rxe_net.c 17.3 KB
Newer Older
M
Moni Shoua 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
/*
 * Copyright (c) 2016 Mellanox Technologies Ltd. All rights reserved.
 * Copyright (c) 2015 System Fabric Works, Inc. 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/skbuff.h>
#include <linux/if_arp.h>
#include <linux/netdevice.h>
#include <linux/if.h>
#include <linux/if_vlan.h>
#include <net/udp_tunnel.h>
#include <net/sch_generic.h>
#include <linux/netfilter.h>
#include <rdma/ib_addr.h>

#include "rxe.h"
#include "rxe_net.h"
#include "rxe_loc.h"

static LIST_HEAD(rxe_dev_list);
49
static DEFINE_SPINLOCK(dev_list_lock); /* spinlock for device list */
M
Moni Shoua 已提交
50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

struct rxe_dev *net_to_rxe(struct net_device *ndev)
{
	struct rxe_dev *rxe;
	struct rxe_dev *found = NULL;

	spin_lock_bh(&dev_list_lock);
	list_for_each_entry(rxe, &rxe_dev_list, list) {
		if (rxe->ndev == ndev) {
			found = rxe;
			break;
		}
	}
	spin_unlock_bh(&dev_list_lock);

	return found;
}

68
struct rxe_dev *get_rxe_by_name(const char *name)
M
Moni Shoua 已提交
69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84
{
	struct rxe_dev *rxe;
	struct rxe_dev *found = NULL;

	spin_lock_bh(&dev_list_lock);
	list_for_each_entry(rxe, &rxe_dev_list, list) {
		if (!strcmp(name, rxe->ib_dev.name)) {
			found = rxe;
			break;
		}
	}
	spin_unlock_bh(&dev_list_lock);
	return found;
}


85
static struct rxe_recv_sockets recv_sockets;
M
Moni Shoua 已提交
86

87
struct device *rxe_dma_device(struct rxe_dev *rxe)
M
Moni Shoua 已提交
88 89 90 91 92
{
	struct net_device *ndev;

	ndev = rxe->ndev;

93
	if (is_vlan_dev(ndev))
M
Moni Shoua 已提交
94 95 96 97 98
		ndev = vlan_dev_real_dev(ndev);

	return ndev->dev.parent;
}

99
int rxe_mcast_add(struct rxe_dev *rxe, union ib_gid *mgid)
M
Moni Shoua 已提交
100 101 102 103 104 105 106 107 108 109
{
	int err;
	unsigned char ll_addr[ETH_ALEN];

	ipv6_eth_mc_map((struct in6_addr *)mgid->raw, ll_addr);
	err = dev_mc_add(rxe->ndev, ll_addr);

	return err;
}

110
int rxe_mcast_delete(struct rxe_dev *rxe, union ib_gid *mgid)
M
Moni Shoua 已提交
111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184
{
	int err;
	unsigned char ll_addr[ETH_ALEN];

	ipv6_eth_mc_map((struct in6_addr *)mgid->raw, ll_addr);
	err = dev_mc_del(rxe->ndev, ll_addr);

	return err;
}

static struct dst_entry *rxe_find_route4(struct net_device *ndev,
				  struct in_addr *saddr,
				  struct in_addr *daddr)
{
	struct rtable *rt;
	struct flowi4 fl = { { 0 } };

	memset(&fl, 0, sizeof(fl));
	fl.flowi4_oif = ndev->ifindex;
	memcpy(&fl.saddr, saddr, sizeof(*saddr));
	memcpy(&fl.daddr, daddr, sizeof(*daddr));
	fl.flowi4_proto = IPPROTO_UDP;

	rt = ip_route_output_key(&init_net, &fl);
	if (IS_ERR(rt)) {
		pr_err_ratelimited("no route to %pI4\n", &daddr->s_addr);
		return NULL;
	}

	return &rt->dst;
}

#if IS_ENABLED(CONFIG_IPV6)
static struct dst_entry *rxe_find_route6(struct net_device *ndev,
					 struct in6_addr *saddr,
					 struct in6_addr *daddr)
{
	struct dst_entry *ndst;
	struct flowi6 fl6 = { { 0 } };

