3 * Many manager configuration settings that are only applicable to user
4 manager or system manager can be always set. It would be better to reject
5 them when parsing config.
9 * Fedora: add an rpmlint check that verifies that all unit files in the RPM are listed in %systemd_post macros.
13 * Rearrange tests so that the various test-xyz.c match a specific src/basic/xyz.c again
15 * rework mount.c and swap.c to follow proper state enumeration/deserialization
16 semantics, like we do for device.c now
20 * repart: support setting up dm-integrity with HMAC
22 * add /etc/integritytab, to support dm-integrity setups. In particular those
23 with HMAC as hash function, so that we can have a protected /home without
24 encryption (leaving encryption to the individual dirs/homed).
26 * complement root=, rootflags=, rootfstype= with rootsubdir= which allows
27 mounting a subdir of the root fs as actual root. This can be used as
28 fstype-agnostic version of btrfs' rootflags=subvol=foobar.
30 * add --copy-from and --copy-to command to systemd-dissect which copies stuff
31 in and out of a disk image
33 * Support ProtectProc= or so, using: https://patchwork.kernel.org/cover/11310197/
35 * if /usr/bin/swapoff fails due to OOM, log a friendly explanatory message about it
37 * build short web pages out of each catalog entry, build them along with man
38 pages, and include hyperlinks to them in the journal output
40 * machined: add API to acquire UID range. add API to mount/dissect loopback
41 file. Both protected by PK. Then make nspawn use these APIs to run
42 unprivileged containers. i.e. push the truly privileged bits into machined,
43 so that the client side can remain entirely unprivileged, with SUID or
46 * journald: do journal file writing out-of-process, with one writer process per
47 client UID, so that synthetic hash table collisions can slow down a specific
48 user's journal stream down but not the others.
50 * add "throttling" to sd-event event sources: optionally, when we wake up too
51 often for one, let's turn it off entirely for a while. Use that for the
52 /proc/self/mountinfo logic.
54 * move our systemd-user PAM snippet to /usr/, which PAM appears to support
57 * nspawn: support time namespaces
59 * systemd-firstboot: make sure to always use chase_symlinks() before
62 * add ConditionSecurity=tpm2
64 * Remove any support for booting without /usr pre-mounted in the initrd entirely.
65 Update INITRD_INTERFACE.md accordingly.
67 * pid1: Move to tracking of main pid/control pid of units per pidfd
69 * pid1: support new clone3() fork-into-cgroup feature
71 * pid1: also remove PID files of a service when the service starts, not just
74 * make us use dynamically fewer deps for containers in general purpose distros:
75 o turn into dlopen() deps:
76 - pcre2 (always) — irrelevant on Fedora, since dep by
77 libselinux, but should benefit Debian
78 - libpwquality (always) - only relevant for homed, and maybe soon
81 - p11-kit-trust (always)
82 - kmod-libs (only when called from PID 1)
83 - cryptsetup-libs (only in RootImage= handling in PID 1, but not in systemd-cryptsetup)
85 - libpam (only when called from PID 1)
86 - bzip2, xz, lz4 (always — gzip and zstd should probably stay static deps the way they are,
87 since they are so basic and our defaults)
88 o move into separate libsystemd-shared-iptables.so .so
89 - iptables-libs (only used by nspawn + networkd)
91 * seccomp: when SystemCallArchitectures=native is set then don't install any
92 other seccomp filters for any of the other archs, in order to reduce the
93 number of seccomp filters we install needlessly.
95 * seccomp: maybe use seccomp_merge() to merge our filters per-arch if we can.
96 Apparently kernel performance is much better with fewer larger seccomp
97 filters than with more smaller seccomp filters.
99 * systemd-path: add ESP and XBOOTLDR path. Add "private" runtime/state/cache dir enum,
100 mapping to $RUNTIME_DIRECTORY, $STATE_DIRECTORY and such
102 * make "systemd-dissect" an official supported tool, i.e. move to /usr/bin/ and
103 provide man page. Given that we now have a tool that can generate images like
104 this, it's useful to have one that can dump contents of them, too.
106 * All tools that support --root= should also learn --image= so that they can
107 operate on disk images directly. Specifically: bootctl, tmpfiles, sysusers,
108 systemctl, repart, journalctl, coredumpctl. (Already done: systemd-nspawn,
111 * seccomp: by default mask x32 ABI system wide on x86-64. it's on its way out
113 * seccomp: don't install filters for ABIs that are masked anyway for the
116 * seccomp: maybe merge all filters we install into one with that libseccomp API that allows merging.
118 * per-service credential system. Specifically: add LoadCredential= (for loading
119 cred from file), AcquireCredential= (for asking user for cred, via
120 ask-password), PassCredential= (for passing on credential systemd itself
121 got). Then, place credentials in a per-service, immutable ramfs instance (so
122 that it cannot be swapped out), destroy after use. Also pass via keyring
123 (with graceful fallback to cover for containers). Define CredentialPath= for
124 defining subdir of /run/credentials/ where to place it. Set $CREDENTIAL_PATH
125 env var for services to the result. Also pass via fd passing (optionally).
127 * homed: add native recovery key support. use 48 lowercase modhex characters
128 (192bit), show qr code of it, include pattern expression in user record.
130 * homed: introduce "degraded" state for home directories that weren't cleanly
131 unmounted (use xattr we add and remove on the loop back file)
133 * homed: during login resize fs automatically towards size goal. Specifically,
134 resize to diskSize if possible, but leave a certain amount (configured by a
135 new value diskLeaveFreeSize) of space free on the backing fs.
137 * homed: permit multiple user record signing keys to be used locally, and pick
138 the right one for signing records automatically depending on a pre-existing
141 * homed: add a way to "adopt" a home directory, i.e. strip foreign signatures
142 and insert a local signature instead.
144 * homed: as an extension to the directory+subvolume backend: if located on
145 especially marked fs, then sync down password into LUKS header of that fs,
146 and always verify passwords against it too. Bootstrapping is a problem
147 though: if no one is logged in (or no other user even exists yet), how do you
148 unlock the volume in order to create the first user and add the first pw.
150 * homed: support new FS_IOC_ADD_ENCRYPTION_KEY ioctl for setting up fscrypt
152 * homed: maybe pre-create ~/.cache as subvol so that it can have separate quota
155 * busctl: maybe expose a verb "ping" for pinging a dbus service to see if it
158 * when systemd-nspawn and suchlike dissect an OS image, and there are multiple
159 root partitions, do an strverscmp() on the partition label and boot
160 first. That is inspired how sd-boot figures out which kernel to boot, and
161 thus allows defining OS images which can be A/B updated and we default to the
162 newest version automatically, both in nspawn and in sd-boot
164 * cryptsetup: support FIDO2 tokens for deriving keys (i.e. do what homed can do
165 also in plain cryptsetup)
167 * systemd-gpt-auto should probably set x-systemd.growfs on the mounts it
170 * homed/userdb: distinguish passwords and recovery keys in the records, since
171 we probably want to use different PBKDF algorithms/settings for them:
172 passwords have low entropy but recovery keys should have good entropy key
173 hence we can make them quicker to work.
