3 * copy.c: set the right chattrs before copying files and others after
5 * Many manager configuration settings that are only applicable to user
6 manager or system manager can be always set. It would be better to reject
7 them when parsing config.
9 * Clarify what IPAddress* matches (source, destination, both?)
13 * Fedora: add an rpmlint check that verifies that all unit files in the RPM are listed in %systemd_post macros.
15 * wiki: update journal format documentation for lz4 additions
19 * Rearrange tests so that the various test-xyz.c match a specific src/basic/xyz.c again
21 * rework mount.c and swap.c to follow proper state enumeration/deserialization
22 semantics, like we do for device.c now
26 * make MAINPID= message reception checks even stricter: if service uses User=,
27 then check sending UID and ignore message if it doesn't match the user or
30 * maybe trigger a uevent "change" on a device if "systemctl reload xyz.device"
33 * when importing an fs tree with machined, optionally apply userns-rec-chown
35 * when importing an fs tree with machined, complain if image is not an OS
37 * Maybe introduce a helper safe_exec() or so, which is to execve() which
38 safe_fork() is to fork(). And then make revert the RLIMIT_NOFILE soft limit
39 to 1K implicitly, unless explicitly opted-out.
41 * rework seccomp/nnp logic that that even if User= is used in combination with
42 a seccomp option we don't have to set NNP. For that, change uid first whil
43 keeping CAP_SYS_ADMIN, then apply seccomp, the drop cap.
45 * add a concept for automatically loading per-unit secrets off disk and
46 inserting them into the kernel keyring. Maybe SecretsDirectory= similar to
47 ConfigurationDirectory=.
49 * systemd-gpt-auto: if we find the root dir mounted read-only and the gpt flag
50 doesn't say so generate job that remounts it writable
52 * when no locale is configured, default to UEFI's PlatformLang variable
54 * When logind.conf contains HandleLidSwitch=suspend-then-hibernate and we can't
55 hibernate because the swap partition isn't large enough, still suspend
57 * bootctl: implement Type #2 boot loader entry discovery
59 * bootctl,sd-boot: actually honour the "architecture" key
61 * when a socket unit is spawned with an AF_UNIX path in /var/run, complain and
62 patch it to use /run instead
64 * set memory.oom.group in cgroupsv2 for all leaf cgroups (kernel v4.19+)
66 * add a new syscall group "@esoteric" for more esoteric stuff such as bpf() and
67 usefaultd() and make systemd-analyze check for it.
69 * paranoia: whenever we process passwords, call mlock() on the memory
70 first. i.e. look for all places we use string_erase()/string_free_erase() and
71 augment them with mlock()
73 * whenever oom_kill memory.event event is triggered print a nice log message
75 * Move RestrictAddressFamily= to the new cgroup create socket
77 * support the bind/connect/sendmsg cgroup stuff for sandboxing, and possibly
80 * maybe implicitly attach monotonic+realtime timestamps to outgoing messages in
81 log.c and sd-journal-send
83 * chown() tty a service is attached to after the service goes down
85 * optionally: turn on cgroup delegation for per-session scope units
87 * introduce per-unit (i.e. per-slice, per-service) journal log size limits.
89 * optionally, if a per-partition GPT flag is set for the root/home/… partitions
90 format the partition on next boot and unset the flag, in order to implement
91 factory reset. also, add a second flag that simply indicates whether such a
92 scheme is supported. then, add a tool (or maybe beef up systemd-dissect) to
93 show state of these flags, and optionally trigger such a factory reset on
94 next boot by setting the flag.
96 * sd-boot: search drop-ins in $BOOT, too
98 * sd-boot: add "oneshot boot timeout" variable support
100 * sd-boot: automatically load EFI modules from some drop-in dir, so that people
101 can add in file system drivers and such
103 * esp generator: also mount $BOOT if found
105 * sd-boot: optionally, show boot menu when previous default boot item has
106 non-zero "tries done" count
108 * logind: add "boot into bootmenu" API, and possibly even "boot into windows"
109 and "boot into macos".
111 * bootspec.c: also enumerate EFI unified kernel images.
113 * maybe set a special xattr on cgroups that have delegate=yes set, to make it
114 easy to mark cut points
116 * introduce an option (or replacement) for "systemctl show" that outputs all
117 properties as JSON, similar to busctl's new JSON output. In contrast to that
118 it should skip the variant type string though.
120 * augment CODE_FILE=, CODE_LINE= with something like CODE_BASE= or so which
121 contains some identifier for the project, which allows us to include
122 clickable links to source files generating these log messages. The identifier
123 could be some abberviated URL prefix or so (taking inspiration from Go
124 imports). For example, for systemd we could use
125 CODE_BASE=github.com/systemd/systemd/blob/98b0b1123cc or so which is
126 sufficient to build a link by prefixing "http://" and suffixing the
129 * when outputting log data with journalctl and the log data includes references
130 to configuration files (CONFIG_FILE=), create a clickable link for it.
132 * Augment MESSAGE_ID with MESSAGE_BASE, in a similar fashion so that we can
133 make clickable links from log messages carrying a MESSAGE_ID, that lead to
134 some explanatory text online.
136 * maybe extend .path units to expose fanotify() per-mount change events
138 * Add a "systemctl list-units --by-slice" mode or so, which rearranges the
139 output of "systemctl list-units" slightly by showing the tree structure of
140 the slices, and the units attached to them.
