]> err.no Git - systemd/log
systemd
13 years agoREADME: update requirements
Lennart Poettering [Wed, 16 Feb 2011 18:09:11 +0000 (19:09 +0100)]
README: update requirements

13 years agoudev: ignore block devices which no known contents, to avoid trying of mounts/swapons...
Lennart Poettering [Wed, 16 Feb 2011 11:57:00 +0000 (12:57 +0100)]
udev: ignore block devices which no known contents, to avoid trying of mounts/swapons when devices aren't set up full yet

13 years agoswap: handle "nofail" from fstab
Kay Sievers [Tue, 15 Feb 2011 17:38:15 +0000 (18:38 +0100)]
swap: handle "nofail" from fstab

13 years agocleanup TODO
Lennart Poettering [Tue, 15 Feb 2011 12:09:00 +0000 (13:09 +0100)]
cleanup TODO

13 years agomount,swap: properly add dependencies to logger if needed
Lennart Poettering [Tue, 15 Feb 2011 11:28:26 +0000 (12:28 +0100)]
mount,swap: properly add dependencies to logger if needed

13 years agoservice: change default std output to inherit
Lennart Poettering [Tue, 15 Feb 2011 11:27:29 +0000 (12:27 +0100)]
service: change default std output to inherit

13 years agoexec: introduce global defaults for the standard output of services
Lennart Poettering [Tue, 15 Feb 2011 10:52:29 +0000 (11:52 +0100)]
exec: introduce global defaults for the standard output of services

13 years agoudev: use SYSTEMD_READY to mask uninitialized DM devices
Lennart Poettering [Tue, 15 Feb 2011 09:59:46 +0000 (10:59 +0100)]
udev: use SYSTEMD_READY to mask uninitialized DM devices

13 years agofsck: output to /dev/console by default in addition to syslog
Lennart Poettering [Tue, 15 Feb 2011 00:28:31 +0000 (01:28 +0100)]
fsck: output to /dev/console by default in addition to syslog

13 years agoexecute: optionally forward program output to /dev/console in addition to syslog...
Lennart Poettering [Tue, 15 Feb 2011 00:27:53 +0000 (01:27 +0100)]
execute: optionally forward program output to /dev/console in addition to syslog/kmsg

13 years agosocket: refuse socket activation for SysV services
Lennart Poettering [Mon, 14 Feb 2011 23:40:17 +0000 (00:40 +0100)]
socket: refuse socket activation for SysV services

Make sure that when a .socket unit is installed without its matching
.service we don't end up activating a legacy SysV/LSB service with the
same name. SysV/LSB style services do not support passing sockets and we
don't want to extend SysV/LSB to ensure we don't break compatibility
with other systems.

13 years agofsck: do not fail boot if fsck returns with an error code that hasn't 2 or 6 set
Lennart Poettering [Mon, 14 Feb 2011 23:30:51 +0000 (00:30 +0100)]
fsck: do not fail boot if fsck returns with an error code that hasn't 2 or 6 set

13 years agoshutdown: execute all binaries in /lib/systemd/system-shutdown as last step before...
Lennart Poettering [Mon, 14 Feb 2011 23:30:11 +0000 (00:30 +0100)]
shutdown: execute all binaries in /lib/systemd/system-shutdown as last step before invoking reboot()

13 years agojob: make status message printing more verbose
Lennart Poettering [Mon, 14 Feb 2011 22:10:31 +0000 (23:10 +0100)]
job: make status message printing more verbose

13 years agocryptsetup: fix unit file description
Lennart Poettering [Mon, 14 Feb 2011 22:09:27 +0000 (23:09 +0100)]
cryptsetup: fix unit file description

13 years agotmpfiles: never delete AF_UNIX sockets that are alive
Lennart Poettering [Mon, 14 Feb 2011 20:55:06 +0000 (21:55 +0100)]
tmpfiles: never delete AF_UNIX sockets that are alive

13 years agogetty: don't parse console= anymore, use /sys/class/tty/console/active instead
Lennart Poettering [Mon, 14 Feb 2011 18:57:18 +0000 (19:57 +0100)]
getty: don't parse console= anymore, use /sys/class/tty/console/active instead

