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.
 
 
 
 
 
 

960 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. * Add AddUser= setting to unit files, similar to DynamicUser=1 which however
  17. creates a static, persistent user rather than a dynamic, transient user. We
  18. can leverage code from sysusers.d for this.
  19. * add some optional flag to ReadWritePaths= and friends, that has the effect
  20. that we create the dir in question when the service is started. Example:
  21. ReadWritePaths=:/var/lib/foobar
  22. * sort generated hwdb files alphabetically when we import them, so that git
  23. diffs remain minimal (in particular: the OUI databases we import are not
  24. sorted, and not stable)
  25. * set SystemCallArchitectures=native on all our services
  26. * maybe add call sd_journal_set_block_timeout() or so to set SO_SNDTIMEO for
  27. the sd-journal logging socket, and, if the timeout is set to 0, sets
  28. O_NONBLOCK on it. That way people can control if and when to block for
  29. logging.
  30. * tighten sd_notify() MAINPID= checks a bit: don't accept foreign PIDs (i.e.
  31. PIDs not managed by the service manager)
  32. * journald: when we recv a log datagram via the native or syslog transports,
  33. search for the PID in the active stream connections, and let's make sure to
  34. always process the datagrams before the streams. Then, cache client metadata
  35. per stream in the stream object. This way we can somewhat fix the race with
  36. quickly exiting processes which log as long as they had their own stream
  37. connection...
  38. * hostnamed: populate form factor data from a new hwdb database, so that old
  39. yogas can be recognized as "convertible" too, even if they predate the DMI
  40. "convertible" form factor
  41. * Maybe add a small tool invoked early at boot, that adds in or resizes
  42. partitions automatically, to be used when the media used is actually larger
  43. than the image written onto it is.
  44. * Maybe add PrivatePIDs= as new unit setting, and do minimal PID namespacing
  45. after all. Be strict however, only support the equivalent of nspawn's
  46. --as-pid2 switch, and sanely proxy sd_notify() messages dropping stuff such
  47. as MAINPID.
  48. * change the dependency Set* objects in Unit structures to become Hashmap*, and
  49. then store a bit mask who created a specific dependency: the source unit via
  50. fragment configuration, the destination unit via fragment configuration, or
  51. the source unit via udev rules (in case of .device units), or any combination
  52. thereof. This information can then be used to flush out old udev-created
  53. dependencies when the udev properties change, and eventually to implement a
  54. "systemctl refresh" operation for reloading the configuration of individual
  55. units without reloading the whole set.
  56. * Add ExecMonitor= setting. May be used multiple times. Forks off a process in
  57. the service cgroup, which is supposed to monitor the service, and when it
  58. exits the service is considered failed by its monitor.
  59. * track the per-service PAM process properly (i.e. as an additional control
  60. process), so that it may be queried on the bus and everything.
  61. * add a new "debug" job mode, that is propagated to unit_start() and for
  62. services results in two things: we raise SIGSTOP right before invoking
  63. execve() and turn off watchdog support. Then, use that to implement
  64. "systemd-gdb" for attaching to the start-up of any system service in its
  65. natural habitat.
  66. * replace all canonicalize_file_name() invocations by chase_symlinks(), in
  67. particulr those where a rootdir is relevant.
  68. * maybe introduce gpt auto discovery for /var/tmp?
  69. * set ProtectSystem=strict for all our usual services.
  70. * fix PrivateNetwork= so that we fall back gracefully on kernels lacking
  71. namespacing support (similar for the other namespacing options)
  72. * maybe add gpt-partition-based user management: each user gets his own
  73. LUKS-encrypted GPT partition with a new GPT type. A small nss module
  74. enumerates users via udev partition enumeration. UIDs are assigned in a fixed
  75. way: the partition index is added as offset to some fixed base uid. User name
  76. is stored in GPT partition name. A PAM module authenticates the user via the
  77. LUKS partition password. Benefits: strong per-user security, compatibility
  78. with stateless/read-only/verity-enabled root. (other idea: do this based on
  79. loopback files in /home, without GPT involvement)
  80. * gpt-auto logic: introduce support for discovering /var matching an image. For
  81. that, use a partition type UUID that is hashed from the OS name (as encoded
  82. in /etc/os-release), the architecture, and 4 new bits from the gpt flags
  83. field of the root partition. This way can easily support multiple OS
  84. installations on the same GPT partition table, without problems with
  85. unmatched /var partitions.
  86. * gpt-auto logic: related to the above, maybe support a "secondary" root
  87. partition, that is mounted to / and is writable, and where the actual root's
  88. /usr is mounted into.
  89. * machined: add apis to query /etc/machine-info data of a container
  90. * .mount and .swap units: add Format=yes|no option that formats the partition before mounting/enabling it, implicitly
  91. * 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
  92. * drop nss-myhostname in favour of nss-resolve?
  93. * drop internal dlopen() based nss-dns fallback in nss-resolve, and rely on the
  94. external nsswitch.conf based one
  95. * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and
  96. then use that for the setting used in user@.service. It should be understood
  97. relative to the configured default value.
