af_packet.c 46.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7
/*
 * INET		An implementation of the TCP/IP protocol suite for the LINUX
 *		operating system.  INET is implemented using the  BSD Socket
 *		interface as the means of communication with the user level.
 *
 *		PACKET - implements raw packet sockets.
 *
8
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
9 10 11
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *		Alan Cox, <gw4pts@gw4pts.ampr.org>
 *
12
 * Fixes:
L
Linus Torvalds 已提交
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
 *		Alan Cox	:	verify_area() now used correctly
 *		Alan Cox	:	new skbuff lists, look ma no backlogs!
 *		Alan Cox	:	tidied skbuff lists.
 *		Alan Cox	:	Now uses generic datagram routines I
 *					added. Also fixed the peek/read crash
 *					from all old Linux datagram code.
 *		Alan Cox	:	Uses the improved datagram code.
 *		Alan Cox	:	Added NULL's for socket options.
 *		Alan Cox	:	Re-commented the code.
 *		Alan Cox	:	Use new kernel side addressing
 *		Rob Janssen	:	Correct MTU usage.
 *		Dave Platt	:	Counter leaks caused by incorrect
 *					interrupt locking and some slightly
 *					dubious gcc output. Can you read
 *					compiler: it said _VOLATILE_
 *	Richard Kooijman	:	Timestamp fixes.
 *		Alan Cox	:	New buffers. Use sk->mac.raw.
 *		Alan Cox	:	sendmsg/recvmsg support.
 *		Alan Cox	:	Protocol setting support
 *	Alexey Kuznetsov	:	Untied from IPv4 stack.
 *	Cyrus Durgin		:	Fixed kerneld for kmod.
 *	Michal Ostrowski        :       Module initialization cleanup.
35
 *         Ulises Alonso        :       Frame number limit removal and
L
Linus Torvalds 已提交
36
 *                                      packet_set_ring memory leak.
37 38 39
 *		Eric Biederman	:	Allow for > 8 byte hardware addresses.
 *					The convention is that longer addresses
 *					will simply extend the hardware address
40
 *					byte arrays at the end of sockaddr_ll
41
 *					and packet_mreq.
L
Linus Torvalds 已提交
42 43 44 45 46 47 48
 *
 *		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.
 *
 */
49

L
Linus Torvalds 已提交
50 51
#include <linux/types.h>
#include <linux/mm.h>
52
#include <linux/capability.h>
L
Linus Torvalds 已提交
53 54 55 56 57 58 59
#include <linux/fcntl.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/if_packet.h>
#include <linux/wireless.h>
60
#include <linux/kernel.h>
L
Linus Torvalds 已提交
61
#include <linux/kmod.h>
62
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
63 64 65 66 67 68 69 70 71 72
#include <net/ip.h>
#include <net/protocol.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <linux/errno.h>
#include <linux/timer.h>
#include <asm/system.h>
#include <asm/uaccess.h>
#include <asm/ioctls.h>
#include <asm/page.h>
A
Al Viro 已提交
73
#include <asm/cacheflush.h>
L
Linus Torvalds 已提交
74 75 76 77 78 79
#include <asm/io.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#include <linux/poll.h>
#include <linux/module.h>
#include <linux/init.h>
80
#include <linux/mutex.h>
L
Linus Torvalds 已提交
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100

#ifdef CONFIG_INET
#include <net/inet_common.h>
#endif

/*
   Assumptions:
   - if device has no dev->hard_header routine, it adds and removes ll header
     inside itself. In this case ll header is invisible outside of device,
     but higher levels still should reserve dev->hard_header_len.
     Some devices are enough clever to reallocate skb, when header
     will not fit to reserved space (tunnel), another ones are silly
     (PPP).
   - packet socket receives packets with pulled ll header,
     so that SOCK_RAW should push it back.

On receive:
-----------

Incoming, dev->hard_header!=NULL
101 102
   mac_header -> ll header
   data       -> data
L
Linus Torvalds 已提交
103 104

Outgoing, dev->hard_header!=NULL
105 106
   mac_header -> ll header
   data       -> ll header
L
Linus Torvalds 已提交
107 108

Incoming, dev->hard_header==NULL
109 110
   mac_header -> UNKNOWN position. It is very likely, that it points to ll
		 header.  PPP makes it, that is wrong, because introduce
111
		 assymetry between rx and tx paths.
112
   data       -> data
L
Linus Torvalds 已提交
113 114

Outgoing, dev->hard_header==NULL
115 116
   mac_header -> data. ll header is still not built!
   data       -> data
L
Linus Torvalds 已提交
117 118 119 120 121 122 123 124 125

Resume
  If dev->hard_header==NULL we are unlikely to restore sensible ll header.


On transmit:
------------

dev->hard_header != NULL
126 127
   mac_header -> ll header
   data       -> ll header
L
Linus Torvalds 已提交
128 129

dev->hard_header == NULL (ll header is added by device, we cannot control it)
130 131
   mac_header -> data
   data       -> data
L
Linus Torvalds 已提交
132 133 134 135 136 137 138 139 140 141 142 143 144 145

   We should set nh.raw on output to correct posistion,
   packet classifier depends on it.
 */

/* Private packet socket structures. */

struct packet_mclist
{
	struct packet_mclist	*next;
	int			ifindex;
	int			count;
	unsigned short		type;
	unsigned short		alen;
146 147 148 149 150 151 152 153 154 155 156
	unsigned char		addr[MAX_ADDR_LEN];
};
/* identical to struct packet_mreq except it has
 * a longer address field.
 */
struct packet_mreq_max
{
	int		mr_ifindex;
	unsigned short	mr_type;
	unsigned short	mr_alen;
	unsigned char	mr_address[MAX_ADDR_LEN];
L
Linus Torvalds 已提交
157
};
158

L
Linus Torvalds 已提交
159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178
#ifdef CONFIG_PACKET_MMAP
static int packet_set_ring(struct sock *sk, struct tpacket_req *req, int closing);
#endif

static void packet_flush_mclist(struct sock *sk);

struct packet_sock {
	/* struct sock has to be the first member of packet_sock */
	struct sock		sk;
	struct tpacket_stats	stats;
#ifdef CONFIG_PACKET_MMAP
	char *			*pg_vec;
	unsigned int		head;
	unsigned int            frames_per_block;
	unsigned int		frame_size;
	unsigned int		frame_max;
	int			copy_thresh;
#endif
	struct packet_type	prot_hook;
	spinlock_t		bind_lock;
179
	struct mutex		pg_vec_lock;
180
	unsigned int		running:1,	/* prot_hook is attached*/
181 182
				auxdata:1,
				origdev:1;
L
Linus Torvalds 已提交
183
	int			ifindex;	/* bound device		*/
A
Al Viro 已提交
184
	__be16			num;
L
Linus Torvalds 已提交
185 186 187 188 189 190
	struct packet_mclist	*mclist;
#ifdef CONFIG_PACKET_MMAP
	atomic_t		mapped;
	unsigned int            pg_vec_order;
	unsigned int		pg_vec_pages;
	unsigned int		pg_vec_len;
191 192
	enum tpacket_versions	tp_version;
	unsigned int		tp_hdrlen;
193
	unsigned int		tp_reserve;
L
Linus Torvalds 已提交
194 195 196
#endif
};

197 198 199 200 201 202 203 204 205
struct packet_skb_cb {
	unsigned int origlen;
	union {
		struct sockaddr_pkt pkt;
		struct sockaddr_ll ll;
	} sa;
};

#define PACKET_SKB_CB(__skb)	((struct packet_skb_cb *)((__skb)->cb))
206

L
Linus Torvalds 已提交
207 208
#ifdef CONFIG_PACKET_MMAP

209 210
static void *packet_lookup_frame(struct packet_sock *po, unsigned int position,
				 int status)
L
Linus Torvalds 已提交
211 212
{
	unsigned int pg_vec_pos, frame_offset;
213 214 215 216 217
	union {
		struct tpacket_hdr *h1;
		struct tpacket2_hdr *h2;
		void *raw;
	} h;
L
Linus Torvalds 已提交
218 219 220 221

	pg_vec_pos = position / po->frames_per_block;
	frame_offset = position % po->frames_per_block;

222 223 224
	h.raw = po->pg_vec[pg_vec_pos] + (frame_offset * po->frame_size);
	switch (po->tp_version) {
	case TPACKET_V1:
225 226
		if (status != (h.h1->tp_status ? TP_STATUS_USER :
						TP_STATUS_KERNEL))
227 228 229
			return NULL;
		break;
	case TPACKET_V2:
230 231
		if (status != (h.h2->tp_status ? TP_STATUS_USER :
						TP_STATUS_KERNEL))
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254
			return NULL;
		break;
	}
	return h.raw;
}

static void __packet_set_status(struct packet_sock *po, void *frame, int status)
{
	union {
		struct tpacket_hdr *h1;
		struct tpacket2_hdr *h2;
		void *raw;
	} h;

	h.raw = frame;
	switch (po->tp_version) {
	case TPACKET_V1:
		h.h1->tp_status = status;
		break;
	case TPACKET_V2:
		h.h2->tp_status = status;
		break;
	}
L
Linus Torvalds 已提交
255 256 257 258 259 260 261 262 263 264
}
#endif

static inline struct packet_sock *pkt_sk(struct sock *sk)
{
	return (struct packet_sock *)sk;
}

static void packet_sock_destruct(struct sock *sk)
{
265 266
	WARN_ON(atomic_read(&sk->sk_rmem_alloc));
	WARN_ON(atomic_read(&sk->sk_wmem_alloc));
L
Linus Torvalds 已提交
267 268 269 270 271 272

	if (!sock_flag(sk, SOCK_DEAD)) {
		printk("Attempt to release alive packet socket: %p\n", sk);
		return;
	}