142 * the a-posteriori stopping of units bound to units that disappeared logic
143 should be reworked: there should be a queue of units, and we should only
144 enqeue stop jobs from a defer event that processes queue instead of
145 right-away when we find a unit that is bound to one that doesn't exist
146 anymore. (similar to how the stop-unneeded queue has been reworked the same
149 * nspawn: make nspawn suitable for shell pipelines: instead of triggering a
150 hangup when input is finished, send ^D, which synthesizes an EOF. Then wait
151 for hangup or ^D before passing on the EOF.
153 * When reloading configuration PID 1 should reset all its properties to the
154 original defaults before calling parse_config()
156 * Add OnTimezoneChange= and OnTimeChange= stanzas to .timer units in order to
157 schedule events based on time and timezone changes.
159 * nspawn: greater control over selinux label?
161 * hibernate/s2h: make this robust and safe to enable in Fedora by default.
164 1. add resume_offset support to the resume code (i.e. support swap files
166 2. check if swap is on weird storage and refuse if so
167 3. add autodetection of hibernation images
169 * cgroups: use inotify to get notified when somebody else modifies cgroups
170 owned by us, then log a friendly warning.
172 * beef up log.c with support for stripping ANSI sequences from strings, so that
173 it is OK to include them in log strings. This would be particularly useful so
174 that our log messages could contain clickable links for example for unit
175 files and suchlike we operate on.
177 * add support for "portablectl attach http://foobar.com/waaa.raw (i.e. importd integration)
179 * add attach --enable and attach --now (for attach+enable+start)
181 * sync dynamic uids/gids between host+portable srvice (i.e. if DynamicUser=1 is set for a service, make sure that the
182 selected user is resolvable in the service even if it ships its own /etc/passwd)
184 * Fix DECIMAL_STR_MAX or DECIMAL_STR_WIDTH. One includes a trailing NUL, the
185 other doesn't. What a desaster. Probably to exclude it. Also
186 DECIMAL_STR_WIDTH should probably add an extra "-" into account for negative
189 * Check that users of inotify's IN_DELETE_SELF flag are using it properly, as
190 usually IN_ATTRIB is the right way to watch deleted files, as the former only
191 fires when a file is actually removed from disk, i.e. the link count drops to
192 zero and is not open anymore, while the latter happens when a file is
193 unlinked from any dir.
195 * port systemctl, busctl, … over to format-table.[ch]'s table formatters
197 * pid1: lock image configured with RootDirectory=/RootImage= using the usual nspawn semantics while the unit is up
199 * add --vacuum-xyz options to coredumpctl, matching those journalctl already has.
201 * SuccessExitStatus= and friends should probably also accept symbolic exit
202 codes names, i.e. error codes from the list maintained in exit-codes.[ch]
204 * introduce Ephemeral= unit file switch, that creates an ephemeral copy of all
205 files and directories that are left writable for a unit, and which are
206 removed after the unit goes down again. A bit like --ephemeral for
207 systemd-nspawn but for system services. If used together with RootImage= this
208 should reflink the image file itself.
210 Related: add Ephemeral=<path1> <path2> … which would allow marking
211 specific paths only like this.
213 * add CopyFile= or so as unit file setting that may be used to copy files or
214 directory trees from the host to te services RootImage= and RootDirectory=
215 environment. Which we can use for /etc/machine-id and in particular
216 /etc/resolv.conf. Should be smart and do something useful on read-only
217 images, for example fallback to read-only bind mounting the file instead.
219 * nspawn's console TTY should be allocated from within the container, not
220 mounted in from the outside
222 * show invocation ID in systemd-run output
224 * bypass SIGTERM state in unit files if KillSignal is SIGKILL
226 * tree-wide: ensure we always block the signals we hook into with
227 sd_event_add_signal() first
229 * add proper dbus APIs for the various sd_notify() commands, such as MAINPID=1
230 and so on, which would mean we could report errors and such.
232 * teach tmpfiles.d q/Q logic something sensible in the context of XFS/ext4
235 * introduce DefaultSlice= or so in system.conf that allows changing where we
236 place our units by default, i.e. change system.slice to something
237 else. Similar, ManagerSlice= should exist so that PID1's own scope unit could
238 be moved somewhere else too. Finally machined and logind should get similar
239 options so that it is possible to move user session scopes and machines to a
240 different slice too by default. Usecase: people who want to put resources on
241 the entire system, with the exception of one specific service. See:
242 https://lists.freedesktop.org/archives/systemd-devel/2018-February/040369.html
244 * maybe rework get_user_creds() to query the user database if $SHELL is used
245 for root, but only then.
247 * be stricter with fds we receive for the fdstore: close them asynchronously
249 * calenderspec: add support for week numbers and day numbers within a
250 year. This would allow us to define "bi-weekly" triggers safely.
252 * sd-bus: add vtable flag, that may be used to request client creds implicitly
253 and asynchronously before dispatching the operation
255 * make use of ethtool veth peer info in machined, for automatically finding out
256 host-side interface pointing to the container.
258 * add some special mode to LogsDirectory=/StateDirectory=… that allows
259 declaring these directories without necessarily pulling in deps for them, or
260 creating them when starting up. That way, we could declare that
261 systemd-journald writes to /var/log/journal, which could be useful when we
262 doing disk usage calculations and so on.
264 * taint systemd if there are fewer than 65536 users assigned (userns) to the system.
266 * deprecate RootDirectoryStartOnly= in favour of a new ExecStart= prefix char
268 * add a new RuntimeDirectoryPreserve= mode that defines a similar lifecycle for
269 the runtime dir as we maintain for the fdstore: i.e. keep it around as long
270 as the unit is running or has a job queued.
