3 * copy.c: set the right chattrs before copying files and others after
7 * Fedora: add an rpmlint check that verifies that all unit files in the RPM are listed in %systemd_post macros.
9 * wiki: update journal format documentation for lz4 additions
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 * set memory.oom.group in cgroupsv2 for all leaf cgroups
22 * whenever oom_kill memory.event event is triggered print a nice log message
24 * Move RestrictAddressFamily= to the new cgroup create socket
26 * support the bind/connect/sendmsg cgroup stuff for sandboxing, and possibly
29 * chown() tty a service is attached to after the service goes down
31 * optionally, if a per-partition GPT flag is set for the root/home/… partitions
32 format the partition on next boot and unset the flag, in order to implement
33 factory reset. also, add a second flag that simply indicates whether such a
34 scheme is supported. then, add a tool (or maybe beef up systemd-dissect) to
35 show state of these flags, and optionally trigger such a factory reset on
36 next boot by setting the flag.
38 * logind: maybe watch utmp asynchronously using inotify, and populate our own
39 tracked session metadata from the fields available therein. Why bother? Right
40 now, all "ssh" sessions will be tracked without their TTY by logind (which is
41 not just unfriendly to users as this means "loginctl session-status" shows
42 less information than "who" in many cases, but also breaks the IdleAction
43 logic, as we never can detect such sessions as idle, as we have no TTY to
44 watch). ssh sets the PAM_TTY field on its PAM sessions to "ssh" rather than
45 the actual pty, because the PAM session is opened early on for new
46 connections, but the PTY only registered much later (if at all). ssh writes
47 the utmp record only after a TTY is actually registered, hence we could use
48 this data then, and use it if it is available. Using utmp for this is ugly of
49 course, and watching things asynchronously even more so, but it should be
50 good enough for the idle detection logic at least.
52 * maybe extend .path units to expose fanotify() per-mount change events
54 * Add a "systemctl list-units --by-slice" mode or so, which rearranges the
55 output of "systemctl list-units" slightly by showing the tree structure of
56 the slices, and the units attached to them.
58 * the a-posteriori stopping of units bound to units that disappeared logic
59 should be reworked: there should be a queue of units, and we should only
60 enqeue stop jobs from a defer event that processes queue instead of
61 right-away when we find a unit that is bound to one that doesn't exist
62 anymore. (similar to how the stop-unneeded queue has been reworked the same
65 * nspawn: make nspawn suitable for shell pipelines: instead of triggering a
66 hangup when input is finished, send ^D, which synthesizes an EOF. Then wait
67 for hangup or ^D before passing on the EOF.
69 * When reloading configuration PID 1 should reset all its properties to the
70 original defaults before calling parse_config()
72 * Add OnTimezoneChange= and OnTimeChange= stanzas to .timer units in order to
73 schedule events based on time and timezone changes.
75 * nspawn: greater control over selinux label?
77 * cgroups: figure out if we can somehow communicate in a cleaner way whether a
78 systemd instance not running in the cgroup root shall or shall not manage the
79 attributes of its top-level cgroup. Currently it assumes it manages all, but
80 then might get EPERM due to permission porblems/userns, which is OK, but this
81 should be revisited to make clearer and also work if the payload systemd runs
82 with full privs and without userns.
84 * hibernate/s2h: make this robust and safe to enable in Fedora by default.
87 1. add resume_offset support to the resume code (i.e. support swap files
89 2. check of swap is on weird storage and refuse if so
90 3. add env-var based option to disable hibernation
91 4. figure out what to do with swap-on-luks
92 5. add autodetection of hibernation images, and add "noresume" to disable
95 * portables: introduce a new unit file directory /etc/systemd/system.attached/
96 or so, where we attach portable services to
98 * cgroups: use inotify to get notified when somebody else modifies cgroups
99 owned by us, then log a friendly warning.
101 * beef up log.c with support for stripping ANSI sequences from strings, so that
102 it is OK to include them in log strings. This would be particularly useful so
103 that our log messages could contain clickable links for example for unit
104 files and suchlike we operate on.
106 * add support for "portablectl attach http://foobar.com/waaa.raw (i.e. importd integration)
108 * add attach --enable and attach --now (for attach+enable+start)
110 * sync dynamic uids/gids between host+portable srvice (i.e. if DynamicUser=1 is set for a service, make sure that the
111 selected user is resolvable in the service even if it ships its own /etc/passwd)
113 * Fix DECIMAL_STR_MAX or DECIMAL_STR_WIDTH. One includes a trailing NUL, the
114 other doesn't. What a desaster. Probably to exclude it. Also
115 DECIMAL_STR_WIDTH should probably add an extra "-" into account for negative
118 * Check that users of inotify's IN_DELETE_SELF flag are using it properly, as
119 usually IN_ATTRIB is the right way to watch deleted files, as the former only
120 fires when a file is actually removed from disk, i.e. the link count drops to
121 zero and is not open anymore, while the latter happens when a file is
122 unlinked from any dir.
124 * port systemctl, busctl, … over to format-table.[ch]'s table formatters
126 * pid1: lock image configured with RootDirectory=/RootImage= using the usual nspawn semantics while the unit is up
128 * add --vacuum-xyz options to coredumpctl, matching those journalctl already has.
