chiark / gitweb /
elogind.git
10 years agoconf-parse: rename config_parse_level() to config_parse_log_level()
Lennart Poettering [Mon, 3 Mar 2014 20:14:07 +0000 (21:14 +0100)]
conf-parse: rename config_parse_level() to config_parse_log_level()

"level" is a bit too generic, let's clarify what kind of level we are
referring to here.

10 years agoupdate TODO
Lennart Poettering [Mon, 3 Mar 2014 20:08:49 +0000 (21:08 +0100)]
update TODO

10 years agologind: ignore lid switch events for 30s after each suspend and 3min after startup
Lennart Poettering [Mon, 3 Mar 2014 19:49:33 +0000 (20:49 +0100)]
logind: ignore lid switch events for 30s after each suspend and 3min after startup

This is needed to give USB docking stations and suchlike time to settle,
so that a display connected to an USB docking station can actually act
as a lid swith inhibitor correctly.

With this change we should have somewhat reliable docking station
support in place.

10 years agologind: fix printf format
Lennart Poettering [Mon, 3 Mar 2014 18:39:51 +0000 (19:39 +0100)]
logind: fix printf format

10 years agologind: ignore lid switch if more than 1 display is connected
Lennart Poettering [Mon, 3 Mar 2014 18:30:16 +0000 (19:30 +0100)]
logind: ignore lid switch if more than 1 display is connected

Previously we expected the desktop environment to take an inhibitor
lock, but this opened a race on boot-up where logind might already be
running but no DE is active.

Hence, let's move checking for additional displays into logind. This
also opens up this logic for other DEs, given that only GNOME
implemented the inhibitor logic so far.

10 years agoman: networkd - mention resolv.conf symlink
Tom Gundersen [Mon, 3 Mar 2014 17:25:37 +0000 (18:25 +0100)]
man: networkd - mention resolv.conf symlink

10 years agocore: introduce new RuntimeDirectory= and RuntimeDirectoryMode= unit settings
Lennart Poettering [Mon, 3 Mar 2014 16:14:07 +0000 (17:14 +0100)]
core: introduce new RuntimeDirectory= and RuntimeDirectoryMode= unit settings

As discussed on the ML these are useful to manage runtime directories
below /run for services.

10 years agoexecute: no need to include seccomp.h from execute.h
Lennart Poettering [Mon, 3 Mar 2014 16:12:56 +0000 (17:12 +0100)]
execute: no need to include seccomp.h from execute.h

10 years agoexecute: free directory path if we fail to remove it because we cannot allocate a...
Lennart Poettering [Mon, 3 Mar 2014 16:11:39 +0000 (17:11 +0100)]
execute: free directory path if we fail to remove it because we cannot allocate a thread

10 years agobuild-sys: prefer using ln --relative -s where appropriate
Lennart Poettering [Mon, 3 Mar 2014 16:10:16 +0000 (17:10 +0100)]
build-sys: prefer using ln --relative -s where appropriate

By using --relative symlinks look nicer when dealing with OS image trees
that are placed in arbitrary places of the OS.

10 years agosd-dhcp-lease: add sd_dhcp_lease_get_next_server()
Tom Gundersen [Mon, 3 Mar 2014 16:13:59 +0000 (17:13 +0100)]
sd-dhcp-lease: add sd_dhcp_lease_get_next_server()

10 years agosd-dhcp-lease: add sd_dhcp_lease_get_server_identifier()
Tom Gundersen [Mon, 3 Mar 2014 15:46:10 +0000 (16:46 +0100)]
sd-dhcp-lease: add sd_dhcp_lease_get_server_identifier()

10 years agosd-dhcp-lease: add Root Path support
Tom Gundersen [Mon, 3 Mar 2014 14:43:02 +0000 (15:43 +0100)]
sd-dhcp-lease: add Root Path support

This is necessary when mounting /dev/nfs based on a DHCP lease.

