Zbigniew Jędrzejewski-Szmek [Thu, 12 Dec 2013 01:34:19 +0000 (20:34 -0500)]
Do not log all assert_return failures
% build/journalctl help
Assertion 'match_is_valid(data, size)' failed at ../src/journal/sd-journal.c:227, function sd_journal_add_match(). Ignoring.
Callers cannot be expect to check all arguments always.
Djalal Harouni [Wed, 11 Dec 2013 23:22:48 +0000 (00:22 +0100)]
journal: pipe journalctl help output into a pager
journalctl help output might run off the screen, so be consistent
as other systemd tools do and pipe it into a pager.
Zbigniew Jędrzejewski-Szmek [Sun, 8 Dec 2013 15:38:33 +0000 (10:38 -0500)]
man: mention SYSTEMD_USER_WANTS
Also split into paragraphs for easier reading and make a few minor
corrections.
Lennart Poettering [Thu, 12 Dec 2013 02:03:03 +0000 (03:03 +0100)]
service: actually leave watchdog enabled in the states where it matters
Lennart Poettering [Thu, 12 Dec 2013 02:00:57 +0000 (03:00 +0100)]
timedated: make sure GetAll() succeeds in systems lacking /dev/rtc (such as containers)
Thomas Hindoe Paaboel Andersen [Wed, 11 Dec 2013 23:06:30 +0000 (00:06 +0100)]
test-fileio: replace mktemp with mkstemp to avoid warnings
This is a fairly useless thing to do but it makes the compilers
and analyzers shut up about the use of mktemp.
Lennart Poettering [Thu, 12 Dec 2013 00:42:41 +0000 (01:42 +0100)]
bus: properly generate NameOwnerChanged messages when we take from/give back to queue/starter
Lennart Poettering [Thu, 12 Dec 2013 00:39:32 +0000 (01:39 +0100)]
bus: fix order in which we check creds
Lennart Poettering [Thu, 12 Dec 2013 00:39:21 +0000 (01:39 +0100)]
update TODO
Lennart Poettering [Wed, 11 Dec 2013 23:35:23 +0000 (00:35 +0100)]
bus: update name listing logic to current kernel interface
Lennart Poettering [Wed, 11 Dec 2013 23:20:11 +0000 (00:20 +0100)]
bus: when checking whether a creds object contains some field, don't use assert_return()
These are not programming errors, so they shouldn't use assert_return()
Lennart Poettering [Wed, 11 Dec 2013 23:07:49 +0000 (00:07 +0100)]
bus: connect directly via kdbus in sd_bus_open_system_container()
kdbus fortunately exposes the container's busses in the host fs, hence
we can access it directly instead of doing the namespacing dance.
Lennart Poettering [Wed, 11 Dec 2013 22:41:21 +0000 (23:41 +0100)]
sd-event: try to move each wakeup to the same spot within every 10s
In addition to the same spot within every 1min, every 1s, every 250s
Lennart Poettering [Wed, 11 Dec 2013 22:36:59 +0000 (23:36 +0100)]
journald: make sure the stdout/stderr stream fd is removed from epoll before we close it
Lennart Poettering [Wed, 11 Dec 2013 22:31:34 +0000 (23:31 +0100)]
units: don't run readahead done timers in containers
We don't run the collector in the container either, hence we don't need
to stop it either.
Lennart Poettering [Wed, 11 Dec 2013 22:31:07 +0000 (23:31 +0100)]
journald: cache cgroup root path, instead of querying it on every incoming log message
Lennart Poettering [Wed, 11 Dec 2013 22:10:36 +0000 (23:10 +0100)]
journal: when listing logs of a container make sure we don't accidentally show messages from host too
Lennart Poettering [Wed, 11 Dec 2013 21:55:57 +0000 (22:55 +0100)]
journald: cache hostname, boot_id and machine_id fields instead of generating them fresh for each log entry
Lennart Poettering [Wed, 11 Dec 2013 21:17:15 +0000 (22:17 +0100)]
journal: don't go belly up when an stdout/stderr client terminates the connection
Lennart Poettering [Wed, 11 Dec 2013 21:04:03 +0000 (22:04 +0100)]
journal: add ability to browse journals of running OS containers
This adds the new library call sd_journal_open_container() and a new
"-M" switch to journalctl. Particular care is taken that journalctl's
"-b" switch resolves to the current boot ID of the container, not the
host.
