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.
9 * Fedora: add an rpmlint check that verifies that all unit files in the RPM are listed in %systemd_post macros.
13 * Rearrange tests so that the various test-xyz.c match a specific src/basic/xyz.c again
15 * rework mount.c and swap.c to follow proper state enumeration/deserialization
16 semantics, like we do for device.c now
20 * add --copy-from and --copy-to command to systemd-dissect which copies stuff
21 in and out of a disk image
23 * Support ProtectProc= or so, using: https://patchwork.kernel.org/cover/11310197/
25 * if /usr/bin/swapoff fails due to OOM, log a friendly explanatory message about it
27 * add loud warning to the logs (with catalog entry) if systemd-udev-settle is
28 pulled into the boot process
30 * build short web pages out of each catalog entry, build them along with man
31 pages, and include hyperlinks to them in the journal output
33 * machined: add API to acquire UID range. add API to mount/dissect loopback
34 file. Both protected by PK. Then make nspawn use these APIs to run
35 unprivileged containers. i.e. push the truly privileged bits into machined,
36 so that the client side can remain entirely unprivileged, with SUID or
39 * journald: do journal file writing out-of-process, with one writer process per
40 client UID, so that synthetic hash table collisions can slow down a specific
41 user's journal stream down but not the others.
43 * add "throttling" to sd-event event sources: optionally, when we wake up too
44 often for one, let's turn it off entirely for a while. Use that for the
45 /proc/self/mountinfo logic.
47 * move our systemd-user PAM snippet to /usr/, which PAM appears to support
50 * nspawn: support time namespaces
52 * add ConditionSecurity=tpm2
54 * Remove any support for booting without /usr pre-mounted in the initrd entirely.
55 Update INITRD_INTERFACE.md accordingly.
57 * pid1: Move to tracking of main pid/control pid of units per pidfd
59 * pid1: support new clone3() fork-into-cgroup feature
61 * pid1: also remove PID files of a service when the service starts, not just
64 * make us use dynamically fewer deps for containers in general purpose distros:
65 o turn into dlopen() deps:
66 - pcre2 (always) — irrelevant on Fedora, since dep by
67 libselinux, but should benefit Debian
68 - libpwquality (always) - only relevant for homed, and maybe soon
71 - p11-kit-trust (always)
72 - kmod-libs (only when called from PID 1)
73 - cryptsetup-libs (only in RootImage= handling in PID 1, but not in systemd-cryptsetup)
75 - libpam (only when called from PID 1)
76 - bzip2, xz, lz4 (always — gzip and zstd should probably stay static deps the way they are,
77 since they are so basic and our defaults)
78 o move into separate libsystemd-shared-iptables.so .so
79 - iptables-libs (only used by nspawn + networkd)
81 * seccomp: when SystemCallArchitectures=native is set then don't install any
82 other seccomp filters for any of the other archs, in order to reduce the
83 number of seccomp filters we install needlessly.
85 * seccomp: maybe use seccomp_merge() to merge our filters per-arch if we can.
86 Apparently kernel performance is much better with fewer larger seccomp
87 filters than with more smaller seccomp filters.
89 * systemd-path: add ESP and XBOOTLDR path. Add "private" runtime/state/cache dir enum,
90 mapping to $RUNTIME_DIRECTORY, $STATE_DIRECTORY and such
92 * make "systemd-dissect" an official supported tool, i.e. move to /usr/bin/ and
93 provide man page. Given that we now have a tool that can generate images like
94 this, it's useful to have one that can dump contents of them, too.
96 * All tools that support --root= should also learn --image= so that they can
97 operate on disk images directly. Specifically: bootctl, firstboot, tmpfiles,
98 sysusers, systemctl, repart, journalctl, coredumpctl.
100 * seccomp: by default mask x32 ABI system wide on x86-64. it's on its way out
102 * seccomp: don't install filters for ABIs that are masked anyway for the
105 * seccomp: maybe merge all filters we install into one with that libseccomp API that allows merging.
107 * per-service credential system. Specifically: add LoadCredential= (for loading
108 cred from file), AcquireCredential= (for asking user for cred, via
109 ask-password), PassCredential= (for passing on credential systemd itself
110 got). Then, place credentials in a per-service, immutable ramfs instance (so
111 that it cannot be swapped out), destroy after use. Also pass via keyring
112 (with graceful fallback to cover for containers). Define CredentialPath= for
113 defining subdir of /run/credentials/ where to place it. Set $CREDENTIAL_PATH
114 env var for services to the result. Also pass via fd passing (optionally).
116 * homed: add native recovery key support. use 48 lowercase modhex characters
117 (192bit), show qr code of it, include pattern expression in user record.
119 * homed: introduce "degraded" state for home directories that weren't cleanly
120 unmounted (use xattr we add and remove on the loop back file)
122 * homed: during login resize fs automatically towards size goal. Specifically,
123 resize to diskSize if possible, but leave a certain amount (configured by a
124 new value diskLeaveFreeSize) of space free on the backing fs.
126 * homed: permit multiple user record signing keys to be used locally, and pick
127 the right one for signing records automatically depending on a pre-existing
130 * homed: add a way to "adopt" a home directory, i.e. strip foreign signatures
131 and insert a local signature instead.
133 * homed: as an extension to the directory+subvolume backend: if located on
134 especially marked fs, then sync down password into LUKS header of that fs,
135 and always verify passwords against it too. Bootstrapping is a problem
136 though: if no one is logged in (or no other user even exists yet), how do you
137 unlock the volume in order to create the first user and add the first pw.
139 * homed: support new FS_IOC_ADD_ENCRYPTION_KEY ioctl for setting up fscrypt
141 * busctl: maybe expose a verb "ping" for pinging a dbus service to see if it
144 * homed: maybe pre-create ~/.cache as subvol so that it can have separate quota
147 * when systemd-nspawn and suchlike dissect an OS image, and there are multiple
148 root partitions, do an strverscmp() on the partition label and boot
149 first. That is inspired how sd-boot figures out which kernel to boot, and
150 thus allows defining OS images which can be A/B updated and we default to the
151 newest version automatically, both in nspawn and in sd-boot
153 * cryptsetup/homed: also support FIDO2 HMAC password logic for unlocking
154 devices. (see: https://github.com/mjec/fido2-hmac-secret)
156 * systemd-gpt-auto should probably set x-systemd.growfs on the mounts it
159 * homed/userdb: distinguish passwords and recovery keys in the records, since
160 we probably want to use different PBKDF algorithms/settings for them:
161 passwords have low entropy but recovery keys should have good entropy key
162 hence we can make them quicker to work.
165 - teach it to prepare an ESP wholesale, i.e. with mkfs.vfat invocation
166 - teach it to copy in unified kernel images and maybe type #1 boot loader spec entries from host
167 - make it operate on loopback files, dissecting enough to find ESP to operate on
169 * Maybe add a separate GPT partition type to the discoverable partition spec
170 for "hibernate" partitions, that are exactly like swap partitions but only
171 activated right before hibernation and thus never used for regular swapping.
