[linux-block.git] / kernel / cgroup / cgroup-internal.h

/* SPDX-License-Identifier: GPL-2.0 */
#ifndef __CGROUP_INTERNAL_H
#define __CGROUP_INTERNAL_H

#include <linux/cgroup.h>
#include <linux/kernfs.h>
#include <linux/workqueue.h>
#include <linux/list.h>
#include <linux/refcount.h>

#define TRACE_CGROUP_PATH_LEN 1024
extern spinlock_t trace_cgroup_path_lock;
extern char trace_cgroup_path[TRACE_CGROUP_PATH_LEN];
extern bool cgroup_debug;
extern void __init enable_debug_cgroup(void);

/*
 * cgroup_path() takes a spin lock. It is good practice not to take
 * spin locks within trace point handlers, as they are mostly hidden
 * from normal view. As cgroup_path() can take the kernfs_rename_lock
 * spin lock, it is best to not call that function from the trace event
 * handler.
 *
 * Note: trace_cgroup_##type##_enabled() is a static branch that will only
 *       be set when the trace event is enabled.
 */
#define TRACE_CGROUP_PATH(type, cgrp, ...)				\
	do {								\
		if (trace_cgroup_##type##_enabled()) {			\
			spin_lock(&trace_cgroup_path_lock);		\
			cgroup_path(cgrp, trace_cgroup_path,		\
				    TRACE_CGROUP_PATH_LEN);		\
			trace_cgroup_##type(cgrp, trace_cgroup_path,	\
					    ##__VA_ARGS__);		\
			spin_unlock(&trace_cgroup_path_lock);		\
		}							\
	} while (0)

/*
 * A cgroup can be associated with multiple css_sets as different tasks may
 * belong to different cgroups on different hierarchies.  In the other
 * direction, a css_set is naturally associated with multiple cgroups.
 * This M:N relationship is represented by the following link structure
 * which exists for each association and allows traversing the associations
 * from both sides.
 */
struct cgrp_cset_link {
	/* the cgroup and css_set this link associates */
	struct cgroup		*cgrp;
	struct css_set		*cset;

	/* list of cgrp_cset_links anchored at cgrp->cset_links */
	struct list_head	cset_link;

	/* list of cgrp_cset_links anchored at css_set->cgrp_links */
	struct list_head	cgrp_link;
};

/* used to track tasks and csets during migration */
struct cgroup_taskset {
	/* the src and dst cset list running through cset->mg_node */
	struct list_head	src_csets;
	struct list_head	dst_csets;

	/* the number of tasks in the set */
	int			nr_tasks;

	/* the subsys currently being processed */
	int			ssid;

	/*
	 * Fields for cgroup_taskset_*() iteration.
	 *
	 * Before migration is committed, the target migration tasks are on
	 * ->mg_tasks of the csets on ->src_csets.  After, on ->mg_tasks of
	 * the csets on ->dst_csets.  ->csets point to either ->src_csets
	 * or ->dst_csets depending on whether migration is committed.
	 *
	 * ->cur_csets and ->cur_task point to the current task position
	 * during iteration.
	 */
	struct list_head	*csets;
	struct css_set		*cur_cset;
	struct task_struct	*cur_task;
};

/* migration context also tracks preloading */
struct cgroup_mgctx {
	/*
	 * Preloaded source and destination csets.  Used to guarantee
	 * atomic success or failure on actual migration.
	 */
	struct list_head	preloaded_src_csets;
	struct list_head	preloaded_dst_csets;

	/* tasks and csets to migrate */
	struct cgroup_taskset	tset;

	/* subsystems affected by migration */
	u16			ss_mask;
};

#define CGROUP_TASKSET_INIT(tset)						\
{										\
	.src_csets		= LIST_HEAD_INIT(tset.src_csets),		\
	.dst_csets		= LIST_HEAD_INIT(tset.dst_csets),		\
	.csets			= &tset.src_csets,				\
}

#define CGROUP_MGCTX_INIT(name)							\
{										\
	LIST_HEAD_INIT(name.preloaded_src_csets),				\
	LIST_HEAD_INIT(name.preloaded_dst_csets),				\
	CGROUP_TASKSET_INIT(name.tset),						\
}

#define DEFINE_CGROUP_MGCTX(name)						\
	struct cgroup_mgctx name = CGROUP_MGCTX_INIT(name)

struct cgroup_sb_opts {
	u16 subsys_mask;
	unsigned int flags;
	char *release_agent;
	bool cpuset_clone_children;
	char *name;
	/* User explicitly requested empty subsystem */
	bool none;
};

extern struct mutex cgroup_mutex;
extern spinlock_t css_set_lock;
extern struct cgroup_subsys *cgroup_subsys[];
extern struct list_head cgroup_roots;
extern struct file_system_type cgroup_fs_type;