Lennart Poettering [Wed, 11 Dec 2013 19:49:43 +0000 (20:49 +0100)]
service: clean up watchdog logic a bit
Lennart Poettering [Wed, 11 Dec 2013 19:13:44 +0000 (20:13 +0100)]
journald: port to sd-event and enable watchdog support
Lennart Poettering [Wed, 11 Dec 2013 17:38:51 +0000 (18:38 +0100)]
man: explain in more detail how SYSTEMD_READY= influences SYSTEMD_WANTS= in udev rules
https://bugzilla.redhat.com/show_bug.cgi?id=1026860
Lennart Poettering [Wed, 11 Dec 2013 17:14:52 +0000 (18:14 +0100)]
event: hook up sd-event with the service watchdog logic
Adds a new call sd_event_set_watchdog() that can be used to hook up the
event loop with the watchdog supervision logic of systemd. If enabled
and $WATCHDOG_USEC is set the event loop will ping the invoking systemd
daemon right after coming back from epoll_wait() but not more often than
$WATCHDOG_USEC/4. The epoll_wait() will sleep no longer than
$WATCHDOG_USEC/4*3, to make sure the service manager is called in time.
This means that setting WatchdogSec= in a .service file and calling
sd_event_set_watchdog() in your daemon is enough to hook it up with the
watchdog logic.
Lennart Poettering [Wed, 11 Dec 2013 03:50:35 +0000 (03:50 +0000)]
event: when handling SIGCHLD of a child process only reap after dispatching event source
That way the even source callback is run with the zombie process still
around so that it can access /proc/$PID/ and similar, and so that it can
be sure that the PID has not been reused yet.
Kay Sievers [Wed, 11 Dec 2013 04:30:46 +0000 (05:30 +0100)]
bus: kdbus - skip only STARTER and IN_QUEUE names for NameOwnerChanged
Zbigniew Jędrzejewski-Szmek [Wed, 11 Dec 2013 02:52:11 +0000 (21:52 -0500)]
activate: clean up inherited descriptors
> [simon@troela server]$ /usr/lib/systemd/systemd-activate -l 9000 main.js
> Assertion 'fd == 3 + count' failed at src/activate/activate.c:115,
> function open_sockets(). Aborting.
> Aborted (core dumped)
> after a bit debuging i found the problem:
> slim appears to leak an fd into all of its children:
> stat /proc/14004/fd/3 (14004 is the pid a random process in my session)
> File: '/proc/14004/fd/3' -> '/var/log/slim.log'
systemd-activate should be robust against the shell (or anything else) leaking
descriptors. Now everything except stdin/stdout/stderr and received sockets
will be closed.
Zbigniew Jędrzejewski-Szmek [Wed, 11 Dec 2013 01:27:14 +0000 (20:27 -0500)]
Update README with test/ requirements
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2013 11:17:01 +0000 (06:17 -0500)]
journald: malloc less when streaming messages
Luke Shumaker [Sat, 23 Nov 2013 09:57:43 +0000 (04:57 -0500)]
ptyfwd: Don't set the output prop of stdin, nor the input props of stdout.
It was calling cfmakeraw(3) on the properties for STDIN_FILENO; cfmakeraw
sets both input and output properties. If (and only if) stdin and stdout
are the same device is this correct. Otherwise, we must change only the
input properties of stdin, and only the output properties of stdout.
Luke Shumaker [Sat, 23 Nov 2013 09:57:42 +0000 (04:57 -0500)]
ptyfwd: Set the size of the PTY base on the size of stdout, not stdin.
David Herrmann [Mon, 2 Dec 2013 22:26:55 +0000 (23:26 +0100)]
core: fix Unit.SetProperties argument parsing
SetProperties has signature "ba(sv)", but the bus_unit_set_properties()
helper already does a enter_container('a', "sv") so we have to skip it in
bus_unit_method_set_properties().
