logind: add infrastructure to keep track of machines, and move to slices

- This changes all logind cgroup objects to use slice objects rather
  than fixed croup locations.

- logind can now collect minimal information about running
  VMs/containers. As fixed cgroup locations can no longer be used we
  need an entity that keeps track of machine cgroups in whatever slice
  they might be located. Since logind already keeps track of users,
  sessions and seats this is a trivial addition.

- nspawn will now register with logind and pass various bits of metadata
  along. A new option "--slice=" has been added to place the container
  in a specific slice.

- loginctl gained commands to list, introspect and terminate machines.

- user.slice and machine.slice will now be pulled in by logind.service,
  since only logind.service requires this slice.

1 files changed, 455 insertions, 0 deletions

diff --git a/src/login/logind-machine.c b/src/login/logind-machine.c
new file mode 100644
index 0000000000..347e5aa022
--- /dev/null
+++ b/src/login/logind-machine.c
@@ -0,0 +1,455 @@
+/*-*- Mode: C; c-basic-offset: 8; indent-tabs-mode: nil -*-*/
+
+/***
+  This file is part of systemd.
+
+  Copyright 2011 Lennart Poettering
+
+  systemd is free software; you can redistribute it and/or modify it
+  under the terms of the GNU Lesser General Public License as published by
+  the Free Software Foundation; either version 2.1 of the License, or
+  (at your option) any later version.
+
+  systemd is distributed in the hope that it will be useful, but
+  WITHOUT ANY WARRANTY; without even the implied warranty of
+  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+  Lesser General Public License for more details.
+
+  You should have received a copy of the GNU Lesser General Public License
+  along with systemd; If not, see <http://www.gnu.org/licenses/>.
+***/
+
+#include <string.h>
+#include <unistd.h>
+#include <errno.h>
+
+#include "logind-machine.h"
+#include "util.h"
+#include "mkdir.h"
+#include "cgroup-util.h"
+#include "hashmap.h"
+#include "strv.h"
+#include "fileio.h"
+#include "special.h"
+#include <systemd/sd-messages.h>
+
+Machine* machine_new(Manager *manager, const char *name) {
+        Machine *m;
+
+        assert(manager);
+        assert(name);
+
+        m = new0(Machine, 1);
+        if (!m)
+                return NULL;
+
+        m->name = strdup(name);
+        if (!m->name)
+                goto fail;
+
+        m->state_file = strappend("/run/systemd/machines/", m->name);
+        if (!m->state_file)
+                goto fail;
+
+        if (hashmap_put(manager->machines, m->name, m) < 0)
+                goto fail;
+
+        m->class = _MACHINE_CLASS_INVALID;
+        m->manager = manager;
+
+        return m;
+
+fail:
+        free(m->state_file);
+        free(m->name);
+        free(m);
+
+        return NULL;
+}
+
+void machine_free(Machine *m) {
+        assert(m);
+
+        if (m->in_gc_queue)
+                LIST_REMOVE(Machine, gc_queue, m->manager->machine_gc_queue, m);
+
+        if (m->cgroup_path) {
+                hashmap_remove(m->manager->machine_cgroups, m->cgroup_path);
+                free(m->cgroup_path);
+        }
+
+        hashmap_remove(m->manager->machines, m->name);
+
+        free(m->name);
+        free(m->state_file);
+        free(m->service);
+        free(m->slice);
+        free(m->root_directory);
+        free(m);
+}
+
+int machine_save(Machine *m) {
+        _cleanup_free_ char *temp_path = NULL;
+        _cleanup_fclose_ FILE *f = NULL;
+        int r;
+
+        assert(m);
+        assert(m->state_file);
+
+        if (!m->started)
+                return 0;
+
+        r = mkdir_safe_label("/run/systemd/machines", 0755, 0, 0);
+        if (r < 0)
+                goto finish;
+
+        r = fopen_temporary(m->state_file, &f, &temp_path);
+        if (r < 0)
+                goto finish;
+
+        fchmod(fileno(f), 0644);
+
+        fprintf(f,
+                "# This is private data. Do not parse.\n"
+                "NAME=%s\n",
+                m->name);
+
+        if (m->cgroup_path)
+                fprintf(f, "CGROUP=%s\n", m->cgroup_path);
+
+        if (m->service)
+                fprintf(f, "SERVICE=%s\n", m->service);
+
+        if (m->slice)
+                fprintf(f, "SLICE=%s\n", m->slice);
+
+        if (m->root_directory)
+                fprintf(f, "ROOT=%s\n", m->root_directory);
+
+        if (!sd_id128_equal(m->id, SD_ID128_NULL))
+                fprintf(f, "ID=" SD_ID128_FORMAT_STR "\n", SD_ID128_FORMAT_VAL(m->id));
+
+        if (m->leader != 0)
+                fprintf(f, "LEADER=%lu\n", (unsigned long) m->leader);
