3 * Should systemctl status \* work on all unit types, not just .service?
5 * Dangling symlinks of .automount unit files in .wants/ directories, set up
6 automount points even when the original .automount file did not exist
7 anymore. Only the .mount unit was still around.
9 * ExecStart with unicode characters fails in strv_split_extract:
12 Environment=ONE='one' "TWO='two two' too" THREE=
13 ExecStart=/bin/python3 -c 'import sys;print(sys.argv)' $ONE $TWO $THREE
17 * Fedora: add an rpmlint check that verifies that all unit files in the RPM are listed in %systemd_post macros.
19 * wiki: update journal format documentation for lz4 additions
23 * Rearrange tests so that the various test-xyz.c match a specific src/basic/xyz.c again
27 * cgroups: figure out if we can somehow communicate in a cleaner way whether a
28 systemd instance not running in the cgroup root shall or shall not manage the
29 attributes of its top-level cgroup. Currently it assumes it manages all, but
30 then might get EPERM due to permission porblems/userns, which is OK, but this
31 should be revisited to make clearer and also work if the payload systemd runs
32 with full privs and without userns.
34 * cgroups: use inotify to get notified when somebody else modifies cgroups
35 owned by us, then log a friendly warning.
37 * beef up log.c with support for stripping ANSI sequences from strings, so that
38 it is OK to include them in log strings. This would be particularly useful so
39 that our log messages could contain clickable links for example for unit
40 files and suchlike we operate on.
42 * Fix DECIMAL_STR_MAX or DECIMAL_STR_WIDTH. One includes a trailing NUL, the
43 other doesn't. What a desaster. Probably to exclude it. Also
44 DECIMAL_STR_WIDTH should probably add an extra "-" into account for negative
47 * port systemctl, systemd-inhibit, busctl, … over to format-table.[ch]'s table formatters
49 * pid1: lock image configured with RootDirectory=/RootImage= using the usual nspawn semantics while the unit is up
51 * add --vacuum-xyz options to coredumpctl, matching those journalctl already has.
53 * enforce that transient units are started exactly once only
55 * list the exit codes from the BSD/glibc <sysexits.h> in our own
56 exit-codes.[ch] tables.
58 * SuccessExitStatus= and friends should probably also accept symbolic exit
59 codes names, i.e. error codes from the list maintained in exit-codes.[ch]
61 * introduce Ephemeral= unit file switch, that creates an ephemeral copy of all
62 files and directories that are left writable for a unit, and which are
63 removed after the unit goes down again. A bit like --ephemeral for
64 systemd-nspawn but for system services. If used together with RootImage= this
65 should reflink the image file itself.
67 Related: add Ephemeral=<path1> <path2> … which would allow marking
68 specific paths only like this.
70 * add CopyFile= or so as unit file setting that may be used to copy files or
71 directory trees from the host to te services RootImage= and RootDirectory=
72 environment. Which we can use for /etc/machine-id and in particular
73 /etc/resolv.conf. Should be smart and do something useful on read-only
74 images, for example fallback to read-only bind mounting the file instead.
76 * nspawn's console TTY should be allocated from within the container, not
77 mounted in from the outside
79 * show invocation ID in systemd-run output
81 * bypass SIGTERM state in unit files if KillSignal is SIGKILL
83 * tree-wide: ensure we always block the signals we hook into with
84 sd_event_add_signal() first
86 * add proper dbus APIs for the various sd_notify() commands, such as MAINPID=1
87 and so on, which would mean we could report errors and such.
89 * teach tmpfiles.d q/Q logic something sensible in the context of XFS/ext4
92 * introduce DefaultSlice= or so in system.conf that allows changing where we
93 place our units by default, i.e. change system.slice to something
94 else. Similar, ManagerSlice= should exist so that PID1's own scope unit could
95 be moved somewhere else too. Finally machined and logind should get similar
96 options so that it is possible to move user session scopes and machines to a
97 different slice too by default. Usecase: people who want to put resources on
98 the entire system, with the exception of one specific service. See:
99 https://lists.freedesktop.org/archives/systemd-devel/2018-February/040369.html
101 * maybe rework get_user_creds() to query the user database if $SHELL is used
102 for root, but only then.
104 * be stricter with fds we receive for the fdstore: close them asynchronously
106 * calenderspec: add support for week numbers and day numbers within a
107 year. This would allow us to define "bi-weekly" triggers safely.
109 * add bpf-based implementation of devices cgroup controller logic for compat
110 with cgroupsv2 as supported by newest kernel
112 * introduce sd_id128_get_boot_app_specific() which is like
113 sd_id128_get_machine_app_specific(). After all on long-running systems both
114 IDs have similar properties.
116 * sd-bus: add vtable flag, that may be used to request client creds implicitly
117 and asynchronously before dispatching the operation
119 * make use of ethtool veth peer info in machined, for automatically finding out
120 host-side interface pointing to the container.
122 * add some special mode to LogsDirectory=/StateDirectory=… that allows
123 declaring these directories without necessarily pulling in deps for them, or
124 creating them when starting up. That way, we could declare that
125 systemd-journald writes to /var/log/journal, which could be useful when we
126 doing disk usage calculations and so on.
