forked from systemd/systemd
-
Notifications
You must be signed in to change notification settings - Fork 0
/
NEWS
14109 lines (11518 loc) · 751 KB
/
NEWS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
systemd System and Service Manager
CHANGES WITH 251 in spe:
Backwards-incompatible changes:
* In v250, a systemd-networkd feature that automatically configures
routes to addresses specified in AllowedIPs= was added and enabled by
default. However, this causes network connectivity issues in many
existing setups. Hence, it has been disabled by default since
systemd-stable 250.3. The feature can still be used by explicitly
configuring RouteTable= setting in .netdev files.
* Jobs started via StartUnitWithFlags() will no longer return 'skipped'
when a Condition*= check does not succeed, restoring the JobRemoved
signal to the behaviour it had before v250.
* The org.freedesktop.portable1 methods GetMetadataWithExtensions() and
GetImageMetadataWithExtensions() have been fixed to provide an extra
return parameter, containing the actual extension release metadata.
The current implementation was judged to be broken and unusable, and
thus the usual procedure of adding a new set of methods was skipped,
and backward compatibility broken instead on the assumption that
nobody can be affected given the current state of this interface.
* All kernels supported by systemd mix RDRAND (or similar) into the
entropy pool at early boot. This means that on those systems, even if
/dev/urandom is not yet initialized, it still returns bytes that that
are at least as high quality as RDRAND. For that reason, we no longer
have reason to invoke RDRAND from systemd itself, which has
historically been a source of bugs. Furthermore, kernels ≥5.6 provide
the getrandom(GRND_INSECURE) interface for returning random bytes
before the entropy pool is initialized without warning into kmsg,
which is what we attempt to use if available. systemd's direct usage
of RDRAND has been removed. x86 systems ≥Broadwell that are running
an older kernel may experience kmsg warnings that were not seen with
250. For newer kernels, non-x86 systems, or older x86 systems, there
should be no visible changes.
* sd-boot will now measure the kernel command line into TPM PCR 12
rather than PCR 8. This improves usefulness of the measurements on
systems where sd-boot is chainloaded from Grub. Grub measures all
commands its executes into PCR 8, which makes it very hard to use
reasonably, hence separate ourselves from that and use PCR 12
instead, which is what certain Ubuntu editions already do. To retain
compatibility with systems running older systemd systems a new Meson
option 'efi-tpm-pcr-compat' has been added (which defaults to false).
If enabled, the measurement is done twice: into the new-style PCR 12
*and* the old-style PCR 8. It's strongly advised to migrate all users
to PCR 12 for this purpose in the long run, as we intend to remove
this compatibility feature in two year's time.
* busctl capture now writes output in the newer pcapng format instead
of pcap.
* An udev rule that imported hwdb matches for USB devices with
lowercase hexadecimal vendor/product ID digits was added in systemd
250. This has been reverted, since uppercase hexadecimal digits are
supposed to be used, and we already had a rule that with the
appropriate match.
Users might need to adjust their local hwdb entries.
* arch_prctl(2) has been moved to the @default set in the syscall filters
(as exposed via the SystemCallFilter= setting in service unit files).
It is apparently used by the linker now.
New functionality and other changes:
* kernel-install's and bootctl's Boot Loader Specification Type #1
entry generation logic has been reworked. The user may now pick
explicitly by which "token" string to name the installation's boot
entries, via the new /etc/kernel/entry-token file or the new
--entry-token= switch to bootctl. By default — as before — the
entries are named after the local machine ID. However, in "golden
image" environments, where the machine ID shall be initialized on
first boot (as opposed to at installation time before first boot) the
machine ID will not be available at build time. In this case the
--entry-token= switch to bootctl (or the /etc/kernel/entry-token
file) may be used to override the "token" for the entries, for
example the IMAGE_ID= or ID= fields from /etc/os-release. This will
make the OS images independent of any machine ID, and ensure that the
images will not carry any identifiable information before first boot,
but on the other hand means that multiple parallel installations of
the very same image on the same disk cannot be supported.
Summary: if you are building golden images that shall acquire
identity information exclusively on first boot, make sure to both
remove /etc/machine-id *and* to write /etc/kernel/entry-token to the
value of the IMAGE_ID= or ID= field of /etc/os-release or another
suitable identifier before deploying the image.
* The Boot Loader Specification has been extended with
/loader/entries.srel file located in the EFI System Partition (ESP)
that disambiguates the format of the entries in the /loader/entries/
directory (in order to discern them from incompatible uses of this
directory by other projects). For entries that follow the
Specification, the string "type1" is stored in this file.
bootctl will now write this file automatically when installing the
systemd-boot boot loader.
* kernel-install supports a new initrd_generator= setting in
/etc/kernel/install.conf, that is exported as
$KERNEL_INSTALL_INITRD_GENERATOR to kernel-install plugins. This
allows choosing different initrd generators.
* kernel-install will now create a "staging area" (an initially-empty
directory to gather files for a Boot Loader Specification Type #1
entry). The path to this directory is exported as
$KERNEL_INSTALL_STAGING_AREA to kernel-install plugins, which should
drop files there instead of writing them directly to the final
location. kernel-install will move them when all files have been
prepared successfully.
