chiark / gitweb /
elogind.git
10 years agolog: support logging into SOCK_STREAM /dev/log sockets for comaptibility with syslog-ng
Lennart Poettering [Wed, 8 Sep 2010 01:07:44 +0000 (03:07 +0200)]
log: support logging into SOCK_STREAM /dev/log sockets for comaptibility with syslog-ng

10 years agounits: On Debian Runlevel 2, 3, 4 and 5 are multi-user
Michael Biebl [Tue, 7 Sep 2010 17:58:14 +0000 (19:58 +0200)]
units: On Debian Runlevel 2, 3, 4 and 5 are multi-user

Patch by Tollef Fog Heen <tfheen@err.no>
Extended to re-create the runlevel5.target symlink by Michael Biebl <biebl@debian.org>

10 years agounits: add umount unit files for Debian
Michael Biebl [Mon, 6 Sep 2010 03:44:00 +0000 (05:44 +0200)]
units: add umount unit files for Debian

Add unit files to call

/etc/init.d/umountnfs.sh stop (network file systems)
/etc/init.d/umountfs stop     (local file systems)
/etc/init.d/umountroot stop   ("/" file system)

in the right order and hook them up in the umount.target so they are run
on shutdown and reboot.

10 years agosd-daemon: optionally, avoid hiding the ELF symbols
Lennart Poettering [Tue, 7 Sep 2010 23:55:17 +0000 (01:55 +0200)]
sd-daemon: optionally, avoid hiding the ELF symbols

10 years agomain: when we encounter 'emergency' on the kernel cmdline start the emergency shell
Lennart Poettering [Tue, 7 Sep 2010 23:52:39 +0000 (01:52 +0200)]
main: when we encounter 'emergency' on the kernel cmdline start the emergency shell

10 years agounits: make sure we always fully write the utmp data
Lennart Poettering [Mon, 6 Sep 2010 13:01:31 +0000 (15:01 +0200)]
units: make sure we always fully write the utmp data

10 years agoman: Fix small typo: s/seperate/separate/
Michael Biebl [Fri, 3 Sep 2010 14:30:48 +0000 (16:30 +0200)]
man: Fix small typo: s/seperate/separate/

10 years agounits: Provide separate fsck.target for Debian
Michael Biebl [Fri, 3 Sep 2010 14:23:05 +0000 (16:23 +0200)]
units: Provide separate fsck.target for Debian

On Debian sysinit is not a single script but a separate runlevel.
Split of fsck.target into separate unit file as otherwise we get an
unbreakable cycle on shutdown/reboot.

10 years agounits: Add halt, killall, poweroff and reboot unit files for Debian
Michael Biebl [Fri, 3 Sep 2010 13:16:35 +0000 (15:16 +0200)]
units: Add halt, killall, poweroff and reboot unit files for Debian

10 years agosystemctl: make --version a little bit more verbose
Lennart Poettering [Mon, 6 Sep 2010 01:11:24 +0000 (03:11 +0200)]
systemctl: make --version a little bit more verbose

10 years agosystemctl: implement --version
Lennart Poettering [Mon, 6 Sep 2010 00:42:49 +0000 (02:42 +0200)]
systemctl: implement --version

https://bugs.freedesktop.org/show_bug.cgi?id=29912

10 years agodbus: downgrade activation failure messages
Lennart Poettering [Mon, 6 Sep 2010 00:42:42 +0000 (02:42 +0200)]
dbus: downgrade activation failure messages

10 years agoautomount: make sure we don't crash if there's an automount unit without mount unit
Lennart Poettering [Sun, 5 Sep 2010 23:04:59 +0000 (01:04 +0200)]
automount: make sure we don't crash if there's an automount unit without mount unit

10 years agowork around busted capability.h
Kay Sievers [Fri, 3 Sep 2010 10:26:30 +0000 (12:26 +0200)]
work around busted capability.h

It contains things like:
  #ifndef _LINUX_TYPES_H
  #define _LINUX_TYPES_H
  #include <stdint.h>
  typedef uint32_t __u32;
  typedef __u32 __le32;
  ...

which result in:
  /usr/include/asm-generic/int-ll64.h:30:42: error: conflicting types for ‘__u64’
  /usr/include/sys/capability.h:39:18: note: previous declaration of ‘__u64’ was here
  make[1]: *** [src/libsystemd_core_la-unit.lo] Error 1
  make[1]: Leaving directory `/home/kay/work/src/systemd'
  make: *** [all] Error 2

So, include linux/types.h to get around that.

