[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->bound_dev_if = dev->ifindex;
	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;

	if (dev_addr->bound_dev_if) {
		dev = dev_get_by_index(&init_net, dev_addr->bound_dev_if);
		if (!dev)
			return -ENODEV;
		ret = rdma_copy_addr(dev_addr, dev, NULL);
		dev_put(dev);
		return ret;
	}

	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;
	fl.oif = addr->bound_dev_if;

	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;
	fl.oif = addr->bound_dev_if;

	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;

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

	if (src_addr) {
		if (src_addr->sa_family != dst_addr->sa_family) {
			ret = -EINVAL;
			goto err;
		}

		memcpy(src_in, src_addr, ip_addr_size(src_addr));
	} else {
		src_in->sa_family = dst_addr->sa_family;
	}

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

	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);
		goto err;
	}
	return ret;
err:
	kfree(req);
	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);
6266ed6e	110	dev_addr->bound_dev_if = dev->ifindex;
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
6266ed6e SH	120	if (dev_addr->bound_dev_if) {
	121	dev = dev_get_by_index(&init_net, dev_addr->bound_dev_if);
	122	if (!dev)
	123	return -ENODEV;
	124	ret = rdma_copy_addr(dev_addr, dev, NULL);
	125	dev_put(dev);
	126	return ret;
	127	}
	128
38617c64 AS	129	switch (addr->sa_family) {
	130	case AF_INET:
	131	dev = ip_dev_find(&init_net,
	132	((struct sockaddr_in *) addr)->sin_addr.s_addr);
	133
	134	if (!dev)
	135	return ret;
7025fcd3	136
38617c64 AS	137	ret = rdma_copy_addr(dev_addr, dev, NULL);
	138	dev_put(dev);
	139	break;
2c4ab624 RD	140
2c4ab624 RD	141	#if defined(CONFIG_IPV6) \|\| defined(CONFIG_IPV6_MODULE)
38617c64	142	case AF_INET6:
0f9ea5d2	143	read_lock(&dev_base_lock);
38617c64 AS	144	for_each_netdev(&init_net, dev) {
	145	if (ipv6_chk_addr(&init_net,
	146	&((struct sockaddr_in6 *) addr)->sin6_addr,
	147	dev, 1)) {
	148	ret = rdma_copy_addr(dev_addr, dev, NULL);
	149	break;
	150	}
	151	}
0f9ea5d2	152	read_unlock(&dev_base_lock);
38617c64	153	break;
2c4ab624	154	#endif
38617c64	155	}
7025fcd3 SH	156	return ret;
	157	}
	158	EXPORT_SYMBOL(rdma_translate_ip);
	159
	160	static void set_timeout(unsigned long time)
	161	{
	162	unsigned long delay;
	163
	164	cancel_delayed_work(&work);
	165
	166	delay = time - jiffies;
	167	if ((long)delay <= 0)
	168	delay = 1;
	169
	170	queue_delayed_work(addr_wq, &work, delay);
	171	}
	172
	173	static void queue_req(struct addr_req *req)
	174	{
	175	struct addr_req *temp_req;
	176
	177	mutex_lock(&lock);
	178	list_for_each_entry_reverse(temp_req, &req_list, list) {
f115db48	179	if (time_after_eq(req->timeout, temp_req->timeout))
7025fcd3 SH	180	break;
	181	}
	182
