]>
Commit | Line | Data |
---|---|---|
afe3ab58 | 1 | Bugfixes: |
c343be28 | 2 | |
54fcb619 ZJS |
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. | |
6 | ||
f38afcd0 | 7 | External: |
f85857df | 8 | |
f38afcd0 | 9 | * Fedora: add an rpmlint check that verifies that all unit files in the RPM are listed in %systemd_post macros. |
bafb15ba | 10 | |
83a1ff25 ZJS |
11 | * wiki: update journal format documentation for lz4 additions |
12 | ||
5e524b40 LP |
13 | Janitorial Clean-ups: |
14 | ||
5e524b40 LP |
15 | * Rearrange tests so that the various test-xyz.c match a specific src/basic/xyz.c again |
16 | ||
bb527e11 LP |
17 | * rework mount.c and swap.c to follow proper state enumeration/deserialization |
18 | semantics, like we do for device.c now | |
19 | ||
c6407108 LP |
20 | Features: |
21 | ||
d200253b LP |
22 | * systemd-gpt-auto should probably set x-systemd.growfs on the mounts it |
23 | creates | |
24 | ||
006c44c1 LP |
25 | * homed/userdb: distuingish passwords and recovery keys in the records, since |
26 | we probably want to use different PBKDF algorithms/settings for them: | |
27 | passwords have low entropy but recovery keys should have good entropy key | |
28 | hence we can make them quicker to work. | |
29 | ||
492f91d8 LP |
30 | * bootctl: |
31 | - teach it to prepare an ESP wholesale, i.e. with mkfs.vfat invocation | |
32 | - teach it to copy in unified kernel images and maybe type #1 boot loader spec entries from host | |
33 | - make it operate on loopback files, dissecting enough to find ESP to operate on | |
34 | ||
e1d32d6e LP |
35 | * by default, in systemd --user service bump the OOMAdjust to 100, as privs |
36 | allow so that systemd survives | |
37 | ||
27cc3c9d LP |
38 | * when dissecting images, warn about unrecognized partition flags |
39 | ||
40 | * honour specifiers in unit files that resolve to some very basic | |
41 | /etc/os-release data, such as ID, VERSION_ID, BUILD_ID, VARIANT_ID. | |
42 | ||
91fc013f LP |
43 | * socket units: allow creating a udev monitor socket with ListenDevices= or so, |
44 | with matches, then actviate app thorugh that passing socket oveer | |
45 | ||
492f91d8 LP |
46 | * unify on openssl: |
47 | - port sd_id128_get_machine_app_specific() over from khash | |
48 | - port resolved over from libgcrypt (DNSSEC code) | |
49 | - port journald + fsprg over from libgcrypt | |
50 | - port importd over from libgcrypt | |
51 | - when that's done: kill khash.c | |
52 | - when that's done: kill gnutls support in resolved | |
53 | ||
0e7e8544 LP |
54 | * kill zenata, all hail weblate? |
55 | ||
492f91d8 LP |
56 | * when we resize disks (homed?) always round up to 4K sectors, not 512K |
57 | ||
58 | * add growvol and makevol options for /etc/crypttab, similar to | |
59 | x-systemd.growfs and x-systemd-makefs. | |
60 | ||
61 | * hook up the TPM to /etc/crypttab, with a new option that is similar to the | |
62 | new PKCS#11 option in crypttab, and allows unlocking a LUKS volume via a key | |
63 | unsealed from the TPM. Optionally, if TPM is not available fall back to | |
64 | TPM-less mode, and set up linear DM mapping instead (inspired by kpartx), so | |
65 | that the device paths stay the same, regardless if crypto is used or not. | |
66 | ||
492f91d8 LP |
67 | * systemd-repart: by default generate minimized partition tables (i.e. tables |
68 | that only covere the space actually used, excluding any free space at the | |
69 | end), in order to maximize dd'ability. Requires libfdisk work, see | |
70 | https://github.com/karelzak/util-linux/issues/907 | |
71 | ||
72 | * systemd-repart: optionally, allow specifiying a path to initialize new | |
73 | partitions from, i.e. an fs image file or a source device node. This would | |
74 | then turn systemd-repart into a simple installer: with a few .repart files | |
006c44c1 LP |
75 | you could replicate the host system on another device. a full installer would |
76 | then be: "systemd-repart /dev/sda && bootctl install /dev/sda && | |
77 | systemd-firstboot --image= …" | |
492f91d8 LP |
78 | |
79 | * systemd-repart: MBR partition table support. Care needs to be taken regarding | |
80 | Type=, so that partition definitions can sanely apply to both the GPT and the | |
81 | MBR case. Idea: accept syntax "Type=gpt:home mbr:0x83" for setting the types | |
82 | for the two partition types explicitly. And provide an internal mapping so | |
83 | that "Type=linux-generic" maps to the right types for both partition tables | |
84 | automatically. | |
85 | ||
86 | * systemd-repart: allow sizing partitions as factor of available RAM, so that | |
87 | we can reasonably size swap partitions for hibernation. | |
88 | ||
006c44c1 LP |
89 | * systemd-repart: allow running mkfs before making partitions pop up + |
90 | encryption via LUKS to allow booting into an empty root with only /usr mounted in | |
91 | ||
92 | * systemd-repart: allow managing the gpt read-only partition flag + auto-mount flag | |
93 | ||
94 | * systemd-repart: allow disabling growing of specific partitions, or making | |
95 | them (think ESP: we don't ever want to grow it, since we cannot resize vfat) | |
96 | ||
97 | * systemd-repart: add specifier expansion, add especifier that refers to root | |
98 | device node of current system, /usr device node, and matching verity, so that | |
99 | an installer can be made a "copy" installer of the booted OS | |
100 | ||
101 | * systemd-repart: make it a static checker during early boot for existance and | |
102 | absence of other partitions for trusted boot environments | |
103 | ||
104 | * systemd-repart: when no configuration is found, exit early do not check | |
105 | partition table, so that it is safe to run in the initrd on any system | |
106 | ||
107 | * systemd-repart: allow config of partition uuid | |
108 | ||
109 | * userdb: allow username prefix searches in varlink API | |
110 | ||
111 | * userdb: allow existance checks | |
112 | ||
113 | * pid: activation by journal search expression | |
114 | ||
492f91d8 LP |
115 | * when switching root from initrd to host, set the machine_id env var so that |
116 | if the host has no machine ID set yet we continue to use the random one the | |
117 | initrd had set. | |
118 | ||
173c7873 LP |
119 | * sd-event: add native support for P_ALL waitid() watching, then move PID 1 to |
120 | it fo reaping assigned but unknown children. This needs to some special care | |
121 | to operate somewhat sensibly in light of priorities: P_ALL will return | |
122 | arbitrary processes, regardless of the priority we want to watch them with, | |
123 | hence on each event loop iteration check all processes which we shall watch | |
124 | with higher prio explicitly, and then watch the entire rest with P_ALL. | |
125 | ||
126 | * tweak sd-event's child watching: keep a prioq of children to watch and use | |
127 | waitid() only on the children with the highest priority until one is waitable | |
128 | and ignore all lower-prio ones from that point on | |
129 | ||
173c7873 LP |
130 | * maybe introduce xattrs that can be set on the root dir of the root fs |
131 | partition that declare the volatility mode to use the image in. Previously I | |
132 | thought marking this via GPT partition flags but that's not ideal since | |
133 | that's outside of the LUKS encryption/verity verification, and we probably | |
134 | shouldn't operate in a volatile mode unless we got told so from a trusted | |
135 | source. | |
136 | ||
173c7873 LP |
137 | * figure out automatic partition discovery when combining writable root dir |
138 | with immutable /usr | |
139 | ||
97e5cc88 LP |
140 | * coredump: maybe when coredumping read a new xattr from /proc/$PID/exe that |
141 | may be used to mark a whole binary as non-coredumpable. Would fix: | |
142 | https://bugs.freedesktop.org/show_bug.cgi?id=69447 | |
143 | ||
f3e361c1 LP |
144 | * teach parse_timestamp() timezones like the calendar spec already knows it |
145 | ||
9ef3376b LP |
146 | * beef up hibernation to optionally do swapon/swapoff immediately before/after |
147 | the hibernation | |
148 | ||
c846b233 LP |
149 | * beef up s2h to implement a battery watch loop: instead of entering |
150 | hibernation unconditionally after coming back from resume make a decision | |
151 | based on the battery load level: if battery level is above a specific | |
152 | threshold, go to suspend again, only hibernate if below it. This means we'd | |
153 | stick to suspend usually, but fall back to hibernation only when battery runs | |
154 | empty (well, subject to our sampling interval). Related to this, check if we | |
155 | can make ACPI _BTP (i.e. /sys/class/power_supply/*/alarm) work for us too, | |
156 | i.e. see if it can wake up machines from suspend, so that we could resume | |
157 | automatically when the system is low on power and move automatically to | |
