xref: /linux/net/netfilter/ipvs/ip_vs_ctl.c (revision 91ec2035134982b98fab0609a9fd8480e8217dc1)

// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * IPVS         An implementation of the IP virtual server support for the
 *              LINUX operating system.  IPVS is now implemented as a module
 *              over the NetFilter framework. IPVS can be used to build a
 *              high-performance and highly available server based on a
 *              cluster of servers.
 *
 * Authors:     Wensong Zhang <wensong@linuxvirtualserver.org>
 *              Peter Kese <peter.kese@ijs.si>
 *              Julian Anastasov <ja@ssi.bg>
 *
 * Changes:
 */

#define pr_fmt(fmt) "IPVS: " fmt

#include <linux/module.h>
#include <linux/init.h>
#include <linux/types.h>
#include <linux/capability.h>
#include <linux/fs.h>
#include <linux/sysctl.h>
#include <linux/proc_fs.h>
#include <linux/workqueue.h>
#include <linux/seq_file.h>
#include <linux/slab.h>

#include <linux/netfilter.h>
#include <linux/netfilter_ipv4.h>
#include <linux/mutex.h>
#include <linux/rcupdate_wait.h>

#include <net/net_namespace.h>
#include <linux/nsproxy.h>
#include <net/ip.h>
#ifdef CONFIG_IP_VS_IPV6
#include <net/ipv6.h>
#include <net/ip6_route.h>
#include <net/netfilter/ipv6/nf_defrag_ipv6.h>
#endif
#include <net/route.h>
#include <net/sock.h>
#include <net/genetlink.h>

#include <linux/uaccess.h>

#include <net/ip_vs.h>

MODULE_ALIAS_GENL_FAMILY(IPVS_GENL_NAME);

static struct lock_class_key __ipvs_service_key;

/* sysctl variables */

#ifdef CONFIG_IP_VS_DEBUG
static int sysctl_ip_vs_debug_level = 0;

int ip_vs_get_debug_level(void)
{
	return sysctl_ip_vs_debug_level;
}
#endif


/*  Protos */
static void __ip_vs_del_service(struct ip_vs_service *svc, bool cleanup);


#ifdef CONFIG_IP_VS_IPV6
/* Taken from rt6_fill_node() in net/ipv6/route.c, is there a better way? */
static bool __ip_vs_addr_is_local_v6(struct net *net,
				     const struct in6_addr *addr)
{
	struct flowi6 fl6 = {
		.daddr = *addr,
	};
	struct dst_entry *dst = ip6_route_output(net, NULL, &fl6);
	bool is_local;

	is_local = !dst->error && dst->dev && (dst->dev->flags & IFF_LOOPBACK);

	dst_release(dst);
	return is_local;
}
#endif

#ifdef CONFIG_SYSCTL
/*
 *	update_defense_level is called from keventd and from sysctl,
 *	so it needs to protect itself from softirqs
 */
static void update_defense_level(struct netns_ipvs *ipvs)
{
	struct sysinfo i;
	int availmem;
	int amemthresh;
	int nomem;
	int to_change = -1;

	/* we only count free and buffered memory (in pages) */
	si_meminfo(&i);
	availmem = i.freeram + i.bufferram;
	/* however in linux 2.5 the i.bufferram is total page cache size,
	   we need adjust it */
	/* si_swapinfo(&i); */
	/* availmem = availmem - (i.totalswap - i.freeswap); */

	amemthresh = max(READ_ONCE(ipvs->sysctl_amemthresh), 0);
	nomem = (availmem < amemthresh);

	local_bh_disable();

	/* drop_entry */
	spin_lock(&ipvs->dropentry_lock);
	switch (ipvs->sysctl_drop_entry) {
	case 0:
		atomic_set(&ipvs->dropentry, 0);
		break;
	case 1:
		if (nomem) {
			atomic_set(&ipvs->dropentry, 1);
			ipvs->sysctl_drop_entry = 2;
		} else {
			atomic_set(&ipvs->dropentry, 0);
		}
		break;
	case 2:
		if (nomem) {
			atomic_set(&ipvs->dropentry, 1);
		} else {
			atomic_set(&ipvs->dropentry, 0);
			ipvs->sysctl_drop_entry = 1;
		}
		break;
	case 3:
		atomic_set(&ipvs->dropentry, 1);
		break;
	}
	spin_unlock(&ipvs->dropentry_lock);

	/* drop_packet */
	spin_lock(&ipvs->droppacket_lock);
	switch (ipvs->sysctl_drop_packet) {
	case 0:
		ipvs->drop_rate = 0;
		break;
	case 1:
		if (nomem) {
			ipvs->drop_counter = amemthresh / (amemthresh - availmem);
			ipvs->drop_rate = ipvs->drop_counter;
			ipvs->sysctl_drop_packet = 2;
		} else {
			ipvs->drop_rate = 0;
		}
		break;
	case 2:
		if (nomem) {
			ipvs->drop_counter = amemthresh / (amemthresh - availmem);
			ipvs->drop_rate = ipvs->drop_counter;
		} else {
			ipvs->drop_rate = 0;
			ipvs->sysctl_drop_packet = 1;
		}
		break;
	case 3:
		ipvs->drop_rate = ipvs->sysctl_am_droprate;
		break;
	}
	spin_unlock(&ipvs->droppacket_lock);

	/* secure_tcp */
	spin_lock(&ipvs->securetcp_lock);
	switch (ipvs->sysctl_secure_tcp) {
	case 0:
		if (ipvs->old_secure_tcp >= 2)
			to_change = 0;
		break;
	case 1:
		if (nomem) {
			if (ipvs->old_secure_tcp < 2)
				to_change = 1;
			ipvs->sysctl_secure_tcp = 2;
		} else {
			if (ipvs->old_secure_tcp >= 2)
				to_change = 0;
		}
		break;
	case 2:
		if (nomem) {
			if (ipvs->old_secure_tcp < 2)
				to_change = 1;
		} else {
			if (ipvs->old_secure_tcp >= 2)
				to_change = 0;
			ipvs->sysctl_secure_tcp = 1;
		}
		break;
	case 3:
		if (ipvs->old_secure_tcp < 2)
			to_change = 1;
		break;
	}
	ipvs->old_secure_tcp = ipvs->sysctl_secure_tcp;
	if (to_change >= 0)
		ip_vs_protocol_timeout_change(ipvs,
					      ipvs->sysctl_secure_tcp > 1);
	spin_unlock(&ipvs->securetcp_lock);

	local_bh_enable();
}

/* Handler for delayed work for expiring no
 * destination connections
 */
static void expire_nodest_conn_handler(struct work_struct *work)
{
	struct netns_ipvs *ipvs;

	ipvs = container_of(work, struct netns_ipvs,
			    expire_nodest_conn_work.work);
	ip_vs_expire_nodest_conn_flush(ipvs);
}

/*
 *	Timer for checking the defense
 */
#define DEFENSE_TIMER_PERIOD	1*HZ

static void defense_work_handler(struct work_struct *work)
{
	struct netns_ipvs *ipvs =
		container_of(work, struct netns_ipvs, defense_work.work);

	update_defense_level(ipvs);
	if (atomic_read(&ipvs->dropentry))
		ip_vs_random_dropentry(ipvs);
	queue_delayed_work(system_dfl_long_wq, &ipvs->defense_work,
			   DEFENSE_TIMER_PERIOD);
}
#endif

static void est_reload_work_handler(struct work_struct *work)
{
	struct netns_ipvs *ipvs =
		container_of(work, struct netns_ipvs, est_reload_work.work);
	int genid_done = atomic_read(&ipvs->est_genid_done);
	unsigned long delay = HZ / 10;	/* repeat startups after failure */
	bool repeat = false;
	int genid;
	int id;

	mutex_lock(&ipvs->est_mutex);
	genid = atomic_read(&ipvs->est_genid);
	for (id = 0; id < ipvs->est_kt_count; id++) {
		struct ip_vs_est_kt_data *kd = ipvs->est_kt_arr[id];

		/* netns clean up started, abort delayed work */
		if (!READ_ONCE(ipvs->enable))
			goto unlock;
		if (!kd)
			continue;
		/* New config ? Stop kthread tasks */
		if (genid != genid_done) {
			if (!id) {
				/* Only we can stop kt 0 but not under mutex */
				mutex_unlock(&ipvs->est_mutex);
				ip_vs_est_kthread_stop(kd);
				mutex_lock(&ipvs->est_mutex);
				if (!READ_ONCE(ipvs->enable))
					goto unlock;
				/* kd for kt 0 is never destroyed */
			} else {
				ip_vs_est_kthread_stop(kd);
			}
		}
		if (!kd->task && !ip_vs_est_stopped(ipvs)) {
			bool start;

			/* Do not start kthreads above 0 in calc phase */
			if (id)
				start = !ipvs->est_calc_phase;
			else
				start = kd->needed;
			if (start && ip_vs_est_kthread_start(ipvs, kd) < 0)
				repeat = true;
		}
	}

	atomic_set(&ipvs->est_genid_done, genid);

	if (repeat)
		queue_delayed_work(system_dfl_long_wq, &ipvs->est_reload_work,
				   delay);

unlock:
	mutex_unlock(&ipvs->est_mutex);
}

static int get_conn_tab_size(struct netns_ipvs *ipvs)
{
	const struct ip_vs_rht *t;
	int size = 0;

	rcu_read_lock();
	t = rcu_dereference(ipvs->conn_tab);
	if (t)
		size = t->size;
	rcu_read_unlock();

	return size;
}

int
ip_vs_use_count_inc(void)
{
	return try_module_get(THIS_MODULE);
}

void
ip_vs_use_count_dec(void)
{
	module_put(THIS_MODULE);
}


/* Service hashing:
 * Operation			Locking order
 * ---------------------------------------------------------------------------
 * add first table		service_mutex
 * attach new table		service_mutex
 * add/del service		service_mutex, RCU, bit lock
 * move between tables (rehash)	svc_resize_sem(W), seqcount_t(W), bit lock
 * replace old with attached	svc_resize_sem(W), svc_replace_sem(W)
 * find service			RCU, seqcount_t(R)
 * walk services(blocking)	service_mutex, svc_resize_sem(R)
 * walk services(non-blocking)	RCU, seqcount_t(R)
 * walk services(non-blocking)	svc_resize_sem(R), RCU, seqcount_t(R)
 * walk services(non-blocking)	svc_replace_sem(R), RCU, seqcount_t(R)
 * del table			service_mutex after stopped work
 *
 * - new table is attached on resizing under service_mutex and all operations
 * can run in parallel in 2 tables until the new table is registered as current
 * one
 * - two contexts can modify buckets: config and table resize (work), both in
 * process context
 * - only table resizer can move entries, so we do not protect t->seqc[]
 * items with t->lock[]
 * - lookups occur under RCU lock and seqcount reader lock to detect if
 * services are moved to new table
 * - move operations may disturb readers: find operation will not miss entries
 * but walkers may see same entry twice if they are forced to retry chains
 * or to walk the newly attached second table
 * - walkers using cond_resched_rcu() on !PREEMPT_RCU may need to check
 * svc_table_changes and repeat the RCU read section if new table is installed
 * - walkers may serialize with the whole resizing process (svc_resize_sem)
 * to prevent seeing same service twice or just with the svc_table
 * replace (svc_replace_sem) when we can see entries twice but we
 * prefer to run concurrently with the rehashing.
 */

/*
 *	Returns hash value for virtual service
 */
static inline u32
ip_vs_svc_hashval(struct ip_vs_rht *t, int af, unsigned int proto,
		  const union nf_inet_addr *addr, __be16 port)
{
	return ip_vs_rht_hash_linfo(t, af, addr, ntohs(port), proto);
}

/*
 *	Returns hash value of fwmark for virtual service lookup
 */
static inline u32 ip_vs_svc_fwm_hashval(struct ip_vs_rht *t, int af,
					__u32 fwmark)
{
	return jhash_2words(fwmark, af, (u32)t->hash_key.key[0]);
}

/* Hashes a service in the svc_table by <proto,addr,port> or by fwmark */
static int ip_vs_svc_hash(struct ip_vs_service *svc)
{
	struct netns_ipvs *ipvs = svc->ipvs;
	struct hlist_bl_head *head;
	struct ip_vs_rht *t;
	u32 hash;

	if (svc->flags & IP_VS_SVC_F_HASHED) {
		pr_err("%s(): request for already hashed, called from %pS\n",
		       __func__, __builtin_return_address(0));
		return 0;
	}

	/* increase its refcnt because it is referenced by the svc table */
	atomic_inc(&svc->refcnt);

	/* We know if new table is attached under service_mutex but rely on
	 * RCU to hold the old table to be freed in resizer
	 */
	rcu_read_lock();

	/* This can be the old or the new table */
	t = rcu_dereference(ipvs->svc_table);

	/* New entries go into recent table */
	t = rcu_dereference(t->new_tbl);

	if (svc->fwmark == 0) {
		/*
		 *  Hash it by <protocol,addr,port>
		 */
		hash = ip_vs_svc_hashval(t, svc->af, svc->protocol,
					 &svc->addr, svc->port);
	} else {
		/*
		 *  Hash it by fwmark
		 */
		hash = ip_vs_svc_fwm_hashval(t, svc->af, svc->fwmark);
	}
	head = t->buckets + (hash & t->mask);
	hlist_bl_lock(head);
	WRITE_ONCE(svc->hash_key, ip_vs_rht_build_hash_key(t, hash));
	svc->flags |= IP_VS_SVC_F_HASHED;
	hlist_bl_add_head_rcu(&svc->s_list, head);
	hlist_bl_unlock(head);

	rcu_read_unlock();

	return 1;
}


/*
 *	Unhashes a service from svc_table.
 *	Should be called with locked tables.
 */
static int ip_vs_svc_unhash(struct ip_vs_service *svc)
{
	struct netns_ipvs *ipvs = svc->ipvs;
	struct hlist_bl_head *head;
	struct ip_vs_rht *t;
	u32 hash_key2;
	u32 hash_key;

	if (!(svc->flags & IP_VS_SVC_F_HASHED)) {
		pr_err("%s(): request for unhash flagged, called from %pS\n",
		       __func__, __builtin_return_address(0));
		return 0;
	}

	/* We know if new table is attached under service_mutex but rely on
	 * RCU to hold the old table to be freed in resizer
	 */
	rcu_read_lock();

	/* This can be the old or the new table */
	t = rcu_dereference(ipvs->svc_table);
	hash_key = READ_ONCE(svc->hash_key);
	/* We need to lock the bucket in the right table */
	if (ip_vs_rht_same_table(t, hash_key)) {
		head = t->buckets + (hash_key & t->mask);
		hlist_bl_lock(head);
		/* Ensure hash_key is read under lock */
		hash_key2 = READ_ONCE(svc->hash_key);
		/* Moved to new table ? */
		if (hash_key != hash_key2) {
			hlist_bl_unlock(head);
			t = rcu_dereference(t->new_tbl);
			head = t->buckets + (hash_key2 & t->mask);
			hlist_bl_lock(head);
		}
	} else {
		/* It is already moved to new table */
		t = rcu_dereference(t->new_tbl);
		head = t->buckets + (hash_key & t->mask);
		hlist_bl_lock(head);
	}
	/* Remove it from svc_table */
	hlist_bl_del_rcu(&svc->s_list);

	svc->flags &= ~IP_VS_SVC_F_HASHED;
	atomic_dec(&svc->refcnt);
	hlist_bl_unlock(head);

	rcu_read_unlock();
	return 1;
}


/*
 *	Get service by {netns, proto,addr,port} in the service table.
 */
static inline struct ip_vs_service *
__ip_vs_service_find(struct netns_ipvs *ipvs, int af, __u16 protocol,
		     const union nf_inet_addr *vaddr, __be16 vport)
{
	DECLARE_IP_VS_RHT_WALK_BUCKET_RCU();
	struct hlist_bl_head *head;
	struct ip_vs_service *svc;
	struct ip_vs_rht *t, *p;
	struct hlist_bl_node *e;
	u32 hash, hash_key;

	ip_vs_rht_for_each_table_rcu(ipvs->svc_table, t, p) {
		/* Check for "full" addressed entries */
		hash = ip_vs_svc_hashval(t, af, protocol, vaddr, vport);

		hash_key = ip_vs_rht_build_hash_key(t, hash);
		ip_vs_rht_walk_bucket_rcu(t, hash_key, head) {
			hlist_bl_for_each_entry_rcu(svc, e, head, s_list) {
				if (READ_ONCE(svc->hash_key) == hash_key &&
				    svc->af == af &&
				    ip_vs_addr_equal(af, &svc->addr, vaddr) &&
				    svc->port == vport &&
				    svc->protocol == protocol && !svc->fwmark) {
					/* HIT */
					return svc;
				}
			}
		}
	}

	return NULL;
}


/*
 *	Get service by {fwmark} in the service table.
 */
static inline struct ip_vs_service *
__ip_vs_svc_fwm_find(struct netns_ipvs *ipvs, int af, __u32 fwmark)
{
	DECLARE_IP_VS_RHT_WALK_BUCKET_RCU();
	struct hlist_bl_head *head;
	struct ip_vs_service *svc;
	struct ip_vs_rht *t, *p;
	struct hlist_bl_node *e;
	u32 hash, hash_key;

	ip_vs_rht_for_each_table_rcu(ipvs->svc_table, t, p) {
		/* Check for fwmark addressed entries */
		hash = ip_vs_svc_fwm_hashval(t, af, fwmark);

		hash_key = ip_vs_rht_build_hash_key(t, hash);
		ip_vs_rht_walk_bucket_rcu(t, hash_key, head) {
			hlist_bl_for_each_entry_rcu(svc, e, head, s_list) {
				if (READ_ONCE(svc->hash_key) == hash_key &&
				    svc->fwmark == fwmark && svc->af == af) {
					/* HIT */
					return svc;
				}
			}
		}
	}

	return NULL;
}

/* Find service, called under RCU lock */
struct ip_vs_service *
ip_vs_service_find(struct netns_ipvs *ipvs, int af, __u32 fwmark, __u16 protocol,
		   const union nf_inet_addr *vaddr, __be16 vport)
{
	struct ip_vs_service *svc = NULL;
	int af_id = ip_vs_af_index(af);

	/*
	 *	Check the table hashed by fwmark first
	 */
	if (fwmark && atomic_read(&ipvs->fwm_services[af_id])) {
		svc = __ip_vs_svc_fwm_find(ipvs, af, fwmark);
		if (svc)
			goto out;
	}

	if (!atomic_read(&ipvs->nonfwm_services[af_id]))
		goto out;

	/*
	 *	Check the table hashed by <protocol,addr,port>
	 *	for "full" addressed entries
	 */
	svc = __ip_vs_service_find(ipvs, af, protocol, vaddr, vport);
	if (svc)
		goto out;

	if (protocol == IPPROTO_TCP &&
	    atomic_read(&ipvs->ftpsvc_counter[af_id]) &&
	    (vport == FTPDATA || !inet_port_requires_bind_service(ipvs->net, ntohs(vport)))) {
		/*
		 * Check if ftp service entry exists, the packet
		 * might belong to FTP data connections.
		 */
		svc = __ip_vs_service_find(ipvs, af, protocol, vaddr, FTPPORT);
		if (svc)
			goto out;
	}

	if (atomic_read(&ipvs->nullsvc_counter[af_id])) {
		/*
		 * Check if the catch-all port (port zero) exists
		 */
		svc = __ip_vs_service_find(ipvs, af, protocol, vaddr, 0);
	}

  out:
	IP_VS_DBG_BUF(9, "lookup service: fwm %u %s %s:%u %s\n",
		      fwmark, ip_vs_proto_name(protocol),
		      IP_VS_DBG_ADDR(af, vaddr), ntohs(vport),
		      svc ? "hit" : "not hit");

	return svc;
}

/* Return the number of registered services */
static int ip_vs_get_num_services(struct netns_ipvs *ipvs)
{
	int ns = 0, ni = IP_VS_AF_MAX;

	while (--ni >= 0)
		ns += atomic_read(&ipvs->num_services[ni]);
	return ns;
}

/* Get default load factor to map num_services/u_thresh to t->size */
static int ip_vs_svc_default_load_factor(struct netns_ipvs *ipvs)
{
	int factor;

	if (net_eq(ipvs->net, &init_net))
		factor = -3;	/* grow if load is above 12.5% */
	else
		factor = -2;	/* grow if load is above 25% */
	return factor;
}

/* Get the desired svc_table size */
static int ip_vs_svc_desired_size(struct netns_ipvs *ipvs, struct ip_vs_rht *t,
				  int lfactor)
{
	return ip_vs_rht_desired_size(ipvs, t, ip_vs_get_num_services(ipvs),
				      lfactor, IP_VS_SVC_TAB_MIN_BITS,
				      IP_VS_SVC_TAB_MAX_BITS);
}

/* Allocate svc_table */
static struct ip_vs_rht *ip_vs_svc_table_alloc(struct netns_ipvs *ipvs,
					       int buckets, int lfactor)
{
	struct ip_vs_rht *t;
	int scounts, locks;

	/* No frequent lookups to race with resizing, so use max of 64
	 * seqcounts. Only resizer moves entries, so use 0 locks.
	 */
	scounts = clamp(buckets >> 4, 1, 64);
	locks = 0;

	t = ip_vs_rht_alloc(buckets, scounts, locks);
	if (!t)
		return NULL;
	t->lfactor = lfactor;
	ip_vs_rht_set_thresholds(t, t->size, lfactor, IP_VS_SVC_TAB_MIN_BITS,
				 IP_VS_SVC_TAB_MAX_BITS);
	return t;
}

/* svc_table resizer work */
static void svc_resize_work_handler(struct work_struct *work)
{
	struct hlist_bl_head *head, *head2;
	struct ip_vs_rht *t_free = NULL;
	unsigned int resched_score = 0;
	struct hlist_bl_node *cn, *nn;
	struct ip_vs_rht *t, *t_new;
	struct ip_vs_service *svc;
	struct netns_ipvs *ipvs;
	bool more_work = true;
	seqcount_t *sc;
	int limit = 0;
	int new_size;
	int lfactor;
	u32 bucket;

	ipvs = container_of(work, struct netns_ipvs, svc_resize_work.work);

	if (!down_write_trylock(&ipvs->svc_resize_sem))
		goto out;
	if (!mutex_trylock(&ipvs->service_mutex))
		goto unlock_sem;
	more_work = false;
	clear_bit(IP_VS_WORK_SVC_RESIZE, &ipvs->work_flags);
	if (!READ_ONCE(ipvs->enable))
		goto unlock_m;
	t = rcu_dereference_protected(ipvs->svc_table, 1);
	/* Do nothing if table is removed */
	if (!t)
		goto unlock_m;
	/* New table already attached? BUG! */
	if (t != rcu_access_pointer(t->new_tbl))
		goto unlock_m;

	lfactor = sysctl_svc_lfactor(ipvs);
	/* Should we resize ? */
	new_size = ip_vs_svc_desired_size(ipvs, t, lfactor);
	if (new_size == t->size && lfactor == t->lfactor)
		goto unlock_m;

	t_new = ip_vs_svc_table_alloc(ipvs, new_size, lfactor);
	if (!t_new) {
		more_work = true;
		goto unlock_m;
	}
	/* Flip the table_id */
	t_new->table_id = t->table_id ^ IP_VS_RHT_TABLE_ID_MASK;

