]>
Commit | Line | Data |
---|---|---|
1 | Bugfixes: | |
2 | * check systemd-tmpfiles for selinux context hookup for mknod(), symlink() and similar | |
3 | ||
4 | * swap units that are activated by one name but shown in the kernel under another are semi-broken | |
5 | ||
6 | * Dangling symlinks of .automount unit files in .wants/ directories, set up | |
7 | automount points even when the original .automount file did not exist | |
8 | anymore. Only the .mount unit was still around. | |
9 | ||
10 | * make polkit checks async | |
11 | ||
12 | * properly handle .mount unit state tracking when two mount points are stacked one on top of another on the exact same mount point. | |
13 | ||
14 | Fedora 19: | |
15 | ||
16 | * fix match logic to add another level of disjunction/conjunction | |
17 | ||
18 | * external: maybe it is time to patch procps so that "ps" links to | |
19 | libsystemd-logind to print a pretty service name, seat name, session | |
20 | name in its output. Currently it only shows cgroup membership, but | |
21 | that's sometimes kinda hard to parse for a human. | |
22 | ||
23 | * cgroup attrs: | |
24 | - update dbus interface docs in wiki | |
25 | ||
26 | * localed: | |
27 | - localectl: support new converted x11→console keymaps | |
28 | ||
29 | * timer logic is confused by units which are skipped due to failing condition | |
30 | http://lists.freedesktop.org/archives/systemd-devel/2013-February/008816.html | |
31 | ||
32 | * timer logic is also confused if a service it triggers hasn't finished when the next timer elapses: | |
33 | http://lists.freedesktop.org/archives/systemd-devel/2013-February/009021.html | |
34 | ||
35 | Features: | |
36 | ||
37 | * add s.th. like "systemctl set-log-level debug" | |
38 | ||
39 | * move session owner code from sd-login.c to cgroup-util.c | |
40 | ||
41 | * decimal_str_max() should become a type-safe macro | |
42 | ||
43 | * show-cgroup.c uses its own ulog10 implementation | |
44 | ||
45 | * cgroup-util: verify syntax of cgroup controllers | |
46 | ||
47 | * cgtop: make cgtop useful in a container | |
48 | ||
49 | * make sure cg_pid_get_path() works properly for co-mounted controllers | |
50 | ||
51 | * nspawn: ensure syntax of --uuid= argument is correct | |
52 | ||
53 | * nspawn: detect whether something is an OS by checking for /etc/os-release | |
54 | ||
55 | * explicitly disallow changing the cgroup path of units in the | |
56 | name=systemd hierarchy, unless it is outside of /system | |
57 | ||
58 | * add 'set -e' to scripts in test/ | |
59 | * make test in test/ work with separate output dir | |
60 | ||
61 | * suppress log output on shutdown when "quiet" is used | |
62 | ||
63 | * systemctl delete x.snapshot leaves no trace in logs (at least at default level). | |
64 | ||
65 | * make the coredump collector tool move itself into the user's cgroup | |
66 | so that the coredump is properly written to the user's own journal | |
67 | file. | |
68 | ||
69 | * move /usr/lib/modules/$(uname -r)/modules.devname parsing from udevd to | |
70 | kmod static-nodes | |
71 | call kmod as an early service, and drop CAP_MKNOD from udevd.service | |
72 | ||
73 | * systemd-delta needs to be made aware of *.d/*.conf drop-in files for | |
74 | units. | |
75 | ||
76 | * seems that when we follow symlinks to units we prefer the symlink | |
77 | destination path over /etc and /usr. We shouldn't do that. Instead | |
78 | /etc should always override /run+/usr and also any symlink | |
79 | destination. | |
80 | ||
81 | * remove duplicate default deps logic from fstab-generator vs. mount.c | |
82 | ||
83 | * when isolating, try to figure out a way how we implicitly can order | |
84 | all units we stop before the isolating unit... | |
85 | ||
86 | * teach udev + logind's uaccess to somehow handle the "dead" device nodes from: | |
87 | /lib/modules/$(uname -r)/modules.devname | |
88 | and apply ACLs to them if they have TAG=="uaccess" in udev rules. | |
89 | ||
90 | * matching against units is currently broken in journalctl. We really | |
91 | need another AND level in the expressions, | |
92 | i.e. sd_journal_add_conjunction(). | |
93 | ||
94 | * add ConditionArchitecture= or so | |
95 | ||
96 | * teach ConditionKernelCommandLine= globs or regexes (in order to match foobar={no,0,off}) | |
97 | ||
98 | * we should log capabilities too | |
99 | ||
100 | * Support SO_REUSEPORT with socket activation: | |