10 years agobuild-sys: fix fedpkg git path
Lennart Poettering [Fri, 3 Sep 2010 03:39:22 +0000 (05:39 +0200)]
build-sys: fix fedpkg git path

10 years agobuild-sys: prepare new release v9
Lennart Poettering [Fri, 3 Sep 2010 02:51:02 +0000 (04:51 +0200)]
build-sys: prepare new release

10 years agoman: extend documentation on systemctl isolate
Matthew Miller [Tue, 31 Aug 2010 16:25:22 +0000 (12:25 -0400)]
man: extend documentation on systemctl isolate

Lennart has convinced me that it's more helpful to participate than to sit
on the sidelines and complain. So, hello everyone.

I'm starting by giving up the battle to change the systemctl "isolate"
command to "switch-to". Can't win them all. :) I've got a suggested patch
to expand the documentation a bit, hopefully making it more clear to new
systemd users.

Is there an easy way to list all units where AllowIsolate is enabled? That
should be included alongside this, I think.

10 years agoservice: calculate sysv startup priority only in start runlevels
Lennart Poettering [Fri, 3 Sep 2010 02:08:16 +0000 (04:08 +0200)]
service: calculate sysv startup priority only in start runlevels

10 years agoupdate fixme
Lennart Poettering [Fri, 3 Sep 2010 01:58:11 +0000 (03:58 +0200)]
update fixme

10 years agoservice: make sure we properly figure out whether a sysv service is enabled before...
Lennart Poettering [Fri, 3 Sep 2010 01:56:05 +0000 (03:56 +0200)]
service: make sure we properly figure out whether a sysv service is enabled before creating $foo wants links

10 years agounits: don't add shutdown conflicts dep to umount.target
Lennart Poettering [Fri, 3 Sep 2010 00:20:39 +0000 (02:20 +0200)]
units: don't add shutdown conflicts dep to umount.target

10 years agounit: don't retroactively start units if we failed to shutdown a unit
Lennart Poettering [Thu, 2 Sep 2010 23:45:35 +0000 (01:45 +0200)]
unit: don't retroactively start units if we failed to shutdown a unit

10 years agodbus: don't send cgroup agent messages directly to system bus to avoid dbus activatio...
Lennart Poettering [Thu, 2 Sep 2010 23:45:14 +0000 (01:45 +0200)]
dbus: don't send cgroup agent messages directly to system bus to avoid dbus activation loop

10 years agotarget: implicitly shutdown targets on shutdown, so that they are pulled into the...
Lennart Poettering [Thu, 2 Sep 2010 23:42:19 +0000 (01:42 +0200)]
target: implicitly shutdown targets on shutdown, so that they are pulled into the final transaction and may be used for synchronization

10 years agoservice: properly handle if the main process is down and we wait for the cgroup to...
Lennart Poettering [Thu, 2 Sep 2010 23:41:40 +0000 (01:41 +0200)]
service: properly handle if the main process is down and we wait for the cgroup to die, too

10 years agodbus: make sure to flush bus connection before disconnecting, so that no messages...
Lennart Poettering [Thu, 2 Sep 2010 21:26:04 +0000 (23:26 +0200)]
dbus: make sure to flush bus connection before disconnecting, so that no messages are lost

10 years agounit: unify some code
Lennart Poettering [Wed, 1 Sep 2010 01:35:04 +0000 (03:35 +0200)]
unit: unify some code

10 years agodbus: don't accept activation requests anymore if we are going down anyway
Lennart Poettering [Wed, 1 Sep 2010 01:30:59 +0000 (03:30 +0200)]
dbus: don't accept activation requests anymore if we are going down anyway

10 years agosystemctl: make sure daemon-reexec and friends return a correct error code
Lennart Poettering [Wed, 1 Sep 2010 00:11:52 +0000 (02:11 +0200)]
systemctl: make sure daemon-reexec and friends return a correct error code

10 years agosystemctl: fix return value of systemctl start and friends
Lennart Poettering [Wed, 1 Sep 2010 00:03:41 +0000 (02:03 +0200)]
systemctl: fix return value of systemctl start and friends

10 years agoservice: recheck PID status when a cgroup runs empty
Lennart Poettering [Tue, 31 Aug 2010 23:33:12 +0000 (01:33 +0200)]
service: recheck PID status when a cgroup runs empty

