tun.c 38.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
/*
 *  TUN - Universal TUN/TAP device driver.
 *  Copyright (C) 1999-2002 Maxim Krasnyansky <maxk@qualcomm.com>
 *
 *  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.
 *
 *  This program is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
 *  GNU General Public License for more details.
 *
 *  $Id: tun.c,v 1.15 2002/03/01 02:44:24 maxk Exp $
 */

/*
 *  Changes:
 *
21 22 23
 *  Mike Kershaw <dragorn@kismetwireless.net> 2005/08/14
 *    Add TUNSETLINK ioctl to set the link encapsulation
 *
L
Linus Torvalds 已提交
24
 *  Mark Smith <markzzzsmith@yahoo.com.au>
J
Joe Perches 已提交
25
 *    Use eth_random_addr() for tap MAC address.
L
Linus Torvalds 已提交
26 27 28 29 30 31 32 33 34 35 36
 *
 *  Harald Roelle <harald.roelle@ifi.lmu.de>  2004/04/20
 *    Fixes in packet dropping, queue length setting and queue wakeup.
 *    Increased default tx queue length.
 *    Added ethtool API.
 *    Minor cleanups
 *
 *  Daniel Podlejski <underley@underley.eu.org>
 *    Modifications for 2.3.99-pre5 kernel.
 */

37 38
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

L
Linus Torvalds 已提交
39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
#define DRV_NAME	"tun"
#define DRV_VERSION	"1.6"
#define DRV_DESCRIPTION	"Universal TUN/TAP device driver"
#define DRV_COPYRIGHT	"(C) 1999-2004 Max Krasnyansky <maxk@qualcomm.com>"

#include <linux/module.h>
#include <linux/errno.h>
#include <linux/kernel.h>
#include <linux/major.h>
#include <linux/slab.h>
#include <linux/poll.h>
#include <linux/fcntl.h>
#include <linux/init.h>
#include <linux/skbuff.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/miscdevice.h>
#include <linux/ethtool.h>
#include <linux/rtnetlink.h>
58
#include <linux/compat.h>
L
Linus Torvalds 已提交
59 60 61 62 63
#include <linux/if.h>
#include <linux/if_arp.h>
#include <linux/if_ether.h>
#include <linux/if_tun.h>
#include <linux/crc32.h>
64
#include <linux/nsproxy.h>
65
#include <linux/virtio_net.h>
M
Michael S. Tsirkin 已提交
66
#include <linux/rcupdate.h>
67
#include <net/net_namespace.h>
68
#include <net/netns/generic.h>
69
#include <net/rtnetlink.h>
70
#include <net/sock.h>
L
Linus Torvalds 已提交
71 72 73

#include <asm/uaccess.h>

74 75 76
/* Uncomment to enable debugging */
/* #define TUN_DEBUG 1 */

L
Linus Torvalds 已提交
77 78
#ifdef TUN_DEBUG
static int debug;
79

80 81 82 83 84 85 86 87 88 89
#define tun_debug(level, tun, fmt, args...)			\
do {								\
	if (tun->debug)						\
		netdev_printk(level, tun->dev, fmt, ##args);	\
} while (0)
#define DBG1(level, fmt, args...)				\
do {								\
	if (debug == 2)						\
		printk(level fmt, ##args);			\
} while (0)
90
#else
91 92 93 94 95 96 97 98 99 100
#define tun_debug(level, tun, fmt, args...)			\
do {								\
	if (0)							\
		netdev_printk(level, tun->dev, fmt, ##args);	\
} while (0)
#define DBG1(level, fmt, args...)				\
do {								\
	if (0)							\
		printk(level fmt, ##args);			\
} while (0)
101 102
#endif

103 104 105 106 107 108 109
#define FLT_EXACT_COUNT 8
struct tap_filter {
	unsigned int    count;    /* Number of addrs. Zero means disabled */
	u32             mask[2];  /* Mask of the hashed addrs */
	unsigned char	addr[FLT_EXACT_COUNT][ETH_ALEN];
};

E
Eric W. Biederman 已提交
110
struct tun_file {
111
	atomic_t count;
E
Eric W. Biederman 已提交
112
	struct tun_struct *tun;
E
Eric W. Biederman 已提交
113
	struct net *net;
E
Eric W. Biederman 已提交
114 115
};

116 117
struct tun_sock;

118
struct tun_struct {
E
Eric W. Biederman 已提交
119
	struct tun_file		*tfile;
120
	unsigned int 		flags;
121 122 123 124
	uid_t			owner;
	gid_t			group;

	struct net_device	*dev;
125
	netdev_features_t	set_features;
126 127
#define TUN_USER_FEATURES (NETIF_F_HW_CSUM|NETIF_F_TSO_ECN|NETIF_F_TSO| \
			  NETIF_F_TSO6|NETIF_F_UFO)
128
	struct fasync_struct	*fasync;
129

130
	struct tap_filter       txflt;
131
	struct socket		socket;
132
	struct socket_wq	wq;
133 134 135

	int			vnet_hdr_sz;

136 137
#ifdef TUN_DEBUG
	int debug;
L
Linus Torvalds 已提交
138
#endif
139
};
L
Linus Torvalds 已提交
140

141 142 143 144 145 146 147 148 149 150
struct tun_sock {
	struct sock		sk;
	struct tun_struct	*tun;
};

static inline struct tun_sock *tun_sk(struct sock *sk)
{
	return container_of(sk, struct tun_sock, sk);
}

151 152
static int tun_attach(struct tun_struct *tun, struct file *file)
{
E
Eric W. Biederman 已提交
153
	struct tun_file *tfile = file->private_data;
154
	int err;
155 156 157

	ASSERT_RTNL();

158 159 160 161 162 163 164 165 166 167 168
	netif_tx_lock_bh(tun->dev);

	err = -EINVAL;
	if (tfile->tun)
		goto out;

	err = -EBUSY;
	if (tun->tfile)
		goto out;

	err = 0;
E
Eric W. Biederman 已提交
169 170
	tfile->tun = tun;
	tun->tfile = tfile;
171
	tun->socket.file = file;
172
	netif_carrier_on(tun->dev);
173
	dev_hold(tun->dev);
174
	sock_hold(tun->socket.sk);
175
	atomic_inc(&tfile->count);
176

177 178 179
out:
	netif_tx_unlock_bh(tun->dev);
	return err;
180 181
}

E
Eric W. Biederman 已提交
182 183 184
static void __tun_detach(struct tun_struct *tun)
{
	/* Detach from net device */
185
	netif_tx_lock_bh(tun->dev);
186
	netif_carrier_off(tun->dev);
E
Eric W. Biederman 已提交
187
	tun->tfile = NULL;
188
	tun->socket.file = NULL;
189
	netif_tx_unlock_bh(tun->dev);
E
Eric W. Biederman 已提交
190 191

	/* Drop read queue */
192
	skb_queue_purge(&tun->socket.sk->sk_receive_queue);
193 194 195 196 197 198 199 200 201 202

	/* Drop the extra count on the net device */
	dev_put(tun->dev);
}

static void tun_detach(struct tun_struct *tun)
{
	rtnl_lock();
	__tun_detach(tun);
	rtnl_unlock();
E
Eric W. Biederman 已提交
203 204 205 206
}

static struct tun_struct *__tun_get(struct tun_file *tfile)
{
207 208 209 210 211 212
	struct tun_struct *tun = NULL;

	if (atomic_inc_not_zero(&tfile->count))
		tun = tfile->tun;

	return tun;
E
Eric W. Biederman 已提交
213 214 215 216 217 218 219 220 221
}

static struct tun_struct *tun_get(struct file *file)
{
	return __tun_get(file->private_data);
}

static void tun_put(struct tun_struct *tun)
{
222 223 224 225
	struct tun_file *tfile = tun->tfile;

	if (atomic_dec_and_test(&tfile->count))
		tun_detach(tfile->tun);
E
Eric W. Biederman 已提交
226 227
}

228
/* TAP filtering */
229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277
static void addr_hash_set(u32 *mask, const u8 *addr)
{
	int n = ether_crc(ETH_ALEN, addr) >> 26;
	mask[n >> 5] |= (1 << (n & 31));
}

static unsigned int addr_hash_test(const u32 *mask, const u8 *addr)
{
	int n = ether_crc(ETH_ALEN, addr) >> 26;
	return mask[n >> 5] & (1 << (n & 31));
}

static int update_filter(struct tap_filter *filter, void __user *arg)
{
	struct { u8 u[ETH_ALEN]; } *addr;
	struct tun_filter uf;
	int err, alen, n, nexact;

	if (copy_from_user(&uf, arg, sizeof(uf)))
		return -EFAULT;

	if (!uf.count) {
		/* Disabled */
		filter->count = 0;
		return 0;
	}

	alen = ETH_ALEN * uf.count;
	addr = kmalloc(alen, GFP_KERNEL);
	if (!addr)
		return -ENOMEM;

	if (copy_from_user(addr, arg + sizeof(uf), alen)) {
		err = -EFAULT;
		goto done;
	}