10 years agoFix systemd-stdio-bridge symlink
Mike Gilbert [Mon, 3 Mar 2014 04:37:39 +0000 (23:37 -0500)]
Fix systemd-stdio-bridge symlink

The symlink is created in bindir (/usr/bin), and points to a binary
which lives in rootlibexecdir (/lib/systemd or /usr/lib/systemd). A
relative symlink does not work here.

10 years agobus: add sd_bus_track object for tracking peers, and port core over to it
Lennart Poettering [Mon, 3 Mar 2014 00:33:45 +0000 (01:33 +0100)]
bus: add sd_bus_track object for tracking peers, and port core over to it

This is primarily useful for services that need to track clients which
reference certain objects they maintain, or which explicitly want to
subscribe to certain events. Something like this is done in a large
number of services, and not trivial to do. Hence, let's unify this at
one place.

This also ports over PID 1 to use this to ensure that subscriptions to
job and manager events are correctly tracked. As a side-effect this
makes sure we properly serialize and restore the track list across
daemon reexec/reload, which didn't work correctly before.

This also simplifies how we distribute messages to broadcast to the
direct busses: we only track subscriptions for the API bus and
implicitly assume that all direct busses are subscribed. This should be
a pretty OK simplification since clients connected via direct bus
connections are shortlived anyway.

10 years agoadd bash completion for systemd-cgtop
Thomas Hindoe Paaboel Andersen [Sun, 2 Mar 2014 21:58:18 +0000 (22:58 +0100)]
add bash completion for systemd-cgtop

10 years agoDisallow sizes with increasing unit size
Zbigniew Jędrzejewski-Szmek [Sun, 2 Mar 2014 18:28:05 +0000 (13:28 -0500)]
Disallow sizes with increasing unit size

Things like 3B4T, 4B50B, 400 100 (meaning 4*1024**4+3, 54, and 500,
respectively) are now disallowed. It is necessary to say 4T3B, 54B,
500 instead. I think this was confusing and error prone.

As a special form, 400B 100 is allowed, i.e. "B" suffix is treated
as different from "", although they mean the same thing.

10 years agoAllow fractional parts in disk sizes
Zbigniew Jędrzejewski-Szmek [Sun, 2 Mar 2014 05:05:16 +0000 (00:05 -0500)]
Allow fractional parts in disk sizes

It seems natural to be able to say SystemMaxUsage=1.5G.

https://bugzilla.redhat.com/show_bug.cgi?id=1047568

10 years agohwdb: Update database of Bluetooth company identifiers
Marcel Holtmann [Sun, 2 Mar 2014 18:02:10 +0000 (10:02 -0800)]
hwdb: Update database of Bluetooth company identifiers

10 years agobuild-sys: fix location of network-util.h
Zbigniew Jędrzejewski-Szmek [Sun, 2 Mar 2014 02:45:27 +0000 (21:45 -0500)]
build-sys: fix location of network-util.h

10 years agonetworkd-wait-online: use automatic cleanup
Thomas Hindoe Paaboel Andersen [Sat, 1 Mar 2014 21:52:25 +0000 (22:52 +0100)]
networkd-wait-online: use automatic cleanup

10 years agoadd bash completion for systemd-cgls
Thomas Hindoe Paaboel Andersen [Sat, 1 Mar 2014 22:08:38 +0000 (23:08 +0100)]
add bash completion for systemd-cgls

10 years agoTODO
Tom Gundersen [Sat, 1 Mar 2014 13:13:18 +0000 (14:13 +0100)]
TODO

10 years agobuild-sys: distribute kdbus documentation
Zbigniew Jędrzejewski-Szmek [Sat, 1 Mar 2014 12:44:47 +0000 (07:44 -0500)]
build-sys: distribute kdbus documentation

10 years agoUpdate gitignore
Zbigniew Jędrzejewski-Szmek [Sat, 1 Mar 2014 12:39:14 +0000 (07:39 -0500)]
Update gitignore

