kernel-pfroute: Make time that is waited for VIPs to appear configurable

[people/ms/strongswan.git] / src / libhydra / plugins / kernel_pfroute / kernel_pfroute_net.c

/*
 * Copyright (C) 2009-2012 Tobias Brunner
 * Hochschule fuer Technik Rapperswil
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License as published by the
 * Free Software Foundation; either version 2 of the License, or (at your
 * option) any later version.  See <http://www.fsf.org/copyleft/gpl.txt>.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
 * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
 * for more details.
 */

#include <sys/types.h>
#include <sys/socket.h>
#include <net/if.h>
#include <net/if_dl.h>
#include <ifaddrs.h>
#include <net/route.h>
#include <unistd.h>
#include <errno.h>

#include "kernel_pfroute_net.h"

#include <hydra.h>
#include <utils/debug.h>
#include <networking/host.h>
#include <networking/tun_device.h>
#include <threading/thread.h>
#include <threading/mutex.h>
#include <threading/condvar.h>
#include <threading/rwlock.h>
#include <collections/hashtable.h>
#include <collections/linked_list.h>
#include <processing/jobs/callback_job.h>

#ifndef HAVE_STRUCT_SOCKADDR_SA_LEN
#error Cannot compile this plugin on systems where 'struct sockaddr' has no sa_len member.
#endif

/** properly align sockaddrs */
#ifdef __APPLE__
/* Apple always uses 4 bytes */
#define SA_ALIGN 4
#else
/* while on other platforms like FreeBSD it depends on the architecture */
#define SA_ALIGN sizeof(long)
#endif
#define SA_LEN(len) ((len) > 0 ? (((len)+SA_ALIGN-1) & ~(SA_ALIGN-1)) : SA_ALIGN)

/** delay before firing roam events (ms) */
#define ROAM_DELAY 100

typedef struct addr_entry_t addr_entry_t;

/**
 * IP address in an inface_entry_t
 */
struct addr_entry_t {

	/** The ip address */
	host_t *ip;

	/** virtual IP managed by us */
	bool virtual;
};

/**
 * destroy a addr_entry_t object
 */
static void addr_entry_destroy(addr_entry_t *this)
{
	this->ip->destroy(this->ip);
	free(this);
}

typedef struct iface_entry_t iface_entry_t;

/**
 * A network interface on this system, containing addr_entry_t's
 */
struct iface_entry_t {

	/** interface index */
	int ifindex;

	/** name of the interface */
	char ifname[IFNAMSIZ];

	/** interface flags, as in netdevice(7) SIOCGIFFLAGS */
	u_int flags;

	/** list of addresses as host_t */
	linked_list_t *addrs;

	/** TRUE if usable by config */
	bool usable;
};

/**
 * destroy an interface entry
 */
static void iface_entry_destroy(iface_entry_t *this)
{
	this->addrs->destroy_function(this->addrs, (void*)addr_entry_destroy);
	free(this);
}

/**
 * check if an interface is up
 */
static inline bool iface_entry_up(iface_entry_t *iface)
{
	return (iface->flags & IFF_UP) == IFF_UP;
}

/**
 * check if an interface is up and usable
 */
static inline bool iface_entry_up_and_usable(iface_entry_t *iface)
{
	return iface->usable && iface_entry_up(iface);
}

typedef struct addr_map_entry_t addr_map_entry_t;

/**
 * Entry that maps an IP address to an interface entry
 */
struct addr_map_entry_t {
	/** The IP address */
	host_t *ip;

	/** The interface this address is installed on */
	iface_entry_t *iface;
};

/**
 * Hash a addr_map_entry_t object, all entries with the same IP address
 * are stored in the same bucket
 */
static u_int addr_map_entry_hash(addr_map_entry_t *this)
{
	return chunk_hash(this->ip->get_address(this->ip));
}

/**
 * Compare two addr_map_entry_t objects, two entries are equal if they are
 * installed on the same interface
 */
static bool addr_map_entry_equals(addr_map_entry_t *a, addr_map_entry_t *b)
{
	return a->iface->ifindex == b->iface->ifindex &&
		   a->ip->ip_equals(a->ip, b->ip);
}

/**
 * Used with get_match this finds an address entry if it is installed on
 * an up and usable interface
 */
static bool addr_map_entry_match_up_and_usable(addr_map_entry_t *a,
											   addr_map_entry_t *b)
{
	return iface_entry_up_and_usable(b->iface) &&
		   a->ip->ip_equals(a->ip, b->ip);
}

/**
 * Used with get_match this finds an address entry if it is installed on
 * any active local interface
 */
static bool addr_map_entry_match_up(addr_map_entry_t *a, addr_map_entry_t *b)
{
	return iface_entry_up(b->iface) && a->ip->ip_equals(a->ip, b->ip);
}

typedef struct private_kernel_pfroute_net_t private_kernel_pfroute_net_t;

/**
 * Private variables and functions of kernel_pfroute class.
 */
struct private_kernel_pfroute_net_t
{
	/**
	 * Public part of the kernel_pfroute_t object.
	 */
	kernel_pfroute_net_t public;

	/**
	 * lock to access lists and maps
	 */
	rwlock_t *lock;

