(The first version of this change, in commit 39fad297fd, broke
`nix-build -A nixosTests.installer.simpleUefiSystemdBoot`. This is the
2nd version, which hopefully does not break anything.)
`nixos-rebuild build-vm-with-bootloader` currently fails with the
default NixOS EFI configuration:
$ cat >configuration.nix <<EOF
{
fileSystems."/".device = "/dev/sda1";
boot.loader.systemd-boot.enable = true;
boot.loader.efi.canTouchEfiVariables = true;
}
EOF
$ nixos-rebuild build-vm-with-bootloader -I nixos-config=$PWD/configuration.nix -I nixpkgs=https://github.com/NixOS/nixpkgs/archive/nixos-20.09.tar.gz
[...]
insmod: ERROR: could not insert module /nix/store/1ibmgfr13r8b6xyn4f0wj115819f359c-linux-5.4.83/lib/modules/5.4.83/kernel/fs/efivarfs/efivarfs.ko.xz: No such device
mount: /sys/firmware/efi/efivars: mount point does not exist.
[ 1.908328] reboot: Power down
builder for '/nix/store/dx2ycclyknvibrskwmii42sgyalagjxa-nixos-boot-disk.drv' failed with exit code 32
[...]
Fix it by setting virtualisation.useEFIBoot = true when needed.
Before:
* release-20.03: successful build, unsuccessful run
* release-20.09 (and master): unsuccessful build
After:
* Successful build and run.
Fixes#107255
This corresponds to agetty's --login-options argument.
With this change, I can set
services.getty.autologinUser = "qyliss";
services.getty.loginOptions = "-- \\u";
and have my username prefilled, but with my password still
required (unlike the normal autologinUser behaviour).
Ports an OpenWRT patch for Atheros wireless drivers (ath*) which allows
the user to change the regulatory domain code to the one which actually
applies.
All Atheros devices have a regulatory domain burned into their EEPROM.
When using a device as AP, this domain is frequently overly restrictive
when compared to the regulation which applies in the country the device
actually operates in; often, this restriction disallows IR on all
channels making it impossible to use the device as an AP at all.
This commit introduces the NixOS config option
networking.wireless.athUserRegulatoryDomain which, if enabled, applies
the patch and sets the kernel config option ATH_USER_REGD.
The original OpenWRT patch targets Linux 5.8.
b478e0043c removed every other instance of uwsgi.service and
httpd.service from Mailman before and requiredBy lists. This one
looks like it was just missed, since I can't see a reason to keep it
but remove the others.
Fixes: b478e0043c ("nixos/mailman: refactor")
* It should be made explicit in the eval-error that the CVE only affects
a component which is turned off by default.
* For more clarity, the default version used by the module is noted in
the manual.
Closes#108419
It's been 8.5 years since NixOS used mingetty, but the option was
never renamed (despite the file definining the module being renamed in
9f5051b76c ("Rename mingetty module to agetty")).
I've chosen to rename it to services.getty here, rather than
services.agetty, because getty is implemantation-neutral and also the
name of the unit that is generated.
By default in Nginx, if you want to override a single fastcgi_param,
you have to override all of them. This is less of a big deal if
you're editing the Nginx configuration directly, but when you're
generating the Nginx configuration with Nix it can be very annoying to
bloat your configuration repeating the default values of FastCGI
parameters every time.
This patch adds a fastcgiParams option to Nginx locations. If any
parameters are set through this, all the default values will be
included as well, so only the ones that are changing need to be
supplied. There's no way to use fastcgiParams to actually override
all parameters if that's what you want, but I think that's a niche use
case and it's still possible using extraConfig, which up until now was
the only option
Nginx allows the fastcgi_param directive in http and server scopes as
well as location, but here I only support location. It would be
possible to support the others, but I don't think it's worth it. It
would be a possible future enhancement if somebody has a need for it.
This plugin is used commonly enough that we should ensure it still
builds (and dovecot works) after loading it.
This is not yet perfect as we aren't testing any of it's functionality
but at least we ensure that dovecot continues to do the regular job.
Enabling the profile can lead to hard-to-debug issues, which should be
warned about in addition to the cost in features and performance.