10 years agoupdate TODO
Thomas Hindoe Paaboel Andersen [Sat, 1 Mar 2014 03:03:52 +0000 (04:03 +0100)]
update TODO

10 years agobash: add completion for systemd-nspawn
Thomas Andersen [Sat, 1 Mar 2014 02:27:49 +0000 (03:27 +0100)]
bash: add completion for systemd-nspawn

10 years agonspawn: fix detection of missing /proc/self/loginuid
Tero Roponen [Tue, 25 Feb 2014 15:19:35 +0000 (17:19 +0200)]
nspawn: fix detection of missing /proc/self/loginuid

Running 'systemd-nspawn -D /srv/Fedora/' gave me this error:
 Failed to read /proc/self/loginuid: No such file or directory

 Container Fedora failed with error code 1.

This patch fixes the problem.

10 years agosd-network: turn states 'unknown' and 'unmanaged' into errnos
Tom Gundersen [Fri, 28 Feb 2014 01:07:29 +0000 (02:07 +0100)]
sd-network: turn states 'unknown' and 'unmanaged' into errnos

10 years agosd-network: fix header
Tom Gundersen [Fri, 28 Feb 2014 00:49:41 +0000 (01:49 +0100)]
sd-network: fix header

10 years agoarchitecture: Add tilegx
Henrik Grindal Bakken [Thu, 27 Feb 2014 20:19:13 +0000 (21:19 +0100)]
architecture: Add tilegx

Add Tilera's TILE-GX processor family support.

10 years agonetworkd: add networkd-wait-online
Tom Gundersen [Fri, 28 Feb 2014 00:02:01 +0000 (01:02 +0100)]
networkd: add networkd-wait-online

This is mostly a proof of concept to try sd-network, so we don't
hook it up with a .service file quite yet. We probably want it to
be more clever about deciding when we are 'online'.

The binary  will wait for at least one network managed by networkd,
and until all networks managed by networkd are configured.

10 years agosd-network: add new library
Tom Gundersen [Thu, 27 Feb 2014 00:24:05 +0000 (01:24 +0100)]
sd-network: add new library

This is similar to sd-login, but exposes the state of networkd rather than logind.

Include it in libsystemd-dhcp and rename it to libsystemd-network.

10 years agoudev/rules: setup tty permissions and group for sclp_line, ttysclp and 3270/tty
Lukas Nykryn [Thu, 27 Feb 2014 10:19:09 +0000 (11:19 +0100)]
udev/rules: setup tty permissions and group for sclp_line, ttysclp and 3270/tty

10 years agocdrom_id: use the old MMC fallback
Lukas Nykryn [Thu, 27 Feb 2014 10:06:37 +0000 (11:06 +0100)]
cdrom_id: use the old MMC fallback

https://bugzilla.redhat.com/show_bug.cgi?id=1038015
The problem seems to be that the your virtual DVD is emulating a really
old DVD device, and doing it kind of strangely.

> dracut:# /lib/udev/cdrom_id --debug /dev/sr0
> probing: '/dev/sr0'
> INQUIRY: [IMM     ][Virtual CD/DVD   ][0316]
> GET CONFIGURATION failed with SK=5h/ASC=24h/ACQ=00h

So your virtual drive rejects the GET CONFIGURATION command as illegal.

Other pre-MMC2 drives that don't accept this command usually return the
error
SK=5h,ASC=20h (invalid/unsupported command code), in which case cdrom_id
tries an older method, and then ID_CDROM_MEDIA_TRACK_COUNT_DATA gets set
and all the /dev/disk/by-label (etc) links get set up.

The virtual drive returns the error SK=5h,ASC=24h (invalid field in
Command Descriptor Block), which cdrom_id doesn't handle, so it gives up
and the links never get made.

The ideal solution would be to make the IMM to emulate a device that's
less than 15 years old, but I'm not going to hold my breath waiting for
that.

