chiark / gitweb /
elogind.git
6 years agobus: don't require that if a child object node exists its parent node must too
Lennart Poettering [Wed, 9 Oct 2013 01:27:22 +0000 (03:27 +0200)]
bus: don't require that if a child object node exists its parent node must too

6 years agobus: fix synthetic error messages
Lennart Poettering [Wed, 9 Oct 2013 01:26:34 +0000 (03:26 +0200)]
bus: fix synthetic error messages

6 years agolibsystemd-bus: add lightweight object vtable implementation for exposing objects...
Lennart Poettering [Wed, 9 Oct 2013 00:37:10 +0000 (02:37 +0200)]
libsystemd-bus: add lightweight object vtable implementation for exposing objects on the bus

This adds a lightweight scheme how to define interfaces in static fixed
arrays which then can be easily registered on a bus connection. This
makes it much easier to write bus services.

This automatically handles implementation of the Properties,
ObjectManager, and Introspection bus interfaces.

6 years agocore: unify the way we denote serialization attributes
Lennart Poettering [Tue, 8 Oct 2013 22:13:55 +0000 (00:13 +0200)]
core: unify the way we denote serialization attributes

6 years agosystemd: serialize/deserialize forbid_restart value
Sylvia Else [Mon, 7 Oct 2013 03:06:35 +0000 (23:06 -0400)]
systemd: serialize/deserialize forbid_restart value

The Service type's forbid_restart field was not preserved by
serialization/deserialization, so the fact that the service should not
be restarted after stopping was lost.

If a systemctl stop foo command has been given, but the foo service
has not yet stopped, and then the systemctl --system daemon-reload was
given, then when the foo service eventually stopped, systemd would
restart it.

https://bugs.freedesktop.org/show_bug.cgi?id=69800

6 years agojournalctl: flip to --full by default
Zbigniew Jędrzejewski-Szmek [Mon, 7 Oct 2013 01:55:18 +0000 (21:55 -0400)]
journalctl: flip to --full by default

We already shew lines in full when using a pager or not on a
tty. The commit disables ellipsization in the sole remaining case,
namely when --follow is used.

This has been a popular request for a long time, and indeed, full
output seems much more useful. Old behaviour can still be requested by
using --no-full. Old options retain their behaviour for compatiblity,
but aren't advertised as much. This change applies only to jornalctl,
not to systemctl, when ellipsization is useful to keep the layout.

https://bugzilla.redhat.com/show_bug.cgi?id=984758

6 years agoudev: support custom Linux Security Module labels for device nodes
Kay Sievers [Mon, 7 Oct 2013 23:59:10 +0000 (01:59 +0200)]
udev: support custom Linux Security Module labels for device nodes

6 years agoRun with a custom SMACK domain (label).
Auke Kok [Wed, 25 Sep 2013 22:49:42 +0000 (15:49 -0700)]
Run with a custom SMACK domain (label).

Allows the systemd --system process to change its current
SMACK label to a predefined custom label (usually "system")
at boot time.

This is needed to have a few system-generated folders and
sockets automatically be created with the right SMACK
label. Without that, processes either cannot communicate with
systemd or systemd fails to perform some actions.

6 years agoMount /run, /dev/shm usable to tasks when using SMACK.
Auke Kok [Thu, 26 Sep 2013 21:41:09 +0000 (14:41 -0700)]
Mount /run, /dev/shm usable to tasks when using SMACK.

Once systemd itself is running in a security domain for SMACK,
it will fail to start countless tasks due to missing privileges
for mounted and created directory structures. For /run and shm
specifically, we grant all tasks access.

These 2 mounts are allowed to fail, which will happen if the
system is not running a SMACK enabled kernel or security=none is
passed to the kernel.

6 years agoshared/util: fix off-by-one error in tag_to_udev_node
Dave Reisner [Sun, 6 Oct 2013 22:26:23 +0000 (18:26 -0400)]
shared/util: fix off-by-one error in tag_to_udev_node

Triggered false negatives when encoding a string which needed every
character to be escaped, e.g. "LABEL=/".

