dev.c 203.1 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
#include <net/sock.h>
E
Eric Dumazet 已提交
99
#include <net/busy_poll.h>
L
Linus Torvalds 已提交
100 101 102
#include <linux/rtnetlink.h>
#include <linux/stat.h>
#include <net/dst.h>
103
#include <net/dst_metadata.h>
L
Linus Torvalds 已提交
104 105
#include <net/pkt_sched.h>
#include <net/checksum.h>
106
#include <net/xfrm.h>
L
Linus Torvalds 已提交
107 108 109 110 111 112 113 114
#include <linux/highmem.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/netpoll.h>
#include <linux/rcupdate.h>
#include <linux/delay.h>
#include <net/iw_handler.h>
#include <asm/current.h>
S
Steve Grubb 已提交
115
#include <linux/audit.h>
116
#include <linux/dmaengine.h>
117
#include <linux/err.h>
118
#include <linux/ctype.h>
119
#include <linux/if_arp.h>
120
#include <linux/if_vlan.h>
121
#include <linux/ip.h>
122
#include <net/ip.h>
123
#include <net/mpls.h>
124 125
#include <linux/ipv6.h>
#include <linux/in.h>
D
David S. Miller 已提交
126 127
#include <linux/jhash.h>
#include <linux/random.h>
128
#include <trace/events/napi.h>
129
#include <trace/events/net.h>
130
#include <trace/events/skb.h>
131
#include <linux/pci.h>
132
#include <linux/inetdevice.h>
133
#include <linux/cpu_rmap.h>
134
#include <linux/static_key.h>
E
Eliezer Tamir 已提交
135
#include <linux/hashtable.h>
136
#include <linux/vmalloc.h>
137
#include <linux/if_macvlan.h>
138
#include <linux/errqueue.h>
139
#include <linux/hrtimer.h>
140
#include <linux/netfilter_ingress.h>
141
#include <linux/sctp.h>
L
Linus Torvalds 已提交
142

143 144
#include "net-sysfs.h"

145 146 147
/* Instead of increasing this, you should create a hash table. */
#define MAX_GRO_SKBS 8

H
Herbert Xu 已提交
148 149 150
/* 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
151
static DEFINE_SPINLOCK(ptype_lock);
152
static DEFINE_SPINLOCK(offload_lock);
153 154
struct list_head ptype_base[PTYPE_HASH_SIZE] __read_mostly;
struct list_head ptype_all __read_mostly;	/* Taps */
155
static struct list_head offload_base __read_mostly;
L
Linus Torvalds 已提交
156

157
static int netif_rx_internal(struct sk_buff *skb);
158 159 160
static int call_netdevice_notifiers_info(unsigned long val,
					 struct net_device *dev,
					 struct netdev_notifier_info *info);
161

L
Linus Torvalds 已提交
162
/*
163
 * The @dev_base_head list is protected by @dev_base_lock and the rtnl
L
Linus Torvalds 已提交
164 165
 * semaphore.
 *
166
 * Pure readers hold dev_base_lock for reading, or rcu_read_lock()
L
Linus Torvalds 已提交
167 168
 *
 * Writers must hold the rtnl semaphore while they loop through the
169
 * dev_base_head list, and hold dev_base_lock for writing when they do the
L
Linus Torvalds 已提交
170 171 172 173 174 175 176 177 178 179 180 181 182 183
 * 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);

E
Eliezer Tamir 已提交
184 185 186
/* protects napi_hash addition/deletion and napi_gen_id */
static DEFINE_SPINLOCK(napi_hash_lock);

187
static unsigned int napi_gen_id = NR_CPUS;
188
static DEFINE_READ_MOSTLY_HASHTABLE(napi_hash, 8);
E
Eliezer Tamir 已提交
189

190
static seqcount_t devnet_rename_seq;
191

192 193 194 195 196
static inline void dev_base_seq_inc(struct net *net)
{
	while (++net->dev_base_seq == 0);
}

197
static inline struct hlist_head *dev_name_hash(struct net *net, const char *name)
L
Linus Torvalds 已提交
198
{
199 200
	unsigned int hash = full_name_hash(name, strnlen(name, IFNAMSIZ));

201
	return &net->dev_name_head[hash_32(hash, NETDEV_HASHBITS)];
L
Linus Torvalds 已提交
202 203
}

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

E
Eric Dumazet 已提交
209
static inline void rps_lock(struct softnet_data *sd)
210 211
{
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
212
	spin_lock(&sd->input_pkt_queue.lock);
213 214 215
#endif
}

E
Eric Dumazet 已提交
216
static inline void rps_unlock(struct softnet_data *sd)
217 218
{
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
219
	spin_unlock(&sd->input_pkt_queue.lock);
220 221 222
#endif
}

223
/* Device list insertion */
224
static void list_netdevice(struct net_device *dev)
225
{
226
	struct net *net = dev_net(dev);
227 228 229 230

	ASSERT_RTNL();

	write_lock_bh(&dev_base_lock);
231
	list_add_tail_rcu(&dev->dev_list, &net->dev_base_head);
232
	hlist_add_head_rcu(&dev->name_hlist, dev_name_hash(net, dev->name));
233 234
	hlist_add_head_rcu(&dev->index_hlist,
			   dev_index_hash(net, dev->ifindex));
235
	write_unlock_bh(&dev_base_lock);
236 237

	dev_base_seq_inc(net);
238 239
}

240 241 242
/* Device list removal
 * caller must respect a RCU grace period before freeing/reusing dev
 */
243 244 245 246 247 248
static void unlist_netdevice(struct net_device *dev)
{
	ASSERT_RTNL();

	/* Unlink dev from the device chain */
	write_lock_bh(&dev_base_lock);
249
	list_del_rcu(&dev->dev_list);
250
	hlist_del_rcu(&dev->name_hlist);
251
	hlist_del_rcu(&dev->index_hlist);
252
	write_unlock_bh(&dev_base_lock);
253 254

	dev_base_seq_inc(dev_net(dev));
255 256
}

L
Linus Torvalds 已提交
257 258 259 260
/*
 *	Our notifier list
 */

261
static RAW_NOTIFIER_HEAD(netdev_chain);
L
Linus Torvalds 已提交
262 263 264 265 266

/*
 *	Device drivers call our routines to queue packets here. We empty the
 *	queue in the local softnet handler.
 */
267

268
DEFINE_PER_CPU_ALIGNED(struct softnet_data, softnet_data);
E
Eric Dumazet 已提交
269
EXPORT_PER_CPU_SYMBOL(softnet_data);
L
Linus Torvalds 已提交
270

271
#ifdef CONFIG_LOCKDEP
272
/*
273
 * register_netdevice() inits txq->_xmit_lock and sets lockdep class
274 275 276 277 278 279 280 281 282 283 284 285 286 287 288
 * 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,
289 290 291
	 ARPHRD_FCFABRIC, ARPHRD_IEEE80211, ARPHRD_IEEE80211_PRISM,
	 ARPHRD_IEEE80211_RADIOTAP, ARPHRD_PHONET, ARPHRD_PHONET_PIPE,
	 ARPHRD_IEEE802154, ARPHRD_VOID, ARPHRD_NONE};
292

293
static const char *const netdev_lock_name[] =
294 295 296 297 298 299 300 301 302 303 304 305
	{"_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",
306 307 308
	 "_xmit_FCFABRIC", "_xmit_IEEE80211", "_xmit_IEEE80211_PRISM",
	 "_xmit_IEEE80211_RADIOTAP", "_xmit_PHONET", "_xmit_PHONET_PIPE",
	 "_xmit_IEEE802154", "_xmit_VOID", "_xmit_NONE"};
309 310

static struct lock_class_key netdev_xmit_lock_key[ARRAY_SIZE(netdev_lock_type)];
311
static struct lock_class_key netdev_addr_lock_key[ARRAY_SIZE(netdev_lock_type)];
312 313 314 315 316 317 318 319 320 321 322 323

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;
}

324 325
static inline void netdev_set_xmit_lockdep_class(spinlock_t *lock,
						 unsigned short dev_type)
326 327 328 329 330 331 332
{
	int i;

	i = netdev_lock_pos(dev_type);
	lockdep_set_class_and_name(lock, &netdev_xmit_lock_key[i],
				   netdev_lock_name[i]);
}
333 334 335 336 337 338 339 340 341 342

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]);
}
343
#else
344 345 346 347 348
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)
349 350 351
{
}
#endif
L
Linus Torvalds 已提交
352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374

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

		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)
 */

375 376 377
static inline struct list_head *ptype_head(const struct packet_type *pt)
{
	if (pt->type == htons(ETH_P_ALL))
378
		return pt->dev ? &pt->dev->ptype_all : &ptype_all;
379
	else
380 381
		return pt->dev ? &pt->dev->ptype_specific :
				 &ptype_base[ntohs(pt->type) & PTYPE_HASH_MASK];
382 383
}

L
Linus Torvalds 已提交
384 385 386 387 388 389 390 391
/**
 *	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.
 *
392
 *	This call does not sleep therefore it can not
L
Linus Torvalds 已提交
393 394 395 396 397 398
 *	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)
{
399
	struct list_head *head = ptype_head(pt);
L
Linus Torvalds 已提交
400

401 402 403
	spin_lock(&ptype_lock);
	list_add_rcu(&pt->list, head);
	spin_unlock(&ptype_lock);
L
Linus Torvalds 已提交
404
}
E
Eric Dumazet 已提交
405
EXPORT_SYMBOL(dev_add_pack);
L
Linus Torvalds 已提交
406 407 408 409 410 411 412 413

/**
 *	__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
414
 *	returns.
L
Linus Torvalds 已提交
415 416 417 418 419 420 421
 *
 *      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)
{
422
	struct list_head *head = ptype_head(pt);
L
Linus Torvalds 已提交
423 424
	struct packet_type *pt1;

425
	spin_lock(&ptype_lock);
L
Linus Torvalds 已提交
426 427 428 429 430 431 432 433

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

434
	pr_warn("dev_remove_pack: %p not found\n", pt);
L
Linus Torvalds 已提交
435
out:
436
	spin_unlock(&ptype_lock);
L
Linus Torvalds 已提交
437
}
E
Eric Dumazet 已提交
438 439
EXPORT_SYMBOL(__dev_remove_pack);

L
Linus Torvalds 已提交
440 441 442 443 444 445 446 447 448 449 450 451 452 453 454
/**
 *	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);
455

L
Linus Torvalds 已提交
456 457
	synchronize_net();
}
E
Eric Dumazet 已提交
458
EXPORT_SYMBOL(dev_remove_pack);
L
Linus Torvalds 已提交
459

460 461 462 463 464 465 466 467 468 469 470 471 472 473 474

/**
 *	dev_add_offload - register offload handlers
 *	@po: protocol offload declaration
 *
 *	Add protocol offload handlers to the networking stack. The passed
 *	&proto_offload is linked into kernel lists and may not be freed until
 *	it has been removed from the kernel lists.
 *
 *	This call does not sleep therefore it can not
 *	guarantee all CPU's that are in middle of receiving packets
 *	will see the new offload handlers (until the next received packet).
 */
void dev_add_offload(struct packet_offload *po)
{
475
	struct packet_offload *elem;
476 477

	spin_lock(&offload_lock);
478 479 480 481 482
	list_for_each_entry(elem, &offload_base, list) {
		if (po->priority < elem->priority)
			break;
	}
	list_add_rcu(&po->list, elem->list.prev);
483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499
	spin_unlock(&offload_lock);
}
EXPORT_SYMBOL(dev_add_offload);

/**
 *	__dev_remove_offload	 - remove offload handler
 *	@po: packet offload declaration
 *
 *	Remove a protocol offload handler that was previously added to the
 *	kernel offload handlers by dev_add_offload(). The passed &offload_type
 *	is removed from the kernel lists and can be freed or reused once this
 *	function returns.
 *
 *      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.
 */
S
stephen hemminger 已提交
500
static void __dev_remove_offload(struct packet_offload *po)
501 502 503 504
{
	struct list_head *head = &offload_base;
	struct packet_offload *po1;

505
	spin_lock(&offload_lock);
506 507 508 509 510 511 512 513 514 515

	list_for_each_entry(po1, head, list) {
		if (po == po1) {
			list_del_rcu(&po->list);
			goto out;
		}
	}

	pr_warn("dev_remove_offload: %p not found\n", po);
out:
516
	spin_unlock(&offload_lock);
517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538
}

/**
 *	dev_remove_offload	 - remove packet offload handler
 *	@po: packet offload declaration
 *
 *	Remove a packet offload handler that was previously added to the kernel
 *	offload handlers by dev_add_offload(). The passed &offload_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_offload(struct packet_offload *po)
{
	__dev_remove_offload(po);

	synchronize_net();
}
EXPORT_SYMBOL(dev_remove_offload);

L
Linus Torvalds 已提交
539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565
/******************************************************************************

		      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));
566
			strlcpy(s[i].name, name, IFNAMSIZ);
L
Linus Torvalds 已提交
567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590
			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] != ' ' &&
591
		    !strcmp(dev->name, s[i].name)) {
L
Linus Torvalds 已提交
592 593 594 595 596 597 598 599 600
			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
601
EXPORT_SYMBOL(netdev_boot_setup_check);
L
Linus Torvalds 已提交
602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625


/**
 *	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
	 */
626
	if (__dev_get_by_name(&init_net, name))
L
Linus Torvalds 已提交
627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669
		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

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

670 671 672 673 674 675 676 677 678 679 680 681 682
/**
 *	dev_get_iflink	- get 'iflink' value of a interface
 *	@dev: targeted interface
 *
 *	Indicates the ifindex the interface is linked to.
 *	Physical interfaces have the same 'ifindex' and 'iflink' values.
 */

int dev_get_iflink(const struct net_device *dev)
{
	if (dev->netdev_ops && dev->netdev_ops->ndo_get_iflink)
		return dev->netdev_ops->ndo_get_iflink(dev);

683
	return dev->ifindex;
684 685 686
}
EXPORT_SYMBOL(dev_get_iflink);

687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712
/**
 *	dev_fill_metadata_dst - Retrieve tunnel egress information.
 *	@dev: targeted interface
 *	@skb: The packet.
 *
 *	For better visibility of tunnel traffic OVS needs to retrieve
 *	egress tunnel information for a packet. Following API allows
 *	user to get this info.
 */
int dev_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb)
{
	struct ip_tunnel_info *info;

	if (!dev->netdev_ops  || !dev->netdev_ops->ndo_fill_metadata_dst)
		return -EINVAL;

	info = skb_tunnel_info_unclone(skb);
	if (!info)
		return -ENOMEM;
	if (unlikely(!(info->mode & IP_TUNNEL_INFO_TX)))
		return -EINVAL;

	return dev->netdev_ops->ndo_fill_metadata_dst(dev, skb);
}
EXPORT_SYMBOL_GPL(dev_fill_metadata_dst);

L
Linus Torvalds 已提交
713 714
/**
 *	__dev_get_by_name	- find a device by its name
715
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
716 717 718 719 720 721 722 723 724
 *	@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.
 */

725
struct net_device *__dev_get_by_name(struct net *net, const char *name)
L
Linus Torvalds 已提交
726
{
E
Eric Dumazet 已提交
727 728
	struct net_device *dev;
	struct hlist_head *head = dev_name_hash(net, name);
L
Linus Torvalds 已提交
729

730
	hlist_for_each_entry(dev, head, name_hlist)
L
Linus Torvalds 已提交
731 732
		if (!strncmp(dev->name, name, IFNAMSIZ))
			return dev;
E
Eric Dumazet 已提交
733

L
Linus Torvalds 已提交
734 735
	return NULL;
}
E
Eric Dumazet 已提交
736
EXPORT_SYMBOL(__dev_get_by_name);
L
Linus Torvalds 已提交
737

738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754
/**
 *	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 net_device *dev;
	struct hlist_head *head = dev_name_hash(net, name);

755
	hlist_for_each_entry_rcu(dev, head, name_hlist)
756 757 758 759 760 761 762
		if (!strncmp(dev->name, name, IFNAMSIZ))
			return dev;

	return NULL;
}
EXPORT_SYMBOL(dev_get_by_name_rcu);

L
Linus Torvalds 已提交
763 764
/**
 *	dev_get_by_name		- find a device by its name
765
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
766 767 768 769 770 771 772 773 774
 *	@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.
 */

775
struct net_device *dev_get_by_name(struct net *net, const char *name)
L
Linus Torvalds 已提交
776 777 778
{
	struct net_device *dev;

779 780
	rcu_read_lock();
	dev = dev_get_by_name_rcu(net, name);
L
Linus Torvalds 已提交
781 782
	if (dev)
		dev_hold(dev);
783
	rcu_read_unlock();
L
Linus Torvalds 已提交
784 785
	return dev;
}
E
Eric Dumazet 已提交
786
EXPORT_SYMBOL(dev_get_by_name);
L
Linus Torvalds 已提交
787 788 789

/**
 *	__dev_get_by_index - find a device by its ifindex
790
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
791 792 793 794 795 796 797 798 799
 *	@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.
 */

800
struct net_device *__dev_get_by_index(struct net *net, int ifindex)
L
Linus Torvalds 已提交
801
{
E
Eric Dumazet 已提交
802 803
	struct net_device *dev;
	struct hlist_head *head = dev_index_hash(net, ifindex);
L
Linus Torvalds 已提交
804

805
	hlist_for_each_entry(dev, head, index_hlist)
L
Linus Torvalds 已提交
806 807
		if (dev->ifindex == ifindex)
			return dev;
E
Eric Dumazet 已提交
808

L
Linus Torvalds 已提交
809 810
	return NULL;
}
E
Eric Dumazet 已提交
811
EXPORT_SYMBOL(__dev_get_by_index);
L
Linus Torvalds 已提交
812

813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828
/**
 *	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 net_device *dev;
	struct hlist_head *head = dev_index_hash(net, ifindex);

