[net-next-2.6.git] / drivers / infiniband / core / addr.c

/*
 * Copyright (c) 2005 Voltaire Inc.  All rights reserved.
 * Copyright (c) 2002-2005, Network Appliance, Inc. All rights reserved.
 * Copyright (c) 1999-2005, Mellanox Technologies, Inc. All rights reserved.
 * Copyright (c) 2005 Intel Corporation.  All rights reserved.
 *
 * This software is available to you under a choice of one of two
 * licenses.  You may choose to be licensed under the terms of the GNU
 * General Public License (GPL) Version 2, available from the file
 * COPYING in the main directory of this source tree, or the
 * OpenIB.org BSD license below:
 *
 *     Redistribution and use in source and binary forms, with or
 *     without modification, are permitted provided that the following
 *     conditions are met:
 *
 *      - Redistributions of source code must retain the above
 *        copyright notice, this list of conditions and the following
 *        disclaimer.
 *
 *      - Redistributions in binary form must reproduce the above
 *        copyright notice, this list of conditions and the following
 *        disclaimer in the documentation and/or other materials
 *        provided with the distribution.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 */

#include <linux/mutex.h>
#include <linux/inetdevice.h>
#include <linux/workqueue.h>
#include <linux/if_arp.h>
#include <net/arp.h>
#include <net/neighbour.h>
#include <net/route.h>
#include <net/netevent.h>
#include <net/addrconf.h>
#include <net/ip6_route.h>
#include <rdma/ib_addr.h>

MODULE_AUTHOR("Sean Hefty");
MODULE_DESCRIPTION("IB Address Translation");
MODULE_LICENSE("Dual BSD/GPL");

struct addr_req {
	struct list_head list;
	struct sockaddr_storage src_addr;
	struct sockaddr_storage dst_addr;
	struct rdma_dev_addr *addr;
	struct rdma_addr_client *client;
	void *context;
	void (*callback)(int status, struct sockaddr *src_addr,
			 struct rdma_dev_addr *addr, void *context);
	unsigned long timeout;
	int status;
};

static void process_req(struct work_struct *work);

static DEFINE_MUTEX(lock);
static LIST_HEAD(req_list);
static DECLARE_DELAYED_WORK(work, process_req);
static struct workqueue_struct *addr_wq;

void rdma_addr_register_client(struct rdma_addr_client *client)
{
	atomic_set(&client->refcount, 1);
	init_completion(&client->comp);
}
EXPORT_SYMBOL(rdma_addr_register_client);

static inline void put_client(struct rdma_addr_client *client)
{
	if (atomic_dec_and_test(&client->refcount))
		complete(&client->comp);
}

void rdma_addr_unregister_client(struct rdma_addr_client *client)
{
	put_client(client);
	wait_for_completion(&client->comp);
}
EXPORT_SYMBOL(rdma_addr_unregister_client);

int rdma_copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
		     const unsigned char *dst_dev_addr)
{
	switch (dev->type) {
	case ARPHRD_INFINIBAND:
		dev_addr->dev_type = RDMA_NODE_IB_CA;
		break;
	case ARPHRD_ETHER:
		dev_addr->dev_type = RDMA_NODE_RNIC;
		break;
	default:
		return -EADDRNOTAVAIL;
	}

	memcpy(dev_addr->src_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	memcpy(dev_addr->broadcast, dev->broadcast, MAX_ADDR_LEN);
	if (dst_dev_addr)
		memcpy(dev_addr->dst_dev_addr, dst_dev_addr, MAX_ADDR_LEN);
	dev_addr->src_dev = dev;
	return 0;
}
EXPORT_SYMBOL(rdma_copy_addr);

int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr)
{
	struct net_device *dev;
	int ret = -EADDRNOTAVAIL;

	switch (addr->sa_family) {
	case AF_INET:
		dev = ip_dev_find(&init_net,
			((struct sockaddr_in *) addr)->sin_addr.s_addr);

		if (!dev)
			return ret;

		ret = rdma_copy_addr(dev_addr, dev, NULL);
		dev_put(dev);
		break;

#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
	case AF_INET6:
		read_lock(&dev_base_lock);
		for_each_netdev(&init_net, dev) {
			if (ipv6_chk_addr(&init_net,
					  &((struct sockaddr_in6 *) addr)->sin6_addr,
					  dev, 1)) {
				ret = rdma_copy_addr(dev_addr, dev, NULL);
				break;
			}
		}
		read_unlock(&dev_base_lock);
		break;
#endif
	}
	return ret;
}
EXPORT_SYMBOL(rdma_translate_ip);

static void set_timeout(unsigned long time)
{
	unsigned long delay;

	cancel_delayed_work(&work);

	delay = time - jiffies;
	if ((long)delay <= 0)
		delay = 1;

	queue_delayed_work(addr_wq, &work, delay);
}

static void queue_req(struct addr_req *req)
{
	struct addr_req *temp_req;

