2 * check systemd-tmpfiles for selinux context hookup for mknod(), symlink() and similar
4 * swap units that are activated by one name but shown in the kernel under another are semi-broken
6 * make anaconda write timeout=0 for encrypted devices
8 * Dangling symlinks of .automount unit files in .wants/ directories, set up
9 automount points even when the original .automount file did not exist
10 anymore. Only the .mount unit was still around.
12 * make polkit checks async
14 * properly handle .mount unit state tracking when two mount points are stacked one on top of another on the exact same mount point.
18 * drop no longer needed modprobe.d udlfb file (it does not belong in /etc anyway)
20 * Retest multi-seat (new kms driver for displaylink)
22 * make sure pkexec works fine with pam_systemd works fine with audit=0
24 * create /var/log/journal/
26 * logind: document new PrepareForSleep(false) semantics and UnlockSessions call in wiki
28 * external: maybe it is time to patch procps so that "ps" links to
29 libsystemd-logind to print a pretty service name, seat name, session
30 name in its output. Currently it only shows cgroup membership, but
31 that's sometimes kinda hard to parse for a human.
34 - don't unconditionally filter out duplicate settings
35 - support high-level cgroup setting syntax in systemctl. Example: "systemctl set-cgroup-attr MemoryLimit 5K"
36 - support writte string mapping even for non-high-level settings
37 - add man page for systemctl commands
38 - make sure we work fine with multi-line strings
40 * kernel cmdline switch to turn off predictable network interface names
42 * journal is not closed properly at shutdown when run in a container?
44 * introduce new "journal" group in place of adm?
47 - localectl: add listing support for X11 keymaps, by parsing /usr/share/X11/xkb/rules/xorg.lst
48 - localectl: support new converted x11→console keymaps
52 * nspawn: investigate whether we can support the same as LXC's
53 lxc.network.type=phys mode, and pass through entire network
54 interfaces to the container
56 * merge service_kill() logic into unit_kill_context() and unify the code paths
58 * when auto-detecting the main pid of a service, only look for those processes with ppid == 1
60 * introduce ExecCondition= in services
62 * if we have systemd-analyze in C "systemctl dot" should move there too
65 - fsck hookup for the ESP mount is missing
66 - write man page for efi boot generator
67 - add configure switch for enabling/disabling efi stuff
68 - honor language efi variables for default language selection (if there are any?)
69 - honor timezone efi variables for default timezone selection (if there are any?)
70 - introduce bootctl (backed by systemd-bootd) to control temporary and persistent default boot goal plus efi variables
72 * polkit: add configure switch to enable/disable polkit client side code
74 * maybe do not install getty@tty1.service symlink in /etc but in /usr?
76 * re-enable "make check" for gtk-doc (broken for unknown reason)
78 * fstab: add new mount option x-systemd-after=/foobar/waldo to allow manual dependencies to other mount points
79 https://bugzilla.redhat.com/show_bug.cgi?id=812826
81 * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word
83 * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units.
86 - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around
87 - When we update the kernel all kind of hibernation should be prohibited until shutdown/reboot
88 - logind: wakelock/opportunistic suspend support
89 - Add pretty name for seats in logind
90 - logind: allow showing logout dialog from system?
91 - logind: spawn user@..service on login
92 - logind: non-local X11 server handling
93 - logind: add equivalent to sd_pid_get_owner_uid() to the D-Bus API
94 - pam_systemd: try to get old session id from cgroup, if audit sessionid cannot be determined
95 - pam: when leaving a session explicitly exclude the ReleaseSession() caller process from the killing spree
97 * exec: when deinitializating a tty device fix the perms and group, too, not only when initializing. Set access mode/gid to 0620/tty.
99 * DeviceAllow/DeviceDeny: disallow everything by default, but whitelist /dev/zero, /dev/null and friends
101 * service: watchdog logic: for testing purposes allow ping, but do not require pong
104 - journald: also get thread ID from client, plus thread name
105 - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups
106 - add API to close/reopen/get fd for journal client fd in libsystemd-journal.
