core: unified cgroup hierarchy support

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.

1 files changed, 60 insertions, 7 deletions

diff --git a/src/core/scope.c b/src/core/scope.c
index 1e94d63561..44cd324f58 100644
--- a/src/core/scope.c
+++ b/src/core/scope.c
@@ -22,12 +22,13 @@
 #include <errno.h>
 #include <unistd.h>
 
-#include "unit.h"
-#include "scope.h"
 #include "log.h"
-#include "dbus-scope.h"
+#include "strv.h"
 #include "special.h"
 #include "unit-name.h"
+#include "unit.h"
+#include "scope.h"
+#include "dbus-scope.h"
 #include "load-dropin.h"
 
 static const UnitActiveState state_translation_table[_SCOPE_STATE_MAX] = {
@@ -136,7 +137,9 @@ static int scope_verify(Scope *s) {
         if (UNIT(s)->load_state != UNIT_LOADED)
                 return 0;
 
-        if (set_isempty(UNIT(s)->pids) && UNIT(s)->manager->n_reloading <= 0) {
+        if (set_isempty(UNIT(s)->pids) &&
+            !manager_is_reloading_or_reexecuting(UNIT(s)->manager) <= 0 &&
+            !unit_has_name(UNIT(s), SPECIAL_INIT_SCOPE)) {
                 log_unit_error(UNIT(s), "Scope has no PIDs. Refusing.");
                 return -EINVAL;
         }
@@ -151,7 +154,7 @@ static int scope_load(Unit *u) {
         assert(s);
         assert(u->load_state == UNIT_STUB);
 
-        if (!u->transient && UNIT(s)->manager->n_reloading <= 0)
+        if (!u->transient && !manager_is_reloading_or_reexecuting(u->manager))
                 return -ENOENT;
 
         u->load_state = UNIT_LOADED;
@@ -279,6 +282,9 @@ static int scope_start(Unit *u) {
 
         assert(s);
 
+        if (unit_has_name(u, SPECIAL_INIT_SCOPE))
+                return -EPERM;
+
         if (s->state == SCOPE_FAILED)
                 return -EPERM;
 
@@ -289,7 +295,7 @@ static int scope_start(Unit *u) {
 
         assert(s->state == SCOPE_DEAD);
 
-        if (!u->transient && UNIT(s)->manager->n_reloading <= 0)
+        if (!u->transient && !manager_is_reloading_or_reexecuting(u->manager))
                 return -ENOENT;
 
         (void) unit_realize_cgroup(u);
@@ -464,6 +470,9 @@ static int scope_dispatch_timer(sd_event_source *source, usec_t usec, void *user
 int scope_abandon(Scope *s) {
         assert(s);
 
+        if (unit_has_name(UNIT(s), SPECIAL_INIT_SCOPE))
+                return -EPERM;
+
         if (!IN_SET(s->state, SCOPE_RUNNING, SCOPE_ABANDONED))
                 return -ESTALE;
 
@@ -499,6 +508,48 @@ _pure_ static const char *scope_sub_state_to_string(Unit *u) {
         return scope_state_to_string(SCOPE(u)->state);
 }
 
+static int scope_enumerate(Manager *m) {
+        Unit *u;
+        int r;
+
+        assert(m);
+
+        /* Let's unconditionally add the "init.scope" special unit
+         * that encapsulates PID 1. Note that PID 1 already is in the
+         * cgroup for this, we hence just need to allocate the object
+         * for it and that's it. */
+
+        u = manager_get_unit(m, SPECIAL_INIT_SCOPE);
+        if (!u) {
+                u = unit_new(m, sizeof(Scope));
+                if (!u)
+                        return log_oom();
+
+                r = unit_add_name(u, SPECIAL_INIT_SCOPE);
+                if (r < 0)  {
+                        unit_free(u);
+                        return log_error_errno(r, "Failed to add init.scope name");
+                }
+        }
+
+        u->transient = true;
+        u->default_dependencies = false;
+        u->no_gc = true;
+        SCOPE(u)->deserialized_state = SCOPE_RUNNING;
+        SCOPE(u)->kill_context.kill_signal = SIGRTMIN+14;
+
+        /* Prettify things, if we can. */
+        if (!u->description)
+                u->description = strdup("System and Service Manager");
+        if (!u->documentation)
+                (void) strv_extend(&u->documentation, "man:systemd(1)");
+
+        unit_add_to_load_queue(u);
+        unit_add_to_dbus_queue(u);
+
+        return 0;
+}
+
 static const char* const scope_state_table[_SCOPE_STATE_MAX] = {
         [SCOPE_DEAD] = "dead",
         [SCOPE_RUNNING] = "running",
@@ -565,5 +616,7 @@ const UnitVTable scope_vtable = {
         .bus_set_property = bus_scope_set_property,
         .bus_commit_properties = bus_scope_commit_properties,
 
-        .can_transient = true
+        .can_transient = true,
+
+        .enumerate = scope_enumerate,
 };