	mutex_lock(&lock);
	list_for_each_entry_reverse(temp_req, &req_list, list) {
		if (time_after_eq(req->timeout, temp_req->timeout))
			break;
	}

	list_add(&req->list, &temp_req->list);

	if (req_list.next == &req->list)
		set_timeout(req->timeout);
	mutex_unlock(&lock);
}

static void addr_send_arp(struct sockaddr *dst_in)
{
	struct rtable *rt;
	struct flowi fl;

	memset(&fl, 0, sizeof fl);

	switch (dst_in->sa_family) {
	case AF_INET:
		fl.nl_u.ip4_u.daddr =
			((struct sockaddr_in *) dst_in)->sin_addr.s_addr;

		if (ip_route_output_key(&init_net, &rt, &fl))
			return;

		neigh_event_send(rt->u.dst.neighbour, NULL);
		ip_rt_put(rt);
		break;

#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
	case AF_INET6:
	{
		struct dst_entry *dst;

		fl.nl_u.ip6_u.daddr =
			((struct sockaddr_in6 *) dst_in)->sin6_addr;

		dst = ip6_route_output(&init_net, NULL, &fl);
		if (!dst)
			return;

		neigh_event_send(dst->neighbour, NULL);
		dst_release(dst);
		break;
	}
#endif
	}
}

static int addr4_resolve_remote(struct sockaddr_in *src_in,
			       struct sockaddr_in *dst_in,
			       struct rdma_dev_addr *addr)
{
	__be32 src_ip = src_in->sin_addr.s_addr;
	__be32 dst_ip = dst_in->sin_addr.s_addr;
	struct flowi fl;
	struct rtable *rt;
	struct neighbour *neigh;
	int ret;

	memset(&fl, 0, sizeof fl);
	fl.nl_u.ip4_u.daddr = dst_ip;
	fl.nl_u.ip4_u.saddr = src_ip;
	ret = ip_route_output_key(&init_net, &rt, &fl);
	if (ret)
		goto out;

	/* If the device does ARP internally, return 'done' */
	if (rt->idev->dev->flags & IFF_NOARP) {
		rdma_copy_addr(addr, rt->idev->dev, NULL);
		goto put;
	}

	neigh = neigh_lookup(&arp_tbl, &rt->rt_gateway, rt->idev->dev);
	if (!neigh) {
		ret = -ENODATA;
		goto put;
	}

	if (!(neigh->nud_state & NUD_VALID)) {
		ret = -ENODATA;
		goto release;
	}

	if (!src_ip) {
		src_in->sin_family = dst_in->sin_family;
		src_in->sin_addr.s_addr = rt->rt_src;
	}

	ret = rdma_copy_addr(addr, neigh->dev, neigh->ha);
release:
	neigh_release(neigh);
put:
	ip_rt_put(rt);
out:
	return ret;
}

#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
static int addr6_resolve_remote(struct sockaddr_in6 *src_in,
			       struct sockaddr_in6 *dst_in,
			       struct rdma_dev_addr *addr)
{
	struct flowi fl;
	struct neighbour *neigh;
	struct dst_entry *dst;
	int ret = -ENODATA;

	memset(&fl, 0, sizeof fl);
	fl.nl_u.ip6_u.daddr = dst_in->sin6_addr;
	fl.nl_u.ip6_u.saddr = src_in->sin6_addr;

	dst = ip6_route_output(&init_net, NULL, &fl);
	if (!dst)
		return ret;

	if (dst->dev->flags & IFF_NOARP) {
		ret = rdma_copy_addr(addr, dst->dev, NULL);
	} else {
		neigh = dst->neighbour;
		if (neigh && (neigh->nud_state & NUD_VALID))
			ret = rdma_copy_addr(addr, neigh->dev, neigh->ha);
	}

	dst_release(dst);
	return ret;
}
#else
static int addr6_resolve_remote(struct sockaddr_in6 *src_in,
			       struct sockaddr_in6 *dst_in,
			       struct rdma_dev_addr *addr)
{
	return -EADDRNOTAVAIL;
}
#endif

static int addr_resolve_remote(struct sockaddr *src_in,
				struct sockaddr *dst_in,
				struct rdma_dev_addr *addr)
{
	if (src_in->sa_family == AF_INET) {
		return addr4_resolve_remote((struct sockaddr_in *) src_in,
			(struct sockaddr_in *) dst_in, addr);
	} else
		return addr6_resolve_remote((struct sockaddr_in6 *) src_in,
			(struct sockaddr_in6 *) dst_in, addr);
}

static void process_req(struct work_struct *work)
{
	struct addr_req *req, *temp_req;
	struct sockaddr *src_in, *dst_in;
	struct list_head done_list;