	/**
	 * Cached list of interfaces and their addresses (iface_entry_t)
	 */
	linked_list_t *ifaces;

	/**
	 * Map for IP addresses to iface_entry_t objects (addr_map_entry_t)
	 */
	hashtable_t *addrs;

	/**
	 * List of tun devices we installed for virtual IPs
	 */
	linked_list_t *tuns;

	/**
	 * mutex to communicate exclusively with PF_KEY
	 */
	mutex_t *mutex;

	/**
	 * condvar to signal if PF_KEY query got a response
	 */
	condvar_t *condvar;

	/**
	 * pid to send PF_ROUTE messages with
	 */
	pid_t pid;

	/**
	 * PF_ROUTE socket to communicate with the kernel
	 */
	int socket;

	/**
	 * sequence number for messages sent to the kernel
	 */
	int seq;

	/**
	 * Sequence number a query is waiting for
	 */
	int waiting_seq;

	/**
	 * Allocated reply message from kernel
	 */
	struct rt_msghdr *reply;

	/**
	 * time of last roam event
	 */
	timeval_t last_roam;

	/**
	 * Time in ms to wait for IP addresses to appear/disappear
	 */
	int vip_wait;
};

/**
 * Add an address map entry
 */
static void addr_map_entry_add(private_kernel_pfroute_net_t *this,
							   addr_entry_t *addr, iface_entry_t *iface)
{
	addr_map_entry_t *entry;

	if (addr->virtual)
	{	/* don't map virtual IPs */
		return;
	}

	INIT(entry,
		.ip = addr->ip,
		.iface = iface,
	);
	entry = this->addrs->put(this->addrs, entry, entry);
	free(entry);
}

/**
 * Remove an address map entry (the argument order is a bit strange because
 * it is also used with linked_list_t.invoke_function)
 */
static void addr_map_entry_remove(addr_entry_t *addr, iface_entry_t *iface,
								  private_kernel_pfroute_net_t *this)
{
	addr_map_entry_t *entry, lookup = {
		.ip = addr->ip,
		.iface = iface,
	};

	if (addr->virtual)
	{	/* these are never mapped, but this check avoid problems if a virtual IP
		 * equals a regular one */
		return;
	}
	entry = this->addrs->remove(this->addrs, &lookup);
	free(entry);
}

/**
 * callback function that raises the delayed roam event
 */
static job_requeue_t roam_event(uintptr_t address)
{
	hydra->kernel_interface->roam(hydra->kernel_interface, address != 0);
	return JOB_REQUEUE_NONE;
}

/**
 * fire a roaming event. we delay it for a bit and fire only one event
 * for multiple calls. otherwise we would create too many events.
 */
static void fire_roam_event(private_kernel_pfroute_net_t *this, bool address)
{
	timeval_t now;
	job_t *job;

	time_monotonic(&now);
	if (timercmp(&now, &this->last_roam, >))
	{
		timeval_add_ms(&now, ROAM_DELAY);
		this->last_roam = now;

		job = (job_t*)callback_job_create((callback_job_cb_t)roam_event,
										  (void*)(uintptr_t)(address ? 1 : 0),
										  NULL, NULL);
		lib->scheduler->schedule_job_ms(lib->scheduler, job, ROAM_DELAY);
	}
}

/**
 * Data for enumerator over rtmsg sockaddrs
 */
typedef struct {
	/** implements enumerator */
	enumerator_t public;
	/** copy of attribute bitfield */
	int types;
	/** bytes remaining in buffer */
	int remaining;
	/** next sockaddr to enumerate */
	struct sockaddr *addr;
} rt_enumerator_t;

METHOD(enumerator_t, rt_enumerate, bool,
	rt_enumerator_t *this, int *xtype, struct sockaddr **addr)
{
	int i, type;

	if (this->remaining < sizeof(this->addr->sa_len) ||
		this->remaining < this->addr->sa_len)
	{
		return FALSE;
	}
	for (i = 0; i < RTAX_MAX; i++)
	{
		type = (1 << i);
		if (this->types & type)
		{
			this->types &= ~type;
			*addr = this->addr;
			*xtype = i;
			this->remaining -= SA_LEN(this->addr->sa_len);
			this->addr = (struct sockaddr*)((char*)this->addr +
											SA_LEN(this->addr->sa_len));
			return TRUE;
		}
	}
	return FALSE;
}

/**
 * Create a safe enumerator over sockaddrs in ifa/ifam/rt_msg
 */
static enumerator_t *create_rtmsg_enumerator(void *hdr, size_t hdrlen)
{
	struct rt_msghdr *rthdr = hdr;
	rt_enumerator_t *this;

	INIT(this,
		.public = {
			.enumerate = (void*)_rt_enumerate,
			.destroy = (void*)free,
		},
		.types = rthdr->rtm_addrs,
		.remaining = rthdr->rtm_msglen - hdrlen,
		.addr = hdr + hdrlen,
	);
	return &this->public;
}