See https://github.com/NixOS/nixpkgs/issues/108262 for an example.
androidenv did not previously write license files, which caused certain
gradle-based Android tools to fail. Restructure androidenv's list of
Android packages into a single repo.json file to prevent duplication
and enable us to extract the EULA texts, which we then hash with
builtins.hashString to produce the license files that Android gradle
tools look for.
Remove includeDocs and lldbVersions, as these have been removed
from the Android package repositories.
Improve documentation and examples.
* Content of `programlisting` shouldn't be indented, otherwise it's
weirdly indented in the output.
* Use `<xref linkend=.../>` in the release notes: then users can
directly go to the option documentation when reading release notes.
* Don't use docbook tags in `mkRemovedOptionModule`: it's only used
during evaluation where docbook isn't rendered.
Previously, after the version bump to v247, we broke journalctl --grep
as libpcre2 was lazily loaded during runtime using dlopen(3). This
ensures that we have a test case that alerts us when it fails again.
As per the in-line comment, this is where distros should configure it.
Not via kernel command line parameters.
As found by looking at the implementation, while exploring the cause of
a bug on the Raspberry Pi 4, it was found that `cma=` on the command
line parameters will overwrite the values a device tree will have
configured for a given platform.
With this, the more recent 5.4 vendor kernel boots just fine on the
Raspberry Pi 4 using our common configuration.
This includes setting up everything for the mainline Raspberry Pi 4
image.
In fact, the only difference left in the Raspberry Pi 4-specific image
is the kernel from the vendor.
This reverts commit f19b7b03a0, reversing
changes made to 572a864d02.
Sorry. I pushed the wrong staging-next (the one that had my master
merged in). This was not intended.
Services that have dynamic users require nscd to resolve users
via pam_systemd. Those services might not even create
their own dynamic users itself i.e. iptables.
To make sure nscd is always started when this is happening we move
nscd to sysinit.target and make sure that it is always started before
starting/reloading/restarting any other service.
There are two use case for this flag:
1. NixOS developer usually use a nixpkgs checkout for development.
Copying nixpkgs everytime when rebuilding NixOS is way to slow, even
with NVME disks.
2. Folks migrating from impure configuration in a sufficient complex
infrastructure need this flag to gradually migrate to NixOS flakes.
- Added an ExecPostStart to acme-$cert.service when webroot is defined to create the acme-challenge
directory and fix required permissions. Lego always tries to create .well-known and acme-challenge,
thus if any permissions in that tree are wrong it will crash and break cert renewal.
- acme-fixperms now configured with acme User and Group, however the script still runs as root. This
ensures the StateDirectories are owned by the acme user.
- Switched to list syntax for systemd options where multiple values are specified.
Instead of treating the sddm config a wall of text that doesn't allow us
to override anything, turn it into an attribute set.
We dump `extraConfig` and instead introduce `settings` that is merged
with the module defaults to provide the final configuration.
There is some additional noise in here due to nixpkgs-fmt.
* nixos/xmonad: xmonad config w/ghc+xmessage
When the "config" option isn't set, we use xmonad-with-packages to
provide xmonad with runtime access to an isolated ghc, ensuring it can
recompile and exec a user's local config (e.g. $HOME/.xmonad/xmonad.hs)
regardless of which ghc (if any) is on PATH.
When the "config" option is set, however, we compile a configured xmonad
executable upfront (during nixos-rebuild), and prior to this commit, it
was not provided with runtime access to an isolated ghc.
As a result, with the "config" option set, it was not possible
to recompile and exec a user's local config unless there was a
compatible version of ghc on PATH with the necessary packages (xmonad,
xmonad-contrib, etc.) in its package database. Adding such a ghc to
environment.systemPackages, e.g.
(haskellPackages.ghcWithPackages (ps: with ps; [xmonad xmonad-contrib]))
is problematic because it adds both ghc and an unconfigured xmonad to
PATH, e.g.
$ ls -l $(which xmonad ghc)
lrwxrwxrwx ... /run/current-system/sw/bin/ghc -> /nix/store/...-ghc-8.10.2-with-packages/bin/ghc
lrwxrwxrwx ... /run/current-system/sw/bin/xmonad -> /nix/store/...-ghc-8.10.2-with-packages/bin/xmonad
Having the unconfigured xmonad on PATH is particularly bad because
restarting xmonad will dump the user into the unconfigured version, and
if no local config exists (e.g. in $HOME/.xmonad/xmonad.hs), they'll be
left in this unconfigured state.
