We use our cgroup APIs in various contexts, including from our libraries
sd-login, sd-bus. As we don#t control those environments we can't rely
that the unified cgroup setup logic succeeds, and hence really shouldn't
assert on it.
This more or less reverts 415fc41cea.
cg_[all_]unified() test whether a specific controller or all controllers are on
the unified hierarchy. While what's being asked is a simple binary question,
the callers must assume that the functions may fail any time, which
unnecessarily complicates their usages. This complication is unnecessary.
Internally, the test result is cached anyway and there are only a few places
where the test actually needs to be performed.
This patch simplifies cg_[all_]unified().
* cg_[all_]unified() are updated to return bool. If the result can't be
decided, assertion failure is triggered. Error handlings from their callers
are dropped.
* cg_unified_flush() is updated to calculate the new result synchrnously and
return whether it succeeded or not. Places which need to flush the test
result are updated to test for failure. This ensures that all the following
cg_[all_]unified() tests succeed.
* Places which expected possible cg_[all_]unified() failures are updated to
call and test cg_unified_flush() before calling cg_[all_]unified(). This
includes functions used while setting up mounts during boot and
manager_setup_cgroup().
Accept AF_VSOCK listen addresses in socket unit files. Both guest and
host can now take advantage of socket activation.
The QEMU guest agent has recently been modified to support socket
activation and can run over AF_VSOCK with this patch.
sockaddr_port() either returns a >= 0 port number or a negative errno.
This works for AF_INET and AF_INET6 because port ranges are only 16-bit.
In AF_VSOCK ports are 32-bit so an int cannot represent all port number
and negative errnos. Separate the port and the return code.
This patch ensures that each system service gets its own session kernel keyring
automatically, and implicitly. Without this a keyring is allocated for it
on-demand, but is then linked with the user's kernel keyring, which is OK
behaviour for logged in users, but not so much for system services.
With this change each service gets a session keyring that is specific to the
service and ceases to exist when the service is shut down. The session keyring
is not linked up with the user keyring and keys hence only search within the
session boundaries by default.
(This is useful in a later commit to store per-service material in the keyring,
for example the invocation ID)
(With input from David Howells)
This monopolizes unit file specifier expansion in load-fragment.c, and removes
it from socket.c + service.c. This way expansion becomes an operation done exclusively at time of loading unit files.
Previously specifiers were resolved for all settings during loading of unit
files with the exception of ExecStart= and friends which were resolved in
socket.c and service.c. With this change the latter is also moved to the
loading of unit files.
Fixes: #3061
We stay in the SERVICE_START while no READY=1 notification message has
been received. When we are in the SERVICE_START_POST state, we have
already received a ready notification. Hence we should not fail when the
cgroup becomes empty in that state.
Before this commit, when the main process of a Type=notify service exits the
service would enter a running state without passing through the startup post
state. This meant ExecStartPost= from being executed and allowed follow-up
units to start too early (before the ready notification).
Additionally, when RemainAfterExit=yes is used on a Type=notify service, the
exit status of the main process would be disregarded.
After this commit, an unsuccessful exit of the main process of a Type=notify
service puts the unit in a failed state. A successful exit is inconsequential
in case RemainAfterExit=yes. Otherwise, when no ready notification has been
received, the unit is put in a failed state because it has never been active.
When all processes in the cgroup of a Type=notify service are gone and no ready
notification has been received yet, the unit is also put in a failed state.
Introduce a SERVICE_FAILURE_PROTOCOL error type for when a service does
not follow the protocol.
This error type is used when a pid file is expected, but not delivered.
It's rather hard to parse the confirmation messages (enabled with
systemd.confirm_spawn=true) amongst the status messages and the kernel
ones (if enabled).
This patch gives the possibility to the user to redirect the confirmation
message to a different virtual console, either by giving its name or its path,
so those messages are separated from the other ones and easier to read.
We don't have plural in the name of any other -util files and this
inconsistency trips me up every time I try to type this file name
from memory. "formats-util" is even hard to pronounce.
Since service_add_fd_store() already does the check, remove the redundant check
from service_add_fd_store_set().
Also, print a warning when repopulating FDStore after daemon-reexec and we hit
the limit. This is a user visible issue, so we should not discard fds silently.
(Note that service_deserialize_item is impacted by the return value from
service_add_fd_store(), but we rely on the general error message, so the caller
does not need to be modified, and does not show up in the diff.)
We would close all the stored fds in service_release_resources(), which of
course broke the whole concept of storing fds over service restart.
Fixes#4408.