/**
 * Process an RTM_*ADDR message from the kernel
 */
static void process_addr(private_kernel_pfroute_net_t *this,
						 struct ifa_msghdr *ifa)
{
	struct sockaddr *sockaddr;
	host_t *host = NULL;
	enumerator_t *ifaces, *addrs;
	iface_entry_t *iface;
	addr_entry_t *addr;
	bool found = FALSE, changed = FALSE, roam = FALSE;
	enumerator_t *enumerator;
	int type;

	enumerator = create_rtmsg_enumerator(ifa, sizeof(*ifa));
	while (enumerator->enumerate(enumerator, &type, &sockaddr))
	{
		if (type == RTAX_IFA)
		{
			host = host_create_from_sockaddr(sockaddr);
			break;
		}
	}
	enumerator->destroy(enumerator);

	if (!host)
	{
		return;
	}

	this->lock->write_lock(this->lock);
	ifaces = this->ifaces->create_enumerator(this->ifaces);
	while (ifaces->enumerate(ifaces, &iface))
	{
		if (iface->ifindex == ifa->ifam_index)
		{
			addrs = iface->addrs->create_enumerator(iface->addrs);
			while (addrs->enumerate(addrs, &addr))
			{
				if (host->ip_equals(host, addr->ip))
				{
					found = TRUE;
					if (ifa->ifam_type == RTM_DELADDR)
					{
						iface->addrs->remove_at(iface->addrs, addrs);
						if (!addr->virtual && iface->usable)
						{
							changed = TRUE;
							DBG1(DBG_KNL, "%H disappeared from %s",
								 host, iface->ifname);
						}
						addr_map_entry_remove(addr, iface, this);
						addr_entry_destroy(addr);
					}
				}
			}
			addrs->destroy(addrs);

			if (!found && ifa->ifam_type == RTM_NEWADDR)
			{
				INIT(addr,
					.ip = host->clone(host),
				);
				changed = TRUE;
				iface->addrs->insert_last(iface->addrs, addr);
				addr_map_entry_add(this, addr, iface);
				if (iface->usable)
				{
					DBG1(DBG_KNL, "%H appeared on %s", host, iface->ifname);
				}
			}

			if (changed && iface_entry_up_and_usable(iface))
			{
				roam = TRUE;
			}
			break;
		}
	}
	ifaces->destroy(ifaces);
	this->lock->unlock(this->lock);
	host->destroy(host);

	if (roam)
	{
		fire_roam_event(this, TRUE);
	}
}

/**
 * Re-initialize address list of an interface if it changes state
 */
static void repopulate_iface(private_kernel_pfroute_net_t *this,
							 iface_entry_t *iface)
{
	struct ifaddrs *ifap, *ifa;
	addr_entry_t *addr;

	while (iface->addrs->remove_last(iface->addrs, (void**)&addr) == SUCCESS)
	{
		addr_map_entry_remove(addr, iface, this);
		addr_entry_destroy(addr);
	}

	if (getifaddrs(&ifap) == 0)
	{
		for (ifa = ifap; ifa != NULL; ifa = ifa->ifa_next)
		{
			if (ifa->ifa_addr && streq(ifa->ifa_name, iface->ifname))
			{
				switch (ifa->ifa_addr->sa_family)
				{
					case AF_INET:
					case AF_INET6:
						INIT(addr,
							.ip = host_create_from_sockaddr(ifa->ifa_addr),
						);
						iface->addrs->insert_last(iface->addrs, addr);
						addr_map_entry_add(this, addr, iface);
						break;
					default:
						break;
				}
			}
		}
		freeifaddrs(ifap);
	}
}

/**
 * Process an RTM_IFINFO message from the kernel
 */
static void process_link(private_kernel_pfroute_net_t *this,
						 struct if_msghdr *msg)
{
	enumerator_t *enumerator;
	iface_entry_t *iface;
	bool roam = FALSE, found = FALSE;

	this->lock->write_lock(this->lock);
	enumerator = this->ifaces->create_enumerator(this->ifaces);
	while (enumerator->enumerate(enumerator, &iface))
	{
		if (iface->ifindex == msg->ifm_index)
		{
			if (iface->usable)
			{
				if (!(iface->flags & IFF_UP) && (msg->ifm_flags & IFF_UP))
				{
					roam = TRUE;
					DBG1(DBG_KNL, "interface %s activated", iface->ifname);
				}
				else if ((iface->flags & IFF_UP) && !(msg->ifm_flags & IFF_UP))
				{
					roam = TRUE;
					DBG1(DBG_KNL, "interface %s deactivated", iface->ifname);
				}
			}
			iface->flags = msg->ifm_flags;
			repopulate_iface(this, iface);
			found = TRUE;
			break;
		}
	}
	enumerator->destroy(enumerator);

	if (!found)
	{
		INIT(iface,
			.ifindex = msg->ifm_index,
			.flags = msg->ifm_flags,
			.addrs = linked_list_create(),
		);
		if (if_indextoname(iface->ifindex, iface->ifname))
		{
			DBG1(DBG_KNL, "interface %s appeared", iface->ifname);
			iface->usable = hydra->kernel_interface->is_interface_usable(
										hydra->kernel_interface, iface->ifname);
			repopulate_iface(this, iface);
			this->ifaces->insert_last(this->ifaces, iface);
		}
		else
		{
			free(iface);
		}
	}
	this->lock->unlock(this->lock);

	if (roam)
	{
		fire_roam_event(this, TRUE);
	}
}

/**
 * Process an RTM_*ROUTE message from the kernel
 */
static void process_route(private_kernel_pfroute_net_t *this,
						  struct rt_msghdr *msg)
{