176 - teach it to prepare an ESP wholesale, i.e. with mkfs.vfat invocation
177 - teach it to copy in unified kernel images and maybe type #1 boot loader spec entries from host
178 - make it operate on loopback files, dissecting enough to find ESP to operate on
180 * Maybe add a separate GPT partition type to the discoverable partition spec
181 for "hibernate" partitions, that are exactly like swap partitions but only
182 activated right before hibernation and thus never used for regular swapping.
184 * by default, in systemd --user service bump the OOMAdjust to 100, as privs
185 allow so that systemd survives
187 * cryptsetup: allow encoding key directly in /etc/crypttab, maybe with a
188 "base64:" prefix. Useful in particular for pkcs11 mode.
190 * cryptsetup: reimplement the mkswap/mke2fs in cryptsetup-generator to use
191 systemd-makefs.service instead.
193 * socket units: allow creating a udev monitor socket with ListenDevices= or so,
194 with matches, then activate app through that passing socket over
196 * unify on openssl (as soon as OpenSSL 3.0 is out, and the Debian license
198 - port sd_id128_get_machine_app_specific() over from khash
199 - port resolved over from libgcrypt (DNSSEC code)
200 - port journald + fsprg over from libgcrypt
201 - port importd over from libgcrypt
202 - when that's done: kill khash.c
203 - when that's done: kill gnutls support in resolved
205 * kill zenata, all hail weblate?
207 * when we resize disks (homed?) always round up to 4K sectors, not 512K
209 * add growvol and makevol options for /etc/crypttab, similar to
210 x-systemd.growfs and x-systemd-makefs.
212 * hook up the TPM to /etc/crypttab, with a new option that is similar to the
213 new PKCS#11 option in crypttab, and allows unlocking a LUKS volume via a key
214 unsealed from the TPM. Optionally, if TPM is not available fall back to
215 TPM-less mode, and set up linear DM mapping instead (inspired by kpartx), so
216 that the device paths stay the same, regardless if crypto is used or not.
218 * systemd-repart: by default generate minimized partition tables (i.e. tables
219 that only cover the space actually used, excluding any free space at the
220 end), in order to maximize dd'ability. Requires libfdisk work, see
221 https://github.com/karelzak/util-linux/issues/907
223 * systemd-repart: MBR partition table support. Care needs to be taken regarding
224 Type=, so that partition definitions can sanely apply to both the GPT and the
225 MBR case. Idea: accept syntax "Type=gpt:home mbr:0x83" for setting the types
226 for the two partition types explicitly. And provide an internal mapping so
227 that "Type=linux-generic" maps to the right types for both partition tables
230 * systemd-repart: allow sizing partitions as factor of available RAM, so that
231 we can reasonably size swap partitions for hibernation.
233 * systemd-repart: allow running mkfs before making partitions pop up +
234 encryption via LUKS to allow booting into an empty root with only /usr mounted in
236 * systemd-repart: allow managing the gpt read-only partition flag + auto-mount flag
238 * systemd-repart: allow boolean option that ensures that if existing partition
239 doesn't exist within the configured size bounds the whole command fails. This
240 is useful to implement ESP vs. XBOOTLDR schemes in installers: have one set
241 of repart files for the case where ESP is large enough and one where it isn't
242 and XBOOTLDR is added in instead. Then apply the former first, and if it
243 fails to apply use the latter.
245 * systemd-repart: add per-partition option to never reuse existing partition
246 and always create anew even if matching partition already exists.
248 * systemd-repart: add per-partition option to fail if partition already exist,
249 i.e. is not added new. Similar, add option to fail if partition does not exist yet.
251 * systemd-repart: add --size=auto for generating/resizing images of minimal
252 size, i.e. where the image file is sized exactly as large as necessary taking
253 SizeMin= into account, but not a single byte larger.
255 * systemd-repart: allow disabling growing of specific partitions, or making
256 them (think ESP: we don't ever want to grow it, since we cannot resize vfat)
258 * systemd-repart: make it a static checker during early boot for existence and
259 absence of other partitions for trusted boot environments
261 * userdb: allow username prefix searches in varlink API, allow realname and
262 realname substr searches in varlink API
264 * userdb: allow uid/gid range checks
266 * userdb: allow existence checks
268 * pid1: activation by journal search expression
270 * when switching root from initrd to host, set the machine_id env var so that
271 if the host has no machine ID set yet we continue to use the random one the
274 * sd-event: add native support for P_ALL waitid() watching, then move PID 1 to
275 it fo reaping assigned but unknown children. This needs to some special care
276 to operate somewhat sensibly in light of priorities: P_ALL will return
277 arbitrary processes, regardless of the priority we want to watch them with,
278 hence on each event loop iteration check all processes which we shall watch
279 with higher prio explicitly, and then watch the entire rest with P_ALL.
281 * tweak sd-event's child watching: keep a prioq of children to watch and use
282 waitid() only on the children with the highest priority until one is waitable
283 and ignore all lower-prio ones from that point on
285 * maybe introduce xattrs that can be set on the root dir of the root fs
286 partition that declare the volatility mode to use the image in. Previously I
287 thought marking this via GPT partition flags but that's not ideal since
288 that's outside of the LUKS encryption/verity verification, and we probably
289 shouldn't operate in a volatile mode unless we got told so from a trusted
292 * figure out automatic partition discovery when combining writable root dir
295 * coredump: maybe when coredumping read a new xattr from /proc/$PID/exe that
296 may be used to mark a whole binary as non-coredumpable. Would fix:
297 https://bugs.freedesktop.org/show_bug.cgi?id=69447
299 * teach parse_timestamp() timezones like the calendar spec already knows it
301 * beef up hibernation to optionally do swapon/swapoff immediately before/after
304 * beef up s2h to implement a battery watch loop: instead of entering
305 hibernation unconditionally after coming back from resume make a decision
306 based on the battery load level: if battery level is above a specific
307 threshold, go to suspend again, only hibernate if below it. This means we'd
308 stick to suspend usually, but fall back to hibernation only when battery runs
309 empty (well, subject to our sampling interval). Related to this, check if we
310 can make ACPI _BTP (i.e. /sys/class/power_supply/*/alarm) work for us too,
311 i.e. see if it can wake up machines from suspend, so that we could resume
312 automatically when the system is low on power and move automatically to
313 hibernation mode. (see
314 https://uefi.org/sites/default/files/resources/ACPI%206_2_A_Sept29.pdf
316 https://docs.microsoft.com/en-us/windows-hardware/design/device-experiences/modern-standby-wake-sources
319 * add an explicit "vertical" mode to format-table, so that "systemctl
320 status"-like outputs (i.e. with a series of field names left and values
321 right) become genuine first class citizens, and we gain automatic, sane JSON
324 * systemd-firstboot: teach it dissector magic, so that you can point it to some
325 disk image and it will just set everything in it all behind the scenes.
327 * We should probably replace /var/log/README, /etc/rc.d/README with symlinks
328 that are linked to these places instead of copied. After all they are
329 constant vendor data.