In this commmit, we give the configured xmonad runtime access to ghc
like xmonad-with-packages does for the unconfigured version. The aim
is to allow the user to switch between the nixos module's config and a
local config (e.g. $HOME/.xmonad/xmonad.hs) at will, so they can try out
config changes without performing a nixos-rebuild.
Since the xmonad on PATH is the configured executable, there's no
danger a user could unwittingly restart into the unconfigured version,
and because xmonad will refuse to recompile when no local config
exists, there's no danger a user could unwittingly recompile into an
unconfigured version.
Given that a local config exists, the recompile/restart behavior depends
on two factors:
- which entry point is used
* 'XMonad.xmonad' (default)
* 'XMonad.launch' (recommended in "config" option description)
- what operation is triggered (i.e. via mod+q)
* `spawn "xmonad --recompile && xmonad --restart"` (default)
* `restart "xmonad" True`
* custom function
If the default 'XMonad.xmonad' entrypoint and default mod+q operation
are used, hitting mod+q will compile and exec the local config, which
will remain in use until next time the display manager is restarted.
If the entrypoint is changed to 'XMonad.launch' but mod+q left with its
default operation, hitting mod+q will have no visible effect. The logs
(as seen by running `journalctl --identifier xmonad --follow`) will show
an error,
X Error of failed request: BadAccess (attempt to access private resource denied)
which indicates that the shell was unable to start xmonad because
another window manager is already running (namely, the nixos-configured
xmonad).
https://wiki.haskell.org/Xmonad/Frequently_asked_questions#X_Error_of_failed_request:_BadAccess_.28attempt_to_access_private_resource_denied.29
Changing the mod+q operation to `restart "xmonad" True` (as recommended
in the "config" option's description) will allow a restart of the
nixos-configured xmonad to be triggeredy by hitting mod+q.
Finally, if the entrypoint is 'XMonad.launch', mod+q has been
bound to `restart "xmonad" True` and another key bound to a custom
recompile/restart function (e.g. `compileRestart` as shown in the
"config" option example), the user can switch between the nixos module's
config and their local config, with the custom key switching to the
local config and mod+q switching back.
* nixos/xmonad: refactor let binding
* nixos/xmonad: refactor (eliminate duplicate code)
* nixos/xmonad: install man pages
Prior to this commit, man pages were not installed if the "config"
option was set.
* nixos/xmonad: comment grammar fixups
* nixos/xmonad: writeStateToFile in example config
Calling writeStateToFile prior to recompiling and restarting allows
state (workspaces, etc.) to be preserved across the restart.
* nixos/xmonad: add ivanbrennan to maintainers
* nixos/xmonad: adjust compileRestart example
* nixos/xmonad: add missing import to example config
Closes#106603
Some webservers (lighttpd) require that the
files they are serving are world readable. We
do our own chmods in the scripts anyway, and
lego has sensible permissions on its output
files, so this change is safe enough.
systemd-tmpfiles is no longer required for
most of the critical paths in the module. The
only one that remains is the webroot
acme-challenge directory since there's no
other good place for this to live and forcing
users to do the right thing alone will only
create more issues.
Closes#106565
When generating multiple certificates which all
share the same server + email, lego will attempt
to create an account multiple times. By adding an
account creation target certificates which share
an account will wait for one service (chosen at
config build time) to complete first.
@poettering decided we only need a limited number of inodes in our /tmp,
so why not limit that for every systemd user? That makes medium-sized nix
builds impossible so this commit restores the old behaviour which is the
kernel default of half the number of physical RAM pages which does not
seem too unreasonable to me.
Since version 5.2.0 there's non-empty stop phase:
ExecStopPost=/usr/bin/env rm -f "/run/knot-resolver/control/%i"
but it's perfectly OK to run that from a different version
(and typically it's no-op anyway). Real-life example where this helps:
https://github.com/NixOS/nixpkgs/issues/49528#issuecomment-747723198
This allows to use files containing only the mpd password without the
permissions, making it easier for other programs connecting to mpd to read the
password from the same password file.
... build-vm-with-bootloader" for EFI systems
This reverts commit 20257280d9, reversing
changes made to 926a1b2094.