10 years agoservice: wait for process exit only if we actually killed somebody
Lennart Poettering [Tue, 31 Aug 2010 22:48:05 +0000 (00:48 +0200)]
service: wait for process exit only if we actually killed somebody

10 years agomanager: reword some error messages regarding "logs"
Lennart Poettering [Tue, 31 Aug 2010 22:32:32 +0000 (00:32 +0200)]
manager: reword some error messages regarding "logs"

https://bugzilla.redhat.com/show_bug.cgi?id=627785

10 years agoutil: leave locked terminal attributes untouched so that we don't interfere with...
Lennart Poettering [Tue, 31 Aug 2010 22:10:41 +0000 (00:10 +0200)]
util: leave locked terminal attributes untouched so that we don't interfere with plymouth

https://bugzilla.redhat.com/show_bug.cgi?id=628913

10 years agoservice: properly pass dbus error structure
Michal Schmidt [Tue, 31 Aug 2010 21:52:51 +0000 (23:52 +0200)]
service: properly pass dbus error structure

10 years agomanager: make sure we free the dbus error only if we actually use one
Lennart Poettering [Tue, 31 Aug 2010 21:51:12 +0000 (23:51 +0200)]
manager: make sure we free the dbus error only if we actually use one

10 years agoservice: rework killing logic so that we always kill the main process, even if it...
Lennart Poettering [Tue, 31 Aug 2010 21:24:47 +0000 (23:24 +0200)]
service: rework killing logic so that we always kill the main process, even if it left our service cgroup

Related to:

http://bugzilla.redhat.com/show_bug.cgi?id=626477

10 years agosystemctl: rework exit codes for all utility programs to follow LSB or other standards
Lennart Poettering [Tue, 31 Aug 2010 19:05:54 +0000 (21:05 +0200)]
systemctl: rework exit codes for all utility programs to follow LSB or other standards

10 years agosystemctl: avoid wrapping of headers/footers
Lennart Poettering [Tue, 31 Aug 2010 14:56:08 +0000 (16:56 +0200)]
systemctl: avoid wrapping of headers/footers

This change makes sure that our output never exceeds 80ch on 80ch
terminals.

https://bugzilla.redhat.com/show_bug.cgi?id=626443

10 years agoutil: look at STDOUT, not STDIN, when determining tty columns
Lennart Poettering [Tue, 31 Aug 2010 14:53:58 +0000 (16:53 +0200)]
util: look at STDOUT, not STDIN, when determining tty columns

https://bugzilla.redhat.com/show_bug.cgi?id=626891

10 years agoexec: replace OOMAdjust= by OOMScoreAdjust= to follow new kernel interface
Lennart Poettering [Mon, 30 Aug 2010 23:33:39 +0000 (01:33 +0200)]
exec: replace OOMAdjust= by OOMScoreAdjust= to follow new kernel interface

This replaces OOMAdjust= by OOMScoreAdjust= in the config files,
breaking compatibility with older unit files. However, this keeps compat
with older kernels which lack the new OOM rework.

10 years agounits: fix a typo in arch's rc-local.service unit file
Lennart Poettering [Mon, 30 Aug 2010 22:43:57 +0000 (00:43 +0200)]
units: fix a typo in arch's rc-local.service unit file

10 years agomanager: add missing second part of s/maintenance/failed/
Matthew Miller [Mon, 30 Aug 2010 22:23:34 +0000 (00:23 +0200)]
manager: add missing second part of s/maintenance/failed/

10 years agomanager: rename 'maintenance' state to 'failed' to avoid user confusion
Matthew Miller [Mon, 30 Aug 2010 22:17:56 +0000 (00:17 +0200)]
manager: rename 'maintenance' state to 'failed' to avoid user confusion

https://bugzilla.redhat.com/show_bug.cgi?id=614619

10 years agomanager: don't try to minimize transaction when using isolate
Lennart Poettering [Mon, 30 Aug 2010 21:51:52 +0000 (23:51 +0200)]
manager: don't try to minimize transaction when using isolate

We should not avoid stopping services when using isolate, since that
kinda defeats the point of it.

http://bugzilla.redhat.com/show_bug.cgi?id=627014

10 years agoexec: make sure O_NONBLOCK is off for all sockets passed as STDIN/STDOUT/STDERR
Lennart Poettering [Mon, 30 Aug 2010 21:31:27 +0000 (23:31 +0200)]
exec: make sure O_NONBLOCK is off for all sockets passed as STDIN/STDOUT/STDERR