158 | hibernation mode. (see | |
159 | https://uefi.org/sites/default/files/resources/ACPI%206_2_A_Sept29.pdf | |
160 | section 10.2.2.8 and | |
161 | https://docs.microsoft.com/en-us/windows-hardware/design/device-experiences/modern-standby-wake-sources | |
162 | at the end). | |
163 | ||
f293fc51 LP |
164 | * add an explicit "vertical" mode to format-table, so that "systemctl |
165 | status"-like outputs (i.e. with a series of field names left and values | |
166 | right) become genuine first class citizens, and we gain automatic, sane JSON | |
167 | output for them. | |
168 | ||
169 | * dissector: invoke fsck on the file systems we encounter, after all ext4 is | |
170 | still pretty popular (and we mount the ESP too with it after all, which is | |
171 | fat) | |
172 | ||
173 | * systemd-firstboot: teach it dissector magic, so that you can point it to some | |
174 | disk image and it will just set everything in it all behind the scenes. | |
175 | ||
176 | * systemd-firstboot: add --force mode that replaces existing configuration. | |
177 | ||
9c230b8f LP |
178 | * We should probably replace /var/log/README, /etc/rc.d/README with symlinks |
179 | that are linked to these places instead of copied. After all they are | |
180 | constant vendor data. | |
181 | ||
312dc153 LP |
182 | * maybe add kernel cmdline params: 1) to force first-boot mode + 2) to force |
183 | random seed crediting | |
184 | ||
185 | * nspawn: on cgroupsv1 issue cgroup empty handler process based on host events, | |
186 | so that we make cgroup agent logic safe | |
187 | ||
188 | * nspawn/machined: add API to invoke binary in container, then use that as | |
189 | fallback in "machinectl shell" | |
72bd8525 | 190 | |
ac5dca64 LP |
191 | * logind: rework pam_logind to also do a bus call in case of invocation from |
192 | user@.service, which returns the XDG_RUNTIME_DIR value, and make this | |
193 | behaviour selectable via pam module option. | |
194 | ||
195 | * introduce a new per-process uuid, similar to the boot id, the machine id, the | |
196 | invocation id, that is derived from process creds, specifically a hashed | |
197 | combination of AT_RANDOM + getpid() + the starttime from | |
198 | /proc/self/status. Then add these ids implicitly when logging. Deriving this | |
199 | uuid from these three things has the benefit that it can be derived easily | |
200 | from /proc/$PID/ in a stable, and unique way that changes on both fork() and | |
201 | exec(). | |
202 | ||
203 | * let's not GC a unit while its ratelimits are still pending | |
204 | ||
5daeeecf LP |
205 | * when killing due to service watchdog timeout maybe detect whether target |
206 | process is under ptracing and then log loudly and continue instead. | |
207 | ||
5802d977 LP |
208 | * introduce a new group to own TPM devices |
209 | ||
210 | * make rfkill uaccess controllable by default, i.e. steal rule from | |
211 | gnome-bluetooth and friends | |
212 | ||
d84520e9 LP |
213 | * tweak journald context caching. In addition to caching per-process attributes |
214 | keyed by PID, cache per-cgroup attributes (i.e. the various xattrs we read) | |
215 | keyed by cgroup path, and guarded by ctime changes. This should provide us | |
216 | with a nice speed-up on services that have many processes running in the same | |
217 | cgroup. | |
218 | ||
03ccc7f0 LP |
219 | * make MAINPID= message reception checks even stricter: if service uses User=, |
220 | then check sending UID and ignore message if it doesn't match the user or | |
221 | root. | |
222 | ||
223 | * maybe trigger a uevent "change" on a device if "systemctl reload xyz.device" | |
224 | is issued. | |
225 | ||
db3cea22 LP |
226 | * when importing an fs tree with machined, optionally apply userns-rec-chown |
227 | ||
228 | * when importing an fs tree with machined, complain if image is not an OS | |
229 | ||
707b3fbd LP |
230 | * Maybe introduce a helper safe_exec() or so, which is to execve() which |
231 | safe_fork() is to fork(). And then make revert the RLIMIT_NOFILE soft limit | |
232 | to 1K implicitly, unless explicitly opted-out. | |
d7b659ef | 233 | |
d238709c | 234 | * rework seccomp/nnp logic that even if User= is used in combination with |
d96c081a LP |
235 | a seccomp option we don't have to set NNP. For that, change uid first whil |
236 | keeping CAP_SYS_ADMIN, then apply seccomp, the drop cap. | |
237 | ||
06bd149a LP |
238 | * add a concept for automatically loading per-unit secrets off disk and |
239 | inserting them into the kernel keyring. Maybe SecretsDirectory= similar to | |
240 | ConfigurationDirectory=. | |
241 | ||
8a7cf157 LP |
242 | * when no locale is configured, default to UEFI's PlatformLang variable |
243 | ||
d3584ec4 | 244 | * bootctl,sd-boot: actually honour the "architecture" key |
fe1bff15 | 245 | |
d96c081a LP |
246 | * add a new syscall group "@esoteric" for more esoteric stuff such as bpf() and |
247 | usefaultd() and make systemd-analyze check for it. | |
248 | ||
06898123 | 249 | * paranoia: whenever we process passwords, call mlock() on the memory |
309c6b19 | 250 | first. i.e. look for all places we use free_and_erasep() and |
9ae4ef49 | 251 | augment them with mlock(). Also use MADV_DONTDUMP. |
06898123 | 252 | |
32875617 LP |
253 | * Move RestrictAddressFamily= to the new cgroup create socket |
254 | ||
255 | * support the bind/connect/sendmsg cgroup stuff for sandboxing, and possibly | |
256 | patching around | |
257 | ||
76853293 LP |
258 | * maybe implicitly attach monotonic+realtime timestamps to outgoing messages in |
259 | log.c and sd-journal-send | |
260 | ||
06898123 LP |
261 | * optionally: turn on cgroup delegation for per-session scope units |
262 | ||
d3aeddb8 LP |
263 | * introduce per-unit (i.e. per-slice, per-service) journal log size limits. |
264 | ||
42e18088 LP |
265 | * sd-boot: automatically load EFI modules from some drop-in dir, so that people |
266 | can add in file system drivers and such | |
267 | ||
42e18088 LP |
268 | * sd-boot: optionally, show boot menu when previous default boot item has |
269 | non-zero "tries done" count | |
270 | ||
8f2eb730 LP |
271 | * introduce an option (or replacement) for "systemctl show" that outputs all |
272 | properties as JSON, similar to busctl's new JSON output. In contrast to that | |
273 | it should skip the variant type string though. | |
274 | ||
275 | * augment CODE_FILE=, CODE_LINE= with something like CODE_BASE= or so which | |
276 | contains some identifier for the project, which allows us to include | |
277 | clickable links to source files generating these log messages. The identifier | |
278 | could be some abberviated URL prefix or so (taking inspiration from Go | |
279 | imports). For example, for systemd we could use | |
280 | CODE_BASE=github.com/systemd/systemd/blob/98b0b1123cc or so which is | |
281 | sufficient to build a link by prefixing "http://" and suffixing the | |
282 | CODE_FILE. | |
283 | ||
8f2eb730 LP |
284 | * Augment MESSAGE_ID with MESSAGE_BASE, in a similar fashion so that we can |
285 | make clickable links from log messages carrying a MESSAGE_ID, that lead to | |
286 | some explanatory text online. | |
287 | ||
7bd4bcf7 LP |
288 | * maybe extend .path units to expose fanotify() per-mount change events |
289 | ||
a33ffd15 LP |
290 | * Add a "systemctl list-units --by-slice" mode or so, which rearranges the |
291 | output of "systemctl list-units" slightly by showing the tree structure of | |
292 | the slices, and the units attached to them. | |
293 | ||
20d4b101 LP |
294 | * the a-posteriori stopping of units bound to units that disappeared logic |
295 | should be reworked: there should be a queue of units, and we should only | |
296 | enqeue stop jobs from a defer event that processes queue instead of | |
297 | right-away when we find a unit that is bound to one that doesn't exist | |
298 | anymore. (similar to how the stop-unneeded queue has been reworked the same | |
299 | way) | |
4e8113bb | 300 | |
126708fa LP |
301 | * nspawn: make nspawn suitable for shell pipelines: instead of triggering a |
302 | hangup when input is finished, send ^D, which synthesizes an EOF. Then wait | |
303 | for hangup or ^D before passing on the EOF. | |
304 | ||
bb527e11 LP |
305 | * When reloading configuration PID 1 should reset all its properties to the |
306 | original defaults before calling parse_config() | |
307 | ||
72d711ef | 308 | * nspawn: greater control over selinux label? |
c7db2a9d | 309 | |
c633b0a6 LP |
310 | * hibernate/s2h: make this robust and safe to enable in Fedora by default. |
311 | Specifically: | |
312 | ||
313 | 1. add resume_offset support to the resume code (i.e. support swap files | |
314 | properly) | |
e83419d0 | 315 | 2. check if swap is on weird storage and refuse if so |
46c41478 | 316 | 3. add auto-detection of hibernation images |
