This patch set adds full support the new unified cgroup hierarchy logic
of modern kernels.
A new kernel command line option "systemd.unified_cgroup_hierarchy=1" is
added. If specified the unified hierarchy is mounted to /sys/fs/cgroup
instead of a tmpfs. No further hierarchies are mounted. The kernel
command line option defaults to off. We can turn it on by default as
soon as the kernel's APIs regarding this are stabilized (but even then
downstream distros might want to turn this off, as this will break any
tools that access cgroupfs directly).
It is possibly to choose for each boot individually whether the unified
or the legacy hierarchy is used. nspawn will by default provide the
legacy hierarchy to containers if the host is using it, and the unified
otherwise. However it is possible to run containers with the unified
hierarchy on a legacy host and vice versa, by setting the
$UNIFIED_CGROUP_HIERARCHY environment variable for nspawn to 1 or 0,
respectively.
The unified hierarchy provides reliable cgroup empty notifications for
the first time, via inotify. To make use of this we maintain one
manager-wide inotify fd, and each cgroup to it.
This patch also removes cg_delete() which is unused now.
On kernel 4.2 only the "memory" controller is compatible with the
unified hierarchy, hence that's the only controller systemd exposes when
booted in unified heirarchy mode.
This introduces a new enum for enumerating supported controllers, plus a
related enum for the mask bits mapping to it. The core is changed to
make use of this everywhere.
This moves PID 1 into a new "init.scope" implicit scope unit in the root
slice. This is necessary since on the unified hierarchy cgroups may
either contain subgroups or processes but not both. PID 1 hence has to
move out of the root cgroup (strictly speaking the root cgroup is the
only one where processes and subgroups are still allowed, but in order
to support containers nicey, we move PID 1 into the new scope in all
cases.) This new unit is also used on legacy hierarchy setups. It's
actually pretty useful on all systems, as it can then be used to filter
journal messages coming from PID 1, and so on.
The root slice ("-.slice") is now implicitly created and started (and
does not require a unit file on disk anymore), since
that's where "init.scope" is located and the slice needs to be started
before the scope can.
To check whether we are in unified or legacy hierarchy mode we use
statfs() on /sys/fs/cgroup. If the .f_type field reports tmpfs we are in
legacy mode, if it reports cgroupfs we are in unified mode.
This patch set carefuly makes sure that cgls and cgtop continue to work
as desired.
When invoking nspawn as a service it will implicitly create two
subcgroups in the cgroup it is using, one to move the nspawn process
into, the other to move the actual container processes into. This is
done because of the requirement that cgroups may either contain
processes or other subgroups.
In all cases where the function (or cg_is_empty_recursive()) ignoring
the calling process is actually wrong, as a process keeps a cgroup busy
regardless if its the current one or another. Hence, let's simplify
things and drop the "ignore_self" parameter.
Let's add a way to get the type-specific D-Bus interface of a unit from
either its type or name to src/basic/unit-name.[ch]. That way we can
share it with the client side, where it is useful in tools like cgls or
machinectl.
Also ports over machinectl to make use of this.
This changes log_unit_info() (and friends) to take a real Unit* object
insted of just a unit name as parameter. The call will now prefix all
logged messages with the unit name, thus allowing the unit name to be
dropped from the various passed romat strings, simplifying invocations
drastically, and unifying log output across messages. Also, UNIT= vs.
USER_UNIT= is now derived from the Manager object attached to the Unit
object, instead of getpid(). This has the benefit of correcting the
field for --test runs.
Also contains a couple of other logging improvements:
- Drops a couple of strerror() invocations in favour of using %m.
- Not only .mount units now warn if a symlinks exist for the mount
point already, .automount units do that too, now.
- A few invocations of log_struct() that didn't actually pass any
additional structured data have been replaced by simpler invocations
of log_unit_info() and friends.
- For structured data a new LOG_UNIT_MESSAGE() macro has been added,
that works like LOG_MESSAGE() but prefixes the message with the unit
name. Similar, there's now LOG_LINK_MESSAGE() and
LOG_NETDEV_MESSAGE().
- For structured data new LOG_UNIT_ID(), LOG_LINK_INTERFACE(),
LOG_NETDEV_INTERFACE() macros have been added that generate the
necessary per object fields. The old log_unit_struct() call has been
removed in favour of these new macros used in raw log_struct()
invocations. In addition to removing one more function call this
allows generated structured log messages that contain two object
fields, as necessary for example for network interfaces that are
joined into another network interface, and whose messages shall be
indexed by both.
- The LOG_ERRNO() macro has been removed, in favour of
log_struct_errno(). The latter has the benefit of ensuring that %m in
format strings is properly resolved to the specified error number.