	/* The filter is updated without holding any locks. Which is
	 * perfectly safe. We disable it first and in the worst
	 * case we'll accept a few undesired packets. */
	filter->count = 0;
	wmb();

	/* Use first set of addresses as an exact filter */
	for (n = 0; n < uf.count && n < FLT_EXACT_COUNT; n++)
		memcpy(filter->addr[n], addr[n].u, ETH_ALEN);

	nexact = n;

278 279
	/* Remaining multicast addresses are hashed,
	 * unicast will leave the filter disabled. */
280
	memset(filter->mask, 0, sizeof(filter->mask));
281 282 283 284 285
	for (; n < uf.count; n++) {
		if (!is_multicast_ether_addr(addr[n].u)) {
			err = 0; /* no filter */
			goto done;
		}
286
		addr_hash_set(filter->mask, addr[n].u);
287
	}
288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315

	/* For ALLMULTI just set the mask to all ones.
	 * This overrides the mask populated above. */
	if ((uf.flags & TUN_FLT_ALLMULTI))
		memset(filter->mask, ~0, sizeof(filter->mask));

	/* Now enable the filter */
	wmb();
	filter->count = nexact;

	/* Return the number of exact filters */
	err = nexact;

done:
	kfree(addr);
	return err;
}

/* Returns: 0 - drop, !=0 - accept */
static int run_filter(struct tap_filter *filter, const struct sk_buff *skb)
{
	/* Cannot use eth_hdr(skb) here because skb_mac_hdr() is incorrect
	 * at this point. */
	struct ethhdr *eh = (struct ethhdr *) skb->data;
	int i;

	/* Exact match */
	for (i = 0; i < filter->count; i++)
316
		if (ether_addr_equal(eh->h_dest, filter->addr[i]))
317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337
			return 1;

	/* Inexact match (multicast only) */
	if (is_multicast_ether_addr(eh->h_dest))
		return addr_hash_test(filter->mask, eh->h_dest);

	return 0;
}

/*
 * Checks whether the packet is accepted or not.
 * Returns: 0 - drop, !=0 - accept
 */
static int check_filter(struct tap_filter *filter, const struct sk_buff *skb)
{
	if (!filter->count)
		return 1;

	return run_filter(filter, skb);
}

L
Linus Torvalds 已提交
338 339
/* Network device part of the driver */

340
static const struct ethtool_ops tun_ethtool_ops;
L
Linus Torvalds 已提交
341

342 343 344 345 346 347 348 349 350
/* Net device detach from fd. */
static void tun_net_uninit(struct net_device *dev)
{
	struct tun_struct *tun = netdev_priv(dev);
	struct tun_file *tfile = tun->tfile;

	/* Inform the methods they need to stop using the dev.
	 */
	if (tfile) {
351
		wake_up_all(&tun->wq.wait);
352 353 354 355 356
		if (atomic_dec_and_test(&tfile->count))
			__tun_detach(tun);
	}
}

357 358 359 360
static void tun_free_netdev(struct net_device *dev)
{
	struct tun_struct *tun = netdev_priv(dev);

361 362
	BUG_ON(!test_bit(SOCK_EXTERNALLY_ALLOCATED, &tun->socket.flags));

363
	sk_release_kernel(tun->socket.sk);
364 365
}

L
Linus Torvalds 已提交
366 367 368 369 370 371 372 373 374 375 376 377 378 379 380
/* Net device open. */
static int tun_net_open(struct net_device *dev)
{
	netif_start_queue(dev);
	return 0;
}

/* Net device close. */
static int tun_net_close(struct net_device *dev)
{
	netif_stop_queue(dev);
	return 0;
}

/* Net device start xmit */
381
static netdev_tx_t tun_net_xmit(struct sk_buff *skb, struct net_device *dev)
L
Linus Torvalds 已提交
382 383 384
{
	struct tun_struct *tun = netdev_priv(dev);

385
	tun_debug(KERN_INFO, tun, "tun_net_xmit %d\n", skb->len);
L
Linus Torvalds 已提交
386 387

	/* Drop packet if interface is not attached */
E
Eric W. Biederman 已提交
388
	if (!tun->tfile)
L
Linus Torvalds 已提交
389 390
		goto drop;

391 392 393 394 395 396
	/* Drop if the filter does not like it.
	 * This is a noop if the filter is disabled.
	 * Filter can be enabled only for the TAP devices. */
	if (!check_filter(&tun->txflt, skb))
		goto drop;

M
Michael S. Tsirkin 已提交
397 398 399 400
	if (tun->socket.sk->sk_filter &&
	    sk_filter(tun->socket.sk, skb))
		goto drop;

401
	if (skb_queue_len(&tun->socket.sk->sk_receive_queue) >= dev->tx_queue_len) {
L
Linus Torvalds 已提交
402 403 404 405 406 407 408
		if (!(tun->flags & TUN_ONE_QUEUE)) {
			/* Normal queueing mode. */
			/* Packet scheduler handles dropping of further packets. */
			netif_stop_queue(dev);

			/* We won't see all dropped packets individually, so overrun
			 * error is more appropriate. */
409
			dev->stats.tx_fifo_errors++;
L
Linus Torvalds 已提交
410 411 412 413 414 415 416
		} else {
			/* Single queue mode.
			 * Driver handles dropping of all packets itself. */
			goto drop;
		}
	}

M
Michael S. Tsirkin 已提交
417 418
	/* Orphan the skb - required as we might hang on to it
	 * for indefinite time. */
M
Michael S. Tsirkin 已提交
419 420
	if (unlikely(skb_orphan_frags(skb, GFP_ATOMIC)))
		goto drop;
M
Michael S. Tsirkin 已提交
421 422
	skb_orphan(skb);

423
	/* Enqueue packet */
424
	skb_queue_tail(&tun->socket.sk->sk_receive_queue, skb);
L
Linus Torvalds 已提交
425 426 427 428

	/* Notify and wake up reader process */
	if (tun->flags & TUN_FASYNC)
		kill_fasync(&tun->fasync, SIGIO, POLL_IN);
429
	wake_up_interruptible_poll(&tun->wq.wait, POLLIN |
430
				   POLLRDNORM | POLLRDBAND);
431
	return NETDEV_TX_OK;
L
Linus Torvalds 已提交
432 433

drop:
434
	dev->stats.tx_dropped++;
L
Linus Torvalds 已提交
435
	kfree_skb(skb);
436
	return NETDEV_TX_OK;
L
Linus Torvalds 已提交
437 438
}

439
static void tun_net_mclist(struct net_device *dev)
L
Linus Torvalds 已提交
440
{
441 442 443 444 445
	/*
	 * This callback is supposed to deal with mc filter in
	 * _rx_ path and has nothing to do with the _tx_ path.
	 * In rx path we always accept everything userspace gives us.
	 */
L
Linus Torvalds 已提交
446 447
}

E
Ed Swierk 已提交
448 449 450 451 452 453 454 455 456 457 458 459
#define MIN_MTU 68
#define MAX_MTU 65535

static int
tun_net_change_mtu(struct net_device *dev, int new_mtu)
{
	if (new_mtu < MIN_MTU || new_mtu + dev->hard_header_len > MAX_MTU)
		return -EINVAL;
	dev->mtu = new_mtu;
	return 0;
}

