In rare cases when vif-route-qubes is called simultaneously with some
other iptables-restore instance, it fails because of missing --wait (and
recent iptables-restore defaults to aborting instead of waiting
for lock). That other call may be from qubes-firewall or user script.
Related to QubesOS/qubes-issues#3665
It may be useful to create AppVM-specific menu entries in AppVM itself.
It may be an application installed there (in /usr/local, or using snap
QubesOS/qubes-issues#2766), but it may be also some user custom
shortcut.
To support this, dom0 will accept menu entries also from
TemplateBasedVMs. But to avoid duplicates, qubes.GetAppmenus service
should send only menu entries actually stored in that VM, not inherited
from its template. To distingush them, first check what type of
persistence this VM has (from qubesdb-read /qubes-vm-persistence). If
it's rw-only, send only entries stored on /rw.
To make it more robust, use $XDG_DATA_DIRS and $XDG_DATA_HOME to
discover directories, instead of looking only for
/usr/{,local/}share/applications. This makes snap and flatpak handled
for free.
FixesQubesOS/qubes-issues#4152
Previously, bind-dirs.sh had a bunch of `true comment goes here` style debug messages (no-ops). Presumably this was done because these messages are intended as debug messages and would only be displayed when calling bind-dirs.sh with xtrace enabled. However, this includes some fatal errors, which are necessary to debug why bind-dirs.sh is ostensibly not working. For example, I tried to mount /var/lib/docker, didn't realize it did not exist (as an empty directory) in my base template, and there was no journalctl output at all. After this change, journalctl will contain the (very helpful) error message.
If xvda is not partitioned, checking size of device mounted as root dev
doesn't detect xvda resize - it will have the new size, while the fs on
it not. Change to checking actual filesystem size.
Also, improve log message to include the size.
FixesQubesOS/qubes-issues#4274
The qubes-core-agent-passwordless-root package ships sudo configuration,
adding to sudo group isn't needed.
Basically revert all changes made by qubes-core-agent-passwordless-root
installation.
FixesQubesOS/qubes-issues#4015
'which' is an external tool, 'type' is a shell builtin. Using the latter
shaves off a bit of latency.
Also use the already open stderr file descriptor for redirection.
- Symlink the other three tools to qvm-copy-to-vm.gnome, use the same
code where possible, and select differing behavior based on invoked
name (like the CLI tools). This brings qvm-move-to-vm.kde up to date
for R4.0 (bugfix on 89183e9).
- Get rid of a window focus race between zenity/kdialog and the
qubes.Filecopy dom0 permission dialog: Only launch the GUI after the
first line has been read from qfile-agent.
- Avoid visual glitches (e.g. for a non-existing file) by special-casing
a no-op progress function for $SIZE == 0.
- Pass -- separator between 'rm -rf' and the files to be removed, in
case someone or something ever invokes the tool on relative file names
starting with a dash.
- Pass -b (implies --apparent-size) and -s to du, to simplify percentage
calculation and to avoid unnecessary output.
The qvm-{copy,move}-to-vm.{gnome,kde} cancel buttons didn't actually
cancel, because qfile-agent ignored EPIPE and - via qfile_pack_init() -
SIGPIPE. So it never noticed when the local PROGRESS_TYPE=gui reader had
shut down.
for qvm-move
Added more descriptive usage text to the
qvm-copy/qvm-move/qvm-copy-to-vm/qvm-move-to-vm family of tools.
Also fixed bug that removed the file being moved for qvm-move-to-vm,
but not for qvm-move.
fixesQubesOS/qubes-issues#3529fixesQubesOS/qubes-issues#4020
As described in QubesOS/qubes-issues#3758, a VM may fail to start
if the volume is large. This because the whole volume is read to
ensure it's empty (=all zeros).
This changes limits the check to the first 1 GiB of the private
volume. As additional safety, a scan with blkid is done.
Fixes https://github.com/QubesOS/qubes-issues/issues/3758
Make sure that qubesdb is configured (including service start) before
executing postinst of qubes-core-agent package, which will communicate
with qubesdb service.
FixesQubesOS/qubes-issues#3951
* view-only:
Add file managers integration for qvm-open-in-dvm --view-only
qvm-open-in-vm: mark file as read-only if opened with --view-only
qvm-open-in-vm: implement --view-only option
qubes-rpc: fix code style - indent with spaces
This will cause most applications to disallow changing the file and also
add some visual indication about the view being read only. This will
avoid making the changes that would be discarded later.
QubesOS/qubes-issues#1118
Implement option to disallow (ignore in fact) modifications of file
opened in another VM (including DispVM).
This commit implements actual services part and handling in wrapping scripts.
FixesQubesOS/qubes-issues#1118
I've run into an issue with incorrectly generated rules for IPv6. I
added some debugging code printing the generated rules and the
resulting error (see below). Turns out "reject with" expects icmpv6
rather than icmp6.
