dev.c 160.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9
/*
 * 	NET3	Protocol independent device support routines.
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or (at your option) any later version.
 *
 *	Derived from the non IP parts of dev.c 1.0.19
10
 * 		Authors:	Ross Biro
L
Linus Torvalds 已提交
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76
 *				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 <linux/bitops.h>
77
#include <linux/capability.h>
L
Linus Torvalds 已提交
78 79 80
#include <linux/cpu.h>
#include <linux/types.h>
#include <linux/kernel.h>
81
#include <linux/hash.h>
82
#include <linux/slab.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 99 100 101 102 103 104 105
#include <net/sock.h>
#include <linux/rtnetlink.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#include <linux/stat.h>
#include <net/dst.h>
#include <net/pkt_sched.h>
#include <net/checksum.h>
106
#include <net/xfrm.h>
L
Linus Torvalds 已提交
107 108 109 110 111 112 113
#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>
114
#include <net/wext.h>
L
Linus Torvalds 已提交
115 116
#include <net/iw_handler.h>
#include <asm/current.h>
S
Steve Grubb 已提交
117
#include <linux/audit.h>
118
#include <linux/dmaengine.h>
119
#include <linux/err.h>
120
#include <linux/ctype.h>
121
#include <linux/if_arp.h>
122
#include <linux/if_vlan.h>
123
#include <linux/ip.h>
124
#include <net/ip.h>
125 126
#include <linux/ipv6.h>
#include <linux/in.h>
D
David S. Miller 已提交
127 128
#include <linux/jhash.h>
#include <linux/random.h>
129
#include <trace/events/napi.h>
130
#include <trace/events/net.h>
131
#include <trace/events/skb.h>
132
#include <linux/pci.h>
133
#include <linux/inetdevice.h>
134
#include <linux/cpu_rmap.h>
135
#include <linux/net_tstamp.h>
136
#include <linux/static_key.h>
137
#include <net/flow_keys.h>
L
Linus Torvalds 已提交
138

139 140
#include "net-sysfs.h"

141 142 143
/* Instead of increasing this, you should create a hash table. */
#define MAX_GRO_SKBS 8

H
Herbert Xu 已提交
144 145 146
/* 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
147 148 149 150 151 152 153 154 155 156 157
/*
 *	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
158
 *             the average user (w/out VLANs) will not be adversely affected.
L
Linus Torvalds 已提交
159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174
 *             --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
 */

175 176 177
#define PTYPE_HASH_SIZE	(16)
#define PTYPE_HASH_MASK	(PTYPE_HASH_SIZE - 1)

L
Linus Torvalds 已提交
178
static DEFINE_SPINLOCK(ptype_lock);
179
static struct list_head ptype_base[PTYPE_HASH_SIZE] __read_mostly;
180
static struct list_head ptype_all __read_mostly;	/* Taps */
L
Linus Torvalds 已提交
181 182

/*
183
 * The @dev_base_head list is protected by @dev_base_lock and the rtnl
L
Linus Torvalds 已提交
184 185
 * semaphore.
 *
186
 * Pure readers hold dev_base_lock for reading, or rcu_read_lock()
L
Linus Torvalds 已提交
187 188
 *
 * Writers must hold the rtnl semaphore while they loop through the
189
 * dev_base_head list, and hold dev_base_lock for writing when they do the
L
Linus Torvalds 已提交
190 191 192 193 194 195 196 197 198 199 200 201 202 203
 * 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);

204 205 206 207 208
static inline void dev_base_seq_inc(struct net *net)
{
	while (++net->dev_base_seq == 0);
}

209
static inline struct hlist_head *dev_name_hash(struct net *net, const char *name)
L
Linus Torvalds 已提交
210
{
211 212
	unsigned int hash = full_name_hash(name, strnlen(name, IFNAMSIZ));

213
	return &net->dev_name_head[hash_32(hash, NETDEV_HASHBITS)];
L
Linus Torvalds 已提交
214 215
}

216
static inline struct hlist_head *dev_index_hash(struct net *net, int ifindex)
L
Linus Torvalds 已提交
217
{
218
	return &net->dev_index_head[ifindex & (NETDEV_HASHENTRIES - 1)];
L
Linus Torvalds 已提交
219 220
}

E
Eric Dumazet 已提交
221
static inline void rps_lock(struct softnet_data *sd)
222 223
{
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
224
	spin_lock(&sd->input_pkt_queue.lock);
225 226 227
#endif
}

E
Eric Dumazet 已提交
228
static inline void rps_unlock(struct softnet_data *sd)
229 230
{
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
231
	spin_unlock(&sd->input_pkt_queue.lock);
232 233 234
#endif
}

235 236 237
/* Device list insertion */
static int list_netdevice(struct net_device *dev)
{
238
	struct net *net = dev_net(dev);
239 240 241 242

	ASSERT_RTNL();

	write_lock_bh(&dev_base_lock);
243
	list_add_tail_rcu(&dev->dev_list, &net->dev_base_head);
244
	hlist_add_head_rcu(&dev->name_hlist, dev_name_hash(net, dev->name));
245 246
	hlist_add_head_rcu(&dev->index_hlist,
			   dev_index_hash(net, dev->ifindex));
247
	write_unlock_bh(&dev_base_lock);
248 249 250

	dev_base_seq_inc(net);

251 252 253
	return 0;
}

254 255 256
/* Device list removal
 * caller must respect a RCU grace period before freeing/reusing dev
 */
257 258 259 260 261 262
static void unlist_netdevice(struct net_device *dev)
{
	ASSERT_RTNL();

	/* Unlink dev from the device chain */
	write_lock_bh(&dev_base_lock);
263
	list_del_rcu(&dev->dev_list);
264
	hlist_del_rcu(&dev->name_hlist);
265
	hlist_del_rcu(&dev->index_hlist);
266
	write_unlock_bh(&dev_base_lock);
267 268

	dev_base_seq_inc(dev_net(dev));
269 270
}

L
Linus Torvalds 已提交
271 272 273 274
/*
 *	Our notifier list
 */

275
static RAW_NOTIFIER_HEAD(netdev_chain);
L
Linus Torvalds 已提交
276 277 278 279 280

/*
 *	Device drivers call our routines to queue packets here. We empty the
 *	queue in the local softnet handler.
 */
281

282
DEFINE_PER_CPU_ALIGNED(struct softnet_data, softnet_data);
E
Eric Dumazet 已提交
283
EXPORT_PER_CPU_SYMBOL(softnet_data);
L
Linus Torvalds 已提交
284

285
#ifdef CONFIG_LOCKDEP
286
/*
287
 * register_netdevice() inits txq->_xmit_lock and sets lockdep class
288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303
 * 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,
304
	 ARPHRD_IEEE80211_PRISM, ARPHRD_IEEE80211_RADIOTAP, ARPHRD_PHONET,
305
	 ARPHRD_PHONET_PIPE, ARPHRD_IEEE802154,
306
	 ARPHRD_VOID, ARPHRD_NONE};
307

308
static const char *const netdev_lock_name[] =
309 310 311 312 313 314 315 316 317 318 319 320 321
	{"_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",
322
	 "_xmit_IEEE80211_PRISM", "_xmit_IEEE80211_RADIOTAP", "_xmit_PHONET",
323
	 "_xmit_PHONET_PIPE", "_xmit_IEEE802154",
324
	 "_xmit_VOID", "_xmit_NONE"};
325 326

static struct lock_class_key netdev_xmit_lock_key[ARRAY_SIZE(netdev_lock_type)];
327
static struct lock_class_key netdev_addr_lock_key[ARRAY_SIZE(netdev_lock_type)];
328 329 330 331 332 333 334 335 336 337 338 339

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;
}

340 341
static inline void netdev_set_xmit_lockdep_class(spinlock_t *lock,
						 unsigned short dev_type)
342 343 344 345 346 347 348
{
	int i;

	i = netdev_lock_pos(dev_type);
	lockdep_set_class_and_name(lock, &netdev_xmit_lock_key[i],
				   netdev_lock_name[i]);
}
349 350 351 352 353 354 355 356 357 358

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]);
}
359
#else
360 361 362 363 364
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)
365 366 367
{
}
#endif
L
Linus Torvalds 已提交
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390

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

		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)
 */

391 392 393 394 395 396 397 398
static inline struct list_head *ptype_head(const struct packet_type *pt)
{
	if (pt->type == htons(ETH_P_ALL))
		return &ptype_all;
	else
		return &ptype_base[ntohs(pt->type) & PTYPE_HASH_MASK];
}

L
Linus Torvalds 已提交
399 400 401 402 403 404 405 406
/**
 *	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.
 *
407
 *	This call does not sleep therefore it can not
L
Linus Torvalds 已提交
408 409 410 411 412 413
 *	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)
{
414
	struct list_head *head = ptype_head(pt);
L
Linus Torvalds 已提交
415

416 417 418
	spin_lock(&ptype_lock);
	list_add_rcu(&pt->list, head);
	spin_unlock(&ptype_lock);
L
Linus Torvalds 已提交
419
}
E
Eric Dumazet 已提交
420
EXPORT_SYMBOL(dev_add_pack);
L
Linus Torvalds 已提交
421 422 423 424 425 426 427 428

/**
 *	__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
429
 *	returns.
L
Linus Torvalds 已提交
430 431 432 433 434 435 436
 *
 *      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)
{
437
	struct list_head *head = ptype_head(pt);
L
Linus Torvalds 已提交
438 439
	struct packet_type *pt1;

440
	spin_lock(&ptype_lock);
L
Linus Torvalds 已提交
441 442 443 444 445 446 447 448

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

449
	pr_warn("dev_remove_pack: %p not found\n", pt);
L
Linus Torvalds 已提交
450
out:
451
	spin_unlock(&ptype_lock);
L
Linus Torvalds 已提交
452
}
E
Eric Dumazet 已提交
453 454
EXPORT_SYMBOL(__dev_remove_pack);

L
Linus Torvalds 已提交
455 456 457 458 459 460 461 462 463 464 465 466 467 468 469
/**
 *	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);
470

L
Linus Torvalds 已提交
471 472
	synchronize_net();
}
E
Eric Dumazet 已提交
473
EXPORT_SYMBOL(dev_remove_pack);
L
Linus Torvalds 已提交
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501

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

		      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));
502
			strlcpy(s[i].name, name, IFNAMSIZ);
L
Linus Torvalds 已提交
503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526
			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] != ' ' &&
527
		    !strcmp(dev->name, s[i].name)) {
L
Linus Torvalds 已提交
528 529 530 531 532 533 534 535 536
			dev->irq 	= s[i].map.irq;
			dev->base_addr 	= s[i].map.base_addr;
			dev->mem_start 	= s[i].map.mem_start;
			dev->mem_end 	= s[i].map.mem_end;
			return 1;
		}
	}
	return 0;
}
E
Eric Dumazet 已提交
537
EXPORT_SYMBOL(netdev_boot_setup_check);
L
Linus Torvalds 已提交
538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561


/**
 *	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
	 */
562
	if (__dev_get_by_name(&init_net, name))
L
Linus Torvalds 已提交
563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607
		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
608
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
609 610 611 612 613 614 615 616 617
 *	@name: name to find
 *
 *	Find an interface by name. 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.
 */

618
struct net_device *__dev_get_by_name(struct net *net, const char *name)
L
Linus Torvalds 已提交
619 620
{
	struct hlist_node *p;
E
Eric Dumazet 已提交
621 622
	struct net_device *dev;
	struct hlist_head *head = dev_name_hash(net, name);
L
Linus Torvalds 已提交
623

E
Eric Dumazet 已提交
624
	hlist_for_each_entry(dev, p, head, name_hlist)
L
Linus Torvalds 已提交
625 626
		if (!strncmp(dev->name, name, IFNAMSIZ))
			return dev;
E
Eric Dumazet 已提交
627

L
Linus Torvalds 已提交
628 629
	return NULL;
}
E
Eric Dumazet 已提交
630
EXPORT_SYMBOL(__dev_get_by_name);
L
Linus Torvalds 已提交
631

632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657
/**
 *	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
658 659
/**
 *	dev_get_by_name		- find a device by its name
660
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
661 662 663 664 665 666 667 668 669
 *	@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.
 */

670
struct net_device *dev_get_by_name(struct net *net, const char *name)
L
Linus Torvalds 已提交
671 672 673
{
	struct net_device *dev;

674 675
	rcu_read_lock();
	dev = dev_get_by_name_rcu(net, name);
L
Linus Torvalds 已提交
676 677
	if (dev)
		dev_hold(dev);
678
	rcu_read_unlock();
L
Linus Torvalds 已提交
679 680
	return dev;
}
E
Eric Dumazet 已提交
681
EXPORT_SYMBOL(dev_get_by_name);
L
Linus Torvalds 已提交
682 683 684

/**
 *	__dev_get_by_index - find a device by its ifindex
685
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
686 687 688 689 690 691 692 693 694
 *	@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.
 */

695
struct net_device *__dev_get_by_index(struct net *net, int ifindex)
L
Linus Torvalds 已提交
696 697
{
	struct hlist_node *p;
E
Eric Dumazet 已提交
698 699
	struct net_device *dev;
	struct hlist_head *head = dev_index_hash(net, ifindex);
L
Linus Torvalds 已提交
700

E
Eric Dumazet 已提交
701
	hlist_for_each_entry(dev, p, head, index_hlist)
L
Linus Torvalds 已提交
702 703
		if (dev->ifindex == ifindex)
			return dev;
E
Eric Dumazet 已提交
704

L
Linus Torvalds 已提交
705 706
	return NULL;
}
E
Eric Dumazet 已提交
707
EXPORT_SYMBOL(__dev_get_by_index);
L
Linus Torvalds 已提交
708

709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733
/**
 *	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
734 735 736

/**
 *	dev_get_by_index - find a device by its ifindex
737
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
738 739 740 741 742 743 744 745
 *	@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.
 */

746
struct net_device *dev_get_by_index(struct net *net, int ifindex)
L
Linus Torvalds 已提交
747 748 749
{
	struct net_device *dev;

750 751
	rcu_read_lock();
	dev = dev_get_by_index_rcu(net, ifindex);
L
Linus Torvalds 已提交
752 753
	if (dev)
		dev_hold(dev);
754
	rcu_read_unlock();
L
Linus Torvalds 已提交
755 756
	return dev;
}
E
Eric Dumazet 已提交
757
EXPORT_SYMBOL(dev_get_by_index);
L
Linus Torvalds 已提交
758 759

/**
760
 *	dev_getbyhwaddr_rcu - find a device by its hardware address
761
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
762 763 764 765
 *	@type: media type of device
 *	@ha: hardware address
 *
 *	Search for an interface by MAC address. Returns NULL if the device
E
Eric Dumazet 已提交
766 767
 *	is not found or a pointer to the device.
 *	The caller must hold RCU or RTNL.
768
 *	The returned device has not had its ref count increased
L
Linus Torvalds 已提交
769 770 771 772
 *	and the caller must therefore be careful about locking
 *
 */

773 774
struct net_device *dev_getbyhwaddr_rcu(struct net *net, unsigned short type,
				       const char *ha)
L
Linus Torvalds 已提交
775 776 777
{
	struct net_device *dev;

778
	for_each_netdev_rcu(net, dev)
L
Linus Torvalds 已提交
779 780
		if (dev->type == type &&
		    !memcmp(dev->dev_addr, ha, dev->addr_len))
781 782 783
			return dev;

	return NULL;
L
Linus Torvalds 已提交
784
}
785
EXPORT_SYMBOL(dev_getbyhwaddr_rcu);
786

787
struct net_device *__dev_getfirstbyhwtype(struct net *net, unsigned short type)
L
Linus Torvalds 已提交
788 789 790
{
	struct net_device *dev;

791
	ASSERT_RTNL();
792
	for_each_netdev(net, dev)
793
		if (dev->type == type)
794 795 796
			return dev;

	return NULL;
797 798 799
}
EXPORT_SYMBOL(__dev_getfirstbyhwtype);

800
struct net_device *dev_getfirstbyhwtype(struct net *net, unsigned short type)
801
{
802
	struct net_device *dev, *ret = NULL;
803

804 805 806 807 808 809 810 811 812
	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
813 814 815 816
}
EXPORT_SYMBOL(dev_getfirstbyhwtype);

/**
E
Eric Dumazet 已提交
817
 *	dev_get_by_flags_rcu - find any device with given flags
818
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
819 820 821 822
 *	@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
E
Eric Dumazet 已提交
823 824
 *	is not found or a pointer to the device. Must be called inside
 *	rcu_read_lock(), and result refcount is unchanged.
L
Linus Torvalds 已提交
825 826
 */

E
Eric Dumazet 已提交
827
struct net_device *dev_get_by_flags_rcu(struct net *net, unsigned short if_flags,
E
Eric Dumazet 已提交
828
				    unsigned short mask)
L
Linus Torvalds 已提交
829
{
830
	struct net_device *dev, *ret;
L
Linus Torvalds 已提交
831

832
	ret = NULL;
833
	for_each_netdev_rcu(net, dev) {
L
Linus Torvalds 已提交
834
		if (((dev->flags ^ if_flags) & mask) == 0) {
835
			ret = dev;
L
Linus Torvalds 已提交
836 837 838
			break;
		}
	}
839
	return ret;
L
Linus Torvalds 已提交
840
}
E
Eric Dumazet 已提交
841
EXPORT_SYMBOL(dev_get_by_flags_rcu);
L
Linus Torvalds 已提交
842 843 844 845 846 847

/**
 *	dev_valid_name - check if name is okay for network device
 *	@name: name string
 *
 *	Network device names need to be valid file names to
848 849
 *	to allow sysfs to work.  We also disallow any kind of
 *	whitespace.
L
Linus Torvalds 已提交
850
 */
851
bool dev_valid_name(const char *name)
L
Linus Torvalds 已提交
852
{
853
	if (*name == '\0')
854
		return false;
855
	if (strlen(name) >= IFNAMSIZ)
856
		return false;
857
	if (!strcmp(name, ".") || !strcmp(name, ".."))
858
		return false;
859 860 861

	while (*name) {
		if (*name == '/' || isspace(*name))
862
			return false;
863 864
		name++;
	}
865
	return true;
L
Linus Torvalds 已提交
866
}
E
Eric Dumazet 已提交
867
EXPORT_SYMBOL(dev_valid_name);
L
Linus Torvalds 已提交
868 869

/**
870 871
 *	__dev_alloc_name - allocate a name for a device
 *	@net: network namespace to allocate the device name in
L
Linus Torvalds 已提交
872
 *	@name: name format string
873
 *	@buf:  scratch buffer and result name string
L
Linus Torvalds 已提交
874 875
 *
 *	Passed a format string - eg "lt%d" it will try and find a suitable
S
Stephen Hemminger 已提交
876 877 878 879 880 881
 *	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
882 883
 */

884
static int __dev_alloc_name(struct net *net, const char *name, char *buf)
L
Linus Torvalds 已提交
885 886 887 888
{
	int i = 0;
	const char *p;
	const int max_netdevices = 8*PAGE_SIZE;
S
Stephen Hemminger 已提交
889
	unsigned long *inuse;
L
Linus Torvalds 已提交
890 891 892 893 894 895 896 897 898 899 900 901 902
	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
903
		inuse = (unsigned long *) get_zeroed_page(GFP_ATOMIC);
L
Linus Torvalds 已提交
904 905 906
		if (!inuse)
			return -ENOMEM;

907
		for_each_netdev(net, d) {
L
Linus Torvalds 已提交
908 909 910 911 912 913
			if (!sscanf(d->name, name, &i))
				continue;
			if (i < 0 || i >= max_netdevices)
				continue;

			/*  avoid cases where sscanf is not exact inverse of printf */
914
			snprintf(buf, IFNAMSIZ, name, i);
L
Linus Torvalds 已提交
915 916 917 918 919 920 921 922
			if (!strncmp(buf, d->name, IFNAMSIZ))
				set_bit(i, inuse);
		}

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

923 924
	if (buf != name)
		snprintf(buf, IFNAMSIZ, name, i);
925
	if (!__dev_get_by_name(net, buf))
L
Linus Torvalds 已提交
926 927 928 929 930 931 932 933 934
		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;
}

935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954
/**
 *	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;

955 956
	BUG_ON(!dev_net(dev));
	net = dev_net(dev);
957 958 959 960 961
	ret = __dev_alloc_name(net, name, buf);
	if (ret >= 0)
		strlcpy(dev->name, buf, IFNAMSIZ);
	return ret;
}
E
Eric Dumazet 已提交
962
EXPORT_SYMBOL(dev_alloc_name);
963

964
static int dev_get_valid_name(struct net_device *dev, const char *name)
965
{
966 967 968 969 970
	struct net *net;

	BUG_ON(!dev_net(dev));
	net = dev_net(dev);

971 972 973
	if (!dev_valid_name(name))
		return -EINVAL;

974
	if (strchr(name, '%'))
975
		return dev_alloc_name(dev, name);
976 977
	else if (__dev_get_by_name(net, name))
		return -EEXIST;
978 979
	else if (dev->name != name)
		strlcpy(dev->name, name, IFNAMSIZ);
980 981 982

	return 0;
}
L
Linus Torvalds 已提交
983 984 985 986 987 988 989 990 991

/**
 *	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.
 */
992
int dev_change_name(struct net_device *dev, const char *newname)
L
Linus Torvalds 已提交
993
{
994
	char oldname[IFNAMSIZ];
L
Linus Torvalds 已提交
995
	int err = 0;
996
	int ret;
997
	struct net *net;
L
Linus Torvalds 已提交
998 999

	ASSERT_RTNL();
1000
	BUG_ON(!dev_net(dev));
L
Linus Torvalds 已提交
1001

1002
	net = dev_net(dev);
L
Linus Torvalds 已提交
1003 1004 1005
	if (dev->flags & IFF_UP)
		return -EBUSY;

1006 1007 1008
	if (strncmp(newname, dev->name, IFNAMSIZ) == 0)
		return 0;

1009 1010
	memcpy(oldname, dev->name, IFNAMSIZ);

1011
	err = dev_get_valid_name(dev, newname);
1012 1013
	if (err < 0)
		return err;
L
Linus Torvalds 已提交
1014

1015
rollback:
1016 1017 1018 1019
	ret = device_rename(&dev->dev, dev->name);
	if (ret) {
		memcpy(dev->name, oldname, IFNAMSIZ);
		return ret;
1020
	}
1021 1022

	write_lock_bh(&dev_base_lock);
1023
	hlist_del_rcu(&dev->name_hlist);
1024 1025 1026 1027 1028 1029
	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));
1030 1031
	write_unlock_bh(&dev_base_lock);

1032
	ret = call_netdevice_notifiers(NETDEV_CHANGENAME, dev);
1033 1034 1035
	ret = notifier_to_errno(ret);

	if (ret) {
1036 1037
		/* err >= 0 after dev_alloc_name() or stores the first errno */
		if (err >= 0) {
1038 1039 1040
			err = ret;
			memcpy(dev->name, oldname, IFNAMSIZ);
			goto rollback;
1041
		} else {
1042
			pr_err("%s: name change rollback failed: %d\n",
1043
			       dev->name, ret);
1044 1045
		}
	}
L
Linus Torvalds 已提交
1046 1047 1048 1049

	return err;
}

1050 1051 1052 1053
/**
 *	dev_set_alias - change ifalias of a device
 *	@dev: device
 *	@alias: name up to IFALIASZ
1054
 *	@len: limit of bytes to copy from info
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064
 *
 *	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;

1065 1066 1067 1068 1069 1070 1071 1072
	if (!len) {
		if (dev->ifalias) {
			kfree(dev->ifalias);
			dev->ifalias = NULL;
		}
		return 0;
	}

E
Eric Dumazet 已提交
1073
	dev->ifalias = krealloc(dev->ifalias, len + 1, GFP_KERNEL);
1074 1075 1076 1077 1078 1079 1080 1081
	if (!dev->ifalias)
		return -ENOMEM;

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


1082
/**
S
Stephen Hemminger 已提交
1083
 *	netdev_features_change - device changes features
1084 1085 1086 1087 1088 1089
 *	@dev: device to cause notification
 *
 *	Called to indicate a device has changed features.
 */
void netdev_features_change(struct net_device *dev)
{
1090
	call_netdevice_notifiers(NETDEV_FEAT_CHANGE, dev);
1091 1092 1093
}
EXPORT_SYMBOL(netdev_features_change);

L
Linus Torvalds 已提交
1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104
/**
 *	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) {
1105
		call_netdevice_notifiers(NETDEV_CHANGE, dev);
L
Linus Torvalds 已提交
1106 1107 1108
		rtmsg_ifinfo(RTM_NEWLINK, dev, 0);
	}
}
E
Eric Dumazet 已提交
1109
EXPORT_SYMBOL(netdev_state_change);
L
Linus Torvalds 已提交
1110