}

/**
 * Receives PF_ROUTE messages from kernel
 */
static job_requeue_t receive_events(private_kernel_pfroute_net_t *this)
{
	struct {
		union {
			struct rt_msghdr rtm;
			struct if_msghdr ifm;
			struct ifa_msghdr ifam;
		};
		char buf[sizeof(struct sockaddr_storage) * RTAX_MAX];
	} msg;
	int len, hdrlen;
	bool oldstate;

	oldstate = thread_cancelability(TRUE);
	len = recv(this->socket, &msg, sizeof(msg), 0);
	thread_cancelability(oldstate);

	if (len < 0)
	{
		switch (errno)
		{
			case EINTR:
			case EAGAIN:
				return JOB_REQUEUE_DIRECT;
			default:
				DBG1(DBG_KNL, "unable to receive from PF_ROUTE event socket");
				sleep(1);
				return JOB_REQUEUE_FAIR;
		}
	}

	if (len < offsetof(struct rt_msghdr, rtm_flags) || len < msg.rtm.rtm_msglen)
	{
		DBG1(DBG_KNL, "received invalid PF_ROUTE message");
		return JOB_REQUEUE_DIRECT;
	}
	if (msg.rtm.rtm_version != RTM_VERSION)
	{
		DBG1(DBG_KNL, "received PF_ROUTE message with unsupported version: %d",
			 msg.rtm.rtm_version);
		return JOB_REQUEUE_DIRECT;
	}
	switch (msg.rtm.rtm_type)
	{
		case RTM_NEWADDR:
		case RTM_DELADDR:
			hdrlen = sizeof(msg.ifam);
			break;
		case RTM_IFINFO:
			hdrlen = sizeof(msg.ifm);
			break;
		case RTM_ADD:
		case RTM_DELETE:
		case RTM_GET:
			hdrlen = sizeof(msg.rtm);
			break;
		default:
			return JOB_REQUEUE_DIRECT;
	}
	if (msg.rtm.rtm_msglen < hdrlen)
	{
		DBG1(DBG_KNL, "ignoring short PF_ROUTE message");
		return JOB_REQUEUE_DIRECT;
	}
	switch (msg.rtm.rtm_type)
	{
		case RTM_NEWADDR:
		case RTM_DELADDR:
			process_addr(this, &msg.ifam);
			break;
		case RTM_IFINFO:
			process_link(this, &msg.ifm);
			break;
		case RTM_ADD:
		case RTM_DELETE:
			process_route(this, &msg.rtm);
			break;
		default:
			break;
	}

	this->mutex->lock(this->mutex);
	if (msg.rtm.rtm_pid == this->pid && msg.rtm.rtm_seq == this->waiting_seq)
	{
		/* seems like the message someone is waiting for, deliver */
		this->reply = realloc(this->reply, msg.rtm.rtm_msglen);
		memcpy(this->reply, &msg, msg.rtm.rtm_msglen);
	}
	/* signal on any event, add_ip()/del_ip() might wait for it */
	this->condvar->broadcast(this->condvar);
	this->mutex->unlock(this->mutex);

	return JOB_REQUEUE_DIRECT;
}


/** enumerator over addresses */
typedef struct {
	private_kernel_pfroute_net_t* this;
	/** which addresses to enumerate */
	kernel_address_type_t which;
} address_enumerator_t;

/**
 * cleanup function for address enumerator
 */
static void address_enumerator_destroy(address_enumerator_t *data)
{
	data->this->lock->unlock(data->this->lock);
	free(data);
}

/**
 * filter for addresses
 */
static bool filter_addresses(address_enumerator_t *data,
							 addr_entry_t** in, host_t** out)
{
	host_t *ip;
	if (!(data->which & ADDR_TYPE_VIRTUAL) && (*in)->virtual)
	{   /* skip virtual interfaces added by us */
		return FALSE;
	}
	if (!(data->which & ADDR_TYPE_REGULAR) && !(*in)->virtual)
	{	/* address is regular, but not requested */
		return FALSE;
	}
	ip = (*in)->ip;
	if (ip->get_family(ip) == AF_INET6)
	{
		struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)ip->get_sockaddr(ip);
		if (IN6_IS_ADDR_LINKLOCAL(&sin6->sin6_addr))
		{   /* skip addresses with a unusable scope */
			return FALSE;
		}
	}
	*out = ip;
	return TRUE;
}

/**
 * enumerator constructor for interfaces
 */
static enumerator_t *create_iface_enumerator(iface_entry_t *iface,
											 address_enumerator_t *data)
{
	return enumerator_create_filter(iface->addrs->create_enumerator(iface->addrs),
									(void*)filter_addresses, data, NULL);
}

