chiark / gitweb /
Jan Janssen [Sat, 13 Jul 2013 11:19:38 +0000 (13:19 +0200)]
cryptsetup: Add tcrypt support
Tcrypt uses a different approach to passphrases/key files. The
passphrase and all key files are incorporated into the "password"
to open the volume. So, the idea of slots that provide a way to
open the volume with different passphrases/key files that are
independent from each other like with LUKS does not apply.
Therefore, we use the key file from /etc/crypttab as the source
for the passphrase. The actual key files that are combined with
the passphrase into a password are provided as a new option in
/etc/crypttab and can be given multiple times if more than one
key file is used by a volume.
Jan Janssen [Sat, 13 Jul 2013 11:19:37 +0000 (13:19 +0200)]
cryptsetup: Move attaching of the device out of main
Jan Janssen [Sat, 13 Jul 2013 11:19:36 +0000 (13:19 +0200)]
cryptsetup: Move password query out of main
Also use _cleanup_free_ where possible.
Frederic Crozat [Mon, 15 Jul 2013 16:33:57 +0000 (18:33 +0200)]
util: recognize 'ncp' as an alias to 'ncpfs'
ncp is also used for Netware mount point, recognize it as such. Fixes
https://bugzilla.novell.com/show_bug.cgi?id=828905.
Lennart Poettering [Mon, 15 Jul 2013 21:54:28 +0000 (23:54 +0200)]
update TODO
Lennart Poettering [Fri, 12 Jul 2013 17:47:08 +0000 (19:47 +0200)]
update TODO
Holger Hans Peter Freyther [Tue, 9 Jul 2013 16:09:08 +0000 (18:09 +0200)]
travis: Add a travis.yml for doing CI after commits
Instruct travis-ci to build systemd and create a tarball. In case
of an error travis-ci will complain on IRC. The systemd testsuite
currently requires the host to have a recent version of systemd
installed and running. This is not the case for the Ubuntu VM of
travis-ci. This means make check and make distcheck will result in
a build failure and to avoid this these commands are not executed.
This requires a one time configuration on travis-ci for the repo
on github by the owner of the repo.
Thomas Hindoe Paaboel Andersen [Mon, 15 Jul 2013 22:07:36 +0000 (00:07 +0200)]
tests: add more tests for shared/util.c
Shawn Landden [Mon, 15 Jul 2013 03:13:09 +0000 (20:13 -0700)]
journalctl: add --force option to recreate FSS
Zbigniew Jędrzejewski-Szmek [Mon, 15 Jul 2013 03:36:14 +0000 (23:36 -0400)]
systemd: fix NULL dereference when disabling a nonexistent instance
Assertion 'p' failed at src/shared/path-util.c:51, function path_get_file_name(). Aborting.
The unit file could not be found, and i->path would not be set.
In
02b9e969 a code path was added which attempts to remove symlinks
to a nonexistent (removed) unit file. This worked OK in case of
non-instance services, but broke in the case of instance services.
Behaviour wrt. to instance units is changed in the way that
02b9e969
changed it for non-instance units: it is now possible to remove
instance symlinks to a template unit that has been removed.
This patch isn't a full fix, because the behaviour wrt. to enabling
and disabling instance units is still broken: e.g it is possible to
start autovt@tty5.service, but it is not possible to enable it,
because autovt@.service is a symlink, and on the other hand, disabling
getty@tty5.service removes all symlinks to getty@.service, which is
wrong too. But segfaults make bad pr, so let's add at least this
partial fix for now.
Kay Sievers [Sun, 14 Jul 2013 22:56:18 +0000 (00:56 +0200)]
hwdb: keyboard update
Kay Sievers [Sun, 14 Jul 2013 22:55:37 +0000 (00:55 +0200)]
hwdb: import data
Zbigniew Jędrzejewski-Szmek [Sun, 14 Jul 2013 03:36:39 +0000 (23:36 -0400)]
man: tweak WantedBy=/RequiredBy= description a bit
https://bugs.freedesktop.org/show_bug.cgi?id=55663
Zbigniew Jędrzejewski-Szmek [Sat, 13 Jul 2013 21:36:03 +0000 (17:36 -0400)]
test: add trivial test for syscall table and extend table tests to error paths
Thomas H.P. Andersen [Thu, 28 Mar 2013 13:54:06 +0000 (14:54 +0100)]
Add test coverage and generate report with lcov
Enable coverage with --enable-coverage.