	INIT_LIST_HEAD(&done_list);

	mutex_lock(&lock);
	list_for_each_entry_safe(req, temp_req, &req_list, list) {
		if (req->status == -ENODATA) {
			src_in = (struct sockaddr *) &req->src_addr;
			dst_in = (struct sockaddr *) &req->dst_addr;
			req->status = addr_resolve_remote(src_in, dst_in,
							  req->addr);
			if (req->status && time_after_eq(jiffies, req->timeout))
				req->status = -ETIMEDOUT;
			else if (req->status == -ENODATA)
				continue;
		}
		list_move_tail(&req->list, &done_list);
	}

	if (!list_empty(&req_list)) {
		req = list_entry(req_list.next, struct addr_req, list);
		set_timeout(req->timeout);
	}
	mutex_unlock(&lock);

	list_for_each_entry_safe(req, temp_req, &done_list, list) {
		list_del(&req->list);
		req->callback(req->status, (struct sockaddr *) &req->src_addr,
			req->addr, req->context);
		put_client(req->client);
		kfree(req);
	}
}

static int addr_resolve_local(struct sockaddr *src_in,
			      struct sockaddr *dst_in,
			      struct rdma_dev_addr *addr)
{
	struct net_device *dev;
	int ret;

	switch (dst_in->sa_family) {
	case AF_INET:
	{
		__be32 src_ip = ((struct sockaddr_in *) src_in)->sin_addr.s_addr;
		__be32 dst_ip = ((struct sockaddr_in *) dst_in)->sin_addr.s_addr;

		dev = ip_dev_find(&init_net, dst_ip);
		if (!dev)
			return -EADDRNOTAVAIL;

		if (ipv4_is_zeronet(src_ip)) {
			src_in->sa_family = dst_in->sa_family;
			((struct sockaddr_in *) src_in)->sin_addr.s_addr = dst_ip;
			ret = rdma_copy_addr(addr, dev, dev->dev_addr);
		} else if (ipv4_is_loopback(src_ip)) {
			ret = rdma_translate_ip(dst_in, addr);
			if (!ret)
				memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
		} else {
			ret = rdma_translate_ip(src_in, addr);
			if (!ret)
				memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
		}
		dev_put(dev);
		break;
	}

#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
	case AF_INET6:
	{
		struct in6_addr *a;

		read_lock(&dev_base_lock);
		for_each_netdev(&init_net, dev)
			if (ipv6_chk_addr(&init_net,
					  &((struct sockaddr_in6 *) dst_in)->sin6_addr,
					  dev, 1))
				break;

		if (!dev) {
			read_unlock(&dev_base_lock);
			return -EADDRNOTAVAIL;
		}

		a = &((struct sockaddr_in6 *) src_in)->sin6_addr;

		if (ipv6_addr_any(a)) {
			src_in->sa_family = dst_in->sa_family;
			((struct sockaddr_in6 *) src_in)->sin6_addr =
				((struct sockaddr_in6 *) dst_in)->sin6_addr;
			ret = rdma_copy_addr(addr, dev, dev->dev_addr);
		} else if (ipv6_addr_loopback(a)) {
			ret = rdma_translate_ip(dst_in, addr);
			if (!ret)
				memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
		} else  {
			ret = rdma_translate_ip(src_in, addr);
			if (!ret)
				memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
		}
		read_unlock(&dev_base_lock);
		break;
	}
#endif

	default:
		ret = -EADDRNOTAVAIL;
		break;
	}

	return ret;
}

int rdma_resolve_ip(struct rdma_addr_client *client,
		    struct sockaddr *src_addr, struct sockaddr *dst_addr,
		    struct rdma_dev_addr *addr, int timeout_ms,
		    void (*callback)(int status, struct sockaddr *src_addr,
				     struct rdma_dev_addr *addr, void *context),
		    void *context)
{
	struct sockaddr *src_in, *dst_in;
	struct addr_req *req;
	int ret = 0;

	req = kzalloc(sizeof *req, GFP_KERNEL);
	if (!req)
		return -ENOMEM;

	if (src_addr)
		memcpy(&req->src_addr, src_addr, ip_addr_size(src_addr));
	memcpy(&req->dst_addr, dst_addr, ip_addr_size(dst_addr));
	req->addr = addr;
	req->callback = callback;
	req->context = context;
	req->client = client;
	atomic_inc(&client->refcount);

	src_in = (struct sockaddr *) &req->src_addr;
	dst_in = (struct sockaddr *) &req->dst_addr;

	req->status = addr_resolve_local(src_in, dst_in, addr);
	if (req->status == -EADDRNOTAVAIL)
		req->status = addr_resolve_remote(src_in, dst_in, addr);

	switch (req->status) {
	case 0:
		req->timeout = jiffies;
		queue_req(req);
		break;
	case -ENODATA:
		req->timeout = msecs_to_jiffies(timeout_ms) + jiffies;
		queue_req(req);
		addr_send_arp(dst_in);
		break;
	default:
		ret = req->status;
		atomic_dec(&client->refcount);
		kfree(req);
		break;
	}
	return ret;
}
EXPORT_SYMBOL(rdma_resolve_ip);