273
	sk_refcnt_debug_dec(sk);
L
Linus Torvalds 已提交
274 275 276
}


277
static const struct proto_ops packet_ops;
L
Linus Torvalds 已提交
278

279
static const struct proto_ops packet_ops_spkt;
L
Linus Torvalds 已提交
280

D
David S. Miller 已提交
281
static int packet_rcv_spkt(struct sk_buff *skb, struct net_device *dev,  struct packet_type *pt, struct net_device *orig_dev)
L
Linus Torvalds 已提交
282 283 284 285 286 287 288 289 290 291
{
	struct sock *sk;
	struct sockaddr_pkt *spkt;

	/*
	 *	When we registered the protocol we saved the socket in the data
	 *	field for just this event.
	 */

	sk = pt->af_packet_priv;
292

L
Linus Torvalds 已提交
293 294 295 296 297 298 299
	/*
	 *	Yank back the headers [hope the device set this
	 *	right or kerboom...]
	 *
	 *	Incoming packets have ll header pulled,
	 *	push it back.
	 *
300
	 *	For outgoing ones skb->data == skb_mac_header(skb)
L
Linus Torvalds 已提交
301 302 303 304 305 306
	 *	so that this procedure is noop.
	 */

	if (skb->pkt_type == PACKET_LOOPBACK)
		goto out;

307
	if (dev_net(dev) != sock_net(sk))
308 309
		goto out;

L
Linus Torvalds 已提交
310 311 312 313 314 315 316
	if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
		goto oom;

	/* drop any routing info */
	dst_release(skb->dst);
	skb->dst = NULL;

317 318 319
	/* drop conntrack reference */
	nf_reset(skb);

320
	spkt = &PACKET_SKB_CB(skb)->sa.pkt;
L
Linus Torvalds 已提交
321

322
	skb_push(skb, skb->data - skb_mac_header(skb));
L
Linus Torvalds 已提交
323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350

	/*
	 *	The SOCK_PACKET socket receives _all_ frames.
	 */

	spkt->spkt_family = dev->type;
	strlcpy(spkt->spkt_device, dev->name, sizeof(spkt->spkt_device));
	spkt->spkt_protocol = skb->protocol;

	/*
	 *	Charge the memory to the socket. This is done specifically
	 *	to prevent sockets using all the memory up.
	 */

	if (sock_queue_rcv_skb(sk,skb) == 0)
		return 0;

out:
	kfree_skb(skb);
oom:
	return 0;
}


/*
 *	Output a raw packet to a device layer. This bypasses all the other
 *	protocol layers and you must therefore supply it with a complete frame
 */
351

L
Linus Torvalds 已提交
352 353 354 355 356 357 358
static int packet_sendmsg_spkt(struct kiocb *iocb, struct socket *sock,
			       struct msghdr *msg, size_t len)
{
	struct sock *sk = sock->sk;
	struct sockaddr_pkt *saddr=(struct sockaddr_pkt *)msg->msg_name;
	struct sk_buff *skb;
	struct net_device *dev;
A
Al Viro 已提交
359
	__be16 proto=0;
L
Linus Torvalds 已提交
360
	int err;
361

L
Linus Torvalds 已提交
362
	/*
363
	 *	Get and verify the address.
L
Linus Torvalds 已提交
364 365 366 367 368 369 370 371 372 373 374 375 376
	 */

	if (saddr)
	{
		if (msg->msg_namelen < sizeof(struct sockaddr))
			return(-EINVAL);
		if (msg->msg_namelen==sizeof(struct sockaddr_pkt))
			proto=saddr->spkt_protocol;
	}
	else
		return(-ENOTCONN);	/* SOCK_PACKET must be sent giving an address */

	/*
377
	 *	Find the device first to size check it
L
Linus Torvalds 已提交
378 379 380
	 */

	saddr->spkt_device[13] = 0;
381
	dev = dev_get_by_name(sock_net(sk), saddr->spkt_device);
L
Linus Torvalds 已提交
382 383 384
	err = -ENODEV;
	if (dev == NULL)
		goto out_unlock;
385

386 387 388 389
	err = -ENETDOWN;
	if (!(dev->flags & IFF_UP))
		goto out_unlock;

L
Linus Torvalds 已提交
390 391 392 393
	/*
	 *	You may not queue a frame bigger than the mtu. This is the lowest level
	 *	raw protocol and you must do your own fragmentation at this level.
	 */
394

L
Linus Torvalds 已提交
395
	err = -EMSGSIZE;
396
	if (len > dev->mtu + dev->hard_header_len)
L
Linus Torvalds 已提交
397 398 399 400 401 402 403 404 405 406
		goto out_unlock;

	err = -ENOBUFS;
	skb = sock_wmalloc(sk, len + LL_RESERVED_SPACE(dev), 0, GFP_KERNEL);

	/*
	 *	If the write buffer is full, then tough. At this level the user gets to
	 *	deal with the problem - do your own algorithmic backoffs. That's far
	 *	more flexible.
	 */
407 408

	if (skb == NULL)
L
Linus Torvalds 已提交
409 410 411
		goto out_unlock;

	/*
412
	 *	Fill it in
L
Linus Torvalds 已提交
413
	 */
414

L
Linus Torvalds 已提交
415 416 417 418 419
	/* FIXME: Save some space for broken drivers that write a
	 * hard header at transmission time by themselves. PPP is the
	 * notable one here. This should really be fixed at the driver level.
	 */
	skb_reserve(skb, LL_RESERVED_SPACE(dev));
420
	skb_reset_network_header(skb);
L
Linus Torvalds 已提交
421 422

	/* Try to align data part correctly */
423
	if (dev->header_ops) {
L
Linus Torvalds 已提交
424 425 426
		skb->data -= dev->hard_header_len;
		skb->tail -= dev->hard_header_len;
		if (len < dev->hard_header_len)
427
			skb_reset_network_header(skb);
L
Linus Torvalds 已提交
428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
	}

	/* Returns -EFAULT on error */
	err = memcpy_fromiovec(skb_put(skb,len), msg->msg_iov, len);
	skb->protocol = proto;
	skb->dev = dev;
	skb->priority = sk->sk_priority;
	if (err)
		goto out_free;

	/*
	 *	Now send it
	 */

	dev_queue_xmit(skb);
	dev_put(dev);
	return(len);

out_free:
	kfree_skb(skb);
out_unlock:
	if (dev)
		dev_put(dev);
	return err;
}

D
David S. Miller 已提交
454 455
static inline unsigned int run_filter(struct sk_buff *skb, struct sock *sk,
				      unsigned int res)
L
Linus Torvalds 已提交
456 457
{
	struct sk_filter *filter;
458 459 460

	rcu_read_lock_bh();
	filter = rcu_dereference(sk->sk_filter);
D
David S. Miller 已提交
461 462
	if (filter != NULL)
		res = sk_run_filter(skb, filter->insns, filter->len);
463
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
464

D
David S. Miller 已提交
465
	return res;
L
Linus Torvalds 已提交
466 467 468 469 470 471 472 473 474 475 476 477 478 479
}

/*
   This function makes lazy skb cloning in hope that most of packets
   are discarded by BPF.

   Note tricky part: we DO mangle shared skb! skb->data, skb->len
   and skb->cb are mangled. It works because (and until) packets
   falling here are owned by current CPU. Output packets are cloned
   by dev_queue_xmit_nit(), input packets are processed by net_bh
   sequencially, so that if we return skb to original state on exit,
   we will not harm anyone.
 */

D
David S. Miller 已提交
480
static int packet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev)
L
Linus Torvalds 已提交
481 482 483 484 485 486
{
	struct sock *sk;
	struct sockaddr_ll *sll;
	struct packet_sock *po;
	u8 * skb_head = skb->data;
	int skb_len = skb->len;
D
David S. Miller 已提交
487
	unsigned int snaplen, res;
L
Linus Torvalds 已提交
488 489 490 491 492 493 494

	if (skb->pkt_type == PACKET_LOOPBACK)
		goto drop;

	sk = pt->af_packet_priv;
	po = pkt_sk(sk);

495
	if (dev_net(dev) != sock_net(sk))
496 497
		goto drop;

L
Linus Torvalds 已提交
498 499
	skb->dev = dev;

500
	if (dev->header_ops) {
L
Linus Torvalds 已提交
501 502 503 504 505 506 507 508
		/* The device has an explicit notion of ll header,
		   exported to higher levels.

