VirtualBox user space binaries now no longer reside in linuxPackages, so
let's use the package for the real user space binaries instead.
Tested using the following command:
nix-build nixos/release.nix -A ova.x86_64-linux
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Switches from the forking service type to simple by running haveged in
the foreground. Also restricts the execution environment a bit (these
are inspired by the Debian service file).
Putting the kernel modules into the same output path as the main
VirtualBox derivation causes all of VirtualBox to be rebuilt on every
single kernel update.
The build process of VirtualBox already outputs the kernel module source
along with the generated files for the configuration of the main
VirtualBox package. We put this into a different output called "modsrc"
which we re-use from linuxPackages.virtualbox, which is now only
containing the resulting kernel modules without the main user space
implementation.
This not only has the advantage of decluttering the Nix expression for
the user space portions but also gets rid of the need to nuke references
and the need to patch out "depmod -a".
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
See #18319 for details. Starting network-online.target manually does not
work as it hangs indefinitely.
Additionally, don't treat avahi and dhcpcd special and sync their systemd units
with the respective upstream suggestion.
The extraOptions option has default values which seems surprising. This
moves those values to startupOptions (which is what gocd-agent uses) and
empties out the default extraOptions.
The gocd-agent startupOptions description was also changed to remove the
mention of the example (given there isn't one).
Systemd upstream provides targets for networking. This also includes a target network-online.target.
In this PR I remove / replace most occurrences since some of them were even wrong and could delay startup.
Radicale can run as a foreground service and will then emits logging and
errors on the standard output. This helps the logging end up in the
systemd journal.
This partially reverts commit ab9537ca22.
From the manpage of systemd-nspawn(1):
Note that systemd-nspawn will mount file systems private to the
container to /dev, /run and similar.
Testing this in a shell turns out:
$ sudo systemd-nspawn --bind-ro=/nix/store "$(readlink "$(which ls)")" /proc
Spawning container aszlig on /home/aszlig.
Press ^] three times within 1s to kill container.
/etc/localtime does not point into /usr/share/zoneinfo/, not updating
container timezone.
1 execdomains kpageflags stat
acpi fb loadavg swaps
asound filesystems locks sys
buddyinfo fs meminfo sysrq-trigger
bus interrupts misc sysvipc
cgroups iomem modules thread-self
cmdline ioports mounts timer_list
config.gz irq mtrr timer_stats
consoles kallsyms net tty
cpuinfo kcore pagetypeinfo uptime
crypto key-users partitions version
devices keys scsi vmallocinfo
diskstats kmsg self vmstat
dma kpagecgroup slabinfo zoneinfo
driver kpagecount softirqs
Container aszlig exited successfully.
So the test on whether PID 1 exists in /proc is enough, because if we
use PID namespaces there actually _is_ a PID 1 (as shown above) and the
special file systems are already mounted. A test on the $containers
variable actually mounts them twice.
This unbreaks NixOS containers and I've tested this against the
containers-imperative NixOS test.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Cc: @rickynils, @shlevy, @edolstra
Follow-up to the following commits:
abdc5961c3cdf9f5893ea1e91ba08ff5089f53a4: Fix starting the firewall
e090701e2d09aec3e8866ab9a8e53c37973ffeb4: Order before sysinit
Solely use sysinit.target here instead of multi-user.target because we
want to make sure that the iptables rules are applied *before* any
socket units are started.
The reason I've dropped the wantedBy on multi-user.target is that
sysinit.target is already a part of the dependency chain of
multi-user.target.
To make sure that this holds true, I've added a small test case to
ensure that during switch of the configuration the firewall.service is
considered as well.
Tested using the firewall NixOS test.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Cc: @edolstra
Probably as a result of 992c514a20, it
was not being started anymore.
My understanding of systemd.special(7) (section "Special passive
system units") is that the firewall should want network-pre.target,
rather than the other way around (not very intuitive...). This in
itself does not cause the firewall to be wanted, which is why the
wanted-by relationship with multi-user.target is necessary.
http://hydra.nixos.org/build/39965589
This partially reverts commit 0aa7520670.
Fine for rsync to be in system path but we still need the explicit path
in nixos-install in case it is invoked from non-NixOS systems and also
to fix OVA test failure
See also 0aa7520670
cc @edolstra
We were pulling in 44 MiB of fonts in the default configuration, which
is a bit excessive for headless configurations like EC2
instances. Note that dejavu_minimal ensures that remote X11-forwarded
applications still have a basic font regardless.
