Before this, it relied on being able to `imports` from the `name`
module argument, which wasn't really necessary and required a
potentially quite breaking change. We can revert that now.
This is a decomposition of the testing-python.nix and build-vms.nix
files into modules.
By refactoring the glue, we accomplish the following:
- NixOS tests can now use `imports` and other module system features.
- Network-wide test setup can now be reusable; example:
- A setup with all VMs configured to use a DNS server
- Split long, slow tests into multiple tests that import a
common module that has most of the setup.
- Type checking for the test arguments
- (TBD) "generated" options reference docs
- Aspects that had to be wired through all the glue are now in their
own files.
- Chief example: interactive.nix.
- Also: network.nix
In rewriting this, I've generally stuck as close as possible to the
existing code; copying pieces of logic and rewiring them, without
changing the logic itself.
I've made two exceptions to this rule
- Introduction of `extraDriverArgs` instead of hardcoded
interactivity logic.
- Incorporation of https://github.com/NixOS/nixpkgs/pull/144110
in testScript.nix.
I might revert the latter and split it into a new commit.
Right now, the latest kernel is version 5.19, for which there is no
compatible upstream release for ZFS. However, our NixOS VM test always
uses linuxPackages_latest and thus will fail with an evaluation error
most of the time when a new mainline kernel is released.
Since we expose a latestCompatibleLinuxPackages attribute for the ZFS
packages, we already know what's the latest kernel version that is
supported so let's use that instead of linuxPackages_latest.
Signed-off-by: aszlig <aszlig@nix.build>
Co-authored-by: Shahar Dawn Or <mightyiampresence@gmail.com>
Co-authored-by: Ctem <c@ctem.me>
Co-authored-by: Ilan Joselevich <personal@ilanjoselevich.com>
Co-authored-by: a-kenji <aks.kenji@protonmail.com>
Co-authored-by: Ctem <c@ctem.me>
Co-authored-by: Brian Leung <leungbk@posteo.net>
Co-authored-by: Shahar Dawn Or <mightyiampresence@gmail.com>
Co-authored-by: Ilan Joselevich <personal@ilanjoselevich.com>
Before this change, the description for
security.wrappers.<name>.capabilities made it seem like you could just
string together the names of capabilities like this:
capabilities = "CAP_SETUID,CAP_SETGID";
In reality, each item in the list must be a full-on capability clause:
capabilities = "CAP_SETUID=ep,CAP_SETGID+i";
This attribute set isn't passed through the NixOS config resolution
mechanism, which means that we can't use lib.mkDefault here.
Instead, just put it before any user overrides so that if the user
specifies this environment variable it'll just override it anyway.
Optional functionality of AusweisApp2 requires an UDP port to be opened.
The module allows for convenient configuration and serves as documentation.
See also https://github.com/NixOS/nixpkgs/issues/136269
deprecate literalDocBook by adding a warning (that will not fire yet) to
its uses and other docbook literal strings by adding optional warning
message to mergeJSON.
This reverts commit 05958b228b.
Issue https://github.com/NixOS/nixpkgs/issues/188998 is concerns quite a
few NixOS users with full disk encryption and custom keymap.
Since there hasn't been a proper fix agreed upon and merged, I am
reverting this.
The changes can be applied again, when it is ensured that they do not
break custom keymaps in initrd.
Summary: fix errors with example code in the manual that shows how to set up DNS-01 verification via the acme protocol, e.g. for those who want to get wildcard certificates from Let's Encrypt.
Fix syntax error in nix arrays (there should not be commas.)
Fix permissions on /var/lib/secrets so it can be read by bind daemon. Without this fix bind won't start.
Add the missing feature: put the generated secret into certs.secret
In order to be able to use the unixd service with the `verify_ca` and
`verify_hostnames` set to `true` it needs to be able to read the
certificate store. This change bind mounts the cacert paths for the
unixd service.
Allow @resources syscalls in the grafana.service unit. While Grafana
itself does not need them, some plugins (incl. first party) crash if
they fail to setrlimit. This was first seen with the official grafana
Clickhouse datasource plugin.
The @resources syscalls set is fairly harmess anyway.
`paperless-ngx.pythonPath` was incomplete due to the missing paperless-ngx
source, so it had to be amended in the service.
Instead of amending it, define it entirely in the service.
This allows an override of `paperless-ngx.propagatedBuildInputs` to be reflected
in the service's PYTHONPATH.
We are building fwupd daemon with polkit support which means
polkit daemon is required.
Previously polkit was enabled by default via udisks2 but that
stopped with f763710065
breaking the fwupd installed tests as a result.
Let’s add the polkit dependency to the fwupd module to ensure polkit is available.
Handing CAP_NET_BIND_SERVICE to the `paperless-web.service` only makes
sense when it actually wants to bind to a port < 1024. Don't hand it out
if that is not the case.