128 * taint systemd if there are fewer than 65536 users assigned (userns) to the system.
130 * deprecate PermissionsStartOnly= and RootDirectoryStartOnly= in favour of the ExecStart= prefix chars
132 * add a new RuntimeDirectoryPreserve= mode that defines a similar lifecycle for
133 the runtime dir as we maintain for the fdstore: i.e. keep it around as long
134 as the unit is running or has a job queued.
136 * support projid-based quota in machinectl for containers, and then drop
137 implicit btrfs loopback magic in machined
139 * Add NetworkNamespacePath= to specify a path to a network namespace
141 * maybe use SOURCE_DATE_EPOCH (i.e. the env var the reproducible builds folks
142 introduced) as the RTC epoch, instead of the mtime of NEWS.
144 * add a way to lock down cgroup migration: a boolean, which when set for a unit
145 makes sure the processes in it can never migrate out of it
147 * blog about fd store and restartable services
149 * document Environment=SYSTEMD_LOG_LEVEL=debug drop-in in debugging document
151 * rework ExecOutput and ExecInput enums so that EXEC_OUTPUT_NULL loses its
152 magic meaning and is no longer upgraded to something else if set explicitly.
154 * in the long run: permit a system with /etc/machine-id linked to /dev/null, to
155 make it lose its identity, i.e. be anonymous. For this we'd have to patch
156 through the whole tree to make all code deal with the case where no machine
159 * optionally, collect cgroup resource data, and store it in per-unit RRD files,
160 suitable for processing with rrdtool. Add bus API to access this data, and
161 possibly implement a CPULoad property based on it.
163 * beef up pam_systemd to take unit file settings such as cgroups properties as
166 * a new "systemd-analyze security" tool outputting a checklist of security
167 features a service does and does not implement
169 * maybe hook of xfs/ext4 quotactl() with services? i.e. automatically manage
170 the quota of a the user indicated in User= via unit file settings, like the
171 other resource management concepts. Would mix nicely with DynamicUser=1. Or
172 alternatively, do this with projids, so that we can also cover services
173 running as root. Quota should probably cover all the special dirs such as
174 StateDirectory=, LogsDirectory=, CacheDirectory=, as well as RootDirectory= if it
175 is set, plus the whole disk space any image configured with RootImage=.
177 * Introduce "exit" as an EmergencyAction value, and allow to configure a
178 per-unit success/failure exit code to configure. This would be useful for
179 running commands inside of services inside of containers, which could then
180 propagate their failure state all the way up.
182 * In DynamicUser= mode: before selecting a UID, use disk quota APIs on relevant
183 disks to see if the UID is already in use.
185 * add "systemctl wait" or so, which does what "systemd-run --wait" does, but
186 for all units. It should be both a way to pin units into memory as well as a
187 wait to retrieve their exit data.
189 * maybe set a new set of env vars for services, based on RuntimeDirectory=,
190 StateDirectory=, LogsDirectory=, CacheDirectory= and ConfigurationDirectory=
191 automatically. For example, there could be $RUNTIME_DIRECTORY,
192 $STATE_DIRECTORY, $LOGS_DIRECTORY=, $CACHE_DIRECTORY and
193 $CONFIGURATION_DIRECTORY or so. This could be useful to write services that
194 can adapt to varying directories for these purposes. Special care has to be
195 taken if multiple dirs are configured. Maybe avoid setting the env vars in
198 * expose IO accounting data on the bus, show it in systemd-run --wait and log
199 about it in the resource log message
201 * add "systemctl purge" for flushing out configuration, state, logs, ... of a
202 unit when it is stopped
204 * show whether a service has out-of-date configuration in "systemctl status" by
205 using mtime data of ConfigurationDirectory=.
207 * replace all uses of fgets() + LINE_MAX by read_line()
209 * Add AddUser= setting to unit files, similar to DynamicUser=1 which however
210 creates a static, persistent user rather than a dynamic, transient user. We
211 can leverage code from sysusers.d for this.
213 * add some optional flag to ReadWritePaths= and friends, that has the effect
214 that we create the dir in question when the service is started. Example:
216 ReadWritePaths=:/var/lib/foobar
218 * maybe add call sd_journal_set_block_timeout() or so to set SO_SNDTIMEO for
219 the sd-journal logging socket, and, if the timeout is set to 0, sets
220 O_NONBLOCK on it. That way people can control if and when to block for
223 * hostnamed: populate form factor data from a new hwdb database, so that old
224 yogas can be recognized as "convertible" too, even if they predate the DMI
225 "convertible" form factor
227 * Maybe add a small tool invoked early at boot, that adds in or resizes
228 partitions automatically, to be used when the media used is actually larger
229 than the image written onto it is.
231 * Maybe add PrivatePIDs= as new unit setting, and do minimal PID namespacing
232 after all. Be strict however, only support the equivalent of nspawn's
233 --as-pid2 switch, and sanely proxy sd_notify() messages dropping stuff such
236 * Add ExecMonitor= setting. May be used multiple times. Forks off a process in
237 the service cgroup, which is supposed to monitor the service, and when it
238 exits the service is considered failed by its monitor.