  98. * on cgroupsv2 add DelegateControllers=, to pick the precise cgroup controllers to delegate
  99. * in networkd, when matching device types, fix up DEVTYPE rubbish the kernel passes to us
  100. * enable LockMLOCK to take a percentage value relative to physical memory
  101. * switch to ProtectSystem=strict for all our long-running services where that's possible
  102. * Permit masking specific netlink APIs with RestrictAddressFamily=
  103. * nspawn: start UID allocation loop from hash of container name
  104. * nspawn: support that /proc, /sys/, /dev are pre-mounted
  105. * define gpt header bits to select volatility mode
  106. * ProtectKernelLogs= (drops CAP_SYSLOG, add seccomp for syslog() syscall, and DeviceAllow to /dev/kmsg) in service files
  107. * ProtectClock= (drops CAP_SYS_TIMES, adds seecomp filters for settimeofday, adjtimex), sets DeviceAllow o /dev/rtc
  108. * ProtectTracing= (drops CAP_SYS_PTRACE, blocks ptrace syscall, makes /sys/kernel/tracing go away)
  109. * ProtectMount= (drop mount/umount/pivot_root from seccomp, disallow fuse via DeviceAllow, imply Mountflags=slave)
  110. * ProtectKeyRing= to take keyring calls away
  111. * RemoveKeyRing= to remove all keyring entries of the specified user
  112. * ProtectReboot= that masks reboot() and kexec_load() syscalls, prohibits kill
  113. on PID 1 with the relevant signals, and makes relevant files in /sys and
  114. /proc (such as the sysrq stuff) unavailable
  115. * DeviceAllow= should also generate seccomp filters for mknod()
  116. * Add DataDirectory=, CacheDirectory= and LogDirectory= to match
  117. RuntimeDirectory=, and create it as necessary when starting a service, owned by the right user.
  118. * make sure the ratelimit object can deal with USEC_INFINITY as way to turn off things
  119. * journalctl: make sure -f ends when the container indicated by -M terminates
  120. * mount: automatically search for "main" partition of an image has multiple
  121. partitions
  122. * expose the "privileged" flag of ExecCommand on the bus, and open it up to
  123. transient units
  124. * in nss-systemd, if we run inside of RootDirectory= with PrivateUsers= set,
  125. find a way to map the User=/Group= of the service to the right name. This way
  126. a user/group for a service only has to exist on the host for the right
  127. mapping to work.
  128. * allow attaching additional journald log fields to cgroups
  129. * add bus API for creating unit files in /etc, reusing the code for transient units
  130. * add bus API to remove unit files from /etc
  131. * add bus API to retrieve current unit file contents (i.e. implement "systemctl cat" on the bus only)
  132. * rework fopen_temporary() to make use of open_tmpfile_linkable() (problem: the
  133. kernel doesn't support linkat() that replaces existing files, currently)
  134. * check if DeviceAllow= should split first, resolve specifiers later
  135. * transient units: don't bother with actually setting unit properties, we
  136. reload the unit file anyway
  137. * journald: sigbus API via a signal-handler safe function that people may call
  138. from the SIGBUS handler
  139. * optionally, also require WATCHDOG=1 notifications during service start-up and shutdown
  140. * resolved: when routing queries, make sure only look for the *longest* suffix...
  141. * delay activation of logind until somebody logs in, or when /dev/tty0 pulls it
  142. in or lingering is on (so that containers don't bother with it until PAM is used). also exit-on-idle
  143. * cache sd_event_now() result from before the first iteration...
  144. * add systemctl stop --job-mode=triggering that follows TRIGGERED_BY deps and adds them to the same transaction
  145. * PID1: find a way how we can reload unit file configuration for
  146. specific units only, without reloading the whole of systemd
  147. * add an explicit parser for LimitRTPRIO= that verifies
  148. the specified range and generates sane error messages for incorrect
  149. specifications.
  150. * do something about "/control" subcgroups in the unified cgroup hierarchy
  151. * when we detect that there are waiting jobs but no running jobs, do something
  152. * push CPUAffinity= also into the "cpuset" cgroup controller (only after the cpuset controller got ported to the unified hierarchy)
  153. * PID 1 should send out sd_notify("WATCHDOG=1") messages (for usage in the --user mode, and when run via nspawn)
  154. * there's probably something wrong with having user mounts below /sys,
  155. as we have for debugfs. for exmaple, src/core/mount.c handles mounts
  156. prefixed with /sys generally special.
  157. http://lists.freedesktop.org/archives/systemd-devel/2015-June/032962.html
  158. * man: document that unless you use StandardError=null the shell >/dev/stderr won't work in shell scripts in services
  159. * fstab-generator: default to tmpfs-as-root if only usr= is specified on the kernel cmdline
  160. * docs: bring http://www.freedesktop.org/wiki/Software/systemd/MyServiceCantGetRealtime up to date
  161. * mounting and unmounting mount points manually with different source
  162. devices will result in collected on all devices used.
  163. http://lists.freedesktop.org/archives/systemd-devel/2015-April/030225.html
  164. * add a job mode that will fail if a transaction would mean stopping
  165. running units. Use this in timedated to manage the NTP service
  166. state.