c633b0a6 | 317 | |
070d0ac9 LP |
318 | * cgroups: use inotify to get notified when somebody else modifies cgroups |
319 | owned by us, then log a friendly warning. | |
320 | ||
e44924f5 LP |
321 | * beef up log.c with support for stripping ANSI sequences from strings, so that |
322 | it is OK to include them in log strings. This would be particularly useful so | |
323 | that our log messages could contain clickable links for example for unit | |
324 | files and suchlike we operate on. | |
070d0ac9 | 325 | |
8c4c2dfc LP |
326 | * add support for "portablectl attach http://foobar.com/waaa.raw (i.e. importd integration) |
327 | ||
328 | * add attach --enable and attach --now (for attach+enable+start) | |
329 | ||
330 | * sync dynamic uids/gids between host+portable srvice (i.e. if DynamicUser=1 is set for a service, make sure that the | |
331 | selected user is resolvable in the service even if it ships its own /etc/passwd) | |
332 | ||
5da19043 | 333 | * Fix DECIMAL_STR_MAX or DECIMAL_STR_WIDTH. One includes a trailing NUL, the |
5238e957 | 334 | other doesn't. What a disaster. Probably to exclude it. Also |
5da19043 LP |
335 | DECIMAL_STR_WIDTH should probably add an extra "-" into account for negative |
336 | numbers. | |
337 | ||
16270697 LP |
338 | * Check that users of inotify's IN_DELETE_SELF flag are using it properly, as |
339 | usually IN_ATTRIB is the right way to watch deleted files, as the former only | |
340 | fires when a file is actually removed from disk, i.e. the link count drops to | |
341 | zero and is not open anymore, while the latter happens when a file is | |
342 | unlinked from any dir. | |
343 | ||
bd1b3f75 | 344 | * port systemctl, busctl, … over to format-table.[ch]'s table formatters |
5da19043 | 345 | |
7bc756ff LP |
346 | * pid1: lock image configured with RootDirectory=/RootImage= using the usual nspawn semantics while the unit is up |
347 | ||
348 | * add --vacuum-xyz options to coredumpctl, matching those journalctl already has. | |
349 | ||
53c70a27 LP |
350 | * introduce Ephemeral= unit file switch, that creates an ephemeral copy of all |
351 | files and directories that are left writable for a unit, and which are | |
352 | removed after the unit goes down again. A bit like --ephemeral for | |
353 | systemd-nspawn but for system services. If used together with RootImage= this | |
354 | should reflink the image file itself. | |
355 | ||
356 | Related: add Ephemeral=<path1> <path2> … which would allow marking | |
357 | specific paths only like this. | |
358 | ||
53c70a27 | 359 | * add CopyFile= or so as unit file setting that may be used to copy files or |
5238e957 | 360 | directory trees from the host to the services RootImage= and RootDirectory= |
53c70a27 LP |
361 | environment. Which we can use for /etc/machine-id and in particular |
362 | /etc/resolv.conf. Should be smart and do something useful on read-only | |
363 | images, for example fallback to read-only bind mounting the file instead. | |
364 | ||
d9b50610 LP |
365 | * show invocation ID in systemd-run output |
366 | ||
367 | * bypass SIGTERM state in unit files if KillSignal is SIGKILL | |
368 | ||
9711b1ad LP |
369 | * add proper dbus APIs for the various sd_notify() commands, such as MAINPID=1 |
370 | and so on, which would mean we could report errors and such. | |
371 | ||
1e784321 LP |
372 | * teach tmpfiles.d q/Q logic something sensible in the context of XFS/ext4 |
373 | project quota | |
0d545655 | 374 | |
6b7b0f39 LP |
375 | * introduce DefaultSlice= or so in system.conf that allows changing where we |
376 | place our units by default, i.e. change system.slice to something | |
377 | else. Similar, ManagerSlice= should exist so that PID1's own scope unit could | |
378 | be moved somewhere else too. Finally machined and logind should get similar | |
379 | options so that it is possible to move user session scopes and machines to a | |
380 | different slice too by default. Usecase: people who want to put resources on | |
381 | the entire system, with the exception of one specific service. See: | |
382 | https://lists.freedesktop.org/archives/systemd-devel/2018-February/040369.html | |
383 | ||
aa79f932 LP |
384 | * maybe rework get_user_creds() to query the user database if $SHELL is used |
385 | for root, but only then. | |
586a8e93 | 386 | |
46099c9e LP |
387 | * be stricter with fds we receive for the fdstore: close them asynchronously |
388 | ||
46099c9e LP |
389 | * calenderspec: add support for week numbers and day numbers within a |
390 | year. This would allow us to define "bi-weekly" triggers safely. | |
391 | ||
c3cd7cc9 LP |
392 | * sd-bus: add vtable flag, that may be used to request client creds implicitly |
393 | and asynchronously before dispatching the operation | |
394 | ||
06345858 LP |
395 | * make use of ethtool veth peer info in machined, for automatically finding out |
396 | host-side interface pointing to the container. | |
397 | ||
48f1b5e5 LP |
398 | * add some special mode to LogsDirectory=/StateDirectory=… that allows |
399 | declaring these directories without necessarily pulling in deps for them, or | |
400 | creating them when starting up. That way, we could declare that | |
401 | systemd-journald writes to /var/log/journal, which could be useful when we | |
402 | doing disk usage calculations and so on. | |
403 | ||
cb77e122 | 404 | * taint systemd if there are fewer than 65536 users assigned (userns) to the system. |
c6009ff0 | 405 | |
899feb72 | 406 | * deprecate RootDirectoryStartOnly= in favour of a new ExecStart= prefix char |
c6009ff0 | 407 | |
3d80d454 LP |
408 | * add a new RuntimeDirectoryPreserve= mode that defines a similar lifecycle for |
409 | the runtime dir as we maintain for the fdstore: i.e. keep it around as long | |
410 | as the unit is running or has a job queued. | |
411 | ||
5f7ecd61 | 412 | * support projid-based quota in machinectl for containers |
5962e9db | 413 | |
17b6f896 LP |
414 | * maybe use SOURCE_DATE_EPOCH (i.e. the env var the reproducible builds folks |
415 | introduced) as the RTC epoch, instead of the mtime of NEWS. | |
416 | ||
17b6f896 LP |
417 | * add a way to lock down cgroup migration: a boolean, which when set for a unit |
418 | makes sure the processes in it can never migrate out of it | |
419 | ||
17b6f896 LP |
420 | * blog about fd store and restartable services |
421 | ||
422 | * document Environment=SYSTEMD_LOG_LEVEL=debug drop-in in debugging document | |
423 | ||
370f9c21 LP |
424 | * rework ExecOutput and ExecInput enums so that EXEC_OUTPUT_NULL loses its |
425 | magic meaning and is no longer upgraded to something else if set explicitly. | |
426 | ||
17b6f896 LP |
427 | * in the long run: permit a system with /etc/machine-id linked to /dev/null, to |
428 | make it lose its identity, i.e. be anonymous. For this we'd have to patch | |
429 | through the whole tree to make all code deal with the case where no machine | |
430 | ID is available. | |
431 | ||
432 | * optionally, collect cgroup resource data, and store it in per-unit RRD files, | |
433 | suitable for processing with rrdtool. Add bus API to access this data, and | |
434 | possibly implement a CPULoad property based on it. | |
435 | ||
e7e4a258 LP |
436 | * beef up pam_systemd to take unit file settings such as cgroups properties as |
437 | parameters | |
438 | ||
6fc373ee LP |
439 | * maybe hook of xfs/ext4 quotactl() with services? i.e. automatically manage |
440 | the quota of a the user indicated in User= via unit file settings, like the | |
5962e9db LP |
441 | other resource management concepts. Would mix nicely with DynamicUser=1. Or |
442 | alternatively, do this with projids, so that we can also cover services | |
443 | running as root. Quota should probably cover all the special dirs such as | |
444 | StateDirectory=, LogsDirectory=, CacheDirectory=, as well as RootDirectory= if it | |
445 | is set, plus the whole disk space any image configured with RootImage=. | |
6fc373ee | 446 | |
9a92a97a LP |
447 | * In DynamicUser= mode: before selecting a UID, use disk quota APIs on relevant |
448 | disks to see if the UID is already in use. | |
449 | ||
2368e974 LP |
450 | * add "systemctl wait" or so, which does what "systemd-run --wait" does, but |
451 | for all units. It should be both a way to pin units into memory as well as a | |
452 | wait to retrieve their exit data. | |
453 | ||
fa991fb7 LP |
454 | * expose IO accounting data on the bus, show it in systemd-run --wait and log |
455 | about it in the resource log message | |
456 | ||
fa991fb7 LP |
457 | * show whether a service has out-of-date configuration in "systemctl status" by |
458 | using mtime data of ConfigurationDirectory=. | |
459 | ||
d73b607d LP |
460 | * Add AddUser= setting to unit files, similar to DynamicUser=1 which however |