"make coverage" will create the report locally,
"make coverage-sync" will upload the report to
http://www.freedesktop.org/software/systemd/coverage/.
Requires lcov version 1.10 to handle naming in systemd and to
use the --no-external option.
[zj: make the coverage at least generate something with
separate build dir, simplify rules a bit: all errors
are mine. ]
Zbigniew Jędrzejewski-Szmek [Sat, 13 Jul 2013 13:59:35 +0000 (09:59 -0400)]
units: do not special-case getty@tty1.service installation
Since a long while we can use "systemctl enable getty@tty1.service"
which does the right thing, so there's no need to abuse Alias=
for installation.
Ross Lagerwall [Mon, 17 Jun 2013 18:11:50 +0000 (19:11 +0100)]
install: make "reenable" work with templated units
Before, "systemctl reenable getty@tty1.service" would fail with:
Failed to issue method call: File exists
To fix this, reimplement "reenable" explicitly as a disable followed by
an enable.
This is shorter and is how the man page documents its behavior.
Jan Engelhardt [Sat, 13 Jul 2013 08:51:35 +0000 (10:51 +0200)]
man: wording and grammar update
Shawn Landden [Sat, 13 Jul 2013 03:57:15 +0000 (20:57 -0700)]
journalctl: have a useful --setup-keys error message when using non-persistant logging
Generating seed...
Generating key pair...
Generating sealing key...
Failed to open /var/log/journal/
33f46101703a10c5fc6fa4f451840101/fss.tmp.k2wDDU: No such file or directory
Tom Gundersen [Fri, 12 Jul 2013 09:14:54 +0000 (11:14 +0200)]
static-nodes: don't hardcode path to mkdir
Zbigniew Jędrzejewski-Szmek [Fri, 12 Jul 2013 17:19:19 +0000 (13:19 -0400)]
Revert "build-sys: don't enable color gcc on dumb terminals"
This reverts commit
cd3069559a09b4e4f85a6f02aa8f0521f48359ca.
Emacs compilation can be fixed by putting
(custom-set-variables
'(compilation-environment (quote ("GCC_COLORS="))))
in ~/.emacs.
Zbigniew Jędrzejewski-Szmek [Fri, 12 Jul 2013 12:15:21 +0000 (08:15 -0400)]
shared/install: fix trivial memleak
We lost the reference when setting path second time.
Zbigniew Jędrzejewski-Szmek [Fri, 12 Jul 2013 12:04:39 +0000 (08:04 -0400)]
test-path-util,test-sched-prio: uninitialize manager to appease valgrind
Kay Sievers [Fri, 12 Jul 2013 11:44:08 +0000 (13:44 +0200)]
hwdb: keyboard -- update comments
Kay Sievers [Fri, 12 Jul 2013 10:20:18 +0000 (12:20 +0200)]
hwdb: keyboard -- add file
Do not enable/install it now, until we switch over from the current
keymap tool.
Zbigniew Jędrzejewski-Szmek [Wed, 10 Jul 2013 05:25:02 +0000 (01:25 -0400)]
man: document the slice and scope units, add systemd.cgroup(5)
Zbigniew Jędrzejewski-Szmek [Fri, 12 Jul 2013 03:39:27 +0000 (23:39 -0400)]
man: use <literal> not <filename> for suffixes
Especially sentences like "filename ends in .suffix" are easier to
parse if the suffix is surrounded by quotes. In sentences like
"requires a .service unit", where the suffix is used as a class
designation, there is no need to use quotes.
Zbigniew Jędrzejewski-Szmek [Thu, 11 Jul 2013 01:43:34 +0000 (21:43 -0400)]
man: remove note about shutdown being legacy
This can only confuse people, because there's no need to
discourage people from using shutdown. It is fully functional
and supported.
