]>
err.no Git - systemd/log
Lennart Poettering [Sun, 21 Nov 2010 18:29:27 +0000 (19:29 +0100)]
man: document locale.conf
Lennart Poettering [Sun, 21 Nov 2010 18:12:44 +0000 (19:12 +0100)]
man: document /etc/vconsole.conf
Lennart Poettering [Sun, 21 Nov 2010 15:29:56 +0000 (16:29 +0100)]
man: document /etc/hostname
Lennart Poettering [Fri, 19 Nov 2010 15:24:21 +0000 (16:24 +0100)]
units: move a couple of units from base.target to sysinit.target
Lennart Poettering [Fri, 19 Nov 2010 15:14:52 +0000 (16:14 +0100)]
man: reorder things to follow the same order everywhere
Lennart Poettering [Fri, 19 Nov 2010 03:22:36 +0000 (04:22 +0100)]
Revert "Revert "fsck: add new -l switch to fsck mount options""
This reverts commit
c8bc83f5eac49f083ae53d965577118c1b5c1116 .
As Karel pointed out the -l switch doesn't actualy conflict with
per-filesystem options, since those can be seperated by -- in the
command line. Since -l is here to stay we hence enable it again for now.
Lennart Poettering [Thu, 18 Nov 2010 22:56:34 +0000 (23:56 +0100)]
build-sys: bump for v13
Lennart Poettering [Thu, 18 Nov 2010 22:40:27 +0000 (23:40 +0100)]
cryptsetup: actually show disk name
Lennart Poettering [Thu, 18 Nov 2010 22:34:42 +0000 (23:34 +0100)]
cryptsetup: show udev device name when asking for password
Lennart Poettering [Thu, 18 Nov 2010 22:03:42 +0000 (23:03 +0100)]
sysctl: don't consider missing kernel options fatal
Lennart Poettering [Thu, 18 Nov 2010 20:52:26 +0000 (21:52 +0100)]
sysctl: implement native sysctl tool to support Debian-style /etc/sysctl.d
Lennart Poettering [Thu, 18 Nov 2010 18:23:17 +0000 (19:23 +0100)]
mount: trivial optimization
Lennart Poettering [Thu, 18 Nov 2010 04:22:25 +0000 (05:22 +0100)]
units: enable console ask-password agent by default
Lennart Poettering [Thu, 18 Nov 2010 03:46:28 +0000 (04:46 +0100)]
util: introduce /etc/os-release so that distros don't need to patch systemd to get a welcome message
Lennart Poettering [Thu, 18 Nov 2010 03:00:38 +0000 (04:00 +0100)]
update TODO
Lennart Poettering [Thu, 18 Nov 2010 02:52:39 +0000 (03:52 +0100)]
job: make sure we don't fail umount.target if a mount unit failed to stop
Lennart Poettering [Thu, 18 Nov 2010 01:45:04 +0000 (02:45 +0100)]
units: make sure prefdm is killed first, user sessions second on shutdown
Lennart Poettering [Thu, 18 Nov 2010 01:37:35 +0000 (02:37 +0100)]
cgroup: after killing cgroup processes, ensure the group is really dead gone. wait for 3s at max
Lennart Poettering [Thu, 18 Nov 2010 01:36:51 +0000 (02:36 +0100)]
cgroup: if we couldn't remove a cgroup after killing evertyhing in it then it's fine
Lennart Poettering [Thu, 18 Nov 2010 01:16:36 +0000 (02:16 +0100)]
cryptsetup: automatically order crypt partitions before cryptsetup.target
Andrey Borzenkov [Wed, 17 Nov 2010 17:38:29 +0000 (20:38 +0300)]
man: trivial BindTo description fix
I guess "unexpected" implies "without involvement".
Lennart Poettering [Wed, 17 Nov 2010 23:42:35 +0000 (00:42 +0100)]
manager: make list of default controllers configurable
Lennart Poettering [Wed, 17 Nov 2010 23:39:09 +0000 (00:39 +0100)]
build: expose libcryptsetup dependency in build string
Lennart Poettering [Wed, 17 Nov 2010 23:38:41 +0000 (00:38 +0100)]
pam: document controllers= switch
Lennart Poettering [Wed, 17 Nov 2010 20:27:53 +0000 (21:27 +0100)]
cgroup: by default, duplicate service cgroup in the cpu hierarchy
Lennart Poettering [Wed, 17 Nov 2010 19:22:07 +0000 (20:22 +0100)]
pam: duplicate cgroup tree in the cpu hierarchy by default, optionally more
Kay Sievers [Wed, 17 Nov 2010 12:21:13 +0000 (13:21 +0100)]
update TODO
Kay Sievers [Wed, 17 Nov 2010 12:19:10 +0000 (13:19 +0100)]
Revert "update TODO"
This reverts commit
afe635e71838c9e62e25a928cbcefc343744dda5 .