	memset(&fl6, 0, sizeof(fl6));
	fl6.flowi6_oif = ndev->ifindex;
	memcpy(&fl6.saddr, saddr, sizeof(*saddr));
	memcpy(&fl6.daddr, daddr, sizeof(*daddr));
	fl6.flowi6_proto = IPPROTO_UDP;

	if (unlikely(ipv6_stub->ipv6_dst_lookup(sock_net(recv_sockets.sk6->sk),
						recv_sockets.sk6->sk, &ndst, &fl6))) {
		pr_err_ratelimited("no route to %pI6\n", daddr);
		goto put;
	}

	if (unlikely(ndst->error)) {
		pr_err("no route to %pI6\n", daddr);
		goto put;
	}

	return ndst;
put:
	dst_release(ndst);
	return NULL;
}

#else

static struct dst_entry *rxe_find_route6(struct net_device *ndev,
					 struct in6_addr *saddr,
					 struct in6_addr *daddr)
{
	return NULL;
}

#endif

185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209
/*
 * Derive the net_device from the av.
 * For physical devices, this will just return rxe->ndev.
 * But for VLAN devices, it will return the vlan dev.
 * Caller should dev_put() the returned net_device.
 */
static struct net_device *rxe_netdev_from_av(struct rxe_dev *rxe,
					     int port_num,
					     struct rxe_av *av)
{
	union ib_gid gid;
	struct ib_gid_attr attr;
	struct net_device *ndev = rxe->ndev;

	if (ib_get_cached_gid(&rxe->ib_dev, port_num, av->grh.sgid_index,
			      &gid, &attr) == 0 &&
	    attr.ndev && attr.ndev != ndev)
		ndev = attr.ndev;
	else
		/* Only to ensure that caller may call dev_put() */
		dev_hold(ndev);

	return ndev;
}

210 211 212 213 214
static struct dst_entry *rxe_find_route(struct rxe_dev *rxe,
					struct rxe_qp *qp,
					struct rxe_av *av)
{
	struct dst_entry *dst = NULL;
215 216 217
	struct net_device *ndev;

	ndev = rxe_netdev_from_av(rxe, qp->attr.port_num, av);
218 219 220 221

	if (qp_type(qp) == IB_QPT_RC)
		dst = sk_dst_get(qp->sk->sk);

222
	if (!dst || !dst_check(dst, qp->dst_cookie)) {
223 224 225 226 227 228 229 230 231
		if (dst)
			dst_release(dst);

		if (av->network_type == RDMA_NETWORK_IPV4) {
			struct in_addr *saddr;
			struct in_addr *daddr;

			saddr = &av->sgid_addr._sockaddr_in.sin_addr;
			daddr = &av->dgid_addr._sockaddr_in.sin_addr;
232
			dst = rxe_find_route4(ndev, saddr, daddr);
233 234 235 236 237 238
		} else if (av->network_type == RDMA_NETWORK_IPV6) {
			struct in6_addr *saddr6;
			struct in6_addr *daddr6;

			saddr6 = &av->sgid_addr._sockaddr_in6.sin6_addr;
			daddr6 = &av->dgid_addr._sockaddr_in6.sin6_addr;
239
			dst = rxe_find_route6(ndev, saddr6, daddr6);
240 241 242 243 244
#if IS_ENABLED(CONFIG_IPV6)
			if (dst)
				qp->dst_cookie =
					rt6_get_cookie((struct rt6_info *)dst);
#endif
245 246 247
		}
	}

248
	dev_put(ndev);
249 250 251
	return dst;
}

M
Moni Shoua 已提交
252 253 254 255
static int rxe_udp_encap_recv(struct sock *sk, struct sk_buff *skb)
{
	struct udphdr *udph;
	struct net_device *ndev = skb->dev;
256
	struct net_device *rdev = ndev;
M
Moni Shoua 已提交
257 258 259
	struct rxe_dev *rxe = net_to_rxe(ndev);
	struct rxe_pkt_info *pkt = SKB_TO_PKT(skb);

