-/***
- This file is part of systemd.
-
- Copyright 2013 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/>.
-***/
+/* SPDX-License-Identifier: LGPL-2.1+ */
#include <fcntl.h>
#include <fnmatch.h>
#include "alloc-util.h"
+//#include "blockdev-util.h"
//#include "bpf-firewall.h"
+//#include "btrfs-util.h"
+//#include "bus-error.h"
#include "cgroup-util.h"
#include "cgroup.h"
#include "fd-util.h"
#include "parse-util.h"
#include "path-util.h"
#include "process-util.h"
+//#include "procfs-util.h"
//#include "special.h"
#include "stdio-util.h"
#include "string-table.h"
#include "string-util.h"
+#include "virt.h"
#define CGROUP_CPU_QUOTA_PERIOD_USEC ((usec_t) 100 * USEC_PER_MSEC)
+bool manager_owns_root_cgroup(Manager *m) {
+ assert(m);
+
+ /* Returns true if we are managing the root cgroup. Note that it isn't sufficient to just check whether the
+ * group root path equals "/" since that will also be the case if CLONE_NEWCGROUP is in the mix. Since there's
+ * appears to be no nice way to detect whether we are in a CLONE_NEWCGROUP namespace we instead just check if
+ * we run in any kind of container virtualization. */
+
+ if (detect_container() > 0)
+ return false;
+
+ return empty_or_root(m->cgroup_root);
+}
+
#if 0 /// UNNEEDED by elogind
+bool unit_has_root_cgroup(Unit *u) {
+ assert(u);
+
+ /* Returns whether this unit manages the root cgroup. This will return true if this unit is the root slice and
+ * the manager manages the root cgroup. */
+
+ if (!manager_owns_root_cgroup(u->manager))
+ return false;
+
+ return unit_has_name(u, SPECIAL_ROOT_SLICE);
+}
+
static void cgroup_compat_warn(void) {
static bool cgroup_compat_warned = false;
if (cgroup_compat_warned)
return;
- log_warning("cgroup compatibility translation between legacy and unified hierarchy settings activated. See cgroup-compat debug messages for details.");
+ log_warning("cgroup compatibility translation between legacy and unified hierarchy settings activated. "
+ "See cgroup-compat debug messages for details.");
+
cgroup_compat_warned = true;
}
(void) cg_mask_to_string(c->delegate_controllers, &t);
- fprintf(f, "%sDelegateController=%s\n",
+ fprintf(f, "%sDelegateControllers=%s\n",
prefix,
strempty(t));
}
}
}
-static int lookup_block_device(const char *p, dev_t *dev) {
+static int lookup_block_device(const char *p, dev_t *ret) {
struct stat st;
int r;
assert(p);
- assert(dev);
+ assert(ret);
- r = stat(p, &st);
- if (r < 0)
- return log_warning_errno(errno, "Couldn't stat device %s: %m", p);
+ if (stat(p, &st) < 0)
+ return log_warning_errno(errno, "Couldn't stat device '%s': %m", p);
if (S_ISBLK(st.st_mode))
- *dev = st.st_rdev;
- else if (major(st.st_dev) != 0) {
- /* If this is not a device node then find the block
- * device this file is stored on */
- *dev = st.st_dev;
-
- /* If this is a partition, try to get the originating
- * block device */
- block_get_whole_disk(*dev, dev);
- } else {
- log_warning("%s is not a block device and file system block device cannot be determined or is not local.", p);
- return -ENODEV;
+ *ret = st.st_rdev;
+ else if (major(st.st_dev) != 0)
+ *ret = st.st_dev; /* If this is not a device node then use the block device this file is stored on */
+ else {
+ /* If this is btrfs, getting the backing block device is a bit harder */
+ r = btrfs_get_block_device(p, ret);
+ if (r < 0 && r != -ENOTTY)
+ return log_warning_errno(r, "Failed to determine block device backing btrfs file system '%s': %m", p);
+ if (r == -ENOTTY) {
+ log_warning("'%s' is not a block device node, and file system block device cannot be determined or is not local.", p);
+ return -ENODEV;
+ }
}
+ /* If this is a LUKS device, try to get the originating block device */
+ (void) block_get_originating(*ret, ret);
+
+ /* If this is a partition, try to get the originating block device */
+ (void) block_get_whole_disk(*ret, ret);
return 0;
}
"Failed to set blkio.weight_device: %m");
}
-static unsigned cgroup_apply_io_device_limit(Unit *u, const char *dev_path, uint64_t *limits) {
