HALF ECHO LEARN / REFERENCE COLLECTION
v262.2NEWS
Git blob: b0d42b491c5b5bbc5168dc8d19d77308851d87df
Lines 2001–3000 of 23,379
protocols, proxies, or connection setup mechanisms.
Changes in systemd-resolved:
* systemd-resolved's .delegate files learnt a new setting FirewallMark=
to set the Linux network stack's "firewall mark" value for all DNS
traffic generated by the delegation.
* resolvectl now uses Varlink to connect to systemd-resolved.
* Queries done through nss-resolve can be limited to a specific
interface with the $SYSTEMD_NSS_RESOLVE_INTERFACE environment
variable.
* systemd-resolved now supports ifindex=0 in the BrowseServices IPC API,
to allow browsing all mDNS interfaces in one call.
Changes in systemd-sysupdate/systemd-sysupdated:
* systemd-sysupdate gained a new 'acquire' verb, allowing the download
and installation or update steps to be done separately.
* systemd-sysupdate will now refuse processing SHA256SUMS manifests if
they list a file BEST-BEFORE- suffixed by a date that is already in
the past, as a simple mechanism to detect freshness.
* systemd-sysupdate now can mark partitions as partially downloaded.
Changes in systemd-vmspawn:
* systemd-vmspawn gained support for registering with systemd-machined
in the user session. New options --user/--system control which
instance is used.
* systemd-vmspawn gained support for ephemeral machines via a new
--ephemeral option. This is similar to the functionality provided via
the same switch in systemd-nspawn.
* systemd-vmspawn gained a new switch --image-format= for selecting the
image format (i.e. support qcow2 in addition to raw) to boot from.
Also --extra-drive= now takes the image format as a colon separated
parameter.
Changes in systemd-nsresourced/systemd-mountfsd:
* The MakeDirectory() Varlink IPC call provided by systemd-mountfsd now
accepts a "mode" parameter for configuring the access mode of the
newly created directory. The MountImage() call gained a new
"mountOptions" parameter for configuring mount options for the
various partitions of a DDI explicitly. The call will now also report
via a new "singleFileSystem" field in the response whether it is
processing a DDI lacking a GPT envelope, and consisting of a raw file
system only. A new input parameter "relaxExtensionReleaseChecks"
controls whether to enforce extension release checks.
* systemd-nsresourced's BPF-LSM based security policy on user
namespaces it delegates UID ranges too is relaxed: processes in such
namespaces may now freely access to inodes owned by UIDs/GIDs outside
of the transient UID range. This reflects the fact that the security
policy exists to ensure ownership of inodes by transient UIDs is
never persisted on disk.
* systemd-nsresourced can now delegate multiples of additional 64K
ranges of UIDs/GIDs to user namespaces, on request. This permits
nesting of user namespace enabled containers with transient UID
ranges.
* systemd-nsresourced now supports a new type user namespace UID
delegation: only the client's UID is mapped. This is very similar to
what the kernel allows anyway as unprivileged delegation without
systemd-nsresourced involvement, however, can be combined with
multiple additional 64K ranges (see above).
* systemd-nsresourced may now optionally map the "foreign" UID/GID
range to itself for user namespaces it delegates transient UIDs/GIDs
to. This opens up the concept for nested containers.
* systemd-nsresourced's and systemd-mountfsd's Varlink sockets may now
be mounted into container trees, to permit nested use of their
functionality. This can be used automatically in systemd-nspawn's
--private-users-delegate= option.
Changes in systemd-logind:
* systemd-logind/systemd-udevd gained support for a new "xaccess"
concept for delegating access to specific devices to users with
specially marked sessions. The augments the "uaccess" logic that
provides device access to users with foreground sessions. The primary
usecase for this is to give access to GPU render devices to local
graphical sessions for remote users, i.e. which are not attached to
any local seat. Sessions are configured via the PAM environment
variable XDG_SESSION_EXTRA_DEVICE_ACCESS= for this logic.
* systemd-inhibit --list option gained support for JSON output
and filtering with --what= , --who=, --why=, and --mode=.
Changes in systemd-portabled:
* systemd-portabled now also runs as a user service. Unprivileged users
can run portable services (on sufficiently fresh
kernels). portablectl gained a pair of switches --user/--system to
explicitly select which service instance to talk to.
* systemd-portabled will now generate a policy and pin the image for a
portable service, so that the image cannot be changed later without a
reattach.
Changes in other components:
* systemd-repart gained basic support for dm-integrity protection of
encrypted volumes. Two new options Integrity= and IntegrityAlgorithm=
can be used to configure integrity checks for LUKS volumes.
* Image dissection policies have been extended to allow restricting
file system types and requiring integrity checks for encrypted
volumes with a new 'encryptedwithintegrity' policy.
* systemd-dissect gained a --copy-ownership= switch to configure
ownership of copied files.
* systemd-keyutil gained an 'extract-certificate' verb to print the
X.509 certificate. The existing 'public' verb has been renamed to
'extract-public' as it works analogously. (The old name remains
available for compatibility.)
* Support for interactive polkit authorization has been added to
systemd-sysext and varlinkctl.
* A polkit policy was added for systemd-ask-password, allowing it to be
used by unprivileged callers.
* journalctl now implements a Varlink interface that exposes a
GetEntries() method, which allows retrieving journal entries.
* systemd-importd gained support for downloading OCI images ("importctl
pull-oci"). They will be stored locally as "mstack" images, which
then can be used by various components, for example be run as system
services via RootMStack= in unit files, or as systemd-nspawn
containers via --mstack= (see below).
* systemd-nspawn gained a new --mstack= parameter to support the new
"mstack" feature for containers.