	183	list_add(&req->list, &temp_req->list);
	184
	185	if (req_list.next == &req->list)
	186	set_timeout(req->timeout);
	187	mutex_unlock(&lock);
	188	}
	189
38617c64	190	static void addr_send_arp(struct sockaddr *dst_in)
7025fcd3 SH	191	{
	192	struct rtable *rt;
	193	struct flowi fl;
7025fcd3 SH	194
7025fcd3 SH	195	memset(&fl, 0, sizeof fl);
2c4ab624 RD	196
	197	switch (dst_in->sa_family) {
	198	case AF_INET:
38617c64 AS	199	fl.nl_u.ip4_u.daddr =
38617c64 AS	200	((struct sockaddr_in *) dst_in)->sin_addr.s_addr;
7025fcd3	201
38617c64 AS	202	if (ip_route_output_key(&init_net, &rt, &fl))
	203	return;
	204
	205	neigh_event_send(rt->u.dst.neighbour, NULL);
	206	ip_rt_put(rt);
2c4ab624 RD	207	break;
	208
	209	#if defined(CONFIG_IPV6) \|\| defined(CONFIG_IPV6_MODULE)
	210	case AF_INET6:
	211	{
	212	struct dst_entry *dst;
38617c64	213
38617c64 AS	214	fl.nl_u.ip6_u.daddr =
	215	((struct sockaddr_in6 *) dst_in)->sin6_addr;
	216
	217	dst = ip6_route_output(&init_net, NULL, &fl);
	218	if (!dst)
	219	return;
	220
	221	neigh_event_send(dst->neighbour, NULL);
	222	dst_release(dst);
2c4ab624 RD	223	break;
	224	}
	225	#endif
38617c64	226	}
7025fcd3 SH	227	}
7025fcd3 SH	228
38617c64	229	static int addr4_resolve_remote(struct sockaddr_in *src_in,
7025fcd3 SH	230	struct sockaddr_in *dst_in,
	231	struct rdma_dev_addr *addr)
	232	{
1b90c137 AV	233	__be32 src_ip = src_in->sin_addr.s_addr;
1b90c137 AV	234	__be32 dst_ip = dst_in->sin_addr.s_addr;
7025fcd3 SH	235	struct flowi fl;
	236	struct rtable *rt;
	237	struct neighbour *neigh;
	238	int ret;
	239
	240	memset(&fl, 0, sizeof fl);
	241	fl.nl_u.ip4_u.daddr = dst_ip;
	242	fl.nl_u.ip4_u.saddr = src_ip;
6266ed6e SH	243	fl.oif = addr->bound_dev_if;
6266ed6e SH	244
f206351a	245	ret = ip_route_output_key(&init_net, &rt, &fl);
7025fcd3 SH	246	if (ret)
	247	goto out;
	248
	249	/* If the device does ARP internally, return 'done' */
	250	if (rt->idev->dev->flags & IFF_NOARP) {
07ebafba	251	rdma_copy_addr(addr, rt->idev->dev, NULL);
7025fcd3 SH	252	goto put;
	253	}
	254
	255	neigh = neigh_lookup(&arp_tbl, &rt->rt_gateway, rt->idev->dev);
	256	if (!neigh) {
	257	ret = -ENODATA;
	258	goto put;
	259	}
	260
	261	if (!(neigh->nud_state & NUD_VALID)) {
	262	ret = -ENODATA;
	263	goto release;
	264	}
	265
	266	if (!src_ip) {
	267	src_in->sin_family = dst_in->sin_family;
	268	src_in->sin_addr.s_addr = rt->rt_src;
	269	}
	270
07ebafba	271	ret = rdma_copy_addr(addr, neigh->dev, neigh->ha);
7025fcd3 SH	272	release:
	273	neigh_release(neigh);
	274	put:
	275	ip_rt_put(rt);
	276	out:
	277	return ret;
	278	}
	279
2c4ab624	280	#if defined(CONFIG_IPV6) \|\| defined(CONFIG_IPV6_MODULE)
38617c64 AS	281	static int addr6_resolve_remote(struct sockaddr_in6 *src_in,
	282	struct sockaddr_in6 *dst_in,
	283	struct rdma_dev_addr *addr)
	284	{
	285	struct flowi fl;