	/* Attach new table */
	rcu_assign_pointer(t->new_tbl, t_new);
	/* Allow add/del to new_tbl while moving from old table */
	mutex_unlock(&ipvs->service_mutex);

	ip_vs_rht_for_each_bucket(t, bucket, head) {
same_bucket:
		if (++limit >= 16) {
			/* Check if work is stopped */
			if (test_bit(IP_VS_WORK_SVC_NORESIZE,
				     &ipvs->work_flags))
				goto unlock_sem;
			if (resched_score >= 100) {
				resched_score = 0;
				cond_resched();
			}
			limit = 0;
		}
		if (hlist_bl_empty(head)) {
			resched_score++;
			continue;
		}
		/* Preemption calls ahead... */
		resched_score = 0;

		sc = &t->seqc[bucket & t->seqc_mask];
		/* seqcount_t usage considering PREEMPT_RT rules:
		 * - we are the only writer => preemption can be allowed
		 * - readers (SoftIRQ) => disable BHs
		 * - readers (processes) => preemption should be disabled
		 */
		local_bh_disable();
		preempt_disable_nested();
		write_seqcount_begin(sc);
		hlist_bl_lock(head);

		hlist_bl_for_each_entry_safe(svc, cn, nn, head, s_list) {
			u32 hash;

			/* New hash for the new table */
			if (svc->fwmark == 0) {
				/*  Hash it by <protocol,addr,port> */
				hash = ip_vs_svc_hashval(t_new, svc->af,
							 svc->protocol,
							 &svc->addr, svc->port);
			} else {
				/* Hash it by fwmark */
				hash = ip_vs_svc_fwm_hashval(t_new, svc->af,
							     svc->fwmark);
			}
			hlist_bl_del_rcu(&svc->s_list);
			head2 = t_new->buckets + (hash & t_new->mask);

			hlist_bl_lock(head2);
			WRITE_ONCE(svc->hash_key,
				   ip_vs_rht_build_hash_key(t_new, hash));
			/* t_new->seqc are not used at this stage, we race
			 * only with add/del, so only lock the bucket.
			 */
			hlist_bl_add_head_rcu(&svc->s_list, head2);
			hlist_bl_unlock(head2);
			/* Too long chain? Do it in steps */
			if (++limit >= 64)
				break;
		}

		hlist_bl_unlock(head);
		write_seqcount_end(sc);
		preempt_enable_nested();
		local_bh_enable();
		if (limit >= 64)
			goto same_bucket;
	}

	/* Serialize with readers that don't like svc_table changes */
	down_write(&ipvs->svc_replace_sem);

	/* Check if work is stopped to avoid synchronize_rcu() */
	if (test_bit(IP_VS_WORK_SVC_NORESIZE, &ipvs->work_flags))
		goto unlock_repl;

	rcu_assign_pointer(ipvs->svc_table, t_new);
	/* Inform readers that new table is installed */
	smp_mb__before_atomic();
	atomic_inc(&ipvs->svc_table_changes);
	t_free = t;

unlock_repl:
	up_write(&ipvs->svc_replace_sem);

unlock_sem:
	up_write(&ipvs->svc_resize_sem);

	if (t_free) {
		/* RCU readers should not see more than two tables in chain.
		 * To prevent new table to be attached wait here instead of
		 * freeing the old table in RCU callback.
		 */
		synchronize_rcu();
		ip_vs_rht_free(t_free);
	}

out:
	if (!READ_ONCE(ipvs->enable) || !more_work ||
	    test_bit(IP_VS_WORK_SVC_NORESIZE, &ipvs->work_flags))
		return;
	queue_delayed_work(system_dfl_long_wq, &ipvs->svc_resize_work, 1);
	return;

unlock_m:
	mutex_unlock(&ipvs->service_mutex);
	goto unlock_sem;
}

static inline void
__ip_vs_bind_svc(struct ip_vs_dest *dest, struct ip_vs_service *svc)
{
	atomic_inc(&svc->refcnt);
	rcu_assign_pointer(dest->svc, svc);
}

static void ip_vs_service_free(struct ip_vs_service *svc)
{
	ip_vs_stats_release(&svc->stats);
	kfree(svc);
}

static void ip_vs_service_rcu_free(struct rcu_head *head)
{
	struct ip_vs_service *svc;

	svc = container_of(head, struct ip_vs_service, rcu_head);
	ip_vs_service_free(svc);
}

static void __ip_vs_svc_put(struct ip_vs_service *svc)
{
	if (atomic_dec_and_test(&svc->refcnt)) {
		IP_VS_DBG_BUF(3, "Removing service %u/%s:%u\n",
			      svc->fwmark,
			      IP_VS_DBG_ADDR(svc->af, &svc->addr),
			      ntohs(svc->port));
		call_rcu(&svc->rcu_head, ip_vs_service_rcu_free);
	}
}


/*
 *	Returns hash value for real service
 */
static inline unsigned int ip_vs_rs_hashkey(int af,
					    const union nf_inet_addr *addr,
					    __be16 port)
{
	unsigned int porth = ntohs(port);
	__be32 addr_fold = addr->ip;

#ifdef CONFIG_IP_VS_IPV6
	if (af == AF_INET6)
		addr_fold = addr->ip6[0]^addr->ip6[1]^
			    addr->ip6[2]^addr->ip6[3];
#endif

	return (ntohl(addr_fold)^(porth>>IP_VS_RTAB_BITS)^porth)
		& IP_VS_RTAB_MASK;
}

/* Hash ip_vs_dest in rs_table by <proto,addr,port>. */
static void ip_vs_rs_hash(struct netns_ipvs *ipvs, struct ip_vs_dest *dest)
{
	unsigned int hash;
	__be16 port;

	if (dest->in_rs_table)
		return;

	switch (IP_VS_DFWD_METHOD(dest)) {
	case IP_VS_CONN_F_MASQ:
		port = dest->port;
		break;
	case IP_VS_CONN_F_TUNNEL:
		switch (dest->tun_type) {
		case IP_VS_CONN_F_TUNNEL_TYPE_GUE:
			port = dest->tun_port;
			break;
		case IP_VS_CONN_F_TUNNEL_TYPE_IPIP:
		case IP_VS_CONN_F_TUNNEL_TYPE_GRE:
			port = 0;
			break;
		default:
			return;
		}
		break;
	default:
		return;
	}

	/*
	 *	Hash by proto,addr,port,
	 *	which are the parameters of the real service.
	 */
	hash = ip_vs_rs_hashkey(dest->af, &dest->addr, port);

	hlist_add_head_rcu(&dest->d_list, &ipvs->rs_table[hash]);
	dest->in_rs_table = 1;
}

/* Unhash ip_vs_dest from rs_table. */
static void ip_vs_rs_unhash(struct ip_vs_dest *dest)
{
	/*
	 * Remove it from the rs_table table.
	 */
	if (dest->in_rs_table) {
		hlist_del_rcu(&dest->d_list);
		dest->in_rs_table = 0;
	}
}

/* Check if real service by <proto,addr,port> is present */
bool ip_vs_has_real_service(struct netns_ipvs *ipvs, int af, __u16 protocol,
			    const union nf_inet_addr *daddr, __be16 dport)
{
	unsigned int hash;
	struct ip_vs_dest *dest;

	/* Check for "full" addressed entries */
	hash = ip_vs_rs_hashkey(af, daddr, dport);

	hlist_for_each_entry_rcu(dest, &ipvs->rs_table[hash], d_list) {
		if (dest->port == dport &&
		    dest->af == af &&
		    ip_vs_addr_equal(af, &dest->addr, daddr) &&
		    (dest->protocol == protocol || dest->vfwmark) &&
		    IP_VS_DFWD_METHOD(dest) == IP_VS_CONN_F_MASQ) {
			/* HIT */
			return true;
		}
	}

	return false;
}

/* Find real service record by <proto,addr,port>.
 * In case of multiple records with the same <proto,addr,port>, only
 * the first found record is returned.
 *
 * To be called under RCU lock.
 */
struct ip_vs_dest *ip_vs_find_real_service(struct netns_ipvs *ipvs, int af,
					   __u16 protocol,
					   const union nf_inet_addr *daddr,
					   __be16 dport)
{
	unsigned int hash;
	struct ip_vs_dest *dest;

	/* Check for "full" addressed entries */
	hash = ip_vs_rs_hashkey(af, daddr, dport);

	hlist_for_each_entry_rcu(dest, &ipvs->rs_table[hash], d_list) {
		if (dest->port == dport &&
		    dest->af == af &&
		    ip_vs_addr_equal(af, &dest->addr, daddr) &&
		    (dest->protocol == protocol || dest->vfwmark) &&
		    IP_VS_DFWD_METHOD(dest) == IP_VS_CONN_F_MASQ) {
			/* HIT */
			return dest;
		}
	}

	return NULL;
}

/* Find real service record by <af,addr,tun_port>.
 * In case of multiple records with the same <af,addr,tun_port>, only
 * the first found record is returned.
 *
 * To be called under RCU lock.
 */
struct ip_vs_dest *ip_vs_find_tunnel(struct netns_ipvs *ipvs, int af,
				     const union nf_inet_addr *daddr,
				     __be16 tun_port)
{
	struct ip_vs_dest *dest;
	unsigned int hash;

	/* Check for "full" addressed entries */
	hash = ip_vs_rs_hashkey(af, daddr, tun_port);

	hlist_for_each_entry_rcu(dest, &ipvs->rs_table[hash], d_list) {
		if (dest->tun_port == tun_port &&
		    dest->af == af &&
		    ip_vs_addr_equal(af, &dest->addr, daddr) &&
		    IP_VS_DFWD_METHOD(dest) == IP_VS_CONN_F_TUNNEL) {
			/* HIT */
			return dest;
		}
	}

	return NULL;
}

/* Lookup destination by {addr,port} in the given service
 * Called under RCU lock.
 */
static struct ip_vs_dest *
ip_vs_lookup_dest(struct ip_vs_service *svc, int dest_af,
		  const union nf_inet_addr *daddr, __be16 dport)
{
	struct ip_vs_dest *dest;

	/*
	 * Find the destination for the given service
	 */
	list_for_each_entry_rcu(dest, &svc->destinations, n_list) {
		if ((dest->af == dest_af) &&
		    ip_vs_addr_equal(dest_af, &dest->addr, daddr) &&
		    (dest->port == dport)) {
			/* HIT */
			return dest;
		}
	}

	return NULL;
}

/*
 * Find destination by {daddr,dport,vaddr,protocol}
 * Created to be used in ip_vs_process_message() in
 * the backup synchronization daemon. It finds the
 * destination to be bound to the received connection
 * on the backup.
 * Called under RCU lock, no refcnt is returned.
 */
struct ip_vs_dest *ip_vs_find_dest(struct netns_ipvs *ipvs, int svc_af, int dest_af,
				   const union nf_inet_addr *daddr,
				   __be16 dport,
				   const union nf_inet_addr *vaddr,
				   __be16 vport, __u16 protocol, __u32 fwmark,
				   __u32 flags)
{
	struct ip_vs_dest *dest;
	struct ip_vs_service *svc;
	__be16 port = dport;

	svc = ip_vs_service_find(ipvs, svc_af, fwmark, protocol, vaddr, vport);
	if (!svc)
		return NULL;
	if (fwmark && (flags & IP_VS_CONN_F_FWD_MASK) != IP_VS_CONN_F_MASQ)
		port = 0;
	dest = ip_vs_lookup_dest(svc, dest_af, daddr, port);
	if (!dest)
		dest = ip_vs_lookup_dest(svc, dest_af, daddr, port ^ dport);
	return dest;
}

void ip_vs_dest_dst_rcu_free(struct rcu_head *head)
{
	struct ip_vs_dest_dst *dest_dst = container_of(head,
						       struct ip_vs_dest_dst,
						       rcu_head);

	dst_release(dest_dst->dst_cache);
	kfree(dest_dst);
}

/* Release dest_dst and dst_cache for dest in user context */
static void __ip_vs_dst_cache_reset(struct ip_vs_dest *dest)
{
	struct ip_vs_dest_dst *old;

	old = rcu_dereference_protected(dest->dest_dst, 1);
	if (old) {
		RCU_INIT_POINTER(dest->dest_dst, NULL);
		call_rcu(&old->rcu_head, ip_vs_dest_dst_rcu_free);
	}
}

/*
 *  Lookup dest by {svc,addr,port} in the destination trash.
 *  The destination trash is used to hold the destinations that are removed
 *  from the service table but are still referenced by some conn entries.
 *  The reason to add the destination trash is when the dest is temporary
 *  down (either by administrator or by monitor program), the dest can be
 *  picked back from the trash, the remaining connections to the dest can
 *  continue, and the counting information of the dest is also useful for
 *  scheduling.
 */
static struct ip_vs_dest *
ip_vs_trash_get_dest(struct ip_vs_service *svc, int dest_af,
		     const union nf_inet_addr *daddr, __be16 dport)
{
	struct ip_vs_dest *dest;
	struct netns_ipvs *ipvs = svc->ipvs;

	/*
	 * Find the destination in trash
	 */
	spin_lock_bh(&ipvs->dest_trash_lock);
	list_for_each_entry(dest, &ipvs->dest_trash, t_list) {
		IP_VS_DBG_BUF(3, "Destination %u/%s:%u still in trash, "
			      "dest->refcnt=%d\n",
			      dest->vfwmark,
			      IP_VS_DBG_ADDR(dest->af, &dest->addr),
			      ntohs(dest->port),
			      refcount_read(&dest->refcnt));
		if (dest->af == dest_af &&
		    ip_vs_addr_equal(dest_af, &dest->addr, daddr) &&
		    dest->port == dport &&
		    dest->vfwmark == svc->fwmark &&
		    dest->protocol == svc->protocol &&
		    (svc->fwmark ||
		     (ip_vs_addr_equal(svc->af, &dest->vaddr, &svc->addr) &&
		      dest->vport == svc->port))) {
			/* HIT */
			list_del(&dest->t_list);
			goto out;
		}
	}

	dest = NULL;

out:
	spin_unlock_bh(&ipvs->dest_trash_lock);

	return dest;
}

/* Put destination in trash */
static void ip_vs_trash_put_dest(struct netns_ipvs *ipvs,
				 struct ip_vs_dest *dest, unsigned long istart,
				 bool cleanup)
{
	spin_lock_bh(&ipvs->dest_trash_lock);
	IP_VS_DBG_BUF(3, "Moving dest %s:%u into trash, dest->refcnt=%d\n",
		      IP_VS_DBG_ADDR(dest->af, &dest->addr), ntohs(dest->port),
		      refcount_read(&dest->refcnt));
	if (list_empty(&ipvs->dest_trash) && !cleanup)
		mod_timer(&ipvs->dest_trash_timer,
			  jiffies + (IP_VS_DEST_TRASH_PERIOD >> 1));
	/* dest lives in trash with reference */
	list_add(&dest->t_list, &ipvs->dest_trash);
	dest->idle_start = istart;
	spin_unlock_bh(&ipvs->dest_trash_lock);
}

static void ip_vs_dest_rcu_free(struct rcu_head *head)
{
	struct ip_vs_dest *dest;

	dest = container_of(head, struct ip_vs_dest, rcu_head);
	ip_vs_stats_release(&dest->stats);
	ip_vs_dest_put_and_free(dest);
}

static void ip_vs_dest_free(struct ip_vs_dest *dest)
{
	struct ip_vs_service *svc = rcu_dereference_protected(dest->svc, 1);

	__ip_vs_svc_put(svc);
	call_rcu(&dest->rcu_head, ip_vs_dest_rcu_free);
}

/*
 *  Clean up all the destinations in the trash
 *  Called by the ip_vs_control_cleanup()
 *
 *  When the ip_vs_control_clearup is activated by ipvs module exit,
 *  the service tables must have been flushed and all the connections
 *  are expired, and the refcnt of each destination in the trash must
 *  be 1, so we simply release them here.
 */
static void ip_vs_trash_cleanup(struct netns_ipvs *ipvs)
{
	struct ip_vs_dest *dest, *nxt;

	timer_delete_sync(&ipvs->dest_trash_timer);
	/* No need to use dest_trash_lock */
	list_for_each_entry_safe(dest, nxt, &ipvs->dest_trash, t_list) {
		list_del(&dest->t_list);
		ip_vs_dest_free(dest);
	}
}

static void ip_vs_stats_rcu_free(struct rcu_head *head)
{
	struct ip_vs_stats_rcu *rs = container_of(head,
						  struct ip_vs_stats_rcu,
						  rcu_head);

	ip_vs_stats_release(&rs->s);
	kfree(rs);
}

static void
ip_vs_copy_stats(struct ip_vs_kstats *dst, struct ip_vs_stats *src)
{
#define IP_VS_SHOW_STATS_COUNTER(c) dst->c = src->kstats.c - src->kstats0.c

	spin_lock(&src->lock);

	IP_VS_SHOW_STATS_COUNTER(conns);
	IP_VS_SHOW_STATS_COUNTER(inpkts);
	IP_VS_SHOW_STATS_COUNTER(outpkts);
	IP_VS_SHOW_STATS_COUNTER(inbytes);
	IP_VS_SHOW_STATS_COUNTER(outbytes);

	ip_vs_read_estimator(dst, src);

	spin_unlock(&src->lock);
}

static void
ip_vs_export_stats_user(struct ip_vs_stats_user *dst, struct ip_vs_kstats *src)
{
	dst->conns = (u32)src->conns;
	dst->inpkts = (u32)src->inpkts;
	dst->outpkts = (u32)src->outpkts;
	dst->inbytes = src->inbytes;
	dst->outbytes = src->outbytes;
	dst->cps = (u32)src->cps;
	dst->inpps = (u32)src->inpps;
	dst->outpps = (u32)src->outpps;
	dst->inbps = (u32)src->inbps;
	dst->outbps = (u32)src->outbps;
}

static void
ip_vs_zero_stats(struct ip_vs_stats *stats)
{
	spin_lock(&stats->lock);

	/* get current counters as zero point, rates are zeroed */

#define IP_VS_ZERO_STATS_COUNTER(c) stats->kstats0.c = stats->kstats.c

	IP_VS_ZERO_STATS_COUNTER(conns);
	IP_VS_ZERO_STATS_COUNTER(inpkts);
	IP_VS_ZERO_STATS_COUNTER(outpkts);
	IP_VS_ZERO_STATS_COUNTER(inbytes);
	IP_VS_ZERO_STATS_COUNTER(outbytes);

	ip_vs_zero_estimator(stats);

	spin_unlock(&stats->lock);
}

/* Allocate fields after kzalloc */
int ip_vs_stats_init_alloc(struct ip_vs_stats *s)
{
	int i;

	spin_lock_init(&s->lock);
	s->cpustats = alloc_percpu(struct ip_vs_cpu_stats);
	if (!s->cpustats)
		return -ENOMEM;

	for_each_possible_cpu(i) {
		struct ip_vs_cpu_stats *cs = per_cpu_ptr(s->cpustats, i);

		u64_stats_init(&cs->syncp);
	}
	return 0;
}

struct ip_vs_stats *ip_vs_stats_alloc(void)
{
	struct ip_vs_stats *s = kzalloc_obj(*s);

	if (s && ip_vs_stats_init_alloc(s) >= 0)
		return s;
	kfree(s);
	return NULL;
}

void ip_vs_stats_release(struct ip_vs_stats *stats)
{
	free_percpu(stats->cpustats);
}

void ip_vs_stats_free(struct ip_vs_stats *stats)
{
	if (stats) {
		ip_vs_stats_release(stats);
		kfree(stats);
	}
}

/* Update overload flag based on number of dest conns and lower/upper
 * connection thresholds:
 * - conns reach u_threshold and exceed it: set the flag
 * - conns go below l_threshold (or 75% of u_threshold): clear the flag
 */
static void __ip_vs_dest_update_overload(struct ip_vs_dest *dest, int mode)
{
	int conns;
	u32 l, u;

	lockdep_assert_held(&dest->dst_lock);
	u = READ_ONCE(dest->u_threshold);
	if (!u)
		goto unset;
	l = READ_ONCE(dest->l_threshold_val);
	conns = atomic_read(&dest->totalconns);
	if (conns >= (mode > 0 ? l : u)) {
		dest->flags |= IP_VS_DEST_F_OVERLOAD;
		return;
	}
	if (conns >= (mode < 0 ? u : l))
		return;

unset:
	dest->flags &= ~IP_VS_DEST_F_OVERLOAD;
}

void ip_vs_dest_update_overload(struct ip_vs_dest *dest, int mode)
{
	spin_lock_bh(&dest->dst_lock);
	__ip_vs_dest_update_overload(dest, mode);
	spin_unlock_bh(&dest->dst_lock);
}

/*
 *	Update a destination in the given service
 */
static void
__ip_vs_update_dest(struct ip_vs_service *svc, struct ip_vs_dest *dest,
		    struct ip_vs_dest_user_kern *udest, int add)
{
	struct netns_ipvs *ipvs = svc->ipvs;
	struct ip_vs_service *old_svc;
	struct ip_vs_scheduler *sched;
	int conn_flags;

	/* We cannot modify an address and change the address family */
	BUG_ON(!add && udest->af != dest->af);

	if (add && udest->af != svc->af)
		ipvs->mixed_address_family_dests++;

	/* keep the last_weight with latest non-0 weight */
	if (add || udest->weight != 0)
		atomic_set(&dest->last_weight, udest->weight);

	/* set the weight and the flags */
	atomic_set(&dest->weight, udest->weight);
	conn_flags = udest->conn_flags & IP_VS_CONN_F_DEST_MASK;
	conn_flags |= IP_VS_CONN_F_INACTIVE;

	/* Need to rehash? */
	if ((udest->conn_flags & IP_VS_CONN_F_FWD_MASK) !=
	    IP_VS_DFWD_METHOD(dest) ||
	    udest->tun_type != dest->tun_type ||
	    udest->tun_port != dest->tun_port)
		ip_vs_rs_unhash(dest);

	/* set the tunnel info */
	dest->tun_type = udest->tun_type;
	dest->tun_port = udest->tun_port;
	dest->tun_flags = udest->tun_flags;

	/* set the IP_VS_CONN_F_NOOUTPUT flag if not masquerading/NAT */
	if ((conn_flags & IP_VS_CONN_F_FWD_MASK) != IP_VS_CONN_F_MASQ) {
		conn_flags |= IP_VS_CONN_F_NOOUTPUT;
	} else {
		/* FTP-NAT requires conntrack for mangling */
		if (svc->port == FTPPORT)
			ip_vs_register_conntrack(svc);
	}
	atomic_set(&dest->conn_flags, conn_flags);
	/* Put the real service in rs_table if not present. */
	ip_vs_rs_hash(ipvs, dest);