101 | - Let systemd maintain a pool of servers. | |
102 | - Use for seamless upgrades, by running the new server before stopping the | |
103 | old. | |
104 | ||
105 | * after all byte-wise realloc() is slow, even on glibc, so i guess we | |
106 | need manual exponential loops after all | |
107 | ||
108 | * BootLoaderSpec: drop allowing ext234 for $BOOT. Clarify that the | |
109 | kernel has to be in $BOOT. Clarify that the boot loader should be | |
110 | installed to the ESP. Define a way how an installer can figure out | |
111 | whether a BLS compliant boot loader is installed. | |
112 | ||
113 | * man: remove .include documentation, and instead push people to use .d/*.conf | |
114 | ||
115 | * disallow .include from included files | |
116 | ||
117 | * safe_atod() is too naive, as it is vulnerable to locale parameters, should be locale independent. | |
118 | ||
119 | * think about requeuing jobs when daemon-reload is issued? usecase: | |
120 | the initrd issues a reload after fstab from the host is accessible | |
121 | and we might want to requeue the mounts local-fs acquired through | |
122 | that automatically. | |
123 | ||
124 | * neither pkexec nor sudo initialize environ[] from the PAM environment? | |
125 | ||
126 | * rework specifier logic so that we can distuingish OOM errors from other errors | |
127 | ||
128 | * systemd-inhibit: make taking delay locks useful: support sending SIGINT or SIGTERM on PrepareForSleep() | |
129 | ||
130 | * journal-or-kmsg is currently broken? See reverted commit 4a01181e460686d8b4a543b1dfa7f77c9e3c5ab8. | |
131 | ||
132 | * remove any syslog support from log.c -- we probably can't do this before split-off udev is gone for good | |
133 | ||
134 | * fedora: connect the timer units of a service to the service via Also= in [Install] | |
135 | ||
136 | * fedora: F20: go timer units all the way, leave cron.daily for cron | |
137 | ||
138 | * add a tool that lists active timer units plus their next elapstion and the time the units ran last | |
139 | ||
140 | * man: document the very specific env the shutdown drop-in tools live in | |
141 | ||
142 | * shutdown logging: store to EFI var, and store to USB stick? | |
143 | ||
144 | * man: extend runlevel(8) to mention that runlevels suck, and are dead. Maybe add runlevel(7) with a note about that too | |
145 | ||
146 | * systemctl: maybe add "systemctl add-wants" or so... | |
147 | ||
148 | * man: add a link to socket activation blog from systemd.socket(5) | |
149 | ||
150 | * man: add more examples to man pages | |
151 | ||
152 | * man: maybe sort directives in man pages, and take sections from --help and apply them to man too | |
153 | ||
154 | * add "# export SYSTEMD_PAGER=" to bash login | |
155 | ||
156 | * /usr/bin/service should actually show the new command line | |
157 | ||
158 | * fedora: suggest auto-restart on failure, but not on sucess and not on coredump. also, ask people to think about changing the start limit logic. Also point people to RestartPreventExitStatus=, SuccessExitStatus= | |
159 | ||
160 | * write UI tool that pops up emergency messages from the journal as notification | |
161 | ||
162 | * think about window-manager-run-as-user-service problem: exit 0 → activate shutdown.target; exit != 0 → restart service | |
163 | ||
164 | * msgcatalog determine default language of entries from the catalog | |
165 | file name. i.e. foobar.de_DE.catalog would set the default entry | |
166 | name for the entries to de_DE if they aren't explicitly suffixed | |
167 | individually. | |
168 | ||
169 | * use "log level" rather than "log priority" everywhere | |
170 | ||
171 | * ensure sd_journal_seek_monotonic actually works properly. | |
172 | ||
173 | * timedate: have global on/off switches for auto-time (NTP), and auto-timezone that connman can subscribe to. | |
174 | ||
175 | * Honour "-" prefix for InaccessibleDirectories= and ReadOnlyDirectories= to | |
176 | suppress errors of the specified path doesn't exist | |
177 | ||
178 | * dev-setup.c: when running in a container, create a tiny stub udev | |
179 | database with the systemd tag set for all network interfaces found, | |
180 | so that libudev reports them as present, and systemd's .device units | |
181 | will be activated. | |
182 | ||
183 | * merge unit_kill_common() and unit_kill_context() | |
184 | ||
