systemd/src/core/cgroup.h

/* SPDX-License-Identifier: LGPL-2.1-or-later */
#pragma once

#include "cgroup-util.h"
#include "core-forward.h"
#include "cpu-set-util.h"
#include "firewall-util.h"
#include "list.h"
#include "psi-util.h"

typedef struct CGroupTasksMax {
        /* If scale == 0, just use value; otherwise, value / scale.
         * See tasks_max_resolve(). */
        uint64_t value;
        uint64_t scale;
} CGroupTasksMax;

#define CGROUP_TASKS_MAX_UNSET ((const CGroupTasksMax) { .value = UINT64_MAX, .scale = 0 })

static inline bool cgroup_tasks_max_isset(const CGroupTasksMax *tasks_max) {
        return tasks_max->value != UINT64_MAX || tasks_max->scale != 0;
}

uint64_t cgroup_tasks_max_resolve(const CGroupTasksMax *tasks_max);

typedef enum CGroupDevicePolicy {
        /* When devices listed, will allow those, plus built-in ones, if none are listed will allow
         * everything. */
        CGROUP_DEVICE_POLICY_AUTO,

        /* Everything forbidden, except built-in ones and listed ones. */
        CGROUP_DEVICE_POLICY_CLOSED,

        /* Everything forbidden, except for the listed devices */
        CGROUP_DEVICE_POLICY_STRICT,

        _CGROUP_DEVICE_POLICY_MAX,
        _CGROUP_DEVICE_POLICY_INVALID = -EINVAL,
} CGroupDevicePolicy;

typedef enum FreezerAction {
        FREEZER_FREEZE,
        FREEZER_PARENT_FREEZE,
        FREEZER_THAW,
        FREEZER_PARENT_THAW,
        _FREEZER_ACTION_MAX,
        _FREEZER_ACTION_INVALID = -EINVAL,
} FreezerAction;

typedef enum CPUSetPartition {
        CPUSET_PARTITION_MEMBER,
        CPUSET_PARTITION_ROOT,
        CPUSET_PARTITION_ISOLATED,
        _CPUSET_PARTITION_MAX,
        _CPUSET_PARTITION_INVALID = -EINVAL,
} CPUSetPartition;

typedef enum CGroupDevicePermissions {
        /* We reuse the same bit meanings the kernel's BPF_DEVCG_ACC_xyz definitions use */
        CGROUP_DEVICE_MKNOD                = 1 << 0,
        CGROUP_DEVICE_READ                 = 1 << 1,
        CGROUP_DEVICE_WRITE                = 1 << 2,
        _CGROUP_DEVICE_PERMISSIONS_MAX     = 1 << 3,
        _CGROUP_DEVICE_PERMISSIONS_ALL     = _CGROUP_DEVICE_PERMISSIONS_MAX - 1,
        _CGROUP_DEVICE_PERMISSIONS_INVALID = -EINVAL,
} CGroupDevicePermissions;

typedef struct CGroupDeviceAllow {
        LIST_FIELDS(CGroupDeviceAllow, device_allow);
        char *path;
        CGroupDevicePermissions permissions;
} CGroupDeviceAllow;

typedef struct CGroupIODeviceWeight {
        LIST_FIELDS(CGroupIODeviceWeight, device_weights);
        char *path;
        uint64_t weight;
} CGroupIODeviceWeight;

typedef struct CGroupIODeviceLimit {
        LIST_FIELDS(CGroupIODeviceLimit, device_limits);
        char *path;
        uint64_t limits[_CGROUP_IO_LIMIT_TYPE_MAX];
} CGroupIODeviceLimit;

typedef struct CGroupIODeviceLatency {
        LIST_FIELDS(CGroupIODeviceLatency, device_latencies);
        char *path;
        usec_t target_usec;
} CGroupIODeviceLatency;

typedef struct CGroupBPFForeignProgram {
        LIST_FIELDS(CGroupBPFForeignProgram, programs);
        uint32_t attach_type;
        char *bpffs_path;
} CGroupBPFForeignProgram;

typedef struct CGroupSocketBindItem {
        LIST_FIELDS(CGroupSocketBindItem, socket_bind_items);
        int address_family;
        int ip_protocol;
        uint16_t nr_ports;
        uint16_t port_min;
} CGroupSocketBindItem;

typedef enum CGroupPressureWatch {
        CGROUP_PRESSURE_WATCH_NO,       /* → tells the service payload explicitly not to watch for pressure */
        CGROUP_PRESSURE_WATCH_YES,
        CGROUP_PRESSURE_WATCH_AUTO,     /* → on if relevant accounting is on anyway for the unit, otherwise off */
        CGROUP_PRESSURE_WATCH_SKIP,     /* → doesn't set up pressure watch, but also doesn't explicitly tell payload to avoid it */
        _CGROUP_PRESSURE_WATCH_MAX,
        _CGROUP_PRESSURE_WATCH_INVALID = -EINVAL,
} CGroupPressureWatch;

typedef struct CGroupPressure {
        CGroupPressureWatch watch;
        usec_t threshold_usec;
} CGroupPressure;

