3 * Many manager configuration settings that are only applicable to user
4 manager or system manager can be always set. It would be better to reject
5 them when parsing config.
7 * userdbctl: "Password OK: yes" is shown even when there are no passwords
8 or the password is locked.
12 * Fedora: add an rpmlint check that verifies that all unit files in the RPM are listed in %systemd_post macros.
16 * Rearrange tests so that the various test-xyz.c match a specific src/basic/xyz.c again
18 * rework mount.c and swap.c to follow proper state enumeration/deserialization
19 semantics, like we do for device.c now
23 * add an Open= setting to service unit files that can open arbitrary file
24 system paths at service startup time and pass them to the service process via
25 our usual socket activation protocol. If passed path refers to AF_UNIX
26 socket: connect() to it.
28 * add a ConnectSocket= setting to service unit files, that may reference a
29 socket unit, and which will connect to the socket defined therein, and pass
30 the resulting fd to the service program via socket activation proto.
32 * Add a concept of ListenStream=anonymous to socket units: listen on a socket
33 that is deleted in the fs. Usecase would be with ConnectSocket= above.
35 * Hook up journald's FSS logic with TPM2: seal the verification disk by
36 time-based policy, so that the verification key can remain on host and ve
39 * sd-event: port to new kernel API epoll_wait2() (new in 5.11), to get more
40 accurate wait timeouts
42 * sd-boot: define a drop-in dir in the ESP that may contain X.509
43 certificates. If the firmware is detected to be in setup mode, automatically
44 enroll them as PK/KEK/db, turn off setup mode and proceed. Optionally,
45 instead of auto-enrolling them add them to the sd-boot menu, giving the user
46 the option to manually enroll them, after selecting the menu entry. This way,
47 installer images can just drop the certfiicates in the ESP, and on first boot
48 can easily enroll the keys without ever booting up.
50 * efi stub: optionally, load initrd from disk as a separate file, HMAC check it
51 with key from TPM, bound to PCR, refusing if failing. This would then allow
52 traditional distros that generate initrds locally to secure them with TPM:
53 after generating the initrd, do the HMAC calculation, put result in initrd
54 filename, done. This would then bind the validity of the initrd to the local
55 host, and used kernel, and means people cannot change initrd or kernel
56 without booting the kernel + initrd.
58 * importd: add ability download images for portabled + sysext
60 * importd: support image signature verification with PKCS#7 + OpenBSD signify
61 logic, as alternative to crummy gpg
63 * sd-boot: add service that automatically runs "bootctl update" on every boot,
64 in a graceful way, so that updated /usr trees automatically propagate into
65 updated boot loaders on reboot.
67 * sysext: optionally, if the merged trees allow it use bind mounts instead of
70 * nspawn: add support for sysext extensions, too. i.e. a new --extension=
71 switch that takes one or more arguments, and applies the extensions already
74 * add "systemd-analyze debug" + AttachDebugger= in unit files: The former
75 specifies a command to execute; the latter specifies that an already running
76 "systemd-analyze debug" instance shall be contacted and execution paused
77 until it gives an OK. That way, tools like gdb or strace can be safely be
78 invoked on processes forked off PID 1.
80 * expose MS_NOSYMFOLLOW in various places
82 * tpm2: support a PIN policy, i.e. allowing windows-style short authentication
83 passwords by using the TPM2 to enforce ratelimiting and such, use for
86 * Add concept for upgrading TPM2 enrollments, maybe a new switch
87 --pcrs=4:<hash> or so, i.e. select a PCR to include in the hash, and then
90 * homed: store PKCS#11 + FIDO2 token info in LUKS2 header, compatible with
91 systemd-cryptsetup, so that it can unlock homed volumes
93 * cryptenroll: politely refuse enrolling new keys to homed volumes, since we
94 we cannot update identity info
96 * TPM2: auto-reenroll in cryptsetup, as fallback for hosed firmware upgrades
99 * cryptsetup: if only recovery keys are registered and no regular passphrases,
100 ask user for "recovery key", not "passphrase"
102 * cyptsetup: add option for automatically removing empty password slot on boot
104 * cryptsetup: optionally, when run during boot-up and password is never
105 entered, and we are on battery power (or so), power off machine again
107 * cryptsetup: when FIDO2/PKCS#11/TPM2 token/chip didn't show up after some
108 time, abort the attempt, fallback to asking for pw
110 * cryptsetup: when waiting for FIDO2/PKCS#11 token, tell plymouth that, and
111 allow plymouth to abort the waiting and enter pw instead
113 * when configuring loopback netif, and it fails due to EPERM, eat up error if
114 it happens to be set up alright already.
116 * at boot: check if battery above some threshold, if not power off again after explanation
118 * userdb: add field for ambient caps, so that a user can have CAP_WAKE_ALARM
119 for example. And add code that resets ambient caps for all services by
122 * homed: try to unmount in regular intervals when home dir was busy when we
125 * sd-bus: when connecting to some dbus server socker, set originating AF_UNIX
126 socket name in abstract namespace to include "description" string, and pick
127 it up from there in sd_bus_creds logic. i.e. we can use the socket peer
128 address as conduit for some minimal connection metainfo, and use it to
129 restore the "description" logic that kdbus used to have.
131 * teach LoadCredential= the ability to load all files from a specified dir as
134 * systemd-analyze netif that explains predictable interface (or networkctl)
136 * port selinux code from mallinfo() to mallinfo2() once added to glibc
138 * Add service setting to run a service within the specified VRF. i.e. do the
139 equivalent of "ip vrf exec".
141 * export action of device object on sd-device, so that monitor becomes useful
143 * add root=tmpfs that mounts a tmpfs to /sysroot (to be used in combination
144 with usr=…, for a similar effect as systemd.volatile=yes but without the
145 "hide-out" effect). Also, add root=gpt-auto-late support or so, that is like
146 root=gpt-auto but initially mounts a tmpfs to /sysroot, and then revisits
147 later after systemd-repart ran. Usecase: let's ship images with only /usr
148 partition, then on first boot create the root partition. In this case we want
149 to read the repart data from /usr before the root partition exists. Add
150 usr=gpt-auto that automatically finds a /usr partition.
152 * change SwitchRoot() implementation in PID 1 to use pivot_root(".", "."), as
153 documented in the pivot_root(2) man page, so that we can drop the /oldroot
156 * special case some calls of chase_symlinks() to use openat2() internally, so
157 that the kernel does what we otherwise do.
159 * homed: keep an fd to the homedir open at all times, to keep the fs pinned
160 (autofs and such) while user is logged in.
162 * nss-systemd: also synthesize shadow records for users/groups
164 * make use of new glibc 2.32 APIs sigabbrev_np() and strerrorname_np().
166 * when main nspawn supervisor process gets suspended due to SIGSTOP/SIGTTOU or
167 so, freeze the payload too.
169 * repart: support setting up dm-integrity with HMAC
171 * add /etc/integritytab, to support dm-integrity setups. In particular those
172 with HMAC as hash function, so that we can have a protected /home without
173 encryption (leaving encryption to the individual dirs/homed).
175 * complement root=, rootflags=, rootfstype= with rootsubdir= which allows
176 mounting a subdir of the root fs as actual root. This can be used as
177 fstype-agnostic version of btrfs' rootflags=subvol=foobar.
