1 systemd System and Service Manager
4 * systemd-logingctl and systemd-journalctl have been renamed
5 to logingctl and journalctl to match systemctl.
7 * The config files: /etc/systemd/systemd-logind.conf and
8 /etc/systemd/systemd-journald.conf have been renamed to
9 logind.conf and journald.conf. Package updates should rename
10 the files to the new names on upgrade.
13 * This is mostly a bugfix release
15 * Support optional initialization of the machine ID from the
16 KVM or container configured UUID.
18 * Support immediate reboots with "systemctl reboot -ff"
20 * Show /etc/os-release data in systemd-analyze output
22 * Many bugfixes for the journal, including endianess fixes and
23 ensuring that disk space enforcement works
25 * sd-login.h is C++ comptaible again
27 * Extend the /etc/os-release format on request of the Debian
30 * We now refuse non-UTF8 strings used in various configuration
31 and unit files. This is done to ensure we don't pass invalid
32 data over D-Bus or expose it elsewhere.
34 * Register Mimo USB Screens as suitable for automatic seat
37 * Read SELinux client context from journal clients in a race
40 * Reorder configuration file lookup order. /etc now always
41 overrides /run in order to allow the administrator to always
42 and unconditionally override vendor supplied or
43 automatically generated data.
45 * The various user visible bits of the journal now have man
46 pages. We still lack man pages for the journal API calls
49 * We now ship all man pages in HTML format again in the
52 Contributions from: Dave Reisner, Dirk Eibach, Frederic
53 Crozat, Harald Hoyer, Kay Sievers, Lennart Poettering, Marti
54 Raudsepp, Michal Schmidt, Shawn Landden, Tero Roponen, Thierry
58 * This is mostly a bugfix release
60 * systems lacking /etc/os-release are no longer supported.
62 * Various functionality updates to libsystemd-login.so
64 * Track class of PAM logins to distuingish greeters from
67 Contributions from: Kay Sievers, Lennart Poettering, Michael
71 * This is an important bugfix release for v41.
73 * Building man pages is now optional which should be useful
74 for those building systemd from git but unwilling to install
77 * Watchdog support for supervising services is now usable. In
78 a future release support for hardware watchdogs
79 (i.e. /dev/watchdog) will be added building on this.
81 * Service start rate limiting is now configurable and can be
82 turned off per service. When a start rate limit is hit a
83 reboot can automatically be triggered.
85 * New CanReboot(), CanPowerOff() bus calls in systemd-logind.
87 Contributions from: Benjamin Franzke, Bill Nottingham,
88 Frederic Crozat, Lennart Poettering, Michael Olbrich, Michal
89 Schmidt, Michał Górny, Piotr Drąg
92 * The systemd binary is installed /usr/lib/systemd/systemd now;
93 An existing /sbin/init symlink needs to be adapted with the
96 * The code that loads kernel modules has been ported to invoke
97 libkmod directly, instead of modprobe. This means we do not
98 support systems with module-init-tools anymore.
100 * Watchdog support is now already useful, but still not
103 * A new kernel command line option systemd.setenv= is
104 understood to set system wide environment variables
107 * We now limit the set of capabilities of systemd-journald.
109 * We now set SIGPIPE to ignore by default, since it only is
110 useful in shell pipelines, and has little use in general
111 code. This can be disabled with IgnoreSIPIPE=no in unit
114 Contributions from: Benjamin Franzke, Kay Sievers, Lennart
115 Poettering, Michael Olbrich, Michal Schmidt, Tom Gundersen,
119 * This is mostly a bugfix release
121 * We now expose the reason why a service failed in the
122 "Result" D-Bus property.
124 * Rudimentary service watchdog support (will be completed over
125 the next few releases.)
127 * When systemd forks off in order execute some service we will
128 now immediately changes its argv[0] to reflect which process
129 it will execute. This is useful to minimize the time window
130 with a generic argv[0], which makes bootcharts more useful
132 Contributions from: Alvaro Soliverez, Chris Paulson-Ellis, Kay
133 Sievers, Lennart Poettering, Michael Olbrich, Michal Schmidt,
134 Mike Kazantsev, Ray Strode
137 * This is mostly a test release, but incorporates many
140 * New systemd-cgtop tool to show control groups by their
143 * Linking against libacl for ACLs is optional again. If
144 disabled, support tracking device access for active logins
145 goes becomes unavailable, and so does access to the user
146 journals by the respective users.
148 * If a group "adm" exists, journal files are automatically
149 owned by them, thus allow members of this group full access
150 to the system journal as well as all user journals.