* Starting with v250 systemd-homed uses UID/GID mapping on the mounts
of activated home directories it manages (if the kernel and selected
file systems support it). So far it mapped three UID ranges: the
range from 0…60000, the user's own UID, and the range 60514…65534,
leaving everything else unmapped (in other words, the 16bit UID range
is mapped almost fully, with the exception of the UID subrange used
for systemd-homed users, with one exception: the user's own UID).
Unmapped UIDs may not be used for file ownership in the home
directory — any chown() attempts with them will fail. With this
release a fourth range is added to these mappings:
524288…1879048191. This range is the UID range intended for container
uses, see:
https://systemd.io/UIDS-GIDS
This range may be used for container managers that place container OS
trees in the home directory (which is a questionable approach, for
quota, permission, SUID handling and network file system
compatibility reasons, but nonetheless apparently commonplace). Note
that this mapping is mapped 1:1 in a pass-through fashion, i.e. the
UID assignments from the range are not managed or mapped by
`systemd-homed`, and must be managed with other mechanisms, in the
context of the local system.
Typically, a better approach to user namespacing in relevant
container managers would be to leave container OS trees on disk at
UID offset 0, but then map them to a dynamically allocated runtime
UID range via another UID mount map at container invocation
time. That way user namespace UID ranges become strictly a runtime
concept, and do not leak into persistent file systems, persistent
user databases or persistent configuration, thus greatly simplifying
handling, and improving compatibility with home directories intended
to be portable like the ones managed by systemd-homed.
* The journal JSON export format has been added to listed of stable
interfaces (https://systemd.io/PORTABILITY_AND_STABILITY/).
* /etc/locale.conf is now populated through tmpfiles.d factory /etc/
handling with the values that were configured during systemd build
(if /etc/locale.conf has not been created through some other
mechanism). This means that /etc/locale.conf should always have
reasonable contents and we avoid a potential mismatch in defaults.
* A new libsystemd-core-<version>.so private shared library is
installed under /usr/lib/systemd/system, mirroring the existing
libsystemd-shared-<version>.so library. This allows the total
installation size to be reduced by binary code reuse.
* The <version> tag used in the name of libsystemd-shared.so and
libsystemd-core.so can be configured. Distributions may build
subsequent versions of the systemd package with unique tags (e.g. the
full package version), thus allowing multiple installations of those
shared libraries to be available at the same time. This is intended
to fix an issue where programs that link to those libraries would
fail to execute because they were installed earlier or later than the
appropriate version of the library.
* A new set of service monitor environment variables will be passed to
OnFailure=/OnSuccess= handlers, but only if exactly one unit lists the
handler unit as OnFailure=/OnSuccess=. The variables are:
$MONITOR_SERVICE_RESULT, $MONITOR_EXIT_CODE, $MONITOR_EXIT_STATUS,
$MONITOR_INVOCATION_ID and $MONITOR_UNIT. For cases when a single
handler needs to watch multiple units, use a templated handler.
* A new ExtensionDirectories= setting in service unit files allows
system extensions to be loaded from a directory. (It is similar to
ExtensionImages=, but takes paths to directories, instead of
disk image files.)
'portablectl attach --extension=' now also accepts directory paths.
* HARDWARE_VENDOR= and HARDWARE_MODEL= can be set in /etc/machine-info
to override the values gleaned from the hwdb.
* A ID_CHASSIS property can be set in the hwdb (for the DMI device
/sys/class/dmi/id) to override the chassis that is reported by
hostnamed.
* hostnamed's D-Bus interface gained a new method GetHardwareSerial()
for reading the hardware serial number, as reportd by DMI.
* Two new hwdb files have been added. One lists "handhelds" (PDAs,
calculators, etc.), the other AV production devices (DJ tables,
keypads, etc.) that should accessible to the seat owner user by
default.
* A new unit systemd-networkd-wait-online@<interface>.service has been
added that can be used to wait for a specific network interface to be
up.
* systemd-resolved is started earlier (in sysinit.target), so it
available earlier and will also be started in the initrd if installed
there.
* udevadm trigger gained a new --prioritized-subsystem= option to
process certain subsystems (and all their parent devices) earlier.
systemd-udev-trigger.service now uses this new option to trigger
block and TPM devices first, hopefully making the boot a bit faster.
* udevadm trigger now implements --type=all, --initialized-match,
--initialized-nomatch to trigger both subsystems and devices, only
already-initialized devices, and only devices which haven't been
initialized yet, respectively.
* systemd-cryptenroll can now control whether to require the user to
enter a PIN when using TPM-based unlocking of a volume via the new
--tpm2-with-pin= option.
Option tpm2-pin= can be used in /etc/crypttab.
* When unlocking devices via TPM, TPM2 parameter encryption is now
used, to ensure that communication between CPU and discrete TPM chips
cannot be eavesdropped to acquire disk encryption keys.
* The user.delegate and user.invocation_id extended attributes on
cgroups are used in addition to trusted.delegate and
trusted.invocation_id. The latter pair requires privileges to set,
but the former doesn't and can be also set by the unprivileged user
manager.
(Only supported on kernels ≥5.6.)
* New option sort-key= has been added to the Boot Loader Specification
to override the sorting order of the entries in the boot menu. It is
read by sd-boot and bootctl, and will be written by kernel-install,
with the default value of IMAGE_ID= or ID= fields from
os-release. Together, this means that on multiboot installations,
entries should be grouped and sorted in a predictable way.
