dev.c 117.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9
/*
 * 	NET3	Protocol independent device support routines.
 *
 *		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.
 *
 *	Derived from the non IP parts of dev.c 1.0.19
10
 * 		Authors:	Ross Biro
L
Linus Torvalds 已提交
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77
 *				Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *				Mark Evans, <evansmp@uhura.aston.ac.uk>
 *
 *	Additional Authors:
 *		Florian la Roche <rzsfl@rz.uni-sb.de>
 *		Alan Cox <gw4pts@gw4pts.ampr.org>
 *		David Hinds <dahinds@users.sourceforge.net>
 *		Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
 *		Adam Sulmicki <adam@cfar.umd.edu>
 *              Pekka Riikonen <priikone@poesidon.pspt.fi>
 *
 *	Changes:
 *              D.J. Barrow     :       Fixed bug where dev->refcnt gets set
 *              			to 2 if register_netdev gets called
 *              			before net_dev_init & also removed a
 *              			few lines of code in the process.
 *		Alan Cox	:	device private ioctl copies fields back.
 *		Alan Cox	:	Transmit queue code does relevant
 *					stunts to keep the queue safe.
 *		Alan Cox	:	Fixed double lock.
 *		Alan Cox	:	Fixed promisc NULL pointer trap
 *		????????	:	Support the full private ioctl range
 *		Alan Cox	:	Moved ioctl permission check into
 *					drivers
 *		Tim Kordas	:	SIOCADDMULTI/SIOCDELMULTI
 *		Alan Cox	:	100 backlog just doesn't cut it when
 *					you start doing multicast video 8)
 *		Alan Cox	:	Rewrote net_bh and list manager.
 *		Alan Cox	: 	Fix ETH_P_ALL echoback lengths.
 *		Alan Cox	:	Took out transmit every packet pass
 *					Saved a few bytes in the ioctl handler
 *		Alan Cox	:	Network driver sets packet type before
 *					calling netif_rx. Saves a function
 *					call a packet.
 *		Alan Cox	:	Hashed net_bh()
 *		Richard Kooijman:	Timestamp fixes.
 *		Alan Cox	:	Wrong field in SIOCGIFDSTADDR
 *		Alan Cox	:	Device lock protection.
 *		Alan Cox	: 	Fixed nasty side effect of device close
 *					changes.
 *		Rudi Cilibrasi	:	Pass the right thing to
 *					set_mac_address()
 *		Dave Miller	:	32bit quantity for the device lock to
 *					make it work out on a Sparc.
 *		Bjorn Ekwall	:	Added KERNELD hack.
 *		Alan Cox	:	Cleaned up the backlog initialise.
 *		Craig Metz	:	SIOCGIFCONF fix if space for under
 *					1 device.
 *	    Thomas Bogendoerfer :	Return ENODEV for dev_open, if there
 *					is no device open function.
 *		Andi Kleen	:	Fix error reporting for SIOCGIFCONF
 *	    Michael Chastain	:	Fix signed/unsigned for SIOCGIFCONF
 *		Cyrus Durgin	:	Cleaned for KMOD
 *		Adam Sulmicki   :	Bug Fix : Network Device Unload
 *					A network device unload needs to purge
 *					the backlog queue.
 *	Paul Rusty Russell	:	SIOCSIFNAME
 *              Pekka Riikonen  :	Netdev boot-time settings code
 *              Andrew Morton   :       Make unregister_netdevice wait
 *              			indefinitely on dev->refcnt
 * 		J Hadi Salim	:	- Backlog queue sampling
 *				        - netif_rx() feedback
 */

#include <asm/uaccess.h>
#include <asm/system.h>
#include <linux/bitops.h>
78
#include <linux/capability.h>
L
Linus Torvalds 已提交
79 80 81 82
#include <linux/cpu.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/sched.h>
A
Arjan van de Ven 已提交
83
#include <linux/mutex.h>
L
Linus Torvalds 已提交
84 85 86 87 88 89 90 91 92
#include <linux/string.h>
#include <linux/mm.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/errno.h>
#include <linux/interrupt.h>
#include <linux/if_ether.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
93
#include <linux/ethtool.h>
L
Linus Torvalds 已提交
94 95
#include <linux/notifier.h>
#include <linux/skbuff.h>
96
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
97 98 99 100 101 102
#include <net/sock.h>
#include <linux/rtnetlink.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#include <linux/stat.h>
#include <linux/if_bridge.h>
P
Patrick McHardy 已提交
103
#include <linux/if_macvlan.h>
L
Linus Torvalds 已提交
104 105 106 107 108 109 110 111 112 113 114
#include <net/dst.h>
#include <net/pkt_sched.h>
#include <net/checksum.h>
#include <linux/highmem.h>
#include <linux/init.h>
#include <linux/kmod.h>
#include <linux/module.h>
#include <linux/kallsyms.h>
#include <linux/netpoll.h>
#include <linux/rcupdate.h>
#include <linux/delay.h>
115
#include <net/wext.h>
L
Linus Torvalds 已提交
116 117
#include <net/iw_handler.h>
#include <asm/current.h>
S
Steve Grubb 已提交
118
#include <linux/audit.h>
119
#include <linux/dmaengine.h>
120
#include <linux/err.h>
121
#include <linux/ctype.h>
122
#include <linux/if_arp.h>
123
#include <linux/if_vlan.h>
124
#include <linux/ip.h>
125
#include <net/ip.h>
126 127
#include <linux/ipv6.h>
#include <linux/in.h>
D
David S. Miller 已提交
128 129
#include <linux/jhash.h>
#include <linux/random.h>
L
Linus Torvalds 已提交
130

131 132
#include "net-sysfs.h"

L
Linus Torvalds 已提交
133 134 135 136 137 138 139 140 141 142 143
/*
 *	The list of packet types we will receive (as opposed to discard)
 *	and the routines to invoke.
 *
 *	Why 16. Because with 16 the only overlap we get on a hash of the
 *	low nibble of the protocol value is RARP/SNAP/X.25.
 *
 *      NOTE:  That is no longer true with the addition of VLAN tags.  Not
 *             sure which should go first, but I bet it won't make much
 *             difference if we are running VLANs.  The good news is that
 *             this protocol won't be in the list unless compiled in, so
S
Stephen Hemminger 已提交
144
 *             the average user (w/out VLANs) will not be adversely affected.
L
Linus Torvalds 已提交
145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160
 *             --BLG
 *
 *		0800	IP
 *		8100    802.1Q VLAN
 *		0001	802.3
 *		0002	AX.25
 *		0004	802.2
 *		8035	RARP
 *		0005	SNAP
 *		0805	X.25
 *		0806	ARP
 *		8137	IPX
 *		0009	Localtalk
 *		86DD	IPv6
 */

161 162 163
#define PTYPE_HASH_SIZE	(16)
#define PTYPE_HASH_MASK	(PTYPE_HASH_SIZE - 1)

L
Linus Torvalds 已提交
164
static DEFINE_SPINLOCK(ptype_lock);
165
static struct list_head ptype_base[PTYPE_HASH_SIZE] __read_mostly;
166
static struct list_head ptype_all __read_mostly;	/* Taps */
L
Linus Torvalds 已提交
167

168
#ifdef CONFIG_NET_DMA
169 170 171 172
struct net_dma {
	struct dma_client client;
	spinlock_t lock;
	cpumask_t channel_mask;
173
	struct dma_chan **channels;
174 175 176 177 178 179 180 181 182 183 184
};

static enum dma_state_client
netdev_dma_event(struct dma_client *client, struct dma_chan *chan,
	enum dma_state state);

static struct net_dma net_dma = {
	.client = {
		.event_callback = netdev_dma_event,
	},
};
185 186
#endif

L
Linus Torvalds 已提交
187
/*
188
 * The @dev_base_head list is protected by @dev_base_lock and the rtnl
L
Linus Torvalds 已提交
189 190 191 192 193
 * semaphore.
 *
 * Pure readers hold dev_base_lock for reading.
 *
 * Writers must hold the rtnl semaphore while they loop through the
194
 * dev_base_head list, and hold dev_base_lock for writing when they do the
L
Linus Torvalds 已提交
195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210
 * actual updates.  This allows pure readers to access the list even
 * while a writer is preparing to update it.
 *
 * To put it another way, dev_base_lock is held for writing only to
 * protect against pure readers; the rtnl semaphore provides the
 * protection against other writers.
 *
 * See, for example usages, register_netdevice() and
 * unregister_netdevice(), which must be called with the rtnl
 * semaphore held.
 */
DEFINE_RWLOCK(dev_base_lock);

EXPORT_SYMBOL(dev_base_lock);

#define NETDEV_HASHBITS	8
211
#define NETDEV_HASHENTRIES (1 << NETDEV_HASHBITS)
L
Linus Torvalds 已提交
212

213
static inline struct hlist_head *dev_name_hash(struct net *net, const char *name)
L
Linus Torvalds 已提交
214 215
{
	unsigned hash = full_name_hash(name, strnlen(name, IFNAMSIZ));
216
	return &net->dev_name_head[hash & ((1 << NETDEV_HASHBITS) - 1)];
L
Linus Torvalds 已提交
217 218
}

219
static inline struct hlist_head *dev_index_hash(struct net *net, int ifindex)
L
Linus Torvalds 已提交
220
{
221
	return &net->dev_index_head[ifindex & ((1 << NETDEV_HASHBITS) - 1)];
L
Linus Torvalds 已提交
222 223
}

224 225 226
/* Device list insertion */
static int list_netdevice(struct net_device *dev)
{
227
	struct net *net = dev_net(dev);
228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251

	ASSERT_RTNL();

	write_lock_bh(&dev_base_lock);
	list_add_tail(&dev->dev_list, &net->dev_base_head);
	hlist_add_head(&dev->name_hlist, dev_name_hash(net, dev->name));
	hlist_add_head(&dev->index_hlist, dev_index_hash(net, dev->ifindex));
	write_unlock_bh(&dev_base_lock);
	return 0;
}

/* Device list removal */
static void unlist_netdevice(struct net_device *dev)
{
	ASSERT_RTNL();

	/* Unlink dev from the device chain */
	write_lock_bh(&dev_base_lock);
	list_del(&dev->dev_list);
	hlist_del(&dev->name_hlist);
	hlist_del(&dev->index_hlist);
	write_unlock_bh(&dev_base_lock);
}

L
Linus Torvalds 已提交
252 253 254 255
/*
 *	Our notifier list
 */

256
static RAW_NOTIFIER_HEAD(netdev_chain);
L
Linus Torvalds 已提交
257 258 259 260 261

/*
 *	Device drivers call our routines to queue packets here. We empty the
 *	queue in the local softnet handler.
 */
262 263

DEFINE_PER_CPU(struct softnet_data, softnet_data);
L
Linus Torvalds 已提交
264

265
#ifdef CONFIG_LOCKDEP
266
/*
267
 * register_netdevice() inits txq->_xmit_lock and sets lockdep class
268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
 * according to dev->type
 */
static const unsigned short netdev_lock_type[] =
	{ARPHRD_NETROM, ARPHRD_ETHER, ARPHRD_EETHER, ARPHRD_AX25,
	 ARPHRD_PRONET, ARPHRD_CHAOS, ARPHRD_IEEE802, ARPHRD_ARCNET,
	 ARPHRD_APPLETLK, ARPHRD_DLCI, ARPHRD_ATM, ARPHRD_METRICOM,
	 ARPHRD_IEEE1394, ARPHRD_EUI64, ARPHRD_INFINIBAND, ARPHRD_SLIP,
	 ARPHRD_CSLIP, ARPHRD_SLIP6, ARPHRD_CSLIP6, ARPHRD_RSRVD,
	 ARPHRD_ADAPT, ARPHRD_ROSE, ARPHRD_X25, ARPHRD_HWX25,
	 ARPHRD_PPP, ARPHRD_CISCO, ARPHRD_LAPB, ARPHRD_DDCMP,
	 ARPHRD_RAWHDLC, ARPHRD_TUNNEL, ARPHRD_TUNNEL6, ARPHRD_FRAD,
	 ARPHRD_SKIP, ARPHRD_LOOPBACK, ARPHRD_LOCALTLK, ARPHRD_FDDI,
	 ARPHRD_BIF, ARPHRD_SIT, ARPHRD_IPDDP, ARPHRD_IPGRE,
	 ARPHRD_PIMREG, ARPHRD_HIPPI, ARPHRD_ASH, ARPHRD_ECONET,
	 ARPHRD_IRDA, ARPHRD_FCPP, ARPHRD_FCAL, ARPHRD_FCPL,
	 ARPHRD_FCFABRIC, ARPHRD_IEEE802_TR, ARPHRD_IEEE80211,
	 ARPHRD_IEEE80211_PRISM, ARPHRD_IEEE80211_RADIOTAP, ARPHRD_VOID,
	 ARPHRD_NONE};

static const char *netdev_lock_name[] =
	{"_xmit_NETROM", "_xmit_ETHER", "_xmit_EETHER", "_xmit_AX25",
	 "_xmit_PRONET", "_xmit_CHAOS", "_xmit_IEEE802", "_xmit_ARCNET",
	 "_xmit_APPLETLK", "_xmit_DLCI", "_xmit_ATM", "_xmit_METRICOM",
	 "_xmit_IEEE1394", "_xmit_EUI64", "_xmit_INFINIBAND", "_xmit_SLIP",
	 "_xmit_CSLIP", "_xmit_SLIP6", "_xmit_CSLIP6", "_xmit_RSRVD",
	 "_xmit_ADAPT", "_xmit_ROSE", "_xmit_X25", "_xmit_HWX25",
	 "_xmit_PPP", "_xmit_CISCO", "_xmit_LAPB", "_xmit_DDCMP",
	 "_xmit_RAWHDLC", "_xmit_TUNNEL", "_xmit_TUNNEL6", "_xmit_FRAD",
	 "_xmit_SKIP", "_xmit_LOOPBACK", "_xmit_LOCALTLK", "_xmit_FDDI",
	 "_xmit_BIF", "_xmit_SIT", "_xmit_IPDDP", "_xmit_IPGRE",
	 "_xmit_PIMREG", "_xmit_HIPPI", "_xmit_ASH", "_xmit_ECONET",
	 "_xmit_IRDA", "_xmit_FCPP", "_xmit_FCAL", "_xmit_FCPL",
	 "_xmit_FCFABRIC", "_xmit_IEEE802_TR", "_xmit_IEEE80211",
	 "_xmit_IEEE80211_PRISM", "_xmit_IEEE80211_RADIOTAP", "_xmit_VOID",
	 "_xmit_NONE"};

static struct lock_class_key netdev_xmit_lock_key[ARRAY_SIZE(netdev_lock_type)];
305
static struct lock_class_key netdev_addr_lock_key[ARRAY_SIZE(netdev_lock_type)];
306 307 308 309 310 311 312 313 314 315 316 317

static inline unsigned short netdev_lock_pos(unsigned short dev_type)
{
	int i;

	for (i = 0; i < ARRAY_SIZE(netdev_lock_type); i++)
		if (netdev_lock_type[i] == dev_type)
			return i;
	/* the last key is used by default */
	return ARRAY_SIZE(netdev_lock_type) - 1;
}

318 319
static inline void netdev_set_xmit_lockdep_class(spinlock_t *lock,
						 unsigned short dev_type)
320 321 322 323 324 325 326
{
	int i;

	i = netdev_lock_pos(dev_type);
	lockdep_set_class_and_name(lock, &netdev_xmit_lock_key[i],
				   netdev_lock_name[i]);
}
327 328 329 330 331 332 333 334 335 336

static inline void netdev_set_addr_lockdep_class(struct net_device *dev)
{
	int i;

	i = netdev_lock_pos(dev->type);
	lockdep_set_class_and_name(&dev->addr_list_lock,
				   &netdev_addr_lock_key[i],
				   netdev_lock_name[i]);
}
337
#else
338 339 340 341 342
static inline void netdev_set_xmit_lockdep_class(spinlock_t *lock,
						 unsigned short dev_type)
{
}
static inline void netdev_set_addr_lockdep_class(struct net_device *dev)
343 344 345
{
}
#endif
L
Linus Torvalds 已提交
346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376

/*******************************************************************************

		Protocol management and registration routines

*******************************************************************************/

/*
 *	Add a protocol ID to the list. Now that the input handler is
 *	smarter we can dispense with all the messy stuff that used to be
 *	here.
 *
 *	BEWARE!!! Protocol handlers, mangling input packets,
 *	MUST BE last in hash buckets and checking protocol handlers
 *	MUST start from promiscuous ptype_all chain in net_bh.
 *	It is true now, do not change it.
 *	Explanation follows: if protocol handler, mangling packet, will
 *	be the first on list, it is not able to sense, that packet
 *	is cloned and should be copied-on-write, so that it will
 *	change it and subsequent readers will get broken packet.
 *							--ANK (980803)
 */

/**
 *	dev_add_pack - add packet handler
 *	@pt: packet type declaration
 *
 *	Add a protocol handler to the networking stack. The passed &packet_type
 *	is linked into kernel lists and may not be freed until it has been
 *	removed from the kernel lists.
 *
377
 *	This call does not sleep therefore it can not
L
Linus Torvalds 已提交
378 379 380 381 382 383 384 385 386
 *	guarantee all CPU's that are in middle of receiving packets
 *	will see the new packet type (until the next received packet).
 */

void dev_add_pack(struct packet_type *pt)
{
	int hash;

	spin_lock_bh(&ptype_lock);
S
Stephen Hemminger 已提交
387
	if (pt->type == htons(ETH_P_ALL))
L
Linus Torvalds 已提交
388
		list_add_rcu(&pt->list, &ptype_all);
S
Stephen Hemminger 已提交
389
	else {
390
		hash = ntohs(pt->type) & PTYPE_HASH_MASK;
L
Linus Torvalds 已提交
391 392 393 394 395 396 397 398 399 400 401 402
		list_add_rcu(&pt->list, &ptype_base[hash]);
	}
	spin_unlock_bh(&ptype_lock);
}

/**
 *	__dev_remove_pack	 - remove packet handler
 *	@pt: packet type declaration
 *
 *	Remove a protocol handler that was previously added to the kernel
 *	protocol handlers by dev_add_pack(). The passed &packet_type is removed
 *	from the kernel lists and can be freed or reused once this function
403
 *	returns.
L
Linus Torvalds 已提交
404 405 406 407 408 409 410 411 412 413 414 415
 *
 *      The packet type might still be in use by receivers
 *	and must not be freed until after all the CPU's have gone
 *	through a quiescent state.
 */
void __dev_remove_pack(struct packet_type *pt)
{
	struct list_head *head;
	struct packet_type *pt1;

	spin_lock_bh(&ptype_lock);

S
Stephen Hemminger 已提交
416
	if (pt->type == htons(ETH_P_ALL))
L
Linus Torvalds 已提交
417
		head = &ptype_all;
S
Stephen Hemminger 已提交
418
	else
419
		head = &ptype_base[ntohs(pt->type) & PTYPE_HASH_MASK];
L
Linus Torvalds 已提交
420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446

	list_for_each_entry(pt1, head, list) {
		if (pt == pt1) {
			list_del_rcu(&pt->list);
			goto out;
		}
	}

	printk(KERN_WARNING "dev_remove_pack: %p not found.\n", pt);
out:
	spin_unlock_bh(&ptype_lock);
}
/**
 *	dev_remove_pack	 - remove packet handler
 *	@pt: packet type declaration
 *
 *	Remove a protocol handler that was previously added to the kernel
 *	protocol handlers by dev_add_pack(). The passed &packet_type is removed
 *	from the kernel lists and can be freed or reused once this function
 *	returns.
 *
 *	This call sleeps to guarantee that no CPU is looking at the packet
 *	type after return.
 */
void dev_remove_pack(struct packet_type *pt)
{
	__dev_remove_pack(pt);
447

L
Linus Torvalds 已提交
448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477
	synchronize_net();
}

/******************************************************************************

		      Device Boot-time Settings Routines

*******************************************************************************/

/* Boot time configuration table */
static struct netdev_boot_setup dev_boot_setup[NETDEV_BOOT_SETUP_MAX];

/**
 *	netdev_boot_setup_add	- add new setup entry
 *	@name: name of the device
 *	@map: configured settings for the device
 *
 *	Adds new setup entry to the dev_boot_setup list.  The function
 *	returns 0 on error and 1 on success.  This is a generic routine to
 *	all netdevices.
 */
static int netdev_boot_setup_add(char *name, struct ifmap *map)
{
	struct netdev_boot_setup *s;
	int i;

	s = dev_boot_setup;
	for (i = 0; i < NETDEV_BOOT_SETUP_MAX; i++) {
		if (s[i].name[0] == '\0' || s[i].name[0] == ' ') {
			memset(s[i].name, 0, sizeof(s[i].name));
478
			strlcpy(s[i].name, name, IFNAMSIZ);
L
Linus Torvalds 已提交
479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502
			memcpy(&s[i].map, map, sizeof(s[i].map));
			break;
		}
	}

	return i >= NETDEV_BOOT_SETUP_MAX ? 0 : 1;
}

/**
 *	netdev_boot_setup_check	- check boot time settings
 *	@dev: the netdevice
 *
 * 	Check boot time settings for the device.
 *	The found settings are set for the device to be used
 *	later in the device probing.
 *	Returns 0 if no settings found, 1 if they are.
 */
int netdev_boot_setup_check(struct net_device *dev)
{
	struct netdev_boot_setup *s = dev_boot_setup;
	int i;

	for (i = 0; i < NETDEV_BOOT_SETUP_MAX; i++) {
		if (s[i].name[0] != '\0' && s[i].name[0] != ' ' &&
503
		    !strcmp(dev->name, s[i].name)) {
L
Linus Torvalds 已提交
504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536
			dev->irq 	= s[i].map.irq;
			dev->base_addr 	= s[i].map.base_addr;
			dev->mem_start 	= s[i].map.mem_start;
			dev->mem_end 	= s[i].map.mem_end;
			return 1;
		}
	}
	return 0;
}


