First, I have to admit this article is not about running vscode on FreeBSD, at least not all. There is already a FreeBSD VSCode port on github which still has some issues. What I am trying to do here is to show you how to run vscode inside a bhyve guest system on the FreeBSD host via ssh X11 Forwarding.
bhyve
The First step is creating a Fedora 31 virtual machine in bhyve, where Visual Studio Code is actually running in.
1. Prepare the host
Runing the following scripts to load the bhyve kernel module and create network interfaces needed by the bhyve guest.
kldload vmm
ifconfig tap0 create
sysctl net.link.tap.up_on_open=1
ifconfig bridge0 create
ifconfig bridge0 addm igb0 addm tap0
ifconfig bridge0 up
Presist the configuration after reboot by adding the following lines to configuration files.
# cat /etc/rc.conf
cloned_interfaces="bridge0 tap0 lo1"
ifconfig_bridge0="addm em0 addm tap0"
kld_list="vmm"
# cat /etc/sysctl.conf
net.link.tap.up_on_open=1
2. Prepare a Fedora Guest
First, create a ZFS volume by:
zfs create -V100G -o volmode=dev zroot/fedora
Then, download an installation image of Fedora 31 to install:
curl -LO https://download.fedoraproject.org/pub/fedora/linux/releases/31/Server/x86_64/iso/Fedora-Server-dvd-x86_64-31-1.9.iso
3. Booting bhyve Guest with UEFI firmware
Install UEFI firmware images from sysutils/bhyve-firmware.
pkg install sysutils/bhyve-firmware
Now we can boot the guest with the image we just downloaded.
bhyve -AHP -s 0:0,hostbridge -s 31:0,lpc \
-s 2:0,virtio-net,tap0 -s 3:0,virtio-blk,/dev/zvol/zroot/fedora \
-s 4:0,ahci-cd,./Fedora-Server-dvd-x86_64-31-1.9.iso
-c 4 -m 8192M \
-s 20,xhci,tablet \
-s 29,fbuf,tcp=0.0.0.0:5900,w=1920,h=1080,wait \
-l bootrom,/usr/local/share/uefi-firmware/BHYVE_UEFI.fd \
fedora
We just create a VNC server for the Guest by adding the -s 29,fbuf,tcp=0.0.0.0:5900,w=1920,h=1080,wait flags. The Guest will boot, we can connect to it via a VNC client and finish the install procedure, when we have installed Fedora 31 on the virtual machine, reboot the Guest. Here we have to remove the instance of the virtual machine otherwise it cannot be started again.
bhyvectl --destroy --vm=fedora
4. Start Fedora VM
Start the Guest by:
bhyve -AHP -s 0:0,hostbridge -s 31:0,lpc \
-s 2:0,virtio-net,tap0 -s 3:0,virtio-blk,/dev/zvol/zroot/fedora \
-c 4 -m 8192M \
-s 20,xhci,tablet \
-s 29,fbuf,tcp=0.0.0.0:5900,w=1920,h=1080 \
-l bootrom,/usr/local/share/uefi-firmware/BHYVE_UEFI.fd \
fedora > /var/log/bhyve 2>&1 &
We can destroy a specified virtual machine by using bhyvectl:
bhyvectl --destroy --vm=fedora
X11 forwarding with Indirect GLX enabled
To enable Indirect GLX, we have to create ~/.xserverrc:
#!/bin/sh
/usr/local/bin/X -iglx "$@"
This time we use the command ssh -X user@fedora connecting to the Fedora Guest, and then input code in the terminal. Ahha, it works.
But there is still an issue, an error that saying “code not responding, wait or quit, blah blah” keeps coming. Using trusted X11 forwarding, aka -Y option, will get rid of this error. Unfortunately Intel HD Graphics 630 on my machine may not support trusted X11 forwarding or whatever, every time I exec ssh -Y user@fedora "code", the system will crash and reboot. So I borrow a GTX 650 from a friend.
pkg install nvidia_driver
And add these lines to /etc/rc.conf:
linux_enable="YES"
kld_list="vmm nvidia-modeset"
nvidia_name="nvidia"
nvidia_modset_name="nvidia-modeset"
Create an X11 configuration file named /usr/local/etc/X11/xorg.conf.d/nvidia.conf.
Section "Device"
Identifier "NVIDIA Card"
VendorName "NVIDIA Corporation"
Driver "nvidia"
EndSection
Reboot. Execute ssh -Y user@fedora "code" again. Finally, it works like a charm.