--- generated rule ---
flush chain ip6 qubes-firewall qbs-fd09-24ef-4179--a89-15
table ip6 qubes-firewall {
chain qbs-fd09-24ef-4179--a89-15 {
ip6 daddr fc00::/8 reject with icmp6 type admin-prohibited
ip6 daddr fd00::/8 reject with icmp6 type admin-prohibited
ip6 daddr fe80::/10 reject with icmp6 type admin-prohibited
accept
reject with icmp6 type admin-prohibited
}
}
--- output ---
/dev/stdin:4:36-40: Error: syntax error, unexpected string, expecting icmp or icmpv6 or tcp or icmpx
^^^^^
/dev/stdin:5:36-40: Error: syntax error, unexpected string, expecting icmp or icmpv6 or tcp or icmpx
^^^^^
/dev/stdin:6:37-41: Error: syntax error, unexpected string, expecting icmp or icmpv6 or tcp or icmpx
^^^^^
/dev/stdin:8:17-21: Error: syntax error, unexpected string, expecting icmp or icmpv6 or tcp or icmpx
^^^^^
qubes-firewall service is meant as mistakes mitigation, not a
high-volume external network filter. Providing feedback (ICMP
admin-prohibited error packet) to the VM reduces timeouts and give
immediate connection failure. This is especially useful when some
website tries to load unwanted (not whitelisted) 3rd party elements -
providing error response avoids long site loading time.
FixesQubesOS/qubes-issues#3607
Some applications complains if compiled version of dconf database is
missing ("dconf-WARNING **: unable to open file '/etc/dconf/db/local':
Failed to open file '/etc/dconf/db/local': open() failed: No such file
or directory; expect degraded performance").
There is only one entry in that database, but generate its binary
version anyway to avoid that warning message.
The dconf call is already included in package scripts, now only make
sure the utility is really installed.
QubesOS/qubes-issues#1951
The /etc/dconf/profile/user file in some distributions is part of dconf
package, in some not. There are even cases where it changes between
package versions (Fedora 27 don't have it, but Fedora 28 do).
Also, base Debian Stretch don't have it, but Kali Linux based on it do.
To avoid overly complex dependency handling, create the file dynamically
on package installation if it's missing in that particular case. The
file content is canonical:
user-db:user
system-db:local
FixesQubesOS/qubes-issues#3834
- BACKEND_VMM may not be available as env variable (mock build), provide
it explicitly
- 'user' group may not exists at package build time, set it at package
installation
qubes-firewall.service have Before=qubes-network.service. The latter
enable ip_forwarding. Make sure the ordering cover not only service
fork, but all its startup sequence, including initial rules and user
scripts.
Reported-by: @tasket
It doesn't make sense in Qubes to rebuild modules at every VM startup.
Especially when modules dir is either read-only, or on non-persistent
root volume (TemplateBasedVM).
This is especially uneeded for dom0-provided kernels, which already have
u2mfn module built, but DKMS try to rebuild it anyway (and fails,
because of missing/broken kernel-devel package).
DKMS already have appropriate (rpm/dpkg) hooks for new kernel
installation, so actually needed modules should be build in template on
kernel update/install anyway.
This saves about 2s from VM startup time.
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.
Since fixing QubesOS/qubes-issues#3213, launch function correctly waits
for some applications exit. This is undesirable for
qubes-session-autostart service, which should just start the
applications and exit.
This is especially important for qubes-desktop-run used inside DispVM.
The DesktopAppInfo.launch() method returns after just launching the
application. In DispVM case it worked by a coincidence - because the
launched application was keeping stdin/out open, which also prevented
DispVM killing. Use DesktopAppInfo.launch_uris_as_manager which at least
allows to learn PIDs of spawned processes, to track them manually.
This still doesn't fix gnome-terminal issue, or any other application
using either DBus activation, or any other client-server model. But at
least fix basic apps like firefox and xterm.
FixesQubesOS/qubes-issues#3213
Since we have proper python package, use it instead of hacky one-file
package. This will ease installation and packaging, including switching
to python3.
Add an option for custom vchan buffer size, to override default 64k (for
each direction). This is especially useful when the other side of
connection is MirageOS based, because of limited memory and default
grant table size (128 entries).
The dconf package isn't required by qubes-core-agent - the package ships
a configuration for it, useful if user have it installed for other
reasons. Don't try to rebuild dconf database if dconf isn't installed -
avoid misleading error message.
FixesQubesOS/qubes-issues#3492
On first VM's boot, setup-rwdev.sh script create filesystem on
/dev/xvdb. But it does so only after checking if /dev/xvdb is really
empty, by comparing it to /dev/zero. Speed up reads from /dev/zero bu
using larger blocks (default of head - 8k, instead of explicit 512).
This speed up the check over 5 times.
* qubesos/pr/94:
Drop fakeroot for list/search actions on Debian
dom0-update: add some approximation of 'list', 'search' and 'reinstall'
dom0-updates: refactor for ease adding new actions with old yum
Previously the script was called through shell as:
execl(shell, "-sh", "-c", "/usr/lib/qubes/qubes-rpc-multiplexer
...", 0);
This tells the shell to load login scripts, including /etc/profile.
Since 5512e4eada this is no longer the
case and the script is called directly. Since most services do expect
proper user session initialized (/etc/profile loaded etc), adjust the
script's shebang to behave like a login shell and load those startup
scripts.
FixesQubesOS/qubes-issues#3615
This avoids duplicating service call parsing in multiple places.
Further improvements to that code (like avoid using shell) can be
implemented in one place.
... instead of requested user. This makes sure that session startup
script will be able to send a signal here to notify about session
startup.
This is especially needed when the first service started in the VM is
called as root (like qubes.InstallUpdatesGUI).
FixesQubesOS/qubes-issues#3526
Update dom0 about all applications installed, not only desktop files for
them. Update also supported features and other things advertised
initially at template installation.
FixesQubesOS/qubes-issues#3579
When VM is set to synchronize the time with the network, to not sync its
time with clockvm.
Besides not having sense, in default configuration it will lead to
loopback qrexec connection (sys-net -> sys-net), which will hang.