130 * SuccessExitStatus= and friends should probably also accept symbolic exit
131 codes names, i.e. error codes from the list maintained in exit-codes.[ch]
133 * introduce Ephemeral= unit file switch, that creates an ephemeral copy of all
134 files and directories that are left writable for a unit, and which are
135 removed after the unit goes down again. A bit like --ephemeral for
136 systemd-nspawn but for system services. If used together with RootImage= this
137 should reflink the image file itself.
139 Related: add Ephemeral=<path1> <path2> … which would allow marking
140 specific paths only like this.
142 * add CopyFile= or so as unit file setting that may be used to copy files or
143 directory trees from the host to te services RootImage= and RootDirectory=
144 environment. Which we can use for /etc/machine-id and in particular
145 /etc/resolv.conf. Should be smart and do something useful on read-only
146 images, for example fallback to read-only bind mounting the file instead.
148 * nspawn's console TTY should be allocated from within the container, not
149 mounted in from the outside
151 * show invocation ID in systemd-run output
153 * bypass SIGTERM state in unit files if KillSignal is SIGKILL
155 * tree-wide: ensure we always block the signals we hook into with
156 sd_event_add_signal() first
158 * add proper dbus APIs for the various sd_notify() commands, such as MAINPID=1
159 and so on, which would mean we could report errors and such.
161 * teach tmpfiles.d q/Q logic something sensible in the context of XFS/ext4
164 * introduce DefaultSlice= or so in system.conf that allows changing where we
165 place our units by default, i.e. change system.slice to something
166 else. Similar, ManagerSlice= should exist so that PID1's own scope unit could
167 be moved somewhere else too. Finally machined and logind should get similar
168 options so that it is possible to move user session scopes and machines to a
169 different slice too by default. Usecase: people who want to put resources on
170 the entire system, with the exception of one specific service. See:
171 https://lists.freedesktop.org/archives/systemd-devel/2018-February/040369.html
173 * maybe rework get_user_creds() to query the user database if $SHELL is used
174 for root, but only then.
176 * be stricter with fds we receive for the fdstore: close them asynchronously
178 * calenderspec: add support for week numbers and day numbers within a
179 year. This would allow us to define "bi-weekly" triggers safely.
181 * add bpf-based implementation of devices cgroup controller logic for compat
182 with cgroupsv2 as supported by newest kernel
184 * introduce sd_id128_get_boot_app_specific() which is like
185 sd_id128_get_machine_app_specific(). After all on long-running systems both
186 IDs have similar properties.
188 * sd-bus: add vtable flag, that may be used to request client creds implicitly
189 and asynchronously before dispatching the operation
191 * make use of ethtool veth peer info in machined, for automatically finding out
192 host-side interface pointing to the container.
194 * add some special mode to LogsDirectory=/StateDirectory=… that allows
195 declaring these directories without necessarily pulling in deps for them, or
196 creating them when starting up. That way, we could declare that
197 systemd-journald writes to /var/log/journal, which could be useful when we
198 doing disk usage calculations and so on.
200 * taint systemd if there are fewer than 65536 users assigned (userns) to the system.
202 * deprecate PermissionsStartOnly= and RootDirectoryStartOnly= in favour of the ExecStart= prefix chars
204 * add a new RuntimeDirectoryPreserve= mode that defines a similar lifecycle for
205 the runtime dir as we maintain for the fdstore: i.e. keep it around as long
206 as the unit is running or has a job queued.
208 * support projid-based quota in machinectl for containers, and then drop
209 implicit btrfs loopback magic in machined
211 * Add NetworkNamespacePath= to specify a path to a network namespace
213 * maybe use SOURCE_DATE_EPOCH (i.e. the env var the reproducible builds folks
214 introduced) as the RTC epoch, instead of the mtime of NEWS.
216 * add a way to lock down cgroup migration: a boolean, which when set for a unit
217 makes sure the processes in it can never migrate out of it
219 * blog about fd store and restartable services
221 * document Environment=SYSTEMD_LOG_LEVEL=debug drop-in in debugging document
223 * rework ExecOutput and ExecInput enums so that EXEC_OUTPUT_NULL loses its
224 magic meaning and is no longer upgraded to something else if set explicitly.
226 * in the long run: permit a system with /etc/machine-id linked to /dev/null, to
227 make it lose its identity, i.e. be anonymous. For this we'd have to patch
228 through the whole tree to make all code deal with the case where no machine
231 * optionally, collect cgroup resource data, and store it in per-unit RRD files,
232 suitable for processing with rrdtool. Add bus API to access this data, and
233 possibly implement a CPULoad property based on it.
235 * beef up pam_systemd to take unit file settings such as cgroups properties as
238 * a new "systemd-analyze security" tool outputting a checklist of security
239 features a service does and does not implement
241 * maybe hook of xfs/ext4 quotactl() with services? i.e. automatically manage
242 the quota of a the user indicated in User= via unit file settings, like the
243 other resource management concepts. Would mix nicely with DynamicUser=1. Or
244 alternatively, do this with projids, so that we can also cover services
245 running as root. Quota should probably cover all the special dirs such as
246 StateDirectory=, LogsDirectory=, CacheDirectory=, as well as RootDirectory= if it
247 is set, plus the whole disk space any image configured with RootImage=.
249 * Introduce "exit" as an EmergencyAction value, and allow to configure a
250 per-unit success/failure exit code to configure. This would be useful for
251 running commands inside of services inside of containers, which could then
252 propagate their failure state all the way up.
254 * In DynamicUser= mode: before selecting a UID, use disk quota APIs on relevant
255 disks to see if the UID is already in use.