/* iterate across the hierarchies */
#define for_each_root(root)						\
	list_for_each_entry((root), &cgroup_roots, root_list)

/**
 * for_each_subsys - iterate all enabled cgroup subsystems
 * @ss: the iteration cursor
 * @ssid: the index of @ss, CGROUP_SUBSYS_COUNT after reaching the end
 */
#define for_each_subsys(ss, ssid)					\
	for ((ssid) = 0; (ssid) < CGROUP_SUBSYS_COUNT &&		\
	     (((ss) = cgroup_subsys[ssid]) || true); (ssid)++)

static inline bool cgroup_is_dead(const struct cgroup *cgrp)
{
	return !(cgrp->self.flags & CSS_ONLINE);
}

static inline bool notify_on_release(const struct cgroup *cgrp)
{
	return test_bit(CGRP_NOTIFY_ON_RELEASE, &cgrp->flags);
}

void put_css_set_locked(struct css_set *cset);

static inline void put_css_set(struct css_set *cset)
{
	unsigned long flags;

	/*
	 * Ensure that the refcount doesn't hit zero while any readers
	 * can see it. Similar to atomic_dec_and_lock(), but for an
	 * rwlock
	 */
	if (refcount_dec_not_one(&cset->refcount))
		return;

	spin_lock_irqsave(&css_set_lock, flags);
	put_css_set_locked(cset);
	spin_unlock_irqrestore(&css_set_lock, flags);
}

/*
 * refcounted get/put for css_set objects
 */
static inline void get_css_set(struct css_set *cset)
{
	refcount_inc(&cset->refcount);
}

bool cgroup_ssid_enabled(int ssid);
bool cgroup_on_dfl(const struct cgroup *cgrp);
bool cgroup_is_thread_root(struct cgroup *cgrp);
bool cgroup_is_threaded(struct cgroup *cgrp);

struct cgroup_root *cgroup_root_from_kf(struct kernfs_root *kf_root);
struct cgroup *task_cgroup_from_root(struct task_struct *task,
				     struct cgroup_root *root);
struct cgroup *cgroup_kn_lock_live(struct kernfs_node *kn, bool drain_offline);
void cgroup_kn_unlock(struct kernfs_node *kn);
int cgroup_path_ns_locked(struct cgroup *cgrp, char *buf, size_t buflen,
			  struct cgroup_namespace *ns);

void cgroup_free_root(struct cgroup_root *root);
void init_cgroup_root(struct cgroup_root *root, struct cgroup_sb_opts *opts);
int cgroup_setup_root(struct cgroup_root *root, u16 ss_mask, int ref_flags);
int rebind_subsystems(struct cgroup_root *dst_root, u16 ss_mask);
struct dentry *cgroup_do_mount(struct file_system_type *fs_type, int flags,
			       struct cgroup_root *root, unsigned long magic,
			       struct cgroup_namespace *ns);

int cgroup_migrate_vet_dst(struct cgroup *dst_cgrp);
void cgroup_migrate_finish(struct cgroup_mgctx *mgctx);
void cgroup_migrate_add_src(struct css_set *src_cset, struct cgroup *dst_cgrp,
			    struct cgroup_mgctx *mgctx);
int cgroup_migrate_prepare_dst(struct cgroup_mgctx *mgctx);
int cgroup_migrate(struct task_struct *leader, bool threadgroup,
		   struct cgroup_mgctx *mgctx);

int cgroup_attach_task(struct cgroup *dst_cgrp, struct task_struct *leader,
		       bool threadgroup);
struct task_struct *cgroup_procs_write_start(char *buf, bool threadgroup)
	__acquires(&cgroup_threadgroup_rwsem);
void cgroup_procs_write_finish(struct task_struct *task)
	__releases(&cgroup_threadgroup_rwsem);

void cgroup_lock_and_drain_offline(struct cgroup *cgrp);

int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode);
int cgroup_rmdir(struct kernfs_node *kn);
int cgroup_show_path(struct seq_file *sf, struct kernfs_node *kf_node,
		     struct kernfs_root *kf_root);

int cgroup_task_count(const struct cgroup *cgrp);