/* The user-supplied cgroup-related configuration options. This remains mostly immutable while the service
 * manager is running (except for an occasional SetProperties() configuration change), outside of reload
 * cycles. */
typedef struct CGroupContext {
        bool io_accounting;
        bool memory_accounting;
        bool tasks_accounting;
        bool ip_accounting;

        /* Configures the memory.oom.group attribute (on unified) */
        bool memory_oom_group;

        bool delegate;
        CGroupMask delegate_controllers;
        CGroupMask disable_controllers;
        char *delegate_subgroup;

        /* For unified hierarchy */
        uint64_t cpu_weight;
        uint64_t startup_cpu_weight;
        usec_t cpu_quota_per_sec_usec;
        usec_t cpu_quota_period_usec;

        CPUSet cpuset_cpus;
        CPUSet startup_cpuset_cpus;
        CPUSet cpuset_mems;
        CPUSet startup_cpuset_mems;
        CPUSetPartition cpuset_partition;

        uint64_t io_weight;
        uint64_t startup_io_weight;
        LIST_HEAD(CGroupIODeviceWeight, io_device_weights);
        LIST_HEAD(CGroupIODeviceLimit, io_device_limits);
        LIST_HEAD(CGroupIODeviceLatency, io_device_latencies);

        uint64_t memory_min;
        uint64_t memory_low;
        uint64_t startup_memory_low;
        uint64_t memory_high;
        uint64_t startup_memory_high;
        uint64_t memory_max;
        uint64_t startup_memory_max;
        uint64_t memory_swap_max;
        uint64_t startup_memory_swap_max;
        uint64_t memory_zswap_max;
        uint64_t startup_memory_zswap_max;

        bool startup_memory_low_set:1;
        bool startup_memory_high_set:1;
        bool startup_memory_max_set:1;
        bool startup_memory_swap_max_set:1;
        bool startup_memory_zswap_max_set:1;

        bool memory_zswap_writeback;

        Set *ip_address_allow;
        Set *ip_address_deny;
        /* These two flags indicate that redundant entries have been removed from
         * ip_address_allow/ip_address_deny, i.e. in_addr_prefixes_reduce() has already been called. */
        bool ip_address_allow_reduced;
        bool ip_address_deny_reduced;

        char **ip_filters_ingress;
        char **ip_filters_egress;
        LIST_HEAD(CGroupBPFForeignProgram, bpf_foreign_programs);

        Set *restrict_network_interfaces;
        bool restrict_network_interfaces_is_allow_list;

        CGroupDevicePolicy device_policy;
        LIST_HEAD(CGroupDeviceAllow, device_allow);

        LIST_HEAD(CGroupSocketBindItem, socket_bind_allow);
        LIST_HEAD(CGroupSocketBindItem, socket_bind_deny);

        char *bind_network_interface;

        /* Common */
        CGroupTasksMax tasks_max;

        /* Settings for systemd-oomd */
        ManagedOOMMode moom_swap;
        ManagedOOMMode moom_mem_pressure;
        uint32_t moom_mem_pressure_limit; /* Normalized to 2^32-1 == 100% */
        usec_t moom_mem_pressure_duration_usec;
        ManagedOOMPreference moom_preference;
        char **moom_rules;

        /* Pressure logic */
        CGroupPressure pressure[_PRESSURE_RESOURCE_MAX];

        NFTSetContext nft_set_context;

        /* Forward coredumps for processes that crash within this cgroup.
         * Requires 'delegate' to also be true. */
        bool coredump_receive;
} CGroupContext;

/* Used when querying IP accounting data */
typedef enum CGroupIPAccountingMetric {
        CGROUP_IP_INGRESS_BYTES,
        CGROUP_IP_INGRESS_PACKETS,
        CGROUP_IP_EGRESS_BYTES,
        CGROUP_IP_EGRESS_PACKETS,
        _CGROUP_IP_ACCOUNTING_METRIC_MAX,
        _CGROUP_IP_ACCOUNTING_METRIC_INVALID = -EINVAL,
} CGroupIPAccountingMetric;