272 * support projid-based quota in machinectl for containers
274 * Add NetworkNamespacePath= to specify a path to a network namespace
276 * maybe use SOURCE_DATE_EPOCH (i.e. the env var the reproducible builds folks
277 introduced) as the RTC epoch, instead of the mtime of NEWS.
279 * add a way to lock down cgroup migration: a boolean, which when set for a unit
280 makes sure the processes in it can never migrate out of it
282 * blog about fd store and restartable services
284 * document Environment=SYSTEMD_LOG_LEVEL=debug drop-in in debugging document
286 * rework ExecOutput and ExecInput enums so that EXEC_OUTPUT_NULL loses its
287 magic meaning and is no longer upgraded to something else if set explicitly.
289 * in the long run: permit a system with /etc/machine-id linked to /dev/null, to
290 make it lose its identity, i.e. be anonymous. For this we'd have to patch
291 through the whole tree to make all code deal with the case where no machine
294 * optionally, collect cgroup resource data, and store it in per-unit RRD files,
295 suitable for processing with rrdtool. Add bus API to access this data, and
296 possibly implement a CPULoad property based on it.
298 * beef up pam_systemd to take unit file settings such as cgroups properties as
301 * maybe hook of xfs/ext4 quotactl() with services? i.e. automatically manage
302 the quota of a the user indicated in User= via unit file settings, like the
303 other resource management concepts. Would mix nicely with DynamicUser=1. Or
304 alternatively, do this with projids, so that we can also cover services
305 running as root. Quota should probably cover all the special dirs such as
306 StateDirectory=, LogsDirectory=, CacheDirectory=, as well as RootDirectory= if it
307 is set, plus the whole disk space any image configured with RootImage=.
309 * In DynamicUser= mode: before selecting a UID, use disk quota APIs on relevant
310 disks to see if the UID is already in use.
312 * add "systemctl wait" or so, which does what "systemd-run --wait" does, but
313 for all units. It should be both a way to pin units into memory as well as a
314 wait to retrieve their exit data.
316 * expose IO accounting data on the bus, show it in systemd-run --wait and log
317 about it in the resource log message
319 * add "systemctl purge" for flushing out configuration, state, logs, ... of a
320 unit when it is stopped
322 * show whether a service has out-of-date configuration in "systemctl status" by
323 using mtime data of ConfigurationDirectory=.
325 * replace all remaining uses of fgets() + LINE_MAX by read_line()
327 * Add AddUser= setting to unit files, similar to DynamicUser=1 which however
328 creates a static, persistent user rather than a dynamic, transient user. We
329 can leverage code from sysusers.d for this.
331 * add some optional flag to ReadWritePaths= and friends, that has the effect
332 that we create the dir in question when the service is started. Example:
334 ReadWritePaths=:/var/lib/foobar
336 * maybe add call sd_journal_set_block_timeout() or so to set SO_SNDTIMEO for
337 the sd-journal logging socket, and, if the timeout is set to 0, sets
338 O_NONBLOCK on it. That way people can control if and when to block for
341 * hostnamed: populate form factor data from a new hwdb database, so that old
342 yogas can be recognized as "convertible" too, even if they predate the DMI
343 "convertible" form factor
345 * Maybe add a small tool invoked early at boot, that adds in or resizes
346 partitions automatically, to be used when the media used is actually larger
347 than the image written onto it is.
349 * Maybe add PrivatePIDs= as new unit setting, and do minimal PID namespacing
350 after all. Be strict however, only support the equivalent of nspawn's
351 --as-pid2 switch, and sanely proxy sd_notify() messages dropping stuff such
354 * Add ExecMonitor= setting. May be used multiple times. Forks off a process in
355 the service cgroup, which is supposed to monitor the service, and when it
356 exits the service is considered failed by its monitor.
358 * track the per-service PAM process properly (i.e. as an additional control
359 process), so that it may be queried on the bus and everything.
361 * add a new "debug" job mode, that is propagated to unit_start() and for
362 services results in two things: we raise SIGSTOP right before invoking
363 execve() and turn off watchdog support. Then, use that to implement
364 "systemd-gdb" for attaching to the start-up of any system service in its
367 * maybe introduce gpt auto discovery for /var/tmp?
369 * maybe add gpt-partition-based user management: each user gets his own
370 LUKS-encrypted GPT partition with a new GPT type. A small nss module
371 enumerates users via udev partition enumeration. UIDs are assigned in a fixed
372 way: the partition index is added as offset to some fixed base uid. User name
373 is stored in GPT partition name. A PAM module authenticates the user via the
374 LUKS partition password. Benefits: strong per-user security, compatibility
375 with stateless/read-only/verity-enabled root. (other idea: do this based on
376 loopback files in /home, without GPT involvement)
378 * gpt-auto logic: introduce support for discovering /var matching an image. For
379 that, use a partition type UUID that is hashed from the OS name (as encoded
380 in /etc/os-release), the architecture, and 4 new bits from the gpt flags
381 field of the root partition. This way can easily support multiple OS
382 installations on the same GPT partition table, without problems with
383 unmatched /var partitions.
385 * gpt-auto logic: related to the above, maybe support a "secondary" root
386 partition, that is mounted to / and is writable, and where the actual root's
387 /usr is mounted into.
389 * 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
391 * drop nss-myhostname in favour of nss-resolve?
393 * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and
394 then use that for the setting used in user@.service. It should be understood
395 relative to the configured default value.