It broke nixosTests.installer.simpleUefiSystemdBoot
and right now channel is lagging behing for two weeks.
This reverts commit 8f177612b1.
Attempting to start any service from udev when systemd-udev-settle is
used at all hangs the boot for 2min. See issue #107341.
Prior to this commit, installation over serial console would requiring
manually having to modify the kernel modeline, as described in
https://github.com/NixOS/nixpkgs/issues/58198 .
This is unnecessarily fiddly, so this commit adds a syslinux boot
entry that has serial enabled.
GRUB already has a serial console entry:
2c07a0800a/nixos/modules/installer/cd-dvd/iso-image.nix (L311-L317)
Why 115200 bps? This is already used in other places, e.g. https://github.com/NixOS/nixpkgs/pull/58196
I tested this change by building the image, booting the image, and
observing the boot process over serial:
$ cd nixos/
$ nix-build -A config.system.build.isoImage -I nixos-config=modules/installer/cd-dvd/installation-cd-minimal.nix default.nix
$ sudo cp /nix/store/arcl702c3z8xlndlvnfplq9yhixjvs9k-nixos-20.09pre-git-x86_64-linux.iso/iso/nixos-20.09pre-git-x86_64-linux.iso /dev/sdb
$ picocom -b 115200 /dev/ttyUSB0
`nixos-rebuild build-vm-with-bootloader` currently fails with the
default NixOS EFI configuration:
$ cat >configuration.nix <<EOF
{
fileSystems."/".device = "/dev/sda1";
boot.loader.systemd-boot.enable = true;
boot.loader.efi.canTouchEfiVariables = true;
}
EOF
$ nixos-rebuild build-vm-with-bootloader -I nixos-config=$PWD/configuration.nix -I nixpkgs=https://github.com/NixOS/nixpkgs/archive/nixos-20.09.tar.gz
[...]
insmod: ERROR: could not insert module /nix/store/1ibmgfr13r8b6xyn4f0wj115819f359c-linux-5.4.83/lib/modules/5.4.83/kernel/fs/efivarfs/efivarfs.ko.xz: No such device
mount: /sys/firmware/efi/efivars: mount point does not exist.
[ 1.908328] reboot: Power down
builder for '/nix/store/dx2ycclyknvibrskwmii42sgyalagjxa-nixos-boot-disk.drv' failed with exit code 32
[...]
Fix it by setting virtualisation.useEFIBoot = true in qemu-vm.nix, when
efi is needed.
And remove the now unneeded configuration in
./nixos/tests/systemd-boot.nix, since it's handled globally.
Before:
* release-20.03: successful build, unsuccessful run
* release-20.09 (and master): unsuccessful build
After:
* Successful build and run.
Fixes https://github.com/NixOS/nixpkgs/issues/107255
Fixes that `containers.<name>.extraVeths.<name>` configuration was not
always applied.
When configuring `containers.<name>.extraVeths.<name>` and not
configuring one of `containers.<name>.localAddress`, `.localAddress6`,
`.hostAddress`, `.hostAddress6` or `.hostBridge` the veth was created,
but otherwise no configuration (i.e. no ip) was applied.
nixos-container always configures the primary veth (when `.localAddress`
or `.hostAddress` is set) to be the containers default gateway, so
this fix is required to create a veth in containers that use a different
default gateway.
To test this patch configure the following container and check if the
addresses are applied:
```
containers.testveth = {
extraVeths.testveth = {
hostAddress = "192.168.13.2";
localAddress = "192.168.13.1";
};
config = {...}:{};
};
```
This means that all systems running from master will trigger
new certificate creation on next rebuild. Race conditions around
multiple account creation are fixed in #106857, not this commit.
Previously the .enable option was used to encode the condition as well,
which lead to some oddness:
- In order to encode an assertion, one had to invert it
- To disable a check, one had to mkForce it
By introducing a separate .check option this is solved because:
- It can be used to encode assertions
- Disabling is done separately with .enable option, whose default can be
overridden without a mkForce
Since slurm-20.11.0.1 the dbd server requires slurmdbd.conf to be
in mode 600 to protect the database password. This change creates
slurmdbd.conf on-the-fly at service startup and thus avoids that
the database password ends up in the nix store.
It's very surprising that services.tor.client.enable would set
services.privoxy.enable. This violates the principle of least
astonishment, because it's Privoxy that can integrate with Tor, rather
than the other way around.