+static void cgroup_apply_io_device_limit(Unit *u, const char *dev_path, uint64_t *limits) {
char limit_bufs[_CGROUP_IO_LIMIT_TYPE_MAX][DECIMAL_STR_MAX(uint64_t)];
char buf[DECIMAL_STR_MAX(dev_t)*2+2+(6+DECIMAL_STR_MAX(uint64_t)+1)*4];
CGroupIOLimitType type;
dev_t dev;
- unsigned n = 0;
int r;
r = lookup_block_device(dev_path, &dev);
if (r < 0)
- return 0;
+ return;
- for (type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++) {
- if (limits[type] != cgroup_io_limit_defaults[type]) {
+ for (type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++)
+ if (limits[type] != cgroup_io_limit_defaults[type])
xsprintf(limit_bufs[type], "%" PRIu64, limits[type]);
- n++;
- } else {
+ else
xsprintf(limit_bufs[type], "%s", limits[type] == CGROUP_LIMIT_MAX ? "max" : "0");
- }
- }
xsprintf(buf, "%u:%u rbps=%s wbps=%s riops=%s wiops=%s\n", major(dev), minor(dev),
limit_bufs[CGROUP_IO_RBPS_MAX], limit_bufs[CGROUP_IO_WBPS_MAX],
if (r < 0)
log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
"Failed to set io.max: %m");
- return n;
}
-static unsigned cgroup_apply_blkio_device_limit(Unit *u, const char *dev_path, uint64_t rbps, uint64_t wbps) {
+static void cgroup_apply_blkio_device_limit(Unit *u, const char *dev_path, uint64_t rbps, uint64_t wbps) {
char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1];
dev_t dev;
- unsigned n = 0;
int r;
r = lookup_block_device(dev_path, &dev);
if (r < 0)
- return 0;
+ return;
- if (rbps != CGROUP_LIMIT_MAX)
- n++;
sprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), rbps);
r = cg_set_attribute("blkio", u->cgroup_path, "blkio.throttle.read_bps_device", buf);
if (r < 0)
log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
"Failed to set blkio.throttle.read_bps_device: %m");
- if (wbps != CGROUP_LIMIT_MAX)
- n++;
sprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), wbps);
r = cg_set_attribute("blkio", u->cgroup_path, "blkio.throttle.write_bps_device", buf);
if (r < 0)
log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
"Failed to set blkio.throttle.write_bps_device: %m");
-
- return n;
}
static bool cgroup_context_has_unified_memory_config(CGroupContext *c) {
"Failed to set %s: %m", file);
}
-static void cgroup_apply_firewall(Unit *u, CGroupContext *c) {
- int r;
+static void cgroup_apply_firewall(Unit *u) {
+ assert(u);
- if (u->type == UNIT_SLICE) /* Skip this for slice units, they are inner cgroup nodes, and since bpf/cgroup is
- * not recursive we don't ever touch the bpf on them */
- return;
+ /* Best-effort: let's apply IP firewalling and/or accounting if that's enabled */
- r = bpf_firewall_compile(u);
- if (r < 0)
+ if (bpf_firewall_compile(u) < 0)
return;
(void) bpf_firewall_install(u);
- return;
}
static void cgroup_context_apply(
assert(u);
- c = unit_get_cgroup_context(u);
- path = u->cgroup_path;
-
- assert(c);
- assert(path);
-
/* Nothing to do? Exit early! */
if (apply_mask == 0 && !apply_bpf)
return;
- /* Some cgroup attributes are not supported on the root cgroup,
- * hence silently ignore */
- is_root = isempty(path) || path_equal(path, "/");
- if (is_root)
- /* Make sure we don't try to display messages with an empty path. */
+ /* Some cgroup attributes are not supported on the root cgroup, hence silently ignore */
+ is_root = unit_has_root_cgroup(u);
+
+ assert_se(c = unit_get_cgroup_context(u));
+ assert_se(path = u->cgroup_path);
+
+ if (is_root) /* Make sure we don't try to display messages with an empty path. */
path = "/";
/* We generally ignore errors caused by read-only mounted
/* Apply limits and free ones without config. */
if (has_io) {
- CGroupIODeviceLimit *l, *next;
+ CGroupIODeviceLimit *l;
+
+ LIST_FOREACH(device_limits, l, c->io_device_limits)
+ cgroup_apply_io_device_limit(u, l->path, l->limits);
- LIST_FOREACH_SAFE(device_limits, l, next, c->io_device_limits) {
- if (!cgroup_apply_io_device_limit(u, l->path, l->limits))
- cgroup_context_free_io_device_limit(c, l);
- }
} else if (has_blockio) {