  167. http://lists.freedesktop.org/archives/systemd-devel/2015-April/030229.html
  168. * Maybe add support for the equivalent of "ethtool advertise" to .link files?
  169. http://lists.freedesktop.org/archives/systemd-devel/2015-April/030112.html
  170. * The udev blkid built-in should expose a property that reflects
  171. whether media was sensed in USB CF/SD card readers. This should then
  172. be used to control SYSTEMD_READY=1/0 so that USB card readers aren't
  173. picked up by systemd unless they contain a medium. This would mirror
  174. the behaviour we already have for CD drives.
  175. * networkd/udev: implement SR_IOV configuration in .link files:
  176. http://lists.freedesktop.org/archives/systemd-devel/2015-January/027451.html
  177. * Rework systemctl's GetAll property parsing to use the generic bus_map_all_properties() API
  178. * implement a per-service firewall based on net_cls
  179. * Port various tools to make use of verbs.[ch], where applicable: busctl,
  180. coredumpctl, hostnamectl, localectl, systemd-analyze, timedatectl
  181. * hostnamectl: show root image uuid
  182. * sysfs set api in libudev is not const
  183. * Find a solution for SMACK capabilities stuff:
  184. http://lists.freedesktop.org/archives/systemd-devel/2014-December/026188.html
  185. * "systemctl preset-all" should probably order the unit files it
  186. operates on lexicographically before starting to work, in order to
  187. ensure deterministic behaviour if two unit files conflict (like DMs
  188. do, for example)
  189. * synchronize console access with BSD locks:
  190. http://lists.freedesktop.org/archives/systemd-devel/2014-October/024582.html
  191. * as soon as we have sender timestamps, revisit coalescing multiple parallel daemon reloads:
  192. http://lists.freedesktop.org/archives/systemd-devel/2014-December/025862.html
  193. * in systemctl list-unit-files: show the install value the presets would suggest for a service in a third column
  194. * figure out when we can use the coarse timers
  195. * add "systemctl start -v foobar.service" that shows logs of a service
  196. while the start command runs. This is non-trivial to do without
  197. races though, since we should flush out all journal messages before
  198. returning from the "systemctl stop".
  199. * 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
  200. * maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment=
  201. * introduce systemd-timesync-wait.service or so to sync on an NTP fix?
  202. * systemd --user should issue sd_notify() upon reaching basic.target, not on becoming idle
  203. * consider showing the unit names during boot up in the status output, not just the unit descriptions
  204. * maybe allow timer units with an empty Units= setting, so that they
  205. can be used for resuming the system but nothing else.
  206. * what to do about udev db binary stability for apps? (raw access is not an option)
  207. * maybe provide an API to allow migration of foreign PIDs into existing scopes.
  208. * man: maybe use the word "inspect" rather than "introspect"?
  209. * systemctl: if some operation fails, show log output?
  210. * systemctl edit: use equvalent of cat() to insert existing config as a comment, prepended with #.
  211. Upon editor exit, lines with one # are removed, lines with two # are left with one #, etc.
  212. * exponential backoff in timesyncd when we cannot reach a server
  213. * timesyncd: add ugly bus calls to set NTP servers per-interface, for usage by NM
  214. * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share....
  215. * systemd.show_status= should probably have a mode where only failed
  216. units are shown.
  217. * add systemd.abort_on_kill or some other such flag to send SIGABRT instead of SIGKILL
  218. (throughout the codebase, not only PID1)
  219. * resolved:
  220. - mDNS/DNS-SD
  221. - service registration
  222. - service/domain/types browsing
  223. - avahi compat
  224. - DNS-SD service registration from socket units
  225. - resolved should optionally register additional per-interface LLMNR
  226. names, so that for the container case we can establish the same name
  227. (maybe "host") for referencing the server, everywhere.
  228. - allow clients to request DNSSEC for a single lookup even if DNSSEC is off (?)
  229. - hook up resolved with machined-based address resolution
  230. * refcounting in sd-resolve is borked
  231. * Add a new verb "systemctl top"
  232. * add new gpt type for btrfs volumes
  233. * support empty /etc boots nicely:
  234. - nspawn/gpt-generator: introduce new gpt partition type for /usr
  235. * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them.
  236. * a way for container managers to turn off getty starting via $container_headless= or so...
  237. * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit
  238. * mount_cgroup_controllers(): symlinks need to get the label applied
  239. * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services
  240. they run added to the initial transaction and thus confuse Type=idle.
  241. * add bus api to query unit file's X fields.
  242. * gpt-auto-generator:
  243. - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap?
  244. - Make /home automount rather than mount?
  245. * add generator that pulls in systemd-network from containers when
  246. CAP_NET_ADMIN is set, more than the loopback device is defined, even
  247. when it is otherwise off
  248. * MessageQueueMessageSize= (and suchlike) should use parse_iec_size().