461 | creates a static, persistent user rather than a dynamic, transient user. We | |
462 | can leverage code from sysusers.d for this. | |
463 | ||
fd63f36c LP |
464 | * add some optional flag to ReadWritePaths= and friends, that has the effect |
465 | that we create the dir in question when the service is started. Example: | |
466 | ||
467 | ReadWritePaths=:/var/lib/foobar | |
468 | ||
3576dddf LP |
469 | * maybe add call sd_journal_set_block_timeout() or so to set SO_SNDTIMEO for |
470 | the sd-journal logging socket, and, if the timeout is set to 0, sets | |
471 | O_NONBLOCK on it. That way people can control if and when to block for | |
472 | logging. | |
473 | ||
fe6a0235 LP |
474 | * hostnamed: populate form factor data from a new hwdb database, so that old |
475 | yogas can be recognized as "convertible" too, even if they predate the DMI | |
476 | "convertible" form factor | |
477 | ||
d543b9f5 LP |
478 | * Maybe add PrivatePIDs= as new unit setting, and do minimal PID namespacing |
479 | after all. Be strict however, only support the equivalent of nspawn's | |
480 | --as-pid2 switch, and sanely proxy sd_notify() messages dropping stuff such | |
481 | as MAINPID. | |
482 | ||
f59d1da8 LP |
483 | * Add ExecMonitor= setting. May be used multiple times. Forks off a process in |
484 | the service cgroup, which is supposed to monitor the service, and when it | |
485 | exits the service is considered failed by its monitor. | |
486 | ||
33bac67b LP |
487 | * track the per-service PAM process properly (i.e. as an additional control |
488 | process), so that it may be queried on the bus and everything. | |
489 | ||
490 | * add a new "debug" job mode, that is propagated to unit_start() and for | |
491 | services results in two things: we raise SIGSTOP right before invoking | |
492 | execve() and turn off watchdog support. Then, use that to implement | |
493 | "systemd-gdb" for attaching to the start-up of any system service in its | |
494 | natural habitat. | |
495 | ||
8eb7383b LP |
496 | * maybe add gpt-partition-based user management: each user gets his own |
497 | LUKS-encrypted GPT partition with a new GPT type. A small nss module | |
498 | enumerates users via udev partition enumeration. UIDs are assigned in a fixed | |
499 | way: the partition index is added as offset to some fixed base uid. User name | |
500 | is stored in GPT partition name. A PAM module authenticates the user via the | |
501 | LUKS partition password. Benefits: strong per-user security, compatibility | |
502 | with stateless/read-only/verity-enabled root. (other idea: do this based on | |
503 | loopback files in /home, without GPT involvement) | |
504 | ||
8eb7383b LP |
505 | * gpt-auto logic: related to the above, maybe support a "secondary" root |
506 | partition, that is mounted to / and is writable, and where the actual root's | |
507 | /usr is mounted into. | |
508 | ||
8eb7383b LP |
509 | * 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 |
510 | ||
67234d21 LP |
511 | * drop nss-myhostname in favour of nss-resolve? |
512 | ||
b5bdbcd5 LP |
513 | * add a percentage syntax for TimeoutStopSec=, e.g. TimeoutStopSec=150%, and |
514 | then use that for the setting used in user@.service. It should be understood | |
515 | relative to the configured default value. | |
516 | ||
d21494ea LP |
517 | * in networkd, when matching device types, fix up DEVTYPE rubbish the kernel passes to us |
518 | ||
519 | * enable LockMLOCK to take a percentage value relative to physical memory | |
520 | ||
04397464 | 521 | * Permit masking specific netlink APIs with RestrictAddressFamily= |
d82047be | 522 | |
d82047be LP |
523 | * nspawn: support that /proc, /sys/, /dev are pre-mounted |
524 | ||
d82047be LP |
525 | * define gpt header bits to select volatility mode |
526 | ||
42d61ded LP |
527 | * ProtectClock= (drops CAP_SYS_TIMES, adds seecomp filters for settimeofday, adjtimex), sets DeviceAllow o /dev/rtc |
528 | ||
04397464 | 529 | * ProtectTracing= (drops CAP_SYS_PTRACE, blocks ptrace syscall, makes /sys/kernel/tracing go away) |
42d61ded | 530 | |
04397464 | 531 | * ProtectMount= (drop mount/umount/pivot_root from seccomp, disallow fuse via DeviceAllow, imply Mountflags=slave) |
563a69f4 | 532 | |
04397464 | 533 | * ProtectKeyRing= to take keyring calls away |
2c5f2958 | 534 | |
8ce9b83a | 535 | * RemoveKeyRing= to remove all keyring entries of the specified user |
a46eac1b LP |
536 | |
537 | * ProtectReboot= that masks reboot() and kexec_load() syscalls, prohibits kill | |
538 | on PID 1 with the relevant signals, and makes relevant files in /sys and | |
539 | /proc (such as the sysrq stuff) unavailable | |
540 | ||
e40a326c | 541 | * make sure the ratelimit object can deal with USEC_INFINITY as way to turn off things |
89f193fa | 542 | |
6809de5b LP |
543 | * journalctl: make sure -f ends when the container indicated by -M terminates |
544 | ||
2c5f2958 LP |
545 | * mount: automatically search for "main" partition of an image has multiple |
546 | partitions | |
547 | ||
8ce9b83a LP |
548 | * in nss-systemd, if we run inside of RootDirectory= with PrivateUsers= set, |
549 | find a way to map the User=/Group= of the service to the right name. This way | |
550 | a user/group for a service only has to exist on the host for the right | |
551 | mapping to work. | |
552 | ||
2c5f2958 LP |
553 | * add bus API for creating unit files in /etc, reusing the code for transient units |
554 | ||
555 | * add bus API to remove unit files from /etc | |
556 | ||
557 | * add bus API to retrieve current unit file contents (i.e. implement "systemctl cat" on the bus only) | |
558 | ||
b8c7afdf LP |
559 | * rework fopen_temporary() to make use of open_tmpfile_linkable() (problem: the |
560 | kernel doesn't support linkat() that replaces existing files, currently) | |
561 | ||
1e555cb5 LP |
562 | * transient units: don't bother with actually setting unit properties, we |
563 | reload the unit file anyway | |
564 | ||
1e555cb5 LP |
565 | * journald: sigbus API via a signal-handler safe function that people may call |
566 | from the SIGBUS handler | |
567 | ||
c8048350 LP |
568 | * optionally, also require WATCHDOG=1 notifications during service start-up and shutdown |
569 | ||
648b1220 LP |
570 | * delay activation of logind until somebody logs in, or when /dev/tty0 pulls it |
571 | in or lingering is on (so that containers don't bother with it until PAM is used). also exit-on-idle | |
572 | ||
3d39e6e5 LP |
573 | * cache sd_event_now() result from before the first iteration... |
574 | ||
bd098bce LP |
575 | * PID1: find a way how we can reload unit file configuration for |
576 | specific units only, without reloading the whole of systemd | |
577 | ||
f9bf1b8f | 578 | * add an explicit parser for LimitRTPRIO= that verifies |
de7399eb | 579 | the specified range and generates sane error messages for incorrect |
f9bf1b8f | 580 | specifications. |
de7399eb | 581 | |
3efc8c72 LP |
582 | * when we detect that there are waiting jobs but no running jobs, do something |
583 | ||
5ebbb45b | 584 | * push CPUAffinity= also into the "cpuset" cgroup controller |
03364e47 | 585 | |
e6a26d8c LP |
586 | * PID 1 should send out sd_notify("WATCHDOG=1") messages (for usage in the --user mode, and when run via nspawn) |
587 | ||
a2088fd0 | 588 | * there's probably something wrong with having user mounts below /sys, |
5238e957 | 589 | as we have for debugfs. for example, src/core/mount.c handles mounts |
a2088fd0 LP |
590 | prefixed with /sys generally special. |
591 | http://lists.freedesktop.org/archives/systemd-devel/2015-June/032962.html | |
592 | ||
8aa20381 LP |
593 | * fstab-generator: default to tmpfs-as-root if only usr= is specified on the kernel cmdline |
594 | ||
b18d23d7 LP |
595 | * docs: bring http://www.freedesktop.org/wiki/Software/systemd/MyServiceCantGetRealtime up to date |
596 | ||
60d17b74 LP |
597 | * add a job mode that will fail if a transaction would mean stopping |
598 | running units. Use this in timedated to manage the NTP service | |
599 | state. | |
600 | http://lists.freedesktop.org/archives/systemd-devel/2015-April/030229.html | |
601 | ||
477e75ef LP |
602 | * The udev blkid built-in should expose a property that reflects |
603 | whether media was sensed in USB CF/SD card readers. This should then | |
604 | be used to control SYSTEMD_READY=1/0 so that USB card readers aren't | |
605 | picked up by systemd unless they contain a medium. This would mirror | |
606 | the behaviour we already have for CD drives. | |
607 | ||
2bc39683 LP |
608 | * networkd/udev: implement SR_IOV configuration in .link files: |
609 | http://lists.freedesktop.org/archives/systemd-devel/2015-January/027451.html | |
610 | ||
c3a0d00d LP |
611 | * hostnamectl: show root image uuid |
612 | ||
d2f81fb0 LP |
613 | * Find a solution for SMACK capabilities stuff: |