	/* bind the service */
	old_svc = rcu_dereference_protected(dest->svc, 1);
	if (!old_svc) {
		__ip_vs_bind_svc(dest, svc);
	} else {
		if (old_svc != svc) {
			ip_vs_zero_stats(&dest->stats);
			__ip_vs_bind_svc(dest, svc);
			__ip_vs_svc_put(old_svc);
		}
	}

	/* set the dest status flags */
	dest->cflags |= IP_VS_DEST_CF_AVAILABLE;

	if (READ_ONCE(dest->u_threshold) != udest->u_threshold ||
	    READ_ONCE(dest->l_threshold) != udest->l_threshold) {
		spin_lock_bh(&dest->dst_lock);
		WRITE_ONCE(dest->u_threshold, udest->u_threshold);
		WRITE_ONCE(dest->l_threshold, udest->l_threshold);
		/* Low threshold defaults to 75% of upper threshold */
		WRITE_ONCE(dest->l_threshold_val,
			   udest->l_threshold ? :
			   (udest->u_threshold -
			    (udest->u_threshold >> 2)));
		__ip_vs_dest_update_overload(dest, 0);
		spin_unlock_bh(&dest->dst_lock);
	}

	dest->af = udest->af;

	if (add) {
		list_add_rcu(&dest->n_list, &svc->destinations);
		svc->num_dests++;
		sched = rcu_dereference_protected(svc->scheduler, 1);
		if (sched && sched->add_dest)
			sched->add_dest(svc, dest);
	} else {
		spin_lock_bh(&dest->dst_lock);
		__ip_vs_dst_cache_reset(dest);
		spin_unlock_bh(&dest->dst_lock);

		sched = rcu_dereference_protected(svc->scheduler, 1);
		if (sched && sched->upd_dest)
			sched->upd_dest(svc, dest);
	}
}


/*
 *	Create a destination for the given service
 */
static int
ip_vs_new_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest)
{
	struct ip_vs_dest *dest;
	unsigned int atype;
	int ret;

#ifdef CONFIG_IP_VS_IPV6
	if (udest->af == AF_INET6) {
		atype = ipv6_addr_type(&udest->addr.in6);
		if ((!(atype & IPV6_ADDR_UNICAST) ||
			atype & IPV6_ADDR_LINKLOCAL) &&
			!__ip_vs_addr_is_local_v6(svc->ipvs->net, &udest->addr.in6))
			return -EINVAL;

		ret = nf_defrag_ipv6_enable(svc->ipvs->net);
		if (ret)
			return ret;
	} else
#endif
	{
		atype = inet_addr_type(svc->ipvs->net, udest->addr.ip);
		if (atype != RTN_LOCAL && atype != RTN_UNICAST)
			return -EINVAL;
	}

	dest = kzalloc_obj(struct ip_vs_dest);
	if (dest == NULL)
		return -ENOMEM;

	ret = ip_vs_stats_init_alloc(&dest->stats);
	if (ret < 0)
		goto err_alloc;

	ret = ip_vs_start_estimator(svc->ipvs, &dest->stats);
	if (ret < 0)
		goto err_stats;

	dest->af = udest->af;
	dest->protocol = svc->protocol;
	dest->vaddr = svc->addr;
	dest->vport = svc->port;
	dest->vfwmark = svc->fwmark;
	ip_vs_addr_copy(udest->af, &dest->addr, &udest->addr);
	dest->port = udest->port;

	atomic_set(&dest->activeconns, 0);
	atomic_set(&dest->totalconns, 0);
	atomic_set(&dest->persistconns, 0);
	refcount_set(&dest->refcnt, 1);

	INIT_HLIST_NODE(&dest->d_list);
	spin_lock_init(&dest->dst_lock);
	__ip_vs_update_dest(svc, dest, udest, 1);

	return 0;

err_stats:
	ip_vs_stats_release(&dest->stats);

err_alloc:
	kfree(dest);
	return ret;
}


/*
 *	Add a destination into an existing service
 */
static int
ip_vs_add_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest)
{
	struct ip_vs_dest *dest;
	union nf_inet_addr daddr;
	__be16 dport = udest->port;
	int ret;

	if (udest->weight < 0) {
		pr_err("%s(): server weight less than zero\n", __func__);
		return -ERANGE;
	}

	if (udest->l_threshold > udest->u_threshold) {
		pr_err("%s(): lower threshold is higher than upper threshold\n",
			__func__);
		return -ERANGE;
	}

	if (udest->u_threshold > INT_MAX)
		return -EINVAL;

	if (udest->tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) {
		if (udest->tun_port == 0) {
			pr_err("%s(): tunnel port is zero\n", __func__);
			return -EINVAL;
		}
	}

	ip_vs_addr_copy(udest->af, &daddr, &udest->addr);

	/* We use function that requires RCU lock */
	rcu_read_lock();
	dest = ip_vs_lookup_dest(svc, udest->af, &daddr, dport);
	rcu_read_unlock();

	if (dest != NULL) {
		IP_VS_DBG(1, "%s(): dest already exists\n", __func__);
		return -EEXIST;
	}

	/*
	 * Check if the dest already exists in the trash and
	 * is from the same service
	 */
	dest = ip_vs_trash_get_dest(svc, udest->af, &daddr, dport);

	if (dest != NULL) {
		IP_VS_DBG_BUF(3, "Get destination %s:%u from trash, "
			      "dest->refcnt=%d, service %u/%s:%u\n",
			      IP_VS_DBG_ADDR(udest->af, &daddr), ntohs(dport),
			      refcount_read(&dest->refcnt),
			      dest->vfwmark,
			      IP_VS_DBG_ADDR(svc->af, &dest->vaddr),
			      ntohs(dest->vport));

		ret = ip_vs_start_estimator(svc->ipvs, &dest->stats);
		/* On error put back dest into the trash */
		if (ret < 0)
			ip_vs_trash_put_dest(svc->ipvs, dest, dest->idle_start,
					     false);
		else
			__ip_vs_update_dest(svc, dest, udest, 1);
	} else {
		/*
		 * Allocate and initialize the dest structure
		 */
		ret = ip_vs_new_dest(svc, udest);
	}

	return ret;
}


/*
 *	Edit a destination in the given service
 */
static int
ip_vs_edit_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest)
{
	struct ip_vs_dest *dest;
	union nf_inet_addr daddr;
	__be16 dport = udest->port;

	if (udest->weight < 0) {
		pr_err("%s(): server weight less than zero\n", __func__);
		return -ERANGE;
	}

	if (udest->l_threshold > udest->u_threshold) {
		pr_err("%s(): lower threshold is higher than upper threshold\n",
			__func__);
		return -ERANGE;
	}

	if (udest->u_threshold > INT_MAX)
		return -EINVAL;

	if (udest->tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) {
		if (udest->tun_port == 0) {
			pr_err("%s(): tunnel port is zero\n", __func__);
			return -EINVAL;
		}
	}

	ip_vs_addr_copy(udest->af, &daddr, &udest->addr);

	/* We use function that requires RCU lock */
	rcu_read_lock();
	dest = ip_vs_lookup_dest(svc, udest->af, &daddr, dport);
	rcu_read_unlock();

	if (dest == NULL) {
		IP_VS_DBG(1, "%s(): dest doesn't exist\n", __func__);
		return -ENOENT;
	}

	__ip_vs_update_dest(svc, dest, udest, 0);

	return 0;
}

/*
 *	Delete a destination (must be already unlinked from the service)
 */
static void __ip_vs_del_dest(struct netns_ipvs *ipvs, struct ip_vs_dest *dest,
			     bool cleanup)
{
	ip_vs_stop_estimator(ipvs, &dest->stats);

	/*
	 *  Remove it from the d-linked list with the real services.
	 */
	ip_vs_rs_unhash(dest);

	ip_vs_trash_put_dest(ipvs, dest, 0, cleanup);

	/* Queue up delayed work to expire all no destination connections.
	 * No-op when CONFIG_SYSCTL is disabled.
	 */
	if (!cleanup)
		ip_vs_enqueue_expire_nodest_conns(ipvs);
}


/*
 *	Unlink a destination from the given service
 */
static void __ip_vs_unlink_dest(struct ip_vs_service *svc,
				struct ip_vs_dest *dest,
				int svcupd)
{
	dest->cflags &= ~IP_VS_DEST_CF_AVAILABLE;

	spin_lock_bh(&dest->dst_lock);
	__ip_vs_dst_cache_reset(dest);
	spin_unlock_bh(&dest->dst_lock);

	/*
	 *  Remove it from the d-linked destination list.
	 */
	list_del_rcu(&dest->n_list);
	svc->num_dests--;

	if (dest->af != svc->af)
		svc->ipvs->mixed_address_family_dests--;

	if (svcupd) {
		struct ip_vs_scheduler *sched;

		sched = rcu_dereference_protected(svc->scheduler, 1);
		if (sched && sched->del_dest)
			sched->del_dest(svc, dest);
	}
}


/*
 *	Delete a destination server in the given service
 */
static int
ip_vs_del_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest)
{
	struct ip_vs_dest *dest;
	__be16 dport = udest->port;

	/* We use function that requires RCU lock */
	rcu_read_lock();
	dest = ip_vs_lookup_dest(svc, udest->af, &udest->addr, dport);
	rcu_read_unlock();

	if (dest == NULL) {
		IP_VS_DBG(1, "%s(): destination not found!\n", __func__);
		return -ENOENT;
	}

	/*
	 *	Unlink dest from the service
	 */
	__ip_vs_unlink_dest(svc, dest, 1);

	/*
	 *	Delete the destination
	 */
	__ip_vs_del_dest(svc->ipvs, dest, false);

	return 0;
}

static void ip_vs_dest_trash_expire(struct timer_list *t)
{
	struct netns_ipvs *ipvs = timer_container_of(ipvs, t,
						     dest_trash_timer);
	struct ip_vs_dest *dest, *next;
	unsigned long now = jiffies;

	spin_lock(&ipvs->dest_trash_lock);
	list_for_each_entry_safe(dest, next, &ipvs->dest_trash, t_list) {
		if (refcount_read(&dest->refcnt) > 1)
			continue;
		if (dest->idle_start) {
			if (time_before(now, dest->idle_start +
					     IP_VS_DEST_TRASH_PERIOD))
				continue;
		} else {
			dest->idle_start = max(1UL, now);
			continue;
		}
		IP_VS_DBG_BUF(3, "Removing destination %u/%s:%u from trash\n",
			      dest->vfwmark,
			      IP_VS_DBG_ADDR(dest->af, &dest->addr),
			      ntohs(dest->port));
		list_del(&dest->t_list);
		ip_vs_dest_free(dest);
	}
	if (!list_empty(&ipvs->dest_trash))
		mod_timer(&ipvs->dest_trash_timer,
			  jiffies + (IP_VS_DEST_TRASH_PERIOD >> 1));
	spin_unlock(&ipvs->dest_trash_lock);
}

/*
 *	Add a service into the service hash table
 */
static int
ip_vs_add_service(struct netns_ipvs *ipvs, struct ip_vs_service_user_kern *u,
		  struct ip_vs_service **svc_p)
{
	struct ip_vs_scheduler *sched = NULL;
	struct ip_vs_rht *tc_new = NULL;
	struct ip_vs_rht *t, *t_new = NULL;
	int af_id = ip_vs_af_index(u->af);
	struct ip_vs_service *svc = NULL;
	struct ip_vs_pe *pe = NULL;
	int ret_hooks = -1;
	int ret = 0;
	bool grow;

	/* increase the module use count */
	if (!ip_vs_use_count_inc())
		return -ENOPROTOOPT;

	/* Lookup the scheduler by 'u->sched_name' */
	if (strcmp(u->sched_name, "none")) {
		sched = ip_vs_scheduler_get(u->sched_name);
		if (!sched) {
			pr_info("Scheduler module ip_vs_%s not found\n",
				u->sched_name);
			ret = -ENOENT;
			goto out_err;
		}
	}

	if (u->pe_name && *u->pe_name) {
		pe = ip_vs_pe_getbyname(u->pe_name);
		if (pe == NULL) {
			pr_info("persistence engine module ip_vs_pe_%s "
				"not found\n", u->pe_name);
			ret = -ENOENT;
			goto out_err;
		}
	}

#ifdef CONFIG_IP_VS_IPV6
	if (u->af == AF_INET6) {
		__u32 plen = (__force __u32) u->netmask;

		if (plen < 1 || plen > 128) {
			ret = -EINVAL;
			goto out_err;
		}

		ret = nf_defrag_ipv6_enable(ipvs->net);
		if (ret)
			goto out_err;
	}
#endif

	/* The old table can be freed, protect it with RCU */
	rcu_read_lock();
	t = rcu_dereference(ipvs->svc_table);
	if (!t) {
		int lfactor = sysctl_svc_lfactor(ipvs);
		int new_size = ip_vs_svc_desired_size(ipvs, NULL, lfactor);

		rcu_read_unlock();
		t_new = ip_vs_svc_table_alloc(ipvs, new_size, lfactor);
		if (!t_new) {
			ret = -ENOMEM;
			goto out_err;
		}
		grow = false;
	} else {
		/* Even the currently attached new table may need to grow */
		t = rcu_dereference(t->new_tbl);
		grow = ip_vs_get_num_services(ipvs) + 1 > t->u_thresh;
		rcu_read_unlock();
	}

	if (!rcu_dereference_protected(ipvs->conn_tab, 1)) {
		int lfactor = sysctl_conn_lfactor(ipvs);
		int new_size = ip_vs_conn_desired_size(ipvs, NULL, lfactor);

		tc_new = ip_vs_conn_tab_alloc(ipvs, new_size, lfactor);
		if (!tc_new) {
			ret = -ENOMEM;
			goto out_err;
		}
	}

	if (!atomic_read(&ipvs->num_services[af_id])) {
		ret = ip_vs_register_hooks(ipvs, u->af);
		if (ret < 0)
			goto out_err;
		ret_hooks = ret;
	}

	svc = kzalloc_obj(struct ip_vs_service);
	if (svc == NULL) {
		IP_VS_DBG(1, "%s(): no memory\n", __func__);
		ret = -ENOMEM;
		goto out_err;
	}
	ret = ip_vs_stats_init_alloc(&svc->stats);
	if (ret < 0)
		goto out_err;

	/* I'm the first user of the service */
	atomic_set(&svc->refcnt, 0);

	svc->af = u->af;
	svc->protocol = u->protocol;
	ip_vs_addr_copy(svc->af, &svc->addr, &u->addr);
	svc->port = u->port;
	svc->fwmark = u->fwmark;
	svc->flags = u->flags & ~IP_VS_SVC_F_HASHED;
	svc->timeout = u->timeout * HZ;
	svc->netmask = u->netmask;
	svc->ipvs = ipvs;

	INIT_LIST_HEAD(&svc->destinations);
	spin_lock_init(&svc->sched_lock);

	/* Bind the scheduler */
	if (sched) {
		ret = ip_vs_bind_scheduler(svc, sched);
		if (ret)
			goto out_err;
	}

	ret = ip_vs_start_estimator(ipvs, &svc->stats);
	if (ret < 0)
		goto out_err;

	if (t_new) {
		/* Add table for first time */
		clear_bit(IP_VS_WORK_SVC_NORESIZE, &ipvs->work_flags);
		rcu_assign_pointer(ipvs->svc_table, t_new);
		t_new = NULL;
	}
	if (tc_new) {
		rcu_assign_pointer(ipvs->conn_tab, tc_new);
		tc_new = NULL;
	}

	/* Update the virtual service counters */
	if (svc->port == FTPPORT)
		atomic_inc(&ipvs->ftpsvc_counter[af_id]);
	else if (!svc->port && !svc->fwmark)
		atomic_inc(&ipvs->nullsvc_counter[af_id]);
	if (pe && pe->conn_out)
		atomic_inc(&ipvs->conn_out_counter[af_id]);

	/* Bind the ct retriever */
	RCU_INIT_POINTER(svc->pe, pe);
	pe = NULL;

	if (svc->fwmark)
		atomic_inc(&ipvs->fwm_services[af_id]);
	else
		atomic_inc(&ipvs->nonfwm_services[af_id]);
	atomic_inc(&ipvs->num_services[af_id]);

	/* Hash the service into the service table */
	ip_vs_svc_hash(svc);

	/* Schedule resize work */
	if (grow && !test_and_set_bit(IP_VS_WORK_SVC_RESIZE, &ipvs->work_flags))
		queue_delayed_work(system_dfl_long_wq, &ipvs->svc_resize_work,
				   1);

	*svc_p = svc;

	if (!READ_ONCE(ipvs->enable)) {
		mutex_lock(&ipvs->est_mutex);

		/* Now there is a service - full throttle */
		WRITE_ONCE(ipvs->enable, 1);

		ipvs->est_max_threads = ip_vs_est_max_threads(ipvs);

		/* Start estimation for first time */
		ip_vs_est_reload_start(ipvs, true);
		mutex_unlock(&ipvs->est_mutex);
	}

	return 0;


 out_err:
	if (tc_new)
		ip_vs_rht_free(tc_new);
	if (t_new)
		ip_vs_rht_free(t_new);
	if (ret_hooks >= 0)
		ip_vs_unregister_hooks(ipvs, u->af);
	if (svc != NULL) {
		ip_vs_unbind_scheduler(svc);
		ip_vs_service_free(svc);
	}
	ip_vs_scheduler_put(sched);
	ip_vs_pe_put(pe);

	/* decrease the module use count */
	ip_vs_use_count_dec();

	return ret;
}


/*
 *	Edit a service and bind it with a new scheduler
 */
static int
ip_vs_edit_service(struct ip_vs_service *svc, struct ip_vs_service_user_kern *u)
{
	struct ip_vs_scheduler *sched = NULL, *old_sched;
	struct ip_vs_pe *pe = NULL, *old_pe = NULL;
	int ret = 0;
	bool new_pe_conn_out, old_pe_conn_out;
	struct netns_ipvs *ipvs = svc->ipvs;
	int af_id = ip_vs_af_index(svc->af);

	/*
	 * Lookup the scheduler, by 'u->sched_name'
	 */
	if (strcmp(u->sched_name, "none")) {
		sched = ip_vs_scheduler_get(u->sched_name);
		if (!sched) {
			pr_info("Scheduler module ip_vs_%s not found\n",
				u->sched_name);
			return -ENOENT;
		}
	}
	old_sched = sched;

	if (u->pe_name && *u->pe_name) {
		pe = ip_vs_pe_getbyname(u->pe_name);
		if (pe == NULL) {
			pr_info("persistence engine module ip_vs_pe_%s "
				"not found\n", u->pe_name);
			ret = -ENOENT;
			goto out;
		}
		old_pe = pe;
	}

#ifdef CONFIG_IP_VS_IPV6
	if (u->af == AF_INET6) {
		__u32 plen = (__force __u32) u->netmask;

		if (plen < 1 || plen > 128) {
			ret = -EINVAL;
			goto out;
		}
	}
#endif

	old_sched = rcu_dereference_protected(svc->scheduler, 1);
	if (sched != old_sched) {
		if (old_sched) {
			ip_vs_unbind_scheduler(svc);
			/* Wait all svc->scheduler/sched_data users */
			synchronize_rcu();
		}
		/* Bind the new scheduler */
		if (sched) {
			ret = ip_vs_bind_scheduler(svc, sched);
			if (ret) {
				ip_vs_scheduler_put(sched);
				/* Try to restore the old_sched */
				if (old_sched &&
				    !ip_vs_bind_scheduler(svc, old_sched))
					old_sched = NULL;
				goto out;
			}
		}
	}

	/*
	 * Set the flags and timeout value
	 */
	svc->flags = u->flags | IP_VS_SVC_F_HASHED;
	svc->timeout = u->timeout * HZ;
	svc->netmask = u->netmask;

	old_pe = rcu_dereference_protected(svc->pe, 1);
	if (pe != old_pe) {
		rcu_assign_pointer(svc->pe, pe);
		/* check for optional methods in new pe */
		new_pe_conn_out = (pe && pe->conn_out) ? true : false;
		old_pe_conn_out = (old_pe && old_pe->conn_out) ? true : false;
		if (new_pe_conn_out && !old_pe_conn_out)
			atomic_inc(&ipvs->conn_out_counter[af_id]);
		if (old_pe_conn_out && !new_pe_conn_out)
			atomic_dec(&ipvs->conn_out_counter[af_id]);
	}

out:
	ip_vs_scheduler_put(old_sched);
	ip_vs_pe_put(old_pe);
	return ret;
}

/*
 *	Delete a service from the service list
 *	- The service must be unlinked, unlocked and not referenced!
 *	- We are called under _bh lock
 */
static void __ip_vs_del_service(struct ip_vs_service *svc, bool cleanup)
{
	struct ip_vs_dest *dest, *nxt;
	struct ip_vs_scheduler *old_sched;
	struct ip_vs_pe *old_pe;
	struct netns_ipvs *ipvs = svc->ipvs;
	int af_id = ip_vs_af_index(svc->af);

	atomic_dec(&ipvs->num_services[af_id]);
	if (!atomic_read(&ipvs->num_services[af_id]))
		ip_vs_unregister_hooks(ipvs, svc->af);
	if (svc->fwmark)
		atomic_dec(&ipvs->fwm_services[af_id]);
	else
		atomic_dec(&ipvs->nonfwm_services[af_id]);

	ip_vs_stop_estimator(svc->ipvs, &svc->stats);

	/* Unbind scheduler */
	old_sched = rcu_dereference_protected(svc->scheduler, 1);
	ip_vs_unbind_scheduler(svc);
	ip_vs_scheduler_put(old_sched);

	/* Unbind persistence engine, keep svc->pe */
	old_pe = rcu_dereference_protected(svc->pe, 1);
	if (old_pe && old_pe->conn_out)
		atomic_dec(&ipvs->conn_out_counter[af_id]);
	ip_vs_pe_put(old_pe);

	/*
	 *    Unlink the whole destination list
	 */
	list_for_each_entry_safe(dest, nxt, &svc->destinations, n_list) {
		__ip_vs_unlink_dest(svc, dest, 0);
		__ip_vs_del_dest(svc->ipvs, dest, cleanup);
	}

	/*
	 *    Update the virtual service counters
	 */
	if (svc->port == FTPPORT)
		atomic_dec(&ipvs->ftpsvc_counter[af_id]);
	else if (!svc->port && !svc->fwmark)
		atomic_dec(&ipvs->nullsvc_counter[af_id]);

	/*
	 *    Free the service if nobody refers to it
	 */
	__ip_vs_svc_put(svc);

	/* decrease the module use count */
	ip_vs_use_count_dec();
}

/*
 * Unlink a service from list and try to delete it if its refcnt reached 0
 */
static void ip_vs_unlink_service(struct ip_vs_service *svc, bool cleanup)
{
	ip_vs_unregister_conntrack(svc);
	/* Hold svc to avoid double release from dest_trash */
	atomic_inc(&svc->refcnt);
	/*
	 * Unhash it from the service table
	 */
	ip_vs_svc_unhash(svc);

