]>
git.ipfire.org Git - thirdparty/systemd.git/log
Lennart Poettering [Fri, 26 Feb 2021 08:36:11 +0000 (09:36 +0100)]
Merge pull request #18801 from keszybz/small-documentation-updates
Small documentation updates
Luca Boccassi [Thu, 25 Feb 2021 22:20:04 +0000 (22:20 +0000)]
Merge pull request #18797 from keszybz/trivial-cleanups
Trivial cleanups
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 20:13:47 +0000 (21:13 +0100)]
man: remove details of ManagedOOMPreference implementation
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 13:54:54 +0000 (14:54 +0100)]
man: advertise shared drop-ins more
systemd.unit(5) is a wall of text. And this particular feature can be very useful
in the context of resource control. Let's avertise this cool feature a bit more.
Fixes #17900.
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 13:30:34 +0000 (14:30 +0100)]
man: fix two issues in udev(7)
Fixed #18050.
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 13:14:39 +0000 (14:14 +0100)]
man: say that hostname must can be a fqdn or not
Fixes #18426
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 13:01:56 +0000 (14:01 +0100)]
timesync: rename variable to match config name
PollIntervalMinSec and PollIntervalMaxSec use the same pattern, but
RootDistanceMaxSec had switched orderd in the code.
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 12:57:20 +0000 (13:57 +0100)]
man: explain settings in timesyncd.conf(5)
Fixes #18508.
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 12:27:49 +0000 (13:27 +0100)]
man: drop misplaced spaces and parenthesization
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 12:25:35 +0000 (13:25 +0100)]
man: use markup more in description of ManagedOOMPreference=
Follow-up for
d8a4d64bc37ad21381f584b6c25792995584c7a2 .
Lennart Poettering [Thu, 25 Feb 2021 16:00:54 +0000 (17:00 +0100)]
Merge pull request #18798 from poettering/getenv-list-fixes
various follow-ups for ExtensionImages= PR
Lennart Poettering [Thu, 25 Feb 2021 14:51:13 +0000 (15:51 +0100)]
update TODO
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 12:16:39 +0000 (13:16 +0100)]
man: fix grammaro
Luca Boccassi [Thu, 25 Feb 2021 10:59:01 +0000 (10:59 +0000)]
NEWS: mention ExtensionImages
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 11:46:34 +0000 (12:46 +0100)]
libudev: fix typo in docstring
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 11:35:26 +0000 (12:35 +0100)]
shared/dissect-image: inline iterator variables
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 11:33:27 +0000 (12:33 +0100)]
core/namespace: inline more iterator variable declarations
Zbigniew Jędrzejewski-Szmek [Mon, 22 Feb 2021 11:20:34 +0000 (12:20 +0100)]
resolved: inline trivial function
Lennart Poettering [Thu, 25 Feb 2021 12:38:20 +0000 (13:38 +0100)]
bus-unit-util: generate proper log message when we fail to parse properties
Lennart Poettering [Thu, 25 Feb 2021 12:36:43 +0000 (13:36 +0100)]
bus-unit-util: don't validate partition designator client side
When we parse properties set via "systemctl set-property" we should
validate to the point where the data we pass over the bus makes rough
sense, but we shouldn't needlessly check whether specified enum values
are among the known enum values. The server side checks that anyway
again, and it's kinda nice if an older systemctl can be used to talk to
a newer systemd.
Lennart Poettering [Thu, 25 Feb 2021 12:36:32 +0000 (13:36 +0100)]
namespace: return correct error code
Lennart Poettering [Thu, 25 Feb 2021 12:16:36 +0000 (13:16 +0100)]
env-util: fix parameter handling of parse_env_extension_hierarchies() + getenv_path_list()
Our coding style dictates we should not clobber return parameters on
failure, and always initialize them on success. Do so here.
This changes getenv_path_list() to return ENXIO if the env var is not
set, which is similar to how we handle this in getenv_bool().
