Created
May 29, 2017 15:16
-
-
Save voutilad/a5080909e88e8dcffd1960312b5f9510 to your computer and use it in GitHub Desktop.
Installing Alpine Linux in OpenBSD's VMM Hypervisor
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# Assuming you're a regular user that has doas allowances for vmctl | |
mkdir -p ~/vmm | |
cd ~/vmm | |
# Grab the the one of the virt iso's of Alpine Linux | |
curl https://nl.alpinelinux.org/alpine/v3.6/releases/x86_64/alpine-virt-3.6.0-x86_64.iso -o alpine-virt-3.6.0-x86_64.iso | |
# Make a new virtual disk image, change the size as needed | |
vmctl create alpine-virt.img -s 6G | |
# Boot Alpine from the ISO. Make sure you have this exact disk order because as of 29 May 2017 | |
# VMM's SeaBIOS will only try to boot from the first one it seems! | |
# Also, this assumes: | |
# - you want 1024M of memory, tune as desired | |
# - you configured a virtual switch called "local" in your /etc/vm.conf | |
# (see: http://www.h-i-r.net/2017/04/openbsd-vmm-hypervisor-part-2.html) | |
doas vmctl start alpine -d alpine-virt-3.6.0-x86_64.iso -d alpine-virt.img -n local -m 1024M -c | |
# You shoud get a serial console connection immediately. Hit enter or whatever to boot Alpine. | |
# Once in Alpine, run: | |
setup-alpine | |
Done. Thanks @voutilad :)
See my reply on the email list: https://marc.info/?l=openbsd-misc&m=167303917804395&w=2
In short, I think this is an issue we will fix in -current
but since I don't think I have newer Intel hardware that supports the TSLEEP
instructions to test my changes it may be a few days to confirm.
Looks like recent patch in 7.2 fixed this. Thank you!
Looks like recent patch in 7.2 fixed this. Thank you!
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Are you on Intel-based hardware for the OpenBSD host? If so, it's probably the Linux kernel probing some expected Intel hardware via MMIO registers.
vmm(4)
andvmd(8)
do not implement support for emulating that access yet, but there is some preliminary support I committed a few months ago that's gated behind an #ifdef. It requires a-current
host and buildingvmd(8)
from source and twiddling theMMIO_NOTYET
definition.It might help the guests boot. If not, check the output and share it on
[email protected]
for me please.