void rdma_addr_cancel(struct rdma_dev_addr *addr)
{
	struct addr_req *req, *temp_req;

	mutex_lock(&lock);
	list_for_each_entry_safe(req, temp_req, &req_list, list) {
		if (req->addr == addr) {
			req->status = -ECANCELED;
			req->timeout = jiffies;
			list_move(&req->list, &req_list);
			set_timeout(req->timeout);
			break;
		}
	}
	mutex_unlock(&lock);
}
EXPORT_SYMBOL(rdma_addr_cancel);

static int netevent_callback(struct notifier_block *self, unsigned long event,
	void *ctx)
{
	if (event == NETEVENT_NEIGH_UPDATE) {
		struct neighbour *neigh = ctx;

		if (neigh->nud_state & NUD_VALID) {
			set_timeout(jiffies);
		}
	}
	return 0;
}

static struct notifier_block nb = {
	.notifier_call = netevent_callback
};

static int __init addr_init(void)
{
	addr_wq = create_singlethread_workqueue("ib_addr");
	if (!addr_wq)
		return -ENOMEM;

	register_netevent_notifier(&nb);
	return 0;
}

static void __exit addr_cleanup(void)
{
	unregister_netevent_notifier(&nb);
	destroy_workqueue(addr_wq);
}

module_init(addr_init);
module_exit(addr_cleanup);
Commit	Line	Data
7025fcd3 SH	1	/*
	2	* Copyright (c) 2005 Voltaire Inc. All rights reserved.
	3	* Copyright (c) 2002-2005, Network Appliance, Inc. All rights reserved.
	4	* Copyright (c) 1999-2005, Mellanox Technologies, Inc. All rights reserved.
	5	* Copyright (c) 2005 Intel Corporation. All rights reserved.
	6	*
a9474917 SH	7	* This software is available to you under a choice of one of two
	8	* licenses. You may choose to be licensed under the terms of the GNU
	9	* General Public License (GPL) Version 2, available from the file
	10	* COPYING in the main directory of this source tree, or the
	11	* OpenIB.org BSD license below:
7025fcd3	12	*
a9474917 SH	13	* Redistribution and use in source and binary forms, with or
	14	* without modification, are permitted provided that the following
	15	* conditions are met:
7025fcd3	16	*
a9474917 SH	17	* - Redistributions of source code must retain the above
	18	* copyright notice, this list of conditions and the following
	19	* disclaimer.
7025fcd3	20	*
a9474917 SH	21	* - Redistributions in binary form must reproduce the above
	22	* copyright notice, this list of conditions and the following
	23	* disclaimer in the documentation and/or other materials
	24	* provided with the distribution.
7025fcd3	25	*
a9474917 SH	26	* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
	27	* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
	28	* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
	29	* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
	30	* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
	31	* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
	32	* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
	33	* SOFTWARE.
7025fcd3 SH	34	*/
	35
	36	#include <linux/mutex.h>
	37	#include <linux/inetdevice.h>
	38	#include <linux/workqueue.h>
	39	#include <linux/if_arp.h>
	40	#include <net/arp.h>
	41	#include <net/neighbour.h>
	42	#include <net/route.h>
e795d092	43	#include <net/netevent.h>
38617c64 AS	44	#include <net/addrconf.h>
38617c64 AS	45	#include <net/ip6_route.h>
7025fcd3 SH	46	#include <rdma/ib_addr.h>
	47
	48	MODULE_AUTHOR("Sean Hefty");
	49	MODULE_DESCRIPTION("IB Address Translation");
	50	MODULE_LICENSE("Dual BSD/GPL");
	51
	52	struct addr_req {
	53	struct list_head list;
38617c64 AS	54	struct sockaddr_storage src_addr;
38617c64 AS	55	struct sockaddr_storage dst_addr;
7025fcd3	56	struct rdma_dev_addr *addr;
7a118df3	57	struct rdma_addr_client *client;
7025fcd3 SH	58	void *context;