/**
 *	netdev_boot_base	- get address from boot time settings
 *	@prefix: prefix for network device
 *	@unit: id for network device
 *
 * 	Check boot time settings for the base address of device.
 *	The found settings are set for the device to be used
 *	later in the device probing.
 *	Returns 0 if no settings found.
 */
unsigned long netdev_boot_base(const char *prefix, int unit)
{
	const struct netdev_boot_setup *s = dev_boot_setup;
	char name[IFNAMSIZ];
	int i;

	sprintf(name, "%s%d", prefix, unit);

	/*
	 * If device already registered then return base of 1
	 * to indicate not to probe for this interface
	 */
537
	if (__dev_get_by_name(&init_net, name))
L
Linus Torvalds 已提交
538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582
		return 1;

	for (i = 0; i < NETDEV_BOOT_SETUP_MAX; i++)
		if (!strcmp(name, s[i].name))
			return s[i].map.base_addr;
	return 0;
}

/*
 * Saves at boot time configured settings for any netdevice.
 */
int __init netdev_boot_setup(char *str)
{
	int ints[5];
	struct ifmap map;

	str = get_options(str, ARRAY_SIZE(ints), ints);
	if (!str || !*str)
		return 0;

	/* Save settings */
	memset(&map, 0, sizeof(map));
	if (ints[0] > 0)
		map.irq = ints[1];
	if (ints[0] > 1)
		map.base_addr = ints[2];
	if (ints[0] > 2)
		map.mem_start = ints[3];
	if (ints[0] > 3)
		map.mem_end = ints[4];

	/* Add new entry to the list */
	return netdev_boot_setup_add(str, &map);
}

__setup("netdev=", netdev_boot_setup);

/*******************************************************************************

			    Device Interface Subroutines

*******************************************************************************/

/**
 *	__dev_get_by_name	- find a device by its name
583
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
584 585 586 587 588 589 590 591 592
 *	@name: name to find
 *
 *	Find an interface by name. Must be called under RTNL semaphore
 *	or @dev_base_lock. If the name is found a pointer to the device
 *	is returned. If the name is not found then %NULL is returned. The
 *	reference counters are not incremented so the caller must be
 *	careful with locks.
 */

593
struct net_device *__dev_get_by_name(struct net *net, const char *name)
L
Linus Torvalds 已提交
594 595 596
{
	struct hlist_node *p;

597
	hlist_for_each(p, dev_name_hash(net, name)) {
L
Linus Torvalds 已提交
598 599 600 601 602 603 604 605 606 607
		struct net_device *dev
			= hlist_entry(p, struct net_device, name_hlist);
		if (!strncmp(dev->name, name, IFNAMSIZ))
			return dev;
	}
	return NULL;
}

/**
 *	dev_get_by_name		- find a device by its name
608
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
609 610 611 612 613 614 615 616 617
 *	@name: name to find
 *
 *	Find an interface by name. This can be called from any
 *	context and does its own locking. The returned handle has
 *	the usage count incremented and the caller must use dev_put() to
 *	release it when it is no longer needed. %NULL is returned if no
 *	matching device is found.
 */

618
struct net_device *dev_get_by_name(struct net *net, const char *name)
L
Linus Torvalds 已提交
619 620 621 622
{
	struct net_device *dev;

	read_lock(&dev_base_lock);
623
	dev = __dev_get_by_name(net, name);
L
Linus Torvalds 已提交
624 625 626 627 628 629 630 631
	if (dev)
		dev_hold(dev);
	read_unlock(&dev_base_lock);
	return dev;
}

/**
 *	__dev_get_by_index - find a device by its ifindex
632
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
633 634 635 636 637 638 639 640 641
 *	@ifindex: index of device
 *
 *	Search for an interface by index. Returns %NULL if the device
 *	is not found or a pointer to the device. The device has not
 *	had its reference counter increased so the caller must be careful
 *	about locking. The caller must hold either the RTNL semaphore
 *	or @dev_base_lock.
 */

642
struct net_device *__dev_get_by_index(struct net *net, int ifindex)
L
Linus Torvalds 已提交
643 644 645
{
	struct hlist_node *p;

646
	hlist_for_each(p, dev_index_hash(net, ifindex)) {
L
Linus Torvalds 已提交
647 648 649 650 651 652 653 654 655 656 657
		struct net_device *dev
			= hlist_entry(p, struct net_device, index_hlist);
		if (dev->ifindex == ifindex)
			return dev;
	}
	return NULL;
}


/**
 *	dev_get_by_index - find a device by its ifindex
658
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
659 660 661 662 663 664 665 666
 *	@ifindex: index of device
 *
 *	Search for an interface by index. Returns NULL if the device
 *	is not found or a pointer to the device. The device returned has
 *	had a reference added and the pointer is safe until the user calls
 *	dev_put to indicate they have finished with it.
 */

667
struct net_device *dev_get_by_index(struct net *net, int ifindex)
L
Linus Torvalds 已提交
668 669 670 671
{
	struct net_device *dev;

	read_lock(&dev_base_lock);
672
	dev = __dev_get_by_index(net, ifindex);
L
Linus Torvalds 已提交
673 674 675 676 677 678 679 680
	if (dev)
		dev_hold(dev);
	read_unlock(&dev_base_lock);
	return dev;
}

/**
 *	dev_getbyhwaddr - find a device by its hardware address
681
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
682 683 684 685 686 687 688 689 690 691 692 693
 *	@type: media type of device
 *	@ha: hardware address
 *
 *	Search for an interface by MAC address. Returns NULL if the device
 *	is not found or a pointer to the device. The caller must hold the
 *	rtnl semaphore. The returned device has not had its ref count increased
 *	and the caller must therefore be careful about locking
 *
 *	BUGS:
 *	If the API was consistent this would be __dev_get_by_hwaddr
 */

694
struct net_device *dev_getbyhwaddr(struct net *net, unsigned short type, char *ha)
L
Linus Torvalds 已提交
695 696 697 698 699
{
	struct net_device *dev;

	ASSERT_RTNL();

700
	for_each_netdev(net, dev)
L
Linus Torvalds 已提交
701 702
		if (dev->type == type &&
		    !memcmp(dev->dev_addr, ha, dev->addr_len))
703 704 705
			return dev;

	return NULL;
L
Linus Torvalds 已提交
706 707
}

708 709
EXPORT_SYMBOL(dev_getbyhwaddr);

710
struct net_device *__dev_getfirstbyhwtype(struct net *net, unsigned short type)
L
Linus Torvalds 已提交
711 712 713
{
	struct net_device *dev;

714
	ASSERT_RTNL();
715
	for_each_netdev(net, dev)
716
		if (dev->type == type)
717 718 719
			return dev;

	return NULL;
720 721 722 723
}

EXPORT_SYMBOL(__dev_getfirstbyhwtype);

724
struct net_device *dev_getfirstbyhwtype(struct net *net, unsigned short type)
725 726 727 728
{
	struct net_device *dev;

	rtnl_lock();
729
	dev = __dev_getfirstbyhwtype(net, type);
730 731
	if (dev)
		dev_hold(dev);
L
Linus Torvalds 已提交
732 733 734 735 736 737 738 739
	rtnl_unlock();
	return dev;
}

EXPORT_SYMBOL(dev_getfirstbyhwtype);

/**
 *	dev_get_by_flags - find any device with given flags
740
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
741 742 743 744
 *	@if_flags: IFF_* values
 *	@mask: bitmask of bits in if_flags to check
 *
 *	Search for any interface with the given flags. Returns NULL if a device
745
 *	is not found or a pointer to the device. The device returned has
L
Linus Torvalds 已提交
746 747 748 749
 *	had a reference added and the pointer is safe until the user calls
 *	dev_put to indicate they have finished with it.
 */

750
struct net_device * dev_get_by_flags(struct net *net, unsigned short if_flags, unsigned short mask)
L
Linus Torvalds 已提交
751
{
752
	struct net_device *dev, *ret;
L
Linus Torvalds 已提交
753

754
	ret = NULL;
L
Linus Torvalds 已提交
755
	read_lock(&dev_base_lock);
756
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
757 758
		if (((dev->flags ^ if_flags) & mask) == 0) {
			dev_hold(dev);
759
			ret = dev;
L
Linus Torvalds 已提交
760 761 762 763
			break;
		}
	}
	read_unlock(&dev_base_lock);
764
	return ret;
L
Linus Torvalds 已提交
765 766 767 768 769 770 771
}

/**
 *	dev_valid_name - check if name is okay for network device
 *	@name: name string
 *
 *	Network device names need to be valid file names to
772 773
 *	to allow sysfs to work.  We also disallow any kind of
 *	whitespace.
L
Linus Torvalds 已提交
774
 */
775
int dev_valid_name(const char *name)
L
Linus Torvalds 已提交
776
{
777 778
	if (*name == '\0')
		return 0;
779 780
	if (strlen(name) >= IFNAMSIZ)
		return 0;
781 782 783 784 785 786 787 788 789
	if (!strcmp(name, ".") || !strcmp(name, ".."))
		return 0;

	while (*name) {
		if (*name == '/' || isspace(*name))
			return 0;
		name++;
	}
	return 1;
L
Linus Torvalds 已提交
790 791 792
}

/**
793 794
 *	__dev_alloc_name - allocate a name for a device
 *	@net: network namespace to allocate the device name in
L
Linus Torvalds 已提交
795
 *	@name: name format string
796
 *	@buf:  scratch buffer and result name string
L
Linus Torvalds 已提交
797 798
 *
 *	Passed a format string - eg "lt%d" it will try and find a suitable
S
Stephen Hemminger 已提交
799 800 801 802 803 804
 *	id. It scans list of devices to build up a free map, then chooses
 *	the first empty slot. The caller must hold the dev_base or rtnl lock
 *	while allocating the name and adding the device in order to avoid
 *	duplicates.
 *	Limited to bits_per_byte * page size devices (ie 32K on most platforms).
 *	Returns the number of the unit assigned or a negative errno code.
L
Linus Torvalds 已提交
805 806
 */

807
static int __dev_alloc_name(struct net *net, const char *name, char *buf)
L
Linus Torvalds 已提交
808 809 810 811
{
	int i = 0;
	const char *p;
	const int max_netdevices = 8*PAGE_SIZE;
S
Stephen Hemminger 已提交
812
	unsigned long *inuse;
L
Linus Torvalds 已提交
813 814 815 816 817 818 819 820 821 822 823 824 825
	struct net_device *d;

	p = strnchr(name, IFNAMSIZ-1, '%');
	if (p) {
		/*
		 * Verify the string as this thing may have come from
		 * the user.  There must be either one "%d" and no other "%"
		 * characters.
		 */
		if (p[1] != 'd' || strchr(p + 2, '%'))
			return -EINVAL;

		/* Use one page as a bit array of possible slots */
S
Stephen Hemminger 已提交
826
		inuse = (unsigned long *) get_zeroed_page(GFP_ATOMIC);
L
Linus Torvalds 已提交
827 828 829
		if (!inuse)
			return -ENOMEM;

830
		for_each_netdev(net, d) {
L
Linus Torvalds 已提交
831 832 833 834 835 836
			if (!sscanf(d->name, name, &i))
				continue;
			if (i < 0 || i >= max_netdevices)
				continue;

			/*  avoid cases where sscanf is not exact inverse of printf */
837
			snprintf(buf, IFNAMSIZ, name, i);
L
Linus Torvalds 已提交
838 839 840 841 842 843 844 845
			if (!strncmp(buf, d->name, IFNAMSIZ))
				set_bit(i, inuse);
		}

		i = find_first_zero_bit(inuse, max_netdevices);
		free_page((unsigned long) inuse);
	}

846 847
	snprintf(buf, IFNAMSIZ, name, i);
	if (!__dev_get_by_name(net, buf))
L
Linus Torvalds 已提交
848 849 850 851 852 853 854 855 856
		return i;

	/* It is possible to run out of possible slots
	 * when the name is long and there isn't enough space left
	 * for the digits, or if all bits are used.
	 */
	return -ENFILE;
}

857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876
/**
 *	dev_alloc_name - allocate a name for a device
 *	@dev: device
 *	@name: name format string
 *
 *	Passed a format string - eg "lt%d" it will try and find a suitable
 *	id. It scans list of devices to build up a free map, then chooses
 *	the first empty slot. The caller must hold the dev_base or rtnl lock
 *	while allocating the name and adding the device in order to avoid
 *	duplicates.
 *	Limited to bits_per_byte * page size devices (ie 32K on most platforms).
 *	Returns the number of the unit assigned or a negative errno code.
 */

int dev_alloc_name(struct net_device *dev, const char *name)
{
	char buf[IFNAMSIZ];
	struct net *net;
	int ret;

877 878
	BUG_ON(!dev_net(dev));
	net = dev_net(dev);
879 880 881 882 883 884
	ret = __dev_alloc_name(net, name, buf);
	if (ret >= 0)
		strlcpy(dev->name, buf, IFNAMSIZ);
	return ret;
}

L
Linus Torvalds 已提交
885 886 887 888 889 890 891 892 893 894 895

/**
 *	dev_change_name - change name of a device
 *	@dev: device
 *	@newname: name (or format string) must be at least IFNAMSIZ
 *
 *	Change name of a device, can pass format strings "eth%d".
 *	for wildcarding.
 */
int dev_change_name(struct net_device *dev, char *newname)
{
896
	char oldname[IFNAMSIZ];
L
Linus Torvalds 已提交
897
	int err = 0;
898
	int ret;
899
	struct net *net;
L
Linus Torvalds 已提交
900 901

	ASSERT_RTNL();
902
	BUG_ON(!dev_net(dev));
L
Linus Torvalds 已提交
903

904
	net = dev_net(dev);
L
Linus Torvalds 已提交
905 906 907 908 909 910
	if (dev->flags & IFF_UP)
		return -EBUSY;

	if (!dev_valid_name(newname))
		return -EINVAL;

911 912 913
	if (strncmp(newname, dev->name, IFNAMSIZ) == 0)
		return 0;

914 915
	memcpy(oldname, dev->name, IFNAMSIZ);

L
Linus Torvalds 已提交
916 917 918 919 920 921
	if (strchr(newname, '%')) {
		err = dev_alloc_name(dev, newname);
		if (err < 0)
			return err;
		strcpy(newname, dev->name);
	}
922
	else if (__dev_get_by_name(net, newname))
L
Linus Torvalds 已提交
923 924 925 926
		return -EEXIST;
	else
		strlcpy(dev->name, newname, IFNAMSIZ);

927
rollback:
928 929 930 931 932
	err = device_rename(&dev->dev, dev->name);
	if (err) {
		memcpy(dev->name, oldname, IFNAMSIZ);
		return err;
	}
933 934

	write_lock_bh(&dev_base_lock);
935
	hlist_del(&dev->name_hlist);
936
	hlist_add_head(&dev->name_hlist, dev_name_hash(net, dev->name));
937 938
	write_unlock_bh(&dev_base_lock);

939
	ret = call_netdevice_notifiers(NETDEV_CHANGENAME, dev);
940 941 942 943 944 945 946 947 948 949 950 951 952
	ret = notifier_to_errno(ret);

	if (ret) {
		if (err) {
			printk(KERN_ERR
			       "%s: name change rollback failed: %d.\n",
			       dev->name, ret);
		} else {
			err = ret;
			memcpy(dev->name, oldname, IFNAMSIZ);
			goto rollback;
		}
	}
L
Linus Torvalds 已提交
953 954 955 956

	return err;
}

957
/**
S
Stephen Hemminger 已提交
958
 *	netdev_features_change - device changes features
959 960 961 962 963 964
 *	@dev: device to cause notification
 *
 *	Called to indicate a device has changed features.
 */
void netdev_features_change(struct net_device *dev)
{
965
	call_netdevice_notifiers(NETDEV_FEAT_CHANGE, dev);
966 967 968
}
EXPORT_SYMBOL(netdev_features_change);

L
Linus Torvalds 已提交
969 970 971 972 973 974 975 976 977 978 979
/**
 *	netdev_state_change - device changes state
 *	@dev: device to cause notification
 *
 *	Called to indicate a device has changed state. This function calls
 *	the notifier chains for netdev_chain and sends a NEWLINK message
 *	to the routing socket.
 */
void netdev_state_change(struct net_device *dev)
{
	if (dev->flags & IFF_UP) {
980
		call_netdevice_notifiers(NETDEV_CHANGE, dev);
L
Linus Torvalds 已提交
981 982 983 984
		rtmsg_ifinfo(RTM_NEWLINK, dev, 0);
	}
}

985 986 987 988 989 990
void netdev_bonding_change(struct net_device *dev)
{
	call_netdevice_notifiers(NETDEV_BONDING_FAILOVER, dev);
}
EXPORT_SYMBOL(netdev_bonding_change);

L
Linus Torvalds 已提交
991 992
/**
 *	dev_load 	- load a network module
993
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
994 995 996 997 998 999 1000
 *	@name: name of interface
 *
 *	If a network interface is not present and the process has suitable
 *	privileges this function loads the module. If module loading is not
 *	available in this kernel then it becomes a nop.
 */

1001
void dev_load(struct net *net, const char *name)
L
Linus Torvalds 已提交
1002
{
1003
	struct net_device *dev;
L
Linus Torvalds 已提交
1004 1005

	read_lock(&dev_base_lock);
1006
	dev = __dev_get_by_name(net, name);
L
Linus Torvalds 已提交
1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028
	read_unlock(&dev_base_lock);

	if (!dev && capable(CAP_SYS_MODULE))
		request_module("%s", name);
}

/**
 *	dev_open	- prepare an interface for use.
 *	@dev:	device to open
 *
 *	Takes a device from down to up state. The device's private open
 *	function is invoked and then the multicast lists are loaded. Finally
 *	the device is moved into the up state and a %NETDEV_UP message is
 *	sent to the netdev notifier chain.
 *
 *	Calling this function on an active interface is a nop. On a failure
 *	a negative errno code is returned.
 */
int dev_open(struct net_device *dev)
{
	int ret = 0;

1029 1030
	ASSERT_RTNL();

L
Linus Torvalds 已提交
1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047
	/*
	 *	Is it already up?
	 */

	if (dev->flags & IFF_UP)
		return 0;

	/*
	 *	Is it even present?
	 */
	if (!netif_device_present(dev))
		return -ENODEV;

	/*
	 *	Call device private open method
	 */
	set_bit(__LINK_STATE_START, &dev->state);
1048 1049 1050 1051 1052

	if (dev->validate_addr)
		ret = dev->validate_addr(dev);

	if (!ret && dev->open)
L
Linus Torvalds 已提交
1053 1054
		ret = dev->open(dev);

1055
	/*
L
Linus Torvalds 已提交
1056 1057 1058
	 *	If it went open OK then:
	 */

1059 1060 1061
	if (ret)
		clear_bit(__LINK_STATE_START, &dev->state);
	else {
L
Linus Torvalds 已提交
1062 1063 1064 1065 1066 1067 1068 1069
		/*
		 *	Set the flags.
		 */
		dev->flags |= IFF_UP;

		/*
		 *	Initialize multicasting status
		 */
1070
		dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
1071 1072 1073 1074 1075 1076 1077 1078 1079

		/*
		 *	Wakeup transmit queue engine
		 */
		dev_activate(dev);

		/*
		 *	... and announce new interface.
		 */
1080
		call_netdevice_notifiers(NETDEV_UP, dev);
L
Linus Torvalds 已提交
1081
	}
1082

L
Linus Torvalds 已提交
1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096
	return ret;
}

/**
 *	dev_close - shutdown an interface.
 *	@dev: device to shutdown
 *
 *	This function moves an active device into down state. A
 *	%NETDEV_GOING_DOWN is sent to the netdev notifier chain. The device
 *	is then deactivated and finally a %NETDEV_DOWN is sent to the notifier
 *	chain.
 */
int dev_close(struct net_device *dev)
{
1097 1098
	ASSERT_RTNL();

1099 1100
	might_sleep();

L
Linus Torvalds 已提交
1101 1102 1103 1104 1105 1106 1107
	if (!(dev->flags & IFF_UP))
		return 0;

	/*
	 *	Tell people we are going down, so that they can
	 *	prepare to death, when device is still operating.
	 */
1108
	call_netdevice_notifiers(NETDEV_GOING_DOWN, dev);
L
Linus Torvalds 已提交
1109 1110 1111 1112

	clear_bit(__LINK_STATE_START, &dev->state);

	/* Synchronize to scheduled poll. We cannot touch poll list,
1113 1114 1115 1116 1117
	 * it can be even on different cpu. So just clear netif_running().
	 *
	 * dev->stop() will invoke napi_disable() on all of it's
	 * napi_struct instances on this device.
	 */
L
Linus Torvalds 已提交
1118 1119
	smp_mb__after_clear_bit(); /* Commit netif_running(). */

1120 1121
	dev_deactivate(dev);

L
Linus Torvalds 已提交
1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140
	/*
	 *	Call the device specific close. This cannot fail.
	 *	Only if device is UP
	 *
	 *	We allow it to be called even after a DETACH hot-plug
	 *	event.
	 */
	if (dev->stop)
		dev->stop(dev);

	/*
	 *	Device is now down.
	 */

	dev->flags &= ~IFF_UP;

	/*
	 * Tell people we are down
	 */
1141
	call_netdevice_notifiers(NETDEV_DOWN, dev);
L
Linus Torvalds 已提交
1142 1143 1144 1145 1146

	return 0;
}


1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169
/**
 *	dev_disable_lro - disable Large Receive Offload on a device
 *	@dev: device
 *
 *	Disable Large Receive Offload (LRO) on a net device.  Must be
 *	called under RTNL.  This is needed if received packets may be
 *	forwarded to another interface.
 */
void dev_disable_lro(struct net_device *dev)
{
	if (dev->ethtool_ops && dev->ethtool_ops->get_flags &&
	    dev->ethtool_ops->set_flags) {
		u32 flags = dev->ethtool_ops->get_flags(dev);
		if (flags & ETH_FLAG_LRO) {
			flags &= ~ETH_FLAG_LRO;
			dev->ethtool_ops->set_flags(dev, flags);
		}
	}
	WARN_ON(dev->features & NETIF_F_LRO);
}
EXPORT_SYMBOL(dev_disable_lro);


1170 1171
static int dev_boot_phase = 1;

L
Linus Torvalds 已提交
1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186
/*
 *	Device change register/unregister. These are not inline or static
 *	as we export them to the world.
 */