10 years agounit: introduce AllowIsolate= switch
Lennart Poettering [Mon, 30 Aug 2010 20:45:46 +0000 (22:45 +0200)]
unit: introduce AllowIsolate= switch

10 years agoconsole: rework automatic getty on kernel console logic again
Lennart Poettering [Mon, 30 Aug 2010 19:31:40 +0000 (21:31 +0200)]
console: rework automatic getty on kernel console logic again

It is essential that the gettys are proper dependencies from
getty.target so that they aren't killed and immediately restarted on
runlevel changes. Hence rework the logic to implicitly add console
gettys to getty.target as dependencies.

This also adds an automatic hvc console for virtualizers.

https://bugzilla.redhat.com/show_bug.cgi?id=501720

10 years agoupdate fixme
Lennart Poettering [Mon, 30 Aug 2010 19:45:19 +0000 (21:45 +0200)]
update fixme

10 years agovar-run: make sure /var/run/user is properly created and labelled
Lennart Poettering [Thu, 26 Aug 2010 22:07:04 +0000 (00:07 +0200)]
var-run: make sure /var/run/user is properly created and labelled

https://bugzilla.redhat.com/show_bug.cgi?id=627457

10 years agounits: don't install runlevel links anymore, since thsoe are maintained in /lib now
Lennart Poettering [Thu, 26 Aug 2010 00:36:04 +0000 (02:36 +0200)]
units: don't install runlevel links anymore, since thsoe are maintained in /lib now

10 years agoservice: properly handle Fedora's broken MTA meta-service name
Lennart Poettering [Thu, 26 Aug 2010 00:28:37 +0000 (02:28 +0200)]
service: properly handle Fedora's broken MTA meta-service name

10 years agoservice: reword warnings a little
Lennart Poettering [Thu, 26 Aug 2010 02:10:17 +0000 (04:10 +0200)]
service: reword warnings a little

10 years agoservice: if we cannot add an alias to a sysv service, ignore this and go on
Lennart Poettering [Thu, 26 Aug 2010 02:08:28 +0000 (04:08 +0200)]
service: if we cannot add an alias to a sysv service, ignore this and go on

10 years agodevice: allow aliases on devices
Lennart Poettering [Thu, 26 Aug 2010 02:08:01 +0000 (04:08 +0200)]
device: allow aliases on devices

10 years agobuild-sys: show audit/selinux in summary
Lennart Poettering [Thu, 26 Aug 2010 01:23:03 +0000 (03:23 +0200)]
build-sys: show audit/selinux in summary

10 years agosystemctl: very very trivial typo patch :)
Matthew Miller [Wed, 25 Aug 2010 20:12:50 +0000 (16:12 -0400)]
systemctl: very very trivial typo patch :)

10 years agobuild-sys: prepare release v8 v8
Lennart Poettering [Wed, 25 Aug 2010 19:28:50 +0000 (21:28 +0200)]
build-sys: prepare release v8

10 years agomount: imply noauto for mount points configured with native mount files
Lennart Poettering [Wed, 25 Aug 2010 19:24:21 +0000 (21:24 +0200)]
mount: imply noauto for mount points configured with native mount files

10 years agoupdate fixme
Lennart Poettering [Wed, 25 Aug 2010 18:39:45 +0000 (20:39 +0200)]
update fixme

10 years agomount: rework automatic mounting to follow the 'nofail' option in fstab
Lennart Poettering [Wed, 25 Aug 2010 18:37:04 +0000 (20:37 +0200)]
mount: rework automatic mounting to follow the 'nofail' option in fstab

10 years agoupdate fixme
Lennart Poettering [Wed, 25 Aug 2010 17:49:52 +0000 (19:49 +0200)]
update fixme

10 years agodbus: simplify some dbus error messages a bit
Lennart Poettering [Wed, 25 Aug 2010 17:49:23 +0000 (19:49 +0200)]
dbus: simplify some dbus error messages a bit

10 years agosd-daemon: drop SCM_CREDENTIALS passing since it is redundant, the kernel fills this...
Lennart Poettering [Wed, 25 Aug 2010 17:37:36 +0000 (19:37 +0200)]
sd-daemon: drop SCM_CREDENTIALS passing since it is redundant, the kernel fills this in anyway