Finding out how to connect paperless to a PostgreSQL database via unix
sockets and peer authentication took me a few minutes, so leaving a hint
in the extraConfig example seems like a good idea to me.
Also remove unnecessary use of literalExpression for attribute set, it
is only required for complex values like functions or values that depend
on other values or packages.
After uploading a document through the webinterface I started seeing
it killed through the SYSBUS signal. Inspecting the call trace led me to
liblapack's memory allocator, that uses the mbind syscall on Linux.
Prior to this change, ffmpeg couldn't be built for an
environment.noXlibs system, because it would fail in:
ffmpeg → SDL2 → libdecor
ffmpeg certainly does not need support for SDL2 windowing on a noXlibs
system.
This fix is important because the minidlna NixOS test, which uses the
minimal profile (and therefore environment.noXlibs) and ffmpeg, can't
currently build.
The primary difference between the standard and minimal variants of
this package is that all the X libraries are removed from the minimal
variant.
I had to switch the order of the definitions in all-packages.nix to
avoid an infinite recursion after the overlay was applied.
The udisks2 service was enabled to fix the test in (c5ebec7ee4).
However, cagebreak doesn't require udisks2, just polkit (which the
udisks2 module enables and which is why the cagebreak test broke after
the udisks2 module was disabled by default).
I've documented why polkit is required in this PR:
https://github.com/NixOS/nixpkgs/pull/156858
In this case the "dependency" chain is basically cagebreak -> wlroots ->
libseat -> logind (with polkit support) -> polkit.
Because of long standing bugs and stability issues & an
uncollaborative upstream there has been talk on the emacs-devel
mailing list to switch the default toolkit to
Lucid (https://lists.gnu.org/archive/html/emacs-devel/2022-08/msg00752.html).
The GTK build also has issues with Xinput2, something that both we and
upstream want to enable by default in Emacs 29.
This situation has prompted me to use both Lucid an no-toolkit (pure X11) Emacs
as a daily driver in recent weeks to evaluate what the
advantages/drawbacks are and I have concluded that, at least for me,
switching the toolkit to Lucid is strictly an upgrade.
It has resulted in better stability (there are far fewer tiny UX
issues that are hard to understand/identify) & a snappier UI.
On top of that the closure size is reduced by ~10%.
In the pure X11 build I noticed some unsharpness around fonts so this
is not a good default choice.
As with everything there is a cost, and that is uglier (I think most
would agree but of course this is subjective) menu bars for
those that use them and no GTK scroll bars.
For anyone who still wants to use GTK they could of course still
choose to do so via the new `emacs-gtk` attribute but I think this
is a bad default.
A note to Wayland users:
This does not affect Wayland compatibility in any way since that will
already need a PGTK build variant in the future.
For example, the wait_for_unit() call in the Moodle test times out for
myself and others[1], so it would be good to be able to increase it to
something less likely to be hit by a test that would otherwise pass.
[1]: https://github.com/NixOS/nixpkgs/pull/177052#issue-1266336706
When switching between different NixOS configurations (with and
without nullmailer and other services), it can happen that the UID of
the nullmailer user changes. When it happens, the nullmailer service
happily starts, but the user cannot send any email, because the
sendmail wrapper doesn't have permission to write them to the queue.
This commit prevents that. Instead of creating the directories by the
nullmailer user, which doesn't have permissions to change ownership,
we now create them by the systemd-tmpfiles, which has sufficient
permissions to adjust ownership.
most of these are hidden because they're either part of a submodule that
doesn't have its type rendered (eg because the submodule type is used in
an either type) or because they are explicitly hidden. some of them are
merely hidden from nix-doc-munge by how their option is put together.
conversions were done using https://github.com/pennae/nix-doc-munge
using (probably) rev f34e145 running
nix-doc-munge nixos/**/*.nix
nix-doc-munge --import nixos/**/*.nix
the tool ensures that only changes that could affect the generated
manual *but don't* are committed, other changes require manual review
and are discarded.
there are sufficiently few variable list around, and they are
sufficiently simple, that it doesn't seem helpful to add another
markdown extension for them. rendering differences are small, except in
the tor module: admonitions inside other blocks cannot be made to work
well with mistune (and likely most other markdown processors), so those
had to be shuffled a bit. we also lose paragraph breaks in the list
items due to how we have to render from markdown to docbook, but once we
remove docbook from the pipeline those paragraph breaks will be restored.
mostly no rendering changes. some lists (like simplelist) don't have an
exact translation to markdown, so we use a comma-separated list of
literals instead.
this mostly means marking options that use markdown already
appropriately and making a few adjustments so they still render
correctly. notable for nftables we have to transform the md links
because the manpage would not render them correctly otherwise.
most of the screen tags used in option docs are actually listings of
some sort. nsd had a notable exception where its screen usage was pretty
much a raw markdown block that made most sense to convert into docbook lists.
the way these are written they introduce lots of whitespace in each
line, which will cause those lines to render as code when converted to
markdown. override the whole description instead.