[linux-2.6-block.git] / kernel / ucount.c

// SPDX-License-Identifier: GPL-2.0-only

#include <linux/stat.h>
#include <linux/sysctl.h>
#include <linux/slab.h>
#include <linux/cred.h>
#include <linux/hash.h>
#include <linux/kmemleak.h>
#include <linux/user_namespace.h>

struct ucounts init_ucounts = {
	.ns    = &init_user_ns,
	.uid   = GLOBAL_ROOT_UID,
	.count = ATOMIC_INIT(1),
};

#define UCOUNTS_HASHTABLE_BITS 10
static struct hlist_head ucounts_hashtable[(1 << UCOUNTS_HASHTABLE_BITS)];
static DEFINE_SPINLOCK(ucounts_lock);

#define ucounts_hashfn(ns, uid)						\
	hash_long((unsigned long)__kuid_val(uid) + (unsigned long)(ns), \
		  UCOUNTS_HASHTABLE_BITS)
#define ucounts_hashentry(ns, uid)	\
	(ucounts_hashtable + ucounts_hashfn(ns, uid))


#ifdef CONFIG_SYSCTL
static struct ctl_table_set *
set_lookup(struct ctl_table_root *root)
{
	return &current_user_ns()->set;
}

static int set_is_seen(struct ctl_table_set *set)
{
	return &current_user_ns()->set == set;
}

static int set_permissions(struct ctl_table_header *head,
				  struct ctl_table *table)
{
	struct user_namespace *user_ns =
		container_of(head->set, struct user_namespace, set);
	int mode;

	/* Allow users with CAP_SYS_RESOURCE unrestrained access */
	if (ns_capable(user_ns, CAP_SYS_RESOURCE))
		mode = (table->mode & S_IRWXU) >> 6;
	else
	/* Allow all others at most read-only access */
		mode = table->mode & S_IROTH;
	return (mode << 6) | (mode << 3) | mode;
}

static struct ctl_table_root set_root = {
	.lookup = set_lookup,
	.permissions = set_permissions,
};

static long ue_zero = 0;
static long ue_int_max = INT_MAX;

#define UCOUNT_ENTRY(name)					\
	{							\
		.procname	= name,				\
		.maxlen		= sizeof(long),			\
		.mode		= 0644,				\
		.proc_handler	= proc_doulongvec_minmax,	\
		.extra1		= &ue_zero,			\
		.extra2		= &ue_int_max,			\
	}
static struct ctl_table user_table[] = {
	UCOUNT_ENTRY("max_user_namespaces"),
	UCOUNT_ENTRY("max_pid_namespaces"),
	UCOUNT_ENTRY("max_uts_namespaces"),
	UCOUNT_ENTRY("max_ipc_namespaces"),
	UCOUNT_ENTRY("max_net_namespaces"),
	UCOUNT_ENTRY("max_mnt_namespaces"),
	UCOUNT_ENTRY("max_cgroup_namespaces"),
	UCOUNT_ENTRY("max_time_namespaces"),
#ifdef CONFIG_INOTIFY_USER
	UCOUNT_ENTRY("max_inotify_instances"),
	UCOUNT_ENTRY("max_inotify_watches"),
#endif
#ifdef CONFIG_FANOTIFY
	UCOUNT_ENTRY("max_fanotify_groups"),
	UCOUNT_ENTRY("max_fanotify_marks"),
#endif
	{ }
};
#endif /* CONFIG_SYSCTL */

bool setup_userns_sysctls(struct user_namespace *ns)
{
#ifdef CONFIG_SYSCTL
	struct ctl_table *tbl;

	BUILD_BUG_ON(ARRAY_SIZE(user_table) != UCOUNT_COUNTS + 1);
	setup_sysctl_set(&ns->set, &set_root, set_is_seen);
	tbl = kmemdup(user_table, sizeof(user_table), GFP_KERNEL);
	if (tbl) {
		int i;
		for (i = 0; i < UCOUNT_COUNTS; i++) {
			tbl[i].data = &ns->ucount_max[i];
		}
		ns->sysctls = __register_sysctl_table(&ns->set, "user", tbl);
	}
	if (!ns->sysctls) {
		kfree(tbl);
		retire_sysctl_set(&ns->set);
		return false;
	}
#endif
	return true;
}