	286	struct neighbour *neigh;
	287	struct dst_entry *dst;
	288	int ret = -ENODATA;
	289
	290	memset(&fl, 0, sizeof fl);
	291	fl.nl_u.ip6_u.daddr = dst_in->sin6_addr;
	292	fl.nl_u.ip6_u.saddr = src_in->sin6_addr;
6266ed6e	293	fl.oif = addr->bound_dev_if;
38617c64 AS	294
	295	dst = ip6_route_output(&init_net, NULL, &fl);
	296	if (!dst)
	297	return ret;
	298
	299	if (dst->dev->flags & IFF_NOARP) {
	300	ret = rdma_copy_addr(addr, dst->dev, NULL);
	301	} else {
	302	neigh = dst->neighbour;
	303	if (neigh && (neigh->nud_state & NUD_VALID))
	304	ret = rdma_copy_addr(addr, neigh->dev, neigh->ha);
	305	}
	306
	307	dst_release(dst);
	308	return ret;
	309	}
2c4ab624 RD	310	#else
	311	static int addr6_resolve_remote(struct sockaddr_in6 *src_in,
	312	struct sockaddr_in6 *dst_in,
	313	struct rdma_dev_addr *addr)
	314	{
	315	return -EADDRNOTAVAIL;
	316	}
	317	#endif
38617c64 AS	318
	319	static int addr_resolve_remote(struct sockaddr *src_in,
	320	struct sockaddr *dst_in,
	321	struct rdma_dev_addr *addr)
	322	{
	323	if (src_in->sa_family == AF_INET) {
	324	return addr4_resolve_remote((struct sockaddr_in *) src_in,
	325	(struct sockaddr_in *) dst_in, addr);
	326	} else
	327	return addr6_resolve_remote((struct sockaddr_in6 *) src_in,
	328	(struct sockaddr_in6 *) dst_in, addr);
	329	}
	330
c4028958	331	static void process_req(struct work_struct *work)
7025fcd3 SH	332	{
7025fcd3 SH	333	struct addr_req req, temp_req;
38617c64	334	struct sockaddr src_in, dst_in;
7025fcd3 SH	335	struct list_head done_list;
	336
	337	INIT_LIST_HEAD(&done_list);
	338
	339	mutex_lock(&lock);
	340	list_for_each_entry_safe(req, temp_req, &req_list, list) {
c78bb844	341	if (req->status == -ENODATA) {
38617c64 AS	342	src_in = (struct sockaddr *) &req->src_addr;
38617c64 AS	343	dst_in = (struct sockaddr *) &req->dst_addr;
7025fcd3 SH	344	req->status = addr_resolve_remote(src_in, dst_in,
7025fcd3 SH	345	req->addr);
c78bb844 KK	346	if (req->status && time_after_eq(jiffies, req->timeout))
	347	req->status = -ETIMEDOUT;
	348	else if (req->status == -ENODATA)
	349	continue;
7025fcd3	350	}
04699a1f	351	list_move_tail(&req->list, &done_list);
7025fcd3 SH	352	}
	353
	354	if (!list_empty(&req_list)) {
	355	req = list_entry(req_list.next, struct addr_req, list);
	356	set_timeout(req->timeout);
	357	}
	358	mutex_unlock(&lock);
	359
	360	list_for_each_entry_safe(req, temp_req, &done_list, list) {
	361	list_del(&req->list);
38617c64 AS	362	req->callback(req->status, (struct sockaddr *) &req->src_addr,
38617c64 AS	363	req->addr, req->context);
7a118df3	364	put_client(req->client);
7025fcd3 SH	365	kfree(req);
	366	}
	367	}
	368
38617c64 AS	369	static int addr_resolve_local(struct sockaddr *src_in,
38617c64 AS	370	struct sockaddr *dst_in,
7025fcd3 SH	371	struct rdma_dev_addr *addr)
	372	{
	373	struct net_device *dev;