- CGroupBlockIODeviceBandwidth *b, *next;
+ CGroupBlockIODeviceBandwidth *b;
- LIST_FOREACH_SAFE(device_bandwidths, b, next, c->blockio_device_bandwidths) {
+ LIST_FOREACH(device_bandwidths, b, c->blockio_device_bandwidths) {
uint64_t limits[_CGROUP_IO_LIMIT_TYPE_MAX];
CGroupIOLimitType type;
log_cgroup_compat(u, "Applying BlockIO{Read|Write}Bandwidth %" PRIu64 " %" PRIu64 " as IO{Read|Write}BandwidthMax for %s",
b->rbps, b->wbps, b->path);
- if (!cgroup_apply_io_device_limit(u, b->path, limits))
- cgroup_context_free_blockio_device_bandwidth(c, b);
+ cgroup_apply_io_device_limit(u, b->path, limits);
}
}
}
/* Apply limits and free ones without config. */
if (has_io) {
- CGroupIODeviceLimit *l, *next;
+ CGroupIODeviceLimit *l;
- LIST_FOREACH_SAFE(device_limits, l, next, c->io_device_limits) {
+ LIST_FOREACH(device_limits, l, c->io_device_limits) {
log_cgroup_compat(u, "Applying IO{Read|Write}Bandwidth %" PRIu64 " %" PRIu64 " as BlockIO{Read|Write}BandwidthMax for %s",
l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX], l->path);
- if (!cgroup_apply_blkio_device_limit(u, l->path, l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX]))
- cgroup_context_free_io_device_limit(c, l);
+ cgroup_apply_blkio_device_limit(u, l->path, l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX]);
}
} else if (has_blockio) {
- CGroupBlockIODeviceBandwidth *b, *next;
+ CGroupBlockIODeviceBandwidth *b;
- LIST_FOREACH_SAFE(device_bandwidths, b, next, c->blockio_device_bandwidths)
- if (!cgroup_apply_blkio_device_limit(u, b->path, b->rbps, b->wbps))
- cgroup_context_free_blockio_device_bandwidth(c, b);
+ LIST_FOREACH(device_bandwidths, b, c->blockio_device_bandwidths)
+ cgroup_apply_blkio_device_limit(u, b->path, b->rbps, b->wbps);
}
}
"/dev/random\0" "rwm\0"
"/dev/urandom\0" "rwm\0"
"/dev/tty\0" "rwm\0"
- "/dev/pts/ptmx\0" "rw\0" /* /dev/pts/ptmx may not be duplicated, but accessed */
- /* Allow /run/elogind/inaccessible/{chr,blk} devices for mapping InaccessiblePaths */
+ "/dev/ptmx\0" "rwm\0"
/* Allow /run/systemd/inaccessible/{chr,blk} devices for mapping InaccessiblePaths */
"-/run/systemd/inaccessible/chr\0" "rwm\0"
"-/run/systemd/inaccessible/blk\0" "rwm\0";
NULSTR_FOREACH_PAIR(x, y, auto_devices)
whitelist_device(path, x, y);
+ /* PTS (/dev/pts) devices may not be duplicated, but accessed */
whitelist_major(path, "pts", 'c', "rw");
}
}
}
- if ((apply_mask & CGROUP_MASK_PIDS) && !is_root) {
+ if (apply_mask & CGROUP_MASK_PIDS) {
+
+ if (is_root) {
+ /* So, the "pids" controller does not expose anything on the root cgroup, in order not to
+ * replicate knobs exposed elsewhere needlessly. We abstract this away here however, and when
+ * the knobs of the root cgroup are modified propagate this to the relevant sysctls. There's a
+ * non-obvious asymmetry however: unlike the cgroup properties we don't really want to take
+ * exclusive ownership of the sysctls, but we still want to honour things if the user sets
+ * limits. Hence we employ sort of a one-way strategy: when the user sets a bounded limit
+ * through us it counts. When the user afterwards unsets it again (i.e. sets it to unbounded)
+ * it also counts. But if the user never set a limit through us (i.e. we are the default of
+ * "unbounded") we leave things unmodified. For this we manage a global boolean that we turn on
+ * the first time we set a limit. Note that this boolean is flushed out on manager reload,
+ * which is desirable so that there's an offical way to release control of the sysctl from
+ * systemd: set the limit to unbounded and reload. */
+
+ if (c->tasks_max != CGROUP_LIMIT_MAX) {
+ u->manager->sysctl_pid_max_changed = true;
+ r = procfs_tasks_set_limit(c->tasks_max);
+ } else if (u->manager->sysctl_pid_max_changed)
+ r = procfs_tasks_set_limit(TASKS_MAX);
+ else
+ r = 0;
- if (c->tasks_max != CGROUP_LIMIT_MAX) {
- char buf[DECIMAL_STR_MAX(uint64_t) + 2];
+ if (r < 0)
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to write to tasks limit sysctls: %m");