		   Otherwise, the device hides datails of it frame
		   structure, so that corresponding packet head
		   never delivered to user.
		 */
		if (sk->sk_type != SOCK_DGRAM)
509
			skb_push(skb, skb->data - skb_mac_header(skb));
L
Linus Torvalds 已提交
510 511
		else if (skb->pkt_type == PACKET_OUTGOING) {
			/* Special case: outgoing packets have ll header at head */
512
			skb_pull(skb, skb_network_offset(skb));
L
Linus Torvalds 已提交
513 514 515 516 517
		}
	}

	snaplen = skb->len;

D
David S. Miller 已提交
518 519
	res = run_filter(skb, sk, snaplen);
	if (!res)
520
		goto drop_n_restore;
D
David S. Miller 已提交
521 522
	if (snaplen > res)
		snaplen = res;
L
Linus Torvalds 已提交
523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540

	if (atomic_read(&sk->sk_rmem_alloc) + skb->truesize >=
	    (unsigned)sk->sk_rcvbuf)
		goto drop_n_acct;

	if (skb_shared(skb)) {
		struct sk_buff *nskb = skb_clone(skb, GFP_ATOMIC);
		if (nskb == NULL)
			goto drop_n_acct;

		if (skb_head != skb->data) {
			skb->data = skb_head;
			skb->len = skb_len;
		}
		kfree_skb(skb);
		skb = nskb;
	}

541 542 543 544
	BUILD_BUG_ON(sizeof(*PACKET_SKB_CB(skb)) + MAX_ADDR_LEN - 8 >
		     sizeof(skb->cb));

	sll = &PACKET_SKB_CB(skb)->sa.ll;
L
Linus Torvalds 已提交
545 546 547 548
	sll->sll_family = AF_PACKET;
	sll->sll_hatype = dev->type;
	sll->sll_protocol = skb->protocol;
	sll->sll_pkttype = skb->pkt_type;
549
	if (unlikely(po->origdev))
550 551 552
		sll->sll_ifindex = orig_dev->ifindex;
	else
		sll->sll_ifindex = dev->ifindex;
L
Linus Torvalds 已提交
553

S
Stephen Hemminger 已提交
554
	sll->sll_halen = dev_parse_header(skb, sll->sll_addr);
L
Linus Torvalds 已提交
555

556
	PACKET_SKB_CB(skb)->origlen = skb->len;
557

L
Linus Torvalds 已提交
558 559 560 561 562 563 564 565
	if (pskb_trim(skb, snaplen))
		goto drop_n_acct;

	skb_set_owner_r(skb, sk);
	skb->dev = NULL;
	dst_release(skb->dst);
	skb->dst = NULL;

566 567 568
	/* drop conntrack reference */
	nf_reset(skb);

L
Linus Torvalds 已提交
569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
	spin_lock(&sk->sk_receive_queue.lock);
	po->stats.tp_packets++;
	__skb_queue_tail(&sk->sk_receive_queue, skb);
	spin_unlock(&sk->sk_receive_queue.lock);
	sk->sk_data_ready(sk, skb->len);
	return 0;

drop_n_acct:
	spin_lock(&sk->sk_receive_queue.lock);
	po->stats.tp_drops++;
	spin_unlock(&sk->sk_receive_queue.lock);

drop_n_restore:
	if (skb_head != skb->data && skb_shared(skb)) {
		skb->data = skb_head;
		skb->len = skb_len;
	}
drop:
587
	consume_skb(skb);
L
Linus Torvalds 已提交
588 589 590 591
	return 0;
}

#ifdef CONFIG_PACKET_MMAP
D
David S. Miller 已提交
592
static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev)
L
Linus Torvalds 已提交
593 594 595 596
{
	struct sock *sk;
	struct packet_sock *po;
	struct sockaddr_ll *sll;
597 598 599 600 601
	union {
		struct tpacket_hdr *h1;
		struct tpacket2_hdr *h2;
		void *raw;
	} h;
L
Linus Torvalds 已提交
602 603
	u8 * skb_head = skb->data;
	int skb_len = skb->len;
D
David S. Miller 已提交
604
	unsigned int snaplen, res;
L
Linus Torvalds 已提交
605
	unsigned long status = TP_STATUS_LOSING|TP_STATUS_USER;
606
	unsigned short macoff, netoff, hdrlen;
L
Linus Torvalds 已提交
607
	struct sk_buff *copy_skb = NULL;
608
	struct timeval tv;
609
	struct timespec ts;
L
Linus Torvalds 已提交
610 611 612 613 614 615 616

	if (skb->pkt_type == PACKET_LOOPBACK)
		goto drop;

	sk = pt->af_packet_priv;
	po = pkt_sk(sk);

617
	if (dev_net(dev) != sock_net(sk))
618 619
		goto drop;

620
	if (dev->header_ops) {
L
Linus Torvalds 已提交
621
		if (sk->sk_type != SOCK_DGRAM)
622
			skb_push(skb, skb->data - skb_mac_header(skb));
L
Linus Torvalds 已提交
623 624
		else if (skb->pkt_type == PACKET_OUTGOING) {
			/* Special case: outgoing packets have ll header at head */
625
			skb_pull(skb, skb_network_offset(skb));
L
Linus Torvalds 已提交
626 627 628
		}
	}

629 630 631
	if (skb->ip_summed == CHECKSUM_PARTIAL)
		status |= TP_STATUS_CSUMNOTREADY;

L
Linus Torvalds 已提交
632 633
	snaplen = skb->len;

D
David S. Miller 已提交
634 635
	res = run_filter(skb, sk, snaplen);
	if (!res)
636
		goto drop_n_restore;
D
David S. Miller 已提交
637 638
	if (snaplen > res)
		snaplen = res;
L
Linus Torvalds 已提交
639 640

	if (sk->sk_type == SOCK_DGRAM) {
641 642
		macoff = netoff = TPACKET_ALIGN(po->tp_hdrlen) + 16 +
				  po->tp_reserve;
L
Linus Torvalds 已提交
643
	} else {
644
		unsigned maclen = skb_network_offset(skb);
645
		netoff = TPACKET_ALIGN(po->tp_hdrlen +
646 647
				       (maclen < 16 ? 16 : maclen)) +
			po->tp_reserve;
L
Linus Torvalds 已提交
648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669
		macoff = netoff - maclen;
	}

	if (macoff + snaplen > po->frame_size) {
		if (po->copy_thresh &&
		    atomic_read(&sk->sk_rmem_alloc) + skb->truesize <
		    (unsigned)sk->sk_rcvbuf) {
			if (skb_shared(skb)) {
				copy_skb = skb_clone(skb, GFP_ATOMIC);
			} else {
				copy_skb = skb_get(skb);
				skb_head = skb->data;
			}
			if (copy_skb)
				skb_set_owner_r(copy_skb, sk);
		}
		snaplen = po->frame_size - macoff;
		if ((int)snaplen < 0)
			snaplen = 0;
	}

	spin_lock(&sk->sk_receive_queue.lock);
670 671
	h.raw = packet_lookup_frame(po, po->head, TP_STATUS_KERNEL);
	if (!h.raw)
L
Linus Torvalds 已提交
672 673 674 675 676 677 678 679 680 681 682
		goto ring_is_full;
	po->head = po->head != po->frame_max ? po->head+1 : 0;
	po->stats.tp_packets++;
	if (copy_skb) {
		status |= TP_STATUS_COPY;
		__skb_queue_tail(&sk->sk_receive_queue, copy_skb);
	}
	if (!po->stats.tp_drops)
		status &= ~TP_STATUS_LOSING;
	spin_unlock(&sk->sk_receive_queue.lock);

683
	skb_copy_bits(skb, 0, h.raw + macoff, snaplen);
L
Linus Torvalds 已提交
684

685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709
	switch (po->tp_version) {
	case TPACKET_V1:
		h.h1->tp_len = skb->len;
		h.h1->tp_snaplen = snaplen;
		h.h1->tp_mac = macoff;
		h.h1->tp_net = netoff;
		if (skb->tstamp.tv64)
			tv = ktime_to_timeval(skb->tstamp);
		else
			do_gettimeofday(&tv);
		h.h1->tp_sec = tv.tv_sec;
		h.h1->tp_usec = tv.tv_usec;
		hdrlen = sizeof(*h.h1);
		break;
	case TPACKET_V2:
		h.h2->tp_len = skb->len;
		h.h2->tp_snaplen = snaplen;
		h.h2->tp_mac = macoff;
		h.h2->tp_net = netoff;
		if (skb->tstamp.tv64)
			ts = ktime_to_timespec(skb->tstamp);
		else
			getnstimeofday(&ts);
		h.h2->tp_sec = ts.tv_sec;
		h.h2->tp_nsec = ts.tv_nsec;
710
		h.h2->tp_vlan_tci = skb->vlan_tci;
711 712 713 714 715
		hdrlen = sizeof(*h.h2);
		break;
	default:
		BUG();
	}
L
Linus Torvalds 已提交
716

717
	sll = h.raw + TPACKET_ALIGN(hdrlen);
S
Stephen Hemminger 已提交
718
	sll->sll_halen = dev_parse_header(skb, sll->sll_addr);
L
Linus Torvalds 已提交
719 720 721 722
	sll->sll_family = AF_PACKET;
	sll->sll_hatype = dev->type;
	sll->sll_protocol = skb->protocol;
	sll->sll_pkttype = skb->pkt_type;
723
	if (unlikely(po->origdev))
724 725 726
		sll->sll_ifindex = orig_dev->ifindex;
	else
		sll->sll_ifindex = dev->ifindex;
L
Linus Torvalds 已提交
727

728
	__packet_set_status(po, h.raw, status);
R
Ralf Baechle 已提交
729
	smp_mb();
L
Linus Torvalds 已提交
730 731 732

	{
		struct page *p_start, *p_end;
733
		u8 *h_end = h.raw + macoff + snaplen - 1;
L
Linus Torvalds 已提交
734

735
		p_start = virt_to_page(h.raw);
L
Linus Torvalds 已提交
736 737 738 739 740 741 742 743 744 745 746 747 748 749 750
		p_end = virt_to_page(h_end);
		while (p_start <= p_end) {
			flush_dcache_page(p_start);
			p_start++;
		}
	}

	sk->sk_data_ready(sk, 0);

drop_n_restore:
	if (skb_head != skb->data && skb_shared(skb)) {
		skb->data = skb_head;
		skb->len = skb_len;
	}
drop:
751
	kfree_skb(skb);
L
Linus Torvalds 已提交
752 753 754 755 756 757 758
	return 0;

ring_is_full:
	po->stats.tp_drops++;
	spin_unlock(&sk->sk_receive_queue.lock);

	sk->sk_data_ready(sk, 0);
759
	kfree_skb(copy_skb);
L
Linus Torvalds 已提交
760 761 762 763 764 765 766 767 768 769 770 771 772
	goto drop_n_restore;
}