- A number of logging messages have been converted to use
log_unit_info() instead of log_info()
- The client code in sysv-generator no longer #includes core code from
src/core/.
- log_unit_full_errno() has been removed, log_unit_full() instead takes
an errno now, too.
- log_unit_info(), log_link_info(), log_netdev_info() and friends, now
avoid double evaluation of their parameters
Because the order of coldplugging is not defined, we can reference a
not-yet-coldplugged unit and read its state while it has not yet been
set to a meaningful value.
This way, already active units may get started again.
We fix this by deferring such actions until all units have been at
least somehow coldplugged.
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=88401
This patch removes includes that are not used. The removals were found with
include-what-you-use which checks if any of the symbols from a header is
in use.
Unit _start() and _stop() implementations can fail with -EAGAIN to delay
execution temporarily. Thus, we should not output status messages before
invoking these calls, but after, and only when we know that the
invocation actually made a change.
As a followup to 086891e5c1 "log: add an "error" parameter to all
low-level logging calls and intrdouce log_error_errno() as log calls
that take error numbers", use sed to convert the simple cases to use
the new macros:
find . -name '*.[ch]' | xargs sed -r -i -e \
's/log_(debug|info|notice|warning|error|emergency)\("(.*)%s"(.*), strerror\(-([a-zA-Z_]+)\)\);/log_\1_errno(-\4, "\2%m"\3);/'
Multi-line log_*() invocations are not covered.
And we also should add log_unit_*_errno().
- Rename log_meta() → log_internal(), to follow naming scheme of most
other log functions that are usually invoked through macros, but never
directly.
- Rename log_info_object() to log_object_info(), simply because the
object should be before any other parameters, to follow OO-style
programming style.
Let's automatically initialize the kill, exec and cgroup contexts of the
various unit types when the object is constructed, instead of
invididually in type-specific code.
Also, when PrivateDevices= is set, set DevicePolicy= to closed.
Previously the returned object of constructor functions where sometimes
returned as last, sometimes as first and sometimes as second parameter.
Let's clean this up a bit. Here are the new rules:
1. The object the new object is derived from is put first, if there is any
2. The object we are creating will be returned in the next arguments
3. This is followed by any additional arguments
Rationale:
For functions that operate on an object we always put that object first.
Constructors should probably not be too different in this regard. Also,
if the additional parameters might want to use varargs which suggests to
put them last.
Note that this new scheme only applies to constructor functions, not to
all other functions. We do give a lot of freedom for those.
Note that this commit only changes the order of the new functions we
added, for old ones we accept the wrong order and leave it like that.
When a process dies that we can associate with a specific unit, start
watching all other processes of that unit, so that we can associate
those processes with the unit too.
Also, for service units start doing this as soon as we get the first
SIGCHLD for either control or main process, so that we can follow the
processes of the service from one to the other, as long as process that
remain are processes of the ones we watched that died and got reassigned
to us as parent.
Similar, for scope units start doing this as soon as the scope
controller abandons the unit, and thus management entirely reverts to
systemd. To abandon a unit introduce a new Abandon() scope unit method
call.
By specifiy a Controller property when creating the scope a client can
specify a bus name that will be notified with a RequestStop bus signal
when the scope has been asked to shut down, instead of sending SIGTERM
to the scope processes themselves.
https://bugzilla.redhat.com/show_bug.cgi?id=1032695
Given that we now have KillMode=mixed where SIGTERM might kill a smaller
set than SIGKILL we need to make sure to always go explicitly throught
the SIGKILL state to get the right end result.
This patch converts PID 1 to libsystemd-bus and thus drops the
dependency on libdbus. The only remaining code using libdbus is a test
case that validates our bus marshalling against libdbus' marshalling,
and this dependency can be turned off.
This patch also adds a couple of things to libsystem-bus, that are
necessary to make the port work:
- Synthesizing of "Disconnected" messages when bus connections are
severed.
- Support for attaching multiple vtables for the same interface on the
same path.
This patch also fixes the SetDefaultTarget() and GetDefaultTarget() bus
calls which used an inappropriate signature.
As a side effect we will now generate PropertiesChanged messages which
carry property contents, rather than just invalidation information.
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.
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.
"Scope" units are very much like service units, however with the
difference that they are created from pre-existing processes, rather
than processes that systemd itself forks off. This means they are
generated programmatically via the bus API as transient units rather
than from static configuration read from disk. Also, they do not provide
execution-time parameters, as at the time systemd adds the processes to
the scope unit they already exist and the parameters cannot be applied
anymore.
The primary benefit of this new unit type is to create arbitrary cgroups
for worker-processes forked off an existing service.
This commit also adds a a new mode to "systemd-run" to run the specified
processes in a scope rather then a transient service.
Lennart Poettering