829
	hlist_for_each_entry_rcu(dev, head, index_hlist)
830 831 832 833 834 835 836
		if (dev->ifindex == ifindex)
			return dev;

	return NULL;
}
EXPORT_SYMBOL(dev_get_by_index_rcu);

L
Linus Torvalds 已提交
837 838 839

/**
 *	dev_get_by_index - find a device by its ifindex
840
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
841 842 843 844 845 846 847 848
 *	@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.
 */

849
struct net_device *dev_get_by_index(struct net *net, int ifindex)
L
Linus Torvalds 已提交
850 851 852
{
	struct net_device *dev;

853 854
	rcu_read_lock();
	dev = dev_get_by_index_rcu(net, ifindex);
L
Linus Torvalds 已提交
855 856
	if (dev)
		dev_hold(dev);
857
	rcu_read_unlock();
L
Linus Torvalds 已提交
858 859
	return dev;
}
E
Eric Dumazet 已提交
860
EXPORT_SYMBOL(dev_get_by_index);
L
Linus Torvalds 已提交
861

862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895
/**
 *	netdev_get_name - get a netdevice name, knowing its ifindex.
 *	@net: network namespace
 *	@name: a pointer to the buffer where the name will be stored.
 *	@ifindex: the ifindex of the interface to get the name from.
 *
 *	The use of raw_seqcount_begin() and cond_resched() before
 *	retrying is required as we want to give the writers a chance
 *	to complete when CONFIG_PREEMPT is not set.
 */
int netdev_get_name(struct net *net, char *name, int ifindex)
{
	struct net_device *dev;
	unsigned int seq;

retry:
	seq = raw_seqcount_begin(&devnet_rename_seq);
	rcu_read_lock();
	dev = dev_get_by_index_rcu(net, ifindex);
	if (!dev) {
		rcu_read_unlock();
		return -ENODEV;
	}

	strcpy(name, dev->name);
	rcu_read_unlock();
	if (read_seqcount_retry(&devnet_rename_seq, seq)) {
		cond_resched();
		goto retry;
	}

	return 0;
}

L
Linus Torvalds 已提交
896
/**
897
 *	dev_getbyhwaddr_rcu - find a device by its hardware address
898
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
899 900 901 902
 *	@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
903 904
 *	is not found or a pointer to the device.
 *	The caller must hold RCU or RTNL.
905
 *	The returned device has not had its ref count increased
L
Linus Torvalds 已提交
906 907 908 909
 *	and the caller must therefore be careful about locking
 *
 */

910 911
struct net_device *dev_getbyhwaddr_rcu(struct net *net, unsigned short type,
				       const char *ha)
L
Linus Torvalds 已提交
912 913 914
{
	struct net_device *dev;

915
	for_each_netdev_rcu(net, dev)
L
Linus Torvalds 已提交
916 917
		if (dev->type == type &&
		    !memcmp(dev->dev_addr, ha, dev->addr_len))
918 919 920
			return dev;

	return NULL;
L
Linus Torvalds 已提交
921
}
922
EXPORT_SYMBOL(dev_getbyhwaddr_rcu);
923

924
struct net_device *__dev_getfirstbyhwtype(struct net *net, unsigned short type)
L
Linus Torvalds 已提交
925 926 927
{
	struct net_device *dev;

928
	ASSERT_RTNL();
929
	for_each_netdev(net, dev)
930
		if (dev->type == type)
931 932 933
			return dev;

	return NULL;
934 935 936
}
EXPORT_SYMBOL(__dev_getfirstbyhwtype);

937
struct net_device *dev_getfirstbyhwtype(struct net *net, unsigned short type)
938
{
939
	struct net_device *dev, *ret = NULL;
940

941 942 943 944 945 946 947 948 949
	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
950 951 952 953
}
EXPORT_SYMBOL(dev_getfirstbyhwtype);

/**
954
 *	__dev_get_by_flags - find any device with given flags
955
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
956 957 958 959
 *	@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
960
 *	is not found or a pointer to the device. Must be called inside
961
 *	rtnl_lock(), and result refcount is unchanged.
L
Linus Torvalds 已提交
962 963
 */

964 965
struct net_device *__dev_get_by_flags(struct net *net, unsigned short if_flags,
				      unsigned short mask)
L
Linus Torvalds 已提交
966
{
967
	struct net_device *dev, *ret;
L
Linus Torvalds 已提交
968

969 970
	ASSERT_RTNL();

971
	ret = NULL;
972
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
973
		if (((dev->flags ^ if_flags) & mask) == 0) {
974
			ret = dev;
L
Linus Torvalds 已提交
975 976 977
			break;
		}
	}
978
	return ret;
L
Linus Torvalds 已提交
979
}
980
EXPORT_SYMBOL(__dev_get_by_flags);
L
Linus Torvalds 已提交
981 982 983 984 985 986

/**
 *	dev_valid_name - check if name is okay for network device
 *	@name: name string
 *
 *	Network device names need to be valid file names to
987 988
 *	to allow sysfs to work.  We also disallow any kind of
 *	whitespace.
L
Linus Torvalds 已提交
989
 */
990
bool dev_valid_name(const char *name)
L
Linus Torvalds 已提交
991
{
992
	if (*name == '\0')
993
		return false;
994
	if (strlen(name) >= IFNAMSIZ)
995
		return false;
996
	if (!strcmp(name, ".") || !strcmp(name, ".."))
997
		return false;
998 999

	while (*name) {
1000
		if (*name == '/' || *name == ':' || isspace(*name))
1001
			return false;
1002 1003
		name++;
	}
1004
	return true;
L
Linus Torvalds 已提交
1005
}
E
Eric Dumazet 已提交
1006
EXPORT_SYMBOL(dev_valid_name);
L
Linus Torvalds 已提交
1007 1008

/**
1009 1010
 *	__dev_alloc_name - allocate a name for a device
 *	@net: network namespace to allocate the device name in
L
Linus Torvalds 已提交
1011
 *	@name: name format string
1012
 *	@buf:  scratch buffer and result name string
L
Linus Torvalds 已提交
1013 1014
 *
 *	Passed a format string - eg "lt%d" it will try and find a suitable
S
Stephen Hemminger 已提交
1015 1016 1017 1018 1019 1020
 *	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
1021 1022
 */

1023
static int __dev_alloc_name(struct net *net, const char *name, char *buf)
L
Linus Torvalds 已提交
1024 1025 1026 1027
{
	int i = 0;
	const char *p;
	const int max_netdevices = 8*PAGE_SIZE;
S
Stephen Hemminger 已提交
1028
	unsigned long *inuse;
L
Linus Torvalds 已提交
1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041
	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
1042
		inuse = (unsigned long *) get_zeroed_page(GFP_ATOMIC);
L
Linus Torvalds 已提交
1043 1044 1045
		if (!inuse)
			return -ENOMEM;

1046
		for_each_netdev(net, d) {
L
Linus Torvalds 已提交
1047 1048 1049 1050 1051 1052
			if (!sscanf(d->name, name, &i))
				continue;
			if (i < 0 || i >= max_netdevices)
				continue;

			/*  avoid cases where sscanf is not exact inverse of printf */
1053
			snprintf(buf, IFNAMSIZ, name, i);
L
Linus Torvalds 已提交
1054 1055 1056 1057 1058 1059 1060 1061
			if (!strncmp(buf, d->name, IFNAMSIZ))
				set_bit(i, inuse);
		}

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

1062 1063
	if (buf != name)
		snprintf(buf, IFNAMSIZ, name, i);
1064
	if (!__dev_get_by_name(net, buf))
L
Linus Torvalds 已提交
1065 1066 1067 1068 1069 1070 1071 1072 1073
		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;
}

1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093
/**
 *	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;

1094 1095
	BUG_ON(!dev_net(dev));
	net = dev_net(dev);
1096 1097 1098 1099 1100
	ret = __dev_alloc_name(net, name, buf);
	if (ret >= 0)
		strlcpy(dev->name, buf, IFNAMSIZ);
	return ret;
}
E
Eric Dumazet 已提交
1101
EXPORT_SYMBOL(dev_alloc_name);
1102

1103 1104 1105
static int dev_alloc_name_ns(struct net *net,
			     struct net_device *dev,
			     const char *name)
1106
{
1107 1108
	char buf[IFNAMSIZ];
	int ret;
1109

1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120
	ret = __dev_alloc_name(net, name, buf);
	if (ret >= 0)
		strlcpy(dev->name, buf, IFNAMSIZ);
	return ret;
}

static int dev_get_valid_name(struct net *net,
			      struct net_device *dev,
			      const char *name)
{
	BUG_ON(!net);
1121

1122 1123 1124
	if (!dev_valid_name(name))
		return -EINVAL;

1125
	if (strchr(name, '%'))
1126
		return dev_alloc_name_ns(net, dev, name);
1127 1128
	else if (__dev_get_by_name(net, name))
		return -EEXIST;
1129 1130
	else if (dev->name != name)
		strlcpy(dev->name, name, IFNAMSIZ);
1131 1132 1133

	return 0;
}
L
Linus Torvalds 已提交
1134 1135 1136 1137 1138 1139 1140 1141 1142

/**
 *	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.
 */
1143
int dev_change_name(struct net_device *dev, const char *newname)
L
Linus Torvalds 已提交
1144
{
1145
	unsigned char old_assign_type;
1146
	char oldname[IFNAMSIZ];
L
Linus Torvalds 已提交
1147
	int err = 0;
1148
	int ret;
1149
	struct net *net;
L
Linus Torvalds 已提交
1150 1151

	ASSERT_RTNL();
1152
	BUG_ON(!dev_net(dev));
L
Linus Torvalds 已提交
1153

1154
	net = dev_net(dev);
L
Linus Torvalds 已提交
1155 1156 1157
	if (dev->flags & IFF_UP)
		return -EBUSY;

1158
	write_seqcount_begin(&devnet_rename_seq);
1159 1160

	if (strncmp(newname, dev->name, IFNAMSIZ) == 0) {
1161
		write_seqcount_end(&devnet_rename_seq);
1162
		return 0;
1163
	}
1164

1165 1166
	memcpy(oldname, dev->name, IFNAMSIZ);

1167
	err = dev_get_valid_name(net, dev, newname);
1168
	if (err < 0) {
1169
		write_seqcount_end(&devnet_rename_seq);
1170
		return err;
1171
	}
L
Linus Torvalds 已提交
1172

1173 1174 1175
	if (oldname[0] && !strchr(oldname, '%'))
		netdev_info(dev, "renamed from %s\n", oldname);

1176 1177 1178
	old_assign_type = dev->name_assign_type;
	dev->name_assign_type = NET_NAME_RENAMED;

1179
rollback:
1180 1181 1182
	ret = device_rename(&dev->dev, dev->name);
	if (ret) {
		memcpy(dev->name, oldname, IFNAMSIZ);
1183
		dev->name_assign_type = old_assign_type;
1184
		write_seqcount_end(&devnet_rename_seq);
1185
		return ret;
1186
	}
1187

1188
	write_seqcount_end(&devnet_rename_seq);
1189

1190 1191
	netdev_adjacent_rename_links(dev, oldname);

1192
	write_lock_bh(&dev_base_lock);
1193
	hlist_del_rcu(&dev->name_hlist);
1194 1195 1196 1197 1198 1199
	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));
1200 1201
	write_unlock_bh(&dev_base_lock);

1202
	ret = call_netdevice_notifiers(NETDEV_CHANGENAME, dev);
1203 1204 1205
	ret = notifier_to_errno(ret);

	if (ret) {
1206 1207
		/* err >= 0 after dev_alloc_name() or stores the first errno */
		if (err >= 0) {
1208
			err = ret;
1209
			write_seqcount_begin(&devnet_rename_seq);
1210
			memcpy(dev->name, oldname, IFNAMSIZ);
1211
			memcpy(oldname, newname, IFNAMSIZ);
1212 1213
			dev->name_assign_type = old_assign_type;
			old_assign_type = NET_NAME_RENAMED;
1214
			goto rollback;
1215
		} else {
1216
			pr_err("%s: name change rollback failed: %d\n",
1217
			       dev->name, ret);
1218 1219
		}
	}
L
Linus Torvalds 已提交
1220 1221 1222 1223

	return err;
}

1224 1225 1226 1227
/**
 *	dev_set_alias - change ifalias of a device
 *	@dev: device
 *	@alias: name up to IFALIASZ
1228
 *	@len: limit of bytes to copy from info
1229 1230 1231 1232 1233
 *
 *	Set ifalias for a device,
 */
int dev_set_alias(struct net_device *dev, const char *alias, size_t len)
{
1234 1235
	char *new_ifalias;

1236 1237 1238 1239 1240
	ASSERT_RTNL();

	if (len >= IFALIASZ)
		return -EINVAL;

1241
	if (!len) {
1242 1243
		kfree(dev->ifalias);
		dev->ifalias = NULL;
1244 1245 1246
		return 0;
	}

1247 1248
	new_ifalias = krealloc(dev->ifalias, len + 1, GFP_KERNEL);
	if (!new_ifalias)
1249
		return -ENOMEM;
1250
	dev->ifalias = new_ifalias;
1251 1252 1253 1254 1255 1256

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


1257
/**
S
Stephen Hemminger 已提交
1258
 *	netdev_features_change - device changes features
1259 1260 1261 1262 1263 1264
 *	@dev: device to cause notification
 *
 *	Called to indicate a device has changed features.
 */
void netdev_features_change(struct net_device *dev)
{
1265
	call_netdevice_notifiers(NETDEV_FEAT_CHANGE, dev);
1266 1267 1268
}
EXPORT_SYMBOL(netdev_features_change);

L
Linus Torvalds 已提交
1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279
/**
 *	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) {
1280 1281 1282 1283 1284
		struct netdev_notifier_change_info change_info;

		change_info.flags_changed = 0;
		call_netdevice_notifiers_info(NETDEV_CHANGE, dev,
					      &change_info.info);
1285
		rtmsg_ifinfo(RTM_NEWLINK, dev, 0, GFP_KERNEL);
L
Linus Torvalds 已提交
1286 1287
	}
}
E
Eric Dumazet 已提交
1288
EXPORT_SYMBOL(netdev_state_change);
L
Linus Torvalds 已提交
1289

1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300
/**
 * 	netdev_notify_peers - notify network peers about existence of @dev
 * 	@dev: network device
 *
 * Generate traffic such that interested network peers are aware of
 * @dev, such as by generating a gratuitous ARP. This may be used when
 * a device wants to inform the rest of the network about some sort of
 * reconfiguration such as a failover event or virtual machine
 * migration.
 */
void netdev_notify_peers(struct net_device *dev)
1301
{
1302 1303 1304
	rtnl_lock();
	call_netdevice_notifiers(NETDEV_NOTIFY_PEERS, dev);
	rtnl_unlock();
1305
}
1306
EXPORT_SYMBOL(netdev_notify_peers);
1307

1308
static int __dev_open(struct net_device *dev)
L
Linus Torvalds 已提交
1309
{
1310
	const struct net_device_ops *ops = dev->netdev_ops;
1311
	int ret;
L
Linus Torvalds 已提交
1312

1313 1314
	ASSERT_RTNL();

L
Linus Torvalds 已提交
1315 1316 1317
	if (!netif_device_present(dev))
		return -ENODEV;

1318 1319 1320 1321
	/* Block netpoll from trying to do any rx path servicing.
	 * If we don't do this there is a chance ndo_poll_controller
	 * or ndo_poll may be running while we open the device
	 */
1322
	netpoll_poll_disable(dev);
1323

1324 1325 1326 1327 1328
	ret = call_netdevice_notifiers(NETDEV_PRE_UP, dev);
	ret = notifier_to_errno(ret);
	if (ret)
		return ret;

L
Linus Torvalds 已提交
1329
	set_bit(__LINK_STATE_START, &dev->state);
1330

1331 1332
	if (ops->ndo_validate_addr)
		ret = ops->ndo_validate_addr(dev);
1333

1334 1335
	if (!ret && ops->ndo_open)
		ret = ops->ndo_open(dev);
L
Linus Torvalds 已提交
1336

1337
	netpoll_poll_enable(dev);
1338

1339 1340 1341
	if (ret)
		clear_bit(__LINK_STATE_START, &dev->state);
	else {
L
Linus Torvalds 已提交
1342
		dev->flags |= IFF_UP;
1343
		dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
1344
		dev_activate(dev);
1345
		add_device_randomness(dev->dev_addr, dev->addr_len);
L
Linus Torvalds 已提交
1346
	}
1347

L
Linus Torvalds 已提交
1348 1349 1350 1351
	return ret;
}

/**
1352 1353
 *	dev_open	- prepare an interface for use.
 *	@dev:	device to open
L
Linus Torvalds 已提交
1354
 *
1355 1356 1357 1358 1359 1360 1361
 *	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
1362
 */
1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373
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;

1374
	rtmsg_ifinfo(RTM_NEWLINK, dev, IFF_UP|IFF_RUNNING, GFP_KERNEL);
1375 1376 1377 1378 1379 1380
	call_netdevice_notifiers(NETDEV_UP, dev);

	return ret;
}
EXPORT_SYMBOL(dev_open);

1381
static int __dev_close_many(struct list_head *head)
L
Linus Torvalds 已提交
1382
{
1383
	struct net_device *dev;
1384

1385
	ASSERT_RTNL();
1386 1387
	might_sleep();

1388
	list_for_each_entry(dev, head, close_list) {
1389
		/* Temporarily disable netpoll until the interface is down */
1390
		netpoll_poll_disable(dev);
1391

1392
		call_netdevice_notifiers(NETDEV_GOING_DOWN, dev);
L
Linus Torvalds 已提交
1393