  249. * implement Distribute= in socket units to allow running multiple
  250. service instances processing the listening socket, and open this up
  251. for ReusePort=
  252. * socket units: support creating sockets in different namespace,
  253. opening it up for JoinsNamespaceOf=. This would require to fork off
  254. a tiny process that joins the namespace and creates/binds the socket
  255. and passes this back to PID1 via SCM_RIGHTS. This also could be used
  256. to allow Chown/chgrp on sockets without requiring NSS in PID 1.
  257. * introduce bus call FreezeUnit(s, b), as well as "systemctl freeze
  258. $UNIT" and "systemctl thaw $UNIT" as wrappers around this. The calls
  259. should SIGSTOP all unit processes in a loop until all processes of
  260. it are fully stopped. This can later be used for app management by
  261. desktop UIs such as gnome-shell to freeze apps that are not visible
  262. on screen, not unlike how job control works on the shell
  263. * cgroups:
  264. - implement per-slice CPUFairScheduling=1 switch
  265. - handle jointly mounted controllers correctly
  266. - introduce high-level settings for RT budget, swappiness
  267. - how to reset dynamically changed unit cgroup attributes sanely?
  268. - when reloading configuration, apply new cgroup configuration
  269. - when recursively showing the cgroup hierarchy, optionally also show
  270. the hierarchies of child processes
  271. * transient units:
  272. - add field to transient units that indicate whether systemd or somebody else saves/restores its settings, for integration with libvirt
  273. - ensure scope units may be started only a single time
  274. * Automatically configure swap partition to use for hibernation by looking for largest swap partition on the root disk?
  275. * when we detect low battery and no AC on boot, show pretty splash and refuse boot
  276. * libsystemd-journal, libsystemd-login, libudev: add calls to easily attach these objects to sd-event event loops
  277. * be more careful what we export on the bus as (usec_t) 0 and (usec_t) -1
  278. * 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
  279. * After coming back from hibernation reset hibernation swap partition using the /dev/snapshot ioctl APIs
  280. * If we try to find a unit via a dangling symlink, generate a clean
  281. error. Currently, we just ignore it and read the unit from the search
  282. path anyway.
  283. * refuse boot if /usr/lib/os-release is missing or /etc/machine-id cannot be set up
  284. * man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted.
  285. * load .d/*.conf dropins for device units
  286. * allow implementation of InaccessibleDirectories=/ plus
  287. ReadOnlyDirectories=... for whitelisting files for a service.
  288. * sd-bus:
  289. - EBADSLT handling
  290. - GetAllProperties() on a non-existing object does not result in a failure currently
  291. - kdbus: process fd=-1 for incoming msgs
  292. - port to sd-resolve for connecting to TCP dbus servers
  293. - kdbus: maybe add controlling tty metadata fields
  294. - 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
  295. - when kdbus does not take our message without memfds, try again with memfds
  296. - see if we can drop more message validation on the sending side
  297. - add API to clone sd_bus_message objects
  298. - make AddMatch calls on dbus1 transports async?
  299. - kdbus: matches against source or destination pids for an "strace -p"-like feel. Problem: The PID info needs to be available in userspace too...
  300. - longer term: priority inheritance
  301. - dbus spec updates:
  302. - kdbus mapping
  303. - NameLost/NameAcquired obsolete
  304. - GVariant
  305. - path escaping
  306. - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now
  307. - test bloom filter generation indexes
  308. - kdbus: introduce a concept of "send-only" connections
  309. - kdbus: add counter for refused unicast messages that is passed out via the RECV ioctl. SImilar to the counter for dropped multicast messages we already have.
  310. * sd-event
  311. - allow multiple signal handlers per signal?
  312. - document chaining of signal handler for SIGCHLD and child handlers
  313. - define more intervals where we will shift wakeup intervals around in, 1h, 6h, 24h, ...
  314. - generate a failure of a default event loop is executed out-of-thread
  315. - maybe add support for inotify events
  316. * investigate endianness issues of UUID vs. GUID
  317. * dbus: when a unit failed to load (i.e. is in UNIT_ERROR state), we
  318. should be able to safely try another attempt when the bus call LoadUnit() is invoked.
  319. * add a pam module that passes the hdd passphrase into the PAM stack and then expires it, for usage by gdm auto-login.
  320. * add a pam module that on password changes updates any LUKS slot where the password matches
  321. * maybe add a generator that looks for "systemd.run=" on the kernel cmdline for container usercases...
  322. * cgtop: make cgtop useful in a container
  323. * test/:
  324. - add 'set -e' to scripts in test/
  325. - make stuff in test/ work with separate output dir
  326. * seems that when we follow symlinks to units we prefer the symlink
  327. destination path over /etc and /usr. We should not do that. Instead
  328. /etc should always override /run+/usr and also any symlink
  329. destination.
  330. * when isolating, try to figure out a way how we implicitly can order
  331. all units we stop before the isolating unit...
  332. * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off})
  333. * BootLoaderSpec: Clarify that the kernel has to be in $BOOT. Clarify
  334. that the boot loader should be installed to the ESP. Define a way
  335. how an installer can figure out whether a BLS compliant boot loader
  336. is installed.