460 461
static netdev_features_t tun_net_fix_features(struct net_device *dev,
	netdev_features_t features)
462 463 464 465 466
{
	struct tun_struct *tun = netdev_priv(dev);

	return (features & tun->set_features) | (features & ~TUN_USER_FEATURES);
}
467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483
#ifdef CONFIG_NET_POLL_CONTROLLER
static void tun_poll_controller(struct net_device *dev)
{
	/*
	 * Tun only receives frames when:
	 * 1) the char device endpoint gets data from user space
	 * 2) the tun socket gets a sendmsg call from user space
	 * Since both of those are syncronous operations, we are guaranteed
	 * never to have pending data when we poll for it
	 * so theres nothing to do here but return.
	 * We need this though so netpoll recognizes us as an interface that
	 * supports polling, which enables bridge devices in virt setups to
	 * still use netconsole
	 */
	return;
}
#endif
S
Stephen Hemminger 已提交
484
static const struct net_device_ops tun_netdev_ops = {
485
	.ndo_uninit		= tun_net_uninit,
S
Stephen Hemminger 已提交
486 487
	.ndo_open		= tun_net_open,
	.ndo_stop		= tun_net_close,
488
	.ndo_start_xmit		= tun_net_xmit,
S
Stephen Hemminger 已提交
489
	.ndo_change_mtu		= tun_net_change_mtu,
490
	.ndo_fix_features	= tun_net_fix_features,
491 492 493
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= tun_poll_controller,
#endif
S
Stephen Hemminger 已提交
494 495 496
};

static const struct net_device_ops tap_netdev_ops = {
497
	.ndo_uninit		= tun_net_uninit,
S
Stephen Hemminger 已提交
498 499
	.ndo_open		= tun_net_open,
	.ndo_stop		= tun_net_close,
500
	.ndo_start_xmit		= tun_net_xmit,
S
Stephen Hemminger 已提交
501
	.ndo_change_mtu		= tun_net_change_mtu,
502
	.ndo_fix_features	= tun_net_fix_features,
503
	.ndo_set_rx_mode	= tun_net_mclist,
S
Stephen Hemminger 已提交
504 505
	.ndo_set_mac_address	= eth_mac_addr,
	.ndo_validate_addr	= eth_validate_addr,
506 507 508
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= tun_poll_controller,
#endif
S
Stephen Hemminger 已提交
509 510
};

L
Linus Torvalds 已提交
511 512 513 514
/* Initialize net device. */
static void tun_net_init(struct net_device *dev)
{
	struct tun_struct *tun = netdev_priv(dev);
515

L
Linus Torvalds 已提交
516 517
	switch (tun->flags & TUN_TYPE_MASK) {
	case TUN_TUN_DEV:
S
Stephen Hemminger 已提交
518 519
		dev->netdev_ops = &tun_netdev_ops;

L
Linus Torvalds 已提交
520 521 522 523 524 525
		/* Point-to-Point TUN Device */
		dev->hard_header_len = 0;
		dev->addr_len = 0;
		dev->mtu = 1500;

		/* Zero header length */
526
		dev->type = ARPHRD_NONE;
L
Linus Torvalds 已提交
527 528 529 530 531
		dev->flags = IFF_POINTOPOINT | IFF_NOARP | IFF_MULTICAST;
		dev->tx_queue_len = TUN_READQ_SIZE;  /* We prefer our own queue length */
		break;

	case TUN_TAP_DEV:
K
Kusanagi Kouichi 已提交
532
		dev->netdev_ops = &tap_netdev_ops;
L
Linus Torvalds 已提交
533 534
		/* Ethernet TAP Device */
		ether_setup(dev);
535
		dev->priv_flags &= ~IFF_TX_SKB_SHARING;
536

537
		eth_hw_addr_random(dev);
538

L
Linus Torvalds 已提交
539 540 541 542 543 544 545 546 547
		dev->tx_queue_len = TUN_READQ_SIZE;  /* We prefer our own queue length */
		break;
	}
}

/* Character device part */

/* Poll */
static unsigned int tun_chr_poll(struct file *file, poll_table * wait)
548
{
549 550
	struct tun_file *tfile = file->private_data;
	struct tun_struct *tun = __tun_get(tfile);
551
	struct sock *sk;
552
	unsigned int mask = 0;
L
Linus Torvalds 已提交
553 554

	if (!tun)
555
		return POLLERR;
L
Linus Torvalds 已提交
556

557
	sk = tun->socket.sk;
558

559
	tun_debug(KERN_INFO, tun, "tun_chr_poll\n");
L
Linus Torvalds 已提交
560

561
	poll_wait(file, &tun->wq.wait, wait);
562

563
	if (!skb_queue_empty(&sk->sk_receive_queue))
L
Linus Torvalds 已提交
564 565
		mask |= POLLIN | POLLRDNORM;

566 567 568 569 570
	if (sock_writeable(sk) ||
	    (!test_and_set_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags) &&
	     sock_writeable(sk)))
		mask |= POLLOUT | POLLWRNORM;

571 572 573
	if (tun->dev->reg_state != NETREG_REGISTERED)
		mask = POLLERR;

E
Eric W. Biederman 已提交
574
	tun_put(tun);
L
Linus Torvalds 已提交
575 576 577
	return mask;
}

578 579
/* prepad is the amount to reserve at front.  len is length after that.
 * linear is a hint as to how much to copy (usually headers). */
580 581 582
static struct sk_buff *tun_alloc_skb(struct tun_struct *tun,
				     size_t prepad, size_t len,
				     size_t linear, int noblock)
583
{
584
	struct sock *sk = tun->socket.sk;
585
	struct sk_buff *skb;
586
	int err;
587

588 589
	sock_update_classid(sk);

590
	/* Under a page?  Don't bother with paged skb. */
H
Herbert Xu 已提交
591
	if (prepad + len < PAGE_SIZE || !linear)
592
		linear = len;
593

594 595
	skb = sock_alloc_send_pskb(sk, prepad + linear, len - linear, noblock,
				   &err);
596
	if (!skb)
597
		return ERR_PTR(err);
598 599 600

	skb_reserve(skb, prepad);
	skb_put(skb, linear);
601 602
	skb->data_len = len - linear;
	skb->len += len - linear;
603 604 605 606

	return skb;
}

L
Linus Torvalds 已提交
607
/* Get packet from user space buffer */
608 609 610
static ssize_t tun_get_user(struct tun_struct *tun,
			    const struct iovec *iv, size_t count,
			    int noblock)
L
Linus Torvalds 已提交
611
{
612
	struct tun_pi pi = { 0, cpu_to_be16(ETH_P_IP) };
L
Linus Torvalds 已提交
613
	struct sk_buff *skb;
614
	size_t len = count, align = NET_SKB_PAD;
615
	struct virtio_net_hdr gso = { 0 };
616
	int offset = 0;
L
Linus Torvalds 已提交
617 618 619 620 621

	if (!(tun->flags & TUN_NO_PI)) {
		if ((len -= sizeof(pi)) > count)
			return -EINVAL;

622
		if (memcpy_fromiovecend((void *)&pi, iv, 0, sizeof(pi)))
L
Linus Torvalds 已提交
623
			return -EFAULT;
624
		offset += sizeof(pi);
L
Linus Torvalds 已提交
625 626
	}

627
	if (tun->flags & TUN_VNET_HDR) {
628
		if ((len -= tun->vnet_hdr_sz) > count)
629 630
			return -EINVAL;

631
		if (memcpy_fromiovecend((void *)&gso, iv, offset, sizeof(gso)))
632 633
			return -EFAULT;

634 635 636 637
		if ((gso.flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) &&
		    gso.csum_start + gso.csum_offset + 2 > gso.hdr_len)
			gso.hdr_len = gso.csum_start + gso.csum_offset + 2;

638 639
		if (gso.hdr_len > len)
			return -EINVAL;
640
		offset += tun->vnet_hdr_sz;
641 642
	}

643
	if ((tun->flags & TUN_TYPE_MASK) == TUN_TAP_DEV) {
644
		align += NET_IP_ALIGN;
H
Herbert Xu 已提交
645 646
		if (unlikely(len < ETH_HLEN ||
			     (gso.hdr_len && gso.hdr_len < ETH_HLEN)))
647 648
			return -EINVAL;
	}
649

650 651 652 653 654
	skb = tun_alloc_skb(tun, align, len, gso.hdr_len, noblock);
	if (IS_ERR(skb)) {
		if (PTR_ERR(skb) != -EAGAIN)
			tun->dev->stats.rx_dropped++;
		return PTR_ERR(skb);
L
Linus Torvalds 已提交
655 656
	}

657
	if (skb_copy_datagram_from_iovec(skb, 0, iv, offset, len)) {
658
		tun->dev->stats.rx_dropped++;
D
Dave Jones 已提交
659
		kfree_skb(skb);
L
Linus Torvalds 已提交
660
		return -EFAULT;
D
Dave Jones 已提交
661
	}
L
Linus Torvalds 已提交
662

663 664 665 666 667 668 669
	if (gso.flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) {
		if (!skb_partial_csum_set(skb, gso.csum_start,
					  gso.csum_offset)) {
			tun->dev->stats.rx_frame_errors++;
			kfree_skb(skb);
			return -EINVAL;
		}
670
	}
671