- sprintf(buf, "%" PRIu64 "\n", c->tasks_max);
- r = cg_set_attribute("pids", path, "pids.max", buf);
- } else
- r = cg_set_attribute("pids", path, "pids.max", "max");
+ } else {
+ if (c->tasks_max != CGROUP_LIMIT_MAX) {
+ char buf[DECIMAL_STR_MAX(uint64_t) + 2];
- if (r < 0)
- log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set pids.max: %m");
+ sprintf(buf, "%" PRIu64 "\n", c->tasks_max);
+ r = cg_set_attribute("pids", path, "pids.max", buf);
+ } else
+ r = cg_set_attribute("pids", path, "pids.max", "max");
+ if (r < 0)
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set pids.max: %m");
+ }
}
if (apply_bpf)
- cgroup_apply_firewall(u, c);
+ cgroup_apply_firewall(u);
}
CGroupMask cgroup_context_get_mask(CGroupContext *c) {
mask |= CGROUP_MASK_DEVICES;
if (c->tasks_accounting ||
- c->tasks_max != (uint64_t) -1)
+ c->tasks_max != CGROUP_LIMIT_MAX)
mask |= CGROUP_MASK_PIDS;
return mask;
if (!c)
return 0;
- return cgroup_context_get_mask(c);
+ return cgroup_context_get_mask(c) | unit_get_delegate_mask(u);
}
CGroupMask unit_get_delegate_mask(Unit *u) {
*
* Note that on the unified hierarchy it is safe to delegate controllers to unprivileged services. */
- if (u->type == UNIT_SLICE)
- return 0;
-
- c = unit_get_cgroup_context(u);
- if (!c)
- return 0;
-
- if (!c->delegate)
+ if (!unit_cgroup_delegate(u))
return 0;
if (cg_all_unified() <= 0) {
return 0;
}
+ assert_se(c = unit_get_cgroup_context(u));
return c->delegate_controllers;
}
if (u->cgroup_members_mask_valid)
return u->cgroup_members_mask;
- u->cgroup_members_mask = unit_get_delegate_mask(u);
+ u->cgroup_members_mask = 0;
if (u->type == UNIT_SLICE) {
void *v;
if (UNIT_ISSET(u->slice))
return unit_get_members_mask(UNIT_DEREF(u->slice));
- return unit_get_subtree_mask(u);
+ return unit_get_subtree_mask(u); /* we are the top-level slice */
}
CGroupMask unit_get_subtree_mask(Unit *u) {
Unit *p;
assert(u);
- /* We never attach BPF to slice units, as they are inner cgroup nodes and cgroup/BPF is not recursive at the
- * moment. */
- if (u->type == UNIT_SLICE)
- return false;
-
c = unit_get_cgroup_context(u);
if (!c)
return false;
}
}
-static const char *migrate_callback(CGroupMask mask, void *userdata) {
- Unit *u = userdata;
+const char *unit_get_realized_cgroup_path(Unit *u, CGroupMask mask) {
- assert(mask != 0);
- assert(u);
+ /* Returns the realized cgroup path of the specified unit where all specified controllers are available. */
while (u) {
+
if (u->cgroup_path &&
u->cgroup_realized &&
- (u->cgroup_realized_mask & mask) == mask)
+ FLAGS_SET(u->cgroup_realized_mask, mask))
return u->cgroup_path;
u = UNIT_DEREF(u->slice);
return NULL;
}
+static const char *migrate_callback(CGroupMask mask, void *userdata) {
+ return unit_get_realized_cgroup_path(userdata, mask);
+}
+
char *unit_default_cgroup_path(Unit *u) {
_cleanup_free_ char *escaped = NULL, *slice = NULL;
int r;
unit_release_cgroup(u);
- u->cgroup_path = p;
- p = NULL;
+ u->cgroup_path = TAKE_PTR(p);
return 1;
}
return 0;
}
+int unit_pick_cgroup_path(Unit *u) {
+ _cleanup_free_ char *path = NULL;
+ int r;
+
+ assert(u);
+
+ if (u->cgroup_path)
+ return 0;
+
+ if (!UNIT_HAS_CGROUP_CONTEXT(u))
+ return -EINVAL;
+
+ path = unit_default_cgroup_path(u);
+ if (!path)
+ return log_oom();
+
+ r = unit_set_cgroup_path(u, path);
+ if (r == -EEXIST)
+ return log_unit_error_errno(u, r, "Control group %s exists already.", path);
+ if (r < 0)
+ return log_unit_error_errno(u, r, "Failed to set unit's control group path to %s: %m", path);
+
+ return 0;
+}
+
static int unit_create_cgroup(
Unit *u,
CGroupMask target_mask,
CGroupContext *c;
int r;
+ bool created;
assert(u);
if (!c)
return 0;
- if (!u->cgroup_path) {
- _cleanup_free_ char *path = NULL;
-
- path = unit_default_cgroup_path(u);
- if (!path)
- return log_oom();
-
- r = unit_set_cgroup_path(u, path);
- if (r == -EEXIST)
- return log_unit_error_errno(u, r, "Control group %s exists already.", path);
- if (r < 0)
- return log_unit_error_errno(u, r, "Failed to set unit's control group path to %s: %m", path);
