You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

975 lines
46 KiB

  1. Bugfixes:
  2. * Should systemctl status \* work on all unit types, not just .service?
  3. * Dangling symlinks of .automount unit files in .wants/ directories, set up
  4. automount points even when the original .automount file did not exist
  5. anymore. Only the .mount unit was still around.
  6. * ExecStart with unicode characters fails in strv_split_extract:
  7. [Service]
  8. Environment=ONE='one' "TWO='two two' too" THREE=
  9. ExecStart=/bin/python3 -c 'import sys;print(sys.argv)' $ONE $TWO $THREE
  10. External:
  11. * Fedora: add an rpmlint check that verifies that all unit files in the RPM are listed in %systemd_post macros.
  12. * wiki: update journal format documentation for lz4 additions
  13. Janitorial Clean-ups:
  14. * Rearrange tests so that the various test-xyz.c match a specific src/basic/xyz.c again
  15. Features:
  16. * make use of ethtool veth peer info in machined, for automatically finding out
  17. host-side interface pointing to the container.
  18. * add some special mode to LogsDirectory=/StateDirectory=… that allows
  19. declaring these directories without necessarily pulling in deps for them, or
  20. creating them when starting up. That way, we could declare that
  21. systemd-journald writes to /var/log/journal, which could be useful when we
  22. doing disk usage calculations and so on.
  23. * taint systemd if there are fewer than 65536 users assigned to the system.
  24. * deprecate PermissionsStartOnly= and RootDirectoryStartOnly= in favour of the ExecStart= prefix chars
  25. * add a new RuntimeDirectoryPreserve= mode that defines a similar lifecycle for
  26. the runtime dir as we maintain for the fdstore: i.e. keep it around as long
  27. as the unit is running or has a job queued.
  28. * hook up sd-bus' creds stuff with SO_PEERGROUPS
  29. * add async version of sd_bus_add_match and make use of that
  30. * support projid-based quota in machinectl for containers, and then drop
  31. implicit btrfs loopback magic in machined
  32. * Add NetworkNamespacePath= to specify a path to a network namespace
  33. * maybe use SOURCE_DATE_EPOCH (i.e. the env var the reproducible builds folks
  34. introduced) as the RTC epoch, instead of the mtime of NEWS.
  35. * add a way to lock down cgroup migration: a boolean, which when set for a unit
  36. makes sure the processes in it can never migrate out of it
  37. * blog about fd store and restartable services
  38. * document Environment=SYSTEMD_LOG_LEVEL=debug drop-in in debugging document
  39. * rework ExecOutput and ExecInput enums so that EXEC_OUTPUT_NULL loses its
  40. magic meaning and is no longer upgraded to something else if set explicitly.
  41. * in the long run: permit a system with /etc/machine-id linked to /dev/null, to
  42. make it lose its identity, i.e. be anonymous. For this we'd have to patch
  43. through the whole tree to make all code deal with the case where no machine
  44. ID is available.
  45. * optionally, collect cgroup resource data, and store it in per-unit RRD files,
  46. suitable for processing with rrdtool. Add bus API to access this data, and
  47. possibly implement a CPULoad property based on it.
  48. * beef up pam_systemd to take unit file settings such as cgroups properties as
  49. parameters
  50. * a new "systemd-analyze security" tool outputting a checklist of security
  51. features a service does and does not implement
  52. * maybe hook of xfs/ext4 quotactl() with services? i.e. automatically manage
  53. the quota of a the user indicated in User= via unit file settings, like the
  54. other resource management concepts. Would mix nicely with DynamicUser=1. Or
  55. alternatively, do this with projids, so that we can also cover services
  56. running as root. Quota should probably cover all the special dirs such as
  57. StateDirectory=, LogsDirectory=, CacheDirectory=, as well as RootDirectory= if it
  58. is set, plus the whole disk space any image configured with RootImage=.
  59. * Introduce "exit" as an EmergencyAction value, and allow to configure a
  60. per-unit success/failure exit code to configure. This would be useful for
  61. running commands inside of services inside of containers, which could then
  62. propagate their failure state all the way up.
  63. * In DynamicUser= mode: before selecting a UID, use disk quota APIs on relevant
  64. disks to see if the UID is already in use.
  65. * add dissect_image_warn() as a wrapper around dissect_image() that prints
  66. friendly log messages for the returned errors, so that we don't have to
  67. duplicate that in nspawn, systemd-dissect and PID 1.
  68. * add "systemctl wait" or so, which does what "systemd-run --wait" does, but
  69. for all units. It should be both a way to pin units into memory as well as a
  70. wait to retrieve their exit data.
  71. * maybe set a new set of env vars for services, based on RuntimeDirectory=,
  72. StateDirectory=, LogsDirectory=, CacheDirectory= and ConfigurationDirectory=
  73. automatically. For example, there could be $RUNTIME_DIRECTORY,
  75. $CONFIGURATION_DIRECTORY or so. This could be useful to write services that
  76. can adapt to varying directories for these purposes. Special care has to be
  77. taken if multiple dirs are configured. Maybe avoid setting the env vars in
  78. that case?