This drops debug logging from parse_env_extension_hierarchies(), since
it's done anyway in getenv_path_list()
Follow-up for: #18018
Lennart Poettering [Thu, 25 Feb 2021 07:56:57 +0000 (08:56 +0100)]
signal-util: make -1 termination of ignore_signals() argument list unnecessary
Clean up ignore_signals() + default_signals() + sigaction_many() a bit:
make it unnecessary to explicitly terminate the signal list with -1.
Merge all three calls into a single function that is just called with
slightly different parameters. And eliminate an unnecessary extra
iteration in its inner for() loop.
No change in behaviour.
Zbigniew Jędrzejewski-Szmek [Thu, 25 Feb 2021 10:31:14 +0000 (11:31 +0100)]
Merge pull request #18018 from bluca/mount_images_overlay
Add ExtensionImages directive to form overlays
Zbigniew Jędrzejewski-Szmek [Wed, 24 Feb 2021 17:02:36 +0000 (18:02 +0100)]
meson: refuse when prefix is not a child of rootprefix
This is most likely to happen when setting one but not the other.
Note that we already warn when rootprefixdir != rootprefix_default,
at the very end.
Dan Streetman [Wed, 24 Feb 2021 12:29:15 +0000 (07:29 -0500)]
test: skip tests if cgroup isn't mounted
Debian builds in a chroot without cgroup mounted, so tests expecting
to access cgroup will fail with ENOMEDIUM, so skip the tests in that
situation.
Lennart Poettering [Wed, 24 Feb 2021 16:49:32 +0000 (17:49 +0100)]
cgroup: don't generate BPF firewall unsupported warning on wrong unit
Let's generate the warning on a unit that actually needs the BPF
firewall, and not confusingly already for a sibling of one.
Lennart Poettering [Wed, 24 Feb 2021 17:54:32 +0000 (18:54 +0100)]
format-table: don't suppress trailing whitespace in first line
We underline the first line, hence we shouldn't suppress the whitespace,
otherwise the line will end early.
Follow-up for:
71894e18313e41a72cecdc77fea5037f95d6903f
Joerg Behrmann [Wed, 24 Feb 2021 20:39:03 +0000 (21:39 +0100)]
treewide: fix spelling
Annika Wickert [Fri, 2 Oct 2020 19:43:05 +0000 (21:43 +0200)]
network: Implement B.A.T.M.A.N. Advanced interface type
Sergey Bugaev [Fri, 19 Feb 2021 12:29:02 +0000 (15:29 +0300)]
resolved: don't discard mDNS queries with bad flags
Even though RFC 6762 specifies these bits MUST be zero, it also says they MUST
be ignored on reception.
Sergey Bugaev [Mon, 15 Feb 2021 16:03:27 +0000 (19:03 +0300)]
resolved: don't discard mDNS packets with ip6.arpa RRs
ip6.arpa is also a valid domain name to put in mDNS packets.
Lennart Poettering [Wed, 24 Feb 2021 16:14:14 +0000 (17:14 +0100)]
Merge pull request #18771 from yuwata/timedate-fix-set-local-rtc-18391
timedate: do not ignore fix_system argument in SetLocalRTC method
Yu Watanabe [Wed, 24 Feb 2021 04:16:15 +0000 (13:16 +0900)]
timedate: do not ignore fix_system argument in SetLocalRTC method
Fixes #18391.
Luca Boccassi [Wed, 24 Feb 2021 13:29:57 +0000 (13:29 +0000)]
test: skip TEST-56-OOMD if systemd-oomd is not installed
Fixes #18775
Yu Watanabe [Wed, 24 Feb 2021 11:29:24 +0000 (20:29 +0900)]
network: use IPMasquerade=both instead of yes
Follow-up for
4c72d851cd007e945a85811f89376a2675daa1a5 .