Lennart Poettering [Tue, 10 Dec 2013 23:11:10 +0000 (23:11 +0000)]
bus: do not dispatch incoming method calls that are broacasted to vtables
Lennart Poettering [Tue, 10 Dec 2013 23:06:50 +0000 (23:06 +0000)]
bus: don't generate assert warnings when we detach a detached bus from an event
Lennart Poettering [Tue, 10 Dec 2013 22:50:19 +0000 (22:50 +0000)]
update TODO
Shawn Landden [Mon, 9 Dec 2013 22:08:59 +0000 (14:08 -0800)]
core: epoll and signalfd has been moved into sd event loop
Thomas Hindoe Paaboel Andersen [Tue, 10 Dec 2013 21:30:46 +0000 (22:30 +0100)]
fix scan-build issues
The static analyzer scan-build had a few issues with analysing
parts of systemd.
gpt-auto-generator.c:
scan-build could not find blkid.h. Whether it should be blkid.h or
blkid/blkid.h seems to depend on the version used. We already use
blkid/blkid.h in udev-builtin-blkid.c so it seems safe to use that
here too.
Makefile.am:
Moved some -D's from CFLAGS to CPPFLAGS. I also simplified them a
bit and got rid of a left over DBUS_CFLAGS.
test-cgroup-mask.c/test-sched-prio.c
A variable was added to store the replaced TEST_DIR. When wrapped
in an assert_se TEST_DIR was not replaced in the logged error.
While not an issue introduced in this patch we might as well fix
it up while we are here.
Lennart Poettering [Tue, 10 Dec 2013 20:38:51 +0000 (20:38 +0000)]
bus: generate debug messages when we cannot deliver messages via kdbus
Lennart Poettering [Tue, 10 Dec 2013 20:38:04 +0000 (20:38 +0000)]
bus: when we _unref() a NULL pointer, don't use assert_return()
We support unreffing NULL pointers just fine and we shouldn't pay the
_unlikely_() price for it, not get a debug message if we do, hence let's
not use assert_return() here.
Lukasz Skalski [Tue, 3 Dec 2013 14:17:55 +0000 (15:17 +0100)]
bus: Add bus_kernel_monitor function
Lennart Poettering [Tue, 10 Dec 2013 19:52:16 +0000 (19:52 +0000)]
hashmap: make gcc shut up on old glibcs that lack getauxval()
Lennart Poettering [Tue, 10 Dec 2013 19:51:47 +0000 (19:51 +0000)]
journald: use a bit more cleanup magic!
Lennart Poettering [Tue, 10 Dec 2013 19:31:10 +0000 (19:31 +0000)]
Revert "libsystemd-bus: use assert_return"
This reverts commit
f7e2bd5a8070ba86cba6bcbf7d1c9a8173d846d4.
Most of these checks are not programming errors, but happen during
normal runtime. For example bus_kernel_pop_memfd() is called all the
time on non-kdbus systems and is supposed to quickly fail if kdbus is
not available. However, assert_return() makes this failure
expensive, and hence has no place here. With the most recent change to
assert_return() it will even log a debug message, which should never
happen here.
Lennart Poettering [Tue, 10 Dec 2013 18:53:08 +0000 (18:53 +0000)]
Revert "systemd: add a start job for all units in SYSTEMD_[USER_]WANTS="
This reverts commit
e775289d56ace2f8d23e62ed79316d71332d6d05.
We really should let the dependency logic add jobs for dependencies here
rather than manually adding in jobs, overtaping the real problem.
Lennart Poettering [Tue, 10 Dec 2013 18:53:03 +0000 (18:53 +0000)]
util: check for overflow in greedy_realloc()
Kay Sievers [Tue, 10 Dec 2013 18:26:51 +0000 (19:26 +0100)]
bus: update kdbus.h
Shawn Landden [Tue, 10 Dec 2013 17:28:26 +0000 (09:28 -0800)]
core/manager: remove infinite loop
Lennart Poettering [Tue, 10 Dec 2013 17:01:10 +0000 (17:01 +0000)]
macro: log assertion at debug level in assert_return()
Lennart Poettering [Tue, 10 Dec 2013 16:49:00 +0000 (16:49 +0000)]
bus: introduce new SD_BUS_VTABLE_HIDDEN flag for vtable members
When this flag is set then its member will not be shown in the
introspection data. Also, properties with this flag set will not be
included in GetAll() responses.
