dev.c 151.0 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
#include <linux/cpu.h>
#include <linux/types.h>
#include <linux/kernel.h>
82
#include <linux/hash.h>
L
Linus Torvalds 已提交
83
#include <linux/sched.h>
A
Arjan van de Ven 已提交
84
#include <linux/mutex.h>
L
Linus Torvalds 已提交
85 86 87 88 89 90 91 92 93
#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>
94
#include <linux/ethtool.h>
L
Linus Torvalds 已提交
95 96
#include <linux/notifier.h>
#include <linux/skbuff.h>
97
#include <net/net_namespace.h>
L
Linus Torvalds 已提交
98 99 100 101 102 103
#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
104
#include <linux/if_macvlan.h>
L
Linus Torvalds 已提交
105 106 107
#include <net/dst.h>
#include <net/pkt_sched.h>
#include <net/checksum.h>
108
#include <net/xfrm.h>
L
Linus Torvalds 已提交
109 110 111 112 113 114 115
#include <linux/highmem.h>
#include <linux/init.h>
#include <linux/kmod.h>
#include <linux/module.h>
#include <linux/netpoll.h>
#include <linux/rcupdate.h>
#include <linux/delay.h>
116
#include <net/wext.h>
L
Linus Torvalds 已提交
117 118
#include <net/iw_handler.h>
#include <asm/current.h>
S
Steve Grubb 已提交
119
#include <linux/audit.h>
120
#include <linux/dmaengine.h>
121
#include <linux/err.h>
122
#include <linux/ctype.h>
123
#include <linux/if_arp.h>
124
#include <linux/if_vlan.h>
125
#include <linux/ip.h>
126
#include <net/ip.h>
127 128
#include <linux/ipv6.h>
#include <linux/in.h>
D
David S. Miller 已提交
129 130
#include <linux/jhash.h>
#include <linux/random.h>
131
#include <trace/events/napi.h>
L
Linus Torvalds 已提交
132

133 134
#include "net-sysfs.h"

135 136 137
/* Instead of increasing this, you should create a hash table. */
#define MAX_GRO_SKBS 8

H
Herbert Xu 已提交
138 139 140
/* This should be increased if a protocol with a bigger head is added. */
#define GRO_MAX_HEAD (MAX_HEADER + 128)

L
Linus Torvalds 已提交
141 142 143 144 145 146 147 148 149 150 151
/*
 *	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
152
 *             the average user (w/out VLANs) will not be adversely affected.
L
Linus Torvalds 已提交
153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168
 *             --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
 */

169 170 171
#define PTYPE_HASH_SIZE	(16)
#define PTYPE_HASH_MASK	(PTYPE_HASH_SIZE - 1)

L
Linus Torvalds 已提交
172
static DEFINE_SPINLOCK(ptype_lock);
173
static struct list_head ptype_base[PTYPE_HASH_SIZE] __read_mostly;
174
static struct list_head ptype_all __read_mostly;	/* Taps */
L
Linus Torvalds 已提交
175 176

/*
177
 * The @dev_base_head list is protected by @dev_base_lock and the rtnl
L
Linus Torvalds 已提交
178 179
 * semaphore.
 *
180
 * Pure readers hold dev_base_lock for reading, or rcu_read_lock()
L
Linus Torvalds 已提交
181 182
 *
 * Writers must hold the rtnl semaphore while they loop through the
183
 * dev_base_head list, and hold dev_base_lock for writing when they do the
L
Linus Torvalds 已提交
184 185 186 187 188 189 190 191 192 193 194 195 196 197
 * 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);

198
static inline struct hlist_head *dev_name_hash(struct net *net, const char *name)
L
Linus Torvalds 已提交
199 200
{
	unsigned hash = full_name_hash(name, strnlen(name, IFNAMSIZ));
201
	return &net->dev_name_head[hash_32(hash, NETDEV_HASHBITS)];
L
Linus Torvalds 已提交
202 203
}

204
static inline struct hlist_head *dev_index_hash(struct net *net, int ifindex)
L
Linus Torvalds 已提交
205
{
206
	return &net->dev_index_head[ifindex & (NETDEV_HASHENTRIES - 1)];
L
Linus Torvalds 已提交
207 208
}

209 210 211
/* Device list insertion */
static int list_netdevice(struct net_device *dev)
{
212
	struct net *net = dev_net(dev);
213 214 215 216

	ASSERT_RTNL();

	write_lock_bh(&dev_base_lock);
217
	list_add_tail_rcu(&dev->dev_list, &net->dev_base_head);
218
	hlist_add_head_rcu(&dev->name_hlist, dev_name_hash(net, dev->name));
219 220
	hlist_add_head_rcu(&dev->index_hlist,
			   dev_index_hash(net, dev->ifindex));
221 222 223 224
	write_unlock_bh(&dev_base_lock);
	return 0;
}

225 226 227
/* Device list removal
 * caller must respect a RCU grace period before freeing/reusing dev
 */
228 229 230 231 232 233
static void unlist_netdevice(struct net_device *dev)
{
	ASSERT_RTNL();

	/* Unlink dev from the device chain */
	write_lock_bh(&dev_base_lock);
234
	list_del_rcu(&dev->dev_list);
235
	hlist_del_rcu(&dev->name_hlist);
236
	hlist_del_rcu(&dev->index_hlist);
237 238 239
	write_unlock_bh(&dev_base_lock);
}

L
Linus Torvalds 已提交
240 241 242 243
/*
 *	Our notifier list
 */

244
static RAW_NOTIFIER_HEAD(netdev_chain);
L
Linus Torvalds 已提交
245 246 247 248 249

/*
 *	Device drivers call our routines to queue packets here. We empty the
 *	queue in the local softnet handler.
 */
250 251

DEFINE_PER_CPU(struct softnet_data, softnet_data);
E
Eric Dumazet 已提交
252
EXPORT_PER_CPU_SYMBOL(softnet_data);
L
Linus Torvalds 已提交
253

254
#ifdef CONFIG_LOCKDEP
255
/*
256
 * register_netdevice() inits txq->_xmit_lock and sets lockdep class
257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
 * 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,
273
	 ARPHRD_IEEE80211_PRISM, ARPHRD_IEEE80211_RADIOTAP, ARPHRD_PHONET,
274
	 ARPHRD_PHONET_PIPE, ARPHRD_IEEE802154,
275
	 ARPHRD_VOID, ARPHRD_NONE};
276

277
static const char *const netdev_lock_name[] =
278 279 280 281 282 283 284 285 286 287 288 289 290
	{"_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",
291
	 "_xmit_IEEE80211_PRISM", "_xmit_IEEE80211_RADIOTAP", "_xmit_PHONET",
292
	 "_xmit_PHONET_PIPE", "_xmit_IEEE802154",
293
	 "_xmit_VOID", "_xmit_NONE"};
294 295

static struct lock_class_key netdev_xmit_lock_key[ARRAY_SIZE(netdev_lock_type)];
296
static struct lock_class_key netdev_addr_lock_key[ARRAY_SIZE(netdev_lock_type)];
297 298 299 300 301 302 303 304 305 306 307 308

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;
}

309 310
static inline void netdev_set_xmit_lockdep_class(spinlock_t *lock,
						 unsigned short dev_type)
311 312 313 314 315 316 317
{
	int i;

	i = netdev_lock_pos(dev_type);
	lockdep_set_class_and_name(lock, &netdev_xmit_lock_key[i],
				   netdev_lock_name[i]);
}
318 319 320 321 322 323 324 325 326 327

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]);
}
328
#else
329 330 331 332 333
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)
334 335 336
{
}
#endif
L
Linus Torvalds 已提交
337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367

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

		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.
 *
368
 *	This call does not sleep therefore it can not
L
Linus Torvalds 已提交
369 370 371 372 373 374 375 376 377
 *	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
378
	if (pt->type == htons(ETH_P_ALL))
L
Linus Torvalds 已提交
379
		list_add_rcu(&pt->list, &ptype_all);
S
Stephen Hemminger 已提交
380
	else {
381
		hash = ntohs(pt->type) & PTYPE_HASH_MASK;
L
Linus Torvalds 已提交
382 383 384 385
		list_add_rcu(&pt->list, &ptype_base[hash]);
	}
	spin_unlock_bh(&ptype_lock);
}
E
Eric Dumazet 已提交
386
EXPORT_SYMBOL(dev_add_pack);
L
Linus Torvalds 已提交
387 388 389 390 391 392 393 394

/**
 *	__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
395
 *	returns.
L
Linus Torvalds 已提交
396 397 398 399 400 401 402 403 404 405 406 407
 *
 *      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
408
	if (pt->type == htons(ETH_P_ALL))
L
Linus Torvalds 已提交
409
		head = &ptype_all;
S
Stephen Hemminger 已提交
410
	else
411
		head = &ptype_base[ntohs(pt->type) & PTYPE_HASH_MASK];
L
Linus Torvalds 已提交
412 413 414 415 416 417 418 419 420 421 422 423

	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);
}
E
Eric Dumazet 已提交
424 425
EXPORT_SYMBOL(__dev_remove_pack);

L
Linus Torvalds 已提交
426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
/**
 *	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);
441

L
Linus Torvalds 已提交
442 443
	synchronize_net();
}
E
Eric Dumazet 已提交
444
EXPORT_SYMBOL(dev_remove_pack);
L
Linus Torvalds 已提交
445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472

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

		      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));
473
			strlcpy(s[i].name, name, IFNAMSIZ);
L
Linus Torvalds 已提交
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497
			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] != ' ' &&
498
		    !strcmp(dev->name, s[i].name)) {
L
Linus Torvalds 已提交
499 500 501 502 503 504 505 506 507
			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;
}
E
Eric Dumazet 已提交
508
EXPORT_SYMBOL(netdev_boot_setup_check);
L
Linus Torvalds 已提交
509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532


/**
 *	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
	 */
533
	if (__dev_get_by_name(&init_net, name))
L
Linus Torvalds 已提交
534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578
		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
579
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
580 581 582 583 584 585 586 587 588
 *	@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.
 */

589
struct net_device *__dev_get_by_name(struct net *net, const char *name)
L
Linus Torvalds 已提交
590 591
{
	struct hlist_node *p;
E
Eric Dumazet 已提交
592 593
	struct net_device *dev;
	struct hlist_head *head = dev_name_hash(net, name);
L
Linus Torvalds 已提交
594

E
Eric Dumazet 已提交
595
	hlist_for_each_entry(dev, p, head, name_hlist)
L
Linus Torvalds 已提交
596 597
		if (!strncmp(dev->name, name, IFNAMSIZ))
			return dev;
E
Eric Dumazet 已提交
598

L
Linus Torvalds 已提交
599 600
	return NULL;
}
E
Eric Dumazet 已提交
601
EXPORT_SYMBOL(__dev_get_by_name);
L
Linus Torvalds 已提交
602

603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628
/**
 *	dev_get_by_name_rcu	- find a device by its name
 *	@net: the applicable net namespace
 *	@name: name to find
 *
 *	Find an interface by name.
 *	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. The caller must hold RCU lock.
 */

struct net_device *dev_get_by_name_rcu(struct net *net, const char *name)
{
	struct hlist_node *p;
	struct net_device *dev;
	struct hlist_head *head = dev_name_hash(net, name);

	hlist_for_each_entry_rcu(dev, p, head, name_hlist)
		if (!strncmp(dev->name, name, IFNAMSIZ))
			return dev;

	return NULL;
}
EXPORT_SYMBOL(dev_get_by_name_rcu);

L
Linus Torvalds 已提交
629 630
/**
 *	dev_get_by_name		- find a device by its name
631
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
632 633 634 635 636 637 638 639 640
 *	@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.
 */

641
struct net_device *dev_get_by_name(struct net *net, const char *name)
L
Linus Torvalds 已提交
642 643 644
{
	struct net_device *dev;

645 646
	rcu_read_lock();
	dev = dev_get_by_name_rcu(net, name);
L
Linus Torvalds 已提交
647 648
	if (dev)
		dev_hold(dev);
649
	rcu_read_unlock();
L
Linus Torvalds 已提交
650 651
	return dev;
}
E
Eric Dumazet 已提交
652
EXPORT_SYMBOL(dev_get_by_name);
L
Linus Torvalds 已提交
653 654 655

/**
 *	__dev_get_by_index - find a device by its ifindex
656
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
657 658 659 660 661 662 663 664 665
 *	@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.
 */

666
struct net_device *__dev_get_by_index(struct net *net, int ifindex)
L
Linus Torvalds 已提交
667 668
{
	struct hlist_node *p;
E
Eric Dumazet 已提交
669 670
	struct net_device *dev;
	struct hlist_head *head = dev_index_hash(net, ifindex);
L
Linus Torvalds 已提交
671

E
Eric Dumazet 已提交
672
	hlist_for_each_entry(dev, p, head, index_hlist)
L
Linus Torvalds 已提交
673 674
		if (dev->ifindex == ifindex)
			return dev;
E
Eric Dumazet 已提交
675

L
Linus Torvalds 已提交
676 677
	return NULL;
}
E
Eric Dumazet 已提交
678
EXPORT_SYMBOL(__dev_get_by_index);
L
Linus Torvalds 已提交
679

680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704
/**
 *	dev_get_by_index_rcu - find a device by its ifindex
 *	@net: the applicable net namespace
 *	@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 RCU lock.
 */

struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex)
{
	struct hlist_node *p;
	struct net_device *dev;
	struct hlist_head *head = dev_index_hash(net, ifindex);

	hlist_for_each_entry_rcu(dev, p, head, index_hlist)
		if (dev->ifindex == ifindex)
			return dev;

	return NULL;
}
EXPORT_SYMBOL(dev_get_by_index_rcu);

L
Linus Torvalds 已提交
705 706 707

/**
 *	dev_get_by_index - find a device by its ifindex
708
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
709 710 711 712 713 714 715 716
 *	@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.
 */

717
struct net_device *dev_get_by_index(struct net *net, int ifindex)
L
Linus Torvalds 已提交
718 719 720
{
	struct net_device *dev;

721 722
	rcu_read_lock();
	dev = dev_get_by_index_rcu(net, ifindex);
L
Linus Torvalds 已提交
723 724
	if (dev)
		dev_hold(dev);
725
	rcu_read_unlock();
L
Linus Torvalds 已提交
726 727
	return dev;
}
E
Eric Dumazet 已提交
728
EXPORT_SYMBOL(dev_get_by_index);
L
Linus Torvalds 已提交
729 730 731

/**
 *	dev_getbyhwaddr - find a device by its hardware address
732
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
733 734 735 736 737 738 739 740 741 742 743 744
 *	@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
 */

745
struct net_device *dev_getbyhwaddr(struct net *net, unsigned short type, char *ha)
L
Linus Torvalds 已提交
746 747 748 749 750
{
	struct net_device *dev;

	ASSERT_RTNL();

751
	for_each_netdev(net, dev)
L
Linus Torvalds 已提交
752 753
		if (dev->type == type &&
		    !memcmp(dev->dev_addr, ha, dev->addr_len))
754 755 756
			return dev;

	return NULL;
L
Linus Torvalds 已提交
757
}
758 759
EXPORT_SYMBOL(dev_getbyhwaddr);

760
struct net_device *__dev_getfirstbyhwtype(struct net *net, unsigned short type)
L
Linus Torvalds 已提交
761 762 763
{
	struct net_device *dev;

764
	ASSERT_RTNL();
765
	for_each_netdev(net, dev)
766
		if (dev->type == type)
767 768 769
			return dev;

	return NULL;
770 771 772
}
EXPORT_SYMBOL(__dev_getfirstbyhwtype);

773
struct net_device *dev_getfirstbyhwtype(struct net *net, unsigned short type)
774
{
775
	struct net_device *dev, *ret = NULL;
776

777 778 779 780 781 782 783 784 785
	rcu_read_lock();
	for_each_netdev_rcu(net, dev)
		if (dev->type == type) {
			dev_hold(dev);
			ret = dev;
			break;
		}
	rcu_read_unlock();
	return ret;
L
Linus Torvalds 已提交
786 787 788 789 790
}
EXPORT_SYMBOL(dev_getfirstbyhwtype);

/**
 *	dev_get_by_flags - find any device with given flags
791
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
792 793 794 795
 *	@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
796
 *	is not found or a pointer to the device. The device returned has
L
Linus Torvalds 已提交
797 798 799 800
 *	had a reference added and the pointer is safe until the user calls
 *	dev_put to indicate they have finished with it.
 */

E
Eric Dumazet 已提交
801 802
struct net_device *dev_get_by_flags(struct net *net, unsigned short if_flags,
				    unsigned short mask)
L
Linus Torvalds 已提交
803
{
804
	struct net_device *dev, *ret;
L
Linus Torvalds 已提交
805

806
	ret = NULL;
807 808
	rcu_read_lock();
	for_each_netdev_rcu(net, dev) {
L
Linus Torvalds 已提交
809 810
		if (((dev->flags ^ if_flags) & mask) == 0) {
			dev_hold(dev);
811
			ret = dev;
L
Linus Torvalds 已提交
812 813 814
			break;
		}
	}
815
	rcu_read_unlock();
816
	return ret;
L
Linus Torvalds 已提交
817
}
E
Eric Dumazet 已提交
818
EXPORT_SYMBOL(dev_get_by_flags);
L
Linus Torvalds 已提交
819 820 821 822 823 824

/**
 *	dev_valid_name - check if name is okay for network device
 *	@name: name string
 *
 *	Network device names need to be valid file names to
825 826
 *	to allow sysfs to work.  We also disallow any kind of
 *	whitespace.
L
Linus Torvalds 已提交
827
 */
828
int dev_valid_name(const char *name)
L
Linus Torvalds 已提交
829
{
830 831
	if (*name == '\0')
		return 0;
832 833
	if (strlen(name) >= IFNAMSIZ)
		return 0;
834 835 836 837 838 839 840 841 842
	if (!strcmp(name, ".") || !strcmp(name, ".."))
		return 0;

	while (*name) {
		if (*name == '/' || isspace(*name))
			return 0;
		name++;
	}
	return 1;
L
Linus Torvalds 已提交
843
}
E
Eric Dumazet 已提交
844
EXPORT_SYMBOL(dev_valid_name);
L
Linus Torvalds 已提交
845 846

/**
847 848
 *	__dev_alloc_name - allocate a name for a device
 *	@net: network namespace to allocate the device name in
L
Linus Torvalds 已提交
849
 *	@name: name format string
850
 *	@buf:  scratch buffer and result name string
L
Linus Torvalds 已提交
851 852
 *
 *	Passed a format string - eg "lt%d" it will try and find a suitable
S
Stephen Hemminger 已提交
853 854 855 856 857 858
 *	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
859 860
 */

861
static int __dev_alloc_name(struct net *net, const char *name, char *buf)
L
Linus Torvalds 已提交
862 863 864 865
{
	int i = 0;
	const char *p;
	const int max_netdevices = 8*PAGE_SIZE;
S
Stephen Hemminger 已提交
866
	unsigned long *inuse;
L
Linus Torvalds 已提交
867 868 869 870 871 872 873 874 875 876 877 878 879
	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
880
		inuse = (unsigned long *) get_zeroed_page(GFP_ATOMIC);
L
Linus Torvalds 已提交
881 882 883
		if (!inuse)
			return -ENOMEM;

884
		for_each_netdev(net, d) {
L
Linus Torvalds 已提交
885 886 887 888 889 890
			if (!sscanf(d->name, name, &i))
				continue;
			if (i < 0 || i >= max_netdevices)
				continue;

			/*  avoid cases where sscanf is not exact inverse of printf */
891
			snprintf(buf, IFNAMSIZ, name, i);
L
Linus Torvalds 已提交
892 893 894 895 896 897 898 899
			if (!strncmp(buf, d->name, IFNAMSIZ))
				set_bit(i, inuse);
		}

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

900 901
	if (buf != name)
		snprintf(buf, IFNAMSIZ, name, i);
902
	if (!__dev_get_by_name(net, buf))
L
Linus Torvalds 已提交
903 904 905 906 907 908 909 910 911
		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;
}

912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931
/**
 *	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;

932 933
	BUG_ON(!dev_net(dev));
	net = dev_net(dev);
934 935 936 937 938
	ret = __dev_alloc_name(net, name, buf);
	if (ret >= 0)
		strlcpy(dev->name, buf, IFNAMSIZ);
	return ret;
}
E
Eric Dumazet 已提交
939
EXPORT_SYMBOL(dev_alloc_name);
940

941 942 943 944 945 946 947 948 949 950 951 952 953 954 955
static int dev_get_valid_name(struct net *net, const char *name, char *buf,
			      bool fmt)
{
	if (!dev_valid_name(name))
		return -EINVAL;

	if (fmt && strchr(name, '%'))
		return __dev_alloc_name(net, name, buf);
	else if (__dev_get_by_name(net, name))
		return -EEXIST;
	else if (buf != name)
		strlcpy(buf, name, IFNAMSIZ);

	return 0;
}
L
Linus Torvalds 已提交
956 957 958 959 960 961 962 963 964

/**
 *	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.
 */
965
int dev_change_name(struct net_device *dev, const char *newname)
L
Linus Torvalds 已提交
966
{
967
	char oldname[IFNAMSIZ];
L
Linus Torvalds 已提交
968
	int err = 0;
969
	int ret;
970
	struct net *net;
L
Linus Torvalds 已提交
971 972

	ASSERT_RTNL();
973
	BUG_ON(!dev_net(dev));
L
Linus Torvalds 已提交
974

975
	net = dev_net(dev);
L
Linus Torvalds 已提交
976 977 978
	if (dev->flags & IFF_UP)
		return -EBUSY;

979 980 981
	if (strncmp(newname, dev->name, IFNAMSIZ) == 0)
		return 0;

982 983
	memcpy(oldname, dev->name, IFNAMSIZ);

984 985 986
	err = dev_get_valid_name(net, newname, dev->name, 1);
	if (err < 0)
		return err;
L
Linus Torvalds 已提交
987