* sd-boot can now optionally beep when the menu is shown and menu
entries are selected, which can be useful on machines without a
working display. (Controllable via a loader.conf setting.)
* In unit files the new %y/%Y specifiers can be used to refer to
normalized unit file path, which is particularly useful for symlinked
unit files.
The new %R specifier resolves to the pretty hostname
(i.e. PRETTY_HOSTNAME= from /etc/machine-info).
The new %d specifier resolves to the credentials directory of a
service (same as $CREDENTIALS_DIRECTORY).
* The RootDirectory=, MountAPIVFS=, ExtensionDirectories=,
*Capabilities*=, ProtectHome=, *Directory=, TemporaryFileSystem=,
PrivateTmp=, PrivateDevices=, PrivateNetwork=, NetworkNamespacePath=,
PrivateIPC=, IPCNamespacePath=, PrivateUsers=, ProtectClock=,
ProtectKernelTunables=, ProtectKernelModules=, ProtectKernelLogs=,
MountFlags= service settings now also work in unprivileged user
services, i.e. those run by the user's --user service manager, as long
as user namespaces are enabled on the system.
* The --make-machine-id-directory= switch to bootctl has been replaced
by --make-entry-directory=, given that the entry directory is not
necessarily named after the machine ID, but after some other suitable
ID as selected via --entry-token= described above. The old name of
the option is still understood to maximize compatibility.
* 'bootctl list' gained support for a new --json= switch to output boot
menu entries in JSON format.
* Services with Restart=always and a failing ExecCondition= will no
longer be restarted, to bring ExecCondition= behaviour in line with
Condition*= settings.
* LoadCredential= now accepts a directory as the argument; all files
from the directory will be loaded as credentials.
* A new D-Bus property ControlGroupId is now exposed on service units,
that encapsulates the service's numeric cgroup ID that newer kernels
assign to each cgroup.
* systemd-networkd gained a new [Bridge] Isolated=true|false setting
that configures the eponymous kernel attribute on the bridge.
* .netdev files now can be used to create virtual WLAN devices, and
configure various settings on them, via the [VirtualWLAN] section.
* .link files gained support for [Match] Firmware= setting to match on
the device firmware description string. By mistake, it was previously
only supported in .network files.
* .link/.network files gained support for [Match] Kind= setting to match
on device kind ("bond", "bridge", "gre", "tun", "veth", etc.)
This value is also shown by 'networkctl status'.
* .link files gained support for setting MDI/MID-X on a link.
* The Local= setting for various virtual network devices gained support
for specifying, in addition to the network address, the name of a
local interface which must have the specified address.
* New [DHCPServer] BootServerName=, BootServerAddress=, and
BootFilename= settings can be used to configure the server address,
server name, and file name sent in the DHCP packet (e.g. to configure
PXE boot).
* journalctl --list-boots now supports JSON output and the --reverse option.
* Under docs/: JOURNAL_EXPORT_FORMATS was imported from the wiki and
updated, BUILDING_IMAGES is new:
https://systemd.io/JOURNAL_EXPORT_FORMATS
https://systemd.io/BUILDING_IMAGES
* The sd-id128 API gained a new call sd_id128_to_uuid_string() that is
similar to sd_id128_to_string() but formats the ID in RFC 4122 UUID
format instead of simple series of hex characters.
* The userdbctl tool will now show UID range information as part of the
list of known users.
* systemctl's --timestamp= option gained a new choice "unix", to show
timestamp as unix times, i.e. seconds since 1970, Jan 1st.
* PID 1 gained support for configuring the "pre-timeout" of watchdog
devices and the associated governor, via the new
RuntimeWatchdogPreSec= and RuntimeWatchdogPreGovernor= configuration
options in /etc/systemd/system.conf.
* The kernel-install tool gained a new 'inspect' verb which shows the
paths and other settings used.
Experimental features:
* sd-boot gained a new *experimental* setting "reboot-for-bitlocker" in
loader.conf that implements booting Microsoft Windows from the
sd-boot in a way that first reboots the system, to reset the TPM
PCRs. This improves compatibility with BitLocker's TPM use, as the
PCRs will only record the Windows boot process, and not sd-boot
itself, thus retaining the PCR measurements not involving sd-boot.
Note that this feature is experimental for now, and is likely going
to be generalized and renamed in a future release, without retaining
compatibility with the current implementation.
* A new systemd-sysupdate component has been added that automatically
discovers, downloads, and installs A/B-style updates for the host
installation itself, or container images, portable service images,
and other assets. See the new systemd-sysupdate man page for updates.
CHANGES WITH 250:
* Support for encrypted and authenticated credentials has been added.
This extends the credential logic introduced with v247 to support
non-interactive symmetric encryption and authentication, based on a
key that is stored on the /var/ file system or in the TPM2 chip (if
available), or the combination of both (by default if a TPM2 chip
exists the combination is used, otherwise the /var/ key only). The
credentials are automatically decrypted at the moment a service is
started, and are made accessible to the service itself in unencrypted
form. A new tool 'systemd-creds' encrypts credentials for this
purpose, and two new service file settings LoadCredentialEncrypted=
and SetCredentialEncrypted= configure such credentials.