void retire_userns_sysctls(struct user_namespace *ns)
{
#ifdef CONFIG_SYSCTL
	struct ctl_table *tbl;

	tbl = ns->sysctls->ctl_table_arg;
	unregister_sysctl_table(ns->sysctls);
	retire_sysctl_set(&ns->set);
	kfree(tbl);
#endif
}

static struct ucounts *find_ucounts(struct user_namespace *ns, kuid_t uid, struct hlist_head *hashent)
{
	struct ucounts *ucounts;

	hlist_for_each_entry(ucounts, hashent, node) {
		if (uid_eq(ucounts->uid, uid) && (ucounts->ns == ns))
			return ucounts;
	}
	return NULL;
}

static void hlist_add_ucounts(struct ucounts *ucounts)
{
	struct hlist_head *hashent = ucounts_hashentry(ucounts->ns, ucounts->uid);
	spin_lock_irq(&ucounts_lock);
	hlist_add_head(&ucounts->node, hashent);
	spin_unlock_irq(&ucounts_lock);
}

static inline bool get_ucounts_or_wrap(struct ucounts *ucounts)
{
	/* Returns true on a successful get, false if the count wraps. */
	return !atomic_add_negative(1, &ucounts->count);
}

struct ucounts *get_ucounts(struct ucounts *ucounts)
{
	if (!get_ucounts_or_wrap(ucounts)) {
		put_ucounts(ucounts);
		ucounts = NULL;
	}
	return ucounts;
}

struct ucounts *alloc_ucounts(struct user_namespace *ns, kuid_t uid)
{
	struct hlist_head *hashent = ucounts_hashentry(ns, uid);
	struct ucounts *ucounts, *new;
	bool wrapped;

	spin_lock_irq(&ucounts_lock);
	ucounts = find_ucounts(ns, uid, hashent);
	if (!ucounts) {
		spin_unlock_irq(&ucounts_lock);

		new = kzalloc(sizeof(*new), GFP_KERNEL);
		if (!new)
			return NULL;

		new->ns = ns;
		new->uid = uid;
		atomic_set(&new->count, 1);

		spin_lock_irq(&ucounts_lock);
		ucounts = find_ucounts(ns, uid, hashent);
		if (ucounts) {
			kfree(new);
		} else {
			hlist_add_head(&new->node, hashent);
			get_user_ns(new->ns);
			spin_unlock_irq(&ucounts_lock);
			return new;
		}
	}
	wrapped = !get_ucounts_or_wrap(ucounts);
	spin_unlock_irq(&ucounts_lock);
	if (wrapped) {
		put_ucounts(ucounts);
		return NULL;
	}
	return ucounts;
}

void put_ucounts(struct ucounts *ucounts)
{
	unsigned long flags;

	if (atomic_dec_and_lock_irqsave(&ucounts->count, &ucounts_lock, flags)) {
		hlist_del_init(&ucounts->node);
		spin_unlock_irqrestore(&ucounts_lock, flags);
		put_user_ns(ucounts->ns);
		kfree(ucounts);
	}
}

static inline bool atomic_long_inc_below(atomic_long_t *v, int u)
{
	long c, old;
	c = atomic_long_read(v);
	for (;;) {
		if (unlikely(c >= u))
			return false;
		old = atomic_long_cmpxchg(v, c, c+1);
		if (likely(old == c))
			return true;
		c = old;
	}
}

struct ucounts *inc_ucount(struct user_namespace *ns, kuid_t uid,
			   enum ucount_type type)
{
	struct ucounts *ucounts, *iter, *bad;
	struct user_namespace *tns;
	ucounts = alloc_ucounts(ns, uid);
	for (iter = ucounts; iter; iter = tns->ucounts) {
		long max;
		tns = iter->ns;
		max = READ_ONCE(tns->ucount_max[type]);
		if (!atomic_long_inc_below(&iter->ucount[type], max))
			goto fail;
	}
	return ucounts;
fail:
	bad = iter;
	for (iter = ucounts; iter != bad; iter = iter->ns->ucounts)
		atomic_long_dec(&iter->ucount[type]);

	put_ucounts(ucounts);
	return NULL;
}

void dec_ucount(struct ucounts *ucounts, enum ucount_type type)
{
	struct ucounts *iter;
	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
		long dec = atomic_long_dec_if_positive(&iter->ucount[type]);
		WARN_ON_ONCE(dec < 0);
	}
	put_ucounts(ucounts);
}

long inc_rlimit_ucounts(struct ucounts *ucounts, enum rlimit_type type, long v)
{
	struct ucounts *iter;
	long max = LONG_MAX;
	long ret = 0;