  79. * expose IO accounting data on the bus, show it in systemd-run --wait and log
  80. about it in the resource log message
  81. * add "systemctl purge" for flushing out configuration, state, logs, ... of a
  82. unit when it is stopped
  83. * show whether a service has out-of-date configuration in "systemctl status" by
  84. using mtime data of ConfigurationDirectory=.
  85. * replace all uses of fgets() + LINE_MAX by read_line()
  86. * Add AddUser= setting to unit files, similar to DynamicUser=1 which however
  87. creates a static, persistent user rather than a dynamic, transient user. We
  88. can leverage code from sysusers.d for this.
  89. * add some optional flag to ReadWritePaths= and friends, that has the effect
  90. that we create the dir in question when the service is started. Example:
  91. ReadWritePaths=:/var/lib/foobar
  92. * maybe add call sd_journal_set_block_timeout() or so to set SO_SNDTIMEO for
  93. the sd-journal logging socket, and, if the timeout is set to 0, sets
  94. O_NONBLOCK on it. That way people can control if and when to block for
  95. logging.
  96. * tighten sd_notify() MAINPID= checks a bit: don't accept foreign PIDs (i.e.
  97. PIDs not managed by the service manager)
  98. * hostnamed: populate form factor data from a new hwdb database, so that old
  99. yogas can be recognized as "convertible" too, even if they predate the DMI
  100. "convertible" form factor
  101. * Maybe add a small tool invoked early at boot, that adds in or resizes
  102. partitions automatically, to be used when the media used is actually larger
  103. than the image written onto it is.
  104. * Maybe add PrivatePIDs= as new unit setting, and do minimal PID namespacing
  105. after all. Be strict however, only support the equivalent of nspawn's
  106. --as-pid2 switch, and sanely proxy sd_notify() messages dropping stuff such
  107. as MAINPID.
  108. * Add ExecMonitor= setting. May be used multiple times. Forks off a process in
  109. the service cgroup, which is supposed to monitor the service, and when it
  110. exits the service is considered failed by its monitor.
  111. * track the per-service PAM process properly (i.e. as an additional control
  112. process), so that it may be queried on the bus and everything.
  113. * add a new "debug" job mode, that is propagated to unit_start() and for
  114. services results in two things: we raise SIGSTOP right before invoking
  115. execve() and turn off watchdog support. Then, use that to implement
  116. "systemd-gdb" for attaching to the start-up of any system service in its
  117. natural habitat.
  118. * replace all canonicalize_file_name() invocations by chase_symlinks(), in
  119. particulr those where a rootdir is relevant.
  120. * maybe introduce gpt auto discovery for /var/tmp?
  121. * maybe add gpt-partition-based user management: each user gets his own
  122. LUKS-encrypted GPT partition with a new GPT type. A small nss module
  123. enumerates users via udev partition enumeration. UIDs are assigned in a fixed
  124. way: the partition index is added as offset to some fixed base uid. User name
  125. is stored in GPT partition name. A PAM module authenticates the user via the
  126. LUKS partition password. Benefits: strong per-user security, compatibility
  127. with stateless/read-only/verity-enabled root. (other idea: do this based on
  128. loopback files in /home, without GPT involvement)
  129. * gpt-auto logic: introduce support for discovering /var matching an image. For
  130. that, use a partition type UUID that is hashed from the OS name (as encoded
  131. in /etc/os-release), the architecture, and 4 new bits from the gpt flags
  132. field of the root partition. This way can easily support multiple OS
  133. installations on the same GPT partition table, without problems with
  134. unmatched /var partitions.
  135. * gpt-auto logic: related to the above, maybe support a "secondary" root
  136. partition, that is mounted to / and is writable, and where the actual root's
  137. /usr is mounted into.
  138. * 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
  139. * drop nss-myhostname in favour of nss-resolve?
  140. * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and
  141. then use that for the setting used in user@.service. It should be understood
  142. relative to the configured default value.
  143. * in networkd, when matching device types, fix up DEVTYPE rubbish the kernel passes to us
  144. * enable LockMLOCK to take a percentage value relative to physical memory
  145. * Permit masking specific netlink APIs with RestrictAddressFamily=
  146. * nspawn: support that /proc, /sys/, /dev are pre-mounted
  147. * define gpt header bits to select volatility mode
  148. * ProtectKernelLogs= (drops CAP_SYSLOG, add seccomp for syslog() syscall, and DeviceAllow to /dev/kmsg) in service files
  149. * ProtectClock= (drops CAP_SYS_TIMES, adds seecomp filters for settimeofday, adjtimex), sets DeviceAllow o /dev/rtc
  150. * ProtectTracing= (drops CAP_SYS_PTRACE, blocks ptrace syscall, makes /sys/kernel/tracing go away)
  151. * ProtectMount= (drop mount/umount/pivot_root from seccomp, disallow fuse via DeviceAllow, imply Mountflags=slave)
  152. * ProtectKeyRing= to take keyring calls away
  153. * RemoveKeyRing= to remove all keyring entries of the specified user
  154. * ProtectReboot= that masks reboot() and kexec_load() syscalls, prohibits kill
  155. on PID 1 with the relevant signals, and makes relevant files in /sys and
  156. /proc (such as the sysrq stuff) unavailable
  157. * DeviceAllow= should also generate seccomp filters for mknod()
  158. * make sure the ratelimit object can deal with USEC_INFINITY as way to turn off things
  159. * journalctl: make sure -f ends when the container indicated by -M terminates
  160. * mount: automatically search for "main" partition of an image has multiple
  161. partitions
  162. * expose the "privileged" flag of ExecCommand on the bus, and open it up to
  163. transient units
  164. * in nss-systemd, if we run inside of RootDirectory= with PrivateUsers= set,
  165. find a way to map the User=/Group= of the service to the right name. This way
  166. a user/group for a service only has to exist on the host for the right
  167. mapping to work.