257 * add "systemctl wait" or so, which does what "systemd-run --wait" does, but
258 for all units. It should be both a way to pin units into memory as well as a
259 wait to retrieve their exit data.
261 * expose IO accounting data on the bus, show it in systemd-run --wait and log
262 about it in the resource log message
264 * add "systemctl purge" for flushing out configuration, state, logs, ... of a
265 unit when it is stopped
267 * show whether a service has out-of-date configuration in "systemctl status" by
268 using mtime data of ConfigurationDirectory=.
270 * replace all uses of fgets() + LINE_MAX by read_line()
272 * Add AddUser= setting to unit files, similar to DynamicUser=1 which however
273 creates a static, persistent user rather than a dynamic, transient user. We
274 can leverage code from sysusers.d for this.
276 * add some optional flag to ReadWritePaths= and friends, that has the effect
277 that we create the dir in question when the service is started. Example:
279 ReadWritePaths=:/var/lib/foobar
281 * maybe add call sd_journal_set_block_timeout() or so to set SO_SNDTIMEO for
282 the sd-journal logging socket, and, if the timeout is set to 0, sets
283 O_NONBLOCK on it. That way people can control if and when to block for
286 * hostnamed: populate form factor data from a new hwdb database, so that old
287 yogas can be recognized as "convertible" too, even if they predate the DMI
288 "convertible" form factor
290 * Maybe add a small tool invoked early at boot, that adds in or resizes
291 partitions automatically, to be used when the media used is actually larger
292 than the image written onto it is.
294 * Maybe add PrivatePIDs= as new unit setting, and do minimal PID namespacing
295 after all. Be strict however, only support the equivalent of nspawn's
296 --as-pid2 switch, and sanely proxy sd_notify() messages dropping stuff such
299 * Add ExecMonitor= setting. May be used multiple times. Forks off a process in
300 the service cgroup, which is supposed to monitor the service, and when it
301 exits the service is considered failed by its monitor.
303 * track the per-service PAM process properly (i.e. as an additional control
304 process), so that it may be queried on the bus and everything.
306 * add a new "debug" job mode, that is propagated to unit_start() and for
307 services results in two things: we raise SIGSTOP right before invoking
308 execve() and turn off watchdog support. Then, use that to implement
309 "systemd-gdb" for attaching to the start-up of any system service in its
312 * maybe introduce gpt auto discovery for /var/tmp?
314 * maybe add gpt-partition-based user management: each user gets his own
315 LUKS-encrypted GPT partition with a new GPT type. A small nss module
316 enumerates users via udev partition enumeration. UIDs are assigned in a fixed
317 way: the partition index is added as offset to some fixed base uid. User name
318 is stored in GPT partition name. A PAM module authenticates the user via the
319 LUKS partition password. Benefits: strong per-user security, compatibility
320 with stateless/read-only/verity-enabled root. (other idea: do this based on
321 loopback files in /home, without GPT involvement)
323 * gpt-auto logic: introduce support for discovering /var matching an image. For
324 that, use a partition type UUID that is hashed from the OS name (as encoded
325 in /etc/os-release), the architecture, and 4 new bits from the gpt flags
326 field of the root partition. This way can easily support multiple OS
327 installations on the same GPT partition table, without problems with
328 unmatched /var partitions.
330 * gpt-auto logic: related to the above, maybe support a "secondary" root
331 partition, that is mounted to / and is writable, and where the actual root's
332 /usr is mounted into.
334 * 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
336 * drop nss-myhostname in favour of nss-resolve?
338 * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and
339 then use that for the setting used in user@.service. It should be understood
340 relative to the configured default value.
342 * in networkd, when matching device types, fix up DEVTYPE rubbish the kernel passes to us
344 * enable LockMLOCK to take a percentage value relative to physical memory
346 * Permit masking specific netlink APIs with RestrictAddressFamily=
348 * nspawn: support that /proc, /sys/, /dev are pre-mounted
350 * define gpt header bits to select volatility mode
352 * ProtectKernelLogs= (drops CAP_SYSLOG, add seccomp for syslog() syscall, and DeviceAllow to /dev/kmsg) in service files
354 * ProtectClock= (drops CAP_SYS_TIMES, adds seecomp filters for settimeofday, adjtimex), sets DeviceAllow o /dev/rtc
356 * ProtectTracing= (drops CAP_SYS_PTRACE, blocks ptrace syscall, makes /sys/kernel/tracing go away)
358 * ProtectMount= (drop mount/umount/pivot_root from seccomp, disallow fuse via DeviceAllow, imply Mountflags=slave)
360 * ProtectKeyRing= to take keyring calls away
362 * RemoveKeyRing= to remove all keyring entries of the specified user
364 * ProtectReboot= that masks reboot() and kexec_load() syscalls, prohibits kill
365 on PID 1 with the relevant signals, and makes relevant files in /sys and
366 /proc (such as the sysrq stuff) unavailable
368 * DeviceAllow= should also generate seccomp filters for mknod()
370 * make sure the ratelimit object can deal with USEC_INFINITY as way to turn off things
372 * journalctl: make sure -f ends when the container indicated by -M terminates
374 * mount: automatically search for "main" partition of an image has multiple
377 * expose the "privileged" flag of ExecCommand on the bus, and open it up to
380 * in nss-systemd, if we run inside of RootDirectory= with PrivateUsers= set,
381 find a way to map the User=/Group= of the service to the right name. This way
382 a user/group for a service only has to exist on the host for the right
385 * add bus API for creating unit files in /etc, reusing the code for transient units
387 * add bus API to remove unit files from /etc
389 * add bus API to retrieve current unit file contents (i.e. implement "systemctl cat" on the bus only)
391 * rework fopen_temporary() to make use of open_tmpfile_linkable() (problem: the
392 kernel doesn't support linkat() that replaces existing files, currently)
394 * transient units: don't bother with actually setting unit properties, we
395 reload the unit file anyway
397 * journald: sigbus API via a signal-handler safe function that people may call
398 from the SIGBUS handler
400 * optionally, also require WATCHDOG=1 notifications during service start-up and shutdown
402 * resolved: when routing queries, make sure only look for the *longest* suffix...