/**
 *	register_netdevice_notifier - register a network notifier block
 *	@nb: notifier
 *
 *	Register a notifier to be called when network device events occur.
 *	The notifier passed is linked into the kernel structures and must
 *	not be reused until it has been unregistered. A negative errno code
 *	is returned on a failure.
 *
 * 	When registered all registration and up events are replayed
1187
 *	to the new notifier to allow device to have a race free
L
Linus Torvalds 已提交
1188 1189 1190 1191 1192 1193
 *	view of the network device list.
 */

int register_netdevice_notifier(struct notifier_block *nb)
{
	struct net_device *dev;
1194
	struct net_device *last;
1195
	struct net *net;
L
Linus Torvalds 已提交
1196 1197 1198
	int err;

	rtnl_lock();
1199
	err = raw_notifier_chain_register(&netdev_chain, nb);
1200 1201
	if (err)
		goto unlock;
1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212
	if (dev_boot_phase)
		goto unlock;
	for_each_net(net) {
		for_each_netdev(net, dev) {
			err = nb->notifier_call(nb, NETDEV_REGISTER, dev);
			err = notifier_to_errno(err);
			if (err)
				goto rollback;

			if (!(dev->flags & IFF_UP))
				continue;
L
Linus Torvalds 已提交
1213

1214 1215
			nb->notifier_call(nb, NETDEV_UP, dev);
		}
L
Linus Torvalds 已提交
1216
	}
1217 1218

unlock:
L
Linus Torvalds 已提交
1219 1220
	rtnl_unlock();
	return err;
1221 1222 1223

rollback:
	last = dev;
1224 1225 1226 1227
	for_each_net(net) {
		for_each_netdev(net, dev) {
			if (dev == last)
				break;
1228

1229 1230 1231 1232 1233
			if (dev->flags & IFF_UP) {
				nb->notifier_call(nb, NETDEV_GOING_DOWN, dev);
				nb->notifier_call(nb, NETDEV_DOWN, dev);
			}
			nb->notifier_call(nb, NETDEV_UNREGISTER, dev);
1234 1235
		}
	}
1236 1237

	raw_notifier_chain_unregister(&netdev_chain, nb);
1238
	goto unlock;
L
Linus Torvalds 已提交
1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252
}

/**
 *	unregister_netdevice_notifier - unregister a network notifier block
 *	@nb: notifier
 *
 *	Unregister a notifier previously registered by
 *	register_netdevice_notifier(). The notifier is unlinked into the
 *	kernel structures and may then be reused. A negative errno code
 *	is returned on a failure.
 */

int unregister_netdevice_notifier(struct notifier_block *nb)
{
1253 1254 1255
	int err;

	rtnl_lock();
1256
	err = raw_notifier_chain_unregister(&netdev_chain, nb);
1257 1258
	rtnl_unlock();
	return err;
L
Linus Torvalds 已提交
1259 1260 1261 1262 1263
}

/**
 *	call_netdevice_notifiers - call all network notifier blocks
 *      @val: value passed unmodified to notifier function
1264
 *      @dev: net_device pointer passed unmodified to notifier function
L
Linus Torvalds 已提交
1265 1266
 *
 *	Call all network notifier blocks.  Parameters and return value
1267
 *	are as for raw_notifier_call_chain().
L
Linus Torvalds 已提交
1268 1269
 */

1270
int call_netdevice_notifiers(unsigned long val, struct net_device *dev)
L
Linus Torvalds 已提交
1271
{
1272
	return raw_notifier_call_chain(&netdev_chain, val, dev);
L
Linus Torvalds 已提交
1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287
}

/* When > 0 there are consumers of rx skb time stamps */
static atomic_t netstamp_needed = ATOMIC_INIT(0);

void net_enable_timestamp(void)
{
	atomic_inc(&netstamp_needed);
}

void net_disable_timestamp(void)
{
	atomic_dec(&netstamp_needed);
}

1288
static inline void net_timestamp(struct sk_buff *skb)
L
Linus Torvalds 已提交
1289 1290
{
	if (atomic_read(&netstamp_needed))
1291
		__net_timestamp(skb);
1292 1293
	else
		skb->tstamp.tv64 = 0;
L
Linus Torvalds 已提交
1294 1295 1296 1297 1298 1299 1300
}

/*
 *	Support routine. Sends outgoing frames to any network
 *	taps currently in use.
 */

1301
static void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev)
L
Linus Torvalds 已提交
1302 1303
{
	struct packet_type *ptype;
1304 1305

	net_timestamp(skb);
L
Linus Torvalds 已提交
1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322

	rcu_read_lock();
	list_for_each_entry_rcu(ptype, &ptype_all, list) {
		/* Never send packets back to the socket
		 * they originated from - MvS (miquels@drinkel.ow.org)
		 */
		if ((ptype->dev == dev || !ptype->dev) &&
		    (ptype->af_packet_priv == NULL ||
		     (struct sock *)ptype->af_packet_priv != skb->sk)) {
			struct sk_buff *skb2= skb_clone(skb, GFP_ATOMIC);
			if (!skb2)
				break;

			/* skb->nh should be correctly
			   set by sender, so that the second statement is
			   just protection against buggy protocols.
			 */
1323
			skb_reset_mac_header(skb2);
L
Linus Torvalds 已提交
1324

1325
			if (skb_network_header(skb2) < skb2->data ||
1326
			    skb2->network_header > skb2->tail) {
L
Linus Torvalds 已提交
1327 1328 1329 1330
				if (net_ratelimit())
					printk(KERN_CRIT "protocol %04x is "
					       "buggy, dev %s\n",
					       skb2->protocol, dev->name);
1331
				skb_reset_network_header(skb2);
L
Linus Torvalds 已提交
1332 1333
			}

1334
			skb2->transport_header = skb2->network_header;
L
Linus Torvalds 已提交
1335
			skb2->pkt_type = PACKET_OUTGOING;
D
David S. Miller 已提交
1336
			ptype->func(skb2, skb->dev, ptype, skb->dev);
L
Linus Torvalds 已提交
1337 1338 1339 1340 1341
		}
	}
	rcu_read_unlock();
}

1342

1343
static inline void __netif_reschedule(struct Qdisc *q)
1344
{
1345 1346
	struct softnet_data *sd;
	unsigned long flags;
1347

1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359
	local_irq_save(flags);
	sd = &__get_cpu_var(softnet_data);
	q->next_sched = sd->output_queue;
	sd->output_queue = q;
	raise_softirq_irqoff(NET_TX_SOFTIRQ);
	local_irq_restore(flags);
}

void __netif_schedule(struct Qdisc *q)
{
	if (!test_and_set_bit(__QDISC_STATE_SCHED, &q->state))
		__netif_reschedule(q);
1360 1361 1362
}
EXPORT_SYMBOL(__netif_schedule);

1363
void dev_kfree_skb_irq(struct sk_buff *skb)
1364
{
1365 1366 1367
	if (atomic_dec_and_test(&skb->users)) {
		struct softnet_data *sd;
		unsigned long flags;
1368

1369 1370 1371 1372 1373 1374 1375
		local_irq_save(flags);
		sd = &__get_cpu_var(softnet_data);
		skb->next = sd->completion_queue;
		sd->completion_queue = skb;
		raise_softirq_irqoff(NET_TX_SOFTIRQ);
		local_irq_restore(flags);
	}
1376
}
1377
EXPORT_SYMBOL(dev_kfree_skb_irq);
1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388

void dev_kfree_skb_any(struct sk_buff *skb)
{
	if (in_irq() || irqs_disabled())
		dev_kfree_skb_irq(skb);
	else
		dev_kfree_skb(skb);
}
EXPORT_SYMBOL(dev_kfree_skb_any);


1389 1390 1391 1392 1393 1394
/**
 * netif_device_detach - mark device as removed
 * @dev: network device
 *
 * Mark device as removed from system and therefore no longer available.
 */
1395 1396 1397 1398 1399 1400 1401 1402 1403
void netif_device_detach(struct net_device *dev)
{
	if (test_and_clear_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
		netif_stop_queue(dev);
	}
}
EXPORT_SYMBOL(netif_device_detach);

1404 1405 1406 1407 1408 1409
/**
 * netif_device_attach - mark device as attached
 * @dev: network device
 *
 * Mark device as attached from system and restart if needed.
 */
1410 1411 1412 1413 1414
void netif_device_attach(struct net_device *dev)
{
	if (!test_and_set_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
		netif_wake_queue(dev);
1415
		__netdev_watchdog_up(dev);
1416 1417 1418 1419
	}
}
EXPORT_SYMBOL(netif_device_attach);

1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442
static bool can_checksum_protocol(unsigned long features, __be16 protocol)
{
	return ((features & NETIF_F_GEN_CSUM) ||
		((features & NETIF_F_IP_CSUM) &&
		 protocol == htons(ETH_P_IP)) ||
		((features & NETIF_F_IPV6_CSUM) &&
		 protocol == htons(ETH_P_IPV6)));
}

static bool dev_can_checksum(struct net_device *dev, struct sk_buff *skb)
{
	if (can_checksum_protocol(dev->features, skb->protocol))
		return true;

	if (skb->protocol == htons(ETH_P_8021Q)) {
		struct vlan_ethhdr *veh = (struct vlan_ethhdr *)skb->data;
		if (can_checksum_protocol(dev->features & dev->vlan_features,
					  veh->h_vlan_encapsulated_proto))
			return true;
	}

	return false;
}
1443

L
Linus Torvalds 已提交
1444 1445 1446 1447
/*
 * Invalidate hardware checksum when packet is to be mangled, and
 * complete checksum manually on outgoing path.
 */
1448
int skb_checksum_help(struct sk_buff *skb)
L
Linus Torvalds 已提交
1449
{
1450
	__wsum csum;
1451
	int ret = 0, offset;
L
Linus Torvalds 已提交
1452

1453
	if (skb->ip_summed == CHECKSUM_COMPLETE)
1454 1455 1456 1457 1458
		goto out_set_summed;

	if (unlikely(skb_shinfo(skb)->gso_size)) {
		/* Let GSO fix up the checksum. */
		goto out_set_summed;
L
Linus Torvalds 已提交
1459 1460
	}

1461 1462 1463 1464 1465 1466 1467 1468 1469
	offset = skb->csum_start - skb_headroom(skb);
	BUG_ON(offset >= skb_headlen(skb));
	csum = skb_checksum(skb, offset, skb->len - offset, 0);

	offset += skb->csum_offset;
	BUG_ON(offset + sizeof(__sum16) > skb_headlen(skb));

	if (skb_cloned(skb) &&
	    !skb_clone_writable(skb, offset + sizeof(__sum16))) {
L
Linus Torvalds 已提交
1470 1471 1472 1473 1474
		ret = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
		if (ret)
			goto out;
	}

1475
	*(__sum16 *)(skb->data + offset) = csum_fold(csum);
1476
out_set_summed:
L
Linus Torvalds 已提交
1477
	skb->ip_summed = CHECKSUM_NONE;
1478
out:
L
Linus Torvalds 已提交
1479 1480 1481
	return ret;
}

1482 1483 1484
/**
 *	skb_gso_segment - Perform segmentation on skb.
 *	@skb: buffer to segment
1485
 *	@features: features for the output path (see dev->features)
1486 1487
 *
 *	This function segments the given skb and returns a list of segments.
1488 1489 1490
 *
 *	It may return NULL if the skb requires no segmentation.  This is
 *	only possible when GSO is used for verifying header integrity.
1491
 */
1492
struct sk_buff *skb_gso_segment(struct sk_buff *skb, int features)
1493 1494 1495
{
	struct sk_buff *segs = ERR_PTR(-EPROTONOSUPPORT);
	struct packet_type *ptype;
A
Al Viro 已提交
1496
	__be16 type = skb->protocol;
1497
	int err;
1498 1499 1500

	BUG_ON(skb_shinfo(skb)->frag_list);

1501
	skb_reset_mac_header(skb);
1502
	skb->mac_len = skb->network_header - skb->mac_header;
1503 1504
	__skb_pull(skb, skb->mac_len);

1505
	if (WARN_ON(skb->ip_summed != CHECKSUM_PARTIAL)) {
1506 1507 1508 1509 1510
		if (skb_header_cloned(skb) &&
		    (err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
			return ERR_PTR(err);
	}

1511
	rcu_read_lock();
1512 1513
	list_for_each_entry_rcu(ptype,
			&ptype_base[ntohs(type) & PTYPE_HASH_MASK], list) {
1514
		if (ptype->type == type && !ptype->dev && ptype->gso_segment) {
1515
			if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) {
1516 1517 1518 1519
				err = ptype->gso_send_check(skb);
				segs = ERR_PTR(err);
				if (err || skb_gso_ok(skb, features))
					break;
1520 1521
				__skb_push(skb, (skb->data -
						 skb_network_header(skb)));
1522
			}
1523
			segs = ptype->gso_segment(skb, features);
1524 1525 1526 1527 1528
			break;
		}
	}
	rcu_read_unlock();

1529
	__skb_push(skb, skb->data - skb_mac_header(skb));
1530

1531 1532 1533 1534 1535
	return segs;
}

EXPORT_SYMBOL(skb_gso_segment);

1536 1537 1538 1539 1540
/* Take action when hardware reception checksum errors are detected. */
#ifdef CONFIG_BUG
void netdev_rx_csum_fault(struct net_device *dev)
{
	if (net_ratelimit()) {
1541
		printk(KERN_ERR "%s: hw csum failure.\n",
1542
			dev ? dev->name : "<unknown>");
1543 1544 1545 1546 1547 1548
		dump_stack();
	}
}
EXPORT_SYMBOL(netdev_rx_csum_fault);
#endif

L
Linus Torvalds 已提交
1549 1550 1551 1552 1553 1554 1555
/* Actually, we should eliminate this check as soon as we know, that:
 * 1. IOMMU is present and allows to map all the memory.
 * 2. No high memory really exists on this machine.
 */

static inline int illegal_highdma(struct net_device *dev, struct sk_buff *skb)
{
1556
#ifdef CONFIG_HIGHMEM
L
Linus Torvalds 已提交
1557 1558 1559 1560 1561 1562 1563 1564 1565
	int i;

	if (dev->features & NETIF_F_HIGHDMA)
		return 0;

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
		if (PageHighMem(skb_shinfo(skb)->frags[i].page))
			return 1;

1566
#endif
L
Linus Torvalds 已提交
1567 1568 1569
	return 0;
}

1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603
struct dev_gso_cb {
	void (*destructor)(struct sk_buff *skb);
};

#define DEV_GSO_CB(skb) ((struct dev_gso_cb *)(skb)->cb)

static void dev_gso_skb_destructor(struct sk_buff *skb)
{
	struct dev_gso_cb *cb;

	do {
		struct sk_buff *nskb = skb->next;

		skb->next = nskb->next;
		nskb->next = NULL;
		kfree_skb(nskb);
	} while (skb->next);

	cb = DEV_GSO_CB(skb);
	if (cb->destructor)
		cb->destructor(skb);
}

/**
 *	dev_gso_segment - Perform emulated hardware segmentation on skb.
 *	@skb: buffer to segment
 *
 *	This function segments the given skb and stores the list of segments
 *	in skb->next.
 */
static int dev_gso_segment(struct sk_buff *skb)
{
	struct net_device *dev = skb->dev;
	struct sk_buff *segs;
1604 1605 1606 1607 1608 1609 1610 1611
	int features = dev->features & ~(illegal_highdma(dev, skb) ?
					 NETIF_F_SG : 0);

	segs = skb_gso_segment(skb, features);

	/* Verifying header integrity only. */
	if (!segs)
		return 0;
1612

1613
	if (IS_ERR(segs))
1614 1615 1616 1617 1618 1619 1620 1621 1622
		return PTR_ERR(segs);

	skb->next = segs;
	DEV_GSO_CB(skb)->destructor = skb->destructor;
	skb->destructor = dev_gso_skb_destructor;

	return 0;
}

1623 1624
int dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
			struct netdev_queue *txq)
1625 1626
{
	if (likely(!skb->next)) {
S
Stephen Hemminger 已提交
1627
		if (!list_empty(&ptype_all))
1628 1629
			dev_queue_xmit_nit(skb, dev);

1630 1631 1632 1633 1634 1635
		if (netif_needs_gso(dev, skb)) {
			if (unlikely(dev_gso_segment(skb)))
				goto out_kfree_skb;
			if (skb->next)
				goto gso;
		}
1636

1637
		return dev->hard_start_xmit(skb, dev);
1638 1639
	}

1640
gso:
1641 1642 1643 1644 1645 1646 1647 1648
	do {
		struct sk_buff *nskb = skb->next;
		int rc;

		skb->next = nskb->next;
		nskb->next = NULL;
		rc = dev->hard_start_xmit(nskb, dev);
		if (unlikely(rc)) {
1649
			nskb->next = skb->next;
1650 1651 1652
			skb->next = nskb;
			return rc;
		}
1653
		if (unlikely(netif_tx_queue_stopped(txq) && skb->next))
1654
			return NETDEV_TX_BUSY;
1655
	} while (skb->next);
1656

1657 1658 1659 1660 1661 1662 1663
	skb->destructor = DEV_GSO_CB(skb)->destructor;

out_kfree_skb:
	kfree_skb(skb);
	return 0;
}

D
David S. Miller 已提交
1664 1665 1666
static u32 simple_tx_hashrnd;
static int simple_tx_hashrnd_initialized = 0;

1667 1668
static u16 simple_tx_hash(struct net_device *dev, struct sk_buff *skb)
{
D
David S. Miller 已提交
1669 1670
	u32 addr1, addr2, ports;
	u32 hash, ihl;
1671
	u8 ip_proto = 0;
D
David S. Miller 已提交
1672 1673 1674 1675 1676

	if (unlikely(!simple_tx_hashrnd_initialized)) {
		get_random_bytes(&simple_tx_hashrnd, 4);
		simple_tx_hashrnd_initialized = 1;
	}
1677 1678 1679

	switch (skb->protocol) {
	case __constant_htons(ETH_P_IP):
1680 1681
		if (!(ip_hdr(skb)->frag_off & htons(IP_MF | IP_OFFSET)))
			ip_proto = ip_hdr(skb)->protocol;
D
David S. Miller 已提交
1682 1683
		addr1 = ip_hdr(skb)->saddr;
		addr2 = ip_hdr(skb)->daddr;
1684 1685 1686 1687
		ihl = ip_hdr(skb)->ihl;
		break;
	case __constant_htons(ETH_P_IPV6):
		ip_proto = ipv6_hdr(skb)->nexthdr;
D
David S. Miller 已提交
1688 1689
		addr1 = ipv6_hdr(skb)->saddr.s6_addr32[3];
		addr2 = ipv6_hdr(skb)->daddr.s6_addr32[3];
1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704
		ihl = (40 >> 2);
		break;
	default:
		return 0;
	}


	switch (ip_proto) {
	case IPPROTO_TCP:
	case IPPROTO_UDP:
	case IPPROTO_DCCP:
	case IPPROTO_ESP:
	case IPPROTO_AH:
	case IPPROTO_SCTP:
	case IPPROTO_UDPLITE:
D
David S. Miller 已提交
1705
		ports = *((u32 *) (skb_network_header(skb) + (ihl * 4)));
1706 1707 1708
		break;

	default:
D
David S. Miller 已提交
1709
		ports = 0;
1710 1711 1712
		break;
	}

D
David S. Miller 已提交
1713 1714 1715
	hash = jhash_3words(addr1, addr2, ports, simple_tx_hashrnd);

	return (u16) (((u64) hash * dev->real_num_tx_queues) >> 32);
1716 1717
}

1718 1719 1720
static struct netdev_queue *dev_pick_tx(struct net_device *dev,
					struct sk_buff *skb)
{
1721 1722
	u16 queue_index = 0;

1723 1724
	if (dev->select_queue)
		queue_index = dev->select_queue(dev, skb);
1725 1726
	else if (dev->real_num_tx_queues > 1)
		queue_index = simple_tx_hash(dev, skb);
1727

1728 1729
	skb_set_queue_mapping(skb, queue_index);
	return netdev_get_tx_queue(dev, queue_index);
1730 1731
}

1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756
/**
 *	dev_queue_xmit - transmit a buffer
 *	@skb: buffer to transmit
 *
 *	Queue a buffer for transmission to a network device. The caller must
 *	have set the device and priority and built the buffer before calling
 *	this function. The function can be called from an interrupt.
 *
 *	A negative errno code is returned on a failure. A success does not
 *	guarantee the frame will be transmitted as it may be dropped due
 *	to congestion or traffic shaping.
 *
 * -----------------------------------------------------------------------------------
 *      I notice this method can also return errors from the queue disciplines,
 *      including NET_XMIT_DROP, which is a positive value.  So, errors can also
 *      be positive.
 *
 *      Regardless of the return value, the skb is consumed, so it is currently
 *      difficult to retry a send to this method.  (You can bump the ref count
 *      before sending to hold a reference for retry if you are careful.)
 *
 *      When calling this method, interrupts MUST be enabled.  This is because
 *      the BH enable code must have IRQs enabled so that it will not deadlock.
 *          --BLG
 */
L
Linus Torvalds 已提交
1757 1758 1759
int dev_queue_xmit(struct sk_buff *skb)
{
	struct net_device *dev = skb->dev;
1760
	struct netdev_queue *txq;
L
Linus Torvalds 已提交
1761 1762 1763
	struct Qdisc *q;
	int rc = -ENOMEM;

1764 1765 1766 1767
	/* GSO will handle the following emulations directly. */
	if (netif_needs_gso(dev, skb))
		goto gso;

L
Linus Torvalds 已提交
1768 1769
	if (skb_shinfo(skb)->frag_list &&
	    !(dev->features & NETIF_F_FRAGLIST) &&
H
Herbert Xu 已提交
1770
	    __skb_linearize(skb))
L
Linus Torvalds 已提交
1771 1772 1773 1774 1775 1776 1777 1778
		goto out_kfree_skb;

	/* Fragmented skb is linearized if device does not support SG,
	 * or if at least one of fragments is in highmem and device
	 * does not support DMA from it.
	 */
	if (skb_shinfo(skb)->nr_frags &&
	    (!(dev->features & NETIF_F_SG) || illegal_highdma(dev, skb)) &&
H
Herbert Xu 已提交
1779
	    __skb_linearize(skb))
L
Linus Torvalds 已提交
1780 1781 1782 1783 1784
		goto out_kfree_skb;

	/* If packet is not checksummed and device does not support
	 * checksumming for this protocol, complete checksumming here.
	 */
1785 1786 1787
	if (skb->ip_summed == CHECKSUM_PARTIAL) {
		skb_set_transport_header(skb, skb->csum_start -
					      skb_headroom(skb));
1788 1789
		if (!dev_can_checksum(dev, skb) && skb_checksum_help(skb))
			goto out_kfree_skb;
1790
	}
L
Linus Torvalds 已提交
1791

1792
gso:
1793 1794
	/* Disable soft irqs for various locks below. Also
	 * stops preemption for RCU.
L
Linus Torvalds 已提交
1795
	 */
1796
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
1797

1798
	txq = dev_pick_tx(dev, skb);
1799
	q = rcu_dereference(txq->qdisc);
1800

L
Linus Torvalds 已提交
1801 1802 1803 1804
#ifdef CONFIG_NET_CLS_ACT
	skb->tc_verd = SET_TC_AT(skb->tc_verd,AT_EGRESS);
#endif
	if (q->enqueue) {
1805
		spinlock_t *root_lock = qdisc_lock(q);
1806 1807 1808

		spin_lock(root_lock);

1809
		if (unlikely(test_bit(__QDISC_STATE_DEACTIVATED, &q->state))) {
1810
			kfree_skb(skb);
1811
			rc = NET_XMIT_DROP;
1812 1813 1814
		} else {
			rc = qdisc_enqueue_root(skb, q);
			qdisc_run(q);
1815
		}
1816 1817 1818
		spin_unlock(root_lock);

		goto out;
L
Linus Torvalds 已提交
1819 1820 1821 1822 1823
	}

	/* The device has no queue. Common case for software devices:
	   loopback, all the sorts of tunnels...