988
rollback:
989 990 991
	/* For now only devices in the initial network namespace
	 * are in sysfs.
	 */
O
Octavian Purdila 已提交
992
	if (net_eq(net, &init_net)) {
993 994 995 996 997
		ret = device_rename(&dev->dev, dev->name);
		if (ret) {
			memcpy(dev->name, oldname, IFNAMSIZ);
			return ret;
		}
998
	}
999 1000

	write_lock_bh(&dev_base_lock);
1001
	hlist_del(&dev->name_hlist);
1002 1003 1004 1005 1006 1007
	write_unlock_bh(&dev_base_lock);

	synchronize_rcu();

	write_lock_bh(&dev_base_lock);
	hlist_add_head_rcu(&dev->name_hlist, dev_name_hash(net, dev->name));
1008 1009
	write_unlock_bh(&dev_base_lock);

1010
	ret = call_netdevice_notifiers(NETDEV_CHANGENAME, dev);
1011 1012 1013
	ret = notifier_to_errno(ret);

	if (ret) {
1014 1015
		/* err >= 0 after dev_alloc_name() or stores the first errno */
		if (err >= 0) {
1016 1017 1018
			err = ret;
			memcpy(dev->name, oldname, IFNAMSIZ);
			goto rollback;
1019 1020 1021 1022
		} else {
			printk(KERN_ERR
			       "%s: name change rollback failed: %d.\n",
			       dev->name, ret);
1023 1024
		}
	}
L
Linus Torvalds 已提交
1025 1026 1027 1028

	return err;
}

1029 1030 1031 1032
/**
 *	dev_set_alias - change ifalias of a device
 *	@dev: device
 *	@alias: name up to IFALIASZ
1033
 *	@len: limit of bytes to copy from info
1034 1035 1036 1037 1038 1039 1040 1041 1042 1043
 *
 *	Set ifalias for a device,
 */
int dev_set_alias(struct net_device *dev, const char *alias, size_t len)
{
	ASSERT_RTNL();

	if (len >= IFALIASZ)
		return -EINVAL;

1044 1045 1046 1047 1048 1049 1050 1051
	if (!len) {
		if (dev->ifalias) {
			kfree(dev->ifalias);
			dev->ifalias = NULL;
		}
		return 0;
	}

E
Eric Dumazet 已提交
1052
	dev->ifalias = krealloc(dev->ifalias, len + 1, GFP_KERNEL);
1053 1054 1055 1056 1057 1058 1059 1060
	if (!dev->ifalias)
		return -ENOMEM;

	strlcpy(dev->ifalias, alias, len+1);
	return len;
}


1061
/**
S
Stephen Hemminger 已提交
1062
 *	netdev_features_change - device changes features
1063 1064 1065 1066 1067 1068
 *	@dev: device to cause notification
 *
 *	Called to indicate a device has changed features.
 */
void netdev_features_change(struct net_device *dev)
{
1069
	call_netdevice_notifiers(NETDEV_FEAT_CHANGE, dev);
1070 1071 1072
}
EXPORT_SYMBOL(netdev_features_change);

L
Linus Torvalds 已提交
1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083
/**
 *	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) {
1084
		call_netdevice_notifiers(NETDEV_CHANGE, dev);
L
Linus Torvalds 已提交
1085 1086 1087
		rtmsg_ifinfo(RTM_NEWLINK, dev, 0);
	}
}
E
Eric Dumazet 已提交
1088
EXPORT_SYMBOL(netdev_state_change);
L
Linus Torvalds 已提交
1089

1090
int netdev_bonding_change(struct net_device *dev, unsigned long event)
1091
{
1092
	return call_netdevice_notifiers(event, dev);
1093 1094 1095
}
EXPORT_SYMBOL(netdev_bonding_change);

L
Linus Torvalds 已提交
1096 1097
/**
 *	dev_load 	- load a network module
1098
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
1099 1100 1101 1102 1103 1104 1105
 *	@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.
 */

1106
void dev_load(struct net *net, const char *name)
L
Linus Torvalds 已提交
1107
{
1108
	struct net_device *dev;
L
Linus Torvalds 已提交
1109

1110 1111 1112
	rcu_read_lock();
	dev = dev_get_by_name_rcu(net, name);
	rcu_read_unlock();
L
Linus Torvalds 已提交
1113

1114
	if (!dev && capable(CAP_NET_ADMIN))
L
Linus Torvalds 已提交
1115 1116
		request_module("%s", name);
}
E
Eric Dumazet 已提交
1117
EXPORT_SYMBOL(dev_load);
L
Linus Torvalds 已提交
1118

1119
static int __dev_open(struct net_device *dev)
L
Linus Torvalds 已提交
1120
{
1121
	const struct net_device_ops *ops = dev->netdev_ops;
1122
	int ret;
L
Linus Torvalds 已提交
1123

1124 1125
	ASSERT_RTNL();

L
Linus Torvalds 已提交
1126 1127 1128 1129 1130 1131
	/*
	 *	Is it even present?
	 */
	if (!netif_device_present(dev))
		return -ENODEV;

1132 1133 1134 1135 1136
	ret = call_netdevice_notifiers(NETDEV_PRE_UP, dev);
	ret = notifier_to_errno(ret);
	if (ret)
		return ret;

L
Linus Torvalds 已提交
1137 1138 1139 1140
	/*
	 *	Call device private open method
	 */
	set_bit(__LINK_STATE_START, &dev->state);
1141

1142 1143
	if (ops->ndo_validate_addr)
		ret = ops->ndo_validate_addr(dev);
1144

1145 1146
	if (!ret && ops->ndo_open)
		ret = ops->ndo_open(dev);
L
Linus Torvalds 已提交
1147

1148
	/*
L
Linus Torvalds 已提交
1149 1150 1151
	 *	If it went open OK then:
	 */

1152 1153 1154
	if (ret)
		clear_bit(__LINK_STATE_START, &dev->state);
	else {
L
Linus Torvalds 已提交
1155 1156 1157 1158 1159
		/*
		 *	Set the flags.
		 */
		dev->flags |= IFF_UP;

1160 1161 1162
		/*
		 *	Enable NET_DMA
		 */
1163
		net_dmaengine_get();
1164

L
Linus Torvalds 已提交
1165 1166 1167
		/*
		 *	Initialize multicasting status
		 */
1168
		dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
1169 1170 1171 1172 1173 1174

		/*
		 *	Wakeup transmit queue engine
		 */
		dev_activate(dev);
	}
1175

L
Linus Torvalds 已提交
1176 1177 1178 1179
	return ret;
}

/**
1180 1181
 *	dev_open	- prepare an interface for use.
 *	@dev:	device to open
L
Linus Torvalds 已提交
1182
 *
1183 1184 1185 1186 1187 1188 1189
 *	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.
L
Linus Torvalds 已提交
1190
 */
1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218
int dev_open(struct net_device *dev)
{
	int ret;

	/*
	 *	Is it already up?
	 */
	if (dev->flags & IFF_UP)
		return 0;

	/*
	 *	Open device
	 */
	ret = __dev_open(dev);
	if (ret < 0)
		return ret;

	/*
	 *	... and announce new interface.
	 */
	rtmsg_ifinfo(RTM_NEWLINK, dev, IFF_UP|IFF_RUNNING);
	call_netdevice_notifiers(NETDEV_UP, dev);

	return ret;
}
EXPORT_SYMBOL(dev_open);

static int __dev_close(struct net_device *dev)
L
Linus Torvalds 已提交
1219
{
1220
	const struct net_device_ops *ops = dev->netdev_ops;
1221

1222
	ASSERT_RTNL();
1223 1224
	might_sleep();

L
Linus Torvalds 已提交
1225 1226 1227 1228
	/*
	 *	Tell people we are going down, so that they can
	 *	prepare to death, when device is still operating.
	 */
1229
	call_netdevice_notifiers(NETDEV_GOING_DOWN, dev);
L
Linus Torvalds 已提交
1230 1231 1232 1233

	clear_bit(__LINK_STATE_START, &dev->state);

	/* Synchronize to scheduled poll. We cannot touch poll list,
1234 1235 1236 1237 1238
	 * 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
1239 1240
	smp_mb__after_clear_bit(); /* Commit netif_running(). */

1241 1242
	dev_deactivate(dev);

L
Linus Torvalds 已提交
1243 1244 1245 1246 1247 1248 1249
	/*
	 *	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.
	 */
1250 1251
	if (ops->ndo_stop)
		ops->ndo_stop(dev);
L
Linus Torvalds 已提交
1252 1253 1254 1255 1256 1257 1258 1259

	/*
	 *	Device is now down.
	 */

	dev->flags &= ~IFF_UP;

	/*
1260
	 *	Shutdown NET_DMA
L
Linus Torvalds 已提交
1261
	 */
1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281
	net_dmaengine_put();

	return 0;
}

/**
 *	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)
{
	if (!(dev->flags & IFF_UP))
		return 0;

	__dev_close(dev);
L
Linus Torvalds 已提交
1282

1283
	/*
1284
	 * Tell people we are down
1285
	 */
1286 1287
	rtmsg_ifinfo(RTM_NEWLINK, dev, IFF_UP|IFF_RUNNING);
	call_netdevice_notifiers(NETDEV_DOWN, dev);
1288

L
Linus Torvalds 已提交
1289 1290
	return 0;
}
E
Eric Dumazet 已提交
1291
EXPORT_SYMBOL(dev_close);
L
Linus Torvalds 已提交
1292 1293


1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316
/**
 *	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);


1317 1318
static int dev_boot_phase = 1;

L
Linus Torvalds 已提交
1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333
/*
 *	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
1334
 *	to the new notifier to allow device to have a race free
L
Linus Torvalds 已提交
1335 1336 1337 1338 1339 1340
 *	view of the network device list.
 */

int register_netdevice_notifier(struct notifier_block *nb)
{
	struct net_device *dev;
1341
	struct net_device *last;
1342
	struct net *net;
L
Linus Torvalds 已提交
1343 1344 1345
	int err;

	rtnl_lock();
1346
	err = raw_notifier_chain_register(&netdev_chain, nb);
1347 1348
	if (err)
		goto unlock;
1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359
	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
1360

1361 1362
			nb->notifier_call(nb, NETDEV_UP, dev);
		}
L
Linus Torvalds 已提交
1363
	}
1364 1365

unlock:
L
Linus Torvalds 已提交
1366 1367
	rtnl_unlock();
	return err;
1368 1369 1370

rollback:
	last = dev;
1371 1372 1373 1374
	for_each_net(net) {
		for_each_netdev(net, dev) {
			if (dev == last)
				break;
1375

1376 1377 1378 1379 1380
			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);
1381
			nb->notifier_call(nb, NETDEV_UNREGISTER_BATCH, dev);
1382 1383
		}
	}
1384 1385

	raw_notifier_chain_unregister(&netdev_chain, nb);
1386
	goto unlock;
L
Linus Torvalds 已提交
1387
}
E
Eric Dumazet 已提交
1388
EXPORT_SYMBOL(register_netdevice_notifier);
L
Linus Torvalds 已提交
1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401

/**
 *	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)
{
1402 1403 1404
	int err;

	rtnl_lock();
1405
	err = raw_notifier_chain_unregister(&netdev_chain, nb);
1406 1407
	rtnl_unlock();
	return err;
L
Linus Torvalds 已提交
1408
}
E
Eric Dumazet 已提交
1409
EXPORT_SYMBOL(unregister_netdevice_notifier);
L
Linus Torvalds 已提交
1410 1411 1412 1413

/**
 *	call_netdevice_notifiers - call all network notifier blocks
 *      @val: value passed unmodified to notifier function
1414
 *      @dev: net_device pointer passed unmodified to notifier function
L
Linus Torvalds 已提交
1415 1416
 *
 *	Call all network notifier blocks.  Parameters and return value
1417
 *	are as for raw_notifier_call_chain().
L
Linus Torvalds 已提交
1418 1419
 */

1420
int call_netdevice_notifiers(unsigned long val, struct net_device *dev)
L
Linus Torvalds 已提交
1421
{
1422
	return raw_notifier_call_chain(&netdev_chain, val, dev);
L
Linus Torvalds 已提交
1423 1424 1425 1426 1427 1428 1429 1430 1431
}

/* 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);
}
E
Eric Dumazet 已提交
1432
EXPORT_SYMBOL(net_enable_timestamp);
L
Linus Torvalds 已提交
1433 1434 1435 1436 1437

void net_disable_timestamp(void)
{
	atomic_dec(&netstamp_needed);
}
E
Eric Dumazet 已提交
1438
EXPORT_SYMBOL(net_disable_timestamp);
L
Linus Torvalds 已提交
1439

1440
static inline void net_timestamp(struct sk_buff *skb)
L
Linus Torvalds 已提交
1441 1442
{
	if (atomic_read(&netstamp_needed))
1443
		__net_timestamp(skb);
1444 1445
	else
		skb->tstamp.tv64 = 0;
L
Linus Torvalds 已提交
1446 1447
}

1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475
/**
 * dev_forward_skb - loopback an skb to another netif
 *
 * @dev: destination network device
 * @skb: buffer to forward
 *
 * return values:
 *	NET_RX_SUCCESS	(no congestion)
 *	NET_RX_DROP     (packet was dropped)
 *
 * dev_forward_skb can be used for injecting an skb from the
 * start_xmit function of one device into the receive queue
 * of another device.
 *
 * The receiving device may be in another namespace, so
 * we have to clear all information in the skb that could
 * impact namespace isolation.
 */
int dev_forward_skb(struct net_device *dev, struct sk_buff *skb)
{
	skb_orphan(skb);

	if (!(dev->flags & IFF_UP))
		return NET_RX_DROP;

	if (skb->len > (dev->mtu + dev->hard_header_len))
		return NET_RX_DROP;

1476
	skb_set_dev(skb, dev);
1477 1478 1479 1480 1481 1482 1483
	skb->tstamp.tv64 = 0;
	skb->pkt_type = PACKET_HOST;
	skb->protocol = eth_type_trans(skb, dev);
	return netif_rx(skb);
}
EXPORT_SYMBOL_GPL(dev_forward_skb);

L
Linus Torvalds 已提交
1484 1485 1486 1487 1488
/*
 *	Support routine. Sends outgoing frames to any network
 *	taps currently in use.
 */

1489
static void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev)
L
Linus Torvalds 已提交
1490 1491
{
	struct packet_type *ptype;
1492

1493 1494 1495 1496
#ifdef CONFIG_NET_CLS_ACT
	if (!(skb->tstamp.tv64 && (G_TC_FROM(skb->tc_verd) & AT_INGRESS)))
		net_timestamp(skb);
#else
1497
	net_timestamp(skb);
1498
#endif
L
Linus Torvalds 已提交
1499 1500 1501 1502 1503 1504 1505 1506 1507

	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)) {
E
Eric Dumazet 已提交
1508
			struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
L
Linus Torvalds 已提交
1509 1510 1511 1512 1513 1514 1515
			if (!skb2)
				break;

			/* skb->nh should be correctly
			   set by sender, so that the second statement is
			   just protection against buggy protocols.
			 */
1516
			skb_reset_mac_header(skb2);
L
Linus Torvalds 已提交
1517

1518
			if (skb_network_header(skb2) < skb2->data ||
1519
			    skb2->network_header > skb2->tail) {
L
Linus Torvalds 已提交
1520 1521 1522 1523
				if (net_ratelimit())
					printk(KERN_CRIT "protocol %04x is "
					       "buggy, dev %s\n",
					       skb2->protocol, dev->name);
1524
				skb_reset_network_header(skb2);
L
Linus Torvalds 已提交
1525 1526
			}

1527
			skb2->transport_header = skb2->network_header;
L
Linus Torvalds 已提交
1528
			skb2->pkt_type = PACKET_OUTGOING;
D
David S. Miller 已提交
1529
			ptype->func(skb2, skb->dev, ptype, skb->dev);
L
Linus Torvalds 已提交
1530 1531 1532 1533 1534
		}
	}
	rcu_read_unlock();
}

1535

1536
static inline void __netif_reschedule(struct Qdisc *q)
1537
{
1538 1539
	struct softnet_data *sd;
	unsigned long flags;
1540

1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552
	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);
1553 1554 1555
}
EXPORT_SYMBOL(__netif_schedule);

1556
void dev_kfree_skb_irq(struct sk_buff *skb)
1557
{
1558 1559 1560
	if (atomic_dec_and_test(&skb->users)) {
		struct softnet_data *sd;
		unsigned long flags;
1561

1562 1563 1564 1565 1566 1567 1568
		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);
	}
1569
}
1570
EXPORT_SYMBOL(dev_kfree_skb_irq);
1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581

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);


1582 1583 1584 1585 1586 1587
/**
 * netif_device_detach - mark device as removed
 * @dev: network device
 *
 * Mark device as removed from system and therefore no longer available.
 */
1588 1589 1590 1591
void netif_device_detach(struct net_device *dev)
{
	if (test_and_clear_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
1592
		netif_tx_stop_all_queues(dev);
1593 1594 1595 1596
	}
}
EXPORT_SYMBOL(netif_device_detach);

1597 1598 1599 1600 1601 1602
/**
 * netif_device_attach - mark device as attached
 * @dev: network device
 *
 * Mark device as attached from system and restart if needed.
 */
1603 1604 1605 1606
void netif_device_attach(struct net_device *dev)
{
	if (!test_and_set_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
1607
		netif_tx_wake_all_queues(dev);
1608
		__netdev_watchdog_up(dev);
1609 1610 1611 1612
	}
}
EXPORT_SYMBOL(netif_device_attach);

1613 1614 1615 1616 1617 1618
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) &&
1619 1620 1621
		 protocol == htons(ETH_P_IPV6)) ||
		((features & NETIF_F_FCOE_CRC) &&
		 protocol == htons(ETH_P_FCOE)));
1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637
}

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;
}
1638

1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668
/**
 * skb_dev_set -- assign a new device to a buffer
 * @skb: buffer for the new device
 * @dev: network device
 *
 * If an skb is owned by a device already, we have to reset
 * all data private to the namespace a device belongs to
 * before assigning it a new device.
 */
#ifdef CONFIG_NET_NS
void skb_set_dev(struct sk_buff *skb, struct net_device *dev)
{
	skb_dst_drop(skb);
	if (skb->dev && !net_eq(dev_net(skb->dev), dev_net(dev))) {
		secpath_reset(skb);
		nf_reset(skb);
		skb_init_secmark(skb);
		skb->mark = 0;
		skb->priority = 0;
		skb->nf_trace = 0;
		skb->ipvs_property = 0;
#ifdef CONFIG_NET_SCHED
		skb->tc_index = 0;
#endif
	}
	skb->dev = dev;
}
EXPORT_SYMBOL(skb_set_dev);
#endif /* CONFIG_NET_NS */

L
Linus Torvalds 已提交
1669 1670 1671 1672
/*
 * Invalidate hardware checksum when packet is to be mangled, and
 * complete checksum manually on outgoing path.
 */
1673
int skb_checksum_help(struct sk_buff *skb)
L
Linus Torvalds 已提交
1674
{
1675
	__wsum csum;
1676
	int ret = 0, offset;
L
Linus Torvalds 已提交
1677

1678
	if (skb->ip_summed == CHECKSUM_COMPLETE)
1679 1680 1681 1682 1683
		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
1684 1685
	}

1686 1687 1688 1689 1690 1691 1692 1693 1694
	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
1695 1696 1697 1698 1699
		ret = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
		if (ret)
			goto out;
	}

1700
	*(__sum16 *)(skb->data + offset) = csum_fold(csum);
1701
out_set_summed:
L
Linus Torvalds 已提交
1702
	skb->ip_summed = CHECKSUM_NONE;
1703
out:
L
Linus Torvalds 已提交
1704 1705
	return ret;
}
E
Eric Dumazet 已提交
1706
EXPORT_SYMBOL(skb_checksum_help);
L
Linus Torvalds 已提交
1707

1708 1709 1710
/**
 *	skb_gso_segment - Perform segmentation on skb.
 *	@skb: buffer to segment
1711
 *	@features: features for the output path (see dev->features)
1712 1713
 *
 *	This function segments the given skb and returns a list of segments.
1714 1715 1716
 *
 *	It may return NULL if the skb requires no segmentation.  This is
 *	only possible when GSO is used for verifying header integrity.
1717
 */
1718
struct sk_buff *skb_gso_segment(struct sk_buff *skb, int features)
1719 1720 1721
{
	struct sk_buff *segs = ERR_PTR(-EPROTONOSUPPORT);
	struct packet_type *ptype;
A
Al Viro 已提交
1722
	__be16 type = skb->protocol;
1723
	int err;
1724

1725
	skb_reset_mac_header(skb);
1726
	skb->mac_len = skb->network_header - skb->mac_header;
1727 1728
	__skb_pull(skb, skb->mac_len);

1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741
	if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) {
		struct net_device *dev = skb->dev;
		struct ethtool_drvinfo info = {};

		if (dev && dev->ethtool_ops && dev->ethtool_ops->get_drvinfo)
			dev->ethtool_ops->get_drvinfo(dev, &info);

		WARN(1, "%s: caps=(0x%lx, 0x%lx) len=%d data_len=%d "
			"ip_summed=%d",
		     info.driver, dev ? dev->features : 0L,
		     skb->sk ? skb->sk->sk_route_caps : 0L,
		     skb->len, skb->data_len, skb->ip_summed);

1742 1743 1744 1745 1746
		if (skb_header_cloned(skb) &&
		    (err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
			return ERR_PTR(err);
	}