331 * maybe add kernel cmdline params: to force random seed crediting
333 * nspawn: on cgroupsv1 issue cgroup empty handler process based on host events,
334 so that we make cgroup agent logic safe
336 * nspawn/machined: add API to invoke binary in container, then use that as
337 fallback in "machinectl shell"
339 * logind: rework pam_logind to also do a bus call in case of invocation from
340 user@.service, which returns the XDG_RUNTIME_DIR value, and make this
341 behaviour selectable via pam module option.
344 - when user tries to log into record signed by unrecognized key, automatically add key to our chain after polkit auth
345 - rollback when resize fails mid-operation
346 - GNOME's side for forget key on suspend (requires rework so that lock screen runs outside of uid)
348 - shrink fs on logout?
349 - update LUKS password on login if we find there's a password that unlocks the JSON record but not the LUKS device.
350 - create on activate?
351 - properties: icon url?, preferred session type?, administrator bool (which translates to 'wheel' membership)?, address?, telephone?, vcard?, samba stuff?, parental controls?
352 - communicate clearly when usb stick is safe to remove. probably involves
353 beefing up logind to make pam session close hook synchronous and wait until
354 systemd --user is shut down.
355 - logind: maybe keep a "busy fd" as long as there's a non-released session around or the user@.service
356 - maybe make automatic, read-only, time-based reflink-copies of LUKS disk
357 images (and btrfs snapshots of subvolumes) (think: time machine)
358 - distinguish destroy / remove (i.e. currently we can unregister a user, unregister+remove their home directory, but not just remove their home directory)
359 - in systemd's PAMName= logic: query passwords with ssh-askpassword, so that we can make "loginctl set-linger" mode work
360 - fingerprint authentication, pattern authentication, …
361 - make sure "classic" user records can also be managed by homed
362 - description field for groups
363 - make size of $XDG_RUNTIME_DIR configurable in user record
364 - reuse pwquality magic in firstboot
365 - query password from kernel keyring first
366 - update even if record is "absent"
367 - add a "access mode" + "fstype" field to the "status" section of json identity records reflecting the actually used access mode and fstype, even on non-luks backends
368 - move acct mgmt stuff from pam_systemd_home to pam_systemd?
369 - when "homectl --pkcs11-token-uri=" is used, synthesize ssh-authorized-keys records for all keys we have private keys on the stick for
370 - make slice for users configurable (requires logind rework)
371 - logind: populate auto-login list bus property from PKCS#11 token
372 - when determining state of a LUKS home directory, check DM suspended sysfs file
373 - introduce API for "making room", that grows/shrinks home directory
374 according to elastic parameters, discards blocks, and removes additional snapshots. Call it
375 either from UI when disk space gets low
377 * introduce a new per-process uuid, similar to the boot id, the machine id, the
378 invocation id, that is derived from process creds, specifically a hashed
379 combination of AT_RANDOM + getpid() + the starttime from
380 /proc/self/status. Then add these ids implicitly when logging. Deriving this
381 uuid from these three things has the benefit that it can be derived easily
382 from /proc/$PID/ in a stable, and unique way that changes on both fork() and
385 * let's not GC a unit while its ratelimits are still pending
387 * when killing due to service watchdog timeout maybe detect whether target
388 process is under ptracing and then log loudly and continue instead.
390 * introduce a new group to own TPM devices
392 * make rfkill uaccess controllable by default, i.e. steal rule from
393 gnome-bluetooth and friends
395 * tweak journald context caching. In addition to caching per-process attributes
396 keyed by PID, cache per-cgroup attributes (i.e. the various xattrs we read)
397 keyed by cgroup path, and guarded by ctime changes. This should provide us
398 with a nice speed-up on services that have many processes running in the same
401 * make MAINPID= message reception checks even stricter: if service uses User=,
402 then check sending UID and ignore message if it doesn't match the user or
405 * maybe trigger a uevent "change" on a device if "systemctl reload xyz.device"
408 * when importing an fs tree with machined, optionally apply userns-rec-chown
410 * when importing an fs tree with machined, complain if image is not an OS
412 * Maybe introduce a helper safe_exec() or so, which is to execve() which
413 safe_fork() is to fork(). And then make revert the RLIMIT_NOFILE soft limit
414 to 1K implicitly, unless explicitly opted-out.
416 * rework seccomp/nnp logic that even if User= is used in combination with
417 a seccomp option we don't have to set NNP. For that, change uid first whil
418 keeping CAP_SYS_ADMIN, then apply seccomp, the drop cap.
420 * add a concept for automatically loading per-unit secrets off disk and
421 inserting them into the kernel keyring. Maybe SecretsDirectory= similar to
422 ConfigurationDirectory=.
424 * when no locale is configured, default to UEFI's PlatformLang variable
426 * bootctl,sd-boot: actually honour the "architecture" key
428 * add a new syscall group "@esoteric" for more esoteric stuff such as bpf() and
429 usefaultd() and make systemd-analyze check for it.
431 * paranoia: whenever we process passwords, call mlock() on the memory
432 first. i.e. look for all places we use free_and_erasep() and
433 augment them with mlock(). Also use MADV_DONTDUMP.
435 * Move RestrictAddressFamily= to the new cgroup create socket
437 * support the bind/connect/sendmsg cgroup stuff for sandboxing, and possibly
440 * maybe implicitly attach monotonic+realtime timestamps to outgoing messages in
441 log.c and sd-journal-send
443 * optionally: turn on cgroup delegation for per-session scope units
445 * introduce per-unit (i.e. per-slice, per-service) journal log size limits.
447 * sd-boot: automatically load EFI modules from some drop-in dir, so that people
448 can add in file system drivers and such
450 * sd-boot: optionally, show boot menu when previous default boot item has
451 non-zero "tries done" count
453 * introduce an option (or replacement) for "systemctl show" that outputs all
454 properties as JSON, similar to busctl's new JSON output. In contrast to that
455 it should skip the variant type string though.
457 * augment CODE_FILE=, CODE_LINE= with something like CODE_BASE= or so which
458 contains some identifier for the project, which allows us to include
459 clickable links to source files generating these log messages. The identifier
460 could be some abberviated URL prefix or so (taking inspiration from Go
461 imports). For example, for systemd we could use
462 CODE_BASE=github.com/systemd/systemd/blob/98b0b1123cc or so which is
463 sufficient to build a link by prefixing "http://" and suffixing the
466 * Augment MESSAGE_ID with MESSAGE_BASE, in a similar fashion so that we can
467 make clickable links from log messages carrying a MESSAGE_ID, that lead to
468 some explanatory text online.
470 * maybe extend .path units to expose fanotify() per-mount change events
472 * Add a "systemctl list-units --by-slice" mode or so, which rearranges the
473 output of "systemctl list-units" slightly by showing the tree structure of
474 the slices, and the units attached to them.