Lukas Nykryn [Tue, 9 Jul 2013 13:45:37 +0000 (15:45 +0200)]
journald-server: r should be checked after journal_file_open_reliably
Lennart Poettering [Thu, 11 Jul 2013 23:18:09 +0000 (01:18 +0200)]
update TODO
Lennart Poettering [Thu, 11 Jul 2013 23:15:52 +0000 (01:15 +0200)]
build-sys: don't enable color gcc on dumb terminals
Guys, we know that emacs is the best editor on earth, but unfortunately
its "M-x compile" terminal cannot do colors (well, it does its own
highlighting of the output anyway), and it will inform the programs it
calls about this with TERM=dumb, and gcc should check for that. But you
guys turned that off. Not cool. Let's turn it on again.
Lukas Nykryn [Tue, 9 Jul 2013 13:45:38 +0000 (15:45 +0200)]
systemctl: remove unused variable
Lukas Nykryn [Tue, 9 Jul 2013 13:45:36 +0000 (15:45 +0200)]
unit: check correct variable after strdup
Lennart Poettering [Thu, 11 Jul 2013 19:30:19 +0000 (21:30 +0200)]
update TODO
Lennart Poettering [Thu, 11 Jul 2013 19:29:33 +0000 (21:29 +0200)]
core: simplify drop-in writing logic a bit
let's make use of some format string magic!
Lennart Poettering [Thu, 11 Jul 2013 19:03:06 +0000 (21:03 +0200)]
core: when writing drop-in files, name them directly after the property we set
Mapping from "FooBar" to "foo-bar" is unnecessary and makes it hard to
handle many different properties with the same code, hence, let's just
not do it.
Lennart Poettering [Thu, 11 Jul 2013 18:40:18 +0000 (20:40 +0200)]
cgroup: split out per-device BlockIOWeight= setting into BlockIODeviceWeight=
This way we can nicely map the configuration directive to properties and
back, without requiring two different signatures for the same property.
Lennart Poettering [Thu, 11 Jul 2013 17:53:46 +0000 (19:53 +0200)]
machined: forward scope properties array from client to systemd
This makes nspawn's --scope= switch work again.
Lennart Poettering [Thu, 11 Jul 2013 17:16:11 +0000 (19:16 +0200)]
update TODO
Lennart Poettering [Thu, 11 Jul 2013 17:15:01 +0000 (19:15 +0200)]
systemctl: show cgroup tree if cgroup is empty but has non-empty children
This makes sure "systemctl status user.slice" shows a nice cgroup tree
of all logged in users.
Lennart Poettering [Thu, 11 Jul 2013 17:14:38 +0000 (19:14 +0200)]
loginctl: suppress cgroup tree output if cgroup is empty
same for machinectl
Lennart Poettering [Thu, 11 Jul 2013 16:49:44 +0000 (18:49 +0200)]
update TODO
Lennart Poettering [Thu, 11 Jul 2013 16:47:20 +0000 (18:47 +0200)]
cgroup: simplify how instantiated units are mapped to cgroups
Previously for an instantiated unit foo@bar.service we created a cgroup
foo@.service/foo@bar.service, in order to place all instances of the
same template inside the same subtree. As we now implicitly add all
instantiated units into one per-template slice we don't need this
complexity anymore, and instance units can map directly to the cgroups
of their full name.
Lennart Poettering [Thu, 11 Jul 2013 16:42:12 +0000 (18:42 +0200)]
cgroup: don't ever try to destroy the cgroup of the root slice
The root slice is after all the root cgroup, so don't attempt to delete
it.
Lennart Poettering [Thu, 11 Jul 2013 16:37:14 +0000 (18:37 +0200)]
cgroup: don't move systemd into systems.slice when running as --user instance
Lennart Poettering [Thu, 11 Jul 2013 01:52:43 +0000 (03:52 +0200)]
core: implicitly create a per-template slice for all instantiated units by default
If no explicit slice is configured for an instantiated unit, create an
implicit one for all instances of the same template.
Dave Reisner [Thu, 11 Jul 2013 14:33:48 +0000 (10:33 -0400)]
configure: split checks for libkmod >= 14
PKG_CHECK_EXISTS won't created a cached variable that later messes with
our PKG_CHECK_MODULES check for an explicit version. Unfortunately,
nesting these checks as the code existed lead to an odd error. Rather,
split the checks apart.