260 261 262 263
	if (!rxe && is_vlan_dev(rdev)) {
		rdev = vlan_dev_real_dev(ndev);
		rxe = net_to_rxe(rdev);
	}
M
Moni Shoua 已提交
264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289
	if (!rxe)
		goto drop;

	if (skb_linearize(skb)) {
		pr_err("skb_linearize failed\n");
		goto drop;
	}

	udph = udp_hdr(skb);
	pkt->rxe = rxe;
	pkt->port_num = 1;
	pkt->hdr = (u8 *)(udph + 1);
	pkt->mask = RXE_GRH_MASK;
	pkt->paylen = be16_to_cpu(udph->len) - sizeof(*udph);

	return rxe_rcv(skb);
drop:
	kfree_skb(skb);
	return 0;
}

static struct socket *rxe_setup_udp_tunnel(struct net *net, __be16 port,
					   bool ipv6)
{
	int err;
	struct socket *sock;
290 291
	struct udp_port_cfg udp_cfg = { };
	struct udp_tunnel_sock_cfg tnl_cfg = { };
M
Moni Shoua 已提交
292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317

	if (ipv6) {
		udp_cfg.family = AF_INET6;
		udp_cfg.ipv6_v6only = 1;
	} else {
		udp_cfg.family = AF_INET;
	}

	udp_cfg.local_udp_port = port;

	/* Create UDP socket */
	err = udp_sock_create(net, &udp_cfg, &sock);
	if (err < 0) {
		pr_err("failed to create udp socket. err = %d\n", err);
		return ERR_PTR(err);
	}

	tnl_cfg.encap_type = 1;
	tnl_cfg.encap_rcv = rxe_udp_encap_recv;

	/* Setup UDP tunnel */
	setup_udp_tunnel_sock(net, sock, &tnl_cfg);

	return sock;
}

318
static void rxe_release_udp_tunnel(struct socket *sk)
M
Moni Shoua 已提交
319
{
320 321
	if (sk)
		udp_tunnel_sock_release(sk);
M
Moni Shoua 已提交
322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347
}

static void prepare_udp_hdr(struct sk_buff *skb, __be16 src_port,
			    __be16 dst_port)
{
	struct udphdr *udph;

	__skb_push(skb, sizeof(*udph));
	skb_reset_transport_header(skb);
	udph = udp_hdr(skb);

	udph->dest = dst_port;
	udph->source = src_port;
	udph->len = htons(skb->len);
	udph->check = 0;
}

static void prepare_ipv4_hdr(struct dst_entry *dst, struct sk_buff *skb,
			     __be32 saddr, __be32 daddr, __u8 proto,
			     __u8 tos, __u8 ttl, __be16 df, bool xnet)
{
	struct iphdr *iph;

	skb_scrub_packet(skb, xnet);

	skb_clear_hash(skb);
348
	skb_dst_set(skb, dst_clone(dst));
M
Moni Shoua 已提交
349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378
	memset(IPCB(skb), 0, sizeof(*IPCB(skb)));

	skb_push(skb, sizeof(struct iphdr));
	skb_reset_network_header(skb);

	iph = ip_hdr(skb);

	iph->version	=	IPVERSION;
	iph->ihl	=	sizeof(struct iphdr) >> 2;
	iph->frag_off	=	df;
	iph->protocol	=	proto;
	iph->tos	=	tos;
	iph->daddr	=	daddr;
	iph->saddr	=	saddr;
	iph->ttl	=	ttl;
	__ip_select_ident(dev_net(dst->dev), iph,
			  skb_shinfo(skb)->gso_segs ?: 1);
	iph->tot_len = htons(skb->len);
	ip_send_check(iph);
}

static void prepare_ipv6_hdr(struct dst_entry *dst, struct sk_buff *skb,
			     struct in6_addr *saddr, struct in6_addr *daddr,
			     __u8 proto, __u8 prio, __u8 ttl)
{
	struct ipv6hdr *ip6h;

	memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
	IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED
			    | IPSKB_REROUTED);
379
	skb_dst_set(skb, dst_clone(dst));
M
Moni Shoua 已提交
380 381 382 383 384 385 386 387 388 389 390 391 392