* A new systemd-mstack command line tool has been introduced to support
the new "mstack" feature interactively.
* New options SYSTEMD_COLORS=auto-16, SYSTEMD_COLORS=auto-256, and
SYSTEMD_COLORS=auto-24bit have been added. They are like
SYSTEMD_COLORS=16, SYSTEMD_COLORS=256, and SYSTEMD_COLORS=24bit
respectively when output is to a non-dumb TTY, and degrade to
SYSTEMD_COLORS=no otherwise.
* Standalone versions of the systemd-sysusers and systemd-tmpfiles
binaries now support full functionality. (Previously, those tools
were compiled without features which would require libmount and pull
in a lot of dependencies. After the conversion to dlopen the option
whether to support features requiring libmount is made by providing
libmount or not, at install time.)
* Internal code dealing with processes has been updated to use pidfds
in many places.
* busctl's 'wait' verb now honours --limit-messages= too.
* systemd-cryptsetup gained support for a new fixate-volume-key=
option, that can be used to pin a specific encrypted volume to an
/etc/crypttab entry via its volume key (more precisely a hash derived
from it). systemd-repart will assist generating this information.
* systemd-sysext/systemd-confext's "refresh" will now by default try to
suppress any operation in case no images where added, removed or
changed. To force a umount/mount operation in this case (i.e. get
back to the status quo ante) a new --always-refresh= option has been
added. Note that the change detection identifies extensions by
identity (verity hash, or else by file handle/inode, mount ID,
creation time and modification time), and does not look at their
contents. In particular, for directory-based extensions, adding,
removing or modifying files inside the directory (or touching the
directory itself) is not detected, and such a refresh is suppressed
by default. Use --always-refresh=yes to force a refresh in these
cases.
* systemd-oomd acquired "prekill hook" functionality, permitting other
system components to synchronously hook into the OOM killing logic,
by registering a Varlink socket in a special directory.
* systemd-analyze learnt a new verb "identify-tpm2" which shows
vendor/model information extracted from the system's TPM.
Changes in units:
* runlevel[0-6].target units that were removed in v258 have been
restored and can be enabled with the new -Dcompat-sysv-interfaces=yes
meson option. The installation of legacy.conf for tmpfiles is now
also conditionalized under the same option.
* getty@.service gained an [Install] and must now be explicitly enabled
to be active.
Contributions from: A S Alam, Adam Williamson, Adrian Vovk,
Alessandro Astone, Alexis-Emmanuel Haeringer, Allison Karlitskaya,
Américo Monteiro, Andrii Zora, André Paiusco, Anton Tiurin,
Antonio Alvarez Feijoo, Arjun-C-S, Artur Kowalski,
AshishKumar Mishra, Baurzhan Muftakhidinov, Ben Boeckel,
Betacentury, Bouke van der Bijl, Carlos Peón Costa,
Carolina Jubran, Cathy Hu, Chris Down, Chris Lindee,
Christian Brauner, Christian Glombek, Christian Hesse,
Christopher Cooper, Christopher Head,
Copilot Autofix powered by AI, Cyrus Xi, Daan De Meyer,
Dan McGregor, Daniel Foster, Daniel Nylander, Daniel Rusek,
David Santamaría Rogado, David Tardon, Derek J. Clark, Dirk Su,
Dmitry V. Levin, Dmytro Bagrii, Dylan M. Taylor,
Efstathios Iosifidis, Eisuke Kawashima, Ettore Atalan, Fergus Dall,
Florian Klink, Franck Bui, Frantisek Sumsal, Govind Venugopal,
Graham Reed, Guiorgy, Han Sol Jin, Hans de Goede, Heran Yang,
IntenseWiggling, Ivan Kruglov, Ivan Shapovalov, James Le Cuirot,
Jan Kuparinen, Jeff Layton, Jeremy Kerr, Jesse Guo, Jian Wen,
Jim Spentzos, Julian Sparber, Jörg Behrmann, Kai Lüke,
Lennart Poettering, Louis Stagg, Luca Boccassi, Lucas Werkmeister,
Luiz Amaral, Léane GRASSER, Malcolm Frazier, Marc Pervaz Boocha,
Marcel Leismann, Mario Limonciello, Mario Limonciello (AMD),
Martin Srebotnjak, Matt Fleming, Matteo Croce, Matthijs Kooijman,
Max Gautier, Maximilian Bosch, Miao Wang, Michael Vogt,
Michal Sekletár, Mike Gilbert, Mike Yuan, Mikhail Novosyolov,
Nandakumar Raghavan, Nick Rosbrook, Nicolas Dorier, Oblivionsage,
Oleksandr Andrushchenko, Oğuz Ersen, Pablo Fraile Alonso,
Peter Oliver, Philip Withnall, Pontus Lundkvist, Popax21,
Rito Rhymes, Rodrigo Campos, Ronan Pigott, Ryan Zeigler,
Salvatore Cocuzza, Sergey A., Skye Soss, Sriman Achanta,
Tabis Kabis, Temuri Doghonadze, The-An0nym, Thomas Weißschuh,
Thorsten Kukuk, Tobias Heider, Tobias Stoeckmann, Ulrich Ölmann,
Usama Arif, Val Markovic, Vitaly Kuznetsov, Vunny Sodhi,
Weixie Cui, Yaping Li, Yaron Shahrabani, Yu Watanabe,
Yuri Chornoivan, ZauberNerd, Zbigniew Jędrzejewski-Szmek,
Zongyuan He, andre4ik3, calm329, cdown, cyclopentane, davidak,
dongshengyuan, francescoza6, gvenugo3, joo es, kiamvdd, lumingzh,
naly zzwd, nikstur, novenary, noxiouz, patrick, ppkramer-hub, r-vdp,
safforddr, scarlet-storm, sd416, seidlerv, smosia, tuhaowen, zefr0x
— Edinburgh, 2026/03/17
CHANGES WITH 259:
Announcements of Future Feature Removals and Incompatible Changes:
* Support for System V service scripts is deprecated and will be
removed in v260. Please make sure to update your software *now* to
include a native systemd unit file instead of a legacy System V
script to retain compatibility with future systemd releases.