L
Linus Torvalds 已提交
672 673
	switch (tun->flags & TUN_TYPE_MASK) {
	case TUN_TUN_DEV:
674 675 676 677 678 679 680 681 682 683 684 685 686 687 688
		if (tun->flags & TUN_NO_PI) {
			switch (skb->data[0] & 0xf0) {
			case 0x40:
				pi.proto = htons(ETH_P_IP);
				break;
			case 0x60:
				pi.proto = htons(ETH_P_IPV6);
				break;
			default:
				tun->dev->stats.rx_dropped++;
				kfree_skb(skb);
				return -EINVAL;
			}
		}

689
		skb_reset_mac_header(skb);
L
Linus Torvalds 已提交
690
		skb->protocol = pi.proto;
691
		skb->dev = tun->dev;
L
Linus Torvalds 已提交
692 693 694 695
		break;
	case TUN_TAP_DEV:
		skb->protocol = eth_type_trans(skb, tun->dev);
		break;
696
	}
L
Linus Torvalds 已提交
697

698 699 700 701 702 703 704 705 706
	if (gso.gso_type != VIRTIO_NET_HDR_GSO_NONE) {
		pr_debug("GSO!\n");
		switch (gso.gso_type & ~VIRTIO_NET_HDR_GSO_ECN) {
		case VIRTIO_NET_HDR_GSO_TCPV4:
			skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
			break;
		case VIRTIO_NET_HDR_GSO_TCPV6:
			skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6;
			break;
707 708 709
		case VIRTIO_NET_HDR_GSO_UDP:
			skb_shinfo(skb)->gso_type = SKB_GSO_UDP;
			break;
710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729
		default:
			tun->dev->stats.rx_frame_errors++;
			kfree_skb(skb);
			return -EINVAL;
		}

		if (gso.gso_type & VIRTIO_NET_HDR_GSO_ECN)
			skb_shinfo(skb)->gso_type |= SKB_GSO_TCP_ECN;

		skb_shinfo(skb)->gso_size = gso.gso_size;
		if (skb_shinfo(skb)->gso_size == 0) {
			tun->dev->stats.rx_frame_errors++;
			kfree_skb(skb);
			return -EINVAL;
		}

		/* Header must be checked, and gso_segs computed. */
		skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
		skb_shinfo(skb)->gso_segs = 0;
	}
730

L
Linus Torvalds 已提交
731
	netif_rx_ni(skb);
732

733 734
	tun->dev->stats.rx_packets++;
	tun->dev->stats.rx_bytes += len;
L
Linus Torvalds 已提交
735 736

	return count;
737
}
L
Linus Torvalds 已提交
738

739 740
static ssize_t tun_chr_aio_write(struct kiocb *iocb, const struct iovec *iv,
			      unsigned long count, loff_t pos)
L
Linus Torvalds 已提交
741
{
742
	struct file *file = iocb->ki_filp;
H
Herbert Xu 已提交
743
	struct tun_struct *tun = tun_get(file);
E
Eric W. Biederman 已提交
744
	ssize_t result;
L
Linus Torvalds 已提交
745 746 747 748

	if (!tun)
		return -EBADFD;

749
	tun_debug(KERN_INFO, tun, "tun_chr_write %ld\n", count);
L
Linus Torvalds 已提交
750

751
	result = tun_get_user(tun, iv, iov_length(iv, count),
752
			      file->f_flags & O_NONBLOCK);
E
Eric W. Biederman 已提交
753 754 755

	tun_put(tun);
	return result;
L
Linus Torvalds 已提交
756 757 758
}

/* Put packet to the user space buffer */
759 760 761
static ssize_t tun_put_user(struct tun_struct *tun,
			    struct sk_buff *skb,
			    const struct iovec *iv, int len)
L
Linus Torvalds 已提交
762 763 764 765 766 767 768 769 770 771 772 773
{
	struct tun_pi pi = { 0, skb->protocol };
	ssize_t total = 0;

	if (!(tun->flags & TUN_NO_PI)) {
		if ((len -= sizeof(pi)) < 0)
			return -EINVAL;

		if (len < skb->len) {
			/* Packet will be striped */
			pi.flags |= TUN_PKT_STRIP;
		}
774

775
		if (memcpy_toiovecend(iv, (void *) &pi, 0, sizeof(pi)))
L
Linus Torvalds 已提交
776 777
			return -EFAULT;
		total += sizeof(pi);
778
	}
L
Linus Torvalds 已提交
779

780 781
	if (tun->flags & TUN_VNET_HDR) {
		struct virtio_net_hdr gso = { 0 }; /* no info leak */
782
		if ((len -= tun->vnet_hdr_sz) < 0)
783 784 785 786 787 788 789 790 791 792 793 794
			return -EINVAL;

		if (skb_is_gso(skb)) {
			struct skb_shared_info *sinfo = skb_shinfo(skb);

			/* This is a hint as to how much should be linear. */
			gso.hdr_len = skb_headlen(skb);
			gso.gso_size = sinfo->gso_size;
			if (sinfo->gso_type & SKB_GSO_TCPV4)
				gso.gso_type = VIRTIO_NET_HDR_GSO_TCPV4;
			else if (sinfo->gso_type & SKB_GSO_TCPV6)
				gso.gso_type = VIRTIO_NET_HDR_GSO_TCPV6;
795 796
			else if (sinfo->gso_type & SKB_GSO_UDP)
				gso.gso_type = VIRTIO_NET_HDR_GSO_UDP;
797
			else {
798
				pr_err("unexpected GSO type: "
799 800 801 802 803 804 805 806 807 808
				       "0x%x, gso_size %d, hdr_len %d\n",
				       sinfo->gso_type, gso.gso_size,
				       gso.hdr_len);
				print_hex_dump(KERN_ERR, "tun: ",
					       DUMP_PREFIX_NONE,
					       16, 1, skb->head,
					       min((int)gso.hdr_len, 64), true);
				WARN_ON_ONCE(1);
				return -EINVAL;
			}
809 810 811 812 813 814 815
			if (sinfo->gso_type & SKB_GSO_TCP_ECN)
				gso.gso_type |= VIRTIO_NET_HDR_GSO_ECN;
		} else
			gso.gso_type = VIRTIO_NET_HDR_GSO_NONE;

		if (skb->ip_summed == CHECKSUM_PARTIAL) {
			gso.flags = VIRTIO_NET_HDR_F_NEEDS_CSUM;
816
			gso.csum_start = skb_checksum_start_offset(skb);
817
			gso.csum_offset = skb->csum_offset;
818 819
		} else if (skb->ip_summed == CHECKSUM_UNNECESSARY) {
			gso.flags = VIRTIO_NET_HDR_F_DATA_VALID;
820 821
		} /* else everything is zero */

822 823
		if (unlikely(memcpy_toiovecend(iv, (void *)&gso, total,
					       sizeof(gso))))
824
			return -EFAULT;
825
		total += tun->vnet_hdr_sz;
826 827
	}

L
Linus Torvalds 已提交
828 829
	len = min_t(int, skb->len, len);

830
	skb_copy_datagram_const_iovec(skb, 0, iv, total, len);
831
	total += skb->len;
L
Linus Torvalds 已提交
832

833 834
	tun->dev->stats.tx_packets++;
	tun->dev->stats.tx_bytes += len;
L
Linus Torvalds 已提交
835 836 837 838

	return total;
}

839 840 841
static ssize_t tun_do_read(struct tun_struct *tun,
			   struct kiocb *iocb, const struct iovec *iv,
			   ssize_t len, int noblock)
L
Linus Torvalds 已提交
842 843 844
{
	DECLARE_WAITQUEUE(wait, current);
	struct sk_buff *skb;
845
	ssize_t ret = 0;
L
Linus Torvalds 已提交
846

847
	tun_debug(KERN_INFO, tun, "tun_chr_read\n");
L
Linus Torvalds 已提交
848

849 850
	if (unlikely(!noblock))
		add_wait_queue(&tun->wq.wait, &wait);
L
Linus Torvalds 已提交
851 852 853 854
	while (len) {
		current->state = TASK_INTERRUPTIBLE;

		/* Read frames from the queue */
855
		if (!(skb=skb_dequeue(&tun->socket.sk->sk_receive_queue))) {
856
			if (noblock) {
L
Linus Torvalds 已提交
857 858 859 860 861 862 863
				ret = -EAGAIN;
				break;
			}
			if (signal_pending(current)) {
				ret = -ERESTARTSYS;
				break;
			}
864 865 866 867
			if (tun->dev->reg_state != NETREG_REGISTERED) {
				ret = -EIO;
				break;
			}
L
Linus Torvalds 已提交
868 869 870 871 872 873 874

			/* Nothing to read, let's sleep */
			schedule();
			continue;
		}
		netif_wake_queue(tun->dev);