	__skb_push(skb, sizeof(*ip6h));
	skb_reset_network_header(skb);
	ip6h		  = ipv6_hdr(skb);
	ip6_flow_hdr(ip6h, prio, htonl(0));
	ip6h->payload_len = htons(skb->len);
	ip6h->nexthdr     = proto;
	ip6h->hop_limit   = ttl;
	ip6h->daddr	  = *daddr;
	ip6h->saddr	  = *saddr;
	ip6h->payload_len = htons(skb->len - sizeof(*ip6h));
}

393 394
static int prepare4(struct rxe_dev *rxe, struct rxe_pkt_info *pkt,
		    struct sk_buff *skb, struct rxe_av *av)
M
Moni Shoua 已提交
395
{
396
	struct rxe_qp *qp = pkt->qp;
M
Moni Shoua 已提交
397 398 399 400 401 402
	struct dst_entry *dst;
	bool xnet = false;
	__be16 df = htons(IP_DF);
	struct in_addr *saddr = &av->sgid_addr._sockaddr_in.sin_addr;
	struct in_addr *daddr = &av->dgid_addr._sockaddr_in.sin_addr;

403
	dst = rxe_find_route(rxe, qp, av);
M
Moni Shoua 已提交
404 405 406 407 408 409 410 411 412 413 414 415 416
	if (!dst) {
		pr_err("Host not reachable\n");
		return -EHOSTUNREACH;
	}

	if (!memcmp(saddr, daddr, sizeof(*daddr)))
		pkt->mask |= RXE_LOOPBACK_MASK;

	prepare_udp_hdr(skb, htons(RXE_ROCE_V2_SPORT),
			htons(ROCE_V2_UDP_DPORT));

	prepare_ipv4_hdr(dst, skb, saddr->s_addr, daddr->s_addr, IPPROTO_UDP,
			 av->grh.traffic_class, av->grh.hop_limit, df, xnet);
417 418 419 420 421 422

	if (qp_type(qp) == IB_QPT_RC)
		sk_dst_set(qp->sk->sk, dst);
	else
		dst_release(dst);

M
Moni Shoua 已提交
423 424 425
	return 0;
}

426 427
static int prepare6(struct rxe_dev *rxe, struct rxe_pkt_info *pkt,
		    struct sk_buff *skb, struct rxe_av *av)
M
Moni Shoua 已提交
428
{
429
	struct rxe_qp *qp = pkt->qp;
430
	struct dst_entry *dst;
M
Moni Shoua 已提交
431 432 433
	struct in6_addr *saddr = &av->sgid_addr._sockaddr_in6.sin6_addr;
	struct in6_addr *daddr = &av->dgid_addr._sockaddr_in6.sin6_addr;

434
	dst = rxe_find_route(rxe, qp, av);
M
Moni Shoua 已提交
435 436 437 438 439 440 441 442 443 444 445 446 447 448
	if (!dst) {
		pr_err("Host not reachable\n");
		return -EHOSTUNREACH;
	}

	if (!memcmp(saddr, daddr, sizeof(*daddr)))
		pkt->mask |= RXE_LOOPBACK_MASK;

	prepare_udp_hdr(skb, htons(RXE_ROCE_V2_SPORT),
			htons(ROCE_V2_UDP_DPORT));

	prepare_ipv6_hdr(dst, skb, saddr, daddr, IPPROTO_UDP,
			 av->grh.traffic_class,
			 av->grh.hop_limit);
449 450 451 452 453 454

	if (qp_type(qp) == IB_QPT_RC)
		sk_dst_set(qp->sk->sk, dst);
	else
		dst_release(dst);