H
Herbert Xu 已提交
1824 1825
	   Really, it is unlikely that netif_tx_lock protection is necessary
	   here.  (f.e. loopback and IP tunnels are clean ignoring statistics
L
Linus Torvalds 已提交
1826 1827 1828 1829 1830 1831 1832 1833 1834 1835
	   counters.)
	   However, it is possible, that they rely on protection
	   made by us here.

	   Check this and shot the lock. It is not prone from deadlocks.
	   Either shot noqueue qdisc, it is even simpler 8)
	 */
	if (dev->flags & IFF_UP) {
		int cpu = smp_processor_id(); /* ok because BHs are off */

1836
		if (txq->xmit_lock_owner != cpu) {
L
Linus Torvalds 已提交
1837

1838
			HARD_TX_LOCK(dev, txq, cpu);
L
Linus Torvalds 已提交
1839

1840
			if (!netif_tx_queue_stopped(txq)) {
L
Linus Torvalds 已提交
1841
				rc = 0;
1842
				if (!dev_hard_start_xmit(skb, dev, txq)) {
1843
					HARD_TX_UNLOCK(dev, txq);
L
Linus Torvalds 已提交
1844 1845 1846
					goto out;
				}
			}
1847
			HARD_TX_UNLOCK(dev, txq);
L
Linus Torvalds 已提交
1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860
			if (net_ratelimit())
				printk(KERN_CRIT "Virtual device %s asks to "
				       "queue packet!\n", dev->name);
		} else {
			/* Recursion is detected! It is possible,
			 * unfortunately */
			if (net_ratelimit())
				printk(KERN_CRIT "Dead loop on virtual device "
				       "%s, fix it urgently!\n", dev->name);
		}
	}

	rc = -ENETDOWN;
1861
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1862 1863 1864 1865 1866

out_kfree_skb:
	kfree_skb(skb);
	return rc;
out:
1867
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1868 1869 1870 1871 1872 1873 1874 1875
	return rc;
}


/*=======================================================================
			Receiver routines
  =======================================================================*/

1876 1877 1878
int netdev_max_backlog __read_mostly = 1000;
int netdev_budget __read_mostly = 300;
int weight_p __read_mostly = 64;            /* old backlog weight */
L
Linus Torvalds 已提交
1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906

DEFINE_PER_CPU(struct netif_rx_stats, netdev_rx_stat) = { 0, };


/**
 *	netif_rx	-	post buffer to the network code
 *	@skb: buffer to post
 *
 *	This function receives a packet from a device driver and queues it for
 *	the upper (protocol) levels to process.  It always succeeds. The buffer
 *	may be dropped during processing for congestion control or by the
 *	protocol layers.
 *
 *	return values:
 *	NET_RX_SUCCESS	(no congestion)
 *	NET_RX_DROP     (packet was dropped)
 *
 */

int netif_rx(struct sk_buff *skb)
{
	struct softnet_data *queue;
	unsigned long flags;

	/* if netpoll wants it, pretend we never saw it */
	if (netpoll_rx(skb))
		return NET_RX_DROP;

1907
	if (!skb->tstamp.tv64)
1908
		net_timestamp(skb);
L
Linus Torvalds 已提交
1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922

	/*
	 * The code is rearranged so that the path is the most
	 * short when CPU is congested, but is still operating.
	 */
	local_irq_save(flags);
	queue = &__get_cpu_var(softnet_data);

	__get_cpu_var(netdev_rx_stat).total++;
	if (queue->input_pkt_queue.qlen <= netdev_max_backlog) {
		if (queue->input_pkt_queue.qlen) {
enqueue:
			__skb_queue_tail(&queue->input_pkt_queue, skb);
			local_irq_restore(flags);
1923
			return NET_RX_SUCCESS;
L
Linus Torvalds 已提交
1924 1925
		}

1926
		napi_schedule(&queue->backlog);
L
Linus Torvalds 已提交
1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967
		goto enqueue;
	}

	__get_cpu_var(netdev_rx_stat).dropped++;
	local_irq_restore(flags);

	kfree_skb(skb);
	return NET_RX_DROP;
}

int netif_rx_ni(struct sk_buff *skb)
{
	int err;

	preempt_disable();
	err = netif_rx(skb);
	if (local_softirq_pending())
		do_softirq();
	preempt_enable();

	return err;
}

EXPORT_SYMBOL(netif_rx_ni);

static void net_tx_action(struct softirq_action *h)
{
	struct softnet_data *sd = &__get_cpu_var(softnet_data);

	if (sd->completion_queue) {
		struct sk_buff *clist;

		local_irq_disable();
		clist = sd->completion_queue;
		sd->completion_queue = NULL;
		local_irq_enable();

		while (clist) {
			struct sk_buff *skb = clist;
			clist = clist->next;

1968
			WARN_ON(atomic_read(&skb->users));
L
Linus Torvalds 已提交
1969 1970 1971 1972 1973
			__kfree_skb(skb);
		}
	}

	if (sd->output_queue) {
1974
		struct Qdisc *head;
L
Linus Torvalds 已提交
1975 1976 1977 1978 1979 1980 1981

		local_irq_disable();
		head = sd->output_queue;
		sd->output_queue = NULL;
		local_irq_enable();

		while (head) {
1982 1983 1984
			struct Qdisc *q = head;
			spinlock_t *root_lock;

L
Linus Torvalds 已提交
1985 1986
			head = head->next_sched;

1987
			root_lock = qdisc_lock(q);
1988
			if (spin_trylock(root_lock)) {
1989 1990 1991
				smp_mb__before_clear_bit();
				clear_bit(__QDISC_STATE_SCHED,
					  &q->state);
1992 1993
				qdisc_run(q);
				spin_unlock(root_lock);
L
Linus Torvalds 已提交
1994
			} else {
1995
				if (!test_bit(__QDISC_STATE_DEACTIVATED,
1996
					      &q->state)) {
1997
					__netif_reschedule(q);
1998 1999 2000 2001 2002
				} else {
					smp_mb__before_clear_bit();
					clear_bit(__QDISC_STATE_SCHED,
						  &q->state);
				}
L
Linus Torvalds 已提交
2003 2004 2005 2006 2007
			}
		}
	}
}

2008 2009 2010
static inline int deliver_skb(struct sk_buff *skb,
			      struct packet_type *pt_prev,
			      struct net_device *orig_dev)
L
Linus Torvalds 已提交
2011 2012
{
	atomic_inc(&skb->users);
D
David S. Miller 已提交
2013
	return pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2014 2015 2016
}

#if defined(CONFIG_BRIDGE) || defined (CONFIG_BRIDGE_MODULE)
2017
/* These hooks defined here for ATM */
L
Linus Torvalds 已提交
2018 2019 2020
struct net_bridge;
struct net_bridge_fdb_entry *(*br_fdb_get_hook)(struct net_bridge *br,
						unsigned char *addr);
2021
void (*br_fdb_put_hook)(struct net_bridge_fdb_entry *ent) __read_mostly;
L
Linus Torvalds 已提交
2022

2023 2024 2025 2026 2027 2028 2029 2030 2031
/*
 * If bridge module is loaded call bridging hook.
 *  returns NULL if packet was consumed.
 */
struct sk_buff *(*br_handle_frame_hook)(struct net_bridge_port *p,
					struct sk_buff *skb) __read_mostly;
static inline struct sk_buff *handle_bridge(struct sk_buff *skb,
					    struct packet_type **pt_prev, int *ret,
					    struct net_device *orig_dev)
L
Linus Torvalds 已提交
2032 2033 2034
{
	struct net_bridge_port *port;

2035 2036 2037
	if (skb->pkt_type == PACKET_LOOPBACK ||
	    (port = rcu_dereference(skb->dev->br_port)) == NULL)
		return skb;
L
Linus Torvalds 已提交
2038 2039

	if (*pt_prev) {
2040
		*ret = deliver_skb(skb, *pt_prev, orig_dev);
L
Linus Torvalds 已提交
2041
		*pt_prev = NULL;
2042 2043
	}

2044
	return br_handle_frame_hook(port, skb);
L
Linus Torvalds 已提交
2045 2046
}
#else
2047
#define handle_bridge(skb, pt_prev, ret, orig_dev)	(skb)
L
Linus Torvalds 已提交
2048 2049
#endif

P
Patrick McHardy 已提交
2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071
#if defined(CONFIG_MACVLAN) || defined(CONFIG_MACVLAN_MODULE)
struct sk_buff *(*macvlan_handle_frame_hook)(struct sk_buff *skb) __read_mostly;
EXPORT_SYMBOL_GPL(macvlan_handle_frame_hook);

static inline struct sk_buff *handle_macvlan(struct sk_buff *skb,
					     struct packet_type **pt_prev,
					     int *ret,
					     struct net_device *orig_dev)
{
	if (skb->dev->macvlan_port == NULL)
		return skb;

	if (*pt_prev) {
		*ret = deliver_skb(skb, *pt_prev, orig_dev);
		*pt_prev = NULL;
	}
	return macvlan_handle_frame_hook(skb);
}
#else
#define handle_macvlan(skb, pt_prev, ret, orig_dev)	(skb)
#endif

L
Linus Torvalds 已提交
2072 2073 2074 2075 2076
#ifdef CONFIG_NET_CLS_ACT
/* TODO: Maybe we should just force sch_ingress to be compiled in
 * when CONFIG_NET_CLS_ACT is? otherwise some useless instructions
 * a compare and 2 stores extra right now if we dont have it on
 * but have CONFIG_NET_CLS_ACT
2077
 * NOTE: This doesnt stop any functionality; if you dont have
L
Linus Torvalds 已提交
2078 2079 2080
 * the ingress scheduler, you just cant add policies on ingress.
 *
 */
2081
static int ing_filter(struct sk_buff *skb)
L
Linus Torvalds 已提交
2082 2083
{
	struct net_device *dev = skb->dev;
2084
	u32 ttl = G_TC_RTTL(skb->tc_verd);
2085 2086 2087
	struct netdev_queue *rxq;
	int result = TC_ACT_OK;
	struct Qdisc *q;
2088

2089 2090 2091 2092 2093 2094
	if (MAX_RED_LOOP < ttl++) {
		printk(KERN_WARNING
		       "Redir loop detected Dropping packet (%d->%d)\n",
		       skb->iif, dev->ifindex);
		return TC_ACT_SHOT;
	}
L
Linus Torvalds 已提交
2095

2096 2097
	skb->tc_verd = SET_TC_RTTL(skb->tc_verd, ttl);
	skb->tc_verd = SET_TC_AT(skb->tc_verd, AT_INGRESS);
L
Linus Torvalds 已提交
2098

2099 2100
	rxq = &dev->rx_queue;

2101
	q = rxq->qdisc;
2102
	if (q != &noop_qdisc) {
2103
		spin_lock(qdisc_lock(q));
2104 2105
		if (likely(!test_bit(__QDISC_STATE_DEACTIVATED, &q->state)))
			result = qdisc_enqueue_root(skb, q);
2106 2107
		spin_unlock(qdisc_lock(q));
	}
2108 2109 2110

	return result;
}
2111

2112 2113 2114 2115
static inline struct sk_buff *handle_ing(struct sk_buff *skb,
					 struct packet_type **pt_prev,
					 int *ret, struct net_device *orig_dev)
{
2116
	if (skb->dev->rx_queue.qdisc == &noop_qdisc)
2117
		goto out;
L
Linus Torvalds 已提交
2118

2119 2120 2121 2122 2123 2124
	if (*pt_prev) {
		*ret = deliver_skb(skb, *pt_prev, orig_dev);
		*pt_prev = NULL;
	} else {
		/* Huh? Why does turning on AF_PACKET affect this? */
		skb->tc_verd = SET_TC_OK2MUNGE(skb->tc_verd);
L
Linus Torvalds 已提交
2125 2126
	}

2127 2128 2129 2130 2131 2132 2133 2134 2135 2136
	switch (ing_filter(skb)) {
	case TC_ACT_SHOT:
	case TC_ACT_STOLEN:
		kfree_skb(skb);
		return NULL;
	}

out:
	skb->tc_verd = 0;
	return skb;
L
Linus Torvalds 已提交
2137 2138 2139
}
#endif

2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166
/*
 * 	netif_nit_deliver - deliver received packets to network taps
 * 	@skb: buffer
 *
 * 	This function is used to deliver incoming packets to network
 * 	taps. It should be used when the normal netif_receive_skb path
 * 	is bypassed, for example because of VLAN acceleration.
 */
void netif_nit_deliver(struct sk_buff *skb)
{
	struct packet_type *ptype;

	if (list_empty(&ptype_all))
		return;

	skb_reset_network_header(skb);
	skb_reset_transport_header(skb);
	skb->mac_len = skb->network_header - skb->mac_header;

	rcu_read_lock();
	list_for_each_entry_rcu(ptype, &ptype_all, list) {
		if (!ptype->dev || ptype->dev == skb->dev)
			deliver_skb(skb, ptype, skb->dev);
	}
	rcu_read_unlock();
}

2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181
/**
 *	netif_receive_skb - process receive buffer from network
 *	@skb: buffer to process
 *
 *	netif_receive_skb() is the main receive data processing function.
 *	It always succeeds. The buffer may be dropped during processing
 *	for congestion control or by the protocol layers.
 *
 *	This function may only be called from softirq context and interrupts
 *	should be enabled.
 *
 *	Return values (usually ignored):
 *	NET_RX_SUCCESS: no congestion
 *	NET_RX_DROP: packet was dropped
 */
L
Linus Torvalds 已提交
2182 2183 2184
int netif_receive_skb(struct sk_buff *skb)
{
	struct packet_type *ptype, *pt_prev;
D
David S. Miller 已提交
2185
	struct net_device *orig_dev;
2186
	struct net_device *null_or_orig;
L
Linus Torvalds 已提交
2187
	int ret = NET_RX_DROP;
A
Al Viro 已提交
2188
	__be16 type;
L
Linus Torvalds 已提交
2189 2190

	/* if we've gotten here through NAPI, check netpoll */
2191
	if (netpoll_receive_skb(skb))
L
Linus Torvalds 已提交
2192 2193
		return NET_RX_DROP;

2194
	if (!skb->tstamp.tv64)
2195
		net_timestamp(skb);
L
Linus Torvalds 已提交
2196

2197 2198
	if (!skb->iif)
		skb->iif = skb->dev->ifindex;
2199

2200
	null_or_orig = NULL;
J
Joe Eykholt 已提交
2201 2202
	orig_dev = skb->dev;
	if (orig_dev->master) {
2203 2204 2205 2206
		if (skb_bond_should_drop(skb))
			null_or_orig = orig_dev; /* deliver only exact match */
		else
			skb->dev = orig_dev->master;
J
Joe Eykholt 已提交
2207
	}
2208

L
Linus Torvalds 已提交
2209 2210
	__get_cpu_var(netdev_rx_stat).total++;

2211
	skb_reset_network_header(skb);
2212
	skb_reset_transport_header(skb);
2213
	skb->mac_len = skb->network_header - skb->mac_header;
L
Linus Torvalds 已提交
2214 2215 2216 2217 2218

	pt_prev = NULL;

	rcu_read_lock();

2219 2220 2221 2222
	/* Don't receive packets in an exiting network namespace */
	if (!net_alive(dev_net(skb->dev)))
		goto out;

L
Linus Torvalds 已提交
2223 2224 2225 2226 2227 2228 2229 2230
#ifdef CONFIG_NET_CLS_ACT
	if (skb->tc_verd & TC_NCLS) {
		skb->tc_verd = CLR_TC_NCLS(skb->tc_verd);
		goto ncls;
	}
#endif

	list_for_each_entry_rcu(ptype, &ptype_all, list) {
2231 2232
		if (ptype->dev == null_or_orig || ptype->dev == skb->dev ||
		    ptype->dev == orig_dev) {
2233
			if (pt_prev)
D
David S. Miller 已提交
2234
				ret = deliver_skb(skb, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2235 2236 2237 2238 2239
			pt_prev = ptype;
		}
	}

#ifdef CONFIG_NET_CLS_ACT
2240 2241
	skb = handle_ing(skb, &pt_prev, &ret, orig_dev);
	if (!skb)
L
Linus Torvalds 已提交
2242 2243 2244 2245
		goto out;
ncls:
#endif

2246
	skb = handle_bridge(skb, &pt_prev, &ret, orig_dev);
P
Patrick McHardy 已提交
2247 2248 2249
	if (!skb)
		goto out;
	skb = handle_macvlan(skb, &pt_prev, &ret, orig_dev);
2250
	if (!skb)
L
Linus Torvalds 已提交
2251 2252 2253
		goto out;

	type = skb->protocol;
2254 2255
	list_for_each_entry_rcu(ptype,
			&ptype_base[ntohs(type) & PTYPE_HASH_MASK], list) {
L
Linus Torvalds 已提交
2256
		if (ptype->type == type &&
2257 2258
		    (ptype->dev == null_or_orig || ptype->dev == skb->dev ||
		     ptype->dev == orig_dev)) {
2259
			if (pt_prev)
D
David S. Miller 已提交
2260
				ret = deliver_skb(skb, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2261 2262 2263 2264 2265
			pt_prev = ptype;
		}
	}

	if (pt_prev) {
D
David S. Miller 已提交
2266
		ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279
	} else {
		kfree_skb(skb);
		/* Jamal, now you will not able to escape explaining
		 * me how you were going to use this. :-)
		 */
		ret = NET_RX_DROP;
	}

out:
	rcu_read_unlock();
	return ret;
}

2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293
/* Network device is going away, flush any packets still pending  */
static void flush_backlog(void *arg)
{
	struct net_device *dev = arg;
	struct softnet_data *queue = &__get_cpu_var(softnet_data);
	struct sk_buff *skb, *tmp;

	skb_queue_walk_safe(&queue->input_pkt_queue, skb, tmp)
		if (skb->dev == dev) {
			__skb_unlink(skb, &queue->input_pkt_queue);
			kfree_skb(skb);
		}
}

2294
static int process_backlog(struct napi_struct *napi, int quota)
L
Linus Torvalds 已提交
2295 2296 2297 2298 2299
{
	int work = 0;
	struct softnet_data *queue = &__get_cpu_var(softnet_data);
	unsigned long start_time = jiffies;

2300 2301
	napi->weight = weight_p;
	do {
L
Linus Torvalds 已提交
2302 2303 2304 2305
		struct sk_buff *skb;

		local_irq_disable();
		skb = __skb_dequeue(&queue->input_pkt_queue);
2306 2307 2308 2309 2310
		if (!skb) {
			__napi_complete(napi);
			local_irq_enable();
			break;
		}
L
Linus Torvalds 已提交
2311 2312 2313
		local_irq_enable();

		netif_receive_skb(skb);
2314
	} while (++work < quota && jiffies == start_time);
L
Linus Torvalds 已提交
2315

2316 2317
	return work;
}
L
Linus Torvalds 已提交
2318

2319 2320
/**
 * __napi_schedule - schedule for receive
2321
 * @n: entry to schedule
2322 2323 2324
 *
 * The entry's receive function will be scheduled to run
 */
H
Harvey Harrison 已提交
2325
void __napi_schedule(struct napi_struct *n)
2326 2327
{
	unsigned long flags;
L
Linus Torvalds 已提交
2328

2329 2330 2331 2332
	local_irq_save(flags);
	list_add_tail(&n->poll_list, &__get_cpu_var(softnet_data).poll_list);
	__raise_softirq_irqoff(NET_RX_SOFTIRQ);
	local_irq_restore(flags);
L
Linus Torvalds 已提交
2333
}
2334 2335
EXPORT_SYMBOL(__napi_schedule);

L
Linus Torvalds 已提交
2336 2337 2338

static void net_rx_action(struct softirq_action *h)
{
2339
	struct list_head *list = &__get_cpu_var(softnet_data).poll_list;
L
Linus Torvalds 已提交
2340
	unsigned long start_time = jiffies;
2341
	int budget = netdev_budget;
2342 2343
	void *have;

L
Linus Torvalds 已提交
2344 2345
	local_irq_disable();

2346 2347 2348
	while (!list_empty(list)) {
		struct napi_struct *n;
		int work, weight;
L
Linus Torvalds 已提交
2349