- }
+ /* Figure out our cgroup path */
+ r = unit_pick_cgroup_path(u);
+ if (r < 0)
+ return r;
/* First, create our own group */
r = cg_create_everywhere(u->manager->cgroup_supported, target_mask, u->cgroup_path);
if (r < 0)
return log_unit_error_errno(u, r, "Failed to create cgroup %s: %m", u->cgroup_path);
+ created = !!r;
/* Start watching it */
(void) unit_watch_cgroup(u);
- /* Enable all controllers we need */
- r = cg_enable_everywhere(u->manager->cgroup_supported, enable_mask, u->cgroup_path);
- if (r < 0)
- log_unit_warning_errno(u, r, "Failed to enable controllers on cgroup %s, ignoring: %m", u->cgroup_path);
+ /* Preserve enabled controllers in delegated units, adjust others. */
+ if (created || !unit_cgroup_delegate(u)) {
+
+ /* Enable all controllers we need */
+ r = cg_enable_everywhere(u->manager->cgroup_supported, enable_mask, u->cgroup_path);
+ if (r < 0)
+ log_unit_warning_errno(u, r, "Failed to enable controllers on cgroup %s, ignoring: %m",
+ u->cgroup_path);
+ }
/* Keep track that this is now realized */
u->cgroup_realized = true;
u->cgroup_enabled_mask = enable_mask;
u->cgroup_bpf_state = needs_bpf ? UNIT_CGROUP_BPF_ON : UNIT_CGROUP_BPF_OFF;
- if (u->type != UNIT_SLICE && !c->delegate) {
+ if (u->type != UNIT_SLICE && !unit_cgroup_delegate(u)) {
/* Then, possibly move things over, but not if
* subgroups may contain processes, which is the case
return 0;
}
-int unit_attach_pids_to_cgroup(Unit *u) {
+static int unit_attach_pid_to_cgroup_via_bus(Unit *u, pid_t pid, const char *suffix_path) {
+ _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL;
+ char *pp;
int r;
+
assert(u);
- r = unit_realize_cgroup(u);
+ if (MANAGER_IS_SYSTEM(u->manager))
+ return -EINVAL;
+
+ if (!u->manager->system_bus)
+ return -EIO;
+
+ if (!u->cgroup_path)
+ return -EINVAL;
+
+ /* Determine this unit's cgroup path relative to our cgroup root */
+ pp = path_startswith(u->cgroup_path, u->manager->cgroup_root);
+ if (!pp)
+ return -EINVAL;
+
+ pp = strjoina("/", pp, suffix_path);
+ path_simplify(pp, false);
+
+ r = sd_bus_call_method(u->manager->system_bus,
+ "org.freedesktop.systemd1",
+ "/org/freedesktop/systemd1",
+ "org.freedesktop.systemd1.Manager",
+ "AttachProcessesToUnit",
+ &error, NULL,
+ "ssau",
+ NULL /* empty unit name means client's unit, i.e. us */, pp, 1, (uint32_t) pid);
if (r < 0)
- return r;
+ return log_unit_debug_errno(u, r, "Failed to attach unit process " PID_FMT " via the bus: %s", pid, bus_error_message(&error, r));
+
+ return 0;
+}
+
+int unit_attach_pids_to_cgroup(Unit *u, Set *pids, const char *suffix_path) {
+ CGroupMask delegated_mask;
+ const char *p;
+ Iterator i;
+ void *pidp;
+ int r, q;
+
+ assert(u);
+
+ if (!UNIT_HAS_CGROUP_CONTEXT(u))
+ return -EINVAL;
- r = cg_attach_many_everywhere(u->manager->cgroup_supported, u->cgroup_path, u->pids, migrate_callback, u);
+ if (set_isempty(pids))
+ return 0;
+
+ r = unit_realize_cgroup(u);
if (r < 0)
return r;
- return 0;
+ if (isempty(suffix_path))
+ p = u->cgroup_path;
+ else
+ p = strjoina(u->cgroup_path, "/", suffix_path);
+
+ delegated_mask = unit_get_delegate_mask(u);
+
+ r = 0;
+ SET_FOREACH(pidp, pids, i) {
+ pid_t pid = PTR_TO_PID(pidp);
+ CGroupController c;
+
+ /* First, attach the PID to the main cgroup hierarchy */
+ q = cg_attach(SYSTEMD_CGROUP_CONTROLLER, p, pid);
+ if (q < 0) {
+ log_unit_debug_errno(u, q, "Couldn't move process " PID_FMT " to requested cgroup '%s': %m", pid, p);
+
+ if (MANAGER_IS_USER(u->manager) && IN_SET(q, -EPERM, -EACCES)) {
+ int z;
+
+ /* If we are in a user instance, and we can't move the process ourselves due to
+ * permission problems, let's ask the system instance about it instead. Since it's more
+ * privileged it might be able to move the process across the leaves of a subtree who's
+ * top node is not owned by us. */
+
+ z = unit_attach_pid_to_cgroup_via_bus(u, pid, suffix_path);
+ if (z < 0)
+ log_unit_debug_errno(u, z, "Couldn't move process " PID_FMT " to requested cgroup '%s' via the system bus either: %m", pid, p);