/*
 * rstat.c
 */
int cgroup_rstat_init(struct cgroup *cgrp);
void cgroup_rstat_exit(struct cgroup *cgrp);
void cgroup_rstat_boot(void);
void cgroup_base_stat_cputime_show(struct seq_file *seq);

/*
 * namespace.c
 */
extern const struct proc_ns_operations cgroupns_operations;

/*
 * cgroup-v1.c
 */
extern struct cftype cgroup1_base_files[];
extern struct kernfs_syscall_ops cgroup1_kf_syscall_ops;

int proc_cgroupstats_show(struct seq_file *m, void *v);
bool cgroup1_ssid_disabled(int ssid);
void cgroup1_pidlist_destroy_all(struct cgroup *cgrp);
void cgroup1_release_agent(struct work_struct *work);
void cgroup1_check_for_release(struct cgroup *cgrp);
struct dentry *cgroup1_mount(struct file_system_type *fs_type, int flags,
			     void *data, unsigned long magic,
			     struct cgroup_namespace *ns);

#endif /* __CGROUP_INTERNAL_H */
Commit	Line	Data
b2441318	1	/* SPDX-License-Identifier: GPL-2.0 */
0a268dbd TH	2	#ifndef __CGROUP_INTERNAL_H
	3	#define __CGROUP_INTERNAL_H
	4
	5	#include <linux/cgroup.h>
	6	#include <linux/kernfs.h>
	7	#include <linux/workqueue.h>
	8	#include <linux/list.h>
4b9502e6	9	#include <linux/refcount.h>
0a268dbd	10
e4f8d81c SRV	11	#define TRACE_CGROUP_PATH_LEN 1024
	12	extern spinlock_t trace_cgroup_path_lock;
	13	extern char trace_cgroup_path[TRACE_CGROUP_PATH_LEN];
5cf8114d WL	14	extern bool cgroup_debug;
5cf8114d WL	15	extern void __init enable_debug_cgroup(void);
e4f8d81c SRV	16
	17	/*
	18	* cgroup_path() takes a spin lock. It is good practice not to take
	19	* spin locks within trace point handlers, as they are mostly hidden
	20	* from normal view. As cgroup_path() can take the kernfs_rename_lock
	21	* spin lock, it is best to not call that function from the trace event
	22	* handler.
	23	*
	24	* Note: trace_cgroup_##type##_enabled() is a static branch that will only
	25	* be set when the trace event is enabled.
	26	*/
	27	#define TRACE_CGROUP_PATH(type, cgrp, ...) \
	28	do { \
	29	if (trace_cgroup_##type##_enabled()) { \
	30	spin_lock(&trace_cgroup_path_lock); \
	31	cgroup_path(cgrp, trace_cgroup_path, \
	32	TRACE_CGROUP_PATH_LEN); \
	33	trace_cgroup_##type(cgrp, trace_cgroup_path, \
	34	##__VA_ARGS__); \
	35	spin_unlock(&trace_cgroup_path_lock); \
	36	} \
	37	} while (0)
	38
0a268dbd TH	39	/*
	40	* A cgroup can be associated with multiple css_sets as different tasks may
	41	* belong to different cgroups on different hierarchies. In the other
	42	* direction, a css_set is naturally associated with multiple cgroups.
	43	* This M:N relationship is represented by the following link structure
	44	* which exists for each association and allows traversing the associations
	45	* from both sides.
	46	*/
	47	struct cgrp_cset_link {
	48	/* the cgroup and css_set this link associates */
	49	struct cgroup *cgrp;
	50	struct css_set *cset;
	51
	52	/* list of cgrp_cset_links anchored at cgrp->cset_links */
	53	struct list_head cset_link;
	54
	55	/* list of cgrp_cset_links anchored at css_set->cgrp_links */
	56	struct list_head cgrp_link;
	57	};
	58
e595cd70 TH	59	/* used to track tasks and csets during migration */
	60	struct cgroup_taskset {
	61	/* the src and dst cset list running through cset->mg_node */
	62	struct list_head src_csets;
	63	struct list_head dst_csets;
	64
61046727 TH	65	/* the number of tasks in the set */
	66	int nr_tasks;
	67
e595cd70 TH	68	/* the subsys currently being processed */
	69	int ssid;
	70
	71	/*
	72	* Fields for cgroup_taskset_*() iteration.
	73	*