1394
		clear_bit(__LINK_STATE_START, &dev->state);
L
Linus Torvalds 已提交
1395

1396 1397 1398 1399 1400 1401
		/* 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.
		 */
1402
		smp_mb__after_atomic(); /* Commit netif_running(). */
1403
	}
L
Linus Torvalds 已提交
1404

1405
	dev_deactivate_many(head);
1406

1407
	list_for_each_entry(dev, head, close_list) {
1408
		const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
1409

1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420
		/*
		 *	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;
1421
		netpoll_poll_enable(dev);
1422 1423 1424 1425 1426 1427 1428
	}

	return 0;
}

static int __dev_close(struct net_device *dev)
{
1429
	int retval;
1430 1431
	LIST_HEAD(single);

1432
	list_add(&dev->close_list, &single);
1433 1434
	retval = __dev_close_many(&single);
	list_del(&single);
1435

1436
	return retval;
1437 1438
}

1439
int dev_close_many(struct list_head *head, bool unlink)
1440 1441
{
	struct net_device *dev, *tmp;
L
Linus Torvalds 已提交
1442

1443 1444
	/* Remove the devices that don't need to be closed */
	list_for_each_entry_safe(dev, tmp, head, close_list)
1445
		if (!(dev->flags & IFF_UP))
1446
			list_del_init(&dev->close_list);
1447 1448

	__dev_close_many(head);
L
Linus Torvalds 已提交
1449

1450
	list_for_each_entry_safe(dev, tmp, head, close_list) {
1451
		rtmsg_ifinfo(RTM_NEWLINK, dev, IFF_UP|IFF_RUNNING, GFP_KERNEL);
1452
		call_netdevice_notifiers(NETDEV_DOWN, dev);
1453 1454
		if (unlink)
			list_del_init(&dev->close_list);
1455
	}
1456 1457 1458

	return 0;
}
1459
EXPORT_SYMBOL(dev_close_many);
1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471

/**
 *	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)
{
1472 1473
	if (dev->flags & IFF_UP) {
		LIST_HEAD(single);
L
Linus Torvalds 已提交
1474

1475
		list_add(&dev->close_list, &single);
1476
		dev_close_many(&single, true);
1477 1478
		list_del(&single);
	}
1479
	return 0;
L
Linus Torvalds 已提交
1480
}
E
Eric Dumazet 已提交
1481
EXPORT_SYMBOL(dev_close);
L
Linus Torvalds 已提交
1482 1483


1484 1485 1486 1487 1488 1489 1490 1491 1492 1493
/**
 *	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)
{
1494 1495
	struct net_device *lower_dev;
	struct list_head *iter;
1496

M
Michał Mirosław 已提交
1497 1498
	dev->wanted_features &= ~NETIF_F_LRO;
	netdev_update_features(dev);
1499

1500 1501
	if (unlikely(dev->features & NETIF_F_LRO))
		netdev_WARN(dev, "failed to disable LRO!\n");
1502 1503 1504

	netdev_for_each_lower_dev(dev, lower_dev, iter)
		dev_disable_lro(lower_dev);
1505 1506 1507
}
EXPORT_SYMBOL(dev_disable_lro);

1508 1509 1510 1511 1512 1513 1514 1515
static int call_netdevice_notifier(struct notifier_block *nb, unsigned long val,
				   struct net_device *dev)
{
	struct netdev_notifier_info info;

	netdev_notifier_info_init(&info, dev);
	return nb->notifier_call(nb, val, &info);
}
1516

1517 1518
static int dev_boot_phase = 1;

L
Linus Torvalds 已提交
1519 1520 1521 1522 1523 1524 1525 1526 1527 1528
/**
 *	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
1529
 *	to the new notifier to allow device to have a race free
L
Linus Torvalds 已提交
1530 1531 1532 1533 1534 1535
 *	view of the network device list.
 */

int register_netdevice_notifier(struct notifier_block *nb)
{
	struct net_device *dev;
1536
	struct net_device *last;
1537
	struct net *net;
L
Linus Torvalds 已提交
1538 1539 1540
	int err;

	rtnl_lock();
1541
	err = raw_notifier_chain_register(&netdev_chain, nb);
1542 1543
	if (err)
		goto unlock;
1544 1545 1546 1547
	if (dev_boot_phase)
		goto unlock;
	for_each_net(net) {
		for_each_netdev(net, dev) {
1548
			err = call_netdevice_notifier(nb, NETDEV_REGISTER, dev);
1549 1550 1551 1552 1553 1554
			err = notifier_to_errno(err);
			if (err)
				goto rollback;

			if (!(dev->flags & IFF_UP))
				continue;
L
Linus Torvalds 已提交
1555

1556
			call_netdevice_notifier(nb, NETDEV_UP, dev);
1557
		}
L
Linus Torvalds 已提交
1558
	}
1559 1560

unlock:
L
Linus Torvalds 已提交
1561 1562
	rtnl_unlock();
	return err;
1563 1564 1565

rollback:
	last = dev;
1566 1567 1568
	for_each_net(net) {
		for_each_netdev(net, dev) {
			if (dev == last)
1569
				goto outroll;
1570

1571
			if (dev->flags & IFF_UP) {
1572 1573 1574
				call_netdevice_notifier(nb, NETDEV_GOING_DOWN,
							dev);
				call_netdevice_notifier(nb, NETDEV_DOWN, dev);
1575
			}
1576
			call_netdevice_notifier(nb, NETDEV_UNREGISTER, dev);
1577 1578
		}
	}
1579

1580
outroll:
1581
	raw_notifier_chain_unregister(&netdev_chain, nb);
1582
	goto unlock;
L
Linus Torvalds 已提交
1583
}
E
Eric Dumazet 已提交
1584
EXPORT_SYMBOL(register_netdevice_notifier);
L
Linus Torvalds 已提交
1585 1586 1587 1588 1589 1590 1591 1592 1593

/**
 *	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.
1594 1595 1596 1597
 *
 * 	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
1598 1599 1600 1601
 */

int unregister_netdevice_notifier(struct notifier_block *nb)
{
1602 1603
	struct net_device *dev;
	struct net *net;
1604 1605 1606
	int err;

	rtnl_lock();
1607
	err = raw_notifier_chain_unregister(&netdev_chain, nb);
1608 1609 1610 1611 1612 1613
	if (err)
		goto unlock;

	for_each_net(net) {
		for_each_netdev(net, dev) {
			if (dev->flags & IFF_UP) {
1614 1615 1616
				call_netdevice_notifier(nb, NETDEV_GOING_DOWN,
							dev);
				call_netdevice_notifier(nb, NETDEV_DOWN, dev);
1617
			}
1618
			call_netdevice_notifier(nb, NETDEV_UNREGISTER, dev);
1619 1620 1621
		}
	}
unlock:
1622 1623
	rtnl_unlock();
	return err;
L
Linus Torvalds 已提交
1624
}
E
Eric Dumazet 已提交
1625
EXPORT_SYMBOL(unregister_netdevice_notifier);
L
Linus Torvalds 已提交
1626

1627 1628 1629 1630 1631 1632 1633 1634 1635 1636
/**
 *	call_netdevice_notifiers_info - call all network notifier blocks
 *	@val: value passed unmodified to notifier function
 *	@dev: net_device pointer passed unmodified to notifier function
 *	@info: notifier information data
 *
 *	Call all network notifier blocks.  Parameters and return value
 *	are as for raw_notifier_call_chain().
 */

S
stephen hemminger 已提交
1637 1638 1639
static int call_netdevice_notifiers_info(unsigned long val,
					 struct net_device *dev,
					 struct netdev_notifier_info *info)
1640 1641 1642 1643 1644 1645
{
	ASSERT_RTNL();
	netdev_notifier_info_init(info, dev);
	return raw_notifier_call_chain(&netdev_chain, val, info);
}

L
Linus Torvalds 已提交
1646 1647 1648
/**
 *	call_netdevice_notifiers - call all network notifier blocks
 *      @val: value passed unmodified to notifier function
1649
 *      @dev: net_device pointer passed unmodified to notifier function
L
Linus Torvalds 已提交
1650 1651
 *
 *	Call all network notifier blocks.  Parameters and return value
1652
 *	are as for raw_notifier_call_chain().
L
Linus Torvalds 已提交
1653 1654
 */

1655
int call_netdevice_notifiers(unsigned long val, struct net_device *dev)
L
Linus Torvalds 已提交
1656
{
1657 1658 1659
	struct netdev_notifier_info info;

	return call_netdevice_notifiers_info(val, dev, &info);
L
Linus Torvalds 已提交
1660
}
1661
EXPORT_SYMBOL(call_netdevice_notifiers);
L
Linus Torvalds 已提交
1662

1663
#ifdef CONFIG_NET_INGRESS
1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678
static struct static_key ingress_needed __read_mostly;

void net_inc_ingress_queue(void)
{
	static_key_slow_inc(&ingress_needed);
}
EXPORT_SYMBOL_GPL(net_inc_ingress_queue);

void net_dec_ingress_queue(void)
{
	static_key_slow_dec(&ingress_needed);
}
EXPORT_SYMBOL_GPL(net_dec_ingress_queue);
#endif

D
Daniel Borkmann 已提交
1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694
#ifdef CONFIG_NET_EGRESS
static struct static_key egress_needed __read_mostly;

void net_inc_egress_queue(void)
{
	static_key_slow_inc(&egress_needed);
}
EXPORT_SYMBOL_GPL(net_inc_egress_queue);

void net_dec_egress_queue(void)
{
	static_key_slow_dec(&egress_needed);
}
EXPORT_SYMBOL_GPL(net_dec_egress_queue);
#endif

1695
static struct static_key netstamp_needed __read_mostly;
1696
#ifdef HAVE_JUMP_LABEL
1697
/* We are not allowed to call static_key_slow_dec() from irq context
1698
 * If net_disable_timestamp() is called from irq context, defer the
1699
 * static_key_slow_dec() calls.
1700 1701 1702
 */
static atomic_t netstamp_needed_deferred;
#endif
L
Linus Torvalds 已提交
1703 1704 1705

void net_enable_timestamp(void)
{
1706 1707 1708 1709 1710
#ifdef HAVE_JUMP_LABEL
	int deferred = atomic_xchg(&netstamp_needed_deferred, 0);

	if (deferred) {
		while (--deferred)
1711
			static_key_slow_dec(&netstamp_needed);
1712 1713 1714
		return;
	}
#endif
1715
	static_key_slow_inc(&netstamp_needed);
L
Linus Torvalds 已提交
1716
}
E
Eric Dumazet 已提交
1717
EXPORT_SYMBOL(net_enable_timestamp);
L
Linus Torvalds 已提交
1718 1719 1720

void net_disable_timestamp(void)
{
1721 1722 1723 1724 1725 1726
#ifdef HAVE_JUMP_LABEL
	if (in_interrupt()) {
		atomic_inc(&netstamp_needed_deferred);
		return;
	}
#endif
1727
	static_key_slow_dec(&netstamp_needed);
L
Linus Torvalds 已提交
1728
}
E
Eric Dumazet 已提交
1729
EXPORT_SYMBOL(net_disable_timestamp);
L
Linus Torvalds 已提交
1730

E
Eric Dumazet 已提交
1731
static inline void net_timestamp_set(struct sk_buff *skb)
L
Linus Torvalds 已提交
1732
{
1733
	skb->tstamp.tv64 = 0;
1734
	if (static_key_false(&netstamp_needed))
1735
		__net_timestamp(skb);
L
Linus Torvalds 已提交
1736 1737
}

1738
#define net_timestamp_check(COND, SKB)			\
1739
	if (static_key_false(&netstamp_needed)) {		\
1740 1741 1742
		if ((COND) && !(SKB)->tstamp.tv64)	\
			__net_timestamp(SKB);		\
	}						\
E
Eric Dumazet 已提交
1743

1744
bool is_skb_forwardable(struct net_device *dev, struct sk_buff *skb)
1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762
{
	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;
}
1763
EXPORT_SYMBOL_GPL(is_skb_forwardable);
1764

H
Herbert Xu 已提交
1765 1766
int __dev_forward_skb(struct net_device *dev, struct sk_buff *skb)
{
1767 1768
	if (skb_orphan_frags(skb, GFP_ATOMIC) ||
	    unlikely(!is_skb_forwardable(dev, skb))) {
H
Herbert Xu 已提交
1769 1770 1771 1772 1773 1774
		atomic_long_inc(&dev->rx_dropped);
		kfree_skb(skb);
		return NET_RX_DROP;
	}

	skb_scrub_packet(skb, true);
1775
	skb->priority = 0;
H
Herbert Xu 已提交
1776
	skb->protocol = eth_type_trans(skb, dev);
1777
	skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN);
H
Herbert Xu 已提交
1778 1779 1780 1781 1782

	return 0;
}
EXPORT_SYMBOL_GPL(__dev_forward_skb);

1783 1784 1785 1786 1787 1788 1789 1790
/**
 * dev_forward_skb - loopback an skb to another netif
 *
 * @dev: destination network device
 * @skb: buffer to forward
 *
 * return values:
 *	NET_RX_SUCCESS	(no congestion)
1791
 *	NET_RX_DROP     (packet was dropped, but freed)
1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802
 *
 * 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)
{
H
Herbert Xu 已提交
1803
	return __dev_forward_skb(dev, skb) ?: netif_rx_internal(skb);
1804 1805 1806
}
EXPORT_SYMBOL_GPL(dev_forward_skb);

1807 1808 1809 1810
static inline int deliver_skb(struct sk_buff *skb,
			      struct packet_type *pt_prev,
			      struct net_device *orig_dev)
{
1811 1812
	if (unlikely(skb_orphan_frags(skb, GFP_ATOMIC)))
		return -ENOMEM;
1813 1814 1815 1816
	atomic_inc(&skb->users);
	return pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
}

1817 1818
static inline void deliver_ptype_list_skb(struct sk_buff *skb,
					  struct packet_type **pt,
1819 1820
					  struct net_device *orig_dev,
					  __be16 type,
1821 1822 1823 1824 1825 1826 1827 1828
					  struct list_head *ptype_list)
{
	struct packet_type *ptype, *pt_prev = *pt;

	list_for_each_entry_rcu(ptype, ptype_list, list) {
		if (ptype->type != type)
			continue;
		if (pt_prev)
1829
			deliver_skb(skb, pt_prev, orig_dev);
1830 1831 1832 1833 1834
		pt_prev = ptype;
	}
	*pt = pt_prev;
}

1835 1836
static inline bool skb_loop_sk(struct packet_type *ptype, struct sk_buff *skb)
{
1837
	if (!ptype->af_packet_priv || !skb->sk)
1838 1839 1840 1841 1842 1843 1844 1845 1846 1847
		return false;

	if (ptype->id_match)
		return ptype->id_match(ptype, skb->sk);
	else if ((struct sock *)ptype->af_packet_priv == skb->sk)
		return true;

	return false;
}

L
Linus Torvalds 已提交
1848 1849 1850 1851 1852
/*
 *	Support routine. Sends outgoing frames to any network
 *	taps currently in use.
 */

1853
static void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev)
L
Linus Torvalds 已提交
1854 1855
{
	struct packet_type *ptype;
1856 1857
	struct sk_buff *skb2 = NULL;
	struct packet_type *pt_prev = NULL;
1858
	struct list_head *ptype_list = &ptype_all;
1859

L
Linus Torvalds 已提交
1860
	rcu_read_lock();
1861 1862
again:
	list_for_each_entry_rcu(ptype, ptype_list, list) {
L
Linus Torvalds 已提交
1863 1864 1865
		/* Never send packets back to the socket
		 * they originated from - MvS (miquels@drinkel.ow.org)
		 */
1866 1867
		if (skb_loop_sk(ptype, skb))
			continue;
1868

1869 1870 1871 1872 1873
		if (pt_prev) {
			deliver_skb(skb2, pt_prev, skb->dev);
			pt_prev = ptype;
			continue;
		}
L
Linus Torvalds 已提交
1874

1875 1876 1877 1878
		/* need to clone skb, done only once */
		skb2 = skb_clone(skb, GFP_ATOMIC);
		if (!skb2)
			goto out_unlock;
1879

1880
		net_timestamp_set(skb2);
L
Linus Torvalds 已提交
1881

1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893
		/* skb->nh should be correctly
		 * set by sender, so that the second statement is
		 * just protection against buggy protocols.
		 */
		skb_reset_mac_header(skb2);

		if (skb_network_header(skb2) < skb2->data ||
		    skb_network_header(skb2) > skb_tail_pointer(skb2)) {
			net_crit_ratelimited("protocol %04x is buggy, dev %s\n",
					     ntohs(skb2->protocol),
					     dev->name);
			skb_reset_network_header(skb2);
L
Linus Torvalds 已提交
1894
		}
1895 1896 1897 1898 1899 1900 1901 1902 1903

		skb2->transport_header = skb2->network_header;
		skb2->pkt_type = PACKET_OUTGOING;
		pt_prev = ptype;
	}

	if (ptype_list == &ptype_all) {
		ptype_list = &dev->ptype_all;
		goto again;
L
Linus Torvalds 已提交
1904
	}
1905
out_unlock:
1906 1907
	if (pt_prev)
		pt_prev->func(skb2, skb->dev, pt_prev, skb->dev);
L
Linus Torvalds 已提交
1908 1909 1910
	rcu_read_unlock();
}

1911 1912
/**
 * netif_setup_tc - Handle tc mappings on real_num_tx_queues change
1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923
 * @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
1924
static void netif_setup_tc(struct net_device *dev, unsigned int txq)
1925 1926 1927 1928 1929 1930
{
	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) {
1931
		pr_warn("Number of in use tx queues changed invalidating tc mappings. Priority traffic classification disabled!\n");
1932 1933 1934 1935 1936 1937 1938 1939 1940 1941
		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) {
1942 1943
			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);
1944 1945 1946 1947 1948
			netdev_set_prio_tc_map(dev, i, 0);
		}
	}
}