It appears that packageOverrides no longer overrides aliases, so
aliases like
dbus_tools = self.dbus.out;
dbus_daemon = self.dbus.daemon;
now use the old, non-overriden version of dbus. That seems like a
pretty serious regression in general, but for this particular problem,
I've fixed it by replacing dbus_daemon by dbus.daemon and dbus_tools
by dbus.
This reverts commit 582313bafe.
Removing rsync is actually pointless because nixos-install depends on
it. So if it's part of the system closure, we may as well provide it
to users.
Probably with the next Nix release we can drop the use of rsync and
use "nix copy" instead.
This one was already merged into release-16.09, so let's not have the
stable branch is ahead of master and confuse things. In addition to
that, currently we have an odd situation that master has less things
actually finished building than in staging.
Conflicts:
pkgs/data/documentation/man-pages/default.nix
Fixes#15512 and #16032
With the multi output, postgresql cannot find at runtime what is its
basedir when looking for libdir and pkglibdir. This commit fixes that.
Before this commit updating /var/setuid-wrappers/ folder introduced
a small window where NixOS activation scripts could be terminated
and resulted into empty /var/setuid-wrappers/ folder.
That's very unfortunate because one might lose sudo binary.
Instead we use two atomic operations mv and ln (as described in
https://axialcorps.com/2013/07/03/atomically-replacing-files-and-directories/)
to achieve atomicity.
Since /var/setuid-wrappers is not a directory anymore, tmpfs mountpoints
were removed in installation scripts and in boot process.
Tested:
- upgrade /var/setuid-wrappers/ from folder to a symlink
- make sure /run/setuid-wrappers-dirs/ legacy symlink is really deleted
Add code to accountsservice that returns an error if the environment
variable NIXOS_USERS_PURE is set. This variable is set from the nixos
accountsservice module if mutableUsers = false
Both btrfs-progs and utillinux are ~5MB, we may discuss in future
to handle this better but I see no better way at the moment than
increaing purity in the install process.
OnCalendar entrys can be specified multiple times in a systemd timer, to
make more complex scheduling possible.
Tested by manually checking the timer generated by the following:
systemd = {
services.huhu = {
description = "meh";
wantedBy = [ "default.target" ];
serviceConfig.ExecStart = "/bin/sh -c 'printf HUHU!'";
startAt = [ "*:*:0/30" "*:0/1:15" ];
};
};
It prints HUHU to the log at seconds 0, 15 and 30 of each minute.
This was one of the ways to build packages, we are trying
hard to minimize different ways so it's easier for newcomers
to learn only one way.
This also:
- removes texLive (old), fixes#14807
- removed upstream-updater, if that code is still used it should be in
separate repo
- changes a few packages like gitit/mit-scheme to use new texlive
Verified that following nixos configuration:
users.users.foo = {
uid = 1000;
name = "foo";
};
users.users.bar = {
name = "bar";
};
Before this commit both users will get uid of 1000, after it's applied
bar will correctly get 1001.
Previously it was only set once, now it is enforced on each start-up of
redis.service. Also set _ownership_ recursively, so that the
/var/lib/redis/dump.rdb file is guaranteed to be accessible by the
currently configured redis user.
Fixes issue #9687, where redis wouldn't start because /var/lib/redis had
wrong owner.
A new internal config option `fileSystems.<name>.early` is added to indicate
that the filesystem needs to be loaded very early (i.e. in initrd). They are
transformed to a shell script in `system.build.earlyMountScript` with calls to
an undefined `specialMount` function, which is expected to be caller-specific.
This option is used by stage-1, stage-2 and activation script to set up and
remount those filesystems. Options for them are updated according to systemd
defaults.
Fixes#13927
cc @edolstra
configFile in make-disk-image clashes with clone-config as the latter does
nothing if it finds a a /etc/nixos/configuration.nix during stage-2.
lustrate /ˈlʌstreɪt/ verb.
purify by expiatory sacrifice, ceremonial washing, or some other
ritual action.
- sudo touch /etc/NIXOS_LUSTRATE
⇒ on next reboot, during stage 1, everything but /nix and /boot
is moved to /old-root
- echo "etc/passwd" | sudo tee -a /etc/NIXOS_LUSTRATE
⇒ on next reboot, during stage 1, everything but /nix and /boot
is moved to /old-root; except /etc/passwd is copied back.