Following components will be removed:
* systemd-rc-local-generator,
* systemd-sysv-generator,
* systemd-sysv-install (hook for systemctl enable/disable/is-enabled).
* Required minimum versions of following components are planned to be
raised in v260:
* Linux kernel >= 5.10 (recommended >= 5.14),
* glibc >= 2.34,
* libxcrypt >= 4.4.0 (libcrypt in glibc will be no longer supported),
* util-linux >= 2.37,
* elfutils >= 0.177,
* openssl >= 3.0.0,
* cryptsetup >= 2.4.0,
* libseccomp >= 2.4.0,
* python >= 3.9.0.
* The parsing of RootImageOptions= and the mount image parameters of
ExtensionImages= and MountImages= will be changed in the next version
so that the last duplicated definition for a given partition wins and
is applied, rather than the first, to keep these options coherent with
other unit settings.
Feature Removals and Incompatible Changes:
* The cgroup2 file system is now mounted with the
"memory_hugetlb_accounting" mount option, supported since kernel 6.6.
This means that HugeTLB memory usage is now counted towards the
cgroup’s overall memory usage for the memory controller.
* The default storage mode for the journal is now 'persistent'.
Previously, the default was 'auto', so the presence or lack of
/var/log/journal determined the default storage mode, if no
overriding configuration was provided. The default can be changed
with -Djournal-storage-default=.
* systemd-networkd and systemd-nspawn no longer support creating NAT
rules via iptables/libiptc APIs; only nftables is now supported.
The corresponding meson option '-Dlibiptc=' is deprecated and will be
dropped in a future release.
* systemd-boot's and systemd-stub's support for TPM 1.2 has been
removed (only TPM 2.0 supported is retained). The security value of
TPM 1.2 support is questionable in 2025, and because we never
supported it in userspace, it was always quite incomplete to the
point of uselessness.
* The image dissection logic will now enforce the VFAT file system type
for XBOOTLDR partitions, similar to how it already does this for the
ESP. This is done for security, since both the ESP and XBOOTLDR must
be directly firmware-accessible and thus cannot be protected by
cryptographic means. Thus it is essential to not mount arbitrarily
complex file systems on them. This restriction only applies if
automatic dissection is used. If other file system types shall be
used for XBOOTLDR (not recommended) this can be achieved via explicit
/etc/fstab entries.
* systemd-machined will now expose "hidden" disk images as read-only by
default (hidden images are those whose name begins with a dot). They
were already used to retain a pristine copy of the downloaded image,
while modifications were made to a 2nd, local writable copy of the
image. Hence, effectively they were read-only already, and this is
now official.
* The LUKS volume label string set by systemd-repart no longer defaults
to the literal same as the partition and file system label, but is
prefixed with "luks-". This is done so that on LUKS enabled images a
conflict between /dev/disk/by-label/ symlinks is removed, as this
symlink is generated both for file system and LUKS superblock
labels. There's a new VolumeLabel= setting for partitions that can be
used to explicitly choose a LUKS superblock label, which can be used
to explicitly revert to the old naming, if required.
Service manager/PID1:
* The service manager's Varlink IPC has been extended considerably. It
now exposes service execution settings and more. Its Unit.List() call
now can filter by cgroup or invocation ID.
* The service manager now exposes Reload() and Reexecute() Varlink IPC
calls, mirroring the calls of the same name accessible via D-Bus.
* The $LISTEN_FDS protocol has been extended to support pidfd inode
IDs. The $LISTEN_PID environment variable is now augmented with a new
$LISTEN_PIDFDID environment variable which contains the inode ID of
the pidfd of the indicated process. This removes any ambiguity
regarding PID recycling: a process which verified that $LISTEN_PID
points to its own PID can now also verify the pidfd inode ID, which
does not recycle IDs.
* The log message made when a service exits will now show the
wallclock time the service took in addition to the previously shown
CPU time.
* A new pair of properties OOMKills and ManagedOOMKills are now exposed
on service units (and other unit types that spawn processes) that
count the number of process kills made by the kernel or systemd-oomd.
* The service manager gained support for a new
RootDirectoryFileDescriptor= property when creating transient service
units. It is similar to RootDirectory= but takes a file descriptor
rather than a path to the new root directory to use.
* The service manager now supports a new UserNamespacePath= setting
which mirrors the existing IPCNamespacePath= and
NetworkNamespacePath= options, but applies to Linux user namespaces.
* The service manager gained a new ExecReloadPost= setting to configure
commands to execute after reloading of the configuration of the
service has completed.
* Service manager job activation transactions now get a per-system
unique 64-bit numeric ID assigned. This ID is logged as an additional
log field for in messages related to the transaction.
* The service manager now keeps track of transactions with ordering
cycles and exposes them in the TransactionsWithOrderingCycle D-Bus
property.
systemd-sysext/systemd-confext:
* systemd-sysext and systemd-confext now support configuration files
/etc/systemd/sysext.conf and /etc/systemd/confext.conf, which can be
used to configure mutability or the image policy to apply to DDI
images.
* systemd-sysext's and systemd-confext's --mutable= switch now accepts
a new value "help" for listing available mutability modes.
