<![CDATA[
DESKTERM SCRIPT # put in ~/.autostart/deskterm
#!/bin/bash
xfce4-terminal --hide-borders --hide-toolbar --hide-menubar --title=deskterm --geometry=110x20+0+0
sleep 3
wmctrl -r deskterm -b add,below,sticky
wmctrl -r deskterm -b add,skip_taskbar,skip_pager
wmctrl -r deskterm -k off
WHISKER MENU ON SUPER KEY
sudo apt install xcape and add xcape -e 'Super_L=Control_L|Escape' to autostart.
from https://spass.neocities.org/tweaks.html
TO CONTINUE USING AN OLD UBUNTU DISTRO
sudo vi /etc/apt/sources.list
and change "archive" to "old-releases"
MAKE LIBVIRT-QEMU NOT SHOW IN USERLIST
# to hide the system user "libvirt-qemu" from login screen
# if running manually (not in laarger script), wrap in "sudo bash -c"
sudo echo -e "[User]\nSystemAccount=true" > /var/lib/AccountsService/users/libvirt-qemu
TO CHANGE THE DESKTOP ICON FOR SHOW DESKTOP
sudo ln -s ~/.icons/user-desktop.png /usr/share/icons/elementary-xfce/places/48/user-desktop.png
MAKE THE USB MOUNT WITH PERMS SO YOU CAN COPY+PASTE TO IT THRU GUI
sudo chown caiser:caiser /media/caiser/97e0788e-3d9f-4262-a6d9-d06bc528615f
# thereafter, it will keep these perms
HADES CANYON GPU PASSTHROUGH WITH KVM
i can't get the host to boot to the igpu instead of the vega
---
theory:
it's getting yoinked and maybe the machine doesn't like that? causing blackscreen?
my /sys/bus rescan stuff doesn't help, but:
even other dude said couldn't get vega to go back to host after the vms shut down
rescannning didn't help him either
solution: (?)
sudo vi /etc/default/grub # to add nomodeset:
GRUB_CMDLINE_LINUX_DEFAULT="nomodeset intel_iommu=on pcie_acs_override=downstream"
update-grub # to rebuild the initrd or whatever
boot to multi-user.target
start the virtmachine
boot to graphical.target
???
lets try it
if fails, can't look at manual ways to tell it to not use the vega
but without disabling it to the extent that its not accessible by the VM!!!
---
did try it
it works for about a second
boots to a command-line environment
checked graphics with:
sudo lshw -c video # note both displays UNCLAIMED
*-display UNCLAIMED
description: VGA compatible controller
product: Polaris 22 [Radeon RX Vega M GL]
vendor: Advanced Micro Devices, Inc. [AMD/ATI]
physical id: 0
bus info: pci@0000:01:00.0
version: c0
width: 64 bits
clock: 33MHz
capabilities: pm pciexpress msi vga_controller bus_master cap_list
configuration: latency=0
resources: memory:90000000-9fffffff memory:a0000000-a01fffff ioport:e000(size=256) memory:db500000-db53ffff memory:c0000-dffff
*-display UNCLAIMED
description: Display controller
product: Intel Corporation
vendor: Intel Corporation
physical id: 2
bus info: pci@0000:00:02.0
version: 04
width: 64 bits
clock: 33MHz
capabilities: pciexpress msi pm cap_list
configuration: latency=0
resources: iomemory:2f0-2ef iomemory:2f0-2ef memory:2ffe000000-2ffeffffff memory:2fa0000000-2fafffffff ioport:f000(size=64)
vm starts and everything shows up in:
virsh start ubuntu18.10
virsh list
and a repeat check of the graphics displays shows the vega is no longer unclaimed
but:
top
shows systemd-journal going out of control and eating all the cpu
and then... the same blackscreen
also tried booting to multi-user.target, then manually isolating graphical.target
(which doesn't autostart the graphics displays, weirdly -- both remained UNCLAIMED)
same results
likewise when I stuck to a pure command-line environment
although that did briefly show an error:
the end trace bit of a kernel panic (couldn't scroll up)
and smth about "recoverable error, must reboot tho"
systemd-journal puts logs in /var/log/journal but they are unreadable binaries???
perhaps the vm and the host are fighting over the gpu
???
]]>