systemd-253

[Douglas R. Reno]

New prerelease version

[Douglas R. Reno]

Now -rc3

[Douglas R. Reno]

Sounds like I get to start off 11.4 with a new systemd. :)

[Xi Ruoyao]

Done at r11.3-6-g770e78952.

[Xi Ruoyao]

CHANGES WITH 253:

Announcements of Future Feature Removals and Incompatible Changes:

• We intend to remove cgroup v1 support from systemd release after the end of 2023. If you run services that make explicit use of cgroup v1 features (i.e. the "legacy hierarchy" with separate hierarchies for each controller), please implement compatibility with cgroup v2 (i.e. the "unified hierarchy") sooner rather than later. Most of Linux userspace has been ported over already.

• We intend to remove support for split-usr (/usr mounted separately during boot) and unmerged-usr (parallel directories /bin and /usr/bin, /lib and /usr/lib, etc). This will happen in the second half of 2023, in the first release that falls into that time window. For more details, see: ​https://lists.freedesktop.org/archives/systemd-devel/2022-September/048352.html

• We intend to change behaviour w.r.t. units of the per-user service manager and sandboxing options, so that they work without having to manually enable PrivateUsers= as well, which is not required for system units. To make this work, we will implicitly enable user namespaces (PrivateUsers=yes) when a sandboxing option is enabled in a user unit. The drawback is that system users will no longer be visible (and appear as 'nobody') to the user unit when a sandboxing option is enabled. By definition a sandboxed user unit should run with reduced privileges, so impact should be small. This will remove a great source of confusion that has been reported by users over the years, due to how these options require an extra setting to be manually enabled when used in the per-user service manager, as opposed as to the system service manager. We plan to enable this change in the next release later this year. For more details, see: ​https://lists.freedesktop.org/archives/systemd-devel/2022-December/048682.html

Deprecations and incompatible changes:

• systemctl will now warn when invoked without /proc/ mounted (e.g. when invoked after chroot() into an directory tree without the API mount points like /proc/ being set up.) Operation in such an environment is not fully supported.

• The return value of 'systemctl is-active|is-enabled|is-failed' for unknown units is changed: previously 1 or 3 were returned, but now 4 (EXIT_PROGRAM_OR_SERVICES_STATUS_UNKNOWN) is used as documented.

• 'udevadm hwdb' subcommand is deprecated and will emit a warning. systemd-hwdb (added in 2014) should be used instead.

• 'bootctl --json' now outputs a single JSON array, instead of a stream of newline-separated JSON objects.

• Udev rules in 60-evdev.rules have been changed to load hwdb properties for all modalias patterns. Previously only the first matching pattern was used. This could change what properties are assigned if the user has more and less specific patterns that could match the same device, but it is expected that the change will have no effect for most users.

• systemd-networkd-wait-online exits successfully when all interfaces are ready or unmanaged. Previously, if neither '--any' nor '--interface=' options were used, at least one interface had to be in configured state. This change allows the case where systemd-networkd is enabled, but no interfaces are configured, to be handled gracefully. It may occur in particular when a different network manager is also enabled and used.

• Some compatibility helpers were dropped: EmergencyAction= in the user manager, as well as measuring kernel command line into PCR 8 in systemd-stub, along with the -Defi-tpm-pcr-compat compile-time option.

• The '-Dupdate-helper-user-timeout=' build-time option has been renamed to '-Dupdate-helper-user-timeout-sec=', and now takes an integer as parameter instead of a string.

• The DDI image dissection logic (which backs RootImage= in service unit files, the --image= switch in various tools such as systemd-nspawn, as well as systemd-dissect) will now only mount file systems of types btrfs, ext4, xfs, erofs, squashfs, vfat. This list can be overridden via the $SYSTEMD_DISSECT_FILE_SYSTEMS environment variable. These file systems are fairly well supported and maintained in current kernels, while others are usually more niche, exotic or legacy and thus typically do not receive the same level of security support and fixes.

• The default per-link multicast DNS mode is changed to "yes" (that was previously "no"). As the default global multicast DNS mode has been "yes" (but can be changed by the build option), now the multicast DNS is enabled on all links by default. You can disable the multicast DNS on all links by setting MulticastDNS= in resolved.conf, or on an interface by calling "resolvectl mdns INTERFACE no".