It's a common pattern, so add a helper for it. A macro is necessary
because a function that takes a pointer to a pointer would be type specific,
similarly to cleanup functions. Seems better to use a macro.
SIGTERM should be considered a clean exit code for daemons (i.e. long-running
processes, as a daemon without SIGTERM handler may be shut down without issues
via SIGTERM still) while it should not be considered a clean exit code for
commands (i.e. short-running processes).
Let's add two different clean checking modes for this, and use the right one at
the appropriate places.
Fixes: #4275
Let's get rid of is_clean_exit_lsb(), let's move the logic for the special
handling of the two LSB exit codes into the sysv-generator by writing out
appropriate SuccessExitStatus= lines if the LSB header exists. This is not only
semantically more correct, bug also fixes a bug as the code in service.c that
chose between is_clean_exit_lsb() and is_clean_exit() based this check on
whether a native unit files was available for the unit. However, that check was
bogus since a long time, since the SysV generator was introduced and native
SysV script support was removed from PID 1, as in that case a unit file always
existed.
This adds a new invocation ID concept to the service manager. The invocation ID
identifies each runtime cycle of a unit uniquely. A new randomized 128bit ID is
generated each time a unit moves from and inactive to an activating or active
state.
The primary usecase for this concept is to connect the runtime data PID 1
maintains about a service with the offline data the journal stores about it.
Previously we'd use the unit name plus start/stop times, which however is
highly racy since the journal will generally process log data after the service
already ended.
The "invocation ID" kinda matches the "boot ID" concept of the Linux kernel,
except that it applies to an individual unit instead of the whole system.
The invocation ID is passed to the activated processes as environment variable.
It is additionally stored as extended attribute on the cgroup of the unit. The
latter is used by journald to automatically retrieve it for each log logged
message and attach it to the log entry. The environment variable is very easily
accessible, even for unprivileged services. OTOH the extended attribute is only
accessible to privileged processes (this is because cgroupfs only supports the
"trusted." xattr namespace, not "user."). The environment variable may be
altered by services, the extended attribute may not be, hence is the better
choice for the journal.
Note that reading the invocation ID off the extended attribute from journald is
racy, similar to the way reading the unit name for a logging process is.
This patch adds APIs to read the invocation ID to sd-id128:
sd_id128_get_invocation() may be used in a similar fashion to
sd_id128_get_boot().
PID1's own logging is updated to always include the invocation ID when it logs
information about a unit.
A new bus call GetUnitByInvocationID() is added that allows retrieving a bus
path to a unit by its invocation ID. The bus path is built using the invocation
ID, thus providing a path for referring to a unit that is valid only for the
current runtime cycleof it.
Outlook for the future: should the kernel eventually allow passing of cgroup
information along AF_UNIX/SOCK_DGRAM messages via a unique cgroup id, then we
can alter the invocation ID to be generated as hash from that rather than
entirely randomly. This way we can derive the invocation race-freely from the
messages.
This adds the boolean RemoveIPC= setting to service, socket, mount and swap
units (i.e. all unit types that may invoke processes). if turned on, and the
unit's user/group is not root, all IPC objects of the user/group are removed
when the service is shut down. The life-cycle of the IPC objects is hence bound
to the unit life-cycle.
This is particularly relevant for units with dynamic users, as it is essential
that no objects owned by the dynamic users survive the service exiting. In
fact, this patch adds code to imply RemoveIPC= if DynamicUser= is set.
In order to communicate the UID/GID of an executed process back to PID 1 this
adds a new "user lookup" socket pair, that is inherited into the forked
processes, and closed before the exec(). This is needed since we cannot do NSS
from PID 1 due to deadlock risks, However need to know the used UID/GID in
order to clean up IPC owned by it if the unit shuts down.
Currently, systemd uses either the legacy hierarchies or the unified hierarchy.
When the legacy hierarchies are used, systemd uses a named legacy hierarchy
mounted on /sys/fs/cgroup/systemd without any kernel controllers for process
management. Due to the shortcomings in the legacy hierarchy, this involves a
lot of workarounds and complexities.
Because the unified hierarchy can be mounted and used in parallel to legacy
hierarchies, there's no reason for systemd to use a legacy hierarchy for
management even if the kernel resource controllers need to be mounted on legacy
hierarchies. It can simply mount the unified hierarchy under
/sys/fs/cgroup/systemd and use it without affecting other legacy hierarchies.
This disables a significant amount of fragile workaround logics and would allow
using features which depend on the unified hierarchy membership such bpf cgroup
v2 membership test. In time, this would also allow deleting the said
complexities.
This patch updates systemd so that it prefers the unified hierarchy for the
systemd cgroup controller hierarchy when legacy hierarchies are used for kernel
resource controllers.