This feature is useful to store sensitive material such as SSL
certificates, passwords and similar securely at rest and only decrypt
them when needed, and in a way that is tied to the local OS
installation or hardware.
* systemd-gpt-auto-generator can now automatically set up discoverable
LUKS2 encrypted swap partitions.
* The GPT Discoverable Partitions Specification has been substantially
extended with support for root and /usr/ partitions for the majority
of architectures systemd supports. This includes platforms that do
not natively support UEFI, because even though GPT is specified under
UEFI umbrella, it is useful on other systems too. Specifically,
systemd-nspawn, systemd-sysext, systemd-gpt-auto-generator and
Portable Services use the concept without requiring UEFI.
* The GPT Discoverable Partitions Specifications has been extended with
a new set of partitions that may carry PKCS#7 signatures for Verity
partitions, encoded in a simple JSON format. This implements a simple
mechanism for building disk images that are fully authenticated and
can be tested against a set of cryptographic certificates. This is
now implemented for the various systemd tools that can operate with
disk images, such as systemd-nspawn, systemd-sysext, systemd-dissect,
Portable services/RootImage=, systemd-tmpfiles, and systemd-sysusers.
The PKCS#7 signatures are passed to the kernel (where they are
checked against certificates from the kernel keyring), or can be
verified against certificates provided in userspace (via a simple
drop-in file mechanism).
* systemd-dissect's inspection logic will now report for which uses a
disk image is intended. Specifically, it will display whether an
image is suitable for booting on UEFI or in a container (using
systemd-nspawn's --image= switch), whether it can be used as portable
service, or attached as system extension.
* The system-extension.d/ drop-in files now support a new field
SYSEXT_SCOPE= that may encode which purpose a system extension image
is for: one of "initrd", "system" or "portable". This is useful to
make images more self-descriptive, and to ensure system extensions
cannot be attached in the wrong contexts.
* The os-release file learnt a new PORTABLE_PREFIXES= field which may
be used in portable service images to indicate which unit prefixes
are supported.
* The GPT image dissection logic in systemd-nspawn/systemd-dissect/…
now is able to decode images for non-native architectures as well.
This allows systemd-nspawn to boot images of non-native architectures
if the corresponding user mode emulator is installed and
systemd-binfmtd is running.
* systemd-logind gained new settings HandlePowerKeyLongPress=,
HandleRebootKeyLongPress=, HandleSuspendKeyLongPress= and
HandleHibernateKeyLongPress= which may be used to configure actions
when the relevant keys are pressed for more than 5s. This is useful
on devices that only have hardware for a subset of these keys. By
default, if the reboot key is pressed long the poweroff operation is
now triggered, and when the suspend key is pressed long the hibernate
operation is triggered. Long pressing the other two keys currently
does not trigger any operation by default.
* When showing unit status updates on the console during boot and
shutdown, and a service is slow to start so that the cylon animation
is shown, the most recent sd_notify() STATUS= text is now shown as
well. Services may use this to make the boot/shutdown output easier
to understand, and to indicate what precisely a service that is slow
to start or stop is waiting for. In particular, the per-user service
manager instance now reports what it is doing and which service it is
waiting for this way to the system service manager.
* The service manager will now re-execute on reception of the
SIGRTMIN+25 signal. It previously already did that on SIGTERM — but
only when running as PID 1. There was no signal to request this when
running as per-user service manager, i.e. as any other PID than 1.
SIGRTMIN+25 works for both system and user managers.
* The hardware watchdog logic in PID 1 gained support for operating
with the default timeout configured in the hardware, instead of
insisting on re-configuring it. Set RuntimeWatchdogSec=default to
request this behavior.
* A new kernel command line option systemd.watchdog_sec= is now
understood which may be used to override the hardware watchdog
time-out for the boot.
* A new setting DefaultOOMScoreAdjust= is now supported in
/etc/systemd/system.conf + /etc/systemd/user.conf that may be used to
set the default process OOM score adjustment value for processes
forked off the service manager. For per-user service managers this
now defaults to 100, but for per-system service managers is left as
is. This means that by default now services forked off the user
service manager are more likely to be killed by the OOM killer than
system services or the managers themselves.
* A new per-service setting RestrictFileSystems= as been added that
restricts the file systems a service has access to by their type.
This is based on the new BPF LSM of the Linux kernel. It provides an
effective way to make certain API file systems unavailable to
services (and thus minimizing attack surface). A new command
"systemd-analyze filesystems" has been added that lists all known
file system types (and how they are grouped together under useful
group handles).
* Services now support a new setting RestrictNetworkInterfaces= for
restricting access to specific network interfaces.
* Service unit files gained new settings StartupAllowedCPUs= and
StartupAllowedMemoryNodes=. These are similar to their counterparts
without the "Startup" prefix and apply during the boot process
only. This is useful to improve boot-time behavior of the system and
assign resources differently during boot than during regular
runtime. This is similar to the preexisting StartupCPUWeight=
vs. CPUWeight.
* Related to this: the various StartupXYZ= settings
(i.e. StartupCPUWeight=, StartupAllowedCPUs=, …) are now also applied
during shutdown. The settings not prefixed with "Startup" hence apply
during regular runtime, and those that are prefixed like that apply
during boot and shutdown.