185 | * introduce ExecCondition= in services | |
186 | ||
187 | * EFI: | |
188 | - fsck hookup for the ESP mount is missing | |
189 | - write man page for efi boot generator | |
190 | - honor language efi variables for default language selection (if there are any?) | |
191 | - honor timezone efi variables for default timezone selection (if there are any?) | |
192 | - introduce bootctl (backed by systemd-bootd) to control temporary and persistent default boot goal plus efi variables | |
193 | ||
194 | * maybe do not install getty@tty1.service symlink in /etc but in /usr? | |
195 | ||
196 | * re-enable "make check" for gtk-doc (broken for unknown reason) | |
197 | ||
198 | * fstab: add new mount option x-systemd-after=/foobar/waldo to allow manual dependencies to other mount points | |
199 | https://bugzilla.redhat.com/show_bug.cgi?id=812826 | |
200 | ||
201 | * print a nicer explanation if people use variable/specifier expansion in ExecStart= for the first word | |
202 | ||
203 | * mount: turn dependency information from /proc/self/mountinfo into dependency information between systemd units. | |
204 | ||
205 | * logind: | |
206 | - logind: optionally, ignore idle-hint logic for autosuspend, block suspend as long as a session is around | |
207 | - When we update the kernel all kind of hibernation should be prohibited until shutdown/reboot | |
208 | - logind: wakelock/opportunistic suspend support | |
209 | - Add pretty name for seats in logind | |
210 | - logind: allow showing logout dialog from system? | |
211 | - logind: spawn user@..service on login | |
212 | - logind: non-local X11 server handling | |
213 | - logind: add equivalent to sd_pid_get_owner_uid() to the D-Bus API | |
214 | - pam: when leaving a session explicitly exclude the ReleaseSession() caller process from the killing spree | |
215 | - logind: GetSessionByPID() should accept 0 as PID value | |
216 | - we should probably handle SIGTERM/SIGINT to not leave dot files around, just in case | |
217 | ||
218 | * exec: when deinitializating a tty device fix the perms and group, too, not only when initializing. Set access mode/gid to 0620/tty. | |
219 | ||
220 | * DeviceAllow/DeviceDeny: disallow everything by default, but whitelist /dev/zero, /dev/null and friends | |
221 | ||
222 | * service: watchdog logic: for testing purposes allow ping, but do not require pong | |
223 | ||
224 | * journal: | |
225 | - do not use magic msync() in src/journal/journal-file.c, just call fsync() | |
226 | <alxchk> poettering: looks like msync just calls vfs_fsync | |
227 | http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/mm/msync.c#n18 | |
228 | - import and delete pstore filesystem content at startup | |
229 | - journald: also get thread ID from client, plus thread name | |
230 | - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups | |
231 | - add API to close/reopen/get fd for journal client fd in libsystemd-journal. | |
232 | - fallback to /dev/log based logging in libsystemd-journal, if we can't log natively? | |
233 | - declare the local journal protocol stable in the wiki interface chart | |
234 | - journal: reuse XZ context | |
235 | - sd-journal: speed up sd_journal_get_data() with transparent hash table in bg | |
236 | - journald: when dropping msgs due to ratelimit make sure to write | |
237 | "dropped %u messages" not only when we are about to print the next | |
238 | message that works, but alraedy after a short tiemout | |
239 | - check if we can make journalctl by default use --follow mode inside of less if called without args? | |
240 | - maybe add API to send pairs of iovecs via sd_journal_send | |
241 | - journal: when writing journal auto-rotate if time jumps backwards | |
242 | - gatewayd: should run under its own UID | |
243 | - journal: add a setgid "systemd-journal" utility to invoke from libsystemd-journal, which passes fds via STDOUT and does PK access | |
244 | - journactl: support negative filtering, i.e. FOOBAR!="waldo", | |
245 | and !FOOBAR for events without FOOBAR. | |
246 | - journal: when rotating, copy over old acls/access mode | |
247 | - journal: send out marker messages every now and then, and immediately sync with fdatasync() afterwards, in order to have hourly guaranteed syncs. | |
248 | - 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. | |
249 | - journal: find a way to allow dropping history early, based on priority, other rules | |