404 * delay activation of logind until somebody logs in, or when /dev/tty0 pulls it
405 in or lingering is on (so that containers don't bother with it until PAM is used). also exit-on-idle
407 * cache sd_event_now() result from before the first iteration...
409 * add systemctl stop --job-mode=triggering that follows TRIGGERED_BY deps and adds them to the same transaction
411 * PID1: find a way how we can reload unit file configuration for
412 specific units only, without reloading the whole of systemd
414 * add an explicit parser for LimitRTPRIO= that verifies
415 the specified range and generates sane error messages for incorrect
418 * when we detect that there are waiting jobs but no running jobs, do something
420 * push CPUAffinity= also into the "cpuset" cgroup controller (only after the cpuset controller got ported to the unified hierarchy)
422 * PID 1 should send out sd_notify("WATCHDOG=1") messages (for usage in the --user mode, and when run via nspawn)
424 * there's probably something wrong with having user mounts below /sys,
425 as we have for debugfs. for exmaple, src/core/mount.c handles mounts
426 prefixed with /sys generally special.
427 http://lists.freedesktop.org/archives/systemd-devel/2015-June/032962.html
429 * fstab-generator: default to tmpfs-as-root if only usr= is specified on the kernel cmdline
431 * docs: bring http://www.freedesktop.org/wiki/Software/systemd/MyServiceCantGetRealtime up to date
433 * add a job mode that will fail if a transaction would mean stopping
434 running units. Use this in timedated to manage the NTP service
436 http://lists.freedesktop.org/archives/systemd-devel/2015-April/030229.html
438 * Maybe add support for the equivalent of "ethtool advertise" to .link files?
439 http://lists.freedesktop.org/archives/systemd-devel/2015-April/030112.html
441 * The udev blkid built-in should expose a property that reflects
442 whether media was sensed in USB CF/SD card readers. This should then
443 be used to control SYSTEMD_READY=1/0 so that USB card readers aren't
444 picked up by systemd unless they contain a medium. This would mirror
445 the behaviour we already have for CD drives.
447 * networkd/udev: implement SR_IOV configuration in .link files:
448 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027451.html
450 * hostnamectl: show root image uuid
452 * Find a solution for SMACK capabilities stuff:
453 http://lists.freedesktop.org/archives/systemd-devel/2014-December/026188.html
455 * "systemctl preset-all" should probably order the unit files it
456 operates on lexicographically before starting to work, in order to
457 ensure deterministic behaviour if two unit files conflict (like DMs
460 * synchronize console access with BSD locks:
461 http://lists.freedesktop.org/archives/systemd-devel/2014-October/024582.html
463 * as soon as we have sender timestamps, revisit coalescing multiple parallel daemon reloads:
464 http://lists.freedesktop.org/archives/systemd-devel/2014-December/025862.html
466 * in systemctl list-unit-files: show the install value the presets would suggest for a service in a third column
468 * figure out when we can use the coarse timers
470 * add "systemctl start -v foobar.service" that shows logs of a service
471 while the start command runs. This is non-trivial to do without
472 races though, since we should flush out all journal messages before
473 returning from the "systemctl stop".
475 * 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
477 * maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment=
479 * introduce systemd-timesync-wait.service or so to sync on an NTP fix?
481 * consider showing the unit names during boot up in the status output, not just the unit descriptions
483 * maybe allow timer units with an empty Units= setting, so that they
484 can be used for resuming the system but nothing else.
486 * what to do about udev db binary stability for apps? (raw access is not an option)
488 * man: maybe use the word "inspect" rather than "introspect"?
490 * systemctl: if some operation fails, show log output?
492 * systemctl edit: use equvalent of cat() to insert existing config as a comment, prepended with #.
493 Upon editor exit, lines with one # are removed, lines with two # are left with one #, etc.
495 * exponential backoff in timesyncd when we cannot reach a server
497 * timesyncd: add ugly bus calls to set NTP servers per-interface, for usage by NM
499 * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share....
501 * systemd.show_status= should probably have a mode where only failed
504 * add systemd.abort_on_kill or some other such flag to send SIGABRT instead of SIGKILL
505 (throughout the codebase, not only PID1)
509 - service registration
510 - service/domain/types browsing
512 - DNS-SD service registration from socket units
513 - resolved should optionally register additional per-interface LLMNR
514 names, so that for the container case we can establish the same name
515 (maybe "host") for referencing the server, everywhere.
516 - allow clients to request DNSSEC for a single lookup even if DNSSEC is off (?)