10 years agosystemd-remount-api-vfs.service: fix typo
Kay Sievers [Wed, 25 Aug 2010 13:06:21 +0000 (15:06 +0200)]
systemd-remount-api-vfs.service: fix typo

10 years agounits: fix auto-serial unit
Lennart Poettering [Wed, 25 Aug 2010 03:02:48 +0000 (05:02 +0200)]
units: fix auto-serial unit

10 years agocgroup: mount cgroup file systems to /sys/fs/cgroup instead of /cgroup
Lennart Poettering [Wed, 25 Aug 2010 02:58:16 +0000 (04:58 +0200)]
cgroup: mount cgroup file systems to /sys/fs/cgroup instead of /cgroup

This requires the /sys/fs/cgroup to exist. Current git kernels have
this, released kernels need a backport of a single patch.

10 years agoutmp: document that the order of preference for runlevel numbers might be something...
Lennart Poettering [Wed, 25 Aug 2010 01:31:59 +0000 (03:31 +0200)]
utmp: document that the order of preference for runlevel numbers might be something to change in other distributions

10 years agounits: make sure prefdm is automatically respawned
Lennart Poettering [Wed, 25 Aug 2010 01:15:32 +0000 (03:15 +0200)]
units: make sure prefdm is automatically respawned

10 years agounits: invoke sulogin instead of /bin/sh
Lennart Poettering [Wed, 25 Aug 2010 01:15:12 +0000 (03:15 +0200)]
units: invoke sulogin instead of /bin/sh

10 years agoutmp: write out runlevels 5 and 3 preferrably if multiples make sense, and S instead...
Lennart Poettering [Wed, 25 Aug 2010 01:14:53 +0000 (03:14 +0200)]
utmp: write out runlevels 5 and 3 preferrably if multiples make sense, and S instead of 1

10 years agosystemctl: rename a few status lines
Lennart Poettering [Wed, 25 Aug 2010 01:14:04 +0000 (03:14 +0200)]
systemctl: rename a few status lines

10 years agosystemctl: show timestamps for state changes
Lennart Poettering [Wed, 25 Aug 2010 01:13:44 +0000 (03:13 +0200)]
systemctl: show timestamps for state changes

10 years agosystemctl: add --sysv-compat
Lennart Poettering [Wed, 25 Aug 2010 01:13:09 +0000 (03:13 +0200)]
systemctl: add --sysv-compat

10 years agolog: SCM_CREDENTIALS will be added by the kernel anyway, so there's no need to write...
Lennart Poettering [Wed, 25 Aug 2010 01:12:19 +0000 (03:12 +0200)]
log: SCM_CREDENTIALS will be added by the kernel anyway, so there's no need to write them manually

10 years agomount: add global configuration options for handling of auto mounts
Lennart Poettering [Wed, 25 Aug 2010 01:11:26 +0000 (03:11 +0200)]
mount: add global configuration options for handling of auto mounts

10 years agoman: minor updates
Lennart Poettering [Wed, 25 Aug 2010 01:10:13 +0000 (03:10 +0200)]
man: minor updates

10 years agoserial: use seperate getty template for serial ttys
Lennart Poettering [Wed, 25 Aug 2010 01:09:57 +0000 (03:09 +0200)]
serial: use seperate getty template for serial ttys

10 years agosyslog: add minimal syslog/kmsg bridge syslogd
Lennart Poettering [Sat, 21 Aug 2010 01:57:47 +0000 (03:57 +0200)]
syslog: add minimal syslog/kmsg bridge syslogd

10 years agomount: don't create depdency for device node of root fs, since it is always there
Lennart Poettering [Fri, 20 Aug 2010 23:51:55 +0000 (01:51 +0200)]
mount: don't create depdency for device node of root fs, since it is always there

10 years agolog: pass SCM_CREDENTIALS when logging to syslog
Lennart Poettering [Fri, 20 Aug 2010 23:51:16 +0000 (01:51 +0200)]
log: pass SCM_CREDENTIALS when logging to syslog

10 years agodevice: handle properly if two devices want to have the same name
Lennart Poettering [Fri, 20 Aug 2010 02:04:05 +0000 (04:04 +0200)]
device: handle properly if two devices want to have the same name

https://bugzilla.redhat.com/show_bug.cgi?id=624539

10 years agounits: add service file that remounts the root file system
Lennart Poettering [Fri, 20 Aug 2010 01:27:23 +0000 (03:27 +0200)]
units: add service file that remounts the root file system