173 * by default, in systemd --user service bump the OOMAdjust to 100, as privs
174 allow so that systemd survives
176 * cryptsetup: allow encoding key directly in /etc/crypttab, maybe with a
177 "base64:" prefix. Useful in particular for pkcs11 mode.
179 * cryptsetup: reimplement the mkswap/mke2fs in cryptsetup-generator to use
180 systemd-makefs.service instead.
182 * socket units: allow creating a udev monitor socket with ListenDevices= or so,
183 with matches, then activate app through that passing socket over
185 * unify on openssl (as soon as OpenSSL 3.0 is out, and the Debian license
187 - port sd_id128_get_machine_app_specific() over from khash
188 - port resolved over from libgcrypt (DNSSEC code)
189 - port journald + fsprg over from libgcrypt
190 - port importd over from libgcrypt
191 - when that's done: kill khash.c
192 - when that's done: kill gnutls support in resolved
194 * kill zenata, all hail weblate?
196 * when we resize disks (homed?) always round up to 4K sectors, not 512K
198 * add growvol and makevol options for /etc/crypttab, similar to
199 x-systemd.growfs and x-systemd-makefs.
201 * hook up the TPM to /etc/crypttab, with a new option that is similar to the
202 new PKCS#11 option in crypttab, and allows unlocking a LUKS volume via a key
203 unsealed from the TPM. Optionally, if TPM is not available fall back to
204 TPM-less mode, and set up linear DM mapping instead (inspired by kpartx), so
205 that the device paths stay the same, regardless if crypto is used or not.
207 * systemd-repart: by default generate minimized partition tables (i.e. tables
208 that only cover the space actually used, excluding any free space at the
209 end), in order to maximize dd'ability. Requires libfdisk work, see
210 https://github.com/karelzak/util-linux/issues/907
212 * systemd-repart: MBR partition table support. Care needs to be taken regarding
213 Type=, so that partition definitions can sanely apply to both the GPT and the
214 MBR case. Idea: accept syntax "Type=gpt:home mbr:0x83" for setting the types
215 for the two partition types explicitly. And provide an internal mapping so
216 that "Type=linux-generic" maps to the right types for both partition tables
219 * systemd-repart: allow sizing partitions as factor of available RAM, so that
220 we can reasonably size swap partitions for hibernation.
222 * systemd-repart: allow running mkfs before making partitions pop up +
223 encryption via LUKS to allow booting into an empty root with only /usr mounted in
225 * systemd-repart: allow managing the gpt read-only partition flag + auto-mount flag
227 * systemd-repart: allow boolean option that ensures that if existing partition
228 doesn't exist within the configured size bounds the whole command fails. This
229 is useful to implement ESP vs. XBOOTLDR schemes in installers: have one set
230 of repart files for the case where ESP is large enough and one where it isn't
231 and XBOOTLDR is added in instead. Then apply the former first, and if it
232 fails to apply use the latter.
234 * systemd-repart: add per-partition option to never reuse existing partition
235 and always create anew even if matching partition already exists.
237 * systemd-repart: add per-partition option to fail if partition already exist,
238 i.e. is not added new. Similar, add option to fail if partition does not exist yet.
240 * systemd-repart: add --size=auto for generating/resizing images of minimal
241 size, i.e. where the image file is sized exactly as large as necessary taking
242 SizeMin= into account, but not a single byte larger.
244 * systemd-repart: allow disabling growing of specific partitions, or making
245 them (think ESP: we don't ever want to grow it, since we cannot resize vfat)
247 * systemd-repart: make it a static checker during early boot for existence and
248 absence of other partitions for trusted boot environments
250 * userdb: allow username prefix searches in varlink API, allow realname and
251 realname substr searches in varlink API
253 * userdb: allow uid/gid range checks
255 * userdb: allow existence checks
257 * pid1: activation by journal search expression
259 * when switching root from initrd to host, set the machine_id env var so that
260 if the host has no machine ID set yet we continue to use the random one the
263 * sd-event: add native support for P_ALL waitid() watching, then move PID 1 to
264 it fo reaping assigned but unknown children. This needs to some special care
265 to operate somewhat sensibly in light of priorities: P_ALL will return
266 arbitrary processes, regardless of the priority we want to watch them with,
267 hence on each event loop iteration check all processes which we shall watch
268 with higher prio explicitly, and then watch the entire rest with P_ALL.
270 * tweak sd-event's child watching: keep a prioq of children to watch and use
271 waitid() only on the children with the highest priority until one is waitable
272 and ignore all lower-prio ones from that point on
274 * maybe introduce xattrs that can be set on the root dir of the root fs
275 partition that declare the volatility mode to use the image in. Previously I
276 thought marking this via GPT partition flags but that's not ideal since
277 that's outside of the LUKS encryption/verity verification, and we probably
278 shouldn't operate in a volatile mode unless we got told so from a trusted
281 * figure out automatic partition discovery when combining writable root dir
284 * coredump: maybe when coredumping read a new xattr from /proc/$PID/exe that
285 may be used to mark a whole binary as non-coredumpable. Would fix:
286 https://bugs.freedesktop.org/show_bug.cgi?id=69447
288 * teach parse_timestamp() timezones like the calendar spec already knows it
290 * beef up hibernation to optionally do swapon/swapoff immediately before/after
293 * beef up s2h to implement a battery watch loop: instead of entering
294 hibernation unconditionally after coming back from resume make a decision
295 based on the battery load level: if battery level is above a specific
296 threshold, go to suspend again, only hibernate if below it. This means we'd
297 stick to suspend usually, but fall back to hibernation only when battery runs
298 empty (well, subject to our sampling interval). Related to this, check if we
299 can make ACPI _BTP (i.e. /sys/class/power_supply/*/alarm) work for us too,
300 i.e. see if it can wake up machines from suspend, so that we could resume
301 automatically when the system is low on power and move automatically to
302 hibernation mode. (see
303 https://uefi.org/sites/default/files/resources/ACPI%206_2_A_Sept29.pdf
305 https://docs.microsoft.com/en-us/windows-hardware/design/device-experiences/modern-standby-wake-sources
308 * add an explicit "vertical" mode to format-table, so that "systemctl
309 status"-like outputs (i.e. with a series of field names left and values
310 right) become genuine first class citizens, and we gain automatic, sane JSON
313 * systemd-firstboot: teach it dissector magic, so that you can point it to some
314 disk image and it will just set everything in it all behind the scenes.
316 * We should probably replace /var/log/README, /etc/rc.d/README with symlinks
317 that are linked to these places instead of copied. After all they are
318 constant vendor data.
320 * maybe add kernel cmdline params: to force random seed crediting
322 * nspawn: on cgroupsv1 issue cgroup empty handler process based on host events,
323 so that we make cgroup agent logic safe
325 * nspawn/machined: add API to invoke binary in container, then use that as
326 fallback in "machinectl shell"
328 * logind: rework pam_logind to also do a bus call in case of invocation from
329 user@.service, which returns the XDG_RUNTIME_DIR value, and make this
330 behaviour selectable via pam module option.