1949 1950 1951 1952 1953
#ifdef CONFIG_XPS
static DEFINE_MUTEX(xps_map_mutex);
#define xmap_dereference(P)		\
	rcu_dereference_protected((P), lockdep_is_held(&xps_map_mutex))

1954 1955
static struct xps_map *remove_xps_queue(struct xps_dev_maps *dev_maps,
					int cpu, u16 index)
1956
{
1957 1958
	struct xps_map *map = NULL;
	int pos;
1959

1960 1961
	if (dev_maps)
		map = xmap_dereference(dev_maps->cpu_map[cpu]);
1962

1963 1964
	for (pos = 0; map && pos < map->len; pos++) {
		if (map->queues[pos] == index) {
1965 1966 1967
			if (map->len > 1) {
				map->queues[pos] = map->queues[--map->len];
			} else {
1968
				RCU_INIT_POINTER(dev_maps->cpu_map[cpu], NULL);
1969 1970 1971
				kfree_rcu(map, rcu);
				map = NULL;
			}
1972
			break;
1973 1974 1975
		}
	}

1976 1977 1978
	return map;
}

1979
static void netif_reset_xps_queues_gt(struct net_device *dev, u16 index)
1980 1981
{
	struct xps_dev_maps *dev_maps;
1982
	int cpu, i;
1983 1984 1985 1986 1987 1988 1989 1990 1991
	bool active = false;

	mutex_lock(&xps_map_mutex);
	dev_maps = xmap_dereference(dev->xps_maps);

	if (!dev_maps)
		goto out_no_maps;

	for_each_possible_cpu(cpu) {
1992 1993 1994 1995 1996
		for (i = index; i < dev->num_tx_queues; i++) {
			if (!remove_xps_queue(dev_maps, cpu, i))
				break;
		}
		if (i == dev->num_tx_queues)
1997 1998 1999 2000
			active = true;
	}

	if (!active) {
2001 2002 2003 2004
		RCU_INIT_POINTER(dev->xps_maps, NULL);
		kfree_rcu(dev_maps, rcu);
	}

2005 2006 2007 2008
	for (i = index; i < dev->num_tx_queues; i++)
		netdev_queue_numa_node_write(netdev_get_tx_queue(dev, i),
					     NUMA_NO_NODE);

2009 2010 2011 2012
out_no_maps:
	mutex_unlock(&xps_map_mutex);
}

2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047
static struct xps_map *expand_xps_map(struct xps_map *map,
				      int cpu, u16 index)
{
	struct xps_map *new_map;
	int alloc_len = XPS_MIN_MAP_ALLOC;
	int i, pos;

	for (pos = 0; map && pos < map->len; pos++) {
		if (map->queues[pos] != index)
			continue;
		return map;
	}

	/* Need to add queue to this CPU's existing map */
	if (map) {
		if (pos < map->alloc_len)
			return map;

		alloc_len = map->alloc_len * 2;
	}

	/* Need to allocate new map to store queue on this CPU's map */
	new_map = kzalloc_node(XPS_MAP_SIZE(alloc_len), GFP_KERNEL,
			       cpu_to_node(cpu));
	if (!new_map)
		return NULL;

	for (i = 0; i < pos; i++)
		new_map->queues[i] = map->queues[i];
	new_map->alloc_len = alloc_len;
	new_map->len = pos;

	return new_map;
}

2048 2049
int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
			u16 index)
2050
{
2051
	struct xps_dev_maps *dev_maps, *new_dev_maps = NULL;
2052 2053
	struct xps_map *map, *new_map;
	int maps_sz = max_t(unsigned int, XPS_DEV_MAPS_SIZE, L1_CACHE_BYTES);
2054 2055
	int cpu, numa_node_id = -2;
	bool active = false;
2056 2057 2058 2059 2060

	mutex_lock(&xps_map_mutex);

	dev_maps = xmap_dereference(dev->xps_maps);

2061 2062 2063 2064 2065 2066 2067
	/* allocate memory for queue storage */
	for_each_online_cpu(cpu) {
		if (!cpumask_test_cpu(cpu, mask))
			continue;

		if (!new_dev_maps)
			new_dev_maps = kzalloc(maps_sz, GFP_KERNEL);
2068 2069
		if (!new_dev_maps) {
			mutex_unlock(&xps_map_mutex);
2070
			return -ENOMEM;
2071
		}
2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085

		map = dev_maps ? xmap_dereference(dev_maps->cpu_map[cpu]) :
				 NULL;

		map = expand_xps_map(map, cpu, index);
		if (!map)
			goto error;

		RCU_INIT_POINTER(new_dev_maps->cpu_map[cpu], map);
	}

	if (!new_dev_maps)
		goto out_no_new_maps;

2086
	for_each_possible_cpu(cpu) {
2087 2088 2089 2090 2091 2092 2093 2094 2095 2096
		if (cpumask_test_cpu(cpu, mask) && cpu_online(cpu)) {
			/* add queue to CPU maps */
			int pos = 0;

			map = xmap_dereference(new_dev_maps->cpu_map[cpu]);
			while ((pos < map->len) && (map->queues[pos] != index))
				pos++;

			if (pos == map->len)
				map->queues[map->len++] = index;
2097 2098 2099 2100 2101 2102
#ifdef CONFIG_NUMA
			if (numa_node_id == -2)
				numa_node_id = cpu_to_node(cpu);
			else if (numa_node_id != cpu_to_node(cpu))
				numa_node_id = -1;
#endif
2103 2104 2105 2106
		} else if (dev_maps) {
			/* fill in the new device map from the old device map */
			map = xmap_dereference(dev_maps->cpu_map[cpu]);
			RCU_INIT_POINTER(new_dev_maps->cpu_map[cpu], map);
2107
		}
2108

2109 2110
	}

2111 2112
	rcu_assign_pointer(dev->xps_maps, new_dev_maps);

2113
	/* Cleanup old maps */
2114 2115 2116 2117 2118 2119 2120
	if (dev_maps) {
		for_each_possible_cpu(cpu) {
			new_map = xmap_dereference(new_dev_maps->cpu_map[cpu]);
			map = xmap_dereference(dev_maps->cpu_map[cpu]);
			if (map && map != new_map)
				kfree_rcu(map, rcu);
		}
2121

2122
		kfree_rcu(dev_maps, rcu);
2123 2124
	}

2125 2126
	dev_maps = new_dev_maps;
	active = true;
2127

2128 2129
out_no_new_maps:
	/* update Tx queue numa node */
2130 2131 2132 2133
	netdev_queue_numa_node_write(netdev_get_tx_queue(dev, index),
				     (numa_node_id >= 0) ? numa_node_id :
				     NUMA_NO_NODE);

2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152
	if (!dev_maps)
		goto out_no_maps;

	/* removes queue from unused CPUs */
	for_each_possible_cpu(cpu) {
		if (cpumask_test_cpu(cpu, mask) && cpu_online(cpu))
			continue;

		if (remove_xps_queue(dev_maps, cpu, index))
			active = true;
	}

	/* free map if not active */
	if (!active) {
		RCU_INIT_POINTER(dev->xps_maps, NULL);
		kfree_rcu(dev_maps, rcu);
	}

out_no_maps:
2153 2154 2155 2156
	mutex_unlock(&xps_map_mutex);

	return 0;
error:
2157 2158 2159 2160 2161 2162 2163 2164 2165
	/* remove any maps that we added */
	for_each_possible_cpu(cpu) {
		new_map = xmap_dereference(new_dev_maps->cpu_map[cpu]);
		map = dev_maps ? xmap_dereference(dev_maps->cpu_map[cpu]) :
				 NULL;
		if (new_map && new_map != map)
			kfree(new_map);
	}

2166 2167 2168 2169 2170 2171 2172 2173
	mutex_unlock(&xps_map_mutex);

	kfree(new_dev_maps);
	return -ENOMEM;
}
EXPORT_SYMBOL(netif_set_xps_queue);

#endif
2174 2175 2176 2177
/*
 * 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.
 */
2178
int netif_set_real_num_tx_queues(struct net_device *dev, unsigned int txq)
2179
{
T
Tom Herbert 已提交
2180 2181
	int rc;

2182 2183
	if (txq < 1 || txq > dev->num_tx_queues)
		return -EINVAL;
2184

2185 2186
	if (dev->reg_state == NETREG_REGISTERED ||
	    dev->reg_state == NETREG_UNREGISTERING) {
2187 2188
		ASSERT_RTNL();

T
Tom Herbert 已提交
2189 2190
		rc = netdev_queue_update_kobjects(dev, dev->real_num_tx_queues,
						  txq);
T
Tom Herbert 已提交
2191 2192 2193
		if (rc)
			return rc;

2194 2195 2196
		if (dev->num_tc)
			netif_setup_tc(dev, txq);

2197
		if (txq < dev->real_num_tx_queues) {
2198
			qdisc_reset_all_tx_gt(dev, txq);
2199 2200 2201 2202
#ifdef CONFIG_XPS
			netif_reset_xps_queues_gt(dev, txq);
#endif
		}
2203
	}
2204 2205 2206

	dev->real_num_tx_queues = txq;
	return 0;
2207 2208
}
EXPORT_SYMBOL(netif_set_real_num_tx_queues);
2209

2210
#ifdef CONFIG_SYSFS
2211 2212 2213 2214 2215 2216 2217
/**
 *	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
2218 2219
 *	negative error code.  If called before registration, it always
 *	succeeds.
2220 2221 2222 2223 2224
 */
int netif_set_real_num_rx_queues(struct net_device *dev, unsigned int rxq)
{
	int rc;

T
Tom Herbert 已提交
2225 2226 2227
	if (rxq < 1 || rxq > dev->num_rx_queues)
		return -EINVAL;

2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242
	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

2243 2244
/**
 * netif_get_num_default_rss_queues - default number of RSS queues
2245 2246 2247 2248
 *
 * This routine should set an upper limit on the number of RSS queues
 * used by default by multiqueue devices.
 */
2249
int netif_get_num_default_rss_queues(void)
2250 2251 2252 2253 2254
{
	return min_t(int, DEFAULT_MAX_NUM_RSS_QUEUES, num_online_cpus());
}
EXPORT_SYMBOL(netif_get_num_default_rss_queues);

2255
static inline void __netif_reschedule(struct Qdisc *q)
2256
{
2257 2258
	struct softnet_data *sd;
	unsigned long flags;
2259

2260
	local_irq_save(flags);
2261
	sd = this_cpu_ptr(&softnet_data);
2262 2263 2264
	q->next_sched = NULL;
	*sd->output_queue_tailp = q;
	sd->output_queue_tailp = &q->next_sched;
2265 2266 2267 2268 2269 2270 2271 2272
	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);
2273 2274 2275
}
EXPORT_SYMBOL(__netif_schedule);

2276 2277 2278 2279 2280
struct dev_kfree_skb_cb {
	enum skb_free_reason reason;
};

static struct dev_kfree_skb_cb *get_kfree_skb_cb(const struct sk_buff *skb)
2281
{
2282 2283 2284
	return (struct dev_kfree_skb_cb *)skb->cb;
}

2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331
void netif_schedule_queue(struct netdev_queue *txq)
{
	rcu_read_lock();
	if (!(txq->state & QUEUE_STATE_ANY_XOFF)) {
		struct Qdisc *q = rcu_dereference(txq->qdisc);

		__netif_schedule(q);
	}
	rcu_read_unlock();
}
EXPORT_SYMBOL(netif_schedule_queue);

/**
 *	netif_wake_subqueue - allow sending packets on subqueue
 *	@dev: network device
 *	@queue_index: sub queue index
 *
 * Resume individual transmit queue of a device with multiple transmit queues.
 */
void netif_wake_subqueue(struct net_device *dev, u16 queue_index)
{
	struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);

	if (test_and_clear_bit(__QUEUE_STATE_DRV_XOFF, &txq->state)) {
		struct Qdisc *q;

		rcu_read_lock();
		q = rcu_dereference(txq->qdisc);
		__netif_schedule(q);
		rcu_read_unlock();
	}
}
EXPORT_SYMBOL(netif_wake_subqueue);

void netif_tx_wake_queue(struct netdev_queue *dev_queue)
{
	if (test_and_clear_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state)) {
		struct Qdisc *q;

		rcu_read_lock();
		q = rcu_dereference(dev_queue->qdisc);
		__netif_schedule(q);
		rcu_read_unlock();
	}
}
EXPORT_SYMBOL(netif_tx_wake_queue);

2332
void __dev_kfree_skb_irq(struct sk_buff *skb, enum skb_free_reason reason)
2333
{
2334
	unsigned long flags;
2335

2336 2337 2338 2339 2340
	if (likely(atomic_read(&skb->users) == 1)) {
		smp_rmb();
		atomic_set(&skb->users, 0);
	} else if (likely(!atomic_dec_and_test(&skb->users))) {
		return;
2341
	}
2342 2343 2344 2345 2346 2347
	get_kfree_skb_cb(skb)->reason = reason;
	local_irq_save(flags);
	skb->next = __this_cpu_read(softnet_data.completion_queue);
	__this_cpu_write(softnet_data.completion_queue, skb);
	raise_softirq_irqoff(NET_TX_SOFTIRQ);
	local_irq_restore(flags);
2348
}
2349
EXPORT_SYMBOL(__dev_kfree_skb_irq);
2350

2351
void __dev_kfree_skb_any(struct sk_buff *skb, enum skb_free_reason reason)
2352 2353
{
	if (in_irq() || irqs_disabled())
2354
		__dev_kfree_skb_irq(skb, reason);
2355 2356 2357
	else
		dev_kfree_skb(skb);
}
2358
EXPORT_SYMBOL(__dev_kfree_skb_any);
2359 2360


2361 2362 2363 2364 2365 2366
/**
 * netif_device_detach - mark device as removed
 * @dev: network device
 *
 * Mark device as removed from system and therefore no longer available.
 */
2367 2368 2369 2370
void netif_device_detach(struct net_device *dev)
{
	if (test_and_clear_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
2371
		netif_tx_stop_all_queues(dev);
2372 2373 2374 2375
	}
}
EXPORT_SYMBOL(netif_device_detach);

2376 2377 2378 2379 2380 2381
/**
 * netif_device_attach - mark device as attached
 * @dev: network device
 *
 * Mark device as attached from system and restart if needed.
 */
2382 2383 2384 2385
void netif_device_attach(struct net_device *dev)
{
	if (!test_and_set_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
2386
		netif_tx_wake_all_queues(dev);
2387
		__netdev_watchdog_up(dev);
2388 2389 2390 2391
	}
}
EXPORT_SYMBOL(netif_device_attach);

J
Jiri Pirko 已提交
2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419
/*
 * 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, struct sk_buff *skb,
		  unsigned int num_tx_queues)
{
	u32 hash;
	u16 qoffset = 0;
	u16 qcount = num_tx_queues;

	if (skb_rx_queue_recorded(skb)) {
		hash = skb_get_rx_queue(skb);
		while (unlikely(hash >= num_tx_queues))
			hash -= num_tx_queues;
		return hash;
	}

	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;
	}

	return (u16) reciprocal_scale(skb_get_hash(skb), qcount) + qoffset;
}
EXPORT_SYMBOL(__skb_tx_hash);

2420 2421
static void skb_warn_bad_offload(const struct sk_buff *skb)
{
2422
	static const netdev_features_t null_features = 0;
2423
	struct net_device *dev = skb->dev;
2424
	const char *name = "";
2425

2426 2427 2428
	if (!net_ratelimit())
		return;

2429 2430 2431 2432 2433 2434
	if (dev) {
		if (dev->dev.parent)
			name = dev_driver_string(dev->dev.parent);
		else
			name = netdev_name(dev);
	}
2435 2436
	WARN(1, "%s: caps=(%pNF, %pNF) len=%d data_len=%d gso_size=%d "
	     "gso_type=%d ip_summed=%d\n",
2437
	     name, dev ? &dev->features : &null_features,
2438
	     skb->sk ? &skb->sk->sk_route_caps : &null_features,
2439 2440 2441 2442
	     skb->len, skb->data_len, skb_shinfo(skb)->gso_size,
	     skb_shinfo(skb)->gso_type, skb->ip_summed);
}

L
Linus Torvalds 已提交
2443 2444 2445 2446
/*
 * Invalidate hardware checksum when packet is to be mangled, and
 * complete checksum manually on outgoing path.
 */
2447
int skb_checksum_help(struct sk_buff *skb)
L
Linus Torvalds 已提交
2448
{
2449
	__wsum csum;
2450
	int ret = 0, offset;
L
Linus Torvalds 已提交
2451

2452
	if (skb->ip_summed == CHECKSUM_COMPLETE)
2453 2454 2455
		goto out_set_summed;

	if (unlikely(skb_shinfo(skb)->gso_size)) {
2456 2457
		skb_warn_bad_offload(skb);
		return -EINVAL;
L
Linus Torvalds 已提交
2458 2459
	}

2460 2461 2462 2463 2464 2465 2466 2467 2468
	/* Before computing a checksum, we should make sure no frag could
	 * be modified by an external entity : checksum could be wrong.
	 */
	if (skb_has_shared_frag(skb)) {
		ret = __skb_linearize(skb);
		if (ret)
			goto out;
	}

2469
	offset = skb_checksum_start_offset(skb);
2470 2471 2472 2473 2474 2475 2476 2477
	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
2478 2479 2480 2481 2482
		ret = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
		if (ret)
			goto out;
	}

2483
	*(__sum16 *)(skb->data + offset) = csum_fold(csum);
2484
out_set_summed:
L
Linus Torvalds 已提交
2485
	skb->ip_summed = CHECKSUM_NONE;
2486
out:
L
Linus Torvalds 已提交
2487 2488
	return ret;
}
E
Eric Dumazet 已提交
2489
EXPORT_SYMBOL(skb_checksum_help);
L
Linus Torvalds 已提交
2490