397 * in networkd, when matching device types, fix up DEVTYPE rubbish the kernel passes to us
399 * enable LockMLOCK to take a percentage value relative to physical memory
401 * Permit masking specific netlink APIs with RestrictAddressFamily=
403 * nspawn: support that /proc, /sys/, /dev are pre-mounted
405 * define gpt header bits to select volatility mode
407 * ProtectKernelLogs= (drops CAP_SYSLOG, add seccomp for syslog() syscall, and DeviceAllow to /dev/kmsg) in service files
409 * ProtectClock= (drops CAP_SYS_TIMES, adds seecomp filters for settimeofday, adjtimex), sets DeviceAllow o /dev/rtc
411 * ProtectTracing= (drops CAP_SYS_PTRACE, blocks ptrace syscall, makes /sys/kernel/tracing go away)
413 * ProtectMount= (drop mount/umount/pivot_root from seccomp, disallow fuse via DeviceAllow, imply Mountflags=slave)
415 * ProtectKeyRing= to take keyring calls away
417 * RemoveKeyRing= to remove all keyring entries of the specified user
419 * ProtectReboot= that masks reboot() and kexec_load() syscalls, prohibits kill
420 on PID 1 with the relevant signals, and makes relevant files in /sys and
421 /proc (such as the sysrq stuff) unavailable
423 * make sure the ratelimit object can deal with USEC_INFINITY as way to turn off things
425 * journalctl: make sure -f ends when the container indicated by -M terminates
427 * mount: automatically search for "main" partition of an image has multiple
430 * expose the "privileged" flag of ExecCommand on the bus, and open it up to
433 * in nss-systemd, if we run inside of RootDirectory= with PrivateUsers= set,
434 find a way to map the User=/Group= of the service to the right name. This way
435 a user/group for a service only has to exist on the host for the right
438 * add bus API for creating unit files in /etc, reusing the code for transient units
440 * add bus API to remove unit files from /etc
442 * add bus API to retrieve current unit file contents (i.e. implement "systemctl cat" on the bus only)
444 * rework fopen_temporary() to make use of open_tmpfile_linkable() (problem: the
445 kernel doesn't support linkat() that replaces existing files, currently)
447 * transient units: don't bother with actually setting unit properties, we
448 reload the unit file anyway
450 * journald: sigbus API via a signal-handler safe function that people may call
451 from the SIGBUS handler
453 * optionally, also require WATCHDOG=1 notifications during service start-up and shutdown
455 * resolved: when routing queries, make sure only look for the *longest* suffix...
457 * delay activation of logind until somebody logs in, or when /dev/tty0 pulls it
458 in or lingering is on (so that containers don't bother with it until PAM is used). also exit-on-idle
460 * cache sd_event_now() result from before the first iteration...
462 * add systemctl stop --job-mode=triggering that follows TRIGGERED_BY deps and adds them to the same transaction
464 * PID1: find a way how we can reload unit file configuration for
465 specific units only, without reloading the whole of systemd
467 * add an explicit parser for LimitRTPRIO= that verifies
468 the specified range and generates sane error messages for incorrect
471 * when we detect that there are waiting jobs but no running jobs, do something
473 * push CPUAffinity= also into the "cpuset" cgroup controller (only after the cpuset controller got ported to the unified hierarchy)
475 * PID 1 should send out sd_notify("WATCHDOG=1") messages (for usage in the --user mode, and when run via nspawn)
477 * there's probably something wrong with having user mounts below /sys,
478 as we have for debugfs. for exmaple, src/core/mount.c handles mounts
479 prefixed with /sys generally special.
480 http://lists.freedesktop.org/archives/systemd-devel/2015-June/032962.html
482 * fstab-generator: default to tmpfs-as-root if only usr= is specified on the kernel cmdline
484 * docs: bring http://www.freedesktop.org/wiki/Software/systemd/MyServiceCantGetRealtime up to date
486 * add a job mode that will fail if a transaction would mean stopping
487 running units. Use this in timedated to manage the NTP service
489 http://lists.freedesktop.org/archives/systemd-devel/2015-April/030229.html
491 * The udev blkid built-in should expose a property that reflects
492 whether media was sensed in USB CF/SD card readers. This should then
493 be used to control SYSTEMD_READY=1/0 so that USB card readers aren't
494 picked up by systemd unless they contain a medium. This would mirror
495 the behaviour we already have for CD drives.
497 * networkd/udev: implement SR_IOV configuration in .link files:
498 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027451.html
500 * hostnamectl: show root image uuid
502 * Find a solution for SMACK capabilities stuff:
503 http://lists.freedesktop.org/archives/systemd-devel/2014-December/026188.html
505 * "systemctl preset-all" should probably order the unit files it
506 operates on lexicographically before starting to work, in order to
507 ensure deterministic behaviour if two unit files conflict (like DMs
510 * synchronize console access with BSD locks:
511 http://lists.freedesktop.org/archives/systemd-devel/2014-October/024582.html
513 * as soon as we have sender timestamps, revisit coalescing multiple parallel daemon reloads:
514 http://lists.freedesktop.org/archives/systemd-devel/2014-December/025862.html
516 * in systemctl list-unit-files: show the install value the presets would suggest for a service in a third column
518 * figure out when we can use the coarse timers
520 * add "systemctl start -v foobar.service" that shows logs of a service
521 while the start command runs. This is non-trivial to do without
522 races though, since we should flush out all journal messages before
523 returning from the "systemctl stop".
525 * 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
527 * maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment=
529 * consider showing the unit names during boot up in the status output, not just the unit descriptions
531 * maybe allow timer units with an empty Units= setting, so that they
532 can be used for resuming the system but nothing else.