Useful for installing NixOS in place on another distro. For instance:
$ nix-env -iE '_: with import <nixpkgs/nixos> { configuration = {}; }; with config.system.build; [ nixos-generate-config manual.manpages ]'
$ sudo mkdir /etc/nixos
$ sudo `which nixos-generate-config`
… edit the configuration files in /etc/nixos using man configuration.nix
if needed
maybe add: users.extraUsers.root.initialHashedPassword = "" ?
… Build the entire NixOS system and link it to the system profile:
$ nix-env -p /nix/var/nix/profiles/system -f '<nixpkgs/nixos>' -A system --set
… If you were using a single user install:
$ sudo chown -R 0.0 /nix
… NixOS is about to take over
$ sudo touch /etc/NIXOS
$ sudo touch /etc/NIXOS_LUSTRATE
… Let's keep the configuration files we just created
$ echo etc/nixos | sudo tee -a /etc/NIXOS_LUSTRATE
$ sudo mv -v /boot /boot.bak &&
sudo /nix/var/nix/profiles/system/bin/switch-to-configuration boot
$ sudo reboot
… NixOS boots, Stage 1 moves all the old distro stuff in /old-root.
Closes#16730. Closes#17770. Closes#17846.
Test plan:
* Check that `fonts.fontconfig.ultimate.preset` changes things;
* Check that `fonts.fontconfig.dpi` changes things;
* Check that `fonts.fontconfig.defaultFonts.monospace` changes things;
Tested with AbiWord, mousepad and Firefox.
The fontconfig-ultimate patches are unmaintained. Since they were
not updated for newer FreeType versions, this removes them and
disables fontconfig-ultimate by default.
This removes our hardcoded presets which weren't updated for quite some time.
Infinality now has new hardcoded presets in freetype, which can be overriden if
desired with environment variables (as before). Accordingly, updated NixOS
module to set the hardcoded preset.
Additionally used a more "right" type for substitutions.
Before commit 54fa0cfe4e, the `redshift`
service was run with the environment variable `DISPLAY` set to `:0`.
Commit 54fa0cfe4e changed this to
instead use the value of the `services.xserver.display` configuration
option in the value of the `DISPLAY` variable. In so doing, no default
value was provided for the case where `services.xserver.display` is
`null`.
While the default value of `services.xserver.display` is `0`, use of
which by the `redshift` module would result in `DISPLAY` again being
set to `:0`, `services.xserver.display` may also be `null`, to which
value it is set by, e.g., the `lightdm` module.
In the case that `services.xserver.display` is `null`, with the change
made in commit 54fa0cfe4e, the `DISPLAY`
variable in the environment of the `redshift` service would be set to
`:` (a single colon), which, according to my personal experience,
would result in —
- the `redshift` service failing to start; and
- systemd repeatedly attempting to restart the `redshift` service,
looping indefinitely, while the hapless `redshift` spews error
messages into the journal.
It can be observed that the malformed value of `DISPLAY` is likely at
fault for this issue by executing the following commands in an
ordinary shell, with a suitable `redshift` executable, and the X11
display not already tinted:
- `redshift -O 2500` — This command should reduce the color
temperature of the display (making it more reddish).
- `DISPLAY=':' redshift -O 6500` — This command should raise the
color temperature back up, were it not for the `DISPLAY`
environment variable being set to `:` for it, which should cause
it to, instead, fail with several error messages.
This commit attempts to fix this issue by having the `DISPLAY`
environment variable for the `redshift` service default to its old
value of `:0` in the case that `services.xserver.display` is `null`.
I have tested this solution on NixOS, albeit without the benefit of a
system with multiple displays.
gdnc is a user process and can't be made into a NixOS module very
easily. It can still be put in the user's login script. According to the
GNUstep documentation it will be started as soon as it is needed.