  168. * add bus API for creating unit files in /etc, reusing the code for transient units
  169. * add bus API to remove unit files from /etc
  170. * add bus API to retrieve current unit file contents (i.e. implement "systemctl cat" on the bus only)
  171. * rework fopen_temporary() to make use of open_tmpfile_linkable() (problem: the
  172. kernel doesn't support linkat() that replaces existing files, currently)
  173. * check if DeviceAllow= should split first, resolve specifiers later
  174. * transient units: don't bother with actually setting unit properties, we
  175. reload the unit file anyway
  176. * journald: sigbus API via a signal-handler safe function that people may call
  177. from the SIGBUS handler
  178. * optionally, also require WATCHDOG=1 notifications during service start-up and shutdown
  179. * resolved: when routing queries, make sure only look for the *longest* suffix...
  180. * delay activation of logind until somebody logs in, or when /dev/tty0 pulls it
  181. in or lingering is on (so that containers don't bother with it until PAM is used). also exit-on-idle
  182. * cache sd_event_now() result from before the first iteration...
  183. * add systemctl stop --job-mode=triggering that follows TRIGGERED_BY deps and adds them to the same transaction
  184. * PID1: find a way how we can reload unit file configuration for
  185. specific units only, without reloading the whole of systemd
  186. * add an explicit parser for LimitRTPRIO= that verifies
  187. the specified range and generates sane error messages for incorrect
  188. specifications.
  189. * when we detect that there are waiting jobs but no running jobs, do something
  190. * push CPUAffinity= also into the "cpuset" cgroup controller (only after the cpuset controller got ported to the unified hierarchy)
  191. * PID 1 should send out sd_notify("WATCHDOG=1") messages (for usage in the --user mode, and when run via nspawn)
  192. * there's probably something wrong with having user mounts below /sys,
  193. as we have for debugfs. for exmaple, src/core/mount.c handles mounts
  194. prefixed with /sys generally special.
  196. * fstab-generator: default to tmpfs-as-root if only usr= is specified on the kernel cmdline
  197. * docs: bring up to date
  198. * mounting and unmounting mount points manually with different source
  199. devices will result in collected on all devices used.
  201. * add a job mode that will fail if a transaction would mean stopping
  202. running units. Use this in timedated to manage the NTP service
  203. state.
  205. * Maybe add support for the equivalent of "ethtool advertise" to .link files?
  207. * The udev blkid built-in should expose a property that reflects
  208. whether media was sensed in USB CF/SD card readers. This should then
  209. be used to control SYSTEMD_READY=1/0 so that USB card readers aren't
  210. picked up by systemd unless they contain a medium. This would mirror
  211. the behaviour we already have for CD drives.
  212. * networkd/udev: implement SR_IOV configuration in .link files:
  214. * Rework systemctl's GetAll property parsing to use the generic bus_map_all_properties() API
  215. * Port various tools to make use of verbs.[ch], where applicable: busctl,
  216. coredumpctl, hostnamectl, localectl, systemd-analyze, timedatectl
  217. * hostnamectl: show root image uuid
  218. * sysfs set api in libudev is not const
  219. * Find a solution for SMACK capabilities stuff:
  221. * "systemctl preset-all" should probably order the unit files it
  222. operates on lexicographically before starting to work, in order to
  223. ensure deterministic behaviour if two unit files conflict (like DMs
  224. do, for example)
  225. * synchronize console access with BSD locks:
  227. * as soon as we have sender timestamps, revisit coalescing multiple parallel daemon reloads:
  229. * in systemctl list-unit-files: show the install value the presets would suggest for a service in a third column
  230. * figure out when we can use the coarse timers
  231. * add "systemctl start -v foobar.service" that shows logs of a service
  232. while the start command runs. This is non-trivial to do without
  233. races though, since we should flush out all journal messages before
  234. returning from the "systemctl stop".
  235. * 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
  236. * maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment=
  237. * introduce systemd-timesync-wait.service or so to sync on an NTP fix?
  238. * consider showing the unit names during boot up in the status output, not just the unit descriptions
  239. * maybe allow timer units with an empty Units= setting, so that they
  240. can be used for resuming the system but nothing else.
  241. * what to do about udev db binary stability for apps? (raw access is not an option)
  242. * maybe provide an API to allow migration of foreign PIDs into existing scopes.
  243. * man: maybe use the word "inspect" rather than "introspect"?