333 - when user tries to log into record signed by unrecognized key, automatically add key to our chain after polkit auth
334 - hook up machined/nspawn users with a varlink user query interface
335 - rollback when resize fails mid-operation
336 - GNOME's side for forget key on suspend (requires rework so that lock screen runs outside of uid)
338 - shrink fs on logout?
339 - update LUKS password on login if we find there's a password that unlocks the JSON record but not the LUKS device.
340 - create on activate?
341 - properties: icon url?, preferred session type?, administrator bool (which translates to 'wheel' membership)?, address?, telephone?, vcard?, samba stuff?, parental controls?
342 - communicate clearly when usb stick is safe to remove. probably involves
343 beefing up logind to make pam session close hook synchronous and wait until
344 systemd --user is shut down.
345 - logind: maybe keep a "busy fd" as long as there's a non-released session around or the user@.service
346 - maybe make automatic, read-only, time-based reflink-copies of LUKS disk images (think: time machine)
347 - distinguish destroy / remove (i.e. currently we can unregister a user, unregister+remove their home directory, but not just remove their home directory)
348 - in systemd's PAMName= logic: query passwords with ssh-askpassword, so that we can make "loginctl set-linger" mode work
349 - fingerprint authentication, pattern authentication, …
350 - make sure "classic" user records can also be managed by homed
351 - description field for groups
352 - make size of $XDG_RUNTIME_DIR configurable in user record
353 - reuse pwquality magic in firstboot
354 - query password from kernel keyring first
355 - update even if record is "absent"
356 - add a "access mode" + "fstype" field to the "status" section of json identity records reflecting the actually used access mode and fstype, even on non-luks backends
357 - move acct mgmt stuff from pam_systemd_home to pam_systemd?
358 - when "homectl --pkcs11-token-uri=" is used, synthesize ssh-authorized-keys records for all keys we have private keys on the stick for
359 - make slice for users configurable (requires logind rework)
360 - logind: populate auto-login list bus property from PKCS#11 token
361 - when determining state of a LUKS home directory, check DM suspended sysfs file
363 * introduce a new per-process uuid, similar to the boot id, the machine id, the
364 invocation id, that is derived from process creds, specifically a hashed
365 combination of AT_RANDOM + getpid() + the starttime from
366 /proc/self/status. Then add these ids implicitly when logging. Deriving this
367 uuid from these three things has the benefit that it can be derived easily
368 from /proc/$PID/ in a stable, and unique way that changes on both fork() and
371 * let's not GC a unit while its ratelimits are still pending
373 * when killing due to service watchdog timeout maybe detect whether target
374 process is under ptracing and then log loudly and continue instead.
376 * introduce a new group to own TPM devices
378 * make rfkill uaccess controllable by default, i.e. steal rule from
379 gnome-bluetooth and friends
381 * tweak journald context caching. In addition to caching per-process attributes
382 keyed by PID, cache per-cgroup attributes (i.e. the various xattrs we read)
383 keyed by cgroup path, and guarded by ctime changes. This should provide us
384 with a nice speed-up on services that have many processes running in the same
387 * make MAINPID= message reception checks even stricter: if service uses User=,
388 then check sending UID and ignore message if it doesn't match the user or
391 * maybe trigger a uevent "change" on a device if "systemctl reload xyz.device"
394 * when importing an fs tree with machined, optionally apply userns-rec-chown
396 * when importing an fs tree with machined, complain if image is not an OS
398 * Maybe introduce a helper safe_exec() or so, which is to execve() which
399 safe_fork() is to fork(). And then make revert the RLIMIT_NOFILE soft limit
400 to 1K implicitly, unless explicitly opted-out.
402 * rework seccomp/nnp logic that even if User= is used in combination with
403 a seccomp option we don't have to set NNP. For that, change uid first whil
404 keeping CAP_SYS_ADMIN, then apply seccomp, the drop cap.
406 * add a concept for automatically loading per-unit secrets off disk and
407 inserting them into the kernel keyring. Maybe SecretsDirectory= similar to
408 ConfigurationDirectory=.
410 * when no locale is configured, default to UEFI's PlatformLang variable
412 * bootctl,sd-boot: actually honour the "architecture" key
414 * add a new syscall group "@esoteric" for more esoteric stuff such as bpf() and
415 usefaultd() and make systemd-analyze check for it.
417 * paranoia: whenever we process passwords, call mlock() on the memory
418 first. i.e. look for all places we use free_and_erasep() and
419 augment them with mlock(). Also use MADV_DONTDUMP.
421 * Move RestrictAddressFamily= to the new cgroup create socket
423 * support the bind/connect/sendmsg cgroup stuff for sandboxing, and possibly
426 * maybe implicitly attach monotonic+realtime timestamps to outgoing messages in
427 log.c and sd-journal-send
429 * optionally: turn on cgroup delegation for per-session scope units
431 * introduce per-unit (i.e. per-slice, per-service) journal log size limits.
433 * sd-boot: automatically load EFI modules from some drop-in dir, so that people
434 can add in file system drivers and such
436 * sd-boot: optionally, show boot menu when previous default boot item has
437 non-zero "tries done" count
439 * introduce an option (or replacement) for "systemctl show" that outputs all
440 properties as JSON, similar to busctl's new JSON output. In contrast to that
441 it should skip the variant type string though.
443 * augment CODE_FILE=, CODE_LINE= with something like CODE_BASE= or so which
444 contains some identifier for the project, which allows us to include
445 clickable links to source files generating these log messages. The identifier
446 could be some abberviated URL prefix or so (taking inspiration from Go
447 imports). For example, for systemd we could use
448 CODE_BASE=github.com/systemd/systemd/blob/98b0b1123cc or so which is
449 sufficient to build a link by prefixing "http://" and suffixing the
452 * Augment MESSAGE_ID with MESSAGE_BASE, in a similar fashion so that we can
453 make clickable links from log messages carrying a MESSAGE_ID, that lead to
454 some explanatory text online.
456 * maybe extend .path units to expose fanotify() per-mount change events
458 * Add a "systemctl list-units --by-slice" mode or so, which rearranges the
459 output of "systemctl list-units" slightly by showing the tree structure of
460 the slices, and the units attached to them.
462 * the a-posteriori stopping of units bound to units that disappeared logic
463 should be reworked: there should be a queue of units, and we should only
464 enqueue stop jobs from a defer event that processes queue instead of
465 right-away when we find a unit that is bound to one that doesn't exist
466 anymore. (similar to how the stop-unneeded queue has been reworked the same
469 * nspawn: make nspawn suitable for shell pipelines: instead of triggering a
470 hangup when input is finished, send ^D, which synthesizes an EOF. Then wait
471 for hangup or ^D before passing on the EOF.