1747
	rcu_read_lock();
1748 1749
	list_for_each_entry_rcu(ptype,
			&ptype_base[ntohs(type) & PTYPE_HASH_MASK], list) {
1750
		if (ptype->type == type && !ptype->dev && ptype->gso_segment) {
1751
			if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) {
1752 1753 1754 1755
				err = ptype->gso_send_check(skb);
				segs = ERR_PTR(err);
				if (err || skb_gso_ok(skb, features))
					break;
1756 1757
				__skb_push(skb, (skb->data -
						 skb_network_header(skb)));
1758
			}
1759
			segs = ptype->gso_segment(skb, features);
1760 1761 1762 1763 1764
			break;
		}
	}
	rcu_read_unlock();

1765
	__skb_push(skb, skb->data - skb_mac_header(skb));
1766

1767 1768 1769 1770
	return segs;
}
EXPORT_SYMBOL(skb_gso_segment);

1771 1772 1773 1774 1775
/* Take action when hardware reception checksum errors are detected. */
#ifdef CONFIG_BUG
void netdev_rx_csum_fault(struct net_device *dev)
{
	if (net_ratelimit()) {
1776
		printk(KERN_ERR "%s: hw csum failure.\n",
1777
			dev ? dev->name : "<unknown>");
1778 1779 1780 1781 1782 1783
		dump_stack();
	}
}
EXPORT_SYMBOL(netdev_rx_csum_fault);
#endif

L
Linus Torvalds 已提交
1784 1785 1786 1787 1788 1789 1790
/* 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)
{
1791
#ifdef CONFIG_HIGHMEM
L
Linus Torvalds 已提交
1792 1793 1794 1795 1796 1797 1798 1799 1800
	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;

1801
#endif
L
Linus Torvalds 已提交
1802 1803 1804
	return 0;
}

1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838
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;
1839 1840 1841 1842 1843 1844 1845 1846
	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;
1847

1848
	if (IS_ERR(segs))
1849 1850 1851 1852 1853 1854 1855 1856 1857
		return PTR_ERR(segs);

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

	return 0;
}

1858 1859
int dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
			struct netdev_queue *txq)
1860
{
1861
	const struct net_device_ops *ops = dev->netdev_ops;
1862
	int rc = NETDEV_TX_OK;
1863

1864
	if (likely(!skb->next)) {
S
Stephen Hemminger 已提交
1865
		if (!list_empty(&ptype_all))
1866 1867
			dev_queue_xmit_nit(skb, dev);

1868 1869 1870 1871 1872 1873
		if (netif_needs_gso(dev, skb)) {
			if (unlikely(dev_gso_segment(skb)))
				goto out_kfree_skb;
			if (skb->next)
				goto gso;
		}
1874

1875 1876 1877 1878
		/*
		 * If device doesnt need skb->dst, release it right now while
		 * its hot in this cpu cache
		 */
E
Eric Dumazet 已提交
1879 1880 1881
		if (dev->priv_flags & IFF_XMIT_DST_RELEASE)
			skb_dst_drop(skb);

1882
		rc = ops->ndo_start_xmit(skb, dev);
1883
		if (rc == NETDEV_TX_OK)
E
Eric Dumazet 已提交
1884
			txq_trans_update(txq);
1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899
		/*
		 * TODO: if skb_orphan() was called by
		 * dev->hard_start_xmit() (for example, the unmodified
		 * igb driver does that; bnx2 doesn't), then
		 * skb_tx_software_timestamp() will be unable to send
		 * back the time stamp.
		 *
		 * How can this be prevented? Always create another
		 * reference to the socket before calling
		 * dev->hard_start_xmit()? Prevent that skb_orphan()
		 * does anything in dev->hard_start_xmit() by clearing
		 * the skb destructor before the call and restoring it
		 * afterwards, then doing the skb_orphan() ourselves?
		 */
		return rc;
1900 1901
	}

1902
gso:
1903 1904 1905 1906 1907
	do {
		struct sk_buff *nskb = skb->next;

		skb->next = nskb->next;
		nskb->next = NULL;
1908 1909 1910 1911 1912 1913 1914 1915

		/*
		 * If device doesnt need nskb->dst, release it right now while
		 * its hot in this cpu cache
		 */
		if (dev->priv_flags & IFF_XMIT_DST_RELEASE)
			skb_dst_drop(nskb);

1916
		rc = ops->ndo_start_xmit(nskb, dev);
1917
		if (unlikely(rc != NETDEV_TX_OK)) {
1918 1919
			if (rc & ~NETDEV_TX_MASK)
				goto out_kfree_gso_skb;
1920
			nskb->next = skb->next;
1921 1922 1923
			skb->next = nskb;
			return rc;
		}
E
Eric Dumazet 已提交
1924
		txq_trans_update(txq);
1925
		if (unlikely(netif_tx_queue_stopped(txq) && skb->next))
1926
			return NETDEV_TX_BUSY;
1927
	} while (skb->next);
1928

1929 1930 1931
out_kfree_gso_skb:
	if (likely(skb->next == NULL))
		skb->destructor = DEV_GSO_CB(skb)->destructor;
1932 1933
out_kfree_skb:
	kfree_skb(skb);
1934
	return rc;
1935 1936
}

T
Tom Herbert 已提交
1937
static u32 hashrnd __read_mostly;
D
David S. Miller 已提交
1938

1939
u16 skb_tx_hash(const struct net_device *dev, const struct sk_buff *skb)
1940
{
1941
	u32 hash;
D
David S. Miller 已提交
1942

1943 1944
	if (skb_rx_queue_recorded(skb)) {
		hash = skb_get_rx_queue(skb);
E
Eric Dumazet 已提交
1945
		while (unlikely(hash >= dev->real_num_tx_queues))
1946 1947 1948
			hash -= dev->real_num_tx_queues;
		return hash;
	}
1949 1950

	if (skb->sk && skb->sk->sk_hash)
1951
		hash = skb->sk->sk_hash;
1952
	else
1953
		hash = skb->protocol;
1954

T
Tom Herbert 已提交
1955
	hash = jhash_1word(hash, hashrnd);
D
David S. Miller 已提交
1956 1957

	return (u16) (((u64) hash * dev->real_num_tx_queues) >> 32);
1958
}
1959
EXPORT_SYMBOL(skb_tx_hash);
1960

1961 1962 1963 1964
static inline u16 dev_cap_txqueue(struct net_device *dev, u16 queue_index)
{
	if (unlikely(queue_index >= dev->real_num_tx_queues)) {
		if (net_ratelimit()) {
T
Tom Herbert 已提交
1965
			netdev_warn(dev, "selects TX queue %d, but "
1966
			     "real number of TX queues is %d\n",
T
Tom Herbert 已提交
1967
			     queue_index, dev->real_num_tx_queues);
1968 1969 1970 1971 1972 1973
		}
		return 0;
	}
	return queue_index;
}

1974 1975 1976
static struct netdev_queue *dev_pick_tx(struct net_device *dev,
					struct sk_buff *skb)
{
1977 1978 1979 1980 1981 1982 1983 1984 1985 1986
	u16 queue_index;
	struct sock *sk = skb->sk;

	if (sk_tx_queue_recorded(sk)) {
		queue_index = sk_tx_queue_get(sk);
	} else {
		const struct net_device_ops *ops = dev->netdev_ops;

		if (ops->ndo_select_queue) {
			queue_index = ops->ndo_select_queue(dev, skb);
1987
			queue_index = dev_cap_txqueue(dev, queue_index);
1988 1989 1990 1991
		} else {
			queue_index = 0;
			if (dev->real_num_tx_queues > 1)
				queue_index = skb_tx_hash(dev, skb);
1992

1993 1994 1995 1996
			if (sk && sk->sk_dst_cache)
				sk_tx_queue_set(sk, queue_index);
		}
	}
1997

1998 1999
	skb_set_queue_mapping(skb, queue_index);
	return netdev_get_tx_queue(dev, queue_index);
2000 2001
}

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035
static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q,
				 struct net_device *dev,
				 struct netdev_queue *txq)
{
	spinlock_t *root_lock = qdisc_lock(q);
	int rc;

	spin_lock(root_lock);
	if (unlikely(test_bit(__QDISC_STATE_DEACTIVATED, &q->state))) {
		kfree_skb(skb);
		rc = NET_XMIT_DROP;
	} else if ((q->flags & TCQ_F_CAN_BYPASS) && !qdisc_qlen(q) &&
		   !test_and_set_bit(__QDISC_STATE_RUNNING, &q->state)) {
		/*
		 * This is a work-conserving queue; there are no old skbs
		 * waiting to be sent out; and the qdisc is not running -
		 * xmit the skb directly.
		 */
		__qdisc_update_bstats(q, skb->len);
		if (sch_direct_xmit(skb, q, dev, txq, root_lock))
			__qdisc_run(q);
		else
			clear_bit(__QDISC_STATE_RUNNING, &q->state);

		rc = NET_XMIT_SUCCESS;
	} else {
		rc = qdisc_enqueue_root(skb, q);
		qdisc_run(q);
	}
	spin_unlock(root_lock);

	return rc;
}

2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050
/*
 * Returns true if either:
 *	1. skb has frag_list and the device doesn't support FRAGLIST, or
 *	2. skb is fragmented and the device does not support SG, or if
 *	   at least one of fragments is in highmem and device does not
 *	   support DMA from it.
 */
static inline int skb_needs_linearize(struct sk_buff *skb,
				      struct net_device *dev)
{
	return (skb_has_frags(skb) && !(dev->features & NETIF_F_FRAGLIST)) ||
	       (skb_shinfo(skb)->nr_frags && (!(dev->features & NETIF_F_SG) ||
					      illegal_highdma(dev, skb)));
}

2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075
/**
 *	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
2076 2077 2078
int dev_queue_xmit(struct sk_buff *skb)
{
	struct net_device *dev = skb->dev;
2079
	struct netdev_queue *txq;
L
Linus Torvalds 已提交
2080 2081 2082
	struct Qdisc *q;
	int rc = -ENOMEM;

2083 2084 2085 2086
	/* GSO will handle the following emulations directly. */
	if (netif_needs_gso(dev, skb))
		goto gso;

2087 2088
	/* Convert a paged skb to linear, if required */
	if (skb_needs_linearize(skb, dev) && __skb_linearize(skb))
L
Linus Torvalds 已提交
2089 2090 2091 2092 2093
		goto out_kfree_skb;

	/* If packet is not checksummed and device does not support
	 * checksumming for this protocol, complete checksumming here.
	 */
2094 2095 2096
	if (skb->ip_summed == CHECKSUM_PARTIAL) {
		skb_set_transport_header(skb, skb->csum_start -
					      skb_headroom(skb));
2097 2098
		if (!dev_can_checksum(dev, skb) && skb_checksum_help(skb))
			goto out_kfree_skb;
2099
	}
L
Linus Torvalds 已提交
2100

2101
gso:
2102 2103
	/* Disable soft irqs for various locks below. Also
	 * stops preemption for RCU.
L
Linus Torvalds 已提交
2104
	 */
2105
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
2106

2107
	txq = dev_pick_tx(dev, skb);
2108
	q = rcu_dereference_bh(txq->qdisc);
2109

L
Linus Torvalds 已提交
2110
#ifdef CONFIG_NET_CLS_ACT
E
Eric Dumazet 已提交
2111
	skb->tc_verd = SET_TC_AT(skb->tc_verd, AT_EGRESS);
L
Linus Torvalds 已提交
2112 2113
#endif
	if (q->enqueue) {
2114
		rc = __dev_xmit_skb(skb, q, dev, txq);
2115
		goto out;
L
Linus Torvalds 已提交
2116 2117 2118 2119 2120
	}

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

H
Herbert Xu 已提交
2121 2122
	   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
2123 2124 2125 2126 2127 2128 2129 2130 2131 2132
	   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 */

2133
		if (txq->xmit_lock_owner != cpu) {
L
Linus Torvalds 已提交
2134

2135
			HARD_TX_LOCK(dev, txq, cpu);
L
Linus Torvalds 已提交
2136

2137
			if (!netif_tx_queue_stopped(txq)) {
2138 2139
				rc = dev_hard_start_xmit(skb, dev, txq);
				if (dev_xmit_complete(rc)) {
2140
					HARD_TX_UNLOCK(dev, txq);
L
Linus Torvalds 已提交
2141 2142 2143
					goto out;
				}
			}
2144
			HARD_TX_UNLOCK(dev, txq);
L
Linus Torvalds 已提交
2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157
			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;
2158
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
2159 2160 2161 2162 2163

out_kfree_skb:
	kfree_skb(skb);
	return rc;
out:
2164
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
2165 2166
	return rc;
}
E
Eric Dumazet 已提交
2167
EXPORT_SYMBOL(dev_queue_xmit);
L
Linus Torvalds 已提交
2168 2169 2170 2171 2172 2173


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

2174 2175 2176
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
2177 2178 2179

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

2180
#ifdef CONFIG_SMP
T
Tom Herbert 已提交
2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299
/*
 * get_rps_cpu is called from netif_receive_skb and returns the target
 * CPU from the RPS map of the receiving queue for a given skb.
 */
static int get_rps_cpu(struct net_device *dev, struct sk_buff *skb)
{
	struct ipv6hdr *ip6;
	struct iphdr *ip;
	struct netdev_rx_queue *rxqueue;
	struct rps_map *map;
	int cpu = -1;
	u8 ip_proto;
	u32 addr1, addr2, ports, ihl;

	rcu_read_lock();

	if (skb_rx_queue_recorded(skb)) {
		u16 index = skb_get_rx_queue(skb);
		if (unlikely(index >= dev->num_rx_queues)) {
			if (net_ratelimit()) {
				netdev_warn(dev, "received packet on queue "
				    "%u, but number of RX queues is %u\n",
				     index, dev->num_rx_queues);
			}
			goto done;
		}
		rxqueue = dev->_rx + index;
	} else
		rxqueue = dev->_rx;

	if (!rxqueue->rps_map)
		goto done;

	if (skb->rxhash)
		goto got_hash; /* Skip hash computation on packet header */

	switch (skb->protocol) {
	case __constant_htons(ETH_P_IP):
		if (!pskb_may_pull(skb, sizeof(*ip)))
			goto done;

		ip = (struct iphdr *) skb->data;
		ip_proto = ip->protocol;
		addr1 = ip->saddr;
		addr2 = ip->daddr;
		ihl = ip->ihl;
		break;
	case __constant_htons(ETH_P_IPV6):
		if (!pskb_may_pull(skb, sizeof(*ip6)))
			goto done;

		ip6 = (struct ipv6hdr *) skb->data;
		ip_proto = ip6->nexthdr;
		addr1 = ip6->saddr.s6_addr32[3];
		addr2 = ip6->daddr.s6_addr32[3];
		ihl = (40 >> 2);
		break;
	default:
		goto done;
	}
	ports = 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:
		if (pskb_may_pull(skb, (ihl * 4) + 4))
			ports = *((u32 *) (skb->data + (ihl * 4)));
		break;

	default:
		break;
	}

	skb->rxhash = jhash_3words(addr1, addr2, ports, hashrnd);
	if (!skb->rxhash)
		skb->rxhash = 1;

got_hash:
	map = rcu_dereference(rxqueue->rps_map);
	if (map) {
		u16 tcpu = map->cpus[((u64) skb->rxhash * map->len) >> 32];

		if (cpu_online(tcpu)) {
			cpu = tcpu;
			goto done;
		}
	}

done:
	rcu_read_unlock();
	return cpu;
}

/*
 * This structure holds the per-CPU mask of CPUs for which IPIs are scheduled
 * to be sent to kick remote softirq processing.  There are two masks since
 * the sending of IPIs must be done with interrupts enabled.  The select field
 * indicates the current mask that enqueue_backlog uses to schedule IPIs.
 * select is flipped before net_rps_action is called while still under lock,
 * net_rps_action then uses the non-selected mask to send the IPIs and clears
 * it without conflicting with enqueue_backlog operation.
 */
struct rps_remote_softirq_cpus {
	cpumask_t mask[2];
	int select;
};
static DEFINE_PER_CPU(struct rps_remote_softirq_cpus, rps_remote_softirq_cpus);

/* Called from hardirq (IPI) context */
static void trigger_softirq(void *data)
{
	struct softnet_data *queue = data;
	__napi_schedule(&queue->backlog);
	__get_cpu_var(netdev_rx_stat).received_rps++;
}
2300
#endif /* CONFIG_SMP */
T
Tom Herbert 已提交
2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327

/*
 * enqueue_to_backlog is called to queue an skb to a per CPU backlog
 * queue (may be a remote CPU queue).
 */
static int enqueue_to_backlog(struct sk_buff *skb, int cpu)
{
	struct softnet_data *queue;
	unsigned long flags;

	queue = &per_cpu(softnet_data, cpu);

	local_irq_save(flags);
	__get_cpu_var(netdev_rx_stat).total++;

	spin_lock(&queue->input_pkt_queue.lock);
	if (queue->input_pkt_queue.qlen <= netdev_max_backlog) {
		if (queue->input_pkt_queue.qlen) {
enqueue:
			__skb_queue_tail(&queue->input_pkt_queue, skb);
			spin_unlock_irqrestore(&queue->input_pkt_queue.lock,
			    flags);
			return NET_RX_SUCCESS;
		}

		/* Schedule NAPI for backlog device */
		if (napi_schedule_prep(&queue->backlog)) {
2328
#ifdef CONFIG_SMP
T
Tom Herbert 已提交
2329 2330 2331 2332 2333 2334 2335 2336
			if (cpu != smp_processor_id()) {
				struct rps_remote_softirq_cpus *rcpus =
				    &__get_cpu_var(rps_remote_softirq_cpus);

				cpu_set(cpu, rcpus->mask[rcpus->select]);
				__raise_softirq_irqoff(NET_RX_SOFTIRQ);
			} else
				__napi_schedule(&queue->backlog);
2337 2338 2339
#else
			__napi_schedule(&queue->backlog);
#endif
T
Tom Herbert 已提交
2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351
		}
		goto enqueue;
	}

	spin_unlock(&queue->input_pkt_queue.lock);

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

	kfree_skb(skb);
	return NET_RX_DROP;
}
L
Linus Torvalds 已提交
2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369

/**
 *	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)
{
T
Tom Herbert 已提交
2370
	int cpu;
L
Linus Torvalds 已提交
2371 2372 2373 2374 2375

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

2376
	if (!skb->tstamp.tv64)
2377
		net_timestamp(skb);
L
Linus Torvalds 已提交
2378

2379
#ifdef CONFIG_SMP
T
Tom Herbert 已提交
2380 2381 2382
	cpu = get_rps_cpu(skb->dev, skb);
	if (cpu < 0)
		cpu = smp_processor_id();
2383 2384 2385
#else
	cpu = smp_processor_id();
#endif
L
Linus Torvalds 已提交
2386

T
Tom Herbert 已提交
2387
	return enqueue_to_backlog(skb, cpu);
L
Linus Torvalds 已提交
2388
}
E
Eric Dumazet 已提交
2389
EXPORT_SYMBOL(netif_rx);
L
Linus Torvalds 已提交
2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420

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;

2421
			WARN_ON(atomic_read(&skb->users));
L
Linus Torvalds 已提交
2422 2423 2424 2425 2426
			__kfree_skb(skb);
		}
	}

	if (sd->output_queue) {
2427
		struct Qdisc *head;
L
Linus Torvalds 已提交
2428 2429 2430 2431 2432 2433 2434

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

		while (head) {
2435 2436 2437
			struct Qdisc *q = head;
			spinlock_t *root_lock;

L
Linus Torvalds 已提交
2438 2439
			head = head->next_sched;

2440
			root_lock = qdisc_lock(q);
2441
			if (spin_trylock(root_lock)) {
2442 2443 2444
				smp_mb__before_clear_bit();
				clear_bit(__QDISC_STATE_SCHED,
					  &q->state);
2445 2446
				qdisc_run(q);
				spin_unlock(root_lock);
L
Linus Torvalds 已提交
2447
			} else {
2448
				if (!test_bit(__QDISC_STATE_DEACTIVATED,
2449
					      &q->state)) {
2450
					__netif_reschedule(q);
2451 2452 2453 2454 2455
				} else {
					smp_mb__before_clear_bit();
					clear_bit(__QDISC_STATE_SCHED,
						  &q->state);
				}
L
Linus Torvalds 已提交
2456 2457 2458 2459 2460
			}
		}
	}
}

2461 2462 2463
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
2464 2465
{
	atomic_inc(&skb->users);
D
David S. Miller 已提交
2466
	return pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2467 2468 2469
}

#if defined(CONFIG_BRIDGE) || defined (CONFIG_BRIDGE_MODULE)
2470 2471 2472 2473 2474

#if defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE)
/* This hook is defined here for ATM LANE */
int (*br_fdb_test_addr_hook)(struct net_device *dev,
			     unsigned char *addr) __read_mostly;
2475
EXPORT_SYMBOL_GPL(br_fdb_test_addr_hook);
2476
#endif
L
Linus Torvalds 已提交
2477

2478 2479 2480 2481 2482 2483
/*
 * 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;
2484
EXPORT_SYMBOL_GPL(br_handle_frame_hook);
2485

2486 2487 2488
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
2489 2490 2491
{
	struct net_bridge_port *port;

2492 2493 2494
	if (skb->pkt_type == PACKET_LOOPBACK ||
	    (port = rcu_dereference(skb->dev->br_port)) == NULL)
		return skb;
L
Linus Torvalds 已提交
2495 2496

	if (*pt_prev) {
2497
		*ret = deliver_skb(skb, *pt_prev, orig_dev);
L
Linus Torvalds 已提交
2498
		*pt_prev = NULL;
2499 2500
	}

2501
	return br_handle_frame_hook(port, skb);
L
Linus Torvalds 已提交
2502 2503
}
#else
2504
#define handle_bridge(skb, pt_prev, ret, orig_dev)	(skb)
L
Linus Torvalds 已提交
2505 2506
#endif

