Lançado o Systemd 252


Em meio a comemoração do Halloween é lançado o Systemd 252 repleto de correções e melhorias.

Lançado o Systemd 252

systemd é um conjunto de softwares que fornecem blocos de construção fundamentais para um sistema operacional Linux. Entre outros recursos, ele inclui o systemd "System and Service Manager", um sistema init usado para inicializar o espaço do usuário e gerenciar processos do sistema após a inicialização. Fonte

 

Lançado em 31 de Outubro de 2022.


 São dezenas de correções e melhorias que você confere no quadro abaixo.

Announcements of Future Feature Removals: * 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 Compatibility Breaks: * ConditionKernelVersion= checks that use the '=' or '!=' operators will now do simple string comparisons (instead of version comparisons á la stverscmp()). Version comparisons are still done for the ordering operators '<', '>', '<=', '>='. Moreover, if no operator is specified, a shell-style glob match is now done. This creates a minor incompatibility compared to older systemd versions when the '*', '?', '[', ']' characters are used, as these will now match as shell globs instead of literally. Given that kernel version strings typically do not include these characters we expect little breakage through this change. * The service manager will now read the SELinux label used for SELinux access checks from the unit file at the time it loads the file. Previously, the label would be read at the moment of the access check, which was problematic since at that time the unit file might already have been updated or removed. New Features: * systemd-measure is a new tool for calculating and signing expected TPM2 PCR values for a given unified kernel image (UKI) booted via sd-stub. The public key used for the signature and the signed expected PCR information can be embedded inside the UKI. This information can be extracted from the UKI by external tools and code in the image itself and is made available to userspace in the booted kernel. systemd-cryptsetup, systemd-cryptenroll, and systemd-creds have been updated to make use of this information if available in the booted kernel: when locking an encrypted volume/credential to the TPM systemd-cryptenroll/systemd-creds will use the public key to bind the volume/credential to any kernel that carries PCR information signed by the same key pair. When unlocking such volumes/credentials systemd-cryptsetup/systemd-creds will use the signature embedded in the booted UKI to gain access. Binding TPM-based disk encryption to public keys/signatures of PCR values — instead of literal PCR values — addresses the inherent "brittleness" of traditional PCR-bound TPM disk encryption schemes: disks remain accessible even if the UKI is updated, without any TPM specific preparation during the OS update — as long as each UKI carries the necessary PCR signature information. Net effect: if you boot a properly prepared kernel, TPM-bound disk encryption now defaults to be locked to kernels which carry PCR signatures from the same key pair. Example: if a hypothetical distro FooOS prepares its UKIs like this, TPM-based disk encryption is now – by default – bound to only FooOS kernels, and encrypted volumes bound to the TPM cannot be unlocked on kernels from other sources. (But do note this behaviour requires preparation/enabling in the UKI, and of course users can always enroll non-TPM ways to unlock the volume.) * systemd-pcrphase is a new tool that is invoked at six places during system runtime, and measures additional words into TPM2 PCR 11, to mark milestones of the boot process. This allows binding access to specific TPM2-encrypted secrets to specific phases of the boot process. (Example: LUKS2 disk encryption key only accessible in the initrd, but not later.) Changes in systemd itself, i.e. the manager and units * The cpu controller is delegated to user manager units by default, and CPUWeight= settings are applied to the top-level user slice units (app.slice, background.slice, session.slice). This provides a degree of resource isolation between different user services competing for the CPU. * Systemd can optionally do a full preset in the "first boot" condition (instead of just enable-only). This behaviour is controlled by the compile-time option -Dfirst-boot-full-preset. Right now it defaults to 'false', but the plan is to switch it to 'true' for the subsequent release. * Drop-ins are now allowed for transient units too. * Systemd will set the taint flag 'support-ended' if it detects that the OS image is past its end-of-support date. This date is declared in a new /etc/os-release field SUPPORT_END= described