	59	void (callback)(int status, struct sockaddr src_addr,
	60	struct rdma_dev_addr addr, void context);
	61	unsigned long timeout;
	62	int status;
	63	};
	64
c4028958	65	static void process_req(struct work_struct *work);
7025fcd3 SH	66
	67	static DEFINE_MUTEX(lock);
	68	static LIST_HEAD(req_list);
c4028958	69	static DECLARE_DELAYED_WORK(work, process_req);
7025fcd3 SH	70	static struct workqueue_struct *addr_wq;
7025fcd3 SH	71
7a118df3 SH	72	void rdma_addr_register_client(struct rdma_addr_client *client)
	73	{
	74	atomic_set(&client->refcount, 1);
	75	init_completion(&client->comp);
	76	}
	77	EXPORT_SYMBOL(rdma_addr_register_client);
	78
	79	static inline void put_client(struct rdma_addr_client *client)
	80	{
	81	if (atomic_dec_and_test(&client->refcount))
	82	complete(&client->comp);
	83	}
	84
	85	void rdma_addr_unregister_client(struct rdma_addr_client *client)
	86	{
	87	put_client(client);
	88	wait_for_completion(&client->comp);
	89	}
	90	EXPORT_SYMBOL(rdma_addr_unregister_client);
	91
07ebafba TT	92	int rdma_copy_addr(struct rdma_dev_addr dev_addr, struct net_device dev,
07ebafba TT	93	const unsigned char *dst_dev_addr)
7025fcd3 SH	94	{
	95	switch (dev->type) {
	96	case ARPHRD_INFINIBAND:
07ebafba TT	97	dev_addr->dev_type = RDMA_NODE_IB_CA;
	98	break;
	99	case ARPHRD_ETHER:
	100	dev_addr->dev_type = RDMA_NODE_RNIC;
7025fcd3 SH	101	break;
	102	default:
	103	return -EADDRNOTAVAIL;
	104	}
	105
	106	memcpy(dev_addr->src_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	107	memcpy(dev_addr->broadcast, dev->broadcast, MAX_ADDR_LEN);
	108	if (dst_dev_addr)
	109	memcpy(dev_addr->dst_dev_addr, dst_dev_addr, MAX_ADDR_LEN);
64c5e613	110	dev_addr->src_dev = dev;
7025fcd3 SH	111	return 0;
7025fcd3 SH	112	}
07ebafba	113	EXPORT_SYMBOL(rdma_copy_addr);
7025fcd3 SH	114
	115	int rdma_translate_ip(struct sockaddr addr, struct rdma_dev_addr dev_addr)
	116	{
	117	struct net_device *dev;
38617c64	118	int ret = -EADDRNOTAVAIL;
7025fcd3	119
38617c64 AS	120	switch (addr->sa_family) {
	121	case AF_INET:
	122	dev = ip_dev_find(&init_net,
	123	((struct sockaddr_in *) addr)->sin_addr.s_addr);
	124
	125	if (!dev)
	126	return ret;
7025fcd3	127
38617c64 AS	128	ret = rdma_copy_addr(dev_addr, dev, NULL);
	129	dev_put(dev);
	130	break;
2c4ab624 RD	131
2c4ab624 RD	132	#if defined(CONFIG_IPV6) \|\| defined(CONFIG_IPV6_MODULE)
38617c64	133	case AF_INET6:
0f9ea5d2	134	read_lock(&dev_base_lock);
38617c64 AS	135	for_each_netdev(&init_net, dev) {
	136	if (ipv6_chk_addr(&init_net,
	137	&((struct sockaddr_in6 *) addr)->sin6_addr,
	138	dev, 1)) {
	139	ret = rdma_copy_addr(dev_addr, dev, NULL);
	140	break;
	141	}
	142	}
0f9ea5d2	143	read_unlock(&dev_base_lock);
38617c64	144	break;
2c4ab624	145	#endif
38617c64	146	}
7025fcd3 SH	147	return ret;
	148	}
	149	EXPORT_SYMBOL(rdma_translate_ip);
	150
	151	static void set_timeout(unsigned long time)
	152	{
	153	unsigned long delay;
	154
	155	cancel_delayed_work(&work);
	156
	157	delay = time - jiffies;
	158	if ((long)delay <= 0)
	159	delay = 1;
	160
	161	queue_delayed_work(addr_wq, &work, delay);
	162	}
	163
	164	static void queue_req(struct addr_req *req)
	165	{
	166	struct addr_req *temp_req;
	167
	168	mutex_lock(&lock);
	169	list_for_each_entry_reverse(temp_req, &req_list, list) {
f115db48	170	if (time_after_eq(req->timeout, temp_req->timeout))
7025fcd3 SH	171	break;
	172	}
	173
	174	list_add(&req->list, &temp_req->list);
	175
	176	if (req_list.next == &req->list)
	177	set_timeout(req->timeout);
	178	mutex_unlock(&lock);
	179	}
	180
38617c64	181	static void addr_send_arp(struct sockaddr *dst_in)