/**
 * filter for interfaces
 */
static bool filter_interfaces(address_enumerator_t *data, iface_entry_t** in,
							  iface_entry_t** out)
{
	if (!(data->which & ADDR_TYPE_IGNORED) && !(*in)->usable)
	{	/* skip interfaces excluded by config */
		return FALSE;
	}
	if (!(data->which & ADDR_TYPE_LOOPBACK) && ((*in)->flags & IFF_LOOPBACK))
	{	/* ignore loopback devices */
		return FALSE;
	}
	if (!(data->which & ADDR_TYPE_DOWN) && !((*in)->flags & IFF_UP))
	{	/* skip interfaces not up */
		return FALSE;
	}
	*out = *in;
	return TRUE;
}

METHOD(kernel_net_t, create_address_enumerator, enumerator_t*,
	private_kernel_pfroute_net_t *this, kernel_address_type_t which)
{
	address_enumerator_t *data;

	INIT(data,
		.this = this,
		.which = which,
	);

	this->lock->read_lock(this->lock);
	return enumerator_create_nested(
				enumerator_create_filter(
					this->ifaces->create_enumerator(this->ifaces),
					(void*)filter_interfaces, data, NULL),
				(void*)create_iface_enumerator, data,
				(void*)address_enumerator_destroy);
}

METHOD(kernel_net_t, get_features, kernel_feature_t,
	private_kernel_pfroute_net_t *this)
{
	return KERNEL_REQUIRE_EXCLUDE_ROUTE;
}

METHOD(kernel_net_t, get_interface_name, bool,
	private_kernel_pfroute_net_t *this, host_t* ip, char **name)
{
	addr_map_entry_t *entry, lookup = {
		.ip = ip,
	};

	if (ip->is_anyaddr(ip))
	{
		return FALSE;
	}
	this->lock->read_lock(this->lock);
	/* first try to find it on an up and usable interface */
	entry = this->addrs->get_match(this->addrs, &lookup,
								  (void*)addr_map_entry_match_up_and_usable);
	if (entry)
	{
		if (name)
		{
			*name = strdup(entry->iface->ifname);
			DBG2(DBG_KNL, "%H is on interface %s", ip, *name);
		}
		this->lock->unlock(this->lock);
		return TRUE;
	}
	/* maybe it is installed on an ignored interface */
	entry = this->addrs->get_match(this->addrs, &lookup,
								  (void*)addr_map_entry_match_up);
	if (!entry)
	{	/* the address does not exist, is on a down interface */
		DBG2(DBG_KNL, "%H is not a local address or the interface is down", ip);
	}
	this->lock->unlock(this->lock);
	return FALSE;
}

METHOD(kernel_net_t, add_ip, status_t,
	private_kernel_pfroute_net_t *this, host_t *vip, int prefix,
	char *ifname)
{
	enumerator_t *ifaces, *addrs;
	iface_entry_t *iface;
	addr_entry_t *addr;
	tun_device_t *tun;
	bool timeout = FALSE;

	tun = tun_device_create(NULL);
	if (!tun)
	{
		return FAILED;
	}
	if (prefix == -1)
	{
		prefix = vip->get_address(vip).len * 8;
	}
	if (!tun->up(tun) || !tun->set_address(tun, vip, prefix))
	{
		tun->destroy(tun);
		return FAILED;
	}

	/* wait until address appears */
	this->mutex->lock(this->mutex);
	while (!timeout && !get_interface_name(this, vip, NULL))
	{
		timeout = this->condvar->timed_wait(this->condvar, this->mutex,
											this->vip_wait);
	}
	this->mutex->unlock(this->mutex);
	if (timeout)
	{
		DBG1(DBG_KNL, "virtual IP %H did not appear on %s",
			 vip, tun->get_name(tun));
		tun->destroy(tun);
		return FAILED;
	}

	this->lock->write_lock(this->lock);
	this->tuns->insert_last(this->tuns, tun);

	ifaces = this->ifaces->create_enumerator(this->ifaces);
	while (ifaces->enumerate(ifaces, &iface))
	{
		if (streq(iface->ifname, tun->get_name(tun)))
		{
			addrs = iface->addrs->create_enumerator(iface->addrs);
			while (addrs->enumerate(addrs, &addr))
			{
				if (addr->ip->ip_equals(addr->ip, vip))
				{
					addr->virtual = TRUE;
				}
			}
			addrs->destroy(addrs);
		}
	}
	ifaces->destroy(ifaces);
	/* lets do this while holding the lock, thus preventing another thread
	 * from deleting the TUN device concurrently, hopefully listeneres are quick
	 * and cause no deadlocks */
	hydra->kernel_interface->tun(hydra->kernel_interface, tun, TRUE);
	this->lock->unlock(this->lock);

	return SUCCESS;
}

METHOD(kernel_net_t, del_ip, status_t,
	private_kernel_pfroute_net_t *this, host_t *vip, int prefix,
	bool wait)
{
	enumerator_t *enumerator;
	tun_device_t *tun;
	host_t *addr;
	bool timeout = FALSE, found = FALSE;

	this->lock->write_lock(this->lock);
	enumerator = this->tuns->create_enumerator(this->tuns);
	while (enumerator->enumerate(enumerator, &tun))
	{
		addr = tun->get_address(tun, NULL);
		if (addr && addr->ip_equals(addr, vip))
		{
			this->tuns->remove_at(this->tuns, enumerator);
			hydra->kernel_interface->tun(hydra->kernel_interface, tun,
										 FALSE);
			tun->destroy(tun);
			found = TRUE;
			break;
		}
	}
	enumerator->destroy(enumerator);
	this->lock->unlock(this->lock);