+
+        if (m->class != _MACHINE_CLASS_INVALID)
+                fprintf(f, "CLASS=%s\n", machine_class_to_string(m->class));
+
+        if (dual_timestamp_is_set(&m->timestamp))
+                fprintf(f,
+                        "REALTIME=%llu\n"
+                        "MONOTONIC=%llu\n",
+                        (unsigned long long) m->timestamp.realtime,
+                        (unsigned long long) m->timestamp.monotonic);
+
+        fflush(f);
+
+        if (ferror(f) || rename(temp_path, m->state_file) < 0) {
+                r = -errno;
+                unlink(m->state_file);
+                unlink(temp_path);
+        }
+
+finish:
+        if (r < 0)
+                log_error("Failed to save machine data for %s: %s", m->name, strerror(-r));
+
+        return r;
+}
+
+int machine_load(Machine *m) {
+        _cleanup_free_ char *realtime = NULL, *monotonic = NULL, *id = NULL, *leader = NULL, *class = NULL;
+        int r;
+
+        assert(m);
+
+        r = parse_env_file(m->state_file, NEWLINE,
+                           "CGROUP",    &m->cgroup_path,
+                           "SERVICE",   &m->service,
+                           "SLICE",     &m->slice,
+                           "ROOT",      &m->root_directory,
+                           "ID",        &id,
+                           "LEADER",    &leader,
+                           "CLASS",     &class,
+                           "REALTIME",  &realtime,
+                           "MONOTONIC", &monotonic,
+                           NULL);
+        if (r < 0) {
+                if (r == -ENOENT)
+                        return 0;
+
+                log_error("Failed to read %s: %s", m->state_file, strerror(-r));
+                return r;
+        }
+
+        if (id)
+                sd_id128_from_string(id, &m->id);
+
+        if (leader)
+                parse_pid(leader, &m->leader);
+
+        if (class) {
+                MachineClass c;
+
+                c = machine_class_from_string(class);
+                if (c >= 0)
+                        m->class = c;
+        }
+
+        if (realtime) {
+                unsigned long long l;
+                if (sscanf(realtime, "%llu", &l) > 0)
+                        m->timestamp.realtime = l;
+        }
+
+        if (monotonic) {
+                unsigned long long l;
+                if (sscanf(monotonic, "%llu", &l) > 0)
+                        m->timestamp.monotonic = l;
+        }
+
+        return r;
+}
+
+static int machine_create_one_group(Machine *m, const char *controller, const char *path) {
+        int r;
+
+        assert(m);
+        assert(path);
+
+        if (m->leader > 0)
+                r = cg_create_and_attach(controller, path, m->leader);
+        else
+                r = -EINVAL;
+
+        if (r < 0) {
+                r = cg_create(controller, path, NULL);
+                if (r < 0)
+                        return r;
+        }
+
+        return 0;
+}
+
+static int machine_create_cgroup(Machine *m) {
+        char **k;
+        int r;
+
+        assert(m);
+
+        if (!m->slice) {
+                m->slice = strdup(SPECIAL_MACHINE_SLICE);
+                if (!m->slice)
+                        return log_oom();
+        }
+
+        if (!m->cgroup_path) {
+                _cleanup_free_ char *escaped = NULL, *slice = NULL;
+                char *name;
+
+                name = strappenda(m->name, ".machine");
+
+                escaped = cg_escape(name);
+                if (!escaped)
+                        return log_oom();
+
+                r = cg_slice_to_path(m->slice, &slice);
+                if (r < 0)
+                        return r;
+
+                m->cgroup_path = strjoin(m->manager->cgroup_root, "/", slice, "/", escaped, NULL);
+                if (!m->cgroup_path)
+                        return log_oom();
+        }
+
+        r = machine_create_one_group(m, SYSTEMD_CGROUP_CONTROLLER, m->cgroup_path);
+        if (r < 0) {
+                log_error("Failed to create cgroup "SYSTEMD_CGROUP_CONTROLLER":%s: %s", m->cgroup_path, strerror(-r));
+                return r;
+        }
+
+        STRV_FOREACH(k, m->manager->controllers) {
+
+                if (strv_contains(m->manager->reset_controllers, *k))
+                        continue;
+
+                r = machine_create_one_group(m, *k, m->cgroup_path);
+                if (r < 0)
+                        log_warning("Failed to create cgroup %s:%s: %s", *k, m->cgroup_path, strerror(-r));
+        }
+
+        if (m->leader > 0) {
+                STRV_FOREACH(k, m->manager->reset_controllers) {
+                        r = cg_attach(*k, "/", m->leader);