7025fcd3 SH	182	{
	183	struct rtable *rt;
	184	struct flowi fl;
7025fcd3 SH	185
7025fcd3 SH	186	memset(&fl, 0, sizeof fl);
2c4ab624 RD	187
	188	switch (dst_in->sa_family) {
	189	case AF_INET:
38617c64 AS	190	fl.nl_u.ip4_u.daddr =
38617c64 AS	191	((struct sockaddr_in *) dst_in)->sin_addr.s_addr;
7025fcd3	192
38617c64 AS	193	if (ip_route_output_key(&init_net, &rt, &fl))
	194	return;
	195
	196	neigh_event_send(rt->u.dst.neighbour, NULL);
	197	ip_rt_put(rt);
2c4ab624 RD	198	break;
	199
	200	#if defined(CONFIG_IPV6) \|\| defined(CONFIG_IPV6_MODULE)
	201	case AF_INET6:
	202	{
	203	struct dst_entry *dst;
38617c64	204
38617c64 AS	205	fl.nl_u.ip6_u.daddr =
	206	((struct sockaddr_in6 *) dst_in)->sin6_addr;
	207
	208	dst = ip6_route_output(&init_net, NULL, &fl);
	209	if (!dst)
	210	return;
	211
	212	neigh_event_send(dst->neighbour, NULL);
	213	dst_release(dst);
2c4ab624 RD	214	break;
	215	}
	216	#endif
38617c64	217	}
7025fcd3 SH	218	}
7025fcd3 SH	219
38617c64	220	static int addr4_resolve_remote(struct sockaddr_in *src_in,
7025fcd3 SH	221	struct sockaddr_in *dst_in,
	222	struct rdma_dev_addr *addr)
	223	{
1b90c137 AV	224	__be32 src_ip = src_in->sin_addr.s_addr;
1b90c137 AV	225	__be32 dst_ip = dst_in->sin_addr.s_addr;
7025fcd3 SH	226	struct flowi fl;
	227	struct rtable *rt;
	228	struct neighbour *neigh;
	229	int ret;
	230
	231	memset(&fl, 0, sizeof fl);
	232	fl.nl_u.ip4_u.daddr = dst_ip;
	233	fl.nl_u.ip4_u.saddr = src_ip;
f206351a	234	ret = ip_route_output_key(&init_net, &rt, &fl);
7025fcd3 SH	235	if (ret)
	236	goto out;
	237
	238	/* If the device does ARP internally, return 'done' */
	239	if (rt->idev->dev->flags & IFF_NOARP) {
07ebafba	240	rdma_copy_addr(addr, rt->idev->dev, NULL);
7025fcd3 SH	241	goto put;
	242	}
	243
	244	neigh = neigh_lookup(&arp_tbl, &rt->rt_gateway, rt->idev->dev);
	245	if (!neigh) {
	246	ret = -ENODATA;
	247	goto put;
	248	}
	249
	250	if (!(neigh->nud_state & NUD_VALID)) {
	251	ret = -ENODATA;
	252	goto release;
	253	}
	254
	255	if (!src_ip) {
	256	src_in->sin_family = dst_in->sin_family;
	257	src_in->sin_addr.s_addr = rt->rt_src;
	258	}
	259
07ebafba	260	ret = rdma_copy_addr(addr, neigh->dev, neigh->ha);
7025fcd3 SH	261	release:
	262	neigh_release(neigh);
	263	put:
	264	ip_rt_put(rt);
	265	out:
	266	return ret;
	267	}
	268
2c4ab624	269	#if defined(CONFIG_IPV6) \|\| defined(CONFIG_IPV6_MODULE)
38617c64 AS	270	static int addr6_resolve_remote(struct sockaddr_in6 *src_in,
	271	struct sockaddr_in6 *dst_in,
	272	struct rdma_dev_addr *addr)
	273	{
	274	struct flowi fl;
	275	struct neighbour *neigh;
	276	struct dst_entry *dst;
	277	int ret = -ENODATA;
	278
	279	memset(&fl, 0, sizeof fl);
	280	fl.nl_u.ip6_u.daddr = dst_in->sin6_addr;
	281	fl.nl_u.ip6_u.saddr = src_in->sin6_addr;
	282
	283	dst = ip6_route_output(&init_net, NULL, &fl);
	284	if (!dst)
	285	return ret;
	286
	287	if (dst->dev->flags & IFF_NOARP) {
	288	ret = rdma_copy_addr(addr, dst->dev, NULL);
	289	} else {
	290	neigh = dst->neighbour;
	291	if (neigh && (neigh->nud_state & NUD_VALID))
	292	ret = rdma_copy_addr(addr, neigh->dev, neigh->ha);
	293	}
	294
	295	dst_release(dst);
	296	return ret;
	297	}
2c4ab624 RD	298	#else
	299	static int addr6_resolve_remote(struct sockaddr_in6 *src_in,
	300	struct sockaddr_in6 *dst_in,
	301	struct rdma_dev_addr *addr)
	302	{
	303	return -EADDRNOTAVAIL;
	304	}
	305	#endif
38617c64 AS	306
	307	static int addr_resolve_remote(struct sockaddr *src_in,
	308	struct sockaddr *dst_in,
	309	struct rdma_dev_addr *addr)
	310	{
	311	if (src_in->sa_family == AF_INET) {