2350 2351 2352 2353 2354 2355 2356 2357
		/* If softirq window is exhuasted then punt.
		 *
		 * Note that this is a slight policy change from the
		 * previous NAPI code, which would allow up to 2
		 * jiffies to pass before breaking out.  The test
		 * used to be "jiffies - start_time > 1".
		 */
		if (unlikely(budget <= 0 || jiffies != start_time))
L
Linus Torvalds 已提交
2358 2359 2360 2361
			goto softnet_break;

		local_irq_enable();

2362 2363 2364 2365 2366 2367
		/* Even though interrupts have been re-enabled, this
		 * access is safe because interrupts can only add new
		 * entries to the tail of this list, and only ->poll()
		 * calls can remove this head entry from the list.
		 */
		n = list_entry(list->next, struct napi_struct, poll_list);
L
Linus Torvalds 已提交
2368

2369 2370 2371 2372
		have = netpoll_poll_lock(n);

		weight = n->weight;

2373 2374 2375 2376 2377 2378 2379 2380 2381
		/* This NAPI_STATE_SCHED test is for avoiding a race
		 * with netpoll's poll_napi().  Only the entity which
		 * obtains the lock and sees NAPI_STATE_SCHED set will
		 * actually make the ->poll() call.  Therefore we avoid
		 * accidently calling ->poll() when NAPI is not scheduled.
		 */
		work = 0;
		if (test_bit(NAPI_STATE_SCHED, &n->state))
			work = n->poll(n, weight);
2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393

		WARN_ON_ONCE(work > weight);

		budget -= work;

		local_irq_disable();

		/* Drivers must not modify the NAPI state if they
		 * consume the entire weight.  In such cases this code
		 * still "owns" the NAPI instance and therefore can
		 * move the instance around on the list at-will.
		 */
2394 2395 2396 2397 2398 2399
		if (unlikely(work == weight)) {
			if (unlikely(napi_disable_pending(n)))
				__napi_complete(n);
			else
				list_move_tail(&n->poll_list, list);
		}
2400 2401

		netpoll_poll_unlock(have);
L
Linus Torvalds 已提交
2402 2403
	}
out:
2404
	local_irq_enable();
2405

2406 2407 2408 2409 2410
#ifdef CONFIG_NET_DMA
	/*
	 * There may not be any more sk_buffs coming right now, so push
	 * any pending DMA copies to hardware
	 */
2411 2412
	if (!cpus_empty(net_dma.channel_mask)) {
		int chan_idx;
2413
		for_each_cpu_mask_nr(chan_idx, net_dma.channel_mask) {
2414 2415 2416 2417
			struct dma_chan *chan = net_dma.channels[chan_idx];
			if (chan)
				dma_async_memcpy_issue_pending(chan);
		}
2418 2419
	}
#endif
2420

L
Linus Torvalds 已提交
2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459
	return;

softnet_break:
	__get_cpu_var(netdev_rx_stat).time_squeeze++;
	__raise_softirq_irqoff(NET_RX_SOFTIRQ);
	goto out;
}

static gifconf_func_t * gifconf_list [NPROTO];

/**
 *	register_gifconf	-	register a SIOCGIF handler
 *	@family: Address family
 *	@gifconf: Function handler
 *
 *	Register protocol dependent address dumping routines. The handler
 *	that is passed must not be freed or reused until it has been replaced
 *	by another handler.
 */
int register_gifconf(unsigned int family, gifconf_func_t * gifconf)
{
	if (family >= NPROTO)
		return -EINVAL;
	gifconf_list[family] = gifconf;
	return 0;
}


/*
 *	Map an interface index to its name (SIOCGIFNAME)
 */

/*
 *	We need this ioctl for efficient implementation of the
 *	if_indextoname() function required by the IPv6 API.  Without
 *	it, we would have to search all the interfaces to find a
 *	match.  --pb
 */

2460
static int dev_ifname(struct net *net, struct ifreq __user *arg)
L
Linus Torvalds 已提交
2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472
{
	struct net_device *dev;
	struct ifreq ifr;

	/*
	 *	Fetch the caller's info block.
	 */

	if (copy_from_user(&ifr, arg, sizeof(struct ifreq)))
		return -EFAULT;

	read_lock(&dev_base_lock);
2473
	dev = __dev_get_by_index(net, ifr.ifr_ifindex);
L
Linus Torvalds 已提交
2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492
	if (!dev) {
		read_unlock(&dev_base_lock);
		return -ENODEV;
	}

	strcpy(ifr.ifr_name, dev->name);
	read_unlock(&dev_base_lock);

	if (copy_to_user(arg, &ifr, sizeof(struct ifreq)))
		return -EFAULT;
	return 0;
}

/*
 *	Perform a SIOCGIFCONF call. This structure will change
 *	size eventually, and there is nothing I can do about it.
 *	Thus we will need a 'compatibility mode'.
 */

2493
static int dev_ifconf(struct net *net, char __user *arg)
L
Linus Torvalds 已提交
2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516
{
	struct ifconf ifc;
	struct net_device *dev;
	char __user *pos;
	int len;
	int total;
	int i;

	/*
	 *	Fetch the caller's info block.
	 */

	if (copy_from_user(&ifc, arg, sizeof(struct ifconf)))
		return -EFAULT;

	pos = ifc.ifc_buf;
	len = ifc.ifc_len;

	/*
	 *	Loop over the interfaces, and write an info block for each.
	 */

	total = 0;
2517
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530
		for (i = 0; i < NPROTO; i++) {
			if (gifconf_list[i]) {
				int done;
				if (!pos)
					done = gifconf_list[i](dev, NULL, 0);
				else
					done = gifconf_list[i](dev, pos + total,
							       len - total);
				if (done < 0)
					return -EFAULT;
				total += done;
			}
		}
2531
	}
L
Linus Torvalds 已提交
2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548

	/*
	 *	All done.  Write the updated control block back to the caller.
	 */
	ifc.ifc_len = total;

	/*
	 * 	Both BSD and Solaris return 0 here, so we do too.
	 */
	return copy_to_user(arg, &ifc, sizeof(struct ifconf)) ? -EFAULT : 0;
}

#ifdef CONFIG_PROC_FS
/*
 *	This is invoked by the /proc filesystem handler to display a device
 *	in detail.
 */
2549
void *dev_seq_start(struct seq_file *seq, loff_t *pos)
2550
	__acquires(dev_base_lock)
L
Linus Torvalds 已提交
2551
{
2552
	struct net *net = seq_file_net(seq);
2553
	loff_t off;
L
Linus Torvalds 已提交
2554 2555
	struct net_device *dev;

2556 2557 2558
	read_lock(&dev_base_lock);
	if (!*pos)
		return SEQ_START_TOKEN;
L
Linus Torvalds 已提交
2559

2560
	off = 1;
2561
	for_each_netdev(net, dev)
2562 2563
		if (off++ == *pos)
			return dev;
L
Linus Torvalds 已提交
2564

2565
	return NULL;
L
Linus Torvalds 已提交
2566 2567 2568 2569
}

void *dev_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
2570
	struct net *net = seq_file_net(seq);
L
Linus Torvalds 已提交
2571
	++*pos;
2572
	return v == SEQ_START_TOKEN ?
2573
		first_net_device(net) : next_net_device((struct net_device *)v);
L
Linus Torvalds 已提交
2574 2575 2576
}

void dev_seq_stop(struct seq_file *seq, void *v)
2577
	__releases(dev_base_lock)
L
Linus Torvalds 已提交
2578 2579 2580 2581 2582 2583
{
	read_unlock(&dev_base_lock);
}

static void dev_seq_printf_stats(struct seq_file *seq, struct net_device *dev)
{
R
Rusty Russell 已提交
2584
	struct net_device_stats *stats = dev->get_stats(dev);
L
Linus Torvalds 已提交
2585

2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602
	seq_printf(seq, "%6s:%8lu %7lu %4lu %4lu %4lu %5lu %10lu %9lu "
		   "%8lu %7lu %4lu %4lu %4lu %5lu %7lu %10lu\n",
		   dev->name, stats->rx_bytes, stats->rx_packets,
		   stats->rx_errors,
		   stats->rx_dropped + stats->rx_missed_errors,
		   stats->rx_fifo_errors,
		   stats->rx_length_errors + stats->rx_over_errors +
		    stats->rx_crc_errors + stats->rx_frame_errors,
		   stats->rx_compressed, stats->multicast,
		   stats->tx_bytes, stats->tx_packets,
		   stats->tx_errors, stats->tx_dropped,
		   stats->tx_fifo_errors, stats->collisions,
		   stats->tx_carrier_errors +
		    stats->tx_aborted_errors +
		    stats->tx_window_errors +
		    stats->tx_heartbeat_errors,
		   stats->tx_compressed);
L
Linus Torvalds 已提交
2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625
}

/*
 *	Called from the PROCfs module. This now uses the new arbitrary sized
 *	/proc/net interface to create /proc/net/dev
 */
static int dev_seq_show(struct seq_file *seq, void *v)
{
	if (v == SEQ_START_TOKEN)
		seq_puts(seq, "Inter-|   Receive                            "
			      "                    |  Transmit\n"
			      " face |bytes    packets errs drop fifo frame "
			      "compressed multicast|bytes    packets errs "
			      "drop fifo colls carrier compressed\n");
	else
		dev_seq_printf_stats(seq, v);
	return 0;
}

static struct netif_rx_stats *softnet_get_online(loff_t *pos)
{
	struct netif_rx_stats *rc = NULL;

2626
	while (*pos < nr_cpu_ids)
2627
		if (cpu_online(*pos)) {
L
Linus Torvalds 已提交
2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654
			rc = &per_cpu(netdev_rx_stat, *pos);
			break;
		} else
			++*pos;
	return rc;
}

static void *softnet_seq_start(struct seq_file *seq, loff_t *pos)
{
	return softnet_get_online(pos);
}

static void *softnet_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
	++*pos;
	return softnet_get_online(pos);
}

static void softnet_seq_stop(struct seq_file *seq, void *v)
{
}

static int softnet_seq_show(struct seq_file *seq, void *v)
{
	struct netif_rx_stats *s = v;

	seq_printf(seq, "%08x %08x %08x %08x %08x %08x %08x %08x %08x\n",
2655
		   s->total, s->dropped, s->time_squeeze, 0,
2656 2657
		   0, 0, 0, 0, /* was fastroute */
		   s->cpu_collision );
L
Linus Torvalds 已提交
2658 2659 2660
	return 0;
}

2661
static const struct seq_operations dev_seq_ops = {
L
Linus Torvalds 已提交
2662 2663 2664 2665 2666 2667 2668 2669
	.start = dev_seq_start,
	.next  = dev_seq_next,
	.stop  = dev_seq_stop,
	.show  = dev_seq_show,
};

static int dev_seq_open(struct inode *inode, struct file *file)
{
2670 2671
	return seq_open_net(inode, file, &dev_seq_ops,
			    sizeof(struct seq_net_private));
L
Linus Torvalds 已提交
2672 2673
}

2674
static const struct file_operations dev_seq_fops = {
L
Linus Torvalds 已提交
2675 2676 2677 2678
	.owner	 = THIS_MODULE,
	.open    = dev_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
2679
	.release = seq_release_net,
L
Linus Torvalds 已提交
2680 2681
};

2682
static const struct seq_operations softnet_seq_ops = {
L
Linus Torvalds 已提交
2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693
	.start = softnet_seq_start,
	.next  = softnet_seq_next,
	.stop  = softnet_seq_stop,
	.show  = softnet_seq_show,
};

static int softnet_seq_open(struct inode *inode, struct file *file)
{
	return seq_open(file, &softnet_seq_ops);
}

2694
static const struct file_operations softnet_seq_fops = {
L
Linus Torvalds 已提交
2695 2696 2697 2698 2699 2700 2701
	.owner	 = THIS_MODULE,
	.open    = softnet_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release,
};

2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713
static void *ptype_get_idx(loff_t pos)
{
	struct packet_type *pt = NULL;
	loff_t i = 0;
	int t;

	list_for_each_entry_rcu(pt, &ptype_all, list) {
		if (i == pos)
			return pt;
		++i;
	}

2714
	for (t = 0; t < PTYPE_HASH_SIZE; t++) {
2715 2716 2717 2718 2719 2720 2721 2722 2723 2724
		list_for_each_entry_rcu(pt, &ptype_base[t], list) {
			if (i == pos)
				return pt;
			++i;
		}
	}
	return NULL;
}

static void *ptype_seq_start(struct seq_file *seq, loff_t *pos)
2725
	__acquires(RCU)
2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748
{
	rcu_read_lock();
	return *pos ? ptype_get_idx(*pos - 1) : SEQ_START_TOKEN;
}

static void *ptype_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
	struct packet_type *pt;
	struct list_head *nxt;
	int hash;

	++*pos;
	if (v == SEQ_START_TOKEN)
		return ptype_get_idx(0);

	pt = v;
	nxt = pt->list.next;
	if (pt->type == htons(ETH_P_ALL)) {
		if (nxt != &ptype_all)
			goto found;
		hash = 0;
		nxt = ptype_base[0].next;
	} else
2749
		hash = ntohs(pt->type) & PTYPE_HASH_MASK;
2750 2751

	while (nxt == &ptype_base[hash]) {
2752
		if (++hash >= PTYPE_HASH_SIZE)
2753 2754 2755 2756 2757 2758 2759 2760
			return NULL;
		nxt = ptype_base[hash].next;
	}
found:
	return list_entry(nxt, struct packet_type, list);
}

static void ptype_seq_stop(struct seq_file *seq, void *v)
2761
	__releases(RCU)
2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796
{
	rcu_read_unlock();
}

static void ptype_seq_decode(struct seq_file *seq, void *sym)
{
#ifdef CONFIG_KALLSYMS
	unsigned long offset = 0, symsize;
	const char *symname;
	char *modname;
	char namebuf[128];

	symname = kallsyms_lookup((unsigned long)sym, &symsize, &offset,
				  &modname, namebuf);

	if (symname) {
		char *delim = ":";

		if (!modname)
			modname = delim = "";
		seq_printf(seq, "%s%s%s%s+0x%lx", delim, modname, delim,
			   symname, offset);
		return;
	}
#endif

	seq_printf(seq, "[%p]", sym);
}

static int ptype_seq_show(struct seq_file *seq, void *v)
{
	struct packet_type *pt = v;

	if (v == SEQ_START_TOKEN)
		seq_puts(seq, "Type Device      Function\n");
2797
	else if (pt->dev == NULL || dev_net(pt->dev) == seq_file_net(seq)) {
2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820
		if (pt->type == htons(ETH_P_ALL))
			seq_puts(seq, "ALL ");
		else
			seq_printf(seq, "%04x", ntohs(pt->type));

		seq_printf(seq, " %-8s ",
			   pt->dev ? pt->dev->name : "");
		ptype_seq_decode(seq,  pt->func);
		seq_putc(seq, '\n');
	}

	return 0;
}

static const struct seq_operations ptype_seq_ops = {
	.start = ptype_seq_start,
	.next  = ptype_seq_next,
	.stop  = ptype_seq_stop,
	.show  = ptype_seq_show,
};

static int ptype_seq_open(struct inode *inode, struct file *file)
{
2821 2822
	return seq_open_net(inode, file, &ptype_seq_ops,
			sizeof(struct seq_net_private));
2823 2824 2825 2826 2827 2828 2829
}

static const struct file_operations ptype_seq_fops = {
	.owner	 = THIS_MODULE,
	.open    = ptype_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
2830
	.release = seq_release_net,
2831 2832 2833
};


2834
static int __net_init dev_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
2835 2836 2837
{
	int rc = -ENOMEM;

2838
	if (!proc_net_fops_create(net, "dev", S_IRUGO, &dev_seq_fops))
L
Linus Torvalds 已提交
2839
		goto out;
2840
	if (!proc_net_fops_create(net, "softnet_stat", S_IRUGO, &softnet_seq_fops))
L
Linus Torvalds 已提交
2841
		goto out_dev;
2842
	if (!proc_net_fops_create(net, "ptype", S_IRUGO, &ptype_seq_fops))
2843
		goto out_softnet;
2844

2845
	if (wext_proc_init(net))
2846
		goto out_ptype;
L
Linus Torvalds 已提交
2847 2848 2849
	rc = 0;
out:
	return rc;
2850
out_ptype:
2851
	proc_net_remove(net, "ptype");
L
Linus Torvalds 已提交
2852
out_softnet:
2853
	proc_net_remove(net, "softnet_stat");
L
Linus Torvalds 已提交
2854
out_dev:
2855
	proc_net_remove(net, "dev");
L
Linus Torvalds 已提交
2856 2857
	goto out;
}
2858

2859
static void __net_exit dev_proc_net_exit(struct net *net)
2860 2861 2862 2863 2864 2865 2866 2867
{
	wext_proc_exit(net);

	proc_net_remove(net, "ptype");
	proc_net_remove(net, "softnet_stat");
	proc_net_remove(net, "dev");
}

2868
static struct pernet_operations __net_initdata dev_proc_ops = {
2869 2870 2871 2872 2873 2874 2875 2876
	.init = dev_proc_net_init,
	.exit = dev_proc_net_exit,
};

static int __init dev_proc_init(void)
{
	return register_pernet_subsys(&dev_proc_ops);
}
L
Linus Torvalds 已提交
2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905
#else
#define dev_proc_init() 0
#endif	/* CONFIG_PROC_FS */


/**
 *	netdev_set_master	-	set up master/slave pair
 *	@slave: slave device
 *	@master: new master device
 *
 *	Changes the master device of the slave. Pass %NULL to break the
 *	bonding. The caller must hold the RTNL semaphore. On a failure
 *	a negative errno code is returned. On success the reference counts
 *	are adjusted, %RTM_NEWLINK is sent to the routing socket and the
 *	function returns zero.
 */
int netdev_set_master(struct net_device *slave, struct net_device *master)
{
	struct net_device *old = slave->master;

	ASSERT_RTNL();

	if (master) {
		if (old)
			return -EBUSY;
		dev_hold(master);
	}

	slave->master = master;
2906

L
Linus Torvalds 已提交
2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920
	synchronize_net();

	if (old)
		dev_put(old);

	if (master)
		slave->flags |= IFF_SLAVE;
	else
		slave->flags &= ~IFF_SLAVE;

	rtmsg_ifinfo(RTM_NEWLINK, slave, IFF_SLAVE);
	return 0;
}

2921 2922 2923 2924 2925 2926
static void dev_change_rx_flags(struct net_device *dev, int flags)
{
	if (dev->flags & IFF_UP && dev->change_rx_flags)
		dev->change_rx_flags(dev, flags);
}

2927
static int __dev_set_promiscuity(struct net_device *dev, int inc)
L
Linus Torvalds 已提交
2928 2929 2930
{
	unsigned short old_flags = dev->flags;

2931 2932
	ASSERT_RTNL();

2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949
	dev->flags |= IFF_PROMISC;
	dev->promiscuity += inc;
	if (dev->promiscuity == 0) {
		/*
		 * Avoid overflow.
		 * If inc causes overflow, untouch promisc and return error.
		 */
		if (inc < 0)
			dev->flags &= ~IFF_PROMISC;
		else {
			dev->promiscuity -= inc;
			printk(KERN_WARNING "%s: promiscuity touches roof, "
				"set promiscuity failed, promiscuity feature "
				"of device might be broken.\n", dev->name);
			return -EOVERFLOW;
		}
	}
2950
	if (dev->flags != old_flags) {
L
Linus Torvalds 已提交
2951 2952
		printk(KERN_INFO "device %s %s promiscuous mode\n",
		       dev->name, (dev->flags & IFF_PROMISC) ? "entered" :
2953
							       "left");
2954 2955 2956 2957 2958 2959 2960 2961 2962
		if (audit_enabled)
			audit_log(current->audit_context, GFP_ATOMIC,
				AUDIT_ANOM_PROMISCUOUS,
				"dev=%s prom=%d old_prom=%d auid=%u uid=%u gid=%u ses=%u",
				dev->name, (dev->flags & IFF_PROMISC),
				(old_flags & IFF_PROMISC),
				audit_get_loginuid(current),
				current->uid, current->gid,
				audit_get_sessionid(current));
2963

2964
		dev_change_rx_flags(dev, IFF_PROMISC);
L
Linus Torvalds 已提交
2965
	}
2966
	return 0;
L
Linus Torvalds 已提交
2967 2968
}

2969 2970 2971 2972 2973 2974 2975 2976 2977
/**
 *	dev_set_promiscuity	- update promiscuity count on a device
 *	@dev: device
 *	@inc: modifier
 *
 *	Add or remove promiscuity from a device. While the count in the device
 *	remains above zero the interface remains promiscuous. Once it hits zero
 *	the device reverts back to normal filtering operation. A negative inc
 *	value is used to drop promiscuity on the device.
2978
 *	Return 0 if successful or a negative errno code on error.
2979
 */
2980
int dev_set_promiscuity(struct net_device *dev, int inc)
2981 2982
{
	unsigned short old_flags = dev->flags;
2983
	int err;
2984

2985
	err = __dev_set_promiscuity(dev, inc);
2986
	if (err < 0)
2987
		return err;
2988 2989
	if (dev->flags != old_flags)
		dev_set_rx_mode(dev);
2990
	return err;
2991 2992
}

L
Linus Torvalds 已提交
2993 2994 2995 2996 2997 2998 2999 3000 3001 3002
/**
 *	dev_set_allmulti	- update allmulti count on a device
 *	@dev: device
 *	@inc: modifier
 *
 *	Add or remove reception of all multicast frames to a device. While the
 *	count in the device remains above zero the interface remains listening
 *	to all interfaces. Once it hits zero the device reverts back to normal
 *	filtering operation. A negative @inc value is used to drop the counter
 *	when releasing a resource needing all multicasts.
3003
 *	Return 0 if successful or a negative errno code on error.
L
Linus Torvalds 已提交
3004 3005
 */

3006
int dev_set_allmulti(struct net_device *dev, int inc)
L
Linus Torvalds 已提交
3007 3008 3009
{
	unsigned short old_flags = dev->flags;

3010 3011
	ASSERT_RTNL();