	__ip_vs_del_service(svc, cleanup);
}

/*
 *	Delete a service from the service list
 */
static int ip_vs_del_service(struct ip_vs_service *svc)
{
	struct netns_ipvs *ipvs;
	struct ip_vs_rht *t, *p;
	int ns;

	if (svc == NULL)
		return -EEXIST;
	ipvs = svc->ipvs;
	ip_vs_unlink_service(svc, false);

	/* Drop the table if no more services */
	ns = ip_vs_get_num_services(ipvs);
	if (!ns) {
		/* Stop the resizer and drop the tables */
		set_bit(IP_VS_WORK_SVC_NORESIZE, &ipvs->work_flags);
		cancel_delayed_work_sync(&ipvs->svc_resize_work);
		t = rcu_dereference_protected(ipvs->svc_table, 1);
		if (t) {
			rcu_assign_pointer(ipvs->svc_table, NULL);
			/* Inform readers that table is removed */
			smp_mb__before_atomic();
			atomic_inc(&ipvs->svc_table_changes);
			while (1) {
				p = rcu_dereference_protected(t->new_tbl, 1);
				call_rcu(&t->rcu_head, ip_vs_rht_rcu_free);
				if (p == t)
					break;
				t = p;
			}
		}
	} else {
		bool shrink;

		rcu_read_lock();
		t = rcu_dereference(ipvs->svc_table);
		/* Even the currently attached new table may need to shrink */
		t = rcu_dereference(t->new_tbl);
		shrink = ns <= t->l_thresh;
		rcu_read_unlock();
		if (shrink && !test_and_set_bit(IP_VS_WORK_SVC_RESIZE,
						&ipvs->work_flags))
			queue_delayed_work(system_dfl_long_wq,
					   &ipvs->svc_resize_work, 1);
	}
	return 0;
}


/*
 *	Flush all the virtual services
 */
static int ip_vs_flush(struct netns_ipvs *ipvs, bool cleanup)
{
	DECLARE_IP_VS_RHT_WALK_BUCKETS();
	struct hlist_bl_head *head;
	struct ip_vs_service *svc;
	struct hlist_bl_node *ne;
	struct hlist_bl_node *e;
	struct ip_vs_rht *t, *p;

	/* Stop the resizer and drop the tables */
	if (!test_and_set_bit(IP_VS_WORK_SVC_NORESIZE, &ipvs->work_flags))
		cancel_delayed_work_sync(&ipvs->svc_resize_work);
	/* No resizer, so now we have exclusive write access */

	if (ip_vs_get_num_services(ipvs)) {
		ip_vs_rht_walk_buckets(ipvs->svc_table, head) {
			hlist_bl_for_each_entry_safe(svc, e, ne, head, s_list)
				ip_vs_unlink_service(svc, cleanup);
		}
	}

	/* Unregister the hash table and release it after RCU grace period */
	t = rcu_dereference_protected(ipvs->svc_table, 1);
	if (t) {
		rcu_assign_pointer(ipvs->svc_table, NULL);
		/* Inform readers that table is removed */
		smp_mb__before_atomic();
		atomic_inc(&ipvs->svc_table_changes);
		while (1) {
			p = rcu_dereference_protected(t->new_tbl, 1);
			call_rcu(&t->rcu_head, ip_vs_rht_rcu_free);
			if (p == t)
				break;
			t = p;
		}
	}
	/* Stop the tot_stats estimator early under service_mutex
	 * to avoid locking it again later.
	 */
	if (cleanup)
		ip_vs_stop_estimator_tot_stats(ipvs);
	return 0;
}

/*
 *	Delete service by {netns} in the service table.
 *	Called by __ip_vs_batch_cleanup()
 */
void ip_vs_service_nets_cleanup(struct list_head *net_list)
{
	struct netns_ipvs *ipvs;
	struct net *net;

	/* Check for "full" addressed entries */
	list_for_each_entry(net, net_list, exit_list) {
		ipvs = net_ipvs(net);
		mutex_lock(&ipvs->service_mutex);
		ip_vs_flush(ipvs, true);
		mutex_unlock(&ipvs->service_mutex);
	}
}

/* Put all references for device (dst_cache) */
static inline void
ip_vs_forget_dev(struct ip_vs_dest *dest, struct net_device *dev)
{
	struct ip_vs_dest_dst *dest_dst;

	spin_lock_bh(&dest->dst_lock);
	dest_dst = rcu_dereference_protected(dest->dest_dst, 1);
	if (dest_dst && dest_dst->dst_cache->dev == dev) {
		IP_VS_DBG_BUF(3, "Reset dev:%s dest %s:%u ,dest->refcnt=%d\n",
			      dev->name,
			      IP_VS_DBG_ADDR(dest->af, &dest->addr),
			      ntohs(dest->port),
			      refcount_read(&dest->refcnt));
		__ip_vs_dst_cache_reset(dest);
	}
	spin_unlock_bh(&dest->dst_lock);

}
/* Netdev event receiver
 * Currently only NETDEV_DOWN is handled to release refs to cached dsts
 */
static int ip_vs_dst_event(struct notifier_block *this, unsigned long event,
			   void *ptr)
{
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
	struct net *net = dev_net(dev);
	struct netns_ipvs *ipvs = net_ipvs(net);
	DECLARE_IP_VS_RHT_WALK_BUCKETS_RCU();
	unsigned int resched_score = 0;
	struct hlist_bl_head *head;
	struct ip_vs_service *svc;
	struct hlist_bl_node *e;
	struct ip_vs_dest *dest;
	int old_gen;

	if (event != NETDEV_DOWN || !ipvs)
		return NOTIFY_DONE;
	IP_VS_DBG(3, "%s() dev=%s\n", __func__, dev->name);

	/* Allow concurrent rehashing on resize but to avoid loop
	 * serialize with installing the new table.
	 */
	down_read(&ipvs->svc_replace_sem);

	old_gen = atomic_read(&ipvs->svc_table_changes);

	rcu_read_lock();

	smp_rmb(); /* ipvs->svc_table and svc_table_changes */
	ip_vs_rht_walk_buckets_rcu(ipvs->svc_table, head) {
		hlist_bl_for_each_entry_rcu(svc, e, head, s_list) {
			list_for_each_entry_rcu(dest, &svc->destinations,
						n_list) {
				ip_vs_forget_dev(dest, dev);
				resched_score += 10;
			}
			resched_score++;
		}
		resched_score++;
		if (resched_score >= 100) {
			cond_resched_rcu();
			/* Flushed? So no more dev refs */
			if (atomic_read(&ipvs->svc_table_changes) != old_gen)
				goto done;
			resched_score = 0;
		}
	}

done:
	rcu_read_unlock();
	up_read(&ipvs->svc_replace_sem);

	return NOTIFY_DONE;
}

/*
 *	Zero counters in a service or all services
 */
static int ip_vs_zero_service(struct ip_vs_service *svc)
{
	struct ip_vs_dest *dest;

	list_for_each_entry(dest, &svc->destinations, n_list) {
		ip_vs_zero_stats(&dest->stats);
	}
	ip_vs_zero_stats(&svc->stats);
	return 0;
}

static int ip_vs_zero_all(struct netns_ipvs *ipvs)
{
	DECLARE_IP_VS_RHT_WALK_BUCKETS_RCU();
	unsigned int resched_score = 0;
	struct hlist_bl_head *head;
	struct ip_vs_service *svc;
	struct hlist_bl_node *e;

	/* svc_table can not be replaced (svc_replace_sem) or
	 * removed (service_mutex)
	 */
	down_read(&ipvs->svc_replace_sem);
	rcu_read_lock();

	ip_vs_rht_walk_buckets_rcu(ipvs->svc_table, head) {
		hlist_bl_for_each_entry_rcu(svc, e, head, s_list) {
			ip_vs_zero_service(svc);
			resched_score += 10;
		}
		resched_score++;
		if (resched_score >= 100) {
			resched_score = 0;
			cond_resched_rcu();
		}
	}

	rcu_read_unlock();
	up_read(&ipvs->svc_replace_sem);

	ip_vs_zero_stats(&ipvs->tot_stats->s);
	return 0;
}

#ifdef CONFIG_SYSCTL

static int
proc_do_conn_max(const struct ctl_table *table, int write,
		 void *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = table->data;
	/* We can not use *valp to check if new value is provided, use INT_MIN
	 * for this because different admins change different limits.
	 */
	int unset = INT_MIN;
	int val = write ? unset : READ_ONCE(*valp);
	int rc;

	const struct ctl_table tmp = {
		.data = &val,
		.maxlen = sizeof(int),
	};

	rc = proc_dointvec(&tmp, write, buffer, lenp, ppos);
	if (write && !rc && val != unset) {
		struct netns_ipvs *ipvs = table->extra2;
		bool priv = capable(CAP_NET_ADMIN);
		int max;

		mutex_lock(&ipvs->service_mutex);
		/* Unprivileged admins can not go above the hard limit */
		max = priv ? IP_VS_CONN_MAX : ipvs->conn_max_limit;
		if (val < 0 || val > max) {
			rc = -EINVAL;
		} else {
			/* Privileged admin changes both limits */
			if (priv)
				ipvs->conn_max_limit = val;
			WRITE_ONCE(*valp, val);
		}
		mutex_unlock(&ipvs->service_mutex);
	}
	return rc;
}

static int
proc_do_defense_mode(const struct ctl_table *table, int write,
		     void *buffer, size_t *lenp, loff_t *ppos)
{
	struct netns_ipvs *ipvs = table->extra2;
	int *valp = table->data;
	int val = *valp;
	int rc;

	struct ctl_table tmp = {
		.data = &val,
		.maxlen = sizeof(int),
		.mode = table->mode,
	};

	rc = proc_dointvec(&tmp, write, buffer, lenp, ppos);
	if (write && (*valp != val)) {
		if (val < 0 || val > 3) {
			rc = -EINVAL;
		} else {
			*valp = val;
			update_defense_level(ipvs);
		}
	}
	return rc;
}

static int
proc_do_sync_threshold(const struct ctl_table *table, int write,
		       void *buffer, size_t *lenp, loff_t *ppos)
{
	struct netns_ipvs *ipvs = table->extra2;
	int *valp = table->data;
	int val[2];
	int rc;
	struct ctl_table tmp = {
		.data = &val,
		.maxlen = table->maxlen,
		.mode = table->mode,
	};

	mutex_lock(&ipvs->sync_mutex);
	memcpy(val, valp, sizeof(val));
	rc = proc_dointvec(&tmp, write, buffer, lenp, ppos);
	if (write) {
		if (val[0] < 0 || val[1] < 0 ||
		    (val[0] >= val[1] && val[1]))
			rc = -EINVAL;
		else
			memcpy(valp, val, sizeof(val));
	}
	mutex_unlock(&ipvs->sync_mutex);
	return rc;
}

static int
proc_do_sync_ports(const struct ctl_table *table, int write,
		   void *buffer, size_t *lenp, loff_t *ppos)
{
	int *valp = table->data;
	int val = *valp;
	int rc;

	struct ctl_table tmp = {
		.data = &val,
		.maxlen = sizeof(int),
		.mode = table->mode,
	};

	rc = proc_dointvec(&tmp, write, buffer, lenp, ppos);
	if (write && (*valp != val)) {
		if (val < 1 || !is_power_of_2(val))
			rc = -EINVAL;
		else
			*valp = val;
	}
	return rc;
}

static int ipvs_proc_est_cpumask_set(const struct ctl_table *table,
				     void *buffer)
{
	struct netns_ipvs *ipvs = table->extra2;
	cpumask_var_t *valp = table->data;
	cpumask_var_t newmask;
	int ret;

	if (!zalloc_cpumask_var(&newmask, GFP_KERNEL))
		return -ENOMEM;

	ret = cpulist_parse(buffer, newmask);
	if (ret)
		goto out;

	mutex_lock(&ipvs->est_mutex);

	if (!ipvs->est_cpulist_valid) {
		if (!zalloc_cpumask_var(valp, GFP_KERNEL)) {
			ret = -ENOMEM;
			goto unlock;
		}
		ipvs->est_cpulist_valid = 1;
	}
	cpumask_and(newmask, newmask, &current->cpus_mask);
	cpumask_copy(*valp, newmask);
	/* est_max_threads may depend on cpulist size */
	ipvs->est_max_threads = ip_vs_est_max_threads(ipvs);
	ipvs->est_calc_phase = 1;
	ip_vs_est_reload_start(ipvs, true);

unlock:
	mutex_unlock(&ipvs->est_mutex);

out:
	free_cpumask_var(newmask);
	return ret;
}

static int ipvs_proc_est_cpumask_get(const struct ctl_table *table,
				     void *buffer, size_t size)
{
	struct netns_ipvs *ipvs = table->extra2;
	cpumask_var_t *valp = table->data;
	struct cpumask *mask;
	int ret;

	mutex_lock(&ipvs->est_mutex);

	/* HK_TYPE_KTHREAD cpumask needs RCU protection */
	scoped_guard(rcu) {
		if (ipvs->est_cpulist_valid)
			mask = *valp;
		else
			mask = (struct cpumask *)housekeeping_cpumask(HK_TYPE_KTHREAD);
		ret = scnprintf(buffer, size, "%*pbl\n", cpumask_pr_args(mask));
	}

	mutex_unlock(&ipvs->est_mutex);

	return ret;
}

static int ipvs_proc_est_cpulist(const struct ctl_table *table, int write,
				 void *buffer, size_t *lenp, loff_t *ppos)
{
	int ret;

	/* Ignore both read and write(append) if *ppos not 0 */
	if (*ppos || !*lenp) {
		*lenp = 0;
		return 0;
	}
	if (write) {
		/* proc_sys_call_handler() appends terminator */
		ret = ipvs_proc_est_cpumask_set(table, buffer);
		if (ret >= 0)
			*ppos += *lenp;
	} else {
		/* proc_sys_call_handler() allocates 1 byte for terminator */
		ret = ipvs_proc_est_cpumask_get(table, buffer, *lenp + 1);
		if (ret >= 0) {
			*lenp = ret;
			*ppos += *lenp;
			ret = 0;
		}
	}
	return ret;
}

static int ipvs_proc_est_nice(const struct ctl_table *table, int write,
			      void *buffer, size_t *lenp, loff_t *ppos)
{
	struct netns_ipvs *ipvs = table->extra2;
	int *valp = table->data;
	int val = *valp;
	int ret;

	struct ctl_table tmp_table = {
		.data = &val,
		.maxlen = sizeof(int),
		.mode = table->mode,
	};

	ret = proc_dointvec(&tmp_table, write, buffer, lenp, ppos);
	if (write && ret >= 0) {
		if (val < MIN_NICE || val > MAX_NICE) {
			ret = -EINVAL;
		} else {
			mutex_lock(&ipvs->est_mutex);
			if (*valp != val) {
				*valp = val;
				ip_vs_est_reload_start(ipvs, true);
			}
			mutex_unlock(&ipvs->est_mutex);
		}
	}
	return ret;
}

static int ipvs_proc_run_estimation(const struct ctl_table *table, int write,
				    void *buffer, size_t *lenp, loff_t *ppos)
{
	struct netns_ipvs *ipvs = table->extra2;
	int *valp = table->data;
	int val = *valp;
	int ret;

	struct ctl_table tmp_table = {
		.data = &val,
		.maxlen = sizeof(int),
		.mode = table->mode,
	};

	ret = proc_dointvec(&tmp_table, write, buffer, lenp, ppos);
	if (write && ret >= 0) {
		mutex_lock(&ipvs->est_mutex);
		if (*valp != val) {
			*valp = val;
			ip_vs_est_reload_start(ipvs, true);
		}
		mutex_unlock(&ipvs->est_mutex);
	}
	return ret;
}

static int ipvs_proc_conn_lfactor(const struct ctl_table *table, int write,
				  void *buffer, size_t *lenp, loff_t *ppos)
{
	struct netns_ipvs *ipvs = table->extra2;
	int *valp = table->data;
	int val = *valp;
	int ret;

	struct ctl_table tmp_table = {
		.data = &val,
		.maxlen = sizeof(int),
	};

	ret = proc_dointvec(&tmp_table, write, buffer, lenp, ppos);
	if (write && ret >= 0) {
		if (val < -8 || val > 8) {
			ret = -EINVAL;
		} else {
			WRITE_ONCE(*valp, val);
			if (rcu_access_pointer(ipvs->conn_tab))
				mod_delayed_work(system_dfl_long_wq,
						 &ipvs->conn_resize_work, 0);
		}
	}
	return ret;
}

static int ipvs_proc_svc_lfactor(const struct ctl_table *table, int write,
				 void *buffer, size_t *lenp, loff_t *ppos)
{
	struct netns_ipvs *ipvs = table->extra2;
	int *valp = table->data;
	int val = *valp;
	int ret;

	struct ctl_table tmp_table = {
		.data = &val,
		.maxlen = sizeof(int),
	};

	ret = proc_dointvec(&tmp_table, write, buffer, lenp, ppos);
	if (write && ret >= 0) {
		if (val < -8 || val > 8) {
			ret = -EINVAL;
		} else {
			mutex_lock(&ipvs->service_mutex);
			WRITE_ONCE(*valp, val);
			/* Make sure the services are present */
			if (rcu_access_pointer(ipvs->svc_table) &&
			    READ_ONCE(ipvs->enable) &&
			    !test_bit(IP_VS_WORK_SVC_NORESIZE,
				      &ipvs->work_flags))
				mod_delayed_work(system_dfl_long_wq,
						 &ipvs->svc_resize_work, 0);
			mutex_unlock(&ipvs->service_mutex);
		}
	}
	return ret;
}

/*
 *	IPVS sysctl table (under the /proc/sys/net/ipv4/vs/)
 *	Do not change order or insert new entries without
 *	align with netns init in ip_vs_control_net_init()
 */

static struct ctl_table vs_vars[] = {
	{
		.procname	= "amemthresh",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "am_droprate",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "conn_max",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_do_conn_max,
	},
	{
		.procname	= "drop_entry",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_do_defense_mode,
	},
	{
		.procname	= "drop_packet",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_do_defense_mode,
	},
#ifdef CONFIG_IP_VS_NFCT
	{
		.procname	= "conntrack",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec,
	},
#endif
	{
		.procname	= "secure_tcp",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_do_defense_mode,
	},
	{
		.procname	= "snat_reroute",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec,
	},
	{
		.procname	= "sync_version",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
	},
	{
		.procname	= "sync_ports",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_do_sync_ports,
	},
	{
		.procname	= "sync_persist_mode",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "sync_qlen_max",
		.maxlen		= sizeof(unsigned long),
		.mode		= 0644,
		.proc_handler	= proc_doulongvec_minmax,
	},
	{
		.procname	= "sync_sock_size",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "cache_bypass",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "expire_nodest_conn",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "sloppy_tcp",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "sloppy_sctp",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "expire_quiescent_template",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "sync_threshold",
		.maxlen		=
			sizeof(((struct netns_ipvs *)0)->sysctl_sync_threshold),
		.mode		= 0644,
		.proc_handler	= proc_do_sync_threshold,
	},
	{
		.procname	= "sync_refresh_period",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
	{
		.procname	= "sync_retries",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_THREE,
	},
	{
		.procname	= "nat_icmp_send",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "pmtu_disc",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "backup_only",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "conn_reuse_mode",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "schedule_icmp",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "ignore_tunneled",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "run_estimation",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= ipvs_proc_run_estimation,
	},
	{
		.procname	= "est_cpulist",
		.maxlen		= NR_CPUS,	/* unused */
		.mode		= 0644,
		.proc_handler	= ipvs_proc_est_cpulist,
	},
	{
		.procname	= "est_nice",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= ipvs_proc_est_nice,
	},
	{
		.procname	= "conn_lfactor",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= ipvs_proc_conn_lfactor,
	},
	{
		.procname	= "svc_lfactor",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= ipvs_proc_svc_lfactor,
	},
#ifdef CONFIG_IP_VS_DEBUG
	{
		.procname	= "debug_level",
		.data		= &sysctl_ip_vs_debug_level,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
#endif
};

#endif

#ifdef CONFIG_PROC_FS

struct ip_vs_iter {
	struct seq_net_private p;  /* Do not move this, netns depends upon it*/
	struct ip_vs_rht *t;
	u32 bucket;
};

/*
 *	Write the contents of the VS rule table to a PROCfs file.
 *	(It is kept just for backward compatibility)
 */
static inline const char *ip_vs_fwd_name(unsigned int flags)
{
	switch (flags & IP_VS_CONN_F_FWD_MASK) {
	case IP_VS_CONN_F_LOCALNODE:
		return "Local";
	case IP_VS_CONN_F_TUNNEL:
		return "Tunnel";
	case IP_VS_CONN_F_DROUTE:
		return "Route";
	default:
		return "Masq";
	}
}

/* Do not expect consistent view during add, del and move(table resize).
 * We may miss entries and even show duplicates.
 */
static struct ip_vs_service *ip_vs_info_array(struct seq_file *seq, loff_t pos)
{
	struct ip_vs_iter *iter = seq->private;
	struct ip_vs_rht *t = iter->t;
	struct ip_vs_service *svc;
	struct hlist_bl_node *e;
	int idx;

	if (!t)
		return NULL;
	for (idx = 0; idx < t->size; idx++) {
		hlist_bl_for_each_entry_rcu(svc, e, &t->buckets[idx], s_list) {
			if (!ip_vs_rht_same_table(t, READ_ONCE(svc->hash_key)))
				break;
			if (pos-- == 0) {
				iter->bucket = idx;
				return svc;
			}
		}
	}
	return NULL;
}

static void *ip_vs_info_seq_start(struct seq_file *seq, loff_t *pos)
	__acquires(RCU)
{
	struct ip_vs_iter *iter = seq->private;
	struct net *net = seq_file_net(seq);
	struct netns_ipvs *ipvs = net_ipvs(net);

	rcu_read_lock();
	iter->t = rcu_dereference(ipvs->svc_table);
	return *pos ? ip_vs_info_array(seq, *pos - 1) : SEQ_START_TOKEN;
}


static void *ip_vs_info_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
	struct ip_vs_service *svc;
	struct ip_vs_iter *iter;
	struct hlist_bl_node *e;
	struct ip_vs_rht *t;