	74	* Before migration is committed, the target migration tasks are on
	75	* ->mg_tasks of the csets on ->src_csets. After, on ->mg_tasks of
	76	* the csets on ->dst_csets. ->csets point to either ->src_csets
	77	* or ->dst_csets depending on whether migration is committed.
	78	*
	79	* ->cur_csets and ->cur_task point to the current task position
	80	* during iteration.
	81	*/
	82	struct list_head *csets;
	83	struct css_set *cur_cset;
	84	struct task_struct *cur_task;
	85	};
	86
	87	/* migration context also tracks preloading */
	88	struct cgroup_mgctx {
	89	/*
	90	* Preloaded source and destination csets. Used to guarantee
	91	* atomic success or failure on actual migration.
	92	*/
	93	struct list_head preloaded_src_csets;
	94	struct list_head preloaded_dst_csets;
	95
	96	/* tasks and csets to migrate */
	97	struct cgroup_taskset tset;
bfc2cf6f TH	98
	99	/* subsystems affected by migration */
	100	u16 ss_mask;
e595cd70 TH	101	};
	102
	103	#define CGROUP_TASKSET_INIT(tset) \
	104	{ \
	105	.src_csets = LIST_HEAD_INIT(tset.src_csets), \
	106	.dst_csets = LIST_HEAD_INIT(tset.dst_csets), \
	107	.csets = &tset.src_csets, \
	108	}
	109
	110	#define CGROUP_MGCTX_INIT(name) \
	111	{ \
	112	LIST_HEAD_INIT(name.preloaded_src_csets), \
	113	LIST_HEAD_INIT(name.preloaded_dst_csets), \
	114	CGROUP_TASKSET_INIT(name.tset), \
	115	}
	116
	117	#define DEFINE_CGROUP_MGCTX(name) \
	118	struct cgroup_mgctx name = CGROUP_MGCTX_INIT(name)
	119
1592c9b2 TH	120	struct cgroup_sb_opts {
	121	u16 subsys_mask;
	122	unsigned int flags;
	123	char *release_agent;
	124	bool cpuset_clone_children;
	125	char *name;
	126	/* User explicitly requested empty subsystem */
	127	bool none;
	128	};
	129
0a268dbd TH	130	extern struct mutex cgroup_mutex;
	131	extern spinlock_t css_set_lock;
	132	extern struct cgroup_subsys *cgroup_subsys[];
	133	extern struct list_head cgroup_roots;
	134	extern struct file_system_type cgroup_fs_type;
	135
	136	/* iterate across the hierarchies */
	137	#define for_each_root(root) \
	138	list_for_each_entry((root), &cgroup_roots, root_list)
	139
	140	/**
	141	* for_each_subsys - iterate all enabled cgroup subsystems
	142	* @ss: the iteration cursor
	143	* @ssid: the index of @ss, CGROUP_SUBSYS_COUNT after reaching the end
	144	*/
	145	#define for_each_subsys(ss, ssid) \
	146	for ((ssid) = 0; (ssid) < CGROUP_SUBSYS_COUNT && \
	147	(((ss) = cgroup_subsys[ssid]) \|\| true); (ssid)++)
	148
	149	static inline bool cgroup_is_dead(const struct cgroup *cgrp)
	150	{
	151	return !(cgrp->self.flags & CSS_ONLINE);
	152	}
	153
	154	static inline bool notify_on_release(const struct cgroup *cgrp)
	155	{
	156	return test_bit(CGRP_NOTIFY_ON_RELEASE, &cgrp->flags);
	157	}
	158
dcfe149b TH	159	void put_css_set_locked(struct css_set *cset);
	160
	161	static inline void put_css_set(struct css_set *cset)
	162	{
	163	unsigned long flags;
	164
	165	/*
	166	* Ensure that the refcount doesn't hit zero while any readers
	167	* can see it. Similar to atomic_dec_and_lock(), but for an
	168	* rwlock
	169	*/
4b9502e6	170	if (refcount_dec_not_one(&cset->refcount))
dcfe149b TH	171	return;
	172
	173	spin_lock_irqsave(&css_set_lock, flags);
	174	put_css_set_locked(cset);
	175	spin_unlock_irqrestore(&css_set_lock, flags);
	176	}
	177
	178	/*
	179	* refcounted get/put for css_set objects
	180	*/
	181	static inline void get_css_set(struct css_set *cset)
	182	{
4b9502e6	183	refcount_inc(&cset->refcount);
dcfe149b TH	184	}
dcfe149b TH	185
0a268dbd TH	186	bool cgroup_ssid_enabled(int ssid);