P
Patrick McHardy 已提交
2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528
#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
2529 2530 2531 2532 2533
#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
2534
 * NOTE: This doesnt stop any functionality; if you dont have
L
Linus Torvalds 已提交
2535 2536 2537
 * the ingress scheduler, you just cant add policies on ingress.
 *
 */
2538
static int ing_filter(struct sk_buff *skb)
L
Linus Torvalds 已提交
2539 2540
{
	struct net_device *dev = skb->dev;
2541
	u32 ttl = G_TC_RTTL(skb->tc_verd);
2542 2543 2544
	struct netdev_queue *rxq;
	int result = TC_ACT_OK;
	struct Qdisc *q;
2545

2546 2547 2548
	if (MAX_RED_LOOP < ttl++) {
		printk(KERN_WARNING
		       "Redir loop detected Dropping packet (%d->%d)\n",
2549
		       skb->skb_iif, dev->ifindex);
2550 2551
		return TC_ACT_SHOT;
	}
L
Linus Torvalds 已提交
2552

2553 2554
	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
2555

2556 2557
	rxq = &dev->rx_queue;

2558
	q = rxq->qdisc;
2559
	if (q != &noop_qdisc) {
2560
		spin_lock(qdisc_lock(q));
2561 2562
		if (likely(!test_bit(__QDISC_STATE_DEACTIVATED, &q->state)))
			result = qdisc_enqueue_root(skb, q);
2563 2564
		spin_unlock(qdisc_lock(q));
	}
2565 2566 2567

	return result;
}
2568

2569 2570 2571 2572
static inline struct sk_buff *handle_ing(struct sk_buff *skb,
					 struct packet_type **pt_prev,
					 int *ret, struct net_device *orig_dev)
{
2573
	if (skb->dev->rx_queue.qdisc == &noop_qdisc)
2574
		goto out;
L
Linus Torvalds 已提交
2575

2576 2577 2578 2579 2580 2581
	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
2582 2583
	}

2584 2585 2586 2587 2588 2589 2590 2591 2592 2593
	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
2594 2595 2596
}
#endif

2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623
/*
 * 	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();
}

T
Tom Herbert 已提交
2624
int __netif_receive_skb(struct sk_buff *skb)
L
Linus Torvalds 已提交
2625 2626
{
	struct packet_type *ptype, *pt_prev;
D
David S. Miller 已提交
2627
	struct net_device *orig_dev;
2628
	struct net_device *master;
2629
	struct net_device *null_or_orig;
2630
	struct net_device *null_or_bond;
L
Linus Torvalds 已提交
2631
	int ret = NET_RX_DROP;
A
Al Viro 已提交
2632
	__be16 type;
L
Linus Torvalds 已提交
2633

2634 2635 2636
	if (!skb->tstamp.tv64)
		net_timestamp(skb);

E
Eric Dumazet 已提交
2637
	if (vlan_tx_tag_present(skb) && vlan_hwaccel_do_receive(skb))
2638 2639
		return NET_RX_SUCCESS;

L
Linus Torvalds 已提交
2640
	/* if we've gotten here through NAPI, check netpoll */
2641
	if (netpoll_receive_skb(skb))
L
Linus Torvalds 已提交
2642 2643
		return NET_RX_DROP;

2644 2645
	if (!skb->skb_iif)
		skb->skb_iif = skb->dev->ifindex;
2646

2647
	null_or_orig = NULL;
J
Joe Eykholt 已提交
2648
	orig_dev = skb->dev;
2649 2650 2651
	master = ACCESS_ONCE(orig_dev->master);
	if (master) {
		if (skb_bond_should_drop(skb, master))
2652 2653
			null_or_orig = orig_dev; /* deliver only exact match */
		else
2654
			skb->dev = master;
J
Joe Eykholt 已提交
2655
	}
2656

L
Linus Torvalds 已提交
2657 2658
	__get_cpu_var(netdev_rx_stat).total++;

2659
	skb_reset_network_header(skb);
2660
	skb_reset_transport_header(skb);
2661
	skb->mac_len = skb->network_header - skb->mac_header;
L
Linus Torvalds 已提交
2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674

	pt_prev = NULL;

	rcu_read_lock();

#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) {
2675 2676
		if (ptype->dev == null_or_orig || ptype->dev == skb->dev ||
		    ptype->dev == orig_dev) {
2677
			if (pt_prev)
D
David S. Miller 已提交
2678
				ret = deliver_skb(skb, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2679 2680 2681 2682 2683
			pt_prev = ptype;
		}
	}

#ifdef CONFIG_NET_CLS_ACT
2684 2685
	skb = handle_ing(skb, &pt_prev, &ret, orig_dev);
	if (!skb)
L
Linus Torvalds 已提交
2686 2687 2688 2689
		goto out;
ncls:
#endif

2690
	skb = handle_bridge(skb, &pt_prev, &ret, orig_dev);
P
Patrick McHardy 已提交
2691 2692 2693
	if (!skb)
		goto out;
	skb = handle_macvlan(skb, &pt_prev, &ret, orig_dev);
2694
	if (!skb)
L
Linus Torvalds 已提交
2695 2696
		goto out;

2697 2698 2699 2700 2701 2702
	/*
	 * Make sure frames received on VLAN interfaces stacked on
	 * bonding interfaces still make their way to any base bonding
	 * device that may have registered for a specific ptype.  The
	 * handler may have to adjust skb->dev and orig_dev.
	 */
2703
	null_or_bond = NULL;
2704 2705
	if ((skb->dev->priv_flags & IFF_802_1Q_VLAN) &&
	    (vlan_dev_real_dev(skb->dev)->priv_flags & IFF_BONDING)) {
2706
		null_or_bond = vlan_dev_real_dev(skb->dev);
2707 2708
	}

L
Linus Torvalds 已提交
2709
	type = skb->protocol;
2710 2711
	list_for_each_entry_rcu(ptype,
			&ptype_base[ntohs(type) & PTYPE_HASH_MASK], list) {
2712
		if (ptype->type == type && (ptype->dev == null_or_orig ||
2713 2714
		     ptype->dev == skb->dev || ptype->dev == orig_dev ||
		     ptype->dev == null_or_bond)) {
2715
			if (pt_prev)
D
David S. Miller 已提交
2716
				ret = deliver_skb(skb, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2717 2718 2719 2720 2721
			pt_prev = ptype;
		}
	}

	if (pt_prev) {
D
David S. Miller 已提交
2722
		ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734
	} 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;
}
T
Tom Herbert 已提交
2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752

/**
 *	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
 */
int netif_receive_skb(struct sk_buff *skb)
{
2753
#ifdef CONFIG_SMP
T
Tom Herbert 已提交
2754 2755 2756 2757 2758 2759 2760 2761
	int cpu;

	cpu = get_rps_cpu(skb->dev, skb);

	if (cpu < 0)
		return __netif_receive_skb(skb);
	else
		return enqueue_to_backlog(skb, cpu);
2762 2763 2764
#else
	return __netif_receive_skb(skb);
#endif
T
Tom Herbert 已提交
2765
}
E
Eric Dumazet 已提交
2766
EXPORT_SYMBOL(netif_receive_skb);
L
Linus Torvalds 已提交
2767

2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781
/* 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);
		}
}

2782 2783 2784 2785 2786 2787 2788
static int napi_gro_complete(struct sk_buff *skb)
{
	struct packet_type *ptype;
	__be16 type = skb->protocol;
	struct list_head *head = &ptype_base[ntohs(type) & PTYPE_HASH_MASK];
	int err = -ENOENT;

2789 2790
	if (NAPI_GRO_CB(skb)->count == 1) {
		skb_shinfo(skb)->gso_size = 0;
2791
		goto out;
2792
	}
2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813

	rcu_read_lock();
	list_for_each_entry_rcu(ptype, head, list) {
		if (ptype->type != type || ptype->dev || !ptype->gro_complete)
			continue;

		err = ptype->gro_complete(skb);
		break;
	}
	rcu_read_unlock();

	if (err) {
		WARN_ON(&ptype->list == head);
		kfree_skb(skb);
		return NET_RX_SUCCESS;
	}

out:
	return netif_receive_skb(skb);
}

2814
static void napi_gro_flush(struct napi_struct *napi)
2815 2816 2817 2818 2819 2820 2821 2822 2823
{
	struct sk_buff *skb, *next;

	for (skb = napi->gro_list; skb; skb = next) {
		next = skb->next;
		skb->next = NULL;
		napi_gro_complete(skb);
	}

2824
	napi->gro_count = 0;
2825 2826 2827
	napi->gro_list = NULL;
}

2828
enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
2829 2830 2831 2832 2833
{
	struct sk_buff **pp = NULL;
	struct packet_type *ptype;
	__be16 type = skb->protocol;
	struct list_head *head = &ptype_base[ntohs(type) & PTYPE_HASH_MASK];
H
Herbert Xu 已提交
2834
	int same_flow;
2835
	int mac_len;
2836
	enum gro_result ret;
2837 2838 2839 2840

	if (!(skb->dev->features & NETIF_F_GRO))
		goto normal;

2841
	if (skb_is_gso(skb) || skb_has_frags(skb))
2842 2843
		goto normal;

2844 2845 2846 2847 2848
	rcu_read_lock();
	list_for_each_entry_rcu(ptype, head, list) {
		if (ptype->type != type || ptype->dev || !ptype->gro_receive)
			continue;

2849
		skb_set_network_header(skb, skb_gro_offset(skb));
2850 2851 2852 2853
		mac_len = skb->network_header - skb->mac_header;
		skb->mac_len = mac_len;
		NAPI_GRO_CB(skb)->same_flow = 0;
		NAPI_GRO_CB(skb)->flush = 0;
H
Herbert Xu 已提交
2854
		NAPI_GRO_CB(skb)->free = 0;
2855 2856 2857 2858 2859 2860 2861 2862 2863

		pp = ptype->gro_receive(&napi->gro_list, skb);
		break;
	}
	rcu_read_unlock();

	if (&ptype->list == head)
		goto normal;

H
Herbert Xu 已提交
2864
	same_flow = NAPI_GRO_CB(skb)->same_flow;
2865
	ret = NAPI_GRO_CB(skb)->free ? GRO_MERGED_FREE : GRO_MERGED;
H
Herbert Xu 已提交
2866

2867 2868 2869 2870 2871 2872
	if (pp) {
		struct sk_buff *nskb = *pp;

		*pp = nskb->next;
		nskb->next = NULL;
		napi_gro_complete(nskb);
2873
		napi->gro_count--;
2874 2875
	}

H
Herbert Xu 已提交
2876
	if (same_flow)
2877 2878
		goto ok;

2879
	if (NAPI_GRO_CB(skb)->flush || napi->gro_count >= MAX_GRO_SKBS)
2880 2881
		goto normal;

2882
	napi->gro_count++;
2883
	NAPI_GRO_CB(skb)->count = 1;
2884
	skb_shinfo(skb)->gso_size = skb_gro_len(skb);
2885 2886
	skb->next = napi->gro_list;
	napi->gro_list = skb;
2887
	ret = GRO_HELD;
2888

2889
pull:
H
Herbert Xu 已提交
2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908
	if (skb_headlen(skb) < skb_gro_offset(skb)) {
		int grow = skb_gro_offset(skb) - skb_headlen(skb);

		BUG_ON(skb->end - skb->tail < grow);

		memcpy(skb_tail_pointer(skb), NAPI_GRO_CB(skb)->frag0, grow);

		skb->tail += grow;
		skb->data_len -= grow;

		skb_shinfo(skb)->frags[0].page_offset += grow;
		skb_shinfo(skb)->frags[0].size -= grow;

		if (unlikely(!skb_shinfo(skb)->frags[0].size)) {
			put_page(skb_shinfo(skb)->frags[0].page);
			memmove(skb_shinfo(skb)->frags,
				skb_shinfo(skb)->frags + 1,
				--skb_shinfo(skb)->nr_frags);
		}
2909 2910
	}

2911
ok:
2912
	return ret;
2913 2914

normal:
2915 2916
	ret = GRO_NORMAL;
	goto pull;
H
Herbert Xu 已提交
2917
}
2918 2919
EXPORT_SYMBOL(dev_gro_receive);

2920 2921
static gro_result_t
__napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
2922 2923 2924
{
	struct sk_buff *p;

2925 2926 2927
	if (netpoll_rx_on(skb))
		return GRO_NORMAL;

2928
	for (p = napi->gro_list; p; p = p->next) {
2929 2930 2931 2932
		NAPI_GRO_CB(p)->same_flow =
			(p->dev == skb->dev) &&
			!compare_ether_header(skb_mac_header(p),
					      skb_gro_mac_header(skb));
2933 2934 2935 2936 2937
		NAPI_GRO_CB(p)->flush = 0;
	}

	return dev_gro_receive(napi, skb);
}
H
Herbert Xu 已提交
2938

2939
gro_result_t napi_skb_finish(gro_result_t ret, struct sk_buff *skb)
H
Herbert Xu 已提交
2940
{
2941 2942
	switch (ret) {
	case GRO_NORMAL:
2943 2944 2945
		if (netif_receive_skb(skb))
			ret = GRO_DROP;
		break;
H
Herbert Xu 已提交
2946

2947 2948
	case GRO_DROP:
	case GRO_MERGED_FREE:
H
Herbert Xu 已提交
2949 2950
		kfree_skb(skb);
		break;
2951 2952 2953 2954

	case GRO_HELD:
	case GRO_MERGED:
		break;
H
Herbert Xu 已提交
2955 2956
	}

2957
	return ret;
2958 2959 2960
}
EXPORT_SYMBOL(napi_skb_finish);

2961 2962 2963 2964
void skb_gro_reset_offset(struct sk_buff *skb)
{
	NAPI_GRO_CB(skb)->data_offset = 0;
	NAPI_GRO_CB(skb)->frag0 = NULL;
2965
	NAPI_GRO_CB(skb)->frag0_len = 0;
2966

2967
	if (skb->mac_header == skb->tail &&
2968
	    !PageHighMem(skb_shinfo(skb)->frags[0].page)) {
2969 2970 2971
		NAPI_GRO_CB(skb)->frag0 =
			page_address(skb_shinfo(skb)->frags[0].page) +
			skb_shinfo(skb)->frags[0].page_offset;
2972 2973
		NAPI_GRO_CB(skb)->frag0_len = skb_shinfo(skb)->frags[0].size;
	}
2974 2975 2976
}
EXPORT_SYMBOL(skb_gro_reset_offset);

2977
gro_result_t napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
2978
{
2979 2980
	skb_gro_reset_offset(skb);

2981
	return napi_skb_finish(__napi_gro_receive(napi, skb), skb);
2982 2983 2984
}
EXPORT_SYMBOL(napi_gro_receive);

2985 2986 2987 2988 2989 2990 2991 2992 2993
void napi_reuse_skb(struct napi_struct *napi, struct sk_buff *skb)
{
	__skb_pull(skb, skb_headlen(skb));
	skb_reserve(skb, NET_IP_ALIGN - skb_headroom(skb));

	napi->skb = skb;
}
EXPORT_SYMBOL(napi_reuse_skb);

2994
struct sk_buff *napi_get_frags(struct napi_struct *napi)
H
Herbert Xu 已提交
2995 2996 2997 2998
{
	struct sk_buff *skb = napi->skb;

	if (!skb) {
2999 3000 3001
		skb = netdev_alloc_skb_ip_align(napi->dev, GRO_MAX_HEAD);
		if (skb)
			napi->skb = skb;
3002
	}
3003 3004
	return skb;
}
3005
EXPORT_SYMBOL(napi_get_frags);
3006

3007 3008
gro_result_t napi_frags_finish(struct napi_struct *napi, struct sk_buff *skb,
			       gro_result_t ret)
3009
{
3010 3011
	switch (ret) {
	case GRO_NORMAL:
3012
	case GRO_HELD:
3013
		skb->protocol = eth_type_trans(skb, skb->dev);
3014

3015 3016 3017 3018
		if (ret == GRO_HELD)
			skb_gro_pull(skb, -ETH_HLEN);
		else if (netif_receive_skb(skb))
			ret = GRO_DROP;
3019
		break;
H
Herbert Xu 已提交
3020

3021 3022 3023 3024
	case GRO_DROP:
	case GRO_MERGED_FREE:
		napi_reuse_skb(napi, skb);
		break;
3025 3026 3027

	case GRO_MERGED:
		break;
3028
	}
H
Herbert Xu 已提交
3029

3030
	return ret;
H
Herbert Xu 已提交
3031
}
3032 3033
EXPORT_SYMBOL(napi_frags_finish);

3034 3035 3036 3037
struct sk_buff *napi_frags_skb(struct napi_struct *napi)
{
	struct sk_buff *skb = napi->skb;
	struct ethhdr *eth;
3038 3039
	unsigned int hlen;
	unsigned int off;
3040 3041 3042 3043 3044 3045

	napi->skb = NULL;

	skb_reset_mac_header(skb);
	skb_gro_reset_offset(skb);

3046 3047 3048 3049 3050 3051 3052 3053 3054 3055
	off = skb_gro_offset(skb);
	hlen = off + sizeof(*eth);
	eth = skb_gro_header_fast(skb, off);
	if (skb_gro_header_hard(skb, hlen)) {
		eth = skb_gro_header_slow(skb, hlen, off);
		if (unlikely(!eth)) {
			napi_reuse_skb(napi, skb);
			skb = NULL;
			goto out;
		}
3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070
	}

	skb_gro_pull(skb, sizeof(*eth));

	/*
	 * This works because the only protocols we care about don't require
	 * special handling.  We'll fix it up properly at the end.
	 */
	skb->protocol = eth->h_proto;

out:
	return skb;
}
EXPORT_SYMBOL(napi_frags_skb);

3071
gro_result_t napi_gro_frags(struct napi_struct *napi)
3072
{
3073
	struct sk_buff *skb = napi_frags_skb(napi);
3074 3075

	if (!skb)
3076
		return GRO_DROP;
3077 3078 3079

	return napi_frags_finish(napi, skb, __napi_gro_receive(napi, skb));
}
H
Herbert Xu 已提交
3080 3081
EXPORT_SYMBOL(napi_gro_frags);

3082
static int process_backlog(struct napi_struct *napi, int quota)
L
Linus Torvalds 已提交
3083 3084 3085 3086 3087
{
	int work = 0;
	struct softnet_data *queue = &__get_cpu_var(softnet_data);
	unsigned long start_time = jiffies;

3088 3089
	napi->weight = weight_p;
	do {
L
Linus Torvalds 已提交
3090 3091
		struct sk_buff *skb;

T
Tom Herbert 已提交
3092
		spin_lock_irq(&queue->input_pkt_queue.lock);
L
Linus Torvalds 已提交
3093
		skb = __skb_dequeue(&queue->input_pkt_queue);
3094
		if (!skb) {
3095
			__napi_complete(napi);
T
Tom Herbert 已提交
3096
			spin_unlock_irq(&queue->input_pkt_queue.lock);
3097
			break;
3098
		}
T
Tom Herbert 已提交
3099
		spin_unlock_irq(&queue->input_pkt_queue.lock);
L
Linus Torvalds 已提交
3100

T
Tom Herbert 已提交
3101
		__netif_receive_skb(skb);
3102
	} while (++work < quota && jiffies == start_time);
L
Linus Torvalds 已提交
3103

3104 3105
	return work;
}
L
Linus Torvalds 已提交
3106

3107 3108
/**
 * __napi_schedule - schedule for receive
3109
 * @n: entry to schedule
3110 3111 3112
 *
 * The entry's receive function will be scheduled to run
 */
H
Harvey Harrison 已提交
3113
void __napi_schedule(struct napi_struct *n)
3114 3115
{
	unsigned long flags;
L
Linus Torvalds 已提交
3116

3117 3118 3119 3120
	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
3121
}
3122 3123
EXPORT_SYMBOL(__napi_schedule);

3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156
void __napi_complete(struct napi_struct *n)
{
	BUG_ON(!test_bit(NAPI_STATE_SCHED, &n->state));
	BUG_ON(n->gro_list);

	list_del(&n->poll_list);
	smp_mb__before_clear_bit();
	clear_bit(NAPI_STATE_SCHED, &n->state);
}
EXPORT_SYMBOL(__napi_complete);

void napi_complete(struct napi_struct *n)
{
	unsigned long flags;

	/*
	 * don't let napi dequeue from the cpu poll list
	 * just in case its running on a different cpu
	 */
	if (unlikely(test_bit(NAPI_STATE_NPSVC, &n->state)))
		return;

	napi_gro_flush(n);
	local_irq_save(flags);
	__napi_complete(n);
	local_irq_restore(flags);
}
EXPORT_SYMBOL(napi_complete);

void netif_napi_add(struct net_device *dev, struct napi_struct *napi,
		    int (*poll)(struct napi_struct *, int), int weight)
{
	INIT_LIST_HEAD(&napi->poll_list);
3157
	napi->gro_count = 0;
3158
	napi->gro_list = NULL;
H
Herbert Xu 已提交
3159
	napi->skb = NULL;
3160 3161 3162 3163
	napi->poll = poll;
	napi->weight = weight;
	list_add(&napi->dev_list, &dev->napi_list);
	napi->dev = dev;
H
Herbert Xu 已提交
3164
#ifdef CONFIG_NETPOLL
3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175
	spin_lock_init(&napi->poll_lock);
	napi->poll_owner = -1;
#endif
	set_bit(NAPI_STATE_SCHED, &napi->state);
}
EXPORT_SYMBOL(netif_napi_add);

void netif_napi_del(struct napi_struct *napi)
{
	struct sk_buff *skb, *next;

3176
	list_del_init(&napi->dev_list);
3177
	napi_free_frags(napi);
3178 3179 3180 3181 3182 3183 3184 3185

	for (skb = napi->gro_list; skb; skb = next) {
		next = skb->next;
		skb->next = NULL;
		kfree_skb(skb);
	}

	napi->gro_list = NULL;
3186
	napi->gro_count = 0;
3187 3188 3189
}
EXPORT_SYMBOL(netif_napi_del);