+                        if (r < 0)
+                                log_warning("Failed to reset controller %s: %s", *k, strerror(-r));
+                }
+        }
+
+        r = hashmap_put(m->manager->machine_cgroups, m->cgroup_path, m);
+        if (r < 0)
+                log_warning("Failed to create mapping between cgroup and machine");
+
+        return 0;
+}
+
+int machine_start(Machine *m) {
+        int r;
+
+        assert(m);
+
+        if (m->started)
+                return 0;
+
+        log_struct(LOG_INFO,
+                   MESSAGE_ID(SD_MESSAGE_MACHINE_START),
+                   "NAME=%s", m->name,
+                   "LEADER=%lu", (unsigned long) m->leader,
+                   "MESSAGE=New machine %s.", m->name,
+                   NULL);
+
+        /* Create cgroup */
+        r = machine_create_cgroup(m);
+        if (r < 0)
+                return r;
+
+        if (!dual_timestamp_is_set(&m->timestamp))
+                dual_timestamp_get(&m->timestamp);
+
+        m->started = true;
+
+        /* Save new machine data */
+        machine_save(m);
+
+        machine_send_signal(m, true);
+
+        return 0;
+}
+
+static int machine_terminate_cgroup(Machine *m) {
+        int r;
+        char **k;
+
+        assert(m);
+
+        if (!m->cgroup_path)
+                return 0;
+
+        cg_trim(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_path, false);
+
+        r = cg_kill_recursive_and_wait(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_path, true);
+        if (r < 0)
+                log_error("Failed to kill machine cgroup: %s", strerror(-r));
+
+        STRV_FOREACH(k, m->manager->controllers)
+                cg_trim(*k, m->cgroup_path, true);
+
+        hashmap_remove(m->manager->machine_cgroups, m->cgroup_path);
+
+        free(m->cgroup_path);
+        m->cgroup_path = NULL;
+
+        return r;
+}
+
+int machine_stop(Machine *m) {
+        int r = 0, k;
+        assert(m);
+
+        if (m->started)
+                log_struct(LOG_INFO,
+                           MESSAGE_ID(SD_MESSAGE_MACHINE_STOP),
+                           "NAME=%s", m->name,
+                           "LEADER=%lu", (unsigned long) m->leader,
+                           "MESSAGE=Machine %s terminated.", m->name,
+                           NULL);
+
+        /* Kill cgroup */
+        k = machine_terminate_cgroup(m);
+        if (k < 0)
+                r = k;
+
+        unlink(m->state_file);
+        machine_add_to_gc_queue(m);
+
+        if (m->started)
+                machine_send_signal(m, false);
+
+        m->started = false;
+
+        return r;
+}
+
+int machine_check_gc(Machine *m, bool drop_not_started) {
+        int r;
+
+        assert(m);
+
+        if (drop_not_started && !m->started)
+                return 0;
+
+        if (m->cgroup_path) {
+                r = cg_is_empty_recursive(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_path, false);
+                if (r < 0)
+                        return r;
+
+                if (r <= 0)
+                        return 1;
+        }
+
+        return 0;
+}
+
+void machine_add_to_gc_queue(Machine *m) {
+        assert(m);
+
+        if (m->in_gc_queue)
+                return;
+
+        LIST_PREPEND(Machine, gc_queue, m->manager->machine_gc_queue, m);
+        m->in_gc_queue = true;
+}
+
+int machine_kill(Machine *m, KillWho who, int signo) {
+        _cleanup_set_free_ Set *pid_set = NULL;
+        int r = 0;
+
+        assert(m);
+
+        if (!m->cgroup_path)
+                return -ESRCH;
+
+        if (m->leader <= 0 && who == KILL_LEADER)
+                return -ESRCH;
+
+        if (m->leader > 0)
+                if (kill(m->leader, signo) < 0)
+                        r = -errno;
+
+        if (who == KILL_ALL) {
+                int q;
+
+                pid_set = set_new(trivial_hash_func, trivial_compare_func);
+                if (!pid_set)
+                        return log_oom();
+
+                if (m->leader > 0) {
+                        q = set_put(pid_set, LONG_TO_PTR(m->leader));
+                        if (q < 0)
+                                r = q;
+                }
+
+                q = cg_kill_recursive(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_path, signo, false, true, false, pid_set);
+                if (q < 0 && (q != -EAGAIN && q != -ESRCH && q != -ENOENT))
+                        r = q;
+        }
+
+        return r;
+}
+
+static const char* const machine_class_table[_MACHINE_CLASS_MAX] = {
+        [MACHINE_CONTAINER] = "container",
+        [MACHINE_VM] = "vm"
+};
+
+DEFINE_STRING_TABLE_LOOKUP(machine_class, MachineClass);