614 | http://lists.freedesktop.org/archives/systemd-devel/2014-December/026188.html | |
98cd2651 | 615 | |
2a441c8a LP |
616 | * "systemctl preset-all" should probably order the unit files it |
617 | operates on lexicographically before starting to work, in order to | |
618 | ensure deterministic behaviour if two unit files conflict (like DMs | |
619 | do, for example) | |
620 | ||
0a86c1a9 LP |
621 | * synchronize console access with BSD locks: |
622 | http://lists.freedesktop.org/archives/systemd-devel/2014-October/024582.html | |
623 | ||
e031c227 | 624 | * as soon as we have sender timestamps, revisit coalescing multiple parallel daemon reloads: |
0a86c1a9 LP |
625 | http://lists.freedesktop.org/archives/systemd-devel/2014-December/025862.html |
626 | ||
8514b677 LP |
627 | * figure out when we can use the coarse timers |
628 | ||
7b0fce61 LP |
629 | * add "systemctl start -v foobar.service" that shows logs of a service |
630 | while the start command runs. This is non-trivial to do without | |
631 | races though, since we should flush out all journal messages before | |
632 | returning from the "systemctl stop". | |
633 | ||
b568ef14 | 634 | * 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 |
c2d5b3c9 | 635 | |
d48b7bd2 LP |
636 | * maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment= |
637 | ||
8483d73f LP |
638 | * maybe allow timer units with an empty Units= setting, so that they |
639 | can be used for resuming the system but nothing else. | |
640 | ||
25e773ee | 641 | * what to do about udev db binary stability for apps? (raw access is not an option) |
b857e042 | 642 | |
279f0366 | 643 | * systemctl: if some operation fails, show log output? |
533bb267 | 644 | |
5238e957 | 645 | * systemctl edit: use equivalent of cat() to insert existing config as a comment, prepended with #. |
b42de08a | 646 | Upon editor exit, lines with one # are removed, lines with two # are left with one #, etc. |
202aea45 | 647 | |
720652b3 | 648 | * exponential backoff in timesyncd when we cannot reach a server |
42aeb14a | 649 | |
720652b3 | 650 | * timesyncd: add ugly bus calls to set NTP servers per-interface, for usage by NM |
e25b5a8d | 651 | |
563b1bdc LP |
652 | * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share.... |
653 | ||
5dcf9838 LP |
654 | * systemd.show_status= should probably have a mode where only failed |
655 | units are shown. | |
656 | ||
6bd7941e TG |
657 | * add systemd.abort_on_kill or some other such flag to send SIGABRT instead of SIGKILL |
658 | (throughout the codebase, not only PID1) | |
659 | ||
9d6db739 | 660 | * resolved: |
9d6db739 | 661 | - mDNS/DNS-SD |
ccc3e8a1 LP |
662 | - service registration |
663 | - service/domain/types browsing | |
0f47ed0a | 664 | - avahi compat |
9d6db739 | 665 | - DNS-SD service registration from socket units |
e25b5a8d DH |
666 | - resolved should optionally register additional per-interface LLMNR |
667 | names, so that for the container case we can establish the same name | |
668 | (maybe "host") for referencing the server, everywhere. | |
720652b3 | 669 | - allow clients to request DNSSEC for a single lookup even if DNSSEC is off (?) |
3efb871a | 670 | - hook up resolved with machined-based address resolution |
3f77a1b1 | 671 | |
e25b5a8d | 672 | * refcounting in sd-resolve is borked |
e2a69298 | 673 | |
2834ffe7 LP |
674 | * Add a new verb "systemctl top" |
675 | ||
a940778f LP |
676 | * add new gpt type for btrfs volumes |
677 | ||
a01647e5 | 678 | * support empty /etc boots nicely: |
3c779fa5 | 679 | - nspawn/gpt-generator: introduce new gpt partition type for /usr |
a01647e5 | 680 | |
3de03738 LP |
681 | * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them. |
682 | ||
37efac5d LP |
683 | * a way for container managers to turn off getty starting via $container_headless= or so... |
684 | ||
7348b3ad LP |
685 | * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit |
686 | ||
81429136 KS |
687 | * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services |
688 | they run added to the initial transaction and thus confuse Type=idle. | |
e107ed18 | 689 | |
edb2935c LP |
690 | * add bus api to query unit file's X fields. |
691 | ||
6a3f892a | 692 | * gpt-auto-generator: |
2a781fc9 | 693 | - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap? |
6a3f892a | 694 | - Make /home automount rather than mount? |
6a3f892a | 695 | |
65026403 LP |
696 | * add generator that pulls in systemd-network from containers when |
697 | CAP_NET_ADMIN is set, more than the loopback device is defined, even | |
698 | when it is otherwise off | |
f8901862 | 699 | |
f9bf1b8f | 700 | * MessageQueueMessageSize= (and suchlike) should use parse_iec_size(). |
eda8f067 | 701 | |
af1082b0 LP |
702 | * implement Distribute= in socket units to allow running multiple |
703 | service instances processing the listening socket, and open this up | |
704 | for ReusePort= | |
705 | ||
09ecd746 LP |
706 | * introduce bus call FreezeUnit(s, b), as well as "systemctl freeze |
707 | $UNIT" and "systemctl thaw $UNIT" as wrappers around this. The calls | |
708 | should SIGSTOP all unit processes in a loop until all processes of | |
709 | it are fully stopped. This can later be used for app management by | |
710 | desktop UIs such as gnome-shell to freeze apps that are not visible | |
711 | on screen, not unlike how job control works on the shell | |
712 | ||
f38afcd0 | 713 | * cgroups: |
f38afcd0 | 714 | - implement per-slice CPUFairScheduling=1 switch |
f38afcd0 LP |
715 | - introduce high-level settings for RT budget, swappiness |
716 | - how to reset dynamically changed unit cgroup attributes sanely? | |
717 | - when reloading configuration, apply new cgroup configuration | |
718 | - when recursively showing the cgroup hierarchy, optionally also show | |
719 | the hierarchies of child processes | |
0bee65f0 | 720 | |
f38afcd0 | 721 | * transient units: |
f38afcd0 | 722 | - add field to transient units that indicate whether systemd or somebody else saves/restores its settings, for integration with libvirt |
ebcf1f97 | 723 | |
718db961 LP |
724 | * when we detect low battery and no AC on boot, show pretty splash and refuse boot |
725 | ||
718db961 LP |
726 | * libsystemd-journal, libsystemd-login, libudev: add calls to easily attach these objects to sd-event event loops |
727 | ||
966204e0 LP |
728 | * be more careful what we export on the bus as (usec_t) 0 and (usec_t) -1 |
729 | ||
41644622 LP |
730 | * 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 |
731 | ||
7f79cd71 | 732 | * After coming back from hibernation reset hibernation swap partition using the /dev/snapshot ioctl APIs |
0aafd43d | 733 | |
19aadacf JE |
734 | * If we try to find a unit via a dangling symlink, generate a clean |
735 | error. Currently, we just ignore it and read the unit from the search | |
df5f6971 LP |
736 | path anyway. |
737 | ||
04397464 | 738 | * refuse boot if /usr/lib/os-release is missing or /etc/machine-id cannot be set up |
fcba531e | 739 | |
bdeeb6b5 LP |
740 | * man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted. |
741 | ||
742 | * load .d/*.conf dropins for device units | |
743 | ||
6bb648a1 | 744 | * sd-bus: |
0d43ffef | 745 | - EBADSLT handling |
fa607802 | 746 | - GetAllProperties() on a non-existing object does not result in a failure currently |
7212c608 | 747 | - port to sd-resolve for connecting to TCP dbus servers |
57f2a947 | 748 | - 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 |
f38afcd0 | 749 | - see if we can drop more message validation on the sending side |
f38afcd0 | 750 | - add API to clone sd_bus_message objects |
ab9716c2 | 751 | - longer term: priority inheritance |
df98a87b | 752 | - dbus spec updates: |
df98a87b LP |
753 | - NameLost/NameAcquired obsolete |
754 | - GVariant | |
2ecfc64e | 755 | - path escaping |
6a8b5fa4 | 756 | - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now |
f38afcd0 LP |
757 | |
758 | * sd-event | |
12179984 | 759 | - allow multiple signal handlers per signal? |
0536ce5d | 760 | - document chaining of signal handler for SIGCHLD and child handlers |
e25b5a8d | 761 | - define more intervals where we will shift wakeup intervals around in, 1h, 6h, 24h, ... |
eb01ba5d | 762 | |
69af4503 | 763 | * investigate endianness issues of UUID vs. GUID |
e8a7a315 | 764 | |
461bd8e4 LP |
765 | * dbus: when a unit failed to load (i.e. is in UNIT_ERROR state), we |
766 | should be able to safely try another attempt when the bus call LoadUnit() is invoked. | |
767 | ||
ab8e074c LP |
768 | * add a pam module that passes the hdd passphrase into the PAM stack and then expires it, for usage by gdm auto-login. |
769 | ||
770 | * add a pam module that on password changes updates any LUKS slot where the password matches | |
771 | ||
fff87a35 | 772 | * test/: |
20d52ab6 | 773 | - add unit tests for config_parse_device_allow() |
