]> git.ipfire.org Git - thirdparty/systemd.git/blob - NEWS
service: replace "failure" boolean by enum, and export it
[thirdparty/systemd.git] / NEWS
1 systemd System and Service Manager
2
3 CHANGES WITH 39:
4 * This is mostly a test release, but incorporates many
5 bugfixes.
6
7 * New systemd-cgtop tool to show control groups by their
8 resource usage.
9
10 * Linking against libacl for ACLs is optional again. If
11 disabled, support tracking device access for active logins
12 goes becomes unavailable, and so does access to the user
13 journals by the respective users.
14
15 * If a group "adm" exists, journal files are automatically
16 owned by them, thus allow members of this group full access
17 to the system journal as well as all user journals.
18
19 * The journal now stores the SELinux context of the logging
20 client for all entries.
21
22 * Add C++ inclusion guards to all public headers
23
24 * New output mode "cat" in the journal to print only text
25 messages, without any meta data like date or time.
26
27 * Include tiny X server wrapper as a temporary stop-gap to
28 teach XOrg udev display enumeration. This is used by display
29 managers such as gdm, and will go away as soon as XOrg
30 learned native udev hotplugging for display devices.
31
32 * Add new systemd-cat tool for executing arbitrary programs
33 with STDERR/STDOUT connected to the journal. Can also act as
34 BSD logger replacement, and does so by default.
35
36 * Optionally store all locally generated coredumps in the
37 journal along with meta data.
38
39 * systemd-tmpfiles learnt four new commands: n, L, c, b, for
40 writing short strings to files (for usage for /sys), and for
41 creating symlinks, character and block device nodes.
42
43 * New unit file option ControlGroupPersistent= to make cgroups
44 persistent, following the mechanisms outlined in
45 http://www.freedesktop.org/wiki/Software/systemd/PaxControlGroups
46
47 * Support multiple local RTCs in a sane way
48
49 * No longer monopolize IO when replaying readahead data on
50 rotating disks, since we might starve non-file-system IO to
51 death, since fanotify() will not see accesses done by blkid,
52 or fsck.
53
54 * Don't show kernel threads in systemd-cgls anymore, unless
55 requested with new -k switch.
56
57 Contributions from: Dan Horák, Kay Sievers, Lennart
58 Poettering, Michal Schmidt
59
60 CHANGES WITH 38:
61 * This is mostly a test release, but incorporates many
62 bugfixes.
63
64 * The git repository moved to:
65 git://anongit.freedesktop.org/systemd/systemd
66 ssh://git.freedesktop.org/git/systemd/systemd
67
68 * First release with the journal
69 http://0pointer.de/blog/projects/the-journal.html
70
71 * The journal replaces both systemd-kmsg-syslogd and
72 systemd-stdout-bridge.
73
74 * New sd_pid_get_unit() API call in libsystemd-logind
75
76 * Many systemadm clean-ups
77
78 * Introduce remote-fs-pre.target which is ordered before all
79 remote mounts and may be used to start services before all
80 remote mounts.
81
82 * Added Mageia support
83
84 * Add bash completion for systemd-loginctl
85
86 * Actively monitor PID file creation for daemons which exit in
87 the parent process before having finished writing the PID
88 file in the daemon process. Daemons which do this need to be
89 fixed (i.e. PID file creation must have finished before the
90 parent exits), but we now react a bit more gracefully to them.
91
92 * Add colourful boot output, mimicking the well-known output
93 of existing distributions.
94
95 * New option PassCredentials= for socket units, for
96 compatibility with a recent kernel ABI breakage.
97
98 * /etc/rc.local is now hooked in via a generator binary, and
99 thus will no longer act as synchronization point during
100 boot.
101
102 * systemctl list-unit-files now supports --root=.
103
104 * systemd-tmpfiles now understands two new commands: z, Z for
105 relabelling files according to the SELinux database. This is
106 useful to apply SELinux labels to specific files in /sys,
107 among other things.
108
109 * Output of SysV services is now forwarded to both the console
110 and the journal by default, not only just the console.
111
112 * New man pages for all APIs from libsystemd-login.
113
114 * The build tree got reorganized and a the build system is a
115 lot more modular allowing embedded setups to specifically
116 select the components of systemd they are interested in.
117
118 * Support for Linux systems lacking the kernel VT subsystem is
119 restored.
120
121 * configure's --with-rootdir= got renamed to
122 --with-rootprefix= to follow the naming used by udev and
123 kmod
124
125 * Unless specified otherwise we'll now install to /usr instead
126 of /usr/local by default.
127
128 * Processes with '@' in argv[0][0] are now excluded from the
129 final shut-down killing spree, following the logic explained
130 in:
131 http://www.freedesktop.org/wiki/Software/systemd/RootStorageDaemons
132
133 * All processes remaining in a service cgroup when we enter
134 the START or START_PRE states are now killed with
135 SIGKILL. That means it is no longer possible to spawn
136 background processes from ExecStart= lines (which was never
137 supported anyway, and bad style).
138
139 * New PropagateReloadTo=/PropagateReloadFrom= options to bind
140 reloading of units together.
141
142 Contributions from: Bill Nottingham, Daniel Walsh, Dave
143 Reisner, Dexter Morgan, Gregs Gregs, Jonathan Nieder, Kay
144 Sievers, Lennart Poettering, Michael Biebl, Michal Schmidt,
145 Michał Górny, Ran Benita, Thomas Jarosch, Tim Waugh, Tollef
146 Fog Heen, Tom Gundersen, Zbigniew Jędrzejewski-Szmek