M
Moni Shoua 已提交
455 456 457
	return 0;
}

458 459
int rxe_prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt,
		struct sk_buff *skb, u32 *crc)
M
Moni Shoua 已提交
460 461 462 463 464
{
	int err = 0;
	struct rxe_av *av = rxe_get_av(pkt);

	if (av->network_type == RDMA_NETWORK_IPV4)
465
		err = prepare4(rxe, pkt, skb, av);
M
Moni Shoua 已提交
466
	else if (av->network_type == RDMA_NETWORK_IPV6)
467
		err = prepare6(rxe, pkt, skb, av);
M
Moni Shoua 已提交
468 469 470 471 472 473 474 475 476 477 478 479 480 481 482

	*crc = rxe_icrc_hdr(pkt, skb);

	return err;
}

static void rxe_skb_tx_dtor(struct sk_buff *skb)
{
	struct sock *sk = skb->sk;
	struct rxe_qp *qp = sk->sk_user_data;
	int skb_out = atomic_dec_return(&qp->skb_out);

	if (unlikely(qp->need_req_skb &&
		     skb_out < RXE_INFLIGHT_SKBS_PER_QP_LOW))
		rxe_run_task(&qp->req.task, 1);
483 484

	rxe_drop_ref(qp);
M
Moni Shoua 已提交
485 486
}

487
int rxe_send(struct rxe_pkt_info *pkt, struct sk_buff *skb)
M
Moni Shoua 已提交
488 489 490 491 492 493
{
	struct rxe_av *av;
	int err;

	av = rxe_get_av(pkt);

494 495
	skb->destructor = rxe_skb_tx_dtor;
	skb->sk = pkt->qp->sk->sk;
M
Moni Shoua 已提交
496

497 498 499
	rxe_add_ref(pkt->qp);
	atomic_inc(&pkt->qp->skb_out);

M
Moni Shoua 已提交
500
	if (av->network_type == RDMA_NETWORK_IPV4) {
501
		err = ip_local_out(dev_net(skb_dst(skb)->dev), skb->sk, skb);
M
Moni Shoua 已提交
502
	} else if (av->network_type == RDMA_NETWORK_IPV6) {
503
		err = ip6_local_out(dev_net(skb_dst(skb)->dev), skb->sk, skb);
M
Moni Shoua 已提交
504 505
	} else {
		pr_err("Unknown layer 3 protocol: %d\n", av->network_type);
506 507
		atomic_dec(&pkt->qp->skb_out);
		rxe_drop_ref(pkt->qp);
508
		kfree_skb(skb);
M
Moni Shoua 已提交
509 510 511 512 513 514 515 516 517 518 519
		return -EINVAL;
	}

	if (unlikely(net_xmit_eval(err))) {
		pr_debug("error sending packet: %d\n", err);
		return -EAGAIN;
	}

	return 0;
}

520
int rxe_loopback(struct sk_buff *skb)
M
Moni Shoua 已提交
521 522 523 524 525 526 527 528 529
{
	return rxe_rcv(skb);
}

static inline int addr_same(struct rxe_dev *rxe, struct rxe_av *av)
{
	return rxe->port.port_guid == av->grh.dgid.global.interface_id;
}

530 531
struct sk_buff *rxe_init_packet(struct rxe_dev *rxe, struct rxe_av *av,
				int paylen, struct rxe_pkt_info *pkt)
M
Moni Shoua 已提交
532 533 534
{
	unsigned int hdr_len;
	struct sk_buff *skb;
535 536 537 538
	struct net_device *ndev;
	const int port_num = 1;

	ndev = rxe_netdev_from_av(rxe, port_num, av);
M
Moni Shoua 已提交
539 540 541 542 543 544 545 546

	if (av->network_type == RDMA_NETWORK_IPV4)
		hdr_len = ETH_HLEN + sizeof(struct udphdr) +
			sizeof(struct iphdr);
	else
		hdr_len = ETH_HLEN + sizeof(struct udphdr) +
			sizeof(struct ipv6hdr);