179 * if /usr/bin/swapoff fails due to OOM, log a friendly explanatory message about it
181 * build short web pages out of each catalog entry, build them along with man
182 pages, and include hyperlinks to them in the journal output
184 * machined: add API to acquire UID range. add API to mount/dissect loopback
185 file. Both protected by PK. Then make nspawn use these APIs to run
186 unprivileged containers. i.e. push the truly privileged bits into machined,
187 so that the client side can remain entirely unprivileged, with SUID or
190 * journald: do journal file writing out-of-process, with one writer process per
191 client UID, so that synthetic hash table collisions can slow down a specific
192 user's journal stream down but not the others.
194 * nspawn: support time namespaces
196 * systemd-firstboot: make sure to always use chase_symlinks() before
197 reading/writing files
199 * Remove any support for booting without /usr pre-mounted in the initrd entirely.
200 Update INITRD_INTERFACE.md accordingly.
202 * pid1: Move to tracking of main pid/control pid of units per pidfd
204 * pid1: support new clone3() fork-into-cgroup feature
206 * pid1: also remove PID files of a service when the service starts, not just
209 * make us use dynamically fewer deps for containers in general purpose distros:
210 o turn into dlopen() deps:
212 - p11-kit-trust (always)
213 - kmod-libs (only when called from PID 1)
214 - libblkid (only in RootImage= handling in PID 1, but not elsewhere)
215 - libpam (only when called from PID 1)
216 - bzip2, xz, lz4 (always — gzip and zstd should probably stay static deps the way they are,
217 since they are so basic and our defaults)
218 o move into separate libsystemd-shared-iptables.so .so
219 - iptables-libs (only used by nspawn + networkd)
221 * seccomp: maybe use seccomp_merge() to merge our filters per-arch if we can.
222 Apparently kernel performance is much better with fewer larger seccomp
223 filters than with more smaller seccomp filters.
225 * systemd-path: add ESP and XBOOTLDR path. Add "private" runtime/state/cache dir enum,
226 mapping to $RUNTIME_DIRECTORY, $STATE_DIRECTORY and such
228 * All tools that support --root= should also learn --image= so that they can
229 operate on disk images directly. Specifically: bootctl, systemctl,
230 coredumpctl. (Already done: systemd-nspawn, systemd-firstboot,
231 systemd-repart, systemd-tmpfiles, systemd-sysusers, journalctl)
233 * seccomp: by default mask x32 ABI system wide on x86-64. it's on its way out
235 * seccomp: don't install filters for ABIs that are masked anyway for the
238 * seccomp: maybe merge all filters we install into one with that libseccomp API that allows merging.
240 * credentials system:
241 - maybe add AcquireCredential= for querying a cred via ask-password
242 - maybe try to acquire creds via keyring?
243 - maybe try to pass creds via keyring?
244 - maybe optionally pass creds via memfd
245 - maybe add support for decrypting creds via TPM
246 - maybe add support for decrypting/importing creds via pkcs11
247 - make systemd-cryptsetup acquire pw via creds logic
248 - make PAMName= acquire pw via creds logic
249 - make macsec/wireguard code in networkd read key via creds logic
250 - make gatwayd/remote read key via creds logic
251 - add sd_notify() command for flushing out creds not needed anymore
253 * homed: during login resize fs automatically towards size goal. Specifically,
254 resize to diskSize if possible, but leave a certain amount (configured by a
255 new value diskLeaveFreeSize) of space free on the backing fs.
257 * homed: permit multiple user record signing keys to be used locally, and pick
258 the right one for signing records automatically depending on a pre-existing
261 * homed: add a way to "adopt" a home directory, i.e. strip foreign signatures
262 and insert a local signature instead.
264 * homed: as an extension to the directory+subvolume backend: if located on
265 especially marked fs, then sync down password into LUKS header of that fs,
266 and always verify passwords against it too. Bootstrapping is a problem
267 though: if no one is logged in (or no other user even exists yet), how do you
268 unlock the volume in order to create the first user and add the first pw.
270 * homed: support new FS_IOC_ADD_ENCRYPTION_KEY ioctl for setting up fscrypt
272 * homed: maybe pre-create ~/.cache as subvol so that it can have separate quota
275 * busctl: maybe expose a verb "ping" for pinging a dbus service to see if it
278 * when systemd-nspawn and suchlike dissect an OS image, and there are multiple
279 root partitions, do an strverscmp() on the partition label and boot
280 first. That is inspired how sd-boot figures out which kernel to boot, and
281 thus allows defining OS images which can be A/B updated and we default to the
282 newest version automatically, both in nspawn and in sd-boot
284 * systemd-gpt-auto should probably set x-systemd.growfs on the mounts it
288 - teach it to prepare an ESP wholesale, i.e. with mkfs.vfat invocation
289 - teach it to copy in unified kernel images and maybe type #1 boot loader spec entries from host
290 - make it operate on loopback files, dissecting enough to find ESP to operate on
292 * Maybe add a separate GPT partition type to the discoverable partition spec
293 for "hibernate" partitions, that are exactly like swap partitions but only
294 activated right before hibernation and thus never used for regular swapping.
296 * by default, in systemd --user service bump the OOMAdjust to 100, as privs
297 allow so that systemd survives
299 * cryptsetup: allow encoding key directly in /etc/crypttab, maybe with a
300 "base64:" prefix. Useful in particular for pkcs11 mode.
302 * cryptsetup: reimplement the mkswap/mke2fs in cryptsetup-generator to use
303 systemd-makefs.service instead.
305 * socket units: allow creating a udev monitor socket with ListenDevices= or so,
306 with matches, then activate app through that passing socket over
308 * unify on openssl (as soon as OpenSSL 3.0 is out, and the Debian license
310 - port sd_id128_get_machine_app_specific() over from khash
311 - port resolved over from libgcrypt (DNSSEC code)
312 - port journald + fsprg over from libgcrypt
313 - port importd over from libgcrypt
314 - when that's done: kill khash.c
315 - when that's done: kill gnutls support in resolved
317 * when we resize disks (homed?) always round up to 4K sectors, not 512K
319 * add growvol and makevol options for /etc/crypttab, similar to
320 x-systemd.growfs and x-systemd-makefs.
322 * systemd-repart: by default generate minimized partition tables (i.e. tables
323 that only cover the space actually used, excluding any free space at the
324 end), in order to maximize dd'ability. Requires libfdisk work, see
325 https://github.com/karelzak/util-linux/issues/907
327 * systemd-repart: MBR partition table support. Care needs to be taken regarding
328 Type=, so that partition definitions can sanely apply to both the GPT and the
329 MBR case. Idea: accept syntax "Type=gpt:home mbr:0x83" for setting the types
330 for the two partition types explicitly. And provide an internal mapping so
331 that "Type=linux-generic" maps to the right types for both partition tables
334 * systemd-repart: allow sizing partitions as factor of available RAM, so that
335 we can reasonably size swap partitions for hibernation.
337 * systemd-repart: allow managing the gpt read-only partition flag + auto-mount flag
339 * systemd-repart: allow boolean option that ensures that if existing partition
340 doesn't exist within the configured size bounds the whole command fails. This
341 is useful to implement ESP vs. XBOOTLDR schemes in installers: have one set
342 of repart files for the case where ESP is large enough and one where it isn't
343 and XBOOTLDR is added in instead. Then apply the former first, and if it
344 fails to apply use the latter.
346 * systemd-repart: add per-partition option to never reuse existing partition
347 and always create anew even if matching partition already exists.
349 * systemd-repart: add per-partition option to fail if partition already exist,
350 i.e. is not added new. Similar, add option to fail if partition does not exist yet.