* cg_unified(@controller) is introduced which tests whether the specific
controller in on unified hierarchy and used to choose the unified hierarchy
code path for process and service management when available. Kernel
controller specific operations remain gated by cg_all_unified().
* "systemd.legacy_systemd_cgroup_controller" kernel argument can be used to
force the use of legacy hierarchy for systemd cgroup controller.
* nspawn: By default nspawn uses the same hierarchies as the host. If
UNIFIED_CGROUP_HIERARCHY is set to 1, unified hierarchy is used for all. If
0, legacy for all.
* nspawn: arg_unified_cgroup_hierarchy is made an enum and now encodes one of
three options - legacy, only systemd controller on unified, and unified. The
value is passed into mount setup functions and controls cgroup configuration.
* nspawn: Interpretation of SYSTEMD_CGROUP_CONTROLLER to the actual mount
option is moved to mount_legacy_cgroup_hierarchy() so that it can take an
appropriate action depending on the configuration of the host.
v2: - CGroupUnified enum replaces open coded integer values to indicate the
cgroup operation mode.
- Various style updates.
v3: Fixed a bug in detect_unified_cgroup_hierarchy() introduced during v2.
v4: Restored legacy container on unified host support and fixed another bug in
detect_unified_cgroup_hierarchy().
A following patch will update cgroup handling so that the systemd controller
(/sys/fs/cgroup/systemd) can use the unified hierarchy even if the kernel
resource controllers are on the legacy hierarchies. This would require
distinguishing whether all controllers are on cgroup v2 or only the systemd
controller is. In preparation, this patch renames cg_unified() to
cg_all_unified().
This patch doesn't cause any functional changes.
This fixes an issue during reexec — the count of connections would be lost:
[zbyszek@fedora-rawhide ~]$ systemctl status testlimit.socket | grep Connected
Accepted: 1; Connected: 1
[zbyszek@fedora-rawhide ~]$ sudo systemctl daemon-reexec
[zbyszek@fedora-rawhide ~]$ systemctl status testlimit.socket | grep Connected
Accepted: 1; Connected: 0
With the patch, Connected count is preserved.
Also add "Accept Socket" to the dump output for services.
Previously, the result value of a unit was overriden with each failure that
took place, so that the result always reported the last failure that took
place.
With this commit this is changed, so that the first failure taking place is
stored instead. This should normally not matter much as multiple failures are
sufficiently uncommon. However, it improves one behaviour: if we send SIGABRT
to a service due to a watchdog timeout, then this currently would be reported
as "coredump" failure, rather than the "watchodg" failure it really is. Hence,
in order to report information about the type of the failure, and not about
the effect of it, let's change this from all unit type to store the first, not
the last failure.
This addresses the issue pointed out here:
https://github.com/systemd/systemd/pull/3818#discussion_r73433520
This should simplify monitoring tools for services, by passing the most basic
information about service result/exit information via environment variables,
thus making it unnecessary to retrieve them explicitly via the bus.
The ExecParameters structure contains a number of bit-flags, that were so far
exposed as bool:1, change this to a proper, single binary bit flag field. This
makes things a bit more expressive, and is helpful as we add more flags, since
these booleans are passed around in various callers, for example
service_spawn(), whose signature can be made much shorter now.
Not all bit booleans from ExecParameters are moved into the flags field for
now, but this can be added later.
This adds a new boolean setting DynamicUser= to service files. If set, a new
user will be allocated dynamically when the unit is started, and released when
it is stopped. The user ID is allocated from the range 61184..65519. The user
will not be added to /etc/passwd (but an NSS module to be added later should
make it show up in getent passwd).
For now, care should be taken that the service writes no files to disk, since
this might result in files owned by UIDs that might get assigned dynamically to
a different service later on. Later patches will tighten sandboxing in order to
ensure that this cannot happen, except for a few selected directories.
A simple way to test this is:
systemd-run -p DynamicUser=1 /bin/sleep 99999
Let's lot at LOG_NOTICE about any processes that we are going to
SIGKILL/SIGABRT because clean termination of them didn't work.
This turns the various boolean flag parameters to cg_kill(), cg_migrate() and
related calls into a single binary flags parameter, simply because the function
now gained even more parameters and the parameter listed shouldn't get too
long.
Logging for killing processes is done either when the kill signal is SIGABRT or
SIGKILL, or on explicit request if KILL_TERMINATE_AND_LOG instead of LOG_TERMINATE
is passed. This isn't used yet in this patch, but is made use of in a later
patch.
Add sd_notify() parameter to change watchdog_usec during runtime.