	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
		long new = atomic_long_add_return(v, &iter->rlimit[type]);
		if (new < 0 || new > max)
			ret = LONG_MAX;
		else if (iter == ucounts)
			ret = new;
		max = get_userns_rlimit_max(iter->ns, type);
	}
	return ret;
}

bool dec_rlimit_ucounts(struct ucounts *ucounts, enum rlimit_type type, long v)
{
	struct ucounts *iter;
	long new = -1; /* Silence compiler warning */
	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
		long dec = atomic_long_sub_return(v, &iter->rlimit[type]);
		WARN_ON_ONCE(dec < 0);
		if (iter == ucounts)
			new = dec;
	}
	return (new == 0);
}

static void do_dec_rlimit_put_ucounts(struct ucounts *ucounts,
				struct ucounts *last, enum rlimit_type type)
{
	struct ucounts *iter, *next;
	for (iter = ucounts; iter != last; iter = next) {
		long dec = atomic_long_sub_return(1, &iter->rlimit[type]);
		WARN_ON_ONCE(dec < 0);
		next = iter->ns->ucounts;
		if (dec == 0)
			put_ucounts(iter);
	}
}

void dec_rlimit_put_ucounts(struct ucounts *ucounts, enum rlimit_type type)
{
	do_dec_rlimit_put_ucounts(ucounts, NULL, type);
}

long inc_rlimit_get_ucounts(struct ucounts *ucounts, enum rlimit_type type)
{
	/* Caller must hold a reference to ucounts */
	struct ucounts *iter;
	long max = LONG_MAX;
	long dec, ret = 0;

	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
		long new = atomic_long_add_return(1, &iter->rlimit[type]);
		if (new < 0 || new > max)
			goto unwind;
		if (iter == ucounts)
			ret = new;
		max = get_userns_rlimit_max(iter->ns, type);
		/*
		 * Grab an extra ucount reference for the caller when
		 * the rlimit count was previously 0.
		 */
		if (new != 1)
			continue;
		if (!get_ucounts(iter))
			goto dec_unwind;
	}
	return ret;
dec_unwind:
	dec = atomic_long_sub_return(1, &iter->rlimit[type]);
	WARN_ON_ONCE(dec < 0);
unwind:
	do_dec_rlimit_put_ucounts(ucounts, iter, type);
	return 0;
}

bool is_rlimit_overlimit(struct ucounts *ucounts, enum rlimit_type type, unsigned long rlimit)
{
	struct ucounts *iter;
	long max = rlimit;
	if (rlimit > LONG_MAX)
		max = LONG_MAX;
	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
		long val = get_rlimit_value(iter, type);
		if (val < 0 || val > max)
			return true;
		max = get_userns_rlimit_max(iter->ns, type);
	}
	return false;
}