2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 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 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625
/* skb_csum_offload_check - Driver helper function to determine if a device
 * with limited checksum offload capabilities is able to offload the checksum
 * for a given packet.
 *
 * Arguments:
 *   skb - sk_buff for the packet in question
 *   spec - contains the description of what device can offload
 *   csum_encapped - returns true if the checksum being offloaded is
 *	      encpasulated. That is it is checksum for the transport header
 *	      in the inner headers.
 *   checksum_help - when set indicates that helper function should
 *	      call skb_checksum_help if offload checks fail
 *
 * Returns:
 *   true: Packet has passed the checksum checks and should be offloadable to
 *	   the device (a driver may still need to check for additional
 *	   restrictions of its device)
 *   false: Checksum is not offloadable. If checksum_help was set then
 *	   skb_checksum_help was called to resolve checksum for non-GSO
 *	   packets and when IP protocol is not SCTP
 */
bool __skb_csum_offload_chk(struct sk_buff *skb,
			    const struct skb_csum_offl_spec *spec,
			    bool *csum_encapped,
			    bool csum_help)
{
	struct iphdr *iph;
	struct ipv6hdr *ipv6;
	void *nhdr;
	int protocol;
	u8 ip_proto;

	if (skb->protocol == htons(ETH_P_8021Q) ||
	    skb->protocol == htons(ETH_P_8021AD)) {
		if (!spec->vlan_okay)
			goto need_help;
	}

	/* We check whether the checksum refers to a transport layer checksum in
	 * the outermost header or an encapsulated transport layer checksum that
	 * corresponds to the inner headers of the skb. If the checksum is for
	 * something else in the packet we need help.
	 */
	if (skb_checksum_start_offset(skb) == skb_transport_offset(skb)) {
		/* Non-encapsulated checksum */
		protocol = eproto_to_ipproto(vlan_get_protocol(skb));
		nhdr = skb_network_header(skb);
		*csum_encapped = false;
		if (spec->no_not_encapped)
			goto need_help;
	} else if (skb->encapsulation && spec->encap_okay &&
		   skb_checksum_start_offset(skb) ==
		   skb_inner_transport_offset(skb)) {
		/* Encapsulated checksum */
		*csum_encapped = true;
		switch (skb->inner_protocol_type) {
		case ENCAP_TYPE_ETHER:
			protocol = eproto_to_ipproto(skb->inner_protocol);
			break;
		case ENCAP_TYPE_IPPROTO:
			protocol = skb->inner_protocol;
			break;
		}
		nhdr = skb_inner_network_header(skb);
	} else {
		goto need_help;
	}

	switch (protocol) {
	case IPPROTO_IP:
		if (!spec->ipv4_okay)
			goto need_help;
		iph = nhdr;
		ip_proto = iph->protocol;
		if (iph->ihl != 5 && !spec->ip_options_okay)
			goto need_help;
		break;
	case IPPROTO_IPV6:
		if (!spec->ipv6_okay)
			goto need_help;
		if (spec->no_encapped_ipv6 && *csum_encapped)
			goto need_help;
		ipv6 = nhdr;
		nhdr += sizeof(*ipv6);
		ip_proto = ipv6->nexthdr;
		break;
	default:
		goto need_help;
	}

ip_proto_again:
	switch (ip_proto) {
	case IPPROTO_TCP:
		if (!spec->tcp_okay ||
		    skb->csum_offset != offsetof(struct tcphdr, check))
			goto need_help;
		break;
	case IPPROTO_UDP:
		if (!spec->udp_okay ||
		    skb->csum_offset != offsetof(struct udphdr, check))
			goto need_help;
		break;
	case IPPROTO_SCTP:
		if (!spec->sctp_okay ||
		    skb->csum_offset != offsetof(struct sctphdr, checksum))
			goto cant_help;
		break;
	case NEXTHDR_HOP:
	case NEXTHDR_ROUTING:
	case NEXTHDR_DEST: {
		u8 *opthdr = nhdr;

		if (protocol != IPPROTO_IPV6 || !spec->ext_hdrs_okay)
			goto need_help;

		ip_proto = opthdr[0];
		nhdr += (opthdr[1] + 1) << 3;

		goto ip_proto_again;
	}
	default:
		goto need_help;
	}

	/* Passed the tests for offloading checksum */
	return true;

need_help:
	if (csum_help && !skb_shinfo(skb)->gso_size)
		skb_checksum_help(skb);
cant_help:
	return false;
}
EXPORT_SYMBOL(__skb_csum_offload_chk);

2626
__be16 skb_network_protocol(struct sk_buff *skb, int *depth)
2627
{
A
Al Viro 已提交
2628
	__be16 type = skb->protocol;
2629

2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640
	/* Tunnel gso handlers can set protocol to ethernet. */
	if (type == htons(ETH_P_TEB)) {
		struct ethhdr *eth;

		if (unlikely(!pskb_may_pull(skb, sizeof(struct ethhdr))))
			return 0;

		eth = (struct ethhdr *)skb_mac_header(skb);
		type = eth->h_proto;
	}

2641
	return __vlan_get_protocol(skb, type, depth);
2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653
}

/**
 *	skb_mac_gso_segment - mac layer segmentation handler.
 *	@skb: buffer to segment
 *	@features: features for the output path (see dev->features)
 */
struct sk_buff *skb_mac_gso_segment(struct sk_buff *skb,
				    netdev_features_t features)
{
	struct sk_buff *segs = ERR_PTR(-EPROTONOSUPPORT);
	struct packet_offload *ptype;
2654 2655
	int vlan_depth = skb->mac_len;
	__be16 type = skb_network_protocol(skb, &vlan_depth);
2656 2657 2658 2659

	if (unlikely(!type))
		return ERR_PTR(-EINVAL);

2660
	__skb_pull(skb, vlan_depth);
2661 2662

	rcu_read_lock();
2663
	list_for_each_entry_rcu(ptype, &offload_base, list) {
2664 2665
		if (ptype->type == type && ptype->callbacks.gso_segment) {
			segs = ptype->callbacks.gso_segment(skb, features);
2666 2667 2668 2669 2670
			break;
		}
	}
	rcu_read_unlock();

2671
	__skb_push(skb, skb->data - skb_mac_header(skb));
2672

2673 2674
	return segs;
}
2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697
EXPORT_SYMBOL(skb_mac_gso_segment);


/* openvswitch calls this on rx path, so we need a different check.
 */
static inline bool skb_needs_check(struct sk_buff *skb, bool tx_path)
{
	if (tx_path)
		return skb->ip_summed != CHECKSUM_PARTIAL;
	else
		return skb->ip_summed == CHECKSUM_NONE;
}

/**
 *	__skb_gso_segment - Perform segmentation on skb.
 *	@skb: buffer to segment
 *	@features: features for the output path (see dev->features)
 *	@tx_path: whether it is called in TX path
 *
 *	This function segments the given skb and returns a list of segments.
 *
 *	It may return NULL if the skb requires no segmentation.  This is
 *	only possible when GSO is used for verifying header integrity.
2698 2699
 *
 *	Segmentation preserves SKB_SGO_CB_OFFSET bytes of previous skb cb.
2700 2701 2702 2703 2704 2705 2706 2707 2708
 */
struct sk_buff *__skb_gso_segment(struct sk_buff *skb,
				  netdev_features_t features, bool tx_path)
{
	if (unlikely(skb_needs_check(skb, tx_path))) {
		int err;

		skb_warn_bad_offload(skb);

2709 2710
		err = skb_cow_head(skb, 0);
		if (err < 0)
2711 2712 2713
			return ERR_PTR(err);
	}

2714 2715 2716
	BUILD_BUG_ON(SKB_SGO_CB_OFFSET +
		     sizeof(*SKB_GSO_CB(skb)) > sizeof(skb->cb));

2717
	SKB_GSO_CB(skb)->mac_offset = skb_headroom(skb);
2718 2719
	SKB_GSO_CB(skb)->encap_level = 0;

2720 2721 2722 2723 2724
	skb_reset_mac_header(skb);
	skb_reset_mac_len(skb);

	return skb_mac_gso_segment(skb, features);
}
2725
EXPORT_SYMBOL(__skb_gso_segment);
2726

2727 2728 2729 2730 2731
/* Take action when hardware reception checksum errors are detected. */
#ifdef CONFIG_BUG
void netdev_rx_csum_fault(struct net_device *dev)
{
	if (net_ratelimit()) {
2732
		pr_err("%s: hw csum failure\n", dev ? dev->name : "<unknown>");
2733 2734 2735 2736 2737 2738
		dump_stack();
	}
}
EXPORT_SYMBOL(netdev_rx_csum_fault);
#endif

L
Linus Torvalds 已提交
2739 2740 2741 2742 2743
/* 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.
 */

2744
static int illegal_highdma(struct net_device *dev, struct sk_buff *skb)
L
Linus Torvalds 已提交
2745
{
2746
#ifdef CONFIG_HIGHMEM
L
Linus Torvalds 已提交
2747
	int i;
2748
	if (!(dev->features & NETIF_F_HIGHDMA)) {
2749 2750 2751
		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)))
2752
				return 1;
2753
		}
2754
	}
L
Linus Torvalds 已提交
2755

2756 2757
	if (PCI_DMA_BUS_IS_PHYS) {
		struct device *pdev = dev->dev.parent;
L
Linus Torvalds 已提交
2758

E
Eric Dumazet 已提交
2759 2760
		if (!pdev)
			return 0;
2761
		for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
2762 2763
			skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
			dma_addr_t addr = page_to_phys(skb_frag_page(frag));
2764 2765 2766 2767
			if (!pdev->dma_mask || addr + PAGE_SIZE - 1 > *pdev->dma_mask)
				return 1;
		}
	}
2768
#endif
L
Linus Torvalds 已提交
2769 2770 2771
	return 0;
}

2772 2773 2774
/* If MPLS offload request, verify we are testing hardware MPLS features
 * instead of standard features for the netdev.
 */
P
Pravin B Shelar 已提交
2775
#if IS_ENABLED(CONFIG_NET_MPLS_GSO)
2776 2777 2778 2779
static netdev_features_t net_mpls_features(struct sk_buff *skb,
					   netdev_features_t features,
					   __be16 type)
{
2780
	if (eth_p_mpls(type))
2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793
		features &= skb->dev->mpls_features;

	return features;
}
#else
static netdev_features_t net_mpls_features(struct sk_buff *skb,
					   netdev_features_t features,
					   __be16 type)
{
	return features;
}
#endif

2794
static netdev_features_t harmonize_features(struct sk_buff *skb,
2795
	netdev_features_t features)
2796
{
2797
	int tmp;
2798 2799 2800 2801
	__be16 type;

	type = skb_network_protocol(skb, &tmp);
	features = net_mpls_features(skb, features, type);
2802

2803
	if (skb->ip_summed != CHECKSUM_NONE &&
2804
	    !can_checksum_protocol(features, type)) {
2805
		features &= ~NETIF_F_CSUM_MASK;
2806
	} else if (illegal_highdma(skb->dev, skb)) {
2807 2808 2809 2810 2811 2812
		features &= ~NETIF_F_SG;
	}

	return features;
}

2813 2814 2815 2816 2817 2818 2819 2820
netdev_features_t passthru_features_check(struct sk_buff *skb,
					  struct net_device *dev,
					  netdev_features_t features)
{
	return features;
}
EXPORT_SYMBOL(passthru_features_check);

2821 2822 2823 2824 2825 2826 2827
static netdev_features_t dflt_features_check(const struct sk_buff *skb,
					     struct net_device *dev,
					     netdev_features_t features)
{
	return vlan_features_check(skb, features);
}

2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850
static netdev_features_t gso_features_check(const struct sk_buff *skb,
					    struct net_device *dev,
					    netdev_features_t features)
{
	u16 gso_segs = skb_shinfo(skb)->gso_segs;

	if (gso_segs > dev->gso_max_segs)
		return features & ~NETIF_F_GSO_MASK;

	/* Make sure to clear the IPv4 ID mangling feature if
	 * the IPv4 header has the potential to be fragmented.
	 */
	if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4) {
		struct iphdr *iph = skb->encapsulation ?
				    inner_ip_hdr(skb) : ip_hdr(skb);

		if (!(iph->frag_off & htons(IP_DF)))
			features &= ~NETIF_F_TSO_MANGLEID;
	}

	return features;
}

2851
netdev_features_t netif_skb_features(struct sk_buff *skb)
2852
{
2853
	struct net_device *dev = skb->dev;
2854
	netdev_features_t features = dev->features;
2855

2856 2857
	if (skb_is_gso(skb))
		features = gso_features_check(skb, dev, features);
2858

2859 2860 2861 2862 2863 2864 2865
	/* If encapsulation offload request, verify we are testing
	 * hardware encapsulation features instead of standard
	 * features for the netdev
	 */
	if (skb->encapsulation)
		features &= dev->hw_enc_features;

2866 2867 2868 2869 2870
	if (skb_vlan_tagged(skb))
		features = netdev_intersect_features(features,
						     dev->vlan_features |
						     NETIF_F_HW_VLAN_CTAG_TX |
						     NETIF_F_HW_VLAN_STAG_TX);
2871

2872 2873 2874
	if (dev->netdev_ops->ndo_features_check)
		features &= dev->netdev_ops->ndo_features_check(skb, dev,
								features);
2875 2876
	else
		features &= dflt_features_check(skb, dev, features);
2877

2878
	return harmonize_features(skb, features);
2879
}
2880
EXPORT_SYMBOL(netif_skb_features);
2881

2882
static int xmit_one(struct sk_buff *skb, struct net_device *dev,
2883
		    struct netdev_queue *txq, bool more)
2884
{
2885 2886
	unsigned int len;
	int rc;
2887

2888
	if (!list_empty(&ptype_all) || !list_empty(&dev->ptype_all))
2889
		dev_queue_xmit_nit(skb, dev);
2890

2891 2892
	len = skb->len;
	trace_net_dev_start_xmit(skb, dev);
2893
	rc = netdev_start_xmit(skb, dev, txq, more);
2894
	trace_net_dev_xmit(skb, rc, dev, len);
E
Eric Dumazet 已提交
2895

2896 2897
	return rc;
}
2898

2899 2900
struct sk_buff *dev_hard_start_xmit(struct sk_buff *first, struct net_device *dev,
				    struct netdev_queue *txq, int *ret)
2901 2902 2903
{
	struct sk_buff *skb = first;
	int rc = NETDEV_TX_OK;
2904

2905 2906
	while (skb) {
		struct sk_buff *next = skb->next;
2907

2908
		skb->next = NULL;
2909
		rc = xmit_one(skb, dev, txq, next != NULL);
2910 2911 2912 2913
		if (unlikely(!dev_xmit_complete(rc))) {
			skb->next = next;
			goto out;
		}
2914

2915 2916 2917 2918
		skb = next;
		if (netif_xmit_stopped(txq) && skb) {
			rc = NETDEV_TX_BUSY;
			break;
2919
		}
2920
	}
2921

2922 2923 2924 2925
out:
	*ret = rc;
	return skb;
}
2926

E
Eric Dumazet 已提交
2927 2928
static struct sk_buff *validate_xmit_vlan(struct sk_buff *skb,
					  netdev_features_t features)
2929
{
2930
	if (skb_vlan_tag_present(skb) &&
2931 2932
	    !vlan_hw_offload_capable(features, skb->vlan_proto))
		skb = __vlan_hwaccel_push_inside(skb);
2933 2934
	return skb;
}
2935

2936
static struct sk_buff *validate_xmit_skb(struct sk_buff *skb, struct net_device *dev)
2937 2938
{
	netdev_features_t features;
2939

2940 2941
	if (skb->next)
		return skb;
2942

2943 2944 2945 2946
	features = netif_skb_features(skb);
	skb = validate_xmit_vlan(skb, features);
	if (unlikely(!skb))
		goto out_null;
2947

2948
	if (netif_needs_gso(skb, features)) {
2949 2950 2951
		struct sk_buff *segs;

		segs = skb_gso_segment(skb, features);
2952
		if (IS_ERR(segs)) {
2953
			goto out_kfree_skb;
2954 2955 2956
		} else if (segs) {
			consume_skb(skb);
			skb = segs;
2957
		}
2958 2959 2960 2961
	} else {
		if (skb_needs_linearize(skb, features) &&
		    __skb_linearize(skb))
			goto out_kfree_skb;
2962

2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973
		/* If packet is not checksummed and device does not
		 * support checksumming for this protocol, complete
		 * checksumming here.
		 */
		if (skb->ip_summed == CHECKSUM_PARTIAL) {
			if (skb->encapsulation)
				skb_set_inner_transport_header(skb,
							       skb_checksum_start_offset(skb));
			else
				skb_set_transport_header(skb,
							 skb_checksum_start_offset(skb));
2974
			if (!(features & NETIF_F_CSUM_MASK) &&
2975 2976
			    skb_checksum_help(skb))
				goto out_kfree_skb;
2977
		}
2978
	}
2979

2980
	return skb;
2981

2982 2983
out_kfree_skb:
	kfree_skb(skb);
2984 2985 2986
out_null:
	return NULL;
}
2987

2988 2989 2990 2991
struct sk_buff *validate_xmit_skb_list(struct sk_buff *skb, struct net_device *dev)
{
	struct sk_buff *next, *head = NULL, *tail;

2992
	for (; skb != NULL; skb = next) {
2993 2994
		next = skb->next;
		skb->next = NULL;
2995 2996 2997 2998

		/* in case skb wont be segmented, point to itself */
		skb->prev = skb;

2999
		skb = validate_xmit_skb(skb, dev);
3000 3001
		if (!skb)
			continue;
3002

3003 3004 3005 3006 3007 3008 3009 3010
		if (!head)
			head = skb;
		else
			tail->next = skb;
		/* If skb was segmented, skb->prev points to
		 * the last segment. If not, it still contains skb.
		 */
		tail = skb->prev;
3011 3012
	}
	return head;
3013 3014
}