Application can change watchdog_usec value by
sd_notify like this. Example. sd_notify(0, "WATCHDOG_USEC=20000000").
To reset watchdog_usec as configured value in service file,
restart service.
Notice.
sd_event is not currently supported. If application uses
sd_event_set_watchdog, or sd_watchdog_enabled, do not use
"WATCHDOG_USEC" option through sd_notify.
Delete the dbus1 generator and some critical wiring. This prevents
kdbus from being loaded or detected. As such, it will never be used,
even if the user still has a useful kdbus module loaded on their system.
Sort of fixes#3480. Not really, but it's better than the current state.
Let's move the enforcement of the per-unit start limit from unit.c into the
type-specific files again. For unit types that know a concept of "result" codes
this allows us to hook up the start limit condition to it with an explicit
result code. Also, this makes sure that the state checks in clal like
service_start() may be done before the start limit is checked, as the start
limit really should be checked last, right before everything has been verified
to be in order.
The generic start limit logic is left in unit.c, but the invocation of it is
moved into the per-type files, in the various xyz_start() functions, so that
they may place the check at the right location.
Note that this change drops the enforcement entirely from device, slice, target
and scope units, since these unit types generally may not fail activation, or
may only be activated a single time. This is also documented now.
Note that restores the "start-limit-hit" result code that existed before
6bf0f408e4 already in the service code. However,
it's not introduced for all units that have a result code concept.
Fixes#3166.
We always call one after the other anyway, and this way service_set_socket_fd()
and service_close_socket_fd() nicely match each other as one undoes the effect
of the other.
In service_set_socket_fd(), let's make sure that if we can't add the requested
dependencies we take no possession of the passed connection fd.
This way, we follow the strict rule: we take possession of the passed fd on
success, but on failure we don't, and the fd remains in possession of the
caller.
This adds a new GetProcesses() bus call to the Unit object which returns an
array consisting of all PIDs, their process names, as well as their full cgroup
paths. This is then used by "systemctl status" to show the per-unit process
tree.
This has the benefit that the client-side no longer needs to access the
cgroupfs directly to show the process tree of a unit. Instead, it now uses this
new API, which means it also works if -H or -M are used correctly, as the
information from the specific host is used, and not the one from the local
system.
Fixes: #2945
This replaces the old function call manager_is_reloading_or_reexecuting() which
was used only at very few places. Use the new macro wherever we check whether
we are reloading. This should hopefully make things a bit more readable, given
the nature of Manager:n_reloading being a counter.
Previously, we had two enums ManagerRunningAs and UnitFileScope, that were
mostly identical and converted from one to the other all the time. The latter
had one more value UNIT_FILE_GLOBAL however.
Let's simplify things, and remove ManagerRunningAs and replace it by
UnitFileScope everywhere, thus making the translation unnecessary. Introduce
two new macros MANAGER_IS_SYSTEM() and MANAGER_IS_USER() to simplify checking
if we are running in one or the user context.
Throughout the tree there's spurious use of spaces separating ++ and --
operators from their respective operands. Make ++ and -- operator
consistent with the majority of existing uses; discard the spaces.
We call dual_timestamp_serialize() only if the s->watchdog_timestamp is
set. But the dual_timestamp_serialize() already checks a given dual
timestamp by the call of the dual_timestamp_is_set(). So we can remove
this check safely.
This moves the StartLimitBurst=, StartLimitInterval=, StartLimitAction=, RebootArgument= from the [Service] section
into the [Unit] section of unit files, and thus support it in all unit types, not just in services.
This way we can enforce the start limit much earlier, in particular before testing the unit conditions, so that
repeated start-up failure due to failed conditions is also considered for the start limit logic.
For compatibility the four options may also be configured in the [Service] section still, but we only document them in
their new section [Unit].
This also renamed the socket unit failure code "service-failed-permanent" into "service-start-limit-hit" to express
more clearly what it is about, after all it's only triggered through the start limit being hit.
Finally, the code in busname_trigger_notify() and socket_trigger_notify() is altered to become more alike.
Fixes: #2467
This makes sure we never run two control processes at the same time, we cannot keep track off.
This introduces a slight change of behaviour but cleans up the definition of ExecStop= and ExecStopPost=. The former is
now invoked only if the service managed to start-up correctly. The latter is called even if start-up failed half-way.
Thus, ExecStopPost= may be used as clean-up step for both successful and failed start-up attempts, but ExecStop='s
purpose is clearly defined as being responsible for shutting down the service and nothing else.
The precise behaviour of this was not documented yet. This commit adds the necessary docs.