* systemd-sysext now supports configuring additional overlayfs mount
settings via the $SYSTEMD_SYSEXT_OVERLAYFS_MOUNT_OPTIONS environment
variable. Similarly systemd-confext now supports
$SYSTEMD_CONFEXT_OVERLAYFS_MOUNT_OPTIONS.
systemd-vmspawn/systemd-nspawn:
* systemd-vmspawn will now initialize the "serial" fields of block
devices attached to VMs to the filename of the file backing them on
the host. This makes it very easy to reference the right media in
case many block devices from files are attached to the same VM via
the /dev/disk/by-id/… links in the VM.
* systemd-nspawn's .nspawn file gained support for a new NamespacePath=
setting in the [Network] section which takes a path to a network
namespace inode, and which ensures the container is run inside that
when booted. (This was previously only available via a command line
switch.)
* systemd-vmspawn gained two new switches
--bind-user=/--bind-user-shell= which mirror the switches of the same
name in systemd-nspawn, and allow sharing a user account from the host
inside the VM in a simple one-step operation.
* systemd-vmspawn and systemd-nspawn gained a new --bind-user-group=
switch to add a user bound via --bind-user= to the specified group
(useful in particular for the 'wheel' or 'empower' groups).
* systemd-vmspawn now configures RSA4096 support in the vTPM, if swtpm
supports it.
* systemd-vmspawn now enables qemu guest agent via the
org.qemu.guest_agent.0 protocol when started with --console=gui.
systemd-repart:
* repart.d/ drop-ins gained support for a new TPM2PCRs= setting, which
can be used to configure the set of TPM2 PCRs to bind disk encryption
to, in case TPM2-bound encryption is used. This was previously only
settable via the systemd-repart command line. Similarly, KeyFile= has
been added to configure a binary LUKS key file to use.
* systemd-repart's functionality is now accessible via Varlink IPC.
* systemd-repart may now be invoked with a device node path specified
as "-". Instead of operating on a block device this will just
determine the minimum block device size required to apply the defined
partitions and exit.
* systemd-repart gained two new switches --defer-partitions-empty=yes
and --defer-partitions-factory-reset=yes which are similar to
--defer-partitions= but instead of expecting a list of partitions to
defer will defer all partitions marked via Format=empty or
FactoryReset=yes. This functionality is useful for installers, as
partitions marked empty or marked for factory reset should typically
be left out at install time, but not on first boot.
* The Subvolumes= values in repart.d/ drop-ins may now be suffixed with
:nodatacow, in order to create subvolumes with data Copy-on-Write
disabled.
systemd-udevd:
* systemd-udevd rules gained support for OPTIONS="dump-json" to dump
the current event status in JSON format. This generates output
similar to "udevadm test --json=short".
* The net_id builtin for systemd-udevd now can generate predictable
interface names for Wifi devices on DeviceTree systems.
* systemd-udevd and systemd-repart will now reread partition tables on
block devices in a more graceful, incremental fashion. Specifically,
they no longer use the kernel BLKRRPART ioctl() which removes all
in-memory partition objects loaded into the kernel and then recreates
them as new objects. Instead they will use the BLKPG ioctl() to make
minimal changes, and individually add, remove, or grow modified
partitions, avoiding removal/re-adding where the partitions were left
unmodified on disk. This should greatly improve behaviour on systems
that make modifications to partition tables on disk while using them.
* A new udev property ID_BLOCK_SUBSYSTEM is now exposed on block devices
reporting a short identifier for the subsystem a block device belongs
to. This only applies to block devices not connected to a regular bus,
i.e. virtual block devices such as loopback, DM, MD, or zram.
* systemd-udevd will now generate /dev/gpio/by-id/… symlinks for GPIO
devices.
systemd-homed/homectl:
* homectl's --recovery-key= option may now be used with the "update"
command to add recovery keys to existing user accounts. Previously,
recovery keys could only be configured during initial user creation.
* Two new --prompt-shell= and --prompt-groups= options have been added
to homectl to control whether to query the user interactively for a
login shell and supplementary groups memberships when interactive
firstboot operation is requested. The invocation in
systemd-homed-firstboot.service now turns both off by default.
* systemd-homed's fscrypt backend gained a new key sealing format that
aims to be more future-proof and robust against advances in brute
forcing techniques. Compatibility with the existing format is maintained,
but new enrollments will happen with the new format. Users enrolling
new systems must thus take care in not downgrading to an older version.
systemd-boot/systemd-stub:
* systemd-boot now supports log levels. The level may be set via
log-level= in loader.conf and via the SMBIOS Type 11 field
'io.systemd.boot.loglevel='.
* systemd-boot's loader.conf file gained support for configuring the
SecureBoot key enrollment time-out via
secure-boot-enroll-timeout-sec=.
* Boot Loader Specification Type #1 entries now support a "profile"
field which may be used to explicitly select a profile in
multi-profile UKIs invoked via the "uki" field.
sd-varlink/varlinkctl:
* sd-varlink's sd_varlink_set_relative_timeout() call will now reset
the timeout to the default if 0 is passed.
* sd-varlink's sd_varlink_server_new() call learned two new flags
SD_VARLINK_SERVER_HANDLE_SIGTERM + SD_VARLINK_SERVER_HANDLE_SIGINT,
which are honoured by sd_varlink_server_loop_auto() and will cause it
to exit processing cleanly once SIGTERM/SIGINT are received.
* varlinkctl in --more mode will now send a READY=1 sd_notify() message
once it receives the first reply. This is useful for tools or scripts
that wrap it (and implement the $NOTIFY_SOCKET protocol) to know when
a first confirmation of success is received.