473 * When reloading configuration PID 1 should reset all its properties to the
474 original defaults before calling parse_config()
476 * nspawn: greater control over selinux label?
478 * hibernate/s2h: make this robust and safe to enable in Fedora by default.
481 1. add resume_offset support to the resume code (i.e. support swap files
483 2. check if swap is on weird storage and refuse if so
484 3. add auto-detection of hibernation images
486 * cgroups: use inotify to get notified when somebody else modifies cgroups
487 owned by us, then log a friendly warning.
489 * beef up log.c with support for stripping ANSI sequences from strings, so that
490 it is OK to include them in log strings. This would be particularly useful so
491 that our log messages could contain clickable links for example for unit
492 files and suchlike we operate on.
494 * add support for "portablectl attach http://foobar.com/waaa.raw (i.e. importd integration)
496 * add attach --enable and attach --now (for attach+enable+start)
498 * sync dynamic uids/gids between host+portable srvice (i.e. if DynamicUser=1 is set for a service, make sure that the
499 selected user is resolvable in the service even if it ships its own /etc/passwd)
501 * Fix DECIMAL_STR_MAX or DECIMAL_STR_WIDTH. One includes a trailing NUL, the
502 other doesn't. What a disaster. Probably to exclude it. Also
503 DECIMAL_STR_WIDTH should probably add an extra "-" into account for negative
506 * Check that users of inotify's IN_DELETE_SELF flag are using it properly, as
507 usually IN_ATTRIB is the right way to watch deleted files, as the former only
508 fires when a file is actually removed from disk, i.e. the link count drops to
509 zero and is not open anymore, while the latter happens when a file is
510 unlinked from any dir.
512 * port systemctl, busctl, … over to format-table.[ch]'s table formatters
514 * pid1: lock image configured with RootDirectory=/RootImage= using the usual nspawn semantics while the unit is up
516 * add --vacuum-xyz options to coredumpctl, matching those journalctl already has.
518 * introduce Ephemeral= unit file switch, that creates an ephemeral copy of all
519 files and directories that are left writable for a unit, and which are
520 removed after the unit goes down again. A bit like --ephemeral for
521 systemd-nspawn but for system services. If used together with RootImage= this
522 should reflink the image file itself.
524 Related: add Ephemeral=<path1> <path2> … which would allow marking
525 specific paths only like this.
527 * add CopyFile= or so as unit file setting that may be used to copy files or
528 directory trees from the host to the services RootImage= and RootDirectory=
529 environment. Which we can use for /etc/machine-id and in particular
530 /etc/resolv.conf. Should be smart and do something useful on read-only
531 images, for example fallback to read-only bind mounting the file instead.
533 * show invocation ID in systemd-run output
535 * bypass SIGTERM state in unit files if KillSignal is SIGKILL
537 * add proper dbus APIs for the various sd_notify() commands, such as MAINPID=1
538 and so on, which would mean we could report errors and such.
540 * teach tmpfiles.d q/Q logic something sensible in the context of XFS/ext4
543 * introduce DefaultSlice= or so in system.conf that allows changing where we
544 place our units by default, i.e. change system.slice to something
545 else. Similar, ManagerSlice= should exist so that PID1's own scope unit could
546 be moved somewhere else too. Finally machined and logind should get similar
547 options so that it is possible to move user session scopes and machines to a
548 different slice too by default. Usecase: people who want to put resources on
549 the entire system, with the exception of one specific service. See:
550 https://lists.freedesktop.org/archives/systemd-devel/2018-February/040369.html
552 * maybe rework get_user_creds() to query the user database if $SHELL is used
553 for root, but only then.
555 * be stricter with fds we receive for the fdstore: close them asynchronously
557 * calenderspec: add support for week numbers and day numbers within a
558 year. This would allow us to define "bi-weekly" triggers safely.
560 * sd-bus: add vtable flag, that may be used to request client creds implicitly
561 and asynchronously before dispatching the operation
563 * make use of ethtool veth peer info in machined, for automatically finding out
564 host-side interface pointing to the container.
566 * add some special mode to LogsDirectory=/StateDirectory=… that allows
567 declaring these directories without necessarily pulling in deps for them, or
568 creating them when starting up. That way, we could declare that
569 systemd-journald writes to /var/log/journal, which could be useful when we
570 doing disk usage calculations and so on.
572 * taint systemd if there are fewer than 65536 users assigned (userns) to the system.
574 * deprecate RootDirectoryStartOnly= in favour of a new ExecStart= prefix char
576 * add a new RuntimeDirectoryPreserve= mode that defines a similar lifecycle for
577 the runtime dir as we maintain for the fdstore: i.e. keep it around as long
578 as the unit is running or has a job queued.
580 * support projid-based quota in machinectl for containers
582 * add a way to lock down cgroup migration: a boolean, which when set for a unit
583 makes sure the processes in it can never migrate out of it
585 * blog about fd store and restartable services
587 * document Environment=SYSTEMD_LOG_LEVEL=debug drop-in in debugging document
589 * rework ExecOutput and ExecInput enums so that EXEC_OUTPUT_NULL loses its
590 magic meaning and is no longer upgraded to something else if set explicitly.
592 * in the long run: permit a system with /etc/machine-id linked to /dev/null, to
593 make it lose its identity, i.e. be anonymous. For this we'd have to patch
594 through the whole tree to make all code deal with the case where no machine
597 * optionally, collect cgroup resource data, and store it in per-unit RRD files,
598 suitable for processing with rrdtool. Add bus API to access this data, and
599 possibly implement a CPULoad property based on it.
601 * beef up pam_systemd to take unit file settings such as cgroups properties as
604 * maybe hook of xfs/ext4 quotactl() with services? i.e. automatically manage
605 the quota of a the user indicated in User= via unit file settings, like the
606 other resource management concepts. Would mix nicely with DynamicUser=1. Or
607 alternatively, do this with projids, so that we can also cover services
608 running as root. Quota should probably cover all the special dirs such as
609 StateDirectory=, LogsDirectory=, CacheDirectory=, as well as RootDirectory= if it
610 is set, plus the whole disk space any image configured with RootImage=.
612 * In DynamicUser= mode: before selecting a UID, use disk quota APIs on relevant
613 disks to see if the UID is already in use.
615 * add "systemctl wait" or so, which does what "systemd-run --wait" does, but
616 for all units. It should be both a way to pin units into memory as well as a
617 wait to retrieve their exit data.
619 * expose IO accounting data on the bus, show it in systemd-run --wait and log
620 about it in the resource log message
622 * show whether a service has out-of-date configuration in "systemctl status" by
623 using mtime data of ConfigurationDirectory=.
625 * Add AddUser= setting to unit files, similar to DynamicUser=1 which however
626 creates a static, persistent user rather than a dynamic, transient user. We
627 can leverage code from sysusers.d for this.