476 * the a-posteriori stopping of units bound to units that disappeared logic
477 should be reworked: there should be a queue of units, and we should only
478 enqueue stop jobs from a defer event that processes queue instead of
479 right-away when we find a unit that is bound to one that doesn't exist
480 anymore. (similar to how the stop-unneeded queue has been reworked the same
483 * nspawn: make nspawn suitable for shell pipelines: instead of triggering a
484 hangup when input is finished, send ^D, which synthesizes an EOF. Then wait
485 for hangup or ^D before passing on the EOF.
487 * When reloading configuration PID 1 should reset all its properties to the
488 original defaults before calling parse_config()
490 * nspawn: greater control over selinux label?
492 * hibernate/s2h: make this robust and safe to enable in Fedora by default.
495 1. add resume_offset support to the resume code (i.e. support swap files
497 2. check if swap is on weird storage and refuse if so
498 3. add auto-detection of hibernation images
500 * cgroups: use inotify to get notified when somebody else modifies cgroups
501 owned by us, then log a friendly warning.
503 * beef up log.c with support for stripping ANSI sequences from strings, so that
504 it is OK to include them in log strings. This would be particularly useful so
505 that our log messages could contain clickable links for example for unit
506 files and suchlike we operate on.
508 * add support for "portablectl attach http://foobar.com/waaa.raw (i.e. importd integration)
510 * add attach --enable and attach --now (for attach+enable+start)
512 * sync dynamic uids/gids between host+portable srvice (i.e. if DynamicUser=1 is set for a service, make sure that the
513 selected user is resolvable in the service even if it ships its own /etc/passwd)
515 * Fix DECIMAL_STR_MAX or DECIMAL_STR_WIDTH. One includes a trailing NUL, the
516 other doesn't. What a disaster. Probably to exclude it. Also
517 DECIMAL_STR_WIDTH should probably add an extra "-" into account for negative
520 * Check that users of inotify's IN_DELETE_SELF flag are using it properly, as
521 usually IN_ATTRIB is the right way to watch deleted files, as the former only
522 fires when a file is actually removed from disk, i.e. the link count drops to
523 zero and is not open anymore, while the latter happens when a file is
524 unlinked from any dir.
526 * port systemctl, busctl, … over to format-table.[ch]'s table formatters
528 * pid1: lock image configured with RootDirectory=/RootImage= using the usual nspawn semantics while the unit is up
530 * add --vacuum-xyz options to coredumpctl, matching those journalctl already has.
532 * introduce Ephemeral= unit file switch, that creates an ephemeral copy of all
533 files and directories that are left writable for a unit, and which are
534 removed after the unit goes down again. A bit like --ephemeral for
535 systemd-nspawn but for system services. If used together with RootImage= this
536 should reflink the image file itself.
538 Related: add Ephemeral=<path1> <path2> … which would allow marking
539 specific paths only like this.
541 * add CopyFile= or so as unit file setting that may be used to copy files or
542 directory trees from the host to the services RootImage= and RootDirectory=
543 environment. Which we can use for /etc/machine-id and in particular
544 /etc/resolv.conf. Should be smart and do something useful on read-only
545 images, for example fallback to read-only bind mounting the file instead.
547 * show invocation ID in systemd-run output
549 * bypass SIGTERM state in unit files if KillSignal is SIGKILL
551 * add proper dbus APIs for the various sd_notify() commands, such as MAINPID=1
552 and so on, which would mean we could report errors and such.
554 * teach tmpfiles.d q/Q logic something sensible in the context of XFS/ext4
557 * introduce DefaultSlice= or so in system.conf that allows changing where we
558 place our units by default, i.e. change system.slice to something
559 else. Similar, ManagerSlice= should exist so that PID1's own scope unit could
560 be moved somewhere else too. Finally machined and logind should get similar
561 options so that it is possible to move user session scopes and machines to a
562 different slice too by default. Usecase: people who want to put resources on
563 the entire system, with the exception of one specific service. See:
564 https://lists.freedesktop.org/archives/systemd-devel/2018-February/040369.html
566 * maybe rework get_user_creds() to query the user database if $SHELL is used
567 for root, but only then.
569 * be stricter with fds we receive for the fdstore: close them asynchronously
571 * calenderspec: add support for week numbers and day numbers within a
572 year. This would allow us to define "bi-weekly" triggers safely.
574 * sd-bus: add vtable flag, that may be used to request client creds implicitly
575 and asynchronously before dispatching the operation
577 * make use of ethtool veth peer info in machined, for automatically finding out
578 host-side interface pointing to the container.
580 * add some special mode to LogsDirectory=/StateDirectory=… that allows
581 declaring these directories without necessarily pulling in deps for them, or
582 creating them when starting up. That way, we could declare that
583 systemd-journald writes to /var/log/journal, which could be useful when we
584 doing disk usage calculations and so on.
586 * taint systemd if there are fewer than 65536 users assigned (userns) to the system.
588 * deprecate RootDirectoryStartOnly= in favour of a new ExecStart= prefix char
590 * add a new RuntimeDirectoryPreserve= mode that defines a similar lifecycle for
591 the runtime dir as we maintain for the fdstore: i.e. keep it around as long
592 as the unit is running or has a job queued.
594 * support projid-based quota in machinectl for containers
596 * add a way to lock down cgroup migration: a boolean, which when set for a unit
597 makes sure the processes in it can never migrate out of it
599 * blog about fd store and restartable services
601 * document Environment=SYSTEMD_LOG_LEVEL=debug drop-in in debugging document
603 * rework ExecOutput and ExecInput enums so that EXEC_OUTPUT_NULL loses its
604 magic meaning and is no longer upgraded to something else if set explicitly.
606 * in the long run: permit a system with /etc/machine-id linked to /dev/null, to
607 make it lose its identity, i.e. be anonymous. For this we'd have to patch
608 through the whole tree to make all code deal with the case where no machine
611 * optionally, collect cgroup resource data, and store it in per-unit RRD files,
612 suitable for processing with rrdtool. Add bus API to access this data, and
613 possibly implement a CPULoad property based on it.
615 * beef up pam_systemd to take unit file settings such as cgroups properties as
618 * maybe hook of xfs/ext4 quotactl() with services? i.e. automatically manage
619 the quota of a the user indicated in User= via unit file settings, like the
620 other resource management concepts. Would mix nicely with DynamicUser=1. Or
621 alternatively, do this with projids, so that we can also cover services
622 running as root. Quota should probably cover all the special dirs such as
623 StateDirectory=, LogsDirectory=, CacheDirectory=, as well as RootDirectory= if it
624 is set, plus the whole disk space any image configured with RootImage=.
626 * In DynamicUser= mode: before selecting a UID, use disk quota APIs on relevant
627 disks to see if the UID is already in use.
629 * add "systemctl wait" or so, which does what "systemd-run --wait" does, but
630 for all units. It should be both a way to pin units into memory as well as a
631 wait to retrieve their exit data.
633 * expose IO accounting data on the bus, show it in systemd-run --wait and log
634 about it in the resource log message
636 * show whether a service has out-of-date configuration in "systemctl status" by
637 using mtime data of ConfigurationDirectory=.