/* Used when querying IO accounting data */
typedef enum CGroupIOAccountingMetric {
        CGROUP_IO_READ_BYTES,
        CGROUP_IO_WRITE_BYTES,
        CGROUP_IO_READ_OPERATIONS,
        CGROUP_IO_WRITE_OPERATIONS,
        _CGROUP_IO_ACCOUNTING_METRIC_MAX,
        _CGROUP_IO_ACCOUNTING_METRIC_INVALID = -EINVAL,
} CGroupIOAccountingMetric;

typedef enum CGroupMemoryAccountingMetric {
        CGROUP_MEMORY_PEAK,
        CGROUP_MEMORY_SWAP_PEAK,
        /* We cache the above attributes, so that they can be fetched even after the cgroup is gone, e.g.
         * when systemd-run exits. */
        _CGROUP_MEMORY_ACCOUNTING_METRIC_CACHED_LAST = CGROUP_MEMORY_SWAP_PEAK,

        /* These attributes are transient, so no need for caching. */
        CGROUP_MEMORY_CURRENT,
        CGROUP_MEMORY_SWAP_CURRENT,
        CGROUP_MEMORY_ZSWAP_CURRENT,

        _CGROUP_MEMORY_ACCOUNTING_METRIC_MAX,
        _CGROUP_MEMORY_ACCOUNTING_METRIC_INVALID = -EINVAL,
} CGroupMemoryAccountingMetric;

/* Used for limits whose value sets have infimum */
typedef enum CGroupLimitType {
        CGROUP_LIMIT_MEMORY_MAX,
        CGROUP_LIMIT_MEMORY_HIGH,
        CGROUP_LIMIT_TASKS_MAX,
        _CGROUP_LIMIT_TYPE_MAX,
        _CGROUP_LIMIT_INVALID = -EINVAL,
} CGroupLimitType;

/* The dynamic, regular updated information about a unit that as a realized cgroup. This is only allocated when a unit is first realized */
typedef struct CGroupRuntime {
        /* Where the cpu.stat or cpuacct.usage was at the time the unit was started */
        nsec_t cpu_usage_base;
        nsec_t cpu_usage_last; /* the most recently read value */

        /* Most recently read value of memory accounting metrics */
        uint64_t memory_accounting_last[_CGROUP_MEMORY_ACCOUNTING_METRIC_CACHED_LAST + 1];

        /* The current counter of OOM kills initiated by systemd-oomd */
        uint64_t managed_oom_kill_last;

        /* The current counter of the oom_kill field in the memory.events cgroup attribute */
        uint64_t oom_kill_last;

        /* Where the io.stat data was at the time the unit was started */
        uint64_t io_accounting_base[_CGROUP_IO_ACCOUNTING_METRIC_MAX];
        uint64_t io_accounting_last[_CGROUP_IO_ACCOUNTING_METRIC_MAX]; /* the most recently read value */

        /* Counterparts in the cgroup filesystem */
        char *cgroup_path;
        uint64_t cgroup_id;
        CGroupMask cgroup_realized_mask;           /* In which hierarchies does this unit's cgroup exist? (only relevant on cgroup v1) */
        CGroupMask cgroup_enabled_mask;            /* Which controllers are enabled (or more correctly: enabled for the children) for this unit's cgroup? (only relevant on cgroup v2) */
        CGroupMask cgroup_invalidated_mask;        /* A mask specifying controllers which shall be considered invalidated, and require re-realization */
        CGroupMask cgroup_members_mask;            /* A cache for the controllers required by all children of this cgroup (only relevant for slice units) */

        /* Inotify watch descriptors for watching cgroup.events and memory.events on cgroupv2 */
        int cgroup_control_inotify_wd;
        int cgroup_memory_inotify_wd;

        /* Device Controller BPF program */
        BPFProgram *bpf_device_control_installed;

        /* IP BPF Firewalling/accounting */
        int ip_accounting_ingress_map_fd;
        int ip_accounting_egress_map_fd;
        uint64_t ip_accounting_extra[_CGROUP_IP_ACCOUNTING_METRIC_MAX];

        int ipv4_allow_map_fd;
        int ipv6_allow_map_fd;
        int ipv4_deny_map_fd;
        int ipv6_deny_map_fd;
        BPFProgram *ip_bpf_ingress, *ip_bpf_ingress_installed;
        BPFProgram *ip_bpf_egress, *ip_bpf_egress_installed;

        Set *ip_bpf_custom_ingress;
        Set *ip_bpf_custom_ingress_installed;
        Set *ip_bpf_custom_egress;
        Set *ip_bpf_custom_egress_installed;