13 years agoutil: properly resolve /dev/console if more than once console= argument was passed...
Lennart Poettering [Mon, 14 Feb 2011 18:54:15 +0000 (19:54 +0100)]
util: properly resolve /dev/console if more than once console= argument was passed on the kernel command line

13 years agogetty: do not auto-spawn getty's on VC ttys if console=ttyN is used
Lennart Poettering [Mon, 14 Feb 2011 17:56:51 +0000 (18:56 +0100)]
getty: do not auto-spawn getty's on VC ttys if console=ttyN is used

https://bugzilla.redhat.com/show_bug.cgi?id=676885

13 years agofsck: skip root fsck if dracut already did it
Lennart Poettering [Mon, 14 Feb 2011 10:42:47 +0000 (11:42 +0100)]
fsck: skip root fsck if dracut already did it

13 years agoupdate TODO
Lennart Poettering [Sun, 13 Feb 2011 18:02:18 +0000 (19:02 +0100)]
update TODO

13 years agoutil: when determining the right TERM for /dev/console consult /sys/class/tty/console...
Lennart Poettering [Sun, 13 Feb 2011 18:01:47 +0000 (19:01 +0100)]
util: when determining the right TERM for /dev/console consult /sys/class/tty/console/active

13 years agoservice: when guessing the main PID don't consider processes that aren't our children
Lennart Poettering [Sun, 13 Feb 2011 17:52:02 +0000 (18:52 +0100)]
service: when guessing the main PID don't consider processes that aren't our children

13 years agoservice: make main pid guessing optional, and reread pid file after reloads
Lennart Poettering [Sun, 13 Feb 2011 17:51:30 +0000 (18:51 +0100)]
service: make main pid guessing optional, and reread pid file after reloads

13 years agopam: introduce whitelist and blacklist user list feature
Lennart Poettering [Sun, 13 Feb 2011 17:21:11 +0000 (18:21 +0100)]
pam: introduce whitelist and blacklist user list feature

This is useful to exclude root from the session logout killings or to
limit killing to the selinux guest users.

13 years agosystemctl: minor optimizations
Lennart Poettering [Sun, 13 Feb 2011 16:18:45 +0000 (17:18 +0100)]
systemctl: minor optimizations

13 years agosystemctl: don't unnecessarily close stdin/stdout/stderr for tty agent so that lockin...
Lennart Poettering [Sun, 13 Feb 2011 16:09:29 +0000 (17:09 +0100)]
systemctl: don't unnecessarily close stdin/stdout/stderr for tty agent so that locking by tty works

13 years agoreadahead: disable readahead in virtual machines
Lennart Poettering [Sun, 13 Feb 2011 14:46:05 +0000 (15:46 +0100)]
readahead: disable readahead in virtual machines

13 years agotmpfiles: simplify default tmpfiles configuration by using globs
Lennart Poettering [Sun, 13 Feb 2011 14:11:28 +0000 (15:11 +0100)]
tmpfiles: simplify default tmpfiles configuration by using globs

13 years agosystemd: document systemd-tmpfiles
Lennart Poettering [Sun, 13 Feb 2011 14:08:15 +0000 (15:08 +0100)]
systemd: document systemd-tmpfiles

13 years agotmpfiles: move binary to /bin to make it publicly available
Lennart Poettering [Sun, 13 Feb 2011 13:04:57 +0000 (14:04 +0100)]
tmpfiles: move binary to /bin to make it publicly available

13 years agotmpfiles: take names of tmpfiles configuration files on the command line
Lennart Poettering [Sun, 13 Feb 2011 13:00:54 +0000 (14:00 +0100)]
tmpfiles: take names of tmpfiles configuration files on the command line

13 years agotmpfiles: log to stderr if possible
Lennart Poettering [Sat, 12 Feb 2011 08:31:38 +0000 (09:31 +0100)]
tmpfiles: log to stderr if possible

13 years agotmpfiles: support globs
Lennart Poettering [Sat, 12 Feb 2011 08:31:25 +0000 (09:31 +0100)]
tmpfiles: support globs

13 years agoupdate TODO
Lennart Poettering [Wed, 9 Feb 2011 21:48:54 +0000 (22:48 +0100)]
update TODO