352 * systemd-repart: allow disabling growing of specific partitions, or making
353 them (think ESP: we don't ever want to grow it, since we cannot resize vfat)
355 * systemd-repart: make it a static checker during early boot for existence and
356 absence of other partitions for trusted boot environments
358 * userdb: allow username prefix searches in varlink API, allow realname and
359 realname substr searches in varlink API
361 * userdb: allow uid/gid range checks
363 * userdb: allow existence checks
365 * pid1: activation by journal search expression
367 * when switching root from initrd to host, set the machine_id env var so that
368 if the host has no machine ID set yet we continue to use the random one the
371 * sd-event: add native support for P_ALL waitid() watching, then move PID 1 to
372 it for reaping assigned but unknown children. This needs to some special care
373 to operate somewhat sensibly in light of priorities: P_ALL will return
374 arbitrary processes, regardless of the priority we want to watch them with,
375 hence on each event loop iteration check all processes which we shall watch
376 with higher prio explicitly, and then watch the entire rest with P_ALL.
378 * tweak sd-event's child watching: keep a prioq of children to watch and use
379 waitid() only on the children with the highest priority until one is waitable
380 and ignore all lower-prio ones from that point on
382 * maybe introduce xattrs that can be set on the root dir of the root fs
383 partition that declare the volatility mode to use the image in. Previously I
384 thought marking this via GPT partition flags but that's not ideal since
385 that's outside of the LUKS encryption/verity verification, and we probably
386 shouldn't operate in a volatile mode unless we got told so from a trusted
389 * figure out automatic partition discovery when combining writable root dir
392 * coredump: maybe when coredumping read a new xattr from /proc/$PID/exe that
393 may be used to mark a whole binary as non-coredumpable. Would fix:
394 https://bugs.freedesktop.org/show_bug.cgi?id=69447
396 * teach parse_timestamp() timezones like the calendar spec already knows it
398 * beef up hibernation to optionally do swapon/swapoff immediately before/after
401 * beef up s2h to implement a battery watch loop: instead of entering
402 hibernation unconditionally after coming back from resume make a decision
403 based on the battery load level: if battery level is above a specific
404 threshold, go to suspend again, only hibernate if below it. This means we'd
405 stick to suspend usually, but fall back to hibernation only when battery runs
406 empty (well, subject to our sampling interval). Related to this, check if we
407 can make ACPI _BTP (i.e. /sys/class/power_supply/*/alarm) work for us too,
408 i.e. see if it can wake up machines from suspend, so that we could resume
409 automatically when the system is low on power and move automatically to
410 hibernation mode. (see
411 https://uefi.org/sites/default/files/resources/ACPI%206_2_A_Sept29.pdf
413 https://docs.microsoft.com/en-us/windows-hardware/design/device-experiences/modern-standby-wake-sources
416 * add an explicit "vertical" mode to format-table, so that "systemctl
417 status"-like outputs (i.e. with a series of field names left and values
418 right) become genuine first class citizens, and we gain automatic, sane JSON
421 * We should probably replace /var/log/README, /etc/rc.d/README with symlinks
422 that are linked to these places instead of copied. After all they are
423 constant vendor data.
425 * maybe add kernel cmdline params: to force random seed crediting
427 * nspawn: on cgroupsv1 issue cgroup empty handler process based on host events,
428 so that we make cgroup agent logic safe
430 * nspawn/machined: add API to invoke binary in container, then use that as
431 fallback in "machinectl shell"
433 * logind: rework pam_logind to also do a bus call in case of invocation from
434 user@.service, which returns the XDG_RUNTIME_DIR value, and make this
435 behaviour selectable via pam module option.
438 - when user tries to log into record signed by unrecognized key, automatically add key to our chain after polkit auth
439 - rollback when resize fails mid-operation
440 - GNOME's side for forget key on suspend (requires rework so that lock screen runs outside of uid)
442 - shrink fs on logout?
443 - update LUKS password on login if we find there's a password that unlocks the JSON record but not the LUKS device.
444 - create on activate?
445 - properties: icon url?, preferred session type?, administrator bool (which translates to 'wheel' membership)?, address?, telephone?, vcard?, samba stuff?, parental controls?
446 - communicate clearly when usb stick is safe to remove. probably involves
447 beefing up logind to make pam session close hook synchronous and wait until
448 systemd --user is shut down.
449 - logind: maybe keep a "busy fd" as long as there's a non-released session around or the user@.service
450 - maybe make automatic, read-only, time-based reflink-copies of LUKS disk
451 images (and btrfs snapshots of subvolumes) (think: time machine)
452 - distinguish destroy / remove (i.e. currently we can unregister a user, unregister+remove their home directory, but not just remove their home directory)
453 - in systemd's PAMName= logic: query passwords with ssh-askpassword, so that we can make "loginctl set-linger" mode work
454 - fingerprint authentication, pattern authentication, …
455 - make sure "classic" user records can also be managed by homed
456 - make size of $XDG_RUNTIME_DIR configurable in user record
457 - query password from kernel keyring first
458 - update even if record is "absent"
459 - add a "access mode" + "fstype" field to the "status" section of json identity records reflecting the actually used access mode and fstype, even on non-luks backends
460 - move acct mgmt stuff from pam_systemd_home to pam_systemd?
461 - when "homectl --pkcs11-token-uri=" is used, synthesize ssh-authorized-keys records for all keys we have private keys on the stick for
462 - make slice for users configurable (requires logind rework)
463 - logind: populate auto-login list bus property from PKCS#11 token
464 - when determining state of a LUKS home directory, check DM suspended sysfs file
465 - introduce API for "making room", that grows/shrinks home directory
466 according to elastic parameters, discards blocks, and removes additional snapshots. Call it
467 either from UI when disk space gets low
469 * introduce a new per-process uuid, similar to the boot id, the machine id, the
470 invocation id, that is derived from process creds, specifically a hashed
471 combination of AT_RANDOM + getpid() + the starttime from
472 /proc/self/status. Then add these ids implicitly when logging. Deriving this
473 uuid from these three things has the benefit that it can be derived easily
474 from /proc/$PID/ in a stable, and unique way that changes on both fork() and
477 * let's not GC a unit while its ratelimits are still pending
479 * when killing due to service watchdog timeout maybe detect whether target
480 process is under ptracing and then log loudly and continue instead.
482 * introduce a new group to own TPM devices
484 * make rfkill uaccess controllable by default, i.e. steal rule from
485 gnome-bluetooth and friends
487 * tweak journald context caching. In addition to caching per-process attributes
488 keyed by PID, cache per-cgroup attributes (i.e. the various xattrs we read)
489 keyed by cgroup path, and guarded by ctime changes. This should provide us
490 with a nice speed-up on services that have many processes running in the same
493 * make MAINPID= message reception checks even stricter: if service uses User=,
494 then check sending UID and ignore message if it doesn't match the user or
497 * maybe trigger a uevent "change" on a device if "systemctl reload xyz.device"
500 * when importing an fs tree with machined, optionally apply userns-rec-chown
502 * when importing an fs tree with machined, complain if image is not an OS
504 * Maybe introduce a helper safe_exec() or so, which is to execve() which
505 safe_fork() is to fork(). And then make revert the RLIMIT_NOFILE soft limit
506 to 1K implicitly, unless explicitly opted-out.
508 * rework seccomp/nnp logic that even if User= is used in combination with
509 a seccomp option we don't have to set NNP. For that, change uid first whil
510 keeping CAP_SYS_ADMIN, then apply seccomp, the drop cap.