- Replace hand-rolled version of nixos-install in make-disk-image by an
actual call to nixos-install
- Required a few cleanups of nixos-install
- nixos-install invokes an activation script which the hand-rolled version
in make-disk-image did not do. We remove /etc/machine-id as that's
a host-specific, impure, output of the activation script
Testing:
nix-build '<nixpkgs/nixos/release.nix>' -A tests.installer.simple passes
Also tried generating an image with:
nix-build -E 'let
pkgs = import <nixpkgs> {};
lib = pkgs.lib;
nixos = import <nixpkgs/nixos> {
configuration = {
fileSystems."/".device = "/dev/disk/by-label/nixos";
boot.loader.grub.devices = [ "/dev/sda" ];
boot.loader.grub.extraEntries = '"''"'
menuentry "Ubuntu" {
insmod ext2
search --set=root --label ubuntu
configfile /boot/grub/grub.cfg
}
'"''"';
};
};
in import <nixpkgs/nixos/lib/make-disk-image.nix> {
inherit pkgs lib;
config = nixos.config;
diskSize = 2000;
partitioned = false;
installBootLoader = false;
}'
Then installed the image:
$ sudo df if=./result/nixos.img of=/dev/sdaX bs=1M
$ sudo resize2fs /dev/disk/by-label/nixos
$ sudo mount /dev/disk/by-label/nixos /mnt
$ sudo mount --rbind /proc /mnt/proc
$ sudo mount --rbind /dev /mnt/dev
$ sudo chroot /mnt /nix/var/nix/profiles/system/bin/switch-to-configuration boot
[ … optionally do something about passwords … ]
and successfully rebooted to that image.
Was doing all this from inside a Ubuntu VM with a single user nix install.
- Fix --no-bootloader which didn't do what it advertised
- Hardcode nixbld GID so that systems which do not have a nixbld user
can still run nixos-install (only with --closure since they can't
build anything)
- Cleanup: get rid of NIX_CONF_DIR(=/tmp)/nix.conf and pass arguments instead
- Cleanup: don't assume that the target system has '<nixpkgs/nixos>' or
'<nixos-config>' to see if config.users.mutableUsers. Instead check if
/var/setuid-wrappers/passwd is there
Installing NixOS now works from a Ubuntu host (using --closure).
nix-build -A tests.installer.simple '<nixpkgs/nixos/release.nix>' succeeds ✓
While useless, some builds may dabble with setuid bits (e.g.,
util-linux), which breaks under grsec. In the interest of user
friendliness, we once again compromise by disabling an otherwise useful
feature ...
Closes https://github.com/NixOS/nixpkgs/issues/17501
It uses import-from-derivation, which is a bad thing, because this
causes hydra-evaluator to build Cassandra at evaluation time.
$ nix-instantiate nixos/release.nix -A tests.cassandra.i686-linux --dry-run
error: cannot read ‘/nix/store/c41blyjz6pfvk9fnvrn6miihq5w3j0l4-cassandra-2.0.16/conf/cassandra-env.sh’, since path ‘/nix/store/0j9ax4z8xhaz5lhrwl3bwj10waxs3hgy-cassandra-2.0.16.drv’ is not valid, at /home/eelco/Dev/nixpkgs/nixos/modules/services/databases/cassandra.nix:373:11
Also, the module is a mess (bad option descriptions, poor indentation,
a gazillion options where a generic "config" option would suffice, it
opens ports in the firewall, it sets vm.swappiness, ...).
- Agent now takes a full URL to the Go.CD server
- Instruct the agent to attempt restart every 30s upon failure
- Test's Accept header did not match the server's expectation
- Replace the tests' complex Awk matches with calls to `jq`
Update gocd-agent package version to 16.6.0-3590 including new sha. Modify heapSize
and maxMemory mkOption to accurately reflect their intended purpose of configuring
initial java heap sizes.
Addresses #17218 in a better way in that it doesn't create a
".git-revision" file on every nixos-rebuild, because we already have
".git" available. Even if we don't nixos-rebuild can't create the
"git-revision" file.
Tested via:
nix-build -E '(import ./nixos/tests/make-test.nix {
name = "foo";
machine = {};
testScript = "startAll; $machine->execute(\"nixos-version >&2\");";
})'
Closes: #17610
Acked-by: @bennofs
Let's first try if we can determine the Git revision from the .git
directory and if that fails, fall back to get the info from the
".git-revision" file... and after that use something generic like
"master".
This should address #17218 in better way, because we don't need to
create another redundant file in the source checkout of nixpkgs.
I'm not going to route of falling back to using .git, because after
55d881e, we already have ".git-revision" files in people's Git
repositories, which in turn means that nixos-version will report that
old file every time even if the working tree has updated.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Cc: @bennofs, Profpatsch
Reported-by: @devhell
Fixes: #17218
This reverts commit 1e534e234b.