static __init int user_namespace_sysctl_init(void)
{
#ifdef CONFIG_SYSCTL
	static struct ctl_table_header *user_header;
	static struct ctl_table empty[1];
	/*
	 * It is necessary to register the user directory in the
	 * default set so that registrations in the child sets work
	 * properly.
	 */
	user_header = register_sysctl("user", empty);
	kmemleak_ignore(user_header);
	BUG_ON(!user_header);
	BUG_ON(!setup_userns_sysctls(&init_user_ns));
#endif
	hlist_add_ucounts(&init_ucounts);
	inc_rlimit_ucounts(&init_ucounts, UCOUNT_RLIMIT_NPROC, 1);
	return 0;
}
subsys_initcall(user_namespace_sysctl_init);
Commit	Line	Data
b886d83c	1	// SPDX-License-Identifier: GPL-2.0-only
dbec2846 EB	2
	3	#include <linux/stat.h>
	4	#include <linux/sysctl.h>
	5	#include <linux/slab.h>
5b825c3a	6	#include <linux/cred.h>
f6b2db1a	7	#include <linux/hash.h>
514c6032	8	#include <linux/kmemleak.h>
dbec2846 EB	9	#include <linux/user_namespace.h>
dbec2846 EB	10
905ae01c AG	11	struct ucounts init_ucounts = {
	12	.ns = &init_user_ns,
	13	.uid = GLOBAL_ROOT_UID,
b6c33652	14	.count = ATOMIC_INIT(1),
905ae01c AG	15	};
905ae01c AG	16
f6b2db1a EB	17	#define UCOUNTS_HASHTABLE_BITS 10
	18	static struct hlist_head ucounts_hashtable[(1 << UCOUNTS_HASHTABLE_BITS)];
	19	static DEFINE_SPINLOCK(ucounts_lock);
	20
	21	#define ucounts_hashfn(ns, uid) \
	22	hash_long((unsigned long)__kuid_val(uid) + (unsigned long)(ns), \
	23	UCOUNTS_HASHTABLE_BITS)
	24	#define ucounts_hashentry(ns, uid) \
	25	(ucounts_hashtable + ucounts_hashfn(ns, uid))
	26
	27
dbec2846 EB	28	#ifdef CONFIG_SYSCTL
	29	static struct ctl_table_set *
	30	set_lookup(struct ctl_table_root *root)
	31	{
	32	return &current_user_ns()->set;
	33	}
	34
	35	static int set_is_seen(struct ctl_table_set *set)
	36	{
	37	return &current_user_ns()->set == set;
	38	}
	39
	40	static int set_permissions(struct ctl_table_header *head,
	41	struct ctl_table *table)
	42	{
	43	struct user_namespace *user_ns =
	44	container_of(head->set, struct user_namespace, set);
	45	int mode;
	46
	47	/* Allow users with CAP_SYS_RESOURCE unrestrained access */
	48	if (ns_capable(user_ns, CAP_SYS_RESOURCE))
	49	mode = (table->mode & S_IRWXU) >> 6;
	50	else
	51	/* Allow all others at most read-only access */
	52	mode = table->mode & S_IROTH;
	53	return (mode << 6) \| (mode << 3) \| mode;
	54	}
	55
	56	static struct ctl_table_root set_root = {
	57	.lookup = set_lookup,
	58	.permissions = set_permissions,
	59	};
	60
f153c224 SS	61	static long ue_zero = 0;
	62	static long ue_int_max = INT_MAX;
	63
	64	#define UCOUNT_ENTRY(name) \
	65	{ \
	66	.procname = name, \
	67	.maxlen = sizeof(long), \
	68	.mode = 0644, \
	69	.proc_handler = proc_doulongvec_minmax, \
	70	.extra1 = &ue_zero, \
	71	.extra2 = &ue_int_max, \
25f9c081	72	}
f6b2db1a	73	static struct ctl_table user_table[] = {
25f9c081	74	UCOUNT_ENTRY("max_user_namespaces"),
f333c700	75	UCOUNT_ENTRY("max_pid_namespaces"),
f7af3d1c	76	UCOUNT_ENTRY("max_uts_namespaces"),
aba35661	77	UCOUNT_ENTRY("max_ipc_namespaces"),
70328660	78	UCOUNT_ENTRY("max_net_namespaces"),
537f7ccb	79	UCOUNT_ENTRY("max_mnt_namespaces"),
d08311dd	80	UCOUNT_ENTRY("max_cgroup_namespaces"),
eeec26d5	81	UCOUNT_ENTRY("max_time_namespaces"),
1cce1eea NB	82	#ifdef CONFIG_INOTIFY_USER