512 * add a concept for automatically loading per-unit secrets off disk and
513 inserting them into the kernel keyring. Maybe SecretsDirectory= similar to
514 ConfigurationDirectory=.
516 * when no locale is configured, default to UEFI's PlatformLang variable
518 * bootctl,sd-boot: actually honour the "architecture" key
520 * add a new syscall group "@esoteric" for more esoteric stuff such as bpf() and
521 usefaultd() and make systemd-analyze check for it.
523 * paranoia: whenever we process passwords, call mlock() on the memory
524 first. i.e. look for all places we use free_and_erasep() and
525 augment them with mlock(). Also use MADV_DONTDUMP.
527 * Move RestrictAddressFamily= to the new cgroup create socket
529 * support the bind/connect/sendmsg cgroup stuff for sandboxing, and possibly
532 * maybe implicitly attach monotonic+realtime timestamps to outgoing messages in
533 log.c and sd-journal-send
535 * optionally: turn on cgroup delegation for per-session scope units
537 * introduce per-unit (i.e. per-slice, per-service) journal log size limits.
539 * sd-boot: automatically load EFI modules from some drop-in dir, so that people
540 can add in file system drivers and such
542 * sd-boot: optionally, show boot menu when previous default boot item has
543 non-zero "tries done" count
545 * introduce an option (or replacement) for "systemctl show" that outputs all
546 properties as JSON, similar to busctl's new JSON output. In contrast to that
547 it should skip the variant type string though.
549 * augment CODE_FILE=, CODE_LINE= with something like CODE_BASE= or so which
550 contains some identifier for the project, which allows us to include
551 clickable links to source files generating these log messages. The identifier
552 could be some abberviated URL prefix or so (taking inspiration from Go
553 imports). For example, for systemd we could use
554 CODE_BASE=github.com/systemd/systemd/blob/98b0b1123cc or so which is
555 sufficient to build a link by prefixing "http://" and suffixing the
558 * Augment MESSAGE_ID with MESSAGE_BASE, in a similar fashion so that we can
559 make clickable links from log messages carrying a MESSAGE_ID, that lead to
560 some explanatory text online.
562 * maybe extend .path units to expose fanotify() per-mount change events
564 * Add a "systemctl list-units --by-slice" mode or so, which rearranges the
565 output of "systemctl list-units" slightly by showing the tree structure of
566 the slices, and the units attached to them.
568 * the a-posteriori stopping of units bound to units that disappeared logic
569 should be reworked: there should be a queue of units, and we should only
570 enqueue stop jobs from a defer event that processes queue instead of
571 right-away when we find a unit that is bound to one that doesn't exist
572 anymore. (similar to how the stop-unneeded queue has been reworked the same
575 * nspawn: make nspawn suitable for shell pipelines: instead of triggering a
576 hangup when input is finished, send ^D, which synthesizes an EOF. Then wait
577 for hangup or ^D before passing on the EOF.
579 * When reloading configuration PID 1 should reset all its properties to the
580 original defaults before calling parse_config()
582 * nspawn: greater control over selinux label?
584 * hibernate/s2h: make this robust and safe to enable in Fedora by default.
587 1. add resume_offset support to the resume code (i.e. support swap files
589 2. check if swap is on weird storage and refuse if so
590 3. add auto-detection of hibernation images
592 * cgroups: use inotify to get notified when somebody else modifies cgroups
593 owned by us, then log a friendly warning.
595 * beef up log.c with support for stripping ANSI sequences from strings, so that
596 it is OK to include them in log strings. This would be particularly useful so
597 that our log messages could contain clickable links for example for unit
598 files and suchlike we operate on.
600 * add support for "portablectl attach http://foobar.com/waaa.raw (i.e. importd integration)
602 * add attach --enable and attach --now (for attach+enable+start)
604 * sync dynamic uids/gids between host+portable srvice (i.e. if DynamicUser=1 is set for a service, make sure that the
605 selected user is resolvable in the service even if it ships its own /etc/passwd)
607 * Fix DECIMAL_STR_MAX or DECIMAL_STR_WIDTH. One includes a trailing NUL, the
608 other doesn't. What a disaster. Probably to exclude it. Also
609 DECIMAL_STR_WIDTH should probably add an extra "-" into account for negative
612 * Check that users of inotify's IN_DELETE_SELF flag are using it properly, as
613 usually IN_ATTRIB is the right way to watch deleted files, as the former only
614 fires when a file is actually removed from disk, i.e. the link count drops to
615 zero and is not open anymore, while the latter happens when a file is
616 unlinked from any dir.
618 * port systemctl, busctl, … over to format-table.[ch]'s table formatters
620 * pid1: lock image configured with RootDirectory=/RootImage= using the usual nspawn semantics while the unit is up
622 * add --vacuum-xyz options to coredumpctl, matching those journalctl already has.
624 * introduce Ephemeral= unit file switch, that creates an ephemeral copy of all
625 files and directories that are left writable for a unit, and which are
626 removed after the unit goes down again. A bit like --ephemeral for
627 systemd-nspawn but for system services. If used together with RootImage= this
628 should reflink the image file itself.
630 Related: add Ephemeral=<path1> <path2> … which would allow marking
631 specific paths only like this.
633 * add CopyFile= or so as unit file setting that may be used to copy files or
634 directory trees from the host to the services RootImage= and RootDirectory=
635 environment. Which we can use for /etc/machine-id and in particular
636 /etc/resolv.conf. Should be smart and do something useful on read-only
637 images, for example fall back to read-only bind mounting the file instead.
639 * show invocation ID in systemd-run output
641 * bypass SIGTERM state in unit files if KillSignal is SIGKILL
643 * add proper dbus APIs for the various sd_notify() commands, such as MAINPID=1
644 and so on, which would mean we could report errors and such.
646 * teach tmpfiles.d q/Q logic something sensible in the context of XFS/ext4
649 * introduce DefaultSlice= or so in system.conf that allows changing where we
650 place our units by default, i.e. change system.slice to something
651 else. Similar, ManagerSlice= should exist so that PID1's own scope unit could
652 be moved somewhere else too. Finally machined and logind should get similar
653 options so that it is possible to move user session scopes and machines to a
654 different slice too by default. Usecase: people who want to put resources on
655 the entire system, with the exception of one specific service. See:
656 https://lists.freedesktop.org/archives/systemd-devel/2018-February/040369.html
658 * maybe rework get_user_creds() to query the user database if $SHELL is used
659 for root, but only then.
661 * be stricter with fds we receive for the fdstore: close them asynchronously
663 * calenderspec: add support for week numbers and day numbers within a
664 year. This would allow us to define "bi-weekly" triggers safely.
666 * sd-bus: add vtable flag, that may be used to request client creds implicitly
667 and asynchronously before dispatching the operation
669 * sd-bus: parse addresses given in sd_bus_set_addresses immediately and not
670 only when used. Add unit tests.
672 * make use of ethtool veth peer info in machined, for automatically finding out
673 host-side interface pointing to the container.
675 * add some special mode to LogsDirectory=/StateDirectory=… that allows
676 declaring these directories without necessarily pulling in deps for them, or
677 creating them when starting up. That way, we could declare that
678 systemd-journald writes to /var/log/journal, which could be useful when we
679 doing disk usage calculations and so on.
681 * taint systemd if there are fewer than 65536 users assigned (userns) to the system.
683 * deprecate RootDirectoryStartOnly= in favour of a new ExecStart= prefix char
685 * add a new RuntimeDirectoryPreserve= mode that defines a similar lifecycle for
686 the runtime dir as we maintain for the fdstore: i.e. keep it around as long
687 as the unit is running or has a job queued.