Oops, libnotify should not have been in here.
Kay Sievers [Wed, 17 Nov 2010 10:58:42 +0000 (11:58 +0100)]
update TODO
Lennart Poettering [Tue, 16 Nov 2010 23:41:14 +0000 (00:41 +0100)]
build-sys: prep release 12
Lennart Poettering [Tue, 16 Nov 2010 23:40:26 +0000 (00:40 +0100)]
update TODO
Lennart Poettering [Tue, 16 Nov 2010 03:28:04 +0000 (04:28 +0100)]
ask-password: add --console mode to ask questions on /dev/console
Lennart Poettering [Tue, 16 Nov 2010 02:47:14 +0000 (03:47 +0100)]
cryptsetup: reword questions a little
Lennart Poettering [Tue, 16 Nov 2010 02:32:54 +0000 (03:32 +0100)]
units: order hwclock after readahead
Lennart Poettering [Tue, 16 Nov 2010 02:27:00 +0000 (03:27 +0100)]
path: don't mention too many inotify msgs
Lennart Poettering [Tue, 16 Nov 2010 02:24:17 +0000 (03:24 +0100)]
cryptsetup: include device name in password question
Lennart Poettering [Tue, 16 Nov 2010 02:23:52 +0000 (03:23 +0100)]
cryptsetup: lock ourselves into memory as long as we deal with passwords
Lennart Poettering [Tue, 16 Nov 2010 02:12:54 +0000 (03:12 +0100)]
plymouth: use updated socket name
Lennart Poettering [Tue, 16 Nov 2010 00:54:07 +0000 (01:54 +0100)]
units: set TERM for gettys again, since they acquire a TTY on their own
Lennart Poettering [Tue, 16 Nov 2010 00:51:57 +0000 (01:51 +0100)]
units: allow start-up of plymouth ask-password agent very early
Lennart Poettering [Tue, 16 Nov 2010 00:25:48 +0000 (01:25 +0100)]
units: enable ask-paswword .path units early enough to be useful for early mounts
Lennart Poettering [Tue, 16 Nov 2010 00:25:19 +0000 (01:25 +0100)]
units: delay getty until logins are allowed
Lennart Poettering [Mon, 15 Nov 2010 23:10:57 +0000 (00:10 +0100)]
pam: always rely on loginuid instead of uid to determine cgroup and XDG_RUNTIME_DIR
Lennart Poettering [Mon, 15 Nov 2010 22:55:53 +0000 (23:55 +0100)]
cgroup: call root cgroup system instead of systemd-1
Lennart Poettering [Mon, 15 Nov 2010 22:49:02 +0000 (23:49 +0100)]
exec: automatically determine right TERM= setting based on tty name
Lennart Poettering [Mon, 15 Nov 2010 22:22:05 +0000 (23:22 +0100)]
pam: rename master user cgroup to 'master'
Lennart Poettering [Mon, 15 Nov 2010 21:12:41 +0000 (22:12 +0100)]
drop support for MANAGER_SESSION, introduce MANAGER_USER instead
Lennart Poettering [Mon, 15 Nov 2010 20:30:52 +0000 (21:30 +0100)]
path: properly free PathSpec structs
Lennart Poettering [Mon, 15 Nov 2010 20:09:18 +0000 (21:09 +0100)]
manager: ignore invalid epoll events
Lennart Poettering [Mon, 15 Nov 2010 19:10:04 +0000 (20:10 +0100)]
units: use ConditionDirectoryNotEmpty= where applicable
Lennart Poettering [Mon, 15 Nov 2010 19:06:49 +0000 (20:06 +0100)]
unit: introduce ConditionDirectoryNotEmpty=
Matthias Clasen [Mon, 15 Nov 2010 18:54:57 +0000 (19:54 +0100)]
build-sys: fix building against libnotify 0.7
Here are two patches I needed to get systemd to build against the
current libnotify and vala releases.
Kay Sievers [Mon, 15 Nov 2010 18:47:25 +0000 (19:47 +0100)]
delete tmp.mount which may conflict with an unrelated fstab entry
Lennart Poettering [Mon, 15 Nov 2010 17:42:34 +0000 (18:42 +0100)]
build-sys: make check for libnotify explicit
Gustavo Sverzut Barbieri [Sun, 14 Nov 2010 10:04:16 +0000 (08:04 -0200)]
build-sys: dbus-glib is just required by gtk's systemadm
do not fail if there is no dbus-glib and we're not building the gtk
frontend.