This also improves to the error message when the requisite version
isn't found, and supplies the literal version systemd needs.
Harald Hoyer [Thu, 11 Jul 2013 11:42:14 +0000 (13:42 +0200)]
kernel-install/90-loaderentry.install: do not fail for missing initrd
Shawn Landden [Mon, 8 Jul 2013 18:28:14 +0000 (18:28 +0000)]
basic SO_REUSEPORT support
Lennart Poettering [Wed, 10 Jul 2013 23:57:09 +0000 (01:57 +0200)]
units: since we auto-spawn user@.service instances now we don need an [Install] section in it
Lennart Poettering [Wed, 10 Jul 2013 23:56:45 +0000 (01:56 +0200)]
update TODO
Lennart Poettering [Wed, 10 Jul 2013 23:56:12 +0000 (01:56 +0200)]
core: grant user@.service instances write access to their own cgroup
Lennart Poettering [Wed, 10 Jul 2013 22:48:52 +0000 (00:48 +0200)]
core: rearrange if blocks a bit
Lennart Poettering [Wed, 10 Jul 2013 22:35:01 +0000 (00:35 +0200)]
shutdown: avoid malloc() if we can
Lennart Poettering [Wed, 10 Jul 2013 22:28:35 +0000 (00:28 +0200)]
update TODO
Lennart Poettering [Wed, 10 Jul 2013 22:27:54 +0000 (00:27 +0200)]
loginctl: restore cgroup display for status output
Same for machinectl.
Kay Sievers [Wed, 10 Jul 2013 22:13:46 +0000 (00:13 +0200)]
shutdown: fix /proc/cmdline reading of 'quiet'
Lennart Poettering [Wed, 10 Jul 2013 21:50:28 +0000 (23:50 +0200)]
core: uninstall cgroup agent only if we are running outside of a container
Since the cgroupfs is currently not virtualized for containers we
shouldn't reset the hosts agent from the container.
Lennart Poettering [Wed, 10 Jul 2013 21:47:15 +0000 (23:47 +0200)]
cgroup: when uninstalling agent, actually turn it off first
Lennart Poettering [Wed, 10 Jul 2013 21:46:30 +0000 (23:46 +0200)]
update TODO
Lennart Poettering [Wed, 10 Jul 2013 21:39:46 +0000 (23:39 +0200)]
user-sessions: rely on PID 1 to kill sessions
As we want to centralized cgroup access we should stop killing the user
sessions directly from the systemd-user-sessions service. Instead, rely
on PID 1 doing this by adding the right ordering dependencies to the
session scope units.
Lennart Poettering [Wed, 10 Jul 2013 21:33:17 +0000 (23:33 +0200)]
logind: when creating the scope job fails, return this immediately to the client that wants to register the session
Otherwise we'll hanging for the job to finish without any job existing.
Similar, for machined.
Lennart Poettering [Wed, 10 Jul 2013 21:31:40 +0000 (23:31 +0200)]
logind: don't misunderstand UnitRemoved signals during reloading
When PID 1 reloads the units logind/machined will see UnitRemoved
signals for all units. Instead of trusting these immediately, let's
check the actual unit state before considering a unit gone, so that
reloading PID 1 is not mistaken as the end of all sessions.
Lennart Poettering [Wed, 10 Jul 2013 19:17:37 +0000 (21:17 +0200)]
unit: when deserializing cgroup path add it back into cgroup hashmap
Also, properly remove cgroup path from hashmap when freeing unit.
Lennart Poettering [Wed, 10 Jul 2013 19:13:56 +0000 (21:13 +0200)]
unit: save description/slice of transient units to /run
This is necessary so that these properties survive a daemon reload.
Lennart Poettering [Wed, 10 Jul 2013 19:10:53 +0000 (21:10 +0200)]
core: send out "Reloading" signal before and after doing a full reload/reexec of PID 1
Since we'll unload all units/job during a reload, and then readd them it
is really useful for clients to be aware of this phase hence sent a
signal out before and after. This signal is called "Reloading" (despite
the fact that it is also sent out during reexecution, which we consider
a special case in this context) and has one boolean parameter which is
true for the signal sent before the reload, and false for the signal
after the reload. The UnitRemoved/JobRremoved and UnitNew/JobNew due to
the reloading are guranteed to be between the pair of Reloading
messages.