	312	return addr4_resolve_remote((struct sockaddr_in *) src_in,
	313	(struct sockaddr_in *) dst_in, addr);
	314	} else
	315	return addr6_resolve_remote((struct sockaddr_in6 *) src_in,
	316	(struct sockaddr_in6 *) dst_in, addr);
	317	}
	318
c4028958	319	static void process_req(struct work_struct *work)
7025fcd3 SH	320	{
7025fcd3 SH	321	struct addr_req req, temp_req;
38617c64	322	struct sockaddr src_in, dst_in;
7025fcd3 SH	323	struct list_head done_list;
	324
	325	INIT_LIST_HEAD(&done_list);
	326
	327	mutex_lock(&lock);
	328	list_for_each_entry_safe(req, temp_req, &req_list, list) {
c78bb844	329	if (req->status == -ENODATA) {
38617c64 AS	330	src_in = (struct sockaddr *) &req->src_addr;
38617c64 AS	331	dst_in = (struct sockaddr *) &req->dst_addr;
7025fcd3 SH	332	req->status = addr_resolve_remote(src_in, dst_in,
7025fcd3 SH	333	req->addr);
c78bb844 KK	334	if (req->status && time_after_eq(jiffies, req->timeout))
	335	req->status = -ETIMEDOUT;
	336	else if (req->status == -ENODATA)
	337	continue;
7025fcd3	338	}
04699a1f	339	list_move_tail(&req->list, &done_list);
7025fcd3 SH	340	}
	341
	342	if (!list_empty(&req_list)) {
	343	req = list_entry(req_list.next, struct addr_req, list);
	344	set_timeout(req->timeout);
	345	}
	346	mutex_unlock(&lock);
	347
	348	list_for_each_entry_safe(req, temp_req, &done_list, list) {
	349	list_del(&req->list);
38617c64 AS	350	req->callback(req->status, (struct sockaddr *) &req->src_addr,
38617c64 AS	351	req->addr, req->context);
7a118df3	352	put_client(req->client);
7025fcd3 SH	353	kfree(req);
	354	}
	355	}
	356
38617c64 AS	357	static int addr_resolve_local(struct sockaddr *src_in,
38617c64 AS	358	struct sockaddr *dst_in,
7025fcd3 SH	359	struct rdma_dev_addr *addr)
	360	{
	361	struct net_device *dev;
7025fcd3 SH	362	int ret;
7025fcd3 SH	363
2c4ab624 RD	364	switch (dst_in->sa_family) {
	365	case AF_INET:
	366	{
38617c64 AS	367	__be32 src_ip = ((struct sockaddr_in *) src_in)->sin_addr.s_addr;
	368	__be32 dst_ip = ((struct sockaddr_in *) dst_in)->sin_addr.s_addr;
	369
	370	dev = ip_dev_find(&init_net, dst_ip);
	371	if (!dev)
	372	return -EADDRNOTAVAIL;
	373
	374	if (ipv4_is_zeronet(src_ip)) {
	375	src_in->sa_family = dst_in->sa_family;
	376	((struct sockaddr_in *) src_in)->sin_addr.s_addr = dst_ip;
	377	ret = rdma_copy_addr(addr, dev, dev->dev_addr);
	378	} else if (ipv4_is_loopback(src_ip)) {
	379	ret = rdma_translate_ip(dst_in, addr);
	380	if (!ret)
	381	memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	382	} else {
	383	ret = rdma_translate_ip(src_in, addr);
	384	if (!ret)
	385	memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	386	}
	387	dev_put(dev);
2c4ab624 RD	388	break;
	389	}
	390
	391	#if defined(CONFIG_IPV6) \|\| defined(CONFIG_IPV6_MODULE)
	392	case AF_INET6:
	393	{
38617c64 AS	394	struct in6_addr *a;
38617c64 AS	395
0f9ea5d2	396	read_lock(&dev_base_lock);
38617c64 AS	397	for_each_netdev(&init_net, dev)
38617c64 AS	398	if (ipv6_chk_addr(&init_net,
85f20b39	399	&((struct sockaddr_in6 *) dst_in)->sin6_addr,
38617c64 AS	400	dev, 1))
	401	break;
	402
0f9ea5d2 ED	403	if (!dev) {
0f9ea5d2 ED	404	read_unlock(&dev_base_lock);
38617c64	405	return -EADDRNOTAVAIL;
0f9ea5d2	406	}
38617c64 AS	407
	408	a = &((struct sockaddr_in6 *) src_in)->sin6_addr;
	409
	410	if (ipv6_addr_any(a)) {
	411	src_in->sa_family = dst_in->sa_family;
	412	((struct sockaddr_in6 *) src_in)->sin6_addr =
	413	((struct sockaddr_in6 *) dst_in)->sin6_addr;
	414	ret = rdma_copy_addr(addr, dev, dev->dev_addr);
	415	} else if (ipv6_addr_loopback(a)) {
	416	ret = rdma_translate_ip(dst_in, addr);
	417	if (!ret)