1111
int netdev_bonding_change(struct net_device *dev, unsigned long event)
1112
{
1113
	return call_netdevice_notifiers(event, dev);
1114 1115 1116
}
EXPORT_SYMBOL(netdev_bonding_change);

L
Linus Torvalds 已提交
1117 1118
/**
 *	dev_load 	- load a network module
1119
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
1120 1121 1122 1123 1124 1125 1126
 *	@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.
 */

1127
void dev_load(struct net *net, const char *name)
L
Linus Torvalds 已提交
1128
{
1129
	struct net_device *dev;
1130
	int no_module;
L
Linus Torvalds 已提交
1131

1132 1133 1134
	rcu_read_lock();
	dev = dev_get_by_name_rcu(net, name);
	rcu_read_unlock();
L
Linus Torvalds 已提交
1135

1136 1137 1138 1139 1140
	no_module = !dev;
	if (no_module && capable(CAP_NET_ADMIN))
		no_module = request_module("netdev-%s", name);
	if (no_module && capable(CAP_SYS_MODULE)) {
		if (!request_module("%s", name))
1141 1142
			pr_err("Loading kernel module for a network device with CAP_SYS_MODULE (deprecated).  Use CAP_NET_ADMIN and alias netdev-%s instead.\n",
			       name);
1143
	}
L
Linus Torvalds 已提交
1144
}
E
Eric Dumazet 已提交
1145
EXPORT_SYMBOL(dev_load);
L
Linus Torvalds 已提交
1146

1147
static int __dev_open(struct net_device *dev)
L
Linus Torvalds 已提交
1148
{
1149
	const struct net_device_ops *ops = dev->netdev_ops;
1150
	int ret;
L
Linus Torvalds 已提交
1151

1152 1153
	ASSERT_RTNL();

L
Linus Torvalds 已提交
1154 1155 1156
	if (!netif_device_present(dev))
		return -ENODEV;

1157 1158 1159 1160 1161
	ret = call_netdevice_notifiers(NETDEV_PRE_UP, dev);
	ret = notifier_to_errno(ret);
	if (ret)
		return ret;

L
Linus Torvalds 已提交
1162
	set_bit(__LINK_STATE_START, &dev->state);
1163

1164 1165
	if (ops->ndo_validate_addr)
		ret = ops->ndo_validate_addr(dev);
1166

1167 1168
	if (!ret && ops->ndo_open)
		ret = ops->ndo_open(dev);
L
Linus Torvalds 已提交
1169

1170 1171 1172
	if (ret)
		clear_bit(__LINK_STATE_START, &dev->state);
	else {
L
Linus Torvalds 已提交
1173
		dev->flags |= IFF_UP;
1174
		net_dmaengine_get();
1175
		dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
1176 1177
		dev_activate(dev);
	}
1178

L
Linus Torvalds 已提交
1179 1180 1181 1182
	return ret;
}

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

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

	ret = __dev_open(dev);
	if (ret < 0)
		return ret;

	rtmsg_ifinfo(RTM_NEWLINK, dev, IFF_UP|IFF_RUNNING);
	call_netdevice_notifiers(NETDEV_UP, dev);

	return ret;
}
EXPORT_SYMBOL(dev_open);

1212
static int __dev_close_many(struct list_head *head)
L
Linus Torvalds 已提交
1213
{
1214
	struct net_device *dev;
1215

1216
	ASSERT_RTNL();
1217 1218
	might_sleep();

1219 1220
	list_for_each_entry(dev, head, unreg_list) {
		call_netdevice_notifiers(NETDEV_GOING_DOWN, dev);
L
Linus Torvalds 已提交
1221

1222
		clear_bit(__LINK_STATE_START, &dev->state);
L
Linus Torvalds 已提交
1223

1224 1225 1226 1227 1228 1229 1230 1231
		/* Synchronize to scheduled poll. We cannot touch poll list, 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.
		 */
		smp_mb__after_clear_bit(); /* Commit netif_running(). */
	}
L
Linus Torvalds 已提交
1232

1233
	dev_deactivate_many(head);
1234

1235 1236
	list_for_each_entry(dev, head, unreg_list) {
		const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
1237

1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256
		/*
		 *	Call the device specific close. This cannot fail.
		 *	Only if device is UP
		 *
		 *	We allow it to be called even after a DETACH hot-plug
		 *	event.
		 */
		if (ops->ndo_stop)
			ops->ndo_stop(dev);

		dev->flags &= ~IFF_UP;
		net_dmaengine_put();
	}

	return 0;
}

static int __dev_close(struct net_device *dev)
{
1257
	int retval;
1258 1259 1260
	LIST_HEAD(single);

	list_add(&dev->unreg_list, &single);
1261 1262 1263
	retval = __dev_close_many(&single);
	list_del(&single);
	return retval;
1264 1265
}

E
Eric Dumazet 已提交
1266
static int dev_close_many(struct list_head *head)
1267 1268 1269
{
	struct net_device *dev, *tmp;
	LIST_HEAD(tmp_list);
L
Linus Torvalds 已提交
1270

1271 1272 1273 1274 1275
	list_for_each_entry_safe(dev, tmp, head, unreg_list)
		if (!(dev->flags & IFF_UP))
			list_move(&dev->unreg_list, &tmp_list);

	__dev_close_many(head);
L
Linus Torvalds 已提交
1276

1277 1278 1279 1280
	list_for_each_entry(dev, head, unreg_list) {
		rtmsg_ifinfo(RTM_NEWLINK, dev, IFF_UP|IFF_RUNNING);
		call_netdevice_notifiers(NETDEV_DOWN, dev);
	}
1281

1282 1283
	/* rollback_registered_many needs the complete original list */
	list_splice(&tmp_list, head);
1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297
	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)
{
1298 1299
	if (dev->flags & IFF_UP) {
		LIST_HEAD(single);
L
Linus Torvalds 已提交
1300

1301 1302 1303 1304
		list_add(&dev->unreg_list, &single);
		dev_close_many(&single);
		list_del(&single);
	}
L
Linus Torvalds 已提交
1305 1306
	return 0;
}
E
Eric Dumazet 已提交
1307
EXPORT_SYMBOL(dev_close);
L
Linus Torvalds 已提交
1308 1309


1310 1311 1312 1313 1314 1315 1316 1317 1318 1319
/**
 *	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)
{
1320 1321 1322 1323 1324 1325 1326
	/*
	 * If we're trying to disable lro on a vlan device
	 * use the underlying physical device instead
	 */
	if (is_vlan_dev(dev))
		dev = vlan_dev_real_dev(dev);

M
Michał Mirosław 已提交
1327 1328
	dev->wanted_features &= ~NETIF_F_LRO;
	netdev_update_features(dev);
1329

1330 1331
	if (unlikely(dev->features & NETIF_F_LRO))
		netdev_WARN(dev, "failed to disable LRO!\n");
1332 1333 1334 1335
}
EXPORT_SYMBOL(dev_disable_lro);


1336 1337
static int dev_boot_phase = 1;

L
Linus Torvalds 已提交
1338 1339 1340 1341 1342 1343 1344 1345 1346 1347
/**
 *	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
1348
 *	to the new notifier to allow device to have a race free
L
Linus Torvalds 已提交
1349 1350 1351 1352 1353 1354
 *	view of the network device list.
 */

int register_netdevice_notifier(struct notifier_block *nb)
{
	struct net_device *dev;
1355
	struct net_device *last;
1356
	struct net *net;
L
Linus Torvalds 已提交
1357 1358 1359
	int err;

	rtnl_lock();
1360
	err = raw_notifier_chain_register(&netdev_chain, nb);
1361 1362
	if (err)
		goto unlock;
1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373
	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
1374

1375 1376
			nb->notifier_call(nb, NETDEV_UP, dev);
		}
L
Linus Torvalds 已提交
1377
	}
1378 1379

unlock:
L
Linus Torvalds 已提交
1380 1381
	rtnl_unlock();
	return err;
1382 1383 1384

rollback:
	last = dev;
1385 1386 1387
	for_each_net(net) {
		for_each_netdev(net, dev) {
			if (dev == last)
1388
				goto outroll;
1389

1390 1391 1392 1393 1394
			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);
1395
			nb->notifier_call(nb, NETDEV_UNREGISTER_BATCH, dev);
1396 1397
		}
	}
1398

1399
outroll:
1400
	raw_notifier_chain_unregister(&netdev_chain, nb);
1401
	goto unlock;
L
Linus Torvalds 已提交
1402
}
E
Eric Dumazet 已提交
1403
EXPORT_SYMBOL(register_netdevice_notifier);
L
Linus Torvalds 已提交
1404 1405 1406 1407 1408 1409 1410 1411 1412

/**
 *	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.
1413 1414 1415 1416
 *
 * 	After unregistering unregister and down device events are synthesized
 *	for all devices on the device list to the removed notifier to remove
 *	the need for special case cleanup code.
L
Linus Torvalds 已提交
1417 1418 1419 1420
 */

int unregister_netdevice_notifier(struct notifier_block *nb)
{
1421 1422
	struct net_device *dev;
	struct net *net;
1423 1424 1425
	int err;

	rtnl_lock();
1426
	err = raw_notifier_chain_unregister(&netdev_chain, nb);
1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440
	if (err)
		goto unlock;

	for_each_net(net) {
		for_each_netdev(net, dev) {
			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);
			nb->notifier_call(nb, NETDEV_UNREGISTER_BATCH, dev);
		}
	}
unlock:
1441 1442
	rtnl_unlock();
	return err;
L
Linus Torvalds 已提交
1443
}
E
Eric Dumazet 已提交
1444
EXPORT_SYMBOL(unregister_netdevice_notifier);
L
Linus Torvalds 已提交
1445 1446 1447 1448

/**
 *	call_netdevice_notifiers - call all network notifier blocks
 *      @val: value passed unmodified to notifier function
1449
 *      @dev: net_device pointer passed unmodified to notifier function
L
Linus Torvalds 已提交
1450 1451
 *
 *	Call all network notifier blocks.  Parameters and return value
1452
 *	are as for raw_notifier_call_chain().
L
Linus Torvalds 已提交
1453 1454
 */

1455
int call_netdevice_notifiers(unsigned long val, struct net_device *dev)
L
Linus Torvalds 已提交
1456
{
1457
	ASSERT_RTNL();
1458
	return raw_notifier_call_chain(&netdev_chain, val, dev);
L
Linus Torvalds 已提交
1459
}
1460
EXPORT_SYMBOL(call_netdevice_notifiers);
L
Linus Torvalds 已提交
1461

1462
static struct static_key netstamp_needed __read_mostly;
1463
#ifdef HAVE_JUMP_LABEL
1464
/* We are not allowed to call static_key_slow_dec() from irq context
1465
 * If net_disable_timestamp() is called from irq context, defer the
1466
 * static_key_slow_dec() calls.
1467 1468 1469
 */
static atomic_t netstamp_needed_deferred;
#endif
L
Linus Torvalds 已提交
1470 1471 1472

void net_enable_timestamp(void)
{
1473 1474 1475 1476 1477
#ifdef HAVE_JUMP_LABEL
	int deferred = atomic_xchg(&netstamp_needed_deferred, 0);

	if (deferred) {
		while (--deferred)
1478
			static_key_slow_dec(&netstamp_needed);
1479 1480 1481 1482
		return;
	}
#endif
	WARN_ON(in_interrupt());
1483
	static_key_slow_inc(&netstamp_needed);
L
Linus Torvalds 已提交
1484
}
E
Eric Dumazet 已提交
1485
EXPORT_SYMBOL(net_enable_timestamp);
L
Linus Torvalds 已提交
1486 1487 1488

void net_disable_timestamp(void)
{
1489 1490 1491 1492 1493 1494
#ifdef HAVE_JUMP_LABEL
	if (in_interrupt()) {
		atomic_inc(&netstamp_needed_deferred);
		return;
	}
#endif
1495
	static_key_slow_dec(&netstamp_needed);
L
Linus Torvalds 已提交
1496
}
E
Eric Dumazet 已提交
1497
EXPORT_SYMBOL(net_disable_timestamp);
L
Linus Torvalds 已提交
1498

E
Eric Dumazet 已提交
1499
static inline void net_timestamp_set(struct sk_buff *skb)
L
Linus Torvalds 已提交
1500
{
1501
	skb->tstamp.tv64 = 0;
1502
	if (static_key_false(&netstamp_needed))
1503
		__net_timestamp(skb);
L
Linus Torvalds 已提交
1504 1505
}

1506
#define net_timestamp_check(COND, SKB)			\
1507
	if (static_key_false(&netstamp_needed)) {		\
1508 1509 1510
		if ((COND) && !(SKB)->tstamp.tv64)	\
			__net_timestamp(SKB);		\
	}						\
E
Eric Dumazet 已提交
1511

1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562
static int net_hwtstamp_validate(struct ifreq *ifr)
{
	struct hwtstamp_config cfg;
	enum hwtstamp_tx_types tx_type;
	enum hwtstamp_rx_filters rx_filter;
	int tx_type_valid = 0;
	int rx_filter_valid = 0;

	if (copy_from_user(&cfg, ifr->ifr_data, sizeof(cfg)))
		return -EFAULT;

	if (cfg.flags) /* reserved for future extensions */
		return -EINVAL;

	tx_type = cfg.tx_type;
	rx_filter = cfg.rx_filter;

	switch (tx_type) {
	case HWTSTAMP_TX_OFF:
	case HWTSTAMP_TX_ON:
	case HWTSTAMP_TX_ONESTEP_SYNC:
		tx_type_valid = 1;
		break;
	}

	switch (rx_filter) {
	case HWTSTAMP_FILTER_NONE:
	case HWTSTAMP_FILTER_ALL:
	case HWTSTAMP_FILTER_SOME:
	case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
	case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
	case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ:
	case HWTSTAMP_FILTER_PTP_V2_L4_EVENT:
	case HWTSTAMP_FILTER_PTP_V2_L4_SYNC:
	case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
	case HWTSTAMP_FILTER_PTP_V2_L2_EVENT:
	case HWTSTAMP_FILTER_PTP_V2_L2_SYNC:
	case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ:
	case HWTSTAMP_FILTER_PTP_V2_EVENT:
	case HWTSTAMP_FILTER_PTP_V2_SYNC:
	case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ:
		rx_filter_valid = 1;
		break;
	}

	if (!tx_type_valid || !rx_filter_valid)
		return -ERANGE;

	return 0;
}

1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583
static inline bool is_skb_forwardable(struct net_device *dev,
				      struct sk_buff *skb)
{
	unsigned int len;

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

	len = dev->mtu + dev->hard_header_len + VLAN_HLEN;
	if (skb->len <= len)
		return true;

	/* if TSO is enabled, we don't care about the length as the packet
	 * could be forwarded without being segmented before
	 */
	if (skb_is_gso(skb))
		return true;

	return false;
}

1584 1585 1586 1587 1588 1589 1590 1591
/**
 * dev_forward_skb - loopback an skb to another netif
 *
 * @dev: destination network device
 * @skb: buffer to forward
 *
 * return values:
 *	NET_RX_SUCCESS	(no congestion)
1592
 *	NET_RX_DROP     (packet was dropped, but freed)
1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603
 *
 * 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)
{
1604 1605 1606 1607 1608 1609 1610 1611
	if (skb_shinfo(skb)->tx_flags & SKBTX_DEV_ZEROCOPY) {
		if (skb_copy_ubufs(skb, GFP_ATOMIC)) {
			atomic_long_inc(&dev->rx_dropped);
			kfree_skb(skb);
			return NET_RX_DROP;
		}
	}

1612
	skb_orphan(skb);
1613
	nf_reset(skb);
1614

1615
	if (unlikely(!is_skb_forwardable(dev, skb))) {
1616
		atomic_long_inc(&dev->rx_dropped);
1617
		kfree_skb(skb);
1618
		return NET_RX_DROP;
1619
	}
1620
	skb->skb_iif = 0;
1621
	skb_set_dev(skb, dev);
1622 1623 1624 1625 1626 1627 1628
	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);

1629 1630 1631 1632 1633 1634 1635 1636
static inline int deliver_skb(struct sk_buff *skb,
			      struct packet_type *pt_prev,
			      struct net_device *orig_dev)
{
	atomic_inc(&skb->users);
	return pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
}

L
Linus Torvalds 已提交
1637 1638 1639 1640 1641
/*
 *	Support routine. Sends outgoing frames to any network
 *	taps currently in use.
 */

1642
static void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev)
L
Linus Torvalds 已提交
1643 1644
{
	struct packet_type *ptype;
1645 1646
	struct sk_buff *skb2 = NULL;
	struct packet_type *pt_prev = NULL;
1647

L
Linus Torvalds 已提交
1648 1649 1650 1651 1652 1653 1654 1655
	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)) {
1656 1657 1658 1659 1660 1661 1662
			if (pt_prev) {
				deliver_skb(skb2, pt_prev, skb->dev);
				pt_prev = ptype;
				continue;
			}

			skb2 = skb_clone(skb, GFP_ATOMIC);
L
Linus Torvalds 已提交
1663 1664 1665
			if (!skb2)
				break;

1666 1667
			net_timestamp_set(skb2);

L
Linus Torvalds 已提交
1668 1669 1670 1671
			/* skb->nh should be correctly
			   set by sender, so that the second statement is
			   just protection against buggy protocols.
			 */
1672
			skb_reset_mac_header(skb2);
L
Linus Torvalds 已提交
1673

1674
			if (skb_network_header(skb2) < skb2->data ||
1675
			    skb2->network_header > skb2->tail) {
L
Linus Torvalds 已提交
1676
				if (net_ratelimit())
1677 1678 1679
					pr_crit("protocol %04x is buggy, dev %s\n",
						ntohs(skb2->protocol),
						dev->name);
1680
				skb_reset_network_header(skb2);
L
Linus Torvalds 已提交
1681 1682
			}

1683
			skb2->transport_header = skb2->network_header;
L
Linus Torvalds 已提交
1684
			skb2->pkt_type = PACKET_OUTGOING;
1685
			pt_prev = ptype;
L
Linus Torvalds 已提交
1686 1687
		}
	}
1688 1689
	if (pt_prev)
		pt_prev->func(skb2, skb->dev, pt_prev, skb->dev);
L
Linus Torvalds 已提交
1690 1691 1692
	rcu_read_unlock();
}

1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704
/* netif_setup_tc - Handle tc mappings on real_num_tx_queues change
 * @dev: Network device
 * @txq: number of queues available
 *
 * If real_num_tx_queues is changed the tc mappings may no longer be
 * valid. To resolve this verify the tc mapping remains valid and if
 * not NULL the mapping. With no priorities mapping to this
 * offset/count pair it will no longer be used. In the worst case TC0
 * is invalid nothing can be done so disable priority mappings. If is
 * expected that drivers will fix this mapping if they can before
 * calling netif_set_real_num_tx_queues.
 */
E
Eric Dumazet 已提交
1705
static void netif_setup_tc(struct net_device *dev, unsigned int txq)
1706 1707 1708 1709 1710 1711
{
	int i;
	struct netdev_tc_txq *tc = &dev->tc_to_txq[0];

	/* If TC0 is invalidated disable TC mapping */
	if (tc->offset + tc->count > txq) {
1712
		pr_warn("Number of in use tx queues changed invalidating tc mappings. Priority traffic classification disabled!\n");
1713 1714 1715 1716 1717 1718 1719 1720 1721 1722
		dev->num_tc = 0;
		return;
	}

	/* Invalidated prio to tc mappings set to TC0 */
	for (i = 1; i < TC_BITMASK + 1; i++) {
		int q = netdev_get_prio_tc_map(dev, i);

		tc = &dev->tc_to_txq[q];
		if (tc->offset + tc->count > txq) {
1723 1724
			pr_warn("Number of in use tx queues changed. Priority %i to tc mapping %i is no longer valid. Setting map to 0\n",
				i, q);
1725 1726 1727 1728 1729
			netdev_set_prio_tc_map(dev, i, 0);
		}
	}
}

1730 1731 1732 1733
/*
 * Routine to help set real_num_tx_queues. To avoid skbs mapped to queues
 * greater then real_num_tx_queues stale skbs on the qdisc must be flushed.
 */
1734
int netif_set_real_num_tx_queues(struct net_device *dev, unsigned int txq)
1735
{
T
Tom Herbert 已提交
1736 1737
	int rc;

1738 1739
	if (txq < 1 || txq > dev->num_tx_queues)
		return -EINVAL;
1740

1741 1742
	if (dev->reg_state == NETREG_REGISTERED ||
	    dev->reg_state == NETREG_UNREGISTERING) {
1743 1744
		ASSERT_RTNL();

T
Tom Herbert 已提交
1745 1746
		rc = netdev_queue_update_kobjects(dev, dev->real_num_tx_queues,
						  txq);
T
Tom Herbert 已提交
1747 1748 1749
		if (rc)
			return rc;

1750 1751 1752
		if (dev->num_tc)
			netif_setup_tc(dev, txq);

1753 1754
		if (txq < dev->real_num_tx_queues)
			qdisc_reset_all_tx_gt(dev, txq);
1755
	}
1756 1757 1758

	dev->real_num_tx_queues = txq;
	return 0;
1759 1760
}
EXPORT_SYMBOL(netif_set_real_num_tx_queues);
1761

1762 1763 1764 1765 1766 1767 1768 1769
#ifdef CONFIG_RPS
/**
 *	netif_set_real_num_rx_queues - set actual number of RX queues used
 *	@dev: Network device
 *	@rxq: Actual number of RX queues
 *
 *	This must be called either with the rtnl_lock held or before
 *	registration of the net device.  Returns 0 on success, or a
1770 1771
 *	negative error code.  If called before registration, it always
 *	succeeds.
1772 1773 1774 1775 1776
 */
int netif_set_real_num_rx_queues(struct net_device *dev, unsigned int rxq)
{
	int rc;

T
Tom Herbert 已提交
1777 1778 1779
	if (rxq < 1 || rxq > dev->num_rx_queues)
		return -EINVAL;

1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794
	if (dev->reg_state == NETREG_REGISTERED) {
		ASSERT_RTNL();

		rc = net_rx_queue_update_kobjects(dev, dev->real_num_rx_queues,
						  rxq);
		if (rc)
			return rc;
	}

	dev->real_num_rx_queues = rxq;
	return 0;
}
EXPORT_SYMBOL(netif_set_real_num_rx_queues);
#endif

1795
static inline void __netif_reschedule(struct Qdisc *q)
1796
{
1797 1798
	struct softnet_data *sd;
	unsigned long flags;
1799

1800 1801
	local_irq_save(flags);
	sd = &__get_cpu_var(softnet_data);
1802 1803 1804
	q->next_sched = NULL;
	*sd->output_queue_tailp = q;
	sd->output_queue_tailp = &q->next_sched;
1805 1806 1807 1808 1809 1810 1811 1812
	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);
1813 1814 1815
}
EXPORT_SYMBOL(__netif_schedule);

1816
void dev_kfree_skb_irq(struct sk_buff *skb)
1817
{
1818
	if (atomic_dec_and_test(&skb->users)) {
1819 1820
		struct softnet_data *sd;
		unsigned long flags;
1821

1822 1823 1824 1825 1826 1827 1828
		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);
	}
1829
}
1830
EXPORT_SYMBOL(dev_kfree_skb_irq);
1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841

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);


1842 1843 1844 1845 1846 1847
/**
 * netif_device_detach - mark device as removed
 * @dev: network device
 *
 * Mark device as removed from system and therefore no longer available.
 */
1848 1849 1850 1851
void netif_device_detach(struct net_device *dev)
{
	if (test_and_clear_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
1852
		netif_tx_stop_all_queues(dev);
1853 1854 1855 1856
	}
}
EXPORT_SYMBOL(netif_device_detach);

1857 1858 1859 1860 1861 1862
/**
 * netif_device_attach - mark device as attached
 * @dev: network device
 *
 * Mark device as attached from system and restart if needed.
 */
1863 1864 1865 1866
void netif_device_attach(struct net_device *dev)
{
	if (!test_and_set_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
1867
		netif_tx_wake_all_queues(dev);
1868
		__netdev_watchdog_up(dev);
1869 1870 1871 1872
	}
}
EXPORT_SYMBOL(netif_device_attach);

1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902
/**
 * 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 */

1903 1904
static void skb_warn_bad_offload(const struct sk_buff *skb)
{
1905
	static const netdev_features_t null_features = 0;
1906 1907 1908 1909 1910 1911 1912 1913
	struct net_device *dev = skb->dev;
	const char *driver = "";

	if (dev && dev->dev.parent)
		driver = dev_driver_string(dev->dev.parent);