New components:

• A tool 'ukify' tool to build, measure, and sign Unified Kernel Images (UKIs) has been added. This replaces functionality provided by 'dracut --uefi' and extends it with automatic calculation of PE file offsets, insertion of signed PCR policies generated by systemd-measure, support for initrd concatenation, signing of the embedded Linux image and the combined image with sbsign, and heuristics to autodetect the kernel uname and verify the splash image.

Changes in systemd and units:

• A new service type Type=notify-reload is defined. When such a unit is reloaded a UNIX process signal (typically SIGHUP) is sent to the main service process. The manager will then wait until it receives a "RELOADING=1" followed by a "READY=1" notification from the unit as response (via sd_notify()). Otherwise, this type is the same as Type=notify. A new setting ReloadSignal= may be used to change the signal to send from the default of SIGHUP.

user@.service, systemd-networkd.service, systemd-udevd.service, and systemd-logind have been updated to this type.

• Initrd environments which are not on a pure memory file system (e.g. overlayfs combination as opposed to tmpfs) are now supported. With this change, during the initrd → host transition ("switch root") systemd will erase all files of the initrd only when the initrd is backed by a memory file system such as tmpfs.

• New per-unit MemoryZSwapMax= option has been added to configure memory.zswap.max cgroup properties (the maximum amount of zswap used).

• A new LogFilterPatterns= option has been added for units. It may be used to specify accept/deny regular expressions for log messages generated by the unit, that shall be enforced by systemd-journald. Rejected messages are neither stored in the journal nor forwarded. This option may be used to suppress noisy or uninteresting messages from units.

• The manager has a new org.freedesktop.systemd1.Manager.GetUnitByPIDFD() D-Bus method to query process ownership via a PIDFD, which is more resilient against PID recycling issues.

• Scope units now support OOMPolicy=. Login session scopes default to OOMPolicy=continue, allowing login scopes to survive the OOM killer terminating some processes in the scope.

• systemd-fstab-generator now supports x-systemd.makefs option for /sysroot/ (in the initrd).

• The maximum rate at which daemon reloads are executed can now be limited with the new ReloadLimitIntervalSec=/ReloadLimitBurst= options. (Or the equivalent on the kernel command line: systemd.reload_limit_interval_sec=/systemd.reload_limit_burst=). In addition, systemd now logs the originating unit and PID when a reload request is received over D-Bus.

• When enabling a swap device systemd will now reinitialize the device when the page size of the swap space does not match the page size of the running kernel. Note that this requires the 'swapon' utility to provide the '--fixpgsz' option, as implemented by util-linux, and it is not supported by busybox at the time of writing.

• systemd now executes generator programs in a mount namespace "sandbox" with most of the file system read-only and write access restricted to the output directories, and with a temporary /tmp/ mount provided. This provides a safeguard against programming errors in the generators, but also fixes here-docs in shells, which previously didn't work in early boot when /tmp/ wasn't available yet. (This feature has no security implications, because the code is still privileged and can trivially exit the sandbox.)

• The system manager will now parse a new "vmm.notify_socket" system credential, which may be supplied to a VM via SMBIOS. If found, the manager will send a "READY=1" notification on the specified socket after boot is complete. This allows readiness notification to be sent from a VM guest to the VM host over a VSOCK socket.

• The sample PAM configuration file for systemd-user@.service now includes a call to pam_namespace. This puts children of user@.service in the expected namespace. (Many distributions replace their file with something custom, so this change has limited effect.)

• A new environment variable $SYSTEMD_DEFAULT_MOUNT_RATE_LIMIT_BURST can be used to override the mount units burst late limit for parsing '/proc/self/mountinfo', which was introduced in v249. Defaults to 5.

• Drop-ins for init.scope changing control group resource limits are now applied, while they were previously ignored.

• New build-time configuration options '-Ddefault-timeout-sec=' and '-Ddefault-user-timeout-sec=' have been added, to let distributions choose the default timeout for starting/stopping/aborting system and user units respectively.