639 * Add AddUser= setting to unit files, similar to DynamicUser=1 which however
640 creates a static, persistent user rather than a dynamic, transient user. We
641 can leverage code from sysusers.d for this.
643 * add some optional flag to ReadWritePaths= and friends, that has the effect
644 that we create the dir in question when the service is started. Example:
646 ReadWritePaths=:/var/lib/foobar
648 * maybe add call sd_journal_set_block_timeout() or so to set SO_SNDTIMEO for
649 the sd-journal logging socket, and, if the timeout is set to 0, sets
650 O_NONBLOCK on it. That way people can control if and when to block for
653 * hostnamed: populate form factor data from a new hwdb database, so that old
654 yogas can be recognized as "convertible" too, even if they predate the DMI
655 "convertible" form factor
657 * Maybe add PrivatePIDs= as new unit setting, and do minimal PID namespacing
658 after all. Be strict however, only support the equivalent of nspawn's
659 --as-pid2 switch, and sanely proxy sd_notify() messages dropping stuff such
662 * Add ExecMonitor= setting. May be used multiple times. Forks off a process in
663 the service cgroup, which is supposed to monitor the service, and when it
664 exits the service is considered failed by its monitor.
666 * track the per-service PAM process properly (i.e. as an additional control
667 process), so that it may be queried on the bus and everything.
669 * add a new "debug" job mode, that is propagated to unit_start() and for
670 services results in two things: we raise SIGSTOP right before invoking
671 execve() and turn off watchdog support. Then, use that to implement
672 "systemd-gdb" for attaching to the start-up of any system service in its
675 * gpt-auto logic: related to the above, maybe support a "secondary" root
676 partition, that is mounted to / and is writable, and where the actual root's
677 /usr is mounted into.
679 * 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
681 * drop nss-myhostname in favour of nss-resolve?
683 * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and
684 then use that for the setting used in user@.service. It should be understood
685 relative to the configured default value.
687 * in networkd, when matching device types, fix up DEVTYPE rubbish the kernel passes to us
689 * enable LockMLOCK to take a percentage value relative to physical memory
691 * Permit masking specific netlink APIs with RestrictAddressFamily=
693 * nspawn: support that /proc, /sys/, /dev are pre-mounted
695 * define gpt header bits to select volatility mode
697 * ProtectClock= (drops CAP_SYS_TIMES, adds seecomp filters for settimeofday, adjtimex), sets DeviceAllow o /dev/rtc
699 * ProtectTracing= (drops CAP_SYS_PTRACE, blocks ptrace syscall, makes /sys/kernel/tracing go away)
701 * ProtectMount= (drop mount/umount/pivot_root from seccomp, disallow fuse via DeviceAllow, imply Mountflags=slave)
703 * ProtectKeyRing= to take keyring calls away
705 * RemoveKeyRing= to remove all keyring entries of the specified user
707 * ProtectReboot= that masks reboot() and kexec_load() syscalls, prohibits kill
708 on PID 1 with the relevant signals, and makes relevant files in /sys and
709 /proc (such as the sysrq stuff) unavailable
711 * make sure the ratelimit object can deal with USEC_INFINITY as way to turn off things
713 * journalctl: make sure -f ends when the container indicated by -M terminates
715 * mount: automatically search for "main" partition of an image has multiple
718 * in nss-systemd, if we run inside of RootDirectory= with PrivateUsers= set,
719 find a way to map the User=/Group= of the service to the right name. This way
720 a user/group for a service only has to exist on the host for the right
723 * add bus API for creating unit files in /etc, reusing the code for transient units
725 * add bus API to remove unit files from /etc
727 * add bus API to retrieve current unit file contents (i.e. implement "systemctl cat" on the bus only)
729 * rework fopen_temporary() to make use of open_tmpfile_linkable() (problem: the
730 kernel doesn't support linkat() that replaces existing files, currently)
732 * transient units: don't bother with actually setting unit properties, we
733 reload the unit file anyway
735 * journald: sigbus API via a signal-handler safe function that people may call
736 from the SIGBUS handler
738 * optionally, also require WATCHDOG=1 notifications during service start-up and shutdown
740 * delay activation of logind until somebody logs in, or when /dev/tty0 pulls it
741 in or lingering is on (so that containers don't bother with it until PAM is used). also exit-on-idle
743 * cache sd_event_now() result from before the first iteration...
745 * PID1: find a way how we can reload unit file configuration for
746 specific units only, without reloading the whole of systemd
748 * add an explicit parser for LimitRTPRIO= that verifies
749 the specified range and generates sane error messages for incorrect
752 * when we detect that there are waiting jobs but no running jobs, do something
754 * push CPUAffinity= also into the "cpuset" cgroup controller
756 * PID 1 should send out sd_notify("WATCHDOG=1") messages (for usage in the --user mode, and when run via nspawn)
758 * there's probably something wrong with having user mounts below /sys,
759 as we have for debugfs. for example, src/core/mount.c handles mounts
760 prefixed with /sys generally special.
761 http://lists.freedesktop.org/archives/systemd-devel/2015-June/032962.html
763 * fstab-generator: default to tmpfs-as-root if only usr= is specified on the kernel cmdline
765 * docs: bring http://www.freedesktop.org/wiki/Software/systemd/MyServiceCantGetRealtime up to date
767 * add a job mode that will fail if a transaction would mean stopping
768 running units. Use this in timedated to manage the NTP service
770 http://lists.freedesktop.org/archives/systemd-devel/2015-April/030229.html
772 * The udev blkid built-in should expose a property that reflects
773 whether media was sensed in USB CF/SD card readers. This should then
774 be used to control SYSTEMD_READY=1/0 so that USB card readers aren't
775 picked up by systemd unless they contain a medium. This would mirror
776 the behaviour we already have for CD drives.
778 * networkd/udev: implement SR_IOV configuration in .link files:
779 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027451.html
781 * hostnamectl: show root image uuid
783 * Find a solution for SMACK capabilities stuff:
784 http://lists.freedesktop.org/archives/systemd-devel/2014-December/026188.html
786 * "systemctl preset-all" should probably order the unit files it
787 operates on lexicographically before starting to work, in order to
788 ensure deterministic behaviour if two unit files conflict (like DMs
791 * synchronize console access with BSD locks:
792 http://lists.freedesktop.org/archives/systemd-devel/2014-October/024582.html
794 * as soon as we have sender timestamps, revisit coalescing multiple parallel daemon reloads:
795 http://lists.freedesktop.org/archives/systemd-devel/2014-December/025862.html
797 * figure out when we can use the coarse timers
799 * add "systemctl start -v foobar.service" that shows logs of a service
800 while the start command runs. This is non-trivial to do without
801 races though, since we should flush out all journal messages before
802 returning from the "systemctl stop".
804 * 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
806 * maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment=
808 * maybe allow timer units with an empty Units= setting, so that they
809 can be used for resuming the system but nothing else.
811 * what to do about udev db binary stability for apps? (raw access is not an option)
813 * systemctl: if some operation fails, show log output?