* sd-varlink gained a new sd_varlink_is_connected() call which reports
whether a Varlink connection is currently connected.
Shared library dependencies:
* Linux audit support is now implemented via dlopen() rather than
regular dynamic library linking. This means the dependency is now
weak, which is useful to reduce footprint inside of containers and
such, where Linux audit doesn't really work anyway.
* Similarly PAM support is now implemented via dlopen() too (except for
the PAM modules pam_systemd + pam_systemd_home + pam_systemd_loadkey,
which are loaded by PAM and hence need PAM anyway to operate).
* Similarly, libmount is now loaded via dlopen().
Note: libmount still must be installed for systemd to manage the
system. However, libsystemd.so no longer requires it, and neither do
various ways to invoke the systemd service manager binary other than
to manage the system.
* Similarly, libacl, libblkid, libseccomp, libselinux are loaded via
dlopen().
* systemd no longer links against libcap at all. The simple system call
wrappers and other APIs it provides have been reimplemented directly
in systemd, which reduced the codebase and the dependency tree.
→ In summary: all direct shared library linking is gone now from
systemd, with the one exception of libc.
systemd-machined/systemd-importd:
* systemd-machined gained support for RegisterMachineEx() +
CreateMachineEx() method calls which operate like their counterparts
without "Ex", but take a number of additional parameters, similar to
what is already supported via the equivalent functionality in the
Varlink APIs of systemd-machined. Most importantly, they support
PIDFDs instead of PIDs.
* systemd-machined may now also run in a per-user instance, in addition
to the per-system instance. systemd-vmspawn and systemd-nspawn have
been updated to register their invocations with both the calling
user's instance of systemd-machined and the system one, if
permissions allow it. machinectl now accepts --user and --system
switches that control which daemon instance to operate on.
systemd-ssh-proxy now will query both instances for the AF_VSOCK CID.
* systemd-machined implements a resolve hook now, so that the names of
local containers and VMs can be resolved locally to their respective
IP addresses.
* systemd-importd's tar extraction logic has been reimplemented based
on libarchive, replacing the previous implementation calling GNU tar.
This completes work begun earlier which already ported
systemd-importd's tar generation.
* systemd-importd now may also be run as a per-user service, in
addition to the existing per-system instance. It will place the
downloaded images in ~/.local/state/machines/ and similar
directories. importctl gained --user/--system switches to control
which instance to talk to.
systemd-firstboot:
* systemd-firstboot's and homectl's interactive boot-time interface
have been updated to show a colored bar at the top and bottom of the
screen, whose color can be configured via /etc/os-release. The bar
can be disabled via the new --chrome= switches to both tools.
* systemd-firstboot's and homectl's interactive boot-time interface
will now temporarily mute the kernel's and PID1's own console output
while running, in order to not mix the tool's own output with the
other sources. This logic can be controlled via the new
--mute-console= switches to both tools. This is implemented via a new
systemd-mute-console component (which provides a simple Varlink
interface).
* systemd-firstboot gained a new switch --prompt-keymap-auto. When
specified, the tool will interactively query the user for a keymap
when running on a real local VT console (i.e. on a user device where
the keymap would actually be respected), but not if invoked on other
TTYs (such as a serial port, hypervisor console, SSH, …), where the
keymap setting would have no effect anyway. The invocation in
systemd-firstboot.service now uses this.
systemd-creds:
* systemd-creds's Varlink IPC API now supports a new "withKey"
parameter on the Encrypt() method call, for selecting what to bind
the encryption to precisely, matching the --with-key= switch on the
command line.
* systemd-creds now allow explicit control of whether to accept
encryption with a NULL key when decrypting, via the --allow-null and
--refuse-null switches. Previously only the former existed, but null
keys were also accepted if UEFI SecureBoot was reported off. This
automatism is retained, but only if neither of the two switches are
specified. The systemd-creds Varlink IPC API learned similar
parameters on the Decrypt() call.
systemd-networkd:
* systemd-networkd's DHCP sever support gained two settings EmitDomain=
and Domain= for controlling whether leases handed out should report a
domain, and which. It also gained a per-static lease Hostname=
setting for the hostname of the client.
* systemd-networkd now exposes a Describe() method call to show network
interface properties.
* systemd-networkd now implements a resolve hook for its internal DHCP
server, so that the hostnames tracked in DHCP leases can be resolved
locally. This is now enabled by default for the DHCP server running
on the host side of local systemd-nspawn or systemd-vmspawn networks.
systemd-resolved:
* systemd-resolved gained a new Varlink IPC method call
DumpDNSConfiguration() which returns the full DNS configuration in
one reply. This is exposed by resolvectl --json=.
* systemd-resolved now allows local, privileged services to hook into
local name resolution requests. For that a new directory
/run/systemd/resolve.hook/ has been introduced. Any privileged local
service can bind an AF_UNIX Varlink socket there, and implement the
simple io.systemd.Resolve.Hook Varlink API on it. If so it will
receive a method call on it for each name resolution request, which
it can then reply to. It can reply positively, deny the request or
let the regular request handling take place.
* DNS0 has been removed from the default fallback DNS server list of
systemd-resolved, since it ceased operations.
TPM2 infrastructure:
* systemd-pcrlock no longer locks to PCR 12 by default, since its own
policy description typically ends up in there, as it is passed into a
UKI via a credential, and such credentials are measured into PCR 12.