below. * Two new settings ConditionCredential= and AssertCredential= can be used to skip or fail units if a certain system credential is not provided. * ConditionMemory= accepts size suffixes (K, M, G, T, …). * DefaultSmackProcessLabel= can be used in system.conf and user.conf to specify the SMACK security label to use when not specified in a unit file. * DefaultDeviceTimeoutSec= can be used in system.conf and user.conf to specify the default timeout when waiting for device units to activate. * C.UTF-8 is used as the default locale if nothing else has been configured. * [Condition|Assert]Firmware= have been extended to support certain SMBIOS fields. For example ConditionFirmware=smbios-field(board_name = "Custom Board") conditionalizes the unit to run only when /sys/class/dmi/id/board_name contains "Custom Board" (without the quotes). * ConditionFirstBoot= now correctly evaluates as true only during the boot phase of the first boot. A unit executed later, after booting has completed, will no longer evaluate this condition as true. * Socket units will now create sockets in the SELinuxContext= of the associated service unit, if any. * Boot phase transitions (start initrd → exit initrd → boot complete → shutdown) will be measured into TPM2 PCR 11, so that secrets can be bound to a specific runtime phase. E.g.: a LUKS encryption key can be unsealed only in the initrd. * Service credentials (i.e. SetCredential=/LoadCredential=/…) will now also be provided to ExecStartPre= processes. * Various units are now correctly ordered against initrd-switch-root.target where previously a conflict without ordering was configured. A stop job for those units would be queued, but without the ordering it could be executed only after initrd-switch-root.service, leading to units not being restarted in the host system as expected. * In order to fully support the IPMI watchdog driver, which has not yet been ported to the new common watchdog device interface, /dev/watchdog0 will be tried first and systemd will silently fallback to /dev/watchdog if it is not found. * New watchdog-related D-Bus properties are now published by systemd: WatchdogDevice, WatchdogLastPingTimestamp, WatchdogLastPingTimestampMonotonic. * At shutdown, API virtual files systems (proc, sys, etc.) will be unmounted lazily. * At shutdown, systemd will now log about processes blocking unmounting of file systems. * A new meson build option 'clock-valid-range-usec-max' was added to allow disabling system time correction if RTC returns a timestamp far in the future. * Propagated restart jobs will no longer be discarded while a unit is activating. * PID 1 will now import system credentials from SMBIOS Type 11 fields ("OEM vendor strings"), in addition to qemu_fwcfg. This provides a simple, fast and generic path for supplying credentials to a VM, without involving external tools such as cloud-init/ignition. * The CPUWeight= setting of unit files now accepts a new special value "idle", which configures "idle" level scheduling for the unit. * Service processes that are activated due to a .timer or .path unit triggering will now receive information about this via environment variables. Note that this is information is lossy, as activation might be coalesced and only one of the activating triggers will be reported. This is hence more suited for debugging or tracing rather than for behaviour decisions. * The riscv_flush_icache(2) system call has been added to the list of system calls allowed by default when SystemCallFilter= is used. * The selinux context derived from the target executable, instead of 'init_t' used for the manager itself, is now used when creating listening sockets for units that specify SELinuxContextFromNet=yes. Changes in sd-boot, bootctl, and the Boot Loader Specification: * The Boot Loader Specification has been cleaned up and clarified. Various corner cases in version string comparisons have been fixed (e.g. comparisons for empty strings). Boot counting is now part of the main specification. * New PCRs measurements are performed during boot: PCR 11 for the the kernel+initrd combo, PCR 13 for any sysext images. If a measurement took place this is now reported to userspace via the new StubPcrKernelImage and StubPcrInitRDSysExts EFI variables. * As before, systemd-stub will measure kernel parameters and system credentials into PCR 12. It will now report this fact via the StubPcrKernelParameters EFI variable to userspace. * The UEFI monotonic boot counter is now included in the updated random seed file maintained by sd-boot, providing some additional entropy. * sd-stub will use LoadImage/StartImage to execute the kernel, instead of arranging