b8b4d3dd | 774 | |
b5c03638 | 775 | * seems that when we follow symlinks to units we prefer the symlink |
d28315e4 | 776 | destination path over /etc and /usr. We should not do that. Instead |
b5c03638 LP |
777 | /etc should always override /run+/usr and also any symlink |
778 | destination. | |
779 | ||
eece8c6f LP |
780 | * when isolating, try to figure out a way how we implicitly can order |
781 | all units we stop before the isolating unit... | |
782 | ||
356ce991 LP |
783 | * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off}) |
784 | ||
c1446921 LP |
785 | * BootLoaderSpec: Clarify that the kernel has to be in $BOOT. Clarify |
786 | that the boot loader should be installed to the ESP. Define a way | |
787 | how an installer can figure out whether a BLS compliant boot loader | |
788 | is installed. | |
795607b2 | 789 | |
0be8342c LP |
790 | * think about requeuing jobs when daemon-reload is issued? usecase: |
791 | the initrd issues a reload after fstab from the host is accessible | |
792 | and we might want to requeue the mounts local-fs acquired through | |
793 | that automatically. | |
794 | ||
e5ec62c5 | 795 | * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep() |
54c31a79 | 796 | |
ccddd104 | 797 | * remove any syslog support from log.c — we probably cannot do this before split-off udev is gone for good |
826872b6 | 798 | |
3679d112 LP |
799 | * shutdown logging: store to EFI var, and store to USB stick? |
800 | ||
356ce991 | 801 | * merge unit_kill_common() and unit_kill_context() |
490b7e47 | 802 | |
b44be3ec | 803 | * EFI: |
b44be3ec LP |
804 | - honor language efi variables for default language selection (if there are any?) |
805 | - honor timezone efi variables for default timezone selection (if there are any?) | |
466784c8 | 806 | - change bootctl to be backed by systemd-bootd to control temporary and persistent default boot goal plus efi variables |
631427d6 | 807 | * bootctl |
631427d6 | 808 | - recognize the case when not booted on EFI |
e4181484 | 809 | |
c0fe5db5 | 810 | * maybe do not install getty@tty1.service symlink in /etc but in /usr? |
e9fd44b7 | 811 | |
b44be3ec | 812 | * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word |
8ed20651 | 813 | |
b44be3ec | 814 | * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units. |
e6c6e7af | 815 | |
b44be3ec LP |
816 | * logind: |
817 | - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around | |
b44be3ec LP |
818 | - logind: wakelock/opportunistic suspend support |
819 | - Add pretty name for seats in logind | |
820 | - logind: allow showing logout dialog from system? | |
f38afcd0 LP |
821 | - add Suspend() bus calls which take timestamps to fix double suspend issues when somebody hits suspend and closes laptop quickly. |
822 | - if pam_systemd is invoked by su from a process that is outside of a | |
823 | any session we should probably just become a NOP, since that's | |
824 | usually not a real user session but just some system code that just | |
825 | needs setuid(). | |
279f0366 LP |
826 | - logind: make the Suspend()/Hibernate() bus calls wait for the for |
827 | the job to be completed. before returning, so that clients can wait | |
828 | for "systemctl suspend" to finish to know when the suspending is | |
829 | complete. | |
830 | - logind: when the power button is pressed short, just popup a | |
831 | logout dialog. If it is pressed for 1s, do the usual | |
832 | shutdown. Inspiration are Macs here. | |
279f0366 | 833 | - expose "Locked" property on logind sesison objects |
e25b5a8d DH |
834 | - maybe allow configuration of the StopTimeout for session scopes |
835 | - rename session scope so that it includes the UID. THat way | |
836 | the session scope can be arranged freely in slices and we don't have | |
837 | make assumptions about their slice anymore. | |
838 | - follow PropertiesChanged state more closely, to deal with quick logouts and | |
839 | relogins | |
e673ad04 | 840 | |
b44be3ec | 841 | * journal: |
57f2a947 | 842 | - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields |
b44be3ec LP |
843 | - journald: also get thread ID from client, plus thread name |
844 | - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups | |
845 | - add API to close/reopen/get fd for journal client fd in libsystemd-journal. | |
d28315e4 | 846 | - fallback to /dev/log based logging in libsystemd-journal, if we cannot log natively? |
b44be3ec | 847 | - declare the local journal protocol stable in the wiki interface chart |
b44be3ec LP |
848 | - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg |
849 | - journald: when dropping msgs due to ratelimit make sure to write | |
850 | "dropped %u messages" not only when we are about to print the next | |
5238e957 | 851 | message that works, but already after a short timeout |
b44be3ec LP |
852 | - check if we can make journalctl by default use --follow mode inside of less if called without args? |
853 | - maybe add API to send pairs of iovecs via sd_journal_send | |
f47ec8eb | 854 | - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access |
b44be3ec LP |
855 | - journactl: support negative filtering, i.e. FOOBAR!="waldo", |
856 | and !FOOBAR for events without FOOBAR. | |
038cf334 ZJS |
857 | - journal: store timestamp of journal_file_set_offline() int he header, |
858 | so it is possible to display when the file was last synced. | |
b44be3ec LP |
859 | - 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. |
860 | - journal: find a way to allow dropping history early, based on priority, other rules | |
861 | - journal: When used on NFS, check payload hashes | |
b44be3ec LP |
862 | - journald: add kernel cmdline option to disable ratelimiting for debug purposes |
863 | - refuse taking lower-case variable names in sd_journal_send() and friends. | |
864 | - journald: we currently rotate only after MaxUse+MaxFilesize has been reached. | |
865 | - journal: deal nicely with byte-by-byte copied files, especially regards header | |
b44be3ec | 866 | - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit |
601d9d6f | 867 | - Replace utmp, wtmp, btmp, and lastlog completely with journal |
f38afcd0 | 868 | - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax? |
f38afcd0 LP |
869 | - when a kernel driver logs in a tight loop, we should ratelimit that too. |
870 | - journald: optionally, log debug messages to /run but everything else to /var | |
871 | - journald: when we drop syslog messages because the syslog socket is | |
872 | full, make sure to write how many messages are lost as first thing | |
873 | to syslog when it works again. | |
f38afcd0 LP |
874 | - change systemd-journal-flush into a service that stays around during |
875 | boot, and causes the journal to be moved back to /run on shutdown, | |
d28315e4 | 876 | so that we do not keep /var busy. This needs to happen synchronously, |
f38afcd0 | 877 | hence doing this via signals is not going to work. |
279f0366 LP |
878 | - optionally support running journald from the command line for testing purposes in external projects |
879 | - journald: allow per-priority and per-service retention times when rotating/vacuuming | |
880 | - journald: make use of uid-range.h to managed uid ranges to split | |
881 | journals in. | |
882 | - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so... | |
883 | - improve journalctl performance by loading journal files | |
884 | lazily. Encode just enough information in the file name, so that we | |
885 | do not have to open it to know that it is not interesting for us, for | |
886 | the most common operations. | |
e25b5a8d | 887 | - man: document that corrupted journal files is nothing to act on |
e25b5a8d DH |
888 | - rework journald sigbus stuff to use mutex |
889 | - Set RLIMIT_NPROC for systemd-journal-xyz, and all other of our | |
890 | services that run under their own user ids, and use User= (but only | |
891 | in a world where userns is ubiquitous since otherwise we cannot | |
892 | invoke those daemons on the host AND in a container anymore). Also, | |
893 | if LimitNPROC= is used without User= we should warn and refuse | |
894 | operation. | |
895 | - journalctl --verify: don't show files that are currently being | |
896 | written to as FAIL, but instead show that their are being written to. | |
897 | - add journalctl -H that talks via ssh to a remote peer and passes through | |
898 | binary logs data | |
e25b5a8d | 899 | - add a version of --merge which also merges /var/log/journal/remote |
e25b5a8d DH |
900 | - journalctl: -m should access container journals directly by enumerating |
901 | them via machined, and also watch containers coming and going. | |
902 | Benefit: nspawn --ephemeral would start working nicely with the journal. | |
903 | - assign MESSAGE_ID to log messages about failed services | |