534 * what to do about udev db binary stability for apps? (raw access is not an option)
536 * man: maybe use the word "inspect" rather than "introspect"?
538 * systemctl: if some operation fails, show log output?
540 * systemctl edit: use equvalent of cat() to insert existing config as a comment, prepended with #.
541 Upon editor exit, lines with one # are removed, lines with two # are left with one #, etc.
543 * exponential backoff in timesyncd when we cannot reach a server
545 * timesyncd: add ugly bus calls to set NTP servers per-interface, for usage by NM
547 * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share....
549 * systemd.show_status= should probably have a mode where only failed
552 * add systemd.abort_on_kill or some other such flag to send SIGABRT instead of SIGKILL
553 (throughout the codebase, not only PID1)
557 - service registration
558 - service/domain/types browsing
560 - DNS-SD service registration from socket units
561 - resolved should optionally register additional per-interface LLMNR
562 names, so that for the container case we can establish the same name
563 (maybe "host") for referencing the server, everywhere.
564 - allow clients to request DNSSEC for a single lookup even if DNSSEC is off (?)
565 - hook up resolved with machined-based address resolution
567 * refcounting in sd-resolve is borked
569 * Add a new verb "systemctl top"
571 * add new gpt type for btrfs volumes
573 * support empty /etc boots nicely:
574 - nspawn/gpt-generator: introduce new gpt partition type for /usr
576 * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them.
578 * a way for container managers to turn off getty starting via $container_headless= or so...
580 * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit
582 * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services
583 they run added to the initial transaction and thus confuse Type=idle.
585 * add bus api to query unit file's X fields.
587 * gpt-auto-generator:
588 - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap?
589 - Make /home automount rather than mount?
591 * add generator that pulls in systemd-network from containers when
592 CAP_NET_ADMIN is set, more than the loopback device is defined, even
593 when it is otherwise off
595 * MessageQueueMessageSize= (and suchlike) should use parse_iec_size().
597 * implement Distribute= in socket units to allow running multiple
598 service instances processing the listening socket, and open this up
601 * socket units: support creating sockets in different namespace,
602 opening it up for JoinsNamespaceOf=. This would require to fork off
603 a tiny process that joins the namespace and creates/binds the socket
604 and passes this back to PID1 via SCM_RIGHTS. This also could be used
605 to allow Chown/chgrp on sockets without requiring NSS in PID 1.
607 * introduce bus call FreezeUnit(s, b), as well as "systemctl freeze
608 $UNIT" and "systemctl thaw $UNIT" as wrappers around this. The calls
609 should SIGSTOP all unit processes in a loop until all processes of
610 it are fully stopped. This can later be used for app management by
611 desktop UIs such as gnome-shell to freeze apps that are not visible
612 on screen, not unlike how job control works on the shell
615 - implement per-slice CPUFairScheduling=1 switch
616 - introduce high-level settings for RT budget, swappiness
617 - how to reset dynamically changed unit cgroup attributes sanely?
618 - when reloading configuration, apply new cgroup configuration
619 - when recursively showing the cgroup hierarchy, optionally also show
620 the hierarchies of child processes
623 - add field to transient units that indicate whether systemd or somebody else saves/restores its settings, for integration with libvirt
625 * Automatically configure swap partition to use for hibernation by looking for largest swap partition on the root disk?
627 * when we detect low battery and no AC on boot, show pretty splash and refuse boot
629 * libsystemd-journal, libsystemd-login, libudev: add calls to easily attach these objects to sd-event event loops
631 * be more careful what we export on the bus as (usec_t) 0 and (usec_t) -1
633 * 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
635 * After coming back from hibernation reset hibernation swap partition using the /dev/snapshot ioctl APIs
637 * If we try to find a unit via a dangling symlink, generate a clean
638 error. Currently, we just ignore it and read the unit from the search
641 * refuse boot if /usr/lib/os-release is missing or /etc/machine-id cannot be set up
643 * man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted.
645 * load .d/*.conf dropins for device units
647 * allow implementation of InaccessibleDirectories=/ plus
648 ReadOnlyDirectories=... for whitelisting files for a service.
652 - GetAllProperties() on a non-existing object does not result in a failure currently
653 - port to sd-resolve for connecting to TCP dbus servers
654 - 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
655 - see if we can drop more message validation on the sending side
656 - add API to clone sd_bus_message objects
657 - longer term: priority inheritance
659 - NameLost/NameAcquired obsolete
662 - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now
665 - allow multiple signal handlers per signal?
666 - document chaining of signal handler for SIGCHLD and child handlers
667 - define more intervals where we will shift wakeup intervals around in, 1h, 6h, 24h, ...
668 - generate a failure of a default event loop is executed out-of-thread
670 * investigate endianness issues of UUID vs. GUID
672 * dbus: when a unit failed to load (i.e. is in UNIT_ERROR state), we
673 should be able to safely try another attempt when the bus call LoadUnit() is invoked.
675 * add a pam module that passes the hdd passphrase into the PAM stack and then expires it, for usage by gdm auto-login.
677 * add a pam module that on password changes updates any LUKS slot where the password matches
680 - add unit tests for config_parse_device_allow()
682 * seems that when we follow symlinks to units we prefer the symlink
683 destination path over /etc and /usr. We should not do that. Instead
684 /etc should always override /run+/usr and also any symlink
687 * when isolating, try to figure out a way how we implicitly can order
688 all units we stop before the isolating unit...