Fixes: #1254
This clean-ups timeout handling in PID 1. Specifically, instead of storing 0 in internal timeout variables as
indication for a disabled timeout, use USEC_INFINITY which is in-line with how we do this in the rest of our code
(following the logic that 0 means "no", and USEC_INFINITY means "never").
This also replace all usec_t additions with invocations to usec_add(), so that USEC_INFINITY is properly propagated,
and sd-event considers it has indication for turning off the event source.
This also alters the deserialization of the units to restart timeouts from the time they were originally started from.
Before this patch timeouts would be restarted beginning with the time of the deserialization, which could lead to
artificially prolonged timeouts if a daemon reload took place.
Finally, a new RuntimeMaxSec= setting is introduced for service units, that specifies a maximum runtime after which a
specific service is forcibly terminated. This is useful to put time limits on time-intensive processing jobs.
This also simplifies the various xyz_spawn() calls of the various types in that explicit distruction of the timers is
removed, as that is done anyway by the state change handlers, and a state change is always done when the xyz_spawn()
calls fail.
Fixes: #2249
During daemon-reload, PID1 temporarly loses its DBus connection, so there's
a small window in which all signals sent by dbus-daemon are lost.
This is a problem, since we rely on the NameOwnerChanged signals in order to
consider a service with Type=dbus fully started or terminated, respectively.
In order to fix this, a rewrite of bus_list_names() is necessary. We used
to walk the current list of names on the bus, and blindly triggered the
bus_name_owner_change() callback on each service, providing the actual name
as current owner. This implementation has a number of problems:
* We cannot detect if the the name was moved from one owner to the other
while we were reloading
* We don't notify services which missed the name loss signal
* Providing the actual name as current owner is a hack, as the comment also
admits.
To fix this, this patch carries the following changes:
* Track the name of the current bus name owner, and (de-)serialize it
during reload. This way, we can detect changes.
* In bus_list_names(), walk the list of bus names we're interested in
first, and then see if the name is active on the bus. If it is,
check it it's still the same as it used to be, and synthesize
NameOwnerChanged signals for the name add and/or loss.
This should fully synchronize the current name list with the internal
state of all services.
When constructing the journal filename to store logs from a remote host, remove the port of the tcp connection, as the port will change with every reboot/connection loss between sender/reveiver machines. Having the port in the filename will cause a new journal file to be created for every reboot or connection loss.
For the implementation, a new argument "bool include_port" is added to the getpeername_pretty() function. This is passed to the sockaddr_pretty() function. The value of the include_port argument is set to true in all calls of getpeername_pretty(), except for 2 calls in journal-remote.c, where it is set to false.
GLIB has recently started to officially support the gcc cleanup
attribute in its public API, hence let's do the same for our APIs.
With this patch we'll define an xyz_unrefp() call for each public
xyz_unref() call, to make it easy to use inside a
__attribute__((cleanup())) expression. Then, all code is ported over to
make use of this.
The new calls are also documented in the man pages, with examples how to
use them (well, I only added docs where the _unref() call itself already
had docs, and the examples, only cover sd_bus_unrefp() and
sd_event_unrefp()).
This also renames sd_lldp_free() to sd_lldp_unref(), since that's how we
tend to call our destructors these days.
Note that this defines no public macro that wraps gcc's attribute and
makes it easier to use. While I think it's our duty in the library to
make our stuff easy to use, I figure it's not our duty to make gcc's own
features easy to use on its own. Most likely, client code which wants to
make use of this should define its own:
#define _cleanup_(function) __attribute__((cleanup(function)))
Or similar, to make the gcc feature easier to use.
Making this logic public has the benefit that we can remove three header
files whose only purpose was to define these functions internally.
See #2008.
Now that we don't have RequiresOverridable= and RequisiteOverridable=
dependencies anymore, we can get rid of tracking the "override" boolean
for jobs in the job engine, as it serves no purpose anymore.
While we are at it, fix some error messages we print when invoking
functions that take the override parameter.
With this change services by default will no longer require
basic.target, but instead only after it it via After=basic.target.
However, they will still Require= on sysinit.target.
This has the benefit that when booting into emergency mode it is
relatively safe to actviate individual services, as this will not pull
the entirety of basic.target anymore, thus avoid everything listed in
sockets.target and suchlike. However, during the usual boot no change
should be noticed.
Snapshots were never useful or used for anything. Many systemd
developers that I spoke to at systemd.conf2015, didn't even know they
existed, so it is fairly safe to assume that this type can be deleted
without harm.
The fundamental problem with snapshots is that the state of the system
is dynamic, devices come and go, users log in and out, timers fire...
and restoring all units to some state from the past would "undo"
those changes, which isn't really possible.