517 - hook up resolved with machined-based address resolution
519 * refcounting in sd-resolve is borked
521 * Add a new verb "systemctl top"
523 * add new gpt type for btrfs volumes
525 * support empty /etc boots nicely:
526 - nspawn/gpt-generator: introduce new gpt partition type for /usr
528 * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them.
530 * a way for container managers to turn off getty starting via $container_headless= or so...
532 * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit
534 * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services
535 they run added to the initial transaction and thus confuse Type=idle.
537 * add bus api to query unit file's X fields.
539 * gpt-auto-generator:
540 - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap?
541 - Make /home automount rather than mount?
543 * add generator that pulls in systemd-network from containers when
544 CAP_NET_ADMIN is set, more than the loopback device is defined, even
545 when it is otherwise off
547 * MessageQueueMessageSize= (and suchlike) should use parse_iec_size().
549 * implement Distribute= in socket units to allow running multiple
550 service instances processing the listening socket, and open this up
553 * socket units: support creating sockets in different namespace,
554 opening it up for JoinsNamespaceOf=. This would require to fork off
555 a tiny process that joins the namespace and creates/binds the socket
556 and passes this back to PID1 via SCM_RIGHTS. This also could be used
557 to allow Chown/chgrp on sockets without requiring NSS in PID 1.
559 * introduce bus call FreezeUnit(s, b), as well as "systemctl freeze
560 $UNIT" and "systemctl thaw $UNIT" as wrappers around this. The calls
561 should SIGSTOP all unit processes in a loop until all processes of
562 it are fully stopped. This can later be used for app management by
563 desktop UIs such as gnome-shell to freeze apps that are not visible
564 on screen, not unlike how job control works on the shell
567 - implement per-slice CPUFairScheduling=1 switch
568 - handle jointly mounted controllers correctly
569 - introduce high-level settings for RT budget, swappiness
570 - how to reset dynamically changed unit cgroup attributes sanely?
571 - when reloading configuration, apply new cgroup configuration
572 - when recursively showing the cgroup hierarchy, optionally also show
573 the hierarchies of child processes
576 - add field to transient units that indicate whether systemd or somebody else saves/restores its settings, for integration with libvirt
578 * Automatically configure swap partition to use for hibernation by looking for largest swap partition on the root disk?
580 * when we detect low battery and no AC on boot, show pretty splash and refuse boot
582 * libsystemd-journal, libsystemd-login, libudev: add calls to easily attach these objects to sd-event event loops
584 * be more careful what we export on the bus as (usec_t) 0 and (usec_t) -1
586 * 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
588 * After coming back from hibernation reset hibernation swap partition using the /dev/snapshot ioctl APIs
590 * If we try to find a unit via a dangling symlink, generate a clean
591 error. Currently, we just ignore it and read the unit from the search
594 * refuse boot if /usr/lib/os-release is missing or /etc/machine-id cannot be set up
596 * man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted.
598 * load .d/*.conf dropins for device units
600 * allow implementation of InaccessibleDirectories=/ plus
601 ReadOnlyDirectories=... for whitelisting files for a service.
605 - GetAllProperties() on a non-existing object does not result in a failure currently
606 - port to sd-resolve for connecting to TCP dbus servers
607 - 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
608 - see if we can drop more message validation on the sending side
609 - add API to clone sd_bus_message objects
610 - longer term: priority inheritance
612 - NameLost/NameAcquired obsolete
615 - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now
618 - allow multiple signal handlers per signal?
619 - document chaining of signal handler for SIGCHLD and child handlers
620 - define more intervals where we will shift wakeup intervals around in, 1h, 6h, 24h, ...
621 - generate a failure of a default event loop is executed out-of-thread
622 - maybe add support for inotify events (which we can do safely now, with O_PATH)
624 * investigate endianness issues of UUID vs. GUID
626 * dbus: when a unit failed to load (i.e. is in UNIT_ERROR state), we
627 should be able to safely try another attempt when the bus call LoadUnit() is invoked.
629 * add a pam module that passes the hdd passphrase into the PAM stack and then expires it, for usage by gdm auto-login.
631 * add a pam module that on password changes updates any LUKS slot where the password matches
633 * maybe add a generator that looks for "systemd.run=" on the kernel cmdline for container usercases...
636 - add unit tests for config_parse_device_allow()
638 * seems that when we follow symlinks to units we prefer the symlink
639 destination path over /etc and /usr. We should not do that. Instead
640 /etc should always override /run+/usr and also any symlink
643 * when isolating, try to figure out a way how we implicitly can order
644 all units we stop before the isolating unit...
646 * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off})
648 * BootLoaderSpec: Clarify that the kernel has to be in $BOOT. Clarify
649 that the boot loader should be installed to the ESP. Define a way
650 how an installer can figure out whether a BLS compliant boot loader
653 * think about requeuing jobs when daemon-reload is issued? usecase:
654 the initrd issues a reload after fstab from the host is accessible
655 and we might want to requeue the mounts local-fs acquired through
658 * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep()
660 * remove any syslog support from log.c — we probably cannot do this before split-off udev is gone for good
662 * shutdown logging: store to EFI var, and store to USB stick?
664 * merge unit_kill_common() and unit_kill_context()
666 * introduce ExecCondition= in services
669 - honor language efi variables for default language selection (if there are any?)
670 - honor timezone efi variables for default timezone selection (if there are any?)
671 - change bootctl to be backed by systemd-bootd to control temporary and persistent default boot goal plus efi variables
673 * maybe do not install getty@tty1.service symlink in /etc but in /usr?