3190
#ifdef CONFIG_SMP
T
Tom Herbert 已提交
3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206
/*
 * net_rps_action sends any pending IPI's for rps.  This is only called from
 * softirq and interrupts must be enabled.
 */
static void net_rps_action(cpumask_t *mask)
{
	int cpu;

	/* Send pending IPI's to kick RPS processing on remote cpus. */
	for_each_cpu_mask_nr(cpu, *mask) {
		struct softnet_data *queue = &per_cpu(softnet_data, cpu);
		if (cpu_online(cpu))
			__smp_call_function_single(cpu, &queue->csd, 0);
	}
	cpus_clear(*mask);
}
3207
#endif
L
Linus Torvalds 已提交
3208 3209 3210

static void net_rx_action(struct softirq_action *h)
{
3211
	struct list_head *list = &__get_cpu_var(softnet_data).poll_list;
3212
	unsigned long time_limit = jiffies + 2;
3213
	int budget = netdev_budget;
3214
	void *have;
3215
#ifdef CONFIG_SMP
T
Tom Herbert 已提交
3216 3217
	int select;
	struct rps_remote_softirq_cpus *rcpus;
3218
#endif
3219

L
Linus Torvalds 已提交
3220 3221
	local_irq_disable();

3222 3223 3224
	while (!list_empty(list)) {
		struct napi_struct *n;
		int work, weight;
L
Linus Torvalds 已提交
3225

3226
		/* If softirq window is exhuasted then punt.
3227 3228
		 * Allow this to run for 2 jiffies since which will allow
		 * an average latency of 1.5/HZ.
3229
		 */
3230
		if (unlikely(budget <= 0 || time_after(jiffies, time_limit)))
L
Linus Torvalds 已提交
3231 3232 3233 3234
			goto softnet_break;

		local_irq_enable();

3235 3236 3237 3238 3239
		/* 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.
		 */
3240
		n = list_first_entry(list, struct napi_struct, poll_list);
L
Linus Torvalds 已提交
3241

3242 3243 3244 3245
		have = netpoll_poll_lock(n);

		weight = n->weight;

3246 3247 3248 3249 3250 3251 3252
		/* 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;
3253
		if (test_bit(NAPI_STATE_SCHED, &n->state)) {
3254
			work = n->poll(n, weight);
3255 3256
			trace_napi_poll(n);
		}
3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268

		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.
		 */
3269
		if (unlikely(work == weight)) {
3270 3271 3272 3273 3274
			if (unlikely(napi_disable_pending(n))) {
				local_irq_enable();
				napi_complete(n);
				local_irq_disable();
			} else
3275 3276
				list_move_tail(&n->poll_list, list);
		}
3277 3278

		netpoll_poll_unlock(have);
L
Linus Torvalds 已提交
3279 3280
	}
out:
3281
#ifdef CONFIG_SMP
T
Tom Herbert 已提交
3282 3283 3284 3285
	rcpus = &__get_cpu_var(rps_remote_softirq_cpus);
	select = rcpus->select;
	rcpus->select ^= 1;

3286
	local_irq_enable();
3287

T
Tom Herbert 已提交
3288
	net_rps_action(&rcpus->mask[select]);
3289 3290 3291
#else
	local_irq_enable();
#endif
T
Tom Herbert 已提交
3292

3293 3294 3295 3296 3297
#ifdef CONFIG_NET_DMA
	/*
	 * There may not be any more sk_buffs coming right now, so push
	 * any pending DMA copies to hardware
	 */
3298
	dma_issue_pending_all();
3299
#endif
3300

L
Linus Torvalds 已提交
3301 3302 3303 3304 3305 3306 3307 3308
	return;

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

E
Eric Dumazet 已提交
3309
static gifconf_func_t *gifconf_list[NPROTO];
L
Linus Torvalds 已提交
3310 3311 3312 3313 3314 3315 3316 3317 3318 3319

/**
 *	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.
 */
E
Eric Dumazet 已提交
3320
int register_gifconf(unsigned int family, gifconf_func_t *gifconf)
L
Linus Torvalds 已提交
3321 3322 3323 3324 3325 3326
{
	if (family >= NPROTO)
		return -EINVAL;
	gifconf_list[family] = gifconf;
	return 0;
}
E
Eric Dumazet 已提交
3327
EXPORT_SYMBOL(register_gifconf);
L
Linus Torvalds 已提交
3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340


/*
 *	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
 */

3341
static int dev_ifname(struct net *net, struct ifreq __user *arg)
L
Linus Torvalds 已提交
3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352
{
	struct net_device *dev;
	struct ifreq ifr;

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

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

3353 3354
	rcu_read_lock();
	dev = dev_get_by_index_rcu(net, ifr.ifr_ifindex);
L
Linus Torvalds 已提交
3355
	if (!dev) {
3356
		rcu_read_unlock();
L
Linus Torvalds 已提交
3357 3358 3359 3360
		return -ENODEV;
	}

	strcpy(ifr.ifr_name, dev->name);
3361
	rcu_read_unlock();
L
Linus Torvalds 已提交
3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373

	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'.
 */

3374
static int dev_ifconf(struct net *net, char __user *arg)
L
Linus Torvalds 已提交
3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397
{
	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;
3398
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411
		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;
			}
		}
3412
	}
L
Linus Torvalds 已提交
3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429

	/*
	 *	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.
 */
3430
void *dev_seq_start(struct seq_file *seq, loff_t *pos)
3431
	__acquires(RCU)
L
Linus Torvalds 已提交
3432
{
3433
	struct net *net = seq_file_net(seq);
3434
	loff_t off;
L
Linus Torvalds 已提交
3435 3436
	struct net_device *dev;

3437
	rcu_read_lock();
3438 3439
	if (!*pos)
		return SEQ_START_TOKEN;
L
Linus Torvalds 已提交
3440

3441
	off = 1;
3442
	for_each_netdev_rcu(net, dev)
3443 3444
		if (off++ == *pos)
			return dev;
L
Linus Torvalds 已提交
3445

3446
	return NULL;
L
Linus Torvalds 已提交
3447 3448 3449 3450
}

void *dev_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
3451 3452 3453 3454
	struct net_device *dev = (v == SEQ_START_TOKEN) ?
				  first_net_device(seq_file_net(seq)) :
				  next_net_device((struct net_device *)v);

L
Linus Torvalds 已提交
3455
	++*pos;
3456
	return rcu_dereference(dev);
L
Linus Torvalds 已提交
3457 3458 3459
}

void dev_seq_stop(struct seq_file *seq, void *v)
3460
	__releases(RCU)
L
Linus Torvalds 已提交
3461
{
3462
	rcu_read_unlock();
L
Linus Torvalds 已提交
3463 3464 3465 3466
}

static void dev_seq_printf_stats(struct seq_file *seq, struct net_device *dev)
{
3467
	const struct net_device_stats *stats = dev_get_stats(dev);
L
Linus Torvalds 已提交
3468

3469
	seq_printf(seq, "%6s: %7lu %7lu %4lu %4lu %4lu %5lu %10lu %9lu "
3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485
		   "%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
3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508
}

/*
 *	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;

3509
	while (*pos < nr_cpu_ids)
3510
		if (cpu_online(*pos)) {
L
Linus Torvalds 已提交
3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536
			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;

T
Tom Herbert 已提交
3537
	seq_printf(seq, "%08x %08x %08x %08x %08x %08x %08x %08x %08x %08x\n",
3538
		   s->total, s->dropped, s->time_squeeze, 0,
3539
		   0, 0, 0, 0, /* was fastroute */
T
Tom Herbert 已提交
3540
		   s->cpu_collision, s->received_rps);
L
Linus Torvalds 已提交
3541 3542 3543
	return 0;
}

3544
static const struct seq_operations dev_seq_ops = {
L
Linus Torvalds 已提交
3545 3546 3547 3548 3549 3550 3551 3552
	.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)
{
3553 3554
	return seq_open_net(inode, file, &dev_seq_ops,
			    sizeof(struct seq_net_private));
L
Linus Torvalds 已提交
3555 3556
}

3557
static const struct file_operations dev_seq_fops = {
L
Linus Torvalds 已提交
3558 3559 3560 3561
	.owner	 = THIS_MODULE,
	.open    = dev_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
3562
	.release = seq_release_net,
L
Linus Torvalds 已提交
3563 3564
};

3565
static const struct seq_operations softnet_seq_ops = {
L
Linus Torvalds 已提交
3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576
	.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);
}

3577
static const struct file_operations softnet_seq_fops = {
L
Linus Torvalds 已提交
3578 3579 3580 3581 3582 3583 3584
	.owner	 = THIS_MODULE,
	.open    = softnet_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release,
};

3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596
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;
	}

3597
	for (t = 0; t < PTYPE_HASH_SIZE; t++) {
3598 3599 3600 3601 3602 3603 3604 3605 3606 3607
		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)
3608
	__acquires(RCU)
3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631
{
	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
3632
		hash = ntohs(pt->type) & PTYPE_HASH_MASK;
3633 3634

	while (nxt == &ptype_base[hash]) {
3635
		if (++hash >= PTYPE_HASH_SIZE)
3636 3637 3638 3639 3640 3641 3642 3643
			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)
3644
	__releases(RCU)
3645 3646 3647 3648 3649 3650 3651 3652 3653 3654
{
	rcu_read_unlock();
}

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");
3655
	else if (pt->dev == NULL || dev_net(pt->dev) == seq_file_net(seq)) {
3656 3657 3658 3659 3660
		if (pt->type == htons(ETH_P_ALL))
			seq_puts(seq, "ALL ");
		else
			seq_printf(seq, "%04x", ntohs(pt->type));

A
Alexey Dobriyan 已提交
3661 3662
		seq_printf(seq, " %-8s %pF\n",
			   pt->dev ? pt->dev->name : "", pt->func);
3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676
	}

	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)
{
3677 3678
	return seq_open_net(inode, file, &ptype_seq_ops,
			sizeof(struct seq_net_private));
3679 3680 3681 3682 3683 3684 3685
}

static const struct file_operations ptype_seq_fops = {
	.owner	 = THIS_MODULE,
	.open    = ptype_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
3686
	.release = seq_release_net,
3687 3688 3689
};


3690
static int __net_init dev_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
3691 3692 3693
{
	int rc = -ENOMEM;

3694
	if (!proc_net_fops_create(net, "dev", S_IRUGO, &dev_seq_fops))
L
Linus Torvalds 已提交
3695
		goto out;
3696
	if (!proc_net_fops_create(net, "softnet_stat", S_IRUGO, &softnet_seq_fops))
L
Linus Torvalds 已提交
3697
		goto out_dev;
3698
	if (!proc_net_fops_create(net, "ptype", S_IRUGO, &ptype_seq_fops))
3699
		goto out_softnet;
3700

3701
	if (wext_proc_init(net))
3702
		goto out_ptype;
L
Linus Torvalds 已提交
3703 3704 3705
	rc = 0;
out:
	return rc;
3706
out_ptype:
3707
	proc_net_remove(net, "ptype");
L
Linus Torvalds 已提交
3708
out_softnet:
3709
	proc_net_remove(net, "softnet_stat");
L
Linus Torvalds 已提交
3710
out_dev:
3711
	proc_net_remove(net, "dev");
L
Linus Torvalds 已提交
3712 3713
	goto out;
}
3714

3715
static void __net_exit dev_proc_net_exit(struct net *net)
3716 3717 3718 3719 3720 3721 3722 3723
{
	wext_proc_exit(net);

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

3724
static struct pernet_operations __net_initdata dev_proc_ops = {
3725 3726 3727 3728 3729 3730 3731 3732
	.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
3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761
#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;
3762

E
Eric Dumazet 已提交
3763 3764
	if (old) {
		synchronize_net();
L
Linus Torvalds 已提交
3765
		dev_put(old);
E
Eric Dumazet 已提交
3766
	}
L
Linus Torvalds 已提交
3767 3768 3769 3770 3771 3772 3773 3774
	if (master)
		slave->flags |= IFF_SLAVE;
	else
		slave->flags &= ~IFF_SLAVE;

	rtmsg_ifinfo(RTM_NEWLINK, slave, IFF_SLAVE);
	return 0;
}
E
Eric Dumazet 已提交
3775
EXPORT_SYMBOL(netdev_set_master);
L
Linus Torvalds 已提交
3776

3777 3778
static void dev_change_rx_flags(struct net_device *dev, int flags)
{
3779 3780 3781 3782
	const struct net_device_ops *ops = dev->netdev_ops;

	if ((dev->flags & IFF_UP) && ops->ndo_change_rx_flags)
		ops->ndo_change_rx_flags(dev, flags);
3783 3784
}

3785
static int __dev_set_promiscuity(struct net_device *dev, int inc)
L
Linus Torvalds 已提交
3786 3787
{
	unsigned short old_flags = dev->flags;
3788 3789
	uid_t uid;
	gid_t gid;
L
Linus Torvalds 已提交
3790

3791 3792
	ASSERT_RTNL();

3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809
	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;
		}
	}
3810
	if (dev->flags != old_flags) {
L
Linus Torvalds 已提交
3811 3812
		printk(KERN_INFO "device %s %s promiscuous mode\n",
		       dev->name, (dev->flags & IFF_PROMISC) ? "entered" :
3813
							       "left");
3814 3815
		if (audit_enabled) {
			current_uid_gid(&uid, &gid);
3816 3817 3818 3819 3820 3821
			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),
3822
				uid, gid,
3823
				audit_get_sessionid(current));
3824
		}
3825

3826
		dev_change_rx_flags(dev, IFF_PROMISC);
L
Linus Torvalds 已提交
3827
	}
3828
	return 0;
L
Linus Torvalds 已提交
3829 3830
}

3831 3832 3833 3834 3835 3836 3837 3838 3839
/**
 *	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.
3840
 *	Return 0 if successful or a negative errno code on error.
3841
 */
3842
int dev_set_promiscuity(struct net_device *dev, int inc)
3843 3844
{
	unsigned short old_flags = dev->flags;
3845
	int err;
3846

3847
	err = __dev_set_promiscuity(dev, inc);
3848
	if (err < 0)
3849
		return err;
3850 3851
	if (dev->flags != old_flags)
		dev_set_rx_mode(dev);
3852
	return err;
3853
}
E
Eric Dumazet 已提交
3854
EXPORT_SYMBOL(dev_set_promiscuity);
3855

L
Linus Torvalds 已提交
3856 3857 3858 3859 3860 3861 3862 3863 3864 3865
/**
 *	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.
3866
 *	Return 0 if successful or a negative errno code on error.
L
Linus Torvalds 已提交
3867 3868
 */

3869
int dev_set_allmulti(struct net_device *dev, int inc)
L
Linus Torvalds 已提交
3870 3871 3872
{
	unsigned short old_flags = dev->flags;

3873 3874
	ASSERT_RTNL();

L
Linus Torvalds 已提交
3875
	dev->flags |= IFF_ALLMULTI;
3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891
	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;
		}
	}
3892
	if (dev->flags ^ old_flags) {
3893
		dev_change_rx_flags(dev, IFF_ALLMULTI);
3894
		dev_set_rx_mode(dev);
3895
	}
3896
	return 0;
3897
}
E
Eric Dumazet 已提交
3898
EXPORT_SYMBOL(dev_set_allmulti);
3899 3900 3901 3902

/*
 *	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
3903
 *	filtering it is put in promiscuous mode while unicast addresses
3904 3905 3906 3907
 *	are present.
 */
void __dev_set_rx_mode(struct net_device *dev)
{
3908 3909
	const struct net_device_ops *ops = dev->netdev_ops;

3910 3911 3912 3913 3914
	/* dev_open will call this function so the list will stay sane. */
	if (!(dev->flags&IFF_UP))
		return;

	if (!netif_device_present(dev))
3915
		return;
3916

3917 3918
	if (ops->ndo_set_rx_mode)
		ops->ndo_set_rx_mode(dev);
3919 3920 3921 3922
	else {
		/* Unicast addresses changes may only happen under the rtnl,
		 * therefore calling __dev_set_promiscuity here is safe.
		 */
3923
		if (!netdev_uc_empty(dev) && !dev->uc_promisc) {
3924 3925
			__dev_set_promiscuity(dev, 1);
			dev->uc_promisc = 1;
3926
		} else if (netdev_uc_empty(dev) && dev->uc_promisc) {
3927 3928 3929 3930
			__dev_set_promiscuity(dev, -1);
			dev->uc_promisc = 0;
		}

3931 3932
		if (ops->ndo_set_multicast_list)
			ops->ndo_set_multicast_list(dev);
3933 3934 3935 3936 3937
	}
}

void dev_set_rx_mode(struct net_device *dev)
{
3938
	netif_addr_lock_bh(dev);
3939
	__dev_set_rx_mode(dev);
3940
	netif_addr_unlock_bh(dev);
L
Linus Torvalds 已提交
3941 3942
}

3943 3944
/* hw addresses list handling functions */

3945 3946
static int __hw_addr_add(struct netdev_hw_addr_list *list, unsigned char *addr,
			 int addr_len, unsigned char addr_type)
3947 3948 3949 3950 3951 3952 3953
{
	struct netdev_hw_addr *ha;
	int alloc_size;

	if (addr_len > MAX_ADDR_LEN)
		return -EINVAL;

3954
	list_for_each_entry(ha, &list->list, list) {
J
Jiri Pirko 已提交
3955 3956 3957 3958 3959 3960 3961 3962
		if (!memcmp(ha->addr, addr, addr_len) &&
		    ha->type == addr_type) {
			ha->refcount++;
			return 0;
		}
	}


3963 3964 3965 3966 3967 3968 3969 3970
	alloc_size = sizeof(*ha);
	if (alloc_size < L1_CACHE_BYTES)
		alloc_size = L1_CACHE_BYTES;
	ha = kmalloc(alloc_size, GFP_ATOMIC);
	if (!ha)
		return -ENOMEM;
	memcpy(ha->addr, addr, addr_len);
	ha->type = addr_type;
J
Jiri Pirko 已提交
3971 3972
	ha->refcount = 1;
	ha->synced = false;
3973 3974
	list_add_tail_rcu(&ha->list, &list->list);
	list->count++;
3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985
	return 0;
}

static void ha_rcu_free(struct rcu_head *head)
{
	struct netdev_hw_addr *ha;

	ha = container_of(head, struct netdev_hw_addr, rcu_head);
	kfree(ha);
}

3986 3987
static int __hw_addr_del(struct netdev_hw_addr_list *list, unsigned char *addr,
			 int addr_len, unsigned char addr_type)
3988 3989 3990
{
	struct netdev_hw_addr *ha;

3991
	list_for_each_entry(ha, &list->list, list) {
J
Jiri Pirko 已提交
3992
		if (!memcmp(ha->addr, addr, addr_len) &&
3993
		    (ha->type == addr_type || !addr_type)) {
J
Jiri Pirko 已提交
3994 3995
			if (--ha->refcount)
				return 0;
3996 3997
			list_del_rcu(&ha->list);
			call_rcu(&ha->rcu_head, ha_rcu_free);
3998
			list->count--;
3999 4000 4001 4002 4003 4004
			return 0;
		}
	}
	return -ENOENT;
}

4005 4006 4007
static int __hw_addr_add_multiple(struct netdev_hw_addr_list *to_list,
				  struct netdev_hw_addr_list *from_list,
				  int addr_len,
J
Jiri Pirko 已提交
4008
				  unsigned char addr_type)
4009 4010 4011 4012 4013
{
	int err;
	struct netdev_hw_addr *ha, *ha2;
	unsigned char type;

4014
	list_for_each_entry(ha, &from_list->list, list) {
4015
		type = addr_type ? addr_type : ha->type;
4016
		err = __hw_addr_add(to_list, ha->addr, addr_len, type);
4017 4018 4019 4020 4021 4022
		if (err)
			goto unroll;
	}
	return 0;

unroll:
4023
	list_for_each_entry(ha2, &from_list->list, list) {
4024 4025 4026
		if (ha2 == ha)
			break;
		type = addr_type ? addr_type : ha2->type;
4027
		__hw_addr_del(to_list, ha2->addr, addr_len, type);
4028 4029 4030 4031
	}
	return err;
}

4032 4033 4034
static void __hw_addr_del_multiple(struct netdev_hw_addr_list *to_list,
				   struct netdev_hw_addr_list *from_list,
				   int addr_len,
J
Jiri Pirko 已提交
4035
				   unsigned char addr_type)
4036 4037 4038 4039
{
	struct netdev_hw_addr *ha;
	unsigned char type;

4040
	list_for_each_entry(ha, &from_list->list, list) {
4041
		type = addr_type ? addr_type : ha->type;
4042
		__hw_addr_del(to_list, ha->addr, addr_len, addr_type);
J
Jiri Pirko 已提交
4043 4044 4045
	}
}

4046 4047
static int __hw_addr_sync(struct netdev_hw_addr_list *to_list,
			  struct netdev_hw_addr_list *from_list,
J
Jiri Pirko 已提交
4048 4049 4050 4051 4052
			  int addr_len)
{
	int err = 0;
	struct netdev_hw_addr *ha, *tmp;

4053
	list_for_each_entry_safe(ha, tmp, &from_list->list, list) {
J
Jiri Pirko 已提交
4054
		if (!ha->synced) {
4055
			err = __hw_addr_add(to_list, ha->addr,
J
Jiri Pirko 已提交
4056 4057 4058 4059 4060 4061
					    addr_len, ha->type);
			if (err)
				break;
			ha->synced = true;
			ha->refcount++;
		} else if (ha->refcount == 1) {
4062 4063
			__hw_addr_del(to_list, ha->addr, addr_len, ha->type);
			__hw_addr_del(from_list, ha->addr, addr_len, ha->type);
J
Jiri Pirko 已提交
4064
		}
4065
	}
J
Jiri Pirko 已提交
4066
	return err;
4067 4068
}