547
	skb = alloc_skb(paylen + hdr_len + LL_RESERVED_SPACE(ndev),
M
Moni Shoua 已提交
548
			GFP_ATOMIC);
549 550 551

	if (unlikely(!skb)) {
		dev_put(ndev);
M
Moni Shoua 已提交
552
		return NULL;
553
	}
M
Moni Shoua 已提交
554 555 556

	skb_reserve(skb, hdr_len + LL_RESERVED_SPACE(rxe->ndev));

557
	skb->dev	= ndev;
M
Moni Shoua 已提交
558 559 560 561 562 563
	if (av->network_type == RDMA_NETWORK_IPV4)
		skb->protocol = htons(ETH_P_IP);
	else
		skb->protocol = htons(ETH_P_IPV6);

	pkt->rxe	= rxe;
564
	pkt->port_num	= port_num;
M
Moni Shoua 已提交
565 566 567 568 569
	pkt->hdr	= skb_put(skb, paylen);
	pkt->mask	|= RXE_GRH_MASK;

	memset(pkt->hdr, 0, paylen);

570
	dev_put(ndev);
M
Moni Shoua 已提交
571 572 573 574 575 576 577
	return skb;
}

/*
 * this is required by rxe_cfg to match rxe devices in
 * /sys/class/infiniband up with their underlying ethernet devices
 */
578
const char *rxe_parent_name(struct rxe_dev *rxe, unsigned int port_num)
M
Moni Shoua 已提交
579 580 581 582
{
	return rxe->ndev->name;
}

583
enum rdma_link_layer rxe_link_layer(struct rxe_dev *rxe, unsigned int port_num)
M
Moni Shoua 已提交
584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605
{
	return IB_LINK_LAYER_ETHERNET;
}

struct rxe_dev *rxe_net_add(struct net_device *ndev)
{
	int err;
	struct rxe_dev *rxe = NULL;

	rxe = (struct rxe_dev *)ib_alloc_device(sizeof(*rxe));
	if (!rxe)
		return NULL;

	rxe->ndev = ndev;

	err = rxe_add(rxe, ndev->mtu);
	if (err) {
		ib_dealloc_device(&rxe->ib_dev);
		return NULL;
	}

	spin_lock_bh(&dev_list_lock);
606
	list_add_tail(&rxe->list, &rxe_dev_list);
M
Moni Shoua 已提交
607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648
	spin_unlock_bh(&dev_list_lock);
	return rxe;
}

void rxe_remove_all(void)
{
	spin_lock_bh(&dev_list_lock);
	while (!list_empty(&rxe_dev_list)) {
		struct rxe_dev *rxe =
			list_first_entry(&rxe_dev_list, struct rxe_dev, list);

		list_del(&rxe->list);
		spin_unlock_bh(&dev_list_lock);
		rxe_remove(rxe);
		spin_lock_bh(&dev_list_lock);
	}
	spin_unlock_bh(&dev_list_lock);
}
EXPORT_SYMBOL(rxe_remove_all);

static void rxe_port_event(struct rxe_dev *rxe,
			   enum ib_event_type event)
{
	struct ib_event ev;

	ev.device = &rxe->ib_dev;
	ev.element.port_num = 1;
	ev.event = event;

	ib_dispatch_event(&ev);
}

/* Caller must hold net_info_lock */
void rxe_port_up(struct rxe_dev *rxe)
{
	struct rxe_port *port;

	port = &rxe->port;
	port->attr.state = IB_PORT_ACTIVE;
	port->attr.phys_state = IB_PHYS_STATE_LINK_UP;

	rxe_port_event(rxe, IB_EVENT_PORT_ACTIVE);
649
	pr_info("set %s active\n", rxe->ib_dev.name);
M
Moni Shoua 已提交
650 651 652 653 654 655 656 657 658 659 660 661
}