Lennart Poettering [Tue, 10 Dec 2013 16:41:39 +0000 (16:41 +0000)]
bus: introduce "trusted" bus concept and encode access control in object vtables
Introduces a new concept of "trusted" vs. "untrusted" busses. For the
latter libsystemd-bus will automatically do per-method access control,
for the former all access is automatically granted. Per-method access
control is encoded in the vtables: by default all methods are only
accessible to privileged clients. If the SD_BUS_VTABLE_UNPRIVILEGED flag
is set for a method it is accessible to unprivileged clients too. By
default whether a client is privileged is determined via checking for
its CAP_SYS_ADMIN capability, but this can be altered via the
SD_BUS_VTABLE_CAPABILITY() macro that can be ORed into the flags field
of the method.
Writable properties are also subject to SD_BUS_VTABLE_UNPRIVILEGED and
SD_BUS_VTABLE_CAPABILITY() for controlling write access to them. Note
however that read access is unrestricted, as PropertiesChanged messages
might send out the values anyway as an unrestricted broadcast.
By default the system bus is set to "untrusted" and the user bus is
"trusted" since per-method access control on the latter is unnecessary.
On dbus1 busses we check the UID of the caller rather than the
configured capability since the capability cannot be determined without
race. On kdbus the capability is checked if possible from the attached
meta-data of a message and otherwise queried from the sending peer.
This also decorates the vtables of the various daemons we ship with
these flags.
Kay Sievers [Tue, 10 Dec 2013 15:36:45 +0000 (16:36 +0100)]
libudev: hwdb - use libudev not systemd logging
Zbigniew Jędrzejewski-Szmek [Tue, 10 Dec 2013 12:39:42 +0000 (07:39 -0500)]
test: wrap $LOOPDEV in quotes
Otherwise things go awry if it is not set ('[ -b ]' is not the same as '[ -b "" ]'!).
Ronny Chevalier [Sat, 7 Dec 2013 23:01:53 +0000 (00:01 +0100)]
test: rework run_qemu
It tries to find a suitable QEMU binary and will use KVM if present.
We can now configure QEMU from outside with 4 variables :
- $QEMU_BIN : path to QEMU's binary
- $KERNEL_APPEND : arguments appended to kernel cmdline
- $KERNEL_BIN : path to a kernel
Default /boot/vmlinuz-$KERNEL_VER
- $INITRD : path to an initramfs
Default /boot/initramfs-${KERNEL_VER}.img
- $QEMU_SMP : number of CPU simulated by QEMU.
Default 1
(from Alexander Graf's script: http://www.spinics.net/lists/kvm/msg72389.html)
Ronny Chevalier [Sat, 7 Dec 2013 23:30:16 +0000 (00:30 +0100)]
test: fix TEST-02-CRYPTSETUP
missing generate_module_dependencies call which prevents dm_mod and
dm_crypt modules to be loaded
Dan McGee [Sun, 8 Dec 2013 17:27:06 +0000 (11:27 -0600)]
Bring bootchart code in line with CODING_STYLE
Use double and not float, as there is little to no benefit.
Dan McGee [Sun, 8 Dec 2013 19:27:05 +0000 (13:27 -0600)]
Ensure unit is journaled for short-lived or oneshot processes
In the time it takes to process incoming log messages, the process we
are logging details for may exit. This means the cgroup data is no
longer available from '/proc'. Unfortunately, the way the code was
structured before, we never log _SYSTEMD_UNIT if we don't have this
cgroup information.
Add an else if case that allows the passed in unit_id to be logged even
if we couldn't capture cgroup information. This ensures a command like
`journalctl -u run-XXX` will return all log messages from a oneshot
process.
Shawn Landden [Mon, 9 Dec 2013 15:04:06 +0000 (07:04 -0800)]
update README to not suggest that systemd works without procfs
Marcel Holtmann [Tue, 10 Dec 2013 11:17:39 +0000 (03:17 -0800)]
hwdb: Update database of Bluetooth company identifiers
Adam Williamson [Tue, 10 Dec 2013 06:02:25 +0000 (22:02 -0800)]
correct name of Tajik kbd layout in kbd-model-map
Adam Williamson [Tue, 10 Dec 2013 05:58:34 +0000 (21:58 -0800)]
drop several entries from kbd-model-map whose kbd layouts do not exist
kbd-model-map was generated from system-config-keyboard's keyboard_models.py.