107 - fallback to /dev/log based logging in libsystemd-journal, if we can't log natively?
108 - declare the local journal protocol stable in the wiki interface chart
109 - journal: reuse XZ context
110 - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg
111 - journald: when dropping msgs due to ratelimit make sure to write
112 "dropped %u messages" not only when we are about to print the next
113 message that works, but alraedy after a short tiemout
114 - check if we can make journalctl by default use --follow mode inside of less if called without args?
115 - maybe add API to send pairs of iovecs via sd_journal_send
116 - journal: when writing journal auto-rotate if time jumps backwards
117 - gatewayd: should run under its own UID
118 - journal: add a setgid "adm" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access
119 - journactl: support negative filtering, i.e. FOOBAR!="waldo",
120 and !FOOBAR for events without FOOBAR.
121 - journal: when rotating, copy over old acls/access mode
122 - journal: document why we do not give ownership to journal files to the user that created them but use FS ACLs for that
123 - journal: send out marker messages every now and then, and immediately sync with fdatasync() afterwards, in order to have hourly guaranteed syncs.
124 - journal: when we haven't written anything in a while, sync to disk and mark file as offline, in order to be more often than not in a clean state
125 - 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.
126 - journal: find a way to allow dropping history early, based on priority, other rules
127 - journal: When used on NFS, check payload hashes
128 - journal: When used on NFS make sure wake up sd_journal_wait() every 2s, to handle missing inotify
129 - document that people can use file system ACLs to manage access to journal files, with example
130 - Introduce journalctl -b <nr> to show journal messages of a previous boot
131 - journald: check whether it is OK if the client can still modify delivered journal entries
132 - journal live copy, based on libneon (client) and libmicrohttpd
133 - journald: add kernel cmdline option to disable ratelimiting for debug purposes
134 - refuse taking lower-case variable names in sd_journal_send() and friends.
135 - journald: we currently rotate only after MaxUse+MaxFilesize has been reached.
136 - journal: deal nicely with byte-by-byte copied files, especially regards header
137 - journalctl: show multiline log messages sanely, expand tabs, and show all valid utf8 messages
138 - journal: store euid in journal if it differs from uid
139 - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit
142 - document unit_name_mangle()
143 - document that deps in [Unit] sections ignore Alias= fileds in
144 [Install] units of other units, unless those units are disabled
145 - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets
146 - Document word splitting syntax for ExecStart= and friends
147 - document that units from /etc override those from /usr and /run
148 - document the exit codes when services fail before they are exec()ed
149 - document that %% can be used to write % in a string that is specifier extended
150 - document that service reload may be implemented as service reexec
153 - systemctl list-jobs - show dependencies
154 - add systemctl switch to dump transaction without executing it
155 - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done
156 - "systemctl disable" on a static unit prints no message and does
157 nothing. "systemctl enable" does nothing, and gives a bad message
158 about it. Should fix both to print nice actionable messages.
159 - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service
160 - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible
161 - systemctl enable: improve the success messages (i.e. more human readable, less shell-like)
162 - systemctl enable: fail if target to alias into doesn't exist? maybe show how many units are enabled afterwards?
163 - systemctl: "Journal has been rotated since unit was started." message is misleading
164 - support "systemctl stop foobar@.service" to stop all units matching a certain template
165 - Something is wrong with symlink handling of "autovt@.service" in "systemctl list-unit-files"
167 * When shutdown.target is queued begin with an asynchronous sync()?
169 * introduce ntp.service (or suchlike) as symlink that is used to arbitrate between various NTP implementations
171 * deal with sendmail/postfix exclusivity
174 - configurable jitter for timer events
175 - timer events with system resume
176 - timer units should get the ability to trigger when:
177 o CLOCK_REALTIME makes jumps (TFD_TIMER_CANCEL_ON_SET)
180 * update the kernel's TZ (sys_tz) when DST changes
182 * sync down the system time to the RTC when:
183 - CLOCK_REALTIME makes jumps (the user explicitely requested a time set)
184 - DST/timezone changes && ntp is active && RTC-in-localtime (never do it without ntp)
185 This takes care of syncing ntpdate updates to the RTC, and DST updates for localtime
186 mode, it will never touch the RTC if the no reliable time source is active or the
187 user did not request anything like it.