  244. * systemctl: if some operation fails, show log output?
  245. * systemctl edit: use equvalent of cat() to insert existing config as a comment, prepended with #.
  246. Upon editor exit, lines with one # are removed, lines with two # are left with one #, etc.
  247. * exponential backoff in timesyncd when we cannot reach a server
  248. * timesyncd: add ugly bus calls to set NTP servers per-interface, for usage by NM
  249. * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share....
  250. * systemd.show_status= should probably have a mode where only failed
  251. units are shown.
  252. * add systemd.abort_on_kill or some other such flag to send SIGABRT instead of SIGKILL
  253. (throughout the codebase, not only PID1)
  254. * resolved:
  255. - mDNS/DNS-SD
  256. - service registration
  257. - service/domain/types browsing
  258. - avahi compat
  259. - DNS-SD service registration from socket units
  260. - resolved should optionally register additional per-interface LLMNR
  261. names, so that for the container case we can establish the same name
  262. (maybe "host") for referencing the server, everywhere.
  263. - allow clients to request DNSSEC for a single lookup even if DNSSEC is off (?)
  264. - hook up resolved with machined-based address resolution
  265. * refcounting in sd-resolve is borked
  266. * Add a new verb "systemctl top"
  267. * add new gpt type for btrfs volumes
  268. * support empty /etc boots nicely:
  269. - nspawn/gpt-generator: introduce new gpt partition type for /usr
  270. * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them.
  271. * a way for container managers to turn off getty starting via $container_headless= or so...
  272. * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit
  273. * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services
  274. they run added to the initial transaction and thus confuse Type=idle.
  275. * add bus api to query unit file's X fields.
  276. * gpt-auto-generator:
  277. - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap?
  278. - Make /home automount rather than mount?
  279. * add generator that pulls in systemd-network from containers when
  280. CAP_NET_ADMIN is set, more than the loopback device is defined, even
  281. when it is otherwise off
  282. * MessageQueueMessageSize= (and suchlike) should use parse_iec_size().
  283. * implement Distribute= in socket units to allow running multiple
  284. service instances processing the listening socket, and open this up
  285. for ReusePort=
  286. * socket units: support creating sockets in different namespace,
  287. opening it up for JoinsNamespaceOf=. This would require to fork off
  288. a tiny process that joins the namespace and creates/binds the socket
  289. and passes this back to PID1 via SCM_RIGHTS. This also could be used
  290. to allow Chown/chgrp on sockets without requiring NSS in PID 1.
  291. * introduce bus call FreezeUnit(s, b), as well as "systemctl freeze
  292. $UNIT" and "systemctl thaw $UNIT" as wrappers around this. The calls
  293. should SIGSTOP all unit processes in a loop until all processes of
  294. it are fully stopped. This can later be used for app management by
  295. desktop UIs such as gnome-shell to freeze apps that are not visible
  296. on screen, not unlike how job control works on the shell
  297. * cgroups:
  298. - implement per-slice CPUFairScheduling=1 switch
  299. - handle jointly mounted controllers correctly
  300. - introduce high-level settings for RT budget, swappiness
  301. - how to reset dynamically changed unit cgroup attributes sanely?
  302. - when reloading configuration, apply new cgroup configuration
  303. - when recursively showing the cgroup hierarchy, optionally also show
  304. the hierarchies of child processes
  305. * transient units:
  306. - add field to transient units that indicate whether systemd or somebody else saves/restores its settings, for integration with libvirt
  307. - ensure scope units may be started only a single time
  308. * Automatically configure swap partition to use for hibernation by looking for largest swap partition on the root disk?
  309. * when we detect low battery and no AC on boot, show pretty splash and refuse boot
  310. * libsystemd-journal, libsystemd-login, libudev: add calls to easily attach these objects to sd-event event loops
  311. * be more careful what we export on the bus as (usec_t) 0 and (usec_t) -1
  312. * 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
  313. * After coming back from hibernation reset hibernation swap partition using the /dev/snapshot ioctl APIs
  314. * If we try to find a unit via a dangling symlink, generate a clean
  315. error. Currently, we just ignore it and read the unit from the search
  316. path anyway.
  317. * refuse boot if /usr/lib/os-release is missing or /etc/machine-id cannot be set up
  318. * man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted.
  319. * load .d/*.conf dropins for device units
  320. * allow implementation of InaccessibleDirectories=/ plus
  321. ReadOnlyDirectories=... for whitelisting files for a service.
  322. * sd-bus:
  323. - EBADSLT handling
  324. - GetAllProperties() on a non-existing object does not result in a failure currently
  325. - port to sd-resolve for connecting to TCP dbus servers
  326. - 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
  327. - see if we can drop more message validation on the sending side
  328. - add API to clone sd_bus_message objects
  329. - make AddMatch calls on dbus1 transports async?
  330. - longer term: priority inheritance
  331. - dbus spec updates:
  332. - NameLost/NameAcquired obsolete
  333. - GVariant
  334. - path escaping
  335. - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now
  336. - test bloom filter generation indexes
  337. * sd-event
  338. - allow multiple signal handlers per signal?