Lennart Poettering [Wed, 10 Jul 2013 18:44:21 +0000 (20:44 +0200)]
systemctl: suppress error message when doing "systemctl daemon-reexec"
When we issue a reexecution request via the private socket we need to
expect a "Disconnected" in addition to "NoReply" when the connection is
terminated.
Lennart Poettering [Wed, 10 Jul 2013 18:37:19 +0000 (20:37 +0200)]
systemctl: suppress error messages when checking whether a unit needs to be reloaded
Lennart Poettering [Wed, 10 Jul 2013 18:35:57 +0000 (20:35 +0200)]
scope: don't require an initialized PIDs set when deserializing
When a scope unit is created due to deserialization rather than client
request don't enforce that the PIDs set must be non-empty, since the
cgroup is already populated.
Lennart Poettering [Wed, 10 Jul 2013 18:33:11 +0000 (20:33 +0200)]
core: while we are reloading don't suppress bus signals
While we are reloading we shouldn't suppress adding units to the bus
queue when there are no subscribers, simply because we might not have
deserialized the subscribers list yet. Hence, during reloading always
assume we have subscribers.
Lennart Poettering [Wed, 10 Jul 2013 17:24:03 +0000 (19:24 +0200)]
core: serialize/deserialize bus subscribers
Lennart Poettering [Wed, 10 Jul 2013 00:01:29 +0000 (02:01 +0200)]
cgroup: downgrade error message when we cannot remove a cgroup to debug
Some units set KillMode=none to survive the initrd→rootfs transition. We
cannot remove their cgroups, but that shouldn't really be considered an
issue, so let's downgrade the error message.
Lennart Poettering [Wed, 10 Jul 2013 00:01:14 +0000 (02:01 +0200)]
update TODO
Kay Sievers [Wed, 10 Jul 2013 14:02:24 +0000 (16:02 +0200)]
udev: add builtin 'keyboard' to manage key mappings
Kay Sievers [Wed, 10 Jul 2013 12:59:24 +0000 (14:59 +0200)]
hwdb: allow list of lookup keys per given record
This allows to specify:
dmi:bvn*:bvr*:bd*:svnVENDOR:pn:Model 231*:pvr*
dmi:bvn*:bvr*:bd*:svnVENDOR:pn:Series 12*:pvr*
KEY_A=value
KEY_B=value
Instead of:
dmi:bvn*:bvr*:bd*:svnVENDOR:pn:Model 231*:pvr*
KEY_A=value
KEY_B=value
dmi:bvn*:bvr*:bd*:svnVENDOR:pn:Series 12*:pvr*
KEY_A=value
KEY_B=value
Zbigniew Jędrzejewski-Szmek [Wed, 10 Jul 2013 05:24:46 +0000 (01:24 -0400)]
man: document systemd-run
Zbigniew Jędrzejewski-Szmek [Wed, 10 Jul 2013 05:33:13 +0000 (01:33 -0400)]
man: document four basic slices
Kay Sievers [Tue, 9 Jul 2013 13:47:11 +0000 (15:47 +0200)]
hwdb: return false if no property is found
Jason St. John [Tue, 9 Jul 2013 11:12:05 +0000 (13:12 +0200)]
man: improve word usage and grammar in logind.conf(5)
Jan Engelhardt [Mon, 8 Jul 2013 16:19:02 +0000 (18:19 +0200)]
man: wording and grammar updates
Tom Gundersen [Mon, 8 Jul 2013 22:12:35 +0000 (00:12 +0200)]
configure: fail if out-of-date kmod found and kmod not disabled
Almost everyone wants kmod support, so don't fail silently if the libs are
out-of-date.
kmod can still be explicitly disabled and if it is not found at all, we still
default to disabling it.
Tom Gundersen [Fri, 14 Jun 2013 20:56:39 +0000 (22:56 +0200)]
static-nodes: move creation of static nodes from udevd to tmpfiles
As of kmod v14, it is possible to export the static node information from
/lib/modules/`uname -r`/modules.devname in tmpfiles.d(5) format.