6 years agoutf8: fix utf8_is_printable
Zbigniew Jędrzejewski-Szmek [Sun, 6 Oct 2013 02:47:52 +0000 (22:47 -0400)]
utf8: fix utf8_is_printable

6 years agocore: do not add "what" to RequiresMountsFor for network mounts
Zbigniew Jędrzejewski-Szmek [Sat, 5 Oct 2013 17:09:43 +0000 (13:09 -0400)]
core: do not add "what" to RequiresMountsFor for network mounts

For cifs mount like //server/share, we would get
RequiresMountsFor=/server/share, which probably isn't
harmful, but quite confusing.

Unfortunately a bunch of static functions had to be moved
up, but patch is really one line.

6 years agomount: check for NULL before reading pm->what
Dave Reisner [Fri, 4 Oct 2013 22:22:40 +0000 (18:22 -0400)]
mount: check for NULL before reading pm->what

Since a57f7e2c828b85, a mount unit with garbage in it would cause
systemd to crash on loading it.

ref: https://bugs.freedesktop.org/show_bug.cgi?id=70148

6 years agologind: fix bus introspection data for TakeControl()
Lennart Poettering [Fri, 4 Oct 2013 19:16:40 +0000 (21:16 +0200)]
logind: fix bus introspection data for TakeControl()

6 years agomanager: when verifying whether clients may change environment using selinux check...
Lennart Poettering [Fri, 4 Oct 2013 15:01:37 +0000 (17:01 +0200)]
manager: when verifying whether clients may change environment using selinux check for "reload" rather "reboot"

This appears to be a copy/paste error.

6 years agoupdate TODO
Lennart Poettering [Fri, 4 Oct 2013 15:01:32 +0000 (17:01 +0200)]
update TODO

6 years agosystemd: order remote mounts from mountinfo before remote-fs.target
Zbigniew Jędrzejewski-Szmek [Fri, 4 Oct 2013 02:15:08 +0000 (22:15 -0400)]
systemd: order remote mounts from mountinfo before remote-fs.target

Usually the network is stopped before filesystems are umounted.
Ordering network filesystems before remote-fs.target means that their
unmounting will be performed earlier, and can terminate sucessfully.

https://bugs.freedesktop.org/show_bug.cgi?id=70002

6 years agoIntroduce _cleanup_endmntent_
Zbigniew Jędrzejewski-Szmek [Fri, 4 Oct 2013 02:13:55 +0000 (22:13 -0400)]
Introduce _cleanup_endmntent_

6 years agogpt-auto-generator: exit immediately if in container
Zbigniew Jędrzejewski-Szmek [Fri, 4 Oct 2013 02:13:01 +0000 (22:13 -0400)]
gpt-auto-generator: exit immediately if in container

Otherwise we get an ugly warning when running systemd in
a container.

6 years agoexecute: more debugging messages
Zbigniew Jędrzejewski-Szmek [Wed, 28 Aug 2013 12:01:30 +0000 (08:01 -0400)]
execute: more debugging messages

6 years agoman: Improve the description of parameter X in tmpfiles.d page
Václav Pavlín [Thu, 3 Oct 2013 13:47:26 +0000 (15:47 +0200)]
man: Improve the description of parameter X in tmpfiles.d page

6 years agoexecute.c: always set $SHELL
Zbigniew Jędrzejewski-Szmek [Wed, 2 Oct 2013 11:23:10 +0000 (13:23 +0200)]
execute.c: always set $SHELL

In e6dca81 $SHELL was added to user@.service. Let's
instead provide it to all units which have a user.

6 years agotmpfiles.d: include setgid perms for /run/log/journal
Dave Reisner [Wed, 2 Oct 2013 19:35:16 +0000 (15:35 -0400)]
tmpfiles.d: include setgid perms for /run/log/journal

4608af4333d0f7f5 set permissions for journal storage on persistent disk
but not the volatile storage.

ref: https://bugs.archlinux.org/task/37170

6 years agonspawn: always copy /etc/resolv.conf rather than bind mount
Lennart Poettering [Wed, 2 Oct 2013 17:40:43 +0000 (19:40 +0200)]
nspawn: always copy /etc/resolv.conf rather than bind mount

We were already creating the file if it was missing, and this way
containers can reconfigure the file without running into problems.