815 * systemctl edit: use equivalent of cat() to insert existing config as a comment, prepended with #.
816 Upon editor exit, lines with one # are removed, lines with two # are left with one #, etc.
818 * exponential backoff in timesyncd when we cannot reach a server
820 * timesyncd: add ugly bus calls to set NTP servers per-interface, for usage by NM
822 * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share....
824 * add systemd.abort_on_kill or some other such flag to send SIGABRT instead of SIGKILL
825 (throughout the codebase, not only PID1)
829 - service registration
830 - service/domain/types browsing
832 - DNS-SD service registration from socket units
833 - resolved should optionally register additional per-interface LLMNR
834 names, so that for the container case we can establish the same name
835 (maybe "host") for referencing the server, everywhere.
836 - allow clients to request DNSSEC for a single lookup even if DNSSEC is off (?)
837 - hook up resolved with machined-based address resolution
839 * refcounting in sd-resolve is borked
841 * Add a new verb "systemctl top"
843 * add new gpt type for btrfs volumes
845 * support empty /etc boots nicely:
846 - nspawn/gpt-generator: introduce new gpt partition type for /usr
848 * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them.
850 * a way for container managers to turn off getty starting via $container_headless= or so...
852 * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit
854 * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services
855 they run added to the initial transaction and thus confuse Type=idle.
857 * add bus api to query unit file's X fields.
859 * gpt-auto-generator:
860 - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap?
861 - Make /home automount rather than mount?
863 * add generator that pulls in systemd-network from containers when
864 CAP_NET_ADMIN is set, more than the loopback device is defined, even
865 when it is otherwise off
867 * MessageQueueMessageSize= (and suchlike) should use parse_iec_size().
869 * implement Distribute= in socket units to allow running multiple
870 service instances processing the listening socket, and open this up
873 * introduce bus call FreezeUnit(s, b), as well as "systemctl freeze
874 $UNIT" and "systemctl thaw $UNIT" as wrappers around this. The calls
875 should SIGSTOP all unit processes in a loop until all processes of
876 it are fully stopped. This can later be used for app management by
877 desktop UIs such as gnome-shell to freeze apps that are not visible
878 on screen, not unlike how job control works on the shell
881 - implement per-slice CPUFairScheduling=1 switch
882 - introduce high-level settings for RT budget, swappiness
883 - how to reset dynamically changed unit cgroup attributes sanely?
884 - when reloading configuration, apply new cgroup configuration
885 - when recursively showing the cgroup hierarchy, optionally also show
886 the hierarchies of child processes
889 - add field to transient units that indicate whether systemd or somebody else saves/restores its settings, for integration with libvirt
891 * when we detect low battery and no AC on boot, show pretty splash and refuse boot
893 * libsystemd-journal, libsystemd-login, libudev: add calls to easily attach these objects to sd-event event loops
895 * be more careful what we export on the bus as (usec_t) 0 and (usec_t) -1
897 * 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
899 * After coming back from hibernation reset hibernation swap partition using the /dev/snapshot ioctl APIs
901 * If we try to find a unit via a dangling symlink, generate a clean
902 error. Currently, we just ignore it and read the unit from the search
905 * refuse boot if /usr/lib/os-release is missing or /etc/machine-id cannot be set up
907 * man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted.
909 * load .d/*.conf dropins for device units
913 - GetAllProperties() on a non-existing object does not result in a failure currently
914 - port to sd-resolve for connecting to TCP dbus servers
915 - 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
916 - see if we can drop more message validation on the sending side
917 - add API to clone sd_bus_message objects
918 - longer term: priority inheritance
920 - NameLost/NameAcquired obsolete
923 - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now
926 - allow multiple signal handlers per signal?
927 - document chaining of signal handler for SIGCHLD and child handlers
928 - define more intervals where we will shift wakeup intervals around in, 1h, 6h, 24h, ...
930 * investigate endianness issues of UUID vs. GUID
932 * dbus: when a unit failed to load (i.e. is in UNIT_ERROR state), we
933 should be able to safely try another attempt when the bus call LoadUnit() is invoked.
935 * add a pam module that passes the hdd passphrase into the PAM stack and then expires it, for usage by gdm auto-login.
937 * add a pam module that on password changes updates any LUKS slot where the password matches
940 - add unit tests for config_parse_device_allow()
942 * seems that when we follow symlinks to units we prefer the symlink
943 destination path over /etc and /usr. We should not do that. Instead
944 /etc should always override /run+/usr and also any symlink
947 * when isolating, try to figure out a way how we implicitly can order
948 all units we stop before the isolating unit...
950 * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off})
952 * Add ConditionDirectoryNotEmpty= handle non-absoute paths as a search path or add
953 ConditionConfigSearchPathNotEmpty= or different syntax? See the discussion starting at
954 https://github.com/systemd/systemd/pull/15109#issuecomment-607740136.
956 * BootLoaderSpec: Clarify that the kernel has to be in $BOOT. Clarify
957 that the boot loader should be installed to the ESP. Define a way
958 how an installer can figure out whether a BLS compliant boot loader
961 * think about requeuing jobs when daemon-reload is issued? usecase:
962 the initrd issues a reload after fstab from the host is accessible
963 and we might want to requeue the mounts local-fs acquired through
966 * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep()
968 * remove any syslog support from log.c — we probably cannot do this before split-off udev is gone for good
970 * shutdown logging: store to EFI var, and store to USB stick?
972 * merge unit_kill_common() and unit_kill_context()
975 - honor language efi variables for default language selection (if there are any?)
976 - honor timezone efi variables for default timezone selection (if there are any?)
977 - change bootctl to be backed by systemd-bootd to control temporary and persistent default boot goal plus efi variables
979 - recognize the case when not booted on EFI
981 * maybe do not install getty@tty1.service symlink in /etc but in /usr?
983 * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word
985 * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units.
988 - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around
989 - logind: wakelock/opportunistic suspend support
990 - Add pretty name for seats in logind
991 - logind: allow showing logout dialog from system?
992 - add Suspend() bus calls which take timestamps to fix double suspend issues when somebody hits suspend and closes laptop quickly.
993 - if pam_systemd is invoked by su from a process that is outside of a
994 any session we should probably just become a NOP, since that's
995 usually not a real user session but just some system code that just
997 - logind: make the Suspend()/Hibernate() bus calls wait for the for
998 the job to be completed. before returning, so that clients can wait
999 for "systemctl suspend" to finish to know when the suspending is
1001 - logind: when the power button is pressed short, just popup a
1002 logout dialog. If it is pressed for 1s, do the usual
1003 shutdown. Inspiration are Macs here.
1004 - expose "Locked" property on logind sesison objects
1005 - maybe allow configuration of the StopTimeout for session scopes
1006 - rename session scope so that it includes the UID. THat way
1007 the session scope can be arranged freely in slices and we don't have
1008 make assumptions about their slice anymore.
1009 - follow PropertiesChanged state more closely, to deal with quick logouts and
1011 - (optionally?) spawn seat-manager@$SEAT.service whenever a seat shows up that as CanGraphical set
1014 - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields
1015 - journald: also get thread ID from client, plus thread name
1016 - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups
1017 - add API to close/reopen/get fd for journal client fd in libsystemd-journal.