190 - implement conditional properties (dmi matches)
191 - hwdb --filter=ID_DRIVE_*
193 * if booted in "quiet" mode, and an error happens, turn on status
194 output again, so that the emergency mode isn't totally
195 surprising. Also, terminate plymouth.
197 * libunwind support for coredump pattern hook, and includes this in
198 the message for coredumps. After all, libunwind is now capable to
199 unwind coredumps since a few weeks ago. This probably requires that
200 we have nice support for multi-line messages on display in
201 logs-show.c. Alternatively: use libelfutil, which seems to be the
202 better supported alternative.
204 * figure out relation of --all and --full in the various tools
206 * add libsystemd-password or so to query passwords during boot using the password agent logic
208 * 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().
210 * fedup: add --unit to systemctl switch-root somehow
211 * fedup: don't delete initrd on switch-root
215 - timedated: export boolean that clarifies whether NTP is even available
216 - timedated: refuse time changes when NTP is on
218 * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed
220 * introduce generic AUGMENT_PID=, AUGMENT_DEVICE= fields
222 * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel
224 * add "provisioning" instructions to setup an empty /etc + /var
225 - used to setup a new container from a shared /usr
226 - superset of tmpfiles model
227 - instructions shipped by packages and stored in /usr/lib/
228 - compose /etc/passwd and /etc/group, copy files
229 - able to create uid + gid used by packages, for file ownership
231 * make repeated alt-ctrl-del presses printing a dump, or even force a reboot without
232 waiting for the timeout
234 * high level net_prio setting in execution context
236 * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ...
238 * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab isn't
240 * WorkingDirectory: support env var replacements like in ExecStart= so that people can use $HOME
242 * refuse boot if /etc/machine-id is not useful (or set taint?)
245 - nspawn: consider changing users for -u with su, so that NSS resolving works correctly
246 - nspawn: implement personality changes a la linux32(8)
247 - nspawn: --read-only is not applied recursively to submounts
248 - nspawn: make use of device cgroup controller by default
249 - bind mount read-only the cgroup tree higher than nspawn
252 - cryptsetup-generator: warn if the password files are world-readable
253 - cryptsetup-generator: add RequiresMountsFor= to cryptseup service files referencing a file, similar for devices
254 - cryptsetup-generator: allow specification of passwords in crypttab itself
255 - move cryptsetup key caching into kernel keyctl?
256 https://bugs.freedesktop.org/show_bug.cgi?id=54982
257 - when key file cannot be found, read it from kbd in cryptsetup
259 * instantiated [Install] for target units
260 https://bugs.freedesktop.org/show_bug.cgi?id=54377
262 * move debug shell to tty6 and make sure this doesn't break the gettys on tty6
264 * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it
265 https://bugs.freedesktop.org/show_bug.cgi?id=54712
267 * after deserializing sockets in socket.c we should reapply sockopts and things
269 * make timer units go away after they elapsed
271 * http://lists.freedesktop.org/archives/systemd-devel/2012-September/006502.html
272 (network and remote-fs on shutdown)
274 * come up with a nice way to write queue/read_ahead_kb for a block device without interfering with readahead
276 * move PID 1 segfaults to /var/lib/systemd/coredump?
278 * create /sbin/init symlinks from the build system
280 * Query Paul Moore about relabelling socket fds while they are open
282 * system.conf should have controls for cgroups
284 * allow writing multiple conditions in unit files on one line
286 * explore multiple service instances per listening socket idea
288 * testing tool for socket activation: some binary that listens on a socket and passes it on using the usual socket activation protocol to some server.
290 * shutdown: don't read-only mount anything when running in container
292 * MountFlags=shared acts as MountFlags=slave right now.
294 * ReadOnlyDirectories= is not applied recursively to submounts
296 * drop PID 1 reloading, only do reexecing (difficult: Reload()
297 currently is properly synchronous, Reexec() is weird, because we
298 can't delay the response properly until we are back, so instead of
299 being properly synchronous we just keep open the fd and close it
300 when done. That means clients don't get a successful method reply,
301 but much rather a disconnect on success.