Yu Watanabe [Wed, 24 Feb 2021 04:06:17 +0000 (13:06 +0900)]
clock-util: rename function argument
Yu Watanabe [Wed, 24 Feb 2021 05:25:24 +0000 (14:25 +0900)]
doc,man: fix-typo (hierachy -> hierarchy)
Zbigniew Jędrzejewski-Szmek [Wed, 24 Feb 2021 08:53:37 +0000 (09:53 +0100)]
Merge pull request #18730 from yuwata/dhcp6-client-do-not-set-t1-t2-18090
dhcp6: do not set T1 and T2 in IA_PD option or friends
Frantisek Sumsal [Tue, 23 Feb 2021 22:40:13 +0000 (23:40 +0100)]
Translated using Weblate (Serbian)
Currently translated at 66.6% (126 of 189 strings)
Co-authored-by: Frantisek Sumsal <frantisek@sumsal.cz>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/master/sr/
Translation: systemd/main
Yu Watanabe [Wed, 24 Feb 2021 00:07:57 +0000 (09:07 +0900)]
Merge pull request #18589 from yuwata/network-nexthop-drop-unnecessary-nexthops
network: drop unnecessary nexthops
Topi Miettinen [Tue, 23 Feb 2021 17:58:28 +0000 (17:58 +0000)]
Update NEWS
Fix typos, improve /dev exec/noexec description
Lennart Poettering [Tue, 23 Feb 2021 14:02:33 +0000 (15:02 +0100)]
bootspec: assume that the root dir is at the top of its file system
Fixes: #17746
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 17:41:53 +0000 (18:41 +0100)]
Merge pull request #18763 from poettering/cov-fixes
A bunch of coverity fixes
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 16:44:13 +0000 (17:44 +0100)]
Merge pull request #18765 from keszybz/more-docs
NEWS and one more man page addition
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 16:18:05 +0000 (17:18 +0100)]
man: add a description of how manager env block is set
Luca Boccassi [Mon, 22 Feb 2021 12:20:33 +0000 (12:20 +0000)]
Add ExtensionImages directive to form overlays
Add support for overlaying images for services on top of their
root fs, using a read-only overlay.
Luca Boccassi [Mon, 22 Feb 2021 12:12:54 +0000 (12:12 +0000)]
core/namespace: reafactor applying mounts in a separate function
The setup_namespace code to apply mounts is a big if block that
keeps growing, so refactor it in a separate function.
Luca Boccassi [Mon, 22 Feb 2021 18:45:36 +0000 (18:45 +0000)]
sysext: split parsing SYSTEMD_SYSEXT_HIERARCHIES in a common helper
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 15:30:13 +0000 (16:30 +0100)]
NEWS: add ConditionCGC=v12 and the hostname stuff
Lennart Poettering [Tue, 23 Feb 2021 13:16:08 +0000 (14:16 +0100)]
efi stub: accept it if our loaded image has no FilePath field set
The firmware spec doesn't really say whether FilePath of the LoadedImage
protocol may be NULL or not. So far we assumed it to be non-NULL, but
apparently the FreeBSD UEFI chainloader sets it to NULL. Handle this
gracefully.
(Noticed and tracked down by Alexander Schreiber)
Fixes: #18733
Lennart Poettering [Tue, 23 Feb 2021 14:43:19 +0000 (15:43 +0100)]
core: fix memory leak in error path
Coverity
1431998
Lennart Poettering [Tue, 23 Feb 2021 14:29:10 +0000 (15:29 +0100)]
resolved: don't dereference ->scope unless set
Coverity
1446388
Lennart Poettering [Tue, 23 Feb 2021 14:26:49 +0000 (15:26 +0100)]
shared: fix memory leak in error path
Coverity
1446676
Lennart Poettering [Tue, 23 Feb 2021 14:13:21 +0000 (15:13 +0100)]
icmp6-util: add missing variable initialization
Coverity
1446712
Lennart Poettering [Tue, 23 Feb 2021 14:11:07 +0000 (15:11 +0100)]
dissect-image: initially fds[] array fully
THe previous code wasn't wrong, but it's certainly nicer to avoid a
partially uninitialized array here, hence Coverity has a point
complaining about this.