250 | - journal: When used on NFS, check payload hashes | |
251 | - journal: When used on NFS make sure wake up sd_journal_wait() every 2s, to handle missing inotify | |
252 | - Introduce journalctl -b <nr> to show journal messages of a previous boot | |
253 | - journald: check whether it is OK if the client can still modify delivered journal entries | |
254 | - journal live copy, based on libneon (client) and libmicrohttpd (server) | |
255 | - journald: add kernel cmdline option to disable ratelimiting for debug purposes | |
256 | - refuse taking lower-case variable names in sd_journal_send() and friends. | |
257 | - journald: we currently rotate only after MaxUse+MaxFilesize has been reached. | |
258 | - journal: deal nicely with byte-by-byte copied files, especially regards header | |
259 | - journalctl: show multiline log messages sanely, expand tabs, and show all valid utf8 messages | |
260 | - journal: store euid in journal if it differs from uid | |
261 | - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit | |
262 | - journalctl: make journalctl smarter, and actually check groups that have access to /var/log/journal before printing message about recomming group membership for journal access | |
263 | ||
264 | * document: | |
265 | - document unit_name_mangle() | |
266 | - document that deps in [Unit] sections ignore Alias= fileds in | |
267 | [Install] units of other units, unless those units are disabled | |
268 | - man: clarify that time-sync.target is not only sysv compat but also useful otherwise. Same for similar targets | |
269 | - Document word splitting syntax for ExecStart= and friends | |
270 | - document that units from /etc override those from /usr and /run | |
271 | - document the exit codes when services fail before they are exec()ed | |
272 | - document that service reload may be implemented as service reexec | |
273 | - document in wiki how to map ical recurrence events to systemd timer unit calendar specifications | |
274 | ||
275 | * systemctl: | |
276 | - systemctl list-jobs - show dependencies | |
277 | - add systemctl switch to dump transaction without executing it | |
278 | - Add a verbose mode to "systemctl start" and friends that explains what is being done or not done | |
279 | - "systemctl disable" on a static unit prints no message and does | |
280 | nothing. "systemctl enable" does nothing, and gives a bad message | |
281 | about it. Should fix both to print nice actionable messages. | |
282 | - print nice message from systemctl --failed if there are no entries shown, and hook that into ExecStartPre of rescue.service/emergency.service | |
283 | - add new command to systemctl: "systemctl system-reexec" which reexecs as many daemons as virtually possible | |
284 | - systemctl enable: improve the success messages (i.e. more human readable, less shell-like) | |
285 | - systemctl enable: fail if target to alias into doesn't exist? maybe show how many units are enabled afterwards? | |
286 | - systemctl: "Journal has been rotated since unit was started." message is misleading | |
287 | - support "systemctl stop foobar@.service" to stop all units matching a certain template | |
288 | - Something is wrong with symlink handling of "autovt@.service" in "systemctl list-unit-files" | |
289 | - add configuration/switches to use | |
290 | freeze (http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git) and | |
291 | standby (https://bugs.freedesktop.org/show_bug.cgi?id=57793) as suspend modes | |
292 | ||
293 | * introduce ntp.service (or suchlike) as symlink that is used to arbitrate between various NTP implementations | |
294 | ||
295 | * deal with sendmail/postfix exclusivity | |
296 | ||
297 | * timer units: | |
298 | - configurable jitter for timer events | |
299 | - timer events with system resume | |
300 | - timer units should get the ability to trigger when: | |
301 | o CLOCK_REALTIME makes jumps (TFD_TIMER_CANCEL_ON_SET) | |
302 | o DST changes | |
303 | - Support 2012-02~4 as syntax for specifying the fourth to last day of the month. | |
304 | ||
305 | * update the kernel's TZ (sys_tz) when DST changes | |
306 | ||
307 | * sync down the system time to the RTC when: | |
308 | - CLOCK_REALTIME makes jumps (the user explicitely requested a time set) | |
309 | - DST/timezone changes && ntp is active && RTC-in-localtime (never do it without ntp) | |