This also makes resolv.conf handling more alike to handling of
/etc/localtime, which is also not a bind mount.

6 years agomkdir: pass a proper function pointer to mkdir_safe_internal
Lennart Poettering [Wed, 2 Oct 2013 17:38:52 +0000 (19:38 +0200)]
mkdir: pass a proper function pointer to mkdir_safe_internal

6 years agostrv: don't access potentially NULL string arrays
Lennart Poettering [Wed, 2 Oct 2013 17:38:28 +0000 (19:38 +0200)]
strv: don't access potentially NULL string arrays

6 years agoefi: never call qsort on potentially NULL arrays
Lennart Poettering [Wed, 2 Oct 2013 17:38:09 +0000 (19:38 +0200)]
efi: never call qsort on potentially NULL arrays

6 years agomodules-load: fix error handling
Lennart Poettering [Wed, 2 Oct 2013 17:37:44 +0000 (19:37 +0200)]
modules-load: fix error handling

6 years agodbus: fix return value of dispatch_rqueue()
Lennart Poettering [Wed, 2 Oct 2013 17:37:30 +0000 (19:37 +0200)]
dbus: fix return value of dispatch_rqueue()

6 years agobus: fix potentially uninitialized memory access
Lennart Poettering [Wed, 2 Oct 2013 17:37:11 +0000 (19:37 +0200)]
bus: fix potentially uninitialized memory access

6 years agojournald: add missing error check
Lennart Poettering [Wed, 2 Oct 2013 17:36:43 +0000 (19:36 +0200)]
journald: add missing error check

6 years agocryptsetup: fix OOM handling when parsing mount options
Lennart Poettering [Wed, 2 Oct 2013 17:36:28 +0000 (19:36 +0200)]
cryptsetup: fix OOM handling when parsing mount options

6 years agosystemctl: fix name mangling for sysv units
Václav Pavlín [Wed, 2 Oct 2013 14:42:42 +0000 (16:42 +0200)]
systemctl: fix name mangling for sysv units

6 years agoacpi: make sure we never free an uninitialized pointer
Lennart Poettering [Wed, 2 Oct 2013 12:03:56 +0000 (14:03 +0200)]
acpi: make sure we never free an uninitialized pointer

6 years agofix lingering references to /var/lib/{backlight,random-seed}
Dave Reisner [Wed, 2 Oct 2013 11:46:24 +0000 (07:46 -0400)]
fix lingering references to /var/lib/{backlight,random-seed}

This should have been part of ef5bfcf668e6029faa78534dfe.

6 years agoacpi-fptd: fix memory leak in acpi_get_boot_usec
Lukas Nykryn [Wed, 2 Oct 2013 11:39:49 +0000 (13:39 +0200)]
acpi-fptd: fix memory leak in acpi_get_boot_usec

6 years agocgroup: there's no point in labelling cgroupfs dirs, so let's not do that v208
Lennart Poettering [Wed, 2 Oct 2013 02:54:07 +0000 (04:54 +0200)]
cgroup: there's no point in labelling cgroupfs dirs, so let's not do that

This allows us to get rid of the dep on libsystemd-label for cgroup
management.

https://bugs.freedesktop.org/show_bug.cgi?id=69966

6 years agobuild-sys: mkdir.[ch] should be in libsystemd-shared
Lennart Poettering [Wed, 2 Oct 2013 02:52:48 +0000 (04:52 +0200)]
build-sys: mkdir.[ch] should be in libsystemd-shared

Otherwise, why is mkdir-label.[ch] split out?

6 years agobuild-ss: prepare new release
Lennart Poettering [Wed, 2 Oct 2013 01:02:25 +0000 (03:02 +0200)]
build-ss: prepare new release

6 years agohashmap: randomize hash functions a bit
Lennart Poettering [Tue, 1 Oct 2013 21:11:23 +0000 (23:11 +0200)]
hashmap: randomize hash functions a bit

6 years agoUpdate TODO
David Herrmann [Tue, 1 Oct 2013 17:21:35 +0000 (19:21 +0200)]
Update TODO

Remove "logind fbdev removal" as it is no longer supported.