303 * remember which condition failed for services, not just the fact that something failed
305 * use opterr = 0 for all getopt tools
307 * properly handle loop back mounts via fstab, especially regards to fsck/passno
309 * allow services with no ExecStart= but with an ExecStop=
311 * dracut-shutdown needs to be ordered before unmounting /boot
313 * initialize the hostname from the fs label of /, if /etc/hostname does not exist?
315 * rename "userspace" to "core-os"
318 - syscall filter: add knowledge about compat syscalls
319 - syscall filter: don't enforce no new privs?
320 - syscall filter: option to return EPERM rather than SIGSYS?
321 - syscall filter: port to libseccomp
322 - system-wide seccomp filter
324 * systemd-analyze: data collection tools should be lightweight (few dependencies); data analysis tools can be heavyweight
326 * .device aliases need to be implemented with the "following" logic, probably.
328 * load-fragment: when loading a unit file via a chain of symlinks
329 verify that it isn't masked via any of the names traversed.
331 * introduce Type=pid-file
333 * maybe allow services with ExecStop= set, but no ExecStart=?
335 * change Requires=basic.target to RequisiteOverride=basic.target
337 * support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator
339 * drop accountsservice's StandardOutput=syslog and Type=dbus fields
341 * when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr
343 * move passno parsing to fstab generator
345 * automount: implement expire:
346 - set superblock timeout AUTOFS_DEV_IOCTL_TIMEOUT_CMD
347 - periodically run AUTOFS_DEV_IOCTL_EXPIRE_CMD
348 - every timeout/4 (original autofs logic)
349 - blocking, needs a thread
351 - receive expire packet on pipe if kernel tells the timeout is over
353 - answer expire packet on pipe with AUTOFS_DEV_IOCTL_{READY,FAIL}_CMD
354 - AUTOFS_DEV_IOCTL_EXPIRE_CMD returns
356 * services which create their own subcgroups break cgroup-empty notification (needs to be fixed in the kernel)
358 * ExecOnFailure=/usr/bin/foo
360 * ConditionSecurity= should learn about IMA and SMACK
363 - remove all (misguided from day 1) userspace firmware_class handling
365 - unify utf8 validator code with shared/
367 - add trigger --subsystem-match=usb/usb_device device
369 * cleanup syslog 'priority' vs. 'level' wording
371 * dbus upstream still refers to dbus.target and shouldn't
373 * when a service has the same env var set twice we actually store it twice and return that in systemctl show -p... We should only show the last setting
375 * support container_ttys=
377 * introduce mix of BindTo and Requisite
379 * add DeleteSocketsOnStop=yes|no option to socket units
381 * There's currently no way to cancel fsck (used to be possible via C-c or c on the console)
383 * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/
385 * default unix qlen is too small (10). bump sysctl? add sockopt?
387 * figure out whether we should leave dbus around during shutdown
389 * dbus: in fedora, make /var/lib/dbus/machine-id a symlink to /etc/machine-id
391 * dbus: move dbus to early boot
393 * save coredump in Windows/Mozilla minidump format
395 * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting)
397 * clean up session cgroups that remain after logout (think sshd), but eventually run empty
399 * when an instanced service exits, remove its parent cgroup too if possible.
401 * default to actual 32bit PIDs, via /proc/sys/kernel/pid_max
403 * be able to specify a forced restart of service A where service B depends on, in case B
404 needs to be auto-respawned?
406 * when a bus name of a service disappears from the bus make sure to queue further activation requests
408 * tmpfiles: apply "x" on "D" too (see patch from William Douglas)
410 * for services: don't set $HOME in services unless requested
412 * hide PAM/TCPWrap options in fragment parser when compile time disabled
414 * when we automatically restart a service, ensure we restart its rdeps, too.
416 * allow Type=simple with PIDFile=
417 https://bugzilla.redhat.com/show_bug.cgi?id=723942
419 * move PAM code into its own binary
421 * implement Register= switch in .socket units to enable registration
422 in Avahi, RPC and other socket registration services.