Tested by creating a snapshot, running the new binary, and checking
that the transition did not cause errors, and the snapshot is gone,
and snapshots cannot be created anymore.
New systemctl says:
Unknown operation snapshot.
Old systemctl says:
Failed to create snapshot: Support for snapshots has been removed.
IgnoreOnSnaphost settings are warned about and ignored:
Support for option IgnoreOnSnapshot= has been removed and it is ignored
http://lists.freedesktop.org/archives/systemd-devel/2015-November/034872.html
Since 5fd9b2c546 passing a pid of 0 to
pid_is_unwaited() and pid_is_live() is considered as a request on the
current process, similar how the other calls in process-util.c handle a
PID of 0. This broke service.c, which passes a 0 PID and expects it to
be considered an unwaited process.
This fix make sure we can boot again.
There are more than enough calls doing string manipulations to deserve
its own files, hence do something about it.
This patch also sorts the #include blocks of all files that needed to be
updated, according to the sorting suggestions from CODING_STYLE. Since
pretty much every file needs our string manipulation functions this
effectively means that most files have sorted #include blocks now.
Also touches a few unrelated include files.
When starting a transient service, allow setting stdin/stdout/stderr fds
for it, by passing them in via the bus.
This also simplifies some of the serialization code for units.
This adds support for naming file descriptors passed using socket
activation. The names are passed in a new $LISTEN_FDNAMES= environment
variable, that matches the existign $LISTEN_FDS= one and contains a
colon-separated list of names.
This also adds support for naming fds submitted to the per-service fd
store using FDNAME= in the sd_notify() message.
This also adds a new FileDescriptorName= setting for socket unit files
to set the name for fds created by socket units.
This also adds a new call sd_listen_fds_with_names(), that is similar to
sd_listen_fds(), but also returns the names of the fds.
systemd-activate gained the new --fdname= switch to specify a name for
testing socket activation.
This is based on #1247 by Maciej Wereski.
Fixes#1247.
Much like the result of the service itself we should not reset the
reload result unless we actually start from the beginning, so that
clients can query it at any time.
Specifically, let's reset the result states only when we begin with a
start operation (for both the main result, and the reload result), when
we begin with a reload operation (only for the load result), or when the
use explicitly asks for that via "systemctl reset-failed".
This is a more generic fix for #1447.
Fixes#1447.
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.
Rework the "service is good" check, to only check the cgroup state if we
really need to instead of always.
This allows us to suppress going to the cgroupfs for an empty check for
the majority of services.
No functional change.
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 reverts commit d4d00020d6. The idea of
the commit is broken and needs to be reworked. We really cannot reduce
the bus-addresses to a single address. We always will have systemd with
native clients and legacy clients at the same time, so we also need both
addresses at the same time.
We should not fall back to dbus-1 and connect to the proxy when kdbus
returns an error that indicates that kdbus is running but just does not
accept new connections because of quota limits or something similar.
Using is_kdbus_available() in libsystemd/ requires it to move from
shared/ to libsystemd/.
Based on a patch from David Herrmann:
https://github.com/systemd/systemd/pull/886
Return 1 from *_reload() methods to signify "we did something", just
like in *_start(). This causes "Reloading foo..." messages to be logged.
"Reloaded foo." messages are already logged.
./configure --enable/disable-kdbus can be used to set the default
behavior regarding kdbus.
If no kdbus kernel support is available, dbus-dameon will be used.
With --enable-kdbus, the kernel command line option "kdbus=0" can
be used to disable kdbus.
With --disable-kdbus, the kernel command line option "kdbus=1" is
required to enable kdbus support.
It's primarily just a property of the Manager object after all, and we
try to refer to PID 1 as "manager" instead of "systemd", hence let's to
stick to this here too.
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
If there's no remaining process to kill, skip the SIGKILL states if
SIGKILL is disabled.
Effectively this doesn't change much since if there's nothing to kill
with SIGTERM or SIGABRT then there's also nothing to kill with SIGKILL.
However, this avoids confusion with the state engine jumping through
SIGKILL states for no reason...
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.
After all it is now much more like strjoin() than strappend(). At the
same time, add support for NULL sentinels, even if they are normally not
necessary.
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.
With this change it is possible to send file descriptors to PID 1, via
sd_pid_notify_with_fds() which PID 1 will store individually for each
service, and pass via the usual fd passing logic on next invocation.
This is useful for enable daemon reload schemes where daemons serialize
their state to /run, push their fds into PID 1 and terminate, restoring
their state on next start from the data in /run and passed in from PID
1.