4069 4070
static void __hw_addr_unsync(struct netdev_hw_addr_list *to_list,
			     struct netdev_hw_addr_list *from_list,
J
Jiri Pirko 已提交
4071 4072 4073 4074
			     int addr_len)
{
	struct netdev_hw_addr *ha, *tmp;

4075
	list_for_each_entry_safe(ha, tmp, &from_list->list, list) {
J
Jiri Pirko 已提交
4076
		if (ha->synced) {
4077
			__hw_addr_del(to_list, ha->addr,
J
Jiri Pirko 已提交
4078 4079
				      addr_len, ha->type);
			ha->synced = false;
4080
			__hw_addr_del(from_list, ha->addr,
J
Jiri Pirko 已提交
4081 4082 4083 4084 4085
				      addr_len, ha->type);
		}
	}
}

4086
static void __hw_addr_flush(struct netdev_hw_addr_list *list)
4087 4088 4089
{
	struct netdev_hw_addr *ha, *tmp;

4090
	list_for_each_entry_safe(ha, tmp, &list->list, list) {
4091 4092 4093
		list_del_rcu(&ha->list);
		call_rcu(&ha->rcu_head, ha_rcu_free);
	}
4094 4095 4096 4097 4098 4099 4100
	list->count = 0;
}

static void __hw_addr_init(struct netdev_hw_addr_list *list)
{
	INIT_LIST_HEAD(&list->list);
	list->count = 0;
4101 4102 4103 4104 4105 4106 4107 4108
}

/* Device addresses handling functions */

static void dev_addr_flush(struct net_device *dev)
{
	/* rtnl_mutex must be held here */

4109
	__hw_addr_flush(&dev->dev_addrs);
4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120
	dev->dev_addr = NULL;
}

static int dev_addr_init(struct net_device *dev)
{
	unsigned char addr[MAX_ADDR_LEN];
	struct netdev_hw_addr *ha;
	int err;

	/* rtnl_mutex must be held here */

4121
	__hw_addr_init(&dev->dev_addrs);
E
Eric Dumazet 已提交
4122
	memset(addr, 0, sizeof(addr));
4123
	err = __hw_addr_add(&dev->dev_addrs, addr, sizeof(addr),
4124 4125 4126 4127 4128 4129
			    NETDEV_HW_ADDR_T_LAN);
	if (!err) {
		/*
		 * Get the first (previously created) address from the list
		 * and set dev_addr pointer to this location.
		 */
4130
		ha = list_first_entry(&dev->dev_addrs.list,
4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154
				      struct netdev_hw_addr, list);
		dev->dev_addr = ha->addr;
	}
	return err;
}

/**
 *	dev_addr_add	- Add a device address
 *	@dev: device
 *	@addr: address to add
 *	@addr_type: address type
 *
 *	Add a device address to the device or increase the reference count if
 *	it already exists.
 *
 *	The caller must hold the rtnl_mutex.
 */
int dev_addr_add(struct net_device *dev, unsigned char *addr,
		 unsigned char addr_type)
{
	int err;

	ASSERT_RTNL();

4155
	err = __hw_addr_add(&dev->dev_addrs, addr, dev->addr_len, addr_type);
4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176
	if (!err)
		call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
	return err;
}
EXPORT_SYMBOL(dev_addr_add);

/**
 *	dev_addr_del	- Release a device address.
 *	@dev: device
 *	@addr: address to delete
 *	@addr_type: address type
 *
 *	Release reference to a device address and remove it from the device
 *	if the reference count drops to zero.
 *
 *	The caller must hold the rtnl_mutex.
 */
int dev_addr_del(struct net_device *dev, unsigned char *addr,
		 unsigned char addr_type)
{
	int err;
J
Jiri Pirko 已提交
4177
	struct netdev_hw_addr *ha;
4178 4179 4180

	ASSERT_RTNL();

J
Jiri Pirko 已提交
4181 4182 4183 4184
	/*
	 * We can not remove the first address from the list because
	 * dev->dev_addr points to that.
	 */
4185 4186
	ha = list_first_entry(&dev->dev_addrs.list,
			      struct netdev_hw_addr, list);
J
Jiri Pirko 已提交
4187 4188 4189
	if (ha->addr == dev->dev_addr && ha->refcount == 1)
		return -ENOENT;

4190
	err = __hw_addr_del(&dev->dev_addrs, addr, dev->addr_len,
J
Jiri Pirko 已提交
4191
			    addr_type);
4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217
	if (!err)
		call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
	return err;
}
EXPORT_SYMBOL(dev_addr_del);

/**
 *	dev_addr_add_multiple	- Add device addresses from another device
 *	@to_dev: device to which addresses will be added
 *	@from_dev: device from which addresses will be added
 *	@addr_type: address type - 0 means type will be used from from_dev
 *
 *	Add device addresses of the one device to another.
 **
 *	The caller must hold the rtnl_mutex.
 */
int dev_addr_add_multiple(struct net_device *to_dev,
			  struct net_device *from_dev,
			  unsigned char addr_type)
{
	int err;

	ASSERT_RTNL();

	if (from_dev->addr_len != to_dev->addr_len)
		return -EINVAL;
4218
	err = __hw_addr_add_multiple(&to_dev->dev_addrs, &from_dev->dev_addrs,
J
Jiri Pirko 已提交
4219
				     to_dev->addr_len, addr_type);
4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243
	if (!err)
		call_netdevice_notifiers(NETDEV_CHANGEADDR, to_dev);
	return err;
}
EXPORT_SYMBOL(dev_addr_add_multiple);

/**
 *	dev_addr_del_multiple	- Delete device addresses by another device
 *	@to_dev: device where the addresses will be deleted
 *	@from_dev: device by which addresses the addresses will be deleted
 *	@addr_type: address type - 0 means type will used from from_dev
 *
 *	Deletes addresses in to device by the list of addresses in from device.
 *
 *	The caller must hold the rtnl_mutex.
 */
int dev_addr_del_multiple(struct net_device *to_dev,
			  struct net_device *from_dev,
			  unsigned char addr_type)
{
	ASSERT_RTNL();

	if (from_dev->addr_len != to_dev->addr_len)
		return -EINVAL;
4244
	__hw_addr_del_multiple(&to_dev->dev_addrs, &from_dev->dev_addrs,
J
Jiri Pirko 已提交
4245
			       to_dev->addr_len, addr_type);
4246 4247 4248 4249 4250
	call_netdevice_notifiers(NETDEV_CHANGEADDR, to_dev);
	return 0;
}
EXPORT_SYMBOL(dev_addr_del_multiple);

4251
/* multicast addresses handling functions */
4252

4253 4254
int __dev_addr_delete(struct dev_addr_list **list, int *count,
		      void *addr, int alen, int glbl)
4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271
{
	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);
4272
			(*count)--;
4273 4274 4275 4276 4277 4278
			return 0;
		}
	}
	return -ENOENT;
}

4279 4280
int __dev_addr_add(struct dev_addr_list **list, int *count,
		   void *addr, int alen, int glbl)
4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297
{
	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;
		}
	}

4298
	da = kzalloc(sizeof(*da), GFP_ATOMIC);
4299 4300 4301 4302 4303 4304 4305 4306
	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;
4307
	(*count)++;
4308 4309 4310
	return 0;
}

4311 4312 4313
/**
 *	dev_unicast_delete	- Release secondary unicast address.
 *	@dev: device
R
Randy Dunlap 已提交
4314
 *	@addr: address to delete
4315 4316
 *
 *	Release reference to a secondary unicast address and remove it
R
Randy Dunlap 已提交
4317
 *	from the device if the reference count drops to zero.
4318 4319 4320
 *
 * 	The caller must hold the rtnl_mutex.
 */
J
Jiri Pirko 已提交
4321
int dev_unicast_delete(struct net_device *dev, void *addr)
4322 4323 4324 4325 4326
{
	int err;

	ASSERT_RTNL();

4327
	netif_addr_lock_bh(dev);
4328 4329
	err = __hw_addr_del(&dev->uc, addr, dev->addr_len,
			    NETDEV_HW_ADDR_T_UNICAST);
4330
	if (!err)
4331
		__dev_set_rx_mode(dev);
4332
	netif_addr_unlock_bh(dev);
4333 4334 4335 4336 4337 4338 4339
	return err;
}
EXPORT_SYMBOL(dev_unicast_delete);

/**
 *	dev_unicast_add		- add a secondary unicast address
 *	@dev: device
4340
 *	@addr: address to add
4341 4342 4343 4344 4345 4346
 *
 *	Add a secondary unicast address to the device or increase
 *	the reference count if it already exists.
 *
 *	The caller must hold the rtnl_mutex.
 */
J
Jiri Pirko 已提交
4347
int dev_unicast_add(struct net_device *dev, void *addr)
4348 4349 4350 4351 4352
{
	int err;

	ASSERT_RTNL();

4353
	netif_addr_lock_bh(dev);
4354 4355
	err = __hw_addr_add(&dev->uc, addr, dev->addr_len,
			    NETDEV_HW_ADDR_T_UNICAST);
4356
	if (!err)
4357
		__dev_set_rx_mode(dev);
4358
	netif_addr_unlock_bh(dev);
4359 4360 4361 4362
	return err;
}
EXPORT_SYMBOL(dev_unicast_add);

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
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;
}
4389
EXPORT_SYMBOL_GPL(__dev_addr_sync);
4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408

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;
	}
}
4409
EXPORT_SYMBOL_GPL(__dev_addr_unsync);
4410 4411 4412 4413 4414 4415 4416

/**
 *	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
4417 4418
 *	addresses that have no users left. The source device must be
 *	locked by netif_tx_lock_bh.
4419 4420 4421 4422 4423 4424 4425 4426
 *
 *	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;

J
Jiri Pirko 已提交
4427 4428 4429
	if (to->addr_len != from->addr_len)
		return -EINVAL;

4430
	netif_addr_lock_bh(to);
4431
	err = __hw_addr_sync(&to->uc, &from->uc, to->addr_len);
4432 4433
	if (!err)
		__dev_set_rx_mode(to);
4434
	netif_addr_unlock_bh(to);
4435 4436 4437 4438 4439
	return err;
}
EXPORT_SYMBOL(dev_unicast_sync);

/**
4440
 *	dev_unicast_unsync - Remove synchronized addresses from the destination device
4441 4442 4443 4444 4445 4446 4447 4448 4449
 *	@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)
{
J
Jiri Pirko 已提交
4450 4451
	if (to->addr_len != from->addr_len)
		return;
4452

4453 4454
	netif_addr_lock_bh(from);
	netif_addr_lock(to);
4455
	__hw_addr_unsync(&to->uc, &from->uc, to->addr_len);
J
Jiri Pirko 已提交
4456
	__dev_set_rx_mode(to);
4457 4458
	netif_addr_unlock(to);
	netif_addr_unlock_bh(from);
4459 4460 4461
}
EXPORT_SYMBOL(dev_unicast_unsync);

4462
void dev_unicast_flush(struct net_device *dev)
J
Jiri Pirko 已提交
4463
{
4464
	netif_addr_lock_bh(dev);
4465
	__hw_addr_flush(&dev->uc);
4466
	netif_addr_unlock_bh(dev);
J
Jiri Pirko 已提交
4467
}
4468
EXPORT_SYMBOL(dev_unicast_flush);
J
Jiri Pirko 已提交
4469 4470 4471

static void dev_unicast_init(struct net_device *dev)
{
4472
	__hw_addr_init(&dev->uc);
J
Jiri Pirko 已提交
4473 4474 4475
}


4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489
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);
	}
}

4490
void dev_addr_discard(struct net_device *dev)
4491
{
4492
	netif_addr_lock_bh(dev);
4493

4494
	__dev_addr_discard(&dev->mc_list);
4495
	netdev_mc_count(dev) = 0;
4496

4497
	netif_addr_unlock_bh(dev);
4498
}
4499
EXPORT_SYMBOL(dev_addr_discard);
4500

4501 4502 4503 4504 4505 4506
/**
 *	dev_get_flags - get flags reported to userspace
 *	@dev: device
 *
 *	Get the combination of flag bits exported through APIs to userspace.
 */
L
Linus Torvalds 已提交
4507 4508 4509 4510 4511 4512
unsigned dev_get_flags(const struct net_device *dev)
{
	unsigned flags;

	flags = (dev->flags & ~(IFF_PROMISC |
				IFF_ALLMULTI |
S
Stefan Rompf 已提交
4513 4514 4515
				IFF_RUNNING |
				IFF_LOWER_UP |
				IFF_DORMANT)) |
L
Linus Torvalds 已提交
4516 4517 4518
		(dev->gflags & (IFF_PROMISC |
				IFF_ALLMULTI));

S
Stefan Rompf 已提交
4519 4520 4521 4522 4523 4524 4525 4526
	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
4527 4528 4529

	return flags;
}
E
Eric Dumazet 已提交
4530
EXPORT_SYMBOL(dev_get_flags);
L
Linus Torvalds 已提交
4531

4532
int __dev_change_flags(struct net_device *dev, unsigned int flags)
L
Linus Torvalds 已提交
4533 4534
{
	int old_flags = dev->flags;
4535
	int ret;
L
Linus Torvalds 已提交
4536

4537 4538
	ASSERT_RTNL();

L
Linus Torvalds 已提交
4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552
	/*
	 *	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.
	 */

4553 4554
	if ((old_flags ^ flags) & IFF_MULTICAST)
		dev_change_rx_flags(dev, IFF_MULTICAST);
4555

4556
	dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
4557 4558 4559 4560 4561 4562 4563 4564 4565

	/*
	 *	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  ? */
4566
		ret = ((old_flags & IFF_UP) ? __dev_close : __dev_open)(dev);
L
Linus Torvalds 已提交
4567 4568

		if (!ret)
4569
			dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
4570 4571 4572
	}

	if ((flags ^ dev->gflags) & IFF_PROMISC) {
E
Eric Dumazet 已提交
4573 4574
		int inc = (flags & IFF_PROMISC) ? 1 : -1;

L
Linus Torvalds 已提交
4575 4576 4577 4578 4579 4580 4581 4582 4583
		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) {
E
Eric Dumazet 已提交
4584 4585
		int inc = (flags & IFF_ALLMULTI) ? 1 : -1;

L
Linus Torvalds 已提交
4586 4587 4588 4589
		dev->gflags ^= IFF_ALLMULTI;
		dev_set_allmulti(dev, inc);
	}

4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626
	return ret;
}

void __dev_notify_flags(struct net_device *dev, unsigned int old_flags)
{
	unsigned int changes = dev->flags ^ old_flags;

	if (changes & IFF_UP) {
		if (dev->flags & IFF_UP)
			call_netdevice_notifiers(NETDEV_UP, dev);
		else
			call_netdevice_notifiers(NETDEV_DOWN, dev);
	}

	if (dev->flags & IFF_UP &&
	    (changes & ~(IFF_UP | IFF_PROMISC | IFF_ALLMULTI | IFF_VOLATILE)))
		call_netdevice_notifiers(NETDEV_CHANGE, dev);
}

/**
 *	dev_change_flags - change device settings
 *	@dev: device
 *	@flags: device state flags
 *
 *	Change settings on device based state flags. The flags are
 *	in the userspace exported format.
 */
int dev_change_flags(struct net_device *dev, unsigned flags)
{
	int ret, changes;
	int old_flags = dev->flags;

	ret = __dev_change_flags(dev, flags);
	if (ret < 0)
		return ret;

	changes = old_flags ^ dev->flags;
4627 4628
	if (changes)
		rtmsg_ifinfo(RTM_NEWLINK, dev, changes);
L
Linus Torvalds 已提交
4629

4630
	__dev_notify_flags(dev, old_flags);
L
Linus Torvalds 已提交
4631 4632
	return ret;
}
E
Eric Dumazet 已提交
4633
EXPORT_SYMBOL(dev_change_flags);
L
Linus Torvalds 已提交
4634

4635 4636 4637 4638 4639 4640 4641
/**
 *	dev_set_mtu - Change maximum transfer unit
 *	@dev: device
 *	@new_mtu: new transfer unit
 *
 *	Change the maximum transfer size of the network device.
 */
L
Linus Torvalds 已提交
4642 4643
int dev_set_mtu(struct net_device *dev, int new_mtu)
{
4644
	const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657
	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;
4658 4659
	if (ops->ndo_change_mtu)
		err = ops->ndo_change_mtu(dev, new_mtu);
L
Linus Torvalds 已提交
4660 4661
	else
		dev->mtu = new_mtu;
4662

L
Linus Torvalds 已提交
4663
	if (!err && dev->flags & IFF_UP)
4664
		call_netdevice_notifiers(NETDEV_CHANGEMTU, dev);
L
Linus Torvalds 已提交
4665 4666
	return err;
}
E
Eric Dumazet 已提交
4667
EXPORT_SYMBOL(dev_set_mtu);
L
Linus Torvalds 已提交
4668

4669 4670 4671 4672 4673 4674 4675
/**
 *	dev_set_mac_address - Change Media Access Control Address
 *	@dev: device
 *	@sa: new address
 *
 *	Change the hardware (MAC) address of the device
 */
L
Linus Torvalds 已提交
4676 4677
int dev_set_mac_address(struct net_device *dev, struct sockaddr *sa)
{
4678
	const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
4679 4680
	int err;

4681
	if (!ops->ndo_set_mac_address)
L
Linus Torvalds 已提交
4682 4683 4684 4685 4686
		return -EOPNOTSUPP;
	if (sa->sa_family != dev->type)
		return -EINVAL;
	if (!netif_device_present(dev))
		return -ENODEV;
4687
	err = ops->ndo_set_mac_address(dev, sa);
L
Linus Torvalds 已提交
4688
	if (!err)
4689
		call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
L
Linus Torvalds 已提交
4690 4691
	return err;
}
E
Eric Dumazet 已提交
4692
EXPORT_SYMBOL(dev_set_mac_address);
L
Linus Torvalds 已提交
4693 4694

/*
E
Eric Dumazet 已提交
4695
 *	Perform the SIOCxIFxxx calls, inside rcu_read_lock()
L
Linus Torvalds 已提交
4696
 */
4697
static int dev_ifsioc_locked(struct net *net, struct ifreq *ifr, unsigned int cmd)
L
Linus Torvalds 已提交
4698 4699
{
	int err;
E
Eric Dumazet 已提交
4700
	struct net_device *dev = dev_get_by_name_rcu(net, ifr->ifr_name);
L
Linus Torvalds 已提交
4701 4702 4703 4704 4705

	if (!dev)
		return -ENODEV;

	switch (cmd) {
E
Eric Dumazet 已提交
4706 4707 4708
	case SIOCGIFFLAGS:	/* Get interface flags */
		ifr->ifr_flags = (short) dev_get_flags(dev);
		return 0;
L
Linus Torvalds 已提交
4709

E
Eric Dumazet 已提交
4710 4711 4712 4713
	case SIOCGIFMETRIC:	/* Get the metric on the interface
				   (currently unused) */
		ifr->ifr_metric = 0;
		return 0;
L
Linus Torvalds 已提交
4714

E
Eric Dumazet 已提交
4715 4716 4717
	case SIOCGIFMTU:	/* Get the MTU of a device */
		ifr->ifr_mtu = dev->mtu;
		return 0;
L
Linus Torvalds 已提交
4718

E
Eric Dumazet 已提交
4719 4720 4721 4722 4723 4724 4725 4726
	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;
L
Linus Torvalds 已提交
4727

E
Eric Dumazet 已提交
4728 4729 4730
	case SIOCGIFSLAVE:
		err = -EINVAL;
		break;
4731

E
Eric Dumazet 已提交
4732 4733 4734 4735 4736 4737 4738 4739
	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;
4740

E
Eric Dumazet 已提交
4741 4742 4743
	case SIOCGIFINDEX:
		ifr->ifr_ifindex = dev->ifindex;
		return 0;
4744

E
Eric Dumazet 已提交
4745 4746 4747
	case SIOCGIFTXQLEN:
		ifr->ifr_qlen = dev->tx_queue_len;
		return 0;
4748

E
Eric Dumazet 已提交
4749 4750 4751 4752 4753 4754 4755
	default:
		/* dev_ioctl() should ensure this case
		 * is never reached
		 */
		WARN_ON(1);
		err = -EINVAL;
		break;
4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767

	}
	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);
J
Jarek Poplawski 已提交
4768
	const struct net_device_ops *ops;
4769 4770 4771 4772

	if (!dev)
		return -ENODEV;

J
Jarek Poplawski 已提交
4773 4774
	ops = dev->netdev_ops;

4775
	switch (cmd) {
E
Eric Dumazet 已提交
4776 4777
	case SIOCSIFFLAGS:	/* Set interface flags */
		return dev_change_flags(dev, ifr->ifr_flags);
4778

E
Eric Dumazet 已提交
4779 4780 4781
	case SIOCSIFMETRIC:	/* Set the metric on the interface
				   (currently unused) */
		return -EOPNOTSUPP;
4782

E
Eric Dumazet 已提交
4783 4784
	case SIOCSIFMTU:	/* Set the MTU of a device */
		return dev_set_mtu(dev, ifr->ifr_mtu);
L
Linus Torvalds 已提交
4785

E
Eric Dumazet 已提交
4786 4787
	case SIOCSIFHWADDR:
		return dev_set_mac_address(dev, &ifr->ifr_hwaddr);
L
Linus Torvalds 已提交
4788

E
Eric Dumazet 已提交
4789 4790 4791 4792 4793 4794 4795
	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));
		call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
		return 0;
L
Linus Torvalds 已提交
4796

E
Eric Dumazet 已提交
4797 4798
	case SIOCSIFMAP:
		if (ops->ndo_set_config) {
L
Linus Torvalds 已提交
4799 4800
			if (!netif_device_present(dev))
				return -ENODEV;
E
Eric Dumazet 已提交
4801 4802 4803
			return ops->ndo_set_config(dev, &ifr->ifr_map);
		}
		return -EOPNOTSUPP;
L
Linus Torvalds 已提交
4804