  337. * think about requeuing jobs when daemon-reload is issued? usecase:
  338. the initrd issues a reload after fstab from the host is accessible
  339. and we might want to requeue the mounts local-fs acquired through
  340. that automatically.
  341. * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep()
  342. * remove any syslog support from log.c — we probably cannot do this before split-off udev is gone for good
  343. * shutdown logging: store to EFI var, and store to USB stick?
  344. * think about window-manager-run-as-user-service problem: exit 0 → activate shutdown.target; exit != 0 → restart service
  345. * merge unit_kill_common() and unit_kill_context()
  346. * introduce ExecCondition= in services
  347. * EFI:
  348. - honor language efi variables for default language selection (if there are any?)
  349. - honor timezone efi variables for default timezone selection (if there are any?)
  350. - change bootctl to be backed by systemd-bootd to control temporary and persistent default boot goal plus efi variables
  351. * maybe do not install getty@tty1.service symlink in /etc but in /usr?
  352. * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word
  353. * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units.
  354. * logind:
  355. - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around
  356. - When we update the kernel all kind of hibernation should be prohibited until shutdown/reboot
  357. - logind: wakelock/opportunistic suspend support
  358. - Add pretty name for seats in logind
  359. - logind: allow showing logout dialog from system?
  360. - we should probably handle SIGTERM/SIGINT to not leave dot files around, just in case
  361. - session scopes/user unit: add RequiresMountsFor for the home directory of the user
  362. - add Suspend() bus calls which take timestamps to fix double suspend issues when somebody hits suspend and closes laptop quickly.
  363. - if pam_systemd is invoked by su from a process that is outside of a
  364. any session we should probably just become a NOP, since that's
  365. usually not a real user session but just some system code that just
  366. needs setuid().
  367. - logind: make the Suspend()/Hibernate() bus calls wait for the for
  368. the job to be completed. before returning, so that clients can wait
  369. for "systemctl suspend" to finish to know when the suspending is
  370. complete.
  371. - logind: when the power button is pressed short, just popup a
  372. logout dialog. If it is pressed for 1s, do the usual
  373. shutdown. Inspiration are Macs here.
  374. - expose "Locked" property on logind sesison objects
  375. - given that logind now lets PID 1 do all nasty work, we can
  376. probably reduce the capability set it retains substantially.
  377. (we need CAP_SYS_ADMIN for drmSetMaster(), so maybe not worth it)
  378. - expose orientation sensors and tablet mode through logind
  379. - maybe allow configuration of the StopTimeout for session scopes
  380. - rename session scope so that it includes the UID. THat way
  381. the session scope can be arranged freely in slices and we don't have
  382. make assumptions about their slice anymore.
  383. - follow PropertiesChanged state more closely, to deal with quick logouts and
  384. relogins
  385. * exec: when deinitializating a tty device fix the perms and group, too, not only when initializing. Set access mode/gid to 0620/tty.
  386. * service: watchdog logic: for testing purposes allow ping, but do not require pong
  387. * journal:
  388. - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields
  389. - import and delete pstore filesystem content at startup
  390. - journald: also get thread ID from client, plus thread name
  391. - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups
  392. - add API to close/reopen/get fd for journal client fd in libsystemd-journal.
  393. - fallback to /dev/log based logging in libsystemd-journal, if we cannot log natively?
  394. - declare the local journal protocol stable in the wiki interface chart
  395. - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg
  396. - journald: when dropping msgs due to ratelimit make sure to write
  397. "dropped %u messages" not only when we are about to print the next
  398. message that works, but alraedy after a short tiemout
  399. - check if we can make journalctl by default use --follow mode inside of less if called without args?
  400. - maybe add API to send pairs of iovecs via sd_journal_send
  401. - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access
  402. - journactl: support negative filtering, i.e. FOOBAR!="waldo",
  403. and !FOOBAR for events without FOOBAR.
  404. - journal: store timestamp of journal_file_set_offline() int he header,
  405. so it is possible to display when the file was last synced.
  406. - 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.
  407. - journal: find a way to allow dropping history early, based on priority, other rules
  408. - journal: When used on NFS, check payload hashes
  409. - journald: add kernel cmdline option to disable ratelimiting for debug purposes
  410. - refuse taking lower-case variable names in sd_journal_send() and friends.
  411. - journald: we currently rotate only after MaxUse+MaxFilesize has been reached.
  412. - journal: deal nicely with byte-by-byte copied files, especially regards header
  413. - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit
  414. - Replace utmp, wtmp, btmp, and lastlog completely with journal
  415. - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax?
  416. - when a kernel driver logs in a tight loop, we should ratelimit that too.
  417. - journald: optionally, log debug messages to /run but everything else to /var
  418. - journald: when we drop syslog messages because the syslog socket is
  419. full, make sure to write how many messages are lost as first thing
  420. to syslog when it works again.
  421. - journald: make sure ratelimit is actually really per-service with the new cgroup changes
  422. - change systemd-journal-flush into a service that stays around during
  423. boot, and causes the journal to be moved back to /run on shutdown,
  424. so that we do not keep /var busy. This needs to happen synchronously,
  425. hence doing this via signals is not going to work.
