Commit Graph
20913 Commits
Author SHA1 Message Date
Lennart Poettering 34f0c1a118 man: some clarifications 2010-09-13 12:05:54 +02:00
Michal Schmidt 54ec68b6f6 manager: report correctly which job was deleted when breaking cycle 2010-09-11 02:00:27 +02:00
Lennart Poettering ee0dd8021b service: don't automatically add D-Bus service name as an alias to all dbus service, rely on manual symlinks to allow disabling of dbus services 2010-09-11 01:31:58 +02:00
Lennart Poettering 88ac58d76b units: pull in syslog.target for synchronization purposes by systemd-logger.service 2010-09-10 18:20:41 +02:00
Lennart Poettering 5a5582b5e5 units: start logger only after syslog is up 2010-09-10 17:47:36 +02:00
Lennart Poettering 1ff5e88f3b units: we should be able to reach syslog.target even if no logger is enabled at all 2010-09-10 17:17:57 +02:00
Lennart Poettering 1fd116f748 units: set working directory to /root for sulogin, since sulogin is crazy 2010-09-10 17:11:49 +02:00
Lennart Poettering b00bad3641 logger: support SOCK_STREAM /dev/log sockets as necessary for syslog-ng 2010-09-10 17:11:48 +02:00
Lennart Poettering 739848d4f3 units: set HOME=/root when running shells 2010-09-10 17:11:48 +02:00
Kay Sievers f27dcdcaa0 update fixme 2010-09-10 15:11:51 +02:00
Kay Sievers 2f4a4e1e54 update fixme 2010-09-10 15:10:39 +02:00
Gustavo Sverzut Barbieri 0d37b36b28 util: add Gentoo welcome message. 2010-09-09 00:17:34 +02:00
Lennart Poettering c31e14954b log: support logging into SOCK_STREAM /dev/log sockets for comaptibility with syslog-ng 2010-09-08 03:07:44 +02:00
Michael Biebl e082d6c8d8 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>
2010-09-08 02:14:58 +02:00
Michael Biebl a5dc0a45fe 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.
2010-09-08 02:12:38 +02:00
Lennart Poettering da5fcb1060 sd-daemon: optionally, avoid hiding the ELF symbols 2010-09-08 01:55:35 +02:00
Lennart Poettering ed370f5daa main: when we encounter 'emergency' on the kernel cmdline start the emergency shell 2010-09-08 01:55:35 +02:00
David Zeuthen 645d40d004 Bump required GLib version to 2.22
This is needed for g_main_context_get_thread_default().

Signed-off-by: David Zeuthen <davidz@redhat.com>
2010-09-07 11:11:11 -04:00
David Zeuthen cbdf255e25 gudev: Deliver ::uevent signal in the thread-default main loop
... that the GUdevClient object was constructed in. This change makes
GUdev follow the GLib guidelines and, more importantly, makes it
possible to actually use the library in a multi-threaded
application. Prior to this patch, signals were emitted in the thread
that ran the "default" main loop.

Signed-off-by: David Zeuthen <davidz@redhat.com>
2010-09-07 11:04:57 -04:00
Kay Sievers cff2f9a3c8 scsi_id: fix compiler warnings 2010-09-07 08:24:12 +02:00
Hannes Reinecke db11b20e41 scsi_id: export target port group
For ALUA support it's useful to have the target port group number
of a device.

Signed-off-by: Hannes Reinecke <hare@suse.de>
2010-09-07 08:22:05 +02:00
Lennart Poettering d3f929f0fc units: make sure we always fully write the utmp data 2010-09-06 15:01:57 +02:00
Michael Biebl 96d4ce0110 man: Fix small typo: s/seperate/separate/ 2010-09-06 03:13:14 +02:00
Michael Biebl f10814d9e0 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.
2010-09-06 03:12:44 +02:00
Michael Biebl ad206afcb9 units: Add halt, killall, poweroff and reboot unit files for Debian 2010-09-06 03:12:28 +02:00
Lennart Poettering 7d568925bd systemctl: make --version a little bit more verbose 2010-09-06 03:11:24 +02:00
Lennart Poettering 35df8f27fa systemctl: implement --version
https://bugs.freedesktop.org/show_bug.cgi?id=29912
2010-09-06 02:42:49 +02:00
Lennart Poettering 2a8cd298f4 dbus: downgrade activation failure messages 2010-09-06 02:42:42 +02:00
Lennart Poettering 7573916f63 automount: make sure we don't crash if there's an automount unit without mount unit 2010-09-06 01:05:38 +02:00
Kay Sievers 397e0169bb version bump 2010-09-03 15:14:36 +02:00
Kay Sievers e7638c038b release 162 2010-09-03 15:07:03 +02:00
Kay Sievers f74310b4a0 Revert "Add alternative KVM MAC address blacklist"
This reverts commit 634afac119.