+ else
+ continue; /* When the bus thing worked via the bus we are fully done for this PID. */
+ }
+
+ if (r >= 0)
+ r = q; /* Remember first error */
+
+ continue;
+ }
+
+ q = cg_all_unified();
+ if (q < 0)
+ return q;
+ if (q > 0)
+ continue;
+
+ /* In the legacy hierarchy, attach the process to the request cgroup if possible, and if not to the
+ * innermost realized one */
+
+ for (c = 0; c < _CGROUP_CONTROLLER_MAX; c++) {
+ CGroupMask bit = CGROUP_CONTROLLER_TO_MASK(c);
+ const char *realized;
+
+ if (!(u->manager->cgroup_supported & bit))
+ continue;
+
+ /* If this controller is delegated and realized, honour the caller's request for the cgroup suffix. */
+ if (delegated_mask & u->cgroup_realized_mask & bit) {
+ q = cg_attach(cgroup_controller_to_string(c), p, pid);
+ if (q >= 0)
+ continue; /* Success! */
+
+ log_unit_debug_errno(u, q, "Failed to attach PID " PID_FMT " to requested cgroup %s in controller %s, falling back to unit's cgroup: %m",
+ pid, p, cgroup_controller_to_string(c));
+ }
+
+ /* So this controller is either not delegate or realized, or something else weird happened. In
+ * that case let's attach the PID at least to the closest cgroup up the tree that is
+ * realized. */
+ realized = unit_get_realized_cgroup_path(u, bit);
+ if (!realized)
+ continue; /* Not even realized in the root slice? Then let's not bother */
+
+ q = cg_attach(cgroup_controller_to_string(c), realized, pid);
+ if (q < 0)
+ log_unit_debug_errno(u, q, "Failed to attach PID " PID_FMT " to realized cgroup %s in controller %s, ignoring: %m",
+ pid, realized, cgroup_controller_to_string(c));
+ }
+ }
+
+ return r;
}
static void cgroup_xattr_apply(Unit *u) {
sd_id128_to_string(u->invocation_id, ids), 32,
0);
if (r < 0)
- log_unit_warning_errno(u, r, "Failed to set invocation ID on control group %s, ignoring: %m", u->cgroup_path);
+ log_unit_debug_errno(u, r, "Failed to set invocation ID on control group %s, ignoring: %m", u->cgroup_path);
}
static bool unit_has_mask_realized(
(!needs_bpf && u->cgroup_bpf_state == UNIT_CGROUP_BPF_OFF));
}
+static void unit_add_to_cgroup_realize_queue(Unit *u) {
+ assert(u);
+
+ if (u->in_cgroup_realize_queue)
+ return;
+
+ LIST_PREPEND(cgroup_realize_queue, u->manager->cgroup_realize_queue, u);
+ u->in_cgroup_realize_queue = true;
+}
+
+static void unit_remove_from_cgroup_realize_queue(Unit *u) {
+ assert(u);
+
+ if (!u->in_cgroup_realize_queue)
+ return;
+
+ LIST_REMOVE(cgroup_realize_queue, u->manager->cgroup_realize_queue, u);
+ u->in_cgroup_realize_queue = false;
+}
+
/* Check if necessary controllers and attributes for a unit are in place.
*
* If so, do nothing.
assert(u);
- if (u->in_cgroup_realize_queue) {
- LIST_REMOVE(cgroup_realize_queue, u->manager->cgroup_realize_queue, u);
- u->in_cgroup_realize_queue = false;
- }
+ unit_remove_from_cgroup_realize_queue(u);
target_mask = unit_get_target_mask(u);
enable_mask = unit_get_enable_mask(u);
return 0;
}
-static void unit_add_to_cgroup_realize_queue(Unit *u) {
- assert(u);
-
- if (u->in_cgroup_realize_queue)
- return;
-
- LIST_PREPEND(cgroup_realize_queue, u->manager->cgroup_realize_queue, u);
- u->in_cgroup_realize_queue = true;
-}
-
unsigned manager_dispatch_cgroup_realize_queue(Manager *m) {
ManagerState state;
unsigned n = 0;
while ((i = m->cgroup_realize_queue)) {
assert(i->in_cgroup_realize_queue);
+ if (UNIT_IS_INACTIVE_OR_FAILED(unit_active_state(i))) {
+ /* Maybe things changed, and the unit is not actually active anymore? */
+ unit_remove_from_cgroup_realize_queue(i);
+ continue;
+ }
+
r = unit_realize_cgroup_now(i, state);
if (r < 0)
log_warning_errno(r, "Failed to realize cgroups for queued unit %s, ignoring: %m", i->id);
return ret;
}
+int unit_synthesize_cgroup_empty_event(Unit *u) {
+ int r;
+
+ assert(u);
+
+ /* Enqueue a synthetic cgroup empty event if this unit doesn't watch any PIDs anymore. This is compatibility
+ * support for non-unified systems where notifications aren't reliable, and hence need to take whatever we can