Use this functionality to let systemd-tmpfilesd create the static device nodes
at boot, and drop the functionality from systemd-udevd.
As an effect of this we can move from systemd-udevd to systemd-tmpfiles-setup-dev:
* the conditional CAP_MKNOD (replaced by checking if /sys is mounted rw)
* ordering before local-fs-pre.target (see
89d09e1b5c65a2d97840f682e0932c8bb499f166)
Kay Sievers [Mon, 8 Jul 2013 09:46:44 +0000 (11:46 +0200)]
hwdb: import data
Kay Sievers [Mon, 8 Jul 2013 09:44:17 +0000 (11:44 +0200)]
hwdb: add --lookup-prefix= option
Kay Sievers [Sun, 7 Jul 2013 23:09:08 +0000 (01:09 +0200)]
hwdb: remove too broad PNP* match from 20-acpi-vendor.hwdb
Zbigniew Jędrzejewski-Szmek [Sun, 7 Jul 2013 03:33:26 +0000 (23:33 -0400)]
man: tweak the description of System/RuntimeMaxUse
https://bugs.freedesktop.org/show_bug.cgi?id=66657
Zbigniew Jędrzejewski-Szmek [Sun, 7 Jul 2013 02:22:05 +0000 (22:22 -0400)]
man: document machinectl and systemd-machined
Zbigniew Jędrzejewski-Szmek [Sat, 6 Jul 2013 02:15:54 +0000 (22:15 -0400)]
systemd-python: wrap sd_notify
_listen_fds() is modified to accept unset_environment arg as keyword,
to match new notify().
Kay Sievers [Fri, 5 Jul 2013 12:59:52 +0000 (14:59 +0200)]
update TODO
Zbigniew Jędrzejewski-Szmek [Fri, 5 Jul 2013 04:25:26 +0000 (00:25 -0400)]
tests: add tests for string lookup tables
The tests check if the tables have entries for all values
in the enum, and that the entries are unique.
Kay Sievers [Thu, 4 Jul 2013 22:32:05 +0000 (00:32 +0200)]
suppress status message output at shutdown when 'quiet' is given
Kay Sievers [Thu, 4 Jul 2013 18:54:40 +0000 (20:54 +0200)]
cgroups-agent: remove ancient fallback code; turn connection error into warning
During re-execution and shutdown cgroups agents might not be able
to connect to systemd's private D-Bus socket, the printed error to
the console is misleding in that case, so turn it into a warning.
Kay Sievers [Thu, 4 Jul 2013 18:31:18 +0000 (20:31 +0200)]
disable the cgroups release agent when shutting down
During shutdown, when we try to clean up all remaining processes, the
kernel will fork new agents every time a cgroup runs empty. These
new processes cause delays in the final SIGTERM, SIGKILL logic.
Apart from that, this should also avoid that the kernel-forked binaries
cause unpredictably timed access to the filesystem which we might need to
unmount.
William Giokas [Wed, 3 Jul 2013 23:00:46 +0000 (18:00 -0500)]
po: add dbus-scope.c to POTFILES.skip
Harald Hoyer [Thu, 4 Jul 2013 09:01:47 +0000 (11:01 +0200)]
core/mount.c:mount_dump(): don't segfault, if mount is not mounted anymore
Don't segfault, if m->from_proc_self_mountinfo and m->from_fragment is
false.
https://bugzilla.redhat.com/show_bug.cgi?id=957783#c9
Lennart Poettering [Wed, 3 Jul 2013 23:09:04 +0000 (01:09 +0200)]
update TODO
Lennart Poettering [Wed, 3 Jul 2013 15:49:43 +0000 (17:49 +0200)]
NEWS: no need to mention journalctl --user/--system twice
Lennart Poettering [Wed, 3 Jul 2013 14:49:54 +0000 (16:49 +0200)]
update TODO
Zbigniew Jędrzejewski-Szmek [Wed, 3 Jul 2013 15:20:17 +0000 (11:20 -0400)]
NEWS: mention set-log-level, --user, --system
Lennart Poettering [Wed, 3 Jul 2013 14:37:39 +0000 (16:37 +0200)]
build-sys: bump/correct library versions