13 years agounits: introduce and hook up sound.target
Lennart Poettering [Wed, 9 Feb 2011 20:07:28 +0000 (21:07 +0100)]
units: introduce and hook up sound.target

13 years agodbus: allow all clients access to org.freedesktop.DBus.Peer
Lennart Poettering [Wed, 9 Feb 2011 14:21:20 +0000 (15:21 +0100)]
dbus: allow all clients access to org.freedesktop.DBus.Peer

13 years agoman: document SYSTEMD_READY=
Lennart Poettering [Wed, 9 Feb 2011 14:02:03 +0000 (15:02 +0100)]
man: document SYSTEMD_READY=

13 years agodevice: consider udev devices with SYSTEMD_READY=0 as unplugged
Lennart Poettering [Wed, 9 Feb 2011 13:48:52 +0000 (14:48 +0100)]
device: consider udev devices with SYSTEMD_READY=0 as unplugged

13 years agosystemd: show status when run with plymouth, and don't start plymouth a second time...
Lennart Poettering [Wed, 9 Feb 2011 13:17:52 +0000 (14:17 +0100)]
systemd: show status when run with plymouth, and don't start plymouth a second time when it is already running

13 years agomanager: if we receive SIGRTMIN+20/21 enable/disable showing of status on the console
Lennart Poettering [Wed, 9 Feb 2011 11:12:30 +0000 (12:12 +0100)]
manager: if we receive SIGRTMIN+20/21 enable/disable showing of status on the console

13 years agoman: clarify a few things
Lennart Poettering [Wed, 9 Feb 2011 10:00:17 +0000 (11:00 +0100)]
man: clarify a few things

13 years agosystemctl: fix opening params
Lennart Poettering [Wed, 9 Feb 2011 00:11:19 +0000 (01:11 +0100)]
systemctl: fix opening params

13 years agosystemctl: don't start agent for --user
Lennart Poettering [Wed, 9 Feb 2011 00:09:09 +0000 (01:09 +0100)]
systemctl: don't start agent for --user

13 years agosystemctl: make sure the tty agent does not retain a copy of stdio
Lennart Poettering [Tue, 8 Feb 2011 23:58:43 +0000 (00:58 +0100)]
systemctl: make sure the tty agent does not retain a copy of stdio

https://bugzilla.redhat.com/show_bug.cgi?id=674916

13 years agoplymouth: move plymouth out of TARGET_FEDORA
Jeff Mahoney [Tue, 8 Feb 2011 17:47:32 +0000 (18:47 +0100)]
plymouth: move plymouth out of TARGET_FEDORA

13 years agolocale: unify some code between fedora and altlinux
Lennart Poettering [Tue, 8 Feb 2011 11:54:06 +0000 (12:54 +0100)]
locale: unify some code between fedora and altlinux

13 years agobuild-sys: initial support ALTLinux
Alexey Shabalin [Tue, 8 Feb 2011 11:50:56 +0000 (12:50 +0100)]
build-sys: initial support ALTLinux

Hi all!
I added basic support for ALTLinux.
Please see patch or
http://git.altlinux.org/people/shaba/packages/systemd.git?p=systemd.git;a=shortlog;h=refs/heads/altlinux
Thanks.

13 years agobuild-sys: make less noise when checking file
Lucas De Marchi [Wed, 26 Jan 2011 20:10:18 +0000 (18:10 -0200)]
build-sys: make less noise when checking file

There's already a condition that gives a warning if user is
cross-compiling an relying on files from the build system. AC_CHECK_FILE
macro only tests for file in the build system. Hence use a simple
'test -f' that does the same job, without the warning.

13 years agoupdate TODO
Lennart Poettering [Tue, 8 Feb 2011 11:12:23 +0000 (12:12 +0100)]
update TODO

13 years agobuild-sys: fix AC_COMPILE_IFELSE tests
Lucas De Marchi [Wed, 26 Jan 2011 19:59:11 +0000 (17:59 -0200)]
build-sys: fix AC_COMPILE_IFELSE tests

This removes some warnings when bootstrapping. The warnings were introduced
in autoconf 2.68. Reference:

http://www.flameeyes.eu/autotools-mythbuster/forwardporting/autoconf.html

The following code

AC_COMPILE_IFELSE([
<MY C CODE HERE>
], ...)

is changed to

AC_COMPILE_IFELSE(
[AC_COMPILE_IFELSE([AC_LANG_SOURCE([
<MY C CODE HERE>
])], ...)