240 * track the per-service PAM process properly (i.e. as an additional control
241 process), so that it may be queried on the bus and everything.
243 * add a new "debug" job mode, that is propagated to unit_start() and for
244 services results in two things: we raise SIGSTOP right before invoking
245 execve() and turn off watchdog support. Then, use that to implement
246 "systemd-gdb" for attaching to the start-up of any system service in its
249 * maybe introduce gpt auto discovery for /var/tmp?
251 * maybe add gpt-partition-based user management: each user gets his own
252 LUKS-encrypted GPT partition with a new GPT type. A small nss module
253 enumerates users via udev partition enumeration. UIDs are assigned in a fixed
254 way: the partition index is added as offset to some fixed base uid. User name
255 is stored in GPT partition name. A PAM module authenticates the user via the
256 LUKS partition password. Benefits: strong per-user security, compatibility
257 with stateless/read-only/verity-enabled root. (other idea: do this based on
258 loopback files in /home, without GPT involvement)
260 * gpt-auto logic: introduce support for discovering /var matching an image. For
261 that, use a partition type UUID that is hashed from the OS name (as encoded
262 in /etc/os-release), the architecture, and 4 new bits from the gpt flags
263 field of the root partition. This way can easily support multiple OS
264 installations on the same GPT partition table, without problems with
265 unmatched /var partitions.
267 * gpt-auto logic: related to the above, maybe support a "secondary" root
268 partition, that is mounted to / and is writable, and where the actual root's
269 /usr is mounted into.
271 * 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
273 * drop nss-myhostname in favour of nss-resolve?
275 * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and
276 then use that for the setting used in user@.service. It should be understood
277 relative to the configured default value.
279 * in networkd, when matching device types, fix up DEVTYPE rubbish the kernel passes to us
281 * enable LockMLOCK to take a percentage value relative to physical memory
283 * Permit masking specific netlink APIs with RestrictAddressFamily=
285 * nspawn: support that /proc, /sys/, /dev are pre-mounted
287 * define gpt header bits to select volatility mode
289 * ProtectKernelLogs= (drops CAP_SYSLOG, add seccomp for syslog() syscall, and DeviceAllow to /dev/kmsg) in service files
291 * ProtectClock= (drops CAP_SYS_TIMES, adds seecomp filters for settimeofday, adjtimex), sets DeviceAllow o /dev/rtc
293 * ProtectTracing= (drops CAP_SYS_PTRACE, blocks ptrace syscall, makes /sys/kernel/tracing go away)
295 * ProtectMount= (drop mount/umount/pivot_root from seccomp, disallow fuse via DeviceAllow, imply Mountflags=slave)
297 * ProtectKeyRing= to take keyring calls away
299 * RemoveKeyRing= to remove all keyring entries of the specified user
301 * ProtectReboot= that masks reboot() and kexec_load() syscalls, prohibits kill
302 on PID 1 with the relevant signals, and makes relevant files in /sys and
303 /proc (such as the sysrq stuff) unavailable
305 * DeviceAllow= should also generate seccomp filters for mknod()
307 * make sure the ratelimit object can deal with USEC_INFINITY as way to turn off things
309 * journalctl: make sure -f ends when the container indicated by -M terminates
311 * mount: automatically search for "main" partition of an image has multiple
314 * expose the "privileged" flag of ExecCommand on the bus, and open it up to
317 * in nss-systemd, if we run inside of RootDirectory= with PrivateUsers= set,
318 find a way to map the User=/Group= of the service to the right name. This way
319 a user/group for a service only has to exist on the host for the right
322 * add bus API for creating unit files in /etc, reusing the code for transient units
324 * add bus API to remove unit files from /etc
326 * add bus API to retrieve current unit file contents (i.e. implement "systemctl cat" on the bus only)
328 * rework fopen_temporary() to make use of open_tmpfile_linkable() (problem: the
329 kernel doesn't support linkat() that replaces existing files, currently)
331 * check if DeviceAllow= should split first, resolve specifiers later
333 * transient units: don't bother with actually setting unit properties, we
334 reload the unit file anyway
336 * journald: sigbus API via a signal-handler safe function that people may call
337 from the SIGBUS handler
339 * optionally, also require WATCHDOG=1 notifications during service start-up and shutdown
341 * resolved: when routing queries, make sure only look for the *longest* suffix...
343 * delay activation of logind until somebody logs in, or when /dev/tty0 pulls it
344 in or lingering is on (so that containers don't bother with it until PAM is used). also exit-on-idle
346 * cache sd_event_now() result from before the first iteration...
348 * add systemctl stop --job-mode=triggering that follows TRIGGERED_BY deps and adds them to the same transaction
350 * PID1: find a way how we can reload unit file configuration for
351 specific units only, without reloading the whole of systemd
353 * add an explicit parser for LimitRTPRIO= that verifies
354 the specified range and generates sane error messages for incorrect
357 * when we detect that there are waiting jobs but no running jobs, do something
359 * push CPUAffinity= also into the "cpuset" cgroup controller (only after the cpuset controller got ported to the unified hierarchy)
361 * PID 1 should send out sd_notify("WATCHDOG=1") messages (for usage in the --user mode, and when run via nspawn)
363 * there's probably something wrong with having user mounts below /sys,
364 as we have for debugfs. for exmaple, src/core/mount.c handles mounts
365 prefixed with /sys generally special.