So probably cdrom_id should also use the old MMC fallback when the error
is SK=5h,ASC=24h, and then all of this would work as expected.

Suggested-by:Luca Miccini <lmiccini@redhat.com>

10 years agoautogen: Allow running autogen from a builddir != srcdir
Stef Walter [Tue, 25 Feb 2014 08:45:46 +0000 (09:45 +0100)]
autogen: Allow running autogen from a builddir != srcdir

While the configure correctly supports being run outside of
the top source directory by default, it's nice if autogen.sh
does too:

 $ git clone git://anongit.freedesktop.org/systemd/systemd
 $ cd systemd
 $ mkdir $(uname -m)
 $ cd $(uname -m)
 $ ../autogen.sh a
 $ make

10 years agobuild-sys: Do not distribute generated udev service files
Armin K [Wed, 26 Feb 2014 14:02:55 +0000 (15:02 +0100)]
build-sys: Do not distribute generated udev service files

They are already in nodist_systemunit_DATA and if they are
shipped, they contain hardcoded paths to udevadm and
systemd-udevd which will cause them to fail to start when
rootprefix != prefix and rootlibdir != libdir.

10 years agohwdb: keymaps for Samsung 900X3E/900X3F
Kieran Clancy [Thu, 27 Feb 2014 05:32:57 +0000 (16:02 +1030)]
hwdb: keymaps for Samsung 900X3E/900X3F

Keymaps for Samsung 900X3E and 900X3F should be the same as for other
models in the series.

I have also moved the comment for '# Series 9' to the right place

10 years agojournalctl: refuse extra arguments with --verify and similar
Zbigniew Jędrzejewski-Szmek [Thu, 27 Feb 2014 04:01:43 +0000 (23:01 -0500)]
journalctl: refuse extra arguments with --verify and similar

Positional arguments only make sense with the default action.
For other actions, complain instead of ignoring them silently.

10 years agologs-show: fix corrupt output with empty messages
Uoti Urpala [Thu, 20 Feb 2014 01:00:09 +0000 (03:00 +0200)]
logs-show: fix corrupt output with empty messages

If a message had zero length, journalctl would print no newline, and
two output lines would be concatenated. Fix. The problem was
introduced in commit 31f7bf199452 ("logs-show: print multiline
messages"). Affected short and verbose output modes.

Before fix:

Feb 09 21:16:17 glyph dhclient[1323]: Feb 09 21:16:17 glyph NetworkManager[788]: <info> (enp4s2): DHCPv4 state changed nbi -> preinit

after:

Feb 09 21:16:17 glyph dhclient[1323]:
Feb 09 21:16:17 glyph NetworkManager[788]: <info> (enp4s2): DHCPv4 state changed nbi -> preinit

10 years agobuild-sys: networkd - simplify makefile fragment
Tom Gundersen [Wed, 26 Feb 2014 22:16:34 +0000 (23:16 +0100)]
build-sys: networkd - simplify makefile fragment

10 years agomount: don't fire PropertiesChanged signals for mounts that are stopped
Lennart Poettering [Wed, 26 Feb 2014 03:28:37 +0000 (04:28 +0100)]
mount: don't fire PropertiesChanged signals for mounts that are stopped

10 years agomount: don't send out PropertiesChanged message if actually nothing got changed
Lennart Poettering [Wed, 26 Feb 2014 03:27:50 +0000 (04:27 +0100)]
mount: don't send out PropertiesChanged message if actually nothing got changed

10 years agomount: minor modernization
Lennart Poettering [Wed, 26 Feb 2014 02:46:04 +0000 (03:46 +0100)]
mount: minor modernization

10 years agoUse /var/run/dbus/system_bus_socket for the D-Bus socket
Zbigniew Jędrzejewski-Szmek [Wed, 26 Feb 2014 02:26:31 +0000 (21:26 -0500)]
Use /var/run/dbus/system_bus_socket for the D-Bus socket