7025fcd3 SH	374	int ret;
7025fcd3 SH	375
2c4ab624 RD	376	switch (dst_in->sa_family) {
	377	case AF_INET:
	378	{
38617c64 AS	379	__be32 src_ip = ((struct sockaddr_in *) src_in)->sin_addr.s_addr;
	380	__be32 dst_ip = ((struct sockaddr_in *) dst_in)->sin_addr.s_addr;
	381
	382	dev = ip_dev_find(&init_net, dst_ip);
	383	if (!dev)
	384	return -EADDRNOTAVAIL;
	385
	386	if (ipv4_is_zeronet(src_ip)) {
	387	src_in->sa_family = dst_in->sa_family;
	388	((struct sockaddr_in *) src_in)->sin_addr.s_addr = dst_ip;
	389	ret = rdma_copy_addr(addr, dev, dev->dev_addr);
	390	} else if (ipv4_is_loopback(src_ip)) {
	391	ret = rdma_translate_ip(dst_in, addr);
	392	if (!ret)
	393	memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	394	} else {
	395	ret = rdma_translate_ip(src_in, addr);
	396	if (!ret)
	397	memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	398	}
	399	dev_put(dev);
2c4ab624 RD	400	break;
	401	}
	402
	403	#if defined(CONFIG_IPV6) \|\| defined(CONFIG_IPV6_MODULE)
	404	case AF_INET6:
	405	{
38617c64 AS	406	struct in6_addr *a;
38617c64 AS	407
0f9ea5d2	408	read_lock(&dev_base_lock);
38617c64 AS	409	for_each_netdev(&init_net, dev)
38617c64 AS	410	if (ipv6_chk_addr(&init_net,
85f20b39	411	&((struct sockaddr_in6 *) dst_in)->sin6_addr,
38617c64 AS	412	dev, 1))
	413	break;
	414
0f9ea5d2 ED	415	if (!dev) {
0f9ea5d2 ED	416	read_unlock(&dev_base_lock);
38617c64	417	return -EADDRNOTAVAIL;
0f9ea5d2	418	}
38617c64 AS	419
	420	a = &((struct sockaddr_in6 *) src_in)->sin6_addr;
	421
	422	if (ipv6_addr_any(a)) {
	423	src_in->sa_family = dst_in->sa_family;
	424	((struct sockaddr_in6 *) src_in)->sin6_addr =
	425	((struct sockaddr_in6 *) dst_in)->sin6_addr;
	426	ret = rdma_copy_addr(addr, dev, dev->dev_addr);
	427	} else if (ipv6_addr_loopback(a)) {
	428	ret = rdma_translate_ip(dst_in, addr);
	429	if (!ret)
	430	memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	431	} else {
	432	ret = rdma_translate_ip(src_in, addr);
	433	if (!ret)
	434	memcpy(addr->dst_dev_addr, dev->dev_addr, MAX_ADDR_LEN);
	435	}
0f9ea5d2	436	read_unlock(&dev_base_lock);
2c4ab624 RD	437	break;
	438	}
	439	#endif
	440
	441	default:
	442	ret = -EADDRNOTAVAIL;
	443	break;
7025fcd3 SH	444	}
7025fcd3 SH	445
7025fcd3 SH	446	return ret;
	447	}
	448
7a118df3 SH	449	int rdma_resolve_ip(struct rdma_addr_client *client,
7a118df3 SH	450	struct sockaddr src_addr, struct sockaddr dst_addr,
7025fcd3 SH	451	struct rdma_dev_addr *addr, int timeout_ms,
	452	void (callback)(int status, struct sockaddr src_addr,
	453	struct rdma_dev_addr addr, void context),
	454	void *context)
	455	{
38617c64	456	struct sockaddr src_in, dst_in;
7025fcd3 SH	457	struct addr_req *req;
	458	int ret = 0;
	459
dd00cc48	460	req = kzalloc(sizeof *req, GFP_KERNEL);
7025fcd3 SH	461	if (!req)
7025fcd3 SH	462	return -ENOMEM;