QubesOS/qubes-issues#3333
NetworkManager reports a bunch of events, reloading DNS at each of them
doesn't make sense and is harmful - systemd have ratelimit on service
restart.
FixesQubesOS/qubes-issues#3135
nullglob must be active before a glob is used on a potentially empty
directory. Call shopt immediately after the shebang line.
FixesQubesOS/qubes-issues#3552
Call it just after creating base chains in iptables/nftables. This allow
the user to modify how those rules are plugged in, add custom rules at
beginning/end etc.
FixesQubesOS/qubes-issues#3260
The most common thing to fix S3 sleep is to unload PCI devices drivers
before suspend. Instead of having every user figuring out what drivers
needs to be blacklisted, detach all drivers from actual PCI devices.
Exclude qemu emulated devices.
FixesQubesOS/qubes-issues#3486
qvm-run-vm cannot make a separate qubes.WaitForSession call for a
DispVM. Instead, pass the new WaitForSession argument to qubes.VMShell,
which will do the equivalent.
GNOME automatically set scaling factor to 2 when HiDPI is detected.
Unfortunately it does it also on not really HiDPI displays, making the
whole UI unusably large. There is no middle ground - scaling factor must
be integer, so 1.5 is not supported. Lets opt on a conservative side and
fallback to scaling factor 1.
Solution by @alyssais, thanks!
FixesQubesOS/qubes-issues#3108
When one use scaling set by gnome tools (gsettings or
gnome-tweak-tool), gsd-xsettings must be running to apply the change
also to other applications.
This include auto scaling on HiDPI screens.
This commit fixes non-uniform behaviour on different VM types.
QubesOS/qubes-issues#3108
There was a bug that interpreted '-' as file type. But convert don't
know how to handle '-' file type, so refused to proceed.
FixesQubesOS/qubes-issues#3085
If fork server is used, proper environment is inherited from the
session. But in other case (like non-default user), it needs to be
created by qrexec-agent itself. PAM provide some variables, but not the
most basic: HOME, SHELL, USER, LOGNAME. Also process should be started
in user home directory (if available).
FixesQubesOS/qubes-issues#3416
I have been using this with a dbus-less Gentoo template since the original
change, and have tested recently on whonix-gw with dbus enabled and running.
(cherry picked from commit bf69335074)
It is more robust, especially handle "# Units below this line will be
re-preset on package upgrade" part of 75-qubes-vm.preset file. This is
needed to fix system configuration without the need to rebuild the whole
template.
QubesOS/qubes-issues#2913
* qubesos/pr/72:
Fix UCA mistake and qvm-actions script
Fix ShellCheck comments
Add debian package support
Disable Thunar thumbnails
Add support for Thunar Qubes VM tools
Chain name in IPv6 cannot be longer than 29 chars, so strip IPv6 prefix
from it.
ICMP on IPv6 is a different protocol than on IPv4 - handle iptables rule
accordingly.
QubesOS/qubes-issues#718
If IPv6 is configured in the VM, and it is providing network to others,
apply IPv6 firewall similar to the IPv4 one (including NAT for outgoing
traffix), instead of blocking everything. Also, enable IP forwarding for
IPv6 in such a case.
FixesQubesOS/qubes-issues#718
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
* qubesos/pr/67:
archlinux fix .service added twice in networking install script
Makefile: install-netvm shouldn't be a dependency of itself.
archlinux: add recently splitted packages as optional dependencies of qubes-vm-core
archlinux: fix incorrect keyring being populated
Makefile: remove invalid reference to network dropins install target
archlinux: fix shellcheck issues
archlinux: create a keyring package to install binary repository automatically
Makefile: add network install targets to install-deb
Makefile: fix typo created when spliting the install targets
Makefile: add basic networking to the new install-corevm target
archlinux: split core-agent from netvm-agent
Makefile: ensure that everything is installed by default for rh based agents
Makefile: split network install target from core agent install target
Start qubes-firewall (which will add "DROP by default" rule) before
enabling IP forwarding, to not leave a time slot where some connection
could go around configured firewall.
QubesOS/qubes-issues#3269
In some cases it may make sense to enfoce outgoing firewall also on
sys-net. If the service is disabled, firewall settings will be
(silently) ignored, so better be on the safe side and enable.
QubesOS/qubes-issues#3290
When qubes-firewall service is started, modify firewall to have "DROP"
policy, so if something goes wrong, no data got leaked.
But keep default action "ACCEPT" in case of legitimate service stop, or
not starting it at all - because one may choose to not use this service
at all.
Achieve this by adding "DROP" rule at the end of QBS-FIREWALL chain and
keep it there while qubes-firewall service is running.
FixesQubesOS/qubes-issues#3269
4.0 template builds use `<package>.install` files with dh_install. The
differences between Debian and Ubuntu packages also need to be represented
in these files.
* qubesos/pr/63:
archlinux: restore setup of pam.d/su-l
archlinux: remove python3 dependency
archlinux: ensure [options] section is present in all pacman drop-ins
archlinux: enforce usage of python2 in all scripts
Makefile: avoid using python interpreter as a static name
archlinux: create user 'user' using bash by default instead of zsh
archlinux: ship pam.d/qrexec as a replacement of using su
archlinux: do not mess with locales in post-install script
archlinux: remove pam configuration for su and su-l
archlinux: remove deprecated setup of pam since v4.0.3
Add the 4.0 repo to the PKGBUILD sources list
Restore the binary pacman repo and update it for QubesOS 4.0
Fix the makefile for archlinux - SBINDIR is already /usr/bin
Update the arch PKGBUILD script for QubesOS 4.0
systemd-timesyncd.service isn't enough, for various reasons:
- it is started too early in the boot process - files in
/var/run/qubes-services are not yet there
- by default it does only one shot synchronization, and there is no
network at that early boot time yet
- by default use-ntp is set to "no"
So, in addition, enable actual ntp client.