689 * support projid-based quota in machinectl for containers
691 * add a way to lock down cgroup migration: a boolean, which when set for a unit
692 makes sure the processes in it can never migrate out of it
694 * blog about fd store and restartable services
696 * document Environment=SYSTEMD_LOG_LEVEL=debug drop-in in debugging document
698 * rework ExecOutput and ExecInput enums so that EXEC_OUTPUT_NULL loses its
699 magic meaning and is no longer upgraded to something else if set explicitly.
701 * in the long run: permit a system with /etc/machine-id linked to /dev/null, to
702 make it lose its identity, i.e. be anonymous. For this we'd have to patch
703 through the whole tree to make all code deal with the case where no machine
706 * optionally, collect cgroup resource data, and store it in per-unit RRD files,
707 suitable for processing with rrdtool. Add bus API to access this data, and
708 possibly implement a CPULoad property based on it.
710 * beef up pam_systemd to take unit file settings such as cgroups properties as
713 * maybe hook of xfs/ext4 quotactl() with services? i.e. automatically manage
714 the quota of the user indicated in User= via unit file settings, like the
715 other resource management concepts. Would mix nicely with DynamicUser=1. Or
716 alternatively, do this with projids, so that we can also cover services
717 running as root. Quota should probably cover all the special dirs such as
718 StateDirectory=, LogsDirectory=, CacheDirectory=, as well as RootDirectory= if it
719 is set, plus the whole disk space any image configured with RootImage=.
721 * In DynamicUser= mode: before selecting a UID, use disk quota APIs on relevant
722 disks to see if the UID is already in use.
724 * add "systemctl wait" or so, which does what "systemd-run --wait" does, but
725 for all units. It should be both a way to pin units into memory as well as a
726 wait to retrieve their exit data.
728 * expose IO accounting data on the bus, show it in systemd-run --wait and log
729 about it in the resource log message
731 * show whether a service has out-of-date configuration in "systemctl status" by
732 using mtime data of ConfigurationDirectory=.
734 * Add AddUser= setting to unit files, similar to DynamicUser=1 which however
735 creates a static, persistent user rather than a dynamic, transient user. We
736 can leverage code from sysusers.d for this.
738 * add some optional flag to ReadWritePaths= and friends, that has the effect
739 that we create the dir in question when the service is started. Example:
741 ReadWritePaths=:/var/lib/foobar
743 * maybe add call sd_journal_set_block_timeout() or so to set SO_SNDTIMEO for
744 the sd-journal logging socket, and, if the timeout is set to 0, sets
745 O_NONBLOCK on it. That way people can control if and when to block for
748 * hostnamed: populate form factor data from a new hwdb database, so that old
749 yogas can be recognized as "convertible" too, even if they predate the DMI
750 "convertible" form factor
752 * Add ExecMonitor= setting. May be used multiple times. Forks off a process in
753 the service cgroup, which is supposed to monitor the service, and when it
754 exits the service is considered failed by its monitor.
756 * track the per-service PAM process properly (i.e. as an additional control
757 process), so that it may be queried on the bus and everything.
759 * add a new "debug" job mode, that is propagated to unit_start() and for
760 services results in two things: we raise SIGSTOP right before invoking
761 execve() and turn off watchdog support. Then, use that to implement
762 "systemd-gdb" for attaching to the start-up of any system service in its
765 * gpt-auto logic: related to the above, maybe support a "secondary" root
766 partition, that is mounted to / and is writable, and where the actual root's
767 /usr is mounted into.
769 * gpt-auto logic: support encrypted swap, add kernel cmdline option to force it, and honour a gpt bit about it, plus maybe a configuration file
771 * drop nss-myhostname in favour of nss-resolve?
773 * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and
774 then use that for the setting used in user@.service. It should be understood
775 relative to the configured default value.
777 * in networkd, when matching device types, fix up DEVTYPE rubbish the kernel passes to us
779 * enable LockMLOCK to take a percentage value relative to physical memory
781 * Permit masking specific netlink APIs with RestrictAddressFamily=
783 * nspawn: support that /proc, /sys/, /dev are pre-mounted
785 * define gpt header bits to select volatility mode
787 * ProtectClock= (drops CAP_SYS_TIMES, adds seecomp filters for settimeofday, adjtimex), sets DeviceAllow o /dev/rtc
789 * ProtectTracing= (drops CAP_SYS_PTRACE, blocks ptrace syscall, makes /sys/kernel/tracing go away)
791 * ProtectMount= (drop mount/umount/pivot_root from seccomp, disallow fuse via DeviceAllow, imply Mountflags=slave)
793 * ProtectKeyRing= to take keyring calls away
795 * RemoveKeyRing= to remove all keyring entries of the specified user
797 * ProtectReboot= that masks reboot() and kexec_load() syscalls, prohibits kill
798 on PID 1 with the relevant signals, and makes relevant files in /sys and
799 /proc (such as the sysrq stuff) unavailable
801 * make sure the ratelimit object can deal with USEC_INFINITY as way to turn off things
803 * journalctl: make sure -f ends when the container indicated by -M terminates
805 * mount: automatically search for "main" partition of an image has multiple
808 * in nss-systemd, if we run inside of RootDirectory= with PrivateUsers= set,
809 find a way to map the User=/Group= of the service to the right name. This way
810 a user/group for a service only has to exist on the host for the right
813 * add bus API for creating unit files in /etc, reusing the code for transient units
815 * add bus API to remove unit files from /etc
817 * add bus API to retrieve current unit file contents (i.e. implement "systemctl cat" on the bus only)
819 * rework fopen_temporary() to make use of open_tmpfile_linkable() (problem: the
820 kernel doesn't support linkat() that replaces existing files, currently)
822 * transient units: don't bother with actually setting unit properties, we
823 reload the unit file anyway
825 * journald: sigbus API via a signal-handler safe function that people may call
826 from the SIGBUS handler
828 * optionally, also require WATCHDOG=1 notifications during service start-up and shutdown
830 * delay activation of logind until somebody logs in, or when /dev/tty0 pulls it
831 in or lingering is on (so that containers don't bother with it until PAM is used). also exit-on-idle
833 * cache sd_event_now() result from before the first iteration...
835 * PID1: find a way how we can reload unit file configuration for
836 specific units only, without reloading the whole of systemd
838 * add an explicit parser for LimitRTPRIO= that verifies
839 the specified range and generates sane error messages for incorrect
842 * when we detect that there are waiting jobs but no running jobs, do something
844 * push CPUAffinity= also into the "cpuset" cgroup controller
846 * PID 1 should send out sd_notify("WATCHDOG=1") messages (for usage in the --user mode, and when run via nspawn)
848 * there's probably something wrong with having user mounts below /sys,
849 as we have for debugfs. for example, src/core/mount.c handles mounts
850 prefixed with /sys generally special.
851 http://lists.freedesktop.org/archives/systemd-devel/2015-June/032962.html
853 * fstab-generator: default to tmpfs-as-root if only usr= is specified on the kernel cmdline
855 * docs: bring http://www.freedesktop.org/wiki/Software/systemd/MyServiceCantGetRealtime up to date
857 * add a job mode that will fail if a transaction would mean stopping
858 running units. Use this in timedated to manage the NTP service
860 http://lists.freedesktop.org/archives/systemd-devel/2015-April/030229.html
862 * The udev blkid built-in should expose a property that reflects
863 whether media was sensed in USB CF/SD card readers. This should then
864 be used to control SYSTEMD_READY=1/0 so that USB card readers aren't
865 picked up by systemd unless they contain a medium. This would mirror
866 the behaviour we already have for CD drives.