310 | This takes care of syncing ntpdate updates to the RTC, and DST updates for localtime | |
311 | mode, it will never touch the RTC if the no reliable time source is active or the | |
312 | user did not request anything like it. | |
313 | ||
314 | * hwdb: | |
315 | - implement conditional properties (dmi matches) | |
316 | - hwdb --filter=ID_DRIVE_* | |
317 | ||
318 | * if booted in "quiet" mode, and an error happens, turn on status | |
319 | output again, so that the emergency mode isn't totally | |
320 | surprising. Also, terminate plymouth. | |
321 | ||
322 | * libunwind support for coredump pattern hook, and includes this in | |
323 | the message for coredumps. After all, libunwind is now capable to | |
324 | unwind coredumps since a few weeks ago. This probably requires that | |
325 | we have nice support for multi-line messages on display in | |
326 | logs-show.c. Alternatively: use libelfutil, which seems to be the | |
327 | better supported alternative. | |
328 | ||
329 | * figure out relation of --all and --full in the various tools | |
330 | ||
331 | * add libsystemd-password or so to query passwords during boot using the password agent logic | |
332 | ||
333 | * 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(). | |
334 | ||
335 | * fedup: add --unit to systemctl switch-root somehow | |
336 | * fedup: don't delete initrd on switch-root | |
337 | * fedup: generator | |
338 | ||
339 | * timedated: | |
340 | - timedated: refuse time changes when NTP is on | |
341 | ||
342 | * clean up date formatting and parsing so that all absolute/relative timestamps we format can also be parsed | |
343 | ||
344 | * introduce generic AUGMENT_PID=, AUGMENT_DEVICE= fields | |
345 | ||
346 | * on shutdown: move utmp, wall, audit logic all into PID 1 (or logind?), get rid of systemd-update-utmp-runlevel | |
347 | ||
348 | * add "provisioning" instructions to setup an empty /etc + /var | |
349 | - used to setup a new container from a shared /usr | |
350 | - superset of tmpfiles model | |
351 | - instructions shipped by packages and stored in /usr/lib/ | |
352 | - compose /etc/passwd and /etc/group, copy files | |
353 | - able to create uid + gid used by packages, for file ownership | |
354 | ||
355 | * make repeated alt-ctrl-del presses printing a dump, or even force a reboot without | |
356 | waiting for the timeout | |
357 | ||
358 | * high level net_prio setting in execution context | |
359 | ||
360 | * hostnamed: before returning information from /etc/machine-info.conf check the modification data and reread. Similar for localed, ... | |
361 | ||
362 | * currently x-systemd.timeout is lost in the initrd, since crypttab is copied into dracut, but fstab isn't | |
363 | ||
364 | * WorkingDirectory: support env var replacements like in ExecStart= so that people can use $HOME | |
365 | ||
366 | * refuse boot if /etc/machine-id is not useful (or set taint?) | |
367 | ||
368 | * nspawn: | |
369 | - nspawn: consider changing users for -u with su, so that NSS resolving works correctly | |
370 | - nspawn: implement personality changes a la linux32(8) | |
371 | - nspawn: --read-only is not applied recursively to submounts | |
372 | - nspawn: make use of device cgroup controller by default | |
373 | - bind mount read-only the cgroup tree higher than nspawn | |
374 | - nspawn: investigate whether we can support the same as LXC's | |
375 | lxc.network.type=phys mode, and pass through entire network | |
376 | interfaces to the container | |
377 | - nspawn: maybe add a way to drop additional caps, in addition to add additional caps | |
378 | - nspawn: maybe explicitly reset loginuid? | |
379 | - nspawn: move all containers to name=systemd:/containers/, and introduce externally visible names for containers | |
380 | - nspawn: make it work for dwalsh and shared /usr containers -- tmpfs mounts as command line parameters, selinux exec context | |
381 | ||
382 | * cryptsetup: | |
383 | - cryptsetup-generator: warn if the password files are world-readable | |
384 | - cryptsetup-generator: allow specification of passwords in crypttab itself | |
385 | - move cryptsetup key caching into kernel keyctl? | |
386 | https://bugs.freedesktop.org/show_bug.cgi?id=54982 | |
387 | - when key file cannot be found, read it from kbd in cryptsetup | |
388 | ||
389 | * move debug shell to tty6 and make sure this doesn't break the gettys on tty6 | |