6 years agologind: remove fbdev session-device support
David Herrmann [Tue, 1 Oct 2013 16:08:54 +0000 (18:08 +0200)]
logind: remove fbdev session-device support

fbdev does not support access-handover so it is quite useless to route it
through logind. If compositors want to use it they ought to open it
themselves. It's highly recommended to be ignored entirely, though. fbdev
is about to be deprecated in the kernel.

6 years agologind: run with CAP_SYS_ADMIN
David Herrmann [Tue, 1 Oct 2013 15:59:44 +0000 (17:59 +0200)]
logind: run with CAP_SYS_ADMIN

DRM Master access requires CAP_SYS_ADMIN, yay! Add it to the capability
bounding set for systemd-logind. As CAP_SYS_ADMIN actually allows a huge
set of actions, this mostly renders the restriction-set useless. Anyway,
patches are already pending to reduce the restriction on the kernel side.
But these won't really make it into any stable-release so for now we're
stuck with CAP_SYS_ADMIN.

6 years agologind: send PropertyChanged during deactivation
David Herrmann [Tue, 1 Oct 2013 15:58:58 +0000 (17:58 +0200)]
logind: send PropertyChanged during deactivation

We only send the PropertyChanged signal for the to-be-activated session
but not for the to-be-deactivated one. Fix that so both listeners get
notified about the new state.

6 years agologind: check whether first drmSetMaster succeeded
David Herrmann [Tue, 1 Oct 2013 15:53:43 +0000 (17:53 +0200)]
logind: check whether first drmSetMaster succeeded

The initial drmSetMaster may fail if there is an active master already. We
must not assume that all existing clients comply to logind rules. We check
for this during session-activation already but didn't during device setup.
Fix this by checking the return code.

As drmSetMaster has had horrible return codes in the past (0 for failure?
EINVAL for denied access, ..) we need to be quite pedantic. To guarantee
an open file-descriptor we need to close the device and reopen it without
master if setting master failed first.

6 years agologind: fix session-device dbus notify
David Herrmann [Tue, 1 Oct 2013 15:48:15 +0000 (17:48 +0200)]
logind: fix session-device dbus notify

Had this fix lying around here for some time. Thanks to missing
type-checking for va-args we passed in the actual major/minor values
instead of pointers to it. Fix it by saving the values on the stack first
and passing in the pointers.

6 years agokernel-install: add compat with 'installkernel'
Tom Gundersen [Wed, 25 Sep 2013 22:38:34 +0000 (00:38 +0200)]
kernel-install: add compat with 'installkernel'

If 'kernel-install' is called as 'installkernel' it will be compatible with the
syntax used by the kernel's build system.

This means it can be called by doing 'make install' in a kernel build
directory, if the correct symlink has been installed (which we don't do by
default yet).

[Edit harald@redhat.com: removed basename and use shift]

6 years agoupdate TODO
Lennart Poettering [Tue, 1 Oct 2013 11:35:37 +0000 (13:35 +0200)]
update TODO

6 years agocore: whenever a new PID is passed to us, make sure we watch it
Lennart Poettering [Tue, 1 Oct 2013 03:06:56 +0000 (05:06 +0200)]
core: whenever a new PID is passed to us, make sure we watch it

6 years agoutil.c: ignore pollfd.revent for loop_read/loop_write
Chen Jie [Thu, 12 Sep 2013 01:21:41 +0000 (09:21 +0800)]
util.c: ignore pollfd.revent for loop_read/loop_write

Let read()/write() report any error/EOF.

6 years agounits: Add SHELL environment variable
Evan Callicoat [Tue, 24 Sep 2013 02:01:04 +0000 (21:01 -0500)]
units: Add SHELL environment variable

With the advent of systemd --user sessions, it's become very interesting to spawn X as a user unit, as well as accompanying processes that may have previously been in a .xinitrc/.xsession, or even just to replace a collection of XDG/GDM/KDM/etc session files with independent systemd --user units. The simplest case here would be to login on a tty, with the traditional /usr/sbin/login "login manager".