Coverity
1446731
Lennart Poettering [Tue, 23 Feb 2021 14:10:14 +0000 (15:10 +0100)]
resolved: fix minor memory leak
Coverity
1446992
Yu Watanabe [Tue, 23 Feb 2021 11:48:51 +0000 (20:48 +0900)]
core: fix stack-use-after-scope
Fixes a bug introduced by
f70e6fb471a4eeed8e2a4a7c608beabbb0aba8be .
Fixes #18753.
AsciiWolf [Tue, 23 Feb 2021 11:08:34 +0000 (12:08 +0100)]
po: remove basic fedora.zanata.org configuration
Zanata is not used anymore and its current configuration is not useful
anyway since the master branch was renamed to main.
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 14:30:33 +0000 (15:30 +0100)]
Merge pull request #18751 from anitazha/oomdconditions
units: add Conditions for systemd-oomd.service
Yu Watanabe [Tue, 23 Feb 2021 09:45:48 +0000 (18:45 +0900)]
test-network: add tests for dropping unnecessary nexthops
Yu Watanabe [Mon, 15 Feb 2021 01:00:14 +0000 (10:00 +0900)]
network: nexthop: drop unnecessary nexthops
Similar to addresses or routes, this makes networkd drops unnecessary
nexthops on configuring links or when a link is dropped.
Yu Watanabe [Mon, 22 Feb 2021 18:10:16 +0000 (03:10 +0900)]
network: nexthop: read protocol in received netlink message
Preparation of later commits.
Yu Watanabe [Mon, 22 Feb 2021 18:05:24 +0000 (03:05 +0900)]
sd-netlink: introduce sd_rtnl_message_nexthop_get_protocol()
Lennart Poettering [Tue, 23 Feb 2021 13:07:46 +0000 (14:07 +0100)]
man: various improvements to systemd.generator(7)
In particular, make clear the .d/*.conf unit file drop-ins are OK to
generate from generators.
Inspired by: https://lists.freedesktop.org/archives/systemd-devel/2021-February/046148.html
Yu Watanabe [Tue, 23 Feb 2021 10:49:49 +0000 (19:49 +0900)]
Revert "meson: remove one more instance of install_dir:bindir"
This reverts commit
5ffa2eaa541b718122a3422e911ef95f4c9981d7 .
It seems that if install_dir is not specified, meson decides install path
based on file type, and non-executable binary files are installed under
/usr/share.
kernel-install is a script. So, we need to set install_dir argument
explicitly.
Fixes #18754.
Yu Watanabe [Mon, 22 Feb 2021 13:39:31 +0000 (22:39 +0900)]
dhcp6: do not set T1 and T2 by dhcp6_option_append_pd() in client
Fixes #18090.
Yu Watanabe [Mon, 22 Feb 2021 12:54:32 +0000 (21:54 +0900)]
dhcp6: do not set T1 and T2 by dhcp6_option_append_ia() in client
Yu Watanabe [Tue, 23 Feb 2021 10:25:17 +0000 (19:25 +0900)]
dhcp6: make dhcp6_option_parse_{address,pdprefix}() return -EINVAL when received address or prefix is refused
And then the caller ignores the error.
Otherwise, `ret_liftime_valid` is not set even if they succeeds.
Anita Zhang [Tue, 23 Feb 2021 10:06:58 +0000 (02:06 -0800)]
README: add config_psi as requirement for systemd-oomd
Anita Zhang [Tue, 23 Feb 2021 08:09:10 +0000 (00:09 -0800)]
units: add Conditions for systemd-oomd.service
Prompted by https://bugzilla.redhat.com/show_bug.cgi?id=
1927148 and
https://bugzilla.redhat.com/show_bug.cgi?id=
1931181 .
Adds condition check for cgroups v2 and PSI, which systemd-oomd cannot
start without.
Yu Watanabe [Mon, 22 Feb 2021 12:59:24 +0000 (21:59 +0900)]
dhcp6: do not use input value before checking
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 00:31:15 +0000 (01:31 +0100)]
resolved: add another explanatory message to stub resolv.conf files
Apparently people do "cat /etc/resolv.conf" and not realize that they
are looking at a file in /run.