	++*pos;
	if (v == SEQ_START_TOKEN)
		return ip_vs_info_array(seq,0);

	svc = v;
	iter = seq->private;
	t = iter->t;
	if (!t)
		return NULL;

	hlist_bl_for_each_entry_continue_rcu(svc, e, s_list) {
		/* Our cursor was moved to new table ? */
		if (!ip_vs_rht_same_table(t, READ_ONCE(svc->hash_key)))
			break;
		return svc;
	}

	while (++iter->bucket < t->size) {
		hlist_bl_for_each_entry_rcu(svc, e, &t->buckets[iter->bucket],
					    s_list) {
			if (!ip_vs_rht_same_table(t, READ_ONCE(svc->hash_key)))
				break;
			return svc;
		}
	}
	return NULL;
}

static void ip_vs_info_seq_stop(struct seq_file *seq, void *v)
	__releases(RCU)
{
	rcu_read_unlock();
}


static int ip_vs_info_seq_show(struct seq_file *seq, void *v)
{
	struct net *net = seq_file_net(seq);
	struct netns_ipvs *ipvs = net_ipvs(net);

	if (v == SEQ_START_TOKEN) {
		seq_printf(seq,
			"IP Virtual Server version %d.%d.%d (size=%d)\n",
			NVERSION(IP_VS_VERSION_CODE), get_conn_tab_size(ipvs));
		seq_puts(seq,
			 "Prot LocalAddress:Port Scheduler Flags\n");
		seq_puts(seq,
			 "  -> RemoteAddress:Port Forward Weight ActiveConn InActConn\n");
	} else {
		const struct ip_vs_service *svc = v;
		const struct ip_vs_dest *dest;
		struct ip_vs_scheduler *sched = rcu_dereference(svc->scheduler);
		char *sched_name = sched ? sched->name : "none";

		if (!svc->fwmark) {
#ifdef CONFIG_IP_VS_IPV6
			if (svc->af == AF_INET6)
				seq_printf(seq, "%s  [%pI6]:%04X %s ",
					   ip_vs_proto_name(svc->protocol),
					   &svc->addr.in6,
					   ntohs(svc->port),
					   sched_name);
			else
#endif
				seq_printf(seq, "%s  %08X:%04X %s %s ",
					   ip_vs_proto_name(svc->protocol),
					   ntohl(svc->addr.ip),
					   ntohs(svc->port),
					   sched_name,
					   (svc->flags & IP_VS_SVC_F_ONEPACKET)?"ops ":"");
		} else {
			seq_printf(seq, "FWM  %08X %s %s",
				   svc->fwmark, sched_name,
				   (svc->flags & IP_VS_SVC_F_ONEPACKET)?"ops ":"");
		}

		if (svc->flags & IP_VS_SVC_F_PERSISTENT)
			seq_printf(seq, "persistent %d %08X\n",
				svc->timeout,
				ntohl(svc->netmask));
		else
			seq_putc(seq, '\n');

		list_for_each_entry_rcu(dest, &svc->destinations, n_list) {
#ifdef CONFIG_IP_VS_IPV6
			if (dest->af == AF_INET6)
				seq_printf(seq,
					   "  -> [%pI6]:%04X"
					   "      %-7s %-6d %-10d %-10d\n",
					   &dest->addr.in6,
					   ntohs(dest->port),
					   ip_vs_fwd_name(atomic_read(&dest->conn_flags)),
					   atomic_read(&dest->weight),
					   atomic_read(&dest->activeconns),
					   ip_vs_dest_inactconns(dest));
			else
#endif
				seq_printf(seq,
					   "  -> %08X:%04X      "
					   "%-7s %-6d %-10d %-10d\n",
					   ntohl(dest->addr.ip),
					   ntohs(dest->port),
					   ip_vs_fwd_name(atomic_read(&dest->conn_flags)),
					   atomic_read(&dest->weight),
					   atomic_read(&dest->activeconns),
					   ip_vs_dest_inactconns(dest));

		}
	}
	return 0;
}

static const struct seq_operations ip_vs_info_seq_ops = {
	.start = ip_vs_info_seq_start,
	.next  = ip_vs_info_seq_next,
	.stop  = ip_vs_info_seq_stop,
	.show  = ip_vs_info_seq_show,
};

static int ip_vs_stats_show(struct seq_file *seq, void *v)
{
	struct net *net = seq_file_single_net(seq);
	struct ip_vs_kstats show;

/*               01234567 01234567 01234567 0123456701234567 0123456701234567 */
	seq_puts(seq,
		 "   Total Incoming Outgoing         Incoming         Outgoing\n");
	seq_puts(seq,
		 "   Conns  Packets  Packets            Bytes            Bytes\n");

	ip_vs_copy_stats(&show, &net_ipvs(net)->tot_stats->s);
	seq_printf(seq, "%8LX %8LX %8LX %16LX %16LX\n\n",
		   (unsigned long long)show.conns,
		   (unsigned long long)show.inpkts,
		   (unsigned long long)show.outpkts,
		   (unsigned long long)show.inbytes,
		   (unsigned long long)show.outbytes);

/*                01234567 01234567 01234567 0123456701234567 0123456701234567*/
	seq_puts(seq,
		 " Conns/s   Pkts/s   Pkts/s          Bytes/s          Bytes/s\n");
	seq_printf(seq, "%8LX %8LX %8LX %16LX %16LX\n",
		   (unsigned long long)show.cps,
		   (unsigned long long)show.inpps,
		   (unsigned long long)show.outpps,
		   (unsigned long long)show.inbps,
		   (unsigned long long)show.outbps);

	return 0;
}

static int ip_vs_stats_percpu_show(struct seq_file *seq, void *v)
{
	struct net *net = seq_file_single_net(seq);
	struct ip_vs_stats *tot_stats = &net_ipvs(net)->tot_stats->s;
	struct ip_vs_cpu_stats __percpu *cpustats = tot_stats->cpustats;
	struct ip_vs_kstats kstats;
	int i;

/*               01234567 01234567 01234567 0123456701234567 0123456701234567 */
	seq_puts(seq,
		 "       Total Incoming Outgoing         Incoming         Outgoing\n");
	seq_puts(seq,
		 "CPU    Conns  Packets  Packets            Bytes            Bytes\n");

	for_each_possible_cpu(i) {
		struct ip_vs_cpu_stats *u = per_cpu_ptr(cpustats, i);
		unsigned int start;
		u64 conns, inpkts, outpkts, inbytes, outbytes;

		do {
			start = u64_stats_fetch_begin(&u->syncp);
			conns = u64_stats_read(&u->cnt.conns);
			inpkts = u64_stats_read(&u->cnt.inpkts);
			outpkts = u64_stats_read(&u->cnt.outpkts);
			inbytes = u64_stats_read(&u->cnt.inbytes);
			outbytes = u64_stats_read(&u->cnt.outbytes);
		} while (u64_stats_fetch_retry(&u->syncp, start));

		seq_printf(seq, "%3X %8LX %8LX %8LX %16LX %16LX\n",
			   i, (u64)conns, (u64)inpkts,
			   (u64)outpkts, (u64)inbytes,
			   (u64)outbytes);
	}

	ip_vs_copy_stats(&kstats, tot_stats);

	seq_printf(seq, "  ~ %8LX %8LX %8LX %16LX %16LX\n\n",
		   (unsigned long long)kstats.conns,
		   (unsigned long long)kstats.inpkts,
		   (unsigned long long)kstats.outpkts,
		   (unsigned long long)kstats.inbytes,
		   (unsigned long long)kstats.outbytes);

/*                ... 01234567 01234567 01234567 0123456701234567 0123456701234567 */
	seq_puts(seq,
		 "     Conns/s   Pkts/s   Pkts/s          Bytes/s          Bytes/s\n");
	seq_printf(seq, "    %8LX %8LX %8LX %16LX %16LX\n",
		   kstats.cps,
		   kstats.inpps,
		   kstats.outpps,
		   kstats.inbps,
		   kstats.outbps);

	return 0;
}

static int ip_vs_status_show(struct seq_file *seq, void *v)
{
	struct net *net = seq_file_single_net(seq);
	struct netns_ipvs *ipvs = net_ipvs(net);
	unsigned int resched_score = 0;
	struct ip_vs_conn_hnode *hn;
	struct hlist_bl_head *head;
	struct ip_vs_service *svc;
	struct ip_vs_rht *t, *pt;
	struct hlist_bl_node *e;
	int old_gen, new_gen;
	u32 counts[8];
	u32 bucket;
	u32 count;
	int loops;
	u32 sum1;
	u32 sum;
	int i;

	/* Info for conns */
	rcu_read_lock();

	t = rcu_dereference(ipvs->conn_tab);

	seq_printf(seq, "Conns:\t%d\n", atomic_read(&ipvs->conn_count));
	seq_printf(seq, "Conn buckets:\t%d (%d bits, lfactor %d)\n",
		   t ? t->size : 0, t ? t->bits : 0, t ? t->lfactor : 0);

	if (!atomic_read(&ipvs->conn_count))
		goto after_conns;
	old_gen = atomic_read(&ipvs->conn_tab_changes);
	loops = 0;

repeat_conn:
	smp_rmb(); /* ipvs->conn_tab and conn_tab_changes */
	memset(counts, 0, sizeof(counts));
	ip_vs_rht_for_each_table_rcu(ipvs->conn_tab, t, pt) {
		for (bucket = 0; bucket < t->size; bucket++) {
			DECLARE_IP_VS_RHT_WALK_BUCKET_RCU();

			count = 0;
			resched_score++;
			ip_vs_rht_walk_bucket_rcu(t, bucket, head) {
				count = 0;
				hlist_bl_for_each_entry_rcu(hn, e, head, node) {
					count++;
					if (count >= ARRAY_SIZE(counts) - 1)
						break;
				}
			}
			resched_score += count;
			if (resched_score >= 100) {
				resched_score = 0;
				cond_resched_rcu();
				new_gen = atomic_read(&ipvs->conn_tab_changes);
				/* New table installed ? */
				if (old_gen != new_gen) {
					/* Too many changes? */
					if (++loops >= 5)
						goto after_conns;
					old_gen = new_gen;
					goto repeat_conn;
				}
			}
			counts[count]++;
		}
	}
	for (sum = 0, i = 0; i < ARRAY_SIZE(counts); i++)
		sum += counts[i];
	sum1 = sum - counts[0];
	seq_printf(seq, "Conn buckets empty:\t%u (%llu%%)\n",
		   counts[0], div_u64((u64)counts[0] * 100U, max(sum, 1U)));
	for (i = 1; i < ARRAY_SIZE(counts); i++) {
		if (!counts[i])
			continue;
		seq_printf(seq, "Conn buckets len-%d:\t%u (%llu%%)\n",
			   i, counts[i],
			   div_u64((u64)counts[i] * 100U, max(sum1, 1U)));
	}

after_conns:
	rcu_read_unlock();

	/* Info for services */
	down_read(&ipvs->svc_replace_sem);
	rcu_read_lock();

	t = rcu_dereference(ipvs->svc_table);

	count = ip_vs_get_num_services(ipvs);
	seq_printf(seq, "Services:\t%u\n", count);
	seq_printf(seq, "Service buckets:\t%d (%d bits, lfactor %d)\n",
		   t ? t->size : 0, t ? t->bits : 0, t ? t->lfactor : 0);

	if (!count)
		goto after_svc;
	old_gen = atomic_read(&ipvs->svc_table_changes);

	smp_rmb(); /* ipvs->svc_table and svc_table_changes */
	memset(counts, 0, sizeof(counts));
	ip_vs_rht_for_each_table_rcu(ipvs->svc_table, t, pt) {
		for (bucket = 0; bucket < t->size; bucket++) {
			DECLARE_IP_VS_RHT_WALK_BUCKET_RCU();

			count = 0;
			resched_score++;
			ip_vs_rht_walk_bucket_rcu(t, bucket, head) {
				count = 0;
				hlist_bl_for_each_entry_rcu(svc, e, head,
							    s_list) {
					count++;
					if (count >= ARRAY_SIZE(counts) - 1)
						break;
				}
			}
			resched_score += count;
			if (resched_score >= 100) {
				resched_score = 0;
				cond_resched_rcu();
				/* Flushed? */
				if (atomic_read(&ipvs->svc_table_changes) !=
				    old_gen)
					goto after_svc;
			}
			counts[count]++;
		}
	}
	for (sum = 0, i = 0; i < ARRAY_SIZE(counts); i++)
		sum += counts[i];
	sum1 = sum - counts[0];
	seq_printf(seq, "Service buckets empty:\t%u (%llu%%)\n",
		   counts[0], div_u64((u64)counts[0] * 100U, max(sum, 1U)));
	for (i = 1; i < ARRAY_SIZE(counts); i++) {
		if (!counts[i])
			continue;
		seq_printf(seq, "Service buckets len-%d:\t%u (%llu%%)\n",
			   i, counts[i],
			   div_u64((u64)counts[i] * 100U, max(sum1, 1U)));
	}

after_svc:
	rcu_read_unlock();
	up_read(&ipvs->svc_replace_sem);

	seq_printf(seq, "Stats thread slots:\t%d (max %lu)\n",
		   ipvs->est_kt_count, ipvs->est_max_threads);
	seq_printf(seq, "Stats chain max len:\t%d\n", ipvs->est_chain_max);
	seq_printf(seq, "Stats thread ests:\t%d\n",
		   ipvs->est_chain_max * IPVS_EST_CHAIN_FACTOR *
		   IPVS_EST_NTICKS);

	return 0;
}

#endif

/*
 *	Set timeout values for tcp tcpfin udp in the timeout_table.
 */
static int ip_vs_set_timeout(struct netns_ipvs *ipvs, struct ip_vs_timeout_user *u)
{
#if defined(CONFIG_IP_VS_PROTO_TCP) || defined(CONFIG_IP_VS_PROTO_UDP)
	struct ip_vs_proto_data *pd;
#endif

	IP_VS_DBG(2, "Setting timeout tcp:%d tcpfin:%d udp:%d\n",
		  u->tcp_timeout,
		  u->tcp_fin_timeout,
		  u->udp_timeout);

#ifdef CONFIG_IP_VS_PROTO_TCP
	if (u->tcp_timeout < 0 || u->tcp_timeout > (INT_MAX / HZ) ||
	    u->tcp_fin_timeout < 0 || u->tcp_fin_timeout > (INT_MAX / HZ)) {
		return -EINVAL;
	}
#endif

#ifdef CONFIG_IP_VS_PROTO_UDP
	if (u->udp_timeout < 0 || u->udp_timeout > (INT_MAX / HZ))
		return -EINVAL;
#endif

#ifdef CONFIG_IP_VS_PROTO_TCP
	if (u->tcp_timeout) {
		pd = ip_vs_proto_data_get(ipvs, IPPROTO_TCP);
		pd->timeout_table[IP_VS_TCP_S_ESTABLISHED]
			= u->tcp_timeout * HZ;
	}

	if (u->tcp_fin_timeout) {
		pd = ip_vs_proto_data_get(ipvs, IPPROTO_TCP);
		pd->timeout_table[IP_VS_TCP_S_FIN_WAIT]
			= u->tcp_fin_timeout * HZ;
	}
#endif

#ifdef CONFIG_IP_VS_PROTO_UDP
	if (u->udp_timeout) {
		pd = ip_vs_proto_data_get(ipvs, IPPROTO_UDP);
		pd->timeout_table[IP_VS_UDP_S_NORMAL]
			= u->udp_timeout * HZ;
	}
#endif
	return 0;
}

#define CMDID(cmd)		(cmd - IP_VS_BASE_CTL)

struct ip_vs_svcdest_user {
	struct ip_vs_service_user	s;
	struct ip_vs_dest_user		d;
};

static const unsigned char set_arglen[CMDID(IP_VS_SO_SET_MAX) + 1] = {
	[CMDID(IP_VS_SO_SET_ADD)]         = sizeof(struct ip_vs_service_user),
	[CMDID(IP_VS_SO_SET_EDIT)]        = sizeof(struct ip_vs_service_user),
	[CMDID(IP_VS_SO_SET_DEL)]         = sizeof(struct ip_vs_service_user),
	[CMDID(IP_VS_SO_SET_ADDDEST)]     = sizeof(struct ip_vs_svcdest_user),
	[CMDID(IP_VS_SO_SET_DELDEST)]     = sizeof(struct ip_vs_svcdest_user),
	[CMDID(IP_VS_SO_SET_EDITDEST)]    = sizeof(struct ip_vs_svcdest_user),
	[CMDID(IP_VS_SO_SET_TIMEOUT)]     = sizeof(struct ip_vs_timeout_user),
	[CMDID(IP_VS_SO_SET_STARTDAEMON)] = sizeof(struct ip_vs_daemon_user),
	[CMDID(IP_VS_SO_SET_STOPDAEMON)]  = sizeof(struct ip_vs_daemon_user),
	[CMDID(IP_VS_SO_SET_ZERO)]        = sizeof(struct ip_vs_service_user),
};

union ip_vs_set_arglen {
	struct ip_vs_service_user	field_IP_VS_SO_SET_ADD;
	struct ip_vs_service_user	field_IP_VS_SO_SET_EDIT;
	struct ip_vs_service_user	field_IP_VS_SO_SET_DEL;
	struct ip_vs_svcdest_user	field_IP_VS_SO_SET_ADDDEST;
	struct ip_vs_svcdest_user	field_IP_VS_SO_SET_DELDEST;
	struct ip_vs_svcdest_user	field_IP_VS_SO_SET_EDITDEST;
	struct ip_vs_timeout_user	field_IP_VS_SO_SET_TIMEOUT;
	struct ip_vs_daemon_user	field_IP_VS_SO_SET_STARTDAEMON;
	struct ip_vs_daemon_user	field_IP_VS_SO_SET_STOPDAEMON;
	struct ip_vs_service_user	field_IP_VS_SO_SET_ZERO;
};

#define MAX_SET_ARGLEN	sizeof(union ip_vs_set_arglen)

static void ip_vs_copy_usvc_compat(struct ip_vs_service_user_kern *usvc,
				  struct ip_vs_service_user *usvc_compat)
{
	memset(usvc, 0, sizeof(*usvc));

	usvc->af		= AF_INET;
	usvc->protocol		= usvc_compat->protocol;
	usvc->addr.ip		= usvc_compat->addr;
	usvc->port		= usvc_compat->port;
	usvc->fwmark		= usvc_compat->fwmark;

	/* Deep copy of sched_name is not needed here */
	usvc->sched_name	= usvc_compat->sched_name;

	usvc->flags		= usvc_compat->flags;
	usvc->timeout		= usvc_compat->timeout;
	usvc->netmask		= usvc_compat->netmask;
}

static void ip_vs_copy_udest_compat(struct ip_vs_dest_user_kern *udest,
				   struct ip_vs_dest_user *udest_compat)
{
	memset(udest, 0, sizeof(*udest));

	udest->addr.ip		= udest_compat->addr;
	udest->port		= udest_compat->port;
	udest->conn_flags	= udest_compat->conn_flags;
	udest->weight		= udest_compat->weight;
	udest->u_threshold	= udest_compat->u_threshold;
	udest->l_threshold	= udest_compat->l_threshold;
	udest->af		= AF_INET;
	udest->tun_type		= IP_VS_CONN_F_TUNNEL_TYPE_IPIP;
}

static int
do_ip_vs_set_ctl(struct sock *sk, int cmd, sockptr_t ptr, unsigned int len)
{
	struct net *net = sock_net(sk);
	int ret;
	unsigned char arg[MAX_SET_ARGLEN];
	struct ip_vs_service_user *usvc_compat;
	struct ip_vs_service_user_kern usvc;
	struct ip_vs_service *svc;
	struct ip_vs_dest_user *udest_compat;
	struct ip_vs_dest_user_kern udest;
	struct netns_ipvs *ipvs = net_ipvs(net);

	BUILD_BUG_ON(sizeof(arg) > 255);
	if (!ns_capable(sock_net(sk)->user_ns, CAP_NET_ADMIN))
		return -EPERM;

	if (cmd < IP_VS_BASE_CTL || cmd > IP_VS_SO_SET_MAX)
		return -EINVAL;
	if (len != set_arglen[CMDID(cmd)]) {
		IP_VS_DBG(1, "set_ctl: len %u != %u\n",
			  len, set_arglen[CMDID(cmd)]);
		return -EINVAL;
	}

	if (copy_from_sockptr(arg, ptr, len) != 0)
		return -EFAULT;

	/* Handle daemons since they have another lock */
	if (cmd == IP_VS_SO_SET_STARTDAEMON ||
	    cmd == IP_VS_SO_SET_STOPDAEMON) {
		struct ip_vs_daemon_user *dm = (struct ip_vs_daemon_user *)arg;

		if (cmd == IP_VS_SO_SET_STARTDAEMON) {
			struct ipvs_sync_daemon_cfg cfg;

			memset(&cfg, 0, sizeof(cfg));
			ret = -EINVAL;
			if (strscpy(cfg.mcast_ifn, dm->mcast_ifn,
				    sizeof(cfg.mcast_ifn)) <= 0)
				return ret;
			cfg.syncid = dm->syncid;
			ret = start_sync_thread(ipvs, &cfg, dm->state);
		} else {
			ret = stop_sync_thread(ipvs, dm->state);
		}
		return ret;
	}

	mutex_lock(&ipvs->service_mutex);
	if (cmd == IP_VS_SO_SET_FLUSH) {
		/* Flush the virtual service */
		ret = ip_vs_flush(ipvs, false);
		goto out_unlock;
	} else if (cmd == IP_VS_SO_SET_TIMEOUT) {
		/* Set timeout values for (tcp tcpfin udp) */
		ret = ip_vs_set_timeout(ipvs, (struct ip_vs_timeout_user *)arg);
		goto out_unlock;
	} else if (!len) {
		/* No more commands with len == 0 below */
		ret = -EINVAL;
		goto out_unlock;
	}

	usvc_compat = (struct ip_vs_service_user *)arg;
	udest_compat = (struct ip_vs_dest_user *)(usvc_compat + 1);

	/* We only use the new structs internally, so copy userspace compat
	 * structs to extended internal versions */
	ip_vs_copy_usvc_compat(&usvc, usvc_compat);
	ip_vs_copy_udest_compat(&udest, udest_compat);

	if (cmd == IP_VS_SO_SET_ZERO) {
		/* if no service address is set, zero counters in all */
		if (!usvc.fwmark && !usvc.addr.ip && !usvc.port) {
			ret = ip_vs_zero_all(ipvs);
			goto out_unlock;
		}
	}

	if ((cmd == IP_VS_SO_SET_ADD || cmd == IP_VS_SO_SET_EDIT) &&
	    strnlen(usvc.sched_name, IP_VS_SCHEDNAME_MAXLEN) ==
	    IP_VS_SCHEDNAME_MAXLEN) {
		ret = -EINVAL;
		goto out_unlock;
	}

	/* Check for valid protocol: TCP or UDP or SCTP, even for fwmark!=0 */
	if (usvc.protocol != IPPROTO_TCP && usvc.protocol != IPPROTO_UDP &&
	    usvc.protocol != IPPROTO_SCTP) {
		pr_err("set_ctl: invalid protocol: %d %pI4:%d\n",
		       usvc.protocol, &usvc.addr.ip,
		       ntohs(usvc.port));
		ret = -EFAULT;
		goto out_unlock;
	}