	83	UCOUNT_ENTRY("max_inotify_instances"),
	84	UCOUNT_ENTRY("max_inotify_watches"),
5b8fea65 AG	85	#endif
	86	#ifdef CONFIG_FANOTIFY
	87	UCOUNT_ENTRY("max_fanotify_groups"),
	88	UCOUNT_ENTRY("max_fanotify_marks"),
1cce1eea	89	#endif
dbec2846 EB	90	{ }
	91	};
	92	#endif /* CONFIG_SYSCTL */
	93
	94	bool setup_userns_sysctls(struct user_namespace *ns)
	95	{
	96	#ifdef CONFIG_SYSCTL
	97	struct ctl_table *tbl;
0f538e3e JK	98
0f538e3e JK	99	BUILD_BUG_ON(ARRAY_SIZE(user_table) != UCOUNT_COUNTS + 1);
dbec2846	100	setup_sysctl_set(&ns->set, &set_root, set_is_seen);
f6b2db1a	101	tbl = kmemdup(user_table, sizeof(user_table), GFP_KERNEL);
dbec2846	102	if (tbl) {
25f9c081 EB	103	int i;
	104	for (i = 0; i < UCOUNT_COUNTS; i++) {
	105	tbl[i].data = &ns->ucount_max[i];
	106	}
f6b2db1a	107	ns->sysctls = __register_sysctl_table(&ns->set, "user", tbl);
dbec2846 EB	108	}
	109	if (!ns->sysctls) {
	110	kfree(tbl);
	111	retire_sysctl_set(&ns->set);
	112	return false;
	113	}
	114	#endif
	115	return true;
	116	}
	117
	118	void retire_userns_sysctls(struct user_namespace *ns)
	119	{
	120	#ifdef CONFIG_SYSCTL
	121	struct ctl_table *tbl;
	122
	123	tbl = ns->sysctls->ctl_table_arg;
	124	unregister_sysctl_table(ns->sysctls);
	125	retire_sysctl_set(&ns->set);
	126	kfree(tbl);
	127	#endif
	128	}
	129
f6b2db1a EB	130	static struct ucounts find_ucounts(struct user_namespace ns, kuid_t uid, struct hlist_head *hashent)
	131	{
	132	struct ucounts *ucounts;
	133
	134	hlist_for_each_entry(ucounts, hashent, node) {
	135	if (uid_eq(ucounts->uid, uid) && (ucounts->ns == ns))
	136	return ucounts;
	137	}
	138	return NULL;
	139	}
	140
905ae01c AG	141	static void hlist_add_ucounts(struct ucounts *ucounts)
	142	{
	143	struct hlist_head *hashent = ucounts_hashentry(ucounts->ns, ucounts->uid);
	144	spin_lock_irq(&ucounts_lock);
	145	hlist_add_head(&ucounts->node, hashent);
	146	spin_unlock_irq(&ucounts_lock);
	147	}
	148
da70d310 EB	149	static inline bool get_ucounts_or_wrap(struct ucounts *ucounts)
	150	{
	151	/* Returns true on a successful get, false if the count wraps. */
	152	return !atomic_add_negative(1, &ucounts->count);
	153	}
	154
b6c33652 AG	155	struct ucounts get_ucounts(struct ucounts ucounts)
b6c33652 AG	156	{
da70d310	157	if (!get_ucounts_or_wrap(ucounts)) {
b6c33652 AG	158	put_ucounts(ucounts);
	159	ucounts = NULL;
	160	}
	161	return ucounts;
	162	}
	163
905ae01c	164	struct ucounts alloc_ucounts(struct user_namespace ns, kuid_t uid)
f6b2db1a EB	165	{
	166	struct hlist_head *hashent = ucounts_hashentry(ns, uid);
	167	struct ucounts ucounts, new;
da70d310	168	bool wrapped;
f6b2db1a	169
880a3854	170	spin_lock_irq(&ucounts_lock);
f6b2db1a EB	171	ucounts = find_ucounts(ns, uid, hashent);
f6b2db1a EB	172	if (!ucounts) {
880a3854	173	spin_unlock_irq(&ucounts_lock);
f6b2db1a EB	174
	175	new = kzalloc(sizeof(*new), GFP_KERNEL);
	176	if (!new)
	177	return NULL;
	178
	179	new->ns = ns;
	180	new->uid = uid;
b6c33652	181	atomic_set(&new->count, 1);
f6b2db1a	182
880a3854	183	spin_lock_irq(&ucounts_lock);
f6b2db1a EB	184	ucounts = find_ucounts(ns, uid, hashent);
	185	if (ucounts) {
	186	kfree(new);
	187	} else {
	188	hlist_add_head(&new->node, hashent);
f9d87929	189	get_user_ns(new->ns);