366 http://lists.freedesktop.org/archives/systemd-devel/2015-June/032962.html
368 * fstab-generator: default to tmpfs-as-root if only usr= is specified on the kernel cmdline
370 * docs: bring http://www.freedesktop.org/wiki/Software/systemd/MyServiceCantGetRealtime up to date
372 * add a job mode that will fail if a transaction would mean stopping
373 running units. Use this in timedated to manage the NTP service
375 http://lists.freedesktop.org/archives/systemd-devel/2015-April/030229.html
377 * Maybe add support for the equivalent of "ethtool advertise" to .link files?
378 http://lists.freedesktop.org/archives/systemd-devel/2015-April/030112.html
380 * The udev blkid built-in should expose a property that reflects
381 whether media was sensed in USB CF/SD card readers. This should then
382 be used to control SYSTEMD_READY=1/0 so that USB card readers aren't
383 picked up by systemd unless they contain a medium. This would mirror
384 the behaviour we already have for CD drives.
386 * networkd/udev: implement SR_IOV configuration in .link files:
387 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027451.html
389 * hostnamectl: show root image uuid
391 * sysfs set api in libudev is not const
393 * Find a solution for SMACK capabilities stuff:
394 http://lists.freedesktop.org/archives/systemd-devel/2014-December/026188.html
396 * "systemctl preset-all" should probably order the unit files it
397 operates on lexicographically before starting to work, in order to
398 ensure deterministic behaviour if two unit files conflict (like DMs
401 * synchronize console access with BSD locks:
402 http://lists.freedesktop.org/archives/systemd-devel/2014-October/024582.html
404 * as soon as we have sender timestamps, revisit coalescing multiple parallel daemon reloads:
405 http://lists.freedesktop.org/archives/systemd-devel/2014-December/025862.html
407 * in systemctl list-unit-files: show the install value the presets would suggest for a service in a third column
409 * figure out when we can use the coarse timers
411 * add "systemctl start -v foobar.service" that shows logs of a service
412 while the start command runs. This is non-trivial to do without
413 races though, since we should flush out all journal messages before
414 returning from the "systemctl stop".
416 * 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
418 * maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment=
420 * introduce systemd-timesync-wait.service or so to sync on an NTP fix?
422 * consider showing the unit names during boot up in the status output, not just the unit descriptions
424 * maybe allow timer units with an empty Units= setting, so that they
425 can be used for resuming the system but nothing else.
427 * what to do about udev db binary stability for apps? (raw access is not an option)
429 * man: maybe use the word "inspect" rather than "introspect"?
431 * systemctl: if some operation fails, show log output?
433 * systemctl edit: use equvalent of cat() to insert existing config as a comment, prepended with #.
434 Upon editor exit, lines with one # are removed, lines with two # are left with one #, etc.
436 * exponential backoff in timesyncd when we cannot reach a server
438 * timesyncd: add ugly bus calls to set NTP servers per-interface, for usage by NM
440 * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share....
442 * systemd.show_status= should probably have a mode where only failed
445 * add systemd.abort_on_kill or some other such flag to send SIGABRT instead of SIGKILL
446 (throughout the codebase, not only PID1)
450 - service registration
451 - service/domain/types browsing
453 - DNS-SD service registration from socket units
454 - resolved should optionally register additional per-interface LLMNR
455 names, so that for the container case we can establish the same name
456 (maybe "host") for referencing the server, everywhere.
457 - allow clients to request DNSSEC for a single lookup even if DNSSEC is off (?)
458 - hook up resolved with machined-based address resolution
460 * refcounting in sd-resolve is borked
462 * Add a new verb "systemctl top"
464 * add new gpt type for btrfs volumes
466 * support empty /etc boots nicely:
467 - nspawn/gpt-generator: introduce new gpt partition type for /usr
469 * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them.
471 * a way for container managers to turn off getty starting via $container_headless= or so...
473 * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit
475 * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services
476 they run added to the initial transaction and thus confuse Type=idle.
478 * add bus api to query unit file's X fields.
480 * gpt-auto-generator:
481 - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap?
482 - Make /home automount rather than mount?
484 * add generator that pulls in systemd-network from containers when
485 CAP_NET_ADMIN is set, more than the loopback device is defined, even
486 when it is otherwise off
488 * MessageQueueMessageSize= (and suchlike) should use parse_iec_size().
490 * implement Distribute= in socket units to allow running multiple
491 service instances processing the listening socket, and open this up
494 * socket units: support creating sockets in different namespace,
495 opening it up for JoinsNamespaceOf=. This would require to fork off
496 a tiny process that joins the namespace and creates/binds the socket
497 and passes this back to PID1 via SCM_RIGHTS. This also could be used
498 to allow Chown/chgrp on sockets without requiring NSS in PID 1.