        /* BPF programs managed (e.g. loaded to kernel) by an entity external to systemd,
         * attached to unit cgroup by provided program fd and attach type. */
        Hashmap *bpf_foreign_by_key;

        FDSet *initial_socket_bind_link_fds;
#if BPF_FRAMEWORK
        /* BPF links to BPF programs attached to cgroup/bind{4|6} hooks and
         * responsible for allowing or denying a unit to bind(2) to a socket
         * address. */
        struct bpf_link *ipv4_socket_bind_link;
        struct bpf_link *ipv6_socket_bind_link;
#endif

        FDSet *initial_restrict_ifaces_link_fds;
#if BPF_FRAMEWORK
        struct bpf_link *restrict_ifaces_ingress_bpf_link;
        struct bpf_link *restrict_ifaces_egress_bpf_link;
#endif

#if BPF_FRAMEWORK
        /* BPF link to BPF programs attached to cgroup/sock_create hooks and
         * responsible for binding created sockets to a given VRF interface. */
        struct bpf_link *bpf_bind_network_interface_link;
#endif
        int initial_bind_network_interface_link_fd;

        bool cgroup_members_mask_valid:1;

        /* Reset cgroup accounting next time we fork something off */
        bool reset_accounting:1;

        /* Whether we warned about clamping the CPU quota period */
        bool warned_clamping_cpu_quota_period:1;

        int deserialized_cgroup_realized; /* tristate, for backwards compat */
} CGroupRuntime;

uint64_t cgroup_context_cpu_weight(CGroupContext *c, ManagerState state);

usec_t cgroup_cpu_adjust_period(usec_t period, usec_t quota, usec_t resolution, usec_t max_period);

void cgroup_context_init(CGroupContext *c);
void cgroup_context_done(CGroupContext *c);

void cgroup_context_dump(Unit *u, FILE* f, const char *prefix);
void cgroup_context_dump_socket_bind_item(const CGroupSocketBindItem *item, FILE *f);
void cgroup_context_dump_socket_bind_items(const CGroupSocketBindItem *items, FILE *f);

void cgroup_context_free_device_allow(CGroupContext *c, CGroupDeviceAllow *a);
void cgroup_context_free_io_device_weight(CGroupContext *c, CGroupIODeviceWeight *w);
void cgroup_context_free_io_device_limit(CGroupContext *c, CGroupIODeviceLimit *l);
void cgroup_context_free_io_device_latency(CGroupContext *c, CGroupIODeviceLatency *l);
void cgroup_context_remove_bpf_foreign_program(CGroupContext *c, CGroupBPFForeignProgram *p);
void cgroup_context_remove_socket_bind(CGroupSocketBindItem **head);

static inline bool cgroup_context_want_pressure(const CGroupContext *c, PressureResource t) {
        assert(c);
        assert(t >= 0 && t < _PRESSURE_RESOURCE_MAX);

        if (c->pressure[t].watch == CGROUP_PRESSURE_WATCH_YES)
                return true;

        if (c->pressure[t].watch != CGROUP_PRESSURE_WATCH_AUTO)
                return false;

        switch (t) {

        case PRESSURE_MEMORY:
                return c->memory_accounting;

        case PRESSURE_CPU:
                return c->cpu_weight != CGROUP_WEIGHT_INVALID ||
                        c->startup_cpu_weight != CGROUP_WEIGHT_INVALID ||
                        c->cpu_quota_per_sec_usec != USEC_INFINITY;

        case PRESSURE_IO:
                return c->io_accounting ||
                        c->io_weight != CGROUP_WEIGHT_INVALID ||
                        c->startup_io_weight != CGROUP_WEIGHT_INVALID ||
                        c->io_device_weights ||
                        c->io_device_latencies ||
                        c->io_device_limits;

        default:
                assert_not_reached();
        }
}

static inline bool cgroup_context_has_device_policy(const CGroupContext *c) {
        assert(c);

        return c->device_policy != CGROUP_DEVICE_POLICY_AUTO ||
                c->device_allow;
}

int cgroup_context_add_device_allow(CGroupContext *c, const char *dev, CGroupDevicePermissions p);
int cgroup_context_add_or_update_device_allow(CGroupContext *c, const char *dev, CGroupDevicePermissions p);
int cgroup_context_add_bpf_foreign_program(CGroupContext *c, uint32_t attach_type, const char *path);

CGroupMask unit_get_own_mask(Unit *u);
CGroupMask unit_get_delegate_mask(Unit *u);
CGroupMask unit_get_members_mask(Unit *u);
CGroupMask unit_get_siblings_mask(Unit *u);