FixesQubesOS/qubes-issues#3210
Since the qubes-download-dom0-updates script executes dnf with fakeroot, some dnf plugins like etckeeper break the update with "Permission denied" errors.
qubes-gui agent calls su-l instead of initializing its own pam
session such as qrexec.
pam.d/su-l qubes specific configuration must be restored to ensure
that the user login session is properly initialized:
https://github.com/QubesOS/qubes-issues/issues/3185
* fixes-20171019:
debian: cleanup after splitting qubes-core-agent
Fix removing temporary file after editing in (Disp)VM
network: fix rules for network setup on new udev
debian: disable timer-based apt-get
Fix removing the file - do not free its filename just before unlink call
(scheduled with atexit function).
At the same time, place the temporary file in a unique directory,
making it possible to edit multiple files with the same name at once.
Remove that directory at exit too.
FixesQubesOS/qubes-issues#3112
New udev have `DRIVERS` matcher, instead of `ENV{ID_NET_DRIVER}`. Add
appropriate rule to the file. Without it, network was working
incidentally, because there is a fallback in qubes-misc-post.service,
but dynamic network change was broken.
This applies at least to Debian stretch.
FixesQubesOS/qubes-issues#3192
Debian stretch in default configuration calls apt-get update every 24h.
And additionally, have automatic unattended security updates enabled.
Generally it would be good thing on standalone system, but in AppVM
which loose its rootfs changes after restart it is a waste of resources.
Especially when it kicks in on multiple VMs simultaneously, while on
battery (apt-daily.service have ConditionACPower=true, but VM don't have
that information...).
It would make some sense on TemplateVM/StandaloneVM, but then it kicks
in just at VM startup. Which conflicts with starting the update manually
then (by clicking "update VM" button in manager for example, or using
salt).
So, disable this feature completely.
The actual solution is based on pkg-manager-no-autoupdate by @adrelanos.
FixesQubesOS/qubes-issues#2621
If root filesystem is the last partition (new layout), resize it
in-place. Use 'parted' tool because it can resize just one partition,
without need to specify the whole new partition table. Since the
partition is mounted, parted is unhappy to modify it. Force it by
answering to its interactive prompts, and add (apparently not
documented) ---pretend-input-tty to use those answers even
though stdin is not a tty. Split the operation into multiple parted
calls, for more reliable interactive prompts handling.
Qubes 3.x disk layout (no partition table) is also supported, but the
one that was used in Qubes 4.0 rc1 (root filesystem as the first
partition) is not.
FixesQubesOS/qubes-issues#3173QubesOS/qubes-issues#3143
* fixes-20171002:
qubes.ResizeDisk: handle dmroot being a symlink
qrexec: use user shell instead of hardcoded /bin/sh
qrexec: code style fix - use spaces for indentation
Add convenient wrappers for qvm-copy-to-vm and qvm-move-to-vm
Default `ask` policy ignore target domain specified by the caller, so it
doesn't make sense to specify one. Provide convenient wrappers not
needing one. Do not change behaviour of existing tools for compatibility
reasons.
FixesQubesOS/qubes-issues#3141
Explicitly block something like "curl http://127.0.0.1:8082" and
return error page in this case. This error page is used in Whonix to
detect if the proxy is torrified. If not blocked, it may happen that
empty response is returned instead of error. See linked ticket for
details.
This was previously done for 10.137.255.254, but since migration to
qrexec-based connection, 127.0.0.1 is used instead.
FixesQubesOS/qubes-issues#1482
* qubesos/pr/53:
Handle fallthrough with attribute(noreturn) for consistancy and compatiblity with older GCC
Add CENTOS/RHEL support (drop fedora-release dependancy as template builder will install it anyway and here it only make harder to support non-fedora builds)
Few reasons for this:
1. new templates use dnf to download packages, so yum.conf is unused
2. dom0 in Qubes 4.0 don't have this file at all (so sed fails here)
3. $OPTS already contains --setopt=reposdir=...
FixesQubesOS/qubes-issues#2945
Some services require GUI access. Make qrexec-agent handling this, based
on per-service configuration, instead of forcing every caller to call
qubes.WaitForSession service first. This is especially important for
Disposable VMs, because those are destroyed after a single service call.
This needs to be done in qrexec-agent (instead of service script, or
qubes-rpc-multiplexer), because agent will behave differently depending
on GUI session being available or not. Namely, will use
qrexec-fork-server (so the process will be a child of session leader),
or will open new session.
Service configuration lives in /etc/qubes/rpc-config/SERVICE_NAME, can
can contain 'key=value' entries (no space around '=' allowed). Currently
the only settings supported is 'wait-for-session', with value either '0'
or '1'.
QubesOS/qubes-issues#2974
Prepend "-" to shell name, to instruct it being a login shell. This way
shell will initialize environment, load /etc/profile etc.
FixesQubesOS/qubes-issues#2903
Since some systemd services are moved to other packages, appropriate
%post/%preun should contain the code to enable/disable them.