500 * introduce bus call FreezeUnit(s, b), as well as "systemctl freeze
501 $UNIT" and "systemctl thaw $UNIT" as wrappers around this. The calls
502 should SIGSTOP all unit processes in a loop until all processes of
503 it are fully stopped. This can later be used for app management by
504 desktop UIs such as gnome-shell to freeze apps that are not visible
505 on screen, not unlike how job control works on the shell
508 - implement per-slice CPUFairScheduling=1 switch
509 - handle jointly mounted controllers correctly
510 - introduce high-level settings for RT budget, swappiness
511 - how to reset dynamically changed unit cgroup attributes sanely?
512 - when reloading configuration, apply new cgroup configuration
513 - when recursively showing the cgroup hierarchy, optionally also show
514 the hierarchies of child processes
517 - add field to transient units that indicate whether systemd or somebody else saves/restores its settings, for integration with libvirt
518 - ensure scope units may be started only a single time
520 * Automatically configure swap partition to use for hibernation by looking for largest swap partition on the root disk?
522 * when we detect low battery and no AC on boot, show pretty splash and refuse boot
524 * libsystemd-journal, libsystemd-login, libudev: add calls to easily attach these objects to sd-event event loops
526 * be more careful what we export on the bus as (usec_t) 0 and (usec_t) -1
528 * 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
530 * After coming back from hibernation reset hibernation swap partition using the /dev/snapshot ioctl APIs
532 * If we try to find a unit via a dangling symlink, generate a clean
533 error. Currently, we just ignore it and read the unit from the search
536 * refuse boot if /usr/lib/os-release is missing or /etc/machine-id cannot be set up
538 * man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted.
540 * load .d/*.conf dropins for device units
542 * allow implementation of InaccessibleDirectories=/ plus
543 ReadOnlyDirectories=... for whitelisting files for a service.
547 - GetAllProperties() on a non-existing object does not result in a failure currently
548 - port to sd-resolve for connecting to TCP dbus servers
549 - 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
550 - see if we can drop more message validation on the sending side
551 - add API to clone sd_bus_message objects
552 - longer term: priority inheritance
554 - NameLost/NameAcquired obsolete
557 - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now
560 - allow multiple signal handlers per signal?
561 - document chaining of signal handler for SIGCHLD and child handlers
562 - define more intervals where we will shift wakeup intervals around in, 1h, 6h, 24h, ...
563 - generate a failure of a default event loop is executed out-of-thread
564 - maybe add support for inotify events
566 * investigate endianness issues of UUID vs. GUID
568 * dbus: when a unit failed to load (i.e. is in UNIT_ERROR state), we
569 should be able to safely try another attempt when the bus call LoadUnit() is invoked.
571 * add a pam module that passes the hdd passphrase into the PAM stack and then expires it, for usage by gdm auto-login.
573 * add a pam module that on password changes updates any LUKS slot where the password matches
575 * maybe add a generator that looks for "systemd.run=" on the kernel cmdline for container usercases...
578 - add unit tests for config_parse_device_allow()
580 * seems that when we follow symlinks to units we prefer the symlink
581 destination path over /etc and /usr. We should not do that. Instead
582 /etc should always override /run+/usr and also any symlink
585 * when isolating, try to figure out a way how we implicitly can order
586 all units we stop before the isolating unit...
588 * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off})
590 * BootLoaderSpec: Clarify that the kernel has to be in $BOOT. Clarify
591 that the boot loader should be installed to the ESP. Define a way
592 how an installer can figure out whether a BLS compliant boot loader
595 * think about requeuing jobs when daemon-reload is issued? usecase:
596 the initrd issues a reload after fstab from the host is accessible
597 and we might want to requeue the mounts local-fs acquired through
600 * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep()
602 * remove any syslog support from log.c — we probably cannot do this before split-off udev is gone for good
604 * shutdown logging: store to EFI var, and store to USB stick?
606 * merge unit_kill_common() and unit_kill_context()
608 * introduce ExecCondition= in services
611 - honor language efi variables for default language selection (if there are any?)
612 - honor timezone efi variables for default timezone selection (if there are any?)
613 - change bootctl to be backed by systemd-bootd to control temporary and persistent default boot goal plus efi variables
615 * maybe do not install getty@tty1.service symlink in /etc but in /usr?
617 * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word
619 * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units.
622 - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around
623 - When we update the kernel all kind of hibernation should be prohibited until shutdown/reboot
624 - logind: wakelock/opportunistic suspend support
625 - Add pretty name for seats in logind
626 - logind: allow showing logout dialog from system?
627 - we should probably handle SIGTERM/SIGINT to not leave dot files around, just in case
628 - session scopes/user unit: add RequiresMountsFor for the home directory of the user
629 - add Suspend() bus calls which take timestamps to fix double suspend issues when somebody hits suspend and closes laptop quickly.
630 - if pam_systemd is invoked by su from a process that is outside of a
631 any session we should probably just become a NOP, since that's
632 usually not a real user session but just some system code that just
634 - logind: make the Suspend()/Hibernate() bus calls wait for the for
635 the job to be completed. before returning, so that clients can wait
636 for "systemctl suspend" to finish to know when the suspending is
638 - logind: when the power button is pressed short, just popup a
639 logout dialog. If it is pressed for 1s, do the usual
640 shutdown. Inspiration are Macs here.