690 * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off})
692 * BootLoaderSpec: Clarify that the kernel has to be in $BOOT. Clarify
693 that the boot loader should be installed to the ESP. Define a way
694 how an installer can figure out whether a BLS compliant boot loader
697 * think about requeuing jobs when daemon-reload is issued? usecase:
698 the initrd issues a reload after fstab from the host is accessible
699 and we might want to requeue the mounts local-fs acquired through
702 * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep()
704 * remove any syslog support from log.c — we probably cannot do this before split-off udev is gone for good
706 * shutdown logging: store to EFI var, and store to USB stick?
708 * merge unit_kill_common() and unit_kill_context()
710 * introduce ExecCondition= in services
713 - honor language efi variables for default language selection (if there are any?)
714 - honor timezone efi variables for default timezone selection (if there are any?)
715 - change bootctl to be backed by systemd-bootd to control temporary and persistent default boot goal plus efi variables
717 * maybe do not install getty@tty1.service symlink in /etc but in /usr?
719 * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word
721 * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units.
724 - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around
725 - logind: wakelock/opportunistic suspend support
726 - Add pretty name for seats in logind
727 - logind: allow showing logout dialog from system?
728 - add Suspend() bus calls which take timestamps to fix double suspend issues when somebody hits suspend and closes laptop quickly.
729 - if pam_systemd is invoked by su from a process that is outside of a
730 any session we should probably just become a NOP, since that's
731 usually not a real user session but just some system code that just
733 - logind: make the Suspend()/Hibernate() bus calls wait for the for
734 the job to be completed. before returning, so that clients can wait
735 for "systemctl suspend" to finish to know when the suspending is
737 - logind: when the power button is pressed short, just popup a
738 logout dialog. If it is pressed for 1s, do the usual
739 shutdown. Inspiration are Macs here.
740 - expose "Locked" property on logind sesison objects
741 - maybe allow configuration of the StopTimeout for session scopes
742 - rename session scope so that it includes the UID. THat way
743 the session scope can be arranged freely in slices and we don't have
744 make assumptions about their slice anymore.
745 - follow PropertiesChanged state more closely, to deal with quick logouts and
748 * exec: when deinitializating a tty device fix the perms and group, too, not only when initializing. Set access mode/gid to 0620/tty.
751 - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields
752 - import and delete pstore filesystem content at startup
753 - journald: also get thread ID from client, plus thread name
754 - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups
755 - add API to close/reopen/get fd for journal client fd in libsystemd-journal.
756 - fallback to /dev/log based logging in libsystemd-journal, if we cannot log natively?
757 - declare the local journal protocol stable in the wiki interface chart
758 - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg
759 - journald: when dropping msgs due to ratelimit make sure to write
760 "dropped %u messages" not only when we are about to print the next
761 message that works, but alraedy after a short tiemout
762 - check if we can make journalctl by default use --follow mode inside of less if called without args?
763 - maybe add API to send pairs of iovecs via sd_journal_send
764 - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access
765 - journactl: support negative filtering, i.e. FOOBAR!="waldo",
766 and !FOOBAR for events without FOOBAR.
767 - journal: store timestamp of journal_file_set_offline() int he header,
768 so it is possible to display when the file was last synced.
769 - 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.
770 - journal: find a way to allow dropping history early, based on priority, other rules
771 - journal: When used on NFS, check payload hashes
772 - journald: add kernel cmdline option to disable ratelimiting for debug purposes
773 - refuse taking lower-case variable names in sd_journal_send() and friends.
774 - journald: we currently rotate only after MaxUse+MaxFilesize has been reached.
775 - journal: deal nicely with byte-by-byte copied files, especially regards header
776 - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit
777 - Replace utmp, wtmp, btmp, and lastlog completely with journal
778 - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax?
779 - when a kernel driver logs in a tight loop, we should ratelimit that too.
780 - journald: optionally, log debug messages to /run but everything else to /var
781 - journald: when we drop syslog messages because the syslog socket is
782 full, make sure to write how many messages are lost as first thing
783 to syslog when it works again.
784 - change systemd-journal-flush into a service that stays around during
785 boot, and causes the journal to be moved back to /run on shutdown,
786 so that we do not keep /var busy. This needs to happen synchronously,
787 hence doing this via signals is not going to work.
788 - optionally support running journald from the command line for testing purposes in external projects
789 - journald: allow per-priority and per-service retention times when rotating/vacuuming
790 - journald: make use of uid-range.h to managed uid ranges to split
792 - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so...
793 - improve journalctl performance by loading journal files
794 lazily. Encode just enough information in the file name, so that we
795 do not have to open it to know that it is not interesting for us, for
796 the most common operations.
797 - man: document that corrupted journal files is nothing to act on
798 - rework journald sigbus stuff to use mutex
799 - Set RLIMIT_NPROC for systemd-journal-xyz, and all other of our
800 services that run under their own user ids, and use User= (but only
801 in a world where userns is ubiquitous since otherwise we cannot
802 invoke those daemons on the host AND in a container anymore). Also,
803 if LimitNPROC= is used without User= we should warn and refuse
805 - journalctl --verify: don't show files that are currently being
806 written to as FAIL, but instead show that their are being written to.
807 - add journalctl -H that talks via ssh to a remote peer and passes through
809 - add a version of --merge which also merges /var/log/journal/remote
810 - journalctl: -m should access container journals directly by enumerating
811 them via machined, and also watch containers coming and going.
812 Benefit: nspawn --ephemeral would start working nicely with the journal.
813 - assign MESSAGE_ID to log messages about failed services
815 * add a test if all entries in the catalog are properly formatted.