* A new per-unit set of conditions/asserts
[Condition|Assert][Memory|CPU|IO]Pressure= have been added to make a
unit skip/fail activation if the system's (or a slice's) memory/cpu/io
pressure is above the configured threshold, using the kernel PSI
feature. For more details see systemd.unit(5) and
https://www.kernel.org/doc/html/latest/accounting/psi.html
* The combination of ProcSubset=pid and ProtectKernelTunables=yes and/or
ProtectKernelLogs=yes can now be used.
* The default maximum numbers of inodes have been raised from 64k to 1M
for /dev/, and from 400k to 1M for /tmp/.
* The per-user service manager learnt support for communicating with
systemd-oomd to acquire OOM kill information.
* A new service setting ExecSearchPath= has been added that allows
changing the search path for executables for services. It affects
where we look for the binaries specified in ExecStart= and similar,
and the specified directories are also added the $PATH environment
variable passed to invoked processes.
* A new setting RuntimeRandomizedExtraSec= has been added for service
and scope units that allows extending the runtime time-out as
configured by RuntimeMaxSec= with a randomized amount.
* The syntax of the service unit settings RuntimeDirectory=,
StateDirectory=, CacheDirectory=, LogsDirectory= has been extended:
if the specified value is now suffixed with a colon, followed by
another filename, the latter will be created as symbolic link to the
specified directory. This allows creating these service directories
together with alias symlinks to make them available under multiple
names.
* Service unit files gained two new settings TTYRows=/TTYColumns= for
configuring rows/columns of the TTY device passed to
stdin/stdout/stderr of the service. This is useful to propagate TTY
dimensions to a virtual machine.
* A new service unit file setting ExitType= has been added that
specifies when to assume a service has exited. By default systemd
only watches the main process of a service. By setting
ExitType=cgroup it can be told to wait for the last process in a
cgroup instead.
* Automount unit files gained a new setting ExtraOptions= that can be
used to configure additional mount options to pass to the kernel when
mounting the autofs instance.
* "Urlification" (generation of ESC sequences that generate clickable
hyperlinks in modern terminals) may now be turned off altogether
during build-time.
* Path units gained new TriggerLimitBurst= and TriggerLimitIntervalSec=
settings that default to 200 and 2 s respectively. The ratelimit
ensures that a path unit cannot cause PID1 to busy-loop when it is
trying to trigger a service that is skipped because of a Condition*=
not being satisfied. This matches the configuration and behaviour of
socket units.
* The TPM2/FIDO2/PKCS11 support in systemd-cryptsetup is now also built
as a plug-in for cryptsetup. This means the plain cryptsetup command
may now be used to unlock volumes set up this way.
* The TPM2 logic in cryptsetup will now automatically detect systems
where the TPM2 chip advertises SHA256 PCR banks but the firmware only
updates the SHA1 banks. In such a case PCR policies will be
automatically bound to the latter, not the former. This makes the PCR
policies reliable, but of course do not provide the same level of
trust as SHA256 banks.
* The TPM2 logic in systemd-cryptsetup/systemd-cryptsetup now supports
RSA primary keys in addition to ECC, improving compatibility with
TPM2 chips that do not support ECC. RSA keys are much slower to use
than ECC, and hence are only used if ECC is not available.
* /etc/crypttab gained support for a new token-timeout= setting for
encrypted volumes that allows configuration of the maximum time to
wait for PKCS#11/FIDO2 tokens to be plugged in. If the time elapses
the logic will query the user for a regular passphrase/recovery key
instead.
* Support for activating dm-integrity volumes at boot via a new file
/etc/integritytab and the tool systemd-integritysetup have been
added. This is similar to /etc/crypttab and /etc/veritytab, but deals
with dm-integrity instead of dm-crypt/dm-verity.
* The systemd-veritysetup-generator now understands a new usrhash=
kernel command line option for specifying the Verity root hash for
the partition backing the /usr/ file system. A matching set of
systemd.verity_usr_* kernel command line options has been added as
well. These all work similar to the corresponding options for the
root partition.
* The sd-device API gained a new API call sd_device_get_diskseq() to
return the DISKSEQ property of a device structure. The "disk
sequence" concept is a new feature recently introduced to the Linux
kernel that allows detecting reuse cycles of block devices, i.e. can
be used to recognize when loopback block devices are reused for a
different purpose or CD-ROM drives get their media changed.
* A new unit systemd-boot-update.service has been added. If enabled
(the default) and the sd-boot loader is detected to be installed, it
is automatically updated to the newest version when out of date. This
is useful to ensure the boot loader remains up-to-date, and updates
automatically propagate from the OS tree in /usr/.
* sd-boot will now build with SBAT by default in order to facilitate
working with recent versions of Shim that require it to be present.
* sd-boot can now parse Microsoft Windows' Boot Configuration Data.
This is used to robustly generate boot entry titles for Windows.
* A new generic target unit factory-reset.target has been added. It is
hooked into systemd-logind similar in fashion to
reboot/poweroff/suspend/hibernate, and is supposed to be used to
initiate a factory reset operation. What precisely this operation
entails is up for the implementer to decide, the primary goal of the
new unit is provide a framework where to plug in the implementation
and how to trigger it.