E
Eric Dumazet 已提交
4805 4806 4807 4808 4809 4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821
	case SIOCADDMULTI:
		if ((!ops->ndo_set_multicast_list && !ops->ndo_set_rx_mode) ||
		    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:
		if ((!ops->ndo_set_multicast_list && !ops->ndo_set_rx_mode) ||
		    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);
L
Linus Torvalds 已提交
4822

E
Eric Dumazet 已提交
4823 4824 4825 4826 4827
	case SIOCSIFTXQLEN:
		if (ifr->ifr_qlen < 0)
			return -EINVAL;
		dev->tx_queue_len = ifr->ifr_qlen;
		return 0;
L
Linus Torvalds 已提交
4828

E
Eric Dumazet 已提交
4829 4830 4831
	case SIOCSIFNAME:
		ifr->ifr_newname[IFNAMSIZ-1] = '\0';
		return dev_change_name(dev, ifr->ifr_newname);
L
Linus Torvalds 已提交
4832

E
Eric Dumazet 已提交
4833 4834 4835 4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846 4847 4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 4858 4859 4860
	/*
	 *	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 == SIOCSHWTSTAMP ||
		    cmd == SIOCWANDEV) {
			err = -EOPNOTSUPP;
			if (ops->ndo_do_ioctl) {
				if (netif_device_present(dev))
					err = ops->ndo_do_ioctl(dev, ifr, cmd);
				else
					err = -ENODEV;
			}
		} else
			err = -EINVAL;
L
Linus Torvalds 已提交
4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871 4872

	}
	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
4873
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
4874 4875 4876 4877 4878 4879 4880 4881 4882
 *	@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.
 */

4883
int dev_ioctl(struct net *net, unsigned int cmd, void __user *arg)
L
Linus Torvalds 已提交
4884 4885 4886 4887 4888 4889 4890 4891 4892 4893 4894
{
	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) {
4895
		rtnl_lock();
4896
		ret = dev_ifconf(net, (char __user *) arg);
4897
		rtnl_unlock();
L
Linus Torvalds 已提交
4898 4899 4900
		return ret;
	}
	if (cmd == SIOCGIFNAME)
4901
		return dev_ifname(net, (struct ifreq __user *)arg);
L
Linus Torvalds 已提交
4902 4903 4904 4905 4906 4907 4908 4909 4910 4911 4912 4913 4914 4915 4916

	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) {
E
Eric Dumazet 已提交
4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931
	/*
	 *	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:
		dev_load(net, ifr.ifr_name);
E
Eric Dumazet 已提交
4932
		rcu_read_lock();
E
Eric Dumazet 已提交
4933
		ret = dev_ifsioc_locked(net, &ifr, cmd);
E
Eric Dumazet 已提交
4934
		rcu_read_unlock();
E
Eric Dumazet 已提交
4935 4936 4937 4938 4939 4940 4941 4942
		if (!ret) {
			if (colon)
				*colon = ':';
			if (copy_to_user(arg, &ifr,
					 sizeof(struct ifreq)))
				ret = -EFAULT;
		}
		return ret;
L
Linus Torvalds 已提交
4943

E
Eric Dumazet 已提交
4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956
	case SIOCETHTOOL:
		dev_load(net, ifr.ifr_name);
		rtnl_lock();
		ret = dev_ethtool(net, &ifr);
		rtnl_unlock();
		if (!ret) {
			if (colon)
				*colon = ':';
			if (copy_to_user(arg, &ifr,
					 sizeof(struct ifreq)))
				ret = -EFAULT;
		}
		return ret;
L
Linus Torvalds 已提交
4957

E
Eric Dumazet 已提交
4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980
	/*
	 *	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;
		dev_load(net, ifr.ifr_name);
		rtnl_lock();
		ret = dev_ifsioc(net, &ifr, cmd);
		rtnl_unlock();
		if (!ret) {
			if (colon)
				*colon = ':';
			if (copy_to_user(arg, &ifr,
					 sizeof(struct ifreq)))
				ret = -EFAULT;
		}
		return ret;
L
Linus Torvalds 已提交
4981

E
Eric Dumazet 已提交
4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996 4997 4998 4999 5000 5001 5002 5003 5004 5005 5006 5007 5008 5009 5010 5011 5012 5013 5014 5015 5016 5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032
	/*
	 *	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:
	case SIOCSHWTSTAMP:
		if (!capable(CAP_NET_ADMIN))
			return -EPERM;
		/* fall through */
	case SIOCBONDSLAVEINFOQUERY:
	case SIOCBONDINFOQUERY:
		dev_load(net, ifr.ifr_name);
		rtnl_lock();
		ret = dev_ifsioc(net, &ifr, cmd);
		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)) {
5033
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
5034
			rtnl_lock();
5035
			ret = dev_ifsioc(net, &ifr, cmd);
L
Linus Torvalds 已提交
5036
			rtnl_unlock();
E
Eric Dumazet 已提交
5037 5038 5039
			if (!ret && copy_to_user(arg, &ifr,
						 sizeof(struct ifreq)))
				ret = -EFAULT;
L
Linus Torvalds 已提交
5040
			return ret;
E
Eric Dumazet 已提交
5041 5042 5043 5044 5045
		}
		/* Take care of Wireless Extensions */
		if (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST)
			return wext_handle_ioctl(net, &ifr, cmd, arg);
		return -EINVAL;
L
Linus Torvalds 已提交
5046 5047 5048 5049 5050 5051
	}
}


/**
 *	dev_new_index	-	allocate an ifindex
5052
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
5053 5054 5055 5056 5057
 *
 *	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.
 */
5058
static int dev_new_index(struct net *net)
L
Linus Torvalds 已提交
5059 5060 5061 5062 5063
{
	static int ifindex;
	for (;;) {
		if (++ifindex <= 0)
			ifindex = 1;
5064
		if (!__dev_get_by_index(net, ifindex))
L
Linus Torvalds 已提交
5065 5066 5067 5068 5069
			return ifindex;
	}
}

/* Delayed registration/unregisteration */
5070
static LIST_HEAD(net_todo_list);
L
Linus Torvalds 已提交
5071

5072
static void net_set_todo(struct net_device *dev)
L
Linus Torvalds 已提交
5073 5074 5075 5076
{
	list_add_tail(&dev->todo_list, &net_todo_list);
}

5077
static void rollback_registered_many(struct list_head *head)
5078
{
5079
	struct net_device *dev, *tmp;
5080

5081 5082 5083
	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

5084
	list_for_each_entry_safe(dev, tmp, head, unreg_list) {
5085
		/* Some devices call without registering
5086 5087
		 * for initialization unwind. Remove those
		 * devices and proceed with the remaining.
5088 5089 5090 5091
		 */
		if (dev->reg_state == NETREG_UNINITIALIZED) {
			pr_debug("unregister_netdevice: device %s/%p never "
				 "was registered\n", dev->name, dev);
5092

5093
			WARN_ON(1);
5094 5095
			list_del(&dev->unreg_list);
			continue;
5096
		}
5097

5098
		BUG_ON(dev->reg_state != NETREG_REGISTERED);
5099

5100 5101
		/* If device is running, close it first. */
		dev_close(dev);
5102

5103 5104
		/* And unlink it from device chain. */
		unlist_netdevice(dev);
5105

5106 5107
		dev->reg_state = NETREG_UNREGISTERING;
	}
5108 5109 5110

	synchronize_net();

5111 5112 5113
	list_for_each_entry(dev, head, unreg_list) {
		/* Shutdown queueing discipline. */
		dev_shutdown(dev);
5114 5115


5116 5117 5118 5119
		/* Notify protocols, that we are about to destroy
		   this device. They should clean all the things.
		*/
		call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
5120

5121 5122 5123 5124
		if (!dev->rtnl_link_ops ||
		    dev->rtnl_link_state == RTNL_LINK_INITIALIZED)
			rtmsg_ifinfo(RTM_DELLINK, dev, ~0U);

5125 5126 5127 5128 5129
		/*
		 *	Flush the unicast and multicast chains
		 */
		dev_unicast_flush(dev);
		dev_addr_discard(dev);
5130

5131 5132
		if (dev->netdev_ops->ndo_uninit)
			dev->netdev_ops->ndo_uninit(dev);
5133

5134 5135
		/* Notifier chain MUST detach us from master device. */
		WARN_ON(dev->master);
5136

5137 5138 5139
		/* Remove entries from kobject tree */
		netdev_unregister_kobject(dev);
	}
5140

5141
	/* Process any work delayed until the end of the batch */
5142
	dev = list_first_entry(head, struct net_device, unreg_list);
5143
	call_netdevice_notifiers(NETDEV_UNREGISTER_BATCH, dev);
5144

5145
	synchronize_net();
5146

5147
	list_for_each_entry(dev, head, unreg_list)
5148 5149 5150 5151 5152 5153 5154 5155 5156
		dev_put(dev);
}

static void rollback_registered(struct net_device *dev)
{
	LIST_HEAD(single);

	list_add(&dev->unreg_list, &single);
	rollback_registered_many(&single);
5157 5158
}

5159 5160 5161
static void __netdev_init_queue_locks_one(struct net_device *dev,
					  struct netdev_queue *dev_queue,
					  void *_unused)
5162 5163
{
	spin_lock_init(&dev_queue->_xmit_lock);
5164
	netdev_set_xmit_lockdep_class(&dev_queue->_xmit_lock, dev->type);
5165 5166 5167 5168 5169
	dev_queue->xmit_lock_owner = -1;
}

static void netdev_init_queue_locks(struct net_device *dev)
{
5170 5171
	netdev_for_each_tx_queue(dev, __netdev_init_queue_locks_one, NULL);
	__netdev_init_queue_locks_one(dev, &dev->rx_queue, NULL);
5172 5173
}

5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186 5187 5188 5189 5190 5191 5192 5193 5194 5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213
unsigned long netdev_fix_features(unsigned long features, const char *name)
{
	/* Fix illegal SG+CSUM combinations. */
	if ((features & NETIF_F_SG) &&
	    !(features & NETIF_F_ALL_CSUM)) {
		if (name)
			printk(KERN_NOTICE "%s: Dropping NETIF_F_SG since no "
			       "checksum feature.\n", name);
		features &= ~NETIF_F_SG;
	}

	/* TSO requires that SG is present as well. */
	if ((features & NETIF_F_TSO) && !(features & NETIF_F_SG)) {
		if (name)
			printk(KERN_NOTICE "%s: Dropping NETIF_F_TSO since no "
			       "SG feature.\n", name);
		features &= ~NETIF_F_TSO;
	}

	if (features & NETIF_F_UFO) {
		if (!(features & NETIF_F_GEN_CSUM)) {
			if (name)
				printk(KERN_ERR "%s: Dropping NETIF_F_UFO "
				       "since no NETIF_F_HW_CSUM feature.\n",
				       name);
			features &= ~NETIF_F_UFO;
		}

		if (!(features & NETIF_F_SG)) {
			if (name)
				printk(KERN_ERR "%s: Dropping NETIF_F_UFO "
				       "since no NETIF_F_SG feature.\n", name);
			features &= ~NETIF_F_UFO;
		}
	}

	return features;
}
EXPORT_SYMBOL(netdev_fix_features);

5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240
/**
 *	netif_stacked_transfer_operstate -	transfer operstate
 *	@rootdev: the root or lower level device to transfer state from
 *	@dev: the device to transfer operstate to
 *
 *	Transfer operational state from root to device. This is normally
 *	called when a stacking relationship exists between the root
 *	device and the device(a leaf device).
 */
void netif_stacked_transfer_operstate(const struct net_device *rootdev,
					struct net_device *dev)
{
	if (rootdev->operstate == IF_OPER_DORMANT)
		netif_dormant_on(dev);
	else
		netif_dormant_off(dev);

	if (netif_carrier_ok(rootdev)) {
		if (!netif_carrier_ok(dev))
			netif_carrier_on(dev);
	} else {
		if (netif_carrier_ok(dev))
			netif_carrier_off(dev);
	}
}
EXPORT_SYMBOL(netif_stacked_transfer_operstate);

L
Linus Torvalds 已提交
5241 5242 5243 5244 5245 5246 5247 5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260
/**
 *	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)
{
	int ret;
5261
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
5262 5263 5264 5265

	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

5266 5267
	might_sleep();

L
Linus Torvalds 已提交
5268 5269
	/* When net_device's are persistent, this will be fatal. */
	BUG_ON(dev->reg_state != NETREG_UNINITIALIZED);
5270
	BUG_ON(!net);
L
Linus Torvalds 已提交
5271

5272
	spin_lock_init(&dev->addr_list_lock);
5273
	netdev_set_addr_lockdep_class(dev);
5274
	netdev_init_queue_locks(dev);
L
Linus Torvalds 已提交
5275 5276 5277

	dev->iflink = -1;

T
Tom Herbert 已提交
5278 5279 5280 5281 5282 5283 5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294
	if (!dev->num_rx_queues) {
		/*
		 * Allocate a single RX queue if driver never called
		 * alloc_netdev_mq
		 */

		dev->_rx = kzalloc(sizeof(struct netdev_rx_queue), GFP_KERNEL);
		if (!dev->_rx) {
			ret = -ENOMEM;
			goto out;
		}

		dev->_rx->first = dev->_rx;
		atomic_set(&dev->_rx->count, 1);
		dev->num_rx_queues = 1;
	}

L
Linus Torvalds 已提交
5295
	/* Init, if this function is available */
5296 5297
	if (dev->netdev_ops->ndo_init) {
		ret = dev->netdev_ops->ndo_init(dev);
L
Linus Torvalds 已提交
5298 5299 5300
		if (ret) {
			if (ret > 0)
				ret = -EIO;
5301
			goto out;
L
Linus Torvalds 已提交
5302 5303
		}
	}
5304

5305 5306
	ret = dev_get_valid_name(net, dev->name, dev->name, 0);
	if (ret)
5307
		goto err_uninit;
L
Linus Torvalds 已提交
5308

5309
	dev->ifindex = dev_new_index(net);
L
Linus Torvalds 已提交
5310 5311 5312
	if (dev->iflink == -1)
		dev->iflink = dev->ifindex;

5313 5314 5315 5316 5317 5318 5319 5320 5321 5322 5323 5324 5325 5326 5327
	/* 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);
	}

5328
	dev->features = netdev_fix_features(dev->features, dev->name);
L
Linus Torvalds 已提交
5329

5330 5331 5332 5333
	/* Enable software GSO if SG is supported. */
	if (dev->features & NETIF_F_SG)
		dev->features |= NETIF_F_GSO;

5334
	netdev_initialize_kobject(dev);
5335 5336 5337 5338 5339 5340

	ret = call_netdevice_notifiers(NETDEV_POST_INIT, dev);
	ret = notifier_to_errno(ret);
	if (ret)
		goto err_uninit;

5341
	ret = netdev_register_kobject(dev);
5342
	if (ret)
5343
		goto err_uninit;
5344 5345
	dev->reg_state = NETREG_REGISTERED;

L
Linus Torvalds 已提交
5346 5347 5348 5349 5350 5351 5352 5353 5354
	/*
	 *	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);
5355
	list_netdevice(dev);
L
Linus Torvalds 已提交
5356 5357

	/* Notify protocols, that a new device appeared. */
5358
	ret = call_netdevice_notifiers(NETDEV_REGISTER, dev);
5359
	ret = notifier_to_errno(ret);
5360 5361 5362 5363
	if (ret) {
		rollback_registered(dev);
		dev->reg_state = NETREG_UNREGISTERED;
	}
5364 5365 5366 5367
	/*
	 *	Prevent userspace races by waiting until the network
	 *	device is fully setup before sending notifications.
	 */
5368 5369 5370
	if (!dev->rtnl_link_ops ||
	    dev->rtnl_link_state == RTNL_LINK_INITIALIZED)
		rtmsg_ifinfo(RTM_NEWLINK, dev, ~0U);
L
Linus Torvalds 已提交
5371 5372 5373

out:
	return ret;
5374 5375

err_uninit:
5376 5377
	if (dev->netdev_ops->ndo_uninit)
		dev->netdev_ops->ndo_uninit(dev);
5378
	goto out;
L
Linus Torvalds 已提交
5379
}
E
Eric Dumazet 已提交
5380
EXPORT_SYMBOL(register_netdevice);
L
Linus Torvalds 已提交
5381

5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392 5393 5394 5395 5396 5397 5398 5399 5400 5401 5402 5403 5404 5405 5406 5407 5408 5409 5410 5411 5412 5413 5414 5415 5416 5417 5418 5419 5420
/**
 *	init_dummy_netdev	- init a dummy network device for NAPI
 *	@dev: device to init
 *
 *	This takes a network device structure and initialize the minimum
 *	amount of fields so it can be used to schedule NAPI polls without
 *	registering a full blown interface. This is to be used by drivers
 *	that need to tie several hardware interfaces to a single NAPI
 *	poll scheduler due to HW limitations.
 */
int init_dummy_netdev(struct net_device *dev)
{
	/* Clear everything. Note we don't initialize spinlocks
	 * are they aren't supposed to be taken by any of the
	 * NAPI code and this dummy netdev is supposed to be
	 * only ever used for NAPI polls
	 */
	memset(dev, 0, sizeof(struct net_device));

	/* make sure we BUG if trying to hit standard
	 * register/unregister code path
	 */
	dev->reg_state = NETREG_DUMMY;

	/* initialize the ref count */
	atomic_set(&dev->refcnt, 1);

	/* NAPI wants this */
	INIT_LIST_HEAD(&dev->napi_list);

	/* a dummy interface is started by default */
	set_bit(__LINK_STATE_PRESENT, &dev->state);
	set_bit(__LINK_STATE_START, &dev->state);

	return 0;
}
EXPORT_SYMBOL_GPL(init_dummy_netdev);


L
Linus Torvalds 已提交
5421 5422 5423 5424 5425 5426 5427 5428 5429
/**
 *	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.
 *
5430
 *	This is a wrapper around register_netdevice that takes the rtnl semaphore
L
Linus Torvalds 已提交
5431 5432 5433 5434 5435 5436 5437 5438 5439 5440 5441 5442 5443 5444 5445 5446 5447 5448
 *	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;
	}
5449

L
Linus Torvalds 已提交
5450 5451 5452 5453 5454 5455 5456 5457 5458 5459 5460 5461 5462 5463 5464 5465
	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
5466
 * call dev_put.
L
Linus Torvalds 已提交
5467 5468 5469 5470 5471
 */
static void netdev_wait_allrefs(struct net_device *dev)
{
	unsigned long rebroadcast_time, warning_time;

5472 5473
	linkwatch_forget_dev(dev);

L
Linus Torvalds 已提交
5474 5475 5476
	rebroadcast_time = warning_time = jiffies;
	while (atomic_read(&dev->refcnt) != 0) {
		if (time_after(jiffies, rebroadcast_time + 1 * HZ)) {
5477
			rtnl_lock();
L
Linus Torvalds 已提交
5478 5479

			/* Rebroadcast unregister notification */
5480
			call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
5481
			/* don't resend NETDEV_UNREGISTER_BATCH, _BATCH users
5482
			 * should have already handle it the first time */
L
Linus Torvalds 已提交
5483 5484 5485 5486 5487 5488 5489 5490 5491 5492 5493 5494

			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();
			}

5495
			__rtnl_unlock();
L
Linus Torvalds 已提交
5496 5497 5498 5499 5500 5501 5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 5519 5520 5521 5522 5523 5524 5525

			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
5526
 * We are invoked by rtnl_unlock().
L
Linus Torvalds 已提交
5527
 * This allows us to deal with problems:
5528
 * 1) We can delete sysfs objects which invoke hotplug
L
Linus Torvalds 已提交
5529 5530 5531
 *    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
5532 5533 5534
 *
 * 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
5535 5536 5537
 */
void netdev_run_todo(void)
{
5538
	struct list_head list;
L
Linus Torvalds 已提交
5539 5540

	/* Snapshot list, allow later requests */
5541
	list_replace_init(&net_todo_list, &list);
H
Herbert Xu 已提交
5542 5543

	__rtnl_unlock();
5544

L
Linus Torvalds 已提交
5545 5546
	while (!list_empty(&list)) {
		struct net_device *dev
5547
			= list_first_entry(&list, struct net_device, todo_list);
L
Linus Torvalds 已提交
5548 5549
		list_del(&dev->todo_list);

5550 5551 5552 5553 5554 5555
		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
5556

5557
		dev->reg_state = NETREG_UNREGISTERED;
L
Linus Torvalds 已提交
5558

5559 5560
		on_each_cpu(flush_backlog, dev, 1);

5561
		netdev_wait_allrefs(dev);
L
Linus Torvalds 已提交
5562

5563 5564
		/* paranoia */
		BUG_ON(atomic_read(&dev->refcnt));
5565 5566 5567
		WARN_ON(dev->ip_ptr);
		WARN_ON(dev->ip6_ptr);
		WARN_ON(dev->dn_ptr);
L
Linus Torvalds 已提交
5568

5569 5570
		if (dev->destructor)
			dev->destructor(dev);
5571 5572 5573

		/* Free network device */
		kobject_put(&dev->dev.kobj);
L
Linus Torvalds 已提交
5574 5575 5576
	}
}

5577 5578 5579 5580 5581 5582 5583 5584 5585 5586 5587 5588 5589 5590 5591 5592 5593 5594 5595 5596 5597 5598 5599 5600 5601 5602
/**
 *	dev_txq_stats_fold - fold tx_queues stats
 *	@dev: device to get statistics from
 *	@stats: struct net_device_stats to hold results
 */
void dev_txq_stats_fold(const struct net_device *dev,
			struct net_device_stats *stats)
{
	unsigned long tx_bytes = 0, tx_packets = 0, tx_dropped = 0;
	unsigned int i;
	struct netdev_queue *txq;

	for (i = 0; i < dev->num_tx_queues; i++) {
		txq = netdev_get_tx_queue(dev, i);
		tx_bytes   += txq->tx_bytes;
		tx_packets += txq->tx_packets;
		tx_dropped += txq->tx_dropped;
	}
	if (tx_bytes || tx_packets || tx_dropped) {
		stats->tx_bytes   = tx_bytes;
		stats->tx_packets = tx_packets;
		stats->tx_dropped = tx_dropped;
	}
}
EXPORT_SYMBOL(dev_txq_stats_fold);