868 * hostnamectl: show root image uuid
870 * Find a solution for SMACK capabilities stuff:
871 http://lists.freedesktop.org/archives/systemd-devel/2014-December/026188.html
873 * "systemctl preset-all" should probably order the unit files it
874 operates on lexicographically before starting to work, in order to
875 ensure deterministic behaviour if two unit files conflict (like DMs
878 * synchronize console access with BSD locks:
879 http://lists.freedesktop.org/archives/systemd-devel/2014-October/024582.html
881 * as soon as we have sender timestamps, revisit coalescing multiple parallel daemon reloads:
882 http://lists.freedesktop.org/archives/systemd-devel/2014-December/025862.html
884 * figure out when we can use the coarse timers
886 * add "systemctl start -v foobar.service" that shows logs of a service
887 while the start command runs. This is non-trivial to do without
888 races though, since we should flush out all journal messages before
889 returning from the "systemctl stop".
891 * firstboot: make it useful to be run immediately after yum --installroot to set up a machine. (most specifically, make --copy-root-password work even if /etc/passwd already exists
893 * maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment=
895 * maybe allow timer units with an empty Units= setting, so that they
896 can be used for resuming the system but nothing else.
898 * what to do about udev db binary stability for apps? (raw access is not an option)
900 * systemctl: if some operation fails, show log output?
902 * exponential backoff in timesyncd when we cannot reach a server
904 * timesyncd: add ugly bus calls to set NTP servers per-interface, for usage by NM
906 * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share....
908 * add systemd.abort_on_kill or some other such flag to send SIGABRT instead of SIGKILL
909 (throughout the codebase, not only PID1)
913 - service registration
914 - service/domain/types browsing
916 - DNS-SD service registration from socket units
917 - resolved should optionally register additional per-interface LLMNR
918 names, so that for the container case we can establish the same name
919 (maybe "host") for referencing the server, everywhere.
920 - allow clients to request DNSSEC for a single lookup even if DNSSEC is off (?)
921 - hook up resolved with machined-based address resolution
923 * refcounting in sd-resolve is borked
925 * Add a new verb "systemctl top"
927 * add new gpt type for btrfs volumes
929 * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them.
931 * a way for container managers to turn off getty starting via $container_headless= or so...
933 * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit
935 * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services
936 they run added to the initial transaction and thus confuse Type=idle.
938 * add bus api to query unit file's X fields.
940 * gpt-auto-generator:
941 - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap?
942 - Make /home automount rather than mount?
944 * add generator that pulls in systemd-network from containers when
945 CAP_NET_ADMIN is set, more than the loopback device is defined, even
946 when it is otherwise off
948 * MessageQueueMessageSize= (and suchlike) should use parse_iec_size().
950 * implement Distribute= in socket units to allow running multiple
951 service instances processing the listening socket, and open this up
955 - implement per-slice CPUFairScheduling=1 switch
956 - introduce high-level settings for RT budget, swappiness
957 - how to reset dynamically changed unit cgroup attributes sanely?
958 - when reloading configuration, apply new cgroup configuration
959 - when recursively showing the cgroup hierarchy, optionally also show
960 the hierarchies of child processes
963 - add field to transient units that indicate whether systemd or somebody else saves/restores its settings, for integration with libvirt
965 * when we detect low battery and no AC on boot, show pretty splash and refuse boot
967 * libsystemd-journal, libsystemd-login, libudev: add calls to easily attach these objects to sd-event event loops
969 * be more careful what we export on the bus as (usec_t) 0 and (usec_t) -1
971 * rfkill,backlight: we probably should run the load tools inside of the udev rules so that the state is properly initialized by the time other software sees it
973 * After coming back from hibernation reset hibernation swap partition using the /dev/snapshot ioctl APIs
975 * If we try to find a unit via a dangling symlink, generate a clean
976 error. Currently, we just ignore it and read the unit from the search
979 * refuse boot if /usr/lib/os-release is missing or /etc/machine-id cannot be set up
981 * man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted.
983 * load .d/*.conf dropins for device units
987 - GetAllProperties() on a non-existing object does not result in a failure currently
988 - port to sd-resolve for connecting to TCP dbus servers
989 - see if we can introduce a new sd_bus_get_owner_machine_id() call to retrieve the machine ID of the machine of the bus itself
990 - see if we can drop more message validation on the sending side
991 - add API to clone sd_bus_message objects
992 - longer term: priority inheritance
994 - NameLost/NameAcquired obsolete
997 - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now
1000 - allow multiple signal handlers per signal?
1001 - document chaining of signal handler for SIGCHLD and child handlers
1002 - define more intervals where we will shift wakeup intervals around in, 1h, 6h, 24h, ...
1003 - maybe support iouring as backend, so that we allow hooking read and write
1004 operations instead of IO ready events into event loops. See considerations
1006 http://blog.vmsplice.net/2020/07/rethinking-event-loop-integration-for.html
1008 * investigate endianness issues of UUID vs. GUID
1010 * dbus: when a unit failed to load (i.e. is in UNIT_ERROR state), we
1011 should be able to safely try another attempt when the bus call LoadUnit() is invoked.
1013 * add a pam module that passes the hdd passphrase into the PAM stack and then expires it, for usage by gdm auto-login.
1015 * add a pam module that on password changes updates any LUKS slot where the password matches
1018 - add unit tests for config_parse_device_allow()
1020 * seems that when we follow symlinks to units we prefer the symlink
1021 destination path over /etc and /usr. We should not do that. Instead
1022 /etc should always override /run+/usr and also any symlink
1025 * when isolating, try to figure out a way how we implicitly can order
1026 all units we stop before the isolating unit...
1028 * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off})
1030 * Add ConditionDirectoryNotEmpty= handle non-absoute paths as a search path or add
1031 ConditionConfigSearchPathNotEmpty= or different syntax? See the discussion starting at
1032 https://github.com/systemd/systemd/pull/15109#issuecomment-607740136.
1034 * BootLoaderSpec: Clarify that the kernel has to be in $BOOT. Clarify
1035 that the boot loader should be installed to the ESP. Define a way
1036 how an installer can figure out whether a BLS compliant boot loader
1039 * think about requeuing jobs when daemon-reload is issued? usecase:
1040 the initrd issues a reload after fstab from the host is accessible
1041 and we might want to requeue the mounts local-fs acquired through
1044 * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep()
1046 * remove any syslog support from log.c — we probably cannot do this before split-off udev is gone for good
1048 * shutdown logging: store to EFI var, and store to USB stick?
1050 * merge unit_kill_common() and unit_kill_context()
1053 - honor language efi variables for default language selection (if there are any?)
1054 - honor timezone efi variables for default timezone selection (if there are any?)
1055 - change bootctl to be backed by systemd-bootd to control temporary and persistent default boot goal plus efi variables
1057 - recognize the case when not booted on EFI
1059 * maybe do not install getty@tty1.service symlink in /etc but in /usr?
1061 * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word
1063 * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units.
1066 - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around
1067 - logind: wakelock/opportunistic suspend support
1068 - Add pretty name for seats in logind
1069 - logind: allow showing logout dialog from system?
1070 - add Suspend() bus calls which take timestamps to fix double suspend issues when somebody hits suspend and closes laptop quickly.