• Service units gained a new setting OpenFile= which may be used to open arbitrary files in the file system (or connect to arbitrary AF_UNIX sockets in the file system), and pass the open file descriptor to the invoked process via the usual file descriptor passing protocol. This is useful to give unprivileged services access to select files which have restrictive access modes that would normally not allow this. It's also useful in case RootDirectory= or RootImage= is used to allow access to files from the host environment (which is after all not visible from the service if these two options are used.)

Changes in udev:

• The new net naming scheme "v253" has been introduced. In the new scheme, ID_NET_NAME_PATH is also set for USB devices not connected via a PCI bus. This extends the coverage of predictable interface names in some embedded systems.

The "amba" bus path is now included in ID_NET_NAME_PATH, resulting in a more informative path on some embedded systems.

• Partition block devices will now also get symlinks in /dev/disk/by-diskseq/<seq>-part<n>, which may be used to reference block device nodes via the kernel's "diskseq" value. Previously those symlinks were only created for the main block device.

• A new operator '-=' is supported for SYMLINK variables. This allows symlinks to be unconfigured even if an earlier rule added them.

• 'udevadm --trigger --settle' now also works for network devices that are being renamed.

Changes in sd-boot, bootctl, and the Boot Loader Specification:

• systemd-boot now passes its random seed directly to the kernel's RNG via the LINUX_EFI_RANDOM_SEED_TABLE_GUID configuration table, which means the RNG gets seeded very early in boot before userspace has started.

• systemd-boot will pass a disk-backed random seed – even when secure boot is enabled – if it can additionally get a random seed from EFI itself (via EFI's RNG protocol), or a prior seed in LINUX_EFI_RANDOM_SEED_TABLE_GUID from a preceding bootloader.

• systemd-boot-system-token.service was renamed to systemd-boot-random-seed.service and extended to always save a random seed to ESP on every boot when a compatible boot loader is used. This allows a refreshed random seed to be used in the boot loader.

• systemd-boot handles various seed inputs using a domain- and field-separated hashing scheme.

• systemd-boot's 'random-seed-mode' option has been removed. A system token is now always required to be present for random seeds to be used.

• systemd-boot now supports being loaded from other locations than the ESP, for example for direct kernel boot under QEMU or when embedded into the firmware.

• systemd-boot now parses SMBIOS information to detect virtualization. This information is used to skip some warnings which are not useful in a VM and to conditionalize other aspects of behaviour.

• systemd-boot now supports a new 'if-safe' mode that will perform UEFI Secure Boot automated certificate enrollment from the ESP only if it is considered 'safe' to do so. At the moment 'safe' means running in a virtual machine.

• systemd-stub now processes random seeds in the same way as systemd-boot already does, in case a unified kernel image is being used from a different bootloader than systemd-boot, or without any boot load at all.

• bootctl will now generate a system token on all EFI systems, even virtualized ones, and is activated in the case that the system token is missing from either sd-boot and sd-stub booted systems.

• bootctl now implements two new verbs: 'kernel-identify' prints the type of a kernel image file, and 'kernel-inspect' provides information about the embedded command line and kernel version of UKIs.

• bootctl now honours $KERNEL_INSTALL_CONF_ROOT with the same meaning as for kernel-install.

• The JSON output of "bootctl list" will now contain two more fields: isDefault and isSelected are boolean fields set to true on the default and currently booted boot menu entries.

• bootctl gained a new verb "unlink" for removing a boot loader entry type #1 file from disk in a safe and robust way.

• bootctl also gained a new verb "cleanup" that automatically removes all files from the ESP's and XBOOTLDR's "entry-token" directory, that is not referenced anymore by any installed Type #1 boot loader specification entry. This is particularly useful in environments where a large number of entries reference the same or partly the same resources (for example, for snapshot-based setups).

Changes in kernel-install:

• A new "installation layout" can be configured as layout=uki. With this setting, a Boot Loader Specification Type#1 entry will not be created. Instead, a new kernel-install plugin 90-uki-copy.install will copy any .efi files from the staging area into the boot partition. A plugin to generate the UKI .efi file must be provided separately.

Changes in systemctl:

• 'systemctl reboot' has dropped support for accepting a positional argument as the argument to the reboot(2) syscall. Please use the --reboot-argument= option instead.