13 years agobuild-sys: ensure selinux configure check follows logic of other optional features
Lennart Poettering [Tue, 8 Feb 2011 10:44:38 +0000 (11:44 +0100)]
build-sys: ensure selinux configure check follows logic of other optional features

13 years agobuild-sys: autodetect and use pkg-config for libselinux
Alexey Shabalin [Thu, 27 Jan 2011 12:44:39 +0000 (15:44 +0300)]
build-sys: autodetect and use pkg-config for libselinux

Hi!
libselinux has a pkg-config file:

$ rpm -qf /usr/lib64/pkgconfig/libselinux.pc
libselinux-devel-2.0.98-alt1

$ cat /usr/lib64/pkgconfig/libselinux.pc
prefix=/usr
exec_prefix=${prefix}
libdir=${exec_prefix}/lib64
includedir=/usr/include

Name: libselinux
Description: SELinux utility library
Version: 2.0.98
URL: http://userspace.selinuxproject.org/
Libs.private: -lsepol
Libs: -lselinux
Cflags:

Patch for use pkg-config in attach.

--
Alexey Shabalin

13 years agodbus: use ControlGroup as property name to match config option
Lennart Poettering [Tue, 8 Feb 2011 10:13:32 +0000 (11:13 +0100)]
dbus: use ControlGroup as property name to match config option

13 years agopam: optionally reset cgroup memberships for login sessions
Lennart Poettering [Fri, 4 Feb 2011 11:46:38 +0000 (12:46 +0100)]
pam: optionally reset cgroup memberships for login sessions

13 years agoload-fragment: properly parse Nice= value
Adrian Spinu [Wed, 2 Feb 2011 13:57:52 +0000 (15:57 +0200)]
load-fragment: properly parse Nice= value

Hello,

     i`ve been using systemd for a while now, and found out that when
using NICE parameter for .service files the varible is not set correctly.
     i`ve found the problem in file *load-fragment.c* function
*config_parse_nice*        variable /*c->nice_set = false;*/  should be
/*c->nice_set = true;*/
     Problem is alsom manifesting on v17 but did not upgrade yet ...

13 years agoautomount: use unit_pending_inactive() where appropriate
Lennart Poettering [Wed, 26 Jan 2011 01:55:35 +0000 (02:55 +0100)]
automount: use unit_pending_inactive() where appropriate

13 years agoTODO
Lennart Poettering [Tue, 25 Jan 2011 03:35:35 +0000 (04:35 +0100)]
TODO

13 years agoclang: fix some issues found with clang-analyzer v17
Lennart Poettering [Sat, 22 Jan 2011 01:18:59 +0000 (02:18 +0100)]
clang: fix some issues found with clang-analyzer

13 years agogcc: make gcc shut up
Lennart Poettering [Sat, 22 Jan 2011 00:47:37 +0000 (01:47 +0100)]
gcc: make gcc shut up

13 years agobuild-sys: prep v17
Lennart Poettering [Sat, 22 Jan 2011 00:44:35 +0000 (01:44 +0100)]
build-sys: prep v17

13 years agoupdate TODO
Lennart Poettering [Sat, 22 Jan 2011 00:26:17 +0000 (01:26 +0100)]
update TODO

13 years agovala 0.10 seem to work fine
Kay Sievers [Fri, 21 Jan 2011 17:22:54 +0000 (18:22 +0100)]
vala 0.10 seem to work fine

13 years agocryptsetup: fix ordering loop when dealing with encrypted swap devices
Lennart Poettering [Fri, 21 Jan 2011 02:30:13 +0000 (03:30 +0100)]
cryptsetup: fix ordering loop when dealing with encrypted swap devices

https://bugzilla.redhat.com/show_bug.cgi?id=657234

13 years agomain: don't warn if /etc/mtab is a symlink to /proc/mounts
Lennart Poettering [Fri, 21 Jan 2011 01:57:58 +0000 (02:57 +0100)]
main: don't warn if /etc/mtab is a symlink to /proc/mounts