FixesQubesOS/qubes-issues#2894
clock synchronization mechanism rewritten to use systemd-timesync instead of NtpDate; at the moment, requires:
- modifying /etc/qubes-rpc/policy/qubes.GetDate to redirect GetDate to designated clockvm
- enabling clocksync service in clockvm ( qvm-features clockvm-name service/clocksync true )
Works as specified in issue listed below, except for:
- each VM synces with clockvm after boot and every 6h
- clockvm synces time with the Internet using systemd-timesync
- dom0 synces itself with clockvm every 1h (using cron)
fixesQubesOS/qubes-issues#1230
When there is only console=hvc0 (i.e. no output to emulated VGA) and
GRUB_TIMEOUT is set to 0, VM startup hangs. This may be very well some
race condition broken by either of console=tty0 or GRUB_TIMEOUT > 0, but
even in such a case, apply this as a workaround for now.
Instead of calling 'su' to switch the user, use own implementation of
this. Thanks to PAM it's pretty simple. The main reason is to have
control over process waiting for session termination (to call
pam_close_sesion/pam_end). Especially we don't want it to keep std* fds
open, which would prevent qrexec-agent from receiving EOF when one of
them will be closed.
Also, this will preserve QREXEC_AGENT_PID environment variable.
FixesQubesOS/qubes-issues#2851
It isn't really needed. It was used to workaround libusb bug (causing
crash when the system does not have any USB controller), but since we
use HVM now which do have some USB controllers it isn't needed anymore.
Also, it is not available in stock Fedora kernels.
When qrexec-client-vm is started with socket on its stdout and no local
process requested, it will try to shutdown(SHUT_WR) this socket when
remote process exists. This is wrong, because this socket may be still
needed by other processes (for example shell from where qrexec-client-vm
was called).
In such a case, simple close() should be used.
1. If local process is started, report its exit code, instaed of remote
one. To get remote exit code, simply start qrexec-client-vm without
third argument (and connect its stdin/stdout with the other process some
other way).
2. Report process terminated by signal.
Don't pretend that process terminated by signal finished successfuly.
Copy shell behaviour of reporting it as 128+signum.
3. Do not wait() for any child process, just the one we expect. In case
of qrexec-client-vm the child process is started differently and
wait()ing on it inside main loop would break its exit code reporting.
FixesQubesOS/qubes-issues#2861
Exit code 1 is very common in all kind of programs, including qrexec
services, so it is hard to distinguish remote failure from service call
refusal. Use something from top of the range here (but not 127, as it is
commonly used to report "Command not found")
QubesOS/qubes-issues#2861
This configuration isn't included twice - it's Debian post-installation
script of grub that copy settings to /etc/default/grub, which results in
parameters being duplicated. Leave it as is for now.
Add GRUB_TIMEOUT=0 for faster VM startup.
QubesOS/qubes-issues#2577
Qubes VM require few config options in grub. Ship appropriate
configuration. Debian have grub.d support, so it can be done cleanly.
On Fedora, /etc/default/grub needs to be modified. Still keep the
options in separate file, but include it manually from
/etc/default/grub.
QubesOS/qubes-issues#2577
- there are many netcat versions (openbsd, nmap, ...), which behave
differently - especially while handling EOF
- Debian jessie doesn't have nmap-ncat (which handle EOFs sufficiently
good)
QubesOS/qubes-issues#1854
While it doesn't make sense to install qubes-core-agent without qrexec,
it may make sense to do the otherway around - install just
qrexec-agent without all the qrexec services and configuration. For
example on some pre-installed system.
QubesOS/qubes-issues#2771
It is expected to be killed by a signal. Exit with returncode 0 anyway.
While at it, adjust it for current service format (executable, with
proper shebang).
When a file is moved to other package, the new package needs Replaces:
and Breaks: dependecies on old package. Otherwise dpkg will refuse to
change file ownership.
QubesOS/qubes-issues#2771
This will save a lot of dependencies if networking is not needed in VMs
based on given template. Thanks to updates proxy over qrexec, template
itself do not need to have network configured too.
QubesOS/qubes-issues#2771
In Fedora it makes little sense, but in Debian it allows to avoid a lot
of dependencies. So split in both, to keep it simple.
QubesOS/qubes-issues#2771
This is meant to notify dom0 about features supported by just-installed
template. This service is called by dom0 just after template
installation.
FixesQubesOS/qubes-issues#1637
Documentation pending: QubesOS/qubes-issues#2829
Configure package manager to use 127.0.0.1:8082 as proxy instead of
"magic" IP intercepted later. The listen on this port and whenever
new connection arrives, spawn qubes.UpdatesProxy service call (to
default target domain - subject to configuration in dom0) and connect
its stdin/out to the local TCP connection. This part use systemd.socket
unit in case of systemd, and ncat --exec otherwise.
On the other end - in target domain - simply pass stdin/out to updates
proxy (tinyproxy) running locally.
It's important to _not_ configure the same VM to both be updates proxy and
use it. In practice such configuration makes little sense - if VM can
access network (which is required to run updates proxy), package manager
can use it directly. Even if this network access is through some
VPN/Tor. If a single VM would be configured as both proxy provider and
proxy user, connection would loop back to itself. Because of this, proxy
connection redirection (to qrexec service) is disabled when the same VM
also run updates proxy.
FixesQubesOS/qubes-issues#1854
...but installed on all Debian versions. This is mostly required by
vebose file list in debian/qubes-core-agent.install. But also make it
use new options when upstream will set them.
QubesOS/qubes-issues#2161
This reverts commit 5dfcf06ef4.
python3-daemon isn't widespread enough yet - for Debian jessie available
only in packports.