the image manually and jumping to the kernel entry point. sd-stub also installs a temporary UEFI SecurityOverride to allow the (unsigned) nested image to be booted. This is safe because the outer (signed) stub+kernel binary must have been verified before the stub was executed. * Booting in EFI mixed mode (a 64-bit kernel over 32-bit UEFI firmware) is now supported by sd-boot. * bootctl gained a bunch of new options: --all-architectures to install binaries for all supported EFI architectures, --root= and --image= options to operate on a directory or disk image, and --install-source= to specify the source for binaries to install, --efi-boot-option-description= to control the name of the boot entry. * The sd-boot stub exports a StubFeatures flag, which is used by bootctl to show features supported by the stub that was used to boot. * The PE section offsets that are used by tools that assemble unified kernel images have historically been hard-coded. This may lead to overlapping PE sections which may break on boot. The UKI will now try to detect and warn about this. Any tools that assemble UKIs must update to calculate these offsets dynamically. Future sd-stub versions may use offsets that will not work with the currently used set of hard-coded offsets! * sd-stub now accepts (and passes to the initrd and then to the full OS) new PE sections '.pcrsig' and '.pcrkey' that can be used to embed signatures of expected PCR values, to allow sealing secrets via the TPM2 against pre-calculated PCR measurements. Changes in the hardware database: * 'systemd-hwdb query' now supports the --root= option. Changes in systemctl: * systemctl now supports --state= and --type= options for the 'show' and 'status' verbs. * systemctl gained a new verb 'list-automounts' to list automount points. * systemctl gained support for a new --image= switch to be able to operate on the specified disk image (similar to the existing --root= which operates relative to some directory). Changes in systemd-networkd: * networkd can set Linux NetLabel labels for integration with the network control in security modules via a new NetLabel= option. * The RapidCommit= is (re-)introduced to enable faster configuration via DHCPv6 (RFC 3315). * networkd gained a new option TCPCongestionControlAlgorithm= that allows setting a per-route TCP algorithm. * networkd gained a new option KeepFileDescriptor= to allow keeping a reference (file descriptor) open on TUN/TAP interfaces, which is useful to avoid link flaps while the underlying service providing the interface is being serviced. * RouteTable= now also accepts route table names. Changes in systemd-nspawn: * The --bind= and --overlay= options now support relative paths. * The --bind= option now supports a 'rootidmap' value, which will use id-mapped mounts to map the root user inside the container to the owner of the mounted directory on the host. Changes in systemd-resolved: * systemd-resolved now persists DNSOverTLS in its state file too. This fixes a problem when used in combination with NetworkManager, which sends the setting only once, causing it to be lost if resolved was restarted at any point. * systemd-resolved now exposes a varlink socket at /run/systemd/resolve/io.systemd.Resolve.Monitor, accessible only for root. Processed DNS requests in a JSON format will be published to any clients connected to this socket. resolvectl gained a 'monitor' verb to make use of this. * systemd-resolved now treats unsupported DNSSEC algorithms as INSECURE instead of returning SERVFAIL, as per RFC: https://datatracker.ietf.org/doc/html/rfc6840#section-5.2 * OpenSSL is the default crypto backend for systemd-resolved. (gnutls is still supported.) Changes in libsystemd and other libraries: * libsystemd now exports sd_bus_error_setfv() (a convenience function for setting bus errors), sd_id128_string_equal (a convenience function for 128bit ID string comparisons), and sd_bus_message_read_strv_extend() (a function to incrementally read string arrays). * libsystemd now exports sd_device_get_child_first()/_next() as a high-level interface for enumerating child devices. It also supports sd_device_new_child() for opening a child device given a device object. * libsystemd now exports sd_device_monitor_set()/get_description() which allow setting a custom description that will be used in log messages by sd_device_monitor*. * Private shared libraries (libsystemd-shared-nnn.so, libsystemd-core-nnn.so) are now installed into arch-specific directories to allow multi-arch installs. * A new sd-gpt.h header is now published, listing GUIDs from the Discoverable Partitions specification. For more details see: https://systemd.io/DISCOVERABLE_PARTITIONS/ * A new function sd_hwdb_new_from_path() has been added to open a hwdb database given an explicit path to the file. * The signal number argument to sd_event_add_signal() now can now be ORed with the SD_EVENT_SIGNAL_PROCMASK flag, causing sigprocmask() to be automatically invoked to block the specified signal. This is useful to simplify invocations as the caller doesn't have to do this manually. * A new convenience call sd_event_set_signal_exit() has been added to sd-event to set up signal handling so that the event loop automatically terminates cleanly on SIGTERM/SIGINT. Changes in other components: * systemd-sysusers, systemd-tmpfiles, and systemd-sysctl configuration can now be provided via the credential mechanism. * systemd-analyze gained a new verb 'compare-versions' that implements comparisons for versions strings (similarly to 'rpmdev-vercmp' and 'dpkg --compare-versions'). * 'systemd-analyze dump' is extended to accept glob patterns for unit names to limit the output to matching units. * tmpfiles.d/ lines can read file contents to write from a credential. The new modifier char '^' is used to specify that the argument is a credential name. This mechanism is used to automatically populate /etc/motd, /etc/issue, and /etc/hosts from credentials. * tmpfiles.d/ may now be configured to avoid changing uid/gid/mode of an inode if the specification is prefixed with ':' and the inode already exists. * Default tmpfiles.d/ configuration now carries a line to automatically use an 'ssh.authorized_keys.root' credential if provided to set up the SSH authorized_keys file for the root user. * systemd-tmpfiles will now gracefully handle absent source of "C" copy lines. * tmpfiles.d/ F/w lines now optionally permit encoding of the payload in base64. This is useful to write arbitrary binary data into files. * The pkgconfig and rpm macros files now export the directory for user units as 'user_tmpfiles_dir' and '%_user_tmpfilesdir'. * Detection of Apple Virtualization and detection of Parallels and KubeVirt virtualization on non-x86 archs have been added. * os-release gained a new field SUPPORT_END=YYYY-MM-DD to inform the user when their system will become unsupported. * When performing suspend-then-hibernate, the system will estimate the discharge rate and use that to set the delay until hibernation and hibernate immediately instead of suspending when running from a battery and the capacity is below 5%. * systemd-sysctl gained a --strict option to fail when a sysctl setting is unknown to the kernel. * machinectl supports --force for the 'copy-to' and 'copy-from' verbs. * coredumpctl gained the --root and --image options to look for journal files under the specified root directory, image, or block device. * 'journalctl -o' and similar commands now implement a new output mode "short-delta". It is similar to "short-monotonic", but also shows the time delta between subsequent messages. * journalctl now respects the --quiet flag when verifying consistency of journal files. * Journal log messages gained a new implicit field _RUNTIME_SCOPE= that will indicate whether a message was logged in the 'initrd' phase or in the 'system' phase of the boot process. * Journal files gained a new compatibility flag 'HEADER_INCOMPATIBLE_COMPACT'. Files with this flag implement changes to the storage format that allow reducing size on disk. As with other compatibility flags, older journalctl versions will not be able to read journal files using this new format. The environment variable 'SYSTEMD_JOURNAL_COMPACT=0' can be passed to systemd-journald to disable this functionality. It is enabled by default. * systemd-run's --working-directory= switch now works when used in combination with --scope. * portablectl gained a --force flag to skip certain sanity checks. This is implemented using new flags accepted by systemd-portabled for the *WithExtensions() D-Bus methods: SD_SYSTEMD_PORTABLE_FORCE_ATTACH flag now means that the attach/detach checks whether the units are already present and running will be skipped. Similarly, SD_SYSTEMD_PORTABLE_FORCE_SYSEXT flag means that the check whether image name matches the name declared inside of the image will be skipped. Callers must be sure to do those checks themselves if appropriate. * systemd-portabled will now use the original filename to check extension-release.NAME for correctness, in case it is passed a symlink. * systemd-portabled now uses PrivateTmp=yes in the 'trusted' profile too. * sysext's extension-release files now support '_any' as a special value for the ID= field, to allow distribution-independent extensions (e.g.: fully statically compiled