13 years agosocket: don't crash if the .service unit for a .socket unit is not found
Lennart Poettering [Fri, 21 Jan 2011 01:49:12 +0000 (02:49 +0100)]
socket: don't crash if the .service unit for a .socket unit is not found

13 years agoupdate TODO: drop issues that cannot be reproduce or have been fixed otherwise
Lennart Poettering [Fri, 21 Jan 2011 00:54:12 +0000 (01:54 +0100)]
update TODO: drop issues that cannot be reproduce or have been fixed otherwise

13 years agomount: ignore if an fsck is requested for a bind mount, so that we don't wait for...
Lennart Poettering [Fri, 21 Jan 2011 00:28:30 +0000 (01:28 +0100)]
mount: ignore if an fsck is requested for a bind mount, so that we don't wait for the bind 'device' to show up

13 years agoautomount: fix segfault when shutting down
Lennart Poettering [Fri, 21 Jan 2011 00:13:09 +0000 (01:13 +0100)]
automount: fix segfault when shutting down

https://bugs.freedesktop.org/show_bug.cgi?id=32076

13 years agoman: give an example for vconsole.conf
Lennart Poettering [Thu, 20 Jan 2011 23:42:47 +0000 (00:42 +0100)]
man: give an example for vconsole.conf

https://bugs.freedesktop.org/show_bug.cgi?id=31955

13 years agodbus: don't try to connect to the system bus before it is actually up
Lennart Poettering [Thu, 20 Jan 2011 18:43:38 +0000 (19:43 +0100)]
dbus: don't try to connect to the system bus before it is actually up

13 years agoservice: make chain of main commands and control commands independent of each other...
Lennart Poettering [Thu, 20 Jan 2011 17:46:38 +0000 (18:46 +0100)]
service: make chain of main commands and control commands independent of each other, so that both can be executed simultaneously and independently

13 years agoservice: don't allow reload operations for oneshot services
Lennart Poettering [Thu, 20 Jan 2011 17:44:06 +0000 (18:44 +0100)]
service: don't allow reload operations for oneshot services

During starting of oneshot we are in START state which we don't want to
combine with the RELOAD state, hence we disallow both at the same time.

13 years agovala: convert from dbus-glib to gdbus
Lennart Poettering [Thu, 20 Jan 2011 17:24:26 +0000 (18:24 +0100)]
vala: convert from dbus-glib to gdbus

13 years agosystemctl: highlight failed processes in systemctl status
Lennart Poettering [Thu, 20 Jan 2011 17:22:03 +0000 (18:22 +0100)]
systemctl: highlight failed processes in systemctl status

13 years agosystemctl: show process type along process exit codes
Lennart Poettering [Thu, 20 Jan 2011 12:24:24 +0000 (13:24 +0100)]
systemctl: show process type along process exit codes

13 years agoservice: when reloading a service fails don't fail the entire service but just the...
Lennart Poettering [Thu, 20 Jan 2011 12:17:22 +0000 (13:17 +0100)]
service: when reloading a service fails don't fail the entire service but just the reload job

13 years agoshutdown: use correct kexec options
Brett Witherspoon [Sat, 15 Jan 2011 03:50:26 +0000 (21:50 -0600)]
shutdown: use correct kexec options

The kexec man page is awful and misleading.

13 years agoserial-getty: invoking /sbin/securetty is no longer necessary since pam_securetty...
Lennart Poettering [Tue, 18 Jan 2011 22:51:06 +0000 (23:51 +0100)]
serial-getty: invoking /sbin/securetty is no longer necessary since pam_securetty can now look for console= in /proc/cmdline

13 years agosystemctl: before spawning pager cache number of columns
Lennart Poettering [Tue, 18 Jan 2011 22:12:49 +0000 (23:12 +0100)]
systemctl: before spawning pager cache number of columns

13 years agopam: optionally keep processes of root user around
Lennart Poettering [Tue, 18 Jan 2011 22:07:06 +0000 (23:07 +0100)]
pam: optionally keep processes of root user around

13 years agoservice: if a reload operation fails, don't shut down the service
Lennart Poettering [Tue, 18 Jan 2011 21:57:46 +0000 (22:57 +0100)]
service: if a reload operation fails, don't shut down the service