However, systemd --user (spawned by user@.service) is at the top level of the slice for the user, and does not inherit any environment variables from the login process. Given the number of common applications which rely on SHELL being set in the environment, it seems like the cleanest way to provide this variable is to set it to %s in the user@.service.

Ideally in the long-term, applications which rely on SHELL being set should be fixed to just grab it from getpwnam() or similar, but until that becomes more common, I propose this simple change to make user sessions a little bit nicer out of the box.

6 years agobuild-sys: don't fallback to upstart defaults
Lennart Poettering [Tue, 1 Oct 2013 01:22:41 +0000 (03:22 +0200)]
build-sys: don't fallback to upstart defaults

6 years agosystemctl: remove legacy upstart compatibility
Cristian Rodríguez [Mon, 30 Sep 2013 02:17:42 +0000 (23:17 -0300)]
systemctl: remove legacy upstart compatibility

6 years agosmack-setup: fix path to Smack/CIPSO mappings
Patrick McCarty [Tue, 1 Oct 2013 00:43:38 +0000 (17:43 -0700)]
smack-setup: fix path to Smack/CIPSO mappings

The correct path to the dir with CIPSO mappings is /etc/smack/cipso.d/;
/etc/smack/cipso is a file that can include these mappings as well,
though it is no longer supported in upstream libsmack.

6 years agoRevert "build-sys: link libsystemd-login with libsystemd-label.la"
Kay Sievers [Mon, 30 Sep 2013 22:19:25 +0000 (00:19 +0200)]
Revert "build-sys: link libsystemd-login with libsystemd-label.la"

Systemd-logind does not pull in cg_create(), if we unconditionally link
this, all users of systemd-logind qill need the label stuff and therefore
link against selinux.

It is probably a build-system issue, or something that need to be sorted
out in a differnt way than linking not needed libs.

This reverts commit ceadabb102b05b237bfab11e1f742975ee4daeb1.

6 years agoUpdate TODO
Lennart Poettering [Mon, 30 Sep 2013 22:15:15 +0000 (00:15 +0200)]
Update TODO

6 years agohashmap: size hashmap bucket array dynamically
Lennart Poettering [Mon, 30 Sep 2013 22:13:18 +0000 (00:13 +0200)]
hashmap: size hashmap bucket array dynamically

Instead of fixing the hashmap bucket array to 127 entries dynamically
size it, starting with a smaller one of 31. As soon as a fill level of
75% is reached, quadruple the size, and so on.

This should siginficantly optimize the lookup time in large tables
(from O(n) back to O(1)), and save memory on smaller tables (which most
are).

6 years agolocal: fix memory leak when putting together locale settings
Lennart Poettering [Mon, 30 Sep 2013 22:08:30 +0000 (00:08 +0200)]
local: fix memory leak when putting together locale settings

Also, we need to use proper strv_env_xyz() calls when putting together
the environment array, since otherwise settings won't be properly
overriden.

And let's get rid of strv_appendf(), is overkill and there was only one
user.

6 years agomain: don't free fds array twice
Lennart Poettering [Mon, 30 Sep 2013 22:06:48 +0000 (00:06 +0200)]
main: don't free fds array twice

6 years agologs-show.c: show all messages for a slice
Zbigniew Jędrzejewski-Szmek [Mon, 30 Sep 2013 21:58:44 +0000 (23:58 +0200)]
logs-show.c: show all messages for a slice

6 years agobuild-sys: link libsystemd-login with libsystemd-label.la
Michał Górny [Mon, 30 Sep 2013 08:08:09 +0000 (10:08 +0200)]
build-sys: link libsystemd-login with libsystemd-label.la

libsystemd-login.la uses cg_create() that currently seems to be a part
of libsystemd-label.la. However, it doesn't link against that library
and it seems that none of the (unconditional) libraries it uses do. In
the end, people end up getting «undefined reference to `cg_create'»
when trying to build e.g. dbus.