1071 - if pam_systemd is invoked by su from a process that is outside of a
1072 any session we should probably just become a NOP, since that's
1073 usually not a real user session but just some system code that just
1075 - logind: make the Suspend()/Hibernate() bus calls wait for the for
1076 the job to be completed. before returning, so that clients can wait
1077 for "systemctl suspend" to finish to know when the suspending is
1079 - logind: when the power button is pressed short, just popup a
1080 logout dialog. If it is pressed for 1s, do the usual
1081 shutdown. Inspiration are Macs here.
1082 - expose "Locked" property on logind session objects
1083 - maybe allow configuration of the StopTimeout for session scopes
1084 - rename session scope so that it includes the UID. THat way
1085 the session scope can be arranged freely in slices and we don't have
1086 make assumptions about their slice anymore.
1087 - follow PropertiesChanged state more closely, to deal with quick logouts and
1089 - (optionally?) spawn seat-manager@$SEAT.service whenever a seat shows up that as CanGraphical set
1092 - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields
1093 - journald: also get thread ID from client, plus thread name
1094 - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups
1095 - add API to close/reopen/get fd for journal client fd in libsystemd-journal.
1096 - fall back to /dev/log based logging in libsystemd-journal, if we cannot log natively?
1097 - declare the local journal protocol stable in the wiki interface chart
1098 - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg
1099 - journald: when dropping msgs due to ratelimit make sure to write
1100 "dropped %u messages" not only when we are about to print the next
1101 message that works, but already after a short timeout
1102 - check if we can make journalctl by default use --follow mode inside of less if called without args?
1103 - maybe add API to send pairs of iovecs via sd_journal_send
1104 - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access
1105 - journactl: support negative filtering, i.e. FOOBAR!="waldo",
1106 and !FOOBAR for events without FOOBAR.
1107 - journal: store timestamp of journal_file_set_offline() in the header,
1108 so it is possible to display when the file was last synced.
1109 - journal-send.c, log.c: when the log socket is clogged, and we drop, count this and write a message about this when it gets unclogged again.
1110 - journal: find a way to allow dropping history early, based on priority, other rules
1111 - journal: When used on NFS, check payload hashes
1112 - journald: add kernel cmdline option to disable ratelimiting for debug purposes
1113 - refuse taking lower-case variable names in sd_journal_send() and friends.
1114 - journald: we currently rotate only after MaxUse+MaxFilesize has been reached.
1115 - journal: deal nicely with byte-by-byte copied files, especially regards header
1116 - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit
1117 - Replace utmp, wtmp, btmp, and lastlog completely with journal
1118 - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax?
1119 - when a kernel driver logs in a tight loop, we should ratelimit that too.
1120 - journald: optionally, log debug messages to /run but everything else to /var
1121 - journald: when we drop syslog messages because the syslog socket is
1122 full, make sure to write how many messages are lost as first thing
1123 to syslog when it works again.
1124 - journald: allow per-priority and per-service retention times when rotating/vacuuming
1125 - journald: make use of uid-range.h to managed uid ranges to split
1127 - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so...
1128 - improve journalctl performance by loading journal files
1129 lazily. Encode just enough information in the file name, so that we
1130 do not have to open it to know that it is not interesting for us, for
1131 the most common operations.
1132 - man: document that corrupted journal files is nothing to act on
1133 - rework journald sigbus stuff to use mutex
1134 - Set RLIMIT_NPROC for systemd-journal-xyz, and all other of our
1135 services that run under their own user ids, and use User= (but only
1136 in a world where userns is ubiquitous since otherwise we cannot
1137 invoke those daemons on the host AND in a container anymore). Also,
1138 if LimitNPROC= is used without User= we should warn and refuse
1140 - journalctl --verify: don't show files that are currently being
1141 written to as FAIL, but instead show that their are being written to.
1142 - add journalctl -H that talks via ssh to a remote peer and passes through
1144 - add a version of --merge which also merges /var/log/journal/remote
1145 - journalctl: -m should access container journals directly by enumerating
1146 them via machined, and also watch containers coming and going.
1147 Benefit: nspawn --ephemeral would start working nicely with the journal.
1148 - assign MESSAGE_ID to log messages about failed services
1149 - check if loop in decompress_blob_xz() is necessary
1151 * add a test if all entries in the catalog are properly formatted.
1152 (Adding dashes in a catalog entry currently results in the catalog entry
1153 being silently skipped. journalctl --update-catalog must warn about this,
1154 and we should also have a unit test to check that all our message are OK.)
1157 - document that deps in [Unit] sections ignore Alias= fields in
1158 [Install] units of other units, unless those units are disabled
1159 - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets
1160 - document that service reload may be implemented as service reexec
1161 - add a man page containing packaging guidelines and recommending usage of things like Documentation=, PrivateTmp=, PrivateNetwork= and ReadOnlyDirectories=/etc /usr.
1162 - document systemd-journal-flush.service properly
1163 - documentation: recommend to connect the timer units of a service to the service via Also= in [Install]
1164 - man: document the very specific env the shutdown drop-in tools live in
1165 - man: add more examples to man pages,
1166 - in particular an example how to do the equivalent of switching runlevels
1167 - man: maybe sort directives in man pages, and take sections from --help and apply them to man too
1168 - document root=gpt-auto properly
1171 - add systemctl switch to dump transaction without executing it
1172 - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done
1173 - "systemctl disable" on a static unit prints no message and does
1174 nothing. "systemctl enable" does nothing, and gives a bad message
1175 about it. Should fix both to print nice actionable messages.
1176 - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service
1177 - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible
1178 - systemctl enable: fail if target to alias into does not exist? maybe show how many units are enabled afterwards?
1179 - systemctl: "Journal has been rotated since unit was started." message is misleading
1180 - systemctl status output should include list of triggering units and their status
1183 - "systemctl mask" should find all names by which a unit is accessible
1184 (i.e. by scanning for symlinks to it) and link them all to /dev/null
1187 - timer units should get the ability to trigger when:
1189 - Modulate timer frequency based on battery state
1191 * add libsystemd-password or so to query passwords during boot using the password agent logic
1193 * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed
1195 * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel
1197 * make repeated alt-ctrl-del presses printing a dump
1199 * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ...
1201 * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab is not
1204 - emulate /dev/kmsg using CUSE and turn off the syslog syscall
1205 with seccomp. That should provide us with a useful log buffer that
1206 systemd can log to during early boot, and disconnect container logs
1207 from the kernel's logs.
1208 - as soon as networkd has a bus interface, hook up --network-interface=,
1209 --network-bridge= with networkd, to trigger netdev creation should an
1210 interface be missing
1211 - a nice way to boot up without machine id set, so that it is set at boot
1212 automatically for supporting --ephemeral. Maybe hash the host machine id
1213 together with the machine name to generate the machine id for the container
1214 - fix logic always print a final newline on output.
1215 https://github.com/systemd/systemd/pull/272#issuecomment-113153176
1216 - should optionally support receiving WATCHDOG=1 messages from its payload
1218 - optionally automatically add FORWARD rules to iptables whenever nspawn is
1219 running, remove them when shut down.
1222 - add an API so that libvirt-lxc can inform us about network interfaces being
1223 removed or added to an existing machine
1224 - "machinectl migrate" or similar to copy a container from or to a
1225 difference host, via ssh
1226 - introduce systemd-nspawn-ephemeral@.service, and hook it into
1227 "machinectl start" with a new --ephemeral switch
1228 - "machinectl status" should also show internal logs of the container in
1230 - "machinectl history"
1232 - "machinectl commit" that takes a writable snapshot of a tree, invokes a
1233 shell in it, and marks it read-only after use
1236 - cryptsetup-generator: allow specification of passwords in crypttab itself
1237 - support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator
1239 * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it
1240 https://bugs.freedesktop.org/show_bug.cgi?id=54712
1242 * add a dependency on standard-conf.xml and other included files to man pages
1244 * MountFlags=shared acts as MountFlags=slave right now.