CGroupMask unit_get_target_mask(Unit *u);
CGroupMask unit_get_enable_mask(Unit *u);

void unit_invalidate_cgroup_members_masks(Unit *u);

void unit_add_family_to_cgroup_realize_queue(Unit *u);

int unit_get_cgroup_path_with_fallback(const Unit *u, char **ret);

int unit_realize_cgroup(Unit *u);
void unit_prune_cgroup(Unit *u);
void unit_add_to_cgroup_realize_queue(Unit *u);

int unit_cgroup_is_empty(Unit *u);
void unit_release_cgroup(Unit *u, bool drop_cgroup_runtime);

int unit_check_oomd_kill(Unit *u);
int unit_check_oom(Unit *u);

int unit_attach_pids_to_cgroup(Unit *u, Set *pids, const char *suffix_path);
int unit_remove_subcgroup(Unit *u, const char *suffix_path);

int manager_setup_cgroup(Manager *m);
void manager_shutdown_cgroup(Manager *m, bool delete);

unsigned manager_dispatch_cgroup_realize_queue(Manager *m);

Unit* manager_get_unit_by_cgroup(Manager *m, const char *cgroup);
Unit* manager_get_unit_by_pidref_cgroup(Manager *m, const PidRef *pid);
Unit* manager_get_unit_by_pidref_watching(Manager *m, const PidRef *pid);
Unit* manager_get_unit_by_pidref(Manager *m, PidRef *pid);

int unit_search_main_pid(Unit *u, PidRef *ret);

int unit_get_memory_available(Unit *u, uint64_t *ret);
int unit_get_memory_accounting(Unit *u, CGroupMemoryAccountingMetric metric, uint64_t *ret);
int unit_get_tasks_current(Unit *u, uint64_t *ret);
int unit_get_cpu_usage(Unit *u, nsec_t *ret);
int unit_get_io_accounting(Unit *u, CGroupIOAccountingMetric metric, uint64_t *ret);
int unit_get_ip_accounting(Unit *u, CGroupIPAccountingMetric metric, uint64_t *ret);
int unit_get_effective_limit(Unit *u, CGroupLimitType type, uint64_t *ret);

int unit_reset_accounting(Unit *u);

#define UNIT_CGROUP_BOOL(u, name)                               \
        ({                                                      \
                CGroupContext *cc = unit_get_cgroup_context(u); \
                cc ? cc->name : false;                          \
        })

bool manager_owns_host_root_cgroup(Manager *m);
bool unit_has_host_root_cgroup(const Unit *u);

bool unit_has_startup_cgroup_constraints(Unit *u);

bool unit_invalidate_cgroup(Unit *u, CGroupMask m);
void unit_invalidate_cgroup_bpf_firewall(Unit *u);

void manager_invalidate_startup_units(Manager *m);

DECLARE_STRING_TABLE_LOOKUP(cgroup_device_policy, CGroupDevicePolicy);

void unit_cgroup_catchup(Unit *u);

bool unit_cgroup_delegate(Unit *u);

void unit_cgroup_disable_all_controllers(Unit *u);

int unit_get_cpuset(Unit *u, CPUSet *cpus, const char *name);

int unit_cgroup_freezer_action(Unit *u, FreezerAction action);

DECLARE_STRING_TABLE_LOOKUP(freezer_action, FreezerAction);
DECLARE_STRING_TABLE_LOOKUP(cpuset_partition, CPUSetPartition);

CGroupRuntime* cgroup_runtime_new(void);
CGroupRuntime* cgroup_runtime_free(CGroupRuntime *crt);
DEFINE_TRIVIAL_CLEANUP_FUNC(CGroupRuntime*, cgroup_runtime_free);

int cgroup_runtime_serialize(Unit *u, FILE *f, FDSet *fds);
int cgroup_runtime_deserialize_one(Unit *u, const char *key, const char *value, FDSet *fds);

DECLARE_STRING_TABLE_LOOKUP(cgroup_pressure_watch, CGroupPressureWatch);

DECLARE_STRING_TABLE_LOOKUP(cgroup_device_permissions, CGroupDevicePermissions);

DECLARE_STRING_TABLE_LOOKUP(cgroup_ip_accounting_metric, CGroupIPAccountingMetric);

DECLARE_STRING_TABLE_LOOKUP(cgroup_io_accounting_metric, CGroupIOAccountingMetric);

DECLARE_STRING_TABLE_LOOKUP(cgroup_effective_limit_type, CGroupLimitType);

DECLARE_STRING_TABLE_LOOKUP(cgroup_memory_accounting_metric, CGroupMemoryAccountingMetric);