#endif


static int packet_sendmsg(struct kiocb *iocb, struct socket *sock,
			  struct msghdr *msg, size_t len)
{
	struct sock *sk = sock->sk;
	struct sockaddr_ll *saddr=(struct sockaddr_ll *)msg->msg_name;
	struct sk_buff *skb;
	struct net_device *dev;
A
Al Viro 已提交
773
	__be16 proto;
L
Linus Torvalds 已提交
774 775 776 777
	unsigned char *addr;
	int ifindex, err, reserve = 0;

	/*
778
	 *	Get and verify the address.
L
Linus Torvalds 已提交
779
	 */
780

L
Linus Torvalds 已提交
781 782 783 784 785 786 787 788 789 790
	if (saddr == NULL) {
		struct packet_sock *po = pkt_sk(sk);

		ifindex	= po->ifindex;
		proto	= po->num;
		addr	= NULL;
	} else {
		err = -EINVAL;
		if (msg->msg_namelen < sizeof(struct sockaddr_ll))
			goto out;
791 792
		if (msg->msg_namelen < (saddr->sll_halen + offsetof(struct sockaddr_ll, sll_addr)))
			goto out;
L
Linus Torvalds 已提交
793 794 795 796 797 798
		ifindex	= saddr->sll_ifindex;
		proto	= saddr->sll_protocol;
		addr	= saddr->sll_addr;
	}


799
	dev = dev_get_by_index(sock_net(sk), ifindex);
L
Linus Torvalds 已提交
800 801 802 803 804 805
	err = -ENXIO;
	if (dev == NULL)
		goto out_unlock;
	if (sock->type == SOCK_RAW)
		reserve = dev->hard_header_len;

806 807 808 809
	err = -ENETDOWN;
	if (!(dev->flags & IFF_UP))
		goto out_unlock;

L
Linus Torvalds 已提交
810 811 812 813
	err = -EMSGSIZE;
	if (len > dev->mtu+reserve)
		goto out_unlock;

814
	skb = sock_alloc_send_skb(sk, len + LL_ALLOCATED_SPACE(dev),
L
Linus Torvalds 已提交
815 816 817 818 819
				msg->msg_flags & MSG_DONTWAIT, &err);
	if (skb==NULL)
		goto out_unlock;

	skb_reserve(skb, LL_RESERVED_SPACE(dev));
820
	skb_reset_network_header(skb);
L
Linus Torvalds 已提交
821

822 823 824 825
	err = -EINVAL;
	if (sock->type == SOCK_DGRAM &&
	    dev_hard_header(skb, dev, ntohs(proto), addr, NULL, len) < 0)
		goto out_free;
L
Linus Torvalds 已提交
826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865

	/* Returns -EFAULT on error */
	err = memcpy_fromiovec(skb_put(skb,len), msg->msg_iov, len);
	if (err)
		goto out_free;

	skb->protocol = proto;
	skb->dev = dev;
	skb->priority = sk->sk_priority;

	/*
	 *	Now send it
	 */

	err = dev_queue_xmit(skb);
	if (err > 0 && (err = net_xmit_errno(err)) != 0)
		goto out_unlock;

	dev_put(dev);

	return(len);

out_free:
	kfree_skb(skb);
out_unlock:
	if (dev)
		dev_put(dev);
out:
	return err;
}

/*
 *	Close a PACKET socket. This is fairly simple. We immediately go
 *	to 'closed' state and remove our protocol entry in the device list.
 */

static int packet_release(struct socket *sock)
{
	struct sock *sk = sock->sk;
	struct packet_sock *po;
866
	struct net *net;
L
Linus Torvalds 已提交
867 868 869 870

	if (!sk)
		return 0;

871
	net = sock_net(sk);
L
Linus Torvalds 已提交
872 873
	po = pkt_sk(sk);

874
	write_lock_bh(&net->packet.sklist_lock);
L
Linus Torvalds 已提交
875
	sk_del_node_init(sk);
876
	sock_prot_inuse_add(net, sk->sk_prot, -1);
877
	write_unlock_bh(&net->packet.sklist_lock);
L
Linus Torvalds 已提交
878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912

	/*
	 *	Unhook packet receive handler.
	 */

	if (po->running) {
		/*
		 *	Remove the protocol hook
		 */
		dev_remove_pack(&po->prot_hook);
		po->running = 0;
		po->num = 0;
		__sock_put(sk);
	}

	packet_flush_mclist(sk);

#ifdef CONFIG_PACKET_MMAP
	if (po->pg_vec) {
		struct tpacket_req req;
		memset(&req, 0, sizeof(req));
		packet_set_ring(sk, &req, 1);
	}
#endif

	/*
	 *	Now the socket is dead. No more input will appear.
	 */

	sock_orphan(sk);
	sock->sk = NULL;

	/* Purge queues */

	skb_queue_purge(&sk->sk_receive_queue);
913
	sk_refcnt_debug_release(sk);
L
Linus Torvalds 已提交
914 915 916 917 918 919 920 921 922

	sock_put(sk);
	return 0;
}

/*
 *	Attach a packet hook.
 */

A
Al Viro 已提交
923
static int packet_do_bind(struct sock *sk, struct net_device *dev, __be16 protocol)
L
Linus Torvalds 已提交
924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950
{
	struct packet_sock *po = pkt_sk(sk);
	/*
	 *	Detach an existing hook if present.
	 */

	lock_sock(sk);

	spin_lock(&po->bind_lock);
	if (po->running) {
		__sock_put(sk);
		po->running = 0;
		po->num = 0;
		spin_unlock(&po->bind_lock);
		dev_remove_pack(&po->prot_hook);
		spin_lock(&po->bind_lock);
	}

	po->num = protocol;
	po->prot_hook.type = protocol;
	po->prot_hook.dev = dev;

	po->ifindex = dev ? dev->ifindex : 0;

	if (protocol == 0)
		goto out_unlock;

951
	if (!dev || (dev->flags & IFF_UP)) {
L
Linus Torvalds 已提交
952 953 954
		dev_add_pack(&po->prot_hook);
		sock_hold(sk);
		po->running = 1;
955 956 957 958
	} else {
		sk->sk_err = ENETDOWN;
		if (!sock_flag(sk, SOCK_DEAD))
			sk->sk_error_report(sk);
L
Linus Torvalds 已提交
959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976
	}

out_unlock:
	spin_unlock(&po->bind_lock);
	release_sock(sk);
	return 0;
}

/*
 *	Bind a packet socket to a device
 */

static int packet_bind_spkt(struct socket *sock, struct sockaddr *uaddr, int addr_len)
{
	struct sock *sk=sock->sk;
	char name[15];
	struct net_device *dev;
	int err = -ENODEV;
977

L
Linus Torvalds 已提交
978 979 980
	/*
	 *	Check legality
	 */
981

982
	if (addr_len != sizeof(struct sockaddr))
L
Linus Torvalds 已提交
983 984 985
		return -EINVAL;
	strlcpy(name,uaddr->sa_data,sizeof(name));

986
	dev = dev_get_by_name(sock_net(sk), name);
L
Linus Torvalds 已提交
987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004
	if (dev) {
		err = packet_do_bind(sk, dev, pkt_sk(sk)->num);
		dev_put(dev);
	}
	return err;
}

static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
{
	struct sockaddr_ll *sll = (struct sockaddr_ll*)uaddr;
	struct sock *sk=sock->sk;
	struct net_device *dev = NULL;
	int err;


	/*
	 *	Check legality
	 */
1005

L
Linus Torvalds 已提交
1006 1007 1008 1009 1010 1011 1012
	if (addr_len < sizeof(struct sockaddr_ll))
		return -EINVAL;
	if (sll->sll_family != AF_PACKET)
		return -EINVAL;

	if (sll->sll_ifindex) {
		err = -ENODEV;
1013
		dev = dev_get_by_index(sock_net(sk), sll->sll_ifindex);
L
Linus Torvalds 已提交
1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031
		if (dev == NULL)
			goto out;
	}
	err = packet_do_bind(sk, dev, sll->sll_protocol ? : pkt_sk(sk)->num);
	if (dev)
		dev_put(dev);

out:
	return err;
}

static struct proto packet_proto = {
	.name	  = "PACKET",
	.owner	  = THIS_MODULE,
	.obj_size = sizeof(struct packet_sock),
};

/*
1032
 *	Create a packet of type SOCK_PACKET.
L
Linus Torvalds 已提交
1033 1034
 */

1035
static int packet_create(struct net *net, struct socket *sock, int protocol)
L
Linus Torvalds 已提交
1036 1037 1038
{
	struct sock *sk;
	struct packet_sock *po;
A
Al Viro 已提交
1039
	__be16 proto = (__force __be16)protocol; /* weird, but documented */
L
Linus Torvalds 已提交
1040 1041 1042 1043
	int err;

	if (!capable(CAP_NET_RAW))
		return -EPERM;
1044 1045
	if (sock->type != SOCK_DGRAM && sock->type != SOCK_RAW &&
	    sock->type != SOCK_PACKET)
L
Linus Torvalds 已提交
1046 1047 1048 1049 1050
		return -ESOCKTNOSUPPORT;

	sock->state = SS_UNCONNECTED;

	err = -ENOBUFS;
1051
	sk = sk_alloc(net, PF_PACKET, GFP_KERNEL, &packet_proto);
L
Linus Torvalds 已提交
1052 1053 1054 1055 1056 1057
	if (sk == NULL)
		goto out;

	sock->ops = &packet_ops;
	if (sock->type == SOCK_PACKET)
		sock->ops = &packet_ops_spkt;
1058

L
Linus Torvalds 已提交
1059 1060 1061 1062
	sock_init_data(sock, sk);

	po = pkt_sk(sk);
	sk->sk_family = PF_PACKET;
A
Al Viro 已提交
1063
	po->num = proto;
L
Linus Torvalds 已提交
1064 1065

	sk->sk_destruct = packet_sock_destruct;
1066
	sk_refcnt_debug_inc(sk);
L
Linus Torvalds 已提交
1067 1068 1069 1070 1071 1072