+ * get as notification source as soon as we stopped having any useful PIDs to watch for. */
+
+ if (!u->cgroup_path)
+ return -ENOENT;
+
+ r = cg_unified_controller(SYSTEMD_CGROUP_CONTROLLER);
+ if (r < 0)
+ return r;
+ if (r > 0) /* On unified we have reliable notifications, and don't need this */
+ return 0;
+
+ if (!set_isempty(u->pids))
+ return 0;
+
+ unit_add_to_cgroup_empty_queue(u);
+ return 0;
+}
+
int unit_watch_all_pids(Unit *u) {
int r;
/* More stuff queued, let's make sure we remain enabled */
r = sd_event_source_set_enabled(s, SD_EVENT_ONESHOT);
if (r < 0)
- log_debug_errno(r, "Failed to reenable cgroup empty event source: %m");
+ log_debug_errno(r, "Failed to reenable cgroup empty event source, ignoring: %m");
}
unit_add_to_gc_queue(u);
const char *scope_path;
CGroupController c;
int r, all_unified;
+#if 0 /// UNNEEDED by elogind
char *e;
+#endif // 0
assert(m);
*e = 0;
#endif // 0
- /* And make sure to store away the root value without trailing
- * slash, even for the root dir, so that we can easily prepend
- * it everywhere. */
- while ((e = endswith(m->cgroup_root, "/")))
- *e = 0;
log_debug_elogind("Cgroup Controller \"%s\" -> root \"%s\"",
SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root);
+ /* And make sure to store away the root value without trailing slash, even for the root dir, so that we can
+ * easily prepend it everywhere. */
+ delete_trailing_chars(m->cgroup_root, "/");
/* 2. Show data */
r = cg_get_path(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, NULL, &path);
return log_error_errno(r, "Couldn't determine if we are running in the unified hierarchy: %m");
all_unified = cg_all_unified();
- if (r < 0)
- return log_error_errno(r, "Couldn't determine whether we are in all unified mode: %m");
- if (r > 0)
+ if (all_unified < 0)
+ return log_error_errno(all_unified, "Couldn't determine whether we are in all unified mode: %m");
+ if (all_unified > 0)
log_debug("Unified cgroup hierarchy is located at %s.", path);
else {
r = cg_unified_controller(SYSTEMD_CGROUP_CONTROLLER);
/* 5. Make sure we are in the special "init.scope" unit in the root slice. */
scope_path = strjoina(m->cgroup_root, "/" SPECIAL_INIT_SCOPE);
r = cg_create_and_attach(SYSTEMD_CGROUP_CONTROLLER, scope_path, 0);
+ if (r >= 0) {
+ /* Also, move all other userspace processes remaining in the root cgroup into that scope. */
+ r = cg_migrate(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, SYSTEMD_CGROUP_CONTROLLER, scope_path, 0);
+ if (r < 0)
+ log_warning_errno(r, "Couldn't move remaining userspace processes, ignoring: %m");
#else
/* Note:
* This method is in core, and normally called by systemd
scope_path = strjoina(m->cgroup_root, "/elogind");
r = cg_create_and_attach(SYSTEMD_CGROUP_CONTROLLER, scope_path, 0);
#endif // 0
- if (r < 0)
- return log_error_errno(r, "Failed to create %s control group: %m", scope_path);
log_debug_elogind("Created control group \"%s\"", scope_path);
-#if 0 /// elogind is not a "sub-controller" like systemd, so migration is not needed.
- /* Also, move all other userspace processes remaining in the root cgroup into that scope. */
- r = cg_migrate(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, SYSTEMD_CGROUP_CONTROLLER, scope_path, 0);
- if (r < 0)
- log_warning_errno(r, "Couldn't move remaining userspace processes, ignoring: %m");
-#endif // 0
+ /* 6. And pin it, so that it cannot be unmounted */
+ safe_close(m->pin_cgroupfs_fd);
+ m->pin_cgroupfs_fd = open(path, O_RDONLY|O_CLOEXEC|O_DIRECTORY|O_NOCTTY|O_NONBLOCK);
+ if (m->pin_cgroupfs_fd < 0)
+ return log_error_errno(errno, "Failed to open pin file: %m");
- /* 6. And pin it, so that it cannot be unmounted */
- safe_close(m->pin_cgroupfs_fd);
- m->pin_cgroupfs_fd = open(path, O_RDONLY|O_CLOEXEC|O_DIRECTORY|O_NOCTTY|O_NONBLOCK);
- if (m->pin_cgroupfs_fd < 0)
- return log_error_errno(errno, "Failed to open pin file: %m");
+#if 0 /// this is from the cgroup migration above that elogind does not need.