b44be3ec | 904 | |
590171d1 ZJS |
905 | * add a test if all entries in the catalog are properly formatted. |
906 | (Adding dashes in a catalog entry currently results in the catalog entry | |
907 | being silently skipped. journalctl --update-catalog must warn about this, | |
908 | and we should also have a unit test to check that all our message are OK.) | |
909 | ||
b44be3ec | 910 | * document: |
8b8f2591 | 911 | - document that deps in [Unit] sections ignore Alias= fields in |
b44be3ec LP |
912 | [Install] units of other units, unless those units are disabled |
913 | - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets | |
b44be3ec | 914 | - document that service reload may be implemented as service reexec |
f38afcd0 LP |
915 | - add a man page containing packaging guidelines and recommending usage of things like Documentation=, PrivateTmp=, PrivateNetwork= and ReadOnlyDirectories=/etc /usr. |
916 | - document systemd-journal-flush.service properly | |
f38afcd0 LP |
917 | - documentation: recommend to connect the timer units of a service to the service via Also= in [Install] |
918 | - man: document the very specific env the shutdown drop-in tools live in | |
f38afcd0 LP |
919 | - man: add more examples to man pages |
920 | - man: maybe sort directives in man pages, and take sections from --help and apply them to man too | |
17ec531f | 921 | - document root=gpt-auto properly |
b44be3ec LP |
922 | |
923 | * systemctl: | |
b44be3ec LP |
924 | - add systemctl switch to dump transaction without executing it |
925 | - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done | |
926 | - "systemctl disable" on a static unit prints no message and does | |
927 | nothing. "systemctl enable" does nothing, and gives a bad message | |
928 | about it. Should fix both to print nice actionable messages. | |
929 | - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service | |
930 | - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible | |
d28315e4 | 931 | - systemctl enable: fail if target to alias into does not exist? maybe show how many units are enabled afterwards? |
b44be3ec | 932 | - systemctl: "Journal has been rotated since unit was started." message is misleading |
61233823 | 933 | - systemctl status output should include list of triggering units and their status |
f38afcd0 LP |
934 | |
935 | * unit install: | |
936 | - "systemctl mask" should find all names by which a unit is accessible | |
937 | (i.e. by scanning for symlinks to it) and link them all to /dev/null | |
8b04b925 | 938 | |
b44be3ec | 939 | * timer units: |
b44be3ec | 940 | - timer units should get the ability to trigger when: |
b44be3ec | 941 | o DST changes |
90e071d1 | 942 | - Modulate timer frequency based on battery state |
d2e83c23 | 943 | |
a19554ed LP |
944 | * add libsystemd-password or so to query passwords during boot using the password agent logic |
945 | ||
a7a3f28b LP |
946 | * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed |
947 | ||
b44be3ec | 948 | * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel |
08f95888 | 949 | |
f59d1da8 | 950 | * make repeated alt-ctrl-del presses printing a dump |
dcfc4b2e | 951 | |
71ef24d0 LP |
952 | * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ... |
953 | ||
d28315e4 | 954 | * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab is not |
1b89884b | 955 | |
b44be3ec | 956 | * nspawn: |
e25b5a8d DH |
957 | - emulate /dev/kmsg using CUSE and turn off the syslog syscall |
958 | with seccomp. That should provide us with a useful log buffer that | |
959 | systemd can log to during early boot, and disconnect container logs | |
960 | from the kernel's logs. | |
961 | - as soon as networkd has a bus interface, hook up --network-interface=, | |
962 | --network-bridge= with networkd, to trigger netdev creation should an | |
963 | interface be missing | |
e25b5a8d DH |
964 | - a nice way to boot up without machine id set, so that it is set at boot |
965 | automatically for supporting --ephemeral. Maybe hash the host machine id | |
966 | together with the machine name to generate the machine id for the container | |
967 | - fix logic always print a final newline on output. | |
968 | https://github.com/systemd/systemd/pull/272#issuecomment-113153176 | |
969 | - should optionally support receiving WATCHDOG=1 messages from its payload | |
970 | PID 1... | |
e25b5a8d DH |
971 | - optionally automatically add FORWARD rules to iptables whenever nspawn is |
972 | running, remove them when shut down. | |
e25b5a8d | 973 | |
347a17d5 ZJS |
974 | * dissect |
975 | - refuse mounting over a mount point | |
976 | - automatically discover .roothash files in dissect, similarly to nspawn | |
977 | ||
e25b5a8d | 978 | * machined: |
e25b5a8d DH |
979 | - add an API so that libvirt-lxc can inform us about network interfaces being |
980 | removed or added to an existing machine | |
981 | - "machinectl migrate" or similar to copy a container from or to a | |
982 | difference host, via ssh | |
e25b5a8d DH |
983 | - introduce systemd-nspawn-ephemeral@.service, and hook it into |
984 | "machinectl start" with a new --ephemeral switch | |
985 | - "machinectl status" should also show internal logs of the container in | |
986 | question | |
e25b5a8d DH |
987 | - "machinectl history" |
988 | - "machinectl diff" | |
989 | - "machinectl commit" that takes a writable snapshot of a tree, invokes a | |
990 | shell in it, and marks it read-only after use | |
991 | ||
b44be3ec | 992 | * cryptsetup: |
b44be3ec | 993 | - cryptsetup-generator: allow specification of passwords in crypttab itself |
5f1dac6b | 994 | - support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator |
f8019684 | 995 | |
424a19f8 | 996 | * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it |
41f9172f | 997 | https://bugs.freedesktop.org/show_bug.cgi?id=54712 |
424a19f8 | 998 | |
83fefc88 ZJS |
999 | * add a dependency on standard-conf.xml and other included files to man pages |
1000 | ||
68f16003 | 1001 | * MountFlags=shared acts as MountFlags=slave right now. |
0790b9fe | 1002 | |
fd4d89b2 LP |
1003 | * properly handle loop back mounts via fstab, especially regards to fsck/passno |
1004 | ||
47ae7201 LP |
1005 | * initialize the hostname from the fs label of /, if /etc/hostname does not exist? |
1006 | ||
abd55b16 | 1007 | * udev: |
abd55b16 | 1008 | - move to LGPL |
abd55b16 KS |
1009 | - kill scsi_id |
1010 | - add trigger --subsystem-match=usb/usb_device device | |
e8d842a0 | 1011 | - reimport udev db after MOVE events for devices without dev_t |
b8217b7b | 1012 | |
7f110ff9 LP |
1013 | * There's currently no way to cancel fsck (used to be possible via C-c or c on the console) |
1014 | ||
d0e5a333 LP |
1015 | * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/ |
1016 | ||
e25b5a8d DH |
1017 | * coredump: |
1018 | - save coredump in Windows/Mozilla minidump format | |
73a99163 | 1019 | - when truncating coredumps, also log the full size that the process had, and make a metadata field so we can report truncated coredumps |
87a8baa3 LP |
1020 | |
1021 | * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting) | |
1022 | ||
f957632b KS |
1023 | * be able to specify a forced restart of service A where service B depends on, in case B |
1024 | needs to be auto-respawned? | |
1025 | ||
f38afcd0 | 1026 | * tmpfiles: |
f38afcd0 | 1027 | - apply "x" on "D" too (see patch from William Douglas) |
614cc34f | 1028 | - instead of ignoring unknown fields, reject them. |
e25b5a8d DH |
1029 | - creating new directories/subvolumes/fifos/device nodes |
1030 | should not follow symlinks. None of the other adjustment or creation | |
1031 | calls follow symlinks. | |
1258097c | 1032 | |
a4c279f8 LP |
1033 | * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early |
1034 | ||
f28f1daf LP |
1035 | * verify that the AF_UNIX sockets of a service in the fs still exist |
1036 | when we start a service in order to avoid confusion when a user | |
1037 | assumes starting a service is enough to make it accessible | |
1038 | ||
88a07670 LP |
1039 | * Make it possible to set the keymap independently from the font on |
1040 | the kernel cmdline. Right now setting one resets also the other. | |
1041 | ||
71092d70 | 1042 | * and a dbus call to generate target from current state |
916abb21 | 1043 | |
68c7d001 | 1044 | * write blog stories about: |
796b06c2 | 1045 | - hwdb: what belongs into it, lsusb |
68c7d001 | 1046 | - enabling dbus services |
7a2a0b90 | 1047 | - how to make changes to sysctl and sysfs attributes |
253ee27a | 1048 | - remote access |
5d0fcd7c | 1049 | - how to pass throw-away units to systemd, or dynamically change properties of existing units |
8bbabc44 | 1050 | - testing with Harald's awesome test kit |
f530371f | 1051 | - auto-restart |
e707c494 | 1052 | - how to develop against journal browsing APIs |