	WARN(1, "%s: caps=(%pNF, %pNF) len=%d data_len=%d gso_size=%d "
	     "gso_type=%d ip_summed=%d\n",
1914 1915
	     driver, dev ? &dev->features : &null_features,
	     skb->sk ? &skb->sk->sk_route_caps : &null_features,
1916 1917 1918 1919
	     skb->len, skb->data_len, skb_shinfo(skb)->gso_size,
	     skb_shinfo(skb)->gso_type, skb->ip_summed);
}

L
Linus Torvalds 已提交
1920 1921 1922 1923
/*
 * Invalidate hardware checksum when packet is to be mangled, and
 * complete checksum manually on outgoing path.
 */
1924
int skb_checksum_help(struct sk_buff *skb)
L
Linus Torvalds 已提交
1925
{
1926
	__wsum csum;
1927
	int ret = 0, offset;
L
Linus Torvalds 已提交
1928

1929
	if (skb->ip_summed == CHECKSUM_COMPLETE)
1930 1931 1932
		goto out_set_summed;

	if (unlikely(skb_shinfo(skb)->gso_size)) {
1933 1934
		skb_warn_bad_offload(skb);
		return -EINVAL;
L
Linus Torvalds 已提交
1935 1936
	}

1937
	offset = skb_checksum_start_offset(skb);
1938 1939 1940 1941 1942 1943 1944 1945
	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
1946 1947 1948 1949 1950
		ret = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
		if (ret)
			goto out;
	}

1951
	*(__sum16 *)(skb->data + offset) = csum_fold(csum);
1952
out_set_summed:
L
Linus Torvalds 已提交
1953
	skb->ip_summed = CHECKSUM_NONE;
1954
out:
L
Linus Torvalds 已提交
1955 1956
	return ret;
}
E
Eric Dumazet 已提交
1957
EXPORT_SYMBOL(skb_checksum_help);
L
Linus Torvalds 已提交
1958

1959 1960 1961
/**
 *	skb_gso_segment - Perform segmentation on skb.
 *	@skb: buffer to segment
1962
 *	@features: features for the output path (see dev->features)
1963 1964
 *
 *	This function segments the given skb and returns a list of segments.
1965 1966 1967
 *
 *	It may return NULL if the skb requires no segmentation.  This is
 *	only possible when GSO is used for verifying header integrity.
1968
 */
1969 1970
struct sk_buff *skb_gso_segment(struct sk_buff *skb,
	netdev_features_t features)
1971 1972 1973
{
	struct sk_buff *segs = ERR_PTR(-EPROTONOSUPPORT);
	struct packet_type *ptype;
A
Al Viro 已提交
1974
	__be16 type = skb->protocol;
1975
	int vlan_depth = ETH_HLEN;
1976
	int err;
1977

1978 1979
	while (type == htons(ETH_P_8021Q)) {
		struct vlan_hdr *vh;
1980

1981
		if (unlikely(!pskb_may_pull(skb, vlan_depth + VLAN_HLEN)))
1982 1983
			return ERR_PTR(-EINVAL);

1984 1985 1986
		vh = (struct vlan_hdr *)(skb->data + vlan_depth);
		type = vh->h_vlan_encapsulated_proto;
		vlan_depth += VLAN_HLEN;
1987 1988
	}

1989
	skb_reset_mac_header(skb);
1990
	skb->mac_len = skb->network_header - skb->mac_header;
1991 1992
	__skb_pull(skb, skb->mac_len);

1993
	if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) {
1994
		skb_warn_bad_offload(skb);
1995

1996 1997 1998 1999 2000
		if (skb_header_cloned(skb) &&
		    (err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
			return ERR_PTR(err);
	}

2001
	rcu_read_lock();
2002 2003
	list_for_each_entry_rcu(ptype,
			&ptype_base[ntohs(type) & PTYPE_HASH_MASK], list) {
2004
		if (ptype->type == type && !ptype->dev && ptype->gso_segment) {
2005
			if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) {
2006 2007 2008 2009
				err = ptype->gso_send_check(skb);
				segs = ERR_PTR(err);
				if (err || skb_gso_ok(skb, features))
					break;
2010 2011
				__skb_push(skb, (skb->data -
						 skb_network_header(skb)));
2012
			}
2013
			segs = ptype->gso_segment(skb, features);
2014 2015 2016 2017 2018
			break;
		}
	}
	rcu_read_unlock();

2019
	__skb_push(skb, skb->data - skb_mac_header(skb));
2020

2021 2022 2023 2024
	return segs;
}
EXPORT_SYMBOL(skb_gso_segment);

2025 2026 2027 2028 2029
/* Take action when hardware reception checksum errors are detected. */
#ifdef CONFIG_BUG
void netdev_rx_csum_fault(struct net_device *dev)
{
	if (net_ratelimit()) {
2030
		pr_err("%s: hw csum failure\n", dev ? dev->name : "<unknown>");
2031 2032 2033 2034 2035 2036
		dump_stack();
	}
}
EXPORT_SYMBOL(netdev_rx_csum_fault);
#endif

L
Linus Torvalds 已提交
2037 2038 2039 2040 2041
/* 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.
 */

E
Eric Dumazet 已提交
2042
static int illegal_highdma(struct net_device *dev, struct sk_buff *skb)
L
Linus Torvalds 已提交
2043
{
2044
#ifdef CONFIG_HIGHMEM
L
Linus Torvalds 已提交
2045
	int i;
2046
	if (!(dev->features & NETIF_F_HIGHDMA)) {
2047 2048 2049
		for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
			skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
			if (PageHighMem(skb_frag_page(frag)))
2050
				return 1;
2051
		}
2052
	}
L
Linus Torvalds 已提交
2053

2054 2055
	if (PCI_DMA_BUS_IS_PHYS) {
		struct device *pdev = dev->dev.parent;
L
Linus Torvalds 已提交
2056

E
Eric Dumazet 已提交
2057 2058
		if (!pdev)
			return 0;
2059
		for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
2060 2061
			skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
			dma_addr_t addr = page_to_phys(skb_frag_page(frag));
2062 2063 2064 2065
			if (!pdev->dma_mask || addr + PAGE_SIZE - 1 > *pdev->dma_mask)
				return 1;
		}
	}
2066
#endif
L
Linus Torvalds 已提交
2067 2068 2069
	return 0;
}

2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095
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
2096
 *	@features: device features as applicable to this skb
2097 2098 2099 2100
 *
 *	This function segments the given skb and stores the list of segments
 *	in skb->next.
 */
2101
static int dev_gso_segment(struct sk_buff *skb, netdev_features_t features)
2102 2103
{
	struct sk_buff *segs;
2104 2105 2106 2107 2108 2109

	segs = skb_gso_segment(skb, features);

	/* Verifying header integrity only. */
	if (!segs)
		return 0;
2110

2111
	if (IS_ERR(segs))
2112 2113 2114 2115 2116 2117 2118 2119 2120
		return PTR_ERR(segs);

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

	return 0;
}

E
Eric Dumazet 已提交
2121 2122
/*
 * Try to orphan skb early, right before transmission by the device.
2123 2124
 * We cannot orphan skb if tx timestamp is requested or the sk-reference
 * is needed on driver level for other reasons, e.g. see net/can/raw.c
E
Eric Dumazet 已提交
2125 2126 2127
 */
static inline void skb_orphan_try(struct sk_buff *skb)
{
2128 2129
	struct sock *sk = skb->sk;

2130
	if (sk && !skb_shinfo(skb)->tx_flags) {
2131 2132 2133 2134 2135
		/* skb_tx_hash() wont be able to get sk.
		 * We copy sk_hash into skb->rxhash
		 */
		if (!skb->rxhash)
			skb->rxhash = sk->sk_hash;
E
Eric Dumazet 已提交
2136
		skb_orphan(skb);
2137
	}
E
Eric Dumazet 已提交
2138 2139
}

2140
static bool can_checksum_protocol(netdev_features_t features, __be16 protocol)
2141 2142 2143 2144 2145 2146 2147 2148 2149 2150
{
	return ((features & NETIF_F_GEN_CSUM) ||
		((features & NETIF_F_V4_CSUM) &&
		 protocol == htons(ETH_P_IP)) ||
		((features & NETIF_F_V6_CSUM) &&
		 protocol == htons(ETH_P_IPV6)) ||
		((features & NETIF_F_FCOE_CRC) &&
		 protocol == htons(ETH_P_FCOE)));
}

2151 2152
static netdev_features_t harmonize_features(struct sk_buff *skb,
	__be16 protocol, netdev_features_t features)
2153
{
2154
	if (!can_checksum_protocol(features, protocol)) {
2155 2156 2157 2158 2159 2160 2161 2162 2163
		features &= ~NETIF_F_ALL_CSUM;
		features &= ~NETIF_F_SG;
	} else if (illegal_highdma(skb->dev, skb)) {
		features &= ~NETIF_F_SG;
	}

	return features;
}

2164
netdev_features_t netif_skb_features(struct sk_buff *skb)
2165 2166
{
	__be16 protocol = skb->protocol;
2167
	netdev_features_t features = skb->dev->features;
2168 2169 2170 2171

	if (protocol == htons(ETH_P_8021Q)) {
		struct vlan_ethhdr *veh = (struct vlan_ethhdr *)skb->data;
		protocol = veh->h_vlan_encapsulated_proto;
2172 2173 2174
	} else if (!vlan_tx_tag_present(skb)) {
		return harmonize_features(skb, protocol, features);
	}
2175

2176
	features &= (skb->dev->vlan_features | NETIF_F_HW_VLAN_TX);
2177 2178 2179 2180 2181

	if (protocol != htons(ETH_P_8021Q)) {
		return harmonize_features(skb, protocol, features);
	} else {
		features &= NETIF_F_SG | NETIF_F_HIGHDMA | NETIF_F_FRAGLIST |
2182
				NETIF_F_GEN_CSUM | NETIF_F_HW_VLAN_TX;
2183 2184
		return harmonize_features(skb, protocol, features);
	}
2185
}
2186
EXPORT_SYMBOL(netif_skb_features);
2187

2188 2189 2190 2191 2192 2193 2194 2195
/*
 * 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,
2196
				      int features)
2197
{
2198 2199 2200
	return skb_is_nonlinear(skb) &&
			((skb_has_frag_list(skb) &&
				!(features & NETIF_F_FRAGLIST)) ||
2201
			(skb_shinfo(skb)->nr_frags &&
2202
				!(features & NETIF_F_SG)));
2203 2204
}

2205 2206
int dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
			struct netdev_queue *txq)
2207
{
2208
	const struct net_device_ops *ops = dev->netdev_ops;
2209
	int rc = NETDEV_TX_OK;
2210
	unsigned int skb_len;
2211

2212
	if (likely(!skb->next)) {
2213
		netdev_features_t features;
2214

2215
		/*
L
Lucas De Marchi 已提交
2216
		 * If device doesn't need skb->dst, release it right now while
2217 2218
		 * its hot in this cpu cache
		 */
E
Eric Dumazet 已提交
2219 2220 2221
		if (dev->priv_flags & IFF_XMIT_DST_RELEASE)
			skb_dst_drop(skb);

2222 2223 2224
		if (!list_empty(&ptype_all))
			dev_queue_xmit_nit(skb, dev);

E
Eric Dumazet 已提交
2225
		skb_orphan_try(skb);
2226

2227 2228
		features = netif_skb_features(skb);

2229
		if (vlan_tx_tag_present(skb) &&
2230
		    !(features & NETIF_F_HW_VLAN_TX)) {
2231 2232 2233 2234 2235 2236 2237
			skb = __vlan_put_tag(skb, vlan_tx_tag_get(skb));
			if (unlikely(!skb))
				goto out;

			skb->vlan_tci = 0;
		}

2238
		if (netif_needs_gso(skb, features)) {
2239
			if (unlikely(dev_gso_segment(skb, features)))
2240 2241 2242
				goto out_kfree_skb;
			if (skb->next)
				goto gso;
2243
		} else {
2244
			if (skb_needs_linearize(skb, features) &&
2245 2246 2247 2248 2249 2250 2251 2252
			    __skb_linearize(skb))
				goto out_kfree_skb;

			/* If packet is not checksummed and device does not
			 * support checksumming for this protocol, complete
			 * checksumming here.
			 */
			if (skb->ip_summed == CHECKSUM_PARTIAL) {
2253 2254
				skb_set_transport_header(skb,
					skb_checksum_start_offset(skb));
2255
				if (!(features & NETIF_F_ALL_CSUM) &&
2256 2257 2258
				     skb_checksum_help(skb))
					goto out_kfree_skb;
			}
2259 2260
		}

2261
		skb_len = skb->len;
2262
		rc = ops->ndo_start_xmit(skb, dev);
2263
		trace_net_dev_xmit(skb, rc, dev, skb_len);
2264
		if (rc == NETDEV_TX_OK)
E
Eric Dumazet 已提交
2265
			txq_trans_update(txq);
2266
		return rc;
2267 2268
	}

2269
gso:
2270 2271 2272 2273 2274
	do {
		struct sk_buff *nskb = skb->next;

		skb->next = nskb->next;
		nskb->next = NULL;
2275 2276

		/*
L
Lucas De Marchi 已提交
2277
		 * If device doesn't need nskb->dst, release it right now while
2278 2279 2280 2281 2282
		 * its hot in this cpu cache
		 */
		if (dev->priv_flags & IFF_XMIT_DST_RELEASE)
			skb_dst_drop(nskb);

2283
		skb_len = nskb->len;
2284
		rc = ops->ndo_start_xmit(nskb, dev);
2285
		trace_net_dev_xmit(nskb, rc, dev, skb_len);
2286
		if (unlikely(rc != NETDEV_TX_OK)) {
2287 2288
			if (rc & ~NETDEV_TX_MASK)
				goto out_kfree_gso_skb;
2289
			nskb->next = skb->next;
2290 2291 2292
			skb->next = nskb;
			return rc;
		}
E
Eric Dumazet 已提交
2293
		txq_trans_update(txq);
2294
		if (unlikely(netif_xmit_stopped(txq) && skb->next))
2295
			return NETDEV_TX_BUSY;
2296
	} while (skb->next);
2297

2298 2299 2300
out_kfree_gso_skb:
	if (likely(skb->next == NULL))
		skb->destructor = DEV_GSO_CB(skb)->destructor;
2301 2302
out_kfree_skb:
	kfree_skb(skb);
2303
out:
2304
	return rc;
2305 2306
}

T
Tom Herbert 已提交
2307
static u32 hashrnd __read_mostly;
D
David S. Miller 已提交
2308

2309 2310 2311 2312 2313 2314
/*
 * Returns a Tx hash based on the given packet descriptor a Tx queues' number
 * to be used as a distribution range.
 */
u16 __skb_tx_hash(const struct net_device *dev, const struct sk_buff *skb,
		  unsigned int num_tx_queues)
2315
{
2316
	u32 hash;
2317 2318
	u16 qoffset = 0;
	u16 qcount = num_tx_queues;
D
David S. Miller 已提交
2319

2320 2321
	if (skb_rx_queue_recorded(skb)) {
		hash = skb_get_rx_queue(skb);
2322 2323
		while (unlikely(hash >= num_tx_queues))
			hash -= num_tx_queues;
2324 2325
		return hash;
	}
2326

2327 2328 2329 2330 2331 2332
	if (dev->num_tc) {
		u8 tc = netdev_get_prio_tc_map(dev, skb->priority);
		qoffset = dev->tc_to_txq[tc].offset;
		qcount = dev->tc_to_txq[tc].count;
	}

2333
	if (skb->sk && skb->sk->sk_hash)
2334
		hash = skb->sk->sk_hash;
2335
	else
2336
		hash = (__force u16) skb->protocol ^ skb->rxhash;
T
Tom Herbert 已提交
2337
	hash = jhash_1word(hash, hashrnd);
D
David S. Miller 已提交
2338

2339
	return (u16) (((u64) hash * qcount) >> 32) + qoffset;
2340
}
2341
EXPORT_SYMBOL(__skb_tx_hash);
2342

2343 2344 2345 2346
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()) {
2347 2348 2349
			pr_warn("%s selects TX queue %d, but real number of TX queues is %d\n",
				dev->name, queue_index,
				dev->real_num_tx_queues);
2350 2351 2352 2353 2354 2355
		}
		return 0;
	}
	return queue_index;
}

T
Tom Herbert 已提交
2356 2357
static inline int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
{
T
Tom Herbert 已提交
2358
#ifdef CONFIG_XPS
T
Tom Herbert 已提交
2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393
	struct xps_dev_maps *dev_maps;
	struct xps_map *map;
	int queue_index = -1;

	rcu_read_lock();
	dev_maps = rcu_dereference(dev->xps_maps);
	if (dev_maps) {
		map = rcu_dereference(
		    dev_maps->cpu_map[raw_smp_processor_id()]);
		if (map) {
			if (map->len == 1)
				queue_index = map->queues[0];
			else {
				u32 hash;
				if (skb->sk && skb->sk->sk_hash)
					hash = skb->sk->sk_hash;
				else
					hash = (__force u16) skb->protocol ^
					    skb->rxhash;
				hash = jhash_1word(hash, hashrnd);
				queue_index = map->queues[
				    ((u64)hash * map->len) >> 32];
			}
			if (unlikely(queue_index >= dev->real_num_tx_queues))
				queue_index = -1;
		}
	}
	rcu_read_unlock();

	return queue_index;
#else
	return -1;
#endif
}

2394 2395 2396
static struct netdev_queue *dev_pick_tx(struct net_device *dev,
					struct sk_buff *skb)
{
2397
	int queue_index;
2398
	const struct net_device_ops *ops = dev->netdev_ops;
2399

2400 2401 2402
	if (dev->real_num_tx_queues == 1)
		queue_index = 0;
	else if (ops->ndo_select_queue) {
2403 2404 2405 2406 2407
		queue_index = ops->ndo_select_queue(dev, skb);
		queue_index = dev_cap_txqueue(dev, queue_index);
	} else {
		struct sock *sk = skb->sk;
		queue_index = sk_tx_queue_get(sk);
2408

2409 2410 2411
		if (queue_index < 0 || skb->ooo_okay ||
		    queue_index >= dev->real_num_tx_queues) {
			int old_index = queue_index;
2412

T
Tom Herbert 已提交
2413 2414 2415
			queue_index = get_xps_queue(dev, skb);
			if (queue_index < 0)
				queue_index = skb_tx_hash(dev, skb);
2416 2417 2418 2419

			if (queue_index != old_index && sk) {
				struct dst_entry *dst =
				    rcu_dereference_check(sk->sk_dst_cache, 1);
E
Eric Dumazet 已提交
2420 2421 2422 2423

				if (dst && skb_dst(skb) == dst)
					sk_tx_queue_set(sk, queue_index);
			}
2424 2425
		}
	}
2426

2427 2428
	skb_set_queue_mapping(skb, queue_index);
	return netdev_get_tx_queue(dev, queue_index);
2429 2430
}

2431 2432 2433 2434 2435
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);
E
Eric Dumazet 已提交
2436
	bool contended;
2437 2438
	int rc;

E
Eric Dumazet 已提交
2439 2440
	qdisc_skb_cb(skb)->pkt_len = skb->len;
	qdisc_calculate_pkt_len(skb, q);
2441 2442 2443 2444 2445 2446
	/*
	 * Heuristic to force contended enqueues to serialize on a
	 * separate lock before trying to get qdisc main lock.
	 * This permits __QDISC_STATE_RUNNING owner to get the lock more often
	 * and dequeue packets faster.
	 */
E
Eric Dumazet 已提交
2447
	contended = qdisc_is_running(q);
2448 2449 2450
	if (unlikely(contended))
		spin_lock(&q->busylock);

2451 2452 2453 2454 2455
	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) &&
2456
		   qdisc_run_begin(q)) {
2457 2458 2459 2460 2461
		/*
		 * 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.
		 */
E
Eric Dumazet 已提交
2462 2463
		if (!(dev->priv_flags & IFF_XMIT_DST_RELEASE))
			skb_dst_force(skb);
2464 2465 2466

		qdisc_bstats_update(q, skb);

2467 2468 2469 2470 2471
		if (sch_direct_xmit(skb, q, dev, txq, root_lock)) {
			if (unlikely(contended)) {
				spin_unlock(&q->busylock);
				contended = false;
			}
2472
			__qdisc_run(q);
2473
		} else
2474
			qdisc_run_end(q);
2475 2476 2477

		rc = NET_XMIT_SUCCESS;
	} else {
E
Eric Dumazet 已提交
2478
		skb_dst_force(skb);
E
Eric Dumazet 已提交
2479
		rc = q->enqueue(skb, q) & NET_XMIT_MASK;
2480 2481 2482 2483 2484 2485 2486
		if (qdisc_run_begin(q)) {
			if (unlikely(contended)) {
				spin_unlock(&q->busylock);
				contended = false;
			}
			__qdisc_run(q);
		}
2487 2488
	}
	spin_unlock(root_lock);
2489 2490
	if (unlikely(contended))
		spin_unlock(&q->busylock);
2491 2492 2493
	return rc;
}

2494 2495 2496
#if IS_ENABLED(CONFIG_NETPRIO_CGROUP)
static void skb_update_prio(struct sk_buff *skb)
{
I
Igor Maravic 已提交
2497
	struct netprio_map *map = rcu_dereference_bh(skb->dev->priomap);
2498 2499 2500 2501 2502 2503 2504 2505

	if ((!skb->priority) && (skb->sk) && map)
		skb->priority = map->priomap[skb->sk->sk_cgrp_prioidx];
}
#else
#define skb_update_prio(skb)
#endif

2506
static DEFINE_PER_CPU(int, xmit_recursion);
2507
#define RECURSION_LIMIT 10
2508

2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533
/**
 *	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
2534 2535 2536
int dev_queue_xmit(struct sk_buff *skb)
{
	struct net_device *dev = skb->dev;
2537
	struct netdev_queue *txq;
L
Linus Torvalds 已提交
2538 2539 2540
	struct Qdisc *q;
	int rc = -ENOMEM;

2541 2542
	/* Disable soft irqs for various locks below. Also
	 * stops preemption for RCU.
L
Linus Torvalds 已提交
2543
	 */
2544
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
2545

2546 2547
	skb_update_prio(skb);

2548
	txq = dev_pick_tx(dev, skb);
2549
	q = rcu_dereference_bh(txq->qdisc);
2550

L
Linus Torvalds 已提交
2551
#ifdef CONFIG_NET_CLS_ACT
E
Eric Dumazet 已提交
2552
	skb->tc_verd = SET_TC_AT(skb->tc_verd, AT_EGRESS);
L
Linus Torvalds 已提交
2553
#endif
2554
	trace_net_dev_queue(skb);
L
Linus Torvalds 已提交
2555
	if (q->enqueue) {
2556
		rc = __dev_xmit_skb(skb, q, dev, txq);
2557
		goto out;
L
Linus Torvalds 已提交
2558 2559 2560 2561 2562
	}

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

H
Herbert Xu 已提交
2563 2564
	   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
2565 2566 2567 2568 2569 2570 2571 2572 2573 2574
	   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 */

2575
		if (txq->xmit_lock_owner != cpu) {
L
Linus Torvalds 已提交
2576

2577 2578 2579
			if (__this_cpu_read(xmit_recursion) > RECURSION_LIMIT)
				goto recursion_alert;

2580
			HARD_TX_LOCK(dev, txq, cpu);
L
Linus Torvalds 已提交
2581

2582
			if (!netif_xmit_stopped(txq)) {
2583
				__this_cpu_inc(xmit_recursion);
2584
				rc = dev_hard_start_xmit(skb, dev, txq);
2585
				__this_cpu_dec(xmit_recursion);
2586
				if (dev_xmit_complete(rc)) {
2587
					HARD_TX_UNLOCK(dev, txq);
L
Linus Torvalds 已提交
2588 2589 2590
					goto out;
				}
			}
2591
			HARD_TX_UNLOCK(dev, txq);
L
Linus Torvalds 已提交
2592
			if (net_ratelimit())
2593 2594
				pr_crit("Virtual device %s asks to queue packet!\n",
					dev->name);
L
Linus Torvalds 已提交
2595 2596
		} else {
			/* Recursion is detected! It is possible,
2597 2598 2599
			 * unfortunately
			 */
recursion_alert:
L
Linus Torvalds 已提交
2600
			if (net_ratelimit())
2601 2602
				pr_crit("Dead loop on virtual device %s, fix it urgently!\n",
					dev->name);
L
Linus Torvalds 已提交
2603 2604 2605 2606
		}
	}

	rc = -ENETDOWN;
2607
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
2608 2609 2610 2611

	kfree_skb(skb);
	return rc;
out:
2612
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
2613 2614
	return rc;
}
E
Eric Dumazet 已提交
2615
EXPORT_SYMBOL(dev_queue_xmit);
L
Linus Torvalds 已提交
2616 2617 2618 2619 2620 2621