  339. - document chaining of signal handler for SIGCHLD and child handlers
  340. - define more intervals where we will shift wakeup intervals around in, 1h, 6h, 24h, ...
  341. - generate a failure of a default event loop is executed out-of-thread
  342. - maybe add support for inotify events
  343. * investigate endianness issues of UUID vs. GUID
  344. * dbus: when a unit failed to load (i.e. is in UNIT_ERROR state), we
  345. should be able to safely try another attempt when the bus call LoadUnit() is invoked.
  346. * add a pam module that passes the hdd passphrase into the PAM stack and then expires it, for usage by gdm auto-login.
  347. * add a pam module that on password changes updates any LUKS slot where the password matches
  348. * maybe add a generator that looks for "" on the kernel cmdline for container usercases...
  349. * test/:
  350. - add 'set -e' to scripts in test/
  351. - make stuff in test/ work with separate output dir
  352. * seems that when we follow symlinks to units we prefer the symlink
  353. destination path over /etc and /usr. We should not do that. Instead
  354. /etc should always override /run+/usr and also any symlink
  355. destination.
  356. * when isolating, try to figure out a way how we implicitly can order
  357. all units we stop before the isolating unit...
  358. * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off})
  359. * BootLoaderSpec: Clarify that the kernel has to be in $BOOT. Clarify
  360. that the boot loader should be installed to the ESP. Define a way
  361. how an installer can figure out whether a BLS compliant boot loader
  362. is installed.
  363. * think about requeuing jobs when daemon-reload is issued? usecase:
  364. the initrd issues a reload after fstab from the host is accessible
  365. and we might want to requeue the mounts local-fs acquired through
  366. that automatically.
  367. * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep()
  368. * remove any syslog support from log.c — we probably cannot do this before split-off udev is gone for good
  369. * shutdown logging: store to EFI var, and store to USB stick?
  370. * merge unit_kill_common() and unit_kill_context()
  371. * introduce ExecCondition= in services
  372. * EFI:
  373. - honor language efi variables for default language selection (if there are any?)
  374. - honor timezone efi variables for default timezone selection (if there are any?)
  375. - change bootctl to be backed by systemd-bootd to control temporary and persistent default boot goal plus efi variables
  376. * maybe do not install getty@tty1.service symlink in /etc but in /usr?
  377. * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word
  378. * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units.
  379. * logind:
  380. - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around
  381. - When we update the kernel all kind of hibernation should be prohibited until shutdown/reboot
  382. - logind: wakelock/opportunistic suspend support
  383. - Add pretty name for seats in logind
  384. - logind: allow showing logout dialog from system?
  385. - we should probably handle SIGTERM/SIGINT to not leave dot files around, just in case
  386. - session scopes/user unit: add RequiresMountsFor for the home directory of the user
  387. - add Suspend() bus calls which take timestamps to fix double suspend issues when somebody hits suspend and closes laptop quickly.
  388. - if pam_systemd is invoked by su from a process that is outside of a
  389. any session we should probably just become a NOP, since that's
  390. usually not a real user session but just some system code that just
  391. needs setuid().
  392. - logind: make the Suspend()/Hibernate() bus calls wait for the for
  393. the job to be completed. before returning, so that clients can wait
  394. for "systemctl suspend" to finish to know when the suspending is
  395. complete.
  396. - logind: when the power button is pressed short, just popup a
  397. logout dialog. If it is pressed for 1s, do the usual
  398. shutdown. Inspiration are Macs here.
  399. - expose "Locked" property on logind sesison objects
  400. - given that logind now lets PID 1 do all nasty work, we can
  401. probably reduce the capability set it retains substantially.
  402. (we need CAP_SYS_ADMIN for drmSetMaster(), so maybe not worth it)
  403. - expose orientation sensors and tablet mode through logind
  404. - maybe allow configuration of the StopTimeout for session scopes
  405. - rename session scope so that it includes the UID. THat way
  406. the session scope can be arranged freely in slices and we don't have
  407. make assumptions about their slice anymore.
  408. - follow PropertiesChanged state more closely, to deal with quick logouts and
  409. relogins
  410. * exec: when deinitializating a tty device fix the perms and group, too, not only when initializing. Set access mode/gid to 0620/tty.
  411. * service: watchdog logic: for testing purposes allow ping, but do not require pong
  412. * journal:
  413. - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields
  414. - import and delete pstore filesystem content at startup
  415. - journald: also get thread ID from client, plus thread name
  416. - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups
  417. - add API to close/reopen/get fd for journal client fd in libsystemd-journal.
  418. - fallback to /dev/log based logging in libsystemd-journal, if we cannot log natively?
  419. - declare the local journal protocol stable in the wiki interface chart
  420. - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg
  421. - journald: when dropping msgs due to ratelimit make sure to write
  422. "dropped %u messages" not only when we are about to print the next
  423. message that works, but alraedy after a short tiemout
  424. - check if we can make journalctl by default use --follow mode inside of less if called without args?
  425. - maybe add API to send pairs of iovecs via sd_journal_send
  426. - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access
  427. - journactl: support negative filtering, i.e. FOOBAR!="waldo",
  428. and !FOOBAR for events without FOOBAR.