629 * add some optional flag to ReadWritePaths= and friends, that has the effect
630 that we create the dir in question when the service is started. Example:
632 ReadWritePaths=:/var/lib/foobar
634 * maybe add call sd_journal_set_block_timeout() or so to set SO_SNDTIMEO for
635 the sd-journal logging socket, and, if the timeout is set to 0, sets
636 O_NONBLOCK on it. That way people can control if and when to block for
639 * hostnamed: populate form factor data from a new hwdb database, so that old
640 yogas can be recognized as "convertible" too, even if they predate the DMI
641 "convertible" form factor
643 * Maybe add PrivatePIDs= as new unit setting, and do minimal PID namespacing
644 after all. Be strict however, only support the equivalent of nspawn's
645 --as-pid2 switch, and sanely proxy sd_notify() messages dropping stuff such
648 * Add ExecMonitor= setting. May be used multiple times. Forks off a process in
649 the service cgroup, which is supposed to monitor the service, and when it
650 exits the service is considered failed by its monitor.
652 * track the per-service PAM process properly (i.e. as an additional control
653 process), so that it may be queried on the bus and everything.
655 * add a new "debug" job mode, that is propagated to unit_start() and for
656 services results in two things: we raise SIGSTOP right before invoking
657 execve() and turn off watchdog support. Then, use that to implement
658 "systemd-gdb" for attaching to the start-up of any system service in its
661 * gpt-auto logic: related to the above, maybe support a "secondary" root
662 partition, that is mounted to / and is writable, and where the actual root's
663 /usr is mounted into.
665 * 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
667 * drop nss-myhostname in favour of nss-resolve?
669 * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and
670 then use that for the setting used in user@.service. It should be understood
671 relative to the configured default value.
673 * in networkd, when matching device types, fix up DEVTYPE rubbish the kernel passes to us
675 * enable LockMLOCK to take a percentage value relative to physical memory
677 * Permit masking specific netlink APIs with RestrictAddressFamily=
679 * nspawn: support that /proc, /sys/, /dev are pre-mounted
681 * define gpt header bits to select volatility mode
683 * ProtectClock= (drops CAP_SYS_TIMES, adds seecomp filters for settimeofday, adjtimex), sets DeviceAllow o /dev/rtc
685 * ProtectTracing= (drops CAP_SYS_PTRACE, blocks ptrace syscall, makes /sys/kernel/tracing go away)
687 * ProtectMount= (drop mount/umount/pivot_root from seccomp, disallow fuse via DeviceAllow, imply Mountflags=slave)
689 * ProtectKeyRing= to take keyring calls away
691 * RemoveKeyRing= to remove all keyring entries of the specified user
693 * ProtectReboot= that masks reboot() and kexec_load() syscalls, prohibits kill
694 on PID 1 with the relevant signals, and makes relevant files in /sys and
695 /proc (such as the sysrq stuff) unavailable
697 * make sure the ratelimit object can deal with USEC_INFINITY as way to turn off things
699 * journalctl: make sure -f ends when the container indicated by -M terminates
701 * mount: automatically search for "main" partition of an image has multiple
704 * in nss-systemd, if we run inside of RootDirectory= with PrivateUsers= set,
705 find a way to map the User=/Group= of the service to the right name. This way
706 a user/group for a service only has to exist on the host for the right
709 * add bus API for creating unit files in /etc, reusing the code for transient units
711 * add bus API to remove unit files from /etc
713 * add bus API to retrieve current unit file contents (i.e. implement "systemctl cat" on the bus only)
715 * rework fopen_temporary() to make use of open_tmpfile_linkable() (problem: the
716 kernel doesn't support linkat() that replaces existing files, currently)
718 * transient units: don't bother with actually setting unit properties, we
719 reload the unit file anyway
721 * journald: sigbus API via a signal-handler safe function that people may call
722 from the SIGBUS handler
724 * optionally, also require WATCHDOG=1 notifications during service start-up and shutdown
726 * delay activation of logind until somebody logs in, or when /dev/tty0 pulls it
727 in or lingering is on (so that containers don't bother with it until PAM is used). also exit-on-idle
729 * cache sd_event_now() result from before the first iteration...
731 * PID1: find a way how we can reload unit file configuration for
732 specific units only, without reloading the whole of systemd
734 * add an explicit parser for LimitRTPRIO= that verifies
735 the specified range and generates sane error messages for incorrect
738 * when we detect that there are waiting jobs but no running jobs, do something
740 * push CPUAffinity= also into the "cpuset" cgroup controller
742 * PID 1 should send out sd_notify("WATCHDOG=1") messages (for usage in the --user mode, and when run via nspawn)
744 * there's probably something wrong with having user mounts below /sys,
745 as we have for debugfs. for example, src/core/mount.c handles mounts
746 prefixed with /sys generally special.
747 http://lists.freedesktop.org/archives/systemd-devel/2015-June/032962.html
749 * fstab-generator: default to tmpfs-as-root if only usr= is specified on the kernel cmdline
751 * docs: bring http://www.freedesktop.org/wiki/Software/systemd/MyServiceCantGetRealtime up to date
753 * add a job mode that will fail if a transaction would mean stopping
754 running units. Use this in timedated to manage the NTP service
756 http://lists.freedesktop.org/archives/systemd-devel/2015-April/030229.html
758 * The udev blkid built-in should expose a property that reflects
759 whether media was sensed in USB CF/SD card readers. This should then
760 be used to control SYSTEMD_READY=1/0 so that USB card readers aren't
761 picked up by systemd unless they contain a medium. This would mirror
762 the behaviour we already have for CD drives.
764 * networkd/udev: implement SR_IOV configuration in .link files:
765 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027451.html
767 * hostnamectl: show root image uuid
769 * Find a solution for SMACK capabilities stuff:
770 http://lists.freedesktop.org/archives/systemd-devel/2014-December/026188.html
772 * "systemctl preset-all" should probably order the unit files it
773 operates on lexicographically before starting to work, in order to
774 ensure deterministic behaviour if two unit files conflict (like DMs
777 * synchronize console access with BSD locks:
778 http://lists.freedesktop.org/archives/systemd-devel/2014-October/024582.html
780 * as soon as we have sender timestamps, revisit coalescing multiple parallel daemon reloads:
781 http://lists.freedesktop.org/archives/systemd-devel/2014-December/025862.html
783 * figure out when we can use the coarse timers
785 * add "systemctl start -v foobar.service" that shows logs of a service
786 while the start command runs. This is non-trivial to do without
787 races though, since we should flush out all journal messages before
788 returning from the "systemctl stop".
790 * 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
792 * maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment=
794 * maybe allow timer units with an empty Units= setting, so that they
795 can be used for resuming the system but nothing else.
797 * what to do about udev db binary stability for apps? (raw access is not an option)
799 * systemctl: if some operation fails, show log output?
801 * systemctl edit: use equivalent of cat() to insert existing config as a comment, prepended with #.