10 years agoREADME: document that /var/run must be a symlink → /run
Lennart Poettering [Wed, 26 Feb 2014 01:54:37 +0000 (02:54 +0100)]
README: document that /var/run must be a symlink → /run

10 years agoRevert back to /var/run at a couple of problems
Lennart Poettering [Wed, 26 Feb 2014 01:47:43 +0000 (02:47 +0100)]
Revert back to /var/run at a couple of problems

This partially reverts 41a55c46ab8fb4ef6727434227071321fc762cce

Some specifications we want to stay compatibility actually document
/var/run, not /run, and we should stay compatible with that. In order to
make sure our D-Bus implementation works on any system, regardless if
running systemd or not, we should always use /var/run which is the
only path mandated by the D-Bus spec.

Similar, glibc hardcodes the utmp location to /var/run, and this is
exposed in _UTMP_PATH in limits.h, hence let's stay in sync with this
public API, too.

We simply do not support systems where /var/run is not a symlink → /run.
Hence both are equivalent. Staying compatible with upstream
specifications hence weighs more than cleaning up superficial
appearance.

10 years agoReplace /var/run with /run in remaining places
Zbigniew Jędrzejewski-Szmek [Wed, 26 Feb 2014 01:11:04 +0000 (20:11 -0500)]
Replace /var/run with /run in remaining places

/run was already used almost everywhere, fix the remaining places
for consistency.

10 years agoexec: imply NoNewPriviliges= only when seccomp filters are used in user mode
Lennart Poettering [Wed, 26 Feb 2014 01:28:52 +0000 (02:28 +0100)]
exec: imply NoNewPriviliges= only when seccomp filters are used in user mode

10 years agoupdate TODO
Lennart Poettering [Wed, 26 Feb 2014 01:19:17 +0000 (02:19 +0100)]
update TODO

10 years agobuild-sys: add missing makefile symlinks
Lennart Poettering [Wed, 26 Feb 2014 01:16:46 +0000 (02:16 +0100)]
build-sys: add missing makefile symlinks

10 years agocore: add new RestrictAddressFamilies= switch
Lennart Poettering [Tue, 25 Feb 2014 19:37:03 +0000 (20:37 +0100)]
core: add new RestrictAddressFamilies= switch

This new unit settings allows restricting which address families are
available to processes. This is an effective way to minimize the attack
surface of services, by turning off entire network stacks for them.

This is based on seccomp, and does not work on x86-32, since seccomp
cannot filter socketcall() syscalls on that platform.

10 years agonspawn: no need for duplicate checks against EEXIST
Lennart Poettering [Tue, 25 Feb 2014 19:33:17 +0000 (20:33 +0100)]
nspawn: no need for duplicate checks against EEXIST

10 years agoseccomp: we should control NO_NEW_PRIVS on our own, not let seccomp do this for us
Lennart Poettering [Tue, 25 Feb 2014 19:32:27 +0000 (20:32 +0100)]
seccomp: we should control NO_NEW_PRIVS on our own, not let seccomp do this for us

10 years agonetworkd: add basic support for MACVLANs
Tom Gundersen [Tue, 25 Feb 2014 20:16:17 +0000 (21:16 +0100)]
networkd: add basic support for MACVLANs

10 years agoman: refer to systemd.net{work,dev}(5) from systemd-networkd(8)
Tom Gundersen [Tue, 25 Feb 2014 18:30:40 +0000 (19:30 +0100)]
man: refer to systemd.net{work,dev}(5) from systemd-networkd(8)

10 years agoman: split out systemd.net{work,dev}(5) from systemd-networkd(8)
Tom Gundersen [Tue, 25 Feb 2014 18:30:40 +0000 (19:30 +0100)]
man: split out systemd.net{work,dev}(5) from systemd-networkd(8)

10 years agoman: split out systemd.link(5) from udev(7)
Tom Gundersen [Tue, 25 Feb 2014 15:56:15 +0000 (16:56 +0100)]
man: split out systemd.link(5) from udev(7)