10 years agoexit-status: add missing files src/exit-status.h src/exit-status.c
Lennart Poettering [Fri, 20 Aug 2010 01:26:37 +0000 (03:26 +0200)]
exit-status: add missing files src/exit-status.h src/exit-status.c

10 years agoremount: add tool that applies /etc/fstab mount options to all api mounts
Lennart Poettering [Fri, 20 Aug 2010 01:26:15 +0000 (03:26 +0200)]
remount: add tool that applies /etc/fstab mount options to all api mounts

10 years agomount: properly handle LABEL="" in fstab
Lennart Poettering [Fri, 20 Aug 2010 00:46:15 +0000 (02:46 +0200)]
mount: properly handle LABEL="" in fstab

10 years agoselinux: properly query policy for FIFO files
Lennart Poettering [Fri, 20 Aug 2010 00:27:05 +0000 (02:27 +0200)]
selinux: properly query policy for FIFO files

10 years agodbus: follow standardized fdo PropertiesChanged signal spec
Lennart Poettering [Fri, 20 Aug 2010 00:26:05 +0000 (02:26 +0200)]
dbus: follow standardized fdo PropertiesChanged signal spec

10 years agoservice/systemctl: don't consider LSB exit codes 5 and 6 as failure, and decode exit...
Lennart Poettering [Thu, 19 Aug 2010 01:18:49 +0000 (03:18 +0200)]
service/systemctl: don't consider LSB exit codes 5 and 6 as failure, and decode exit codes in systemctl

10 years agoservice: make sure sysv services marked as interactive in the LSB header get output...
Lennart Poettering [Thu, 19 Aug 2010 00:42:07 +0000 (02:42 +0200)]
service: make sure sysv services marked as interactive in the LSB header get output on the TTY in all cases

10 years agoupdate fixme
Kay Sievers [Thu, 19 Aug 2010 08:02:42 +0000 (10:02 +0200)]
update fixme

10 years agounits: fix After= setting in var-lock.service
Lennart Poettering [Thu, 19 Aug 2010 00:23:49 +0000 (02:23 +0200)]
units: fix After= setting in var-lock.service

10 years agokmod: downgrade modprobe log message for since ipv6 might be compiled as module for...
Lennart Poettering [Thu, 19 Aug 2010 01:30:36 +0000 (03:30 +0200)]
kmod: downgrade modprobe log message for since ipv6 might be compiled as module for blacklisting

10 years agohostname: if no hostname is configured use localhost
Lennart Poettering [Thu, 19 Aug 2010 01:29:43 +0000 (03:29 +0200)]
hostname: if no hostname is configured use localhost

10 years agohostname: on all distros make the name configured in /etc/hostname take precedence...
Lennart Poettering [Thu, 19 Aug 2010 01:02:22 +0000 (03:02 +0200)]
hostname: on all distros make the name configured in /etc/hostname take precedence over distro-specific configuration

In order to unify configuration across distributions we pick the
simple-most option by default (Debian's /etc/hostname) and then fall
back to distro-specific hacks if that doesn't exist.

10 years agoupdate fixme
Lennart Poettering [Thu, 19 Aug 2010 00:04:47 +0000 (02:04 +0200)]
update fixme

10 years agomodules-load: add systemd-modules-load tool that can load a configured list of module...
Lennart Poettering [Thu, 19 Aug 2010 00:03:03 +0000 (02:03 +0200)]
modules-load: add systemd-modules-load tool that can load a configured list of modules in /etc/modules.d/ on boot, replacing distro-dependent shell hacks for this

10 years agounits: make sure getty@xxx.service is always run before gettys.target
Lennart Poettering [Wed, 18 Aug 2010 21:10:27 +0000 (23:10 +0200)]
units: make sure getty@xxx.service is always run before gettys.target

10 years agodbus: small typo in introspection data
Matthias Clasen [Wed, 18 Aug 2010 20:27:58 +0000 (16:27 -0400)]
dbus: small typo in introspection data

Hey,

here is a patch to fix a small typo in the systemd dbus introspection data.
I found this while trying to explore the systemd dbus API via bash
completion with gdbus. You should try this, it is really fun, and,
evidently, helps to find some bugs...

Matthias

10 years agorandom-seed: honour kernel pool size when saving/restoring seed
Lennart Poettering [Wed, 18 Aug 2010 17:38:45 +0000 (19:38 +0200)]
random-seed: honour kernel pool size when saving/restoring seed