VirtualBox: black screen (with movable pointer) after login

Need support for GhostBSD. Ask here if your question does not fit elsewhere.
Post Reply
grahamperrin
Posts: 81
Joined: Sun Sep 04, 2016 6:06 am

VirtualBox: black screen (with movable pointer) after login

Post by grahamperrin »

Host: VirtualBox 6.1.22, FreeBSD 14.0-CURRENT.

Yesterday with what I assume was 13.0-STABLE 1300500, I signed in without difficulty and began the Update Station routine. During installation, I closed the virtual machine and saved its state.

This morning when I started the virtual machine I noticed a start of the operating system, which half-surprised me because I had no recollection of completing the installation. (Maybe I did complete it, in a previous VirtualBox session, and forgot.) An attempt to login (as grahamperrin) produced nothing more than a movable pointer on a black screen. I noted 13.0-STABLE 1300504 but did not tell the version of guest additions.

I activated then booted the one and only backup boot environment,

backup-2021-06-12-06-33

I had the same problem with an attempt to login as grahamperrin. First guess: a transient issue involving something in my home directory, and/or the mismatch between 6.1.22 on the host and 6.1.20 in the guest. Second guess: the symptom (black screen + movable pointer) is comparable to some transient problems recently observed with FreeBSD 13.0-RELEASE-p2 in VirtualBox – https://forums.freebsd.org/threads/80860/ – although honestly, this guess is stretching things (the topic there is inconclusive, I should not jump to conclusions here or there).

Without restarting the system, from ttyv1 I ran startx as root and got a working desktop environment. I created, activated then booted a new environment:

1300504

If I recall correctly: at this point (with the environment at 1300500, not yet updated to 1300504), boot in safe mode failed – there were two or more unexpected restarts of the OS (kernel panics, presumably). If I want to reproduce and diagnose this, later, I can use the outdated non-default Default boot environment. For now, let's assume that kernel panics were symptomatic of FreeBSD bug 254412 – emulators/virtualbox-ose-additions - Boot time crash - Sleeping thread owns a non-sleepable lock. Afterthought: the first boot, alone, would have been safe mode; subsequent (unexpected) boots would have been normal mode.

I interrupted the boot loop, started in normal mode, login as grahamperrin succeeded ☑

After waiting a few minutes with no automated hint of available updates, I manually opened Software Station, fetches began.

I clicked the update alert, not realising that a second instance of Software Station would open. I closed the second instance, leaving the first instance processing candidates.

Installation of updates is proceeding, I'll follow up after completion.
grahamperrin
Posts: 81
Joined: Sun Sep 04, 2016 6:06 am

pkg exited on signal 6 (core dumped)

Post by grahamperrin »

Software Station did not complete.

Image

Tail of /root/update.failed

[197/794] Reinstalling gsettings-desktop-schemas-3.38.0...
[197/794] Extracting gsettings-desktop-schemas-3.38.0: .......... done
[198/794] Reinstalling gnupg-2.3.1...

/root/update.failed in its entirety: https://pastebin.com/vEuBpQq3

The machine has 2,048 MB memory at the time of writing, I assume that this is enough for Software Station and MATE Terminal (with the terminal doing nothing more than tailing /var/log/messages …). An earlier screenshot:

Image

– plenty of swap space.

Code: Select all

root@mowa219-gjp4-ghostbsd-13-vm:/usr/home/grahamperrin # gdb file pkg.core
GNU gdb (GDB) 10.2 [GDB v10.2 for FreeBSD]
Copyright (C) 2021 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Type "show copying" and "show warranty" for details.
This GDB was configured as "x86_64-portbld-freebsd13.0".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<https://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
    <http://www.gnu.org/software/gdb/documentation/>.

For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from file...
(No debugging symbols found in file)

warning: core file may not match specified executable file.
[New LWP 100768]
Core was generated by `pkg upgrade -nf'.
Program terminated with signal SIGABRT, Aborted.
#0  0x0000000800c521aa in ?? ()
(gdb) bt
#0  0x0000000800c521aa in ?? ()
#1  0x0000000800bc8354 in ?? ()
#2  0x00000000000189a0 in ?? ()
#3  0x44d4103eb3d26515 in ?? ()
#4  0x00007fffffffdf34 in ?? ()
#5  0x000000080131fe00 in ?? ()
#6  0x00007fffffffdf50 in ?? ()
#7  0x0000000800c7a499 in ?? ()
#8  0x44d4103eb3d26515 in ?? ()
#9  0x000000000000007c in ?? ()
#10 0xffffffdf0000007c in ?? ()
#11 0xffffffffffffffff in ?? ()
#12 0x00000008ffffffff in ?? ()
#13 0x0000000809e6df00 in ?? ()
#14 0x00007fffffffdf60 in ?? ()
#15 0x0000000800bab351 in ?? ()
#16 0x00007fffffffe080 in ?? ()
#17 0x00000000004625a3 in ?? ()
#18 0x00007fffffffdfe0 in ?? ()
#19 0x0000000809dfa8a0 in ?? ()
#20 0x0000000000000000 in ?? ()
(gdb) q
I'll log out from the desktop environment, not restart, switch to ttyv1 and run pkg upgrade as root.
grahamperrin
Posts: 81
Joined: Sun Sep 04, 2016 6:06 am

pkg upgrade, not including gnupg

Post by grahamperrin »

Switch to ttyv1 failed:

Image

– so I switched back to ttyv8 then again logged in as grahamperrin.