10 years agoman: networkd - clarify that multiple addresses/routes are supported
Tom Gundersen [Tue, 25 Feb 2014 15:06:45 +0000 (16:06 +0100)]
man: networkd - clarify that multiple addresses/routes are supported

10 years agofix typo in iDRAC network interface name: irdac->idrac
Tomasz Torcz [Tue, 25 Feb 2014 11:43:55 +0000 (12:43 +0100)]
fix typo in iDRAC network interface name: irdac->idrac

10 years agologin: Allow calling org.freedesktop.login1.Seat.SwitchTo
Jasper St. Pierre [Fri, 21 Feb 2014 23:23:17 +0000 (18:23 -0500)]
login: Allow calling org.freedesktop.login1.Seat.SwitchTo

10 years agologin: set pos-slot to fallback on pos-eviction
David Herrmann [Tue, 25 Feb 2014 12:08:24 +0000 (13:08 +0100)]
login: set pos-slot to fallback on pos-eviction

If we evict a session position, we open the position slot for future
sessions. However, there might already be another session on the same
position if both were started on the same VT. This is currently done if
gdm spawns the session on its own Xserver.

Hence, look for such a session on pos-eviction and claim the new slot
immediately.

10 years agoNEWS: fix typo
Ansgar Burchardt [Tue, 25 Feb 2014 09:53:49 +0000 (10:53 +0100)]
NEWS: fix typo

10 years agologin: fix pos-array allocation
David Herrmann [Tue, 25 Feb 2014 11:20:25 +0000 (12:20 +0100)]
login: fix pos-array allocation

GREEDY_REALLOC takes a pointer to the real size, not the array-width as
argument. Therefore, our array is currently way to small to keep the seat
positions.

Introduce GREEDY_REALLOC0_T() as typed version of GREEDY_REALLOC and store
the array-width instead of array-size.

10 years agoupdate TODO
Lennart Poettering [Tue, 25 Feb 2014 01:37:34 +0000 (02:37 +0100)]
update TODO

10 years agoNEWS: fix typo
Lennart Poettering [Tue, 25 Feb 2014 01:36:12 +0000 (02:36 +0100)]
NEWS: fix typo

10 years agonspawn: add new switch --network-macvlan= to add a macvlan device to the container
Lennart Poettering [Tue, 25 Feb 2014 01:27:39 +0000 (02:27 +0100)]
nspawn: add new switch --network-macvlan= to add a macvlan device to the container

10 years agoRemove dead lines in various places
Zbigniew Jędrzejewski-Szmek [Sun, 23 Feb 2014 00:26:27 +0000 (19:26 -0500)]
Remove dead lines in various places

As pointed-out by clang -Wunreachable-code.

No behaviour changes.

10 years agohostnamectl: read virtualization/architecture from remote side
Lennart Poettering [Tue, 25 Feb 2014 00:03:25 +0000 (01:03 +0100)]
hostnamectl: read virtualization/architecture from remote side

This way we make this more network/container transparent and also do not
require any client side privileges.

10 years agocore: expose architecture as a bus property, so that we can easily query it remotely
Lennart Poettering [Mon, 24 Feb 2014 23:50:38 +0000 (00:50 +0100)]
core: expose architecture as a bus property, so that we can easily query it remotely

10 years agocore: add global settings for enabling CPUAccounting=, MemoryAccounting=, BlockIOAcco...
Lennart Poettering [Mon, 24 Feb 2014 22:50:10 +0000 (23:50 +0100)]
core: add global settings for enabling CPUAccounting=, MemoryAccounting=, BlockIOAccounting= for all units at once

10 years agoupdate TODO v210
Lennart Poettering [Mon, 24 Feb 2014 20:23:04 +0000 (21:23 +0100)]
update TODO