• 'systemctl disable' will now warn when called on units without install information. A new --no-warn option has been added that silences this warning.

• New option '--drop-in=' can be used to tell 'systemctl edit' the name of the drop-in to edit. (Previously, 'override.conf' was always used.)

• 'systemctl list-dependencies' now respects --type= and --state=.

• 'systemctl kexec' now supports XEN VMM environments.

• 'systemctl edit' will now tell the invoked editor to jump into the first line with actual unit file data, skipping over synthesized comments.

Changes in systemd-networkd and related tools:

• The [DHCPv4] section in .network file gained new SocketPriority= setting that assigns the Linux socket priority used by the DHCPv4 raw socket. This may be used in conjunction with the EgressQOSMaps=setting in [VLAN] section of .netdev file to send the desired ethernet 802.1Q frame priority for DHCPv4 initial packets. This cannot be achieved with netfilter mangle tables because of the raw socket bypass.

• The [DHCPv4] and [IPv6AcceptRA] sections in .network file gained a new QuickAck= boolean setting that enables the TCP quick ACK mode for the routes configured by the acquired DHCPv4 lease or received router advertisements (RAs).

• The RouteMetric= option (for DHCPv4, DHCPv6, and IPv6 advertised routes) now accepts three values, for high, medium, and low preference of the router (which can be set with the RouterPreference=) setting.

• systemd-networkd-wait-online now supports matching via alternative interface names.

• The [DHCPv6] section in .network file gained new SendRelease= setting which enables the DHCPv6 client to send release when it stops. This is the analog of the [DHCPv4] SendRelease= setting. It is enabled by default.

• If the Address= setting in [Network] or [Address] sections in .network specified without its prefix length, then now systemd-networkd assumes /32 for IPv4 or /128 for IPv6 addresses.

• networkctl shows network and link file dropins in status output.

Changes in systemd-dissect:

• systemd-dissect gained a new option --list, to print the paths of all files and directories in a DDI.

• systemd-dissect gained a new option --mtree, to generate a file manifest compatible with BSD mtree(5) of a DDI

• systemd-dissect gained a new option --with, to execute a command with the specified DDI temporarily mounted and used as working directory. This is for example useful to convert a DDI to "tar" simply by running it within a "systemd-dissect --with" invocation.

• systemd-dissect gained a new option --discover, to search for Discoverable Disk Images (DDIs) in well-known directories of the system. This will list machine, portable service and system extension disk images.

• systemd-dissect now understands 2nd stage initrd images stored as a Discoverable Disk Image (DDI).

• systemd-dissect will now display the main UUID of GPT DDIs (i.e. the disk UUID stored in the GPT header) among the other data it can show.

• systemd-dissect gained a new --in-memory switch to operate on an in-memory copy of the specified DDI file. This is useful to access a DDI with write access without persisting any changes. It's also useful for accessing a DDI without keeping the originating file system busy.

• The DDI dissection logic will now automatically detect the intended sector size of disk images stored in files, based on the GPT partition table arrangement. Loopback block devices for such DDIs will then be configured automatically for the right sector size. This is useful to make dealing with modern 4K sector size DDIs fully automatic. The systemd-dissect tool will now show the detected sector size among the other DDI information in its output.

Changes in systemd-repart:

• systemd-repart gained new options --include-partitions= and --exclude-partitions= to filter operation on partitions by type UUID. This allows systemd-repart to be used to build images in which the type of one partition is set based on the contents of another partition (for example when the boot partition shall include a verity hash of the root partition).

• systemd-repart also gained a --defer-partitions= option that is similar to --exclude-partitions=, but the size of the partition is still taken into account when sizing partitions, but without populating it.

• systemd-repart gained a new --sector-size= option to specify what sector size should be used when an image is created.

• systemd-repart now supports generating erofs file systems via CopyFiles= (a read-only file system similar to squashfs).

• The Minimize= option was extended to accept "best" (which means the most minimal image possible, but may require multiple attempts) and "guess" (which means a reasonably small image).

• The systemd-growfs binary now comes with a regular unit file template systemd-growfs@.service which can be instantiated directly for any desired file system. (Previously, the unit was generated dynamically by various generators, but no regular unit file template was available.)