* A new meson build-time option 'clock-valid-range-usec-max' has been
added which takes a time in µs and defaults to 15 years. If the RTC
time is noticed to be more than the specified time ahead of the
built-in epoch of systemd (which by default is the release timestamp
of systemd) it is assumed that the RTC is not working correctly, and
the RTC is reset to the epoch. (It already is reset to the epoch when
noticed to be before it.) This should increase the chance that time
doesn't accidentally jump too far ahead due to faulty hardware or
batteries.
* A new setting SaveIntervalSec= has been added to systemd-timesyncd,
which may be used to automatically save the current system time to
disk in regular intervals. This is useful to maintain a roughly
monotonic clock even without RTC hardware and with some robustness
against abnormal system shutdown.
* systemd-analyze verify gained support for a pair of new --image= +
--root= switches for verifying units below a specific root
directory/image instead of on the host.
* systemd-analyze verify gained support for verifying unit files under
an explicitly specified unit name, independently of what the filename
actually is.
* systemd-analyze verify gained a new switch --recursive-errors= which
controls whether to only fail on errors found in the specified units
or recursively any dependent units.
* systemd-analyze security now supports a new --offline mode for
analyzing unit files stored on disk instead of loaded units. It may
be combined with --root=/--image to analyze unit files under a root
directory or disk image. It also learnt a new --threshold= parameter
for specifying an exposure level threshold: if the exposure level
exceeds the specified value the call will fail. It also gained a new
--security-policy= switch for configuring security policies to
enforce on the units. A policy is a JSON file that lists which tests
shall be weighted how much to determine the overall exposure
level. Altogether these new features are useful for fully automatic
analysis and enforcement of security policies on unit files.
* systemd-analyze security gain a new --json= switch for JSON output.
* systemd-analyze learnt a new --quiet switch for reducing
non-essential output. It's honored by the "dot", "syscall-filter",
"filesystems" commands.
* systemd-analyze security gained a --profile= option that can be used
to take into account a portable profile when analyzing portable
services, since a lot of the security-related settings are enabled
through them.
* systemd-analyze learnt a new inspect-elf verb that parses ELF core
files, binaries and executables and prints metadata information,
including the build-id and other info described on:
https://systemd.io/COREDUMP_PACKAGE_METADATA/
* .network files gained a new UplinkInterface= in the [IPv6SendRA]
section, for automatically propagating DNS settings from other
interfaces.
* The static lease DHCP server logic in systemd-networkd may now serve
IP addresses outside of the configured IP pool range for the server.
* CAN support in systemd-networkd gained four new settings Loopback=,
OneShot=, PresumeAck=, ClassicDataLengthCode= for tweaking CAN
control modes. It gained a number of further settings for tweaking
CAN timing quanta.
* The [CAN] section in .network file gained new TimeQuantaNSec=,
PropagationSegment=, PhaseBufferSegment1=, PhaseBufferSegment2=,
SyncJumpWidth=, DataTimeQuantaNSec=, DataPropagationSegment=,
DataPhaseBufferSegment1=, DataPhaseBufferSegment2=, and
DataSyncJumpWidth= settings to control bit-timing processed by the
CAN interface.
* DHCPv4 client support in systemd-networkd learnt a new Label= option
for configuring the address label to apply to configure IPv4
addresses.
* The [IPv6AcceptRA] section of .network files gained support for a new
UseMTU= setting that may be used to control whether to apply the
announced MTU settings to the local interface.
* The [DHCPv4] section in .network file gained a new Use6RD= boolean
setting to control whether the DHCPv4 client request and process the
DHCP 6RD option.
* The [DHCPv6PrefixDelegation] section in .network file is renamed to
[DHCPPrefixDelegation], as now the prefix delegation is also supported
with DHCPv4 protocol by enabling the Use6RD= setting.
* The [DHCPPrefixDelegation] section in .network file gained a new
setting UplinkInterface= to specify the upstream interface.
* The [DHCPv6] section in .network file gained a new setting
UseDelegatedPrefix= to control whether the delegated prefixes will be
propagated to the downstream interfaces.
* The [IPv6AcceptRA] section of .network files now understands two new
settings UseGateway=/UseRoutePrefix= for explicitly configuring
whether to use the relevant fields from the IPv6 Router Advertisement
records.
* The ForceDHCPv6PDOtherInformation= setting in the [DHCPv6] section
has been removed. Please use the WithoutRA= and UseDelegatedPrefix=
settings in the [DHCPv6] section and the DHCPv6Client= setting in the
[IPv6AcceptRA] section to control when the DHCPv6 client is started
and how the delegated prefixes are handled by the DHCPv6 client.
* The IPv6Token= section in the [Network] section is deprecated, and
the [IPv6AcceptRA] section gained the Token= setting for its
replacement. The [IPv6Prefix] section also gained the Token= setting.
The Token= setting gained 'eui64' mode to explicitly configure an
address with the EUI64 algorithm based on the interface MAC address.
The 'prefixstable' mode can now optionally take a secret key. The
Token= setting in the [DHCPPrefixDelegation] section now supports all
algorithms supported by the same settings in the other sections.
* The [RoutingPolicyRule] section of .network file gained a new
SuppressInterfaceGroup= setting.
* The IgnoreCarrierLoss= setting in the [Network] section of .network
files now allows a duration to be specified, controlling how long to
wait before reacting to carrier loss.
* The [DHCPServer] section of .network file gained a new Router=
setting to specify the router address.