641 - expose "Locked" property on logind sesison objects
642 - given that logind now lets PID 1 do all nasty work, we can
643 probably reduce the capability set it retains substantially.
644 (we need CAP_SYS_ADMIN for drmSetMaster(), so maybe not worth it)
645 - expose orientation sensors and tablet mode through logind
646 - maybe allow configuration of the StopTimeout for session scopes
647 - rename session scope so that it includes the UID. THat way
648 the session scope can be arranged freely in slices and we don't have
649 make assumptions about their slice anymore.
650 - follow PropertiesChanged state more closely, to deal with quick logouts and
653 * exec: when deinitializating a tty device fix the perms and group, too, not only when initializing. Set access mode/gid to 0620/tty.
656 - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields
657 - import and delete pstore filesystem content at startup
658 - journald: also get thread ID from client, plus thread name
659 - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups
660 - add API to close/reopen/get fd for journal client fd in libsystemd-journal.
661 - fallback to /dev/log based logging in libsystemd-journal, if we cannot log natively?
662 - declare the local journal protocol stable in the wiki interface chart
663 - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg
664 - journald: when dropping msgs due to ratelimit make sure to write
665 "dropped %u messages" not only when we are about to print the next
666 message that works, but alraedy after a short tiemout
667 - check if we can make journalctl by default use --follow mode inside of less if called without args?
668 - maybe add API to send pairs of iovecs via sd_journal_send
669 - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access
670 - journactl: support negative filtering, i.e. FOOBAR!="waldo",
671 and !FOOBAR for events without FOOBAR.
672 - journal: store timestamp of journal_file_set_offline() int he header,
673 so it is possible to display when the file was last synced.
674 - 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.
675 - journal: find a way to allow dropping history early, based on priority, other rules
676 - journal: When used on NFS, check payload hashes
677 - journald: add kernel cmdline option to disable ratelimiting for debug purposes
678 - refuse taking lower-case variable names in sd_journal_send() and friends.
679 - journald: we currently rotate only after MaxUse+MaxFilesize has been reached.
680 - journal: deal nicely with byte-by-byte copied files, especially regards header
681 - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit
682 - Replace utmp, wtmp, btmp, and lastlog completely with journal
683 - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax?
684 - when a kernel driver logs in a tight loop, we should ratelimit that too.
685 - journald: optionally, log debug messages to /run but everything else to /var
686 - journald: when we drop syslog messages because the syslog socket is
687 full, make sure to write how many messages are lost as first thing
688 to syslog when it works again.
689 - change systemd-journal-flush into a service that stays around during
690 boot, and causes the journal to be moved back to /run on shutdown,
691 so that we do not keep /var busy. This needs to happen synchronously,
692 hence doing this via signals is not going to work.
693 - optionally support running journald from the command line for testing purposes in external projects
694 - journald: allow per-priority and per-service retention times when rotating/vacuuming
695 - journald: make use of uid-range.h to managed uid ranges to split
697 - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so...
698 - improve journalctl performance by loading journal files
699 lazily. Encode just enough information in the file name, so that we
700 do not have to open it to know that it is not interesting for us, for
701 the most common operations.
702 - journal-or-kmsg is currently broken? See reverted
703 commit 4a01181e460686d8b4a543b1dfa7f77c9e3c5ab8.
704 - man: document that corrupted journal files is nothing to act on
705 - rework journald sigbus stuff to use mutex
706 - Set RLIMIT_NPROC for systemd-journal-xyz, and all other of our
707 services that run under their own user ids, and use User= (but only
708 in a world where userns is ubiquitous since otherwise we cannot
709 invoke those daemons on the host AND in a container anymore). Also,
710 if LimitNPROC= is used without User= we should warn and refuse
712 - journalctl --verify: don't show files that are currently being
713 written to as FAIL, but instead show that their are being written to.
714 - add journalctl -H that talks via ssh to a remote peer and passes through
716 - add a version of --merge which also merges /var/log/journal/remote
717 - journalctl: -m should access container journals directly by enumerating
718 them via machined, and also watch containers coming and going.
719 Benefit: nspawn --ephemeral would start working nicely with the journal.
720 - assign MESSAGE_ID to log messages about failed services
722 * add a test if all entries in the catalog are properly formatted.
723 (Adding dashes in a catalog entry currently results in the catalog entry
724 being silently skipped. journalctl --update-catalog must warn about this,
725 and we should also have a unit test to check that all our message are OK.)
728 - document that deps in [Unit] sections ignore Alias= fields in
729 [Install] units of other units, unless those units are disabled
730 - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets
731 - document that service reload may be implemented as service reexec
732 - add a man page containing packaging guidelines and recommending usage of things like Documentation=, PrivateTmp=, PrivateNetwork= and ReadOnlyDirectories=/etc /usr.
733 - document systemd-journal-flush.service properly
734 - documentation: recommend to connect the timer units of a service to the service via Also= in [Install]
735 - man: document the very specific env the shutdown drop-in tools live in
736 - man: add more examples to man pages
737 - man: maybe sort directives in man pages, and take sections from --help and apply them to man too
740 - add systemctl switch to dump transaction without executing it
741 - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done
742 - "systemctl disable" on a static unit prints no message and does
743 nothing. "systemctl enable" does nothing, and gives a bad message
744 about it. Should fix both to print nice actionable messages.