3015 3016 3017 3018 3019 3020 3021 3022 3023 3024
static void qdisc_pkt_len_init(struct sk_buff *skb)
{
	const struct skb_shared_info *shinfo = skb_shinfo(skb);

	qdisc_skb_cb(skb)->pkt_len = skb->len;

	/* To get more precise estimation of bytes sent on wire,
	 * we add to pkt_len the headers size of all segments
	 */
	if (shinfo->gso_size)  {
3025
		unsigned int hdr_len;
3026
		u16 gso_segs = shinfo->gso_segs;
3027

3028 3029 3030 3031
		/* mac layer + network layer */
		hdr_len = skb_transport_header(skb) - skb_mac_header(skb);

		/* + transport layer */
3032 3033 3034 3035
		if (likely(shinfo->gso_type & (SKB_GSO_TCPV4 | SKB_GSO_TCPV6)))
			hdr_len += tcp_hdrlen(skb);
		else
			hdr_len += sizeof(struct udphdr);
3036 3037 3038 3039 3040 3041

		if (shinfo->gso_type & SKB_GSO_DODGY)
			gso_segs = DIV_ROUND_UP(skb->len - hdr_len,
						shinfo->gso_size);

		qdisc_skb_cb(skb)->pkt_len += (gso_segs - 1) * hdr_len;
3042 3043 3044
	}
}

3045 3046 3047 3048 3049
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
3050
	bool contended;
3051 3052
	int rc;

E
Eric Dumazet 已提交
3053
	qdisc_calculate_pkt_len(skb, q);
3054 3055 3056
	/*
	 * Heuristic to force contended enqueues to serialize on a
	 * separate lock before trying to get qdisc main lock.
Y
Ying Xue 已提交
3057 3058
	 * This permits __QDISC___STATE_RUNNING owner to get the lock more
	 * often and dequeue packets faster.
3059
	 */
E
Eric Dumazet 已提交
3060
	contended = qdisc_is_running(q);
3061 3062 3063
	if (unlikely(contended))
		spin_lock(&q->busylock);

3064 3065 3066 3067 3068
	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) &&
3069
		   qdisc_run_begin(q)) {
3070 3071 3072 3073 3074
		/*
		 * 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.
		 */
3075 3076 3077

		qdisc_bstats_update(q, skb);

3078
		if (sch_direct_xmit(skb, q, dev, txq, root_lock, true)) {
3079 3080 3081 3082
			if (unlikely(contended)) {
				spin_unlock(&q->busylock);
				contended = false;
			}
3083
			__qdisc_run(q);
3084
		} else
3085
			qdisc_run_end(q);
3086 3087 3088

		rc = NET_XMIT_SUCCESS;
	} else {
E
Eric Dumazet 已提交
3089
		rc = q->enqueue(skb, q) & NET_XMIT_MASK;
3090 3091 3092 3093 3094 3095 3096
		if (qdisc_run_begin(q)) {
			if (unlikely(contended)) {
				spin_unlock(&q->busylock);
				contended = false;
			}
			__qdisc_run(q);
		}
3097 3098
	}
	spin_unlock(root_lock);
3099 3100
	if (unlikely(contended))
		spin_unlock(&q->busylock);
3101 3102 3103
	return rc;
}

3104
#if IS_ENABLED(CONFIG_CGROUP_NET_PRIO)
3105 3106
static void skb_update_prio(struct sk_buff *skb)
{
I
Igor Maravic 已提交
3107
	struct netprio_map *map = rcu_dereference_bh(skb->dev->priomap);
3108

3109
	if (!skb->priority && skb->sk && map) {
3110 3111
		unsigned int prioidx =
			sock_cgroup_prioidx(&skb->sk->sk_cgrp_data);
3112 3113 3114 3115

		if (prioidx < map->priomap_len)
			skb->priority = map->priomap[prioidx];
	}
3116 3117 3118 3119 3120
}
#else
#define skb_update_prio(skb)
#endif

3121 3122 3123
DEFINE_PER_CPU(int, xmit_recursion);
EXPORT_SYMBOL(xmit_recursion);

3124
#define RECURSION_LIMIT 10
3125

3126 3127
/**
 *	dev_loopback_xmit - loop back @skb
3128 3129
 *	@net: network namespace this loopback is happening in
 *	@sk:  sk needed to be a netfilter okfn
3130 3131
 *	@skb: buffer to transmit
 */
3132
int dev_loopback_xmit(struct net *net, struct sock *sk, struct sk_buff *skb)
3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144
{
	skb_reset_mac_header(skb);
	__skb_pull(skb, skb_network_offset(skb));
	skb->pkt_type = PACKET_LOOPBACK;
	skb->ip_summed = CHECKSUM_UNNECESSARY;
	WARN_ON(!skb_dst(skb));
	skb_dst_force(skb);
	netif_rx_ni(skb);
	return 0;
}
EXPORT_SYMBOL(dev_loopback_xmit);

D
Daniel Borkmann 已提交
3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187
#ifdef CONFIG_NET_EGRESS
static struct sk_buff *
sch_handle_egress(struct sk_buff *skb, int *ret, struct net_device *dev)
{
	struct tcf_proto *cl = rcu_dereference_bh(dev->egress_cl_list);
	struct tcf_result cl_res;

	if (!cl)
		return skb;

	/* skb->tc_verd and qdisc_skb_cb(skb)->pkt_len were already set
	 * earlier by the caller.
	 */
	qdisc_bstats_cpu_update(cl->q, skb);

	switch (tc_classify(skb, cl, &cl_res, false)) {
	case TC_ACT_OK:
	case TC_ACT_RECLASSIFY:
		skb->tc_index = TC_H_MIN(cl_res.classid);
		break;
	case TC_ACT_SHOT:
		qdisc_qstats_cpu_drop(cl->q);
		*ret = NET_XMIT_DROP;
		goto drop;
	case TC_ACT_STOLEN:
	case TC_ACT_QUEUED:
		*ret = NET_XMIT_SUCCESS;
drop:
		kfree_skb(skb);
		return NULL;
	case TC_ACT_REDIRECT:
		/* No need to push/pop skb's mac_header here on egress! */
		skb_do_redirect(skb);
		*ret = NET_XMIT_SUCCESS;
		return NULL;
	default:
		break;
	}

	return skb;
}
#endif /* CONFIG_NET_EGRESS */

3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229
static inline int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
{
#ifdef CONFIG_XPS
	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[skb->sender_cpu - 1]);
		if (map) {
			if (map->len == 1)
				queue_index = map->queues[0];
			else
				queue_index = map->queues[reciprocal_scale(skb_get_hash(skb),
									   map->len)];
			if (unlikely(queue_index >= dev->real_num_tx_queues))
				queue_index = -1;
		}
	}
	rcu_read_unlock();

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

static u16 __netdev_pick_tx(struct net_device *dev, struct sk_buff *skb)
{
	struct sock *sk = skb->sk;
	int queue_index = sk_tx_queue_get(sk);

	if (queue_index < 0 || skb->ooo_okay ||
	    queue_index >= dev->real_num_tx_queues) {
		int new_index = get_xps_queue(dev, skb);
		if (new_index < 0)
			new_index = skb_tx_hash(dev, skb);

		if (queue_index != new_index && sk &&
3230
		    sk_fullsock(sk) &&
3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246
		    rcu_access_pointer(sk->sk_dst_cache))
			sk_tx_queue_set(sk, new_index);

		queue_index = new_index;
	}

	return queue_index;
}

struct netdev_queue *netdev_pick_tx(struct net_device *dev,
				    struct sk_buff *skb,
				    void *accel_priv)
{
	int queue_index = 0;

#ifdef CONFIG_XPS
3247 3248 3249
	u32 sender_cpu = skb->sender_cpu - 1;

	if (sender_cpu >= (u32)NR_CPUS)
3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268
		skb->sender_cpu = raw_smp_processor_id() + 1;
#endif

	if (dev->real_num_tx_queues != 1) {
		const struct net_device_ops *ops = dev->netdev_ops;
		if (ops->ndo_select_queue)
			queue_index = ops->ndo_select_queue(dev, skb, accel_priv,
							    __netdev_pick_tx);
		else
			queue_index = __netdev_pick_tx(dev, skb);

		if (!accel_priv)
			queue_index = netdev_cap_txqueue(dev, queue_index);
	}

	skb_set_queue_mapping(skb, queue_index);
	return netdev_get_tx_queue(dev, queue_index);
}

3269
/**
3270
 *	__dev_queue_xmit - transmit a buffer
3271
 *	@skb: buffer to transmit
3272
 *	@accel_priv: private data used for L2 forwarding offload
3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294
 *
 *	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
 */
3295
static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv)
L
Linus Torvalds 已提交
3296 3297
{
	struct net_device *dev = skb->dev;
3298
	struct netdev_queue *txq;
L
Linus Torvalds 已提交
3299 3300 3301
	struct Qdisc *q;
	int rc = -ENOMEM;

3302 3303
	skb_reset_mac_header(skb);

3304 3305 3306
	if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_SCHED_TSTAMP))
		__skb_tstamp_tx(skb, NULL, skb->sk, SCM_TSTAMP_SCHED);

3307 3308
	/* Disable soft irqs for various locks below. Also
	 * stops preemption for RCU.
L
Linus Torvalds 已提交
3309
	 */
3310
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
3311

3312 3313
	skb_update_prio(skb);

D
Daniel Borkmann 已提交
3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324
	qdisc_pkt_len_init(skb);
#ifdef CONFIG_NET_CLS_ACT
	skb->tc_verd = SET_TC_AT(skb->tc_verd, AT_EGRESS);
# ifdef CONFIG_NET_EGRESS
	if (static_key_false(&egress_needed)) {
		skb = sch_handle_egress(skb, &rc, dev);
		if (!skb)
			goto out;
	}
# endif
#endif
3325 3326 3327 3328 3329 3330 3331 3332
	/* If device/qdisc don't need skb->dst, release it right now while
	 * its hot in this cpu cache.
	 */
	if (dev->priv_flags & IFF_XMIT_DST_RELEASE)
		skb_dst_drop(skb);
	else
		skb_dst_force(skb);

3333 3334 3335 3336 3337 3338 3339 3340 3341 3342
#ifdef CONFIG_NET_SWITCHDEV
	/* Don't forward if offload device already forwarded */
	if (skb->offload_fwd_mark &&
	    skb->offload_fwd_mark == dev->offload_fwd_mark) {
		consume_skb(skb);
		rc = NET_XMIT_SUCCESS;
		goto out;
	}
#endif

3343
	txq = netdev_pick_tx(dev, skb, accel_priv);
3344
	q = rcu_dereference_bh(txq->qdisc);
3345

3346
	trace_net_dev_queue(skb);
L
Linus Torvalds 已提交
3347
	if (q->enqueue) {
3348
		rc = __dev_xmit_skb(skb, q, dev, txq);
3349
		goto out;
L
Linus Torvalds 已提交
3350 3351 3352 3353 3354
	}

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

H
Herbert Xu 已提交
3355 3356
	   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
3357 3358 3359 3360 3361 3362 3363 3364 3365 3366
	   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 */

3367
		if (txq->xmit_lock_owner != cpu) {
L
Linus Torvalds 已提交
3368

3369 3370 3371
			if (__this_cpu_read(xmit_recursion) > RECURSION_LIMIT)
				goto recursion_alert;

3372 3373 3374 3375
			skb = validate_xmit_skb(skb, dev);
			if (!skb)
				goto drop;

3376
			HARD_TX_LOCK(dev, txq, cpu);
L
Linus Torvalds 已提交
3377

3378
			if (!netif_xmit_stopped(txq)) {
3379
				__this_cpu_inc(xmit_recursion);
3380
				skb = dev_hard_start_xmit(skb, dev, txq, &rc);
3381
				__this_cpu_dec(xmit_recursion);
3382
				if (dev_xmit_complete(rc)) {
3383
					HARD_TX_UNLOCK(dev, txq);
L
Linus Torvalds 已提交
3384 3385 3386
					goto out;
				}
			}
3387
			HARD_TX_UNLOCK(dev, txq);
3388 3389
			net_crit_ratelimited("Virtual device %s asks to queue packet!\n",
					     dev->name);
L
Linus Torvalds 已提交
3390 3391
		} else {
			/* Recursion is detected! It is possible,
3392 3393 3394
			 * unfortunately
			 */
recursion_alert:
3395 3396
			net_crit_ratelimited("Dead loop on virtual device %s, fix it urgently!\n",
					     dev->name);
L
Linus Torvalds 已提交
3397 3398 3399 3400
		}
	}

	rc = -ENETDOWN;
3401
drop:
3402
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
3403

3404
	atomic_long_inc(&dev->tx_dropped);
3405
	kfree_skb_list(skb);
L
Linus Torvalds 已提交
3406 3407
	return rc;
out:
3408
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
3409 3410
	return rc;
}
3411

3412
int dev_queue_xmit(struct sk_buff *skb)
3413 3414 3415
{
	return __dev_queue_xmit(skb, NULL);
}
3416
EXPORT_SYMBOL(dev_queue_xmit);
L
Linus Torvalds 已提交
3417

3418 3419 3420 3421 3422 3423
int dev_queue_xmit_accel(struct sk_buff *skb, void *accel_priv)
{
	return __dev_queue_xmit(skb, accel_priv);
}
EXPORT_SYMBOL(dev_queue_xmit_accel);

L
Linus Torvalds 已提交
3424 3425 3426 3427 3428

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

3429
int netdev_max_backlog __read_mostly = 1000;
E
Eric Dumazet 已提交
3430 3431
EXPORT_SYMBOL(netdev_max_backlog);

E
Eric Dumazet 已提交
3432
int netdev_tstamp_prequeue __read_mostly = 1;
3433 3434
int netdev_budget __read_mostly = 300;
int weight_p __read_mostly = 64;            /* old backlog weight */
L
Linus Torvalds 已提交
3435

E
Eric Dumazet 已提交
3436 3437 3438 3439 3440 3441 3442 3443
/* 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);
}

3444 3445 3446
#ifdef CONFIG_RPS

/* One global table that all flow-based protocols share. */
E
Eric Dumazet 已提交
3447
struct rps_sock_flow_table __rcu *rps_sock_flow_table __read_mostly;
3448
EXPORT_SYMBOL(rps_sock_flow_table);
3449 3450
u32 rps_cpu_mask __read_mostly;
EXPORT_SYMBOL(rps_cpu_mask);
3451

3452
struct static_key rps_needed __read_mostly;
3453

3454 3455 3456 3457
static struct rps_dev_flow *
set_rps_cpu(struct net_device *dev, struct sk_buff *skb,
	    struct rps_dev_flow *rflow, u16 next_cpu)
{
3458
	if (next_cpu < nr_cpu_ids) {
3459 3460 3461 3462 3463 3464 3465 3466 3467
#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? */
3468 3469
		if (!skb_rx_queue_recorded(skb) || !dev->rx_cpu_rmap ||
		    !(dev->features & NETIF_F_NTUPLE))
3470 3471 3472 3473 3474 3475 3476 3477 3478
			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;
3479
		flow_id = skb_get_hash(skb) & flow_table->mask;
3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491
		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 =
3492
			per_cpu(softnet_data, next_cpu).input_queue_head;
3493 3494
	}

3495
	rflow->cpu = next_cpu;
3496 3497 3498
	return rflow;
}

3499 3500 3501 3502 3503 3504 3505 3506
/*
 * 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)
{
3507 3508
	const struct rps_sock_flow_table *sock_flow_table;
	struct netdev_rx_queue *rxqueue = dev->_rx;
3509
	struct rps_dev_flow_table *flow_table;
3510
	struct rps_map *map;
3511
	int cpu = -1;
3512
	u32 tcpu;
3513
	u32 hash;
3514 3515 3516

	if (skb_rx_queue_recorded(skb)) {
		u16 index = skb_get_rx_queue(skb);
3517

3518 3519 3520 3521 3522
		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);
3523 3524
			goto done;
		}
3525 3526
		rxqueue += index;
	}
3527

3528 3529 3530
	/* Avoid computing hash if RFS/RPS is not active for this rxqueue */

	flow_table = rcu_dereference(rxqueue->rps_flow_table);
E
Eric Dumazet 已提交
3531
	map = rcu_dereference(rxqueue->rps_map);
3532
	if (!flow_table && !map)
3533 3534
		goto done;

3535
	skb_reset_network_header(skb);
3536 3537
	hash = skb_get_hash(skb);
	if (!hash)
3538 3539
		goto done;

T
Tom Herbert 已提交
3540 3541 3542
	sock_flow_table = rcu_dereference(rps_sock_flow_table);
	if (flow_table && sock_flow_table) {
		struct rps_dev_flow *rflow;
3543 3544 3545 3546 3547 3548 3549
		u32 next_cpu;
		u32 ident;

		/* First check into global flow table if there is a match */
		ident = sock_flow_table->ents[hash & sock_flow_table->mask];
		if ((ident ^ hash) & ~rps_cpu_mask)
			goto try_rps;
T
Tom Herbert 已提交
3550

3551 3552 3553 3554 3555
		next_cpu = ident & rps_cpu_mask;

		/* OK, now we know there is a match,
		 * we can look at the local (per receive queue) flow table
		 */
3556
		rflow = &flow_table->flows[hash & flow_table->mask];
T
Tom Herbert 已提交
3557 3558 3559 3560 3561 3562
		tcpu = rflow->cpu;

		/*
		 * 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:
3563
		 *   - Current CPU is unset (>= nr_cpu_ids).
T
Tom Herbert 已提交
3564 3565 3566 3567 3568 3569 3570
		 *   - 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) &&
3571
		    (tcpu >= nr_cpu_ids || !cpu_online(tcpu) ||
T
Tom Herbert 已提交
3572
		     ((int)(per_cpu(softnet_data, tcpu).input_queue_head -
3573 3574
		      rflow->last_qtail)) >= 0)) {
			tcpu = next_cpu;
3575
			rflow = set_rps_cpu(dev, skb, rflow, next_cpu);
3576
		}
3577

