My command is:
kvm -m 4G -cpu host \
-device e1000,netdev=net0 \
-netdev user,id=net0,hostfwd=tcp::10022-:22,hostfwd=tcp::3389-:3389 \
-device intel-hda \
-device hda-duplex \
-spice port=5930,disable-ticketing=on \
-device virtio-serial \
-device virtserialport,chardev=vdagent,name=com.redhat.spice.0 \
-chardev spicevmc,debug=0,id=vdagent,name=vdagent \
-vga qxl \
-virtfs local,mount_tag=home,path=/home/arno,security_model=none \
-drive file=debian-testing-amd64-netinst.img,format=raw \
-boot d -cdrom debian-testing-amd64-netinst.iso
I use remote-viewer and SPICE protocol to interact with the VM.
Maybe of interest, I see some errror messages in /var/log/syslog, after
I confirm the second screen. Here's a good chunk of logs for context,
the actual errors are the lines with "main-menu[527]: (process:xxxx):":
Nov 23 00:55:38 debconf: Setting debconf/language to en
Nov 23 00:55:38 main-menu[527]: INFO: Falling back to the package
description for brltty-udeb
Nov 23 00:55:38 main-menu[527]: INFO: Menu item 'localechooser' selected
Nov 23 00:55:40 debconf: Setting debconf/language to en
Nov 23 00:55:47 init: starting pid 507, tty '/dev/tty2': '-/bin/sh'
Nov 23 00:56:05 localechooser: info: Language = 'en'
Nov 23 00:56:05 localechooser: info: line=en;0;US;en_US.UTF-8;;console-setup
Nov 23 00:56:05 localechooser: info: Set debian-installer/language = 'en'
Nov 23 00:56:05 localechooser: info: Default country = 'US'
Nov 23 00:56:05 localechooser: info: Default locale = 'en_US.UTF-8'
Nov 23 00:56:05 localechooser: info: Set debian-installer/consoledisplay
= 'console-setup'
Nov 23 00:56:05 debconf: Setting debconf/language to en
Nov 23 00:56:08 localechooser: info: Set debian-installer/country = 'US'
Nov 23 00:56:08 localechooser: info: Set debian-installer/locale =
'en_US.UTF-8'
Nov 23 00:56:08 localechooser: info: System locale
(debian-installer/locale) = 'en_US.UTF-8'
Nov 23 00:56:08 main-menu[527]: (process:538): ls:
/usr/share/mbrola/en[1-9]/en[1-9]: No such file or directory
Nov 23 00:56:08 main-menu[527]: (process:538): sh: missing ]
Nov 23 00:56:08 main-menu[527]: (process:538): ls:
/usr/share/mbrola/en[1-9]/en[1-9]: No such file or directory
Nov 23 00:56:08 main-menu[527]: (process:538): sh: missing ]
Nov 23 00:56:08 main-menu[527]: INFO: Falling back to the package
description for brltty-udeb
Nov 23 00:56:08 main-menu[527]: INFO: Falling back to the package
description for brltty-udeb
Nov 23 00:56:08 main-menu[527]: INFO: Falling back to the package
description for brltty-udeb
Nov 23 00:56:08 main-menu[527]: INFO: Menu item 'brltty-udeb' selected
Nov 23 00:56:08 main-menu[527]: WARNING **: Unable to set title for
brltty-udeb.
Nov 23 00:56:08 main-menu[527]: INFO: Falling back to the package
description for brltty-udeb
Nov 23 00:56:08 main-menu[527]: INFO: Falling back to the package
description for brltty-udeb
Nov 23 00:56:08 main-menu[527]: INFO: Menu item 'espeakup-udeb' selected
Nov 23 00:56:10 main-menu[527]: (process:1753): ls:
/usr/share/mbrola/en[1-9]/en[1-9]: No such file or directory
Nov 23 00:56:10 main-menu[527]: (process:1753): sh: missing ]
Nov 23 00:56:10 main-menu[527]: INFO: Falling back to the package
description for brltty-udeb
Nov 23 00:56:10 main-menu[527]: INFO: Falling back to the package
description for brltty-udeb
Nov 23 00:56:10 main-menu[527]: INFO: Menu item 'console-setup-udeb'
selected
The errors related to mbrola and sh come from
debian/espeakup-udeb.restart in src:espeakup. I don't know if it's
relevant to the issue at hand though.
Just to confirm, I grabbed the debian-testing-amd64-netinst.iso from
today (2022-11-22) and I can confirm that I can still reproduce the issue.
Another interesting detail is that restarting espeakup makes it speaks
again, like if nothing happens. So I think it could be useful to have
something monitor the daemon and restart it in case it crashes. Would
make it more reliable overall, I suppose. Unfortunately I'm not familiar
with d-i and I couldn't even find from *where* espeakup is started (no
etc/init.d, no systemd).
Cheers,