745 - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service
746 - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible
747 - systemctl enable: fail if target to alias into does not exist? maybe show how many units are enabled afterwards?
748 - systemctl: "Journal has been rotated since unit was started." message is misleading
749 - systemctl status output should include list of triggering units and their status
752 - "systemctl mask" should find all names by which a unit is accessible
753 (i.e. by scanning for symlinks to it) and link them all to /dev/null
756 - timer units should get the ability to trigger when:
757 o CLOCK_REALTIME makes jumps (TFD_TIMER_CANCEL_ON_SET)
759 - Modulate timer frequency based on battery state
761 * add libsystemd-password or so to query passwords during boot using the password agent logic
763 * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed
765 * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel
767 * make repeated alt-ctrl-del presses printing a dump
769 * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ...
771 * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab is not
774 - emulate /dev/kmsg using CUSE and turn off the syslog syscall
775 with seccomp. That should provide us with a useful log buffer that
776 systemd can log to during early boot, and disconnect container logs
777 from the kernel's logs.
778 - as soon as networkd has a bus interface, hook up --network-interface=,
779 --network-bridge= with networkd, to trigger netdev creation should an
781 - a nice way to boot up without machine id set, so that it is set at boot
782 automatically for supporting --ephemeral. Maybe hash the host machine id
783 together with the machine name to generate the machine id for the container
784 - fix logic always print a final newline on output.
785 https://github.com/systemd/systemd/pull/272#issuecomment-113153176
786 - should optionally support receiving WATCHDOG=1 messages from its payload
788 - optionally automatically add FORWARD rules to iptables whenever nspawn is
789 running, remove them when shut down.
790 - maybe make copying of /etc/resolv.conf optional, and skip it if --read-only
794 - refuse mounting over a mount point
795 - automatically discover .roothash files in dissect, similarly to nspawn
798 - add an API so that libvirt-lxc can inform us about network interfaces being
799 removed or added to an existing machine
800 - "machinectl migrate" or similar to copy a container from or to a
801 difference host, via ssh
802 - introduce systemd-nspawn-ephemeral@.service, and hook it into
803 "machinectl start" with a new --ephemeral switch
804 - "machinectl status" should also show internal logs of the container in
806 - "machinectl list-images" should show os-release data, as well as
807 machine-info data (including deployment level)
808 - "machinectl history"
810 - "machinectl commit" that takes a writable snapshot of a tree, invokes a
811 shell in it, and marks it read-only after use
814 - generate a nice warning if mkfs.btrfs is missing
817 - cryptsetup-generator: allow specification of passwords in crypttab itself
818 - support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator
820 * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it
821 https://bugs.freedesktop.org/show_bug.cgi?id=54712
823 * add a dependency on standard-conf.xml and other included files to man pages
825 * MountFlags=shared acts as MountFlags=slave right now.
827 * properly handle loop back mounts via fstab, especially regards to fsck/passno
829 * initialize the hostname from the fs label of /, if /etc/hostname does not exist?
834 - add trigger --subsystem-match=usb/usb_device device
835 - reimport udev db after MOVE events for devices without dev_t
837 * There's currently no way to cancel fsck (used to be possible via C-c or c on the console)
839 * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/
842 - save coredump in Windows/Mozilla minidump format
843 - when truncating coredumps, also log the full size that the process had, and make a metadata field so we can report truncated coredumps
845 * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting)
847 * default to actual 32-bit PIDs, via /proc/sys/kernel/pid_max
849 * be able to specify a forced restart of service A where service B depends on, in case B
850 needs to be auto-respawned?
853 - apply "x" on "D" too (see patch from William Douglas)
855 - instead of ignoring unknown fields, reject them.
856 - creating new directories/subvolumes/fifos/device nodes
857 should not follow symlinks. None of the other adjustment or creation
858 calls follow symlinks.
860 * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early
862 * verify that the AF_UNIX sockets of a service in the fs still exist
863 when we start a service in order to avoid confusion when a user
864 assumes starting a service is enough to make it accessible
866 * Make it possible to set the keymap independently from the font on
867 the kernel cmdline. Right now setting one resets also the other.
869 * and a dbus call to generate target from current state
871 * write blog stories about:
872 - hwdb: what belongs into it, lsusb
873 - enabling dbus services
874 - how to make changes to sysctl and sysfs attributes
876 - how to pass throw-away units to systemd, or dynamically change properties of existing units
877 - testing with Harald's awesome test kit
879 - how to develop against journal browsing APIs
880 - the journal HTTP iface
881 - non-cgroup resource management
882 - dynamic resource management with cgroups
883 - refreshed, longer missions statement
884 - calendar time events
885 - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell
886 - how to create your own target
887 - instantiated apache, dovecot and so on
888 - hooking a script into various stages of shutdown/rearly booot
890 * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support.
892 * dot output for --test showing the 'initial transaction'
895 - When logging about multiple units (stopping BoundTo units, conflicts, etc.),
896 log both units as UNIT=, so that journalctl -u triggers on both.