	/* Lookup the exact service by <protocol, addr, port> or fwmark */
	rcu_read_lock();
	if (usvc.fwmark == 0)
		svc = __ip_vs_service_find(ipvs, usvc.af, usvc.protocol,
					   &usvc.addr, usvc.port);
	else
		svc = __ip_vs_svc_fwm_find(ipvs, usvc.af, usvc.fwmark);
	rcu_read_unlock();

	if (cmd != IP_VS_SO_SET_ADD
	    && (svc == NULL || svc->protocol != usvc.protocol)) {
		ret = -ESRCH;
		goto out_unlock;
	}

	switch (cmd) {
	case IP_VS_SO_SET_ADD:
		if (svc != NULL)
			ret = -EEXIST;
		else
			ret = ip_vs_add_service(ipvs, &usvc, &svc);
		break;
	case IP_VS_SO_SET_EDIT:
		ret = ip_vs_edit_service(svc, &usvc);
		break;
	case IP_VS_SO_SET_DEL:
		ret = ip_vs_del_service(svc);
		if (!ret)
			goto out_unlock;
		break;
	case IP_VS_SO_SET_ZERO:
		ret = ip_vs_zero_service(svc);
		break;
	case IP_VS_SO_SET_ADDDEST:
		ret = ip_vs_add_dest(svc, &udest);
		break;
	case IP_VS_SO_SET_EDITDEST:
		ret = ip_vs_edit_dest(svc, &udest);
		break;
	case IP_VS_SO_SET_DELDEST:
		ret = ip_vs_del_dest(svc, &udest);
		break;
	default:
		WARN_ON_ONCE(1);
		ret = -EINVAL;
		break;
	}

  out_unlock:
	mutex_unlock(&ipvs->service_mutex);
	return ret;
}


static void
ip_vs_copy_service(struct ip_vs_service_entry *dst, struct ip_vs_service *src)
{
	struct ip_vs_scheduler *sched;
	struct ip_vs_kstats kstats;
	char *sched_name;

	sched = rcu_dereference_protected(src->scheduler, 1);
	sched_name = sched ? sched->name : "none";
	dst->protocol = src->protocol;
	dst->addr = src->addr.ip;
	dst->port = src->port;
	dst->fwmark = src->fwmark;
	strscpy(dst->sched_name, sched_name, sizeof(dst->sched_name));
	dst->flags = src->flags;
	dst->timeout = src->timeout / HZ;
	dst->netmask = src->netmask;
	dst->num_dests = src->num_dests;
	ip_vs_copy_stats(&kstats, &src->stats);
	ip_vs_export_stats_user(&dst->stats, &kstats);
}

static inline int
__ip_vs_get_service_entries(struct netns_ipvs *ipvs,
			    const struct ip_vs_get_services *get,
			    struct ip_vs_get_services __user *uptr)
{
	struct ip_vs_service_entry entry;
	DECLARE_IP_VS_RHT_WALK_BUCKETS();
	struct hlist_bl_head *head;
	struct ip_vs_service *svc;
	struct hlist_bl_node *e;
	int count = 0;
	int ret = 0;

	lockdep_assert_held(&ipvs->svc_resize_sem);
	/* All svc_table modifications are disabled, go ahead */
	ip_vs_rht_walk_buckets(ipvs->svc_table, head) {
		hlist_bl_for_each_entry(svc, e, head, s_list) {
			/* Only expose IPv4 entries to old interface */
			if (svc->af != AF_INET)
				continue;

			if (count >= get->num_services)
				goto out;
			memset(&entry, 0, sizeof(entry));
			ip_vs_copy_service(&entry, svc);
			if (copy_to_user(&uptr->entrytable[count],
					 &entry, sizeof(entry))) {
				ret = -EFAULT;
				goto out;
			}
			count++;
		}
	}

out:
	return ret;
}

static inline int
__ip_vs_get_dest_entries(struct netns_ipvs *ipvs, const struct ip_vs_get_dests *get,
			 struct ip_vs_get_dests __user *uptr)
{
	struct ip_vs_service *svc;
	union nf_inet_addr addr = { .ip = get->addr };
	int ret = 0;

	rcu_read_lock();
	if (get->fwmark)
		svc = __ip_vs_svc_fwm_find(ipvs, AF_INET, get->fwmark);
	else
		svc = __ip_vs_service_find(ipvs, AF_INET, get->protocol, &addr,
					   get->port);
	rcu_read_unlock();

	if (svc) {
		int count = 0;
		struct ip_vs_dest *dest;
		struct ip_vs_dest_entry entry;
		struct ip_vs_kstats kstats;

		memset(&entry, 0, sizeof(entry));
		list_for_each_entry(dest, &svc->destinations, n_list) {
			if (count >= get->num_dests)
				break;

			/* Cannot expose heterogeneous members via sockopt
			 * interface
			 */
			if (dest->af != svc->af)
				continue;

			entry.addr = dest->addr.ip;
			entry.port = dest->port;
			entry.conn_flags = atomic_read(&dest->conn_flags);
			entry.weight = atomic_read(&dest->weight);
			entry.u_threshold = READ_ONCE(dest->u_threshold);
			entry.l_threshold = READ_ONCE(dest->l_threshold);
			entry.activeconns = atomic_read(&dest->activeconns);
			entry.inactconns = ip_vs_dest_inactconns(dest);
			entry.persistconns = atomic_read(&dest->persistconns);
			ip_vs_copy_stats(&kstats, &dest->stats);
			ip_vs_export_stats_user(&entry.stats, &kstats);
			if (copy_to_user(&uptr->entrytable[count],
					 &entry, sizeof(entry))) {
				ret = -EFAULT;
				break;
			}
			count++;
		}
	} else
		ret = -ESRCH;
	return ret;
}

static inline void
__ip_vs_get_timeouts(struct netns_ipvs *ipvs, struct ip_vs_timeout_user *u)
{
#if defined(CONFIG_IP_VS_PROTO_TCP) || defined(CONFIG_IP_VS_PROTO_UDP)
	struct ip_vs_proto_data *pd;
#endif

	memset(u, 0, sizeof (*u));

#ifdef CONFIG_IP_VS_PROTO_TCP
	pd = ip_vs_proto_data_get(ipvs, IPPROTO_TCP);
	u->tcp_timeout = pd->timeout_table[IP_VS_TCP_S_ESTABLISHED] / HZ;
	u->tcp_fin_timeout = pd->timeout_table[IP_VS_TCP_S_FIN_WAIT] / HZ;
#endif
#ifdef CONFIG_IP_VS_PROTO_UDP
	pd = ip_vs_proto_data_get(ipvs, IPPROTO_UDP);
	u->udp_timeout =
			pd->timeout_table[IP_VS_UDP_S_NORMAL] / HZ;
#endif
}

static const unsigned char get_arglen[CMDID(IP_VS_SO_GET_MAX) + 1] = {
	[CMDID(IP_VS_SO_GET_VERSION)]  = 64,
	[CMDID(IP_VS_SO_GET_INFO)]     = sizeof(struct ip_vs_getinfo),
	[CMDID(IP_VS_SO_GET_SERVICES)] = sizeof(struct ip_vs_get_services),
	[CMDID(IP_VS_SO_GET_SERVICE)]  = sizeof(struct ip_vs_service_entry),
	[CMDID(IP_VS_SO_GET_DESTS)]    = sizeof(struct ip_vs_get_dests),
	[CMDID(IP_VS_SO_GET_TIMEOUT)]  = sizeof(struct ip_vs_timeout_user),
	[CMDID(IP_VS_SO_GET_DAEMON)]   = 2 * sizeof(struct ip_vs_daemon_user),
};

union ip_vs_get_arglen {
	char				field_IP_VS_SO_GET_VERSION[64];
	struct ip_vs_getinfo		field_IP_VS_SO_GET_INFO;
	struct ip_vs_get_services	field_IP_VS_SO_GET_SERVICES;
	struct ip_vs_service_entry	field_IP_VS_SO_GET_SERVICE;
	struct ip_vs_get_dests		field_IP_VS_SO_GET_DESTS;
	struct ip_vs_timeout_user	field_IP_VS_SO_GET_TIMEOUT;
	struct ip_vs_daemon_user	field_IP_VS_SO_GET_DAEMON[2];
};

#define MAX_GET_ARGLEN	sizeof(union ip_vs_get_arglen)

static int
do_ip_vs_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
{
	unsigned char arg[MAX_GET_ARGLEN];
	int ret = 0;
	unsigned int copylen;
	struct net *net = sock_net(sk);
	struct netns_ipvs *ipvs = net_ipvs(net);

	BUG_ON(!net);
	BUILD_BUG_ON(sizeof(arg) > 255);
	if (!ns_capable(sock_net(sk)->user_ns, CAP_NET_ADMIN))
		return -EPERM;

	if (cmd < IP_VS_BASE_CTL || cmd > IP_VS_SO_GET_MAX)
		return -EINVAL;

	copylen = get_arglen[CMDID(cmd)];
	if (*len < (int) copylen) {
		IP_VS_DBG(1, "get_ctl: len %d < %u\n", *len, copylen);
		return -EINVAL;
	}

	if (copy_from_user(arg, user, copylen) != 0)
		return -EFAULT;
	/*
	 * Handle daemons first since it has its own locking
	 */
	if (cmd == IP_VS_SO_GET_DAEMON) {
		struct ip_vs_daemon_user d[2];

		memset(&d, 0, sizeof(d));
		mutex_lock(&ipvs->sync_mutex);
		if (ipvs->sync_state & IP_VS_STATE_MASTER) {
			d[0].state = IP_VS_STATE_MASTER;
			strscpy(d[0].mcast_ifn, ipvs->mcfg.mcast_ifn,
				sizeof(d[0].mcast_ifn));
			d[0].syncid = ipvs->mcfg.syncid;
		}
		if (ipvs->sync_state & IP_VS_STATE_BACKUP) {
			d[1].state = IP_VS_STATE_BACKUP;
			strscpy(d[1].mcast_ifn, ipvs->bcfg.mcast_ifn,
				sizeof(d[1].mcast_ifn));
			d[1].syncid = ipvs->bcfg.syncid;
		}
		if (copy_to_user(user, &d, sizeof(d)) != 0)
			ret = -EFAULT;
		mutex_unlock(&ipvs->sync_mutex);
		return ret;
	}

	if (cmd == IP_VS_SO_GET_SERVICES) {
		struct ip_vs_get_services *get;
		size_t size;

		get = (struct ip_vs_get_services *)arg;
		size = struct_size(get, entrytable, get->num_services);
		if (*len != size) {
			pr_err("length: %u != %zu\n", *len, size);
			return -EINVAL;
		}
		/* Prevent modifications to the list with services.
		 * Try reverse locking, so that we do not hold the mutex
		 * while waiting for semaphore.
		 */
		while (1) {
			ret = down_read_killable(&ipvs->svc_resize_sem);
			if (ret < 0)
				return ret;
			if (mutex_trylock(&ipvs->service_mutex))
				break;
			up_read(&ipvs->svc_resize_sem);
			cond_resched();
		}
		ret = __ip_vs_get_service_entries(ipvs, get, user);
		up_read(&ipvs->svc_resize_sem);
		mutex_unlock(&ipvs->service_mutex);
		return ret;
	}

	mutex_lock(&ipvs->service_mutex);
	switch (cmd) {
	case IP_VS_SO_GET_VERSION:
	{
		char buf[64];

		sprintf(buf, "IP Virtual Server version %d.%d.%d (size=%d)",
			NVERSION(IP_VS_VERSION_CODE), get_conn_tab_size(ipvs));
		if (copy_to_user(user, buf, strlen(buf)+1) != 0) {
			ret = -EFAULT;
			goto out;
		}
		*len = strlen(buf)+1;
	}
	break;

	case IP_VS_SO_GET_INFO:
	{
		struct ip_vs_getinfo info;

		info.version = IP_VS_VERSION_CODE;
		info.size = get_conn_tab_size(ipvs);
		info.num_services =
			atomic_read(&ipvs->num_services[IP_VS_AF_INET]);
		if (copy_to_user(user, &info, sizeof(info)) != 0)
			ret = -EFAULT;
	}
	break;

	case IP_VS_SO_GET_SERVICE:
	{
		struct ip_vs_service_entry *entry;
		struct ip_vs_service *svc;
		union nf_inet_addr addr;

		entry = (struct ip_vs_service_entry *)arg;
		addr.ip = entry->addr;
		rcu_read_lock();
		if (entry->fwmark)
			svc = __ip_vs_svc_fwm_find(ipvs, AF_INET, entry->fwmark);
		else
			svc = __ip_vs_service_find(ipvs, AF_INET,
						   entry->protocol, &addr,
						   entry->port);
		rcu_read_unlock();
		if (svc) {
			ip_vs_copy_service(entry, svc);
			if (copy_to_user(user, entry, sizeof(*entry)) != 0)
				ret = -EFAULT;
		} else
			ret = -ESRCH;
	}
	break;

	case IP_VS_SO_GET_DESTS:
	{
		struct ip_vs_get_dests *get;
		size_t size;

		get = (struct ip_vs_get_dests *)arg;
		size = struct_size(get, entrytable, get->num_dests);
		if (*len != size) {
			pr_err("length: %u != %zu\n", *len, size);
			ret = -EINVAL;
			goto out;
		}
		ret = __ip_vs_get_dest_entries(ipvs, get, user);
	}
	break;

	case IP_VS_SO_GET_TIMEOUT:
	{
		struct ip_vs_timeout_user t;

		__ip_vs_get_timeouts(ipvs, &t);
		if (copy_to_user(user, &t, sizeof(t)) != 0)
			ret = -EFAULT;
	}
	break;

	default:
		ret = -EINVAL;
	}

out:
	mutex_unlock(&ipvs->service_mutex);
	return ret;
}


static struct nf_sockopt_ops ip_vs_sockopts = {
	.pf		= PF_INET,
	.set_optmin	= IP_VS_BASE_CTL,
	.set_optmax	= IP_VS_SO_SET_MAX+1,
	.set		= do_ip_vs_set_ctl,
	.get_optmin	= IP_VS_BASE_CTL,
	.get_optmax	= IP_VS_SO_GET_MAX+1,
	.get		= do_ip_vs_get_ctl,
	.owner		= THIS_MODULE,
};

/*
 * Generic Netlink interface
 */

/* IPVS genetlink family */
static struct genl_family ip_vs_genl_family;

/* Policy used for first-level command attributes */
static const struct nla_policy ip_vs_cmd_policy[IPVS_CMD_ATTR_MAX + 1] = {
	[IPVS_CMD_ATTR_SERVICE]		= { .type = NLA_NESTED },
	[IPVS_CMD_ATTR_DEST]		= { .type = NLA_NESTED },
	[IPVS_CMD_ATTR_DAEMON]		= { .type = NLA_NESTED },
	[IPVS_CMD_ATTR_TIMEOUT_TCP]	= { .type = NLA_U32 },
	[IPVS_CMD_ATTR_TIMEOUT_TCP_FIN]	= { .type = NLA_U32 },
	[IPVS_CMD_ATTR_TIMEOUT_UDP]	= { .type = NLA_U32 },
};

/* Policy used for attributes in nested attribute IPVS_CMD_ATTR_DAEMON */
static const struct nla_policy ip_vs_daemon_policy[IPVS_DAEMON_ATTR_MAX + 1] = {
	[IPVS_DAEMON_ATTR_STATE]	= { .type = NLA_U32 },
	[IPVS_DAEMON_ATTR_MCAST_IFN]	= { .type = NLA_NUL_STRING,
					    .len = IP_VS_IFNAME_MAXLEN - 1 },
	[IPVS_DAEMON_ATTR_SYNC_ID]	= { .type = NLA_U32 },
	[IPVS_DAEMON_ATTR_SYNC_MAXLEN]	= { .type = NLA_U16 },
	[IPVS_DAEMON_ATTR_MCAST_GROUP]	= { .type = NLA_U32 },
	[IPVS_DAEMON_ATTR_MCAST_GROUP6]	= { .len = sizeof(struct in6_addr) },
	[IPVS_DAEMON_ATTR_MCAST_PORT]	= { .type = NLA_U16 },
	[IPVS_DAEMON_ATTR_MCAST_TTL]	= { .type = NLA_U8 },
};

/* Policy used for attributes in nested attribute IPVS_CMD_ATTR_SERVICE */
static const struct nla_policy ip_vs_svc_policy[IPVS_SVC_ATTR_MAX + 1] = {
	[IPVS_SVC_ATTR_AF]		= { .type = NLA_U16 },
	[IPVS_SVC_ATTR_PROTOCOL]	= { .type = NLA_U16 },
	[IPVS_SVC_ATTR_ADDR]		= { .type = NLA_BINARY,
					    .len = sizeof(union nf_inet_addr) },
	[IPVS_SVC_ATTR_PORT]		= { .type = NLA_U16 },
	[IPVS_SVC_ATTR_FWMARK]		= { .type = NLA_U32 },
	[IPVS_SVC_ATTR_SCHED_NAME]	= { .type = NLA_NUL_STRING,
					    .len = IP_VS_SCHEDNAME_MAXLEN - 1 },
	[IPVS_SVC_ATTR_PE_NAME]		= { .type = NLA_NUL_STRING,
					    .len = IP_VS_PENAME_MAXLEN },
	[IPVS_SVC_ATTR_FLAGS]		= { .type = NLA_BINARY,
					    .len = sizeof(struct ip_vs_flags) },
	[IPVS_SVC_ATTR_TIMEOUT]		= { .type = NLA_U32 },
	[IPVS_SVC_ATTR_NETMASK]		= { .type = NLA_U32 },
	[IPVS_SVC_ATTR_STATS]		= { .type = NLA_NESTED },
};

/* Policy used for attributes in nested attribute IPVS_CMD_ATTR_DEST */
static const struct nla_policy ip_vs_dest_policy[IPVS_DEST_ATTR_MAX + 1] = {
	[IPVS_DEST_ATTR_ADDR]		= { .type = NLA_BINARY,
					    .len = sizeof(union nf_inet_addr) },
	[IPVS_DEST_ATTR_PORT]		= { .type = NLA_U16 },
	[IPVS_DEST_ATTR_FWD_METHOD]	= { .type = NLA_U32 },
	[IPVS_DEST_ATTR_WEIGHT]		= { .type = NLA_U32 },
	[IPVS_DEST_ATTR_U_THRESH]	= { .type = NLA_U32 },
	[IPVS_DEST_ATTR_L_THRESH]	= { .type = NLA_U32 },
	[IPVS_DEST_ATTR_ACTIVE_CONNS]	= { .type = NLA_U32 },
	[IPVS_DEST_ATTR_INACT_CONNS]	= { .type = NLA_U32 },
	[IPVS_DEST_ATTR_PERSIST_CONNS]	= { .type = NLA_U32 },
	[IPVS_DEST_ATTR_STATS]		= { .type = NLA_NESTED },
	[IPVS_DEST_ATTR_ADDR_FAMILY]	= { .type = NLA_U16 },
	[IPVS_DEST_ATTR_TUN_TYPE]	= { .type = NLA_U8 },
	[IPVS_DEST_ATTR_TUN_PORT]	= { .type = NLA_U16 },
	[IPVS_DEST_ATTR_TUN_FLAGS]	= { .type = NLA_U16 },
};

static int ip_vs_genl_fill_stats(struct sk_buff *skb, int container_type,
				 struct ip_vs_kstats *kstats)
{
	struct nlattr *nl_stats = nla_nest_start_noflag(skb, container_type);

	if (!nl_stats)
		return -EMSGSIZE;

	if (nla_put_u32(skb, IPVS_STATS_ATTR_CONNS, (u32)kstats->conns) ||
	    nla_put_u32(skb, IPVS_STATS_ATTR_INPKTS, (u32)kstats->inpkts) ||
	    nla_put_u32(skb, IPVS_STATS_ATTR_OUTPKTS, (u32)kstats->outpkts) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_INBYTES, kstats->inbytes,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_OUTBYTES, kstats->outbytes,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u32(skb, IPVS_STATS_ATTR_CPS, (u32)kstats->cps) ||
	    nla_put_u32(skb, IPVS_STATS_ATTR_INPPS, (u32)kstats->inpps) ||
	    nla_put_u32(skb, IPVS_STATS_ATTR_OUTPPS, (u32)kstats->outpps) ||
	    nla_put_u32(skb, IPVS_STATS_ATTR_INBPS, (u32)kstats->inbps) ||
	    nla_put_u32(skb, IPVS_STATS_ATTR_OUTBPS, (u32)kstats->outbps))
		goto nla_put_failure;
	nla_nest_end(skb, nl_stats);

	return 0;

nla_put_failure:
	nla_nest_cancel(skb, nl_stats);
	return -EMSGSIZE;
}

static int ip_vs_genl_fill_stats64(struct sk_buff *skb, int container_type,
				   struct ip_vs_kstats *kstats)
{
	struct nlattr *nl_stats = nla_nest_start_noflag(skb, container_type);

	if (!nl_stats)
		return -EMSGSIZE;

	if (nla_put_u64_64bit(skb, IPVS_STATS_ATTR_CONNS, kstats->conns,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_INPKTS, kstats->inpkts,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_OUTPKTS, kstats->outpkts,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_INBYTES, kstats->inbytes,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_OUTBYTES, kstats->outbytes,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_CPS, kstats->cps,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_INPPS, kstats->inpps,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_OUTPPS, kstats->outpps,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_INBPS, kstats->inbps,
			      IPVS_STATS_ATTR_PAD) ||
	    nla_put_u64_64bit(skb, IPVS_STATS_ATTR_OUTBPS, kstats->outbps,
			      IPVS_STATS_ATTR_PAD))
		goto nla_put_failure;
	nla_nest_end(skb, nl_stats);

	return 0;

nla_put_failure:
	nla_nest_cancel(skb, nl_stats);
	return -EMSGSIZE;
}

static int ip_vs_genl_fill_service(struct sk_buff *skb,
				   struct ip_vs_service *svc)
{
	struct ip_vs_scheduler *sched;
	struct ip_vs_pe *pe;
	struct nlattr *nl_service;
	struct ip_vs_flags flags = { .flags = svc->flags,
				     .mask = ~0 };
	struct ip_vs_kstats kstats;
	char *sched_name;

	nl_service = nla_nest_start_noflag(skb, IPVS_CMD_ATTR_SERVICE);
	if (!nl_service)
		return -EMSGSIZE;

	if (nla_put_u16(skb, IPVS_SVC_ATTR_AF, svc->af))
		goto nla_put_failure;
	if (svc->fwmark) {
		if (nla_put_u32(skb, IPVS_SVC_ATTR_FWMARK, svc->fwmark))
			goto nla_put_failure;
	} else {
		if (nla_put_u16(skb, IPVS_SVC_ATTR_PROTOCOL, svc->protocol) ||
		    nla_put(skb, IPVS_SVC_ATTR_ADDR, sizeof(svc->addr), &svc->addr) ||
		    nla_put_be16(skb, IPVS_SVC_ATTR_PORT, svc->port))
			goto nla_put_failure;
	}

	sched = rcu_dereference(svc->scheduler);
	sched_name = sched ? sched->name : "none";
	pe = rcu_dereference(svc->pe);
	if (nla_put_string(skb, IPVS_SVC_ATTR_SCHED_NAME, sched_name) ||
	    (pe && nla_put_string(skb, IPVS_SVC_ATTR_PE_NAME, pe->name)) ||
	    nla_put(skb, IPVS_SVC_ATTR_FLAGS, sizeof(flags), &flags) ||
	    nla_put_u32(skb, IPVS_SVC_ATTR_TIMEOUT, svc->timeout / HZ) ||
	    nla_put_be32(skb, IPVS_SVC_ATTR_NETMASK, svc->netmask))
		goto nla_put_failure;
	ip_vs_copy_stats(&kstats, &svc->stats);
	if (ip_vs_genl_fill_stats(skb, IPVS_SVC_ATTR_STATS, &kstats))
		goto nla_put_failure;
	if (ip_vs_genl_fill_stats64(skb, IPVS_SVC_ATTR_STATS64, &kstats))
		goto nla_put_failure;

	nla_nest_end(skb, nl_service);

	return 0;

nla_put_failure:
	nla_nest_cancel(skb, nl_service);
	return -EMSGSIZE;
}

static int ip_vs_genl_dump_service(struct sk_buff *skb,
				   struct ip_vs_service *svc,
				   struct netlink_callback *cb)
{
	void *hdr;

	hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
			  &ip_vs_genl_family, NLM_F_MULTI,
			  IPVS_CMD_NEW_SERVICE);
	if (!hdr)
		return -EMSGSIZE;

	if (ip_vs_genl_fill_service(skb, svc) < 0)
		goto nla_put_failure;

	genlmsg_end(skb, hdr);
	return 0;

nla_put_failure:
	genlmsg_cancel(skb, hdr);
	return -EMSGSIZE;
}

static int ip_vs_genl_dump_services(struct sk_buff *skb,
				    struct netlink_callback *cb)
{
	DECLARE_IP_VS_RHT_WALK_BUCKETS_SAFE_RCU();
	struct net *net = sock_net(skb->sk);
	struct netns_ipvs *ipvs = net_ipvs(net);
	struct hlist_bl_head *head;
	struct ip_vs_service *svc;
	struct hlist_bl_node *e;
	int start = cb->args[0];
	int idx = 0;