a19554ed | 1053 | - the journal HTTP iface |
a73d88fa | 1054 | - non-cgroup resource management |
a74a8793 | 1055 | - dynamic resource management with cgroups |
3040728b | 1056 | - refreshed, longer missions statement |
3679d112 LP |
1057 | - calendar time events |
1058 | - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell | |
1059 | - how to create your own target | |
1060 | - instantiated apache, dovecot and so on | |
1061 | - hooking a script into various stages of shutdown/rearly booot | |
73090dc8 | 1062 | |
21bdae12 LP |
1063 | * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support. |
1064 | ||
59fee421 LP |
1065 | * dot output for --test showing the 'initial transaction' |
1066 | ||
e25b5a8d | 1067 | * pid1: |
e25b5a8d DH |
1068 | - When logging about multiple units (stopping BoundTo units, conflicts, etc.), |
1069 | log both units as UNIT=, so that journalctl -u triggers on both. | |
1070 | - generate better errors when people try to set transient properties | |
1071 | that are not supported... | |
1072 | http://lists.freedesktop.org/archives/systemd-devel/2015-February/028076.html | |
1073 | - maybe introduce WantsMountsFor=? Usecase: | |
1074 | http://lists.freedesktop.org/archives/systemd-devel/2015-January/027729.html | |
1075 | - recreate systemd's D-Bus private socket file on SIGUSR2 | |
e25b5a8d DH |
1076 | - move PAM code into its own binary |
1077 | - when we automatically restart a service, ensure we restart its rdeps, too. | |
e25b5a8d DH |
1078 | - hide PAM options in fragment parser when compile time disabled |
1079 | - Support --test based on current system state | |
1080 | - 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(). | |
1081 | - after deserializing sockets in socket.c we should reapply sockopts and things | |
e25b5a8d DH |
1082 | - drop PID 1 reloading, only do reexecing (difficult: Reload() |
1083 | currently is properly synchronous, Reexec() is weird, because we | |
1084 | cannot delay the response properly until we are back, so instead of | |
1085 | being properly synchronous we just keep open the fd and close it | |
1086 | when done. That means clients do not get a successful method reply, | |
1087 | but much rather a disconnect on success. | |
1088 | - when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr | |
1089 | - when a bus name of a service disappears from the bus make sure to queue further activation requests | |
1090 | ||
1091 | * unit files: | |
1092 | - allow port=0 in .socket units | |
1093 | - maybe introduce ExecRestartPre= | |
1094 | - add ReloadSignal= for configuring a reload signal to use | |
1095 | - implement Register= switch in .socket units to enable registration | |
1096 | in Avahi, RPC and other socket registration services. | |
1097 | - allow Type=simple with PIDFile= | |
1098 | https://bugzilla.redhat.com/show_bug.cgi?id=723942 | |
1099 | - allow writing multiple conditions in unit files on one line | |
e25b5a8d | 1100 | - introduce Type=pid-file |
e25b5a8d DH |
1101 | - introduce mix of BindTo and Requisite |
1102 | - add a concept of RemainAfterExit= to scope units | |
e25b5a8d | 1103 | - Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely |
e5171296 | 1104 | - add verification of [Install] section to systemd-analyze verify |
e25b5a8d | 1105 | |
af6f0d42 TG |
1106 | * udev-link-config: |
1107 | - Make sure ID_PATH is always exported and complete for | |
1108 | network devices where possible, so we can safely rely | |
1109 | on Path= matching | |
af6f0d42 | 1110 | |
88e4d1d7 | 1111 | * sd-rtnl: |
88e4d1d7 | 1112 | - add support for more attribute types |
c589a0e6 | 1113 | - inbuilt piping support (essentially degenerate async)? see loopback-setup.c and other places |
88e4d1d7 | 1114 | |
0a4b9a07 | 1115 | * networkd: |
c74ecd71 TG |
1116 | - add more keys to [Route] and [Address] sections |
1117 | - add support for more DHCPv4 options (and, longer term, other kinds of dynamic config) | |
0a4b9a07 | 1118 | - add proper initrd support (in particular generate .network/.link files based on /proc/cmdline) |
e8d842a0 | 1119 | - add reduced [Link] support to .network files |
5c1d3fc9 | 1120 | - add Scope= parsing option for [Network] |
798e174a | 1121 | - properly handle routerless dhcp leases |
a8eaaee7 | 1122 | - work with non-Ethernet devices |
c9fa3670 | 1123 | - add support for more bond options |
e25b5a8d | 1124 | - dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from? |
e25b5a8d DH |
1125 | - the DHCP lease data (such as NTP/DNS) is still made available when |
1126 | a carrier is lost on a link. It should be removed instantly. | |
1127 | - expose in the API the following bits: | |
1128 | - option 15, domain name and/or option 119, search list | |
1129 | - option 12, host name and/or option 81, fqdn | |
1130 | - option 123, 144, geolocation | |
1131 | - option 252, configure http proxy (PAC/wpad) | |
1132 | - provide a way to define a per-network interface default metric value | |
1133 | for all routes to it. possibly a second default for DHCP routes. | |
1134 | - allow Name= to be specified repeatedly in the [Match] section. Maybe also | |
1135 | support Name=foo*|bar*|baz ? | |
1136 | - duplicate address check for static IPs (like ARPCHECK in network-scripts) | |
1137 | - allow DUID/IAID to be customized, see issue #394. | |
e25b5a8d | 1138 | - whenever uplink info changes, make DHCP server send out FORCERENEW |
155e8b9a | 1139 | |
473dfd7b TG |
1140 | * networkd-wait-online: |
1141 | - make operstates to wait for configurable? | |
8f5675cc | 1142 | |
ac976532 | 1143 | * dhcp: |
424a8732 | 1144 | - figure out how much we can increase Maximum Message Size |
ac976532 | 1145 | |
37d8b536 PF |
1146 | * dhcp6: |
1147 | - add functions to set previously stored IPv6 addresses on startup and get | |
1148 | them at shutdown; store them in client->ia_na | |
1149 | - write more test cases | |
37d8b536 | 1150 | - implement reconfigure support, see 5.3., 15.11. and 22.20. |
37d8b536 | 1151 | - implement support for temporary adressess (IA_TA) |
37d8b536 PF |
1152 | - implement dhcpv6 authentication |
1153 | - investigate the usefulness of Confirm messages; i.e. are there any | |
1154 | situations where the link changes without any loss in carrier detection | |
1155 | or interface down | |
1156 | - some servers don't do rapid commit without a filled in IA_NA, verify | |
1157 | this behavior | |
4a77c53d | 1158 | - RouteTable= ? |
37d8b536 | 1159 | |
2f8cd170 | 1160 | External: |
74fe1fe3 | 1161 | |
cc9784c6 | 1162 | * dbus: |
3377af3e | 1163 | - natively watch for dbus-*.service symlinks (PENDING) |
b44be3ec | 1164 | - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service |
cc9784c6 | 1165 | |
71092d70 LP |
1166 | * fix alsa mixer restore to not print error when no config is stored |
1167 | ||
74fe1fe3 LP |
1168 | * make cryptsetup lower --iter-time |
1169 | ||
9096ad36 | 1170 | * patch kernel for xattr support in /dev, /proc/, /sys? |
06ae4bfe | 1171 | |
71092d70 | 1172 | * kernel: add device_type = "fb", "fbcon" to class "graphics" |
14bd37fe | 1173 | |
8973790e LP |
1174 | * /usr/bin/service should actually show the new command line |
1175 | ||
d2bbca68 | 1176 | * 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= |
8973790e LP |
1177 | |
1178 | * fedora: F20: go timer units all the way, leave cron.daily for cron | |
1179 | ||
1180 | * neither pkexec nor sudo initialize environ[] from the PAM environment? | |
b1ce67bf | 1181 | |
4641a16b | 1182 | * fedora: update policy to declare access mode and ownership of unit files to root:root 0644, and add an rpmlint check for it |
19aa7c4f | 1183 | |
baf167ee ZJS |
1184 | * register catalog database signature as file magic |
1185 | ||
1cf3c30c ZJS |
1186 | * zsh shell completion: |
1187 | - <command> <verb> -<TAB> should complete options, but currently does not | |
409886c4 | 1188 | - systemctl add-wants,add-requires |
1cf3c30c | 1189 | |
90137b62 | 1190 | * systemctl status should know about 'systemd-analyze calendar ... --iterations=' |
90137b62 ZJS |
1191 | * If timer has just OnInactiveSec=..., it should fire after a specified time |
1192 | after being started. | |
2c12a402 | 1193 | |
129126f3 | 1194 | Regularly: |
874aa2cc | 1195 | |
129126f3 | 1196 | * look for close() vs. close_nointr() vs. close_nointr_nofail() |
5021be21 | 1197 | |
129126f3 | 1198 | * check for strerror(r) instead of strerror(-r) |
5b6319dc | 1199 | |
444a79d3 LP |
1200 | * pahole |
1201 | ||
d28315e4 | 1202 | * set_put(), hashmap_put() return values check. i.e. == 0 does not free()! |
a3a3e5b6 | 1203 | |
4db17f29 | 1204 | * use secure_getenv() instead of getenv() where appropriate |
88fae6e0 | 1205 | |
d05c556b | 1206 | * link up selected blog stories from man pages and unit files Documentation= fields |
62353f69 | 1207 | String is not UTF-8 clean, ignoring assignment |