897 - generate better errors when people try to set transient properties
898 that are not supported...
899 http://lists.freedesktop.org/archives/systemd-devel/2015-February/028076.html
900 - maybe introduce WantsMountsFor=? Usecase:
901 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027729.html
902 - recreate systemd's D-Bus private socket file on SIGUSR2
903 - move PAM code into its own binary
904 - when we automatically restart a service, ensure we restart its rdeps, too.
905 - hide PAM options in fragment parser when compile time disabled
906 - Support --test based on current system state
907 - 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().
908 - after deserializing sockets in socket.c we should reapply sockopts and things
909 - drop PID 1 reloading, only do reexecing (difficult: Reload()
910 currently is properly synchronous, Reexec() is weird, because we
911 cannot delay the response properly until we are back, so instead of
912 being properly synchronous we just keep open the fd and close it
913 when done. That means clients do not get a successful method reply,
914 but much rather a disconnect on success.
915 - when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr
916 - when a bus name of a service disappears from the bus make sure to queue further activation requests
919 - allow port=0 in .socket units
920 - maybe introduce ExecRestartPre=
921 - add ReloadSignal= for configuring a reload signal to use
922 - implement Register= switch in .socket units to enable registration
923 in Avahi, RPC and other socket registration services.
924 - allow Type=simple with PIDFile=
925 https://bugzilla.redhat.com/show_bug.cgi?id=723942
926 - allow writing multiple conditions in unit files on one line
927 - load-fragment: when loading a unit file via a chain of symlinks
928 verify that it is not masked via any of the names traversed.
929 - introduce Type=pid-file
930 - introduce mix of BindTo and Requisite
931 - add a concept of RemainAfterExit= to scope units
932 - Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely
935 - Make sure ID_PATH is always exported and complete for
936 network devices where possible, so we can safely rely
938 - check MTUBytes parsing (expecting size_t but we are using unsigned)
941 - add support for more attribute types
942 - inbuilt piping support (essentially degenerate async)? see loopback-setup.c and other places
945 - add more keys to [Route] and [Address] sections
946 - add support for more DHCPv4 options (and, longer term, other kinds of dynamic config)
947 - add proper initrd support (in particular generate .network/.link files based on /proc/cmdline)
948 - add reduced [Link] support to .network files
949 - add Scope= parsing option for [Network]
950 - properly handle routerless dhcp leases
951 - work with non-Ethernet devices
952 - add support for more bond options
953 - dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from?
954 - the DHCP lease data (such as NTP/DNS) is still made available when
955 a carrier is lost on a link. It should be removed instantly.
956 - expose in the API the following bits:
957 - option 15, domain name and/or option 119, search list
958 - option 12, host name and/or option 81, fqdn
959 - option 123, 144, geolocation
960 - option 252, configure http proxy (PAC/wpad)
961 - provide a way to define a per-network interface default metric value
962 for all routes to it. possibly a second default for DHCP routes.
963 - allow Name= to be specified repeatedly in the [Match] section. Maybe also
964 support Name=foo*|bar*|baz ?
965 - duplicate address check for static IPs (like ARPCHECK in network-scripts)
966 - allow DUID/IAID to be customized, see issue #394.
967 - whenever uplink info changes, make DHCP server send out FORCERENEW
969 * networkd-wait-online:
970 - make operstates to wait for configurable?
973 - figure out how much we can increase Maximum Message Size
976 - add functions to set previously stored IPv6 addresses on startup and get
977 them at shutdown; store them in client->ia_na
978 - write more test cases
979 - implement reconfigure support, see 5.3., 15.11. and 22.20.
980 - implement support for temporary adressess (IA_TA)
981 - implement dhcpv6 authentication
982 - investigate the usefulness of Confirm messages; i.e. are there any
983 situations where the link changes without any loss in carrier detection
985 - some servers don't do rapid commit without a filled in IA_NA, verify
992 - natively watch for dbus-*.service symlinks (PENDING)
993 - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service
995 * fix alsa mixer restore to not print error when no config is stored
997 * make cryptsetup lower --iter-time
999 * patch kernel for xattr support in /dev, /proc/, /sys?
1001 * kernel: add device_type = "fb", "fbcon" to class "graphics"
1003 * drop accountsservice's StandardOutput=syslog and Type=dbus fields
1005 * /usr/bin/service should actually show the new command line
1007 * 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=
1009 * fedora: F20: go timer units all the way, leave cron.daily for cron
1011 * neither pkexec nor sudo initialize environ[] from the PAM environment?
1013 * fedora: update policy to declare access mode and ownership of unit files to root:root 0644, and add an rpmlint check for it
1015 * register catalog database signature as file magic
1017 * zsh shell completion:
1018 - <command> <verb> -<TAB> should complete options, but currently does not
1019 - systemctl add-wants,add-requires
1024 * look for close() vs. close_nointr() vs. close_nointr_nofail()
1026 * check for strerror(r) instead of strerror(-r)
1030 * set_put(), hashmap_put() return values check. i.e. == 0 does not free()!
1032 * use secure_getenv() instead of getenv() where appropriate
1034 * link up selected blog stories from man pages and unit files Documentation= fields