3578
		if (tcpu < nr_cpu_ids && cpu_online(tcpu)) {
T
Tom Herbert 已提交
3579 3580 3581 3582 3583 3584
			*rflowp = rflow;
			cpu = tcpu;
			goto done;
		}
	}

3585 3586
try_rps:

T
Tom Herbert 已提交
3587
	if (map) {
3588
		tcpu = map->cpus[reciprocal_scale(hash, map->len)];
T
Tom Herbert 已提交
3589 3590 3591 3592 3593 3594 3595 3596 3597 3598
		if (cpu_online(tcpu)) {
			cpu = tcpu;
			goto done;
		}
	}

done:
	return cpu;
}

3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618
#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;
3619
	unsigned int cpu;
3620 3621 3622 3623 3624 3625

	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);
3626
		if (rflow->filter == filter_id && cpu < nr_cpu_ids &&
3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638
		    ((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
3639
/* Called from hardirq (IPI) context */
E
Eric Dumazet 已提交
3640
static void rps_trigger_softirq(void *data)
T
Tom Herbert 已提交
3641
{
E
Eric Dumazet 已提交
3642 3643
	struct softnet_data *sd = data;

E
Eric Dumazet 已提交
3644
	____napi_schedule(sd, &sd->backlog);
C
Changli Gao 已提交
3645
	sd->received_rps++;
T
Tom Herbert 已提交
3646
}
E
Eric Dumazet 已提交
3647

T
Tom Herbert 已提交
3648
#endif /* CONFIG_RPS */
T
Tom Herbert 已提交
3649

E
Eric Dumazet 已提交
3650 3651 3652 3653 3654 3655 3656 3657
/*
 * 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
3658
	struct softnet_data *mysd = this_cpu_ptr(&softnet_data);
E
Eric Dumazet 已提交
3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670

	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;
}

3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684
#ifdef CONFIG_NET_FLOW_LIMIT
int netdev_flow_limit_table_len __read_mostly = (1 << 12);
#endif

static bool skb_flow_limit(struct sk_buff *skb, unsigned int qlen)
{
#ifdef CONFIG_NET_FLOW_LIMIT
	struct sd_flow_limit *fl;
	struct softnet_data *sd;
	unsigned int old_flow, new_flow;

	if (qlen < (netdev_max_backlog >> 1))
		return false;

3685
	sd = this_cpu_ptr(&softnet_data);
3686 3687 3688 3689

	rcu_read_lock();
	fl = rcu_dereference(sd->flow_limit);
	if (fl) {
3690
		new_flow = skb_get_hash(skb) & (fl->num_buckets - 1);
3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710
		old_flow = fl->history[fl->history_head];
		fl->history[fl->history_head] = new_flow;

		fl->history_head++;
		fl->history_head &= FLOW_LIMIT_HISTORY - 1;

		if (likely(fl->buckets[old_flow]))
			fl->buckets[old_flow]--;

		if (++fl->buckets[new_flow] > (FLOW_LIMIT_HISTORY >> 1)) {
			fl->count++;
			rcu_read_unlock();
			return true;
		}
	}
	rcu_read_unlock();
#endif
	return false;
}

T
Tom Herbert 已提交
3711 3712 3713 3714
/*
 * 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
3715 3716
static int enqueue_to_backlog(struct sk_buff *skb, int cpu,
			      unsigned int *qtail)
T
Tom Herbert 已提交
3717
{
E
Eric Dumazet 已提交
3718
	struct softnet_data *sd;
T
Tom Herbert 已提交
3719
	unsigned long flags;
3720
	unsigned int qlen;
T
Tom Herbert 已提交
3721

E
Eric Dumazet 已提交
3722
	sd = &per_cpu(softnet_data, cpu);
T
Tom Herbert 已提交
3723 3724 3725

	local_irq_save(flags);

E
Eric Dumazet 已提交
3726
	rps_lock(sd);
3727 3728
	if (!netif_running(skb->dev))
		goto drop;
3729 3730
	qlen = skb_queue_len(&sd->input_pkt_queue);
	if (qlen <= netdev_max_backlog && !skb_flow_limit(skb, qlen)) {
3731
		if (qlen) {
T
Tom Herbert 已提交
3732
enqueue:
E
Eric Dumazet 已提交
3733
			__skb_queue_tail(&sd->input_pkt_queue, skb);
3734
			input_queue_tail_incr_save(sd, qtail);
E
Eric Dumazet 已提交
3735
			rps_unlock(sd);
3736
			local_irq_restore(flags);
T
Tom Herbert 已提交
3737 3738 3739
			return NET_RX_SUCCESS;
		}

3740 3741 3742 3743
		/* 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
3744
			if (!rps_ipi_queued(sd))
E
Eric Dumazet 已提交
3745
				____napi_schedule(sd, &sd->backlog);
T
Tom Herbert 已提交
3746 3747 3748 3749
		}
		goto enqueue;
	}

3750
drop:
C
Changli Gao 已提交
3751
	sd->dropped++;
E
Eric Dumazet 已提交
3752
	rps_unlock(sd);
T
Tom Herbert 已提交
3753 3754 3755

	local_irq_restore(flags);

3756
	atomic_long_inc(&skb->dev->rx_dropped);
T
Tom Herbert 已提交
3757 3758 3759
	kfree_skb(skb);
	return NET_RX_DROP;
}
L
Linus Torvalds 已提交
3760

3761
static int netif_rx_internal(struct sk_buff *skb)
L
Linus Torvalds 已提交
3762
{
3763
	int ret;
L
Linus Torvalds 已提交
3764

3765
	net_timestamp_check(netdev_tstamp_prequeue, skb);
L
Linus Torvalds 已提交
3766

3767
	trace_netif_rx(skb);
E
Eric Dumazet 已提交
3768
#ifdef CONFIG_RPS
3769
	if (static_key_false(&rps_needed)) {
T
Tom Herbert 已提交
3770
		struct rps_dev_flow voidflow, *rflow = &voidflow;
3771 3772
		int cpu;

3773
		preempt_disable();
3774
		rcu_read_lock();
T
Tom Herbert 已提交
3775 3776

		cpu = get_rps_cpu(skb->dev, skb, &rflow);
3777 3778
		if (cpu < 0)
			cpu = smp_processor_id();
T
Tom Herbert 已提交
3779 3780 3781

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

3782
		rcu_read_unlock();
3783
		preempt_enable();
3784 3785
	} else
#endif
T
Tom Herbert 已提交
3786 3787 3788 3789 3790
	{
		unsigned int qtail;
		ret = enqueue_to_backlog(skb, get_cpu(), &qtail);
		put_cpu();
	}
3791
	return ret;
L
Linus Torvalds 已提交
3792
}
3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814

/**
 *	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)
{
	trace_netif_rx_entry(skb);

	return netif_rx_internal(skb);
}
E
Eric Dumazet 已提交
3815
EXPORT_SYMBOL(netif_rx);
L
Linus Torvalds 已提交
3816 3817 3818 3819 3820

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

3821 3822
	trace_netif_rx_ni_entry(skb);

L
Linus Torvalds 已提交
3823
	preempt_disable();
3824
	err = netif_rx_internal(skb);
L
Linus Torvalds 已提交
3825 3826 3827 3828 3829 3830 3831 3832 3833 3834
	if (local_softirq_pending())
		do_softirq();
	preempt_enable();

	return err;
}
EXPORT_SYMBOL(netif_rx_ni);

static void net_tx_action(struct softirq_action *h)
{
3835
	struct softnet_data *sd = this_cpu_ptr(&softnet_data);
L
Linus Torvalds 已提交
3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848

	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;

3849
			WARN_ON(atomic_read(&skb->users));
3850 3851 3852 3853
			if (likely(get_kfree_skb_cb(skb)->reason == SKB_REASON_CONSUMED))
				trace_consume_skb(skb);
			else
				trace_kfree_skb(skb, net_tx_action);
3854 3855 3856 3857 3858

			if (skb->fclone != SKB_FCLONE_UNAVAILABLE)
				__kfree_skb(skb);
			else
				__kfree_skb_defer(skb);
L
Linus Torvalds 已提交
3859
		}
3860 3861

		__kfree_skb_flush();
L
Linus Torvalds 已提交
3862 3863 3864
	}

	if (sd->output_queue) {
3865
		struct Qdisc *head;
L
Linus Torvalds 已提交
3866 3867 3868 3869

		local_irq_disable();
		head = sd->output_queue;
		sd->output_queue = NULL;
3870
		sd->output_queue_tailp = &sd->output_queue;
L
Linus Torvalds 已提交
3871 3872 3873
		local_irq_enable();

		while (head) {
3874 3875 3876
			struct Qdisc *q = head;
			spinlock_t *root_lock;

L
Linus Torvalds 已提交
3877 3878
			head = head->next_sched;

3879
			root_lock = qdisc_lock(q);
3880
			if (spin_trylock(root_lock)) {
3881
				smp_mb__before_atomic();
3882 3883
				clear_bit(__QDISC_STATE_SCHED,
					  &q->state);
3884 3885
				qdisc_run(q);
				spin_unlock(root_lock);
L
Linus Torvalds 已提交
3886
			} else {
3887
				if (!test_bit(__QDISC_STATE_DEACTIVATED,
3888
					      &q->state)) {
3889
					__netif_reschedule(q);
3890
				} else {
3891
					smp_mb__before_atomic();
3892 3893 3894
					clear_bit(__QDISC_STATE_SCHED,
						  &q->state);
				}
L
Linus Torvalds 已提交
3895 3896 3897 3898 3899
			}
		}
	}
}

3900 3901
#if (defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)) && \
    (defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE))
3902 3903 3904
/* This hook is defined here for ATM LANE */
int (*br_fdb_test_addr_hook)(struct net_device *dev,
			     unsigned char *addr) __read_mostly;
3905
EXPORT_SYMBOL_GPL(br_fdb_test_addr_hook);
3906
#endif
L
Linus Torvalds 已提交
3907

D
Daniel Borkmann 已提交
3908 3909 3910
static inline struct sk_buff *
sch_handle_ingress(struct sk_buff *skb, struct packet_type **pt_prev, int *ret,
		   struct net_device *orig_dev)
3911
{
3912
#ifdef CONFIG_NET_CLS_ACT
3913 3914
	struct tcf_proto *cl = rcu_dereference_bh(skb->dev->ingress_cl_list);
	struct tcf_result cl_res;
3915

3916 3917 3918
	/* If there's at least one ingress present somewhere (so
	 * we get here via enabled static key), remaining devices
	 * that are not configured with an ingress qdisc will bail
3919
	 * out here.
3920
	 */
3921
	if (!cl)
3922
		return skb;
3923 3924 3925
	if (*pt_prev) {
		*ret = deliver_skb(skb, *pt_prev, orig_dev);
		*pt_prev = NULL;
L
Linus Torvalds 已提交
3926 3927
	}

3928
	qdisc_skb_cb(skb)->pkt_len = skb->len;
3929
	skb->tc_verd = SET_TC_AT(skb->tc_verd, AT_INGRESS);
3930
	qdisc_bstats_cpu_update(cl->q, skb);
3931

3932
	switch (tc_classify(skb, cl, &cl_res, false)) {
3933 3934 3935 3936 3937
	case TC_ACT_OK:
	case TC_ACT_RECLASSIFY:
		skb->tc_index = TC_H_MIN(cl_res.classid);
		break;
	case TC_ACT_SHOT:
3938
		qdisc_qstats_cpu_drop(cl->q);
3939 3940 3941 3942
	case TC_ACT_STOLEN:
	case TC_ACT_QUEUED:
		kfree_skb(skb);
		return NULL;
3943 3944 3945 3946 3947 3948 3949 3950
	case TC_ACT_REDIRECT:
		/* skb_mac_header check was done by cls/act_bpf, so
		 * we can safely push the L2 header back before
		 * redirecting to another netdev
		 */
		__skb_push(skb, skb->mac_len);
		skb_do_redirect(skb);
		return NULL;
3951 3952
	default:
		break;
3953
	}
3954
#endif /* CONFIG_NET_CLS_ACT */
3955 3956
	return skb;
}
L
Linus Torvalds 已提交
3957

3958 3959 3960 3961
/**
 *	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
3962
 *	@rx_handler_data: data pointer that is used by rx handler
3963
 *
3964
 *	Register a receive handler for a device. This handler will then be
3965 3966 3967 3968
 *	called from __netif_receive_skb. A negative errno code is returned
 *	on a failure.
 *
 *	The caller must hold the rtnl_mutex.
3969 3970
 *
 *	For a general description of rx_handler, see enum rx_handler_result.
3971 3972
 */
int netdev_rx_handler_register(struct net_device *dev,
J
Jiri Pirko 已提交
3973 3974
			       rx_handler_func_t *rx_handler,
			       void *rx_handler_data)
3975 3976 3977 3978 3979 3980
{
	ASSERT_RTNL();

	if (dev->rx_handler)
		return -EBUSY;

3981
	/* Note: rx_handler_data must be set before rx_handler */
J
Jiri Pirko 已提交
3982
	rcu_assign_pointer(dev->rx_handler_data, rx_handler_data);
3983 3984 3985 3986 3987 3988 3989 3990 3991 3992
	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
 *
K
Kusanagi Kouichi 已提交
3993
 *	Unregister a receive handler from a device.
3994 3995 3996 3997 3998 3999 4000
 *
 *	The caller must hold the rtnl_mutex.
 */
void netdev_rx_handler_unregister(struct net_device *dev)
{

	ASSERT_RTNL();
4001
	RCU_INIT_POINTER(dev->rx_handler, NULL);
4002 4003 4004 4005 4006
	/* a reader seeing a non NULL rx_handler in a rcu_read_lock()
	 * section has a guarantee to see a non NULL rx_handler_data
	 * as well.
	 */
	synchronize_net();
4007
	RCU_INIT_POINTER(dev->rx_handler_data, NULL);
4008 4009 4010
}
EXPORT_SYMBOL_GPL(netdev_rx_handler_unregister);

4011 4012 4013 4014 4015 4016 4017
/*
 * Limit the use of PFMEMALLOC reserves to those protocols that implement
 * the special handling of PFMEMALLOC skbs.
 */
static bool skb_pfmemalloc_protocol(struct sk_buff *skb)
{
	switch (skb->protocol) {
4018 4019 4020 4021 4022
	case htons(ETH_P_ARP):
	case htons(ETH_P_IP):
	case htons(ETH_P_IPV6):
	case htons(ETH_P_8021Q):
	case htons(ETH_P_8021AD):
4023 4024 4025 4026 4027 4028
		return true;
	default:
		return false;
	}
}

4029 4030 4031
static inline int nf_ingress(struct sk_buff *skb, struct packet_type **pt_prev,
			     int *ret, struct net_device *orig_dev)
{
4032
#ifdef CONFIG_NETFILTER_INGRESS
4033 4034 4035 4036 4037 4038 4039 4040
	if (nf_hook_ingress_active(skb)) {
		if (*pt_prev) {
			*ret = deliver_skb(skb, *pt_prev, orig_dev);
			*pt_prev = NULL;
		}

		return nf_hook_ingress(skb);
	}
4041
#endif /* CONFIG_NETFILTER_INGRESS */
4042 4043 4044
	return 0;
}

4045
static int __netif_receive_skb_core(struct sk_buff *skb, bool pfmemalloc)
L
Linus Torvalds 已提交
4046 4047
{
	struct packet_type *ptype, *pt_prev;
4048
	rx_handler_func_t *rx_handler;
D
David S. Miller 已提交
4049
	struct net_device *orig_dev;
4050
	bool deliver_exact = false;
L
Linus Torvalds 已提交
4051
	int ret = NET_RX_DROP;
A
Al Viro 已提交
4052
	__be16 type;
L
Linus Torvalds 已提交
4053

4054
	net_timestamp_check(!netdev_tstamp_prequeue, skb);
4055

4056
	trace_netif_receive_skb(skb);
4057

J
Joe Eykholt 已提交
4058
	orig_dev = skb->dev;
4059

4060
	skb_reset_network_header(skb);
4061 4062
	if (!skb_transport_header_was_set(skb))
		skb_reset_transport_header(skb);
4063
	skb_reset_mac_len(skb);
L
Linus Torvalds 已提交
4064 4065 4066

	pt_prev = NULL;

4067
another_round:
4068
	skb->skb_iif = skb->dev->ifindex;
4069 4070 4071

	__this_cpu_inc(softnet_data.processed);

P
Patrick McHardy 已提交
4072 4073
	if (skb->protocol == cpu_to_be16(ETH_P_8021Q) ||
	    skb->protocol == cpu_to_be16(ETH_P_8021AD)) {
4074
		skb = skb_vlan_untag(skb);
4075
		if (unlikely(!skb))
4076
			goto out;
4077 4078
	}

L
Linus Torvalds 已提交
4079 4080 4081 4082 4083 4084 4085
#ifdef CONFIG_NET_CLS_ACT
	if (skb->tc_verd & TC_NCLS) {
		skb->tc_verd = CLR_TC_NCLS(skb->tc_verd);
		goto ncls;
	}
#endif

4086
	if (pfmemalloc)
4087 4088
		goto skip_taps;

L
Linus Torvalds 已提交
4089
	list_for_each_entry_rcu(ptype, &ptype_all, list) {
4090 4091 4092 4093 4094 4095 4096 4097 4098
		if (pt_prev)
			ret = deliver_skb(skb, pt_prev, orig_dev);
		pt_prev = ptype;
	}

	list_for_each_entry_rcu(ptype, &skb->dev->ptype_all, list) {
		if (pt_prev)
			ret = deliver_skb(skb, pt_prev, orig_dev);
		pt_prev = ptype;
L
Linus Torvalds 已提交
4099 4100
	}