	418	memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	419	} else {
	420	ret = rdma_translate_ip(src_in, addr);
	421	if (!ret)
	422	memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	423	}
0f9ea5d2	424	read_unlock(&dev_base_lock);
2c4ab624 RD	425	break;
	426	}
	427	#endif
	428
	429	default:
	430	ret = -EADDRNOTAVAIL;
	431	break;
7025fcd3 SH	432	}
7025fcd3 SH	433
7025fcd3 SH	434	return ret;
	435	}
	436
7a118df3 SH	437	int rdma_resolve_ip(struct rdma_addr_client *client,
7a118df3 SH	438	struct sockaddr src_addr, struct sockaddr dst_addr,
7025fcd3 SH	439	struct rdma_dev_addr *addr, int timeout_ms,
	440	void (callback)(int status, struct sockaddr src_addr,
	441	struct rdma_dev_addr addr, void context),
	442	void *context)
	443	{
38617c64	444	struct sockaddr src_in, dst_in;
7025fcd3 SH	445	struct addr_req *req;
	446	int ret = 0;
	447
dd00cc48	448	req = kzalloc(sizeof *req, GFP_KERNEL);
7025fcd3 SH	449	if (!req)
7025fcd3 SH	450	return -ENOMEM;
7025fcd3 SH	451
	452	if (src_addr)
	453	memcpy(&req->src_addr, src_addr, ip_addr_size(src_addr));
	454	memcpy(&req->dst_addr, dst_addr, ip_addr_size(dst_addr));
	455	req->addr = addr;
	456	req->callback = callback;
	457	req->context = context;
7a118df3 SH	458	req->client = client;
7a118df3 SH	459	atomic_inc(&client->refcount);
7025fcd3	460
38617c64 AS	461	src_in = (struct sockaddr *) &req->src_addr;
38617c64 AS	462	dst_in = (struct sockaddr *) &req->dst_addr;
7025fcd3 SH	463
	464	req->status = addr_resolve_local(src_in, dst_in, addr);
	465	if (req->status == -EADDRNOTAVAIL)
	466	req->status = addr_resolve_remote(src_in, dst_in, addr);
	467
	468	switch (req->status) {
	469	case 0:
	470	req->timeout = jiffies;
	471	queue_req(req);
	472	break;
	473	case -ENODATA:
	474	req->timeout = msecs_to_jiffies(timeout_ms) + jiffies;
	475	queue_req(req);
	476	addr_send_arp(dst_in);
	477	break;
	478	default:
	479	ret = req->status;
7a118df3	480	atomic_dec(&client->refcount);
7025fcd3 SH	481	kfree(req);
	482	break;
	483	}
	484	return ret;
	485	}
	486	EXPORT_SYMBOL(rdma_resolve_ip);
	487
	488	void rdma_addr_cancel(struct rdma_dev_addr *addr)
	489	{
	490	struct addr_req req, temp_req;
	491
	492	mutex_lock(&lock);
	493	list_for_each_entry_safe(req, temp_req, &req_list, list) {
	494	if (req->addr == addr) {
	495	req->status = -ECANCELED;
	496	req->timeout = jiffies;
04699a1f	497	list_move(&req->list, &req_list);
7025fcd3 SH	498	set_timeout(req->timeout);
	499	break;
	500	}
	501	}
	502	mutex_unlock(&lock);
	503	}
	504	EXPORT_SYMBOL(rdma_addr_cancel);
	505
3cd96564	506	static int netevent_callback(struct notifier_block *self, unsigned long event,
e795d092	507	void *ctx)
7025fcd3	508	{
3cd96564	509	if (event == NETEVENT_NEIGH_UPDATE) {
e795d092	510	struct neighbour *neigh = ctx;
7025fcd3	511
1f126670	512	if (neigh->nud_state & NUD_VALID) {
e795d092 TT	513	set_timeout(jiffies);
	514	}
	515	}
7025fcd3 SH	516	return 0;
	517	}
	518
e795d092 TT	519	static struct notifier_block nb = {
e795d092 TT	520	.notifier_call = netevent_callback
7025fcd3 SH	521	};
7025fcd3 SH	522
716abb1f	523	static int __init addr_init(void)
7025fcd3	524	{
c7f743a6	525	addr_wq = create_singlethread_workqueue("ib_addr");
7025fcd3 SH	526	if (!addr_wq)
	527	return -ENOMEM;
	528
e795d092	529	register_netevent_notifier(&nb);
7025fcd3 SH	530	return 0;
	531	}
	532
716abb1f	533	static void __exit addr_cleanup(void)
7025fcd3	534	{
e795d092	535	unregister_netevent_notifier(&nb);
7025fcd3 SH	536	destroy_workqueue(addr_wq);
	537	}
	538
	539	module_init(addr_init);
	540	module_exit(addr_cleanup);