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 00:56:21 +0000 (01:56 +0100)]
meson: bump version numbers
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 00:55:11 +0000 (01:55 +0100)]
NEWS: add contributors
Lennart Poettering [Mon, 22 Feb 2021 23:07:12 +0000 (00:07 +0100)]
resolved: actually use the specified rrsig argument
The argument only exists to be used to override/fill in the RRSIG record
of the answer item. Hence actually use it instead of ignore it.
(Not sure how this got lost earlier.)
Fixes: #18714
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 00:52:01 +0000 (01:52 +0100)]
Merge pull request #18741 from poettering/stub-no-cname
resolved: don't follow CNAMEs in the stub anymore
Zbigniew Jędrzejewski-Szmek [Mon, 22 Feb 2021 21:09:38 +0000 (22:09 +0100)]
shared/condition: add ConditionControlGroupController=v1|v2
Before, we only allowed conditionalizing on controllers, not the hierarchy.
This commit extends this to allow a simple check for v1 (i.e. classic or hybrid),
and v2 (full unified).
An alternative approach would be to add a separate Condition for this, but I'm
not too keen on that, considering that v1 is already being deprecrecated
(c.f.
82f3063218 ).
Luca Boccassi [Mon, 22 Feb 2021 21:31:44 +0000 (21:31 +0000)]
MountImages: parse unit specifiers after permissive hyphen
The permissive bit it not something a specifier might synthetise
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 00:10:38 +0000 (01:10 +0100)]
Merge pull request #18745 from keszybz/stop-using-fstrings
Stop using fstrings
Zbigniew Jędrzejewski-Szmek [Tue, 23 Feb 2021 00:06:08 +0000 (01:06 +0100)]
test-dns-packet: fix leak of DnsResourceRecord object
Zbigniew Jędrzejewski-Szmek [Mon, 22 Feb 2021 23:39:08 +0000 (00:39 +0100)]
NEWS: a few more entries for v248-rc1
I left the stuff related to [NextHop] out. There are still
patches outstanding, and we can add a comprehensive entry once
things reached the final form.
Zbigniew Jędrzejewski-Szmek [Mon, 22 Feb 2021 23:41:27 +0000 (00:41 +0100)]
Merge pull request #18704 from keszybz/fallback-hostame-override
Allow overriding of fallback hostname through envvar and os-release field
Lennart Poettering [Mon, 22 Feb 2021 22:07:57 +0000 (23:07 +0100)]
resolved: stick CNAME targets into main answer section in stub replies
Lennart Poettering [Mon, 22 Feb 2021 22:07:01 +0000 (23:07 +0100)]
resolved: add dns_resource_record_get_cname_target() helper
This determines the redirection target from a CNAME or DNAME RR given it
matches some given RR key.
Lennart Poettering [Mon, 22 Feb 2021 22:07:22 +0000 (23:07 +0100)]
resolved: remove duplicate comment
Practically the same comment is a few lines up covering both parts
anyway, let's remove one.
Lennart Poettering [Mon, 22 Feb 2021 21:51:32 +0000 (22:51 +0100)]
resolved: update comment in DNS stub code a bit
There's no "answer_auxiliary" object anymore, it's all one "answer"
object, and we have per-item flags that tell us which section things are
from, i.e. from the main answer section, or the additional or
authoritative ones.
Lennart Poettering [Mon, 22 Feb 2021 20:32:13 +0000 (21:32 +0100)]
resolved: don't follow CNAMEs in the stub anymore
CNAME following was broken by
775ae35403f8f3c01b7ac13387fe8aac1759993f
where we'd not properly collect RRs along the CNAME path. Good thing
though is that we don't have to anymore: since we nowadays propagate all
sections of the upstream replies into the cache and back to stub clients
all the information should already be available anyway, and there's no
need for us to collect it.
