can't open /dev/ttyX: No such device or address" messages during installation

Bug #1704929 reported by Frank Heimes
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Ubuntu on IBM z Systems
Fix Released
Medium
Unassigned
console-setup (Ubuntu)
Fix Released
Undecided
Unassigned

Bug Description

[Impact]

console-setup continuously tries to open /dev/tty[1-6] on s390x, when such consoles do not exist on s390x.

This can be seen on boot in the output from the initramfs, and during the installer.

[Cause]

It seems to me that the postinst of the console-setup is incorrect for s390, since on s390 Linux tty[1-6] do not exist in any modes (LPAR, z/VM, KVM)

[Solution]
I do not know what ACTIVE_CONSOLES should be set as, my guess is to set it to... "guess". Usually it should be slcp, but that depends on which consoles are configured/activated in the given KVM.

[Testcase]
On boot, scroll all the messages and makesure there are no error messages about inability to open /dev/tty*

[Original Bug Report]

During the installation (z/VM guest and KVM virtual machine) of Ubuntu Server 16.04.1 (and 16.04.2) repeating messages of the form:

"
Select and install software ... 10% can't open /dev/tty4: No such device or address
can't open /dev/tty2: No such device or address
can't open /dev/tty3: No such device or address
can't open /dev/tty4: No such device or address
can't open /dev/tty2: No such device or address
can't open /dev/tty3: No such device or address
can't open /dev/tty2: No such device or address
can't open /dev/tty4: No such device or address
can't open /dev/tty4: No such device or address
can't open /dev/tty3: No such device or address
can't open /dev/tty2: No such device or address
can't open /dev/tty2: No such device or address
can't open /dev/tty3: No such device or address
... 20% can't open /dev/tty4: No such device or address
can't open /dev/tty2: No such device or address
can't open /dev/tty3: No such device or address
... 30%... 40% can't open /dev/tty3: No such device or address
can't open /dev/tty4: No such device or address
can't open /dev/tty2: No such device or address
... 50% can't open /dev/tty3: No such device or address
can't open /dev/tty4: No such device or address
can't open /dev/tty2: No such device or address
... 60% can't open /dev/tty2: No such device or address
can't open /dev/tty3: No such device or address
can't open /dev/tty4: No such device or address
can't open /dev/tty2: No such device or address
...
Finishing the installation ... 13%... 22%... 31% can't open /dev/tty3: No such device or address
can't open /dev/tty2: No such device or address
can't open /dev/tty4: No such device or address
... 40%... 50%... 63%... 72%... 81% can't open /dev/tty4: No such device or address
can't open /dev/tty2: No such device or address
can't open /dev/tty3: No such device or address
... 90% The system is going down NOW
 Sent SIGTERM to all processes
 Sent SIGKILL to all processes
 Requesting system reboot
01: HCPGSP2629I The virtual machine is placed in CP mode due to a SIGP stop CPU 00.
02: HCPGSP2629I The virtual machine is placed in CP mode due to a SIGP stop CPU 00.
03: HCPGSP2629I The virtual machine is placed in CP mode due to a SIGP stop CPU 00.
00 Storage cleared - system reset.
00 zIPL ......
"

They start to occur when the software gets installed:

"Select and install software ... 10% can't open /dev/tty4: No such device or address"

And only stop with the final system reboot at the end of the installation.

description: updated
Revision history for this message
Andrew McLeod (admcleod) wrote :

I'm seeing this installing 17.10 on s390x also.

Frank Heimes (fheimes)
Changed in ubuntu-z-systems:
status: New → Confirmed
Changed in console-setup (Ubuntu Xenial):
status: New → Confirmed
Changed in console-setup (Ubuntu Artful):
status: New → Confirmed
no longer affects: console-setup (Ubuntu Xenial)
no longer affects: console-setup (Ubuntu Zesty)
no longer affects: console-setup (Ubuntu Artful)
Changed in console-setup (Ubuntu):
status: Confirmed → Incomplete
Changed in linux (Ubuntu):
status: New → Confirmed
Revision history for this message
Dimitri John Ledkov (xnox) wrote :

/dev/tty1 file exists on s390x ubuntu, however, one cannot open that file doing that results in ENXIO.

console-setup is an architecture independent package which tries to detect and configure all consoles. Ideally, I do not want to encode if `uname -p` is s390x or some such to filter-out and skip tty[0-9].