152 * The journal now stores the SELinux context of the logging
153 client for all entries.
155 * Add C++ inclusion guards to all public headers
157 * New output mode "cat" in the journal to print only text
158 messages, without any meta data like date or time.
160 * Include tiny X server wrapper as a temporary stop-gap to
161 teach XOrg udev display enumeration. This is used by display
162 managers such as gdm, and will go away as soon as XOrg
163 learned native udev hotplugging for display devices.
165 * Add new systemd-cat tool for executing arbitrary programs
166 with STDERR/STDOUT connected to the journal. Can also act as
167 BSD logger replacement, and does so by default.
169 * Optionally store all locally generated coredumps in the
170 journal along with meta data.
172 * systemd-tmpfiles learnt four new commands: n, L, c, b, for
173 writing short strings to files (for usage for /sys), and for
174 creating symlinks, character and block device nodes.
176 * New unit file option ControlGroupPersistent= to make cgroups
177 persistent, following the mechanisms outlined in
178 http://www.freedesktop.org/wiki/Software/systemd/PaxControlGroups
180 * Support multiple local RTCs in a sane way
182 * No longer monopolize IO when replaying readahead data on
183 rotating disks, since we might starve non-file-system IO to
184 death, since fanotify() will not see accesses done by blkid,
187 * Don't show kernel threads in systemd-cgls anymore, unless
188 requested with new -k switch.
190 Contributions from: Dan Horák, Kay Sievers, Lennart
191 Poettering, Michal Schmidt
194 * This is mostly a test release, but incorporates many
197 * The git repository moved to:
198 git://anongit.freedesktop.org/systemd/systemd
199 ssh://git.freedesktop.org/git/systemd/systemd
201 * First release with the journal
202 http://0pointer.de/blog/projects/the-journal.html
204 * The journal replaces both systemd-kmsg-syslogd and
205 systemd-stdout-bridge.
207 * New sd_pid_get_unit() API call in libsystemd-logind
209 * Many systemadm clean-ups
211 * Introduce remote-fs-pre.target which is ordered before all
212 remote mounts and may be used to start services before all
215 * Added Mageia support
217 * Add bash completion for systemd-loginctl
219 * Actively monitor PID file creation for daemons which exit in
220 the parent process before having finished writing the PID
221 file in the daemon process. Daemons which do this need to be
222 fixed (i.e. PID file creation must have finished before the
223 parent exits), but we now react a bit more gracefully to them.
225 * Add colourful boot output, mimicking the well-known output
226 of existing distributions.
228 * New option PassCredentials= for socket units, for
229 compatibility with a recent kernel ABI breakage.
231 * /etc/rc.local is now hooked in via a generator binary, and
232 thus will no longer act as synchronization point during
235 * systemctl list-unit-files now supports --root=.
237 * systemd-tmpfiles now understands two new commands: z, Z for
238 relabelling files according to the SELinux database. This is
239 useful to apply SELinux labels to specific files in /sys,
242 * Output of SysV services is now forwarded to both the console
243 and the journal by default, not only just the console.
245 * New man pages for all APIs from libsystemd-login.
247 * The build tree got reorganized and a the build system is a
248 lot more modular allowing embedded setups to specifically
249 select the components of systemd they are interested in.
251 * Support for Linux systems lacking the kernel VT subsystem is
254 * configure's --with-rootdir= got renamed to
255 --with-rootprefix= to follow the naming used by udev and
258 * Unless specified otherwise we'll now install to /usr instead
259 of /usr/local by default.
261 * Processes with '@' in argv[0][0] are now excluded from the
262 final shut-down killing spree, following the logic explained
264 http://www.freedesktop.org/wiki/Software/systemd/RootStorageDaemons
266 * All processes remaining in a service cgroup when we enter
267 the START or START_PRE states are now killed with
268 SIGKILL. That means it is no longer possible to spawn
269 background processes from ExecStart= lines (which was never
270 supported anyway, and bad style).
272 * New PropagateReloadTo=/PropagateReloadFrom= options to bind
273 reloading of units together.
275 Contributions from: Bill Nottingham, Daniel Walsh, Dave
276 Reisner, Dexter Morgan, Gregs Gregs, Jonathan Nieder, Kay
277 Sievers, Lennart Poettering, Michael Biebl, Michal Schmidt,
278 Michał Górny, Ran Benita, Thomas Jarosch, Tim Waugh, Tollef
279 Fog Heen, Tom Gundersen, Zbigniew Jędrzejewski-Szmek