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

2622
int netdev_max_backlog __read_mostly = 1000;
E
Eric Dumazet 已提交
2623
int netdev_tstamp_prequeue __read_mostly = 1;
2624 2625
int netdev_budget __read_mostly = 300;
int weight_p __read_mostly = 64;            /* old backlog weight */
L
Linus Torvalds 已提交
2626

E
Eric Dumazet 已提交
2627 2628 2629 2630 2631 2632 2633 2634
/* Called with irq disabled */
static inline void ____napi_schedule(struct softnet_data *sd,
				     struct napi_struct *napi)
{
	list_add_tail(&napi->poll_list, &sd->poll_list);
	__raise_softirq_irqoff(NET_RX_SOFTIRQ);
}

T
Tom Herbert 已提交
2635
/*
2636
 * __skb_get_rxhash: calculate a flow hash based on src/dst addresses
2637 2638 2639
 * and src/dst port numbers.  Sets rxhash in skb to non-zero hash value
 * on success, zero indicates no valid hash.  Also, sets l4_rxhash in skb
 * if hash is a canonical 4-tuple hash over transport ports.
T
Tom Herbert 已提交
2640
 */
2641
void __skb_get_rxhash(struct sk_buff *skb)
T
Tom Herbert 已提交
2642
{
2643 2644
	struct flow_keys keys;
	u32 hash;
2645

2646 2647
	if (!skb_flow_dissect(skb, &keys))
		return;
2648

2649 2650 2651 2652
	if (keys.ports) {
		if ((__force u16)keys.port16[1] < (__force u16)keys.port16[0])
			swap(keys.port16[0], keys.port16[1]);
		skb->l4_rxhash = 1;
T
Tom Herbert 已提交
2653 2654
	}

E
Eric Dumazet 已提交
2655
	/* get a consistent hash (same value on both flow directions) */
2656 2657
	if ((__force u32)keys.dst < (__force u32)keys.src)
		swap(keys.dst, keys.src);
T
Tom Herbert 已提交
2658

2659 2660 2661
	hash = jhash_3words((__force u32)keys.dst,
			    (__force u32)keys.src,
			    (__force u32)keys.ports, hashrnd);
2662 2663 2664
	if (!hash)
		hash = 1;

2665
	skb->rxhash = hash;
2666 2667 2668 2669 2670 2671
}
EXPORT_SYMBOL(__skb_get_rxhash);

#ifdef CONFIG_RPS

/* One global table that all flow-based protocols share. */
E
Eric Dumazet 已提交
2672
struct rps_sock_flow_table __rcu *rps_sock_flow_table __read_mostly;
2673 2674
EXPORT_SYMBOL(rps_sock_flow_table);

2675
struct static_key rps_needed __read_mostly;
2676

2677 2678 2679 2680
static struct rps_dev_flow *
set_rps_cpu(struct net_device *dev, struct sk_buff *skb,
	    struct rps_dev_flow *rflow, u16 next_cpu)
{
2681
	if (next_cpu != RPS_NO_CPU) {
2682 2683 2684 2685 2686 2687 2688 2689 2690
#ifdef CONFIG_RFS_ACCEL
		struct netdev_rx_queue *rxqueue;
		struct rps_dev_flow_table *flow_table;
		struct rps_dev_flow *old_rflow;
		u32 flow_id;
		u16 rxq_index;
		int rc;

		/* Should we steer this flow to a different hardware queue? */
2691 2692
		if (!skb_rx_queue_recorded(skb) || !dev->rx_cpu_rmap ||
		    !(dev->features & NETIF_F_NTUPLE))
2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714
			goto out;
		rxq_index = cpu_rmap_lookup_index(dev->rx_cpu_rmap, next_cpu);
		if (rxq_index == skb_get_rx_queue(skb))
			goto out;

		rxqueue = dev->_rx + rxq_index;
		flow_table = rcu_dereference(rxqueue->rps_flow_table);
		if (!flow_table)
			goto out;
		flow_id = skb->rxhash & flow_table->mask;
		rc = dev->netdev_ops->ndo_rx_flow_steer(dev, skb,
							rxq_index, flow_id);
		if (rc < 0)
			goto out;
		old_rflow = rflow;
		rflow = &flow_table->flows[flow_id];
		rflow->filter = rc;
		if (old_rflow->filter == rflow->filter)
			old_rflow->filter = RPS_NO_FILTER;
	out:
#endif
		rflow->last_qtail =
2715
			per_cpu(softnet_data, next_cpu).input_queue_head;
2716 2717
	}

2718
	rflow->cpu = next_cpu;
2719 2720 2721
	return rflow;
}

2722 2723 2724 2725 2726 2727 2728 2729 2730
/*
 * 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.
 * rcu_read_lock must be held on entry.
 */
static int get_rps_cpu(struct net_device *dev, struct sk_buff *skb,
		       struct rps_dev_flow **rflowp)
{
	struct netdev_rx_queue *rxqueue;
E
Eric Dumazet 已提交
2731
	struct rps_map *map;
2732 2733 2734 2735 2736 2737 2738
	struct rps_dev_flow_table *flow_table;
	struct rps_sock_flow_table *sock_flow_table;
	int cpu = -1;
	u16 tcpu;

	if (skb_rx_queue_recorded(skb)) {
		u16 index = skb_get_rx_queue(skb);
2739 2740 2741 2742 2743
		if (unlikely(index >= dev->real_num_rx_queues)) {
			WARN_ONCE(dev->real_num_rx_queues > 1,
				  "%s received packet on queue %u, but number "
				  "of RX queues is %u\n",
				  dev->name, index, dev->real_num_rx_queues);
2744 2745 2746 2747 2748 2749
			goto done;
		}
		rxqueue = dev->_rx + index;
	} else
		rxqueue = dev->_rx;

E
Eric Dumazet 已提交
2750 2751
	map = rcu_dereference(rxqueue->rps_map);
	if (map) {
2752
		if (map->len == 1 &&
2753
		    !rcu_access_pointer(rxqueue->rps_flow_table)) {
2754 2755 2756 2757 2758
			tcpu = map->cpus[0];
			if (cpu_online(tcpu))
				cpu = tcpu;
			goto done;
		}
2759
	} else if (!rcu_access_pointer(rxqueue->rps_flow_table)) {
2760
		goto done;
2761
	}
2762

2763
	skb_reset_network_header(skb);
2764 2765 2766
	if (!skb_get_rxhash(skb))
		goto done;

T
Tom Herbert 已提交
2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792
	flow_table = rcu_dereference(rxqueue->rps_flow_table);
	sock_flow_table = rcu_dereference(rps_sock_flow_table);
	if (flow_table && sock_flow_table) {
		u16 next_cpu;
		struct rps_dev_flow *rflow;

		rflow = &flow_table->flows[skb->rxhash & flow_table->mask];
		tcpu = rflow->cpu;

		next_cpu = sock_flow_table->ents[skb->rxhash &
		    sock_flow_table->mask];

		/*
		 * If the desired CPU (where last recvmsg was done) is
		 * different from current CPU (one in the rx-queue flow
		 * table entry), switch if one of the following holds:
		 *   - Current CPU is unset (equal to RPS_NO_CPU).
		 *   - Current CPU is offline.
		 *   - The current CPU's queue tail has advanced beyond the
		 *     last packet that was enqueued using this table entry.
		 *     This guarantees that all previous packets for the flow
		 *     have been dequeued, thus preserving in order delivery.
		 */
		if (unlikely(tcpu != next_cpu) &&
		    (tcpu == RPS_NO_CPU || !cpu_online(tcpu) ||
		     ((int)(per_cpu(softnet_data, tcpu).input_queue_head -
2793 2794 2795
		      rflow->last_qtail)) >= 0))
			rflow = set_rps_cpu(dev, skb, rflow, next_cpu);

T
Tom Herbert 已提交
2796 2797 2798 2799 2800 2801 2802
		if (tcpu != RPS_NO_CPU && cpu_online(tcpu)) {
			*rflowp = rflow;
			cpu = tcpu;
			goto done;
		}
	}

T
Tom Herbert 已提交
2803
	if (map) {
T
Tom Herbert 已提交
2804
		tcpu = map->cpus[((u64) skb->rxhash * map->len) >> 32];
T
Tom Herbert 已提交
2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815

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

done:
	return cpu;
}

2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855
#ifdef CONFIG_RFS_ACCEL

/**
 * rps_may_expire_flow - check whether an RFS hardware filter may be removed
 * @dev: Device on which the filter was set
 * @rxq_index: RX queue index
 * @flow_id: Flow ID passed to ndo_rx_flow_steer()
 * @filter_id: Filter ID returned by ndo_rx_flow_steer()
 *
 * Drivers that implement ndo_rx_flow_steer() should periodically call
 * this function for each installed filter and remove the filters for
 * which it returns %true.
 */
bool rps_may_expire_flow(struct net_device *dev, u16 rxq_index,
			 u32 flow_id, u16 filter_id)
{
	struct netdev_rx_queue *rxqueue = dev->_rx + rxq_index;
	struct rps_dev_flow_table *flow_table;
	struct rps_dev_flow *rflow;
	bool expire = true;
	int cpu;

	rcu_read_lock();
	flow_table = rcu_dereference(rxqueue->rps_flow_table);
	if (flow_table && flow_id <= flow_table->mask) {
		rflow = &flow_table->flows[flow_id];
		cpu = ACCESS_ONCE(rflow->cpu);
		if (rflow->filter == filter_id && cpu != RPS_NO_CPU &&
		    ((int)(per_cpu(softnet_data, cpu).input_queue_head -
			   rflow->last_qtail) <
		     (int)(10 * flow_table->mask)))
			expire = false;
	}
	rcu_read_unlock();
	return expire;
}
EXPORT_SYMBOL(rps_may_expire_flow);

#endif /* CONFIG_RFS_ACCEL */

T
Tom Herbert 已提交
2856
/* Called from hardirq (IPI) context */
E
Eric Dumazet 已提交
2857
static void rps_trigger_softirq(void *data)
T
Tom Herbert 已提交
2858
{
E
Eric Dumazet 已提交
2859 2860
	struct softnet_data *sd = data;

E
Eric Dumazet 已提交
2861
	____napi_schedule(sd, &sd->backlog);
C
Changli Gao 已提交
2862
	sd->received_rps++;
T
Tom Herbert 已提交
2863
}
E
Eric Dumazet 已提交
2864

T
Tom Herbert 已提交
2865
#endif /* CONFIG_RPS */
T
Tom Herbert 已提交
2866

E
Eric Dumazet 已提交
2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887
/*
 * Check if this softnet_data structure is another cpu one
 * If yes, queue it to our IPI list and return 1
 * If no, return 0
 */
static int rps_ipi_queued(struct softnet_data *sd)
{
#ifdef CONFIG_RPS
	struct softnet_data *mysd = &__get_cpu_var(softnet_data);

	if (sd != mysd) {
		sd->rps_ipi_next = mysd->rps_ipi_list;
		mysd->rps_ipi_list = sd;

		__raise_softirq_irqoff(NET_RX_SOFTIRQ);
		return 1;
	}
#endif /* CONFIG_RPS */
	return 0;
}

T
Tom Herbert 已提交
2888 2889 2890 2891
/*
 * enqueue_to_backlog is called to queue an skb to a per CPU backlog
 * queue (may be a remote CPU queue).
 */
T
Tom Herbert 已提交
2892 2893
static int enqueue_to_backlog(struct sk_buff *skb, int cpu,
			      unsigned int *qtail)
T
Tom Herbert 已提交
2894
{
E
Eric Dumazet 已提交
2895
	struct softnet_data *sd;
T
Tom Herbert 已提交
2896 2897
	unsigned long flags;

E
Eric Dumazet 已提交
2898
	sd = &per_cpu(softnet_data, cpu);
T
Tom Herbert 已提交
2899 2900 2901

	local_irq_save(flags);

E
Eric Dumazet 已提交
2902
	rps_lock(sd);
2903 2904
	if (skb_queue_len(&sd->input_pkt_queue) <= netdev_max_backlog) {
		if (skb_queue_len(&sd->input_pkt_queue)) {
T
Tom Herbert 已提交
2905
enqueue:
E
Eric Dumazet 已提交
2906
			__skb_queue_tail(&sd->input_pkt_queue, skb);
2907
			input_queue_tail_incr_save(sd, qtail);
E
Eric Dumazet 已提交
2908
			rps_unlock(sd);
2909
			local_irq_restore(flags);
T
Tom Herbert 已提交
2910 2911 2912
			return NET_RX_SUCCESS;
		}

2913 2914 2915 2916
		/* Schedule NAPI for backlog device
		 * We can use non atomic operation since we own the queue lock
		 */
		if (!__test_and_set_bit(NAPI_STATE_SCHED, &sd->backlog.state)) {
E
Eric Dumazet 已提交
2917
			if (!rps_ipi_queued(sd))
E
Eric Dumazet 已提交
2918
				____napi_schedule(sd, &sd->backlog);
T
Tom Herbert 已提交
2919 2920 2921 2922
		}
		goto enqueue;
	}

C
Changli Gao 已提交
2923
	sd->dropped++;
E
Eric Dumazet 已提交
2924
	rps_unlock(sd);
T
Tom Herbert 已提交
2925 2926 2927

	local_irq_restore(flags);

2928
	atomic_long_inc(&skb->dev->rx_dropped);
T
Tom Herbert 已提交
2929 2930 2931
	kfree_skb(skb);
	return NET_RX_DROP;
}
L
Linus Torvalds 已提交
2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949

/**
 *	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)
{
2950
	int ret;
L
Linus Torvalds 已提交
2951 2952 2953 2954 2955

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

2956
	net_timestamp_check(netdev_tstamp_prequeue, skb);
L
Linus Torvalds 已提交
2957

2958
	trace_netif_rx(skb);
E
Eric Dumazet 已提交
2959
#ifdef CONFIG_RPS
2960
	if (static_key_false(&rps_needed)) {
T
Tom Herbert 已提交
2961
		struct rps_dev_flow voidflow, *rflow = &voidflow;
2962 2963
		int cpu;

2964
		preempt_disable();
2965
		rcu_read_lock();
T
Tom Herbert 已提交
2966 2967

		cpu = get_rps_cpu(skb->dev, skb, &rflow);
2968 2969
		if (cpu < 0)
			cpu = smp_processor_id();
T
Tom Herbert 已提交
2970 2971 2972

		ret = enqueue_to_backlog(skb, cpu, &rflow->last_qtail);

2973
		rcu_read_unlock();
2974
		preempt_enable();
2975 2976
	} else
#endif
T
Tom Herbert 已提交
2977 2978 2979 2980 2981
	{
		unsigned int qtail;
		ret = enqueue_to_backlog(skb, get_cpu(), &qtail);
		put_cpu();
	}
2982
	return ret;
L
Linus Torvalds 已提交
2983
}
E
Eric Dumazet 已提交
2984
EXPORT_SYMBOL(netif_rx);
L
Linus Torvalds 已提交
2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015

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;

3016
			WARN_ON(atomic_read(&skb->users));
3017
			trace_kfree_skb(skb, net_tx_action);
L
Linus Torvalds 已提交
3018 3019 3020 3021 3022
			__kfree_skb(skb);
		}
	}

	if (sd->output_queue) {
3023
		struct Qdisc *head;
L
Linus Torvalds 已提交
3024 3025 3026 3027

		local_irq_disable();
		head = sd->output_queue;
		sd->output_queue = NULL;
3028
		sd->output_queue_tailp = &sd->output_queue;
L
Linus Torvalds 已提交
3029 3030 3031
		local_irq_enable();

		while (head) {
3032 3033 3034
			struct Qdisc *q = head;
			spinlock_t *root_lock;

L
Linus Torvalds 已提交
3035 3036
			head = head->next_sched;

3037
			root_lock = qdisc_lock(q);
3038
			if (spin_trylock(root_lock)) {
3039 3040 3041
				smp_mb__before_clear_bit();
				clear_bit(__QDISC_STATE_SCHED,
					  &q->state);
3042 3043
				qdisc_run(q);
				spin_unlock(root_lock);
L
Linus Torvalds 已提交
3044
			} else {
3045
				if (!test_bit(__QDISC_STATE_DEACTIVATED,
3046
					      &q->state)) {
3047
					__netif_reschedule(q);
3048 3049 3050 3051 3052
				} else {
					smp_mb__before_clear_bit();
					clear_bit(__QDISC_STATE_SCHED,
						  &q->state);
				}
L
Linus Torvalds 已提交
3053 3054 3055 3056 3057
			}
		}
	}
}

3058 3059
#if (defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)) && \
    (defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE))
3060 3061 3062
/* This hook is defined here for ATM LANE */
int (*br_fdb_test_addr_hook)(struct net_device *dev,
			     unsigned char *addr) __read_mostly;
3063
EXPORT_SYMBOL_GPL(br_fdb_test_addr_hook);
3064
#endif
L
Linus Torvalds 已提交
3065 3066 3067 3068 3069 3070

#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
L
Lucas De Marchi 已提交
3071 3072
 * NOTE: This doesn't stop any functionality; if you dont have
 * the ingress scheduler, you just can't add policies on ingress.
L
Linus Torvalds 已提交
3073 3074
 *
 */
3075
static int ing_filter(struct sk_buff *skb, struct netdev_queue *rxq)
L
Linus Torvalds 已提交
3076 3077
{
	struct net_device *dev = skb->dev;
3078
	u32 ttl = G_TC_RTTL(skb->tc_verd);
3079 3080
	int result = TC_ACT_OK;
	struct Qdisc *q;
3081

3082 3083
	if (unlikely(MAX_RED_LOOP < ttl++)) {
		if (net_ratelimit())
3084 3085
			pr_warn("Redir loop detected Dropping packet (%d->%d)\n",
				skb->skb_iif, dev->ifindex);
3086 3087
		return TC_ACT_SHOT;
	}
L
Linus Torvalds 已提交
3088

3089 3090
	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
3091

3092
	q = rxq->qdisc;
3093
	if (q != &noop_qdisc) {
3094
		spin_lock(qdisc_lock(q));
3095 3096
		if (likely(!test_bit(__QDISC_STATE_DEACTIVATED, &q->state)))
			result = qdisc_enqueue_root(skb, q);
3097 3098
		spin_unlock(qdisc_lock(q));
	}
3099 3100 3101

	return result;
}
3102

3103 3104 3105 3106
static inline struct sk_buff *handle_ing(struct sk_buff *skb,
					 struct packet_type **pt_prev,
					 int *ret, struct net_device *orig_dev)
{
3107 3108 3109
	struct netdev_queue *rxq = rcu_dereference(skb->dev->ingress_queue);

	if (!rxq || rxq->qdisc == &noop_qdisc)
3110
		goto out;
L
Linus Torvalds 已提交
3111

3112 3113 3114
	if (*pt_prev) {
		*ret = deliver_skb(skb, *pt_prev, orig_dev);
		*pt_prev = NULL;
L
Linus Torvalds 已提交
3115 3116
	}

3117
	switch (ing_filter(skb, rxq)) {
3118 3119 3120 3121 3122 3123 3124 3125 3126
	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
3127 3128 3129
}
#endif

3130 3131 3132 3133
/**
 *	netdev_rx_handler_register - register receive handler
 *	@dev: device to register a handler for
 *	@rx_handler: receive handler to register
J
Jiri Pirko 已提交
3134
 *	@rx_handler_data: data pointer that is used by rx handler
3135 3136 3137 3138 3139 3140
 *
 *	Register a receive hander for a device. This handler will then be
 *	called from __netif_receive_skb. A negative errno code is returned
 *	on a failure.
 *
 *	The caller must hold the rtnl_mutex.
3141 3142
 *
 *	For a general description of rx_handler, see enum rx_handler_result.
3143 3144
 */
int netdev_rx_handler_register(struct net_device *dev,
J
Jiri Pirko 已提交
3145 3146
			       rx_handler_func_t *rx_handler,
			       void *rx_handler_data)
3147 3148 3149 3150 3151 3152
{
	ASSERT_RTNL();

	if (dev->rx_handler)
		return -EBUSY;

J
Jiri Pirko 已提交
3153
	rcu_assign_pointer(dev->rx_handler_data, rx_handler_data);
3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171
	rcu_assign_pointer(dev->rx_handler, rx_handler);

	return 0;
}
EXPORT_SYMBOL_GPL(netdev_rx_handler_register);

/**
 *	netdev_rx_handler_unregister - unregister receive handler
 *	@dev: device to unregister a handler from
 *
 *	Unregister a receive hander from a device.
 *
 *	The caller must hold the rtnl_mutex.
 */
void netdev_rx_handler_unregister(struct net_device *dev)
{

	ASSERT_RTNL();
3172 3173
	RCU_INIT_POINTER(dev->rx_handler, NULL);
	RCU_INIT_POINTER(dev->rx_handler_data, NULL);
3174 3175 3176
}
EXPORT_SYMBOL_GPL(netdev_rx_handler_unregister);

3177
static int __netif_receive_skb(struct sk_buff *skb)
L
Linus Torvalds 已提交
3178 3179
{
	struct packet_type *ptype, *pt_prev;
3180
	rx_handler_func_t *rx_handler;
D
David S. Miller 已提交
3181
	struct net_device *orig_dev;
3182
	struct net_device *null_or_dev;
3183
	bool deliver_exact = false;
L
Linus Torvalds 已提交
3184
	int ret = NET_RX_DROP;
A
Al Viro 已提交
3185
	__be16 type;
L
Linus Torvalds 已提交
3186

3187
	net_timestamp_check(!netdev_tstamp_prequeue, skb);
3188

3189
	trace_netif_receive_skb(skb);
3190

L
Linus Torvalds 已提交
3191
	/* if we've gotten here through NAPI, check netpoll */
3192
	if (netpoll_receive_skb(skb))
L
Linus Torvalds 已提交
3193 3194
		return NET_RX_DROP;

3195 3196
	if (!skb->skb_iif)
		skb->skb_iif = skb->dev->ifindex;
J
Joe Eykholt 已提交
3197
	orig_dev = skb->dev;
3198

3199
	skb_reset_network_header(skb);
3200
	skb_reset_transport_header(skb);
3201
	skb_reset_mac_len(skb);
L
Linus Torvalds 已提交
3202 3203 3204 3205 3206

	pt_prev = NULL;

	rcu_read_lock();

3207 3208 3209 3210
another_round:

	__this_cpu_inc(softnet_data.processed);

3211 3212 3213 3214 3215 3216
	if (skb->protocol == cpu_to_be16(ETH_P_8021Q)) {
		skb = vlan_untag(skb);
		if (unlikely(!skb))
			goto out;
	}

L
Linus Torvalds 已提交
3217 3218 3219 3220 3221 3222 3223 3224
#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) {
3225
		if (!ptype->dev || ptype->dev == skb->dev) {
3226
			if (pt_prev)
D
David S. Miller 已提交
3227
				ret = deliver_skb(skb, pt_prev, orig_dev);
L
Linus Torvalds 已提交
3228 3229 3230 3231 3232
			pt_prev = ptype;
		}
	}

#ifdef CONFIG_NET_CLS_ACT
3233 3234
	skb = handle_ing(skb, &pt_prev, &ret, orig_dev);
	if (!skb)
L
Linus Torvalds 已提交
3235 3236 3237 3238
		goto out;
ncls:
#endif

3239
	rx_handler = rcu_dereference(skb->dev->rx_handler);
3240 3241 3242 3243 3244
	if (vlan_tx_tag_present(skb)) {
		if (pt_prev) {
			ret = deliver_skb(skb, pt_prev, orig_dev);
			pt_prev = NULL;
		}
3245
		if (vlan_do_receive(&skb, !rx_handler))
3246 3247 3248 3249 3250
			goto another_round;
		else if (unlikely(!skb))
			goto out;
	}

3251 3252 3253 3254 3255
	if (rx_handler) {
		if (pt_prev) {
			ret = deliver_skb(skb, pt_prev, orig_dev);
			pt_prev = NULL;
		}
3256 3257
		switch (rx_handler(&skb)) {
		case RX_HANDLER_CONSUMED:
3258
			goto out;
3259
		case RX_HANDLER_ANOTHER:
3260
			goto another_round;
3261 3262 3263 3264 3265 3266 3267
		case RX_HANDLER_EXACT:
			deliver_exact = true;
		case RX_HANDLER_PASS:
			break;
		default:
			BUG();
		}
3268
	}
L
Linus Torvalds 已提交
3269