	if (!found)
	{
		return NOT_FOUND;
	}
	/* wait until address disappears */
	if (wait)
	{
		this->mutex->lock(this->mutex);
		while (!timeout && get_interface_name(this, vip, NULL))
		{
			timeout = this->condvar->timed_wait(this->condvar, this->mutex,
												this->vip_wait);
		}
		this->mutex->unlock(this->mutex);
		if (timeout)
		{
			DBG1(DBG_KNL, "virtual IP %H did not disappear from tun", vip);
			return FAILED;
		}
	}
	return SUCCESS;
}

/**
 * Append a sockaddr_in/in6 of given type to routing message
 */
static void add_rt_addr(struct rt_msghdr *hdr, int type, host_t *addr)
{
	if (addr)
	{
		int len;

		len = *addr->get_sockaddr_len(addr);
		memcpy((char*)hdr + hdr->rtm_msglen, addr->get_sockaddr(addr), len);
		hdr->rtm_msglen += SA_LEN(len);
		hdr->rtm_addrs |= type;
	}
}

/**
 * Append a subnet mask sockaddr using the given prefix to routing message
 */
static void add_rt_mask(struct rt_msghdr *hdr, int type, int family, int prefix)
{
	host_t *mask;

	mask = host_create_netmask(family, prefix);
	if (mask)
	{
		add_rt_addr(hdr, type, mask);
		mask->destroy(mask);
	}
}

/**
 * Append an interface name sockaddr_dl to routing message
 */
static void add_rt_ifname(struct rt_msghdr *hdr, int type, char *name)
{
	struct sockaddr_dl sdl = {
		.sdl_len = sizeof(struct sockaddr_dl),
		.sdl_family = AF_LINK,
		.sdl_nlen = strlen(name),
	};

	if (strlen(name) <= sizeof(sdl.sdl_data))
	{
		memcpy(sdl.sdl_data, name, sdl.sdl_nlen);
		memcpy((char*)hdr + hdr->rtm_msglen, &sdl, sdl.sdl_len);
		hdr->rtm_msglen += SA_LEN(sdl.sdl_len);
		hdr->rtm_addrs |= type;
	}
}

/**
 * Add or remove a route
 */
static status_t manage_route(private_kernel_pfroute_net_t *this, int op,
							 chunk_t dst_net, u_int8_t prefixlen,
							 host_t *gateway, char *if_name)
{
	struct {
		struct rt_msghdr hdr;
		char buf[sizeof(struct sockaddr_storage) * RTAX_MAX];
	} msg = {
		.hdr = {
			.rtm_version = RTM_VERSION,
			.rtm_type = op,
			.rtm_flags = RTF_UP | RTF_STATIC,
			.rtm_pid = this->pid,
			.rtm_seq = ++this->seq,
		},
	};
	host_t *dst;
	int type;

	if (prefixlen == 0 && dst_net.len)
	{
		status_t status;
		chunk_t half;

		half = chunk_clonea(dst_net);
		half.ptr[0] |= 0x80;
		prefixlen = 1;
		status = manage_route(this, op, half, prefixlen, gateway, if_name);
		if (status != SUCCESS)
		{
			return status;
		}
	}

	dst = host_create_from_chunk(AF_UNSPEC, dst_net, 0);
	if (!dst)
	{
		return FAILED;
	}

	if ((dst->get_family(dst) == AF_INET && prefixlen == 32) ||
		(dst->get_family(dst) == AF_INET6 && prefixlen == 128))
	{
		msg.hdr.rtm_flags |= RTF_HOST | RTF_GATEWAY;
	}

	msg.hdr.rtm_msglen = sizeof(struct rt_msghdr);
	for (type = 0; type < RTAX_MAX; type++)
	{
		switch (type)
		{
			case RTAX_DST:
				add_rt_addr(&msg.hdr, RTA_DST, dst);
				break;
			case RTAX_NETMASK:
				if (!(msg.hdr.rtm_flags & RTF_HOST))
				{
					add_rt_mask(&msg.hdr, RTA_NETMASK,
								dst->get_family(dst), prefixlen);
				}
				break;
			case RTAX_IFP:
				if (if_name)
				{
					add_rt_ifname(&msg.hdr, RTA_IFP, if_name);
				}
				break;
			case RTAX_GATEWAY:
				if (gateway)
				{
					add_rt_addr(&msg.hdr, RTA_GATEWAY, gateway);
				}
				break;
			default:
				break;
		}
	}
	dst->destroy(dst);

	if (send(this->socket, &msg, msg.hdr.rtm_msglen, 0) != msg.hdr.rtm_msglen)
	{
		DBG1(DBG_KNL, "%s PF_ROUTE route failed: %s",
			 op == RTM_ADD ? "adding" : "deleting", strerror(errno));
		return FAILED;
	}
	return SUCCESS;
}