L
Linus Torvalds 已提交
3012
	dev->flags |= IFF_ALLMULTI;
3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028
	dev->allmulti += inc;
	if (dev->allmulti == 0) {
		/*
		 * Avoid overflow.
		 * If inc causes overflow, untouch allmulti and return error.
		 */
		if (inc < 0)
			dev->flags &= ~IFF_ALLMULTI;
		else {
			dev->allmulti -= inc;
			printk(KERN_WARNING "%s: allmulti touches roof, "
				"set allmulti failed, allmulti feature of "
				"device might be broken.\n", dev->name);
			return -EOVERFLOW;
		}
	}
3029
	if (dev->flags ^ old_flags) {
3030
		dev_change_rx_flags(dev, IFF_ALLMULTI);
3031
		dev_set_rx_mode(dev);
3032
	}
3033
	return 0;
3034 3035 3036 3037 3038
}

/*
 *	Upload unicast and multicast address lists to device and
 *	configure RX filtering. When the device doesn't support unicast
J
Joe Perches 已提交
3039
 *	filtering it is put in promiscuous mode while unicast addresses
3040 3041 3042 3043 3044 3045 3046 3047 3048
 *	are present.
 */
void __dev_set_rx_mode(struct net_device *dev)
{
	/* dev_open will call this function so the list will stay sane. */
	if (!(dev->flags&IFF_UP))
		return;

	if (!netif_device_present(dev))
3049
		return;
3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071

	if (dev->set_rx_mode)
		dev->set_rx_mode(dev);
	else {
		/* Unicast addresses changes may only happen under the rtnl,
		 * therefore calling __dev_set_promiscuity here is safe.
		 */
		if (dev->uc_count > 0 && !dev->uc_promisc) {
			__dev_set_promiscuity(dev, 1);
			dev->uc_promisc = 1;
		} else if (dev->uc_count == 0 && dev->uc_promisc) {
			__dev_set_promiscuity(dev, -1);
			dev->uc_promisc = 0;
		}

		if (dev->set_multicast_list)
			dev->set_multicast_list(dev);
	}
}

void dev_set_rx_mode(struct net_device *dev)
{
3072
	netif_addr_lock_bh(dev);
3073
	__dev_set_rx_mode(dev);
3074
	netif_addr_unlock_bh(dev);
L
Linus Torvalds 已提交
3075 3076
}

3077 3078
int __dev_addr_delete(struct dev_addr_list **list, int *count,
		      void *addr, int alen, int glbl)
3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095
{
	struct dev_addr_list *da;

	for (; (da = *list) != NULL; list = &da->next) {
		if (memcmp(da->da_addr, addr, da->da_addrlen) == 0 &&
		    alen == da->da_addrlen) {
			if (glbl) {
				int old_glbl = da->da_gusers;
				da->da_gusers = 0;
				if (old_glbl == 0)
					break;
			}
			if (--da->da_users)
				return 0;

			*list = da->next;
			kfree(da);
3096
			(*count)--;
3097 3098 3099 3100 3101 3102
			return 0;
		}
	}
	return -ENOENT;
}

3103 3104
int __dev_addr_add(struct dev_addr_list **list, int *count,
		   void *addr, int alen, int glbl)
3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121
{
	struct dev_addr_list *da;

	for (da = *list; da != NULL; da = da->next) {
		if (memcmp(da->da_addr, addr, da->da_addrlen) == 0 &&
		    da->da_addrlen == alen) {
			if (glbl) {
				int old_glbl = da->da_gusers;
				da->da_gusers = 1;
				if (old_glbl)
					return 0;
			}
			da->da_users++;
			return 0;
		}
	}

3122
	da = kzalloc(sizeof(*da), GFP_ATOMIC);
3123 3124 3125 3126 3127 3128 3129 3130
	if (da == NULL)
		return -ENOMEM;
	memcpy(da->da_addr, addr, alen);
	da->da_addrlen = alen;
	da->da_users = 1;
	da->da_gusers = glbl ? 1 : 0;
	da->next = *list;
	*list = da;
3131
	(*count)++;
3132 3133 3134
	return 0;
}

3135 3136 3137
/**
 *	dev_unicast_delete	- Release secondary unicast address.
 *	@dev: device
R
Randy Dunlap 已提交
3138 3139
 *	@addr: address to delete
 *	@alen: length of @addr
3140 3141
 *
 *	Release reference to a secondary unicast address and remove it
R
Randy Dunlap 已提交
3142
 *	from the device if the reference count drops to zero.
3143 3144 3145 3146 3147 3148 3149 3150 3151
 *
 * 	The caller must hold the rtnl_mutex.
 */
int dev_unicast_delete(struct net_device *dev, void *addr, int alen)
{
	int err;

	ASSERT_RTNL();

3152
	netif_addr_lock_bh(dev);
3153 3154
	err = __dev_addr_delete(&dev->uc_list, &dev->uc_count, addr, alen, 0);
	if (!err)
3155
		__dev_set_rx_mode(dev);
3156
	netif_addr_unlock_bh(dev);
3157 3158 3159 3160 3161 3162 3163
	return err;
}
EXPORT_SYMBOL(dev_unicast_delete);

/**
 *	dev_unicast_add		- add a secondary unicast address
 *	@dev: device
3164
 *	@addr: address to add
R
Randy Dunlap 已提交
3165
 *	@alen: length of @addr
3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177
 *
 *	Add a secondary unicast address to the device or increase
 *	the reference count if it already exists.
 *
 *	The caller must hold the rtnl_mutex.
 */
int dev_unicast_add(struct net_device *dev, void *addr, int alen)
{
	int err;

	ASSERT_RTNL();

3178
	netif_addr_lock_bh(dev);
3179 3180
	err = __dev_addr_add(&dev->uc_list, &dev->uc_count, addr, alen, 0);
	if (!err)
3181
		__dev_set_rx_mode(dev);
3182
	netif_addr_unlock_bh(dev);
3183 3184 3185 3186
	return err;
}
EXPORT_SYMBOL(dev_unicast_add);

3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248
int __dev_addr_sync(struct dev_addr_list **to, int *to_count,
		    struct dev_addr_list **from, int *from_count)
{
	struct dev_addr_list *da, *next;
	int err = 0;

	da = *from;
	while (da != NULL) {
		next = da->next;
		if (!da->da_synced) {
			err = __dev_addr_add(to, to_count,
					     da->da_addr, da->da_addrlen, 0);
			if (err < 0)
				break;
			da->da_synced = 1;
			da->da_users++;
		} else if (da->da_users == 1) {
			__dev_addr_delete(to, to_count,
					  da->da_addr, da->da_addrlen, 0);
			__dev_addr_delete(from, from_count,
					  da->da_addr, da->da_addrlen, 0);
		}
		da = next;
	}
	return err;
}

void __dev_addr_unsync(struct dev_addr_list **to, int *to_count,
		       struct dev_addr_list **from, int *from_count)
{
	struct dev_addr_list *da, *next;

	da = *from;
	while (da != NULL) {
		next = da->next;
		if (da->da_synced) {
			__dev_addr_delete(to, to_count,
					  da->da_addr, da->da_addrlen, 0);
			da->da_synced = 0;
			__dev_addr_delete(from, from_count,
					  da->da_addr, da->da_addrlen, 0);
		}
		da = next;
	}
}

/**
 *	dev_unicast_sync - Synchronize device's unicast list to another device
 *	@to: destination device
 *	@from: source device
 *
 *	Add newly added addresses to the destination device and release
 *	addresses that have no users left. The source device must be
 *	locked by netif_tx_lock_bh.
 *
 *	This function is intended to be called from the dev->set_rx_mode
 *	function of layered software devices.
 */
int dev_unicast_sync(struct net_device *to, struct net_device *from)
{
	int err = 0;

3249
	netif_addr_lock_bh(to);
3250 3251 3252 3253
	err = __dev_addr_sync(&to->uc_list, &to->uc_count,
			      &from->uc_list, &from->uc_count);
	if (!err)
		__dev_set_rx_mode(to);
3254
	netif_addr_unlock_bh(to);
3255 3256 3257 3258 3259
	return err;
}
EXPORT_SYMBOL(dev_unicast_sync);

/**
3260
 *	dev_unicast_unsync - Remove synchronized addresses from the destination device
3261 3262 3263 3264 3265 3266 3267 3268 3269
 *	@to: destination device
 *	@from: source device
 *
 *	Remove all addresses that were added to the destination device by
 *	dev_unicast_sync(). This function is intended to be called from the
 *	dev->stop function of layered software devices.
 */
void dev_unicast_unsync(struct net_device *to, struct net_device *from)
{
3270
	netif_addr_lock_bh(from);
3271
	netif_addr_lock(to);
3272 3273 3274 3275 3276

	__dev_addr_unsync(&to->uc_list, &to->uc_count,
			  &from->uc_list, &from->uc_count);
	__dev_set_rx_mode(to);

3277
	netif_addr_unlock(to);
3278
	netif_addr_unlock_bh(from);
3279 3280 3281
}
EXPORT_SYMBOL(dev_unicast_unsync);

3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295
static void __dev_addr_discard(struct dev_addr_list **list)
{
	struct dev_addr_list *tmp;

	while (*list != NULL) {
		tmp = *list;
		*list = tmp->next;
		if (tmp->da_users > tmp->da_gusers)
			printk("__dev_addr_discard: address leakage! "
			       "da_users=%d\n", tmp->da_users);
		kfree(tmp);
	}
}

3296
static void dev_addr_discard(struct net_device *dev)
3297
{
3298
	netif_addr_lock_bh(dev);
3299

3300 3301 3302
	__dev_addr_discard(&dev->uc_list);
	dev->uc_count = 0;

3303 3304
	__dev_addr_discard(&dev->mc_list);
	dev->mc_count = 0;
3305

3306
	netif_addr_unlock_bh(dev);
3307 3308
}

L
Linus Torvalds 已提交
3309 3310 3311 3312 3313 3314
unsigned dev_get_flags(const struct net_device *dev)
{
	unsigned flags;

	flags = (dev->flags & ~(IFF_PROMISC |
				IFF_ALLMULTI |
S
Stefan Rompf 已提交
3315 3316 3317
				IFF_RUNNING |
				IFF_LOWER_UP |
				IFF_DORMANT)) |
L
Linus Torvalds 已提交
3318 3319 3320
		(dev->gflags & (IFF_PROMISC |
				IFF_ALLMULTI));

S
Stefan Rompf 已提交
3321 3322 3323 3324 3325 3326 3327 3328
	if (netif_running(dev)) {
		if (netif_oper_up(dev))
			flags |= IFF_RUNNING;
		if (netif_carrier_ok(dev))
			flags |= IFF_LOWER_UP;
		if (netif_dormant(dev))
			flags |= IFF_DORMANT;
	}
L
Linus Torvalds 已提交
3329 3330 3331 3332 3333 3334

	return flags;
}

int dev_change_flags(struct net_device *dev, unsigned flags)
{
3335
	int ret, changes;
L
Linus Torvalds 已提交
3336 3337
	int old_flags = dev->flags;

3338 3339
	ASSERT_RTNL();

L
Linus Torvalds 已提交
3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353
	/*
	 *	Set the flags on our device.
	 */

	dev->flags = (flags & (IFF_DEBUG | IFF_NOTRAILERS | IFF_NOARP |
			       IFF_DYNAMIC | IFF_MULTICAST | IFF_PORTSEL |
			       IFF_AUTOMEDIA)) |
		     (dev->flags & (IFF_UP | IFF_VOLATILE | IFF_PROMISC |
				    IFF_ALLMULTI));

	/*
	 *	Load in the correct multicast list now the flags have changed.
	 */

3354 3355
	if ((old_flags ^ flags) & IFF_MULTICAST)
		dev_change_rx_flags(dev, IFF_MULTICAST);
3356

3357
	dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369

	/*
	 *	Have we downed the interface. We handle IFF_UP ourselves
	 *	according to user attempts to set it, rather than blindly
	 *	setting it.
	 */

	ret = 0;
	if ((old_flags ^ flags) & IFF_UP) {	/* Bit is different  ? */
		ret = ((old_flags & IFF_UP) ? dev_close : dev_open)(dev);

		if (!ret)
3370
			dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
3371 3372 3373 3374 3375
	}

	if (dev->flags & IFF_UP &&
	    ((old_flags ^ dev->flags) &~ (IFF_UP | IFF_PROMISC | IFF_ALLMULTI |
					  IFF_VOLATILE)))
3376
		call_netdevice_notifiers(NETDEV_CHANGE, dev);
L
Linus Torvalds 已提交
3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393

	if ((flags ^ dev->gflags) & IFF_PROMISC) {
		int inc = (flags & IFF_PROMISC) ? +1 : -1;
		dev->gflags ^= IFF_PROMISC;
		dev_set_promiscuity(dev, inc);
	}

	/* NOTE: order of synchronization of IFF_PROMISC and IFF_ALLMULTI
	   is important. Some (broken) drivers set IFF_PROMISC, when
	   IFF_ALLMULTI is requested not asking us and not reporting.
	 */
	if ((flags ^ dev->gflags) & IFF_ALLMULTI) {
		int inc = (flags & IFF_ALLMULTI) ? +1 : -1;
		dev->gflags ^= IFF_ALLMULTI;
		dev_set_allmulti(dev, inc);
	}

3394 3395 3396 3397
	/* Exclude state transition flags, already notified */
	changes = (old_flags ^ dev->flags) & ~(IFF_UP | IFF_RUNNING);
	if (changes)
		rtmsg_ifinfo(RTM_NEWLINK, dev, changes);
L
Linus Torvalds 已提交
3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421

	return ret;
}

int dev_set_mtu(struct net_device *dev, int new_mtu)
{
	int err;

	if (new_mtu == dev->mtu)
		return 0;

	/*	MTU must be positive.	 */
	if (new_mtu < 0)
		return -EINVAL;

	if (!netif_device_present(dev))
		return -ENODEV;

	err = 0;
	if (dev->change_mtu)
		err = dev->change_mtu(dev, new_mtu);
	else
		dev->mtu = new_mtu;
	if (!err && dev->flags & IFF_UP)
3422
		call_netdevice_notifiers(NETDEV_CHANGEMTU, dev);
L
Linus Torvalds 已提交
3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437
	return err;
}

int dev_set_mac_address(struct net_device *dev, struct sockaddr *sa)
{
	int err;

	if (!dev->set_mac_address)
		return -EOPNOTSUPP;
	if (sa->sa_family != dev->type)
		return -EINVAL;
	if (!netif_device_present(dev))
		return -ENODEV;
	err = dev->set_mac_address(dev, sa);
	if (!err)
3438
		call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
L
Linus Torvalds 已提交
3439 3440 3441 3442
	return err;
}

/*
3443
 *	Perform the SIOCxIFxxx calls, inside read_lock(dev_base_lock)
L
Linus Torvalds 已提交
3444
 */
3445
static int dev_ifsioc_locked(struct net *net, struct ifreq *ifr, unsigned int cmd)
L
Linus Torvalds 已提交
3446 3447
{
	int err;
3448
	struct net_device *dev = __dev_get_by_name(net, ifr->ifr_name);
L
Linus Torvalds 已提交
3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475

	if (!dev)
		return -ENODEV;

	switch (cmd) {
		case SIOCGIFFLAGS:	/* Get interface flags */
			ifr->ifr_flags = dev_get_flags(dev);
			return 0;

		case SIOCGIFMETRIC:	/* Get the metric on the interface
					   (currently unused) */
			ifr->ifr_metric = 0;
			return 0;

		case SIOCGIFMTU:	/* Get the MTU of a device */
			ifr->ifr_mtu = dev->mtu;
			return 0;

		case SIOCGIFHWADDR:
			if (!dev->addr_len)
				memset(ifr->ifr_hwaddr.sa_data, 0, sizeof ifr->ifr_hwaddr.sa_data);
			else
				memcpy(ifr->ifr_hwaddr.sa_data, dev->dev_addr,
				       min(sizeof ifr->ifr_hwaddr.sa_data, (size_t) dev->addr_len));
			ifr->ifr_hwaddr.sa_family = dev->type;
			return 0;

3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530
		case SIOCGIFSLAVE:
			err = -EINVAL;
			break;

		case SIOCGIFMAP:
			ifr->ifr_map.mem_start = dev->mem_start;
			ifr->ifr_map.mem_end   = dev->mem_end;
			ifr->ifr_map.base_addr = dev->base_addr;
			ifr->ifr_map.irq       = dev->irq;
			ifr->ifr_map.dma       = dev->dma;
			ifr->ifr_map.port      = dev->if_port;
			return 0;

		case SIOCGIFINDEX:
			ifr->ifr_ifindex = dev->ifindex;
			return 0;

		case SIOCGIFTXQLEN:
			ifr->ifr_qlen = dev->tx_queue_len;
			return 0;

		default:
			/* dev_ioctl() should ensure this case
			 * is never reached
			 */
			WARN_ON(1);
			err = -EINVAL;
			break;

	}
	return err;
}

/*
 *	Perform the SIOCxIFxxx calls, inside rtnl_lock()
 */
static int dev_ifsioc(struct net *net, struct ifreq *ifr, unsigned int cmd)
{
	int err;
	struct net_device *dev = __dev_get_by_name(net, ifr->ifr_name);

	if (!dev)
		return -ENODEV;

	switch (cmd) {
		case SIOCSIFFLAGS:	/* Set interface flags */
			return dev_change_flags(dev, ifr->ifr_flags);

		case SIOCSIFMETRIC:	/* Set the metric on the interface
					   (currently unused) */
			return -EOPNOTSUPP;

		case SIOCSIFMTU:	/* Set the MTU of a device */
			return dev_set_mtu(dev, ifr->ifr_mtu);

L
Linus Torvalds 已提交
3531 3532 3533 3534 3535 3536 3537 3538
		case SIOCSIFHWADDR:
			return dev_set_mac_address(dev, &ifr->ifr_hwaddr);

		case SIOCSIFHWBROADCAST:
			if (ifr->ifr_hwaddr.sa_family != dev->type)
				return -EINVAL;
			memcpy(dev->broadcast, ifr->ifr_hwaddr.sa_data,
			       min(sizeof ifr->ifr_hwaddr.sa_data, (size_t) dev->addr_len));
3539
			call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
L
Linus Torvalds 已提交
3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550
			return 0;

		case SIOCSIFMAP:
			if (dev->set_config) {
				if (!netif_device_present(dev))
					return -ENODEV;
				return dev->set_config(dev, &ifr->ifr_map);
			}
			return -EOPNOTSUPP;

		case SIOCADDMULTI:
3551
			if ((!dev->set_multicast_list && !dev->set_rx_mode) ||
L
Linus Torvalds 已提交
3552 3553 3554 3555 3556 3557 3558 3559
			    ifr->ifr_hwaddr.sa_family != AF_UNSPEC)
				return -EINVAL;
			if (!netif_device_present(dev))
				return -ENODEV;
			return dev_mc_add(dev, ifr->ifr_hwaddr.sa_data,
					  dev->addr_len, 1);

		case SIOCDELMULTI:
3560
			if ((!dev->set_multicast_list && !dev->set_rx_mode) ||
L
Linus Torvalds 已提交
3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618
			    ifr->ifr_hwaddr.sa_family != AF_UNSPEC)
				return -EINVAL;
			if (!netif_device_present(dev))
				return -ENODEV;
			return dev_mc_delete(dev, ifr->ifr_hwaddr.sa_data,
					     dev->addr_len, 1);

		case SIOCSIFTXQLEN:
			if (ifr->ifr_qlen < 0)
				return -EINVAL;
			dev->tx_queue_len = ifr->ifr_qlen;
			return 0;

		case SIOCSIFNAME:
			ifr->ifr_newname[IFNAMSIZ-1] = '\0';
			return dev_change_name(dev, ifr->ifr_newname);

		/*
		 *	Unknown or private ioctl
		 */

		default:
			if ((cmd >= SIOCDEVPRIVATE &&
			    cmd <= SIOCDEVPRIVATE + 15) ||
			    cmd == SIOCBONDENSLAVE ||
			    cmd == SIOCBONDRELEASE ||
			    cmd == SIOCBONDSETHWADDR ||
			    cmd == SIOCBONDSLAVEINFOQUERY ||
			    cmd == SIOCBONDINFOQUERY ||
			    cmd == SIOCBONDCHANGEACTIVE ||
			    cmd == SIOCGMIIPHY ||
			    cmd == SIOCGMIIREG ||
			    cmd == SIOCSMIIREG ||
			    cmd == SIOCBRADDIF ||
			    cmd == SIOCBRDELIF ||
			    cmd == SIOCWANDEV) {
				err = -EOPNOTSUPP;
				if (dev->do_ioctl) {
					if (netif_device_present(dev))
						err = dev->do_ioctl(dev, ifr,
								    cmd);
					else
						err = -ENODEV;
				}
			} else
				err = -EINVAL;

	}
	return err;
}

/*
 *	This function handles all "interface"-type I/O control requests. The actual
 *	'doing' part of this is dev_ifsioc above.
 */

/**
 *	dev_ioctl	-	network device ioctl
3619
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
3620 3621 3622 3623 3624 3625 3626 3627 3628
 *	@cmd: command to issue
 *	@arg: pointer to a struct ifreq in user space
 *
 *	Issue ioctl functions to devices. This is normally called by the
 *	user space syscall interfaces but can sometimes be useful for
 *	other purposes. The return value is the return from the syscall if
 *	positive or a negative errno code on error.
 */

3629
int dev_ioctl(struct net *net, unsigned int cmd, void __user *arg)
L
Linus Torvalds 已提交
3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640
{
	struct ifreq ifr;
	int ret;
	char *colon;

	/* One special case: SIOCGIFCONF takes ifconf argument
	   and requires shared lock, because it sleeps writing
	   to user space.
	 */

	if (cmd == SIOCGIFCONF) {
3641
		rtnl_lock();
3642
		ret = dev_ifconf(net, (char __user *) arg);
3643
		rtnl_unlock();
L
Linus Torvalds 已提交
3644 3645 3646
		return ret;
	}
	if (cmd == SIOCGIFNAME)
3647
		return dev_ifname(net, (struct ifreq __user *)arg);
L
Linus Torvalds 已提交
3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676

	if (copy_from_user(&ifr, arg, sizeof(struct ifreq)))
		return -EFAULT;

	ifr.ifr_name[IFNAMSIZ-1] = 0;

	colon = strchr(ifr.ifr_name, ':');
	if (colon)
		*colon = 0;