3270
	/* deliver only exact match when indicated */
3271
	null_or_dev = deliver_exact ? skb->dev : NULL;
3272

L
Linus Torvalds 已提交
3273
	type = skb->protocol;
3274 3275
	list_for_each_entry_rcu(ptype,
			&ptype_base[ntohs(type) & PTYPE_HASH_MASK], list) {
3276
		if (ptype->type == type &&
3277 3278
		    (ptype->dev == null_or_dev || ptype->dev == skb->dev ||
		     ptype->dev == orig_dev)) {
3279
			if (pt_prev)
D
David S. Miller 已提交
3280
				ret = deliver_skb(skb, pt_prev, orig_dev);
L
Linus Torvalds 已提交
3281 3282 3283 3284 3285
			pt_prev = ptype;
		}
	}

	if (pt_prev) {
D
David S. Miller 已提交
3286
		ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
L
Linus Torvalds 已提交
3287
	} else {
3288
		atomic_long_inc(&skb->dev->rx_dropped);
L
Linus Torvalds 已提交
3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299
		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
3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317

/**
 *	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)
{
3318
	net_timestamp_check(netdev_tstamp_prequeue, skb);
E
Eric Dumazet 已提交
3319

3320 3321 3322
	if (skb_defer_rx_timestamp(skb))
		return NET_RX_SUCCESS;

E
Eric Dumazet 已提交
3323
#ifdef CONFIG_RPS
3324
	if (static_key_false(&rps_needed)) {
E
Eric Dumazet 已提交
3325 3326
		struct rps_dev_flow voidflow, *rflow = &voidflow;
		int cpu, ret;
T
Tom Herbert 已提交
3327

E
Eric Dumazet 已提交
3328 3329 3330
		rcu_read_lock();

		cpu = get_rps_cpu(skb->dev, skb, &rflow);
T
Tom Herbert 已提交
3331

E
Eric Dumazet 已提交
3332 3333 3334
		if (cpu >= 0) {
			ret = enqueue_to_backlog(skb, cpu, &rflow->last_qtail);
			rcu_read_unlock();
3335
			return ret;
E
Eric Dumazet 已提交
3336
		}
3337
		rcu_read_unlock();
T
Tom Herbert 已提交
3338
	}
3339
#endif
3340
	return __netif_receive_skb(skb);
T
Tom Herbert 已提交
3341
}
E
Eric Dumazet 已提交
3342
EXPORT_SYMBOL(netif_receive_skb);
L
Linus Torvalds 已提交
3343

E
Eric Dumazet 已提交
3344 3345 3346
/* Network device is going away, flush any packets still pending
 * Called with irqs disabled.
 */
3347
static void flush_backlog(void *arg)
3348
{
3349
	struct net_device *dev = arg;
E
Eric Dumazet 已提交
3350
	struct softnet_data *sd = &__get_cpu_var(softnet_data);
3351 3352
	struct sk_buff *skb, *tmp;

E
Eric Dumazet 已提交
3353
	rps_lock(sd);
3354
	skb_queue_walk_safe(&sd->input_pkt_queue, skb, tmp) {
3355
		if (skb->dev == dev) {
E
Eric Dumazet 已提交
3356
			__skb_unlink(skb, &sd->input_pkt_queue);
3357
			kfree_skb(skb);
3358
			input_queue_head_incr(sd);
3359
		}
3360
	}
E
Eric Dumazet 已提交
3361
	rps_unlock(sd);
3362 3363 3364 3365 3366

	skb_queue_walk_safe(&sd->process_queue, skb, tmp) {
		if (skb->dev == dev) {
			__skb_unlink(skb, &sd->process_queue);
			kfree_skb(skb);
3367
			input_queue_head_incr(sd);
3368 3369
		}
	}
3370 3371
}

3372 3373 3374 3375 3376 3377 3378
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;

3379 3380
	if (NAPI_GRO_CB(skb)->count == 1) {
		skb_shinfo(skb)->gso_size = 0;
3381
		goto out;
3382
	}
3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403

	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);
}

E
Eric Dumazet 已提交
3404
inline void napi_gro_flush(struct napi_struct *napi)
3405 3406 3407 3408 3409 3410 3411 3412 3413
{
	struct sk_buff *skb, *next;

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

3414
	napi->gro_count = 0;
3415 3416
	napi->gro_list = NULL;
}
E
Eric Dumazet 已提交
3417
EXPORT_SYMBOL(napi_gro_flush);
3418

3419
enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
3420 3421 3422 3423 3424
{
	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
3425
	int same_flow;
3426
	int mac_len;
3427
	enum gro_result ret;
3428

3429
	if (!(skb->dev->features & NETIF_F_GRO) || netpoll_rx_on(skb))
3430 3431
		goto normal;

3432
	if (skb_is_gso(skb) || skb_has_frag_list(skb))
3433 3434
		goto normal;

3435 3436 3437 3438 3439
	rcu_read_lock();
	list_for_each_entry_rcu(ptype, head, list) {
		if (ptype->type != type || ptype->dev || !ptype->gro_receive)
			continue;

3440
		skb_set_network_header(skb, skb_gro_offset(skb));
3441 3442 3443 3444
		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
3445
		NAPI_GRO_CB(skb)->free = 0;
3446 3447 3448 3449 3450 3451 3452 3453 3454

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

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

H
Herbert Xu 已提交
3455
	same_flow = NAPI_GRO_CB(skb)->same_flow;
3456
	ret = NAPI_GRO_CB(skb)->free ? GRO_MERGED_FREE : GRO_MERGED;
H
Herbert Xu 已提交
3457

3458 3459 3460 3461 3462 3463
	if (pp) {
		struct sk_buff *nskb = *pp;

		*pp = nskb->next;
		nskb->next = NULL;
		napi_gro_complete(nskb);
3464
		napi->gro_count--;
3465 3466
	}

H
Herbert Xu 已提交
3467
	if (same_flow)
3468 3469
		goto ok;

3470
	if (NAPI_GRO_CB(skb)->flush || napi->gro_count >= MAX_GRO_SKBS)
3471 3472
		goto normal;

3473
	napi->gro_count++;
3474
	NAPI_GRO_CB(skb)->count = 1;
3475
	skb_shinfo(skb)->gso_size = skb_gro_len(skb);
3476 3477
	skb->next = napi->gro_list;
	napi->gro_list = skb;
3478
	ret = GRO_HELD;
3479

3480
pull:
H
Herbert Xu 已提交
3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491
	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;
E
Eric Dumazet 已提交
3492
		skb_frag_size_sub(&skb_shinfo(skb)->frags[0], grow);
H
Herbert Xu 已提交
3493

E
Eric Dumazet 已提交
3494
		if (unlikely(!skb_frag_size(&skb_shinfo(skb)->frags[0]))) {
3495
			skb_frag_unref(skb, 0);
H
Herbert Xu 已提交
3496 3497
			memmove(skb_shinfo(skb)->frags,
				skb_shinfo(skb)->frags + 1,
3498
				--skb_shinfo(skb)->nr_frags * sizeof(skb_frag_t));
H
Herbert Xu 已提交
3499
		}
3500 3501
	}

3502
ok:
3503
	return ret;
3504 3505

normal:
3506 3507
	ret = GRO_NORMAL;
	goto pull;
H
Herbert Xu 已提交
3508
}
3509 3510
EXPORT_SYMBOL(dev_gro_receive);

3511
static inline gro_result_t
3512
__napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
3513 3514
{
	struct sk_buff *p;
E
Eric Dumazet 已提交
3515
	unsigned int maclen = skb->dev->hard_header_len;
3516 3517

	for (p = napi->gro_list; p; p = p->next) {
3518 3519 3520
		unsigned long diffs;

		diffs = (unsigned long)p->dev ^ (unsigned long)skb->dev;
3521
		diffs |= p->vlan_tci ^ skb->vlan_tci;
E
Eric Dumazet 已提交
3522 3523 3524 3525 3526 3527 3528
		if (maclen == ETH_HLEN)
			diffs |= compare_ether_header(skb_mac_header(p),
						      skb_gro_mac_header(skb));
		else if (!diffs)
			diffs = memcmp(skb_mac_header(p),
				       skb_gro_mac_header(skb),
				       maclen);
3529
		NAPI_GRO_CB(p)->same_flow = !diffs;
3530 3531 3532 3533 3534
		NAPI_GRO_CB(p)->flush = 0;
	}

	return dev_gro_receive(napi, skb);
}
H
Herbert Xu 已提交
3535

3536
gro_result_t napi_skb_finish(gro_result_t ret, struct sk_buff *skb)
H
Herbert Xu 已提交
3537
{
3538 3539
	switch (ret) {
	case GRO_NORMAL:
3540 3541 3542
		if (netif_receive_skb(skb))
			ret = GRO_DROP;
		break;
H
Herbert Xu 已提交
3543

3544
	case GRO_DROP:
H
Herbert Xu 已提交
3545 3546
		kfree_skb(skb);
		break;
3547

3548
	case GRO_MERGED_FREE:
3549 3550 3551 3552
		if (NAPI_GRO_CB(skb)->free == NAPI_GRO_FREE_STOLEN_HEAD)
			kmem_cache_free(skbuff_head_cache, skb);
		else
			__kfree_skb(skb);
3553 3554
		break;

3555 3556 3557
	case GRO_HELD:
	case GRO_MERGED:
		break;
H
Herbert Xu 已提交
3558 3559
	}

3560
	return ret;
3561 3562 3563
}
EXPORT_SYMBOL(napi_skb_finish);

3564 3565 3566 3567
void skb_gro_reset_offset(struct sk_buff *skb)
{
	NAPI_GRO_CB(skb)->data_offset = 0;
	NAPI_GRO_CB(skb)->frag0 = NULL;
3568
	NAPI_GRO_CB(skb)->frag0_len = 0;
3569

3570
	if (skb->mac_header == skb->tail &&
3571
	    !PageHighMem(skb_frag_page(&skb_shinfo(skb)->frags[0]))) {
3572
		NAPI_GRO_CB(skb)->frag0 =
3573
			skb_frag_address(&skb_shinfo(skb)->frags[0]);
E
Eric Dumazet 已提交
3574
		NAPI_GRO_CB(skb)->frag0_len = skb_frag_size(&skb_shinfo(skb)->frags[0]);
3575
	}
3576 3577 3578
}
EXPORT_SYMBOL(skb_gro_reset_offset);

3579
gro_result_t napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
3580
{
3581 3582
	skb_gro_reset_offset(skb);

3583
	return napi_skb_finish(__napi_gro_receive(napi, skb), skb);
3584 3585 3586
}
EXPORT_SYMBOL(napi_gro_receive);

S
stephen hemminger 已提交
3587
static void napi_reuse_skb(struct napi_struct *napi, struct sk_buff *skb)
3588 3589
{
	__skb_pull(skb, skb_headlen(skb));
3590 3591
	/* restore the reserve we had after netdev_alloc_skb_ip_align() */
	skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN - skb_headroom(skb));
3592
	skb->vlan_tci = 0;
H
Herbert Xu 已提交
3593
	skb->dev = napi->dev;
A
Andy Gospodarek 已提交
3594
	skb->skb_iif = 0;
3595 3596 3597 3598

	napi->skb = skb;
}

3599
struct sk_buff *napi_get_frags(struct napi_struct *napi)
H
Herbert Xu 已提交
3600 3601 3602 3603
{
	struct sk_buff *skb = napi->skb;

	if (!skb) {
3604 3605 3606
		skb = netdev_alloc_skb_ip_align(napi->dev, GRO_MAX_HEAD);
		if (skb)
			napi->skb = skb;
3607
	}
3608 3609
	return skb;
}
3610
EXPORT_SYMBOL(napi_get_frags);
3611

3612 3613
gro_result_t napi_frags_finish(struct napi_struct *napi, struct sk_buff *skb,
			       gro_result_t ret)
3614
{
3615 3616
	switch (ret) {
	case GRO_NORMAL:
3617
	case GRO_HELD:
3618
		skb->protocol = eth_type_trans(skb, skb->dev);
3619

3620 3621 3622 3623
		if (ret == GRO_HELD)
			skb_gro_pull(skb, -ETH_HLEN);
		else if (netif_receive_skb(skb))
			ret = GRO_DROP;
3624
		break;
H
Herbert Xu 已提交
3625

3626 3627 3628 3629
	case GRO_DROP:
	case GRO_MERGED_FREE:
		napi_reuse_skb(napi, skb);
		break;
3630 3631 3632

	case GRO_MERGED:
		break;
3633
	}
H
Herbert Xu 已提交
3634

3635
	return ret;
H
Herbert Xu 已提交
3636
}
3637 3638
EXPORT_SYMBOL(napi_frags_finish);

3639 3640 3641 3642
struct sk_buff *napi_frags_skb(struct napi_struct *napi)
{
	struct sk_buff *skb = napi->skb;
	struct ethhdr *eth;
3643 3644
	unsigned int hlen;
	unsigned int off;
3645 3646 3647 3648 3649 3650

	napi->skb = NULL;

	skb_reset_mac_header(skb);
	skb_gro_reset_offset(skb);

3651 3652 3653 3654 3655 3656 3657 3658 3659 3660
	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;
		}
3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675
	}

	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);

3676
gro_result_t napi_gro_frags(struct napi_struct *napi)
3677
{
3678
	struct sk_buff *skb = napi_frags_skb(napi);
3679 3680

	if (!skb)
3681
		return GRO_DROP;
3682 3683 3684

	return napi_frags_finish(napi, skb, __napi_gro_receive(napi, skb));
}
H
Herbert Xu 已提交
3685 3686
EXPORT_SYMBOL(napi_gro_frags);

3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714
/*
 * net_rps_action sends any pending IPI's for rps.
 * Note: called with local irq disabled, but exits with local irq enabled.
 */
static void net_rps_action_and_irq_enable(struct softnet_data *sd)
{
#ifdef CONFIG_RPS
	struct softnet_data *remsd = sd->rps_ipi_list;

	if (remsd) {
		sd->rps_ipi_list = NULL;

		local_irq_enable();

		/* Send pending IPI's to kick RPS processing on remote cpus. */
		while (remsd) {
			struct softnet_data *next = remsd->rps_ipi_next;

			if (cpu_online(remsd->cpu))
				__smp_call_function_single(remsd->cpu,
							   &remsd->csd, 0);
			remsd = next;
		}
	} else
#endif
		local_irq_enable();
}

3715
static int process_backlog(struct napi_struct *napi, int quota)
L
Linus Torvalds 已提交
3716 3717
{
	int work = 0;
E
Eric Dumazet 已提交
3718
	struct softnet_data *sd = container_of(napi, struct softnet_data, backlog);
L
Linus Torvalds 已提交
3719

3720 3721 3722 3723 3724 3725 3726 3727 3728
#ifdef CONFIG_RPS
	/* Check if we have pending ipi, its better to send them now,
	 * not waiting net_rx_action() end.
	 */
	if (sd->rps_ipi_list) {
		local_irq_disable();
		net_rps_action_and_irq_enable(sd);
	}
#endif
3729
	napi->weight = weight_p;
3730 3731
	local_irq_disable();
	while (work < quota) {
L
Linus Torvalds 已提交
3732
		struct sk_buff *skb;
3733 3734 3735 3736 3737 3738
		unsigned int qlen;

		while ((skb = __skb_dequeue(&sd->process_queue))) {
			local_irq_enable();
			__netif_receive_skb(skb);
			local_irq_disable();
3739 3740 3741 3742 3743
			input_queue_head_incr(sd);
			if (++work >= quota) {
				local_irq_enable();
				return work;
			}
3744
		}
L
Linus Torvalds 已提交
3745

E
Eric Dumazet 已提交
3746
		rps_lock(sd);
3747
		qlen = skb_queue_len(&sd->input_pkt_queue);
3748
		if (qlen)
3749 3750
			skb_queue_splice_tail_init(&sd->input_pkt_queue,
						   &sd->process_queue);
3751

3752
		if (qlen < quota - work) {
E
Eric Dumazet 已提交
3753 3754 3755 3756 3757 3758 3759 3760 3761 3762
			/*
			 * Inline a custom version of __napi_complete().
			 * only current cpu owns and manipulates this napi,
			 * and NAPI_STATE_SCHED is the only possible flag set on backlog.
			 * we can use a plain write instead of clear_bit(),
			 * and we dont need an smp_mb() memory barrier.
			 */
			list_del(&napi->poll_list);
			napi->state = 0;

3763
			quota = work + qlen;
3764
		}
E
Eric Dumazet 已提交
3765
		rps_unlock(sd);
3766 3767
	}
	local_irq_enable();
L
Linus Torvalds 已提交
3768

3769 3770
	return work;
}
L
Linus Torvalds 已提交
3771

3772 3773
/**
 * __napi_schedule - schedule for receive
3774
 * @n: entry to schedule
3775 3776 3777
 *
 * The entry's receive function will be scheduled to run
 */
H
Harvey Harrison 已提交
3778
void __napi_schedule(struct napi_struct *n)
3779 3780
{
	unsigned long flags;
L
Linus Torvalds 已提交
3781

3782
	local_irq_save(flags);
E
Eric Dumazet 已提交
3783
	____napi_schedule(&__get_cpu_var(softnet_data), n);
3784
	local_irq_restore(flags);
L
Linus Torvalds 已提交
3785
}
3786 3787
EXPORT_SYMBOL(__napi_schedule);

3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820
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);
3821
	napi->gro_count = 0;
3822
	napi->gro_list = NULL;
H
Herbert Xu 已提交
3823
	napi->skb = NULL;
3824 3825 3826 3827
	napi->poll = poll;
	napi->weight = weight;
	list_add(&napi->dev_list, &dev->napi_list);
	napi->dev = dev;
H
Herbert Xu 已提交
3828
#ifdef CONFIG_NETPOLL
3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839
	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;

3840
	list_del_init(&napi->dev_list);
3841
	napi_free_frags(napi);
3842 3843 3844 3845 3846 3847 3848 3849

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

	napi->gro_list = NULL;
3850
	napi->gro_count = 0;
3851 3852 3853
}
EXPORT_SYMBOL(netif_napi_del);

L
Linus Torvalds 已提交
3854 3855
static void net_rx_action(struct softirq_action *h)
{
3856
	struct softnet_data *sd = &__get_cpu_var(softnet_data);
3857
	unsigned long time_limit = jiffies + 2;
3858
	int budget = netdev_budget;
3859 3860
	void *have;

L
Linus Torvalds 已提交
3861 3862
	local_irq_disable();

3863
	while (!list_empty(&sd->poll_list)) {
3864 3865
		struct napi_struct *n;
		int work, weight;
L
Linus Torvalds 已提交
3866

3867
		/* If softirq window is exhuasted then punt.
3868 3869
		 * Allow this to run for 2 jiffies since which will allow
		 * an average latency of 1.5/HZ.
3870
		 */
3871
		if (unlikely(budget <= 0 || time_after(jiffies, time_limit)))
L
Linus Torvalds 已提交
3872 3873 3874 3875
			goto softnet_break;

		local_irq_enable();

3876 3877 3878 3879 3880
		/* 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.
		 */
3881
		n = list_first_entry(&sd->poll_list, struct napi_struct, poll_list);
L
Linus Torvalds 已提交
3882

3883 3884 3885 3886
		have = netpoll_poll_lock(n);

		weight = n->weight;

3887 3888 3889 3890
		/* 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
L
Lucas De Marchi 已提交
3891
		 * accidentally calling ->poll() when NAPI is not scheduled.
3892 3893
		 */
		work = 0;
3894
		if (test_bit(NAPI_STATE_SCHED, &n->state)) {
3895
			work = n->poll(n, weight);
3896 3897
			trace_napi_poll(n);
		}
3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909

		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.
		 */
3910
		if (unlikely(work == weight)) {
3911 3912 3913 3914 3915
			if (unlikely(napi_disable_pending(n))) {
				local_irq_enable();
				napi_complete(n);
				local_irq_disable();
			} else
3916
				list_move_tail(&n->poll_list, &sd->poll_list);
3917
		}
3918 3919

		netpoll_poll_unlock(have);
L
Linus Torvalds 已提交
3920 3921
	}
out:
3922
	net_rps_action_and_irq_enable(sd);
T
Tom Herbert 已提交
3923

3924 3925 3926 3927 3928
#ifdef CONFIG_NET_DMA
	/*
	 * There may not be any more sk_buffs coming right now, so push
	 * any pending DMA copies to hardware
	 */
3929
	dma_issue_pending_all();
3930
#endif
3931

L
Linus Torvalds 已提交
3932 3933 3934
	return;

softnet_break:
C
Changli Gao 已提交
3935
	sd->time_squeeze++;
L
Linus Torvalds 已提交
3936 3937 3938 3939
	__raise_softirq_irqoff(NET_RX_SOFTIRQ);
	goto out;
}

E
Eric Dumazet 已提交
3940
static gifconf_func_t *gifconf_list[NPROTO];
L
Linus Torvalds 已提交
3941 3942 3943 3944 3945 3946 3947 3948 3949 3950

/**
 *	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
3951
int register_gifconf(unsigned int family, gifconf_func_t *gifconf)
L
Linus Torvalds 已提交
3952 3953 3954 3955 3956 3957
{
	if (family >= NPROTO)
		return -EINVAL;
	gifconf_list[family] = gifconf;
	return 0;
}
E
Eric Dumazet 已提交
3958
EXPORT_SYMBOL(register_gifconf);
L
Linus Torvalds 已提交
3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971


/*
 *	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
 */

3972
static int dev_ifname(struct net *net, struct ifreq __user *arg)
L
Linus Torvalds 已提交
3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983
{
	struct net_device *dev;
	struct ifreq ifr;

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

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

3984 3985
	rcu_read_lock();
	dev = dev_get_by_index_rcu(net, ifr.ifr_ifindex);
L
Linus Torvalds 已提交
3986
	if (!dev) {
3987
		rcu_read_unlock();
L
Linus Torvalds 已提交
3988 3989 3990 3991
		return -ENODEV;
	}

	strcpy(ifr.ifr_name, dev->name);
3992
	rcu_read_unlock();
L
Linus Torvalds 已提交
3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004

	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'.
 */

4005
static int dev_ifconf(struct net *net, char __user *arg)
L
Linus Torvalds 已提交
4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028
{
	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;
4029
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042
		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;
			}
		}
4043
	}
L
Linus Torvalds 已提交
4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056

	/*
	 *	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
4057

E
Eric Dumazet 已提交
4058
#define BUCKET_SPACE (32 - NETDEV_HASHBITS - 1)
4059 4060 4061 4062 4063

#define get_bucket(x) ((x) >> BUCKET_SPACE)
#define get_offset(x) ((x) & ((1 << BUCKET_SPACE) - 1))
#define set_bucket_offset(b, o) ((b) << BUCKET_SPACE | (o))

E
Eric Dumazet 已提交
4064
static inline struct net_device *dev_from_same_bucket(struct seq_file *seq, loff_t *pos)
4065 4066 4067 4068 4069
{
	struct net *net = seq_file_net(seq);
	struct net_device *dev;
	struct hlist_node *p;
	struct hlist_head *h;
E
Eric Dumazet 已提交
4070
	unsigned int count = 0, offset = get_offset(*pos);
4071

E
Eric Dumazet 已提交
4072
	h = &net->dev_name_head[get_bucket(*pos)];
4073
	hlist_for_each_entry_rcu(dev, p, h, name_hlist) {
E
Eric Dumazet 已提交
4074
		if (++count == offset)
4075 4076 4077 4078 4079 4080
			return dev;
	}

	return NULL;
}

E
Eric Dumazet 已提交
4081
static inline struct net_device *dev_from_bucket(struct seq_file *seq, loff_t *pos)
4082 4083 4084 4085 4086
{
	struct net_device *dev;
	unsigned int bucket;

	do {
E
Eric Dumazet 已提交
4087
		dev = dev_from_same_bucket(seq, pos);
4088 4089 4090
		if (dev)
			return dev;

E
Eric Dumazet 已提交
4091 4092
		bucket = get_bucket(*pos) + 1;
		*pos = set_bucket_offset(bucket, 1);
4093 4094 4095 4096 4097
	} while (bucket < NETDEV_HASHENTRIES);

	return NULL;
}

L
Linus Torvalds 已提交
4098 4099 4100 4101
/*
 *	This is invoked by the /proc filesystem handler to display a device
 *	in detail.
 */
4102
void *dev_seq_start(struct seq_file *seq, loff_t *pos)
4103
	__acquires(RCU)
L
Linus Torvalds 已提交
4104
{
4105
	rcu_read_lock();
4106 4107
	if (!*pos)
		return SEQ_START_TOKEN;
L
Linus Torvalds 已提交
4108