* The TPM2 infrastructure gained support for additional PCRs
implemented via TPM2 NV Indexes in TPM2_NT_EXTEND mode. These
additional PCRs are called "NvPCRs" in our documentation (even though
they are very much volatile, much like the value of TPM2_NT_EXTEND NV
indexes, from which we inherit the confusing nomenclature). By
introducing NvPCRs the scarcity of PCRs is addressed, which allows us
to measure more resources later without affecting the definition and
current use of the scarce regular PCRs. Note that NvPCRs have
different semantics than PCRs: they are not available pre-userspace
(i.e. initrd userspace creates them and initializes them), including
in the pre-kernel firmware world; moreover, they require an explicit
"anchor" initialization of a privileged per-system secret (in order
to prevent attackers from removing/recreating the backing NV indexes
to reset them). This makes them predictable only if the result of the
anchor measurement is known ahead of time, which will differ on each
installed system. Initialization of defined NvPCRs is done in
systemd-tpm2-setup.service in the initrd. Information about the
initialization of NvPCRs is measured into PCR 9, and finalized by a
separator measurement. The NV index base handle is configurable at
build time via the "tpm2-nvpcr-base" meson setting. It currently
defaults to a value the TCG has shown intent to assign to Linux, but
this has not officially been done yet. systemd-pcrextend and its
Varlink APIs have been extended to optionally measure into an NvPCR
instead of a classic PCR.
* A new service systemd-pcrproduct.service is added which is similar to
systemd-pcrmachine.service but instead of the machine ID
(i.e. /etc/machine-id) measures the product ID (as reported by SMBIOS
or Devicetree). It uses a new NvPCR called "hardware" for this.
* systemd-pcrlock has been updated to generate CEL event log data
covering NvPCRs too.
systemd-analyze:
* systemd-analyze gained a new verb "dlopen-metadata" which can show
the dlopen() weak dependency metadata of an ELF binary that declares
that.
* A new verb "nvpcrs" has been added to systemd-analyze, which lists
NvPCRs with their names and values, similar to the existing "pcrs"
operation which does the same for classic PCRs.
systemd-run/run0:
* run0 gained a new --empower switch. It will invoke a new session with
elevated privileges – without switching to the root user.
Specifically, it sets the full ambient capabilities mask (including
CAP_SYS_ADMIN), which ensures that privileged system calls will
typically be permitted. Moreover, it adds the session processes to
the new "empower" system group, which is respected by polkit and
allows privileged access to most polkit actions. This provides a much
less invasive way to acquire privileges, as it will not change $HOME
or the UID and hence risk creation of files owned by the wrong UID in
the user's home. (Note that --empower might not work in all cases, as
many programs still do access checks purely based on the UID, without
Linux process capabilities or polkit policies having any effect on
them.)
* systemd-run gained support for --root-directory= to invoke the service
in the specified root directory. It also gained --same-root-dir (with
a short switch -R) for invoking the new service in the same root
directory as the caller's. --same-root-dir has also been added to run0.
sd-event:
* sd-event's sd_event_add_child() and sd_event_add_child_pidfd() calls
now support the WNOWAIT flag which tells sd-event to not reap the
child process.
* sd-event gained two new calls sd_event_set_exit_on_idle() and
sd_event_get_exit_on_idle(), which enable automatic exit from the
event loop if no enabled (non-exit) event sources remain.
Other:
* User records gained a new UUID field, and the userdbctl tool gained
the ability to search for user records by UUID, via the new --uuid=
switch. The userdb Varlink API has been extended to allow server-side
searches for UUIDs.
* systemd-sysctl gained a new --inline switch, similar to the switch of
the same name systemd-sysusers already supports.
* systemd-cryptsetup has been updated to understand a new
tpm2-measure-keyslot-nvpcr= option which takes an NvPCR name to
measure information about the used LUKS keyslot into.
systemd-gpt-auto-generator now uses this for a new "cryptsetup"
NvPCR.
* systemd will now ignore configuration file drop-ins suffixed with
".ignore" in most places, similar to how it already ignores files
with suffixes such as ".rpmsave". Unlike those suffixes, ".ignore" is
package manager agnostic.
* systemd-modules-load will now load configured kernel modules in
parallel.
* systemd-integrity-setup now supports HMAC-SHA256, PHMAC-SHA256,
PHMAC-SHA512.
* systemd-stdio-bridge gained a new --quiet option.
* systemd-mountfsd's MountImage() call gained support for explicitly
controlling whether to share dm-verity volumes between images that
have the same root hashes. It also learned support for setting up
bare file system images with separate Verity data files and
signatures.
* journalctl learned a new short switch "-W" for the existing long
switch "--no-hostname".
* system-alloc-{uid,gid}-min are now exported in systemd.pc.
* Incomplete support for musl libc is now available by setting the
"libc" meson option to "musl". Note that systemd compiled with musl
has various limitations: since NSS or equivalent functionality is not
available, nss-systemd, nss-resolve, DynamicUser=, systemd-homed,
systemd-userdbd, the foreign UID ID, unprivileged systemd-nspawn,
systemd-nsresourced, and so on will not work. Also, the usual memory
pressure behaviour of long-running systemd services has no effect on
musl. We also implemented a bunch of shims and workarounds to
support compiling and running with musl. Caveat emptor.
This support for musl is provided without a promise of continued
support in future releases. We'll make the decision based on the
amount of work required to maintain the compatibility layer in
systemd, how many musl-specific bugs are reported, and feedback on
the desirability of this effort provided by users and distributions.