Imho, if /dev/tty[0-9] devices do not exist, why does kernel create such files? Can it be escalated to IBM linux team to not have those files at all? E.g. make a upstream kernel patch to not create these devices. Maybe I am crazy, I don't understand things - e.g. is it possible to somehow get a working /dev/tty1?

Revision history for this message
Frank Heimes (fheimes) wrote :

Reverse proxy initiated ...

Changed in ubuntu-z-systems:
assignee: nobody → bugproxy (bugproxy)
Revision history for this message
bugproxy (bugproxy) wrote : Comment bridged from LTC Bugzilla

------- Comment From <email address hidden> 2017-09-27 12:20 EDT-------
Will be covered by original Launchpad bug

tags: added: architecture-s39064 bugnameltc-159373 severity-high targetmilestone-inin1710
Frank Heimes (fheimes)
Changed in ubuntu-z-systems:
importance: Undecided → Medium
tags: added: id-597a82ade5d5089d05d20931
Mathew Hodson (mhodson)
description: updated
Changed in linux (Ubuntu):
importance: Undecided → High
Revision history for this message
bugproxy (bugproxy) wrote :

------- Comment From <email address hidden> 2018-01-25 10:12 EDT-------
(In reply to comment #4)
> During the installation (z/VM guest and KVM virtual machine) of Ubuntu
> Server 16.04.1 (and 16.04.2) repeating messages of the form:
>
> "
> Select and install software ... 10% can't open /dev/tty4: No such device or
> address
> can't open /dev/tty2: No such device or address
> can't open /dev/tty3: No such device or address
> can't open /dev/tty4: No such device or address

...

> /dev/tty1 file exists on s390x ubuntu, however, one cannot open that file
> doing that results in ENXIO.

...

> Imho, if /dev/tty[0-9] devices do not exist, why does kernel create such

The /dev/tty[0-9] devices represent the x86 VT consoles. They do not exist on s390x nor should the tty terminal device driver for them be active.

So I suspect that these are not created by the kernel and, further, I guess that they are manually created by those utilities that create the install image. That means, these are static created device files rather then dynamically ones handled by udev.

> files? Can it be escalated to IBM linux team to not have those files at all?
> E.g. make a upstream kernel patch to not create these devices. Maybe I am
> crazy, I don't understand things - e.g. is it possible to somehow get a
> working /dev/tty1?

No there isn't. The only special case is /dev/tty (without a number). This refers to the current terminal of a process (it is not a VT console).

Frank Heimes (fheimes)
tags: added: reverse-proxy-bugzilla
Revision history for this message
Frank Heimes (fheimes) wrote :

I just did a pure bionic LPAR install using exclusively the 'Operating System Messages' tasks at the HMC and interestingly I did not came across a single ttyX message there.

tags: added: id-5a7c414e913451969b0ad7c9
Changed in ubuntu-z-systems:
assignee: bugproxy (bugproxy) → nobody
no longer affects: linux (Ubuntu)
Revision history for this message
Frank Heimes (fheimes) wrote :

Since this only seems to happen with xenial, and not with any other Ubuntu release (Y to E) and since it only happens during install time, and even then only with using the console for the entire installation (and not using the 2nd stage via ssh), I closing this as Won't Fix.

Changed in ubuntu-z-systems:
status: Confirmed → Won't Fix
Revision history for this message
Launchpad Janitor (janitor) wrote :

[Expired for console-setup (Ubuntu) because there has been no activity for 60 days.]

Changed in console-setup (Ubuntu):
status: Incomplete → Expired
Revision history for this message
Frank Heimes (fheimes) wrote :

Since disco these messages are no longer visible.
Hence closing this ticket.

Changed in console-setup (Ubuntu):
status: Expired → Fix Released
Changed in ubuntu-z-systems:
status: Won't Fix → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.