Fixes: #18690
Zbigniew Jędrzejewski-Szmek [Mon, 22 Feb 2021 22:08:13 +0000 (23:08 +0100)]
Merge pull request #18575 from bugaevc/aa
Set the AA bit in answers for synthetic records & mDNS
Zbigniew Jędrzejewski-Szmek [Mon, 22 Feb 2021 21:42:01 +0000 (22:42 +0100)]
Revert "generate-dns_type-gperf: modernize python syntax"
This reverts commit
b0a336a66929ed2a146888178157bf1af5c8598c .
Fixes #18708.
Zbigniew Jędrzejewski-Szmek [Mon, 22 Feb 2021 21:41:52 +0000 (22:41 +0100)]
Stop using f-strings in generate-sym-test.py
This partially reverts
7857b6e8383f5debab9544ef3abb15a27830fafa .
Debian 9 has python3.5 which does not have f-strings yet.
Partially fixes #18708.
Zbigniew Jędrzejewski-Szmek [Mon, 22 Feb 2021 20:10:04 +0000 (21:10 +0100)]
docs: align tables vertically to top
Fixes #18706.
Frantisek Sumsal [Mon, 22 Feb 2021 20:21:52 +0000 (21:21 +0100)]
Translated using Weblate (Slovak)
Currently translated at 24.8% (47 of 189 strings)
Co-authored-by: Frantisek Sumsal <frantisek@sumsal.cz>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/master/sk/
Translation: systemd/main
Yu Watanabe [Mon, 22 Feb 2021 19:44:12 +0000 (04:44 +0900)]
Merge pull request #18718 from yuwata/network-nexthop-blackhole
network: introduce Blackhole= setting in [NextHop] section
Yu Watanabe [Mon, 22 Feb 2021 19:43:22 +0000 (04:43 +0900)]
Merge pull request #18734 from poettering/cryptsetup-description-escape
cryptsetup: unescape ID_PART_ENTRY_NAME udev field
Luca Boccassi [Mon, 22 Feb 2021 10:12:34 +0000 (10:12 +0000)]
MountImages: fix exec_context_dump printf
Use strempty as options might not be set, and add the separator
for each option tuple
Yu Watanabe [Mon, 22 Feb 2021 19:42:09 +0000 (04:42 +0900)]
Merge pull request #18731 from yuwata/backlight-trivial-cleanups
backlight: trivial cleanups
Zbigniew Jędrzejewski-Szmek [Sat, 20 Feb 2021 17:54:24 +0000 (18:54 +0100)]
hostnamed: rename FallbackHostname to DefaultHostname
This follows the addition of DEFAULT_HOSTNAME= in os-release.
The distinction between the value from os-release or the env var and
the compile-time setting is not made in the api: HostnameSource is
"default" is all cases. I think that this level of detail is not needed,
because the users of this mostly care whether the hostname was set by
user configuration or not.
Zbigniew Jędrzejewski-Szmek [Fri, 19 Feb 2021 16:48:20 +0000 (17:48 +0100)]
Use the DEFAULT_HOSTNAME field from os-release
This provides a fairly comprehensible fix for
https://bugzilla.redhat.com/show_bug.cgi?id=
1893417 .
This adds yet-another level of configuration:
- /etc/hostname
- transient hostname
- $SYSTEMD_DEFAULT_HOSTNAME
- DEFAULT_HOSTNAME is os-release
- -Dfallback-hostname=
- "linux"
It's a lot of layers, but each has it's own justification.
Zbigniew Jędrzejewski-Szmek [Mon, 15 Feb 2021 10:50:55 +0000 (11:50 +0100)]
Allow the fallback hostname to be overriden using an environment variable
See https://bugzilla.redhat.com/show_bug.cgi?id=
1893417 for the back story:
the fallback hostname matters a lot in certain environments. Right now the only
way to configure the fallback hostname is by recompiling systemd, which is
obviously problematic in case when the fallback hostname shall differ between
different editions of the same distro that share a single compiled rpm.
By making this overridable through an envvar, we're providing an escape hatch
without making this a top-level api. Later on a way to set this through
os-release is added, but I think the approach with the variable is still
useful. It it very convenient for testing, or to override settings only in a
particular service, etc.