In addition to the revert itself, adjust packaging for this change
(mostly for Debian).
Add --install-layout=deb option to setup.py, so files will not land in
/usr/local.
Also, explicitly list packaged files - make it easier to split the
package later.
This way:
- VM prompt do know VM list, the list may be filtered based on policy
- source VM don't learn name of target VM
FixesQubesOS/qubes-issues#910
glib-compile-schemas recommend naming override files with nn_ prefix,
where nn is a number. Lets use 20, to allow both higher and lower
priority files.
QubesOS/qubes-issues#1108
When `qubes-dom0-update --refresh` was called, the script checked
metadata twice - once to check updates availability, then to actually
download them. This two stage approach is needed only on Debian, when
--downloadonly option is not supported. Rearrange code accordingly.
Also, drop --doit option (ignore it), as the same (but more readable)
can be achieved with --check-only.
When /etc/yum.conf is not present (yum-deprecated not installed), don't
try to append to it. It would result in invalid configuration file -
without any section header, and break yum when installed later.
This looks like an overkill. And more importantly - required modules are
not installed by default, making it impossible to use in default minimal
template.
FixesQubesOS/qubes-issues#1610
qubes-early-vm-config.service depend on qubes-db daemon running, to
retrieve various configuration parameters. But the startup ordering
declaration was missing
FixesQubesOS/qubes-issues#2750
Qubes-specific options must be added in NetworkManager.conf, even if
those lines are commented out by default. The problem is solved by using
extended regular expressions.
Before, the script skipped none existing ro paths even if the path
existed below /rw. This would require someone who wants to use bind-dirs
only in TemplateBasedVM to ensure that the paths exist before the
bind-dirs script gets called.
This patch changes this behavior so that if the path exists below /rw, it
is ensured that an (empty) file/directory is present in ro (where the
corresponding path from /rw is then bind mounted over).
Requires: Docs update. I can open a PR when this PRs looks good.
Fixes limitation: "Does not work if the file / folder in question does
not already exist in the root image. I.e. a file that does not exist in
the root image cannot be bind mounted in the TemplateBasedVM."
Example use case: https://github.com/debops/ansible-persistent_paths
Tested on: Qubes OS 3.2; Debian 8 TemplateBasedVM (and Template)
Related to: https://github.com/QubesOS/qubes-issues/issues/2661
When trying offline resize2fs, it require running fsck first, which
takes time, especially on large volumes. And in most cases, resize2fs
will notice that no action is needed - after wasting some time on fsck.
To remedy this, use resize2fs in online mode (on mounted filesystem).
And drop fsck call if it fails (filesystem is already mounted
read-write, running fsck isn't good idea).
But do not remove fsck call completely - still call it, but without '-f'
flag, so it run actual check only when really needed (unclean shutdown,
last check far in the past etc).
FixesQubesOS/qubes-issues#979FixesQubesOS/qubes-issues#2583
dnf stdout messages differ from yum. Handle this particular difference
(info about last metadata check time), but in addition properly use its
exit code - 0 means no updates, 100 means some updates.
FixesQubesOS/qubes-issues#2096
* qubesos/pr/25:
Add systemd override for haveged in xenial and stretch. (#2161) Reenable haveged.service after debian package installation
FixesQubesOS/qubes-issues#2161
Use /32 inside network namespace too. Otherwise inter-VM traffic is
broken - as all VMs seems to be in a single /24 subnet, but in fact are
not.
QubesOS/qubes-issues#1143
Core3 no longer reuse netvm own IP for primary DNS. At the same time,
disable dropping traffic to netvm itself because it breaks DNS (as one
of blocked things). This allows VM to learn real netvm IP, but:
- this mechanism is not intended to avoid detection from already
compromised VM, only about unintentional leaks
- this can be prevented using vif-qubes-nat.sh on the netvm itself (so
it will also have hidden its own IP)
QubesOS/qubes-issues#1143
Since 'script' xenstore entry no longer allows passing arguments
(actually this always was a side effect, not intended behaviour), we
need to pass additional parameters some other way. Natural choice for
Qubes-specific script is to use QubesDB.
And since those parameters are passed some other way, it is no longer
necessary to keep it as separate script.
FixesQubesOS/qubes-issues#1143
Keep "main" IP (the one in xenstore) as the one seen by the netvm, and
pass the "fake" one (the one seen by the VM) as script parameter.
FixesQubesOS/qubes-issues#1143
Up until today, Qubes OS would insist on either masking or disabling
or activating units that should get their state properly changed
but only on first package install (when the template is built).
This commit adds the possibility of having two types of unit presets:
* Initial presets: these are only changed state during first package
installs.
* Upgrade presets: these get their state changed during first
package installs as well as during upgrades.
All the maintainer has to do is abide by the instructions in the
preset file. Nothing else is necessary.
Namely, this allows users to enable SSHD on their templates or
standalone VMs and still keep it enabled even after the
qubes-core-vm-systemd package is upgraded.
Matt really wanted that, and so did I, so now we can do it!
:-)
qubes-setup-dnat-to-ns is called multiple times during boot. Of particular interest are the two invocations done by:
1. `/usr/lib/qubes/init/network-proxy.setup.sh` (`qubes-network.service`)
2. `/usr/lib/qubes/init/misc-post.sh` (`qubes-misc-post.service`)
These can, and do often, run in parallel. Often enough that the `PR-QBS` `nat` chain can end up with eight rules instead of four, or (worse) zero rules.