Several of the kbd layouts referred in that file do not exist and, so far as I
can tell, never did. I believe these entries existed simply to provide the xkb
configuration information for those layouts, and there never were matching kbd
entries; the kbd names were entirely notional, to satisfy the need for some
entry or other in that field.
For systemd, the only function of kbd-model-map is to 'match' kbd and xkb
configurations, so it does not make any sense to maintain entries for cases
where only one or the other exists in this context.
Jason St. John [Tue, 10 Dec 2013 05:10:03 +0000 (00:10 -0500)]
man: improve wording and comma usage in systemd.journal-fields(7)
Improve wording under "Description" and "_KERNEL_DEVICE="
Lukasz Skalski [Mon, 9 Dec 2013 13:09:25 +0000 (14:09 +0100)]
libsystemd-bus: use assert_return
Sébastien Luttringer [Thu, 5 Dec 2013 01:55:05 +0000 (02:55 +0100)]
kernel-install: fix help output
Kernel install doesn't need the second argument on his command line when
removing.
This is correctly documented in the man page.
Lennart Poettering [Tue, 10 Dec 2013 02:31:52 +0000 (03:31 +0100)]
test: add gvariant test for empty messages
Lennart Poettering [Tue, 10 Dec 2013 02:30:51 +0000 (03:30 +0100)]
bus: properly handle empty messages
Lennart Poettering [Tue, 10 Dec 2013 02:04:16 +0000 (03:04 +0100)]
bus: empty gvariant arrays should at least get size 1
Lennart Poettering [Tue, 10 Dec 2013 01:57:46 +0000 (02:57 +0100)]
bus: it's OK to send messages with an empty payload but non-empty signature
THis might happen when sending arrays with 0 enttries, hence this is not
an indication for a problem.
Lennart Poettering [Tue, 10 Dec 2013 01:33:33 +0000 (02:33 +0100)]
bus: update kdbus.h
Lennart Poettering [Tue, 10 Dec 2013 01:33:13 +0000 (02:33 +0100)]
bus: make gcc shut up
Lennart Poettering [Tue, 10 Dec 2013 00:11:43 +0000 (01:11 +0100)]
update TODO
Lennart Poettering [Tue, 10 Dec 2013 00:07:32 +0000 (01:07 +0100)]
bus: beef up zero copy test case
Lennart Poettering [Tue, 10 Dec 2013 00:07:09 +0000 (01:07 +0100)]
bus: only accept gvariant native endian messages via kdbus
Lennart Poettering [Tue, 10 Dec 2013 00:06:52 +0000 (01:06 +0100)]
bus: fix rewinding in gvariant messages
Lennart Poettering [Tue, 10 Dec 2013 00:06:29 +0000 (01:06 +0100)]
bus: fix signature handling when exiting container
Lennart Poettering [Tue, 10 Dec 2013 00:05:23 +0000 (01:05 +0100)]
bus: properly deserialize gvariant fixed size arrays
Lennart Poettering [Tue, 10 Dec 2013 00:04:36 +0000 (01:04 +0100)]
bus: handler empty messages correctly when using gvariant marshalling
Lennart Poettering [Tue, 10 Dec 2013 00:03:41 +0000 (01:03 +0100)]
bus: suppress creating empty parts in messages
Lennart Poettering [Mon, 9 Dec 2013 22:24:17 +0000 (23:24 +0100)]
bus: rely on explicit eof check instead of checking when mapping payload
This allows us to drop quite a bit of code.