	/*
	 *	Attach a protocol block
	 */

	spin_lock_init(&po->bind_lock);
1073
	mutex_init(&po->pg_vec_lock);
L
Linus Torvalds 已提交
1074
	po->prot_hook.func = packet_rcv;
1075

L
Linus Torvalds 已提交
1076 1077
	if (sock->type == SOCK_PACKET)
		po->prot_hook.func = packet_rcv_spkt;
1078

L
Linus Torvalds 已提交
1079 1080
	po->prot_hook.af_packet_priv = sk;

A
Al Viro 已提交
1081 1082
	if (proto) {
		po->prot_hook.type = proto;
L
Linus Torvalds 已提交
1083 1084 1085 1086 1087
		dev_add_pack(&po->prot_hook);
		sock_hold(sk);
		po->running = 1;
	}

1088 1089
	write_lock_bh(&net->packet.sklist_lock);
	sk_add_node(sk, &net->packet.sklist);
1090
	sock_prot_inuse_add(net, &packet_proto, 1);
1091
	write_unlock_bh(&net->packet.sklist_lock);
L
Linus Torvalds 已提交
1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
	return(0);
out:
	return err;
}

/*
 *	Pull a packet from our receive queue and hand it to the user.
 *	If necessary we block.
 */

static int packet_recvmsg(struct kiocb *iocb, struct socket *sock,
			  struct msghdr *msg, size_t len, int flags)
{
	struct sock *sk = sock->sk;
	struct sk_buff *skb;
	int copied, err;
1108
	struct sockaddr_ll *sll;
L
Linus Torvalds 已提交
1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131

	err = -EINVAL;
	if (flags & ~(MSG_PEEK|MSG_DONTWAIT|MSG_TRUNC|MSG_CMSG_COMPAT))
		goto out;

#if 0
	/* What error should we return now? EUNATTACH? */
	if (pkt_sk(sk)->ifindex < 0)
		return -ENODEV;
#endif

	/*
	 *	Call the generic datagram receiver. This handles all sorts
	 *	of horrible races and re-entrancy so we can forget about it
	 *	in the protocol layers.
	 *
	 *	Now it will return ENETDOWN, if device have just gone down,
	 *	but then it will block.
	 */

	skb=skb_recv_datagram(sk,flags,flags&MSG_DONTWAIT,&err);

	/*
1132
	 *	An error occurred so return it. Because skb_recv_datagram()
L
Linus Torvalds 已提交
1133 1134 1135 1136
	 *	handles the blocking we don't see and worry about blocking
	 *	retries.
	 */

1137
	if (skb == NULL)
L
Linus Torvalds 已提交
1138 1139
		goto out;

1140 1141 1142 1143 1144
	/*
	 *	If the address length field is there to be filled in, we fill
	 *	it in now.
	 */

1145
	sll = &PACKET_SKB_CB(skb)->sa.ll;
1146 1147 1148 1149 1150
	if (sock->type == SOCK_PACKET)
		msg->msg_namelen = sizeof(struct sockaddr_pkt);
	else
		msg->msg_namelen = sll->sll_halen + offsetof(struct sockaddr_ll, sll_addr);

L
Linus Torvalds 已提交
1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169
	/*
	 *	You lose any data beyond the buffer you gave. If it worries a
	 *	user program they can ask the device for its MTU anyway.
	 */

	copied = skb->len;
	if (copied > len)
	{
		copied=len;
		msg->msg_flags|=MSG_TRUNC;
	}

	err = skb_copy_datagram_iovec(skb, 0, msg->msg_iov, copied);
	if (err)
		goto out_free;

	sock_recv_timestamp(msg, sk, skb);

	if (msg->msg_name)
1170 1171
		memcpy(msg->msg_name, &PACKET_SKB_CB(skb)->sa,
		       msg->msg_namelen);
L
Linus Torvalds 已提交
1172

1173
	if (pkt_sk(sk)->auxdata) {
1174 1175 1176 1177 1178 1179 1180 1181
		struct tpacket_auxdata aux;

		aux.tp_status = TP_STATUS_USER;
		if (skb->ip_summed == CHECKSUM_PARTIAL)
			aux.tp_status |= TP_STATUS_CSUMNOTREADY;
		aux.tp_len = PACKET_SKB_CB(skb)->origlen;
		aux.tp_snaplen = skb->len;
		aux.tp_mac = 0;
1182
		aux.tp_net = skb_network_offset(skb);
1183
		aux.tp_vlan_tci = skb->vlan_tci;
1184 1185

		put_cmsg(msg, SOL_PACKET, PACKET_AUXDATA, sizeof(aux), &aux);
1186 1187
	}

L
Linus Torvalds 已提交
1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209
	/*
	 *	Free or return the buffer as appropriate. Again this
	 *	hides all the races and re-entrancy issues from us.
	 */
	err = (flags&MSG_TRUNC) ? skb->len : copied;

out_free:
	skb_free_datagram(sk, skb);
out:
	return err;
}

static int packet_getname_spkt(struct socket *sock, struct sockaddr *uaddr,
			       int *uaddr_len, int peer)
{
	struct net_device *dev;
	struct sock *sk	= sock->sk;

	if (peer)
		return -EOPNOTSUPP;

	uaddr->sa_family = AF_PACKET;
1210
	dev = dev_get_by_index(sock_net(sk), pkt_sk(sk)->ifindex);
L
Linus Torvalds 已提交
1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234
	if (dev) {
		strlcpy(uaddr->sa_data, dev->name, 15);
		dev_put(dev);
	} else
		memset(uaddr->sa_data, 0, 14);
	*uaddr_len = sizeof(*uaddr);

	return 0;
}

static int packet_getname(struct socket *sock, struct sockaddr *uaddr,
			  int *uaddr_len, int peer)
{
	struct net_device *dev;
	struct sock *sk = sock->sk;
	struct packet_sock *po = pkt_sk(sk);
	struct sockaddr_ll *sll = (struct sockaddr_ll*)uaddr;

	if (peer)
		return -EOPNOTSUPP;

	sll->sll_family = AF_PACKET;
	sll->sll_ifindex = po->ifindex;
	sll->sll_protocol = po->num;
1235
	dev = dev_get_by_index(sock_net(sk), po->ifindex);
L
Linus Torvalds 已提交
1236 1237 1238 1239 1240 1241 1242 1243 1244
	if (dev) {
		sll->sll_hatype = dev->type;
		sll->sll_halen = dev->addr_len;
		memcpy(sll->sll_addr, dev->dev_addr, dev->addr_len);
		dev_put(dev);
	} else {
		sll->sll_hatype = 0;	/* Bad: we have no ARPHRD_UNSPEC */
		sll->sll_halen = 0;
	}
1245
	*uaddr_len = offsetof(struct sockaddr_ll, sll_addr) + sll->sll_halen;
L
Linus Torvalds 已提交
1246 1247 1248 1249

	return 0;
}

1250 1251
static int packet_dev_mc(struct net_device *dev, struct packet_mclist *i,
			 int what)
L
Linus Torvalds 已提交
1252 1253 1254 1255 1256 1257 1258 1259 1260
{
	switch (i->type) {
	case PACKET_MR_MULTICAST:
		if (what > 0)
			dev_mc_add(dev, i->addr, i->alen, 0);
		else
			dev_mc_delete(dev, i->addr, i->alen, 0);
		break;
	case PACKET_MR_PROMISC:
1261
		return dev_set_promiscuity(dev, what);
L
Linus Torvalds 已提交
1262 1263
		break;
	case PACKET_MR_ALLMULTI:
1264
		return dev_set_allmulti(dev, what);
L
Linus Torvalds 已提交
1265 1266 1267
		break;
	default:;
	}
1268
	return 0;
L
Linus Torvalds 已提交
1269 1270 1271 1272 1273 1274 1275 1276 1277 1278
}

static void packet_dev_mclist(struct net_device *dev, struct packet_mclist *i, int what)
{
	for ( ; i; i=i->next) {
		if (i->ifindex == dev->ifindex)
			packet_dev_mc(dev, i, what);
	}
}

1279
static int packet_mc_add(struct sock *sk, struct packet_mreq_max *mreq)
L
Linus Torvalds 已提交
1280 1281 1282 1283 1284 1285 1286 1287 1288
{
	struct packet_sock *po = pkt_sk(sk);
	struct packet_mclist *ml, *i;
	struct net_device *dev;
	int err;

	rtnl_lock();

	err = -ENODEV;
1289
	dev = __dev_get_by_index(sock_net(sk), mreq->mr_ifindex);
L
Linus Torvalds 已提交
1290 1291 1292 1293 1294 1295 1296 1297
	if (!dev)
		goto done;

	err = -EINVAL;
	if (mreq->mr_alen > dev->addr_len)
		goto done;

	err = -ENOBUFS;
1298
	i = kmalloc(sizeof(*i), GFP_KERNEL);
L
Linus Torvalds 已提交
1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321
	if (i == NULL)
		goto done;

	err = 0;
	for (ml = po->mclist; ml; ml = ml->next) {
		if (ml->ifindex == mreq->mr_ifindex &&
		    ml->type == mreq->mr_type &&
		    ml->alen == mreq->mr_alen &&
		    memcmp(ml->addr, mreq->mr_address, ml->alen) == 0) {
			ml->count++;
			/* Free the new element ... */
			kfree(i);
			goto done;
		}
	}