675 * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word
677 * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units.
680 - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around
681 - When we update the kernel all kind of hibernation should be prohibited until shutdown/reboot
682 - logind: wakelock/opportunistic suspend support
683 - Add pretty name for seats in logind
684 - logind: allow showing logout dialog from system?
685 - session scopes/user unit: add RequiresMountsFor for the home directory of the user
686 - add Suspend() bus calls which take timestamps to fix double suspend issues when somebody hits suspend and closes laptop quickly.
687 - if pam_systemd is invoked by su from a process that is outside of a
688 any session we should probably just become a NOP, since that's
689 usually not a real user session but just some system code that just
691 - logind: make the Suspend()/Hibernate() bus calls wait for the for
692 the job to be completed. before returning, so that clients can wait
693 for "systemctl suspend" to finish to know when the suspending is
695 - logind: when the power button is pressed short, just popup a
696 logout dialog. If it is pressed for 1s, do the usual
697 shutdown. Inspiration are Macs here.
698 - expose "Locked" property on logind sesison objects
699 - maybe allow configuration of the StopTimeout for session scopes
700 - rename session scope so that it includes the UID. THat way
701 the session scope can be arranged freely in slices and we don't have
702 make assumptions about their slice anymore.
703 - follow PropertiesChanged state more closely, to deal with quick logouts and
706 * exec: when deinitializating a tty device fix the perms and group, too, not only when initializing. Set access mode/gid to 0620/tty.
709 - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields
710 - import and delete pstore filesystem content at startup
711 - journald: also get thread ID from client, plus thread name
712 - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups
713 - add API to close/reopen/get fd for journal client fd in libsystemd-journal.
714 - fallback to /dev/log based logging in libsystemd-journal, if we cannot log natively?
715 - declare the local journal protocol stable in the wiki interface chart
716 - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg
717 - journald: when dropping msgs due to ratelimit make sure to write
718 "dropped %u messages" not only when we are about to print the next
719 message that works, but alraedy after a short tiemout
720 - check if we can make journalctl by default use --follow mode inside of less if called without args?
721 - maybe add API to send pairs of iovecs via sd_journal_send
722 - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access
723 - journactl: support negative filtering, i.e. FOOBAR!="waldo",
724 and !FOOBAR for events without FOOBAR.
725 - journal: store timestamp of journal_file_set_offline() int he header,
726 so it is possible to display when the file was last synced.
727 - 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.
728 - journal: find a way to allow dropping history early, based on priority, other rules
729 - journal: When used on NFS, check payload hashes
730 - journald: add kernel cmdline option to disable ratelimiting for debug purposes
731 - refuse taking lower-case variable names in sd_journal_send() and friends.
732 - journald: we currently rotate only after MaxUse+MaxFilesize has been reached.
733 - journal: deal nicely with byte-by-byte copied files, especially regards header
734 - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit
735 - Replace utmp, wtmp, btmp, and lastlog completely with journal
736 - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax?
737 - when a kernel driver logs in a tight loop, we should ratelimit that too.
738 - journald: optionally, log debug messages to /run but everything else to /var
739 - journald: when we drop syslog messages because the syslog socket is
740 full, make sure to write how many messages are lost as first thing
741 to syslog when it works again.
742 - change systemd-journal-flush into a service that stays around during
743 boot, and causes the journal to be moved back to /run on shutdown,
744 so that we do not keep /var busy. This needs to happen synchronously,
745 hence doing this via signals is not going to work.
746 - optionally support running journald from the command line for testing purposes in external projects
747 - journald: allow per-priority and per-service retention times when rotating/vacuuming
748 - journald: make use of uid-range.h to managed uid ranges to split
750 - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so...
751 - improve journalctl performance by loading journal files
752 lazily. Encode just enough information in the file name, so that we
753 do not have to open it to know that it is not interesting for us, for
754 the most common operations.
755 - man: document that corrupted journal files is nothing to act on
756 - rework journald sigbus stuff to use mutex
757 - Set RLIMIT_NPROC for systemd-journal-xyz, and all other of our
758 services that run under their own user ids, and use User= (but only
759 in a world where userns is ubiquitous since otherwise we cannot
760 invoke those daemons on the host AND in a container anymore). Also,
761 if LimitNPROC= is used without User= we should warn and refuse
763 - journalctl --verify: don't show files that are currently being
764 written to as FAIL, but instead show that their are being written to.
765 - add journalctl -H that talks via ssh to a remote peer and passes through
767 - add a version of --merge which also merges /var/log/journal/remote
768 - journalctl: -m should access container journals directly by enumerating
769 them via machined, and also watch containers coming and going.
770 Benefit: nspawn --ephemeral would start working nicely with the journal.
771 - assign MESSAGE_ID to log messages about failed services
773 * add a test if all entries in the catalog are properly formatted.
774 (Adding dashes in a catalog entry currently results in the catalog entry
775 being silently skipped. journalctl --update-catalog must warn about this,
776 and we should also have a unit test to check that all our message are OK.)
779 - document that deps in [Unit] sections ignore Alias= fields in
780 [Install] units of other units, unless those units are disabled
781 - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets
782 - document that service reload may be implemented as service reexec
783 - add a man page containing packaging guidelines and recommending usage of things like Documentation=, PrivateTmp=, PrivateNetwork= and ReadOnlyDirectories=/etc /usr.