Lennart Poettering [Mon, 15 Nov 2010 17:26:37 +0000 (18:26 +0100)]
units: get rid of no_requires unit flag, and make crypto disks require cryptsetup service
Lennart Poettering [Mon, 15 Nov 2010 03:24:04 +0000 (04:24 +0100)]
load-dropin: add support for .requires directories
Lennart Poettering [Mon, 15 Nov 2010 03:11:42 +0000 (04:11 +0100)]
cryptsetup: bind cryptsetup service to both source and destination device
Lennart Poettering [Mon, 15 Nov 2010 02:44:11 +0000 (03:44 +0100)]
manager: consider jobs already installed as redundant when reducing new transactions
Lennart Poettering [Sun, 14 Nov 2010 23:49:21 +0000 (00:49 +0100)]
path: avoid immediate restarting of units triggered by paths if nothing actually changed on disk
Lennart Poettering [Sun, 14 Nov 2010 23:47:29 +0000 (00:47 +0100)]
path: always look for IN_ATTRIB since deletion is signalled that way
Lennart Poettering [Sun, 14 Nov 2010 22:47:53 +0000 (23:47 +0100)]
manager: always pull 'following' units into transaction
Lennart Poettering [Sun, 14 Nov 2010 22:27:27 +0000 (23:27 +0100)]
util: always highlight distro name
Lennart Poettering [Sun, 14 Nov 2010 22:26:53 +0000 (23:26 +0100)]
units: make use of agetty mandatory
Lennart Poettering [Sun, 14 Nov 2010 21:09:57 +0000 (22:09 +0100)]
manager: don't fail transaction if adding CONFLICTED_BY job fails
Lennart Poettering [Sun, 14 Nov 2010 20:59:25 +0000 (21:59 +0100)]
pam: rename 'no-session' to 'user' cgroup
Lennart Poettering [Sun, 14 Nov 2010 20:57:10 +0000 (21:57 +0100)]
job: make it possible to wait for devices to be unplugged
Michael Biebl [Sun, 14 Nov 2010 21:14:39 +0000 (22:14 +0100)]
Remove umount*.service files for Debian
With the introduction of native shutdown/reboot, the killall.service was
removed (as this functionality was moved into systemd-shutdown).
Without killall.service though, the umount*.service files no longer work
correctly.
Wit native mount support those files are also no longer necessary, so
remove them.
Lennart Poettering [Sun, 14 Nov 2010 19:12:51 +0000 (20:12 +0100)]
tmpfiles: ignore files marked with the sticky bit
Lennart Poettering [Sun, 14 Nov 2010 18:58:33 +0000 (19:58 +0100)]
load-fragment: properly do comparison of words
Lennart Poettering [Sun, 14 Nov 2010 18:24:41 +0000 (19:24 +0100)]
udev: ignore temporary udev devices
Lennart Poettering [Sun, 14 Nov 2010 01:08:31 +0000 (02:08 +0100)]
cryptsetup: handle password=none properly
Lennart Poettering [Sun, 14 Nov 2010 01:01:29 +0000 (02:01 +0100)]
cryptsetup: properly parse cipher= switch
Lennart Poettering [Sun, 14 Nov 2010 00:53:46 +0000 (01:53 +0100)]
cryptsetup: support non-LUKS crypto partitions
Kay Sievers [Fri, 12 Nov 2010 18:04:07 +0000 (19:04 +0100)]
update TODO
Lennart Poettering [Fri, 12 Nov 2010 02:56:43 +0000 (03:56 +0100)]
update fixme
Lennart Poettering [Fri, 12 Nov 2010 02:42:13 +0000 (03:42 +0100)]
ask-password: leave printing of final prompt colon to message
Lennart Poettering [Fri, 12 Nov 2010 02:33:08 +0000 (03:33 +0100)]
ask-password: don't show wall message on ttys we are already running a tty agent on
Lennart Poettering [Fri, 12 Nov 2010 02:05:20 +0000 (03:05 +0100)]
ask-password: refer to right binary name in wall message
Lennart Poettering [Fri, 12 Nov 2010 02:05:03 +0000 (03:05 +0100)]
ask-password: ignore unknown query file fields
Lennart Poettering [Fri, 12 Nov 2010 02:04:47 +0000 (03:04 +0100)]
ask-password: properly NULL terminate table
Lennart Poettering [Fri, 12 Nov 2010 02:04:29 +0000 (03:04 +0100)]
manager: be a bit more verbose if we receive unknown epoll event
Lennart Poettering [Fri, 12 Nov 2010 02:04:10 +0000 (03:04 +0100)]
cryptsetup: automatically start cryptsetup when looking for mount source
Lennart Poettering [Fri, 12 Nov 2010 00:02:03 +0000 (01:02 +0100)]
log: fix build
Lennart Poettering [Fri, 12 Nov 2010 00:01:04 +0000 (01:01 +0100)]
log: add automatic log target
Lennart Poettering [Thu, 11 Nov 2010 23:39:17 +0000 (00:39 +0100)]
cryptsetup: hook up tool with ask-password
Lennart Poettering [Thu, 11 Nov 2010 20:28:33 +0000 (21:28 +0100)]
manager: hookup generators
Lennart Poettering [Thu, 11 Nov 2010 02:04:45 +0000 (03:04 +0100)]
build-sys: fix usage of path macros
Lennart Poettering [Thu, 11 Nov 2010 00:22:42 +0000 (01:22 +0100)]
mount: don't relabel /dev twice in a row
Kay Sievers [Thu, 11 Nov 2010 15:55:28 +0000 (16:55 +0100)]
update TODO
Andrew Edmunds [Thu, 11 Nov 2010 13:24:47 +0000 (14:24 +0100)]
man: Fix various typos
Kay Sievers [Thu, 11 Nov 2010 10:15:16 +0000 (11:15 +0100)]
split mount_point_is_api() and mount_point_ignore()
We should not handle the ignore list as API mounts, as
systemd itself never touches them.