Installed xterm, I'm now using xterm as root to run pkg upgrade.

With this run, the absence of gnupg is remarkable – it was in the partial last logged line for the prior (Update Station) upgrade routine.

I'll await completion of the current run then, to tell whether any file was missing:

pkg delete -f -y gnupg && pkg install -q -y --automatic security/gnupg
grahamperrin
Posts: 81
Joined: Sun Sep 04, 2016 6:06 am

gnupg was OK

Post by grahamperrin »

No missing files when I deleted (then reinstalled) gnupg.

Let's assume that when pkg crashed earlier, the reinstallation was done; that the log (without the word done) was incomplete due to the crash.

Code: Select all

grahamperrin@mowa219-gjp4-ghostbsd-13-vm:~ % su -
Password:
root@mowa219-gjp4-ghostbsd-13-vm:~ # date ; uptime
Sun Jun 13 13:44:54 BST 2021
 1:44PM  up  2:03, 1 user, load averages: 0.77, 1.50, 1.09
root@mowa219-gjp4-ghostbsd-13-vm:~ # bectl list -c creation
BE                      Active Mountpoint Space Created
default                 -      -          9.33G 2021-05-01 04:46
1300504                 NR     /          17.9G 2021-06-13 11:36
backup-2021-06-13-11-53 -      -          57.0M 2021-06-13 11:53
root@mowa219-gjp4-ghostbsd-13-vm:~ # uname -KrU
13.0-STABLE 1300500 1300504
root@mowa219-gjp4-ghostbsd-13-vm:~ #  pkg delete -f -y gnupg && pkg install -q -y --automatic devel/gnupg
Checking integrity... done (0 conflicting)
Deinstallation has been requested for the following 1 packages (of 0 packages in the universe):

Installed packages to be REMOVED:
	gnupg: 2.3.1

Number of packages to be removed: 1

The operation will free 11 MiB.
[1/1] Deinstalling gnupg-2.3.1...
[1/1] Deleting files for gnupg-2.3.1: 100%
pkg: No packages available to install matching 'devel/gnupg' have been found in the repositories
root@mowa219-gjp4-ghostbsd-13-vm:~ # pkg install -q -y --automatic security/gnupg
=====
Message from gnupg-2.3.1:

--
When run on hosts without IPv6 connectivity, GnuPG may fail to connect to
dual-stack hkp servers [1]. As a workaround, add:

disable-ipv6

to:

/usr/local/etc/dirmngr.conf

[1] https://dev.gnupg.org/rGecfc4db3a2f8bc2652ba4ac4de5ca1cd13bfcbec
root@mowa219-gjp4-ghostbsd-13-vm:~ # 
grahamperrin
Posts: 81
Joined: Sun Sep 04, 2016 6:06 am

Randomness

Post by grahamperrin »

Unfortunately, there's a sense of randomness. Whatever's wrong – I suspect that it's not an issue with GhostBSD, it's not (or not yet) easily reproducible.

With the machine successfully updated and booted in safe mode with 2 GB memory and 32 MB video memory: black screen + movable pointer after login.‡ Rebooted in safe mode: the same problem after login. Rebooted in normal mode: OK after login. Rebooted in safe mode: OK after login.

With 4 GB memory, 24 MB video memory and USB disabled: OK after login.

Reverted memory to 2 GB, video memory to 32 GB, USB re-enabled, booted in safe mode: OK after login. VirtualBox snapshot taken.‡‡

----

‡ Most remarkable about this session, for which I have a VirtualBox snapshot, was misbehaviour in ttyv1. If I recall correctly: when I typed root and keyed Enter, the cursor moved to the next line (as if there was input) but there was no visible character (no sight of the word 'root') and no prompt to enter a password.
grahamperrin
Posts: 81
Joined: Sun Sep 04, 2016 6:06 am

Evening summary

Post by grahamperrin »

I restored then started this VirtualBox snapshot, recorded the screen (of the KDE Plasma host, with the Ghost BSD guest in sight) in a way that might demonstrate some of what was wrong with ttyv1; and wrong after logging in with LightDM. https://photos.app.goo.gl/WvVWRYstce1Nk8YLA

----

‡‡ Given the good state of GhostBSD at the time of this VirtualBox snapshot, I don't plan to investigate further here.

Instead, I'll probably focus on comparable cases (me and three other people) in FreeBSD Forums.

For the GhostBSD guest, I'm purging some older snapshots, to regain free space for the host.

Note to self, aim to retain this named snapshot for as long as possible:
  • Safe boot succeeded, black screen + movable pointer after login.
Post Reply