E
Eric Dumazet 已提交
4109
	if (get_bucket(*pos) >= NETDEV_HASHENTRIES)
4110
		return NULL;
L
Linus Torvalds 已提交
4111

E
Eric Dumazet 已提交
4112
	return dev_from_bucket(seq, pos);
L
Linus Torvalds 已提交
4113 4114 4115 4116
}

void *dev_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
4117
	++*pos;
E
Eric Dumazet 已提交
4118
	return dev_from_bucket(seq, pos);
L
Linus Torvalds 已提交
4119 4120 4121
}

void dev_seq_stop(struct seq_file *seq, void *v)
4122
	__releases(RCU)
L
Linus Torvalds 已提交
4123
{
4124
	rcu_read_unlock();
L
Linus Torvalds 已提交
4125 4126 4127 4128
}

static void dev_seq_printf_stats(struct seq_file *seq, struct net_device *dev)
{
4129 4130
	struct rtnl_link_stats64 temp;
	const struct rtnl_link_stats64 *stats = dev_get_stats(dev, &temp);
L
Linus Torvalds 已提交
4131

4132 4133
	seq_printf(seq, "%6s: %7llu %7llu %4llu %4llu %4llu %5llu %10llu %9llu "
		   "%8llu %7llu %4llu %4llu %4llu %5llu %7llu %10llu\n",
4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148
		   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
4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167
}

/*
 *	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;
}

C
Changli Gao 已提交
4168
static struct softnet_data *softnet_get_online(loff_t *pos)
L
Linus Torvalds 已提交
4169
{
C
Changli Gao 已提交
4170
	struct softnet_data *sd = NULL;
L
Linus Torvalds 已提交
4171

4172
	while (*pos < nr_cpu_ids)
4173
		if (cpu_online(*pos)) {
C
Changli Gao 已提交
4174
			sd = &per_cpu(softnet_data, *pos);
L
Linus Torvalds 已提交
4175 4176 4177
			break;
		} else
			++*pos;
C
Changli Gao 已提交
4178
	return sd;
L
Linus Torvalds 已提交
4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197
}

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)
{
C
Changli Gao 已提交
4198
	struct softnet_data *sd = v;
L
Linus Torvalds 已提交
4199

T
Tom Herbert 已提交
4200
	seq_printf(seq, "%08x %08x %08x %08x %08x %08x %08x %08x %08x %08x\n",
C
Changli Gao 已提交
4201
		   sd->processed, sd->dropped, sd->time_squeeze, 0,
4202
		   0, 0, 0, 0, /* was fastroute */
C
Changli Gao 已提交
4203
		   sd->cpu_collision, sd->received_rps);
L
Linus Torvalds 已提交
4204 4205 4206
	return 0;
}

4207
static const struct seq_operations dev_seq_ops = {
L
Linus Torvalds 已提交
4208 4209 4210 4211 4212 4213 4214 4215
	.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)
{
4216
	return seq_open_net(inode, file, &dev_seq_ops,
E
Eric Dumazet 已提交
4217
			    sizeof(struct seq_net_private));
4218 4219
}

4220
static const struct file_operations dev_seq_fops = {
L
Linus Torvalds 已提交
4221 4222 4223 4224
	.owner	 = THIS_MODULE,
	.open    = dev_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
4225
	.release = seq_release_net,
L
Linus Torvalds 已提交
4226 4227
};

4228
static const struct seq_operations softnet_seq_ops = {
L
Linus Torvalds 已提交
4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239
	.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);
}

4240
static const struct file_operations softnet_seq_fops = {
L
Linus Torvalds 已提交
4241 4242 4243 4244 4245 4246 4247
	.owner	 = THIS_MODULE,
	.open    = softnet_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release,
};

4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259
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;
	}

4260
	for (t = 0; t < PTYPE_HASH_SIZE; t++) {
4261 4262 4263 4264 4265 4266 4267 4268 4269 4270
		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)
4271
	__acquires(RCU)
4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294
{
	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
4295
		hash = ntohs(pt->type) & PTYPE_HASH_MASK;
4296 4297

	while (nxt == &ptype_base[hash]) {
4298
		if (++hash >= PTYPE_HASH_SIZE)
4299 4300 4301 4302 4303 4304 4305 4306
			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)
4307
	__releases(RCU)
4308 4309 4310 4311 4312 4313 4314 4315 4316 4317
{
	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");
4318
	else if (pt->dev == NULL || dev_net(pt->dev) == seq_file_net(seq)) {
4319 4320 4321 4322 4323
		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
4324 4325
		seq_printf(seq, " %-8s %pF\n",
			   pt->dev ? pt->dev->name : "", pt->func);
4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339
	}

	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)
{
4340 4341
	return seq_open_net(inode, file, &ptype_seq_ops,
			sizeof(struct seq_net_private));
4342 4343 4344 4345 4346 4347 4348
}

static const struct file_operations ptype_seq_fops = {
	.owner	 = THIS_MODULE,
	.open    = ptype_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
4349
	.release = seq_release_net,
4350 4351 4352
};


4353
static int __net_init dev_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
4354 4355 4356
{
	int rc = -ENOMEM;

4357
	if (!proc_net_fops_create(net, "dev", S_IRUGO, &dev_seq_fops))
L
Linus Torvalds 已提交
4358
		goto out;
4359
	if (!proc_net_fops_create(net, "softnet_stat", S_IRUGO, &softnet_seq_fops))
L
Linus Torvalds 已提交
4360
		goto out_dev;
4361
	if (!proc_net_fops_create(net, "ptype", S_IRUGO, &ptype_seq_fops))
4362
		goto out_softnet;
4363

4364
	if (wext_proc_init(net))
4365
		goto out_ptype;
L
Linus Torvalds 已提交
4366 4367 4368
	rc = 0;
out:
	return rc;
4369
out_ptype:
4370
	proc_net_remove(net, "ptype");
L
Linus Torvalds 已提交
4371
out_softnet:
4372
	proc_net_remove(net, "softnet_stat");
L
Linus Torvalds 已提交
4373
out_dev:
4374
	proc_net_remove(net, "dev");
L
Linus Torvalds 已提交
4375 4376
	goto out;
}
4377

4378
static void __net_exit dev_proc_net_exit(struct net *net)
4379 4380 4381 4382 4383 4384 4385 4386
{
	wext_proc_exit(net);

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

4387
static struct pernet_operations __net_initdata dev_proc_ops = {
4388 4389 4390 4391 4392 4393 4394 4395
	.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
4396 4397 4398 4399 4400 4401
#else
#define dev_proc_init() 0
#endif	/* CONFIG_PROC_FS */


/**
J
Jiri Pirko 已提交
4402
 *	netdev_set_master	-	set up master pointer
L
Linus Torvalds 已提交
4403 4404 4405 4406 4407 4408
 *	@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
J
Jiri Pirko 已提交
4409
 *	are adjusted and the function returns zero.
L
Linus Torvalds 已提交
4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423
 */
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;
4424

4425
	if (old)
L
Linus Torvalds 已提交
4426
		dev_put(old);
J
Jiri Pirko 已提交
4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449
	return 0;
}
EXPORT_SYMBOL(netdev_set_master);

/**
 *	netdev_set_bond_master	-	set up bonding 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 %RTM_NEWLINK is sent
 *	to the routing socket and the function returns zero.
 */
int netdev_set_bond_master(struct net_device *slave, struct net_device *master)
{
	int err;

	ASSERT_RTNL();

	err = netdev_set_master(slave, master);
	if (err)
		return err;
L
Linus Torvalds 已提交
4450 4451 4452 4453 4454 4455 4456 4457
	if (master)
		slave->flags |= IFF_SLAVE;
	else
		slave->flags &= ~IFF_SLAVE;

	rtmsg_ifinfo(RTM_NEWLINK, slave, IFF_SLAVE);
	return 0;
}
J
Jiri Pirko 已提交
4458
EXPORT_SYMBOL(netdev_set_bond_master);
L
Linus Torvalds 已提交
4459

4460 4461
static void dev_change_rx_flags(struct net_device *dev, int flags)
{
4462 4463 4464 4465
	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);
4466 4467
}

4468
static int __dev_set_promiscuity(struct net_device *dev, int inc)
L
Linus Torvalds 已提交
4469
{
4470
	unsigned int old_flags = dev->flags;
4471 4472
	uid_t uid;
	gid_t gid;
L
Linus Torvalds 已提交
4473

4474 4475
	ASSERT_RTNL();

4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486
	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;
4487 4488
			pr_warn("%s: promiscuity touches roof, set promiscuity failed. promiscuity feature of device might be broken.\n",
				dev->name);
4489 4490 4491
			return -EOVERFLOW;
		}
	}
4492
	if (dev->flags != old_flags) {
4493 4494 4495
		pr_info("device %s %s promiscuous mode\n",
			dev->name,
			dev->flags & IFF_PROMISC ? "entered" : "left");
4496 4497
		if (audit_enabled) {
			current_uid_gid(&uid, &gid);
4498 4499 4500 4501 4502 4503
			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),
4504
				uid, gid,
4505
				audit_get_sessionid(current));
4506
		}
4507

4508
		dev_change_rx_flags(dev, IFF_PROMISC);
L
Linus Torvalds 已提交
4509
	}
4510
	return 0;
L
Linus Torvalds 已提交
4511 4512
}

4513 4514 4515 4516 4517 4518 4519 4520 4521
/**
 *	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.
4522
 *	Return 0 if successful or a negative errno code on error.
4523
 */
4524
int dev_set_promiscuity(struct net_device *dev, int inc)
4525
{
4526
	unsigned int old_flags = dev->flags;
4527
	int err;
4528

4529
	err = __dev_set_promiscuity(dev, inc);
4530
	if (err < 0)
4531
		return err;
4532 4533
	if (dev->flags != old_flags)
		dev_set_rx_mode(dev);
4534
	return err;
4535
}
E
Eric Dumazet 已提交
4536
EXPORT_SYMBOL(dev_set_promiscuity);
4537

L
Linus Torvalds 已提交
4538 4539 4540 4541 4542 4543 4544 4545 4546 4547
/**
 *	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.
4548
 *	Return 0 if successful or a negative errno code on error.
L
Linus Torvalds 已提交
4549 4550
 */

4551
int dev_set_allmulti(struct net_device *dev, int inc)
L
Linus Torvalds 已提交
4552
{
4553
	unsigned int old_flags = dev->flags;
L
Linus Torvalds 已提交
4554

4555 4556
	ASSERT_RTNL();

L
Linus Torvalds 已提交
4557
	dev->flags |= IFF_ALLMULTI;
4558 4559 4560 4561 4562 4563 4564 4565 4566 4567
	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;
4568 4569
			pr_warn("%s: allmulti touches roof, set allmulti failed. allmulti feature of device might be broken.\n",
				dev->name);
4570 4571 4572
			return -EOVERFLOW;
		}
	}
4573
	if (dev->flags ^ old_flags) {
4574
		dev_change_rx_flags(dev, IFF_ALLMULTI);
4575
		dev_set_rx_mode(dev);
4576
	}
4577
	return 0;
4578
}
E
Eric Dumazet 已提交
4579
EXPORT_SYMBOL(dev_set_allmulti);
4580 4581 4582 4583

/*
 *	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
4584
 *	filtering it is put in promiscuous mode while unicast addresses
4585 4586 4587 4588
 *	are present.
 */
void __dev_set_rx_mode(struct net_device *dev)
{
4589 4590
	const struct net_device_ops *ops = dev->netdev_ops;

4591 4592 4593 4594 4595
	/* dev_open will call this function so the list will stay sane. */
	if (!(dev->flags&IFF_UP))
		return;

	if (!netif_device_present(dev))
4596
		return;
4597

4598
	if (!(dev->priv_flags & IFF_UNICAST_FLT)) {
4599 4600 4601
		/* Unicast addresses changes may only happen under the rtnl,
		 * therefore calling __dev_set_promiscuity here is safe.
		 */
4602
		if (!netdev_uc_empty(dev) && !dev->uc_promisc) {
4603
			__dev_set_promiscuity(dev, 1);
4604
			dev->uc_promisc = true;
4605
		} else if (netdev_uc_empty(dev) && dev->uc_promisc) {
4606
			__dev_set_promiscuity(dev, -1);
4607
			dev->uc_promisc = false;
4608 4609
		}
	}
4610 4611 4612

	if (ops->ndo_set_rx_mode)
		ops->ndo_set_rx_mode(dev);
4613 4614 4615 4616
}

void dev_set_rx_mode(struct net_device *dev)
{
4617
	netif_addr_lock_bh(dev);
4618
	__dev_set_rx_mode(dev);
4619
	netif_addr_unlock_bh(dev);
L
Linus Torvalds 已提交
4620 4621
}

4622 4623 4624 4625 4626 4627
/**
 *	dev_get_flags - get flags reported to userspace
 *	@dev: device
 *
 *	Get the combination of flag bits exported through APIs to userspace.
 */
4628
unsigned int dev_get_flags(const struct net_device *dev)
L
Linus Torvalds 已提交
4629
{
4630
	unsigned int flags;
L
Linus Torvalds 已提交
4631 4632 4633

	flags = (dev->flags & ~(IFF_PROMISC |
				IFF_ALLMULTI |
S
Stefan Rompf 已提交
4634 4635 4636
				IFF_RUNNING |
				IFF_LOWER_UP |
				IFF_DORMANT)) |
L
Linus Torvalds 已提交
4637 4638 4639
		(dev->gflags & (IFF_PROMISC |
				IFF_ALLMULTI));

S
Stefan Rompf 已提交
4640 4641 4642 4643 4644 4645 4646 4647
	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
4648 4649 4650

	return flags;
}
E
Eric Dumazet 已提交
4651
EXPORT_SYMBOL(dev_get_flags);
L
Linus Torvalds 已提交
4652

4653
int __dev_change_flags(struct net_device *dev, unsigned int flags)
L
Linus Torvalds 已提交
4654
{
4655
	unsigned int old_flags = dev->flags;
4656
	int ret;
L
Linus Torvalds 已提交
4657

4658 4659
	ASSERT_RTNL();

L
Linus Torvalds 已提交
4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672 4673
	/*
	 *	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.
	 */

4674 4675
	if ((old_flags ^ flags) & IFF_MULTICAST)
		dev_change_rx_flags(dev, IFF_MULTICAST);
4676

4677
	dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
4678 4679 4680 4681 4682 4683 4684 4685 4686

	/*
	 *	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  ? */
4687
		ret = ((old_flags & IFF_UP) ? __dev_close : __dev_open)(dev);
L
Linus Torvalds 已提交
4688 4689

		if (!ret)
4690
			dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
4691 4692 4693
	}

	if ((flags ^ dev->gflags) & IFF_PROMISC) {
E
Eric Dumazet 已提交
4694 4695
		int inc = (flags & IFF_PROMISC) ? 1 : -1;

L
Linus Torvalds 已提交
4696 4697 4698 4699 4700 4701 4702 4703 4704
		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
4705 4706
		int inc = (flags & IFF_ALLMULTI) ? 1 : -1;

L
Linus Torvalds 已提交
4707 4708 4709 4710
		dev->gflags ^= IFF_ALLMULTI;
		dev_set_allmulti(dev, inc);
	}

4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737
	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.
 */
4738
int dev_change_flags(struct net_device *dev, unsigned int flags)
4739
{
4740 4741
	int ret;
	unsigned int changes, old_flags = dev->flags;
4742 4743 4744 4745 4746 4747

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

	changes = old_flags ^ dev->flags;
4748 4749
	if (changes)
		rtmsg_ifinfo(RTM_NEWLINK, dev, changes);
L
Linus Torvalds 已提交
4750

4751
	__dev_notify_flags(dev, old_flags);
L
Linus Torvalds 已提交
4752 4753
	return ret;
}
E
Eric Dumazet 已提交
4754
EXPORT_SYMBOL(dev_change_flags);
L
Linus Torvalds 已提交
4755

4756 4757 4758 4759 4760 4761 4762
/**
 *	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
4763 4764
int dev_set_mtu(struct net_device *dev, int new_mtu)
{
4765
	const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
4766 4767 4768 4769 4770 4771 4772 4773 4774 4775 4776 4777 4778
	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;
4779 4780
	if (ops->ndo_change_mtu)
		err = ops->ndo_change_mtu(dev, new_mtu);
L
Linus Torvalds 已提交
4781 4782
	else
		dev->mtu = new_mtu;
4783

L
Linus Torvalds 已提交
4784
	if (!err && dev->flags & IFF_UP)
4785
		call_netdevice_notifiers(NETDEV_CHANGEMTU, dev);
L
Linus Torvalds 已提交
4786 4787
	return err;
}
E
Eric Dumazet 已提交
4788
EXPORT_SYMBOL(dev_set_mtu);
L
Linus Torvalds 已提交
4789

4790 4791 4792 4793 4794 4795 4796 4797 4798 4799 4800
/**
 *	dev_set_group - Change group this device belongs to
 *	@dev: device
 *	@new_group: group this device should belong to
 */
void dev_set_group(struct net_device *dev, int new_group)
{
	dev->group = new_group;
}
EXPORT_SYMBOL(dev_set_group);

4801 4802 4803 4804 4805 4806 4807
/**
 *	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
4808 4809
int dev_set_mac_address(struct net_device *dev, struct sockaddr *sa)
{
4810
	const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
4811 4812
	int err;

4813
	if (!ops->ndo_set_mac_address)
L
Linus Torvalds 已提交
4814 4815 4816 4817 4818
		return -EOPNOTSUPP;
	if (sa->sa_family != dev->type)
		return -EINVAL;
	if (!netif_device_present(dev))
		return -ENODEV;
4819
	err = ops->ndo_set_mac_address(dev, sa);
L
Linus Torvalds 已提交
4820
	if (!err)
4821
		call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
L
Linus Torvalds 已提交
4822 4823
	return err;
}
E
Eric Dumazet 已提交
4824
EXPORT_SYMBOL(dev_set_mac_address);
L
Linus Torvalds 已提交
4825 4826

/*
E
Eric Dumazet 已提交
4827
 *	Perform the SIOCxIFxxx calls, inside rcu_read_lock()
L
Linus Torvalds 已提交
4828
 */
4829
static int dev_ifsioc_locked(struct net *net, struct ifreq *ifr, unsigned int cmd)
L
Linus Torvalds 已提交
4830 4831
{
	int err;
E
Eric Dumazet 已提交
4832
	struct net_device *dev = dev_get_by_name_rcu(net, ifr->ifr_name);
L
Linus Torvalds 已提交
4833 4834 4835 4836 4837

	if (!dev)
		return -ENODEV;

	switch (cmd) {
E
Eric Dumazet 已提交
4838 4839 4840
	case SIOCGIFFLAGS:	/* Get interface flags */
		ifr->ifr_flags = (short) dev_get_flags(dev);
		return 0;
L
Linus Torvalds 已提交
4841

E
Eric Dumazet 已提交
4842 4843 4844 4845
	case SIOCGIFMETRIC:	/* Get the metric on the interface
				   (currently unused) */
		ifr->ifr_metric = 0;
		return 0;
L
Linus Torvalds 已提交
4846

E
Eric Dumazet 已提交
4847 4848 4849
	case SIOCGIFMTU:	/* Get the MTU of a device */
		ifr->ifr_mtu = dev->mtu;
		return 0;
L
Linus Torvalds 已提交
4850

E
Eric Dumazet 已提交
4851 4852 4853 4854 4855 4856 4857 4858
	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
4859

E
Eric Dumazet 已提交
4860 4861 4862
	case SIOCGIFSLAVE:
		err = -EINVAL;
		break;
4863

E
Eric Dumazet 已提交
4864 4865 4866 4867 4868 4869 4870 4871
	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;
4872

E
Eric Dumazet 已提交
4873 4874 4875
	case SIOCGIFINDEX:
		ifr->ifr_ifindex = dev->ifindex;
		return 0;
4876

E
Eric Dumazet 已提交
4877 4878 4879
	case SIOCGIFTXQLEN:
		ifr->ifr_qlen = dev->tx_queue_len;
		return 0;
4880

E
Eric Dumazet 已提交
4881 4882 4883 4884 4885
	default:
		/* dev_ioctl() should ensure this case
		 * is never reached
		 */
		WARN_ON(1);
4886
		err = -ENOTTY;
E
Eric Dumazet 已提交
4887
		break;
4888 4889 4890 4891 4892 4893 4894 4895 4896 4897 4898 4899

	}
	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
4900
	const struct net_device_ops *ops;
4901 4902 4903 4904

	if (!dev)
		return -ENODEV;

J
Jarek Poplawski 已提交
4905 4906
	ops = dev->netdev_ops;

4907
	switch (cmd) {
E
Eric Dumazet 已提交
4908 4909
	case SIOCSIFFLAGS:	/* Set interface flags */
		return dev_change_flags(dev, ifr->ifr_flags);
4910

E
Eric Dumazet 已提交
4911 4912 4913
	case SIOCSIFMETRIC:	/* Set the metric on the interface
				   (currently unused) */
		return -EOPNOTSUPP;
4914

E
Eric Dumazet 已提交
4915 4916
	case SIOCSIFMTU:	/* Set the MTU of a device */
		return dev_set_mtu(dev, ifr->ifr_mtu);
L
Linus Torvalds 已提交
4917

E
Eric Dumazet 已提交
4918 4919
	case SIOCSIFHWADDR:
		return dev_set_mac_address(dev, &ifr->ifr_hwaddr);
L
Linus Torvalds 已提交
4920

E
Eric Dumazet 已提交
4921 4922 4923 4924 4925 4926 4927
	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
4928

E
Eric Dumazet 已提交
4929 4930
	case SIOCSIFMAP:
		if (ops->ndo_set_config) {
L
Linus Torvalds 已提交
4931 4932
			if (!netif_device_present(dev))
				return -ENODEV;
E
Eric Dumazet 已提交
4933 4934 4935
			return ops->ndo_set_config(dev, &ifr->ifr_map);
		}
		return -EOPNOTSUPP;
L
Linus Torvalds 已提交
4936

E
Eric Dumazet 已提交
4937
	case SIOCADDMULTI:
4938
		if (!ops->ndo_set_rx_mode ||
E
Eric Dumazet 已提交
4939 4940 4941 4942
		    ifr->ifr_hwaddr.sa_family != AF_UNSPEC)
			return -EINVAL;
		if (!netif_device_present(dev))
			return -ENODEV;
4943
		return dev_mc_add_global(dev, ifr->ifr_hwaddr.sa_data);
E
Eric Dumazet 已提交
4944 4945

	case SIOCDELMULTI:
4946
		if (!ops->ndo_set_rx_mode ||
E
Eric Dumazet 已提交
4947 4948 4949 4950
		    ifr->ifr_hwaddr.sa_family != AF_UNSPEC)
			return -EINVAL;
		if (!netif_device_present(dev))
			return -ENODEV;
4951
		return dev_mc_del_global(dev, ifr->ifr_hwaddr.sa_data);
L
Linus Torvalds 已提交
4952

E
Eric Dumazet 已提交
4953 4954 4955 4956 4957
	case SIOCSIFTXQLEN:
		if (ifr->ifr_qlen < 0)
			return -EINVAL;
		dev->tx_queue_len = ifr->ifr_qlen;
		return 0;
L
Linus Torvalds 已提交
4958

E
Eric Dumazet 已提交
4959 4960 4961
	case SIOCSIFNAME:
		ifr->ifr_newname[IFNAMSIZ-1] = '\0';
		return dev_change_name(dev, ifr->ifr_newname);
L
Linus Torvalds 已提交
4962

4963 4964 4965 4966 4967 4968
	case SIOCSHWTSTAMP:
		err = net_hwtstamp_validate(ifr);
		if (err)
			return err;
		/* fall through */

E
Eric Dumazet 已提交
4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996
	/*
	 *	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
4997 4998 4999 5000 5001 5002 5003 5004 5005 5006 5007 5008

	}
	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
5009
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
5010 5011 5012 5013 5014 5015 5016 5017 5018
 *	@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.
 */

5019
int dev_ioctl(struct net *net, unsigned int cmd, void __user *arg)
L
Linus Torvalds 已提交
5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030
{
	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) {
5031
		rtnl_lock();
5032
		ret = dev_ifconf(net, (char __user *) arg);
5033
		rtnl_unlock();
L
Linus Torvalds 已提交
5034 5035 5036
		return ret;
	}
	if (cmd == SIOCGIFNAME)
5037
		return dev_ifname(net, (struct ifreq __user *)arg);
L
Linus Torvalds 已提交
5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050 5051 5052