Changes in journal tools:

• Various systemd tools will append extra fields to log messages when in debug mode, or when SYSTEMD_ENABLE_LOG_CONTEXT=1 is set. Currently this includes information about D-Bus messages when sd-bus is used, e.g. DBUS_SENDER=, DBUS_DESTINATION=, and DBUS_PATH=, and information about devices when sd-device is used, e.g. DEVNAME= and DRIVER=. Details of what is logged and when are subject to change.

• The systemd-journald-audit.socket can now be disabled via the usual "systemctl disable" mechanism to stop collection of audit messages. Please note that it is not enabled statically anymore and must be handled by the preset/enablement logic in package installation scripts.

• New options MaxUse=, KeepFree=, MaxFileSize=, and MaxFiles= can be used to curtail disk use by systemd-journal-remote. This is similar to the options supported by systemd-journald.

Changes in systemd-cryptenroll, systemd-cryptsetup, and related components:

• When enrolling new keys systemd-cryptenroll now supports unlocking via FIDO2 tokens (option --unlock-fido2-device=). Previously, a password was strictly required to be specified.

• systemd-cryptsetup now supports pre-flight requests for FIDO2 tokens (except for tokens with user verification, UV) to identify tokens before authentication. Multiple FIDO2 tokens can now be enrolled at the same time, and systemd-cryptsetup will automatically select one that corresponds to one of the available LUKS key slots.

• systemd-cryptsetup now supports new options tpm2-measure-bank= and tpm2-measure-pcr= in crypttab(5). These allow specifying the TPM2 PCR bank and number into which the volume key should be measured. This is automatically enabled for the encrypted root volume discovered and activated by systemd-gpt-auto-generator.

• systemd-gpt-auto-generator mounts the ESP and XBOOTLDR partitions with "noexec,nosuid,nodev".

• systemd-gpt-auto-generator will now honour the rootfstype= and rootflags= kernel command line switches for root file systems it discovers, to match behaviour in case an explicit root fs is specified via root=.

• systemd-pcrphase gained new options --machine-id and --file-system= to measure the machine-id and mount point information into PCR 15. New service unit files systemd-pcrmachine.service and systemd-pcrfs@.service have been added that invoke the tool with these switches during early boot.

• systemd-pcrphase gained a --graceful switch will make it exit cleanly with a success exit code even if no TPM device is detected.

• systemd-cryptenroll now stores the user-supplied PIN with a salt, making it harder to brute-force.

Changes in other tools:

• systemd-homed gained support for luksPbkdfForceIterations (the intended number of iterations for the PBKDF operation on LUKS).

• Environment variables $SYSTEMD_HOME_MKFS_OPTIONS_BTRFS, $SYSTEMD_HOME_MKFS_OPTIONS_EXT4, and $SYSTEMD_HOME_MKFS_OPTIONS_XFS may now be used to specify additional arguments for mkfs when systemd-homed formats a file system.

• systemd-hostnamed now exports the contents of /sys/class/dmi/id/bios_vendor and /sys/class/dmi/id/bios_date via two new D-Bus properties: FirmwareVendor and FirmwareDate. This allows unprivileged code to access those values.

systemd-hostnamed also exports the SUPPORT_END= field from os-release(5) as OperatingSystemSupportEnd. hostnamectl make uses of this to show the status of the installed system.

• systemd-measure gained an --append= option to sign multiple phase paths with different signing keys. This allows secrets to be accessible only in certain parts of the boot sequence. Note that 'ukify' provides similar functionality in a more accessible form.

• systemd-timesyncd will now write a structured log message with MESSAGE_ID set to SD_MESSAGE_TIME_BUMP when it bumps the clock based on a on-disk timestamp, similarly to what it did when reaching synchronization via NTP.

• systemd-timesyncd will now update the on-disk timestamp file on each boot at least once, making it more likely that the system time increases in subsequent boots.

• systemd-vconsole-setup gained support for system/service credentials: vconsole.keymap/vconsole.keymap_toggle and vconsole.font/vconsole.font_map/vconsole.font_unimap are analogous the similarly-named options in vconsole.conf.