	/*
	 *	See which interface the caller is talking about.
	 */

	switch (cmd) {
		/*
		 *	These ioctl calls:
		 *	- can be done by all.
		 *	- atomic and do not require locking.
		 *	- return a value
		 */
		case SIOCGIFFLAGS:
		case SIOCGIFMETRIC:
		case SIOCGIFMTU:
		case SIOCGIFHWADDR:
		case SIOCGIFSLAVE:
		case SIOCGIFMAP:
		case SIOCGIFINDEX:
		case SIOCGIFTXQLEN:
3677
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3678
			read_lock(&dev_base_lock);
3679
			ret = dev_ifsioc_locked(net, &ifr, cmd);
L
Linus Torvalds 已提交
3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690
			read_unlock(&dev_base_lock);
			if (!ret) {
				if (colon)
					*colon = ':';
				if (copy_to_user(arg, &ifr,
						 sizeof(struct ifreq)))
					ret = -EFAULT;
			}
			return ret;

		case SIOCETHTOOL:
3691
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3692
			rtnl_lock();
3693
			ret = dev_ethtool(net, &ifr);
L
Linus Torvalds 已提交
3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714
			rtnl_unlock();
			if (!ret) {
				if (colon)
					*colon = ':';
				if (copy_to_user(arg, &ifr,
						 sizeof(struct ifreq)))
					ret = -EFAULT;
			}
			return ret;

		/*
		 *	These ioctl calls:
		 *	- require superuser power.
		 *	- require strict serialization.
		 *	- return a value
		 */
		case SIOCGMIIPHY:
		case SIOCGMIIREG:
		case SIOCSIFNAME:
			if (!capable(CAP_NET_ADMIN))
				return -EPERM;
3715
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3716
			rtnl_lock();
3717
			ret = dev_ifsioc(net, &ifr, cmd);
L
Linus Torvalds 已提交
3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752
			rtnl_unlock();
			if (!ret) {
				if (colon)
					*colon = ':';
				if (copy_to_user(arg, &ifr,
						 sizeof(struct ifreq)))
					ret = -EFAULT;
			}
			return ret;

		/*
		 *	These ioctl calls:
		 *	- require superuser power.
		 *	- require strict serialization.
		 *	- do not return a value
		 */
		case SIOCSIFFLAGS:
		case SIOCSIFMETRIC:
		case SIOCSIFMTU:
		case SIOCSIFMAP:
		case SIOCSIFHWADDR:
		case SIOCSIFSLAVE:
		case SIOCADDMULTI:
		case SIOCDELMULTI:
		case SIOCSIFHWBROADCAST:
		case SIOCSIFTXQLEN:
		case SIOCSMIIREG:
		case SIOCBONDENSLAVE:
		case SIOCBONDRELEASE:
		case SIOCBONDSETHWADDR:
		case SIOCBONDCHANGEACTIVE:
		case SIOCBRADDIF:
		case SIOCBRDELIF:
			if (!capable(CAP_NET_ADMIN))
				return -EPERM;
3753 3754 3755
			/* fall through */
		case SIOCBONDSLAVEINFOQUERY:
		case SIOCBONDINFOQUERY:
3756
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3757
			rtnl_lock();
3758
			ret = dev_ifsioc(net, &ifr, cmd);
L
Linus Torvalds 已提交
3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777
			rtnl_unlock();
			return ret;

		case SIOCGIFMEM:
			/* Get the per device memory space. We can add this but
			 * currently do not support it */
		case SIOCSIFMEM:
			/* Set the per device memory buffer space.
			 * Not applicable in our case */
		case SIOCSIFLINK:
			return -EINVAL;

		/*
		 *	Unknown or private ioctl.
		 */
		default:
			if (cmd == SIOCWANDEV ||
			    (cmd >= SIOCDEVPRIVATE &&
			     cmd <= SIOCDEVPRIVATE + 15)) {
3778
				dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3779
				rtnl_lock();
3780
				ret = dev_ifsioc(net, &ifr, cmd);
L
Linus Torvalds 已提交
3781 3782 3783 3784 3785 3786 3787
				rtnl_unlock();
				if (!ret && copy_to_user(arg, &ifr,
							 sizeof(struct ifreq)))
					ret = -EFAULT;
				return ret;
			}
			/* Take care of Wireless Extensions */
3788
			if (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST)
3789
				return wext_handle_ioctl(net, &ifr, cmd, arg);
L
Linus Torvalds 已提交
3790 3791 3792 3793 3794 3795 3796
			return -EINVAL;
	}
}


/**
 *	dev_new_index	-	allocate an ifindex
3797
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
3798 3799 3800 3801 3802
 *
 *	Returns a suitable unique value for a new device interface
 *	number.  The caller must hold the rtnl semaphore or the
 *	dev_base_lock to be sure it remains unique.
 */
3803
static int dev_new_index(struct net *net)
L
Linus Torvalds 已提交
3804 3805 3806 3807 3808
{
	static int ifindex;
	for (;;) {
		if (++ifindex <= 0)
			ifindex = 1;
3809
		if (!__dev_get_by_index(net, ifindex))
L
Linus Torvalds 已提交
3810 3811 3812 3813 3814
			return ifindex;
	}
}

/* Delayed registration/unregisteration */
3815
static LIST_HEAD(net_todo_list);
L
Linus Torvalds 已提交
3816

3817
static void net_set_todo(struct net_device *dev)
L
Linus Torvalds 已提交
3818 3819 3820 3821
{
	list_add_tail(&dev->todo_list, &net_todo_list);
}

3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865
static void rollback_registered(struct net_device *dev)
{
	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

	/* Some devices call without registering for initialization unwind. */
	if (dev->reg_state == NETREG_UNINITIALIZED) {
		printk(KERN_DEBUG "unregister_netdevice: device %s/%p never "
				  "was registered\n", dev->name, dev);

		WARN_ON(1);
		return;
	}

	BUG_ON(dev->reg_state != NETREG_REGISTERED);

	/* If device is running, close it first. */
	dev_close(dev);

	/* And unlink it from device chain. */
	unlist_netdevice(dev);

	dev->reg_state = NETREG_UNREGISTERING;

	synchronize_net();

	/* Shutdown queueing discipline. */
	dev_shutdown(dev);


	/* Notify protocols, that we are about to destroy
	   this device. They should clean all the things.
	*/
	call_netdevice_notifiers(NETDEV_UNREGISTER, dev);

	/*
	 *	Flush the unicast and multicast chains
	 */
	dev_addr_discard(dev);

	if (dev->uninit)
		dev->uninit(dev);

	/* Notifier chain MUST detach us from master device. */
3866
	WARN_ON(dev->master);
3867 3868 3869 3870 3871 3872 3873 3874 3875

	/* Remove entries from kobject tree */
	netdev_unregister_kobject(dev);

	synchronize_net();

	dev_put(dev);
}

3876 3877 3878
static void __netdev_init_queue_locks_one(struct net_device *dev,
					  struct netdev_queue *dev_queue,
					  void *_unused)
3879 3880
{
	spin_lock_init(&dev_queue->_xmit_lock);
3881
	netdev_set_xmit_lockdep_class(&dev_queue->_xmit_lock, dev->type);
3882 3883 3884 3885 3886
	dev_queue->xmit_lock_owner = -1;
}

static void netdev_init_queue_locks(struct net_device *dev)
{
3887 3888
	netdev_for_each_tx_queue(dev, __netdev_init_queue_locks_one, NULL);
	__netdev_init_queue_locks_one(dev, &dev->rx_queue, NULL);
3889 3890
}

L
Linus Torvalds 已提交
3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912
/**
 *	register_netdevice	- register a network device
 *	@dev: device to register
 *
 *	Take a completed network device structure and add it to the kernel
 *	interfaces. A %NETDEV_REGISTER message is sent to the netdev notifier
 *	chain. 0 is returned on success. A negative errno code is returned
 *	on a failure to set up the device, or if the name is a duplicate.
 *
 *	Callers must hold the rtnl semaphore. You may want
 *	register_netdev() instead of this.
 *
 *	BUGS:
 *	The locking appears insufficient to guarantee two parallel registers
 *	will not get the same name.
 */

int register_netdevice(struct net_device *dev)
{
	struct hlist_head *head;
	struct hlist_node *p;
	int ret;
3913
	struct net *net;
L
Linus Torvalds 已提交
3914 3915 3916 3917

	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

3918 3919
	might_sleep();

L
Linus Torvalds 已提交
3920 3921
	/* When net_device's are persistent, this will be fatal. */
	BUG_ON(dev->reg_state != NETREG_UNINITIALIZED);
3922 3923
	BUG_ON(!dev_net(dev));
	net = dev_net(dev);
L
Linus Torvalds 已提交
3924

3925
	spin_lock_init(&dev->addr_list_lock);
3926
	netdev_set_addr_lockdep_class(dev);
3927
	netdev_init_queue_locks(dev);
L
Linus Torvalds 已提交
3928 3929 3930 3931 3932 3933 3934 3935 3936

	dev->iflink = -1;

	/* Init, if this function is available */
	if (dev->init) {
		ret = dev->init(dev);
		if (ret) {
			if (ret > 0)
				ret = -EIO;
3937
			goto out;
L
Linus Torvalds 已提交
3938 3939
		}
	}
3940

L
Linus Torvalds 已提交
3941 3942
	if (!dev_valid_name(dev->name)) {
		ret = -EINVAL;
3943
		goto err_uninit;
L
Linus Torvalds 已提交
3944 3945
	}

3946
	dev->ifindex = dev_new_index(net);
L
Linus Torvalds 已提交
3947 3948 3949 3950
	if (dev->iflink == -1)
		dev->iflink = dev->ifindex;

	/* Check for existence of name */
3951
	head = dev_name_hash(net, dev->name);
L
Linus Torvalds 已提交
3952 3953 3954 3955 3956
	hlist_for_each(p, head) {
		struct net_device *d
			= hlist_entry(p, struct net_device, name_hlist);
		if (!strncmp(d->name, dev->name, IFNAMSIZ)) {
			ret = -EEXIST;
3957
			goto err_uninit;
L
Linus Torvalds 已提交
3958
		}
3959
	}
L
Linus Torvalds 已提交
3960

3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976
	/* Fix illegal checksum combinations */
	if ((dev->features & NETIF_F_HW_CSUM) &&
	    (dev->features & (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))) {
		printk(KERN_NOTICE "%s: mixed HW and IP checksum settings.\n",
		       dev->name);
		dev->features &= ~(NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM);
	}

	if ((dev->features & NETIF_F_NO_CSUM) &&
	    (dev->features & (NETIF_F_HW_CSUM|NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))) {
		printk(KERN_NOTICE "%s: mixed no checksumming and other settings.\n",
		       dev->name);
		dev->features &= ~(NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM|NETIF_F_HW_CSUM);
	}


L
Linus Torvalds 已提交
3977 3978
	/* Fix illegal SG+CSUM combinations. */
	if ((dev->features & NETIF_F_SG) &&
3979
	    !(dev->features & NETIF_F_ALL_CSUM)) {
3980
		printk(KERN_NOTICE "%s: Dropping NETIF_F_SG since no checksum feature.\n",
L
Linus Torvalds 已提交
3981 3982 3983 3984 3985 3986 3987
		       dev->name);
		dev->features &= ~NETIF_F_SG;
	}

	/* TSO requires that SG is present as well. */
	if ((dev->features & NETIF_F_TSO) &&
	    !(dev->features & NETIF_F_SG)) {
3988
		printk(KERN_NOTICE "%s: Dropping NETIF_F_TSO since no SG feature.\n",
L
Linus Torvalds 已提交
3989 3990 3991
		       dev->name);
		dev->features &= ~NETIF_F_TSO;
	}
3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005
	if (dev->features & NETIF_F_UFO) {
		if (!(dev->features & NETIF_F_HW_CSUM)) {
			printk(KERN_ERR "%s: Dropping NETIF_F_UFO since no "
					"NETIF_F_HW_CSUM feature.\n",
							dev->name);
			dev->features &= ~NETIF_F_UFO;
		}
		if (!(dev->features & NETIF_F_SG)) {
			printk(KERN_ERR "%s: Dropping NETIF_F_UFO since no "
					"NETIF_F_SG feature.\n",
					dev->name);
			dev->features &= ~NETIF_F_UFO;
		}
	}
L
Linus Torvalds 已提交
4006

4007 4008 4009 4010
	/* Enable software GSO if SG is supported. */
	if (dev->features & NETIF_F_SG)
		dev->features |= NETIF_F_GSO;

4011
	netdev_initialize_kobject(dev);
4012
	ret = netdev_register_kobject(dev);
4013
	if (ret)
4014
		goto err_uninit;
4015 4016
	dev->reg_state = NETREG_REGISTERED;

L
Linus Torvalds 已提交
4017 4018 4019 4020 4021 4022 4023 4024 4025
	/*
	 *	Default initial state at registry is that the
	 *	device is present.
	 */

	set_bit(__LINK_STATE_PRESENT, &dev->state);

	dev_init_scheduler(dev);
	dev_hold(dev);
4026
	list_netdevice(dev);
L
Linus Torvalds 已提交
4027 4028

	/* Notify protocols, that a new device appeared. */
4029
	ret = call_netdevice_notifiers(NETDEV_REGISTER, dev);
4030
	ret = notifier_to_errno(ret);
4031 4032 4033 4034
	if (ret) {
		rollback_registered(dev);
		dev->reg_state = NETREG_UNREGISTERED;
	}
L
Linus Torvalds 已提交
4035 4036 4037

out:
	return ret;
4038 4039 4040 4041 4042

err_uninit:
	if (dev->uninit)
		dev->uninit(dev);
	goto out;
L
Linus Torvalds 已提交
4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053
}

/**
 *	register_netdev	- register a network device
 *	@dev: device to register
 *
 *	Take a completed network device structure and add it to the kernel
 *	interfaces. A %NETDEV_REGISTER message is sent to the netdev notifier
 *	chain. 0 is returned on success. A negative errno code is returned
 *	on a failure to set up the device, or if the name is a duplicate.
 *
4054
 *	This is a wrapper around register_netdevice that takes the rtnl semaphore
L
Linus Torvalds 已提交
4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072
 *	and expands the device name if you passed a format string to
 *	alloc_netdev.
 */
int register_netdev(struct net_device *dev)
{
	int err;

	rtnl_lock();

	/*
	 * If the name is a format string the caller wants us to do a
	 * name allocation.
	 */
	if (strchr(dev->name, '%')) {
		err = dev_alloc_name(dev, dev->name);
		if (err < 0)
			goto out;
	}
4073

L
Linus Torvalds 已提交
4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089
	err = register_netdevice(dev);
out:
	rtnl_unlock();
	return err;
}
EXPORT_SYMBOL(register_netdev);

/*
 * netdev_wait_allrefs - wait until all references are gone.
 *
 * This is called when unregistering network devices.
 *
 * Any protocol or device that holds a reference should register
 * for netdevice notification, and cleanup and put back the
 * reference if they receive an UNREGISTER event.
 * We can get stuck here if buggy protocols don't correctly
4090
 * call dev_put.
L
Linus Torvalds 已提交
4091 4092 4093 4094 4095 4096 4097 4098
 */
static void netdev_wait_allrefs(struct net_device *dev)
{
	unsigned long rebroadcast_time, warning_time;

	rebroadcast_time = warning_time = jiffies;
	while (atomic_read(&dev->refcnt) != 0) {
		if (time_after(jiffies, rebroadcast_time + 1 * HZ)) {
4099
			rtnl_lock();
L
Linus Torvalds 已提交
4100 4101

			/* Rebroadcast unregister notification */
4102
			call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
L
Linus Torvalds 已提交
4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114

			if (test_bit(__LINK_STATE_LINKWATCH_PENDING,
				     &dev->state)) {
				/* We must not have linkwatch events
				 * pending on unregister. If this
				 * happens, we simply run the queue
				 * unscheduled, resulting in a noop
				 * for this device.
				 */
				linkwatch_run_queue();
			}

4115
			__rtnl_unlock();
L
Linus Torvalds 已提交
4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145

			rebroadcast_time = jiffies;
		}

		msleep(250);

		if (time_after(jiffies, warning_time + 10 * HZ)) {
			printk(KERN_EMERG "unregister_netdevice: "
			       "waiting for %s to become free. Usage "
			       "count = %d\n",
			       dev->name, atomic_read(&dev->refcnt));
			warning_time = jiffies;
		}
	}
}

/* The sequence is:
 *
 *	rtnl_lock();
 *	...
 *	register_netdevice(x1);
 *	register_netdevice(x2);
 *	...
 *	unregister_netdevice(y1);
 *	unregister_netdevice(y2);
 *      ...
 *	rtnl_unlock();
 *	free_netdev(y1);
 *	free_netdev(y2);
 *
H
Herbert Xu 已提交
4146
 * We are invoked by rtnl_unlock().
L
Linus Torvalds 已提交
4147
 * This allows us to deal with problems:
4148
 * 1) We can delete sysfs objects which invoke hotplug
L
Linus Torvalds 已提交
4149 4150 4151
 *    without deadlocking with linkwatch via keventd.
 * 2) Since we run with the RTNL semaphore not held, we can sleep
 *    safely in order to wait for the netdev refcnt to drop to zero.
H
Herbert Xu 已提交
4152 4153 4154
 *
 * We must not return until all unregister events added during
 * the interval the lock was held have been completed.
L
Linus Torvalds 已提交
4155 4156 4157
 */
void netdev_run_todo(void)
{
4158
	struct list_head list;
L
Linus Torvalds 已提交
4159 4160

	/* Snapshot list, allow later requests */
4161
	list_replace_init(&net_todo_list, &list);
H
Herbert Xu 已提交
4162 4163

	__rtnl_unlock();
4164

L
Linus Torvalds 已提交
4165 4166 4167 4168 4169
	while (!list_empty(&list)) {
		struct net_device *dev
			= list_entry(list.next, struct net_device, todo_list);
		list_del(&dev->todo_list);

4170 4171 4172 4173 4174 4175
		if (unlikely(dev->reg_state != NETREG_UNREGISTERING)) {
			printk(KERN_ERR "network todo '%s' but state %d\n",
			       dev->name, dev->reg_state);
			dump_stack();
			continue;
		}
L
Linus Torvalds 已提交
4176

4177
		dev->reg_state = NETREG_UNREGISTERED;
L
Linus Torvalds 已提交
4178

4179 4180
		on_each_cpu(flush_backlog, dev, 1);

4181
		netdev_wait_allrefs(dev);
L
Linus Torvalds 已提交
4182

4183 4184
		/* paranoia */
		BUG_ON(atomic_read(&dev->refcnt));
4185 4186 4187
		WARN_ON(dev->ip_ptr);
		WARN_ON(dev->ip6_ptr);
		WARN_ON(dev->dn_ptr);
L
Linus Torvalds 已提交
4188

4189 4190
		if (dev->destructor)
			dev->destructor(dev);
4191 4192 4193

		/* Free network device */
		kobject_put(&dev->dev.kobj);
L
Linus Torvalds 已提交
4194 4195 4196
	}
}

4197
static struct net_device_stats *internal_stats(struct net_device *dev)
R
Rusty Russell 已提交
4198
{
4199
	return &dev->stats;
R
Rusty Russell 已提交
4200 4201
}

4202
static void netdev_init_one_queue(struct net_device *dev,
4203 4204
				  struct netdev_queue *queue,
				  void *_unused)
4205 4206 4207 4208
{
	queue->dev = dev;
}

4209 4210
static void netdev_init_queues(struct net_device *dev)
{
4211 4212
	netdev_init_one_queue(dev, &dev->rx_queue, NULL);
	netdev_for_each_tx_queue(dev, netdev_init_one_queue, NULL);
4213
	spin_lock_init(&dev->tx_global_lock);
4214 4215
}

L
Linus Torvalds 已提交
4216
/**
4217
 *	alloc_netdev_mq - allocate network device
L
Linus Torvalds 已提交
4218 4219 4220
 *	@sizeof_priv:	size of private data to allocate space for
 *	@name:		device name format string
 *	@setup:		callback to initialize device
4221
 *	@queue_count:	the number of subqueues to allocate
L
Linus Torvalds 已提交
4222 4223
 *
 *	Allocates a struct net_device with private data area for driver use
4224 4225
 *	and performs basic initialization.  Also allocates subquue structs
 *	for each queue on the device at the end of the netdevice.
L
Linus Torvalds 已提交
4226
 */
4227 4228
struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name,
		void (*setup)(struct net_device *), unsigned int queue_count)
L
Linus Torvalds 已提交
4229
{
4230
	struct netdev_queue *tx;
L
Linus Torvalds 已提交
4231
	struct net_device *dev;
4232
	size_t alloc_size;
4233
	void *p;
L
Linus Torvalds 已提交
4234

4235 4236
	BUG_ON(strlen(name) >= sizeof(dev->name));

4237
	alloc_size = sizeof(struct net_device);
4238 4239 4240 4241 4242 4243 4244
	if (sizeof_priv) {
		/* ensure 32-byte alignment of private area */
		alloc_size = (alloc_size + NETDEV_ALIGN_CONST) & ~NETDEV_ALIGN_CONST;
		alloc_size += sizeof_priv;
	}
	/* ensure 32-byte alignment of whole construct */
	alloc_size += NETDEV_ALIGN_CONST;
L
Linus Torvalds 已提交
4245

4246
	p = kzalloc(alloc_size, GFP_KERNEL);
L
Linus Torvalds 已提交
4247
	if (!p) {
4248
		printk(KERN_ERR "alloc_netdev: Unable to allocate device.\n");
L
Linus Torvalds 已提交
4249 4250 4251
		return NULL;
	}

4252
	tx = kcalloc(queue_count, sizeof(struct netdev_queue), GFP_KERNEL);
4253 4254 4255 4256 4257 4258 4259
	if (!tx) {
		printk(KERN_ERR "alloc_netdev: Unable to allocate "
		       "tx qdiscs.\n");
		kfree(p);
		return NULL;
	}

L
Linus Torvalds 已提交
4260 4261 4262
	dev = (struct net_device *)
		(((long)p + NETDEV_ALIGN_CONST) & ~NETDEV_ALIGN_CONST);
	dev->padded = (char *)dev - (char *)p;
4263
	dev_net_set(dev, &init_net);
L
Linus Torvalds 已提交
4264