b6c33652 AG	190	spin_unlock_irq(&ucounts_lock);
b6c33652 AG	191	return new;
f6b2db1a EB	192	}
f6b2db1a EB	193	}
da70d310	194	wrapped = !get_ucounts_or_wrap(ucounts);
880a3854	195	spin_unlock_irq(&ucounts_lock);
da70d310	196	if (wrapped) {
345daff2 AG	197	put_ucounts(ucounts);
	198	return NULL;
	199	}
f6b2db1a EB	200	return ucounts;
	201	}
	202
905ae01c	203	void put_ucounts(struct ucounts *ucounts)
f6b2db1a	204	{
880a3854 NB	205	unsigned long flags;
880a3854 NB	206
345daff2	207	if (atomic_dec_and_lock_irqsave(&ucounts->count, &ucounts_lock, flags)) {
f6b2db1a	208	hlist_del_init(&ucounts->node);
b6c33652	209	spin_unlock_irqrestore(&ucounts_lock, flags);
f9d87929	210	put_user_ns(ucounts->ns);
b6c33652 AG	211	kfree(ucounts);
b6c33652 AG	212	}
f6b2db1a EB	213	}
f6b2db1a EB	214
f9c82a4e	215	static inline bool atomic_long_inc_below(atomic_long_t *v, int u)
b376c3e1	216	{
f9c82a4e AG	217	long c, old;
f9c82a4e AG	218	c = atomic_long_read(v);
b376c3e1 EB	219	for (;;) {
	220	if (unlikely(c >= u))
	221	return false;
f9c82a4e	222	old = atomic_long_cmpxchg(v, c, c+1);
b376c3e1 EB	223	if (likely(old == c))
	224	return true;
	225	c = old;
	226	}
	227	}
	228
25f9c081 EB	229	struct ucounts inc_ucount(struct user_namespace ns, kuid_t uid,
25f9c081 EB	230	enum ucount_type type)
b376c3e1	231	{
f6b2db1a EB	232	struct ucounts ucounts, iter, *bad;
f6b2db1a EB	233	struct user_namespace *tns;
905ae01c	234	ucounts = alloc_ucounts(ns, uid);
f6b2db1a	235	for (iter = ucounts; iter; iter = tns->ucounts) {
f9c82a4e	236	long max;
f6b2db1a	237	tns = iter->ns;
25f9c081	238	max = READ_ONCE(tns->ucount_max[type]);
f9c82a4e	239	if (!atomic_long_inc_below(&iter->ucount[type], max))
b376c3e1 EB	240	goto fail;
b376c3e1 EB	241	}
f6b2db1a	242	return ucounts;
b376c3e1	243	fail:
f6b2db1a EB	244	bad = iter;
f6b2db1a EB	245	for (iter = ucounts; iter != bad; iter = iter->ns->ucounts)
f9c82a4e	246	atomic_long_dec(&iter->ucount[type]);
b376c3e1	247
f6b2db1a EB	248	put_ucounts(ucounts);
f6b2db1a EB	249	return NULL;
b376c3e1 EB	250	}
b376c3e1 EB	251
25f9c081	252	void dec_ucount(struct ucounts *ucounts, enum ucount_type type)
b376c3e1	253	{
f6b2db1a EB	254	struct ucounts *iter;
f6b2db1a EB	255	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
f9c82a4e	256	long dec = atomic_long_dec_if_positive(&iter->ucount[type]);
b376c3e1 EB	257	WARN_ON_ONCE(dec < 0);
b376c3e1 EB	258	}
f6b2db1a	259	put_ucounts(ucounts);
b376c3e1 EB	260	}
b376c3e1 EB	261
de399236	262	long inc_rlimit_ucounts(struct ucounts *ucounts, enum rlimit_type type, long v)
21d1c5e3 AG	263	{
21d1c5e3 AG	264	struct ucounts *iter;
59ec7157	265	long max = LONG_MAX;
21d1c5e3 AG	266	long ret = 0;
	267
	268	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
de399236	269	long new = atomic_long_add_return(v, &iter->rlimit[type]);
21d1c5e3 AG	270	if (new < 0 \|\| new > max)
	271	ret = LONG_MAX;
	272	else if (iter == ucounts)
	273	ret = new;
de399236	274	max = get_userns_rlimit_max(iter->ns, type);
21d1c5e3 AG	275	}
	276	return ret;
	277	}
	278
de399236	279	bool dec_rlimit_ucounts(struct ucounts *ucounts, enum rlimit_type type, long v)
21d1c5e3 AG	280	{
21d1c5e3 AG	281	struct ucounts *iter;
f928ef68	282	long new = -1; /* Silence compiler warning */