  426. - optionally support running journald from the command line for testing purposes in external projects
  427. - journald: allow per-priority and per-service retention times when rotating/vacuuming
  428. - journald: make use of uid-range.h to managed uid ranges to split
  429. journals in.
  430. - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so...
  431. - improve journalctl performance by loading journal files
  432. lazily. Encode just enough information in the file name, so that we
  433. do not have to open it to know that it is not interesting for us, for
  434. the most common operations.
  435. - journal-or-kmsg is currently broken? See reverted
  436. commit 4a01181e460686d8b4a543b1dfa7f77c9e3c5ab8.
  437. - man: document that corrupted journal files is nothing to act on
  438. - rework journald sigbus stuff to use mutex
  439. - Set RLIMIT_NPROC for systemd-journal-xyz, and all other of our
  440. services that run under their own user ids, and use User= (but only
  441. in a world where userns is ubiquitous since otherwise we cannot
  442. invoke those daemons on the host AND in a container anymore). Also,
  443. if LimitNPROC= is used without User= we should warn and refuse
  444. operation.
  445. - journalctl --verify: don't show files that are currently being
  446. written to as FAIL, but instead show that their are being written to.
  447. - add journalctl -H that talks via ssh to a remote peer and passes through
  448. binary logs data
  449. - add a version of --merge which also merges /var/log/journal/remote
  450. - log accumulated resource usage after each service invocation
  451. - journalctl: -m should access container journals directly by enumerating
  452. them via machined, and also watch containers coming and going.
  453. Benefit: nspawn --ephemeral would start working nicely with the journal.
  454. - assign MESSAGE_ID to log messages about failed services
  455. * document:
  456. - document that deps in [Unit] sections ignore Alias= fields in
  457. [Install] units of other units, unless those units are disabled
  458. - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets
  459. - document the exit codes when services fail before they are exec()ed
  460. - document that service reload may be implemented as service reexec
  461. - document in wiki how to map ical recurrence events to systemd timer unit calendar specifications
  462. - add a man page containing packaging guidelines and recommending usage of things like Documentation=, PrivateTmp=, PrivateNetwork= and ReadOnlyDirectories=/etc /usr.
  463. - document systemd-journal-flush.service properly
  464. - documentation: recommend to connect the timer units of a service to the service via Also= in [Install]
  465. - man: document the very specific env the shutdown drop-in tools live in
  466. - man: add more examples to man pages
  467. - man: maybe sort directives in man pages, and take sections from --help and apply them to man too
  468. * systemctl:
  469. - add systemctl switch to dump transaction without executing it
  470. - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done
  471. - "systemctl disable" on a static unit prints no message and does
  472. nothing. "systemctl enable" does nothing, and gives a bad message
  473. about it. Should fix both to print nice actionable messages.
  474. - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service
  475. - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible
  476. - systemctl enable: fail if target to alias into does not exist? maybe show how many units are enabled afterwards?
  477. - systemctl: "Journal has been rotated since unit was started." message is misleading
  478. - better error message if you run systemctl without systemd running
  479. - systemctl status output should include list of triggering units and their status
  480. * unit install:
  481. - "systemctl mask" should find all names by which a unit is accessible
  482. (i.e. by scanning for symlinks to it) and link them all to /dev/null
  483. * timer units:
  484. - timer units should get the ability to trigger when:
  485. o CLOCK_REALTIME makes jumps (TFD_TIMER_CANCEL_ON_SET)
  486. o DST changes
  487. - Modulate timer frequency based on battery state
  488. * add libsystemd-password or so to query passwords during boot using the password agent logic
  489. * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed
  490. * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel
  491. * make repeated alt-ctrl-del presses printing a dump
  492. * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ...
  493. * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab is not
  494. * nspawn:
  495. - emulate /dev/kmsg using CUSE and turn off the syslog syscall
  496. with seccomp. That should provide us with a useful log buffer that
  497. systemd can log to during early boot, and disconnect container logs
  498. from the kernel's logs.
  499. - as soon as networkd has a bus interface, hook up --network-interface=,
  500. --network-bridge= with networkd, to trigger netdev creation should an
  501. interface be missing
  502. - a nice way to boot up without machine id set, so that it is set at boot
  503. automatically for supporting --ephemeral. Maybe hash the host machine id
  504. together with the machine name to generate the machine id for the container
  505. - fix logic always print a final newline on output.
  506. https://github.com/systemd/systemd/pull/272#issuecomment-113153176
  507. - should optionally support receiving WATCHDOG=1 messages from its payload
  508. PID 1...
  509. - should send out sd_notify("WATCHDOG=1") messages
  510. - optionally automatically add FORWARD rules to iptables whenever nspawn is
  511. running, remove them when shut down.