	i->type = mreq->mr_type;
	i->ifindex = mreq->mr_ifindex;
	i->alen = mreq->mr_alen;
	memcpy(i->addr, mreq->mr_address, i->alen);
	i->count = 1;
	i->next = po->mclist;
	po->mclist = i;
1322 1323 1324 1325 1326
	err = packet_dev_mc(dev, i, 1);
	if (err) {
		po->mclist = i->next;
		kfree(i);
	}
L
Linus Torvalds 已提交
1327 1328 1329 1330 1331 1332

done:
	rtnl_unlock();
	return err;
}

1333
static int packet_mc_drop(struct sock *sk, struct packet_mreq_max *mreq)
L
Linus Torvalds 已提交
1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346
{
	struct packet_mclist *ml, **mlp;

	rtnl_lock();

	for (mlp = &pkt_sk(sk)->mclist; (ml = *mlp) != NULL; mlp = &ml->next) {
		if (ml->ifindex == mreq->mr_ifindex &&
		    ml->type == mreq->mr_type &&
		    ml->alen == mreq->mr_alen &&
		    memcmp(ml->addr, mreq->mr_address, ml->alen) == 0) {
			if (--ml->count == 0) {
				struct net_device *dev;
				*mlp = ml->next;
1347
				dev = dev_get_by_index(sock_net(sk), ml->ifindex);
L
Linus Torvalds 已提交
1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374
				if (dev) {
					packet_dev_mc(dev, ml, -1);
					dev_put(dev);
				}
				kfree(ml);
			}
			rtnl_unlock();
			return 0;
		}
	}
	rtnl_unlock();
	return -EADDRNOTAVAIL;
}

static void packet_flush_mclist(struct sock *sk)
{
	struct packet_sock *po = pkt_sk(sk);
	struct packet_mclist *ml;

	if (!po->mclist)
		return;

	rtnl_lock();
	while ((ml = po->mclist) != NULL) {
		struct net_device *dev;

		po->mclist = ml->next;
1375
		if ((dev = dev_get_by_index(sock_net(sk), ml->ifindex)) != NULL) {
L
Linus Torvalds 已提交
1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387
			packet_dev_mc(dev, ml, -1);
			dev_put(dev);
		}
		kfree(ml);
	}
	rtnl_unlock();
}

static int
packet_setsockopt(struct socket *sock, int level, int optname, char __user *optval, int optlen)
{
	struct sock *sk = sock->sk;
1388
	struct packet_sock *po = pkt_sk(sk);
L
Linus Torvalds 已提交
1389 1390 1391 1392 1393 1394
	int ret;

	if (level != SOL_PACKET)
		return -ENOPROTOOPT;

	switch(optname)	{
1395
	case PACKET_ADD_MEMBERSHIP:
L
Linus Torvalds 已提交
1396 1397
	case PACKET_DROP_MEMBERSHIP:
	{
1398 1399 1400 1401
		struct packet_mreq_max mreq;
		int len = optlen;
		memset(&mreq, 0, sizeof(mreq));
		if (len < sizeof(struct packet_mreq))
L
Linus Torvalds 已提交
1402
			return -EINVAL;
1403 1404 1405
		if (len > sizeof(mreq))
			len = sizeof(mreq);
		if (copy_from_user(&mreq,optval,len))
L
Linus Torvalds 已提交
1406
			return -EFAULT;
1407 1408
		if (len < (mreq.mr_alen + offsetof(struct packet_mreq, mr_address)))
			return -EINVAL;
L
Linus Torvalds 已提交
1409 1410 1411 1412 1413 1414
		if (optname == PACKET_ADD_MEMBERSHIP)
			ret = packet_mc_add(sk, &mreq);
		else
			ret = packet_mc_drop(sk, &mreq);
		return ret;
	}
1415

L
Linus Torvalds 已提交
1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438
#ifdef CONFIG_PACKET_MMAP
	case PACKET_RX_RING:
	{
		struct tpacket_req req;

		if (optlen<sizeof(req))
			return -EINVAL;
		if (copy_from_user(&req,optval,sizeof(req)))
			return -EFAULT;
		return packet_set_ring(sk, &req, 0);
	}
	case PACKET_COPY_THRESH:
	{
		int val;

		if (optlen!=sizeof(val))
			return -EINVAL;
		if (copy_from_user(&val,optval,sizeof(val)))
			return -EFAULT;

		pkt_sk(sk)->copy_thresh = val;
		return 0;
	}
1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457
	case PACKET_VERSION:
	{
		int val;

		if (optlen != sizeof(val))
			return -EINVAL;
		if (po->pg_vec)
			return -EBUSY;
		if (copy_from_user(&val, optval, sizeof(val)))
			return -EFAULT;
		switch (val) {
		case TPACKET_V1:
		case TPACKET_V2:
			po->tp_version = val;
			return 0;
		default:
			return -EINVAL;
		}
	}
1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470
	case PACKET_RESERVE:
	{
		unsigned int val;

		if (optlen != sizeof(val))
			return -EINVAL;
		if (po->pg_vec)
			return -EBUSY;
		if (copy_from_user(&val, optval, sizeof(val)))
			return -EFAULT;
		po->tp_reserve = val;
		return 0;
	}
L
Linus Torvalds 已提交
1471
#endif
1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483
	case PACKET_AUXDATA:
	{
		int val;

		if (optlen < sizeof(val))
			return -EINVAL;
		if (copy_from_user(&val, optval, sizeof(val)))
			return -EFAULT;

		po->auxdata = !!val;
		return 0;
	}
1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495
	case PACKET_ORIGDEV:
	{
		int val;

		if (optlen < sizeof(val))
			return -EINVAL;
		if (copy_from_user(&val, optval, sizeof(val)))
			return -EFAULT;

		po->origdev = !!val;
		return 0;
	}
L
Linus Torvalds 已提交
1496 1497 1498 1499 1500 1501 1502 1503 1504
	default:
		return -ENOPROTOOPT;
	}
}

static int packet_getsockopt(struct socket *sock, int level, int optname,
			     char __user *optval, int __user *optlen)
{
	int len;
1505
	int val;
L
Linus Torvalds 已提交
1506 1507
	struct sock *sk = sock->sk;
	struct packet_sock *po = pkt_sk(sk);
1508 1509
	void *data;
	struct tpacket_stats st;
L
Linus Torvalds 已提交
1510 1511 1512 1513

	if (level != SOL_PACKET)
		return -ENOPROTOOPT;

1514 1515
	if (get_user(len, optlen))
		return -EFAULT;
L
Linus Torvalds 已提交
1516 1517 1518

	if (len < 0)
		return -EINVAL;
1519

L
Linus Torvalds 已提交
1520 1521 1522 1523 1524 1525 1526 1527 1528 1529
	switch(optname)	{
	case PACKET_STATISTICS:
		if (len > sizeof(struct tpacket_stats))
			len = sizeof(struct tpacket_stats);
		spin_lock_bh(&sk->sk_receive_queue.lock);
		st = po->stats;
		memset(&po->stats, 0, sizeof(st));
		spin_unlock_bh(&sk->sk_receive_queue.lock);
		st.tp_packets += st.tp_drops;

1530 1531 1532 1533 1534 1535 1536
		data = &st;
		break;
	case PACKET_AUXDATA:
		if (len > sizeof(int))
			len = sizeof(int);
		val = po->auxdata;

1537 1538 1539 1540 1541 1542 1543
		data = &val;
		break;
	case PACKET_ORIGDEV:
		if (len > sizeof(int))
			len = sizeof(int);
		val = po->origdev;

1544
		data = &val;
L
Linus Torvalds 已提交
1545
		break;
1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569
#ifdef CONFIG_PACKET_MMAP
	case PACKET_VERSION:
		if (len > sizeof(int))
			len = sizeof(int);
		val = po->tp_version;
		data = &val;
		break;
	case PACKET_HDRLEN:
		if (len > sizeof(int))
			len = sizeof(int);
		if (copy_from_user(&val, optval, len))
			return -EFAULT;
		switch (val) {
		case TPACKET_V1:
			val = sizeof(struct tpacket_hdr);
			break;
		case TPACKET_V2:
			val = sizeof(struct tpacket2_hdr);
			break;
		default:
			return -EINVAL;
		}
		data = &val;
		break;
1570 1571 1572 1573 1574 1575
	case PACKET_RESERVE:
		if (len > sizeof(unsigned int))
			len = sizeof(unsigned int);
		val = po->tp_reserve;
		data = &val;
		break;
1576
#endif
L
Linus Torvalds 已提交
1577 1578 1579 1580
	default:
		return -ENOPROTOOPT;
	}

1581 1582
	if (put_user(len, optlen))
		return -EFAULT;
1583 1584
	if (copy_to_user(optval, data, len))
		return -EFAULT;
1585
	return 0;
L
Linus Torvalds 已提交
1586 1587 1588 1589 1590 1591 1592
}


static int packet_notifier(struct notifier_block *this, unsigned long msg, void *data)
{
	struct sock *sk;
	struct hlist_node *node;
1593
	struct net_device *dev = data;
1594
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
1595

1596 1597
	read_lock(&net->packet.sklist_lock);
	sk_for_each(sk, node, &net->packet.sklist) {
L
Linus Torvalds 已提交
1598 1599 1600 1601 1602 1603
		struct packet_sock *po = pkt_sk(sk);

		switch (msg) {
		case NETDEV_UNREGISTER:
			if (po->mclist)
				packet_dev_mclist(dev, po->mclist, -1);
1604 1605
			/* fallthrough */