1018 - fallback to /dev/log based logging in libsystemd-journal, if we cannot log natively?
1019 - declare the local journal protocol stable in the wiki interface chart
1020 - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg
1021 - journald: when dropping msgs due to ratelimit make sure to write
1022 "dropped %u messages" not only when we are about to print the next
1023 message that works, but already after a short timeout
1024 - check if we can make journalctl by default use --follow mode inside of less if called without args?
1025 - maybe add API to send pairs of iovecs via sd_journal_send
1026 - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access
1027 - journactl: support negative filtering, i.e. FOOBAR!="waldo",
1028 and !FOOBAR for events without FOOBAR.
1029 - journal: store timestamp of journal_file_set_offline() in the header,
1030 so it is possible to display when the file was last synced.
1031 - 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.
1032 - journal: find a way to allow dropping history early, based on priority, other rules
1033 - journal: When used on NFS, check payload hashes
1034 - journald: add kernel cmdline option to disable ratelimiting for debug purposes
1035 - refuse taking lower-case variable names in sd_journal_send() and friends.
1036 - journald: we currently rotate only after MaxUse+MaxFilesize has been reached.
1037 - journal: deal nicely with byte-by-byte copied files, especially regards header
1038 - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit
1039 - Replace utmp, wtmp, btmp, and lastlog completely with journal
1040 - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax?
1041 - when a kernel driver logs in a tight loop, we should ratelimit that too.
1042 - journald: optionally, log debug messages to /run but everything else to /var
1043 - journald: when we drop syslog messages because the syslog socket is
1044 full, make sure to write how many messages are lost as first thing
1045 to syslog when it works again.
1046 - journald: allow per-priority and per-service retention times when rotating/vacuuming
1047 - journald: make use of uid-range.h to managed uid ranges to split
1049 - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so...
1050 - improve journalctl performance by loading journal files
1051 lazily. Encode just enough information in the file name, so that we
1052 do not have to open it to know that it is not interesting for us, for
1053 the most common operations.
1054 - man: document that corrupted journal files is nothing to act on
1055 - rework journald sigbus stuff to use mutex
1056 - Set RLIMIT_NPROC for systemd-journal-xyz, and all other of our
1057 services that run under their own user ids, and use User= (but only
1058 in a world where userns is ubiquitous since otherwise we cannot
1059 invoke those daemons on the host AND in a container anymore). Also,
1060 if LimitNPROC= is used without User= we should warn and refuse
1062 - journalctl --verify: don't show files that are currently being
1063 written to as FAIL, but instead show that their are being written to.
1064 - add journalctl -H that talks via ssh to a remote peer and passes through
1066 - add a version of --merge which also merges /var/log/journal/remote
1067 - journalctl: -m should access container journals directly by enumerating
1068 them via machined, and also watch containers coming and going.
1069 Benefit: nspawn --ephemeral would start working nicely with the journal.
1070 - assign MESSAGE_ID to log messages about failed services
1071 - check if loop in decompress_blob_xz() is necessary
1073 * add a test if all entries in the catalog are properly formatted.
1074 (Adding dashes in a catalog entry currently results in the catalog entry
1075 being silently skipped. journalctl --update-catalog must warn about this,
1076 and we should also have a unit test to check that all our message are OK.)
1079 - document that deps in [Unit] sections ignore Alias= fields in
1080 [Install] units of other units, unless those units are disabled
1081 - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets
1082 - document that service reload may be implemented as service reexec
1083 - add a man page containing packaging guidelines and recommending usage of things like Documentation=, PrivateTmp=, PrivateNetwork= and ReadOnlyDirectories=/etc /usr.
1084 - document systemd-journal-flush.service properly
1085 - documentation: recommend to connect the timer units of a service to the service via Also= in [Install]
1086 - man: document the very specific env the shutdown drop-in tools live in
1087 - man: add more examples to man pages
1088 - man: maybe sort directives in man pages, and take sections from --help and apply them to man too
1089 - document root=gpt-auto properly
1092 - add systemctl switch to dump transaction without executing it
1093 - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done
1094 - "systemctl disable" on a static unit prints no message and does
1095 nothing. "systemctl enable" does nothing, and gives a bad message
1096 about it. Should fix both to print nice actionable messages.
1097 - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service
1098 - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible
1099 - systemctl enable: fail if target to alias into does not exist? maybe show how many units are enabled afterwards?
1100 - systemctl: "Journal has been rotated since unit was started." message is misleading
1101 - systemctl status output should include list of triggering units and their status
1104 - "systemctl mask" should find all names by which a unit is accessible
1105 (i.e. by scanning for symlinks to it) and link them all to /dev/null
1108 - timer units should get the ability to trigger when:
1110 - Modulate timer frequency based on battery state
1112 * add libsystemd-password or so to query passwords during boot using the password agent logic
1114 * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed
1116 * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel
1118 * make repeated alt-ctrl-del presses printing a dump
1120 * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ...
1122 * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab is not
1125 - emulate /dev/kmsg using CUSE and turn off the syslog syscall
1126 with seccomp. That should provide us with a useful log buffer that
1127 systemd can log to during early boot, and disconnect container logs
1128 from the kernel's logs.
1129 - as soon as networkd has a bus interface, hook up --network-interface=,
1130 --network-bridge= with networkd, to trigger netdev creation should an
1131 interface be missing
1132 - a nice way to boot up without machine id set, so that it is set at boot
1133 automatically for supporting --ephemeral. Maybe hash the host machine id
1134 together with the machine name to generate the machine id for the container
1135 - fix logic always print a final newline on output.
1136 https://github.com/systemd/systemd/pull/272#issuecomment-113153176
1137 - should optionally support receiving WATCHDOG=1 messages from its payload
1139 - optionally automatically add FORWARD rules to iptables whenever nspawn is
1140 running, remove them when shut down.
1143 - refuse mounting over a mount point
1144 - automatically discover .roothash files in dissect, similarly to nspawn
1147 - add an API so that libvirt-lxc can inform us about network interfaces being
1148 removed or added to an existing machine
1149 - "machinectl migrate" or similar to copy a container from or to a
1150 difference host, via ssh
1151 - introduce systemd-nspawn-ephemeral@.service, and hook it into
1152 "machinectl start" with a new --ephemeral switch
1153 - "machinectl status" should also show internal logs of the container in
1155 - "machinectl history"
1157 - "machinectl commit" that takes a writable snapshot of a tree, invokes a
1158 shell in it, and marks it read-only after use
1161 - cryptsetup-generator: allow specification of passwords in crypttab itself
1162 - support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator
1164 * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it
1165 https://bugs.freedesktop.org/show_bug.cgi?id=54712
1167 * add a dependency on standard-conf.xml and other included files to man pages
1169 * MountFlags=shared acts as MountFlags=slave right now.
1171 * properly handle loop back mounts via fstab, especially regards to fsck/passno
1173 * initialize the hostname from the fs label of /, if /etc/hostname does not exist?