816 (Adding dashes in a catalog entry currently results in the catalog entry
817 being silently skipped. journalctl --update-catalog must warn about this,
818 and we should also have a unit test to check that all our message are OK.)
821 - document that deps in [Unit] sections ignore Alias= fields in
822 [Install] units of other units, unless those units are disabled
823 - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets
824 - document that service reload may be implemented as service reexec
825 - add a man page containing packaging guidelines and recommending usage of things like Documentation=, PrivateTmp=, PrivateNetwork= and ReadOnlyDirectories=/etc /usr.
826 - document systemd-journal-flush.service properly
827 - documentation: recommend to connect the timer units of a service to the service via Also= in [Install]
828 - man: document the very specific env the shutdown drop-in tools live in
829 - man: add more examples to man pages
830 - man: maybe sort directives in man pages, and take sections from --help and apply them to man too
833 - add systemctl switch to dump transaction without executing it
834 - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done
835 - "systemctl disable" on a static unit prints no message and does
836 nothing. "systemctl enable" does nothing, and gives a bad message
837 about it. Should fix both to print nice actionable messages.
838 - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service
839 - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible
840 - systemctl enable: fail if target to alias into does not exist? maybe show how many units are enabled afterwards?
841 - systemctl: "Journal has been rotated since unit was started." message is misleading
842 - systemctl status output should include list of triggering units and their status
845 - "systemctl mask" should find all names by which a unit is accessible
846 (i.e. by scanning for symlinks to it) and link them all to /dev/null
849 - timer units should get the ability to trigger when:
850 o CLOCK_REALTIME makes jumps (TFD_TIMER_CANCEL_ON_SET)
853 - Modulate timer frequency based on battery state
855 * add libsystemd-password or so to query passwords during boot using the password agent logic
857 * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed
859 * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel
861 * make repeated alt-ctrl-del presses printing a dump
863 * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ...
865 * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab is not
868 - emulate /dev/kmsg using CUSE and turn off the syslog syscall
869 with seccomp. That should provide us with a useful log buffer that
870 systemd can log to during early boot, and disconnect container logs
871 from the kernel's logs.
872 - as soon as networkd has a bus interface, hook up --network-interface=,
873 --network-bridge= with networkd, to trigger netdev creation should an
875 - a nice way to boot up without machine id set, so that it is set at boot
876 automatically for supporting --ephemeral. Maybe hash the host machine id
877 together with the machine name to generate the machine id for the container
878 - fix logic always print a final newline on output.
879 https://github.com/systemd/systemd/pull/272#issuecomment-113153176
880 - should optionally support receiving WATCHDOG=1 messages from its payload
882 - optionally automatically add FORWARD rules to iptables whenever nspawn is
883 running, remove them when shut down.
886 - refuse mounting over a mount point
887 - automatically discover .roothash files in dissect, similarly to nspawn
890 - add an API so that libvirt-lxc can inform us about network interfaces being
891 removed or added to an existing machine
892 - "machinectl migrate" or similar to copy a container from or to a
893 difference host, via ssh
894 - introduce systemd-nspawn-ephemeral@.service, and hook it into
895 "machinectl start" with a new --ephemeral switch
896 - "machinectl status" should also show internal logs of the container in
898 - "machinectl history"
900 - "machinectl commit" that takes a writable snapshot of a tree, invokes a
901 shell in it, and marks it read-only after use
904 - cryptsetup-generator: allow specification of passwords in crypttab itself
905 - support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator
907 * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it
908 https://bugs.freedesktop.org/show_bug.cgi?id=54712
910 * add a dependency on standard-conf.xml and other included files to man pages
912 * MountFlags=shared acts as MountFlags=slave right now.
914 * properly handle loop back mounts via fstab, especially regards to fsck/passno
916 * initialize the hostname from the fs label of /, if /etc/hostname does not exist?
921 - add trigger --subsystem-match=usb/usb_device device
922 - reimport udev db after MOVE events for devices without dev_t
924 * There's currently no way to cancel fsck (used to be possible via C-c or c on the console)
926 * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/
929 - save coredump in Windows/Mozilla minidump format
930 - when truncating coredumps, also log the full size that the process had, and make a metadata field so we can report truncated coredumps
932 * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting)
934 * default to actual 32-bit PIDs, via /proc/sys/kernel/pid_max
936 * be able to specify a forced restart of service A where service B depends on, in case B
937 needs to be auto-respawned?
940 - apply "x" on "D" too (see patch from William Douglas)
942 - instead of ignoring unknown fields, reject them.
943 - creating new directories/subvolumes/fifos/device nodes
944 should not follow symlinks. None of the other adjustment or creation
945 calls follow symlinks.
947 * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early
949 * verify that the AF_UNIX sockets of a service in the fs still exist
950 when we start a service in order to avoid confusion when a user
951 assumes starting a service is enough to make it accessible
953 * Make it possible to set the keymap independently from the font on
954 the kernel cmdline. Right now setting one resets also the other.
956 * and a dbus call to generate target from current state
958 * write blog stories about:
959 - hwdb: what belongs into it, lsusb
960 - enabling dbus services
961 - how to make changes to sysctl and sysfs attributes
963 - how to pass throw-away units to systemd, or dynamically change properties of existing units
964 - testing with Harald's awesome test kit
966 - how to develop against journal browsing APIs
967 - the journal HTTP iface
968 - non-cgroup resource management
969 - dynamic resource management with cgroups
970 - refreshed, longer missions statement
971 - calendar time events
972 - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell
973 - how to create your own target
974 - instantiated apache, dovecot and so on
975 - hooking a script into various stages of shutdown/rearly booot
977 * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support.