424 * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early
426 * support sd_notify() style notification when reload begins (RELOADING=1), reload is finished (READY=1), and add ReloadSignal= then to use in combination
428 * support sd_notify() style notification when shutting down, to make auto-exit bus services work (STOPPING=1)
430 * verify that the AF_UNIX sockets of a service in the fs still exist
431 when we start a service in order to avoid confusion when a user
432 assumes starting a service is enough to make it accessible
434 * Make it possible to set the keymap independently from the font on
435 the kernel cmdline. Right now setting one resets also the other.
437 * and a dbus call to generate target from current state
440 - drop /.readahead on bigger upgrades with yum
441 - move readahead files into /var (look for them with .path units?)
442 - readahead: use BTRFS_IOC_DEFRAG_RANGE instead of BTRFS_IOC_DEFRAG ioctl, with START_IO
443 - readahead: when bumping /sys readahead variable save mtime and compare later to detect changes
444 - readahead: make use of EXT4_IOC_MOVE_EXT, as used by http://e4rat.sourceforge.net/
446 * add support for /bin/mount -s
448 * GC unreferenced jobs (such as .device jobs)
450 * write blog stories about:
451 - hwdb: what belongs into it, lsusb
452 - enabling dbus services
454 - how to make changes to sysctl and sysfs attributes
456 - how to pass throw-away units to systemd, or dynamically change properties of existing units
457 - how to integrate cgconfig and suchlike with systemd
458 - testing with Harald's awesome test kit
460 - how to develop against journal browsing APIs
461 - the journal HTTP iface
462 - non-cgroup resource management
463 - dynamic resource management with cgroups
464 - refreshed, longer missions statement
465 - celendar time events
467 * allow port=0 in .socket units
469 * support systemd.mask= on the kernel command line.
471 * reuse mkdtemp namespace dirs in /tmp?
473 * recreate systemd's D-Bus private socket file on SIGUSR2
475 * Support --test based on current system state
477 * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support.
479 * maybe introduce ExecRestartPre=
481 * dot output for --test showing the 'initial transaction'
483 * port over to LISTEN_FDS/LISTEN_PID:
484 - rpcbind (/var/run/rpcbind.sock!) HAVEPATCH
488 - libvirtd (/var/run/libvirt/libvirt-sock-ro)
489 - bluetoothd (/var/run/sdp! @/org/bluez/audio!)
492 * fingerprint.target, wireless.target, gps.target, netdevice.target
494 * io priority during initialization
496 * drop cap bounding set in readahead and other services
502 - natively watch for dbus-*.service symlinks (PENDING)
503 - allow specification of socket mode/umask when allocating DBusServer
504 - allow disabling of fd passing when connecting a AF_UNIX connection
505 - allow disabling of UID passing for AUTH EXTERNAL
506 - always pass cred data along each message
507 - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service
509 * fix alsa mixer restore to not print error when no config is stored
511 * make cryptsetup lower --iter-time
513 * patch kernel for xattr support in /dev, /proc/, /sys?
515 * NTP: the kernel's 11-minutes-mode syncs the system time to the RTC, but only
516 in an ~30 minutes window. It does not adjust larger differences. Find a way
517 to tell the kernel, to always do a full time sync when the RTC is in UTC and
518 we are in 11-minutes-mode. When we trust the system time to NTP we also want
521 * kernel: add device_type = "fb", "fbcon" to class "graphics"
525 * look for close() vs. close_nointr() vs. close_nointr_nofail()
527 * check for strerror(r) instead of strerror(-r)
529 * Use PR_SET_PROCTITLE_AREA if it becomes available in the kernel
531 * %m in printf() instead of strerror(errno);
535 * set_put(), hashmap_put() return values check. i.e. == 0 doesn't free()!
537 * use secure_getenv() instead of getenv() where appropriate
539 * link up selected blog stories from man pages and unit files Documentation= fields
541 Scheduled for removal or fixing:
543 * xxxOverridable dependencies (probably: fix)