L
Linus Torvalds 已提交
1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635
		case NETDEV_DOWN:
			if (dev->ifindex == po->ifindex) {
				spin_lock(&po->bind_lock);
				if (po->running) {
					__dev_remove_pack(&po->prot_hook);
					__sock_put(sk);
					po->running = 0;
					sk->sk_err = ENETDOWN;
					if (!sock_flag(sk, SOCK_DEAD))
						sk->sk_error_report(sk);
				}
				if (msg == NETDEV_UNREGISTER) {
					po->ifindex = -1;
					po->prot_hook.dev = NULL;
				}
				spin_unlock(&po->bind_lock);
			}
			break;
		case NETDEV_UP:
			spin_lock(&po->bind_lock);
			if (dev->ifindex == po->ifindex && po->num &&
			    !po->running) {
				dev_add_pack(&po->prot_hook);
				sock_hold(sk);
				po->running = 1;
			}
			spin_unlock(&po->bind_lock);
			break;
		}
	}
1636
	read_unlock(&net->packet.sklist_lock);
L
Linus Torvalds 已提交
1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665
	return NOTIFY_DONE;
}


static int packet_ioctl(struct socket *sock, unsigned int cmd,
			unsigned long arg)
{
	struct sock *sk = sock->sk;

	switch(cmd) {
		case SIOCOUTQ:
		{
			int amount = atomic_read(&sk->sk_wmem_alloc);
			return put_user(amount, (int __user *)arg);
		}
		case SIOCINQ:
		{
			struct sk_buff *skb;
			int amount = 0;

			spin_lock_bh(&sk->sk_receive_queue.lock);
			skb = skb_peek(&sk->sk_receive_queue);
			if (skb)
				amount = skb->len;
			spin_unlock_bh(&sk->sk_receive_queue.lock);
			return put_user(amount, (int __user *)arg);
		}
		case SIOCGSTAMP:
			return sock_get_timestamp(sk, (struct timeval __user *)arg);
1666 1667
		case SIOCGSTAMPNS:
			return sock_get_timestampns(sk, (struct timespec __user *)arg);
1668

L
Linus Torvalds 已提交
1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683
#ifdef CONFIG_INET
		case SIOCADDRT:
		case SIOCDELRT:
		case SIOCDARP:
		case SIOCGARP:
		case SIOCSARP:
		case SIOCGIFADDR:
		case SIOCSIFADDR:
		case SIOCGIFBRDADDR:
		case SIOCSIFBRDADDR:
		case SIOCGIFNETMASK:
		case SIOCSIFNETMASK:
		case SIOCGIFDSTADDR:
		case SIOCSIFDSTADDR:
		case SIOCSIFFLAGS:
1684
			if (!net_eq(sock_net(sk), &init_net))
1685
				return -ENOIOCTLCMD;
L
Linus Torvalds 已提交
1686 1687 1688 1689
			return inet_dgram_ops.ioctl(sock, cmd, arg);
#endif

		default:
1690
			return -ENOIOCTLCMD;
L
Linus Torvalds 已提交
1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710
	}
	return 0;
}

#ifndef CONFIG_PACKET_MMAP
#define packet_mmap sock_no_mmap
#define packet_poll datagram_poll
#else

static unsigned int packet_poll(struct file * file, struct socket *sock,
				poll_table *wait)
{
	struct sock *sk = sock->sk;
	struct packet_sock *po = pkt_sk(sk);
	unsigned int mask = datagram_poll(file, sock, wait);

	spin_lock_bh(&sk->sk_receive_queue.lock);
	if (po->pg_vec) {
		unsigned last = po->head ? po->head-1 : po->frame_max;

1711
		if (packet_lookup_frame(po, last, TP_STATUS_USER))
L
Linus Torvalds 已提交
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725
			mask |= POLLIN | POLLRDNORM;
	}
	spin_unlock_bh(&sk->sk_receive_queue.lock);
	return mask;
}


/* Dirty? Well, I still did not learn better way to account
 * for user mmaps.
 */

static void packet_mm_open(struct vm_area_struct *vma)
{
	struct file *file = vma->vm_file;
1726
	struct socket * sock = file->private_data;
L
Linus Torvalds 已提交
1727
	struct sock *sk = sock->sk;
1728

L
Linus Torvalds 已提交
1729 1730 1731 1732 1733 1734 1735
	if (sk)
		atomic_inc(&pkt_sk(sk)->mapped);
}

static void packet_mm_close(struct vm_area_struct *vma)
{
	struct file *file = vma->vm_file;
1736
	struct socket * sock = file->private_data;
L
Linus Torvalds 已提交
1737
	struct sock *sk = sock->sk;
1738

L
Linus Torvalds 已提交
1739 1740 1741 1742 1743 1744 1745 1746 1747
	if (sk)
		atomic_dec(&pkt_sk(sk)->mapped);
}

static struct vm_operations_struct packet_mmap_ops = {
	.open =	packet_mm_open,
	.close =packet_mm_close,
};

1748
static void free_pg_vec(char **pg_vec, unsigned int order, unsigned int len)
L
Linus Torvalds 已提交
1749 1750 1751
{
	int i;

1752 1753 1754
	for (i = 0; i < len; i++) {
		if (likely(pg_vec[i]))
			free_pages((unsigned long) pg_vec[i], order);
L
Linus Torvalds 已提交
1755 1756 1757 1758
	}
	kfree(pg_vec);
}

1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788
static inline char *alloc_one_pg_vec_page(unsigned long order)
{
	return (char *) __get_free_pages(GFP_KERNEL | __GFP_COMP | __GFP_ZERO,
					 order);
}

static char **alloc_pg_vec(struct tpacket_req *req, int order)
{
	unsigned int block_nr = req->tp_block_nr;
	char **pg_vec;
	int i;

	pg_vec = kzalloc(block_nr * sizeof(char *), GFP_KERNEL);
	if (unlikely(!pg_vec))
		goto out;

	for (i = 0; i < block_nr; i++) {
		pg_vec[i] = alloc_one_pg_vec_page(order);
		if (unlikely(!pg_vec[i]))
			goto out_free_pgvec;
	}

out:
	return pg_vec;

out_free_pgvec:
	free_pg_vec(pg_vec, order, block_nr);
	pg_vec = NULL;
	goto out;
}
L
Linus Torvalds 已提交
1789 1790 1791 1792 1793

static int packet_set_ring(struct sock *sk, struct tpacket_req *req, int closing)
{
	char **pg_vec = NULL;
	struct packet_sock *po = pkt_sk(sk);
A
Al Viro 已提交
1794 1795
	int was_running, order = 0;
	__be16 num;
L
Linus Torvalds 已提交
1796
	int err = 0;
1797

L
Linus Torvalds 已提交
1798
	if (req->tp_block_nr) {
J
Jiri Olsa 已提交
1799
		int i;
L
Linus Torvalds 已提交
1800 1801 1802

		/* Sanity tests and some calculations */

1803
		if (unlikely(po->pg_vec))
L
Linus Torvalds 已提交
1804 1805
			return -EBUSY;

1806 1807 1808 1809 1810 1811 1812 1813 1814
		switch (po->tp_version) {
		case TPACKET_V1:
			po->tp_hdrlen = TPACKET_HDRLEN;
			break;
		case TPACKET_V2:
			po->tp_hdrlen = TPACKET2_HDRLEN;
			break;
		}

1815
		if (unlikely((int)req->tp_block_size <= 0))
L
Linus Torvalds 已提交
1816
			return -EINVAL;
1817
		if (unlikely(req->tp_block_size & (PAGE_SIZE - 1)))
L
Linus Torvalds 已提交
1818
			return -EINVAL;
1819 1820
		if (unlikely(req->tp_frame_size < po->tp_hdrlen +
						  po->tp_reserve))
L
Linus Torvalds 已提交
1821
			return -EINVAL;
1822
		if (unlikely(req->tp_frame_size & (TPACKET_ALIGNMENT - 1)))
L
Linus Torvalds 已提交
1823 1824 1825
			return -EINVAL;

		po->frames_per_block = req->tp_block_size/req->tp_frame_size;
1826
		if (unlikely(po->frames_per_block <= 0))
L
Linus Torvalds 已提交
1827
			return -EINVAL;
1828 1829
		if (unlikely((po->frames_per_block * req->tp_block_nr) !=
			     req->tp_frame_nr))
L
Linus Torvalds 已提交
1830 1831 1832
			return -EINVAL;

		err = -ENOMEM;
1833 1834 1835
		order = get_order(req->tp_block_size);
		pg_vec = alloc_pg_vec(req, order);
		if (unlikely(!pg_vec))
L
Linus Torvalds 已提交
1836 1837
			goto out;

1838
		for (i = 0; i < req->tp_block_nr; i++) {
1839
			void *ptr = pg_vec[i];
L
Linus Torvalds 已提交
1840 1841
			int k;

1842
			for (k = 0; k < po->frames_per_block; k++) {
1843
				__packet_set_status(po, ptr, TP_STATUS_KERNEL);
L
Linus Torvalds 已提交
1844 1845 1846 1847 1848
				ptr += req->tp_frame_size;
			}
		}
		/* Done */
	} else {
1849
		if (unlikely(req->tp_frame_nr))
L
Linus Torvalds 已提交
1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865
			return -EINVAL;
	}

	lock_sock(sk);

	/* Detach socket from network */
	spin_lock(&po->bind_lock);
	was_running = po->running;
	num = po->num;
	if (was_running) {
		__dev_remove_pack(&po->prot_hook);
		po->num = 0;
		po->running = 0;
		__sock_put(sk);
	}
	spin_unlock(&po->bind_lock);
1866