390 | ||
391 | * hw watchdog: optionally try to use the preset watchdog timeout instead of always overriding it | |
392 | https://bugs.freedesktop.org/show_bug.cgi?id=54712 | |
393 | ||
394 | * after deserializing sockets in socket.c we should reapply sockopts and things | |
395 | ||
396 | * make timer units go away after they elapsed | |
397 | ||
398 | * http://lists.freedesktop.org/archives/systemd-devel/2012-September/006502.html | |
399 | (network and remote-fs on shutdown) | |
400 | ||
401 | * come up with a nice way to write queue/read_ahead_kb for a block device without interfering with readahead | |
402 | ||
403 | * move PID 1 segfaults to /var/lib/systemd/coredump? | |
404 | ||
405 | * create /sbin/init symlinks from the build system | |
406 | ||
407 | * Query Paul Moore about relabelling socket fds while they are open | |
408 | ||
409 | * system.conf should have controls for cgroups | |
410 | ||
411 | * allow writing multiple conditions in unit files on one line | |
412 | ||
413 | * explore multiple service instances per listening socket idea | |
414 | ||
415 | * shutdown: don't read-only mount anything when running in container | |
416 | ||
417 | * MountFlags=shared acts as MountFlags=slave right now. | |
418 | ||
419 | * ReadOnlyDirectories= is not applied recursively to submounts | |
420 | ||
421 | * drop PID 1 reloading, only do reexecing (difficult: Reload() | |
422 | currently is properly synchronous, Reexec() is weird, because we | |
423 | can't delay the response properly until we are back, so instead of | |
424 | being properly synchronous we just keep open the fd and close it | |
425 | when done. That means clients don't get a successful method reply, | |
426 | but much rather a disconnect on success. | |
427 | ||
428 | * remember which condition failed for services, not just the fact that something failed | |
429 | ||
430 | * use opterr = 0 for all getopt tools | |
431 | ||
432 | * properly handle loop back mounts via fstab, especially regards to fsck/passno | |
433 | ||
434 | * allow services with no ExecStart= but with an ExecStop= | |
435 | ||
436 | * dracut-shutdown needs to be ordered before unmounting /boot | |
437 | ||
438 | * initialize the hostname from the fs label of /, if /etc/hostname does not exist? | |
439 | ||
440 | * rename "userspace" to "core-os" | |
441 | ||
442 | * syscall filter: | |
443 | - syscall filter: add knowledge about compat syscalls | |
444 | - syscall filter: don't enforce no new privs? | |
445 | - syscall filter: option to return EPERM rather than SIGSYS? | |
446 | - syscall filter: port to libseccomp | |
447 | - system-wide seccomp filter | |
448 | ||
449 | * .device aliases need to be implemented with the "following" logic, probably. | |
450 | ||
451 | * load-fragment: when loading a unit file via a chain of symlinks | |
452 | verify that it isn't masked via any of the names traversed. | |
453 | ||
454 | * introduce Type=pid-file | |
455 | ||
456 | * maybe allow services with ExecStop= set, but no ExecStart=? | |
457 | ||
458 | * change Requires=basic.target to RequisiteOverride=basic.target | |
459 | ||
460 | * support rd.luks.allow-discards= kernel cmdline params in cryptsetup generator | |
461 | ||
462 | * drop accountsservice's StandardOutput=syslog and Type=dbus fields | |
463 | ||
464 | * when breaking cycles drop sysv services first, then services from /run, then from /etc, then from /usr | |
465 | ||
466 | * move passno parsing to fstab generator | |
467 | ||
468 | * automount: implement expire: | |
469 | - set superblock timeout AUTOFS_DEV_IOCTL_TIMEOUT_CMD | |
470 | - periodically run AUTOFS_DEV_IOCTL_EXPIRE_CMD | |
471 | - every timeout/4 (original autofs logic) | |
472 | - blocking, needs a thread | |
473 | - run until -EAGAIN | |
474 | - receive expire packet on pipe if kernel tells the timeout is over | |
475 | - call umount | |
476 | - answer expire packet on pipe with AUTOFS_DEV_IOCTL_{READY,FAIL}_CMD | |
477 | - AUTOFS_DEV_IOCTL_EXPIRE_CMD returns | |
478 | ||
479 | * services which create their own subcgroups break cgroup-empty notification (needs to be fixed in the kernel) | |
480 | ||
481 | * ExecOnFailure=/usr/bin/foo | |
482 | ||
483 | * ConditionSecurity= should learn about IMA and SMACK | |
484 | ||
485 | * udev: | |
486 | - remove src/udev/udev-builtin-firmware.c (CONFIG_FW_LOADER_USER_HELPER=n) | |