875
		ret = tun_put_user(tun, skb, iv, len);
876 877
		kfree_skb(skb);
		break;
L
Linus Torvalds 已提交
878 879 880
	}

	current->state = TASK_RUNNING;
881 882
	if (unlikely(!noblock))
		remove_wait_queue(&tun->wq.wait, &wait);
L
Linus Torvalds 已提交
883

884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904
	return ret;
}

static ssize_t tun_chr_aio_read(struct kiocb *iocb, const struct iovec *iv,
			    unsigned long count, loff_t pos)
{
	struct file *file = iocb->ki_filp;
	struct tun_file *tfile = file->private_data;
	struct tun_struct *tun = __tun_get(tfile);
	ssize_t len, ret;

	if (!tun)
		return -EBADFD;
	len = iov_length(iv, count);
	if (len < 0) {
		ret = -EINVAL;
		goto out;
	}

	ret = tun_do_read(tun, iocb, iv, len, file->f_flags & O_NONBLOCK);
	ret = min_t(ssize_t, ret, len);
E
Eric W. Biederman 已提交
905 906
out:
	tun_put(tun);
L
Linus Torvalds 已提交
907 908 909 910 911 912 913 914
	return ret;
}

static void tun_setup(struct net_device *dev)
{
	struct tun_struct *tun = netdev_priv(dev);

	tun->owner = -1;
915
	tun->group = -1;
L
Linus Torvalds 已提交
916 917

	dev->ethtool_ops = &tun_ethtool_ops;
918
	dev->destructor = tun_free_netdev;
L
Linus Torvalds 已提交
919 920
}

921 922 923 924 925 926 927 928 929 930 931 932 933 934 935
/* Trivial set of netlink ops to allow deleting tun or tap
 * device with netlink.
 */
static int tun_validate(struct nlattr *tb[], struct nlattr *data[])
{
	return -EINVAL;
}

static struct rtnl_link_ops tun_link_ops __read_mostly = {
	.kind		= DRV_NAME,
	.priv_size	= sizeof(struct tun_struct),
	.setup		= tun_setup,
	.validate	= tun_validate,
};

936 937 938
static void tun_sock_write_space(struct sock *sk)
{
	struct tun_struct *tun;
939
	wait_queue_head_t *wqueue;
940 941 942 943 944 945 946

	if (!sock_writeable(sk))
		return;

	if (!test_and_clear_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags))
		return;

947 948 949
	wqueue = sk_sleep(sk);
	if (wqueue && waitqueue_active(wqueue))
		wake_up_interruptible_sync_poll(wqueue, POLLOUT |
950
						POLLWRNORM | POLLWRBAND);
H
Herbert Xu 已提交
951

952
	tun = tun_sk(sk)->tun;
953 954 955 956 957
	kill_fasync(&tun->fasync, SIGIO, POLL_OUT);
}

static void tun_sock_destruct(struct sock *sk)
{
958
	free_netdev(tun_sk(sk)->tun->dev);
959 960
}

961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985
static int tun_sendmsg(struct kiocb *iocb, struct socket *sock,
		       struct msghdr *m, size_t total_len)
{
	struct tun_struct *tun = container_of(sock, struct tun_struct, socket);
	return tun_get_user(tun, m->msg_iov, total_len,
			    m->msg_flags & MSG_DONTWAIT);
}

static int tun_recvmsg(struct kiocb *iocb, struct socket *sock,
		       struct msghdr *m, size_t total_len,
		       int flags)
{
	struct tun_struct *tun = container_of(sock, struct tun_struct, socket);
	int ret;
	if (flags & ~(MSG_DONTWAIT|MSG_TRUNC))
		return -EINVAL;
	ret = tun_do_read(tun, iocb, m->msg_iov, total_len,
			  flags & MSG_DONTWAIT);
	if (ret > total_len) {
		m->msg_flags |= MSG_TRUNC;
		ret = flags & MSG_TRUNC ? ret : total_len;
	}
	return ret;
}

986 987 988 989 990 991 992
static int tun_release(struct socket *sock)
{
	if (sock->sk)
		sock_put(sock->sk);
	return 0;
}

993 994 995 996
/* Ops structure to mimic raw sockets with tun */
static const struct proto_ops tun_socket_ops = {
	.sendmsg = tun_sendmsg,
	.recvmsg = tun_recvmsg,
997
	.release = tun_release,
998 999
};

1000 1001 1002 1003 1004
static struct proto tun_proto = {
	.name		= "tun",
	.owner		= THIS_MODULE,
	.obj_size	= sizeof(struct tun_sock),
};
1005

1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051
static int tun_flags(struct tun_struct *tun)
{
	int flags = 0;

	if (tun->flags & TUN_TUN_DEV)
		flags |= IFF_TUN;
	else
		flags |= IFF_TAP;

	if (tun->flags & TUN_NO_PI)
		flags |= IFF_NO_PI;

	if (tun->flags & TUN_ONE_QUEUE)
		flags |= IFF_ONE_QUEUE;

	if (tun->flags & TUN_VNET_HDR)
		flags |= IFF_VNET_HDR;

	return flags;
}

static ssize_t tun_show_flags(struct device *dev, struct device_attribute *attr,
			      char *buf)
{
	struct tun_struct *tun = netdev_priv(to_net_dev(dev));
	return sprintf(buf, "0x%x\n", tun_flags(tun));
}

static ssize_t tun_show_owner(struct device *dev, struct device_attribute *attr,
			      char *buf)
{
	struct tun_struct *tun = netdev_priv(to_net_dev(dev));
	return sprintf(buf, "%d\n", tun->owner);
}

static ssize_t tun_show_group(struct device *dev, struct device_attribute *attr,
			      char *buf)
{
	struct tun_struct *tun = netdev_priv(to_net_dev(dev));
	return sprintf(buf, "%d\n", tun->group);
}

static DEVICE_ATTR(tun_flags, 0444, tun_show_flags, NULL);
static DEVICE_ATTR(owner, 0444, tun_show_owner, NULL);
static DEVICE_ATTR(group, 0444, tun_show_group, NULL);

1052
static int tun_set_iff(struct net *net, struct file *file, struct ifreq *ifr)
L
Linus Torvalds 已提交
1053
{
1054
	struct sock *sk;
L
Linus Torvalds 已提交
1055 1056 1057 1058
	struct tun_struct *tun;
	struct net_device *dev;
	int err;

1059 1060
	dev = __dev_get_by_name(net, ifr->ifr_name);
	if (dev) {
P
Paul Moore 已提交
1061 1062
		const struct cred *cred = current_cred();

1063 1064
		if (ifr->ifr_flags & IFF_TUN_EXCL)
			return -EBUSY;
1065 1066 1067 1068 1069 1070 1071
		if ((ifr->ifr_flags & IFF_TUN) && dev->netdev_ops == &tun_netdev_ops)
			tun = netdev_priv(dev);
		else if ((ifr->ifr_flags & IFF_TAP) && dev->netdev_ops == &tap_netdev_ops)
			tun = netdev_priv(dev);
		else
			return -EINVAL;

P
Paul Moore 已提交
1072 1073 1074 1075
		if (((tun->owner != -1 && cred->euid != tun->owner) ||
		     (tun->group != -1 && !in_egroup_p(tun->group))) &&
		    !capable(CAP_NET_ADMIN))
			return -EPERM;
1076
		err = security_tun_dev_attach(tun->socket.sk);
P
Paul Moore 已提交
1077 1078 1079
		if (err < 0)
			return err;

1080 1081 1082
		err = tun_attach(tun, file);
		if (err < 0)
			return err;
1083
	}
L
Linus Torvalds 已提交
1084 1085 1086 1087
	else {
		char *name;
		unsigned long flags = 0;

1088 1089
		if (!capable(CAP_NET_ADMIN))
			return -EPERM;
P
Paul Moore 已提交
1090 1091 1092
		err = security_tun_dev_create();
		if (err < 0)
			return err;
1093

L
Linus Torvalds 已提交
1094 1095 1096 1097 1098 1099 1100 1101 1102
		/* Set dev type */
		if (ifr->ifr_flags & IFF_TUN) {
			/* TUN device */
			flags |= TUN_TUN_DEV;
			name = "tun%d";
		} else if (ifr->ifr_flags & IFF_TAP) {
			/* TAP device */
			flags |= TUN_TAP_DEV;
			name = "tap%d";
1103
		} else
1104
			return -EINVAL;
1105