L
Linus Torvalds 已提交
1867 1868 1869
	synchronize_net();

	err = -EBUSY;
1870
	mutex_lock(&po->pg_vec_lock);
L
Linus Torvalds 已提交
1871 1872 1873 1874 1875 1876
	if (closing || atomic_read(&po->mapped) == 0) {
		err = 0;
#define XC(a, b) ({ __typeof__ ((a)) __t; __t = (a); (a) = (b); __t; })

		spin_lock_bh(&sk->sk_receive_queue.lock);
		pg_vec = XC(po->pg_vec, pg_vec);
1877
		po->frame_max = (req->tp_frame_nr - 1);
L
Linus Torvalds 已提交
1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891
		po->head = 0;
		po->frame_size = req->tp_frame_size;
		spin_unlock_bh(&sk->sk_receive_queue.lock);

		order = XC(po->pg_vec_order, order);
		req->tp_block_nr = XC(po->pg_vec_len, req->tp_block_nr);

		po->pg_vec_pages = req->tp_block_size/PAGE_SIZE;
		po->prot_hook.func = po->pg_vec ? tpacket_rcv : packet_rcv;
		skb_queue_purge(&sk->sk_receive_queue);
#undef XC
		if (atomic_read(&po->mapped))
			printk(KERN_DEBUG "packet_mmap: vma is busy: %d\n", atomic_read(&po->mapped));
	}
1892
	mutex_unlock(&po->pg_vec_lock);
L
Linus Torvalds 已提交
1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924

	spin_lock(&po->bind_lock);
	if (was_running && !po->running) {
		sock_hold(sk);
		po->running = 1;
		po->num = num;
		dev_add_pack(&po->prot_hook);
	}
	spin_unlock(&po->bind_lock);

	release_sock(sk);

	if (pg_vec)
		free_pg_vec(pg_vec, order, req->tp_block_nr);
out:
	return err;
}

static int packet_mmap(struct file *file, struct socket *sock, struct vm_area_struct *vma)
{
	struct sock *sk = sock->sk;
	struct packet_sock *po = pkt_sk(sk);
	unsigned long size;
	unsigned long start;
	int err = -EINVAL;
	int i;

	if (vma->vm_pgoff)
		return -EINVAL;

	size = vma->vm_end - vma->vm_start;

1925
	mutex_lock(&po->pg_vec_lock);
L
Linus Torvalds 已提交
1926 1927 1928 1929 1930 1931
	if (po->pg_vec == NULL)
		goto out;
	if (size != po->pg_vec_len*po->pg_vec_pages*PAGE_SIZE)
		goto out;

	start = vma->vm_start;
1932 1933 1934 1935 1936 1937 1938 1939 1940 1941
	for (i = 0; i < po->pg_vec_len; i++) {
		struct page *page = virt_to_page(po->pg_vec[i]);
		int pg_num;

		for (pg_num = 0; pg_num < po->pg_vec_pages; pg_num++, page++) {
			err = vm_insert_page(vma, start, page);
			if (unlikely(err))
				goto out;
			start += PAGE_SIZE;
		}
L
Linus Torvalds 已提交
1942
	}
1943
	atomic_inc(&po->mapped);
L
Linus Torvalds 已提交
1944 1945 1946 1947
	vma->vm_ops = &packet_mmap_ops;
	err = 0;

out:
1948
	mutex_unlock(&po->pg_vec_lock);
L
Linus Torvalds 已提交
1949 1950 1951 1952 1953
	return err;
}
#endif


1954
static const struct proto_ops packet_ops_spkt = {
L
Linus Torvalds 已提交
1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974
	.family =	PF_PACKET,
	.owner =	THIS_MODULE,
	.release =	packet_release,
	.bind =		packet_bind_spkt,
	.connect =	sock_no_connect,
	.socketpair =	sock_no_socketpair,
	.accept =	sock_no_accept,
	.getname =	packet_getname_spkt,
	.poll =		datagram_poll,
	.ioctl =	packet_ioctl,
	.listen =	sock_no_listen,
	.shutdown =	sock_no_shutdown,
	.setsockopt =	sock_no_setsockopt,
	.getsockopt =	sock_no_getsockopt,
	.sendmsg =	packet_sendmsg_spkt,
	.recvmsg =	packet_recvmsg,
	.mmap =		sock_no_mmap,
	.sendpage =	sock_no_sendpage,
};

1975
static const struct proto_ops packet_ops = {
L
Linus Torvalds 已提交
1976 1977 1978 1979 1980 1981 1982
	.family =	PF_PACKET,
	.owner =	THIS_MODULE,
	.release =	packet_release,
	.bind =		packet_bind,
	.connect =	sock_no_connect,
	.socketpair =	sock_no_socketpair,
	.accept =	sock_no_accept,
1983
	.getname =	packet_getname,
L
Linus Torvalds 已提交
1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
	.poll =		packet_poll,
	.ioctl =	packet_ioctl,
	.listen =	sock_no_listen,
	.shutdown =	sock_no_shutdown,
	.setsockopt =	packet_setsockopt,
	.getsockopt =	packet_getsockopt,
	.sendmsg =	packet_sendmsg,
	.recvmsg =	packet_recvmsg,
	.mmap =		packet_mmap,
	.sendpage =	sock_no_sendpage,
};

static struct net_proto_family packet_family_ops = {
	.family =	PF_PACKET,
	.create =	packet_create,
	.owner	=	THIS_MODULE,
};

static struct notifier_block packet_netdev_notifier = {
	.notifier_call =packet_notifier,
};

#ifdef CONFIG_PROC_FS
2007
static inline struct sock *packet_seq_idx(struct net *net, loff_t off)
L
Linus Torvalds 已提交
2008 2009 2010 2011
{
	struct sock *s;
	struct hlist_node *node;

2012
	sk_for_each(s, node, &net->packet.sklist) {
L
Linus Torvalds 已提交
2013 2014 2015 2016 2017 2018 2019
		if (!off--)
			return s;
	}
	return NULL;
}

static void *packet_seq_start(struct seq_file *seq, loff_t *pos)
2020
	__acquires(seq_file_net(seq)->packet.sklist_lock)
L
Linus Torvalds 已提交
2021
{
2022
	struct net *net = seq_file_net(seq);
2023
	read_lock(&net->packet.sklist_lock);
2024
	return *pos ? packet_seq_idx(net, *pos - 1) : SEQ_START_TOKEN;
L
Linus Torvalds 已提交
2025 2026 2027 2028
}

static void *packet_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
2029
	struct net *net = seq_file_net(seq);
L
Linus Torvalds 已提交
2030
	++*pos;
2031
	return  (v == SEQ_START_TOKEN)
2032
		? sk_head(&net->packet.sklist)
L
Linus Torvalds 已提交
2033 2034 2035 2036
		: sk_next((struct sock*)v) ;
}

static void packet_seq_stop(struct seq_file *seq, void *v)
2037
	__releases(seq_file_net(seq)->packet.sklist_lock)
L
Linus Torvalds 已提交
2038
{
2039
	struct net *net = seq_file_net(seq);
2040
	read_unlock(&net->packet.sklist_lock);
L
Linus Torvalds 已提交
2041 2042
}

2043
static int packet_seq_show(struct seq_file *seq, void *v)
L
Linus Torvalds 已提交
2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066
{
	if (v == SEQ_START_TOKEN)
		seq_puts(seq, "sk       RefCnt Type Proto  Iface R Rmem   User   Inode\n");
	else {
		struct sock *s = v;
		const struct packet_sock *po = pkt_sk(s);

		seq_printf(seq,
			   "%p %-6d %-4d %04x   %-5d %1d %-6u %-6u %-6lu\n",
			   s,
			   atomic_read(&s->sk_refcnt),
			   s->sk_type,
			   ntohs(po->num),
			   po->ifindex,
			   po->running,
			   atomic_read(&s->sk_rmem_alloc),
			   sock_i_uid(s),
			   sock_i_ino(s) );
	}

	return 0;
}

2067
static const struct seq_operations packet_seq_ops = {
L
Linus Torvalds 已提交
2068 2069 2070 2071 2072 2073 2074 2075
	.start	= packet_seq_start,
	.next	= packet_seq_next,
	.stop	= packet_seq_stop,
	.show	= packet_seq_show,
};

static int packet_seq_open(struct inode *inode, struct file *file)
{
2076 2077
	return seq_open_net(inode, file, &packet_seq_ops,
			    sizeof(struct seq_net_private));
L
Linus Torvalds 已提交
2078 2079
}

2080
static const struct file_operations packet_seq_fops = {
L
Linus Torvalds 已提交
2081 2082 2083 2084
	.owner		= THIS_MODULE,
	.open		= packet_seq_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
2085
	.release	= seq_release_net,
L
Linus Torvalds 已提交
2086 2087 2088 2089
};

#endif

2090 2091
static int packet_net_init(struct net *net)
{
2092 2093
	rwlock_init(&net->packet.sklist_lock);
	INIT_HLIST_HEAD(&net->packet.sklist);
2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111

	if (!proc_net_fops_create(net, "packet", 0, &packet_seq_fops))
		return -ENOMEM;

	return 0;
}

static void packet_net_exit(struct net *net)
{
	proc_net_remove(net, "packet");
}

static struct pernet_operations packet_net_ops = {
	.init = packet_net_init,
	.exit = packet_net_exit,
};


L
Linus Torvalds 已提交
2112 2113 2114
static void __exit packet_exit(void)
{
	unregister_netdevice_notifier(&packet_netdev_notifier);
2115
	unregister_pernet_subsys(&packet_net_ops);
L
Linus Torvalds 已提交
2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127
	sock_unregister(PF_PACKET);
	proto_unregister(&packet_proto);
}

static int __init packet_init(void)
{
	int rc = proto_register(&packet_proto, 0);

	if (rc != 0)
		goto out;

	sock_register(&packet_family_ops);
2128
	register_pernet_subsys(&packet_net_ops);
L
Linus Torvalds 已提交
2129 2130 2131 2132 2133 2134 2135 2136 2137
	register_netdevice_notifier(&packet_netdev_notifier);
out:
	return rc;
}

module_init(packet_init);
module_exit(packet_exit);
MODULE_LICENSE("GPL");
MODULE_ALIAS_NETPROTO(PF_PACKET);