487 | - move to LGPL | |
488 | - unify utf8 validator code with shared/ | |
489 | - kill scsi_id | |
490 | - add trigger --subsystem-match=usb/usb_device device | |
491 | ||
492 | * cleanup syslog 'priority' vs. 'level' wording | |
493 | ||
494 | * dbus upstream still refers to dbus.target and shouldn't | |
495 | ||
496 | * when a service has the same env var set twice we actually store it twice and return that in systemctl show -p... We should only show the last setting | |
497 | ||
498 | * support container_ttys= | |
499 | ||
500 | * introduce mix of BindTo and Requisite | |
501 | ||
502 | * add DeleteSocketsOnStop=yes|no option to socket units | |
503 | ||
504 | * There's currently no way to cancel fsck (used to be possible via C-c or c on the console) | |
505 | ||
506 | * add option to sockets to avoid activation. Instead just drop packets/connections, see http://cyberelk.net/tim/2012/02/15/portreserve-systemd-solution/ | |
507 | ||
508 | * default unix qlen is too small (10). bump sysctl? add sockopt? | |
509 | ||
510 | * figure out whether we should leave dbus around during shutdown | |
511 | ||
512 | * dbus: in fedora, make /var/lib/dbus/machine-id a symlink to /etc/machine-id | |
513 | ||
514 | * dbus: move dbus to early boot | |
515 | ||
516 | * save coredump in Windows/Mozilla minidump format | |
517 | ||
518 | * support crash reporting operation modes (https://live.gnome.org/GnomeOS/Design/Whiteboards/ProblemReporting) | |
519 | ||
520 | * clean up session cgroups that remain after logout (think sshd), but eventually run empty | |
521 | ||
522 | * when an instanced service exits, remove its parent cgroup too if possible. | |
523 | ||
524 | * default to actual 32bit PIDs, via /proc/sys/kernel/pid_max | |
525 | ||
526 | * be able to specify a forced restart of service A where service B depends on, in case B | |
527 | needs to be auto-respawned? | |
528 | ||
529 | * when a bus name of a service disappears from the bus make sure to queue further activation requests | |
530 | ||
531 | * tmpfiles: apply "x" on "D" too (see patch from William Douglas) | |
532 | ||
533 | * for services: don't set $HOME in services unless requested | |
534 | ||
535 | * hide PAM/TCPWrap options in fragment parser when compile time disabled | |
536 | ||
537 | * when we automatically restart a service, ensure we restart its rdeps, too. | |
538 | ||
539 | * allow Type=simple with PIDFile= | |
540 | https://bugzilla.redhat.com/show_bug.cgi?id=723942 | |
541 | ||
542 | * move PAM code into its own binary | |
543 | ||
544 | * implement Register= switch in .socket units to enable registration | |
545 | in Avahi, RPC and other socket registration services. | |
546 | ||
547 | * make sure systemd-ask-password-wall does not shutdown systemd-ask-password-console too early | |
548 | ||
549 | * support sd_notify() style notification when reload begins (RELOADING=1), reload is finished (READY=1), and add ReloadSignal= then to use in combination | |
550 | ||
551 | * support sd_notify() style notification when shutting down, to make auto-exit bus services work (STOPPING=1) | |
552 | ||
553 | * verify that the AF_UNIX sockets of a service in the fs still exist | |
554 | when we start a service in order to avoid confusion when a user | |
555 | assumes starting a service is enough to make it accessible | |
556 | ||
557 | * Make it possible to set the keymap independently from the font on | |
558 | the kernel cmdline. Right now setting one resets also the other. | |
559 | ||
560 | * and a dbus call to generate target from current state | |
561 | ||
562 | * readahead: | |
563 | - drop /.readahead on bigger upgrades with yum | |
564 | - move readahead files into /var (look for them with .path units?) | |
565 | - readahead: use BTRFS_IOC_DEFRAG_RANGE instead of BTRFS_IOC_DEFRAG ioctl, with START_IO | |
566 | - readahead: when bumping /sys readahead variable save mtime and compare later to detect changes | |
567 | - readahead: make use of EXT4_IOC_MOVE_EXT, as used by http://e4rat.sourceforge.net/ | |
568 | ||
569 | * add support for /bin/mount -s | |
570 | ||
571 | * GC unreferenced jobs (such as .device jobs) | |
572 | ||
573 | * write blog stories about: | |
574 | - hwdb: what belongs into it, lsusb | |
575 | - enabling dbus services | |
576 | - status update | |
577 | - how to make changes to sysctl and sysfs attributes | |