	/* Make sure we do not see same service twice during resize */
	down_read(&ipvs->svc_resize_sem);
	rcu_read_lock();
	ip_vs_rht_walk_buckets_safe_rcu(ipvs->svc_table, head) {
		hlist_bl_for_each_entry_rcu(svc, e, head, s_list) {
			if (++idx <= start)
				continue;
			if (ip_vs_genl_dump_service(skb, svc, cb) < 0) {
				idx--;
				goto nla_put_failure;
			}
		}
	}

nla_put_failure:
	rcu_read_unlock();
	up_read(&ipvs->svc_resize_sem);
	cb->args[0] = idx;

	return skb->len;
}

static bool ip_vs_is_af_valid(int af)
{
	if (af == AF_INET)
		return true;
#ifdef CONFIG_IP_VS_IPV6
	if (af == AF_INET6 && ipv6_mod_enabled())
		return true;
#endif
	return false;
}

static int ip_vs_genl_parse_service(struct netns_ipvs *ipvs,
				    struct ip_vs_service_user_kern *usvc,
				    struct nlattr *nla, bool full_entry,
				    struct ip_vs_service **ret_svc)
{
	struct nlattr *attrs[IPVS_SVC_ATTR_MAX + 1];
	struct nlattr *nla_af, *nla_port, *nla_fwmark, *nla_protocol, *nla_addr;
	struct ip_vs_service *svc;

	/* Parse mandatory identifying service fields first */
	if (nla == NULL ||
	    nla_parse_nested_deprecated(attrs, IPVS_SVC_ATTR_MAX, nla, ip_vs_svc_policy, NULL))
		return -EINVAL;

	nla_af		= attrs[IPVS_SVC_ATTR_AF];
	nla_protocol	= attrs[IPVS_SVC_ATTR_PROTOCOL];
	nla_addr	= attrs[IPVS_SVC_ATTR_ADDR];
	nla_port	= attrs[IPVS_SVC_ATTR_PORT];
	nla_fwmark	= attrs[IPVS_SVC_ATTR_FWMARK];

	if (!(nla_af && (nla_fwmark || (nla_port && nla_protocol && nla_addr))))
		return -EINVAL;

	memset(usvc, 0, sizeof(*usvc));

	usvc->af = nla_get_u16(nla_af);
	if (!ip_vs_is_af_valid(usvc->af))
		return -EAFNOSUPPORT;

	if (nla_fwmark) {
		usvc->protocol = IPPROTO_TCP;
		usvc->fwmark = nla_get_u32(nla_fwmark);
	} else {
		usvc->protocol = nla_get_u16(nla_protocol);
		nla_memcpy(&usvc->addr, nla_addr, sizeof(usvc->addr));
		usvc->port = nla_get_be16(nla_port);
		usvc->fwmark = 0;
	}

	if (usvc->fwmark)
		svc = __ip_vs_svc_fwm_find(ipvs, usvc->af, usvc->fwmark);
	else
		svc = __ip_vs_service_find(ipvs, usvc->af, usvc->protocol,
					   &usvc->addr, usvc->port);
	*ret_svc = svc;

	/* If a full entry was requested, check for the additional fields */
	if (full_entry) {
		struct nlattr *nla_sched, *nla_flags, *nla_pe, *nla_timeout,
			      *nla_netmask;
		struct ip_vs_flags flags;

		nla_sched = attrs[IPVS_SVC_ATTR_SCHED_NAME];
		nla_pe = attrs[IPVS_SVC_ATTR_PE_NAME];
		nla_flags = attrs[IPVS_SVC_ATTR_FLAGS];
		nla_timeout = attrs[IPVS_SVC_ATTR_TIMEOUT];
		nla_netmask = attrs[IPVS_SVC_ATTR_NETMASK];

		if (!(nla_sched && nla_flags && nla_timeout && nla_netmask))
			return -EINVAL;

		nla_memcpy(&flags, nla_flags, sizeof(flags));

		/* prefill flags from service if it already exists */
		if (svc)
			usvc->flags = svc->flags;

		/* set new flags from userland */
		usvc->flags = (usvc->flags & ~flags.mask) |
			      (flags.flags & flags.mask);
		usvc->sched_name = nla_data(nla_sched);
		usvc->pe_name = nla_pe ? nla_data(nla_pe) : NULL;
		usvc->timeout = nla_get_u32(nla_timeout);
		usvc->netmask = nla_get_be32(nla_netmask);
	}

	return 0;
}

static struct ip_vs_service *ip_vs_genl_find_service(struct netns_ipvs *ipvs,
						     struct nlattr *nla)
{
	struct ip_vs_service_user_kern usvc;
	struct ip_vs_service *svc;
	int ret;

	ret = ip_vs_genl_parse_service(ipvs, &usvc, nla, false, &svc);
	return ret ? ERR_PTR(ret) : svc;
}

static int ip_vs_genl_fill_dest(struct sk_buff *skb, struct ip_vs_dest *dest)
{
	struct nlattr *nl_dest;
	struct ip_vs_kstats kstats;

	nl_dest = nla_nest_start_noflag(skb, IPVS_CMD_ATTR_DEST);
	if (!nl_dest)
		return -EMSGSIZE;

	if (nla_put(skb, IPVS_DEST_ATTR_ADDR, sizeof(dest->addr), &dest->addr) ||
	    nla_put_be16(skb, IPVS_DEST_ATTR_PORT, dest->port) ||
	    nla_put_u32(skb, IPVS_DEST_ATTR_FWD_METHOD,
			(atomic_read(&dest->conn_flags) &
			 IP_VS_CONN_F_FWD_MASK)) ||
	    nla_put_u32(skb, IPVS_DEST_ATTR_WEIGHT,
			atomic_read(&dest->weight)) ||
	    nla_put_u8(skb, IPVS_DEST_ATTR_TUN_TYPE,
		       dest->tun_type) ||
	    nla_put_be16(skb, IPVS_DEST_ATTR_TUN_PORT,
			 dest->tun_port) ||
	    nla_put_u16(skb, IPVS_DEST_ATTR_TUN_FLAGS,
			dest->tun_flags) ||
	    nla_put_u32(skb, IPVS_DEST_ATTR_U_THRESH,
			READ_ONCE(dest->u_threshold)) ||
	    nla_put_u32(skb, IPVS_DEST_ATTR_L_THRESH,
			READ_ONCE(dest->l_threshold)) ||
	    nla_put_u32(skb, IPVS_DEST_ATTR_ACTIVE_CONNS,
			atomic_read(&dest->activeconns)) ||
	    nla_put_u32(skb, IPVS_DEST_ATTR_INACT_CONNS,
			ip_vs_dest_inactconns(dest)) ||
	    nla_put_u32(skb, IPVS_DEST_ATTR_PERSIST_CONNS,
			atomic_read(&dest->persistconns)) ||
	    nla_put_u16(skb, IPVS_DEST_ATTR_ADDR_FAMILY, dest->af))
		goto nla_put_failure;
	ip_vs_copy_stats(&kstats, &dest->stats);
	if (ip_vs_genl_fill_stats(skb, IPVS_DEST_ATTR_STATS, &kstats))
		goto nla_put_failure;
	if (ip_vs_genl_fill_stats64(skb, IPVS_DEST_ATTR_STATS64, &kstats))
		goto nla_put_failure;

	nla_nest_end(skb, nl_dest);

	return 0;

nla_put_failure:
	nla_nest_cancel(skb, nl_dest);
	return -EMSGSIZE;
}

static int ip_vs_genl_dump_dest(struct sk_buff *skb, struct ip_vs_dest *dest,
				struct netlink_callback *cb)
{
	void *hdr;

	hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
			  &ip_vs_genl_family, NLM_F_MULTI,
			  IPVS_CMD_NEW_DEST);
	if (!hdr)
		return -EMSGSIZE;

	if (ip_vs_genl_fill_dest(skb, dest) < 0)
		goto nla_put_failure;

	genlmsg_end(skb, hdr);
	return 0;

nla_put_failure:
	genlmsg_cancel(skb, hdr);
	return -EMSGSIZE;
}

static int ip_vs_genl_dump_dests(struct sk_buff *skb,
				 struct netlink_callback *cb)
{
	int idx = 0;
	int start = cb->args[0];
	struct ip_vs_service *svc;
	struct ip_vs_dest *dest;
	struct nlattr *attrs[IPVS_CMD_ATTR_MAX + 1];
	struct net *net = sock_net(skb->sk);
	struct netns_ipvs *ipvs = net_ipvs(net);

	rcu_read_lock();

	/* Try to find the service for which to dump destinations */
	if (nlmsg_parse_deprecated(cb->nlh, GENL_HDRLEN, attrs, IPVS_CMD_ATTR_MAX, ip_vs_cmd_policy, cb->extack))
		goto out_err;


	svc = ip_vs_genl_find_service(ipvs, attrs[IPVS_CMD_ATTR_SERVICE]);
	if (IS_ERR_OR_NULL(svc))
		goto out_err;

	/* Dump the destinations */
	list_for_each_entry_rcu(dest, &svc->destinations, n_list) {
		if (++idx <= start)
			continue;
		if (ip_vs_genl_dump_dest(skb, dest, cb) < 0) {
			idx--;
			goto nla_put_failure;
		}
	}

nla_put_failure:
	cb->args[0] = idx;

out_err:
	rcu_read_unlock();

	return skb->len;
}

static int ip_vs_genl_parse_dest(struct ip_vs_dest_user_kern *udest,
				 struct nlattr *nla, bool full_entry)
{
	struct nlattr *attrs[IPVS_DEST_ATTR_MAX + 1];
	struct nlattr *nla_addr, *nla_port;
	struct nlattr *nla_addr_family;

	/* Parse mandatory identifying destination fields first */
	if (nla == NULL ||
	    nla_parse_nested_deprecated(attrs, IPVS_DEST_ATTR_MAX, nla, ip_vs_dest_policy, NULL))
		return -EINVAL;

	nla_addr	= attrs[IPVS_DEST_ATTR_ADDR];
	nla_port	= attrs[IPVS_DEST_ATTR_PORT];
	nla_addr_family	= attrs[IPVS_DEST_ATTR_ADDR_FAMILY];

	if (!(nla_addr && nla_port))
		return -EINVAL;

	memset(udest, 0, sizeof(*udest));

	nla_memcpy(&udest->addr, nla_addr, sizeof(udest->addr));
	udest->port = nla_get_be16(nla_port);

	udest->af = nla_get_u16_default(nla_addr_family, 0);

	/* If a full entry was requested, check for the additional fields */
	if (full_entry) {
		struct nlattr *nla_fwd, *nla_weight, *nla_u_thresh,
			      *nla_l_thresh, *nla_tun_type, *nla_tun_port,
			      *nla_tun_flags;

		nla_fwd		= attrs[IPVS_DEST_ATTR_FWD_METHOD];
		nla_weight	= attrs[IPVS_DEST_ATTR_WEIGHT];
		nla_u_thresh	= attrs[IPVS_DEST_ATTR_U_THRESH];
		nla_l_thresh	= attrs[IPVS_DEST_ATTR_L_THRESH];
		nla_tun_type	= attrs[IPVS_DEST_ATTR_TUN_TYPE];
		nla_tun_port	= attrs[IPVS_DEST_ATTR_TUN_PORT];
		nla_tun_flags	= attrs[IPVS_DEST_ATTR_TUN_FLAGS];

		if (!(nla_fwd && nla_weight && nla_u_thresh && nla_l_thresh))
			return -EINVAL;

		udest->conn_flags = nla_get_u32(nla_fwd)
				    & IP_VS_CONN_F_FWD_MASK;
		udest->weight = nla_get_u32(nla_weight);
		udest->u_threshold = nla_get_u32(nla_u_thresh);
		udest->l_threshold = nla_get_u32(nla_l_thresh);

		if (nla_tun_type)
			udest->tun_type = nla_get_u8(nla_tun_type);

		if (nla_tun_port)
			udest->tun_port = nla_get_be16(nla_tun_port);

		if (nla_tun_flags)
			udest->tun_flags = nla_get_u16(nla_tun_flags);
	}

	return 0;
}

static int ip_vs_genl_fill_daemon(struct sk_buff *skb, __u32 state,
				  struct ipvs_sync_daemon_cfg *c)
{
	struct nlattr *nl_daemon;

	nl_daemon = nla_nest_start_noflag(skb, IPVS_CMD_ATTR_DAEMON);
	if (!nl_daemon)
		return -EMSGSIZE;

	if (nla_put_u32(skb, IPVS_DAEMON_ATTR_STATE, state) ||
	    nla_put_string(skb, IPVS_DAEMON_ATTR_MCAST_IFN, c->mcast_ifn) ||
	    nla_put_u32(skb, IPVS_DAEMON_ATTR_SYNC_ID, c->syncid) ||
	    nla_put_u16(skb, IPVS_DAEMON_ATTR_SYNC_MAXLEN, c->sync_maxlen) ||
	    nla_put_u16(skb, IPVS_DAEMON_ATTR_MCAST_PORT, c->mcast_port) ||
	    nla_put_u8(skb, IPVS_DAEMON_ATTR_MCAST_TTL, c->mcast_ttl))
		goto nla_put_failure;
#ifdef CONFIG_IP_VS_IPV6
	if (c->mcast_af == AF_INET6) {
		if (nla_put_in6_addr(skb, IPVS_DAEMON_ATTR_MCAST_GROUP6,
				     &c->mcast_group.in6))
			goto nla_put_failure;
	} else
#endif
		if (c->mcast_af == AF_INET &&
		    nla_put_in_addr(skb, IPVS_DAEMON_ATTR_MCAST_GROUP,
				    c->mcast_group.ip))
			goto nla_put_failure;
	nla_nest_end(skb, nl_daemon);

	return 0;

nla_put_failure:
	nla_nest_cancel(skb, nl_daemon);
	return -EMSGSIZE;
}

static int ip_vs_genl_dump_daemon(struct sk_buff *skb, __u32 state,
				  struct ipvs_sync_daemon_cfg *c,
				  struct netlink_callback *cb)
{
	void *hdr;
	hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
			  &ip_vs_genl_family, NLM_F_MULTI,
			  IPVS_CMD_NEW_DAEMON);
	if (!hdr)
		return -EMSGSIZE;

	if (ip_vs_genl_fill_daemon(skb, state, c))
		goto nla_put_failure;

	genlmsg_end(skb, hdr);
	return 0;

nla_put_failure:
	genlmsg_cancel(skb, hdr);
	return -EMSGSIZE;
}

static int ip_vs_genl_dump_daemons(struct sk_buff *skb,
				   struct netlink_callback *cb)
{
	struct net *net = sock_net(skb->sk);
	struct netns_ipvs *ipvs = net_ipvs(net);

	mutex_lock(&ipvs->sync_mutex);
	if ((ipvs->sync_state & IP_VS_STATE_MASTER) && !cb->args[0]) {
		if (ip_vs_genl_dump_daemon(skb, IP_VS_STATE_MASTER,
					   &ipvs->mcfg, cb) < 0)
			goto nla_put_failure;

		cb->args[0] = 1;
	}

	if ((ipvs->sync_state & IP_VS_STATE_BACKUP) && !cb->args[1]) {
		if (ip_vs_genl_dump_daemon(skb, IP_VS_STATE_BACKUP,
					   &ipvs->bcfg, cb) < 0)
			goto nla_put_failure;

		cb->args[1] = 1;
	}

nla_put_failure:
	mutex_unlock(&ipvs->sync_mutex);

	return skb->len;
}

static int ip_vs_genl_new_daemon(struct netns_ipvs *ipvs, struct nlattr **attrs)
{
	struct ipvs_sync_daemon_cfg c;
	struct nlattr *a;
	int ret;

	memset(&c, 0, sizeof(c));
	if (!(attrs[IPVS_DAEMON_ATTR_STATE] &&
	      attrs[IPVS_DAEMON_ATTR_MCAST_IFN] &&
	      attrs[IPVS_DAEMON_ATTR_SYNC_ID]))
		return -EINVAL;
	strscpy(c.mcast_ifn, nla_data(attrs[IPVS_DAEMON_ATTR_MCAST_IFN]),
		sizeof(c.mcast_ifn));
	c.syncid = nla_get_u32(attrs[IPVS_DAEMON_ATTR_SYNC_ID]);

	a = attrs[IPVS_DAEMON_ATTR_SYNC_MAXLEN];
	if (a)
		c.sync_maxlen = nla_get_u16(a);

	a = attrs[IPVS_DAEMON_ATTR_MCAST_GROUP];
	if (a) {
		c.mcast_af = AF_INET;
		c.mcast_group.ip = nla_get_in_addr(a);
		if (!ipv4_is_multicast(c.mcast_group.ip))
			return -EINVAL;
	} else {
		a = attrs[IPVS_DAEMON_ATTR_MCAST_GROUP6];
		if (a) {
#ifdef CONFIG_IP_VS_IPV6
			int addr_type;

			c.mcast_af = AF_INET6;
			c.mcast_group.in6 = nla_get_in6_addr(a);
			addr_type = ipv6_addr_type(&c.mcast_group.in6);
			if (!(addr_type & IPV6_ADDR_MULTICAST))
				return -EINVAL;
#else
			return -EAFNOSUPPORT;
#endif
		}
	}

	a = attrs[IPVS_DAEMON_ATTR_MCAST_PORT];
	if (a)
		c.mcast_port = nla_get_u16(a);

	a = attrs[IPVS_DAEMON_ATTR_MCAST_TTL];
	if (a)
		c.mcast_ttl = nla_get_u8(a);

	/* The synchronization protocol is incompatible with mixed family
	 * services
	 */
	if (ipvs->mixed_address_family_dests > 0)
		return -EINVAL;

	ret = start_sync_thread(ipvs, &c,
				nla_get_u32(attrs[IPVS_DAEMON_ATTR_STATE]));
	return ret;
}

static int ip_vs_genl_del_daemon(struct netns_ipvs *ipvs, struct nlattr **attrs)
{
	int ret;

	if (!attrs[IPVS_DAEMON_ATTR_STATE])
		return -EINVAL;

	ret = stop_sync_thread(ipvs,
			       nla_get_u32(attrs[IPVS_DAEMON_ATTR_STATE]));
	return ret;
}

static int ip_vs_genl_set_config(struct netns_ipvs *ipvs, struct nlattr **attrs)
{
	struct ip_vs_timeout_user t;

	__ip_vs_get_timeouts(ipvs, &t);

	if (attrs[IPVS_CMD_ATTR_TIMEOUT_TCP])
		t.tcp_timeout = nla_get_u32(attrs[IPVS_CMD_ATTR_TIMEOUT_TCP]);

	if (attrs[IPVS_CMD_ATTR_TIMEOUT_TCP_FIN])
		t.tcp_fin_timeout =
			nla_get_u32(attrs[IPVS_CMD_ATTR_TIMEOUT_TCP_FIN]);

	if (attrs[IPVS_CMD_ATTR_TIMEOUT_UDP])
		t.udp_timeout = nla_get_u32(attrs[IPVS_CMD_ATTR_TIMEOUT_UDP]);

	return ip_vs_set_timeout(ipvs, &t);
}

static int ip_vs_genl_set_daemon(struct sk_buff *skb, struct genl_info *info)
{
	int ret = -EINVAL, cmd;
	struct net *net = sock_net(skb->sk);
	struct netns_ipvs *ipvs = net_ipvs(net);

	cmd = info->genlhdr->cmd;

	if (cmd == IPVS_CMD_NEW_DAEMON || cmd == IPVS_CMD_DEL_DAEMON) {
		struct nlattr *daemon_attrs[IPVS_DAEMON_ATTR_MAX + 1];

		if (!info->attrs[IPVS_CMD_ATTR_DAEMON] ||
		    nla_parse_nested_deprecated(daemon_attrs, IPVS_DAEMON_ATTR_MAX, info->attrs[IPVS_CMD_ATTR_DAEMON], ip_vs_daemon_policy, info->extack))
			goto out;

		if (cmd == IPVS_CMD_NEW_DAEMON)
			ret = ip_vs_genl_new_daemon(ipvs, daemon_attrs);
		else
			ret = ip_vs_genl_del_daemon(ipvs, daemon_attrs);
	}

out:
	return ret;
}

static int ip_vs_genl_set_cmd(struct sk_buff *skb, struct genl_info *info)
{
	bool need_full_svc = false, need_full_dest = false;
	struct ip_vs_service *svc = NULL;
	struct ip_vs_service_user_kern usvc;
	struct ip_vs_dest_user_kern udest;
	int ret = 0, cmd;
	struct net *net = sock_net(skb->sk);
	struct netns_ipvs *ipvs = net_ipvs(net);

	cmd = info->genlhdr->cmd;

	mutex_lock(&ipvs->service_mutex);

	if (cmd == IPVS_CMD_FLUSH) {
		ret = ip_vs_flush(ipvs, false);
		goto out;
	} else if (cmd == IPVS_CMD_SET_CONFIG) {
		ret = ip_vs_genl_set_config(ipvs, info->attrs);
		goto out;
	} else if (cmd == IPVS_CMD_ZERO &&
		   !info->attrs[IPVS_CMD_ATTR_SERVICE]) {
		ret = ip_vs_zero_all(ipvs);
		goto out;
	}