METHOD(kernel_net_t, add_route, status_t,
	private_kernel_pfroute_net_t *this, chunk_t dst_net, u_int8_t prefixlen,
	host_t *gateway, host_t *src_ip, char *if_name)
{
	return manage_route(this, RTM_ADD, dst_net, prefixlen, gateway, if_name);
}

METHOD(kernel_net_t, del_route, status_t,
	private_kernel_pfroute_net_t *this, chunk_t dst_net, u_int8_t prefixlen,
	host_t *gateway, host_t *src_ip, char *if_name)
{
	return manage_route(this, RTM_DELETE, dst_net, prefixlen, gateway, if_name);
}

/**
 * Do a route lookup for dest and return either the nexthop or the source
 * address.
 */
static host_t *get_route(private_kernel_pfroute_net_t *this, bool nexthop,
						 host_t *dest, host_t *src)
{
	struct {
		struct rt_msghdr hdr;
		char buf[sizeof(struct sockaddr_storage) * RTAX_MAX];
	} msg = {
		.hdr = {
			.rtm_version = RTM_VERSION,
			.rtm_type = RTM_GET,
			.rtm_pid = this->pid,
			.rtm_seq = ++this->seq,
		},
	};
	host_t *host = NULL;
	enumerator_t *enumerator;
	struct sockaddr *addr;
	bool failed = FALSE;
	int type;

retry:
	msg.hdr.rtm_msglen = sizeof(struct rt_msghdr);
	for (type = 0; type < RTAX_MAX; type++)
	{
		switch (type)
		{
			case RTAX_DST:
				add_rt_addr(&msg.hdr, RTA_DST, dest);
				break;
			case RTAX_IFA:
				add_rt_addr(&msg.hdr, RTA_IFA, src);
				break;
			case RTAX_IFP:
				if (!nexthop)
				{	/* add an empty IFP to ensure we get a source address */
					add_rt_ifname(&msg.hdr, RTA_IFP, "");
				}
				break;
			default:
				break;
		}
	}
	this->mutex->lock(this->mutex);

	while (this->waiting_seq)
	{
		this->condvar->wait(this->condvar, this->mutex);
	}
	this->waiting_seq = msg.hdr.rtm_seq;
	if (send(this->socket, &msg, msg.hdr.rtm_msglen, 0) == msg.hdr.rtm_msglen)
	{
		while (TRUE)
		{
			if (this->condvar->timed_wait(this->condvar, this->mutex, 1000))
			{	/* timed out? */
				break;
			}
			if (this->reply->rtm_msglen < sizeof(*this->reply) ||
				msg.hdr.rtm_seq != this->reply->rtm_seq)
			{
				continue;
			}
			enumerator = create_rtmsg_enumerator(this->reply,
												 sizeof(*this->reply));
			while (enumerator->enumerate(enumerator, &type, &addr))
			{
				if (nexthop)
				{
					if (type == RTAX_DST && this->reply->rtm_flags & RTF_HOST)
					{	/* probably a cloned/cached direct route, only use that
						 * as fallback if no gateway is found */
						host = host ?: host_create_from_sockaddr(addr);
					}
					if (type == RTAX_GATEWAY)
					{	/* could actually be a MAC address */
						host_t *gtw = host_create_from_sockaddr(addr);
						if (gtw)
						{
							DESTROY_IF(host);
							host = gtw;
						}
					}
				}
				else
				{
					if (type == RTAX_IFA)
					{
						host = host_create_from_sockaddr(addr);
					}
				}
			}
			enumerator->destroy(enumerator);
			break;
		}
	}
	else
	{
		failed = TRUE;
	}
	/* signal completion of query to a waiting thread */
	this->waiting_seq = 0;
	this->condvar->signal(this->condvar);
	this->mutex->unlock(this->mutex);

	if (failed)
	{
		if (src)
		{	/* the given source address might be gone, try again without */
			src = NULL;
			msg.hdr.rtm_seq = ++this->seq;
			msg.hdr.rtm_addrs = 0;
			memset(msg.buf, sizeof(msg.buf), 0);
			goto retry;
		}
		DBG1(DBG_KNL, "PF_ROUTE lookup failed: %s", strerror(errno));
	}

	if (host)
	{
		DBG2(DBG_KNL, "using %H as %s to reach %H", host,
			 nexthop ? "nexthop" : "address", dest);
	}
	return host;
}

METHOD(kernel_net_t, get_source_addr, host_t*,
	private_kernel_pfroute_net_t *this, host_t *dest, host_t *src)
{
	return get_route(this, FALSE, dest, src);
}

METHOD(kernel_net_t, get_nexthop, host_t*,
	private_kernel_pfroute_net_t *this, host_t *dest, host_t *src)
{
	return get_route(this, TRUE, dest, src);
}

/**
 * Initialize a list of local addresses.
 */
static status_t init_address_list(private_kernel_pfroute_net_t *this)
{
	struct ifaddrs *ifap, *ifa;
	iface_entry_t *iface, *current;
	addr_entry_t *addr;
	enumerator_t *ifaces, *addrs;