6 years agoman: mention the systemd homepage from systemd(1)
Lennart Poettering [Mon, 30 Sep 2013 16:56:34 +0000 (18:56 +0200)]
man: mention the systemd homepage from systemd(1)

6 years agoupdate TODO
Lennart Poettering [Mon, 30 Sep 2013 16:54:12 +0000 (18:54 +0200)]
update TODO

6 years agoman: link cgroups api docs from relevant man pages
Lennart Poettering [Mon, 30 Sep 2013 16:54:05 +0000 (18:54 +0200)]
man: link cgroups api docs from relevant man pages

6 years agoset IgnoreOnIsolate=true for systemd-cryptsetup@.service
Andrey Borzenkov [Sun, 29 Sep 2013 11:37:30 +0000 (15:37 +0400)]
set IgnoreOnIsolate=true for systemd-cryptsetup@.service

When crypttab contains noauto, cryptsetup service does not have any
explicit dependencies. If service is started later manually (directly or via
mount dependency) it will be stopped on isolate.

mount units already have IgnoreOnIsolate set by default. Set it by
default for cryptsetup units as well.

6 years agoFix buffer overrun when enumerating files
Zbigniew Jędrzejewski-Szmek [Sun, 29 Sep 2013 12:40:58 +0000 (14:40 +0200)]
Fix buffer overrun when enumerating files

https://bugs.freedesktop.org/show_bug.cgi?id=69887

Based-on-a-patch-by: Hans Petter Jansson <hpj@copyleft.no>
6 years agokmod-static-nodes: condition execution on modules.devname
Dave Reisner [Sat, 28 Sep 2013 18:51:39 +0000 (14:51 -0400)]
kmod-static-nodes: condition execution on modules.devname

6 years agoAdd a bit more explicit message, to help confused users
Michael Scherer [Fri, 27 Sep 2013 09:43:28 +0000 (11:43 +0200)]
Add a bit more explicit message, to help confused users

Seeing http://www.happyassassin.net/2013/09/27/further-sysadmin-adventures-wheres-my-freeipa-badge/
it seems that the default message is a bit confusing for people
who never encountered it before, so adding a link to the manpage could
help them.

6 years agokernel-install: avoid using 'cp --preserve'
Tom Gundersen [Wed, 25 Sep 2013 20:58:00 +0000 (22:58 +0200)]
kernel-install: avoid using 'cp --preserve'

Force 0644 and root:root instead, to avoid problems with fat filesystems.

6 years agobuild-sys: restore detection of sphinx
Zbigniew Jędrzejewski-Szmek [Fri, 27 Sep 2013 06:24:00 +0000 (08:24 +0200)]
build-sys: restore detection of sphinx

6 years agoDo not use unitialized variable and remove duplicated line
Zbigniew Jędrzejewski-Szmek [Fri, 27 Sep 2013 05:58:57 +0000 (07:58 +0200)]
Do not use unitialized variable and remove duplicated line

6 years agoUpdate TODO
Lennart Poettering [Fri, 27 Sep 2013 00:59:50 +0000 (02:59 +0200)]
Update TODO

6 years agoswap: properly expose timeout property on the bus
Lennart Poettering [Fri, 27 Sep 2013 00:02:21 +0000 (02:02 +0200)]
swap: properly expose timeout property on the bus

6 years agodbus: fix introspection for TimerSlackNSec
Lennart Poettering [Thu, 26 Sep 2013 22:35:36 +0000 (00:35 +0200)]
dbus: fix introspection for TimerSlackNSec

6 years agosystemctl: make sure set-property mangles unit names
Lennart Poettering [Thu, 26 Sep 2013 22:11:54 +0000 (00:11 +0200)]
systemctl: make sure set-property mangles unit names

6 years agoman: drop references to "cgroup" wher appropriate
Lennart Poettering [Thu, 26 Sep 2013 22:05:07 +0000 (00:05 +0200)]
man: drop references to "cgroup" wher appropriate

Since cgroups are mostly now an implementation detail of systemd lets
deemphasize it a bit in the man pages. This renames systemd.cgroup(5) to
systemd.resource-control(5) and uses the term "resource control" rather
than "cgroup" where appropriate.