  429. - journal: store timestamp of journal_file_set_offline() int he header,
  430. so it is possible to display when the file was last synced.
  431. - 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.
  432. - journal: find a way to allow dropping history early, based on priority, other rules
  433. - journal: When used on NFS, check payload hashes
  434. - journald: add kernel cmdline option to disable ratelimiting for debug purposes
  435. - refuse taking lower-case variable names in sd_journal_send() and friends.
  436. - journald: we currently rotate only after MaxUse+MaxFilesize has been reached.
  437. - journal: deal nicely with byte-by-byte copied files, especially regards header
  438. - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit
  439. - Replace utmp, wtmp, btmp, and lastlog completely with journal
  440. - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax?
  441. - when a kernel driver logs in a tight loop, we should ratelimit that too.
  442. - journald: optionally, log debug messages to /run but everything else to /var
  443. - journald: when we drop syslog messages because the syslog socket is
  444. full, make sure to write how many messages are lost as first thing
  445. to syslog when it works again.
  446. - change systemd-journal-flush into a service that stays around during
  447. boot, and causes the journal to be moved back to /run on shutdown,
  448. so that we do not keep /var busy. This needs to happen synchronously,
  449. hence doing this via signals is not going to work.
  450. - optionally support running journald from the command line for testing purposes in external projects
  451. - journald: allow per-priority and per-service retention times when rotating/vacuuming
  452. - journald: make use of uid-range.h to managed uid ranges to split
  453. journals in.
  454. - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so...
  455. - improve journalctl performance by loading journal files
  456. lazily. Encode just enough information in the file name, so that we
  457. do not have to open it to know that it is not interesting for us, for
  458. the most common operations.
  459. - journal-or-kmsg is currently broken? See reverted
  460. commit 4a01181e460686d8b4a543b1dfa7f77c9e3c5ab8.
  461. - man: document that corrupted journal files is nothing to act on
  462. - rework journald sigbus stuff to use mutex
  463. - Set RLIMIT_NPROC for systemd-journal-xyz, and all other of our
  464. services that run under their own user ids, and use User= (but only
  465. in a world where userns is ubiquitous since otherwise we cannot
  466. invoke those daemons on the host AND in a container anymore). Also,
  467. if LimitNPROC= is used without User= we should warn and refuse
  468. operation.
  469. - journalctl --verify: don't show files that are currently being
  470. written to as FAIL, but instead show that their are being written to.
  471. - add journalctl -H that talks via ssh to a remote peer and passes through
  472. binary logs data
  473. - add a version of --merge which also merges /var/log/journal/remote
  474. - journalctl: -m should access container journals directly by enumerating
  475. them via machined, and also watch containers coming and going.
  476. Benefit: nspawn --ephemeral would start working nicely with the journal.
  477. - assign MESSAGE_ID to log messages about failed services
  478. * add a test if all entries in the catalog are properly formatted.
  479. (Adding dashes in a catalog entry currently results in the catalog entry
  480. being silently skipped. journalctl --update-catalog must warn about this,
  481. and we should also have a unit test to check that all our message are OK.)
  482. * document:
  483. - document that deps in [Unit] sections ignore Alias= fields in
  484. [Install] units of other units, unless those units are disabled
  485. - man: clarify that is not only sysv compat but also useful otherwise. Same for similar targets
  486. - document that service reload may be implemented as service reexec
  487. - add a man page containing packaging guidelines and recommending usage of things like Documentation=, PrivateTmp=, PrivateNetwork= and ReadOnlyDirectories=/etc /usr.
  488. - document systemd-journal-flush.service properly
  489. - documentation: recommend to connect the timer units of a service to the service via Also= in [Install]
  490. - man: document the very specific env the shutdown drop-in tools live in
  491. - man: add more examples to man pages
  492. - man: maybe sort directives in man pages, and take sections from --help and apply them to man too
  493. * systemctl:
  494. - add systemctl switch to dump transaction without executing it
  495. - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done
  496. - "systemctl disable" on a static unit prints no message and does
  497. nothing. "systemctl enable" does nothing, and gives a bad message
  498. about it. Should fix both to print nice actionable messages.
  499. - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service
  500. - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible
  501. - systemctl enable: fail if target to alias into does not exist? maybe show how many units are enabled afterwards?
  502. - systemctl: "Journal has been rotated since unit was started." message is misleading
  503. - systemctl status output should include list of triggering units and their status
  504. * unit install:
  505. - "systemctl mask" should find all names by which a unit is accessible
  506. (i.e. by scanning for symlinks to it) and link them all to /dev/null
  507. * timer units:
  508. - timer units should get the ability to trigger when:
  510. o DST changes
  511. - Modulate timer frequency based on battery state
  512. * add libsystemd-password or so to query passwords during boot using the password agent logic
  513. * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed
  514. * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel
  515. * make repeated alt-ctrl-del presses printing a dump
  516. * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ...