binaries, scripts). It also gained support for a new ARCHITECTURE= field that may be used to explicitly restrict an image to hosts of a specific architecture. * systemd-repart now supports creating squashfs partitions. This requires mksquashfs from squashfs-tools. * systemd-repart gained a --split flag to also generate split artifacts, i.e. a separate file for each partition. This is useful in conjunction with systemd-sysupdate or other tools, or to generate split dm-verity artifacts. * systemd-repart is now able to generate dm-verity partitions, including signatures. * systemd-repart can now set a partition UUID to zero, allowing it to be filled in later, such as when using verity partitions. * systemd-repart now supports drop-ins for its configuration files. * Package metadata logged by systemd-coredump in the system journal is now more compact. * xdg-autostart-service now expands 'tilde' characters in Exec lines. * systemd-oomd now automatically links against libatomic, if available. * systemd-oomd now sends out a 'Killed' D-Bus signal when a cgroup is killed. * scope units now also provide oom-kill status. * systemd-pstore will now try to load only the efi_pstore kernel module before running, ensuring that pstore can be used. * systemd-logind gained a new StopIdleSessionSec= option to stop an idle session after a preconfigure timeout. * systemd-homed will now wait up to 30 seconds for workers to terminate, rather than indefinitely. * homectl gained a new '--luks-sector-size=' flag that allows users to select the preferred LUKS sector size. Must be a power of 2 between 512 and 4096. systemd-userdbd records gained a corresponding field. * systemd-sysusers will now respect the 'SOURCE_DATE_EPOCH' environment variable when generating the 'sp_lstchg' field, to ensure an image build can be reproducible. * 'udevadm wait' will now listen to kernel uevents too when called with --initialized=no. * When naming network devices udev will now consult the Devicetree "alias" fields for the device. * systemd-udev will now create infiniband/by-path and infiniband/by-ibdev links for Infiniband verbs devices. * systemd-udev-trigger.service will now also prioritize input devices. * ConditionACPower= and systemd-ac-power will now assume the system is running on AC power if no battery can be found. * All features and tools using the TPM2 will now communicate with it using a bind key. Beforehand, the tpm2 support used encrypted sessions by creating a primary key that was used to encrypt traffic. This creates a problem as the key created for encrypting the traffic could be faked by an active interposer on the bus. In cases when a pin is used, a bind key will be used. The pin is used as the auth value for the seal key, aka the disk encryption key, and that auth value will be used in the session establishment. An attacker would need the pin value to create the secure session and thus an active interposer without the pin cannot interpose on TPM2 traffic. * systemd-growfs no longer requires udev to run. * systemd-backlight now will better support systems with multiple graphic cards. * systemd-cryptsetup's keyfile-timeout= option now also works when a device is used as a keyfile. * systemd-cryptenroll gained a new --unlock-key-file= option to get the unlocking key from a key file (instead of prompting the user). Note that this is the key for unlocking the volume in order to be able to enroll a new key, but it is not the key that is enrolled. * systemd-dissect gained a new --umount switch that will safely and synchronously unmount all partitions of an image previously mounted with 'systemd-dissect --mount'. * When using gcrypt, all systemd tools and services will now configure it to prefer the OS random number generator if present. * All example code shipped with documentation has been relicensed from CC0 to MIT-0. * Unit tests will no longer fail when running on a system without /etc/machine-id. Experimental features: * BPF programs can now be compiled with bpf-gcc (requires libbpf >= 1.0 and bpftool >= 7.0). * sd-boot can automatically enroll SecureBoot keys from files found on the ESP. This enrollment can be either automatic ('force' mode) or controlled by the user ('manual' mode). It is sufficient to place the SecureBoot keys in the right place in the ESP and they will be picked up by sd-boot and shown in the boot menu. * The mkosi config in systemd gained support for automatically compiling a kernel with the configuration appropriate for testing systemd. This may be useful when developing or testing systemd in tandem with the kernel.

 

Fonte

 

Comentários

Você precisa ver isso

Todos os arquivos do blog

Mostrar mais