We already should have a .git directory if it is managed via Git,
otherwise there is no way to get the Git revision if neither
.git-revision or .git is present.
But having .git-revision _and_ .git present seems very much redundant to
me.
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
Cc: @bennofs, @Profpatsch
Issue: #17218
The builder has this convoluted `while` loop which just replicates
`readlink -e`. I'm sure there was a reason at one point, because the
loop has been there since time immemorial. It kept getting copied
around, I suspect because nobody bothered to understand what it actually
did.
Incidentally, this fixes#17513, but I have no idea why.
From @fpletz: Keep poolConfigs option for backwards-compatibility.
The original commit 6b3f5b5a42 was previously
reverted by c7860cae1a but the issues were
resolved.
Closes#17460
Changed the wrapper derivation to produce a second output containing the sandbox.
Add a launch wrapper to try and locate the sandbox (either in /var/setuid-wrappers or in /nix/store).
This launch wrapper also sheds libredirect.so from LD_PRELOAD as Chromium does not tolerate it.
Does not trigger a Chromium rebuild.
cc @cleverca22 @joachifm @jasom
The varnish tools (varnishstat, varnishlog, ...) tried to load the VSM
file from a spurious var directory in the Nix store. Fix the default so
the tools "just work" when also keeping services.varnish.stateDir at the
default.
Notes:
- The tools use $localstatedir/$HOSTNAME so I've adapted the default for
stateDir as well to contain hostName.
- Added postStop action to remove the localstatedir. There is no point
in keeping it around when varnish does not run, as it regenerates it
on startup anyway.
Fixes#7495
The name gitlab-runner clashes with a component of Gitlab CI with the
same name and only confuses people. It's now called gitlab-bundle and
a convenience-script gitlab-rake for easier invocation of rake tasks
was added. This was the primary use case of gitlab-runner.
The module will configure a Cassandra server with common options being
tweakable. Included is also a test which will spin up 3 nodes and
verify that the cluster can be formed, broken, and repaired.
In light of Emacs packaging improvements such as those mentioned
in #11503, and with the addition of a systemd service (#15807
and #16356), and considering that the wiki page is completely
out of date (#13217), it seems that some documentation is in order.
Enabling EFI runtime services provides a venue for injecting code into
the kernel.
When grsecurity is enabled, we close this by default by disabling access
to EFI runtime services. The upshot of this is that
/sys/firmware/efi/efivars will be unavailable by default (and attempts
to mount it will fail).
This is not strictly a grsecurity related option, it could be made into
a general option, but it seems to be of particular interest to
grsecurity users (for non-grsecurity users, there are other, more
immediate kernel injection attack dangers to contend with anyway).
Adds a new chain in the raw table for reverse path filtering and optional
logging. A rule to allow serving DHCPv4 was also added as it is commonly
needed and poses no security risk even when no DHCPv4 server is running.
Fixes#10101.
Adds options for tcp streaming and avahi zeroconf support (so that the
server can be easily found by clients).
There is also an option to allow anonymous clients to stream to the
server (by default pulseaudio uses a cookie mechanism, see manpage).
Instead of one package `extra-cmake-modules`, there is now `ecm` and
`ecmNoHooks`. The latter is used when one does not want to incur a Qt 5
dependency; it is also available as a top-level package
`extra-cmake-modules`.
With these changes, a container can have more then one veth-pair. This allows for example to have LAN and DMZ as bridges on the host and add dedicated containers for proxies, ipv4-firewall and ipv6-firewall. Or to have a bridge for normal WAN, one bridge for administration and one bridge for customer-internal communication. So that web-server containers can be reached from outside per http, from the management via ssh and can talk to their database via the customer network.
The scripts to set up the containers are now rendered several times instead of just one template. The scripts now contain per-container code to configure the extra veth interfaces. The default template without support for extra-veths is still rendered for the imperative containers.
Also a test is there to see if extra veths can be placed into host-bridges or can be reached via routing.
This makes the container a bit more secure, by preventing root
creating device nodes to access the host file system, for
instance. (Reference: systemd-nspawn@.service in systemd.)
Revert httpConfig its old behaviour and make it mutually exclusive to
the new structured configuration. Adds appendHttpConfig to have the
ability to write custom config in the generated http block.