802 Upon editor exit, lines with one # are removed, lines with two # are left with one #, etc.
804 * exponential backoff in timesyncd when we cannot reach a server
806 * timesyncd: add ugly bus calls to set NTP servers per-interface, for usage by NM
808 * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share....
810 * add systemd.abort_on_kill or some other such flag to send SIGABRT instead of SIGKILL
811 (throughout the codebase, not only PID1)
815 - service registration
816 - service/domain/types browsing
818 - DNS-SD service registration from socket units
819 - resolved should optionally register additional per-interface LLMNR
820 names, so that for the container case we can establish the same name
821 (maybe "host") for referencing the server, everywhere.
822 - allow clients to request DNSSEC for a single lookup even if DNSSEC is off (?)
823 - hook up resolved with machined-based address resolution
825 * refcounting in sd-resolve is borked
827 * Add a new verb "systemctl top"
829 * add new gpt type for btrfs volumes
831 * support empty /etc boots nicely:
832 - nspawn/gpt-generator: introduce new gpt partition type for /usr
834 * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them.
836 * a way for container managers to turn off getty starting via $container_headless= or so...
838 * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit
840 * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services
841 they run added to the initial transaction and thus confuse Type=idle.
843 * add bus api to query unit file's X fields.
845 * gpt-auto-generator:
846 - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap?
847 - Make /home automount rather than mount?
849 * add generator that pulls in systemd-network from containers when
850 CAP_NET_ADMIN is set, more than the loopback device is defined, even
851 when it is otherwise off
853 * MessageQueueMessageSize= (and suchlike) should use parse_iec_size().
855 * implement Distribute= in socket units to allow running multiple
856 service instances processing the listening socket, and open this up
859 * introduce bus call FreezeUnit(s, b), as well as "systemctl freeze
860 $UNIT" and "systemctl thaw $UNIT" as wrappers around this. The calls
861 should SIGSTOP all unit processes in a loop until all processes of
862 it are fully stopped. This can later be used for app management by
863 desktop UIs such as gnome-shell to freeze apps that are not visible
864 on screen, not unlike how job control works on the shell
867 - implement per-slice CPUFairScheduling=1 switch
868 - introduce high-level settings for RT budget, swappiness
869 - how to reset dynamically changed unit cgroup attributes sanely?
870 - when reloading configuration, apply new cgroup configuration
871 - when recursively showing the cgroup hierarchy, optionally also show
872 the hierarchies of child processes
875 - add field to transient units that indicate whether systemd or somebody else saves/restores its settings, for integration with libvirt
877 * when we detect low battery and no AC on boot, show pretty splash and refuse boot
879 * libsystemd-journal, libsystemd-login, libudev: add calls to easily attach these objects to sd-event event loops
881 * be more careful what we export on the bus as (usec_t) 0 and (usec_t) -1
883 * 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
885 * After coming back from hibernation reset hibernation swap partition using the /dev/snapshot ioctl APIs
887 * If we try to find a unit via a dangling symlink, generate a clean
888 error. Currently, we just ignore it and read the unit from the search
891 * refuse boot if /usr/lib/os-release is missing or /etc/machine-id cannot be set up
893 * man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted.
895 * load .d/*.conf dropins for device units
899 - GetAllProperties() on a non-existing object does not result in a failure currently
900 - port to sd-resolve for connecting to TCP dbus servers
901 - 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
902 - see if we can drop more message validation on the sending side
903 - add API to clone sd_bus_message objects
904 - longer term: priority inheritance
906 - NameLost/NameAcquired obsolete
909 - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now
912 - allow multiple signal handlers per signal?
913 - document chaining of signal handler for SIGCHLD and child handlers
914 - define more intervals where we will shift wakeup intervals around in, 1h, 6h, 24h, ...
916 * investigate endianness issues of UUID vs. GUID
918 * dbus: when a unit failed to load (i.e. is in UNIT_ERROR state), we
919 should be able to safely try another attempt when the bus call LoadUnit() is invoked.
921 * add a pam module that passes the hdd passphrase into the PAM stack and then expires it, for usage by gdm auto-login.
923 * add a pam module that on password changes updates any LUKS slot where the password matches
926 - add unit tests for config_parse_device_allow()
928 * seems that when we follow symlinks to units we prefer the symlink
929 destination path over /etc and /usr. We should not do that. Instead
930 /etc should always override /run+/usr and also any symlink
933 * when isolating, try to figure out a way how we implicitly can order
934 all units we stop before the isolating unit...
936 * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off})
938 * Add ConditionDirectoryNotEmpty= handle non-absoute paths as a search path or add
939 ConditionConfigSearchPathNotEmpty= or different syntax? See the discussion starting at
940 https://github.com/systemd/systemd/pull/15109#issuecomment-607740136.
942 * BootLoaderSpec: Clarify that the kernel has to be in $BOOT. Clarify
943 that the boot loader should be installed to the ESP. Define a way
944 how an installer can figure out whether a BLS compliant boot loader
947 * think about requeuing jobs when daemon-reload is issued? usecase:
948 the initrd issues a reload after fstab from the host is accessible
949 and we might want to requeue the mounts local-fs acquired through
952 * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep()
954 * remove any syslog support from log.c — we probably cannot do this before split-off udev is gone for good
956 * shutdown logging: store to EFI var, and store to USB stick?
958 * merge unit_kill_common() and unit_kill_context()
961 - honor language efi variables for default language selection (if there are any?)
962 - honor timezone efi variables for default timezone selection (if there are any?)
963 - change bootctl to be backed by systemd-bootd to control temporary and persistent default boot goal plus efi variables
965 - recognize the case when not booted on EFI
967 * maybe do not install getty@tty1.service symlink in /etc but in /usr?
969 * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word
971 * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units.
974 - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around
975 - logind: wakelock/opportunistic suspend support
976 - Add pretty name for seats in logind
977 - logind: allow showing logout dialog from system?
978 - add Suspend() bus calls which take timestamps to fix double suspend issues when somebody hits suspend and closes laptop quickly.
979 - if pam_systemd is invoked by su from a process that is outside of a
980 any session we should probably just become a NOP, since that's
981 usually not a real user session but just some system code that just
983 - logind: make the Suspend()/Hibernate() bus calls wait for the for
984 the job to be completed. before returning, so that clients can wait
985 for "systemctl suspend" to finish to know when the suspending is
987 - logind: when the power button is pressed short, just popup a
988 logout dialog. If it is pressed for 1s, do the usual
989 shutdown. Inspiration are Macs here.
990 - expose "Locked" property on logind sesison objects
991 - maybe allow configuration of the StopTimeout for session scopes
992 - rename session scope so that it includes the UID. THat way
993 the session scope can be arranged freely in slices and we don't have
994 make assumptions about their slice anymore.
995 - follow PropertiesChanged state more closely, to deal with quick logouts and
997 - (optionally?) spawn seat-manager@$SEAT.service whenever a seat shows up that as CanGraphical set
1000 - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields
1001 - journald: also get thread ID from client, plus thread name
1002 - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups
1003 - add API to close/reopen/get fd for journal client fd in libsystemd-journal.