Contributions from: 0x06, Abílio Costa, Alan Brady, Alberto Planas,
Aleksandr Mezin, Alexandru Tocar, Alexis-Emmanuel Haeringer,
Allison Karlitskaya, Andreas Schneider, Andrew Halaney,
Anton Tiurin, Antonio Alvarez Feijoo, Antonio Álvarez Feijoo,
Arian van Putten, Armin Brauns, Armin Wolf, Bastian Almendras,
Charlie Le, Chen Qi, Chris Down, Christian Hesse,
Christoph Anton Mitterer, Colin Walters, Craig McLure,
Daan De Meyer, Daniel Brackenbury, Daniel Foster, Daniel Hast,
Daniel Rusek, Danilo Spinella, David Santamaría Rogado,
David Tardon, Dimitri John Ledkov, Dr. David Alan Gilbert,
Duy Nguyen Van, Emanuele Giuseppe Esposito, Emil Renner Berthing,
Eric Curtin, Erin Shepherd, Evgeny Vereshchagin,
Fco. Javier F. Serrador, Felix Pehla, Fletcher Woodruff, Florian,
Francesco Valla, Franck Bui, Frantisek Sumsal, Gero Schwäricke,
Goffredo Baroncelli, Govind Venugopal, Guido Günther, Haiyue Wang,
Hans de Goede, Henri Aunin, Igor Opaniuk, Ingo Franzki, Itxaka,
Ivan Kruglov, Jelle van der Waa, Jeremy Kerr, Jesse Guo,
Jim Spentzos, Joshua Krusell, João Rodrigues, Justin Kromlinger,
Jörg Behrmann, Kai Lueke, Kai Wohlfahrt, Le_Futuriste,
Lennart Poettering, Luca Boccassi, Lucas Adriano Salles,
Lukáš Nykrýn, Lukáš Zaoral, Managor, Mantas Mikulėnas,
Marc-Antoine Riou, Marcel Leismann, Marcos Alano, Marien Zwart,
Markus Boehme, Martin Hundebøll, Martin Srebotnjak, Masanari Iida,
Matteo Croce, Maximilian Bosch, Michal Sekletár, Mike Gilbert,
Mike Yuan, Miroslav Lichvar, Moisticules, Morgan, Natalie Vock,
Nick Labich, Nick Rosbrook, Nils K, Osama Abdelkader, Oğuz Ersen,
Pascal Bachor, Pasquale van Heumen, Pavel Borecki, Peter Hutterer,
Philip Withnall, Pranay Pawar, Priit Jõerüüt, Quentin Deslandes,
QuickSwift315490, Rafael Fontenelle, Rebecca Cran, Ricardo Salveti,
Ronan Pigott, Ryan Brue, Sebastian Gross, Septatrix, Simon Barth,
Stephanie Wilde-Hobbs, Taylan Kammer, Temuri Doghonadze,
Thomas Blume, Thomas Mühlbacher, Tobias Heider, Vivian Wang,
Xarblu, Yu Watanabe, Zbigniew Jędrzejewski-Szmek, anthisfan, cvlc12,
dgengtek, dramforever, gvenugo3, helpvisa, huyubiao, jouyouyun, jsks,
kanitha chim, lumingzh, n0099, ners, nkraetzschmar, nl6720, q66,
theSillywhat, val4oss, 雪叶
— Edinburgh, 2025/12/17
CHANGES WITH 258:
Incompatible changes:
* Support for cgroup v1 ('legacy' and 'hybrid' hierarchies) has been
removed. cgroup v2 ('unified' hierarchy) will always be mounted
during system bootup and systemd-nspawn container initialization.
* The minimum kernel baseline version has been bumped to v5.4 (released
in 2019), with the recommended version now going up to v5.7. Consult
the README file for a list of required kernel APIs.
* The default access mode of tty/pts device nodes has been changed to
0600, which was 0620 in the older releases, due to general security
concerns about terminals being written to by other users. To restore
the old default access mode, use the '-Dtty-mode=0620' meson build
option. (This effectively means "mesg n" is now the default, rather
than "mesg y", see mesg(1) man page for help.)
* ACLs for device nodes requested by "uaccess" udev tag are now always
applied/updated by systemd-udevd through "uaccess" udev builtin, and
systemd-logind no longer applies/updates ACLs but triggers "change"
uevents to make systemd-udevd apply/update ACLs. Hence, the "uaccess"
udev tag should be set not only on "add" action but also on "change"
action, and it is highly recommended that the rule is applied all
actions except for "remove" action.
Recommended example:
ACTION!="remove", SUBSYSTEM=="hidraw", TAG+="uaccess"
The following example does not work since v258:
ACTION=="add", SUBSYSTEM=="hidraw", TAG+="uaccess"
Udev rules files setting the "uaccess" tag need to have a file name
starting with a number smaller than 73 to be applied.
* systemd-run's --expand-environment= switch, which was disabled
by default when combined with --scope, has been changed to be
enabled by default. This brings cmdline expansion of transient
scopes on par with services.
* systemd-logind PAM sessions that previously were automatically
determined to be of class "background", and which are owned by root
or system accounts, will now automatically be set to class
"background-light" instead. PAM sessions that previously were
automatically determined to be of class "user", and which are owned
by non-root system users, will now automatically be set to class
"user-light" instead. This effectively means that cron jobs or FTP
sessions (i.e. all PAM sessions that have no TTY assigned and neither
are graphical) for system users no longer pull in a service manager
by default. This behaviour can be changed by explicitly setting the
session class (for example via the class= parameter to
pam_systemd.so, or by setting the XDG_SESSION_CLASS environment
variable as input for the service's PAM stack). This change does not
affect graphical sessions, nor does it affect regular users. This is
an incompatible change of sorts, since per-user services will
typically not be available for such PAM sessions of system users.
* systemd-udevd ignores OWNER=/GROUP= settings with a non-system
user/group specified in udev rules files, to avoid device nodes being
owned by a non-system user/group. It is recommended to check udev
rules files with 'udevadm verify' and/or 'udevadm test' commands if
the specified user/group in OWNER=/GROUP= are valid.