L
Linus Torvalds 已提交
1106 1107 1108 1109 1110 1111 1112 1113
		if (*ifr->ifr_name)
			name = ifr->ifr_name;

		dev = alloc_netdev(sizeof(struct tun_struct), name,
				   tun_setup);
		if (!dev)
			return -ENOMEM;

1114
		dev_net_set(dev, net);
1115
		dev->rtnl_link_ops = &tun_link_ops;
S
Stephen Hemminger 已提交
1116

L
Linus Torvalds 已提交
1117 1118 1119
		tun = netdev_priv(dev);
		tun->dev = dev;
		tun->flags = flags;
1120
		tun->txflt.count = 0;
1121
		tun->vnet_hdr_sz = sizeof(struct virtio_net_hdr);
1122
		set_bit(SOCK_EXTERNALLY_ALLOCATED, &tun->socket.flags);
L
Linus Torvalds 已提交
1123

1124
		err = -ENOMEM;
1125
		sk = sk_alloc(&init_net, AF_UNSPEC, GFP_KERNEL, &tun_proto);
1126 1127 1128
		if (!sk)
			goto err_free_dev;

1129
		sk_change_net(sk, net);
1130 1131
		tun->socket.wq = &tun->wq;
		init_waitqueue_head(&tun->wq.wait);
1132
		tun->socket.ops = &tun_socket_ops;
1133 1134 1135 1136
		sock_init_data(&tun->socket, sk);
		sk->sk_write_space = tun_sock_write_space;
		sk->sk_sndbuf = INT_MAX;

1137
		tun_sk(sk)->tun = tun;
1138

P
Paul Moore 已提交
1139 1140
		security_tun_dev_post_create(sk);

L
Linus Torvalds 已提交
1141 1142
		tun_net_init(dev);

1143 1144 1145 1146
		dev->hw_features = NETIF_F_SG | NETIF_F_FRAGLIST |
			TUN_USER_FEATURES;
		dev->features = dev->hw_features;

L
Linus Torvalds 已提交
1147 1148
		err = register_netdevice(tun->dev);
		if (err < 0)
1149 1150
			goto err_free_sk;

1151 1152 1153
		if (device_create_file(&tun->dev->dev, &dev_attr_tun_flags) ||
		    device_create_file(&tun->dev->dev, &dev_attr_owner) ||
		    device_create_file(&tun->dev->dev, &dev_attr_group))
1154
			pr_err("Failed to create tun sysfs files\n");
1155

1156
		sk->sk_destruct = tun_sock_destruct;
1157 1158 1159

		err = tun_attach(tun, file);
		if (err < 0)
1160
			goto failed;
L
Linus Torvalds 已提交
1161 1162
	}

1163
	tun_debug(KERN_INFO, tun, "tun_set_iff\n");
L
Linus Torvalds 已提交
1164 1165 1166

	if (ifr->ifr_flags & IFF_NO_PI)
		tun->flags |= TUN_NO_PI;
1167 1168
	else
		tun->flags &= ~TUN_NO_PI;
L
Linus Torvalds 已提交
1169 1170 1171

	if (ifr->ifr_flags & IFF_ONE_QUEUE)
		tun->flags |= TUN_ONE_QUEUE;
1172 1173
	else
		tun->flags &= ~TUN_ONE_QUEUE;
L
Linus Torvalds 已提交
1174

1175 1176 1177 1178 1179
	if (ifr->ifr_flags & IFF_VNET_HDR)
		tun->flags |= TUN_VNET_HDR;
	else
		tun->flags &= ~TUN_VNET_HDR;

1180 1181 1182 1183 1184 1185
	/* Make sure persistent devices do not get stuck in
	 * xoff state.
	 */
	if (netif_running(tun->dev))
		netif_wake_queue(tun->dev);

L
Linus Torvalds 已提交
1186 1187 1188
	strcpy(ifr->ifr_name, tun->dev->name);
	return 0;

1189
 err_free_sk:
1190
	tun_free_netdev(dev);
L
Linus Torvalds 已提交
1191 1192 1193 1194 1195 1196
 err_free_dev:
	free_netdev(dev);
 failed:
	return err;
}

1197 1198
static int tun_get_iff(struct net *net, struct tun_struct *tun,
		       struct ifreq *ifr)
1199
{
1200
	tun_debug(KERN_INFO, tun, "tun_get_iff\n");
1201 1202 1203

	strcpy(ifr->ifr_name, tun->dev->name);

1204
	ifr->ifr_flags = tun_flags(tun);
1205 1206 1207 1208

	return 0;
}

1209 1210
/* This is like a cut-down ethtool ops, except done via tun fd so no
 * privs required. */
1211
static int set_offload(struct tun_struct *tun, unsigned long arg)
1212
{
1213
	netdev_features_t features = 0;
1214 1215

	if (arg & TUN_F_CSUM) {
1216
		features |= NETIF_F_HW_CSUM;
1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229
		arg &= ~TUN_F_CSUM;

		if (arg & (TUN_F_TSO4|TUN_F_TSO6)) {
			if (arg & TUN_F_TSO_ECN) {
				features |= NETIF_F_TSO_ECN;
				arg &= ~TUN_F_TSO_ECN;
			}
			if (arg & TUN_F_TSO4)
				features |= NETIF_F_TSO;
			if (arg & TUN_F_TSO6)
				features |= NETIF_F_TSO6;
			arg &= ~(TUN_F_TSO4|TUN_F_TSO6);
		}
1230 1231 1232 1233 1234

		if (arg & TUN_F_UFO) {
			features |= NETIF_F_UFO;
			arg &= ~TUN_F_UFO;
		}
1235 1236 1237 1238 1239 1240 1241
	}

	/* This gives the user a way to test for new features in future by
	 * trying to set them. */
	if (arg)
		return -EINVAL;

1242 1243
	tun->set_features = features;
	netdev_update_features(tun->dev);
1244 1245 1246 1247

	return 0;
}

1248 1249
static long __tun_chr_ioctl(struct file *file, unsigned int cmd,
			    unsigned long arg, int ifreq_len)
L
Linus Torvalds 已提交
1250
{
E
Eric W. Biederman 已提交
1251
	struct tun_file *tfile = file->private_data;
E
Eric W. Biederman 已提交
1252
	struct tun_struct *tun;
L
Linus Torvalds 已提交
1253
	void __user* argp = (void __user*)arg;
M
Michael S. Tsirkin 已提交
1254
	struct sock_fprog fprog;
L
Linus Torvalds 已提交
1255
	struct ifreq ifr;
1256
	int sndbuf;
1257
	int vnet_hdr_sz;
1258
	int ret;
L
Linus Torvalds 已提交
1259 1260

	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89)
1261
		if (copy_from_user(&ifr, argp, ifreq_len))
L
Linus Torvalds 已提交
1262 1263
			return -EFAULT;

E
Eric W. Biederman 已提交
1264 1265 1266 1267 1268 1269 1270 1271 1272
	if (cmd == TUNGETFEATURES) {
		/* Currently this just means: "what IFF flags are valid?".
		 * This is needed because we never checked for invalid flags on
		 * TUNSETIFF. */
		return put_user(IFF_TUN | IFF_TAP | IFF_NO_PI | IFF_ONE_QUEUE |
				IFF_VNET_HDR,
				(unsigned int __user*)argp);
	}

1273 1274
	rtnl_lock();

E
Eric W. Biederman 已提交
1275
	tun = __tun_get(tfile);
L
Linus Torvalds 已提交
1276 1277 1278
	if (cmd == TUNSETIFF && !tun) {
		ifr.ifr_name[IFNAMSIZ-1] = '\0';

1279
		ret = tun_set_iff(tfile->net, file, &ifr);
L
Linus Torvalds 已提交
1280

1281 1282
		if (ret)
			goto unlock;
L
Linus Torvalds 已提交
1283

1284
		if (copy_to_user(argp, &ifr, ifreq_len))
1285 1286
			ret = -EFAULT;
		goto unlock;
L
Linus Torvalds 已提交
1287 1288
	}

1289
	ret = -EBADFD;
L
Linus Torvalds 已提交
1290
	if (!tun)
1291
		goto unlock;
L
Linus Torvalds 已提交
1292

1293
	tun_debug(KERN_INFO, tun, "tun_chr_ioctl cmd %d\n", cmd);
L
Linus Torvalds 已提交
1294

E
Eric W. Biederman 已提交
1295
	ret = 0;
L
Linus Torvalds 已提交
1296
	switch (cmd) {
1297
	case TUNGETIFF:
1298
		ret = tun_get_iff(current->nsproxy->net_ns, tun, &ifr);
1299
		if (ret)
E
Eric W. Biederman 已提交
1300
			break;
1301