• systemd-localed will now save the XKB keyboard configuration to /etc/vconsole.conf, and also read it from there with a higher preference than the /etc/X11/xorg.conf.d/00-keyboard.conf config file. Previously, this information was stored in the former file in converted form, and only in latter file in the original form. Tools which want to access keyboard configuration can now do so from a standard location.

• systemd-resolved gained support for configuring the nameservers and search domains via kernel command line (nameserver=, domain=) and credentials (network.dns, network.search_domains).

• systemd-resolved will now synthesize host names for the DNS stub addresses it supports. Specifically when "_localdnsstub" is resolved, 127.0.0.53 is returned, and if "_localdnsproxy" is resolved 127.0.0.54 is returned.

• systemd-notify will now send a "RELOADING=1" notification when called with --reloading, and "STOPPING=1" when called with --stopping. This can be used to implement notifications from units where it's easier to call a program than to use the sd-daemon library.

• systemd-analyze's 'plot' command can now output its information in JSON, controlled via the --json= switch. Also, new --table, and --no-legend options have been added.

• 'machinectl enable' will now automatically enable machines.target unit in addition to adding the machine unit to the target.

Similarly, 'machinectl start|stop' gained a --now option to enable or disable the machine unit when starting or stopping it.

• systemd-sysusers will now create /etc/ if it is missing.

• systemd-sleep 'HibernateDelaySec=' setting is changed back to pre-v252's behaviour, and a new 'SuspendEstimationSec=' setting is added to provide the new initial value for the new automated battery estimation functionality. If 'HibernateDelaySec=' is set to any value, the automated estimate (and thus the automated hibernation on low battery to avoid data loss) functionality will be disabled.

• Default tmpfiles.d/ configuration will now automatically create credentials storage directory '/etc/credstore/' with the appropriate, secure permissions. If '/run/credstore/' exists, its permissions will be fixed too in case they are not correct.

Changes in libsystemd and shared code:

• sd-bus gained new convenience functions sd_bus_emit_signal_to(), sd_bus_emit_signal_tov(), and sd_bus_message_new_signal_to().

• sd-id128 functions now return -EUCLEAN (instead of -EIO) when the 128bit ID in files such as /etc/machine-id has an invalid format. They also accept NULL as output parameter in more places, which is useful when the caller only wants to validate the inputs and does not need the output value.

• sd-login gained new functions sd_pidfd_get_session(), sd_pidfd_get_owner_uid(), sd_pidfd_get_unit(), sd_pidfd_get_user_unit(), sd_pidfd_get_slice(), sd_pidfd_get_user_slice(), sd_pidfd_get_machine_name(), and sd_pidfd_get_cgroup(), that are analogous to sd_pid_get_*(), but accept a PIDFD instead of a PID.

• sd-path (and systemd-path) now export four new paths: SD_PATH_SYSTEMD_SYSTEM_ENVIRONMENT_GENERATOR, SD_PATH_SYSTEMD_USER_ENVIRONMENT_GENERATOR, SD_PATH_SYSTEMD_SEARCH_SYSTEM_ENVIRONMENT_GENERATOR, and SD_PATH_SYSTEMD_SEARCH_USER_ENVIRONMENT_GENERATOR,

• sd_notify() now supports AF_VSOCK as transport for notification messages (in addition to the existing AF_UNIX support). This is enabled if $NOTIFY_SOCKET is set in a "vsock:CID:port" format.

• Detection of chroot() environments now works if /proc/ is not mounted. This affects systemd-detect-virt --chroot, but also means that systemd tools will silently skip various operations in such an environment.

• "Lockheed Martin Hardened Security for Intel Processors" (HS SRE) virtualization is now detected.

Changes in the build system:

• Standalone variants of systemd-repart and systemd-shutdown may now be built (if -Dstandalone=true).

• systemd-ac-power has been moved from /usr/lib/ to /usr/bin/, to, for example, allow scripts to conditionalize execution on AC power supply.

• The libp11kit library is now loaded through dlopen(3).

Changes in the documentation:

• Specifications that are not closely tied to systemd have moved to ​https://uapi-group.org/specifications/: the Boot Loader Specification and the Discoverable Partitions Specification.

[Xi Ruoyao]

I'll add the short description of ukify later.

[Bruce Dubbs]

Milestone renamed