+ } else if (r < 0 && !m->test_run_flags)
+ return log_error_errno(r, "Failed to create %s control group: %m", scope_path);
+#endif // 0
/* 7. Always enable hierarchical support if it exists... */
if (!all_unified && m->test_run_flags == 0)
void manager_shutdown_cgroup(Manager *m, bool delete) {
assert(m);
+#if 0 /// elogind is not init
/* We can't really delete the group, since we are in it. But
* let's trim it. */
- if (delete && m->cgroup_root)
+ if (delete && m->cgroup_root && m->test_run_flags != MANAGER_TEST_RUN_MINIMAL)
(void) cg_trim(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, false);
-#if 0 /// elogind is not init
m->cgroup_empty_event_source = sd_event_source_unref(m->cgroup_empty_event_source);
m->cgroup_inotify_wd_unit = hashmap_free(m->cgroup_inotify_wd_unit);
Unit *manager_get_unit_by_pid_cgroup(Manager *m, pid_t pid) {
_cleanup_free_ char *cgroup = NULL;
- int r;
assert(m);
- if (pid <= 0)
+ if (!pid_is_valid(pid))
return NULL;
- r = cg_pid_get_path(SYSTEMD_CGROUP_CONTROLLER, pid, &cgroup);
- if (r < 0)
+ if (cg_pid_get_path(SYSTEMD_CGROUP_CONTROLLER, pid, &cgroup) < 0)
return NULL;
return manager_get_unit_by_cgroup(m, cgroup);
}
Unit *manager_get_unit_by_pid(Manager *m, pid_t pid) {
- Unit *u;
+ Unit *u, **array;
assert(m);
- if (pid <= 0)
+ /* Note that a process might be owned by multiple units, we return only one here, which is good enough for most
+ * cases, though not strictly correct. We prefer the one reported by cgroup membership, as that's the most
+ * relevant one as children of the process will be assigned to that one, too, before all else. */
+
+ if (!pid_is_valid(pid))
return NULL;
- if (pid == 1)
+ if (pid == getpid_cached())
return hashmap_get(m->units, SPECIAL_INIT_SCOPE);
- u = hashmap_get(m->watch_pids1, PID_TO_PTR(pid));
+ u = manager_get_unit_by_pid_cgroup(m, pid);
if (u)
return u;
- u = hashmap_get(m->watch_pids2, PID_TO_PTR(pid));
+ u = hashmap_get(m->watch_pids, PID_TO_PTR(pid));
if (u)
return u;
- return manager_get_unit_by_pid_cgroup(m, pid);
+ array = hashmap_get(m->watch_pids, PID_TO_PTR(-pid));
+ if (array)
+ return array[0];
+
+ return NULL;
}
#endif // 0
if (!u->cgroup_path)
return -ENODATA;
+ /* The root cgroup doesn't expose this information, let's get it from /proc instead */
+ if (unit_has_root_cgroup(u))
+ return procfs_memory_get_current(ret);
+
if ((u->cgroup_realized_mask & CGROUP_MASK_MEMORY) == 0)
return -ENODATA;
if (!u->cgroup_path)
return -ENODATA;
+ /* The root cgroup doesn't expose this information, let's get it from /proc instead */
+ if (unit_has_root_cgroup(u))
+ return procfs_tasks_get_current(ret);
+
if ((u->cgroup_realized_mask & CGROUP_MASK_PIDS) == 0)
return -ENODATA;
if (!u->cgroup_path)
return -ENODATA;
+ /* The root cgroup doesn't expose this information, let's get it from /proc instead */
+ if (unit_has_root_cgroup(u))
+ return procfs_cpu_get_usage(ret);
+
r = cg_all_unified();
if (r < 0)
return r;
if (r > 0) {
- const char *keys[] = { "usage_usec", NULL };
_cleanup_free_ char *val = NULL;
uint64_t us;
if ((u->cgroup_realized_mask & CGROUP_MASK_CPU) == 0)
return -ENODATA;
- r = cg_get_keyed_attribute("cpu", u->cgroup_path, "cpu.stat", keys, &val);
+ r = cg_get_keyed_attribute("cpu", u->cgroup_path, "cpu.stat", STRV_MAKE("usage_usec"), &val);
if (r < 0)
return r;
+ if (IN_SET(r, -ENOENT, -ENXIO))
+ return -ENODATA;
r = safe_atou64(val, &us);
if (r < 0)
assert(metric < _CGROUP_IP_ACCOUNTING_METRIC_MAX);
assert(ret);
- /* IP accounting is currently not recursive, and hence we refuse to return any data for slice nodes. Slices are
- * inner cgroup nodes and hence have no processes directly attached, hence their counters would be zero
- * anyway. And if we block this now we can later open this up, if the kernel learns recursive BPF cgroup
- * filters. */
- if (u->type == UNIT_SLICE)
- return -ENODATA;
-
if (!UNIT_CGROUP_BOOL(u, ip_accounting))
return -ENODATA;
fd = IN_SET(metric, CGROUP_IP_INGRESS_BYTES, CGROUP_IP_INGRESS_PACKETS) ?
u->ip_accounting_ingress_map_fd :
u->ip_accounting_egress_map_fd;
-
if (fd < 0)
return -ENODATA;
if (m & (CGROUP_MASK_CPU | CGROUP_MASK_CPUACCT))
m |= CGROUP_MASK_CPU | CGROUP_MASK_CPUACCT;
- if ((u->cgroup_realized_mask & m) == 0)
+ if ((u->cgroup_realized_mask & m) == 0) /* NOP? */
return;
u->cgroup_realized_mask &= ~m;
if (!UNIT_HAS_CGROUP_CONTEXT(u))
return;
- if (u->cgroup_bpf_state == UNIT_CGROUP_BPF_INVALIDATED)
+ if (u->cgroup_bpf_state == UNIT_CGROUP_BPF_INVALIDATED) /* NOP? */
return;
u->cgroup_bpf_state = UNIT_CGROUP_BPF_INVALIDATED;
}
}
+bool unit_cgroup_delegate(Unit *u) {
+ CGroupContext *c;
+
+ assert(u);
+
+ if (!UNIT_VTABLE(u)->can_delegate)
+ return false;
+
+ c = unit_get_cgroup_context(u);
+ if (!c)
+ return false;
+
+ return c->delegate;
+}
+
void manager_invalidate_startup_units(Manager *m) {
Iterator i;
Unit *u;
~ianmdlvl / elogind.git / blobdiff