Similarly, systemd-networkd refuses User=/Group= settings with a
non-system user/group specified in .netdev files for Tun/Tap
interfaces.
NOTE: this change was partially reverted in v260 and the patch may be
backported to the v259-stable branch.
* systemd-cryptenroll, systemd-repart and systemd-creds no longer
default to locking TPM2 enrollments to the current, literal value of
PCR 7, i.e. the PCR the SecureBoot policy is measured into by the
firmware. This change reflects the fact that nowadays SecureBoot
policies are updated (at least) as frequently as firmware code
(simply because SecureBoot policy updates are typically managed by
fwupd these days). The new default PCR mask for new TPM2 enrollments
is thus empty by default. It is recommended to use managed
systemd-pcrlock policies for binding to PCR 7 instead (as well as
combining such policies with signed policies for PCR 11). Or in other
words, it's recommended to make more use of the logic behind the
--tpm2-public-key=, --tpm2-public-key-pcrs= and --tpm2-pcrlock=
switches of the mentioned tools in place of --tpm2-pcrs=.
* Support for the SystemdOptions EFI variable has been removed.
* Meson options '-Dsplit-usr=', '-Drootlibdir=', '-Drootprefix='
(deprecated in v255), '-Ddefault-hierarchy=' (deprecated in v256),
and '-Dnscd=' (deprecated in v257) have been removed.
* OpenSSL is now the only supported cryptography backend for
systemd-resolved and systemd-importd, and support for gnutls and
gcrypt has been removed. Hence, 'gnutls' setting for the
'-Ddns-over-tls=' meson option has been deprecated. Also, the
'-Dcryptolib=' meson option has been deprecated. They will be removed
in a future release.
* systemd-logind's session tracking, which used to be performed via a
FIFO installed in the client, now uses PIDFDs. The file descriptor
returned by CreateSession() and related calls is therefore unused.
Moreover, the exit of the session leader process will immediately
cause the session to be stopped.
* To work around limitations of X11's keyboard handling systemd's
keyboard mapping hardware database (hwdb.d/60-keyboard.hwdb) so far
mapped the microphone mute and touchpad on/off/toggle keys to the
function keys F20, F21, F22, F23 instead of their correct key codes.
This key code mangling has been removed from udev.
To maintain compatibility with X11 applications that rely on the old
function key code mappings, this mangling has now been added to the
relevant X11 keyboard driver modules. In order to ensure these keys
continue to work, update to xf86-input-evdev >= 2.11.0 and
xf86-input-libinput >= 1.5.0 before updating to systemd >= 258.
* The D-Bus method org.freedesktop.systemd1.StartAuxiliaryScope() has
been removed, which was deprecated since v257.
* systemd-networkd previously emitted the machine ID as chassis ID
through LLDP protocol, but now emits a deterministic ID,
cryptographically derived from the machine ID as chassis ID. If you
want to use the previous behavior, please set
SYSTEMD_LLDP_SEND_MACHINE_ID=1 environment variable for
systemd-networkd.
* Support for the !! command line prefix on ExecStart= lines (and
related) has been removed, and if specified will be ignored. The
concept was supposed to provide compatibility with kernels that
predated the introduction of "ambient" process capabilities. However,
the kernel baseline of the systemd project is now far beyond any
kernels that lacked support for it, hence the prefix serves no
purpose anymore.
* The default keyring for systemd-importd and related tools, shipped
in /usr/lib/systemd/, has been renamed from import-pubring.gpg to
import-pubring.pgp, as it is supported by other PGP tools as well as
GPG. The local keyring /etc/systemd/import-pubring.gpg is still parsed
if present, to preserve backward compatibility.
* Normally, per-user encrypted credentials are decrypted via the the
systemd-creds.socket Varlink service, while the per-system ones are
directly encrypted within the execution context of the intended
service (which hence typically required access to /dev/tpmrm0). This
has been changed: units that enable either PrivateDevices= or use
DeviceAllow=/DevicePolicy= (and thus restrict access to device nodes)
will now also make use of the systemd-creds.socket Varlink
functionality, and will not attempt to decrypt the credentials
in-process (and attempt to try to talk to the TPM for that).
Previously, encrypted credentials for per-system services were
incompatible with PrivateDevices= and resulted in automatic extension
of the DeviceAllow= list. The latter behaviour has been removed.
* The command 'journalctl --follow' now exits with success on
SIGTERM/SIGINT and when the pipe it is writing to is disconnected.
* Support for System V style system state control has been removed:
- The /dev/initctl device node has been removed.
- The initctl, runlevel, and telinit commands have been removed.
- Support for system state control via the init command (e.g.
'init 3') has been removed.
- The units runlevel[0-6].target have been removed.
- The concept of runlevels has been removed, so runlevel transitions
are no longer recorded in the utmp/wtmp databases.
* Support for traditional /forcefsck and /fastboot files to control
execution mode of fsck on boot has been removed from systemd-fsck. To
control the mode, please use the fsck.mode= kernel command line option
or newly introduced fsck.mode credential.
* Support for traditional /forcequotacheck file to control execution
mode of quotacheck on boot has been removed from systemd-quotacheck.
To control the mode, please use the quotacheck.mode= kernel command
line option of newly introduced quotacheck.mode credential.
* systemd-stub v258 requires ukify v257.9 or v258 or newer when
building a UKI. Due to an incompatible change necessary in order to
fix a bug related to embedding a .sbat section larger than 512 bytes,
ukify v257.8 or older will not be able to use systemd-stub v258 or
newer.