578 | - remote access | |
579 | - how to pass throw-away units to systemd, or dynamically change properties of existing units | |
580 | - how to integrate cgconfig and suchlike with systemd | |
581 | - testing with Harald's awesome test kit | |
582 | - auto-restart | |
583 | - how to develop against journal browsing APIs | |
584 | - the journal HTTP iface | |
585 | - non-cgroup resource management | |
586 | - dynamic resource management with cgroups | |
587 | - refreshed, longer missions statement | |
588 | - calendar time events | |
589 | - init=/bin/sh vs. "emergency" mode, vs. "rescue" mode, vs. "multi-user" mode, vs. "graphical" mode, and the debug shell | |
590 | - how to create your own target | |
591 | - instantiated apache, dovecot and so on | |
592 | - hooking a script into various stages of shutdown/rearly booot | |
593 | ||
594 | * allow port=0 in .socket units | |
595 | ||
596 | * support systemd.mask= on the kernel command line. | |
597 | ||
598 | * reuse mkdtemp namespace dirs in /tmp? | |
599 | ||
600 | * recreate systemd's D-Bus private socket file on SIGUSR2 | |
601 | ||
602 | * Support --test based on current system state | |
603 | ||
604 | * investigate whether the gnome pty helper should be moved into systemd, to provide cgroup support. | |
605 | ||
606 | * maybe introduce ExecRestartPre= | |
607 | ||
608 | * dot output for --test showing the 'initial transaction' | |
609 | ||
610 | * port over to LISTEN_FDS/LISTEN_PID: | |
611 | - rpcbind (/var/run/rpcbind.sock!) HAVEPATCH | |
612 | - cups HAVEPATCH | |
613 | - postfix, saslauthd | |
614 | - apache/samba | |
615 | - libvirtd (/var/run/libvirt/libvirt-sock-ro) | |
616 | - bluetoothd (/var/run/sdp! @/org/bluez/audio!) | |
617 | - distccd | |
618 | ||
619 | * fingerprint.target, wireless.target, gps.target, netdevice.target | |
620 | ||
621 | * io priority during initialization | |
622 | ||
623 | * drop cap bounding set in readahead and other services | |
624 | ||
625 | * systemd-python: | |
626 | - figure out a simple way to wait for journal events in a way that | |
627 | works with ^C | |
628 | - add documentation to systemd.daemon | |
629 | ||
630 | External: | |
631 | ||
632 | * dbus: | |
633 | - dbus --user | |
634 | - natively watch for dbus-*.service symlinks (PENDING) | |
635 | - allow specification of socket mode/umask when allocating DBusServer | |
636 | - allow disabling of fd passing when connecting a AF_UNIX connection | |
637 | - allow disabling of UID passing for AUTH EXTERNAL | |
638 | - always pass cred data along each message | |
639 | - teach dbus to activate all services it finds in /etc/systemd/services/org-*.service | |
640 | ||
641 | * fix alsa mixer restore to not print error when no config is stored | |
642 | ||
643 | * make cryptsetup lower --iter-time | |
644 | ||
645 | * patch kernel for xattr support in /dev, /proc/, /sys? | |
646 | ||
647 | * NTP: the kernel's 11-minutes-mode syncs the system time to the RTC, but only | |
648 | in an ~30 minutes window. It does not adjust larger differences. Find a way | |
649 | to tell the kernel, to always do a full time sync when the RTC is in UTC and | |
650 | we are in 11-minutes-mode. When we trust the system time to NTP we also want | |
651 | the RTC to sync up. | |
652 | ||
653 | * kernel: add device_type = "fb", "fbcon" to class "graphics" | |
654 | ||
655 | * bootchart: | |
656 | - plot per-process IO utilization | |
657 | - group processes based on service association (cgroups) | |
658 | - document initcall_debug | |
659 | - put bootcharts in the journal | |
660 | - kernel cmdline "bootchart" option for simplicity? | |
661 | ||
662 | Regularly: | |
663 | ||
664 | * look for close() vs. close_nointr() vs. close_nointr_nofail() | |
665 | ||
666 | * check for strerror(r) instead of strerror(-r) | |
667 | ||
668 | * Use PR_SET_PROCTITLE_AREA if it becomes available in the kernel | |
669 | ||
670 | * %m in printf() instead of strerror(errno); | |
671 | ||
672 | * pahole | |
673 | ||
674 | * set_put(), hashmap_put() return values check. i.e. == 0 doesn't free()! | |
675 | ||
676 | * use secure_getenv() instead of getenv() where appropriate | |
677 | ||
678 | * link up selected blog stories from man pages and unit files Documentation= fields | |
679 | ||
680 | Scheduled for removal or fixing: | |
681 | ||
682 | * xxxOverridable dependencies (probably: fix) |