1302
		if (copy_to_user(argp, &ifr, ifreq_len))
E
Eric W. Biederman 已提交
1303
			ret = -EFAULT;
1304 1305
		break;

L
Linus Torvalds 已提交
1306 1307 1308
	case TUNSETNOCSUM:
		/* Disable/Enable checksum */

1309 1310
		/* [unimplemented] */
		tun_debug(KERN_INFO, tun, "ignored: set checksum %s\n",
1311
			  arg ? "disabled" : "enabled");
L
Linus Torvalds 已提交
1312 1313 1314 1315 1316 1317 1318 1319 1320
		break;

	case TUNSETPERSIST:
		/* Disable/Enable persist mode */
		if (arg)
			tun->flags |= TUN_PERSIST;
		else
			tun->flags &= ~TUN_PERSIST;

1321 1322
		tun_debug(KERN_INFO, tun, "persist %s\n",
			  arg ? "enabled" : "disabled");
L
Linus Torvalds 已提交
1323 1324 1325 1326 1327 1328
		break;

	case TUNSETOWNER:
		/* Set owner of the device */
		tun->owner = (uid_t) arg;

1329
		tun_debug(KERN_INFO, tun, "owner set to %d\n", tun->owner);
L
Linus Torvalds 已提交
1330 1331
		break;

1332 1333 1334 1335
	case TUNSETGROUP:
		/* Set group of the device */
		tun->group= (gid_t) arg;

1336
		tun_debug(KERN_INFO, tun, "group set to %d\n", tun->group);
1337 1338
		break;

1339 1340 1341
	case TUNSETLINK:
		/* Only allow setting the type when the interface is down */
		if (tun->dev->flags & IFF_UP) {
1342 1343
			tun_debug(KERN_INFO, tun,
				  "Linktype set failed because interface is up\n");
1344
			ret = -EBUSY;
1345 1346
		} else {
			tun->dev->type = (int) arg;
1347 1348
			tun_debug(KERN_INFO, tun, "linktype set to %d\n",
				  tun->dev->type);
1349
			ret = 0;
1350
		}
E
Eric W. Biederman 已提交
1351
		break;
1352

L
Linus Torvalds 已提交
1353 1354 1355 1356 1357
#ifdef TUN_DEBUG
	case TUNSETDEBUG:
		tun->debug = arg;
		break;
#endif
1358
	case TUNSETOFFLOAD:
1359
		ret = set_offload(tun, arg);
E
Eric W. Biederman 已提交
1360
		break;
1361

1362 1363
	case TUNSETTXFILTER:
		/* Can be set only for TAPs */
E
Eric W. Biederman 已提交
1364
		ret = -EINVAL;
1365
		if ((tun->flags & TUN_TYPE_MASK) != TUN_TAP_DEV)
E
Eric W. Biederman 已提交
1366
			break;
H
Harvey Harrison 已提交
1367
		ret = update_filter(&tun->txflt, (void __user *)arg);
E
Eric W. Biederman 已提交
1368
		break;
L
Linus Torvalds 已提交
1369 1370

	case SIOCGIFHWADDR:
1371
		/* Get hw address */
1372 1373
		memcpy(ifr.ifr_hwaddr.sa_data, tun->dev->dev_addr, ETH_ALEN);
		ifr.ifr_hwaddr.sa_family = tun->dev->type;
1374
		if (copy_to_user(argp, &ifr, ifreq_len))
E
Eric W. Biederman 已提交
1375 1376
			ret = -EFAULT;
		break;
L
Linus Torvalds 已提交
1377 1378

	case SIOCSIFHWADDR:
1379
		/* Set hw address */
1380 1381
		tun_debug(KERN_DEBUG, tun, "set hw address: %pM\n",
			  ifr.ifr_hwaddr.sa_data);
1382 1383

		ret = dev_set_mac_address(tun->dev, &ifr.ifr_hwaddr);
E
Eric W. Biederman 已提交
1384
		break;
1385 1386

	case TUNGETSNDBUF:
1387
		sndbuf = tun->socket.sk->sk_sndbuf;
1388 1389 1390 1391 1392 1393 1394 1395 1396 1397
		if (copy_to_user(argp, &sndbuf, sizeof(sndbuf)))
			ret = -EFAULT;
		break;

	case TUNSETSNDBUF:
		if (copy_from_user(&sndbuf, argp, sizeof(sndbuf))) {
			ret = -EFAULT;
			break;
		}

1398
		tun->socket.sk->sk_sndbuf = sndbuf;
1399 1400
		break;

1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419
	case TUNGETVNETHDRSZ:
		vnet_hdr_sz = tun->vnet_hdr_sz;
		if (copy_to_user(argp, &vnet_hdr_sz, sizeof(vnet_hdr_sz)))
			ret = -EFAULT;
		break;

	case TUNSETVNETHDRSZ:
		if (copy_from_user(&vnet_hdr_sz, argp, sizeof(vnet_hdr_sz))) {
			ret = -EFAULT;
			break;
		}
		if (vnet_hdr_sz < (int)sizeof(struct virtio_net_hdr)) {
			ret = -EINVAL;
			break;
		}

		tun->vnet_hdr_sz = vnet_hdr_sz;
		break;

M
Michael S. Tsirkin 已提交
1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439
	case TUNATTACHFILTER:
		/* Can be set only for TAPs */
		ret = -EINVAL;
		if ((tun->flags & TUN_TYPE_MASK) != TUN_TAP_DEV)
			break;
		ret = -EFAULT;
		if (copy_from_user(&fprog, argp, sizeof(fprog)))
			break;

		ret = sk_attach_filter(&fprog, tun->socket.sk);
		break;

	case TUNDETACHFILTER:
		/* Can be set only for TAPs */
		ret = -EINVAL;
		if ((tun->flags & TUN_TYPE_MASK) != TUN_TAP_DEV)
			break;
		ret = sk_detach_filter(tun->socket.sk);
		break;

L
Linus Torvalds 已提交
1440
	default:
E
Eric W. Biederman 已提交
1441 1442
		ret = -EINVAL;
		break;
1443
	}
L
Linus Torvalds 已提交
1444

1445 1446 1447 1448
unlock:
	rtnl_unlock();
	if (tun)
		tun_put(tun);
E
Eric W. Biederman 已提交
1449
	return ret;
L
Linus Torvalds 已提交
1450 1451
}

1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486
static long tun_chr_ioctl(struct file *file,
			  unsigned int cmd, unsigned long arg)
{
	return __tun_chr_ioctl(file, cmd, arg, sizeof (struct ifreq));
}

#ifdef CONFIG_COMPAT
static long tun_chr_compat_ioctl(struct file *file,
			 unsigned int cmd, unsigned long arg)
{
	switch (cmd) {
	case TUNSETIFF:
	case TUNGETIFF:
	case TUNSETTXFILTER:
	case TUNGETSNDBUF:
	case TUNSETSNDBUF:
	case SIOCGIFHWADDR:
	case SIOCSIFHWADDR:
		arg = (unsigned long)compat_ptr(arg);
		break;
	default:
		arg = (compat_ulong_t)arg;
		break;
	}

	/*
	 * compat_ifreq is shorter than ifreq, so we must not access beyond
	 * the end of that structure. All fields that are used in this
	 * driver are compatible though, we don't need to convert the
	 * contents.
	 */
	return __tun_chr_ioctl(file, cmd, arg, sizeof(struct compat_ifreq));
}
#endif /* CONFIG_COMPAT */

L
Linus Torvalds 已提交
1487 1488
static int tun_chr_fasync(int fd, struct file *file, int on)
{
E
Eric W. Biederman 已提交
1489
	struct tun_struct *tun = tun_get(file);
L
Linus Torvalds 已提交
1490 1491 1492 1493 1494
	int ret;

	if (!tun)
		return -EBADFD;

1495
	tun_debug(KERN_INFO, tun, "tun_chr_fasync %d\n", on);
L
Linus Torvalds 已提交
1496 1497

	if ((ret = fasync_helper(fd, file, on, &tun->fasync)) < 0)
J
Jonathan Corbet 已提交
1498
		goto out;
1499

L
Linus Torvalds 已提交
1500
	if (on) {
1501
		ret = __f_setown(file, task_pid(current), PIDTYPE_PID, 0);
L
Linus Torvalds 已提交
1502
		if (ret)
J
Jonathan Corbet 已提交
1503
			goto out;
L
Linus Torvalds 已提交
1504
		tun->flags |= TUN_FASYNC;
1505
	} else
L
Linus Torvalds 已提交
1506
		tun->flags &= ~TUN_FASYNC;
J
Jonathan Corbet 已提交
1507 1508
	ret = 0;
out:
E
Eric W. Biederman 已提交
1509
	tun_put(tun);
J
Jonathan Corbet 已提交
1510
	return ret;
L
Linus Torvalds 已提交
1511 1512 1513 1514
}

static int tun_chr_open(struct inode *inode, struct file * file)
{
E
Eric W. Biederman 已提交
1515
	struct tun_file *tfile;
T
Thomas Gleixner 已提交
1516