So this patch moves the Privoxy Tor integration to the Privoxy module,
and it also disables it by default. This change is documented in the
release notes.
Reported-by: V <v@anomalous.eu>
configuration.nix(1) states
users.extraUsers.<name>.createHome
[...] If [...] the home directory already exists but is not
owned by the user, directory owner and group will be changed to
match the user.
i.e. ownership would change only if the user mismatched; the code
however ignores the owner, it is sufficient to enable `createHome`:
if ($u->{createHome}) {
make_path($u->{home}, { mode => 0700 }) if ! -e $u->{home};
chown $u->{uid}, $u->{gid}, $u->{home};
}
Furthermore, permissions are ignored on already existing directories and
therefore may allow others to read private data eventually.
Given that createHome already acts as switch to not only create but
effectively own the home directory, manage permissions in the same
manner to ensure the intended default and cover all primary attributes.
Avoid yet another configuration option to have administrators make a
clear and simple choice between securely managing home directories
and optionally defering management to own code (taking care of custom
location, ownership, mode, extended attributes, etc.).
While here, simplify and thereby fix misleading documentation.
Not everyone has a suitable remote builder set up, so the cross-compilation
tests that _include_ running the result are separate. That way, most people
can run the majority of the test suite without the extra setup.
Docker (via containerd) and the the OCI Image Configuration imply and
suggest, respectfully, that the architecture set in images matches those
of GOARCH in the Go Language document.
This changeset updates the implimentation of getArch in dockerTools to
return GOARCH values, to satisfy Docker.
Fixes: #106695
Mailman can now work with MTAs other than Postfix. You'll have to configure
it yourself using the options in `services.mailman.settings.mta`.
This addition is reflected in the release notes for 21.03.
This partially reverts bf3d3dd19b.
I don't know why we weren't getting a default logfile back then but Xorg
definitely provides one now ($XDG_DATA_HOME for regular users and /var/log for
root, see `man Xorg`)
Only set Before=display-manager.service if it is actually present.
On headless systems, `systemctl list-units --state not-found` will
otherwise show display-manager.service.
Reported-In: https://github.com/NixOS/nixpkgs/issues/88597
Dnscrypt-proxy needs some options to be set before it can do anything useful.
Currently, we only apply what the user configured which, by default, is nothing.
This leads to the dnscrypt-proxy2 service failing to start when you only set
`enable = true;` which is not a great user experience.
This patch makes the module take the example config from the upstream repo as a
base on top of which the user-specified settings are applied (it contains sane
defaults).
An option has been added to restore the old behaviour.
`file_exists` also returns `FALSE` if the file is in a directory that
can't be read by the user. This e.g. happens if permissions for
`nixops(1)`-deployment keys aren't configured correctly.
This patch improves the error message for invalid files to avoid
confusion[1].
[1] https://discourse.nixos.org/t/nixops-deploy-secrets-to-nextcloud/10414/4
This patch:
* Removes an invalid/useless classpath element;
* Removes an unnecessary environment variable;
* Creates the required '/version-2' data subdirectory;
* Redirects audit logging to the "console" (systemd) by default.
A big jump, but the structure hasn't changed much.
This recipe is still based on a binary release provided by upstream.
(It might be interesting to start doing our own builds at some point,
to split client from server, and/or to create packages for removed
"contribs" such as 'zooInspector'. Upstream intends to further slim
down its release tarballs as most deployments only need specific assets.)
Unbound throws the following error:
--8<---------------cut here---------------start------------->8---
error: failed to list interfaces: getifaddrs: Address family not supported by protocol
fatal error: could not open ports
--8<---------------cut here---------------end--------------->8---
The solution is pulled from upstream:
https://github.com/NLnetLabs/unbound/pull/351
Unfortunately, I had a use-case where `services.nginx.config` was
necessary quite recently. While working on that config I had to look up
the module's code to understand which options can be used and which
don't.
To slightly improve the situation, I changed the documentation like
this:
* Added `types.str` as type since `config` is not mergeable on purpose.
It must be a string as it's rendered verbatim into `nginx.conf` and if
the type is `unspecified`, it can be confused with RFC42-like options.
* Mention which config options that don't generate config in
`nginx.conf` are NOT mutually exclusive.
