Previously, network uplink (eth0) was configured in two places:
- udev (asynchronously)
- qubes-misc-post.service - at the very end of the boot process
This caused multiple issues:
1. Depending on udev event processing (non-deterministic), network
uplink could be enabled too early, for example before setting up
firewall.
2. Again depending on udev processing, it can be enabled quite late in
the boot process, after network.target is up and services assume
network already configured. This for example causes qubes-firewall to
fail DNS queries.
3. If udev happen try to enable enable networking even earlier, it may
happend before qubesdb-daemon is started, in which case network setup
fill fail. For this case, there was network re-setup in
qubes-misc-post service - much later in the boot.
Fix the above by placing network uplink setup in a dedicated
qubes-network-uplink@${INTERFACE}.service unit ordered after
network-pre.target and pulled in by udev based on vif device existence,
to handle also dynamic network attach/detach.
Then, create qubes-network-uplink.service unit waiting for appropriate
interface-specific unit (if one is expected!) and order it before
network.target.
QubesOS/qubes-issues#5576
If IPv6 gateway address provided by dom0 isn't a link local address, add
a /128 route to it. Also, add this address on backend interfaces (vif*).
This is to allow proper ICMP host unreachable packets forwarding - if
gateway (address on vif* interface) have only fe80: address, it will be
used as a source for ICMP reply. It will be properly delivered to the VM
directly connected there (for example from sys-net to sys-firewall), but
because of being link-local address, it will not be forwarded any
further.
This results timeouts if host doesn't have IPv6 connectivity.
If dom0 expose IPv6 address settings, configure it on the interface.
Both backend and frontend side. If no IPv6 configuration is provided,
block IPv6 as it was before.
FixesQubesOS/qubes-issues#718
For a long time the DNS address was the same as default gateway. This is
still the case in R3.x, but using `qubes-gateway` configuration
parameter for it is misleading. It should be up to dom0 to provide DNS
address (whether the value is the same as gateway or not).
FixesQubesOS/qubes-issues#1817
Apparently unmanaged devices are loaded only from main
NetworkManager.conf. Exactly the same line pasted (not typed!) to main
NetworkManager.conf works, but in
/etc/NetworkManager/conf.d/30-qubes.conf it doesn't.
BTW There was a typo in option name ("unmanaged_devices" instead of
"unmanaged-devices", but it wasn't the cause).
This reverts commit 6c4831339c.
QubesOS/qubes-issues#1176
Previously even if NetworkManager was enabled, our script manually
configured network parameters. This apparently have negative effects,
because NetworkManager tries to configure some things differently - for
example use metric 1024 for default gateway.
FixesQubesOS/qubes-issues#1052
According to the specification[1], the setting name is 'addresses', not
'address'. The later apparently worked on some NetworkManager versions,
but for example not on the one in Debian wheezy. Also fix value
format (IP;netmask;gateway).
[1] htts://developer.gnome.org/NetworkManager/unstable/ref-settings.html
FixesQubesOS/qubes-issues#1280
Do not modify main /etc/NetworkManager/NetworkManager.conf as it would
cause conflicts during updates. Use
/etc/NetworkManager/conf.d/30-qubes.conf instead.
Also remove some dead code for dynamically generated parts (no longer
required to "blacklist" eth0 in VMs - we have proper connection
generated for it). It was commented out for some time already
FixesQubesOS/qubes-issues#1176
A file is created in /var/lib/qubes/protected-files. Scripts can grep this file before modifying
known files to be protected and skip any modifications if the file path is within protected-files.
Usage Example:
if ! grep -q "^/etc/hostname$" "${PROTECTED_FILE_LIST}" 2>/dev/null; then
Also cleaned up maintainer scripts removing unneeded systemd status functions and streamlined
the enable/disable systemd unit files functions
This patch introduces two new qvm-services:
- disable-default-route
- disable-dns-server
Both disabled by default. You can enable any of them to not set default
route and/or DNS servers in the VM. Those settings have no effect on
NetVM, where such settings are controlled by NetworkManager.
This is based on patch sent by Joonas Lehtonen
<joonas.lehtonen@openmailbox.org>
https://groups.google.com/d/msgid/qubes-devel/54C7FB59.2020603%40openmailbox.org
Conflicts:
network/setup-ip
vm-init.d/qubes-core
vm-systemd/qubes-sysinit.sh
This patch introduces two new qvm-services:
- set-default-route
- set-dns-server
Both enabled by default. You can disable any of them to not set default
route and/or DNS servers in the VM. Those settings have no effect on
NetVM, where such settings are controlled by NetworkManager.
This is based on patch sent by Joonas Lehtonen
<joonas.lehtonen@openmailbox.org>
https://groups.google.com/d/msgid/qubes-devel/54C39656.3090303%40openmailbox.org
Conflicts:
network/setup-ip
vm-init.d/qubes-core
vm-systemd/qubes-sysinit.sh
Backend domain generates its IP address based on frontend IP, not
settings given from dom0. So change frontend method to the same (for
DispVM it makes a difference). Now "qubes-gateway" xenstore entry is
basically primary DNS address only.
Get rid of underscores in filenames, use dashes instead.
This is first part of cleanup in filenames.
"qubes_rpc" still untouched - will be in separate commit.