7025fcd3	463
d2e08862 SH	464	src_in = (struct sockaddr *) &req->src_addr;
	465	dst_in = (struct sockaddr *) &req->dst_addr;
	466
	467	if (src_addr) {
	468	if (src_addr->sa_family != dst_addr->sa_family) {
	469	ret = -EINVAL;
	470	goto err;
	471	}
	472
	473	memcpy(src_in, src_addr, ip_addr_size(src_addr));
	474	} else {
	475	src_in->sa_family = dst_addr->sa_family;
	476	}
	477
	478	memcpy(dst_in, dst_addr, ip_addr_size(dst_addr));
7025fcd3 SH	479	req->addr = addr;
	480	req->callback = callback;
	481	req->context = context;
7a118df3 SH	482	req->client = client;
7a118df3 SH	483	atomic_inc(&client->refcount);
7025fcd3	484
7025fcd3 SH	485	req->status = addr_resolve_local(src_in, dst_in, addr);
	486	if (req->status == -EADDRNOTAVAIL)
	487	req->status = addr_resolve_remote(src_in, dst_in, addr);
	488
	489	switch (req->status) {
	490	case 0:
	491	req->timeout = jiffies;
	492	queue_req(req);
	493	break;
	494	case -ENODATA:
	495	req->timeout = msecs_to_jiffies(timeout_ms) + jiffies;
	496	queue_req(req);
	497	addr_send_arp(dst_in);
	498	break;
	499	default:
	500	ret = req->status;
7a118df3	501	atomic_dec(&client->refcount);
d2e08862	502	goto err;
7025fcd3 SH	503	}
7025fcd3 SH	504	return ret;
d2e08862 SH	505	err:
	506	kfree(req);
	507	return ret;
7025fcd3 SH	508	}
	509	EXPORT_SYMBOL(rdma_resolve_ip);
	510
	511	void rdma_addr_cancel(struct rdma_dev_addr *addr)
	512	{
	513	struct addr_req req, temp_req;
	514
	515	mutex_lock(&lock);
	516	list_for_each_entry_safe(req, temp_req, &req_list, list) {
	517	if (req->addr == addr) {
	518	req->status = -ECANCELED;
	519	req->timeout = jiffies;
04699a1f	520	list_move(&req->list, &req_list);
7025fcd3 SH	521	set_timeout(req->timeout);
	522	break;
	523	}
	524	}
	525	mutex_unlock(&lock);
	526	}
	527	EXPORT_SYMBOL(rdma_addr_cancel);
	528
3cd96564	529	static int netevent_callback(struct notifier_block *self, unsigned long event,
e795d092	530	void *ctx)
7025fcd3	531	{
3cd96564	532	if (event == NETEVENT_NEIGH_UPDATE) {
e795d092	533	struct neighbour *neigh = ctx;
7025fcd3	534
1f126670	535	if (neigh->nud_state & NUD_VALID) {
e795d092 TT	536	set_timeout(jiffies);
	537	}
	538	}
7025fcd3 SH	539	return 0;
	540	}
	541
e795d092 TT	542	static struct notifier_block nb = {
e795d092 TT	543	.notifier_call = netevent_callback
7025fcd3 SH	544	};
7025fcd3 SH	545
716abb1f	546	static int __init addr_init(void)
7025fcd3	547	{
c7f743a6	548	addr_wq = create_singlethread_workqueue("ib_addr");
7025fcd3 SH	549	if (!addr_wq)
	550	return -ENOMEM;
	551
e795d092	552	register_netevent_notifier(&nb);
7025fcd3 SH	553	return 0;
	554	}
	555
716abb1f	556	static void __exit addr_cleanup(void)
7025fcd3	557	{
e795d092	558	unregister_netevent_notifier(&nb);
7025fcd3 SH	559	destroy_workqueue(addr_wq);
	560	}
	561
	562	module_init(addr_init);
	563	module_exit(addr_cleanup);