  517. * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab is not
  518. * nspawn:
  519. - emulate /dev/kmsg using CUSE and turn off the syslog syscall
  520. with seccomp. That should provide us with a useful log buffer that
  521. systemd can log to during early boot, and disconnect container logs
  522. from the kernel's logs.
  523. - as soon as networkd has a bus interface, hook up --network-interface=,
  524. --network-bridge= with networkd, to trigger netdev creation should an
  525. interface be missing
  526. - a nice way to boot up without machine id set, so that it is set at boot
  527. automatically for supporting --ephemeral. Maybe hash the host machine id
  528. together with the machine name to generate the machine id for the container
  529. - fix logic always print a final newline on output.
  531. - should optionally support receiving WATCHDOG=1 messages from its payload
  532. PID 1...
  533. - optionally automatically add FORWARD rules to iptables whenever nspawn is
  534. running, remove them when shut down.
  535. - maybe make copying of /etc/resolv.conf optional, and skip it if --read-only
  536. is used
  537. * dissect
  538. - refuse mounting over a mount point
  539. - automatically discover .roothash files in dissect, similarly to nspawn
  540. * machined:
  541. - add an API so that libvirt-lxc can inform us about network interfaces being
  542. removed or added to an existing machine
  543. - "machinectl migrate" or similar to copy a container from or to a
  544. difference host, via ssh
  545. - introduce systemd-nspawn-ephemeral@.service, and hook it into
  546. "machinectl start" with a new --ephemeral switch
  547. - "machinectl status" should also show internal logs of the container in
  548. question
  549. - "machinectl list-images" should show os-release data, as well as
  550. machine-info data (including deployment level)
  551. - "machinectl history"
  552. - "machinectl diff"
  553. - "machinectl commit" that takes a writable snapshot of a tree, invokes a
  554. shell in it, and marks it read-only after use
  555. * importd:
  556. - generate a nice warning if mkfs.btrfs is missing
  557. * cryptsetup:
  558. - cryptsetup-generator: allow specification of passwords in crypttab itself
  559. - support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator
  560. * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it
  562. * create /sbin/init symlinks from the build system
  563. * add a dependency on standard-conf.xml and other included files to man pages
  564. * MountFlags=shared acts as MountFlags=slave right now.
  565. * properly handle loop back mounts via fstab, especially regards to fsck/passno
  566. * initialize the hostname from the fs label of /, if /etc/hostname does not exist?
  567. * udev:
  568. - move to LGPL
  569. - kill scsi_id
  570. - add trigger --subsystem-match=usb/usb_device device
  571. - reimport udev db after MOVE events for devices without dev_t
  572. * There's currently no way to cancel fsck (used to be possible via C-c or c on the console)
  573. * add option to sockets to avoid activation. Instead just drop packets/connections, see
  574. * coredump:
  575. - save coredump in Windows/Mozilla minidump format
  576. - when truncating coredumps, also log the full size that the process had, and make a metadata field so we can report truncated coredumps
  577. * support crash reporting operation modes (
  578. * default to actual 32-bit PIDs, via /proc/sys/kernel/pid_max
  579. * be able to specify a forced restart of service A where service B depends on, in case B
  580. needs to be auto-respawned?
  581. * tmpfiles:
  582. - apply "x" on "D" too (see patch from William Douglas)
  583. - replace F with f+.
  584. - instead of ignoring unknown fields, reject them.
  585. - creating new directories/subvolumes/fifos/device nodes
  586. should not follow symlinks. None of the other adjustment or creation
  587. calls follow symlinks.
  588. * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early
  589. * verify that the AF_UNIX sockets of a service in the fs still exist
  590. when we start a service in order to avoid confusion when a user
  591. assumes starting a service is enough to make it accessible
  592. * Make it possible to set the keymap independently from the font on
  593. the kernel cmdline. Right now setting one resets also the other.
  594. * and a dbus call to generate target from current state
  595. * write blog stories about:
  596. - hwdb: what belongs into it, lsusb
  597. - enabling dbus services
  598. - how to make changes to sysctl and sysfs attributes
  599. - remote access
  600. - how to pass throw-away units to systemd, or dynamically change properties of existing units
  601. - testing with Harald's awesome test kit
  602. - auto-restart
  603. - how to develop against journal browsing APIs
  604. - the journal HTTP iface
  605. - non-cgroup resource management
  606. - dynamic resource management with cgroups
  607. - refreshed, longer missions statement
  608. - calendar time events
  609. - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell
  610. - how to create your own target
  611. - instantiated apache, dovecot and so on
  612. - hooking a script into various stages of shutdown/rearly booot
  613. * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support.
  614. * dot output for --test showing the 'initial transaction'
  615. * pid1:
  616. - When logging about multiple units (stopping BoundTo units, conflicts, etc.),
  617. log both units as UNIT=, so that journalctl -u triggers on both.
  618. - generate better errors when people try to set transient properties
  619. that are not supported...
  621. - maybe introduce WantsMountsFor=? Usecase:
  623. - recreate systemd's D-Bus private socket file on SIGUSR2
  624. - move PAM code into its own binary
  625. - when we automatically restart a service, ensure we restart its rdeps, too.
  626. - hide PAM options in fragment parser when compile time disabled
  627. - Support --test based on current system state
  628. - 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().