The fds are kept by PID 1 as long as no POLLHUP or POLLERR is seen on
them, and the service they belong to are either not dead or failed, or
have a job queued.
- 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.
For priviliged units this resource control property ensures that the
processes have all controllers systemd manages enabled.
For unpriviliged services (those with User= set) this ensures that
access rights to the service cgroup is granted to the user in question,
to create further subgroups. Note that this only applies to the
name=systemd hierarchy though, as access to other controllers is not
safe for unpriviliged processes.
Delegate=yes should be set for container scopes where a systemd instance
inside the container shall manage the hierarchies below its own cgroup
and have access to all controllers.
Delegate=yes should also be set for user@.service, so that systemd
--user can run, controlling its own cgroup tree.
This commit changes machined, systemd-nspawn@.service and user@.service
to set this boolean, in order to ensure that container management will
just work, and the user systemd instance can run fine.
Clean up the function namespace by renaming the following:
sd_bus_get_owner_uid() → sd_bus_get_name_creds_uid()
sd_bus_get_owner_machine_id() → sd_bus_get_name_machine_id()
sd_bus_get_peer_creds() → sd_bus_get_owner_creds()
SERVICE_STOP would mean we're running the ExecStop command. That's not
the case with "STOPPING=1".
Instead we should enter the same state as if we just told the service
to stop, i.e. SERVICE_STOP_SIGTERM.
This fixes a bug where voluntarily exiting services would hang around in
deactivating state until timeout.
This makes possible to spawn service instances triggered by socket with
MLS/MCS SELinux labels which are created based on information provided by
connected peer.
Implementation of label_get_child_mls_label derived from xinetd.
Reviewed-by: Paul Moore <pmoore@redhat.com>
If BusPolicy= was passed, the parser function will have created
an ExecContext->bus_endpoint object, along with policy information.
In that case, create a kdbus endpoint, and pass its path name to the
namespace logic, to it will be mounted over the actual 'bus' node.
At endpoint creation time, no policy is updloaded. That is done after
fork(), through a separate call. This is necessary because we don't
know the real uid of the process earlier than that.
This way, the list of arguments to that function gets more comprehensive,
and we can get around passing lots of NULL and 0 arguments from socket.c,
swap.c and mount.c.
It also allows for splitting up the code in exec_spawn().
While at it, make ExecContext const in execute.c.
When this system-wide start-up timeout is hit we execute one of the
failure actions already implemented for services that fail.
This should not only be useful on embedded devices, but also on laptops
which have the power-button reachable when the lid is closed. This
devices, when in a backpack might get powered on by accident due to the
easily reachable power button. We want to make sure that the system
turns itself off if it starts up due this after a while.
When the system manages to fully start-up logind will suspend the
machine by default if the lid is closed. However, in some cases we don't
even get as far as logind, and the boot hangs much earlier, for example
because we ask for a LUKS password that nobody ever enters.
Yeah, this is a real-life problem on my Yoga 13, which has one of those
easily accessible power buttons, even if the device is closed.
This is useful for services that simply want to run something on
shutdown, but not at bootup. They should only set ExecStop= but leave
ExecStart= unset.
Unlike strv_find_prefix() the new call will return a pointer to the
suffix of the item we found, instead of the whole item. This is more
closer inline with what startswith() does, and allows us to simplify a
couple of invocations.
Parsing sysv files was moved to the sysv-generator in the previous commit.
This patch removes the sysv parsing from serivce.c.
Note that this patch drops the following now unused sysv-specific info
from service dump:
"SysV Init Script has LSB Header: (yes/no)"
"SysVEnabled: (yes/no)"
"SysVRunLevels: (levels)"
Restart=on-abnormal is similar to Restart=on-failure, but avoids
restarts on unclean exit codes (but still doing restarts on all
obviously unclean exits, such as timeouts, signals, coredumps, watchdog
timeouts).
Also see:
https://fedorahosted.org/fpc/ticket/191
Most likely the facility needed is actual connectivity, rather than whether or not the
network managment daemon is running.
We also need to explicitly pull in the network-online.target, as it is not active by
default.
This means {systemd-networkd,NetworkManager}-wait-online.service, can be enabled by default
as part of network-online.target, and only delay boot when some service actively pulls it in.
See: <https://bugzilla.gnome.org/show_bug.cgi?id=728965>
Cc: Pavel Šimerda <psimerda@redhat.com>
Cc: Michal Sekletar <msekleta@redhat.com>
When rebooting with systemctl, an optional argument can be passed to the
reboot system call. This makes it possible the specify the argument in a
service file and use it when the service triggers a restart.