1004 - fallback to /dev/log based logging in libsystemd-journal, if we cannot log natively?
1005 - declare the local journal protocol stable in the wiki interface chart
1006 - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg
1007 - journald: when dropping msgs due to ratelimit make sure to write
1008 "dropped %u messages" not only when we are about to print the next
1009 message that works, but already after a short timeout
1010 - check if we can make journalctl by default use --follow mode inside of less if called without args?
1011 - maybe add API to send pairs of iovecs via sd_journal_send
1012 - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access
1013 - journactl: support negative filtering, i.e. FOOBAR!="waldo",
1014 and !FOOBAR for events without FOOBAR.
1015 - journal: store timestamp of journal_file_set_offline() int he header,
1016 so it is possible to display when the file was last synced.
1017 - 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.
1018 - journal: find a way to allow dropping history early, based on priority, other rules
1019 - journal: When used on NFS, check payload hashes
1020 - journald: add kernel cmdline option to disable ratelimiting for debug purposes
1021 - refuse taking lower-case variable names in sd_journal_send() and friends.
1022 - journald: we currently rotate only after MaxUse+MaxFilesize has been reached.
1023 - journal: deal nicely with byte-by-byte copied files, especially regards header
1024 - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit
1025 - Replace utmp, wtmp, btmp, and lastlog completely with journal
1026 - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax?
1027 - when a kernel driver logs in a tight loop, we should ratelimit that too.
1028 - journald: optionally, log debug messages to /run but everything else to /var
1029 - journald: when we drop syslog messages because the syslog socket is
1030 full, make sure to write how many messages are lost as first thing
1031 to syslog when it works again.
1032 - journald: allow per-priority and per-service retention times when rotating/vacuuming
1033 - journald: make use of uid-range.h to managed uid ranges to split
1035 - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so...
1036 - improve journalctl performance by loading journal files
1037 lazily. Encode just enough information in the file name, so that we
1038 do not have to open it to know that it is not interesting for us, for
1039 the most common operations.
1040 - man: document that corrupted journal files is nothing to act on
1041 - rework journald sigbus stuff to use mutex
1042 - Set RLIMIT_NPROC for systemd-journal-xyz, and all other of our
1043 services that run under their own user ids, and use User= (but only
1044 in a world where userns is ubiquitous since otherwise we cannot
1045 invoke those daemons on the host AND in a container anymore). Also,
1046 if LimitNPROC= is used without User= we should warn and refuse
1048 - journalctl --verify: don't show files that are currently being
1049 written to as FAIL, but instead show that their are being written to.
1050 - add journalctl -H that talks via ssh to a remote peer and passes through
1052 - add a version of --merge which also merges /var/log/journal/remote
1053 - journalctl: -m should access container journals directly by enumerating
1054 them via machined, and also watch containers coming and going.
1055 Benefit: nspawn --ephemeral would start working nicely with the journal.
1056 - assign MESSAGE_ID to log messages about failed services
1058 * add a test if all entries in the catalog are properly formatted.
1059 (Adding dashes in a catalog entry currently results in the catalog entry
1060 being silently skipped. journalctl --update-catalog must warn about this,
1061 and we should also have a unit test to check that all our message are OK.)
1064 - document that deps in [Unit] sections ignore Alias= fields in
1065 [Install] units of other units, unless those units are disabled
1066 - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets
1067 - document that service reload may be implemented as service reexec
1068 - add a man page containing packaging guidelines and recommending usage of things like Documentation=, PrivateTmp=, PrivateNetwork= and ReadOnlyDirectories=/etc /usr.
1069 - document systemd-journal-flush.service properly
1070 - documentation: recommend to connect the timer units of a service to the service via Also= in [Install]
1071 - man: document the very specific env the shutdown drop-in tools live in
1072 - man: add more examples to man pages
1073 - man: maybe sort directives in man pages, and take sections from --help and apply them to man too
1074 - document root=gpt-auto properly
1077 - add systemctl switch to dump transaction without executing it
1078 - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done
1079 - "systemctl disable" on a static unit prints no message and does
1080 nothing. "systemctl enable" does nothing, and gives a bad message
1081 about it. Should fix both to print nice actionable messages.
1082 - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service
1083 - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible
1084 - systemctl enable: fail if target to alias into does not exist? maybe show how many units are enabled afterwards?
1085 - systemctl: "Journal has been rotated since unit was started." message is misleading
1086 - systemctl status output should include list of triggering units and their status
1089 - "systemctl mask" should find all names by which a unit is accessible
1090 (i.e. by scanning for symlinks to it) and link them all to /dev/null
1093 - timer units should get the ability to trigger when:
1095 - Modulate timer frequency based on battery state
1097 * add libsystemd-password or so to query passwords during boot using the password agent logic
1099 * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed
1101 * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel
1103 * make repeated alt-ctrl-del presses printing a dump
1105 * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ...
1107 * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab is not
1110 - emulate /dev/kmsg using CUSE and turn off the syslog syscall
1111 with seccomp. That should provide us with a useful log buffer that
1112 systemd can log to during early boot, and disconnect container logs
1113 from the kernel's logs.
1114 - as soon as networkd has a bus interface, hook up --network-interface=,
1115 --network-bridge= with networkd, to trigger netdev creation should an
1116 interface be missing
1117 - a nice way to boot up without machine id set, so that it is set at boot
1118 automatically for supporting --ephemeral. Maybe hash the host machine id
1119 together with the machine name to generate the machine id for the container
1120 - fix logic always print a final newline on output.
1121 https://github.com/systemd/systemd/pull/272#issuecomment-113153176
1122 - should optionally support receiving WATCHDOG=1 messages from its payload
1124 - optionally automatically add FORWARD rules to iptables whenever nspawn is
1125 running, remove them when shut down.
1128 - refuse mounting over a mount point
1129 - automatically discover .roothash files in dissect, similarly to nspawn
1132 - add an API so that libvirt-lxc can inform us about network interfaces being
1133 removed or added to an existing machine
1134 - "machinectl migrate" or similar to copy a container from or to a
1135 difference host, via ssh
1136 - introduce systemd-nspawn-ephemeral@.service, and hook it into
1137 "machinectl start" with a new --ephemeral switch
1138 - "machinectl status" should also show internal logs of the container in
1140 - "machinectl history"
1142 - "machinectl commit" that takes a writable snapshot of a tree, invokes a
1143 shell in it, and marks it read-only after use
1146 - cryptsetup-generator: allow specification of passwords in crypttab itself
1147 - support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator
1149 * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it
1150 https://bugs.freedesktop.org/show_bug.cgi?id=54712
1152 * add a dependency on standard-conf.xml and other included files to man pages
1154 * MountFlags=shared acts as MountFlags=slave right now.
1156 * properly handle loop back mounts via fstab, especially regards to fsck/passno
1158 * initialize the hostname from the fs label of /, if /etc/hostname does not exist?