4265 4266
	dev->_tx = tx;
	dev->num_tx_queues = queue_count;
4267
	dev->real_num_tx_queues = queue_count;
4268

4269 4270
	if (sizeof_priv) {
		dev->priv = ((char *)dev +
4271
			     ((sizeof(struct net_device) + NETDEV_ALIGN_CONST)
4272 4273 4274
			      & ~NETDEV_ALIGN_CONST));
	}

4275
	dev->gso_max_size = GSO_MAX_SIZE;
L
Linus Torvalds 已提交
4276

4277 4278
	netdev_init_queues(dev);

4279
	dev->get_stats = internal_stats;
4280
	netpoll_netdev_init(dev);
L
Linus Torvalds 已提交
4281 4282 4283 4284
	setup(dev);
	strcpy(dev->name, name);
	return dev;
}
4285
EXPORT_SYMBOL(alloc_netdev_mq);
L
Linus Torvalds 已提交
4286 4287 4288 4289 4290

/**
 *	free_netdev - free network device
 *	@dev: device
 *
4291 4292
 *	This function does the last stage of destroying an allocated device
 * 	interface. The reference to the device object is released.
L
Linus Torvalds 已提交
4293 4294 4295 4296
 *	If this is the last reference then it will be freed.
 */
void free_netdev(struct net_device *dev)
{
4297 4298
	release_net(dev_net(dev));

4299 4300
	kfree(dev->_tx);

S
Stephen Hemminger 已提交
4301
	/*  Compatibility with error handling in drivers */
L
Linus Torvalds 已提交
4302 4303 4304 4305 4306 4307 4308 4309
	if (dev->reg_state == NETREG_UNINITIALIZED) {
		kfree((char *)dev - dev->padded);
		return;
	}

	BUG_ON(dev->reg_state != NETREG_UNREGISTERED);
	dev->reg_state = NETREG_RELEASED;

4310 4311
	/* will free via device release */
	put_device(&dev->dev);
L
Linus Torvalds 已提交
4312
}
4313

L
Linus Torvalds 已提交
4314
/* Synchronize with packet receive processing. */
4315
void synchronize_net(void)
L
Linus Torvalds 已提交
4316 4317
{
	might_sleep();
4318
	synchronize_rcu();
L
Linus Torvalds 已提交
4319 4320 4321 4322 4323 4324 4325
}

/**
 *	unregister_netdevice - remove device from the kernel
 *	@dev: device
 *
 *	This function shuts down a device interface and removes it
4326
 *	from the kernel tables.
L
Linus Torvalds 已提交
4327 4328 4329 4330 4331
 *
 *	Callers must hold the rtnl semaphore.  You may want
 *	unregister_netdev() instead of this.
 */

4332
void unregister_netdevice(struct net_device *dev)
L
Linus Torvalds 已提交
4333
{
4334 4335
	ASSERT_RTNL();

4336
	rollback_registered(dev);
L
Linus Torvalds 已提交
4337 4338 4339 4340 4341 4342 4343 4344 4345
	/* Finish processing unregister after unlock */
	net_set_todo(dev);
}

/**
 *	unregister_netdev - remove device from the kernel
 *	@dev: device
 *
 *	This function shuts down a device interface and removes it
4346
 *	from the kernel tables.
L
Linus Torvalds 已提交
4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360
 *
 *	This is just a wrapper for unregister_netdevice that takes
 *	the rtnl semaphore.  In general you want to use this and not
 *	unregister_netdevice.
 */
void unregister_netdev(struct net_device *dev)
{
	rtnl_lock();
	unregister_netdevice(dev);
	rtnl_unlock();
}

EXPORT_SYMBOL(unregister_netdev);

4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394
/**
 *	dev_change_net_namespace - move device to different nethost namespace
 *	@dev: device
 *	@net: network namespace
 *	@pat: If not NULL name pattern to try if the current device name
 *	      is already taken in the destination network namespace.
 *
 *	This function shuts down a device interface and moves it
 *	to a new network namespace. On success 0 is returned, on
 *	a failure a netagive errno code is returned.
 *
 *	Callers must hold the rtnl semaphore.
 */

int dev_change_net_namespace(struct net_device *dev, struct net *net, const char *pat)
{
	char buf[IFNAMSIZ];
	const char *destname;
	int err;

	ASSERT_RTNL();

	/* Don't allow namespace local devices to be moved. */
	err = -EINVAL;
	if (dev->features & NETIF_F_NETNS_LOCAL)
		goto out;

	/* Ensure the device has been registrered */
	err = -EINVAL;
	if (dev->reg_state != NETREG_REGISTERED)
		goto out;

	/* Get out if there is nothing todo */
	err = 0;
4395
	if (net_eq(dev_net(dev), net))
4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423
		goto out;

	/* Pick the destination device name, and ensure
	 * we can use it in the destination network namespace.
	 */
	err = -EEXIST;
	destname = dev->name;
	if (__dev_get_by_name(net, destname)) {
		/* We get here if we can't use the current device name */
		if (!pat)
			goto out;
		if (!dev_valid_name(pat))
			goto out;
		if (strchr(pat, '%')) {
			if (__dev_alloc_name(net, pat, buf) < 0)
				goto out;
			destname = buf;
		} else
			destname = pat;
		if (__dev_get_by_name(net, destname))
			goto out;
	}

	/*
	 * And now a mini version of register_netdevice unregister_netdevice.
	 */

	/* If device is running close it first. */
4424
	dev_close(dev);
4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445

	/* And unlink it from device chain */
	err = -ENODEV;
	unlist_netdevice(dev);

	synchronize_net();

	/* Shutdown queueing discipline. */
	dev_shutdown(dev);

	/* Notify protocols, that we are about to destroy
	   this device. They should clean all the things.
	*/
	call_netdevice_notifiers(NETDEV_UNREGISTER, dev);

	/*
	 *	Flush the unicast and multicast chains
	 */
	dev_addr_discard(dev);

	/* Actually switch the network namespace */
4446
	dev_net_set(dev, net);
4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459

	/* Assign the new device name */
	if (destname != dev->name)
		strcpy(dev->name, destname);

	/* If there is an ifindex conflict assign a new one */
	if (__dev_get_by_index(net, dev->ifindex)) {
		int iflink = (dev->iflink == dev->ifindex);
		dev->ifindex = dev_new_index(net);
		if (iflink)
			dev->iflink = dev->ifindex;
	}

4460
	/* Fixup kobjects */
4461 4462
	netdev_unregister_kobject(dev);
	err = netdev_register_kobject(dev);
4463
	WARN_ON(err);
4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476

	/* Add the device back in the hashes */
	list_netdevice(dev);

	/* Notify protocols, that a new device appeared. */
	call_netdevice_notifiers(NETDEV_REGISTER, dev);

	synchronize_net();
	err = 0;
out:
	return err;
}

L
Linus Torvalds 已提交
4477 4478 4479 4480 4481
static int dev_cpu_callback(struct notifier_block *nfb,
			    unsigned long action,
			    void *ocpu)
{
	struct sk_buff **list_skb;
4482
	struct Qdisc **list_net;
L
Linus Torvalds 已提交
4483 4484 4485 4486
	struct sk_buff *skb;
	unsigned int cpu, oldcpu = (unsigned long)ocpu;
	struct softnet_data *sd, *oldsd;

4487
	if (action != CPU_DEAD && action != CPU_DEAD_FROZEN)
L
Linus Torvalds 已提交
4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520
		return NOTIFY_OK;

	local_irq_disable();
	cpu = smp_processor_id();
	sd = &per_cpu(softnet_data, cpu);
	oldsd = &per_cpu(softnet_data, oldcpu);

	/* Find end of our completion_queue. */
	list_skb = &sd->completion_queue;
	while (*list_skb)
		list_skb = &(*list_skb)->next;
	/* Append completion queue from offline CPU. */
	*list_skb = oldsd->completion_queue;
	oldsd->completion_queue = NULL;

	/* Find end of our output_queue. */
	list_net = &sd->output_queue;
	while (*list_net)
		list_net = &(*list_net)->next_sched;
	/* Append output queue from offline CPU. */
	*list_net = oldsd->output_queue;
	oldsd->output_queue = NULL;

	raise_softirq_irqoff(NET_TX_SOFTIRQ);
	local_irq_enable();

	/* Process offline CPU's input_pkt_queue */
	while ((skb = __skb_dequeue(&oldsd->input_pkt_queue)))
		netif_rx(skb);

	return NOTIFY_OK;
}

4521 4522
#ifdef CONFIG_NET_DMA
/**
R
Randy Dunlap 已提交
4523 4524 4525 4526 4527
 * net_dma_rebalance - try to maintain one DMA channel per CPU
 * @net_dma: DMA client and associated data (lock, channels, channel_mask)
 *
 * This is called when the number of channels allocated to the net_dma client
 * changes.  The net_dma client tries to have one DMA channel per CPU.
4528
 */
4529 4530

static void net_dma_rebalance(struct net_dma *net_dma)
4531
{
4532
	unsigned int cpu, i, n, chan_idx;
4533 4534
	struct dma_chan *chan;

4535
	if (cpus_empty(net_dma->channel_mask)) {
4536
		for_each_online_cpu(cpu)
A
Alexey Dobriyan 已提交
4537
			rcu_assign_pointer(per_cpu(softnet_data, cpu).net_dma, NULL);
4538 4539 4540 4541 4542 4543
		return;
	}

	i = 0;
	cpu = first_cpu(cpu_online_map);

4544
	for_each_cpu_mask_nr(chan_idx, net_dma->channel_mask) {
4545 4546 4547 4548 4549
		chan = net_dma->channels[chan_idx];

		n = ((num_online_cpus() / cpus_weight(net_dma->channel_mask))
		   + (i < (num_online_cpus() %
			cpus_weight(net_dma->channel_mask)) ? 1 : 0));
4550 4551

		while(n) {
A
Alexey Dobriyan 已提交
4552
			per_cpu(softnet_data, cpu).net_dma = chan;
4553 4554 4555 4556 4557 4558 4559 4560 4561 4562
			cpu = next_cpu(cpu, cpu_online_map);
			n--;
		}
		i++;
	}
}

/**
 * netdev_dma_event - event callback for the net_dma_client
 * @client: should always be net_dma_client
R
Randy Dunlap 已提交
4563
 * @chan: DMA channel for the event
R
Randy Dunlap 已提交
4564
 * @state: DMA state to be handled
4565
 */
4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577
static enum dma_state_client
netdev_dma_event(struct dma_client *client, struct dma_chan *chan,
	enum dma_state state)
{
	int i, found = 0, pos = -1;
	struct net_dma *net_dma =
		container_of(client, struct net_dma, client);
	enum dma_state_client ack = DMA_DUP; /* default: take no action */

	spin_lock(&net_dma->lock);
	switch (state) {
	case DMA_RESOURCE_AVAILABLE:
4578
		for (i = 0; i < nr_cpu_ids; i++)
4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590
			if (net_dma->channels[i] == chan) {
				found = 1;
				break;
			} else if (net_dma->channels[i] == NULL && pos < 0)
				pos = i;

		if (!found && pos >= 0) {
			ack = DMA_ACK;
			net_dma->channels[pos] = chan;
			cpu_set(pos, net_dma->channel_mask);
			net_dma_rebalance(net_dma);
		}
4591 4592
		break;
	case DMA_RESOURCE_REMOVED:
4593
		for (i = 0; i < nr_cpu_ids; i++)
4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605
			if (net_dma->channels[i] == chan) {
				found = 1;
				pos = i;
				break;
			}

		if (found) {
			ack = DMA_ACK;
			cpu_clear(pos, net_dma->channel_mask);
			net_dma->channels[i] = NULL;
			net_dma_rebalance(net_dma);
		}
4606 4607 4608 4609
		break;
	default:
		break;
	}
4610 4611 4612
	spin_unlock(&net_dma->lock);

	return ack;
4613 4614 4615 4616 4617 4618 4619
}

/**
 * netdev_dma_regiser - register the networking subsystem as a DMA client
 */
static int __init netdev_dma_register(void)
{
4620 4621 4622 4623 4624 4625 4626
	net_dma.channels = kzalloc(nr_cpu_ids * sizeof(struct net_dma),
								GFP_KERNEL);
	if (unlikely(!net_dma.channels)) {
		printk(KERN_NOTICE
				"netdev_dma: no memory for net_dma.channels\n");
		return -ENOMEM;
	}
4627 4628 4629 4630
	spin_lock_init(&net_dma.lock);
	dma_cap_set(DMA_MEMCPY, net_dma.client.cap_mask);
	dma_async_client_register(&net_dma.client);
	dma_async_client_chan_request(&net_dma.client);
4631 4632 4633 4634 4635 4636
	return 0;
}

#else
static int __init netdev_dma_register(void) { return -ENODEV; }
#endif /* CONFIG_NET_DMA */
L
Linus Torvalds 已提交
4637

4638 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676
/**
 *	netdev_compute_feature - compute conjunction of two feature sets
 *	@all: first feature set
 *	@one: second feature set
 *
 *	Computes a new feature set after adding a device with feature set
 *	@one to the master device with current feature set @all.  Returns
 *	the new feature set.
 */
int netdev_compute_features(unsigned long all, unsigned long one)
{
	/* if device needs checksumming, downgrade to hw checksumming */
	if (all & NETIF_F_NO_CSUM && !(one & NETIF_F_NO_CSUM))
		all ^= NETIF_F_NO_CSUM | NETIF_F_HW_CSUM;

	/* if device can't do all checksum, downgrade to ipv4/ipv6 */
	if (all & NETIF_F_HW_CSUM && !(one & NETIF_F_HW_CSUM))
		all ^= NETIF_F_HW_CSUM
			| NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM;

	if (one & NETIF_F_GSO)
		one |= NETIF_F_GSO_SOFTWARE;
	one |= NETIF_F_GSO;

	/* If even one device supports robust GSO, enable it for all. */
	if (one & NETIF_F_GSO_ROBUST)
		all |= NETIF_F_GSO_ROBUST;

	all &= one | NETIF_F_LLTX;

	if (!(all & NETIF_F_ALL_CSUM))
		all &= ~NETIF_F_SG;
	if (!(all & NETIF_F_SG))
		all &= ~NETIF_F_GSO_MASK;

	return all;
}
EXPORT_SYMBOL(netdev_compute_features);

4677 4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689
static struct hlist_head *netdev_create_hash(void)
{
	int i;
	struct hlist_head *hash;

	hash = kmalloc(sizeof(*hash) * NETDEV_HASHENTRIES, GFP_KERNEL);
	if (hash != NULL)
		for (i = 0; i < NETDEV_HASHENTRIES; i++)
			INIT_HLIST_HEAD(&hash[i]);

	return hash;
}

4690
/* Initialize per network namespace state */
4691
static int __net_init netdev_init(struct net *net)
4692 4693 4694
{
	INIT_LIST_HEAD(&net->dev_base_head);

4695 4696 4697
	net->dev_name_head = netdev_create_hash();
	if (net->dev_name_head == NULL)
		goto err_name;
4698

4699 4700 4701
	net->dev_index_head = netdev_create_hash();
	if (net->dev_index_head == NULL)
		goto err_idx;
4702 4703

	return 0;
4704 4705 4706 4707 4708

err_idx:
	kfree(net->dev_name_head);
err_name:
	return -ENOMEM;
4709 4710
}

4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730
char *netdev_drivername(struct net_device *dev, char *buffer, int len)
{
	struct device_driver *driver;
	struct device *parent;

	if (len <= 0 || !buffer)
		return buffer;
	buffer[0] = 0;

	parent = dev->dev.parent;

	if (!parent)
		return buffer;

	driver = parent->driver;
	if (driver && driver->name)
		strlcpy(buffer, driver->name, len);
	return buffer;
}

4731
static void __net_exit netdev_exit(struct net *net)
4732 4733 4734 4735 4736
{
	kfree(net->dev_name_head);
	kfree(net->dev_index_head);
}

4737
static struct pernet_operations __net_initdata netdev_net_ops = {
4738 4739 4740 4741
	.init = netdev_init,
	.exit = netdev_exit,
};

4742
static void __net_exit default_device_exit(struct net *net)
4743 4744 4745 4746 4747 4748 4749 4750 4751
{
	struct net_device *dev, *next;
	/*
	 * Push all migratable of the network devices back to the
	 * initial network namespace
	 */
	rtnl_lock();
	for_each_netdev_safe(net, dev, next) {
		int err;
4752
		char fb_name[IFNAMSIZ];
4753 4754 4755 4756 4757 4758

		/* Ignore unmoveable devices (i.e. loopback) */
		if (dev->features & NETIF_F_NETNS_LOCAL)
			continue;

		/* Push remaing network devices to init_net */
4759 4760
		snprintf(fb_name, IFNAMSIZ, "dev%d", dev->ifindex);
		err = dev_change_net_namespace(dev, &init_net, fb_name);
4761
		if (err) {
4762
			printk(KERN_EMERG "%s: failed to move %s to init_net: %d\n",
4763
				__func__, dev->name, err);
4764
			BUG();
4765 4766 4767 4768 4769
		}
	}
	rtnl_unlock();
}

4770
static struct pernet_operations __net_initdata default_device_ops = {
4771 4772 4773
	.exit = default_device_exit,
};

L
Linus Torvalds 已提交
4774 4775 4776 4777 4778 4779 4780 4781 4782 4783 4784 4785 4786 4787 4788 4789 4790 4791 4792 4793
/*
 *	Initialize the DEV module. At boot time this walks the device list and
 *	unhooks any devices that fail to initialise (normally hardware not
 *	present) and leaves us with a valid list of present and active devices.
 *
 */

/*
 *       This is called single threaded during boot, so no need
 *       to take the rtnl semaphore.
 */
static int __init net_dev_init(void)
{
	int i, rc = -ENOMEM;

	BUG_ON(!dev_boot_phase);

	if (dev_proc_init())
		goto out;

4794
	if (netdev_kobject_init())
L
Linus Torvalds 已提交
4795 4796 4797
		goto out;

	INIT_LIST_HEAD(&ptype_all);
4798
	for (i = 0; i < PTYPE_HASH_SIZE; i++)
L
Linus Torvalds 已提交
4799 4800
		INIT_LIST_HEAD(&ptype_base[i]);

4801 4802
	if (register_pernet_subsys(&netdev_net_ops))
		goto out;
L
Linus Torvalds 已提交
4803

4804 4805 4806
	if (register_pernet_device(&default_device_ops))
		goto out;

L
Linus Torvalds 已提交
4807 4808 4809 4810
	/*
	 *	Initialise the packet receive queues.
	 */

4811
	for_each_possible_cpu(i) {
L
Linus Torvalds 已提交
4812 4813 4814 4815 4816 4817
		struct softnet_data *queue;

		queue = &per_cpu(softnet_data, i);
		skb_queue_head_init(&queue->input_pkt_queue);
		queue->completion_queue = NULL;
		INIT_LIST_HEAD(&queue->poll_list);
4818 4819 4820

		queue->backlog.poll = process_backlog;
		queue->backlog.weight = weight_p;
L
Linus Torvalds 已提交
4821 4822
	}

4823 4824
	netdev_dma_register();

L
Linus Torvalds 已提交
4825 4826
	dev_boot_phase = 0;

4827 4828
	open_softirq(NET_TX_SOFTIRQ, net_tx_action);
	open_softirq(NET_RX_SOFTIRQ, net_rx_action);
L
Linus Torvalds 已提交
4829 4830 4831 4832 4833 4834 4835 4836 4837 4838 4839 4840 4841 4842

	hotcpu_notifier(dev_cpu_callback, 0);
	dst_init();
	dev_mcast_init();
	rc = 0;
out:
	return rc;
}

subsys_initcall(net_dev_init);

EXPORT_SYMBOL(__dev_get_by_index);
EXPORT_SYMBOL(__dev_get_by_name);
EXPORT_SYMBOL(__dev_remove_pack);
4843
EXPORT_SYMBOL(dev_valid_name);
L
Linus Torvalds 已提交
4844 4845 4846 4847 4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 4858 4859 4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871 4872 4873 4874 4875 4876 4877 4878 4879 4880 4881 4882 4883 4884 4885
EXPORT_SYMBOL(dev_add_pack);
EXPORT_SYMBOL(dev_alloc_name);
EXPORT_SYMBOL(dev_close);
EXPORT_SYMBOL(dev_get_by_flags);
EXPORT_SYMBOL(dev_get_by_index);
EXPORT_SYMBOL(dev_get_by_name);
EXPORT_SYMBOL(dev_open);
EXPORT_SYMBOL(dev_queue_xmit);
EXPORT_SYMBOL(dev_remove_pack);
EXPORT_SYMBOL(dev_set_allmulti);
EXPORT_SYMBOL(dev_set_promiscuity);
EXPORT_SYMBOL(dev_change_flags);
EXPORT_SYMBOL(dev_set_mtu);
EXPORT_SYMBOL(dev_set_mac_address);
EXPORT_SYMBOL(free_netdev);
EXPORT_SYMBOL(netdev_boot_setup_check);
EXPORT_SYMBOL(netdev_set_master);
EXPORT_SYMBOL(netdev_state_change);
EXPORT_SYMBOL(netif_receive_skb);
EXPORT_SYMBOL(netif_rx);
EXPORT_SYMBOL(register_gifconf);
EXPORT_SYMBOL(register_netdevice);
EXPORT_SYMBOL(register_netdevice_notifier);
EXPORT_SYMBOL(skb_checksum_help);
EXPORT_SYMBOL(synchronize_net);
EXPORT_SYMBOL(unregister_netdevice);
EXPORT_SYMBOL(unregister_netdevice_notifier);
EXPORT_SYMBOL(net_enable_timestamp);
EXPORT_SYMBOL(net_disable_timestamp);
EXPORT_SYMBOL(dev_get_flags);

#if defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)
EXPORT_SYMBOL(br_handle_frame_hook);
EXPORT_SYMBOL(br_fdb_get_hook);
EXPORT_SYMBOL(br_fdb_put_hook);
#endif

#ifdef CONFIG_KMOD
EXPORT_SYMBOL(dev_load);
#endif

EXPORT_PER_CPU_SYMBOL(softnet_data);