* The [CAKE] section of .network files gained various new settings
AutoRateIngress=, CompensationMode=, FlowIsolationMode=, NAT=,
MPUBytes=, PriorityQueueingPreset=, FirewallMark=, Wash=, SplitGSO=,
and UseRawPacketSize= for configuring CAKE.
* systemd-networkd now ships with new default .network files:
80-container-vb.network which matches host-side network bridge device
created by systemd-nspawn's --network-bridge or --network-zone
switch, and 80-6rd-tunnel.network which matches automatically created
sit tunnel with 6rd prefix when the DHCP 6RD option is received.
* systemd-networkd's handling of Endpoint= resolution for WireGuard
interfaces has been improved.
* systemd-networkd will now automatically configure routes to addresses
specified in AllowedIPs=. This feature can be controlled via
RouteTable= and RouteMetric= settings in [WireGuard] or
[WireGuardPeer] sections.
* systemd-networkd will now once again automatically generate persistent
MAC addresses for batadv and bridge interfaces. Users can disable this
by using MACAddress=none in .netdev files.
* systemd-networkd and systemd-udevd now support IP over InfiniBand
interfaces. The Kind= setting in .netdev file accepts "ipoib". And
systemd.netdev files gained the [IPoIB] section.
* systemd-networkd and systemd-udevd now support net.ifname-policy=
option on the kernel command-line. This is implemented through the
systemd-network-generator service that automatically generates
appropriate .link, .network, and .netdev files.
* The various systemd-udevd "ethtool" buffer settings now understand
the special value "max" to configure the buffers to the maximum the
hardware supports.
* systemd-udevd's .link files may now configure a large variety of
NIC coalescing settings, plus more hardware offload settings.
* .link files gained a new WakeOnLanPassword= setting in the [Link]
section that allows to specify a WoL "SecureOn" password on hardware
that supports this.
* systemd-nspawn's --setenv= switch now supports an additional syntax:
if only a variable name is specified (i.e. without being suffixed by
a '=' character and a value) the current value of the environment
variable is propagated to the container. e.g. --setenv=FOO will
lookup the current value of $FOO in the environment, and pass it down
to the container. Similar behavior has been added to homectl's,
machinectl's and systemd-run's --setenv= switch.
* systemd-nspawn gained a new switch --suppress-sync= which may be used
to optionally suppress the effect of the sync()/fsync()/fdatasync()
system calls for the container payload. This is useful for build
system environments where safety against abnormal system shutdown is
not essential as all build artifacts can be regenerated any time, but
the performance win is beneficial.
* systemd-nspawn will now raise the RLIMIT_NOFILE hard limit to the
same value that PID 1 uses for most forked off processes.
* systemd-nspawn's --bind=/--bind-ro= switches now optionally take
uidmap/nouidmap options as last parameter. If "uidmap" is used the
bind mounts are created with UID mapping taking place that ensures
the host's file ownerships are mapped 1:1 to container file
ownerships, even if user namespacing is used. This way
files/directories bound into containers will no longer show up as
owned by the nobody user as they typically did if no special care was
taken to shift them manually.
* When discovering Windows installations sd-boot will now attempt to
show the Windows version.
* The color scheme to use in sd-boot may now be configured at
build-time.
* sd-boot gained the ability to change screen resolution during
boot-time, by hitting the "r" key. This will cycle through available
resolutions and save the last selection.
* sd-boot learnt a new hotkey "f". When pressed the system will enter
firmware setup. This is useful in environments where it is difficult
to hit the right keys early enough to enter the firmware, and works
on any firmware regardless which key it natively uses.
* sd-boot gained support for automatically booting into the menu item
selected on the last boot (using the "@saved" identifier for menu
items).
* sd-boot gained support for automatically loading all EFI drivers
placed in the /EFI/systemd/drivers/ subdirectory of the EFI System
Partition (ESP). These drivers are loaded before the menu entries are
loaded. This is useful e.g. to load additional file system drivers
for the XBOOTLDR partition.
* systemd-boot will now paint the input cursor on its own instead of
relying on the firmware to do so, increasing compatibility with broken
firmware that doesn't make the cursor reasonably visible.
* sd-boot now embeds a .osrel PE section like we expect from Boot
Loader Specification Type #2 Unified Kernels. This means sd-boot
itself may be used in place of a Type #2 Unified Kernel. This is
useful for debugging purposes as it allows chain-loading one a
(development) sd-boot instance from another.
* sd-boot now supports a new "devicetree" field in Boot Loader
Specification Type #1 entries: if configured the specified device
tree file is installed before the kernel is invoked. This is useful
for installing/applying new devicetree files without updating the
kernel image.
* Similarly, sd-stub now can read devicetree data from a PE section
".dtb" and apply it before invoking the kernel.
* sd-stub (the EFI stub that can be glued in front of a Linux kernel)
gained the ability to pick up credentials and sysext files, wrap them
in a cpio archive, and pass as an additional initrd to the invoked
Linux kernel, in effect placing those files in the /.extra/ directory
of the initrd environment. This is useful to implement trusted initrd
environments which are fully authenticated but still can be extended
(via sysexts) and parameterized (via encrypted/authenticated
credentials, see above).
Credentials can be located next to the kernel image file (credentials
specific to a single boot entry), or in one of the shared directories
(credentials applicable to multiple boot entries).