10 years agoNEWS: fix typos
Kay Sievers [Mon, 24 Feb 2014 19:20:25 +0000 (20:20 +0100)]
NEWS: fix typos

10 years agobuild-sys: bump revisions and version
Lennart Poettering [Mon, 24 Feb 2014 18:25:00 +0000 (19:25 +0100)]
build-sys: bump revisions and version

10 years agoarchitecture: apparently there is LE ppc now
Lennart Poettering [Mon, 24 Feb 2014 18:21:18 +0000 (19:21 +0100)]
architecture: apparently there is LE ppc now

10 years agoNEWS: add note about ifunc
Lennart Poettering [Mon, 24 Feb 2014 18:13:39 +0000 (19:13 +0100)]
NEWS: add note about ifunc

10 years agoNEWS: prepare for release of 210
Lennart Poettering [Mon, 24 Feb 2014 17:34:12 +0000 (18:34 +0100)]
NEWS: prepare for release of 210

10 years agosmack: rework security labeling for multiple frameworks
Łukasz Stelmach [Wed, 19 Feb 2014 15:56:34 +0000 (16:56 +0100)]
smack: rework security labeling for multiple frameworks

10 years agosmack: set loaded_policy in smack_setup()
Łukasz Stelmach [Thu, 19 Dec 2013 14:15:54 +0000 (15:15 +0100)]
smack: set loaded_policy in smack_setup()

With loaded_policy set to true mount_setup() relabels /dev properly.

10 years agosmack: relabel directories and files created by systemd
Łukasz Stelmach [Wed, 19 Feb 2014 15:39:04 +0000 (16:39 +0100)]
smack: relabel directories and files created by systemd

Systemd creates directories in /dev. These directories will
get the label of systemd, which is the label of the System
domain, which is not accessable to everyone. Relabel the
directories, files and symlinks created so that they can be
generally used.

Based on a patch by Casey Schaufler <casey@schaufler-ca.com>.

10 years agoupdate TODO
Lennart Poettering [Mon, 24 Feb 2014 16:34:37 +0000 (17:34 +0100)]
update TODO

10 years agohostnamed: correct error message
Michal Sekletar [Mon, 24 Feb 2014 14:59:29 +0000 (15:59 +0100)]
hostnamed: correct error message

We are not parsing timezone data.

10 years agologind: detect whether the system is docked, and if it is inhibit lid switch processing
Lennart Poettering [Mon, 24 Feb 2014 15:22:23 +0000 (16:22 +0100)]
logind: detect whether the system is docked, and if it is inhibit lid switch processing

This should make operation nicer with docking stations, but will not
cover anything that does not implement SW_DOCK.

10 years agoupdate TODO
Lennart Poettering [Mon, 24 Feb 2014 14:29:36 +0000 (15:29 +0100)]
update TODO

10 years agounits/serial-getty@.service: add [Install] section
Zbigniew Jędrzejewski-Szmek [Mon, 24 Feb 2014 13:02:28 +0000 (08:02 -0500)]
units/serial-getty@.service: add [Install] section

This makes it easier to manually enable and disable
specific gettys, and also mirrors getty@.service.

http://lists.freedesktop.org/archives/systemd-devel/2014-February/017329.html

10 years agovirt: make Virtualization an anonymous enum
Thomas Hindoe Paaboel Andersen [Sun, 23 Feb 2014 21:05:18 +0000 (22:05 +0100)]
virt: make Virtualization an anonymous enum

This makes llvm happy when we assign an error code to the variable.

10 years agoupdate TODO
Lennart Poettering [Mon, 24 Feb 2014 12:21:55 +0000 (13:21 +0100)]
update TODO

10 years agoconfigure: Do not require xsltproc for installation of man pages
Mike Gilbert [Sun, 23 Feb 2014 16:21:13 +0000 (11:21 -0500)]
configure: Do not require xsltproc for installation of man pages

The release tarballs ship with pre-generated man pages, so we do not
need xsltproc for a typical end-user build.