21d1c5e3	283	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
de399236	284	long dec = atomic_long_sub_return(v, &iter->rlimit[type]);
21d1c5e3 AG	285	WARN_ON_ONCE(dec < 0);
	286	if (iter == ucounts)
	287	new = dec;
	288	}
	289	return (new == 0);
	290	}
	291
15bc01ef	292	static void do_dec_rlimit_put_ucounts(struct ucounts *ucounts,
de399236	293	struct ucounts *last, enum rlimit_type type)
15bc01ef EB	294	{
	295	struct ucounts iter, next;
	296	for (iter = ucounts; iter != last; iter = next) {
de399236	297	long dec = atomic_long_sub_return(1, &iter->rlimit[type]);
15bc01ef EB	298	WARN_ON_ONCE(dec < 0);
	299	next = iter->ns->ucounts;
	300	if (dec == 0)
	301	put_ucounts(iter);
	302	}
	303	}
	304
de399236	305	void dec_rlimit_put_ucounts(struct ucounts *ucounts, enum rlimit_type type)
15bc01ef EB	306	{
	307	do_dec_rlimit_put_ucounts(ucounts, NULL, type);
	308	}
	309
de399236	310	long inc_rlimit_get_ucounts(struct ucounts *ucounts, enum rlimit_type type)
15bc01ef EB	311	{
	312	/* Caller must hold a reference to ucounts */
	313	struct ucounts *iter;
59ec7157	314	long max = LONG_MAX;
15bc01ef EB	315	long dec, ret = 0;
	316
	317	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
de399236	318	long new = atomic_long_add_return(1, &iter->rlimit[type]);
15bc01ef EB	319	if (new < 0 \|\| new > max)
	320	goto unwind;
	321	if (iter == ucounts)
	322	ret = new;
de399236	323	max = get_userns_rlimit_max(iter->ns, type);
15bc01ef EB	324	/*
	325	* Grab an extra ucount reference for the caller when
	326	* the rlimit count was previously 0.
	327	*/
	328	if (new != 1)
	329	continue;
	330	if (!get_ucounts(iter))
	331	goto dec_unwind;
	332	}
	333	return ret;
	334	dec_unwind:
de399236	335	dec = atomic_long_sub_return(1, &iter->rlimit[type]);
15bc01ef EB	336	WARN_ON_ONCE(dec < 0);
	337	unwind:
	338	do_dec_rlimit_put_ucounts(ucounts, iter, type);
	339	return 0;
	340	}
	341
de399236	342	bool is_rlimit_overlimit(struct ucounts *ucounts, enum rlimit_type type, unsigned long rlimit)
21d1c5e3 AG	343	{
21d1c5e3 AG	344	struct ucounts *iter;
59ec7157 AG	345	long max = rlimit;
	346	if (rlimit > LONG_MAX)
	347	max = LONG_MAX;
21d1c5e3	348	for (iter = ucounts; iter; iter = iter->ns->ucounts) {
de399236	349	long val = get_rlimit_value(iter, type);
0cbae9e2	350	if (val < 0 \|\| val > max)
21d1c5e3	351	return true;
de399236	352	max = get_userns_rlimit_max(iter->ns, type);
21d1c5e3 AG	353	}
	354	return false;
	355	}
	356
dbec2846 EB	357	static __init int user_namespace_sysctl_init(void)
	358	{
	359	#ifdef CONFIG_SYSCTL
f6b2db1a	360	static struct ctl_table_header *user_header;
dbec2846 EB	361	static struct ctl_table empty[1];
dbec2846 EB	362	/*
f6b2db1a	363	* It is necessary to register the user directory in the
dbec2846 EB	364	* default set so that registrations in the child sets work
	365	* properly.
	366	*/
f6b2db1a	367	user_header = register_sysctl("user", empty);
ed5bd7dc	368	kmemleak_ignore(user_header);
f6b2db1a	369	BUG_ON(!user_header);
dbec2846 EB	370	BUG_ON(!setup_userns_sysctls(&init_user_ns));
dbec2846 EB	371	#endif
905ae01c	372	hlist_add_ucounts(&init_ucounts);
21d1c5e3	373	inc_rlimit_ucounts(&init_ucounts, UCOUNT_RLIMIT_NPROC, 1);
dbec2846 EB	374	return 0;
	375	}
	376	subsys_initcall(user_namespace_sysctl_init);