* sd-stub now comes with a full man page, that explains its feature set
and how to combine a kernel image, an initrd and the stub to build a
complete EFI unified kernel image, implementing Boot Loader
Specification Type #2.
* sd-stub may now provide the initrd to the executed kernel via the
LINUX_EFI_INITRD_MEDIA_GUID EFI protocol, adding compatibility for
non-x86 architectures.
* bootctl learnt new set-timeout and set-timeout-oneshot commands that
may be used to set the boot menu time-out of the boot loader (for all
or just the subsequent boot).
* bootctl and kernel-install will now read variables
KERNEL_INSTALL_LAYOUT= from /etc/machine-info and layout= from
/etc/kernel/install.conf. When set, it specifies the layout to use
for installation directories on the boot partition, so that tools
don't need to guess it based on the already-existing directories. The
only value that is defined natively is "bls", corresponding to the
layout specified in
https://systemd.io/BOOT_LOADER_SPECIFICATION/. Plugins for
kernel-install that implement a different layout can declare other
values for this variable.
'bootctl install' will now write KERNEL_INSTALL_LAYOUT=bls, on the
assumption that if the user installed sd-boot to the ESP, they intend
to use the entry layout understood by sd-boot. It'll also write
KERNEL_INSTALL_MACHINE_ID= if it creates any directories using the ID
(and it wasn't specified in the config file yet). Similarly,
kernel-install will now write KERNEL_INSTALL_MACHINE_ID= (if it
wasn't specified in the config file yet). Effectively, those changes
mean that the machine-id used for boot loader entry installation is
"frozen" upon first use and becomes independent of the actual
machine-id.
Configuring KERNEL_INSTALL_MACHINE_ID fixes the following problem:
images created for distribution ("golden images") are built with no
machine-id, so that a unique machine-id can be created on the first
boot. But those images may contain boot loader entries with the
machine-id used during build included in paths. Using a "frozen"
value allows unambiguously identifying entries that match the
specific installation, while still permitting parallel installations
without conflict.
Configuring KERNEL_INSTALL_LAYOUT obviates the need for
kernel-install to guess the installation layout. This fixes the
problem where a (possibly empty) directory in the boot partition is
created from a different layout causing kernel-install plugins to
assume the wrong layout. A particular example of how this may happen
is the grub2 package in Fedora which includes directories under /boot
directly in its file list. Various other packages pull in grub2 as a
dependency, so it may be installed even if unused, breaking
installations that use the bls layout.
* bootctl and systemd-bless-boot can now be linked statically.
* systemd-sysext now optionally doesn't insist on extension-release.d/
files being placed in the image under the image's file name. If the
file system xattr user.extension-release.strict is set on the
extension release file, it is accepted regardless of its name. This
relaxes security restrictions a bit, as system extension may be
attached under a wrong name this way.
* udevadm's test-builtin command learnt a new --action= switch for
testing the built-in with the specified action (in place of the
default 'add').
* udevadm info gained new switches --property=/--value for showing only
specific udev properties/values instead of all.
* A new hwdb database has been added that contains matches for various
types of signal analyzers (protocol analyzers, logic analyzers,
oscilloscopes, multimeters, bench power supplies, etc.) that should
be accessible to regular users.
* A new hwdb database entry has been added that carries information
about types of cameras (regular or infrared), and in which direction
they point (front or back).
* A new rule to allow console users access to rfkill by default has been
added to hwdb.
* Device nodes for the Software Guard eXtension enclaves (sgx_vepc) are
now also owned by the system group "sgx".
* A new build-time meson option "extra-net-naming-schemes=" has been
added to define additional naming schemes schemes for udev's network
interface naming logic. This is useful for enterprise distributions
and similar which want to pin the schemes of certain distribution
releases under a specific name and previously had to patch the
sources to introduce new named schemes.
* The predictable naming logic for network interfaces has been extended
to generate stable names from Xen netfront device information.
* hostnamed's chassis property can now be sourced from chassis-type
field encoded in devicetree (in addition to the existing DMI
support).
* systemd-cgls now optionally displays cgroup IDs and extended
attributes for each cgroup. (Controllable via the new --xattr= +
--cgroup-id= switches.)
* coredumpctl gained a new --all switch for operating on all
Journal files instead of just the local ones.
* systemd-coredump will now use libdw/libelf via dlopen() rather than
directly linking, allowing users to easily opt-out of backtrace/metadata
analysis of core files, and reduce image sizes when this is not needed.
* systemd-coredump will now analyze core files with libdw/libelf in a
forked, sandboxed process.
* systemd-homed will now try to unmount an activate home area in
regular intervals once the user logged out fully. Previously this was
attempted exactly once but if the home directory was busy for some
reason it was not tried again.
* systemd-homed's LUKS2 home area backend will now create a BSD file
system lock on the image file while the home area is active
(i.e. mounted). If a home area is found to be locked, logins are
politely refused. This should improve behavior when using home areas
images that are accessible via the network from multiple clients, and
reduce the chance of accidental file system corruption in that case.
* Optionally, systemd-homed will now drop the kernel buffer cache once
a user has fully logged out, configurable via the new --drop-caches=
homectl switch.
* systemd-homed now makes use of UID mapped mounts for the home areas.
If the kernel and used file system support it, files are now
internally owned by the "nobody" user (i.e. the user typically used