Developers will probably have xsltproc anyway, but if not they will now
encounter a build-time failure instead of an error in configure.

10 years agosystemctl: move next elapse calculation to its own function
Djalal Harouni [Mon, 24 Feb 2014 00:42:24 +0000 (01:42 +0100)]
systemctl: move next elapse calculation to its own function

10 years agodbus-timer: fix bus_timer_vtable to have the correct times
Djalal Harouni [Mon, 24 Feb 2014 00:42:23 +0000 (01:42 +0100)]
dbus-timer: fix bus_timer_vtable to have the correct times

next_elapse_monotonic() should map to the "NextElapseUSecMonotonic"
property and next_elapse_realtime() to "NextElapseUSecRealtime" one.

This makes "systemctl list-timers" compute and show the correct times.

https://bugs.freedesktop.org/show_bug.cgi?id=75272

10 years agoupdate TODO
Lennart Poettering [Mon, 24 Feb 2014 03:04:23 +0000 (04:04 +0100)]
update TODO

10 years agogetty-generator: verify ttys before we make use of them
Lennart Poettering [Mon, 24 Feb 2014 02:59:50 +0000 (03:59 +0100)]
getty-generator: verify ttys before we make use of them

The ttyS[0-3] devices are weird. They may be enumerated, but when one
actually tries to open and use them they return EIO, because they don't
actually exist. Because they may be enumerated they may be specified on
the kernel command line as console=. And some people do that as default.
As response to that we'll spawn a getty on the tty that will quickly
fail, and we retry a couple of time before giving up. That is quite
noisy.

With this new change we will validate all serial terminals configured
with console= on the kernel cmdline before adding gettys on them, and
remove the invalid ones. THis should remove the noise later on.

This should make Eric Paris happy!

10 years agocgroup: certain cgroup attributes are not available in the root cgroup, hence don...
Lennart Poettering [Mon, 24 Feb 2014 02:38:24 +0000 (03:38 +0100)]
cgroup: certain cgroup attributes are not available in the root cgroup, hence don't bother

10 years agocore: gc half-created stub units
Lennart Poettering [Mon, 24 Feb 2014 02:20:54 +0000 (03:20 +0100)]
core: gc half-created stub units

10 years agocore: expose root control group on the bus
Lennart Poettering [Mon, 24 Feb 2014 02:20:24 +0000 (03:20 +0100)]
core: expose root control group on the bus

We expose the control group of the units on the bus, so let's also
expose the root control group.

10 years agoupdate TODO
Lennart Poettering [Mon, 24 Feb 2014 02:03:05 +0000 (03:03 +0100)]
update TODO

10 years agonspawn: make use of the devices cgroup controller by default
Lennart Poettering [Mon, 24 Feb 2014 02:00:38 +0000 (03:00 +0100)]
nspawn: make use of the devices cgroup controller by default

10 years agocgroup: parse array cgroup properties correctly when they aren't at the end of the...
Lennart Poettering [Mon, 24 Feb 2014 01:59:57 +0000 (02:59 +0100)]
cgroup: parse array cgroup properties correctly when they aren't at the end of the message

10 years agosd-dhcp-client: respect TP_STATUS_CSUMNOTREADY
Tom Gundersen [Mon, 24 Feb 2014 00:09:21 +0000 (01:09 +0100)]
sd-dhcp-client: respect TP_STATUS_CSUMNOTREADY

If an UDP packet has not passed through a hardware device, its checksum may not
have been computed. This is exposed through the TP_STATUS_CSUMNOTREADY sockopt.

When using raw sockets, skip checksum validation when TP_STATUS_CSUMNOTREADY
is set.

This is necessary for dhcp to work directly over a veth tunnel, e.g. as done
in systemd-nspawn.

10 years agoupdate TODO
Lennart Poettering [Mon, 24 Feb 2014 01:31:01 +0000 (02:31 +0100)]
update TODO