	DBG2(DBG_KNL, "known interfaces and IP addresses:");

	if (getifaddrs(&ifap) < 0)
	{
		DBG1(DBG_KNL, "  failed to get interfaces!");
		return FAILED;
	}

	for (ifa = ifap; ifa != NULL; ifa = ifa->ifa_next)
	{
		if (ifa->ifa_addr == NULL)
		{
			continue;
		}
		switch(ifa->ifa_addr->sa_family)
		{
			case AF_LINK:
			case AF_INET:
			case AF_INET6:
			{
				iface = NULL;
				ifaces = this->ifaces->create_enumerator(this->ifaces);
				while (ifaces->enumerate(ifaces, &current))
				{
					if (streq(current->ifname, ifa->ifa_name))
					{
						iface = current;
						break;
					}
				}
				ifaces->destroy(ifaces);

				if (!iface)
				{
					INIT(iface,
						.ifindex = if_nametoindex(ifa->ifa_name),
						.flags = ifa->ifa_flags,
						.addrs = linked_list_create(),
						.usable = hydra->kernel_interface->is_interface_usable(
										hydra->kernel_interface, ifa->ifa_name),
					);
					memcpy(iface->ifname, ifa->ifa_name, IFNAMSIZ);
					this->ifaces->insert_last(this->ifaces, iface);
				}

				if (ifa->ifa_addr->sa_family != AF_LINK)
				{
					INIT(addr,
						.ip = host_create_from_sockaddr(ifa->ifa_addr),
					);
					iface->addrs->insert_last(iface->addrs, addr);
					addr_map_entry_add(this, addr, iface);
				}
			}
		}
	}
	freeifaddrs(ifap);

	ifaces = this->ifaces->create_enumerator(this->ifaces);
	while (ifaces->enumerate(ifaces, &iface))
	{
		if (iface->usable && iface->flags & IFF_UP)
		{
			DBG2(DBG_KNL, "  %s", iface->ifname);
			addrs = iface->addrs->create_enumerator(iface->addrs);
			while (addrs->enumerate(addrs, (void**)&addr))
			{
				DBG2(DBG_KNL, "    %H", addr->ip);
			}
			addrs->destroy(addrs);
		}
	}
	ifaces->destroy(ifaces);

	return SUCCESS;
}

METHOD(kernel_net_t, destroy, void,
	private_kernel_pfroute_net_t *this)
{
	enumerator_t *enumerator;
	addr_entry_t *addr;

	if (this->socket != -1)
	{
		close(this->socket);
	}
	enumerator = this->addrs->create_enumerator(this->addrs);
	while (enumerator->enumerate(enumerator, NULL, (void**)&addr))
	{
		free(addr);
	}
	enumerator->destroy(enumerator);
	this->addrs->destroy(this->addrs);
	this->ifaces->destroy_function(this->ifaces, (void*)iface_entry_destroy);
	this->tuns->destroy(this->tuns);
	this->lock->destroy(this->lock);
	this->mutex->destroy(this->mutex);
	this->condvar->destroy(this->condvar);
	free(this->reply);
	free(this);
}

/*
 * Described in header.
 */
kernel_pfroute_net_t *kernel_pfroute_net_create()
{
	private_kernel_pfroute_net_t *this;

	INIT(this,
		.public = {
			.interface = {
				.get_features = _get_features,
				.get_interface = _get_interface_name,
				.create_address_enumerator = _create_address_enumerator,
				.get_source_addr = _get_source_addr,
				.get_nexthop = _get_nexthop,
				.add_ip = _add_ip,
				.del_ip = _del_ip,
				.add_route = _add_route,
				.del_route = _del_route,
				.destroy = _destroy,
			},
		},
		.pid = getpid(),
		.ifaces = linked_list_create(),
		.addrs = hashtable_create(
								(hashtable_hash_t)addr_map_entry_hash,
								(hashtable_equals_t)addr_map_entry_equals, 16),
		.tuns = linked_list_create(),
		.lock = rwlock_create(RWLOCK_TYPE_DEFAULT),
		.mutex = mutex_create(MUTEX_TYPE_DEFAULT),
		.condvar = condvar_create(CONDVAR_TYPE_DEFAULT),
		.vip_wait = lib->settings->get_int(lib->settings,
					"%s.plugins.kernel-pfroute.vip_wait", 1000, hydra->daemon),
	);

	/* create a PF_ROUTE socket to communicate with the kernel */
	this->socket = socket(PF_ROUTE, SOCK_RAW, AF_UNSPEC);
	if (this->socket == -1)
	{
		DBG1(DBG_KNL, "unable to create PF_ROUTE socket");
		destroy(this);
		return NULL;
	}

	if (streq(hydra->daemon, "starter"))
	{
		/* starter has no threads, so we do not register for kernel events */
		if (shutdown(this->socket, SHUT_RD) != 0)
		{
			DBG1(DBG_KNL, "closing read end of PF_ROUTE socket failed: %s",
				 strerror(errno));
		}
	}
	else
	{
		lib->processor->queue_job(lib->processor,
			(job_t*)callback_job_create_with_prio(
					(callback_job_cb_t)receive_events, this, NULL,
					(callback_job_cancel_t)return_false, JOB_PRIO_CRITICAL));
	}
	if (init_address_list(this) != SUCCESS)
	{
		DBG1(DBG_KNL, "unable to get interface list");
		destroy(this);
		return NULL;
	}

	return &this->public;
}