54:52:00 was just a bug in libvirt, and that's better fixed locally,
and we should not carry it in udev rules.
2010-09-03 14:08:14 +02:00
Kay Sievers a04879b564 systemd: enable all udev services unconditionally 2010-09-03 13:30:57 +02:00
Kay Sievers dacdf15367 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.
2010-09-03 12:26:30 +02:00
Michal Soltys 8937f371fb ChangeLog fix
In v141 -> v142 entry, there's a note about udevd creating
/dev/{null,kmsg,console}. It was added in commit 540f46698d,
but shortly after that removed in a00bdfa16b before v142
release.

Signed-off-by: Michal Soltys <soltys@ziu.info>
2010-09-03 10:53:26 +02:00
Lennart Poettering 53eff0c762 build-sys: fix fedpkg git path 2010-09-03 05:39:22 +02:00
Lennart Poettering a9c72fe933 build-sys: prepare new release 2010-09-03 04:51:02 +02:00
Matthew Miller 25f63d4671 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.
2010-09-03 04:48:00 +02:00
Lennart Poettering f73d93a4d9 service: calculate sysv startup priority only in start runlevels 2010-09-03 04:08:16 +02:00
Lennart Poettering 31091693ec update fixme 2010-09-03 03:58:11 +02:00
Lennart Poettering c68364b790 service: make sure we properly figure out whether a sysv service is enabled before creating $foo wants links 2010-09-03 03:56:05 +02:00
Lennart Poettering 8f78c3aafb units: don't add shutdown conflicts dep to umount.target 2010-09-03 02:20:39 +02:00
Lennart Poettering 5ed9f5d65e unit: don't retroactively start units if we failed to shutdown a unit 2010-09-03 01:45:53 +02:00
Lennart Poettering 53c6a358a8 dbus: don't send cgroup agent messages directly to system bus to avoid dbus activation loop 2010-09-03 01:45:53 +02:00
Lennart Poettering b401e1fb9d target: implicitly shutdown targets on shutdown, so that they are pulled into the final transaction and may be used for synchronization 2010-09-03 01:42:57 +02:00
Lennart Poettering 7f97f0fee4 service: properly handle if the main process is down and we wait for the cgroup to die, too 2010-09-03 01:42:57 +02:00
Lennart Poettering 5d452f9cec dbus: make sure to flush bus connection before disconnecting, so that no messages are lost 2010-09-02 23:26:04 +02:00
Martin Pitt 634afac119 Add alternative KVM MAC address blacklist
Reportedly, older KVM/Qemu instances indeed do use 54:52:00:*,
so add this as an alternative.
2010-09-02 11:11:32 +02:00
Michael Forney 620343ddd6 Don't install systemd scripts with --without-systemdsystemunitdir
Currently, the scripts get installed to /no/ if that option is
specified.

Signed-off-by: Martin Pitt <martin.pitt@ubuntu.com>
2010-09-02 08:54:37 +02:00
Martin Pitt d4de0a0321 do not create persistent name rules for VMWare network interfaces
Not generating persistent MAC address rules will significantly ease cloning of
VMs.  The kernel reliably sorts eth* enumeration by bus number, so as long as
you only have cards from one vendor (or more precisely, drivers), the
enumeration will be stable. Having cards from different vendors is very
unlikely in VMs.

KVM was already covered in the previous commit, this is the equivalent
blacklist for VMWare:

  http://www.coffer.com/mac_find/?string=005056
  http://www.coffer.com/mac_find/?string=000c29

https://launchpad.net/bugs/341006
2010-09-02 08:39:12 +02:00