1163 - add trigger --subsystem-match=usb/usb_device device
1164 - reimport udev db after MOVE events for devices without dev_t
1166 * There's currently no way to cancel fsck (used to be possible via C-c or c on the console)
1168 * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/
1171 - save coredump in Windows/Mozilla minidump format
1172 - when truncating coredumps, also log the full size that the process had, and make a metadata field so we can report truncated coredumps
1174 * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting)
1176 * be able to specify a forced restart of service A where service B depends on, in case B
1177 needs to be auto-respawned?
1180 - apply "x" on "D" too (see patch from William Douglas)
1181 - instead of ignoring unknown fields, reject them.
1182 - creating new directories/subvolumes/fifos/device nodes
1183 should not follow symlinks. None of the other adjustment or creation
1184 calls follow symlinks.
1186 * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early
1188 * verify that the AF_UNIX sockets of a service in the fs still exist
1189 when we start a service in order to avoid confusion when a user
1190 assumes starting a service is enough to make it accessible
1192 * Make it possible to set the keymap independently from the font on
1193 the kernel cmdline. Right now setting one resets also the other.
1195 * and a dbus call to generate target from current state
1197 * write blog stories about:
1198 - hwdb: what belongs into it, lsusb
1199 - enabling dbus services
1200 - how to make changes to sysctl and sysfs attributes
1202 - how to pass throw-away units to systemd, or dynamically change properties of existing units
1203 - testing with Harald's awesome test kit
1205 - how to develop against journal browsing APIs
1206 - the journal HTTP iface
1207 - non-cgroup resource management
1208 - dynamic resource management with cgroups
1209 - refreshed, longer missions statement
1210 - calendar time events
1211 - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell
1212 - how to create your own target
1213 - instantiated apache, dovecot and so on
1214 - hooking a script into various stages of shutdown/rearly booot
1216 * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support.
1218 * dot output for --test showing the 'initial transaction'
1221 - When logging about multiple units (stopping BoundTo units, conflicts, etc.),
1222 log both units as UNIT=, so that journalctl -u triggers on both.
1223 - generate better errors when people try to set transient properties
1224 that are not supported...
1225 http://lists.freedesktop.org/archives/systemd-devel/2015-February/028076.html
1226 - maybe introduce WantsMountsFor=? Usecase:
1227 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027729.html
1228 - recreate systemd's D-Bus private socket file on SIGUSR2
1229 - move PAM code into its own binary
1230 - when we automatically restart a service, ensure we restart its rdeps, too.
1231 - hide PAM options in fragment parser when compile time disabled
1232 - Support --test based on current system state
1233 - 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().
1234 - after deserializing sockets in socket.c we should reapply sockopts and things
1235 - drop PID 1 reloading, only do reexecing (difficult: Reload()
1236 currently is properly synchronous, Reexec() is weird, because we
1237 cannot delay the response properly until we are back, so instead of
1238 being properly synchronous we just keep open the fd and close it
1239 when done. That means clients do not get a successful method reply,
1240 but much rather a disconnect on success.
1241 - when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr
1242 - when a bus name of a service disappears from the bus make sure to queue further activation requests
1245 - allow port=0 in .socket units
1246 - maybe introduce ExecRestartPre=
1247 - add ReloadSignal= for configuring a reload signal to use
1248 - implement Register= switch in .socket units to enable registration
1249 in Avahi, RPC and other socket registration services.
1250 - allow Type=simple with PIDFile=
1251 https://bugzilla.redhat.com/show_bug.cgi?id=723942
1252 - allow writing multiple conditions in unit files on one line
1253 - introduce Type=pid-file
1254 - introduce mix of BindTo and Requisite
1255 - add a concept of RemainAfterExit= to scope units
1256 - Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely
1257 - add verification of [Install] section to systemd-analyze verify
1260 - Make sure ID_PATH is always exported and complete for
1261 network devices where possible, so we can safely rely
1265 - add support for more attribute types
1266 - inbuilt piping support (essentially degenerate async)? see loopback-setup.c and other places
1269 - add more keys to [Route] and [Address] sections
1270 - add support for more DHCPv4 options (and, longer term, other kinds of dynamic config)
1271 - add reduced [Link] support to .network files
1272 - properly handle routerless dhcp leases
1273 - work with non-Ethernet devices
1274 - dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from?
1275 - the DHCP lease data (such as NTP/DNS) is still made available when
1276 a carrier is lost on a link. It should be removed instantly.
1277 - expose in the API the following bits:
1278 - option 15, domain name
1279 - option 12, hostname and/or option 81, fqdn
1280 - option 123, 144, geolocation
1281 - option 252, configure http proxy (PAC/wpad)
1282 - provide a way to define a per-network interface default metric value
1283 for all routes to it. possibly a second default for DHCP routes.
1284 - allow Name= to be specified repeatedly in the [Match] section. Maybe also
1285 support Name=foo*|bar*|baz ?
1286 - duplicate address check for static IPs (like ARPCHECK in network-scripts)
1287 - whenever uplink info changes, make DHCP server send out FORCERENEW
1289 * Figure out how to do unittests of networkd's state serialization
1292 - figure out how much we can increase Maximum Message Size
1295 - add functions to set previously stored IPv6 addresses on startup and get
1296 them at shutdown; store them in client->ia_na
1297 - write more test cases
1298 - implement reconfigure support, see 5.3., 15.11. and 22.20.
1299 - implement support for temporary adressess (IA_TA)
1300 - implement dhcpv6 authentication
1301 - investigate the usefulness of Confirm messages; i.e. are there any
1302 situations where the link changes without any loss in carrier detection
1304 - some servers don't do rapid commit without a filled in IA_NA, verify
1311 - natively watch for dbus-*.service symlinks (PENDING)
1312 - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service
1314 * make cryptsetup lower --iter-time
1316 * kernel: add device_type = "fb", "fbcon" to class "graphics"
1318 * /usr/bin/service should actually show the new command line
1320 * 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=
1322 * neither pkexec nor sudo initialize environ[] from the PAM environment?
1324 * fedora: update policy to declare access mode and ownership of unit files to root:root 0644, and add an rpmlint check for it
1326 * register catalog database signature as file magic
1328 * zsh shell completion:
1329 - <command> <verb> -<TAB> should complete options, but currently does not
1330 - systemctl add-wants,add-requires
1332 * systemctl status should know about 'systemd-analyze calendar ... --iterations='
1333 * If timer has just OnInactiveSec=..., it should fire after a specified time
1334 after being started.
1338 * look for close() vs. close_nointr() vs. close_nointr_nofail()
1340 * check for strerror(r) instead of strerror(-r)
1344 * set_put(), hashmap_put() return values check. i.e. == 0 does not free()!
1346 * use secure_getenv() instead of getenv() where appropriate
1348 * link up selected blog stories from man pages and unit files Documentation= fields
1349 String is not UTF-8 clean, ignoring assignment