/* Caller must hold net_info_lock */
void rxe_port_down(struct rxe_dev *rxe)
{
	struct rxe_port *port;

	port = &rxe->port;
	port->attr.state = IB_PORT_DOWN;
	port->attr.phys_state = IB_PHYS_STATE_LINK_DOWN;

	rxe_port_event(rxe, IB_EVENT_PORT_ERR);
662
	pr_info("set %s down\n", rxe->ib_dev.name);
M
Moni Shoua 已提交
663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686
}

static int rxe_notify(struct notifier_block *not_blk,
		      unsigned long event,
		      void *arg)
{
	struct net_device *ndev = netdev_notifier_info_to_dev(arg);
	struct rxe_dev *rxe = net_to_rxe(ndev);

	if (!rxe)
		goto out;

	switch (event) {
	case NETDEV_UNREGISTER:
		list_del(&rxe->list);
		rxe_remove(rxe);
		break;
	case NETDEV_UP:
		rxe_port_up(rxe);
		break;
	case NETDEV_DOWN:
		rxe_port_down(rxe);
		break;
	case NETDEV_CHANGEMTU:
687
		pr_info("%s changed mtu to %d\n", ndev->name, ndev->mtu);
M
Moni Shoua 已提交
688 689 690
		rxe_set_mtu(rxe, ndev->mtu);
		break;
	case NETDEV_CHANGE:
691 692 693 694 695 696
		if (netif_running(ndev) && netif_carrier_ok(ndev))
			rxe_port_up(rxe);
		else
			rxe_port_down(rxe);
		break;
	case NETDEV_REBOOT:
M
Moni Shoua 已提交
697 698 699 700 701
	case NETDEV_GOING_DOWN:
	case NETDEV_CHANGEADDR:
	case NETDEV_CHANGENAME:
	case NETDEV_FEAT_CHANGE:
	default:
702
		pr_info("ignoring netdev event = %ld for %s\n",
M
Moni Shoua 已提交
703 704 705 706 707 708 709
			event, ndev->name);
		break;
	}
out:
	return NOTIFY_OK;
}

Z
Zhu Yanjun 已提交
710
static struct notifier_block rxe_net_notifier = {
M
Moni Shoua 已提交
711 712 713
	.notifier_call = rxe_notify,
};

714
static int rxe_net_ipv4_init(void)
M
Moni Shoua 已提交
715 716
{
	recv_sockets.sk4 = rxe_setup_udp_tunnel(&init_net,
717
				htons(ROCE_V2_UDP_DPORT), false);
M
Moni Shoua 已提交
718 719
	if (IS_ERR(recv_sockets.sk4)) {
		recv_sockets.sk4 = NULL;
720
		pr_err("Failed to create IPv4 UDP tunnel\n");
M
Moni Shoua 已提交
721 722 723
		return -1;
	}

724
	return 0;
M
Moni Shoua 已提交
725 726
}

727
static int rxe_net_ipv6_init(void)
M
Moni Shoua 已提交
728
{
729
#if IS_ENABLED(CONFIG_IPV6)
M
Moni Shoua 已提交
730

731 732 733 734
	recv_sockets.sk6 = rxe_setup_udp_tunnel(&init_net,
						htons(ROCE_V2_UDP_DPORT), true);
	if (IS_ERR(recv_sockets.sk6)) {
		recv_sockets.sk6 = NULL;
735
		pr_err("Failed to create IPv6 UDP tunnel\n");
736 737 738 739 740 741 742 743 744 745
		return -1;
	}
#endif
	return 0;
}

void rxe_net_exit(void)
{
	rxe_release_udp_tunnel(recv_sockets.sk6);
	rxe_release_udp_tunnel(recv_sockets.sk4);
M
Moni Shoua 已提交
746 747
	unregister_netdevice_notifier(&rxe_net_notifier);
}
748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770

int rxe_net_init(void)
{
	int err;

	recv_sockets.sk6 = NULL;

	err = rxe_net_ipv4_init();
	if (err)
		return err;
	err = rxe_net_ipv6_init();
	if (err)
		goto err_out;
	err = register_netdevice_notifier(&rxe_net_notifier);
	if (err) {
		pr_err("Failed to register netdev notifier\n");
		goto err_out;
	}
	return 0;
err_out:
	rxe_net_exit();
	return err;
}