5603 5604 5605 5606 5607 5608 5609 5610 5611
/**
 *	dev_get_stats	- get network device statistics
 *	@dev: device to get statistics from
 *
 *	Get network statistics from device. The device driver may provide
 *	its own method by setting dev->netdev_ops->get_stats; otherwise
 *	the internal statistics structure is used.
 */
const struct net_device_stats *dev_get_stats(struct net_device *dev)
5612
{
5613 5614 5615 5616
	const struct net_device_ops *ops = dev->netdev_ops;

	if (ops->ndo_get_stats)
		return ops->ndo_get_stats(dev);
5617 5618 5619

	dev_txq_stats_fold(dev, &dev->stats);
	return &dev->stats;
R
Rusty Russell 已提交
5620
}
5621
EXPORT_SYMBOL(dev_get_stats);
R
Rusty Russell 已提交
5622

5623
static void netdev_init_one_queue(struct net_device *dev,
5624 5625
				  struct netdev_queue *queue,
				  void *_unused)
5626 5627 5628 5629
{
	queue->dev = dev;
}

5630 5631
static void netdev_init_queues(struct net_device *dev)
{
5632 5633
	netdev_init_one_queue(dev, &dev->rx_queue, NULL);
	netdev_for_each_tx_queue(dev, netdev_init_one_queue, NULL);
5634
	spin_lock_init(&dev->tx_global_lock);
5635 5636
}

L
Linus Torvalds 已提交
5637
/**
5638
 *	alloc_netdev_mq - allocate network device
L
Linus Torvalds 已提交
5639 5640 5641
 *	@sizeof_priv:	size of private data to allocate space for
 *	@name:		device name format string
 *	@setup:		callback to initialize device
5642
 *	@queue_count:	the number of subqueues to allocate
L
Linus Torvalds 已提交
5643 5644
 *
 *	Allocates a struct net_device with private data area for driver use
5645 5646
 *	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
5647
 */
5648 5649
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
5650
{
5651
	struct netdev_queue *tx;
T
Tom Herbert 已提交
5652
	struct netdev_rx_queue *rx;
L
Linus Torvalds 已提交
5653
	struct net_device *dev;
5654
	size_t alloc_size;
5655
	struct net_device *p;
T
Tom Herbert 已提交
5656
	int i;
L
Linus Torvalds 已提交
5657

5658 5659
	BUG_ON(strlen(name) >= sizeof(dev->name));

5660
	alloc_size = sizeof(struct net_device);
5661 5662
	if (sizeof_priv) {
		/* ensure 32-byte alignment of private area */
5663
		alloc_size = ALIGN(alloc_size, NETDEV_ALIGN);
5664 5665 5666
		alloc_size += sizeof_priv;
	}
	/* ensure 32-byte alignment of whole construct */
5667
	alloc_size += NETDEV_ALIGN - 1;
L
Linus Torvalds 已提交
5668

5669
	p = kzalloc(alloc_size, GFP_KERNEL);
L
Linus Torvalds 已提交
5670
	if (!p) {
5671
		printk(KERN_ERR "alloc_netdev: Unable to allocate device.\n");
L
Linus Torvalds 已提交
5672 5673 5674
		return NULL;
	}

5675
	tx = kcalloc(queue_count, sizeof(struct netdev_queue), GFP_KERNEL);
5676 5677 5678
	if (!tx) {
		printk(KERN_ERR "alloc_netdev: Unable to allocate "
		       "tx qdiscs.\n");
5679
		goto free_p;
5680 5681
	}

T
Tom Herbert 已提交
5682 5683 5684 5685 5686 5687 5688 5689 5690 5691 5692 5693 5694 5695 5696 5697
	rx = kcalloc(queue_count, sizeof(struct netdev_rx_queue), GFP_KERNEL);
	if (!rx) {
		printk(KERN_ERR "alloc_netdev: Unable to allocate "
		       "rx queues.\n");
		goto free_tx;
	}

	atomic_set(&rx->count, queue_count);

	/*
	 * Set a pointer to first element in the array which holds the
	 * reference count.
	 */
	for (i = 0; i < queue_count; i++)
		rx[i].first = rx;

5698
	dev = PTR_ALIGN(p, NETDEV_ALIGN);
L
Linus Torvalds 已提交
5699
	dev->padded = (char *)dev - (char *)p;
5700 5701

	if (dev_addr_init(dev))
T
Tom Herbert 已提交
5702
		goto free_rx;
5703

J
Jiri Pirko 已提交
5704 5705
	dev_unicast_init(dev);

5706
	dev_net_set(dev, &init_net);
L
Linus Torvalds 已提交
5707

5708 5709
	dev->_tx = tx;
	dev->num_tx_queues = queue_count;
5710
	dev->real_num_tx_queues = queue_count;
5711

T
Tom Herbert 已提交
5712 5713 5714
	dev->_rx = rx;
	dev->num_rx_queues = queue_count;

5715
	dev->gso_max_size = GSO_MAX_SIZE;
L
Linus Torvalds 已提交
5716

5717 5718
	netdev_init_queues(dev);

5719 5720
	INIT_LIST_HEAD(&dev->ethtool_ntuple_list.list);
	dev->ethtool_ntuple_list.count = 0;
5721
	INIT_LIST_HEAD(&dev->napi_list);
5722
	INIT_LIST_HEAD(&dev->unreg_list);
5723
	INIT_LIST_HEAD(&dev->link_watch_list);
5724
	dev->priv_flags = IFF_XMIT_DST_RELEASE;
L
Linus Torvalds 已提交
5725 5726 5727
	setup(dev);
	strcpy(dev->name, name);
	return dev;
5728

T
Tom Herbert 已提交
5729 5730
free_rx:
	kfree(rx);
5731 5732 5733 5734 5735
free_tx:
	kfree(tx);
free_p:
	kfree(p);
	return NULL;
L
Linus Torvalds 已提交
5736
}
5737
EXPORT_SYMBOL(alloc_netdev_mq);
L
Linus Torvalds 已提交
5738 5739 5740 5741 5742

/**
 *	free_netdev - free network device
 *	@dev: device
 *
5743 5744
 *	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
5745 5746 5747 5748
 *	If this is the last reference then it will be freed.
 */
void free_netdev(struct net_device *dev)
{
5749 5750
	struct napi_struct *p, *n;

5751 5752
	release_net(dev_net(dev));

5753 5754
	kfree(dev->_tx);

5755 5756 5757
	/* Flush device addresses */
	dev_addr_flush(dev);

5758 5759 5760
	/* Clear ethtool n-tuple list */
	ethtool_ntuple_flush(dev);

5761 5762 5763
	list_for_each_entry_safe(p, n, &dev->napi_list, dev_list)
		netif_napi_del(p);

S
Stephen Hemminger 已提交
5764
	/*  Compatibility with error handling in drivers */
L
Linus Torvalds 已提交
5765 5766 5767 5768 5769 5770 5771 5772
	if (dev->reg_state == NETREG_UNINITIALIZED) {
		kfree((char *)dev - dev->padded);
		return;
	}

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

5773 5774
	/* will free via device release */
	put_device(&dev->dev);
L
Linus Torvalds 已提交
5775
}
E
Eric Dumazet 已提交
5776
EXPORT_SYMBOL(free_netdev);
5777

5778 5779 5780 5781 5782 5783
/**
 *	synchronize_net -  Synchronize with packet receive processing
 *
 *	Wait for packets currently being received to be done.
 *	Does not block later packets from starting.
 */
5784
void synchronize_net(void)
L
Linus Torvalds 已提交
5785 5786
{
	might_sleep();
5787
	synchronize_rcu();
L
Linus Torvalds 已提交
5788
}
E
Eric Dumazet 已提交
5789
EXPORT_SYMBOL(synchronize_net);
L
Linus Torvalds 已提交
5790 5791

/**
5792
 *	unregister_netdevice_queue - remove device from the kernel
L
Linus Torvalds 已提交
5793
 *	@dev: device
5794
 *	@head: list
5795
 *
L
Linus Torvalds 已提交
5796
 *	This function shuts down a device interface and removes it
5797
 *	from the kernel tables.
5798
 *	If head not NULL, device is queued to be unregistered later.
L
Linus Torvalds 已提交
5799 5800 5801 5802 5803
 *
 *	Callers must hold the rtnl semaphore.  You may want
 *	unregister_netdev() instead of this.
 */

5804
void unregister_netdevice_queue(struct net_device *dev, struct list_head *head)
L
Linus Torvalds 已提交
5805
{
5806 5807
	ASSERT_RTNL();

5808
	if (head) {
5809
		list_move_tail(&dev->unreg_list, head);
5810 5811 5812 5813 5814
	} else {
		rollback_registered(dev);
		/* Finish processing unregister after unlock */
		net_set_todo(dev);
	}
L
Linus Torvalds 已提交
5815
}
5816
EXPORT_SYMBOL(unregister_netdevice_queue);
L
Linus Torvalds 已提交
5817

5818 5819 5820 5821 5822 5823 5824 5825 5826 5827 5828 5829 5830 5831
/**
 *	unregister_netdevice_many - unregister many devices
 *	@head: list of devices
 */
void unregister_netdevice_many(struct list_head *head)
{
	struct net_device *dev;

	if (!list_empty(head)) {
		rollback_registered_many(head);
		list_for_each_entry(dev, head, unreg_list)
			net_set_todo(dev);
	}
}
5832
EXPORT_SYMBOL(unregister_netdevice_many);
5833

L
Linus Torvalds 已提交
5834 5835 5836 5837 5838
/**
 *	unregister_netdev - remove device from the kernel
 *	@dev: device
 *
 *	This function shuts down a device interface and removes it
5839
 *	from the kernel tables.
L
Linus Torvalds 已提交
5840 5841 5842 5843 5844 5845 5846 5847 5848 5849 5850 5851 5852
 *
 *	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);

5853 5854 5855 5856 5857 5858 5859 5860 5861 5862 5863 5864 5865 5866 5867 5868 5869 5870 5871 5872 5873 5874 5875 5876 5877
/**
 *	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)
{
	int err;

	ASSERT_RTNL();

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

5878 5879 5880 5881 5882 5883 5884 5885 5886
#ifdef CONFIG_SYSFS
	/* Don't allow real devices to be moved when sysfs
	 * is enabled.
	 */
	err = -EINVAL;
	if (dev->dev.parent)
		goto out;
#endif

5887 5888 5889 5890 5891 5892 5893
	/* 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;
5894
	if (net_eq(dev_net(dev), net))
5895 5896 5897 5898 5899 5900
		goto out;

	/* Pick the destination device name, and ensure
	 * we can use it in the destination network namespace.
	 */
	err = -EEXIST;
5901
	if (__dev_get_by_name(net, dev->name)) {
5902 5903 5904
		/* We get here if we can't use the current device name */
		if (!pat)
			goto out;
5905
		if (dev_get_valid_name(net, pat, dev->name, 1))
5906 5907 5908 5909 5910 5911 5912 5913
			goto out;
	}

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

	/* If device is running close it first. */
5914
	dev_close(dev);
5915 5916 5917 5918 5919 5920 5921 5922 5923 5924 5925 5926 5927 5928

	/* 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);
5929
	call_netdevice_notifiers(NETDEV_UNREGISTER_BATCH, dev);
5930 5931 5932 5933

	/*
	 *	Flush the unicast and multicast chains
	 */
J
Jiri Pirko 已提交
5934
	dev_unicast_flush(dev);
5935 5936
	dev_addr_discard(dev);

5937 5938
	netdev_unregister_kobject(dev);

5939
	/* Actually switch the network namespace */
5940
	dev_net_set(dev, net);
5941 5942 5943 5944 5945 5946 5947 5948 5949

	/* 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;
	}

5950
	/* Fixup kobjects */
5951
	err = netdev_register_kobject(dev);
5952
	WARN_ON(err);
5953 5954 5955 5956 5957 5958 5959

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

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

5960 5961 5962 5963 5964 5965
	/*
	 *	Prevent userspace races by waiting until the network
	 *	device is fully setup before sending notifications.
	 */
	rtmsg_ifinfo(RTM_NEWLINK, dev, ~0U);

5966 5967 5968 5969 5970
	synchronize_net();
	err = 0;
out:
	return err;
}
5971
EXPORT_SYMBOL_GPL(dev_change_net_namespace);
5972

L
Linus Torvalds 已提交
5973 5974 5975 5976 5977
static int dev_cpu_callback(struct notifier_block *nfb,
			    unsigned long action,
			    void *ocpu)
{
	struct sk_buff **list_skb;
5978
	struct Qdisc **list_net;
L
Linus Torvalds 已提交
5979 5980 5981 5982
	struct sk_buff *skb;
	unsigned int cpu, oldcpu = (unsigned long)ocpu;
	struct softnet_data *sd, *oldsd;

5983
	if (action != CPU_DEAD && action != CPU_DEAD_FROZEN)
L
Linus Torvalds 已提交
5984 5985 5986 5987 5988 5989 5990 5991 5992 5993 5994 5995 5996 5997 5998 5999 6000 6001 6002 6003 6004 6005 6006 6007 6008 6009 6010 6011 6012 6013 6014 6015 6016 6017
		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;
}


6018
/**
6019 6020 6021 6022
 *	netdev_increment_features - increment feature set by one
 *	@all: current feature set
 *	@one: new feature set
 *	@mask: mask feature set
6023 6024
 *
 *	Computes a new feature set after adding a device with feature set
6025 6026
 *	@one to the master device with current feature set @all.  Will not
 *	enable anything that is off in @mask. Returns the new feature set.
6027
 */
6028 6029 6030 6031
unsigned long netdev_increment_features(unsigned long all, unsigned long one,
					unsigned long mask)
{
	/* If device needs checksumming, downgrade to it. */
E
Eric Dumazet 已提交
6032
	if (all & NETIF_F_NO_CSUM && !(one & NETIF_F_NO_CSUM))
6033 6034 6035 6036 6037 6038 6039 6040
		all ^= NETIF_F_NO_CSUM | (one & NETIF_F_ALL_CSUM);
	else if (mask & NETIF_F_ALL_CSUM) {
		/* If one device supports v4/v6 checksumming, set for all. */
		if (one & (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM) &&
		    !(all & NETIF_F_GEN_CSUM)) {
			all &= ~NETIF_F_ALL_CSUM;
			all |= one & (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM);
		}
6041

6042 6043 6044 6045 6046 6047
		/* If one device supports hw checksumming, set for all. */
		if (one & NETIF_F_GEN_CSUM && !(all & NETIF_F_GEN_CSUM)) {
			all &= ~NETIF_F_ALL_CSUM;
			all |= NETIF_F_HW_CSUM;
		}
	}
6048

6049
	one |= NETIF_F_ALL_CSUM;
6050

6051
	one |= all & NETIF_F_ONE_FOR_ALL;
6052
	all &= one | NETIF_F_LLTX | NETIF_F_GSO | NETIF_F_UFO;
6053
	all |= one & mask & NETIF_F_ONE_FOR_ALL;
6054 6055 6056

	return all;
}
6057
EXPORT_SYMBOL(netdev_increment_features);
6058

6059 6060 6061 6062 6063 6064 6065 6066 6067 6068 6069 6070 6071
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;
}

6072
/* Initialize per network namespace state */
6073
static int __net_init netdev_init(struct net *net)
6074 6075 6076
{
	INIT_LIST_HEAD(&net->dev_base_head);

6077 6078 6079
	net->dev_name_head = netdev_create_hash();
	if (net->dev_name_head == NULL)
		goto err_name;
6080

6081 6082 6083
	net->dev_index_head = netdev_create_hash();
	if (net->dev_index_head == NULL)
		goto err_idx;
6084 6085

	return 0;
6086 6087 6088 6089 6090

err_idx:
	kfree(net->dev_name_head);
err_name:
	return -ENOMEM;
6091 6092
}

6093 6094 6095 6096 6097 6098 6099 6100
/**
 *	netdev_drivername - network driver for the device
 *	@dev: network device
 *	@buffer: buffer for resulting name
 *	@len: size of buffer
 *
 *	Determine network driver for device.
 */
6101
char *netdev_drivername(const struct net_device *dev, char *buffer, int len)
6102
{
6103 6104
	const struct device_driver *driver;
	const struct device *parent;
6105 6106 6107 6108 6109 6110 6111 6112 6113 6114 6115 6116 6117 6118 6119 6120

	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;
}

6121
static void __net_exit netdev_exit(struct net *net)
6122 6123 6124 6125 6126
{
	kfree(net->dev_name_head);
	kfree(net->dev_index_head);
}

6127
static struct pernet_operations __net_initdata netdev_net_ops = {
6128 6129 6130 6131
	.init = netdev_init,
	.exit = netdev_exit,
};

6132
static void __net_exit default_device_exit(struct net *net)
6133
{
6134
	struct net_device *dev, *aux;
6135
	/*
6136
	 * Push all migratable network devices back to the
6137 6138 6139
	 * initial network namespace
	 */
	rtnl_lock();
6140
	for_each_netdev_safe(net, dev, aux) {
6141
		int err;
6142
		char fb_name[IFNAMSIZ];
6143 6144 6145 6146 6147

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

6148 6149 6150
		/* Leave virtual devices for the generic cleanup */
		if (dev->rtnl_link_ops)
			continue;
6151

6152
		/* Push remaing network devices to init_net */
6153 6154
		snprintf(fb_name, IFNAMSIZ, "dev%d", dev->ifindex);
		err = dev_change_net_namespace(dev, &init_net, fb_name);
6155
		if (err) {
6156
			printk(KERN_EMERG "%s: failed to move %s to init_net: %d\n",
6157
				__func__, dev->name, err);
6158
			BUG();
6159 6160 6161 6162 6163
		}
	}
	rtnl_unlock();
}

6164 6165 6166 6167 6168 6169 6170 6171 6172 6173 6174 6175 6176 6177 6178 6179 6180 6181 6182 6183 6184 6185 6186 6187
static void __net_exit default_device_exit_batch(struct list_head *net_list)
{
	/* At exit all network devices most be removed from a network
	 * namespace.  Do this in the reverse order of registeration.
	 * Do this across as many network namespaces as possible to
	 * improve batching efficiency.
	 */
	struct net_device *dev;
	struct net *net;
	LIST_HEAD(dev_kill_list);

	rtnl_lock();
	list_for_each_entry(net, net_list, exit_list) {
		for_each_netdev_reverse(net, dev) {
			if (dev->rtnl_link_ops)
				dev->rtnl_link_ops->dellink(dev, &dev_kill_list);
			else
				unregister_netdevice_queue(dev, &dev_kill_list);
		}
	}
	unregister_netdevice_many(&dev_kill_list);
	rtnl_unlock();
}

6188
static struct pernet_operations __net_initdata default_device_ops = {
6189
	.exit = default_device_exit,
6190
	.exit_batch = default_device_exit_batch,
6191 6192
};

L
Linus Torvalds 已提交
6193 6194 6195 6196 6197 6198 6199 6200 6201 6202 6203 6204 6205 6206 6207 6208 6209 6210 6211 6212
/*
 *	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;

6213
	if (netdev_kobject_init())
L
Linus Torvalds 已提交
6214 6215 6216
		goto out;

	INIT_LIST_HEAD(&ptype_all);
6217
	for (i = 0; i < PTYPE_HASH_SIZE; i++)
L
Linus Torvalds 已提交
6218 6219
		INIT_LIST_HEAD(&ptype_base[i]);

6220 6221
	if (register_pernet_subsys(&netdev_net_ops))
		goto out;
L
Linus Torvalds 已提交
6222 6223 6224 6225 6226

	/*
	 *	Initialise the packet receive queues.
	 */

6227
	for_each_possible_cpu(i) {
L
Linus Torvalds 已提交
6228 6229 6230 6231 6232 6233
		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);
6234

6235
#ifdef CONFIG_SMP
T
Tom Herbert 已提交
6236 6237 6238
		queue->csd.func = trigger_softirq;
		queue->csd.info = queue;
		queue->csd.flags = 0;
6239
#endif
T
Tom Herbert 已提交
6240

6241 6242
		queue->backlog.poll = process_backlog;
		queue->backlog.weight = weight_p;
6243
		queue->backlog.gro_list = NULL;
6244
		queue->backlog.gro_count = 0;
L
Linus Torvalds 已提交
6245 6246 6247 6248
	}

	dev_boot_phase = 0;

6249 6250 6251 6252 6253 6254 6255 6256 6257 6258 6259 6260 6261 6262 6263
	/* The loopback device is special if any other network devices
	 * is present in a network namespace the loopback device must
	 * be present. Since we now dynamically allocate and free the
	 * loopback device ensure this invariant is maintained by
	 * keeping the loopback device as the first device on the
	 * list of network devices.  Ensuring the loopback devices
	 * is the first device that appears and the last network device
	 * that disappears.
	 */
	if (register_pernet_device(&loopback_net_ops))
		goto out;

	if (register_pernet_device(&default_device_ops))
		goto out;

6264 6265
	open_softirq(NET_TX_SOFTIRQ, net_tx_action);
	open_softirq(NET_RX_SOFTIRQ, net_rx_action);
L
Linus Torvalds 已提交
6266 6267 6268 6269 6270 6271 6272 6273 6274 6275 6276

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

subsys_initcall(net_dev_init);

6277 6278
static int __init initialize_hashrnd(void)
{
T
Tom Herbert 已提交
6279
	get_random_bytes(&hashrnd, sizeof(hashrnd));
6280 6281 6282 6283 6284
	return 0;
}

late_initcall_sync(initialize_hashrnd);