This is useful to distinguish between manual reboots and reboots caused by
failing services.
Given that native services do not carry a sysv priority anyway it is
pointless reading them from chkconfig headers, and pretend they'd work.
So let's drop this.
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.
safe_close() automatically becomes a NOP when a negative fd is passed,
and returns -1 unconditionally. This makes it easy to write lines like
this:
fd = safe_close(fd);
Which will close an fd if it is open, and reset the fd variable
correctly.
By making use of this new scheme we can drop a > 200 lines of code that
was required to test for non-negative fds or to reset the closed fd
variable afterwards.
GCC optimizes strlen("string constant") to a constant, even with -O0.
Thus, replace patterns like sizeof("string constant")-1 with
strlen("string constant") where possible, for clarity. In particular,
for expressions intended to add up the lengths of components going into
a string, this often makes it clearer that the expression counts the
trailing '\0' exactly once, by putting the +1 for the '\0' at the end of
the expression, rather than hidden in a sizeof in the middle of the
expression.
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.
Commit 5ba6985b moves the UNIT_VTABLE(u)->sigchld_event before systemd
actually reaps the zombie. Which leads to service_load_pid_file accepting
zombie as a valid pid.
This fixes timeouts like:
[ 2746.602243] systemd[1]: chronyd.service stop-sigterm timed out. Killing.
[ 2836.852545] systemd[1]: chronyd.service still around after SIGKILL. Ignoring.
[ 2927.102187] systemd[1]: chronyd.service stop-final-sigterm timed out. Killing.
[ 3017.352560] systemd[1]: chronyd.service still around after final SIGKILL. Entering failed mode.
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.
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.
Sockets are ordered before sockets.target anyway, and sockets.target
is ordered before basic.target, and hence all bus services end up
being ordered after dbus.socket anyway. Since for kdbus clients
dbus.socket is obsolete, let's not add this dependency explicitly.
Also, it's hot in Australia and we are going for breakfast now.
It is nicer to predefine patterns using configure time check instead of
using casts everywhere.
Since we do not need to use any flags, include "%" in the format instead
of excluding it like PRI* macros.
Also, introduce a new environment variable named $WATCHDOG_PID which
cotnains the PID of the process that is supposed to send the keep-alive
events. This is similar how $LISTEN_FDS and $LISTEN_PID work together,
and protects against confusing processes further down the process tree
due to inherited environment.
In some circumstances, for example when start-up times out we
immediately jump into the final state, at which point we still should
try to watch the main pid so that the SIGCHLD allows us to quickly
move into dead state.
We need to properly initialize all error structs before use and free
them after use.
Also, there's no point in flushing stdout if we output a \n anyway...
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.
This way we can unify handling of credentials that are attached to
messages, or can be queried for bus name owners or connection peers.
This also adds the ability to extend incomplete credential information
with data from /proc,
Also, provide a convenience call that will automatically determine the
most appropriate credential object for an incoming message, by using the
the attached information if possible, the sending name information if
available and otherwise the peer's credentials.
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.
When a service exits succesfully and has RemainAfterExit set, its hold
on the console (in m->n_on_console) wasn't released since the unit state
didn't change.
We now treat passno as boleans in the generators, and don't need this any more. fsck itself
is able to sequentialize checks on the same local media, so in the common case the ordering
is redundant.
It is still possible to force an order by using .d fragments, in case that is desired.
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
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.
Previously the specifier calls could only indicate OOM by returning
NULL. With this change they will return negative errno-style error codes
like everything else.
"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.
Transient units can be created via the bus API. They are configured via
the method call parameters rather than on-disk files. They are subject
to normal GC. Transient units currently may only be created for
services (however, we will extend this), and currently only ExecStart=
and the cgroup parameters can be configured (also to be extended).
Transient units require a unique name, that previously had no
configuration file on disk.
A tool systemd-run is added that makes use of this functionality to run
arbitrary command lines as transient services:
$ systemd-run /bin/ping www.heise.de
Will cause systemd to create a new transient service and run ping in it.
Replace the very generic cgroup hookup with a much simpler one. With
this change only the high-level cgroup settings remain, the ability to
set arbitrary cgroup attributes is removed, so is support for adding
units to arbitrary cgroup controllers or setting arbitrary paths for
them (especially paths that are different for the various controllers).
This also introduces a new -.slice root slice, that is the parent of
system.slice and friends. This enables easy admin configuration of
root-level cgrouo properties.
This replaces DeviceDeny= by DevicePolicy=, and implicitly adds in
/dev/null, /dev/zero and friends if DeviceAllow= is used (unless this is
turned off by DevicePolicy=).