979 * dot output for --test showing the 'initial transaction'
982 - When logging about multiple units (stopping BoundTo units, conflicts, etc.),
983 log both units as UNIT=, so that journalctl -u triggers on both.
984 - generate better errors when people try to set transient properties
985 that are not supported...
986 http://lists.freedesktop.org/archives/systemd-devel/2015-February/028076.html
987 - maybe introduce WantsMountsFor=? Usecase:
988 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027729.html
989 - recreate systemd's D-Bus private socket file on SIGUSR2
990 - move PAM code into its own binary
991 - when we automatically restart a service, ensure we restart its rdeps, too.
992 - hide PAM options in fragment parser when compile time disabled
993 - Support --test based on current system state
994 - 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().
995 - after deserializing sockets in socket.c we should reapply sockopts and things
996 - drop PID 1 reloading, only do reexecing (difficult: Reload()
997 currently is properly synchronous, Reexec() is weird, because we
998 cannot delay the response properly until we are back, so instead of
999 being properly synchronous we just keep open the fd and close it
1000 when done. That means clients do not get a successful method reply,
1001 but much rather a disconnect on success.
1002 - when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr
1003 - when a bus name of a service disappears from the bus make sure to queue further activation requests
1006 - allow port=0 in .socket units
1007 - maybe introduce ExecRestartPre=
1008 - add ReloadSignal= for configuring a reload signal to use
1009 - implement Register= switch in .socket units to enable registration
1010 in Avahi, RPC and other socket registration services.
1011 - allow Type=simple with PIDFile=
1012 https://bugzilla.redhat.com/show_bug.cgi?id=723942
1013 - allow writing multiple conditions in unit files on one line
1014 - load-fragment: when loading a unit file via a chain of symlinks
1015 verify that it is not masked via any of the names traversed.
1016 - introduce Type=pid-file
1017 - introduce mix of BindTo and Requisite
1018 - add a concept of RemainAfterExit= to scope units
1019 - Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely
1022 - Make sure ID_PATH is always exported and complete for
1023 network devices where possible, so we can safely rely
1025 - check MTUBytes parsing (expecting size_t but we are using unsigned)
1028 - add support for more attribute types
1029 - inbuilt piping support (essentially degenerate async)? see loopback-setup.c and other places
1032 - add more keys to [Route] and [Address] sections
1033 - add support for more DHCPv4 options (and, longer term, other kinds of dynamic config)
1034 - add proper initrd support (in particular generate .network/.link files based on /proc/cmdline)
1035 - add reduced [Link] support to .network files
1036 - add Scope= parsing option for [Network]
1037 - properly handle routerless dhcp leases
1038 - work with non-Ethernet devices
1039 - add support for more bond options
1040 - dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from?
1041 - the DHCP lease data (such as NTP/DNS) is still made available when
1042 a carrier is lost on a link. It should be removed instantly.
1043 - expose in the API the following bits:
1044 - option 15, domain name and/or option 119, search list
1045 - option 12, host name and/or option 81, fqdn
1046 - option 123, 144, geolocation
1047 - option 252, configure http proxy (PAC/wpad)
1048 - provide a way to define a per-network interface default metric value
1049 for all routes to it. possibly a second default for DHCP routes.
1050 - allow Name= to be specified repeatedly in the [Match] section. Maybe also
1051 support Name=foo*|bar*|baz ?
1052 - duplicate address check for static IPs (like ARPCHECK in network-scripts)
1053 - allow DUID/IAID to be customized, see issue #394.
1054 - whenever uplink info changes, make DHCP server send out FORCERENEW
1056 * networkd-wait-online:
1057 - make operstates to wait for configurable?
1060 - figure out how much we can increase Maximum Message Size
1063 - add functions to set previously stored IPv6 addresses on startup and get
1064 them at shutdown; store them in client->ia_na
1065 - write more test cases
1066 - implement reconfigure support, see 5.3., 15.11. and 22.20.
1067 - implement support for temporary adressess (IA_TA)
1068 - implement dhcpv6 authentication
1069 - investigate the usefulness of Confirm messages; i.e. are there any
1070 situations where the link changes without any loss in carrier detection
1072 - some servers don't do rapid commit without a filled in IA_NA, verify
1079 - natively watch for dbus-*.service symlinks (PENDING)
1080 - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service
1082 * fix alsa mixer restore to not print error when no config is stored
1084 * make cryptsetup lower --iter-time
1086 * patch kernel for xattr support in /dev, /proc/, /sys?
1088 * kernel: add device_type = "fb", "fbcon" to class "graphics"
1090 * /usr/bin/service should actually show the new command line
1092 * 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=
1094 * fedora: F20: go timer units all the way, leave cron.daily for cron
1096 * neither pkexec nor sudo initialize environ[] from the PAM environment?
1098 * fedora: update policy to declare access mode and ownership of unit files to root:root 0644, and add an rpmlint check for it
1100 * register catalog database signature as file magic
1102 * zsh shell completion:
1103 - <command> <verb> -<TAB> should complete options, but currently does not
1104 - systemctl add-wants,add-requires
1109 * look for close() vs. close_nointr() vs. close_nointr_nofail()
1111 * check for strerror(r) instead of strerror(-r)
1115 * set_put(), hashmap_put() return values check. i.e. == 0 does not free()!
1117 * use secure_getenv() instead of getenv() where appropriate
1119 * link up selected blog stories from man pages and unit files Documentation= fields