On Thu, Nov 11, 2010 at 10:34, Andreas Jaeger <aj@novell.com> wrote:
>
> I noticed for some time that systemd-remount-api-vfs is in the
> failed state and found now the following in the log files
>
> systemd-remount-api-vfs[467]: /bin/mount for /proc/bus/usb exited with exit status 32.
> systemd-remount-api-vfs[467]: mount: /proc/bus/usb not mounted already, or bad option
> systemd[1]: systemd-remount-api-vfs.service: main process exited, code=exited, status=1
> systemd[1]: Unit systemd-remount-api-vfs.service entered failed state.
Lennart Poettering [Wed, 10 Nov 2010 23:30:00 +0000 (00:30 +0100)]
units: deal properly with the fact that agetty is called getty on Debian
Lennart Poettering [Wed, 10 Nov 2010 22:58:01 +0000 (23:58 +0100)]
tmpfiles: include reference to man page in tmpfiles files
Andrey Borzenkov [Sat, 6 Nov 2010 16:11:03 +0000 (19:11 +0300)]
vconsole: support additional keymap for toggling layouts
At least Mandriva offers configuring characters to toggle
keyboard layout independently from main keymap. This functions
much the same as XkbOptions for X11 and actually is configured
together. The patch adds support for additional keymap,
KEYMAP_TOGGLE, to /etc/vconsole.conf, that is intended to be
used for the same purpose.
Lennart Poettering [Wed, 10 Nov 2010 22:18:34 +0000 (23:18 +0100)]
main: warn if /etc/mtab is not a symlink
Ran Benita [Mon, 8 Nov 2010 23:03:27 +0000 (01:03 +0200)]
add bash completion for systemctl --system
I've been playing recently with systemd on Arch, and had much fun. But
soon, alas, my fingers started to ache from repeatedly writing
systemctl restart some-long-service.service. So, I wrote a completion
script. I figured other people may want to use it, so I prepared a
patch against systemd-git (attached).
There are some notes/disclaimers, however:
- It requires bash>=4.0, sed, grep and awk. A bash-completion package
is not strictly needed; sourcing the file is enough.
- It wouldn't work properly with --session, as I had no way to test it.
- It uses the output of systemctl list-units directly when that's
enough, but also runs systemctl show when completing on some verbs
(for example, to check for AllowIsolate=yes). This /may/ be somewhat
slow once there are many units, since it calls a dbus method on each
one. Is there a faster way to have that information?
- The code is perhaps a bit long and messy; honestly, I blame the tool ;)
One way to improve on the situation is to integrate some completion
code in systemctl itself, the way e.g. gdbus, gsettings and django do
it. This will allow for finer grained and faster completions, and it
won't be necessary to keep the verb/option tables in sync with some
other file. But it does mean adding all of this code in C. If this is
acceptable, I'll try to have a go at it.
Finally, a couple of completion tips I run into:
- If you alias systemctl to, say, sctl, you get completions on that
too by running to following command:
complete -F _systemctl sctl
- Add the following line to your .inputrc, to have the completion show
after only a single tab press:
set show-all-if-ambiguous on
It makes the shell quite more pleasant.
Hope it's good enough!
Ran
Lennart Poettering [Wed, 10 Nov 2010 22:01:20 +0000 (23:01 +0100)]
man: minor tmpfiles(5) updates and reindenting
Lennart Poettering [Wed, 10 Nov 2010 21:35:05 +0000 (22:35 +0100)]
main: rename process on startup to 'systemd' to avoid confusion
Lennart Poettering [Wed, 10 Nov 2010 21:28:19 +0000 (22:28 +0100)]
unit: add ConditionNull= condition