784 - document systemd-journal-flush.service properly
785 - documentation: recommend to connect the timer units of a service to the service via Also= in [Install]
786 - man: document the very specific env the shutdown drop-in tools live in
787 - man: add more examples to man pages
788 - man: maybe sort directives in man pages, and take sections from --help and apply them to man too
791 - add systemctl switch to dump transaction without executing it
792 - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done
793 - "systemctl disable" on a static unit prints no message and does
794 nothing. "systemctl enable" does nothing, and gives a bad message
795 about it. Should fix both to print nice actionable messages.
796 - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service
797 - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible
798 - systemctl enable: fail if target to alias into does not exist? maybe show how many units are enabled afterwards?
799 - systemctl: "Journal has been rotated since unit was started." message is misleading
800 - systemctl status output should include list of triggering units and their status
803 - "systemctl mask" should find all names by which a unit is accessible
804 (i.e. by scanning for symlinks to it) and link them all to /dev/null
807 - timer units should get the ability to trigger when:
808 o CLOCK_REALTIME makes jumps (TFD_TIMER_CANCEL_ON_SET)
811 - Modulate timer frequency based on battery state
813 * add libsystemd-password or so to query passwords during boot using the password agent logic
815 * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed
817 * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel
819 * make repeated alt-ctrl-del presses printing a dump
821 * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ...
823 * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab is not
826 - emulate /dev/kmsg using CUSE and turn off the syslog syscall
827 with seccomp. That should provide us with a useful log buffer that
828 systemd can log to during early boot, and disconnect container logs
829 from the kernel's logs.
830 - as soon as networkd has a bus interface, hook up --network-interface=,
831 --network-bridge= with networkd, to trigger netdev creation should an
833 - a nice way to boot up without machine id set, so that it is set at boot
834 automatically for supporting --ephemeral. Maybe hash the host machine id
835 together with the machine name to generate the machine id for the container
836 - fix logic always print a final newline on output.
837 https://github.com/systemd/systemd/pull/272#issuecomment-113153176
838 - should optionally support receiving WATCHDOG=1 messages from its payload
840 - optionally automatically add FORWARD rules to iptables whenever nspawn is
841 running, remove them when shut down.
842 - maybe make copying of /etc/resolv.conf optional, and skip it if --read-only
846 - refuse mounting over a mount point
847 - automatically discover .roothash files in dissect, similarly to nspawn
850 - add an API so that libvirt-lxc can inform us about network interfaces being
851 removed or added to an existing machine
852 - "machinectl migrate" or similar to copy a container from or to a
853 difference host, via ssh
854 - introduce systemd-nspawn-ephemeral@.service, and hook it into
855 "machinectl start" with a new --ephemeral switch
856 - "machinectl status" should also show internal logs of the container in
858 - "machinectl list-images" should show os-release data, as well as
859 machine-info data (including deployment level)
860 - "machinectl history"
862 - "machinectl commit" that takes a writable snapshot of a tree, invokes a
863 shell in it, and marks it read-only after use
866 - generate a nice warning if mkfs.btrfs is missing
869 - cryptsetup-generator: allow specification of passwords in crypttab itself
870 - support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator
872 * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it
873 https://bugs.freedesktop.org/show_bug.cgi?id=54712
875 * add a dependency on standard-conf.xml and other included files to man pages
877 * MountFlags=shared acts as MountFlags=slave right now.
879 * properly handle loop back mounts via fstab, especially regards to fsck/passno
881 * initialize the hostname from the fs label of /, if /etc/hostname does not exist?
886 - add trigger --subsystem-match=usb/usb_device device
887 - reimport udev db after MOVE events for devices without dev_t
889 * There's currently no way to cancel fsck (used to be possible via C-c or c on the console)
891 * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/
894 - save coredump in Windows/Mozilla minidump format
895 - when truncating coredumps, also log the full size that the process had, and make a metadata field so we can report truncated coredumps
897 * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting)
899 * default to actual 32-bit PIDs, via /proc/sys/kernel/pid_max
901 * be able to specify a forced restart of service A where service B depends on, in case B
902 needs to be auto-respawned?
905 - apply "x" on "D" too (see patch from William Douglas)
907 - instead of ignoring unknown fields, reject them.
908 - creating new directories/subvolumes/fifos/device nodes
909 should not follow symlinks. None of the other adjustment or creation
910 calls follow symlinks.
912 * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early
914 * verify that the AF_UNIX sockets of a service in the fs still exist
915 when we start a service in order to avoid confusion when a user
916 assumes starting a service is enough to make it accessible
918 * Make it possible to set the keymap independently from the font on
919 the kernel cmdline. Right now setting one resets also the other.
921 * and a dbus call to generate target from current state
923 * write blog stories about:
924 - hwdb: what belongs into it, lsusb
925 - enabling dbus services
926 - how to make changes to sysctl and sysfs attributes
928 - how to pass throw-away units to systemd, or dynamically change properties of existing units
929 - testing with Harald's awesome test kit
931 - how to develop against journal browsing APIs
932 - the journal HTTP iface
933 - non-cgroup resource management
934 - dynamic resource management with cgroups
935 - refreshed, longer missions statement
936 - calendar time events
937 - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell
938 - how to create your own target
939 - instantiated apache, dovecot and so on
940 - hooking a script into various stages of shutdown/rearly booot
942 * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support.