	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
5053 5054 5055 5056 5057 5058 5059 5060 5061 5062 5063 5064 5065 5066 5067
	/*
	 *	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
5068
		rcu_read_lock();
E
Eric Dumazet 已提交
5069
		ret = dev_ifsioc_locked(net, &ifr, cmd);
E
Eric Dumazet 已提交
5070
		rcu_read_unlock();
E
Eric Dumazet 已提交
5071 5072 5073 5074 5075 5076 5077 5078
		if (!ret) {
			if (colon)
				*colon = ':';
			if (copy_to_user(arg, &ifr,
					 sizeof(struct ifreq)))
				ret = -EFAULT;
		}
		return ret;
L
Linus Torvalds 已提交
5079

E
Eric Dumazet 已提交
5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092
	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
5093

E
Eric Dumazet 已提交
5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116
	/*
	 *	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
5117

E
Eric Dumazet 已提交
5118 5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138 5139 5140 5141 5142 5143 5144 5145 5146 5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159
	/*
	 *	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:
5160
		return -ENOTTY;
E
Eric Dumazet 已提交
5161 5162 5163 5164 5165 5166 5167 5168

	/*
	 *	Unknown or private ioctl.
	 */
	default:
		if (cmd == SIOCWANDEV ||
		    (cmd >= SIOCDEVPRIVATE &&
		     cmd <= SIOCDEVPRIVATE + 15)) {
5169
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
5170
			rtnl_lock();
5171
			ret = dev_ifsioc(net, &ifr, cmd);
L
Linus Torvalds 已提交
5172
			rtnl_unlock();
E
Eric Dumazet 已提交
5173 5174 5175
			if (!ret && copy_to_user(arg, &ifr,
						 sizeof(struct ifreq)))
				ret = -EFAULT;
L
Linus Torvalds 已提交
5176
			return ret;
E
Eric Dumazet 已提交
5177 5178 5179 5180
		}
		/* Take care of Wireless Extensions */
		if (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST)
			return wext_handle_ioctl(net, &ifr, cmd, arg);
5181
		return -ENOTTY;
L
Linus Torvalds 已提交
5182 5183 5184 5185 5186 5187
	}
}


/**
 *	dev_new_index	-	allocate an ifindex
5188
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
5189 5190 5191 5192 5193
 *
 *	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.
 */
5194
static int dev_new_index(struct net *net)
L
Linus Torvalds 已提交
5195 5196 5197 5198 5199
{
	static int ifindex;
	for (;;) {
		if (++ifindex <= 0)
			ifindex = 1;
5200
		if (!__dev_get_by_index(net, ifindex))
L
Linus Torvalds 已提交
5201 5202 5203 5204 5205
			return ifindex;
	}
}

/* Delayed registration/unregisteration */
5206
static LIST_HEAD(net_todo_list);
L
Linus Torvalds 已提交
5207

5208
static void net_set_todo(struct net_device *dev)
L
Linus Torvalds 已提交
5209 5210 5211 5212
{
	list_add_tail(&dev->todo_list, &net_todo_list);
}

5213
static void rollback_registered_many(struct list_head *head)
5214
{
5215
	struct net_device *dev, *tmp;
5216

5217 5218 5219
	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

5220
	list_for_each_entry_safe(dev, tmp, head, unreg_list) {
5221
		/* Some devices call without registering
5222 5223
		 * for initialization unwind. Remove those
		 * devices and proceed with the remaining.
5224 5225
		 */
		if (dev->reg_state == NETREG_UNINITIALIZED) {
5226 5227
			pr_debug("unregister_netdevice: device %s/%p never was registered\n",
				 dev->name, dev);
5228

5229
			WARN_ON(1);
5230 5231
			list_del(&dev->unreg_list);
			continue;
5232
		}
5233
		dev->dismantle = true;
5234
		BUG_ON(dev->reg_state != NETREG_REGISTERED);
5235
	}
5236

5237 5238
	/* If device is running, close it first. */
	dev_close_many(head);
5239

5240
	list_for_each_entry(dev, head, unreg_list) {
5241 5242
		/* And unlink it from device chain. */
		unlist_netdevice(dev);
5243

5244 5245
		dev->reg_state = NETREG_UNREGISTERING;
	}
5246 5247 5248

	synchronize_net();

5249 5250 5251
	list_for_each_entry(dev, head, unreg_list) {
		/* Shutdown queueing discipline. */
		dev_shutdown(dev);
5252 5253


5254 5255 5256 5257
		/* Notify protocols, that we are about to destroy
		   this device. They should clean all the things.
		*/
		call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
5258

5259 5260 5261 5262
		if (!dev->rtnl_link_ops ||
		    dev->rtnl_link_state == RTNL_LINK_INITIALIZED)
			rtmsg_ifinfo(RTM_DELLINK, dev, ~0U);

5263 5264 5265
		/*
		 *	Flush the unicast and multicast chains
		 */
5266
		dev_uc_flush(dev);
5267
		dev_mc_flush(dev);
5268

5269 5270
		if (dev->netdev_ops->ndo_uninit)
			dev->netdev_ops->ndo_uninit(dev);
5271

5272 5273
		/* Notifier chain MUST detach us from master device. */
		WARN_ON(dev->master);
5274

5275 5276 5277
		/* Remove entries from kobject tree */
		netdev_unregister_kobject(dev);
	}
5278

5279
	/* Process any work delayed until the end of the batch */
5280
	dev = list_first_entry(head, struct net_device, unreg_list);
5281
	call_netdevice_notifiers(NETDEV_UNREGISTER_BATCH, dev);
5282

5283
	synchronize_net();
5284

5285
	list_for_each_entry(dev, head, unreg_list)
5286 5287 5288 5289 5290 5291 5292 5293 5294
		dev_put(dev);
}

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

	list_add(&dev->unreg_list, &single);
	rollback_registered_many(&single);
E
Eric Dumazet 已提交
5295
	list_del(&single);
5296 5297
}

5298 5299
static netdev_features_t netdev_fix_features(struct net_device *dev,
	netdev_features_t features)
5300
{
5301 5302 5303
	/* Fix illegal checksum combinations */
	if ((features & NETIF_F_HW_CSUM) &&
	    (features & (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))) {
5304
		netdev_warn(dev, "mixed HW and IP checksum settings.\n");
5305 5306 5307
		features &= ~(NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM);
	}

5308 5309 5310
	/* Fix illegal SG+CSUM combinations. */
	if ((features & NETIF_F_SG) &&
	    !(features & NETIF_F_ALL_CSUM)) {
5311 5312
		netdev_dbg(dev,
			"Dropping NETIF_F_SG since no checksum feature.\n");
5313 5314 5315 5316
		features &= ~NETIF_F_SG;
	}

	/* TSO requires that SG is present as well. */
5317
	if ((features & NETIF_F_ALL_TSO) && !(features & NETIF_F_SG)) {
5318
		netdev_dbg(dev, "Dropping TSO features since no SG feature.\n");
5319
		features &= ~NETIF_F_ALL_TSO;
5320 5321
	}

5322 5323 5324 5325
	/* TSO ECN requires that TSO is present as well. */
	if ((features & NETIF_F_ALL_TSO) == NETIF_F_TSO_ECN)
		features &= ~NETIF_F_TSO_ECN;

5326 5327
	/* Software GSO depends on SG. */
	if ((features & NETIF_F_GSO) && !(features & NETIF_F_SG)) {
5328
		netdev_dbg(dev, "Dropping NETIF_F_GSO since no SG feature.\n");
5329 5330 5331
		features &= ~NETIF_F_GSO;
	}

5332
	/* UFO needs SG and checksumming */
5333
	if (features & NETIF_F_UFO) {
5334 5335 5336 5337
		/* maybe split UFO into V4 and V6? */
		if (!((features & NETIF_F_GEN_CSUM) ||
		    (features & (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))
			    == (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))) {
5338
			netdev_dbg(dev,
5339
				"Dropping NETIF_F_UFO since no checksum offload features.\n");
5340 5341 5342 5343
			features &= ~NETIF_F_UFO;
		}

		if (!(features & NETIF_F_SG)) {
5344
			netdev_dbg(dev,
5345
				"Dropping NETIF_F_UFO since no NETIF_F_SG feature.\n");
5346 5347 5348 5349 5350 5351 5352
			features &= ~NETIF_F_UFO;
		}
	}

	return features;
}

5353
int __netdev_update_features(struct net_device *dev)
5354
{
5355
	netdev_features_t features;
5356 5357
	int err = 0;

5358 5359
	ASSERT_RTNL();

5360 5361 5362 5363 5364 5365 5366 5367 5368
	features = netdev_get_wanted_features(dev);

	if (dev->netdev_ops->ndo_fix_features)
		features = dev->netdev_ops->ndo_fix_features(dev, features);

	/* driver might be less strict about feature dependencies */
	features = netdev_fix_features(dev, features);

	if (dev->features == features)
5369
		return 0;
5370

5371 5372
	netdev_dbg(dev, "Features changed: %pNF -> %pNF\n",
		&dev->features, &features);
5373 5374 5375 5376

	if (dev->netdev_ops->ndo_set_features)
		err = dev->netdev_ops->ndo_set_features(dev, features);

5377
	if (unlikely(err < 0)) {
5378
		netdev_err(dev,
5379 5380
			"set_features() failed (%d); wanted %pNF, left %pNF\n",
			err, &features, &dev->features);
5381 5382 5383 5384 5385 5386 5387 5388 5389
		return -1;
	}

	if (!err)
		dev->features = features;

	return 1;
}

5390 5391 5392 5393 5394 5395 5396 5397
/**
 *	netdev_update_features - recalculate device features
 *	@dev: the device to check
 *
 *	Recalculate dev->features set and send notifications if it
 *	has changed. Should be called after driver or hardware dependent
 *	conditions might have changed that influence the features.
 */
5398 5399 5400 5401
void netdev_update_features(struct net_device *dev)
{
	if (__netdev_update_features(dev))
		netdev_features_change(dev);
5402 5403 5404
}
EXPORT_SYMBOL(netdev_update_features);

5405 5406 5407 5408 5409 5410 5411 5412 5413 5414 5415 5416 5417 5418 5419 5420 5421
/**
 *	netdev_change_features - recalculate device features
 *	@dev: the device to check
 *
 *	Recalculate dev->features set and send notifications even
 *	if they have not changed. Should be called instead of
 *	netdev_update_features() if also dev->vlan_features might
 *	have changed to allow the changes to be propagated to stacked
 *	VLAN devices.
 */
void netdev_change_features(struct net_device *dev)
{
	__netdev_update_features(dev);
	netdev_features_change(dev);
}
EXPORT_SYMBOL(netdev_change_features);

5422 5423 5424 5425 5426 5427 5428 5429 5430 5431 5432 5433 5434 5435 5436 5437 5438 5439 5440 5441 5442 5443 5444 5445 5446 5447 5448
/**
 *	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);

T
Tom Herbert 已提交
5449
#ifdef CONFIG_RPS
5450 5451 5452
static int netif_alloc_rx_queues(struct net_device *dev)
{
	unsigned int i, count = dev->num_rx_queues;
T
Tom Herbert 已提交
5453
	struct netdev_rx_queue *rx;
5454

T
Tom Herbert 已提交
5455
	BUG_ON(count < 1);
5456

T
Tom Herbert 已提交
5457 5458
	rx = kcalloc(count, sizeof(struct netdev_rx_queue), GFP_KERNEL);
	if (!rx) {
5459
		pr_err("netdev: Unable to allocate %u rx queues\n", count);
T
Tom Herbert 已提交
5460
		return -ENOMEM;
5461
	}
T
Tom Herbert 已提交
5462 5463 5464
	dev->_rx = rx;

	for (i = 0; i < count; i++)
T
Tom Herbert 已提交
5465
		rx[i].dev = dev;
5466 5467
	return 0;
}
T
Tom Herbert 已提交
5468
#endif
5469

C
Changli Gao 已提交
5470 5471 5472 5473 5474 5475 5476
static void netdev_init_one_queue(struct net_device *dev,
				  struct netdev_queue *queue, void *_unused)
{
	/* Initialize queue lock */
	spin_lock_init(&queue->_xmit_lock);
	netdev_set_xmit_lockdep_class(&queue->_xmit_lock, dev->type);
	queue->xmit_lock_owner = -1;
5477
	netdev_queue_numa_node_write(queue, NUMA_NO_NODE);
C
Changli Gao 已提交
5478
	queue->dev = dev;
T
Tom Herbert 已提交
5479 5480 5481
#ifdef CONFIG_BQL
	dql_init(&queue->dql, HZ);
#endif
C
Changli Gao 已提交
5482 5483
}

5484 5485 5486 5487 5488 5489 5490 5491 5492
static int netif_alloc_netdev_queues(struct net_device *dev)
{
	unsigned int count = dev->num_tx_queues;
	struct netdev_queue *tx;

	BUG_ON(count < 1);

	tx = kcalloc(count, sizeof(struct netdev_queue), GFP_KERNEL);
	if (!tx) {
5493
		pr_err("netdev: Unable to allocate %u tx queues\n", count);
5494 5495 5496
		return -ENOMEM;
	}
	dev->_tx = tx;
T
Tom Herbert 已提交
5497

5498 5499
	netdev_for_each_tx_queue(dev, netdev_init_one_queue, NULL);
	spin_lock_init(&dev->tx_global_lock);
C
Changli Gao 已提交
5500 5501

	return 0;
5502 5503
}

L
Linus Torvalds 已提交
5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 5519 5520 5521 5522 5523
/**
 *	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;
5524
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
5525 5526 5527 5528

	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

5529 5530
	might_sleep();

L
Linus Torvalds 已提交
5531 5532
	/* When net_device's are persistent, this will be fatal. */
	BUG_ON(dev->reg_state != NETREG_UNINITIALIZED);
5533
	BUG_ON(!net);
L
Linus Torvalds 已提交
5534

5535
	spin_lock_init(&dev->addr_list_lock);
5536
	netdev_set_addr_lockdep_class(dev);
L
Linus Torvalds 已提交
5537 5538 5539

	dev->iflink = -1;

5540 5541 5542 5543
	ret = dev_get_valid_name(dev, dev->name);
	if (ret < 0)
		goto out;

L
Linus Torvalds 已提交
5544
	/* Init, if this function is available */
5545 5546
	if (dev->netdev_ops->ndo_init) {
		ret = dev->netdev_ops->ndo_init(dev);
L
Linus Torvalds 已提交
5547 5548 5549
		if (ret) {
			if (ret > 0)
				ret = -EIO;
5550
			goto out;
L
Linus Torvalds 已提交
5551 5552
		}
	}
5553

5554
	dev->ifindex = dev_new_index(net);
L
Linus Torvalds 已提交
5555 5556 5557
	if (dev->iflink == -1)
		dev->iflink = dev->ifindex;

5558 5559 5560 5561
	/* Transfer changeable features to wanted_features and enable
	 * software offloads (GSO and GRO).
	 */
	dev->hw_features |= NETIF_F_SOFT_FEATURES;
5562 5563
	dev->features |= NETIF_F_SOFT_FEATURES;
	dev->wanted_features = dev->features & dev->hw_features;
L
Linus Torvalds 已提交
5564

5565
	/* Turn on no cache copy if HW is doing checksum */
5566 5567 5568 5569 5570 5571
	if (!(dev->flags & IFF_LOOPBACK)) {
		dev->hw_features |= NETIF_F_NOCACHE_COPY;
		if (dev->features & NETIF_F_ALL_CSUM) {
			dev->wanted_features |= NETIF_F_NOCACHE_COPY;
			dev->features |= NETIF_F_NOCACHE_COPY;
		}
5572 5573
	}

5574
	/* Make NETIF_F_HIGHDMA inheritable to VLAN devices.
5575
	 */
5576
	dev->vlan_features |= NETIF_F_HIGHDMA;
5577

5578 5579 5580 5581 5582
	ret = call_netdevice_notifiers(NETDEV_POST_INIT, dev);
	ret = notifier_to_errno(ret);
	if (ret)
		goto err_uninit;

5583
	ret = netdev_register_kobject(dev);
5584
	if (ret)
5585
		goto err_uninit;
5586 5587
	dev->reg_state = NETREG_REGISTERED;

5588
	__netdev_update_features(dev);
5589

L
Linus Torvalds 已提交
5590 5591 5592 5593 5594 5595 5596 5597 5598
	/*
	 *	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);
5599
	list_netdevice(dev);
L
Linus Torvalds 已提交
5600 5601

	/* Notify protocols, that a new device appeared. */
5602
	ret = call_netdevice_notifiers(NETDEV_REGISTER, dev);
5603
	ret = notifier_to_errno(ret);
5604 5605 5606 5607
	if (ret) {
		rollback_registered(dev);
		dev->reg_state = NETREG_UNREGISTERED;
	}
5608 5609 5610 5611
	/*
	 *	Prevent userspace races by waiting until the network
	 *	device is fully setup before sending notifications.
	 */
5612 5613 5614
	if (!dev->rtnl_link_ops ||
	    dev->rtnl_link_state == RTNL_LINK_INITIALIZED)
		rtmsg_ifinfo(RTM_NEWLINK, dev, ~0U);
L
Linus Torvalds 已提交
5615 5616 5617

out:
	return ret;
5618 5619

err_uninit:
5620 5621
	if (dev->netdev_ops->ndo_uninit)
		dev->netdev_ops->ndo_uninit(dev);
5622
	goto out;
L
Linus Torvalds 已提交
5623
}
E
Eric Dumazet 已提交
5624
EXPORT_SYMBOL(register_netdevice);
L
Linus Torvalds 已提交
5625

5626 5627 5628 5629 5630 5631 5632 5633 5634 5635 5636 5637 5638 5639 5640 5641 5642 5643 5644 5645 5646 5647 5648 5649 5650 5651 5652 5653 5654 5655 5656
/**
 *	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;

	/* 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);

E
Eric Dumazet 已提交
5657 5658 5659 5660 5661
	/* Note : We dont allocate pcpu_refcnt for dummy devices,
	 * because users of this 'device' dont need to change
	 * its refcount.
	 */

5662 5663 5664 5665 5666
	return 0;
}
EXPORT_SYMBOL_GPL(init_dummy_netdev);


L
Linus Torvalds 已提交
5667 5668 5669 5670 5671 5672 5673 5674 5675
/**
 *	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.
 *
5676
 *	This is a wrapper around register_netdevice that takes the rtnl semaphore
L
Linus Torvalds 已提交
5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687 5688 5689 5690
 *	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();
	err = register_netdevice(dev);
	rtnl_unlock();
	return err;
}
EXPORT_SYMBOL(register_netdev);

E
Eric Dumazet 已提交
5691 5692 5693 5694 5695 5696 5697 5698 5699 5700
int netdev_refcnt_read(const struct net_device *dev)
{
	int i, refcnt = 0;

	for_each_possible_cpu(i)
		refcnt += *per_cpu_ptr(dev->pcpu_refcnt, i);
	return refcnt;
}
EXPORT_SYMBOL(netdev_refcnt_read);

L
Linus Torvalds 已提交
5701 5702 5703 5704 5705 5706 5707 5708 5709
/*
 * 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
5710
 * call dev_put.
L
Linus Torvalds 已提交
5711 5712 5713 5714
 */
static void netdev_wait_allrefs(struct net_device *dev)
{
	unsigned long rebroadcast_time, warning_time;
E
Eric Dumazet 已提交
5715
	int refcnt;
L
Linus Torvalds 已提交
5716

5717 5718
	linkwatch_forget_dev(dev);

L
Linus Torvalds 已提交
5719
	rebroadcast_time = warning_time = jiffies;
E
Eric Dumazet 已提交
5720 5721 5722
	refcnt = netdev_refcnt_read(dev);

	while (refcnt != 0) {
L
Linus Torvalds 已提交
5723
		if (time_after(jiffies, rebroadcast_time + 1 * HZ)) {
5724
			rtnl_lock();
L
Linus Torvalds 已提交
5725 5726

			/* Rebroadcast unregister notification */
5727
			call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
5728
			/* don't resend NETDEV_UNREGISTER_BATCH, _BATCH users
5729
			 * should have already handle it the first time */
L
Linus Torvalds 已提交
5730 5731 5732 5733 5734 5735 5736 5737 5738 5739 5740 5741

			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();
			}

5742
			__rtnl_unlock();
L
Linus Torvalds 已提交
5743 5744 5745 5746 5747 5748

			rebroadcast_time = jiffies;
		}

		msleep(250);

E
Eric Dumazet 已提交
5749 5750
		refcnt = netdev_refcnt_read(dev);

L
Linus Torvalds 已提交
5751
		if (time_after(jiffies, warning_time + 10 * HZ)) {
5752 5753
			pr_emerg("unregister_netdevice: waiting for %s to become free. Usage count = %d\n",
				 dev->name, refcnt);
L
Linus Torvalds 已提交
5754 5755 5756 5757 5758 5759 5760 5761 5762 5763 5764 5765 5766 5767 5768 5769 5770 5771 5772
			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
5773
 * We are invoked by rtnl_unlock().
L
Linus Torvalds 已提交
5774
 * This allows us to deal with problems:
5775
 * 1) We can delete sysfs objects which invoke hotplug
L
Linus Torvalds 已提交
5776 5777 5778
 *    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
5779 5780 5781
 *
 * 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
5782 5783 5784
 */
void netdev_run_todo(void)
{
5785
	struct list_head list;
L
Linus Torvalds 已提交
5786 5787

	/* Snapshot list, allow later requests */
5788
	list_replace_init(&net_todo_list, &list);
H
Herbert Xu 已提交
5789 5790

	__rtnl_unlock();
5791

5792 5793 5794 5795 5796 5797
	/* Wait for rcu callbacks to finish before attempting to drain
	 * the device list.  This usually avoids a 250ms wait.
	 */
	if (!list_empty(&list))
		rcu_barrier();

L
Linus Torvalds 已提交
5798 5799
	while (!list_empty(&list)) {
		struct net_device *dev
5800
			= list_first_entry(&list, struct net_device, todo_list);
L
Linus Torvalds 已提交
5801 5802
		list_del(&dev->todo_list);

5803
		if (unlikely(dev->reg_state != NETREG_UNREGISTERING)) {
5804
			pr_err("network todo '%s' but state %d\n",
5805 5806 5807 5808
			       dev->name, dev->reg_state);
			dump_stack();
			continue;
		}
L
Linus Torvalds 已提交
5809

5810
		dev->reg_state = NETREG_UNREGISTERED;
L
Linus Torvalds 已提交
5811

5812
		on_each_cpu(flush_backlog, dev, 1);
5813

5814
		netdev_wait_allrefs(dev);
L
Linus Torvalds 已提交
5815

5816
		/* paranoia */
E
Eric Dumazet 已提交
5817
		BUG_ON(netdev_refcnt_read(dev));
5818 5819
		WARN_ON(rcu_access_pointer(dev->ip_ptr));
		WARN_ON(rcu_access_pointer(dev->ip6_ptr));
5820
		WARN_ON(dev->dn_ptr);
L
Linus Torvalds 已提交
5821

5822 5823
		if (dev->destructor)
			dev->destructor(dev);
5824 5825 5826

		/* Free network device */
		kobject_put(&dev->dev.kobj);
L
Linus Torvalds 已提交
5827 5828 5829
	}
}

5830 5831 5832
/* Convert net_device_stats to rtnl_link_stats64.  They have the same
 * fields in the same order, with only the type differing.
 */
5833 5834
void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64,
			     const struct net_device_stats *netdev_stats)
5835 5836
{
#if BITS_PER_LONG == 64
5837 5838
	BUILD_BUG_ON(sizeof(*stats64) != sizeof(*netdev_stats));
	memcpy(stats64, netdev_stats, sizeof(*stats64));
5839 5840 5841 5842 5843 5844 5845 5846 5847 5848 5849
#else
	size_t i, n = sizeof(*stats64) / sizeof(u64);
	const unsigned long *src = (const unsigned long *)netdev_stats;
	u64 *dst = (u64 *)stats64;

	BUILD_BUG_ON(sizeof(*netdev_stats) / sizeof(unsigned long) !=
		     sizeof(*stats64) / sizeof(u64));
	for (i = 0; i < n; i++)
		dst[i] = src[i];
#endif
}
5850
EXPORT_SYMBOL(netdev_stats_to_stats64);
5851

5852 5853 5854
/**
 *	dev_get_stats	- get network device statistics
 *	@dev: device to get statistics from
5855
 *	@storage: place to store stats
5856
 *
5857 5858 5859 5860
 *	Get network statistics from device. Return @storage.
 *	The device driver may provide its own method by setting
 *	dev->netdev_ops->get_stats64 or dev->netdev_ops->get_stats;
 *	otherwise the internal statistics structure is used.
5861
 */
5862 5863
struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev,
					struct rtnl_link_stats64 *storage)
5864
{
5865 5866
	const struct net_device_ops *ops = dev->netdev_ops;