This commit represents the proper boot ordering of these services, where the post startup *must* happen after Qubes has already started its iptables, firewall, network setup and netwatcher.
This eliminates the race.
Do not use a symlink there, as it will be left after NetworkManager
shutdown - as a broken link then
FixesQubesOS/qubes-issues#2320
Reported by Achim Patzner <noses@noses.com>
This rewrite is mainly to adopt new interface for Qubes 4.x.
Main changes:
- change language from bash to python, introduce qubesagent python package
- support both nftables (preferred) and iptables
- new interface (https://qubes-os.org/doc/vm-interface/)
- IPv6 support
- unit tests included
- nftables version support running along with other firewall loaded
FixesQubesOS/qubes-issues#1815QubesOS/qubes-issues#718
f4d367a6 dropped the check if the bind target exists and added
"--no-clobber" to the cp call. For directories this does not work as
desired: cp checks per (recursive) file instead of once for the
specified directory.
The systemctl in Debian unstable fails when trying to disable a removed
service. The manpage do not mention a switch to change this behaviour.
But it says:
Note that this operation creates only the suggested symlinks for
the units. While this command is the recommended way to manipulate
the unit configuration directory, the administrator is free to make
additional changes manually by placing or removing symlinks in the
directory.
So a simple rm should be fine.
- qubes-misc-post.service is no longer responsible for mounting /rw
- both qubes-sysinit.service and qubes-mount-dirs.service are part of
basic.target, so no need to mention them explicitly (as long as
DefaultDependencies=yes)
QubesOS/qubes-issues#2198
It is needed for vchan communication. It was loaded implicitly by mount
/proc/xen, but since we're moving away from this legacy interface, load
it explicitly.
QubesOS/qubes-issues#2194
The service is really responsible for mounting /rw and /home, so should
be ordered before local-fs.target - this will allow other services
to use standard ordering targets.
This probably makes Before=qubes-gui-agent.service not needed anymore,
but do not remove it yet without extensive testing to not risk
regression.
FixesQubesOS/qubes-issues#2194
Even if update check fails for some reason (network problem, apt-get lock
being held etc), don't mark the service as failed. The update check
mechanism is designed this way to not worry about such single failures
- other VM(s) may still check and report updates availability.
FixesQubesOS/qubes-issues#1889
This doesn't help when xen update is installed after this one. So, deal
with it in xen %post itself.
This reverts commit f2257e1e3b.
QubesOS/qubes-issues#2141
* origin/pr/77:
archlinux: fix update-proxy-configs to use pacman.d drop-ins
archlinux: ensure repositories are the last pacman.d files included
archlinux: Setup default package repository
archlinux: switch to usage of pacman.d drop-ins
systemctl preset output lengthy warning when trying to operate on
non-existing unit. This preset action is meant to disable unit, so it's
even better it doesn't exists.
* qubesos/pr/18:
Enable xendriverdomain.service in 75-qubes-vm.preset
Remove 'if true' wrapper from 06a0d30d50
*Do* block until good random is available again
dvm, then xendriverdomain, then qrexec-agent
Network management software should order itself after network-pre.target
(man 7 systemd.special) so that other units can order themselves before
the *beginning* of network initialization. (qubes-misc-post too because
it calls setup-ip.)
Relevant for QubesOS/qubes-issues#2108
Since yum-deprecated is slowly removed from Fedora (in Fedora 23 is not
installed by default), we're forced to migrate to dnf. The main problem
with dnf here is lack of --downloaddir option
(https://bugzilla.redhat.com/show_bug.cgi?id=1279001). As nobody is
going to implement it, simply extract downloaded packages from cache
directory (thanks to provided config file, it is always /var/cache/yum).
This basically replaces "dom0-updates: use yum-deprecated instead of dnf
in all calls" with a set of workarounds for dnf missing parts.
Related to QubesOS/qubes-issues#1574
This have many advantages:
- prevent XSS (QubesOS/qubes-issues#1462)
- use default browser instead of default HTML viewer
- better qrexec policy control
- easier to control where are opened files vs URLs
For now allow only http(s):// and ftp:// addresses (especially prevent
file://). But this list can be easily extended.
QubesOS/qubes-issues#1462FixesQubesOS/qubes-issues#1487
Files in /var/run/qubes-service are created by qubes-sysinit.service. So
defer that condition check after that service start.
Thanks @adrelanos for the report.
FixesQubesOS/qubes-issues#1985
rsvg-convert doesn't scale the image. Do it with convert, only when
really needed. Don't upscale the icon after converting to raster
version.
FixesQubesOS/qubes-issues#1884
Always set QREXEC_AGENT_PID variable, setup SIGUSR1 handler. And do that
before starting child process to avoid race conditions.
Required for QubesOS/qubes-issues#
FixesQubesOS/qubes-issues#1863
* qubesos/pr/11:
Properly handle case of empty domain name.
Use proper quoting around variables.
Move usage information printing to separate function, and print usage to stderr; also added some spacing.
Use proper space-expanded tabs, as per the coding guidelines.
* origin/pr/58:
refactoring / code simplification
fixed broken file copy for files in multi level directories
also exit from bind-directories if file /var/run/qubes-service/qubes-dvm exists
use symlink_level_max rather than hardcoding 10; comment
run /usr/lib/qubes/bind-dirs.sh from mount-dirs.sh
renamed: bind-dirs -> bind-dirs.sh
renamed: misc/bind-dirs -> vm-systemd/bind-dirs
work on bind-dirs
work on bind-dirs
work on bind-dirs https://phabricator.whonix.org/T414
Those are currently barely useful, since starting the application isn't
the biggest bottleneck of DispVM. And since stderr is now visible on
qvm-open-in-dvm output, not scare the user with it.