13 years agoexecute: make sending of SIGKILL on shutdown optional
Lennart Poettering [Tue, 18 Jan 2011 21:55:54 +0000 (22:55 +0100)]
execute: make sending of SIGKILL on shutdown optional

13 years agomount: FAT uuids are uppercase
Lennart Poettering [Tue, 18 Jan 2011 21:32:15 +0000 (22:32 +0100)]
mount: FAT uuids are uppercase

13 years agoman: document missing KillSignal= and swap options
Lennart Poettering [Mon, 17 Jan 2011 23:40:10 +0000 (00:40 +0100)]
man: document missing KillSignal= and swap options

13 years agoupdate TODO
Lennart Poettering [Mon, 17 Jan 2011 21:41:53 +0000 (22:41 +0100)]
update TODO

13 years agoarch: remove rc-local.service
Tom Gundersen [Sun, 16 Jan 2011 00:46:34 +0000 (01:46 +0100)]
arch: remove rc-local.service

This is now shipped downstream.

13 years agoarch: remove support for legacy daemons
Tom Gundersen [Sun, 16 Jan 2011 00:46:33 +0000 (01:46 +0100)]
arch: remove support for legacy daemons

This is now done by a generator, shipped downstream.

13 years agobuild-sys: prepare v16 v16
Lennart Poettering [Sat, 8 Jan 2011 01:34:55 +0000 (02:34 +0100)]
build-sys: prepare v16

13 years agodevice: don't warn if we cannot bump the socket buffer
Lennart Poettering [Sat, 8 Jan 2011 01:30:07 +0000 (02:30 +0100)]
device: don't warn if we cannot bump the socket buffer

13 years agoupdate TODO
Lennart Poettering [Fri, 7 Jan 2011 00:12:27 +0000 (01:12 +0100)]
update TODO

13 years agologger: when passing on PID info, fall back to our own if originating process is...
Lennart Poettering [Thu, 6 Jan 2011 23:18:51 +0000 (00:18 +0100)]
logger: when passing on PID info, fall back to our own if originating process is already gone

13 years agoservice: don't hit an assert if information in LSB headers is incorrectly formatted
Lennart Poettering [Thu, 6 Jan 2011 22:52:58 +0000 (23:52 +0100)]
service: don't hit an assert if information in LSB headers is incorrectly formatted

https://bugzilla.redhat.com/show_bug.cgi?id=667665

13 years agoexecute,util: fix two small memory leaks
Lennart Poettering [Thu, 6 Jan 2011 22:52:17 +0000 (23:52 +0100)]
execute,util: fix two small memory leaks

13 years agospecifier: at minimal test
Lennart Poettering [Thu, 6 Jan 2011 22:51:52 +0000 (23:51 +0100)]
specifier: at minimal test

13 years agogetty: unset locale before execution
Lennart Poettering [Thu, 6 Jan 2011 19:38:06 +0000 (20:38 +0100)]
getty: unset locale before execution

On the console indian characters cannot be displayed, hence it is
advisable to disable indian locales on the console, which most
distributions traditionally did from a shell fragment executed post
login. If getty gets started with locale settings passed it would itself
however be translated without the no-indian-on-console fixup applied.

Hence, for now don't pass any locale settings to getty/login, and thus
rely on the classic post-login script fragment to set and fix the
locale.

Eventually we probably want to drop this again since the system locale
should be read and set at one place, and not at multiple, and that one
place should be PID 1.

https://bugzilla.redhat.com/show_bug.cgi?id=663900

13 years agoexecute: drop empty assignments from env blocks on execution but keep them around...
Lennart Poettering [Thu, 6 Jan 2011 19:38:02 +0000 (20:38 +0100)]
execute: drop empty assignments from env blocks on execution but keep them around otherwise to make them visible

13 years agoumount: don't try to detach the dm device the root dir is on, to minimize warning...
Lennart Poettering [Thu, 6 Jan 2011 18:10:17 +0000 (19:10 +0100)]
umount: don't try to detach the dm device the root dir is on, to minimize warning messages

13 years agoupdate TODO
Lennart Poettering [Thu, 6 Jan 2011 17:54:10 +0000 (18:54 +0100)]
update TODO

13 years agolocale: fix variable names
Lennart Poettering [Thu, 6 Jan 2011 00:39:25 +0000 (01:39 +0100)]
locale: fix variable names