	/* All following commands require a service argument, so check if we
	 * received a valid one. We need a full service specification when
	 * adding / editing a service. Only identifying members otherwise. */
	if (cmd == IPVS_CMD_NEW_SERVICE || cmd == IPVS_CMD_SET_SERVICE)
		need_full_svc = true;

	/* We use function that requires RCU lock (hlist_bl) */
	rcu_read_lock();
	ret = ip_vs_genl_parse_service(ipvs, &usvc,
				       info->attrs[IPVS_CMD_ATTR_SERVICE],
				       need_full_svc, &svc);
	rcu_read_unlock();
	if (ret)
		goto out;

	/* Unless we're adding a new service, the service must already exist */
	if ((cmd != IPVS_CMD_NEW_SERVICE) && (svc == NULL)) {
		ret = -ESRCH;
		goto out;
	}

	/* Destination commands require a valid destination argument. For
	 * adding / editing a destination, we need a full destination
	 * specification. */
	if (cmd == IPVS_CMD_NEW_DEST || cmd == IPVS_CMD_SET_DEST ||
	    cmd == IPVS_CMD_DEL_DEST) {
		if (cmd != IPVS_CMD_DEL_DEST)
			need_full_dest = true;

		ret = ip_vs_genl_parse_dest(&udest,
					    info->attrs[IPVS_CMD_ATTR_DEST],
					    need_full_dest);
		if (ret)
			goto out;

		/* Old protocols did not allow the user to specify address
		 * family, so we set it to zero instead.  We also didn't
		 * allow heterogeneous pools in the old code, so it's safe
		 * to assume that this will have the same address family as
		 * the service.
		 */
		if (udest.af == 0)
			udest.af = svc->af;

		if (!ip_vs_is_af_valid(udest.af)) {
			ret = -EAFNOSUPPORT;
			goto out;
		}

		if (udest.af != svc->af && cmd != IPVS_CMD_DEL_DEST) {
			/* The synchronization protocol is incompatible
			 * with mixed family services
			 */
			if (ipvs->sync_state) {
				ret = -EINVAL;
				goto out;
			}

			/* Which connection types do we support? */
			switch (udest.conn_flags) {
			case IP_VS_CONN_F_TUNNEL:
				/* We are able to forward this */
				break;
			default:
				ret = -EINVAL;
				goto out;
			}
		}
	}

	switch (cmd) {
	case IPVS_CMD_NEW_SERVICE:
		if (svc == NULL)
			ret = ip_vs_add_service(ipvs, &usvc, &svc);
		else
			ret = -EEXIST;
		break;
	case IPVS_CMD_SET_SERVICE:
		ret = ip_vs_edit_service(svc, &usvc);
		break;
	case IPVS_CMD_DEL_SERVICE:
		ret = ip_vs_del_service(svc);
		/* do not use svc, it can be freed */
		break;
	case IPVS_CMD_NEW_DEST:
		ret = ip_vs_add_dest(svc, &udest);
		break;
	case IPVS_CMD_SET_DEST:
		ret = ip_vs_edit_dest(svc, &udest);
		break;
	case IPVS_CMD_DEL_DEST:
		ret = ip_vs_del_dest(svc, &udest);
		break;
	case IPVS_CMD_ZERO:
		ret = ip_vs_zero_service(svc);
		break;
	default:
		ret = -EINVAL;
	}

out:
	mutex_unlock(&ipvs->service_mutex);

	return ret;
}

static int ip_vs_genl_get_cmd(struct sk_buff *skb, struct genl_info *info)
{
	struct sk_buff *msg;
	void *reply;
	int ret, cmd, reply_cmd;
	struct net *net = sock_net(skb->sk);
	struct netns_ipvs *ipvs = net_ipvs(net);

	cmd = info->genlhdr->cmd;

	if (cmd == IPVS_CMD_GET_SERVICE)
		reply_cmd = IPVS_CMD_NEW_SERVICE;
	else if (cmd == IPVS_CMD_GET_INFO)
		reply_cmd = IPVS_CMD_SET_INFO;
	else if (cmd == IPVS_CMD_GET_CONFIG)
		reply_cmd = IPVS_CMD_SET_CONFIG;
	else {
		pr_err("unknown Generic Netlink command\n");
		return -EINVAL;
	}

	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
	if (!msg)
		return -ENOMEM;

	rcu_read_lock();

	reply = genlmsg_put_reply(msg, info, &ip_vs_genl_family, 0, reply_cmd);
	if (reply == NULL)
		goto nla_put_failure;

	switch (cmd) {
	case IPVS_CMD_GET_SERVICE:
	{
		struct ip_vs_service *svc;

		svc = ip_vs_genl_find_service(ipvs,
					      info->attrs[IPVS_CMD_ATTR_SERVICE]);
		if (IS_ERR(svc)) {
			ret = PTR_ERR(svc);
			goto out_err;
		} else if (svc) {
			ret = ip_vs_genl_fill_service(msg, svc);
			if (ret)
				goto nla_put_failure;
		} else {
			ret = -ESRCH;
			goto out_err;
		}

		break;
	}

	case IPVS_CMD_GET_CONFIG:
	{
		struct ip_vs_timeout_user t;

		__ip_vs_get_timeouts(ipvs, &t);
#ifdef CONFIG_IP_VS_PROTO_TCP
		if (nla_put_u32(msg, IPVS_CMD_ATTR_TIMEOUT_TCP,
				t.tcp_timeout) ||
		    nla_put_u32(msg, IPVS_CMD_ATTR_TIMEOUT_TCP_FIN,
				t.tcp_fin_timeout))
			goto nla_put_failure;
#endif
#ifdef CONFIG_IP_VS_PROTO_UDP
		if (nla_put_u32(msg, IPVS_CMD_ATTR_TIMEOUT_UDP, t.udp_timeout))
			goto nla_put_failure;
#endif

		break;
	}

	case IPVS_CMD_GET_INFO:
		if (nla_put_u32(msg, IPVS_INFO_ATTR_VERSION,
				IP_VS_VERSION_CODE) ||
		    nla_put_u32(msg, IPVS_INFO_ATTR_CONN_TAB_SIZE,
				get_conn_tab_size(ipvs)))
			goto nla_put_failure;
		break;
	}

	genlmsg_end(msg, reply);
	ret = genlmsg_reply(msg, info);
	goto out;

nla_put_failure:
	pr_err("not enough space in Netlink message\n");
	ret = -EMSGSIZE;

out_err:
	nlmsg_free(msg);
out:
	rcu_read_unlock();

	return ret;
}


static const struct genl_small_ops ip_vs_genl_ops[] = {
	{
		.cmd	= IPVS_CMD_NEW_SERVICE,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_cmd,
	},
	{
		.cmd	= IPVS_CMD_SET_SERVICE,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_cmd,
	},
	{
		.cmd	= IPVS_CMD_DEL_SERVICE,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_cmd,
	},
	{
		.cmd	= IPVS_CMD_GET_SERVICE,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_get_cmd,
		.dumpit	= ip_vs_genl_dump_services,
	},
	{
		.cmd	= IPVS_CMD_NEW_DEST,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_cmd,
	},
	{
		.cmd	= IPVS_CMD_SET_DEST,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_cmd,
	},
	{
		.cmd	= IPVS_CMD_DEL_DEST,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_cmd,
	},
	{
		.cmd	= IPVS_CMD_GET_DEST,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.dumpit	= ip_vs_genl_dump_dests,
	},
	{
		.cmd	= IPVS_CMD_NEW_DAEMON,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_daemon,
	},
	{
		.cmd	= IPVS_CMD_DEL_DAEMON,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_daemon,
	},
	{
		.cmd	= IPVS_CMD_GET_DAEMON,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.dumpit	= ip_vs_genl_dump_daemons,
	},
	{
		.cmd	= IPVS_CMD_SET_CONFIG,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_cmd,
	},
	{
		.cmd	= IPVS_CMD_GET_CONFIG,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_get_cmd,
	},
	{
		.cmd	= IPVS_CMD_GET_INFO,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_get_cmd,
	},
	{
		.cmd	= IPVS_CMD_ZERO,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_cmd,
	},
	{
		.cmd	= IPVS_CMD_FLUSH,
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
		.flags	= GENL_ADMIN_PERM,
		.doit	= ip_vs_genl_set_cmd,
	},
};

static struct genl_family ip_vs_genl_family __ro_after_init = {
	.hdrsize	= 0,
	.name		= IPVS_GENL_NAME,
	.version	= IPVS_GENL_VERSION,
	.maxattr	= IPVS_CMD_ATTR_MAX,
	.policy = ip_vs_cmd_policy,
	.netnsok        = true,         /* Make ipvsadm to work on netns */
	.module		= THIS_MODULE,
	.small_ops	= ip_vs_genl_ops,
	.n_small_ops	= ARRAY_SIZE(ip_vs_genl_ops),
	.resv_start_op	= IPVS_CMD_FLUSH + 1,
	.parallel_ops	= 1,
};

static int __init ip_vs_genl_register(void)
{
	return genl_register_family(&ip_vs_genl_family);
}

static void ip_vs_genl_unregister(void)
{
	genl_unregister_family(&ip_vs_genl_family);
}

/* End of Generic Netlink interface definitions */

/*
 * per netns intit/exit func.
 */
#ifdef CONFIG_SYSCTL
static int __net_init ip_vs_control_net_init_sysctl(struct netns_ipvs *ipvs)
{
	struct net *net = ipvs->net;
	struct ctl_table *tbl;
	int idx, ret;
	size_t ctl_table_size = ARRAY_SIZE(vs_vars);
	bool unpriv = net->user_ns != &init_user_ns;

	atomic_set(&ipvs->dropentry, 0);
	spin_lock_init(&ipvs->dropentry_lock);
	spin_lock_init(&ipvs->droppacket_lock);
	spin_lock_init(&ipvs->securetcp_lock);
	INIT_DELAYED_WORK(&ipvs->defense_work, defense_work_handler);
	INIT_DELAYED_WORK(&ipvs->expire_nodest_conn_work,
			  expire_nodest_conn_handler);
	ipvs->est_stopped = 0;

	if (!net_eq(net, &init_net)) {
		tbl = kmemdup(vs_vars, sizeof(vs_vars), GFP_KERNEL);
		if (tbl == NULL)
			return -ENOMEM;
	} else
		tbl = vs_vars;
	/* Initialize sysctl defaults */
	for (idx = 0; idx < ARRAY_SIZE(vs_vars); idx++) {
		if (tbl[idx].proc_handler == proc_do_defense_mode)
			tbl[idx].extra2 = ipvs;
	}
	idx = 0;
	ipvs->sysctl_amemthresh = 1024;
	tbl[idx++].data = &ipvs->sysctl_amemthresh;
	ipvs->sysctl_am_droprate = 10;
	tbl[idx++].data = &ipvs->sysctl_am_droprate;

	/* Inherit both limits from init_net:conn_max */
	ipvs->conn_max_limit = net_eq(net, &init_net) ? IP_VS_CONN_MAX :
			       READ_ONCE(*(int *)vs_vars[idx].data);
	ipvs->sysctl_conn_max = ipvs->conn_max_limit;
	tbl[idx].extra2 = ipvs;
	tbl[idx++].data = &ipvs->sysctl_conn_max;

	tbl[idx++].data = &ipvs->sysctl_drop_entry;
	tbl[idx++].data = &ipvs->sysctl_drop_packet;
#ifdef CONFIG_IP_VS_NFCT
	tbl[idx++].data = &ipvs->sysctl_conntrack;
#endif
	tbl[idx++].data = &ipvs->sysctl_secure_tcp;
	ipvs->sysctl_snat_reroute = 1;
	tbl[idx++].data = &ipvs->sysctl_snat_reroute;
	ipvs->sysctl_sync_ver = 1;
	tbl[idx++].data = &ipvs->sysctl_sync_ver;
	ipvs->sysctl_sync_ports = 1;
	tbl[idx++].data = &ipvs->sysctl_sync_ports;
	tbl[idx++].data = &ipvs->sysctl_sync_persist_mode;

	ipvs->sysctl_sync_qlen_max = nr_free_buffer_pages() / 32;
	if (unpriv)
		tbl[idx].mode = 0444;
	tbl[idx++].data = &ipvs->sysctl_sync_qlen_max;

	ipvs->sysctl_sync_sock_size = 0;
	if (unpriv)
		tbl[idx].mode = 0444;
	tbl[idx++].data = &ipvs->sysctl_sync_sock_size;

	tbl[idx++].data = &ipvs->sysctl_cache_bypass;
	tbl[idx++].data = &ipvs->sysctl_expire_nodest_conn;
	tbl[idx++].data = &ipvs->sysctl_sloppy_tcp;
	tbl[idx++].data = &ipvs->sysctl_sloppy_sctp;
	tbl[idx++].data = &ipvs->sysctl_expire_quiescent_template;
	ipvs->sysctl_sync_threshold[0] = DEFAULT_SYNC_THRESHOLD;
	ipvs->sysctl_sync_threshold[1] = DEFAULT_SYNC_PERIOD;
	tbl[idx].data = &ipvs->sysctl_sync_threshold;
	tbl[idx].extra2 = ipvs;
	tbl[idx++].maxlen = sizeof(ipvs->sysctl_sync_threshold);
	ipvs->sysctl_sync_refresh_period = DEFAULT_SYNC_REFRESH_PERIOD;
	tbl[idx++].data = &ipvs->sysctl_sync_refresh_period;
	ipvs->sysctl_sync_retries = clamp_t(int, DEFAULT_SYNC_RETRIES, 0, 3);
	tbl[idx++].data = &ipvs->sysctl_sync_retries;
	tbl[idx++].data = &ipvs->sysctl_nat_icmp_send;
	ipvs->sysctl_pmtu_disc = 1;
	tbl[idx++].data = &ipvs->sysctl_pmtu_disc;
	tbl[idx++].data = &ipvs->sysctl_backup_only;
	ipvs->sysctl_conn_reuse_mode = 1;
	tbl[idx++].data = &ipvs->sysctl_conn_reuse_mode;
	tbl[idx++].data = &ipvs->sysctl_schedule_icmp;
	tbl[idx++].data = &ipvs->sysctl_ignore_tunneled;

	ipvs->sysctl_run_estimation = 1;
	if (unpriv)
		tbl[idx].mode = 0444;
	tbl[idx].extra2 = ipvs;
	tbl[idx++].data = &ipvs->sysctl_run_estimation;

	ipvs->est_cpulist_valid = 0;
	if (unpriv)
		tbl[idx].mode = 0444;
	tbl[idx].extra2 = ipvs;
	tbl[idx++].data = &ipvs->sysctl_est_cpulist;

	ipvs->sysctl_est_nice = IPVS_EST_NICE;
	if (unpriv)
		tbl[idx].mode = 0444;
	tbl[idx].extra2 = ipvs;
	tbl[idx++].data = &ipvs->sysctl_est_nice;

	if (unpriv)
		tbl[idx].mode = 0444;
	tbl[idx].extra2 = ipvs;
	tbl[idx++].data = &ipvs->sysctl_conn_lfactor;

	if (unpriv)
		tbl[idx].mode = 0444;
	tbl[idx].extra2 = ipvs;
	tbl[idx++].data = &ipvs->sysctl_svc_lfactor;

#ifdef CONFIG_IP_VS_DEBUG
	/* Global sysctls must be ro in non-init netns */
	if (!net_eq(net, &init_net))
		tbl[idx++].mode = 0444;
#endif

	ret = -ENOMEM;
	ipvs->sysctl_hdr = register_net_sysctl_sz(net, "net/ipv4/vs", tbl,
						  ctl_table_size);
	if (!ipvs->sysctl_hdr)
		goto err;
	ipvs->sysctl_tbl = tbl;

	ret = ip_vs_start_estimator(ipvs, &ipvs->tot_stats->s);
	if (ret < 0)
		goto err;

	/* Schedule defense work */
	queue_delayed_work(system_dfl_long_wq, &ipvs->defense_work,
			   DEFENSE_TIMER_PERIOD);

	return 0;

err:
	unregister_net_sysctl_table(ipvs->sysctl_hdr);
	if (!net_eq(net, &init_net))
		kfree(tbl);
	return ret;
}

static void __net_exit ip_vs_control_net_cleanup_sysctl(struct netns_ipvs *ipvs)
{
	struct net *net = ipvs->net;

	cancel_delayed_work_sync(&ipvs->expire_nodest_conn_work);
	cancel_delayed_work_sync(&ipvs->defense_work);
	cancel_work_sync(&ipvs->defense_work.work);
	unregister_net_sysctl_table(ipvs->sysctl_hdr);
	if (ipvs->tot_stats->s.est.ktid != -2) {
		/* Not stopped yet? This happens only on netns init error and
		 * we even do not need to lock the service_mutex for this case.
		 */
		mutex_lock(&ipvs->service_mutex);
		ip_vs_stop_estimator(ipvs, &ipvs->tot_stats->s);
		mutex_unlock(&ipvs->service_mutex);
	}

	if (ipvs->est_cpulist_valid)
		free_cpumask_var(ipvs->sysctl_est_cpulist);

	if (!net_eq(net, &init_net))
		kfree(ipvs->sysctl_tbl);
}

#else

static int __net_init ip_vs_control_net_init_sysctl(struct netns_ipvs *ipvs) { return 0; }
static void __net_exit ip_vs_control_net_cleanup_sysctl(struct netns_ipvs *ipvs) { }

#endif

static struct notifier_block ip_vs_dst_notifier = {
	.notifier_call = ip_vs_dst_event,
#ifdef CONFIG_IP_VS_IPV6
	.priority = ADDRCONF_NOTIFY_PRIORITY + 5,
#endif
};

int __net_init ip_vs_control_net_init(struct netns_ipvs *ipvs)
{
	int ret = -ENOMEM;
	int idx;

	/* Initialize service_mutex, svc_table per netns */
	__mutex_init(&ipvs->service_mutex, "ipvs->service_mutex", &__ipvs_service_key);
	init_rwsem(&ipvs->svc_resize_sem);
	init_rwsem(&ipvs->svc_replace_sem);
	INIT_DELAYED_WORK(&ipvs->svc_resize_work, svc_resize_work_handler);
	atomic_set(&ipvs->svc_table_changes, 0);
	RCU_INIT_POINTER(ipvs->svc_table, NULL);

	/* Initialize rs_table */
	for (idx = 0; idx < IP_VS_RTAB_SIZE; idx++)
		INIT_HLIST_HEAD(&ipvs->rs_table[idx]);

	INIT_LIST_HEAD(&ipvs->dest_trash);
	spin_lock_init(&ipvs->dest_trash_lock);
	timer_setup(&ipvs->dest_trash_timer, ip_vs_dest_trash_expire, 0);
	for (idx = 0; idx < IP_VS_AF_MAX; idx++) {
		atomic_set(&ipvs->num_services[idx], 0);
		atomic_set(&ipvs->fwm_services[idx], 0);
		atomic_set(&ipvs->nonfwm_services[idx], 0);
		atomic_set(&ipvs->ftpsvc_counter[idx], 0);
		atomic_set(&ipvs->nullsvc_counter[idx], 0);
		atomic_set(&ipvs->conn_out_counter[idx], 0);
	}

	INIT_DELAYED_WORK(&ipvs->est_reload_work, est_reload_work_handler);
	ipvs->sysctl_svc_lfactor = ip_vs_svc_default_load_factor(ipvs);

	/* procfs stats */
	ipvs->tot_stats = kzalloc_obj(*ipvs->tot_stats);
	if (!ipvs->tot_stats)
		goto out;
	if (ip_vs_stats_init_alloc(&ipvs->tot_stats->s) < 0)
		goto err_tot_stats;

#ifdef CONFIG_PROC_FS
	if (!proc_create_net("ip_vs", 0, ipvs->net->proc_net,
			     &ip_vs_info_seq_ops, sizeof(struct ip_vs_iter)))
		goto err_vs;
	if (!proc_create_net_single("ip_vs_stats", 0, ipvs->net->proc_net,
				    ip_vs_stats_show, NULL))
		goto err_stats;
	if (!proc_create_net_single("ip_vs_stats_percpu", 0,
				    ipvs->net->proc_net,
				    ip_vs_stats_percpu_show, NULL))
		goto err_percpu;
	if (!proc_create_net_single("ip_vs_status", 0440, ipvs->net->proc_net,
				    ip_vs_status_show, NULL))
		goto err_status;
#endif

	ret = ip_vs_control_net_init_sysctl(ipvs);
	if (ret < 0)
		goto err;

	return 0;

err:
#ifdef CONFIG_PROC_FS
	remove_proc_entry("ip_vs_status", ipvs->net->proc_net);

err_status:
	remove_proc_entry("ip_vs_stats_percpu", ipvs->net->proc_net);

err_percpu:
	remove_proc_entry("ip_vs_stats", ipvs->net->proc_net);

err_stats:
	remove_proc_entry("ip_vs", ipvs->net->proc_net);

err_vs:
#endif
	ip_vs_stats_release(&ipvs->tot_stats->s);

err_tot_stats:
	kfree(ipvs->tot_stats);

out:
	return ret;
}

void __net_exit ip_vs_control_net_cleanup(struct netns_ipvs *ipvs)
{
	ip_vs_trash_cleanup(ipvs);
	ip_vs_control_net_cleanup_sysctl(ipvs);
	cancel_delayed_work_sync(&ipvs->est_reload_work);
#ifdef CONFIG_PROC_FS
	remove_proc_entry("ip_vs_status", ipvs->net->proc_net);
	remove_proc_entry("ip_vs_stats_percpu", ipvs->net->proc_net);
	remove_proc_entry("ip_vs_stats", ipvs->net->proc_net);
	remove_proc_entry("ip_vs", ipvs->net->proc_net);
#endif
	call_rcu(&ipvs->tot_stats->rcu_head, ip_vs_stats_rcu_free);
}

int __init ip_vs_register_nl_ioctl(void)
{
	int ret;

	ret = nf_register_sockopt(&ip_vs_sockopts);
	if (ret) {
		pr_err("cannot register sockopt.\n");
		goto err_sock;
	}

	ret = ip_vs_genl_register();
	if (ret) {
		pr_err("cannot register Generic Netlink interface.\n");
		goto err_genl;
	}
	return 0;

err_genl:
	nf_unregister_sockopt(&ip_vs_sockopts);
err_sock:
	return ret;
}

void ip_vs_unregister_nl_ioctl(void)
{
	ip_vs_genl_unregister();
	nf_unregister_sockopt(&ip_vs_sockopts);
}

int __init ip_vs_control_init(void)
{
	int ret;

	ret = register_netdevice_notifier(&ip_vs_dst_notifier);
	if (ret < 0)
		return ret;

	return 0;
}


void ip_vs_control_cleanup(void)
{
	unregister_netdevice_notifier(&ip_vs_dst_notifier);
	/* relying on common rcu_barrier() in ip_vs_cleanup() */
}