  512. - Improve error message when --bind= is used on a non-existing source
  513. directory
  514. - maybe make copying of /etc/resolv.conf optional, and skip it if --read-only
  515. is used
  516. * dissect
  517. - refuse mounting over a mount point
  518. - automatically discover .roothash files in dissect, similarly to nspawn
  519. * machined:
  520. - add an API so that libvirt-lxc can inform us about network interfaces being
  521. removed or added to an existing machine
  522. - "machinectl migrate" or similar to copy a container from or to a
  523. difference host, via ssh
  524. - introduce systemd-nspawn-ephemeral@.service, and hook it into
  525. "machinectl start" with a new --ephemeral switch
  526. - "machinectl status" should also show internal logs of the container in
  527. question
  528. - "machinectl list-images" should show os-release data, as well as
  529. machine-info data (including deployment level)
  530. - "machinectl history"
  531. - "machinectl diff"
  532. - "machinectl commit" that takes a writable snapshot of a tree, invokes a
  533. shell in it, and marks it read-only after use
  534. * importd:
  535. - generate a nice warning if mkfs.btrfs is missing
  536. * cryptsetup:
  537. - cryptsetup-generator: allow specification of passwords in crypttab itself
  538. - move cryptsetup key caching into kernel keyctl?
  539. https://bugs.freedesktop.org/show_bug.cgi?id=54982
  540. - support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator
  541. * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it
  542. https://bugs.freedesktop.org/show_bug.cgi?id=54712
  543. * create /sbin/init symlinks from the build system
  544. * MountFlags=shared acts as MountFlags=slave right now.
  545. * properly handle loop back mounts via fstab, especially regards to fsck/passno
  546. * initialize the hostname from the fs label of /, if /etc/hostname does not exist?
  547. * udev:
  548. - move to LGPL
  549. - kill scsi_id
  550. - add trigger --subsystem-match=usb/usb_device device
  551. - reimport udev db after MOVE events for devices without dev_t
  552. * 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
  553. * There's currently no way to cancel fsck (used to be possible via C-c or c on the console)
  554. * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/
  555. * coredump:
  556. - save coredump in Windows/Mozilla minidump format
  557. - when truncating coredumps, also log the full size that the process had, and make a metadata field so we can report truncated coredumps
  558. * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting)
  559. * default to actual 32-bit PIDs, via /proc/sys/kernel/pid_max
  560. * be able to specify a forced restart of service A where service B depends on, in case B
  561. needs to be auto-respawned?
  562. * tmpfiles:
  563. - apply "x" on "D" too (see patch from William Douglas)
  564. - replace F with f+.
  565. - instead of ignoring unknown fields, reject them.
  566. - creating new directories/subvolumes/fifos/device nodes
  567. should not follow symlinks. None of the other adjustment or creation
  568. calls follow symlinks.
  569. * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early
  570. * verify that the AF_UNIX sockets of a service in the fs still exist
  571. when we start a service in order to avoid confusion when a user
  572. assumes starting a service is enough to make it accessible
  573. * Make it possible to set the keymap independently from the font on
  574. the kernel cmdline. Right now setting one resets also the other.
  575. * and a dbus call to generate target from current state
  576. * write blog stories about:
  577. - hwdb: what belongs into it, lsusb
  578. - enabling dbus services
  579. - status update
  580. - how to make changes to sysctl and sysfs attributes
  581. - remote access
  582. - how to pass throw-away units to systemd, or dynamically change properties of existing units
  583. - testing with Harald's awesome test kit
  584. - auto-restart
  585. - how to develop against journal browsing APIs
  586. - the journal HTTP iface
  587. - non-cgroup resource management
  588. - dynamic resource management with cgroups
  589. - refreshed, longer missions statement
  590. - calendar time events
  591. - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell
  592. - how to create your own target
  593. - instantiated apache, dovecot and so on
  594. - hooking a script into various stages of shutdown/rearly booot
  595. * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support.
  596. * dot output for --test showing the 'initial transaction'
  597. * fingerprint.target, wireless.target, gps.target, netdevice.target
  598. * pid1:
  599. - .timer units should optionally support CLOCK_BOOTTIME in addition to CLOCK_MONOTONIC
  600. - When logging about multiple units (stopping BoundTo units, conflicts, etc.),
  601. log both units as UNIT=, so that journalctl -u triggers on both.
  602. - generate better errors when people try to set transient properties
  603. that are not supported...
  604. http://lists.freedesktop.org/archives/systemd-devel/2015-February/028076.html
  605. - maybe introduce WantsMountsFor=? Usecase:
  606. http://lists.freedesktop.org/archives/systemd-devel/2015-January/027729.html
  607. - recreate systemd's D-Bus private socket file on SIGUSR2
  608. - move PAM code into its own binary
  609. - when we automatically restart a service, ensure we restart its rdeps, too.
  610. - hide PAM options in fragment parser when compile time disabled
  611. - Support --test based on current system state
  612. - 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().
  613. - after deserializing sockets in socket.c we should reapply sockopts and things
  614. - drop PID 1 reloading, only do reexecing (difficult: Reload()
  615. currently is properly synchronous, Reexec() is weird, because we
  616. cannot delay the response properly until we are back, so instead of
  617. being properly synchronous we just keep open the fd and close it
  618. when done. That means clients do not get a successful method reply,
  619. but much rather a disconnect on success.