5867 5868
	if (ops->ndo_get_stats64) {
		memset(storage, 0, sizeof(*storage));
5869 5870
		ops->ndo_get_stats64(dev, storage);
	} else if (ops->ndo_get_stats) {
5871
		netdev_stats_to_stats64(storage, ops->ndo_get_stats(dev));
5872 5873
	} else {
		netdev_stats_to_stats64(storage, &dev->stats);
5874
	}
5875
	storage->rx_dropped += atomic_long_read(&dev->rx_dropped);
5876
	return storage;
R
Rusty Russell 已提交
5877
}
5878
EXPORT_SYMBOL(dev_get_stats);
R
Rusty Russell 已提交
5879

5880
struct netdev_queue *dev_ingress_queue_create(struct net_device *dev)
5881
{
5882
	struct netdev_queue *queue = dev_ingress_queue(dev);
5883

5884 5885 5886 5887 5888 5889 5890 5891 5892 5893 5894 5895
#ifdef CONFIG_NET_CLS_ACT
	if (queue)
		return queue;
	queue = kzalloc(sizeof(*queue), GFP_KERNEL);
	if (!queue)
		return NULL;
	netdev_init_one_queue(dev, queue, NULL);
	queue->qdisc = &noop_qdisc;
	queue->qdisc_sleeping = &noop_qdisc;
	rcu_assign_pointer(dev->ingress_queue, queue);
#endif
	return queue;
5896 5897
}

L
Linus Torvalds 已提交
5898
/**
T
Tom Herbert 已提交
5899
 *	alloc_netdev_mqs - allocate network device
L
Linus Torvalds 已提交
5900 5901 5902
 *	@sizeof_priv:	size of private data to allocate space for
 *	@name:		device name format string
 *	@setup:		callback to initialize device
T
Tom Herbert 已提交
5903 5904
 *	@txqs:		the number of TX subqueues to allocate
 *	@rxqs:		the number of RX subqueues to allocate
L
Linus Torvalds 已提交
5905 5906
 *
 *	Allocates a struct net_device with private data area for driver use
5907
 *	and performs basic initialization.  Also allocates subquue structs
T
Tom Herbert 已提交
5908
 *	for each queue on the device.
L
Linus Torvalds 已提交
5909
 */
T
Tom Herbert 已提交
5910 5911 5912
struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
		void (*setup)(struct net_device *),
		unsigned int txqs, unsigned int rxqs)
L
Linus Torvalds 已提交
5913 5914
{
	struct net_device *dev;
5915
	size_t alloc_size;
5916
	struct net_device *p;
L
Linus Torvalds 已提交
5917

5918 5919
	BUG_ON(strlen(name) >= sizeof(dev->name));

T
Tom Herbert 已提交
5920
	if (txqs < 1) {
5921
		pr_err("alloc_netdev: Unable to allocate device with zero queues\n");
5922 5923 5924
		return NULL;
	}

T
Tom Herbert 已提交
5925 5926
#ifdef CONFIG_RPS
	if (rxqs < 1) {
5927
		pr_err("alloc_netdev: Unable to allocate device with zero RX queues\n");
T
Tom Herbert 已提交
5928 5929 5930 5931
		return NULL;
	}
#endif

5932
	alloc_size = sizeof(struct net_device);
5933 5934
	if (sizeof_priv) {
		/* ensure 32-byte alignment of private area */
5935
		alloc_size = ALIGN(alloc_size, NETDEV_ALIGN);
5936 5937 5938
		alloc_size += sizeof_priv;
	}
	/* ensure 32-byte alignment of whole construct */
5939
	alloc_size += NETDEV_ALIGN - 1;
L
Linus Torvalds 已提交
5940

5941
	p = kzalloc(alloc_size, GFP_KERNEL);
L
Linus Torvalds 已提交
5942
	if (!p) {
5943
		pr_err("alloc_netdev: Unable to allocate device\n");
L
Linus Torvalds 已提交
5944 5945 5946
		return NULL;
	}

5947
	dev = PTR_ALIGN(p, NETDEV_ALIGN);
L
Linus Torvalds 已提交
5948
	dev->padded = (char *)dev - (char *)p;
5949

E
Eric Dumazet 已提交
5950 5951
	dev->pcpu_refcnt = alloc_percpu(int);
	if (!dev->pcpu_refcnt)
5952
		goto free_p;
5953 5954

	if (dev_addr_init(dev))
E
Eric Dumazet 已提交
5955
		goto free_pcpu;
5956

5957
	dev_mc_init(dev);
5958
	dev_uc_init(dev);
J
Jiri Pirko 已提交
5959

5960
	dev_net_set(dev, &init_net);
L
Linus Torvalds 已提交
5961

5962 5963 5964 5965 5966 5967 5968 5969
	dev->gso_max_size = GSO_MAX_SIZE;

	INIT_LIST_HEAD(&dev->napi_list);
	INIT_LIST_HEAD(&dev->unreg_list);
	INIT_LIST_HEAD(&dev->link_watch_list);
	dev->priv_flags = IFF_XMIT_DST_RELEASE;
	setup(dev);

T
Tom Herbert 已提交
5970 5971
	dev->num_tx_queues = txqs;
	dev->real_num_tx_queues = txqs;
5972
	if (netif_alloc_netdev_queues(dev))
5973
		goto free_all;
5974

E
Eric Dumazet 已提交
5975
#ifdef CONFIG_RPS
T
Tom Herbert 已提交
5976 5977
	dev->num_rx_queues = rxqs;
	dev->real_num_rx_queues = rxqs;
T
Tom Herbert 已提交
5978
	if (netif_alloc_rx_queues(dev))
5979
		goto free_all;
E
Eric Dumazet 已提交
5980
#endif
T
Tom Herbert 已提交
5981

L
Linus Torvalds 已提交
5982
	strcpy(dev->name, name);
5983
	dev->group = INIT_NETDEV_GROUP;
L
Linus Torvalds 已提交
5984
	return dev;
5985

5986 5987 5988 5989
free_all:
	free_netdev(dev);
	return NULL;

E
Eric Dumazet 已提交
5990 5991
free_pcpu:
	free_percpu(dev->pcpu_refcnt);
5992
	kfree(dev->_tx);
T
Tom Herbert 已提交
5993 5994 5995 5996
#ifdef CONFIG_RPS
	kfree(dev->_rx);
#endif

5997 5998 5999
free_p:
	kfree(p);
	return NULL;
L
Linus Torvalds 已提交
6000
}
T
Tom Herbert 已提交
6001
EXPORT_SYMBOL(alloc_netdev_mqs);
L
Linus Torvalds 已提交
6002 6003 6004 6005 6006

/**
 *	free_netdev - free network device
 *	@dev: device
 *
6007 6008
 *	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
6009 6010 6011 6012
 *	If this is the last reference then it will be freed.
 */
void free_netdev(struct net_device *dev)
{
6013 6014
	struct napi_struct *p, *n;

6015 6016
	release_net(dev_net(dev));

6017
	kfree(dev->_tx);
T
Tom Herbert 已提交
6018 6019 6020
#ifdef CONFIG_RPS
	kfree(dev->_rx);
#endif
6021

6022
	kfree(rcu_dereference_protected(dev->ingress_queue, 1));
6023

6024 6025 6026
	/* Flush device addresses */
	dev_addr_flush(dev);

6027 6028 6029
	list_for_each_entry_safe(p, n, &dev->napi_list, dev_list)
		netif_napi_del(p);

E
Eric Dumazet 已提交
6030 6031 6032
	free_percpu(dev->pcpu_refcnt);
	dev->pcpu_refcnt = NULL;

S
Stephen Hemminger 已提交
6033
	/*  Compatibility with error handling in drivers */
L
Linus Torvalds 已提交
6034 6035 6036 6037 6038 6039 6040 6041
	if (dev->reg_state == NETREG_UNINITIALIZED) {
		kfree((char *)dev - dev->padded);
		return;
	}

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

6042 6043
	/* will free via device release */
	put_device(&dev->dev);
L
Linus Torvalds 已提交
6044
}
E
Eric Dumazet 已提交
6045
EXPORT_SYMBOL(free_netdev);
6046

6047 6048 6049 6050 6051 6052
/**
 *	synchronize_net -  Synchronize with packet receive processing
 *
 *	Wait for packets currently being received to be done.
 *	Does not block later packets from starting.
 */
6053
void synchronize_net(void)
L
Linus Torvalds 已提交
6054 6055
{
	might_sleep();
6056 6057 6058 6059
	if (rtnl_is_locked())
		synchronize_rcu_expedited();
	else
		synchronize_rcu();
L
Linus Torvalds 已提交
6060
}
E
Eric Dumazet 已提交
6061
EXPORT_SYMBOL(synchronize_net);
L
Linus Torvalds 已提交
6062 6063

/**
6064
 *	unregister_netdevice_queue - remove device from the kernel
L
Linus Torvalds 已提交
6065
 *	@dev: device
6066
 *	@head: list
6067
 *
L
Linus Torvalds 已提交
6068
 *	This function shuts down a device interface and removes it
6069
 *	from the kernel tables.
6070
 *	If head not NULL, device is queued to be unregistered later.
L
Linus Torvalds 已提交
6071 6072 6073 6074 6075
 *
 *	Callers must hold the rtnl semaphore.  You may want
 *	unregister_netdev() instead of this.
 */

6076
void unregister_netdevice_queue(struct net_device *dev, struct list_head *head)
L
Linus Torvalds 已提交
6077
{
6078 6079
	ASSERT_RTNL();

6080
	if (head) {
6081
		list_move_tail(&dev->unreg_list, head);
6082 6083 6084 6085 6086
	} else {
		rollback_registered(dev);
		/* Finish processing unregister after unlock */
		net_set_todo(dev);
	}
L
Linus Torvalds 已提交
6087
}
6088
EXPORT_SYMBOL(unregister_netdevice_queue);
L
Linus Torvalds 已提交
6089

6090 6091 6092 6093 6094 6095 6096 6097 6098 6099 6100 6101 6102 6103
/**
 *	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);
	}
}
6104
EXPORT_SYMBOL(unregister_netdevice_many);
6105

L
Linus Torvalds 已提交
6106 6107 6108 6109 6110
/**
 *	unregister_netdev - remove device from the kernel
 *	@dev: device
 *
 *	This function shuts down a device interface and removes it
6111
 *	from the kernel tables.
L
Linus Torvalds 已提交
6112 6113 6114 6115 6116 6117 6118 6119 6120 6121 6122 6123 6124
 *
 *	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);

6125 6126 6127 6128 6129 6130 6131 6132 6133 6134 6135 6136 6137 6138 6139 6140 6141 6142 6143 6144 6145 6146 6147 6148 6149 6150 6151 6152 6153 6154 6155 6156
/**
 *	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;

	/* 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;
6157
	if (net_eq(dev_net(dev), net))
6158 6159 6160 6161 6162 6163
		goto out;

	/* Pick the destination device name, and ensure
	 * we can use it in the destination network namespace.
	 */
	err = -EEXIST;
6164
	if (__dev_get_by_name(net, dev->name)) {
6165 6166 6167
		/* We get here if we can't use the current device name */
		if (!pat)
			goto out;
6168
		if (dev_get_valid_name(dev, pat) < 0)
6169 6170 6171 6172 6173 6174 6175 6176
			goto out;
	}

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

	/* If device is running close it first. */
6177
	dev_close(dev);
6178 6179 6180 6181 6182 6183 6184 6185 6186 6187 6188 6189

	/* 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.
6190 6191 6192 6193

	   Note that dev->reg_state stays at NETREG_REGISTERED.
	   This is wanted because this way 8021q and macvlan know
	   the device is just moving and can keep their slaves up.
6194 6195
	*/
	call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
6196
	call_netdevice_notifiers(NETDEV_UNREGISTER_BATCH, dev);
6197
	rtmsg_ifinfo(RTM_DELLINK, dev, ~0U);
6198 6199 6200 6201

	/*
	 *	Flush the unicast and multicast chains
	 */
6202
	dev_uc_flush(dev);
6203
	dev_mc_flush(dev);
6204 6205

	/* Actually switch the network namespace */
6206
	dev_net_set(dev, net);
6207 6208 6209 6210 6211 6212 6213 6214 6215

	/* 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;
	}

6216
	/* Fixup kobjects */
6217
	err = device_rename(&dev->dev, dev->name);
6218
	WARN_ON(err);
6219 6220 6221 6222 6223 6224 6225

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

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

6226 6227 6228 6229 6230 6231
	/*
	 *	Prevent userspace races by waiting until the network
	 *	device is fully setup before sending notifications.
	 */
	rtmsg_ifinfo(RTM_NEWLINK, dev, ~0U);

6232 6233 6234 6235 6236
	synchronize_net();
	err = 0;
out:
	return err;
}
6237
EXPORT_SYMBOL_GPL(dev_change_net_namespace);
6238

L
Linus Torvalds 已提交
6239 6240 6241 6242 6243 6244 6245 6246 6247
static int dev_cpu_callback(struct notifier_block *nfb,
			    unsigned long action,
			    void *ocpu)
{
	struct sk_buff **list_skb;
	struct sk_buff *skb;
	unsigned int cpu, oldcpu = (unsigned long)ocpu;
	struct softnet_data *sd, *oldsd;

6248
	if (action != CPU_DEAD && action != CPU_DEAD_FROZEN)
L
Linus Torvalds 已提交
6249 6250 6251 6252 6253 6254 6255 6256 6257 6258 6259 6260 6261 6262 6263 6264
		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;

	/* Append output queue from offline CPU. */
6265 6266 6267 6268 6269 6270
	if (oldsd->output_queue) {
		*sd->output_queue_tailp = oldsd->output_queue;
		sd->output_queue_tailp = oldsd->output_queue_tailp;
		oldsd->output_queue = NULL;
		oldsd->output_queue_tailp = &oldsd->output_queue;
	}
6271 6272 6273 6274 6275
	/* Append NAPI poll list from offline CPU. */
	if (!list_empty(&oldsd->poll_list)) {
		list_splice_init(&oldsd->poll_list, &sd->poll_list);
		raise_softirq_irqoff(NET_RX_SOFTIRQ);
	}
L
Linus Torvalds 已提交
6276 6277 6278 6279 6280

	raise_softirq_irqoff(NET_TX_SOFTIRQ);
	local_irq_enable();

	/* Process offline CPU's input_pkt_queue */
6281
	while ((skb = __skb_dequeue(&oldsd->process_queue))) {
L
Linus Torvalds 已提交
6282
		netif_rx(skb);
6283
		input_queue_head_incr(oldsd);
T
Tom Herbert 已提交
6284
	}
6285
	while ((skb = __skb_dequeue(&oldsd->input_pkt_queue))) {
6286
		netif_rx(skb);
6287 6288
		input_queue_head_incr(oldsd);
	}
L
Linus Torvalds 已提交
6289 6290 6291 6292 6293

	return NOTIFY_OK;
}


6294
/**
6295 6296 6297 6298
 *	netdev_increment_features - increment feature set by one
 *	@all: current feature set
 *	@one: new feature set
 *	@mask: mask feature set
6299 6300
 *
 *	Computes a new feature set after adding a device with feature set
6301 6302
 *	@one to the master device with current feature set @all.  Will not
 *	enable anything that is off in @mask. Returns the new feature set.
6303
 */
6304 6305
netdev_features_t netdev_increment_features(netdev_features_t all,
	netdev_features_t one, netdev_features_t mask)
6306
{
6307 6308 6309
	if (mask & NETIF_F_GEN_CSUM)
		mask |= NETIF_F_ALL_CSUM;
	mask |= NETIF_F_VLAN_CHALLENGED;
6310

6311 6312
	all |= one & (NETIF_F_ONE_FOR_ALL|NETIF_F_ALL_CSUM) & mask;
	all &= one | ~NETIF_F_ALL_FOR_ALL;
6313

6314 6315 6316
	/* If one device supports hw checksumming, set for all. */
	if (all & NETIF_F_GEN_CSUM)
		all &= ~(NETIF_F_ALL_CSUM & ~NETIF_F_GEN_CSUM);
6317 6318 6319

	return all;
}
6320
EXPORT_SYMBOL(netdev_increment_features);
6321

6322 6323 6324 6325 6326 6327 6328 6329 6330 6331 6332 6333 6334
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;
}

6335
/* Initialize per network namespace state */
6336
static int __net_init netdev_init(struct net *net)
6337 6338 6339
{
	INIT_LIST_HEAD(&net->dev_base_head);

6340 6341 6342
	net->dev_name_head = netdev_create_hash();
	if (net->dev_name_head == NULL)
		goto err_name;
6343

6344 6345 6346
	net->dev_index_head = netdev_create_hash();
	if (net->dev_index_head == NULL)
		goto err_idx;
6347 6348

	return 0;
6349 6350 6351 6352 6353

err_idx:
	kfree(net->dev_name_head);
err_name:
	return -ENOMEM;
6354 6355
}

6356 6357 6358 6359 6360 6361
/**
 *	netdev_drivername - network driver for the device
 *	@dev: network device
 *
 *	Determine network driver for device.
 */
6362
const char *netdev_drivername(const struct net_device *dev)
6363
{
6364 6365
	const struct device_driver *driver;
	const struct device *parent;
6366
	const char *empty = "";
6367 6368 6369

	parent = dev->dev.parent;
	if (!parent)
6370
		return empty;
6371 6372 6373

	driver = parent->driver;
	if (driver && driver->name)
6374 6375
		return driver->name;
	return empty;
6376 6377
}

6378
int __netdev_printk(const char *level, const struct net_device *dev,
6379 6380 6381 6382 6383 6384 6385 6386 6387 6388 6389 6390 6391 6392
			   struct va_format *vaf)
{
	int r;

	if (dev && dev->dev.parent)
		r = dev_printk(level, dev->dev.parent, "%s: %pV",
			       netdev_name(dev), vaf);
	else if (dev)
		r = printk("%s%s: %pV", level, netdev_name(dev), vaf);
	else
		r = printk("%s(NULL net_device): %pV", level, vaf);

	return r;
}
6393
EXPORT_SYMBOL(__netdev_printk);
6394 6395 6396 6397 6398 6399 6400 6401 6402 6403 6404 6405 6406 6407 6408 6409 6410 6411 6412 6413 6414 6415 6416 6417 6418 6419 6420 6421 6422 6423 6424 6425 6426 6427 6428 6429 6430 6431 6432 6433 6434 6435 6436 6437 6438 6439 6440

int netdev_printk(const char *level, const struct net_device *dev,
		  const char *format, ...)
{
	struct va_format vaf;
	va_list args;
	int r;

	va_start(args, format);

	vaf.fmt = format;
	vaf.va = &args;

	r = __netdev_printk(level, dev, &vaf);
	va_end(args);

	return r;
}
EXPORT_SYMBOL(netdev_printk);

#define define_netdev_printk_level(func, level)			\
int func(const struct net_device *dev, const char *fmt, ...)	\
{								\
	int r;							\
	struct va_format vaf;					\
	va_list args;						\
								\
	va_start(args, fmt);					\
								\
	vaf.fmt = fmt;						\
	vaf.va = &args;						\
								\
	r = __netdev_printk(level, dev, &vaf);			\
	va_end(args);						\
								\
	return r;						\
}								\
EXPORT_SYMBOL(func);

define_netdev_printk_level(netdev_emerg, KERN_EMERG);
define_netdev_printk_level(netdev_alert, KERN_ALERT);
define_netdev_printk_level(netdev_crit, KERN_CRIT);
define_netdev_printk_level(netdev_err, KERN_ERR);
define_netdev_printk_level(netdev_warn, KERN_WARNING);
define_netdev_printk_level(netdev_notice, KERN_NOTICE);
define_netdev_printk_level(netdev_info, KERN_INFO);

6441
static void __net_exit netdev_exit(struct net *net)
6442 6443 6444 6445 6446
{
	kfree(net->dev_name_head);
	kfree(net->dev_index_head);
}

6447
static struct pernet_operations __net_initdata netdev_net_ops = {
6448 6449 6450 6451
	.init = netdev_init,
	.exit = netdev_exit,
};

6452
static void __net_exit default_device_exit(struct net *net)
6453
{
6454
	struct net_device *dev, *aux;
6455
	/*
6456
	 * Push all migratable network devices back to the
6457 6458 6459
	 * initial network namespace
	 */
	rtnl_lock();
6460
	for_each_netdev_safe(net, dev, aux) {
6461
		int err;
6462
		char fb_name[IFNAMSIZ];
6463 6464 6465 6466 6467

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

6468 6469 6470
		/* Leave virtual devices for the generic cleanup */
		if (dev->rtnl_link_ops)
			continue;
6471

L
Lucas De Marchi 已提交
6472
		/* Push remaining network devices to init_net */
6473 6474
		snprintf(fb_name, IFNAMSIZ, "dev%d", dev->ifindex);
		err = dev_change_net_namespace(dev, &init_net, fb_name);
6475
		if (err) {
6476 6477
			pr_emerg("%s: failed to move %s to init_net: %d\n",
				 __func__, dev->name, err);
6478
			BUG();
6479 6480 6481 6482 6483
		}
	}
	rtnl_unlock();
}

6484 6485 6486
static void __net_exit default_device_exit_batch(struct list_head *net_list)
{
	/* At exit all network devices most be removed from a network
6487
	 * namespace.  Do this in the reverse order of registration.
6488 6489 6490 6491 6492 6493 6494 6495 6496 6497 6498 6499 6500 6501 6502 6503 6504
	 * 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);
E
Eric Dumazet 已提交
6505
	list_del(&dev_kill_list);
6506 6507 6508
	rtnl_unlock();
}

6509
static struct pernet_operations __net_initdata default_device_ops = {
6510
	.exit = default_device_exit,
6511
	.exit_batch = default_device_exit_batch,
6512 6513
};

L
Linus Torvalds 已提交
6514 6515 6516 6517 6518 6519 6520 6521 6522 6523 6524 6525 6526 6527 6528 6529 6530 6531 6532 6533
/*
 *	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;

6534
	if (netdev_kobject_init())
L
Linus Torvalds 已提交
6535 6536 6537
		goto out;

	INIT_LIST_HEAD(&ptype_all);
6538
	for (i = 0; i < PTYPE_HASH_SIZE; i++)
L
Linus Torvalds 已提交
6539 6540
		INIT_LIST_HEAD(&ptype_base[i]);

6541 6542
	if (register_pernet_subsys(&netdev_net_ops))
		goto out;
L
Linus Torvalds 已提交
6543 6544 6545 6546 6547

	/*
	 *	Initialise the packet receive queues.
	 */

6548
	for_each_possible_cpu(i) {
E
Eric Dumazet 已提交
6549
		struct softnet_data *sd = &per_cpu(softnet_data, i);
L
Linus Torvalds 已提交
6550

C
Changli Gao 已提交
6551
		memset(sd, 0, sizeof(*sd));
E
Eric Dumazet 已提交
6552
		skb_queue_head_init(&sd->input_pkt_queue);
6553
		skb_queue_head_init(&sd->process_queue);
E
Eric Dumazet 已提交
6554 6555
		sd->completion_queue = NULL;
		INIT_LIST_HEAD(&sd->poll_list);
6556 6557
		sd->output_queue = NULL;
		sd->output_queue_tailp = &sd->output_queue;
E
Eric Dumazet 已提交
6558
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
6559 6560 6561 6562
		sd->csd.func = rps_trigger_softirq;
		sd->csd.info = sd;
		sd->csd.flags = 0;
		sd->cpu = i;
6563
#endif
T
Tom Herbert 已提交
6564

E
Eric Dumazet 已提交
6565 6566 6567 6568
		sd->backlog.poll = process_backlog;
		sd->backlog.weight = weight_p;
		sd->backlog.gro_list = NULL;
		sd->backlog.gro_count = 0;
L
Linus Torvalds 已提交
6569 6570 6571 6572
	}

	dev_boot_phase = 0;

6573 6574 6575 6576 6577 6578 6579 6580 6581 6582 6583 6584 6585 6586 6587
	/* 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;

6588 6589
	open_softirq(NET_TX_SOFTIRQ, net_tx_action);
	open_softirq(NET_RX_SOFTIRQ, net_rx_action);
L
Linus Torvalds 已提交
6590 6591 6592 6593 6594 6595 6596 6597 6598 6599 6600

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

subsys_initcall(net_dev_init);

6601 6602
static int __init initialize_hashrnd(void)
{
T
Tom Herbert 已提交
6603
	get_random_bytes(&hashrnd, sizeof(hashrnd));
6604 6605 6606 6607 6608
	return 0;
}

late_initcall_sync(initialize_hashrnd);