This leaves the word "cgroup" in at a couple of places though, like for
example systemd-cgtop and systemd-cgls where cgroup stuff is at the core
of what is happening.

6 years agocore: drop some out-of-date references to cgroup settings
Lennart Poettering [Thu, 26 Sep 2013 20:49:44 +0000 (22:49 +0200)]
core: drop some out-of-date references to cgroup settings

6 years agologind: never consider a closing session relevant for PK checks
Lennart Poettering [Thu, 26 Sep 2013 19:32:47 +0000 (21:32 +0200)]
logind: never consider a closing session relevant for PK checks

https://bugzilla.redhat.com/show_bug.cgi?id=1010215

6 years agomain: set umask before creating any files
Mike Gilbert [Thu, 26 Sep 2013 18:39:41 +0000 (20:39 +0200)]
main: set umask before creating any files

This avoids a problem when we inherit a non-zero umask from the
initramfs. This would cause /run/systemd to be created with the wrong
mode.

6 years agologind: if a user is sitting in front of the computer and can shutdown the machine...
Lennart Poettering [Thu, 26 Sep 2013 18:31:37 +0000 (20:31 +0200)]
logind: if a user is sitting in front of the computer and can shutdown the machine anyway he should also be able to reboot it

6 years agocore: rework how we match mount units against each other
Lennart Poettering [Thu, 26 Sep 2013 18:14:24 +0000 (20:14 +0200)]
core: rework how we match mount units against each other

Previously to automatically create dependencies between mount units we
matched every mount unit agains all others resulting in O(n^2)
complexity. On setups with large amounts of mount units this might make
things slow.

This change replaces the matching code to use a hashtable that is keyed
by a path prefix, and points to a set of units that require that path to
be around. When a new mount unit is installed it is hence sufficient to
simply look up this set of units via its own file system paths to know
which units to order after itself.

This patch also changes all unit types to only create automatic mount
dependencies via the RequiresMountsFor= logic, and this is exposed to
the outside to make things more transparent.

With this change we still have some O(n) complexities in place when
handling mounts, but that's currently unavoidable due to kernel APIs,
and still substantially better than O(n^2) as before.

https://bugs.freedesktop.org/show_bug.cgi?id=69740

6 years agounit-name: when escaping a path consider the empty path identical to the root dir
Lennart Poettering [Thu, 26 Sep 2013 18:03:20 +0000 (20:03 +0200)]
unit-name: when escaping a path consider the empty path identical to the root dir

6 years agoutil: properly handle the root dir in PATH_FOREACH_PREFIX
Lennart Poettering [Thu, 26 Sep 2013 17:58:33 +0000 (19:58 +0200)]
util: properly handle the root dir in PATH_FOREACH_PREFIX

Also add PATH_FOREACH_PREFIX_MORE which includes the specified dir
itself in the iteration

6 years agocgroup: when referencing cgroup controller trees allow omission of the path
Lennart Poettering [Thu, 26 Sep 2013 17:57:58 +0000 (19:57 +0200)]
cgroup: when referencing cgroup controller trees allow omission of the path

6 years agoMove part of logind.c into a separate file
Zbigniew Jędrzejewski-Szmek [Wed, 25 Sep 2013 15:04:41 +0000 (17:04 +0200)]
Move part of logind.c into a separate file

liblogind-core.la was underlinked, missing a few functions
defined in logind.c. They are moved to a new file, logind-core.c,
and this file is linked into liblogind-core.la.
In addition, logind-acl.c is attached to the liblogind-core.la,
instead of systemd-logind directly.

6 years agoMove functions around to fix underlinking in test-machine-tables
Zbigniew Jędrzejewski-Szmek [Wed, 25 Sep 2013 11:26:08 +0000 (13:26 +0200)]
Move functions around to fix underlinking in test-machine-tables

6 years agobuild-sys: add ./configure --enable-address-sanitizer
Zbigniew Jędrzejewski-Szmek [Thu, 26 Sep 2013 08:31:44 +0000 (10:31 +0200)]
build-sys: add ./configure --enable-address-sanitizer

Enabling address sanitizer seems like a useful thing, but is quite
tricky. Proper flags have to be passed to CPPFLAGS, CFLAGS and
LDFLAGS, but passing them on the commandline doesn't work because
we tests are done with ld directly, and not with libtool like in
real linking. We might want to fix this, but let's add a handy
way to enable address checking anyway.