0a268dbd TH	187	bool cgroup_on_dfl(const struct cgroup *cgrp);
7a0cf0e7 WL	188	bool cgroup_is_thread_root(struct cgroup *cgrp);
7a0cf0e7 WL	189	bool cgroup_is_threaded(struct cgroup *cgrp);
0a268dbd TH	190
	191	struct cgroup_root cgroup_root_from_kf(struct kernfs_root kf_root);
	192	struct cgroup task_cgroup_from_root(struct task_struct task,
	193	struct cgroup_root *root);
	194	struct cgroup cgroup_kn_lock_live(struct kernfs_node kn, bool drain_offline);
	195	void cgroup_kn_unlock(struct kernfs_node *kn);
	196	int cgroup_path_ns_locked(struct cgroup cgrp, char buf, size_t buflen,
	197	struct cgroup_namespace *ns);
	198
1592c9b2 TH	199	void cgroup_free_root(struct cgroup_root *root);
1592c9b2 TH	200	void init_cgroup_root(struct cgroup_root root, struct cgroup_sb_opts opts);
9732adc5	201	int cgroup_setup_root(struct cgroup_root *root, u16 ss_mask, int ref_flags);
0a268dbd	202	int rebind_subsystems(struct cgroup_root *dst_root, u16 ss_mask);
1592c9b2 TH	203	struct dentry cgroup_do_mount(struct file_system_type fs_type, int flags,
	204	struct cgroup_root *root, unsigned long magic,
	205	struct cgroup_namespace *ns);
0a268dbd	206
8cfd8147	207	int cgroup_migrate_vet_dst(struct cgroup *dst_cgrp);
e595cd70 TH	208	void cgroup_migrate_finish(struct cgroup_mgctx *mgctx);
	209	void cgroup_migrate_add_src(struct css_set src_cset, struct cgroup dst_cgrp,
	210	struct cgroup_mgctx *mgctx);
	211	int cgroup_migrate_prepare_dst(struct cgroup_mgctx *mgctx);
0a268dbd	212	int cgroup_migrate(struct task_struct *leader, bool threadgroup,
bfc2cf6f	213	struct cgroup_mgctx *mgctx);
0a268dbd TH	214
	215	int cgroup_attach_task(struct cgroup dst_cgrp, struct task_struct leader,
	216	bool threadgroup);
715c809d TH	217	struct task_struct cgroup_procs_write_start(char buf, bool threadgroup)
	218	__acquires(&cgroup_threadgroup_rwsem);
	219	void cgroup_procs_write_finish(struct task_struct *task)
	220	__releases(&cgroup_threadgroup_rwsem);
0a268dbd TH	221
	222	void cgroup_lock_and_drain_offline(struct cgroup *cgrp);
	223
1592c9b2 TH	224	int cgroup_mkdir(struct kernfs_node parent_kn, const char name, umode_t mode);
	225	int cgroup_rmdir(struct kernfs_node *kn);
	226	int cgroup_show_path(struct seq_file sf, struct kernfs_node kf_node,
	227	struct kernfs_root *kf_root);
	228
a28f8f5e WL	229	int cgroup_task_count(const struct cgroup *cgrp);
a28f8f5e WL	230
041cd640	231	/*
c58632b3	232	* rstat.c
041cd640	233	*/
c58632b3 TH	234	int cgroup_rstat_init(struct cgroup *cgrp);
c58632b3 TH	235	void cgroup_rstat_exit(struct cgroup *cgrp);
c58632b3	236	void cgroup_rstat_boot(void);
a17556f8	237	void cgroup_base_stat_cputime_show(struct seq_file *seq);
041cd640	238
dcfe149b TH	239	/*
	240	* namespace.c
	241	*/
	242	extern const struct proc_ns_operations cgroupns_operations;
	243
0a268dbd TH	244	/*
	245	* cgroup-v1.c
	246	*/
d62beb7f	247	extern struct cftype cgroup1_base_files[];
1592c9b2	248	extern struct kernfs_syscall_ops cgroup1_kf_syscall_ops;
0a268dbd	249
3f3942ac	250	int proc_cgroupstats_show(struct seq_file m, void v);
d62beb7f TH	251	bool cgroup1_ssid_disabled(int ssid);
	252	void cgroup1_pidlist_destroy_all(struct cgroup *cgrp);
	253	void cgroup1_release_agent(struct work_struct *work);
	254	void cgroup1_check_for_release(struct cgroup *cgrp);
1592c9b2 TH	255	struct dentry cgroup1_mount(struct file_system_type fs_type, int flags,
	256	void *data, unsigned long magic,
	257	struct cgroup_namespace *ns);
0a268dbd TH	258
0a268dbd TH	259	#endif /* __CGROUP_INTERNAL_H */