1517
	DBG1(KERN_INFO, "tunX: tun_chr_open\n");
E
Eric W. Biederman 已提交
1518 1519 1520 1521

	tfile = kmalloc(sizeof(*tfile), GFP_KERNEL);
	if (!tfile)
		return -ENOMEM;
1522
	atomic_set(&tfile->count, 0);
E
Eric W. Biederman 已提交
1523
	tfile->tun = NULL;
E
Eric W. Biederman 已提交
1524
	tfile->net = get_net(current->nsproxy->net_ns);
E
Eric W. Biederman 已提交
1525
	file->private_data = tfile;
L
Linus Torvalds 已提交
1526 1527 1528 1529 1530
	return 0;
}

static int tun_chr_close(struct inode *inode, struct file *file)
{
E
Eric W. Biederman 已提交
1531
	struct tun_file *tfile = file->private_data;
E
Eric W. Biederman 已提交
1532
	struct tun_struct *tun;
L
Linus Torvalds 已提交
1533

E
Eric W. Biederman 已提交
1534
	tun = __tun_get(tfile);
E
Eric W. Biederman 已提交
1535
	if (tun) {
H
Herbert Xu 已提交
1536 1537
		struct net_device *dev = tun->dev;

1538
		tun_debug(KERN_INFO, tun, "tun_chr_close\n");
L
Linus Torvalds 已提交
1539

E
Eric W. Biederman 已提交
1540
		__tun_detach(tun);
L
Linus Torvalds 已提交
1541

D
Daniel Mack 已提交
1542
		/* If desirable, unregister the netdevice. */
H
Herbert Xu 已提交
1543 1544 1545 1546 1547 1548
		if (!(tun->flags & TUN_PERSIST)) {
			rtnl_lock();
			if (dev->reg_state == NETREG_REGISTERED)
				unregister_netdevice(dev);
			rtnl_unlock();
		}
E
Eric W. Biederman 已提交
1549
	}
L
Linus Torvalds 已提交
1550

1551 1552
	tun = tfile->tun;
	if (tun)
1553
		sock_put(tun->socket.sk);
1554

E
Eric W. Biederman 已提交
1555
	put_net(tfile->net);
E
Eric W. Biederman 已提交
1556
	kfree(tfile);
L
Linus Torvalds 已提交
1557 1558 1559 1560

	return 0;
}

1561
static const struct file_operations tun_fops = {
1562
	.owner	= THIS_MODULE,
L
Linus Torvalds 已提交
1563
	.llseek = no_llseek,
1564 1565 1566 1567
	.read  = do_sync_read,
	.aio_read  = tun_chr_aio_read,
	.write = do_sync_write,
	.aio_write = tun_chr_aio_write,
L
Linus Torvalds 已提交
1568
	.poll	= tun_chr_poll,
1569 1570 1571 1572
	.unlocked_ioctl	= tun_chr_ioctl,
#ifdef CONFIG_COMPAT
	.compat_ioctl = tun_chr_compat_ioctl,
#endif
L
Linus Torvalds 已提交
1573 1574
	.open	= tun_chr_open,
	.release = tun_chr_close,
1575
	.fasync = tun_chr_fasync
L
Linus Torvalds 已提交
1576 1577 1578 1579 1580
};

static struct miscdevice tun_miscdev = {
	.minor = TUN_MINOR,
	.name = "tun",
1581
	.nodename = "net/tun",
L
Linus Torvalds 已提交
1582 1583 1584 1585 1586 1587 1588 1589 1590
	.fops = &tun_fops,
};

/* ethtool interface */

static int tun_get_settings(struct net_device *dev, struct ethtool_cmd *cmd)
{
	cmd->supported		= 0;
	cmd->advertising	= 0;
1591
	ethtool_cmd_speed_set(cmd, SPEED_10);
L
Linus Torvalds 已提交
1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605
	cmd->duplex		= DUPLEX_FULL;
	cmd->port		= PORT_TP;
	cmd->phy_address	= 0;
	cmd->transceiver	= XCVR_INTERNAL;
	cmd->autoneg		= AUTONEG_DISABLE;
	cmd->maxtxpkt		= 0;
	cmd->maxrxpkt		= 0;
	return 0;
}

static void tun_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info)
{
	struct tun_struct *tun = netdev_priv(dev);

1606 1607
	strlcpy(info->driver, DRV_NAME, sizeof(info->driver));
	strlcpy(info->version, DRV_VERSION, sizeof(info->version));
L
Linus Torvalds 已提交
1608 1609 1610

	switch (tun->flags & TUN_TYPE_MASK) {
	case TUN_TUN_DEV:
1611
		strlcpy(info->bus_info, "tun", sizeof(info->bus_info));
L
Linus Torvalds 已提交
1612 1613
		break;
	case TUN_TAP_DEV:
1614
		strlcpy(info->bus_info, "tap", sizeof(info->bus_info));
L
Linus Torvalds 已提交
1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636
		break;
	}
}

static u32 tun_get_msglevel(struct net_device *dev)
{
#ifdef TUN_DEBUG
	struct tun_struct *tun = netdev_priv(dev);
	return tun->debug;
#else
	return -EOPNOTSUPP;
#endif
}

static void tun_set_msglevel(struct net_device *dev, u32 value)
{
#ifdef TUN_DEBUG
	struct tun_struct *tun = netdev_priv(dev);
	tun->debug = value;
#endif
}

1637
static const struct ethtool_ops tun_ethtool_ops = {
L
Linus Torvalds 已提交
1638 1639 1640 1641
	.get_settings	= tun_get_settings,
	.get_drvinfo	= tun_get_drvinfo,
	.get_msglevel	= tun_get_msglevel,
	.set_msglevel	= tun_set_msglevel,
1642
	.get_link	= ethtool_op_get_link,
L
Linus Torvalds 已提交
1643 1644
};

1645

L
Linus Torvalds 已提交
1646 1647 1648 1649
static int __init tun_init(void)
{
	int ret = 0;

1650 1651
	pr_info("%s, %s\n", DRV_DESCRIPTION, DRV_VERSION);
	pr_info("%s\n", DRV_COPYRIGHT);
L
Linus Torvalds 已提交
1652

1653
	ret = rtnl_link_register(&tun_link_ops);
1654
	if (ret) {
1655
		pr_err("Can't register link_ops\n");
1656
		goto err_linkops;
1657 1658
	}

L
Linus Torvalds 已提交
1659
	ret = misc_register(&tun_miscdev);
1660
	if (ret) {
1661
		pr_err("Can't register misc device %d\n", TUN_MINOR);
1662 1663
		goto err_misc;
	}
1664
	return  0;
1665
err_misc:
1666 1667
	rtnl_link_unregister(&tun_link_ops);
err_linkops:
L
Linus Torvalds 已提交
1668 1669 1670 1671 1672
	return ret;
}

static void tun_cleanup(void)
{
1673
	misc_deregister(&tun_miscdev);
1674
	rtnl_link_unregister(&tun_link_ops);
L
Linus Torvalds 已提交
1675 1676
}

1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693
/* Get an underlying socket object from tun file.  Returns error unless file is
 * attached to a device.  The returned object works like a packet socket, it
 * can be used for sock_sendmsg/sock_recvmsg.  The caller is responsible for
 * holding a reference to the file for as long as the socket is in use. */
struct socket *tun_get_socket(struct file *file)
{
	struct tun_struct *tun;
	if (file->f_op != &tun_fops)
		return ERR_PTR(-EINVAL);
	tun = tun_get(file);
	if (!tun)
		return ERR_PTR(-EBADFD);
	tun_put(tun);
	return &tun->socket;
}
EXPORT_SYMBOL_GPL(tun_get_socket);

L
Linus Torvalds 已提交
1694 1695 1696 1697 1698 1699
module_init(tun_init);
module_exit(tun_cleanup);
MODULE_DESCRIPTION(DRV_DESCRIPTION);
MODULE_AUTHOR(DRV_COPYRIGHT);
MODULE_LICENSE("GPL");
MODULE_ALIAS_MISCDEV(TUN_MINOR);
1700
MODULE_ALIAS("devname:net/tun");