1246 * properly handle loop back mounts via fstab, especially regards to fsck/passno
1248 * initialize the hostname from the fs label of /, if /etc/hostname does not exist?
1253 - add trigger --subsystem-match=usb/usb_device device
1254 - reimport udev db after MOVE events for devices without dev_t
1256 * There's currently no way to cancel fsck (used to be possible via C-c or c on the console)
1258 * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/
1261 - save coredump in Windows/Mozilla minidump format
1262 - when truncating coredumps, also log the full size that the process had, and make a metadata field so we can report truncated coredumps
1264 * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting)
1266 * be able to specify a forced restart of service A where service B depends on, in case B
1267 needs to be auto-respawned?
1270 - apply "x" on "D" too (see patch from William Douglas)
1271 - instead of ignoring unknown fields, reject them.
1272 - creating new directories/subvolumes/fifos/device nodes
1273 should not follow symlinks. None of the other adjustment or creation
1274 calls follow symlinks.
1276 * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early
1278 * verify that the AF_UNIX sockets of a service in the fs still exist
1279 when we start a service in order to avoid confusion when a user
1280 assumes starting a service is enough to make it accessible
1282 * Make it possible to set the keymap independently from the font on
1283 the kernel cmdline. Right now setting one resets also the other.
1285 * and a dbus call to generate target from current state
1287 * write blog stories about:
1288 - hwdb: what belongs into it, lsusb
1289 - enabling dbus services
1290 - how to make changes to sysctl and sysfs attributes
1292 - how to pass throw-away units to systemd, or dynamically change properties of existing units
1293 - testing with Harald's awesome test kit
1295 - how to develop against journal browsing APIs
1296 - the journal HTTP iface
1297 - non-cgroup resource management
1298 - dynamic resource management with cgroups
1299 - refreshed, longer missions statement
1300 - calendar time events
1301 - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell
1302 - how to create your own target
1303 - instantiated apache, dovecot and so on
1304 - hooking a script into various stages of shutdown/rearly booot
1306 * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support.
1308 * dot output for --test showing the 'initial transaction'
1311 - When logging about multiple units (stopping BoundTo units, conflicts, etc.),
1312 log both units as UNIT=, so that journalctl -u triggers on both.
1313 - generate better errors when people try to set transient properties
1314 that are not supported...
1315 http://lists.freedesktop.org/archives/systemd-devel/2015-February/028076.html
1316 - maybe introduce WantsMountsFor=? Usecase:
1317 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027729.html
1318 - recreate systemd's D-Bus private socket file on SIGUSR2
1319 - move PAM code into its own binary
1320 - when we automatically restart a service, ensure we restart its rdeps, too.
1321 - hide PAM options in fragment parser when compile time disabled
1322 - Support --test based on current system state
1323 - If we show an error about a unit (such as not showing up) and it has no Description string, then show a description string generated form the reverse of unit_name_mangle().
1324 - after deserializing sockets in socket.c we should reapply sockopts and things
1325 - drop PID 1 reloading, only do reexecing (difficult: Reload()
1326 currently is properly synchronous, Reexec() is weird, because we
1327 cannot delay the response properly until we are back, so instead of
1328 being properly synchronous we just keep open the fd and close it
1329 when done. That means clients do not get a successful method reply,
1330 but much rather a disconnect on success.
1331 - when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr
1332 - when a bus name of a service disappears from the bus make sure to queue further activation requests
1335 - allow port=0 in .socket units
1336 - maybe introduce ExecRestartPre=
1337 - add ReloadSignal= for configuring a reload signal to use
1338 - implement Register= switch in .socket units to enable registration
1339 in Avahi, RPC and other socket registration services.
1340 - allow Type=simple with PIDFile=
1341 https://bugzilla.redhat.com/show_bug.cgi?id=723942
1342 - allow writing multiple conditions in unit files on one line
1343 - introduce Type=pid-file
1344 - introduce mix of BindTo and Requisite
1345 - add a concept of RemainAfterExit= to scope units
1346 - Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely
1347 - add verification of [Install] section to systemd-analyze verify
1350 - Make sure ID_PATH is always exported and complete for
1351 network devices where possible, so we can safely rely
1355 - add support for more attribute types
1356 - inbuilt piping support (essentially degenerate async)? see loopback-setup.c and other places
1359 - add more keys to [Route] and [Address] sections
1360 - add support for more DHCPv4 options (and, longer term, other kinds of dynamic config)
1361 - add reduced [Link] support to .network files
1362 - properly handle routerless dhcp leases
1363 - work with non-Ethernet devices
1364 - dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from?
1365 - the DHCP lease data (such as NTP/DNS) is still made available when
1366 a carrier is lost on a link. It should be removed instantly.
1367 - expose in the API the following bits:
1368 - option 15, domain name
1369 - option 12, hostname and/or option 81, fqdn
1370 - option 123, 144, geolocation
1371 - option 252, configure http proxy (PAC/wpad)
1372 - provide a way to define a per-network interface default metric value
1373 for all routes to it. possibly a second default for DHCP routes.
1374 - allow Name= to be specified repeatedly in the [Match] section. Maybe also
1375 support Name=foo*|bar*|baz ?
1376 - whenever uplink info changes, make DHCP server send out FORCERENEW
1378 * Figure out how to do unittests of networkd's state serialization
1381 - figure out how much we can increase Maximum Message Size
1384 - add functions to set previously stored IPv6 addresses on startup and get
1385 them at shutdown; store them in client->ia_na
1386 - write more test cases
1387 - implement reconfigure support, see 5.3., 15.11. and 22.20.
1388 - implement support for temporary adressess (IA_TA)
1389 - implement dhcpv6 authentication
1390 - investigate the usefulness of Confirm messages; i.e. are there any
1391 situations where the link changes without any loss in carrier detection
1393 - some servers don't do rapid commit without a filled in IA_NA, verify
1400 - natively watch for dbus-*.service symlinks (PENDING)
1401 - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service
1403 * make cryptsetup lower --iter-time
1405 * kernel: add device_type = "fb", "fbcon" to class "graphics"
1407 * /usr/bin/service should actually show the new command line
1409 * fedora: suggest auto-restart on failure, but not on success and not on coredump. also, ask people to think about changing the start limit logic. Also point people to RestartPreventExitStatus=, SuccessExitStatus=
1411 * neither pkexec nor sudo initialize environ[] from the PAM environment?
1413 * fedora: update policy to declare access mode and ownership of unit files to root:root 0644, and add an rpmlint check for it
1415 * register catalog database signature as file magic
1417 * zsh shell completion:
1418 - <command> <verb> -<TAB> should complete options, but currently does not
1419 - systemctl add-wants,add-requires
1421 * systemctl status should know about 'systemd-analyze calendar ... --iterations='
1422 * If timer has just OnInactiveSec=..., it should fire after a specified time
1423 after being started.
1427 * look for close() vs. close_nointr() vs. close_nointr_nofail()
1429 * check for strerror(r) instead of strerror(-r)
1433 * set_put(), hashmap_put() return values check. i.e. == 0 does not free()!
1435 * use secure_getenv() instead of getenv() where appropriate
1437 * link up selected blog stories from man pages and unit files Documentation= fields
1438 String is not UTF-8 clean, ignoring assignment