  629. - after deserializing sockets in socket.c we should reapply sockopts and things
  630. - drop PID 1 reloading, only do reexecing (difficult: Reload()
  631. currently is properly synchronous, Reexec() is weird, because we
  632. cannot delay the response properly until we are back, so instead of
  633. being properly synchronous we just keep open the fd and close it
  634. when done. That means clients do not get a successful method reply,
  635. but much rather a disconnect on success.
  636. - when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr
  637. - when a bus name of a service disappears from the bus make sure to queue further activation requests
  638. * unit files:
  639. - allow port=0 in .socket units
  640. - maybe introduce ExecRestartPre=
  641. - add ReloadSignal= for configuring a reload signal to use
  642. - implement Register= switch in .socket units to enable registration
  643. in Avahi, RPC and other socket registration services.
  644. - allow Type=simple with PIDFile=
  646. - allow writing multiple conditions in unit files on one line
  647. - load-fragment: when loading a unit file via a chain of symlinks
  648. verify that it is not masked via any of the names traversed.
  649. - introduce Type=pid-file
  650. - introduce mix of BindTo and Requisite
  651. - add a concept of RemainAfterExit= to scope units
  652. - Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely
  653. * udev-link-config:
  654. - Make sure ID_PATH is always exported and complete for
  655. network devices where possible, so we can safely rely
  656. on Path= matching
  657. - check MTUBytes parsing (expecting size_t but we are using unsigned)
  658. * sd-rtnl:
  659. - add support for more attribute types
  660. - inbuilt piping support (essentially degenerate async)? see loopback-setup.c and other places
  661. * networkd:
  662. - add more keys to [Route] and [Address] sections
  663. - add support for more DHCPv4 options (and, longer term, other kinds of dynamic config)
  664. - add proper initrd support (in particular generate .network/.link files based on /proc/cmdline)
  665. - add reduced [Link] support to .network files
  666. - add Scope= parsing option for [Network]
  667. - properly handle routerless dhcp leases
  668. - work with non-Ethernet devices
  669. - add support for more bond options
  670. - dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from?
  671. - the DHCP lease data (such as NTP/DNS) is still made available when
  672. a carrier is lost on a link. It should be removed instantly.
  673. - expose in the API the following bits:
  674. - option 15, domain name and/or option 119, search list
  675. - option 12, host name and/or option 81, fqdn
  676. - option 123, 144, geolocation
  677. - option 252, configure http proxy (PAC/wpad)
  678. - provide a way to define a per-network interface default metric value
  679. for all routes to it. possibly a second default for DHCP routes.
  680. - allow Name= to be specified repeatedly in the [Match] section. Maybe also
  681. support Name=foo*|bar*|baz ?
  682. - duplicate address check for static IPs (like ARPCHECK in network-scripts)
  683. - allow DUID/IAID to be customized, see issue #394.
  684. - whenever uplink info changes, make DHCP server send out FORCERENEW
  685. * networkd-wait-online:
  686. - make operstates to wait for configurable?
  687. * dhcp:
  688. - figure out how much we can increase Maximum Message Size
  689. * dhcp6:
  690. - add functions to set previously stored IPv6 addresses on startup and get
  691. them at shutdown; store them in client->ia_na
  692. - write more test cases
  693. - implement reconfigure support, see 5.3., 15.11. and 22.20.
  694. - implement support for temporary adressess (IA_TA)
  695. - implement dhcpv6 authentication
  696. - investigate the usefulness of Confirm messages; i.e. are there any
  697. situations where the link changes without any loss in carrier detection
  698. or interface down
  699. - some servers don't do rapid commit without a filled in IA_NA, verify
  700. this behavior
  701. - RouteTable= ?
  702. External:
  703. * dbus:
  704. - natively watch for dbus-*.service symlinks (PENDING)
  705. - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service
  706. * fix alsa mixer restore to not print error when no config is stored
  707. * make cryptsetup lower --iter-time
  708. * patch kernel for xattr support in /dev, /proc/, /sys?
  709. * kernel: add device_type = "fb", "fbcon" to class "graphics"
  710. * drop accountsservice's StandardOutput=syslog and Type=dbus fields
  711. * /usr/bin/service should actually show the new command line
  712. * 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=
  713. * fedora: F20: go timer units all the way, leave cron.daily for cron
  714. * neither pkexec nor sudo initialize environ[] from the PAM environment?
  715. * fedora: update policy to declare access mode and ownership of unit files to root:root 0644, and add an rpmlint check for it
  716. * register catalog database signature as file magic
  717. * zsh shell completion:
  718. - <command> <verb> -<TAB> should complete options, but currently does not
  719. - systemctl add-wants,add-requires
  720. Regularly:
  721. * look for close() vs. close_nointr() vs. close_nointr_nofail()
  722. * check for strerror(r) instead of strerror(-r)
  723. * pahole
  724. * set_put(), hashmap_put() return values check. i.e. == 0 does not free()!
  725. * use secure_getenv() instead of getenv() where appropriate
  726. * link up selected blog stories from man pages and unit files Documentation= fields