In case of some services it makes much sense for caller to receive also
stderr in addition to stdout. For example:
- qubes.VMShell (stderr required for salt-ssh over qrexec)
- qubes.OpenInVM - especially when called to DispVM - otherwise
diagnosing errors can be hard
And generally all sort of error reporting (the purpose of stderr). It
would ease debugging - instead of message "error occurred, check here and
there for more details", it could be "error occurred: the reason".
FixesQubesOS/qubes-issues#1808
sed 's:^#!/usr/bin/python.*:#!/usr/bin/python2:' -i misc/*
sed 's:^#!/usr/bin/env python.*:#!/usr/bin/env python2:' -i misc/*
sed 's:^#!/usr/bin/python.*:#!/usr/bin/python2:' -i qubes-rpc/*
sed 's:^#!/usr/bin/env python.*:#!/usr/bin/env python2:' -i qubes-rpc/*
# Force running all scripts with python2
sed 's:#!/usr/bin/python:#!/usr/bin/python2:' -i misc/*
sed 's:#!/usr/bin/env python:#!/usr/bin/env python2:' -i misc/*
sed 's:#!/usr/bin/python:#!/usr/bin/python2:' -i qubes-rpc/*
sed 's:#!/usr/bin/env python:#!/usr/bin/env python2:' -i qubes-rpc/*
# Fix for archlinux sbindir
sed 's:/usr/sbin/ntpdate:/usr/bin/ntpdate:g' -i qubes-rpc/sync-ntp-clock
sed 's:/usr/sbin/qubes-netwatcher:/usr/bin/qubes-netwatcher:g' -i vm-systemd/qubes-netwatcher.service
sed 's:/usr/sbin/qubes-firewall:/usr/bin/qubes-firewall:g' -i vm-systemd/qubes-firewall.service
for dir in qubes-rpc qrexec misc;do
(cd$dir; make)
done
# Fix for archlinux sbindir
sed 's:/usr/sbin/ntpdate:/usr/bin/ntpdate:g' -i qubes-rpc/sync-ntp-clock
sed 's:/usr/sbin/qubes-firewall:/usr/bin/qubes-firewall:g' -i vm-systemd/qubes-firewall.service
for dir in qubes-rpc qrexec misc;do
make -C "$dir"
done
}
package(){
# Note: Archlinux removed use of directory such as /sbin /bin /usr/sbin (https://mailman.archlinux.org/pipermail/arch-dev-public/2012-March/022625.html)
(cd qrexec; make install DESTDIR=$pkgdirSBINDIR=/usr/bin LIBDIR=/usr/lib SYSLIBDIR=/usr/lib)
# Note: Archlinux removed use of directory such as /sbin /bin /usr/sbin (https://mailman.archlinux.org/pipermail/arch-dev-public/2012-March/022625.html)
# shellcheck disable=SC2154
make -C qrexec install DESTDIR="$pkgdir"SBINDIR=/usr/bin LIBDIR=/usr/lib SYSLIBDIR=/usr/lib
make install-vm DESTDIR=$pkgdirSBINDIR=/usr/bin LIBDIR=/usr/lib SYSLIBDIR=/usr/lib DROPIN_DIR=usr/lib/systemd DIST=archlinux
PYTHON=python2 make install-corevm DESTDIR="$pkgdir"SBINDIR=/usr/bin LIBDIR=/usr/lib SYSLIBDIR=/usr/lib SYSTEM_DROPIN_DIR=/usr/lib/systemd/system USER_DROPIN_DIR=/usr/lib/systemd/userDIST=archlinux
# Remove things non wanted in archlinux
rm -r $pkgdir/etc/yum*
rm -r $pkgdir/etc/init.d
# Remove fedora specific scripts
rm $pkgdir/etc/fstab
# Remove things non wanted in archlinux
rm -r "$pkgdir/etc/yum"*
rm -r "$pkgdir/etc/dnf"*
rm -r "$pkgdir/etc/init.d"
# Remove fedora specific scripts
rm "$pkgdir/etc/fstab"
# Install systemd script allowing to automount /lib/modules
# Archlinux bash version has a 'bug' when running su -c, /etc/profile is not loaded because bash consider there is no interactive pty when running 'su - user -c' or something like this.
# See https://bugs.archlinux.org/task/31831
id -u 'user' >/dev/null 2>&1 || {
useradd --user-group --create-home --shell /bin/zsh user
}
usermod -a --groups qubes user
# Archlinux bash version has a 'bug' when running su -c, /etc/profile is not loaded because bash consider there is no interactive pty when running 'su - user -c' or something like this.
# See https://bugs.archlinux.org/task/31831
id -u 'user' >/dev/null 2>&1 || {
useradd --user-group --create-home --shell /bin/bash user
}
usermod -a --groups qubes user
}
## arg 1: the new package version
pre_install() {
echo "Pre install..."
echo "Pre install..."
update_default_user
update_default_user
# do this whole %pre thing only when updating for the first time...
# do this whole %pre thing only when updating for the first time...
mkdir -p /var/lib/qubes
mkdir -p /var/lib/qubes
# Backup fstab / But use archlinux defaults (cp instead of mv)
if [ -e /etc/fstab ] ; then
cp /etc/fstab /var/lib/qubes/fstab.orig
fi
# Backup fstab / But use archlinux defaults (cp instead of mv)
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.