1178 - add trigger --subsystem-match=usb/usb_device device
1179 - reimport udev db after MOVE events for devices without dev_t
1181 * There's currently no way to cancel fsck (used to be possible via C-c or c on the console)
1183 * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/
1186 - save coredump in Windows/Mozilla minidump format
1187 - when truncating coredumps, also log the full size that the process had, and make a metadata field so we can report truncated coredumps
1189 * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting)
1191 * be able to specify a forced restart of service A where service B depends on, in case B
1192 needs to be auto-respawned?
1195 - apply "x" on "D" too (see patch from William Douglas)
1196 - instead of ignoring unknown fields, reject them.
1197 - creating new directories/subvolumes/fifos/device nodes
1198 should not follow symlinks. None of the other adjustment or creation
1199 calls follow symlinks.
1201 * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early
1203 * verify that the AF_UNIX sockets of a service in the fs still exist
1204 when we start a service in order to avoid confusion when a user
1205 assumes starting a service is enough to make it accessible
1207 * Make it possible to set the keymap independently from the font on
1208 the kernel cmdline. Right now setting one resets also the other.
1210 * and a dbus call to generate target from current state
1212 * write blog stories about:
1213 - hwdb: what belongs into it, lsusb
1214 - enabling dbus services
1215 - how to make changes to sysctl and sysfs attributes
1217 - how to pass throw-away units to systemd, or dynamically change properties of existing units
1218 - testing with Harald's awesome test kit
1220 - how to develop against journal browsing APIs
1221 - the journal HTTP iface
1222 - non-cgroup resource management
1223 - dynamic resource management with cgroups
1224 - refreshed, longer missions statement
1225 - calendar time events
1226 - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell
1227 - how to create your own target
1228 - instantiated apache, dovecot and so on
1229 - hooking a script into various stages of shutdown/rearly booot
1231 * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support.
1233 * dot output for --test showing the 'initial transaction'
1236 - When logging about multiple units (stopping BoundTo units, conflicts, etc.),
1237 log both units as UNIT=, so that journalctl -u triggers on both.
1238 - generate better errors when people try to set transient properties
1239 that are not supported...
1240 http://lists.freedesktop.org/archives/systemd-devel/2015-February/028076.html
1241 - maybe introduce WantsMountsFor=? Usecase:
1242 http://lists.freedesktop.org/archives/systemd-devel/2015-January/027729.html
1243 - recreate systemd's D-Bus private socket file on SIGUSR2
1244 - move PAM code into its own binary
1245 - when we automatically restart a service, ensure we restart its rdeps, too.
1246 - hide PAM options in fragment parser when compile time disabled
1247 - Support --test based on current system state
1248 - 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().
1249 - after deserializing sockets in socket.c we should reapply sockopts and things
1250 - drop PID 1 reloading, only do reexecing (difficult: Reload()
1251 currently is properly synchronous, Reexec() is weird, because we
1252 cannot delay the response properly until we are back, so instead of
1253 being properly synchronous we just keep open the fd and close it
1254 when done. That means clients do not get a successful method reply,
1255 but much rather a disconnect on success.
1256 - when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr
1257 - when a bus name of a service disappears from the bus make sure to queue further activation requests
1260 - allow port=0 in .socket units
1261 - maybe introduce ExecRestartPre=
1262 - add ReloadSignal= for configuring a reload signal to use
1263 - implement Register= switch in .socket units to enable registration
1264 in Avahi, RPC and other socket registration services.
1265 - allow Type=simple with PIDFile=
1266 https://bugzilla.redhat.com/show_bug.cgi?id=723942
1267 - allow writing multiple conditions in unit files on one line
1268 - introduce Type=pid-file
1269 - introduce mix of BindTo and Requisite
1270 - add a concept of RemainAfterExit= to scope units
1271 - Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely
1272 - add verification of [Install] section to systemd-analyze verify
1275 - Make sure ID_PATH is always exported and complete for
1276 network devices where possible, so we can safely rely
1280 - add support for more attribute types
1281 - inbuilt piping support (essentially degenerate async)? see loopback-setup.c and other places
1284 - add more keys to [Route] and [Address] sections
1285 - add support for more DHCPv4 options (and, longer term, other kinds of dynamic config)
1286 - add reduced [Link] support to .network files
1287 - properly handle routerless dhcp leases
1288 - work with non-Ethernet devices
1289 - dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from?
1290 - the DHCP lease data (such as NTP/DNS) is still made available when
1291 a carrier is lost on a link. It should be removed instantly.
1292 - expose in the API the following bits:
1293 - option 15, domain name
1294 - option 12, hostname and/or option 81, fqdn
1295 - option 123, 144, geolocation
1296 - option 252, configure http proxy (PAC/wpad)
1297 - provide a way to define a per-network interface default metric value
1298 for all routes to it. possibly a second default for DHCP routes.
1299 - allow Name= to be specified repeatedly in the [Match] section. Maybe also
1300 support Name=foo*|bar*|baz ?
1301 - duplicate address check for static IPs (like ARPCHECK in network-scripts)
1302 - whenever uplink info changes, make DHCP server send out FORCERENEW
1304 * Figure out how to do unittests of networkd's state serialization
1307 - figure out how much we can increase Maximum Message Size
1310 - add functions to set previously stored IPv6 addresses on startup and get
1311 them at shutdown; store them in client->ia_na
1312 - write more test cases
1313 - implement reconfigure support, see 5.3., 15.11. and 22.20.
1314 - implement support for temporary adressess (IA_TA)
1315 - implement dhcpv6 authentication
1316 - investigate the usefulness of Confirm messages; i.e. are there any
1317 situations where the link changes without any loss in carrier detection
1319 - some servers don't do rapid commit without a filled in IA_NA, verify
1326 - natively watch for dbus-*.service symlinks (PENDING)
1327 - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service
1329 * make cryptsetup lower --iter-time
1331 * kernel: add device_type = "fb", "fbcon" to class "graphics"
1333 * /usr/bin/service should actually show the new command line
1335 * 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=
1337 * neither pkexec nor sudo initialize environ[] from the PAM environment?
1339 * fedora: update policy to declare access mode and ownership of unit files to root:root 0644, and add an rpmlint check for it
1341 * register catalog database signature as file magic
1343 * zsh shell completion:
1344 - <command> <verb> -<TAB> should complete options, but currently does not
1345 - systemctl add-wants,add-requires
1347 * systemctl status should know about 'systemd-analyze calendar ... --iterations='
1348 * If timer has just OnInactiveSec=..., it should fire after a specified time
1349 after being started.
1353 * look for close() vs. close_nointr() vs. close_nointr_nofail()
1355 * check for strerror(r) instead of strerror(-r)
1359 * set_put(), hashmap_put() return values check. i.e. == 0 does not free()!
1361 * use secure_getenv() instead of getenv() where appropriate
1363 * link up selected blog stories from man pages and unit files Documentation= fields
1364 String is not UTF-8 clean, ignoring assignment