Lennart Poettering [Mon, 9 Dec 2013 22:16:14 +0000 (23:16 +0100)]
bus: rearrange bus-message.c function order to keep read and write calls together
Lennart Poettering [Mon, 9 Dec 2013 22:03:21 +0000 (23:03 +0100)]
bus: demarshal gvariant
Lennart Poettering [Mon, 9 Dec 2013 21:51:35 +0000 (22:51 +0100)]
extend CODING_STYLE document a bit
Zbigniew Jędrzejewski-Szmek [Sun, 8 Dec 2013 23:56:16 +0000 (18:56 -0500)]
man: document 'is-enabled' output
https://bugzilla.redhat.com/show_bug.cgi?id=953077
Zbigniew Jędrzejewski-Szmek [Sun, 8 Dec 2013 22:51:25 +0000 (17:51 -0500)]
systemctl: fix 'is-enabled'
Zbigniew Jędrzejewski-Szmek [Sun, 8 Dec 2013 22:50:25 +0000 (17:50 -0500)]
systemctl: simplify argument parsing
Zbigniew Jędrzejewski-Szmek [Sun, 8 Dec 2013 13:16:59 +0000 (08:16 -0500)]
Help output spring cleaning
Use [brackets] only for optional elements.
Use <optional> in XML sources.
Zbigniew Jędrzejewski-Szmek [Sun, 8 Dec 2013 12:46:46 +0000 (07:46 -0500)]
kernel-install: add -h/--help
Zbigniew Jędrzejewski-Szmek [Sun, 8 Dec 2013 11:46:07 +0000 (06:46 -0500)]
Trim TODO
Thomas Hindoe Paaboel Andersen [Sun, 8 Dec 2013 21:28:37 +0000 (22:28 +0100)]
bootchart: remove dead assignment
Dan McGee [Sun, 8 Dec 2013 20:33:45 +0000 (14:33 -0600)]
Fix memory leak in stdout journal streams
Just as 'identifier' is strdup-ed and freed, we need to do the same for
unit_id.
Kay Sievers [Sun, 8 Dec 2013 05:36:39 +0000 (06:36 +0100)]
test: cgroup-util - do not fail if cpu controller is not available
Thomas Hindoe Paaboel Andersen [Sat, 7 Dec 2013 20:53:37 +0000 (21:53 +0100)]
systemctl: fix and refactor wait_for_jobs
wait_for_jobs was ignoring the errors from the jobs stored in r.
It would only ever return whether the call to sd_bus_remove_filter
went ok. This patch changes it to return the first job related error
encountered. If a job related error is found, then the result of the
call to sd_bus_remove_filter is ignored.
wait_for_jobs was a bit hard to read so I split it up to avoid
the goto and deep nesting.
Zbigniew Jędrzejewski-Szmek [Sat, 7 Dec 2013 02:29:55 +0000 (21:29 -0500)]
systemctl: allow globbing in list-<whatever> commands
It is nicer to say 'systemctl list-units ssh\*' then to use grep,
because colouring is preserved and it is easier to match just against
the unit name.
Zbigniew Jędrzejewski-Szmek [Sat, 7 Dec 2013 02:29:55 +0000 (21:29 -0500)]
Get rid of our reimplementation of basename
The only problem is that libgen.h #defines basename to point to it's
own broken implementation instead of the GNU one. This can be fixed
by #undefining basename.
Thomas Hindoe Paaboel Andersen [Fri, 6 Dec 2013 18:48:08 +0000 (19:48 +0100)]
systemctl: check the value from start_unit_one
introduced in
f459b6025f9368116d8c410376546c157314c205
Kay Sievers [Fri, 6 Dec 2013 20:11:37 +0000 (21:11 +0100)]
bus: catch up with latest kdbus changes
Kay Sievers [Fri, 6 Dec 2013 19:19:20 +0000 (20:19 +0100)]
bus: catch up with latest kdbus changes
Bastien Nocera [Fri, 6 Dec 2013 14:51:02 +0000 (15:51 +0100)]
rfkill: Avoid error when state restore is disabled
When the state restore is disabled, we would print:
"Unknown verb: load" instead of simply skipping loading the
state.
Kay Sievers [Fri, 6 Dec 2013 13:31:43 +0000 (14:31 +0100)]
bus: catch up with latest kdbus changes
Kay Sievers [Fri, 6 Dec 2013 12:56:23 +0000 (13:56 +0100)]
busctl: remove ','; it's all one string
Shawn Landden [Thu, 5 Dec 2013 14:20:08 +0000 (06:20 -0800)]
nspawn: fix buggy mount_binds, now works for bind-mounted files