  620. - when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr
  621. - when a bus name of a service disappears from the bus make sure to queue further activation requests
  622. * unit files:
  623. - allow port=0 in .socket units
  624. - maybe introduce ExecRestartPre=
  625. - add ReloadSignal= for configuring a reload signal to use
  626. - implement Register= switch in .socket units to enable registration
  627. in Avahi, RPC and other socket registration services.
  628. - allow Type=simple with PIDFile=
  629. https://bugzilla.redhat.com/show_bug.cgi?id=723942
  630. - allow writing multiple conditions in unit files on one line
  631. - load-fragment: when loading a unit file via a chain of symlinks
  632. verify that it is not masked via any of the names traversed.
  633. - introduce Type=pid-file
  634. - ExecOnFailure=/usr/bin/foo
  635. - introduce mix of BindTo and Requisite
  636. - add a concept of RemainAfterExit= to scope units
  637. - Set NoNewPrivileges= on all of our own services, where that makes sense
  638. - Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely
  639. - consider adding RuntimeDirectoryUser= + RuntimeDirectoryGroup=
  640. * udev-link-config:
  641. - Make sure ID_PATH is always exported and complete for
  642. network devices where possible, so we can safely rely
  643. on Path= matching
  644. - check MTUBytes parsing (expecting size_t but we are using unsigned)
  645. * sd-rtnl:
  646. - add support for more attribute types
  647. - inbuilt piping support (essentially degenerate async)? see loopback-setup.c and other places
  648. * networkd:
  649. - add more keys to [Route] and [Address] sections
  650. - add support for more DHCPv4 options (and, longer term, other kinds of dynamic config)
  651. - add proper initrd support (in particular generate .network/.link files based on /proc/cmdline)
  652. - add reduced [Link] support to .network files
  653. - add Scope= parsing option for [Network]
  654. - properly handle routerless dhcp leases
  655. - work with non-Ethernet devices
  656. - add support for more bond options
  657. - dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from?
  658. - the DHCP lease data (such as NTP/DNS) is still made available when
  659. a carrier is lost on a link. It should be removed instantly.
  660. - expose in the API the following bits:
  661. - option 15, domain name and/or option 119, search list
  662. - option 12, host name and/or option 81, fqdn
  663. - option 123, 144, geolocation
  664. - option 252, configure http proxy (PAC/wpad)
  665. - provide a way to define a per-network interface default metric value
  666. for all routes to it. possibly a second default for DHCP routes.
  667. - allow Name= to be specified repeatedly in the [Match] section. Maybe also
  668. support Name=foo*|bar*|baz ?
  669. - duplicate address check for static IPs (like ARPCHECK in network-scripts)
  670. - allow DUID/IAID to be customized, see issue #394.
  671. - whenever uplink info changes, make DHCP server send out FORCERENEW
  672. * networkd-wait-online:
  673. - make operstates to wait for configurable?
  674. * dhcp:
  675. - figure out how much we can increase Maximum Message Size
  676. - support RFC4702 (pass FQDN)
  677. * dhcp6:
  678. - add functions to set previously stored IPv6 addresses on startup and get
  679. them at shutdown; store them in client->ia_na
  680. - write more test cases
  681. - implement reconfigure support, see 5.3., 15.11. and 22.20.
  682. - implement support for temporary adressess (IA_TA)
  683. - implement dhcpv6 authentication
  684. - investigate the usefulness of Confirm messages; i.e. are there any
  685. situations where the link changes without any loss in carrier detection
  686. or interface down
  687. - some servers don't do rapid commit without a filled in IA_NA, verify
  688. this behavior
  689. - RouteTable= ?
  690. External:
  691. * dbus:
  692. - natively watch for dbus-*.service symlinks (PENDING)
  693. - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service
  694. * fix alsa mixer restore to not print error when no config is stored
  695. * make cryptsetup lower --iter-time
  696. * patch kernel for xattr support in /dev, /proc/, /sys?
  697. * kernel: add device_type = "fb", "fbcon" to class "graphics"
  698. * drop accountsservice's StandardOutput=syslog and Type=dbus fields
  699. * dbus: in fedora, make /var/lib/dbus/machine-id a symlink to /etc/machine-id
  700. * /usr/bin/service should actually show the new command line
  701. * 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=
  702. * fedora: F20: go timer units all the way, leave cron.daily for cron
  703. * neither pkexec nor sudo initialize environ[] from the PAM environment?
  704. * fedora: update policy to declare access mode and ownership of unit files to root:root 0644, and add an rpmlint check for it
  705. * register catalog database signature as file magic
  706. * zsh shell completion:
  707. - <command> <verb> -<TAB> should complete options, but currently does not
  708. - systemctl add-wants,add-requires
  709. Regularly:
  710. * look for close() vs. close_nointr() vs. close_nointr_nofail()
  711. * check for strerror(r) instead of strerror(-r)
  712. * Use PR_SET_PROCTITLE_AREA if it becomes available in the kernel
  713. * pahole
  714. * set_put(), hashmap_put() return values check. i.e. == 0 does not free()!
  715. * use secure_getenv() instead of getenv() where appropriate
  716. * link up selected blog stories from man pages and unit files Documentation= fields