6 years agojournald: accept EPOLLERR from /dev/kmsg
Zbigniew Jędrzejewski-Szmek [Thu, 19 Sep 2013 21:57:57 +0000 (16:57 -0500)]
journald: accept EPOLLERR from /dev/kmsg

Also print out unexpected epoll events explictly.

6 years agotest-hashmap: fix access to uninitialized memory
Zbigniew Jędrzejewski-Szmek [Wed, 25 Sep 2013 15:52:43 +0000 (17:52 +0200)]
test-hashmap: fix access to uninitialized memory

6 years agoexecute.c: little modernization
Zbigniew Jędrzejewski-Szmek [Wed, 28 Aug 2013 11:54:43 +0000 (07:54 -0400)]
execute.c: little modernization

6 years agoutil: add macro for iterating through all prefixes of a path
Lennart Poettering [Wed, 25 Sep 2013 18:58:23 +0000 (20:58 +0200)]
util: add macro for iterating through all prefixes of a path

Syntactic sugar in a macro PATH_FOREACH_PREFIX.

6 years agokeymap: Add Samsung Ativ 9 Plus
Jimmie Tauriainen [Wed, 25 Sep 2013 05:30:23 +0000 (07:30 +0200)]
keymap: Add Samsung Ativ 9 Plus

https://launchpad.net/bugs/1229936

6 years agocgroup: if we do a cgroup operation then do something on all supported controllers
Lennart Poettering [Tue, 24 Sep 2013 02:56:05 +0000 (21:56 -0500)]
cgroup: if we do a cgroup operation then do something on all supported controllers

Previously we did operations like attach, trim or migrate only on the
controllers that were enabled for a specific unit. With this changes we
will now do them for all supproted controllers, and fall back to all
possible prefix paths if the specified paths do not exist.

This fixes issues if a controller is being disabled for a unit where it
was previously enabled, and makes sure that all processes stay as "far
down" the tree as groups exist.

6 years agojournalctl(1): s/adm/systemd-journal/
Dave Reisner [Tue, 24 Sep 2013 18:39:40 +0000 (14:39 -0400)]
journalctl(1): s/adm/systemd-journal/

6 years agologind: return -EINVAL when PID is wrong
Lukas Nykryn [Tue, 24 Sep 2013 14:27:20 +0000 (16:27 +0200)]
logind: return -EINVAL when PID is wrong

dbus-send --print-reply --system --dest=org.freedesktop.login1
/org/freedesktop/login1 org.freedesktop.login1.Manager.GetUserByPID
uint32:0
causes
systemd-logind[29843]: Assertion 'pid >= 1' failed at
src/login/logind.c:938, function manager_get_user_by_pid(). Aborting.

6 years agoTODO: add header back
Kay Sievers [Tue, 24 Sep 2013 13:47:42 +0000 (15:47 +0200)]
TODO: add header back

6 years agohwdb: update
Kay Sievers [Tue, 24 Sep 2013 13:47:12 +0000 (15:47 +0200)]
hwdb: update

6 years agosupport acpi firmware performance data (FPDT)
Kay Sievers [Mon, 23 Sep 2013 22:23:35 +0000 (17:23 -0500)]
support acpi firmware performance data (FPDT)

Prefer firmware-provided performance data over loader-exported ones; if
ACPI data is available, always use it, otherwise try to read the loader
data.

The firmware-provided variables start at the time the first EFI image
is executed and end when the operating system exits the boot services;
the (loader) time calculated in systemd-analyze increases.

6 years agoshared: device-nodes - add include guard
Kay Sievers [Mon, 23 Sep 2013 21:53:19 +0000 (16:53 -0500)]
shared: device-nodes - add include guard

6 years agoTODO: update
Kay Sievers [Mon, 23 Sep 2013 21:52:58 +0000 (16:52 -0500)]
TODO: update