944 * dot output for --test showing the 'initial transaction'
947 - When logging about multiple units (stopping BoundTo units, conflicts, etc.),
948 log both units as UNIT=, so that journalctl -u triggers on both.
949 - generate better errors when people try to set transient properties
950 that are not supported...
951 http://lists.freedesktop.org/archives/systemd-devel/2015-February/028076.html
952 - maybe introduce WantsMountsFor=? Usecase:
953 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027729.html
954 - recreate systemd's D-Bus private socket file on SIGUSR2
955 - move PAM code into its own binary
956 - when we automatically restart a service, ensure we restart its rdeps, too.
957 - hide PAM options in fragment parser when compile time disabled
958 - Support --test based on current system state
959 - 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().
960 - after deserializing sockets in socket.c we should reapply sockopts and things
961 - drop PID 1 reloading, only do reexecing (difficult: Reload()
962 currently is properly synchronous, Reexec() is weird, because we
963 cannot delay the response properly until we are back, so instead of
964 being properly synchronous we just keep open the fd and close it
965 when done. That means clients do not get a successful method reply,
966 but much rather a disconnect on success.
967 - when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr
968 - when a bus name of a service disappears from the bus make sure to queue further activation requests
971 - allow port=0 in .socket units
972 - maybe introduce ExecRestartPre=
973 - add ReloadSignal= for configuring a reload signal to use
974 - implement Register= switch in .socket units to enable registration
975 in Avahi, RPC and other socket registration services.
976 - allow Type=simple with PIDFile=
977 https://bugzilla.redhat.com/show_bug.cgi?id=723942
978 - allow writing multiple conditions in unit files on one line
979 - load-fragment: when loading a unit file via a chain of symlinks
980 verify that it is not masked via any of the names traversed.
981 - introduce Type=pid-file
982 - introduce mix of BindTo and Requisite
983 - add a concept of RemainAfterExit= to scope units
984 - Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely
987 - Make sure ID_PATH is always exported and complete for
988 network devices where possible, so we can safely rely
990 - check MTUBytes parsing (expecting size_t but we are using unsigned)
993 - add support for more attribute types
994 - inbuilt piping support (essentially degenerate async)? see loopback-setup.c and other places
997 - add more keys to [Route] and [Address] sections
998 - add support for more DHCPv4 options (and, longer term, other kinds of dynamic config)
999 - add proper initrd support (in particular generate .network/.link files based on /proc/cmdline)
1000 - add reduced [Link] support to .network files
1001 - add Scope= parsing option for [Network]
1002 - properly handle routerless dhcp leases
1003 - work with non-Ethernet devices
1004 - add support for more bond options
1005 - dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from?
1006 - the DHCP lease data (such as NTP/DNS) is still made available when
1007 a carrier is lost on a link. It should be removed instantly.
1008 - expose in the API the following bits:
1009 - option 15, domain name and/or option 119, search list
1010 - option 12, host name and/or option 81, fqdn
1011 - option 123, 144, geolocation
1012 - option 252, configure http proxy (PAC/wpad)
1013 - provide a way to define a per-network interface default metric value
1014 for all routes to it. possibly a second default for DHCP routes.
1015 - allow Name= to be specified repeatedly in the [Match] section. Maybe also
1016 support Name=foo*|bar*|baz ?
1017 - duplicate address check for static IPs (like ARPCHECK in network-scripts)
1018 - allow DUID/IAID to be customized, see issue #394.
1019 - whenever uplink info changes, make DHCP server send out FORCERENEW
1021 * networkd-wait-online:
1022 - make operstates to wait for configurable?
1025 - figure out how much we can increase Maximum Message Size
1028 - add functions to set previously stored IPv6 addresses on startup and get
1029 them at shutdown; store them in client->ia_na
1030 - write more test cases
1031 - implement reconfigure support, see 5.3., 15.11. and 22.20.
1032 - implement support for temporary adressess (IA_TA)
1033 - implement dhcpv6 authentication
1034 - investigate the usefulness of Confirm messages; i.e. are there any
1035 situations where the link changes without any loss in carrier detection
1037 - some servers don't do rapid commit without a filled in IA_NA, verify
1044 - natively watch for dbus-*.service symlinks (PENDING)
1045 - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service
1047 * fix alsa mixer restore to not print error when no config is stored
1049 * make cryptsetup lower --iter-time
1051 * patch kernel for xattr support in /dev, /proc/, /sys?
1053 * kernel: add device_type = "fb", "fbcon" to class "graphics"
1055 * drop accountsservice's StandardOutput=syslog and Type=dbus fields
1057 * /usr/bin/service should actually show the new command line
1059 * 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=
1061 * fedora: F20: go timer units all the way, leave cron.daily for cron
1063 * neither pkexec nor sudo initialize environ[] from the PAM environment?
1065 * fedora: update policy to declare access mode and ownership of unit files to root:root 0644, and add an rpmlint check for it
1067 * register catalog database signature as file magic
1069 * zsh shell completion:
1070 - <command> <verb> -<TAB> should complete options, but currently does not
1071 - systemctl add-wants,add-requires
1076 * look for close() vs. close_nointr() vs. close_nointr_nofail()
1078 * check for strerror(r) instead of strerror(-r)
1082 * set_put(), hashmap_put() return values check. i.e. == 0 does not free()!
1084 * use secure_getenv() instead of getenv() where appropriate
1086 * link up selected blog stories from man pages and unit files Documentation= fields