4101
skip_taps:
4102
#ifdef CONFIG_NET_INGRESS
4103
	if (static_key_false(&ingress_needed)) {
D
Daniel Borkmann 已提交
4104
		skb = sch_handle_ingress(skb, &pt_prev, &ret, orig_dev);
4105
		if (!skb)
4106
			goto out;
4107 4108

		if (nf_ingress(skb, &pt_prev, &ret, orig_dev) < 0)
4109
			goto out;
4110
	}
4111 4112
#endif
#ifdef CONFIG_NET_CLS_ACT
4113
	skb->tc_verd = 0;
L
Linus Torvalds 已提交
4114 4115
ncls:
#endif
4116
	if (pfmemalloc && !skb_pfmemalloc_protocol(skb))
4117 4118
		goto drop;

4119
	if (skb_vlan_tag_present(skb)) {
4120 4121 4122 4123
		if (pt_prev) {
			ret = deliver_skb(skb, pt_prev, orig_dev);
			pt_prev = NULL;
		}
4124
		if (vlan_do_receive(&skb))
4125 4126
			goto another_round;
		else if (unlikely(!skb))
4127
			goto out;
4128 4129
	}

4130
	rx_handler = rcu_dereference(skb->dev->rx_handler);
4131 4132 4133 4134 4135
	if (rx_handler) {
		if (pt_prev) {
			ret = deliver_skb(skb, pt_prev, orig_dev);
			pt_prev = NULL;
		}
4136 4137
		switch (rx_handler(&skb)) {
		case RX_HANDLER_CONSUMED:
4138
			ret = NET_RX_SUCCESS;
4139
			goto out;
4140
		case RX_HANDLER_ANOTHER:
4141
			goto another_round;
4142 4143 4144 4145 4146 4147 4148
		case RX_HANDLER_EXACT:
			deliver_exact = true;
		case RX_HANDLER_PASS:
			break;
		default:
			BUG();
		}
4149
	}
L
Linus Torvalds 已提交
4150

4151 4152
	if (unlikely(skb_vlan_tag_present(skb))) {
		if (skb_vlan_tag_get_id(skb))
E
Eric Dumazet 已提交
4153 4154 4155 4156 4157 4158 4159
			skb->pkt_type = PACKET_OTHERHOST;
		/* Note: we might in the future use prio bits
		 * and set skb->priority like in vlan_do_receive()
		 * For the time being, just ignore Priority Code Point
		 */
		skb->vlan_tci = 0;
	}
4160

4161 4162
	type = skb->protocol;

4163
	/* deliver only exact match when indicated */
4164 4165 4166 4167 4168
	if (likely(!deliver_exact)) {
		deliver_ptype_list_skb(skb, &pt_prev, orig_dev, type,
				       &ptype_base[ntohs(type) &
						   PTYPE_HASH_MASK]);
	}
4169

4170 4171 4172 4173 4174 4175
	deliver_ptype_list_skb(skb, &pt_prev, orig_dev, type,
			       &orig_dev->ptype_specific);

	if (unlikely(skb->dev != orig_dev)) {
		deliver_ptype_list_skb(skb, &pt_prev, orig_dev, type,
				       &skb->dev->ptype_specific);
L
Linus Torvalds 已提交
4176 4177 4178
	}

	if (pt_prev) {
4179
		if (unlikely(skb_orphan_frags(skb, GFP_ATOMIC)))
4180
			goto drop;
4181 4182
		else
			ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
L
Linus Torvalds 已提交
4183
	} else {
4184
drop:
4185 4186 4187 4188
		if (!deliver_exact)
			atomic_long_inc(&skb->dev->rx_dropped);
		else
			atomic_long_inc(&skb->dev->rx_nohandler);
L
Linus Torvalds 已提交
4189 4190 4191 4192 4193 4194 4195
		kfree_skb(skb);
		/* Jamal, now you will not able to escape explaining
		 * me how you were going to use this. :-)
		 */
		ret = NET_RX_DROP;
	}

4196
out:
4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221
	return ret;
}

static int __netif_receive_skb(struct sk_buff *skb)
{
	int ret;

	if (sk_memalloc_socks() && skb_pfmemalloc(skb)) {
		unsigned long pflags = current->flags;

		/*
		 * PFMEMALLOC skbs are special, they should
		 * - be delivered to SOCK_MEMALLOC sockets only
		 * - stay away from userspace
		 * - have bounded memory usage
		 *
		 * Use PF_MEMALLOC as this saves us from propagating the allocation
		 * context down to all allocation sites.
		 */
		current->flags |= PF_MEMALLOC;
		ret = __netif_receive_skb_core(skb, true);
		tsk_restore_flags(current, pflags, PF_MEMALLOC);
	} else
		ret = __netif_receive_skb_core(skb, false);

L
Linus Torvalds 已提交
4222 4223
	return ret;
}
T
Tom Herbert 已提交
4224

4225
static int netif_receive_skb_internal(struct sk_buff *skb)
T
Tom Herbert 已提交
4226
{
4227 4228
	int ret;

4229
	net_timestamp_check(netdev_tstamp_prequeue, skb);
E
Eric Dumazet 已提交
4230

4231 4232 4233
	if (skb_defer_rx_timestamp(skb))
		return NET_RX_SUCCESS;

4234 4235
	rcu_read_lock();

E
Eric Dumazet 已提交
4236
#ifdef CONFIG_RPS
4237
	if (static_key_false(&rps_needed)) {
E
Eric Dumazet 已提交
4238
		struct rps_dev_flow voidflow, *rflow = &voidflow;
4239
		int cpu = get_rps_cpu(skb->dev, skb, &rflow);
T
Tom Herbert 已提交
4240

E
Eric Dumazet 已提交
4241 4242 4243
		if (cpu >= 0) {
			ret = enqueue_to_backlog(skb, cpu, &rflow->last_qtail);
			rcu_read_unlock();
4244
			return ret;
E
Eric Dumazet 已提交
4245
		}
T
Tom Herbert 已提交
4246
	}
4247
#endif
4248 4249 4250
	ret = __netif_receive_skb(skb);
	rcu_read_unlock();
	return ret;
T
Tom Herbert 已提交
4251
}
4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267

/**
 *	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
 */
4268
int netif_receive_skb(struct sk_buff *skb)
4269 4270 4271 4272 4273
{
	trace_netif_receive_skb_entry(skb);

	return netif_receive_skb_internal(skb);
}
4274
EXPORT_SYMBOL(netif_receive_skb);
L
Linus Torvalds 已提交
4275

E
Eric Dumazet 已提交
4276 4277 4278
/* Network device is going away, flush any packets still pending
 * Called with irqs disabled.
 */
4279
static void flush_backlog(void *arg)
4280
{
4281
	struct net_device *dev = arg;
4282
	struct softnet_data *sd = this_cpu_ptr(&softnet_data);
4283 4284
	struct sk_buff *skb, *tmp;

E
Eric Dumazet 已提交
4285
	rps_lock(sd);
4286
	skb_queue_walk_safe(&sd->input_pkt_queue, skb, tmp) {
4287
		if (skb->dev == dev) {
E
Eric Dumazet 已提交
4288
			__skb_unlink(skb, &sd->input_pkt_queue);
4289
			kfree_skb(skb);
4290
			input_queue_head_incr(sd);
4291
		}
4292
	}
E
Eric Dumazet 已提交
4293
	rps_unlock(sd);
4294 4295 4296 4297 4298

	skb_queue_walk_safe(&sd->process_queue, skb, tmp) {
		if (skb->dev == dev) {
			__skb_unlink(skb, &sd->process_queue);
			kfree_skb(skb);
4299
			input_queue_head_incr(sd);
4300 4301
		}
	}
4302 4303
}

4304 4305
static int napi_gro_complete(struct sk_buff *skb)
{
4306
	struct packet_offload *ptype;
4307
	__be16 type = skb->protocol;
4308
	struct list_head *head = &offload_base;
4309 4310
	int err = -ENOENT;

4311 4312
	BUILD_BUG_ON(sizeof(struct napi_gro_cb) > sizeof(skb->cb));

4313 4314
	if (NAPI_GRO_CB(skb)->count == 1) {
		skb_shinfo(skb)->gso_size = 0;
4315
		goto out;
4316
	}
4317 4318 4319

	rcu_read_lock();
	list_for_each_entry_rcu(ptype, head, list) {
4320
		if (ptype->type != type || !ptype->callbacks.gro_complete)
4321 4322
			continue;

4323
		err = ptype->callbacks.gro_complete(skb, 0);
4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334
		break;
	}
	rcu_read_unlock();

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

out:
4335
	return netif_receive_skb_internal(skb);
4336 4337
}

4338 4339 4340 4341 4342
/* napi->gro_list contains packets ordered by age.
 * youngest packets at the head of it.
 * Complete skbs in reverse order to reduce latencies.
 */
void napi_gro_flush(struct napi_struct *napi, bool flush_old)
4343
{
4344
	struct sk_buff *skb, *prev = NULL;
4345

4346 4347 4348 4349 4350 4351 4352
	/* scan list and build reverse chain */
	for (skb = napi->gro_list; skb != NULL; skb = skb->next) {
		skb->prev = prev;
		prev = skb;
	}

	for (skb = prev; skb; skb = prev) {
4353
		skb->next = NULL;
4354 4355 4356 4357 4358

		if (flush_old && NAPI_GRO_CB(skb)->age == jiffies)
			return;

		prev = skb->prev;
4359
		napi_gro_complete(skb);
4360
		napi->gro_count--;
4361 4362 4363 4364
	}

	napi->gro_list = NULL;
}
E
Eric Dumazet 已提交
4365
EXPORT_SYMBOL(napi_gro_flush);
4366

E
Eric Dumazet 已提交
4367 4368 4369 4370
static void gro_list_prepare(struct napi_struct *napi, struct sk_buff *skb)
{
	struct sk_buff *p;
	unsigned int maclen = skb->dev->hard_header_len;
4371
	u32 hash = skb_get_hash_raw(skb);
E
Eric Dumazet 已提交
4372 4373 4374 4375

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

4376 4377 4378 4379 4380 4381 4382
		NAPI_GRO_CB(p)->flush = 0;

		if (hash != skb_get_hash_raw(p)) {
			NAPI_GRO_CB(p)->same_flow = 0;
			continue;
		}

E
Eric Dumazet 已提交
4383 4384
		diffs = (unsigned long)p->dev ^ (unsigned long)skb->dev;
		diffs |= p->vlan_tci ^ skb->vlan_tci;
4385
		diffs |= skb_metadata_dst_cmp(p, skb);
E
Eric Dumazet 已提交
4386 4387
		if (maclen == ETH_HLEN)
			diffs |= compare_ether_header(skb_mac_header(p),
4388
						      skb_mac_header(skb));
E
Eric Dumazet 已提交
4389 4390
		else if (!diffs)
			diffs = memcmp(skb_mac_header(p),
4391
				       skb_mac_header(skb),
E
Eric Dumazet 已提交
4392 4393 4394 4395 4396
				       maclen);
		NAPI_GRO_CB(p)->same_flow = !diffs;
	}
}

4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410
static void skb_gro_reset_offset(struct sk_buff *skb)
{
	const struct skb_shared_info *pinfo = skb_shinfo(skb);
	const skb_frag_t *frag0 = &pinfo->frags[0];

	NAPI_GRO_CB(skb)->data_offset = 0;
	NAPI_GRO_CB(skb)->frag0 = NULL;
	NAPI_GRO_CB(skb)->frag0_len = 0;

	if (skb_mac_header(skb) == skb_tail_pointer(skb) &&
	    pinfo->nr_frags &&
	    !PageHighMem(skb_frag_page(frag0))) {
		NAPI_GRO_CB(skb)->frag0 = skb_frag_address(frag0);
		NAPI_GRO_CB(skb)->frag0_len = skb_frag_size(frag0);
E
Eric Dumazet 已提交
4411 4412 4413
	}
}

4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434
static void gro_pull_from_frag0(struct sk_buff *skb, int grow)
{
	struct skb_shared_info *pinfo = skb_shinfo(skb);

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

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

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

	pinfo->frags[0].page_offset += grow;
	skb_frag_size_sub(&pinfo->frags[0], grow);

	if (unlikely(!skb_frag_size(&pinfo->frags[0]))) {
		skb_frag_unref(skb, 0);
		memmove(pinfo->frags, pinfo->frags + 1,
			--pinfo->nr_frags * sizeof(pinfo->frags[0]));
	}
}

R
Rami Rosen 已提交
4435
static enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
4436 4437
{
	struct sk_buff **pp = NULL;
4438
	struct packet_offload *ptype;
4439
	__be16 type = skb->protocol;
4440
	struct list_head *head = &offload_base;
H
Herbert Xu 已提交
4441
	int same_flow;
4442
	enum gro_result ret;
4443
	int grow;
4444

4445
	if (!(skb->dev->features & NETIF_F_GRO))
4446 4447
		goto normal;

4448
	if (skb_is_gso(skb) || skb_has_frag_list(skb) || skb->csum_bad)
4449 4450
		goto normal;

E
Eric Dumazet 已提交
4451 4452
	gro_list_prepare(napi, skb);

4453 4454
	rcu_read_lock();
	list_for_each_entry_rcu(ptype, head, list) {
4455
		if (ptype->type != type || !ptype->callbacks.gro_receive)
4456 4457
			continue;

4458
		skb_set_network_header(skb, skb_gro_offset(skb));
4459
		skb_reset_mac_len(skb);
4460 4461
		NAPI_GRO_CB(skb)->same_flow = 0;
		NAPI_GRO_CB(skb)->flush = 0;
H
Herbert Xu 已提交
4462
		NAPI_GRO_CB(skb)->free = 0;
4463
		NAPI_GRO_CB(skb)->encap_mark = 0;
4464
		NAPI_GRO_CB(skb)->is_fou = 0;
4465
		NAPI_GRO_CB(skb)->is_atomic = 1;
4466
		NAPI_GRO_CB(skb)->gro_remcsum_start = 0;
4467

4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482
		/* Setup for GRO checksum validation */
		switch (skb->ip_summed) {
		case CHECKSUM_COMPLETE:
			NAPI_GRO_CB(skb)->csum = skb->csum;
			NAPI_GRO_CB(skb)->csum_valid = 1;
			NAPI_GRO_CB(skb)->csum_cnt = 0;
			break;
		case CHECKSUM_UNNECESSARY:
			NAPI_GRO_CB(skb)->csum_cnt = skb->csum_level + 1;
			NAPI_GRO_CB(skb)->csum_valid = 0;
			break;
		default:
			NAPI_GRO_CB(skb)->csum_cnt = 0;
			NAPI_GRO_CB(skb)->csum_valid = 0;
		}
4483

4484
		pp = ptype->callbacks.gro_receive(&napi->gro_list, skb);
4485 4486 4487 4488 4489 4490 4491
		break;
	}
	rcu_read_unlock();

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

H
Herbert Xu 已提交
4492
	same_flow = NAPI_GRO_CB(skb)->same_flow;
4493
	ret = NAPI_GRO_CB(skb)->free ? GRO_MERGED_FREE : GRO_MERGED;
H
Herbert Xu 已提交
4494

4495 4496 4497 4498 4499 4500
	if (pp) {
		struct sk_buff *nskb = *pp;

		*pp = nskb->next;
		nskb->next = NULL;
		napi_gro_complete(nskb);
4501
		napi->gro_count--;
4502 4503
	}

H
Herbert Xu 已提交
4504
	if (same_flow)
4505 4506
		goto ok;

4507
	if (NAPI_GRO_CB(skb)->flush)
4508 4509
		goto normal;

4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523
	if (unlikely(napi->gro_count >= MAX_GRO_SKBS)) {
		struct sk_buff *nskb = napi->gro_list;

		/* locate the end of the list to select the 'oldest' flow */
		while (nskb->next) {
			pp = &nskb->next;
			nskb = *pp;
		}
		*pp = NULL;
		nskb->next = NULL;
		napi_gro_complete(nskb);
	} else {
		napi->gro_count++;
	}
4524
	NAPI_GRO_CB(skb)->count = 1;
4525
	NAPI_GRO_CB(skb)->age = jiffies;
4526
	NAPI_GRO_CB(skb)->last = skb;
4527
	skb_shinfo(skb)->gso_size = skb_gro_len(skb);
4528 4529
	skb->next = napi->gro_list;
	napi->gro_list = skb;
4530
	ret = GRO_HELD;
4531

4532
pull:
4533 4534 4535
	grow = skb_gro_offset(skb) - skb_headlen(skb);
	if (grow > 0)
		gro_pull_from_frag0(skb, grow);
4536
ok:
4537
	return ret;
4538 4539

normal:
4540 4541
	ret = GRO_NORMAL;
	goto pull;
H
Herbert Xu 已提交
4542
}
4543

4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555
struct packet_offload *gro_find_receive_by_type(__be16 type)
{
	struct list_head *offload_head = &offload_base;
	struct packet_offload *ptype;

	list_for_each_entry_rcu(ptype, offload_head, list) {
		if (ptype->type != type || !ptype->callbacks.gro_receive)
			continue;
		return ptype;
	}
	return NULL;
}
4556
EXPORT_SYMBOL(gro_find_receive_by_type);
4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569

struct packet_offload *gro_find_complete_by_type(__be16 type)
{
	struct list_head *offload_head = &offload_base;
	struct packet_offload *ptype;

	list_for_each_entry_rcu(ptype, offload_head, list) {
		if (ptype->type != type || !ptype->callbacks.gro_complete)
			continue;
		return ptype;
	}
	return NULL;
}
4570
EXPORT_SYMBOL(gro_find_complete_by_type);
H
Herbert Xu 已提交
4571

R
Rami Rosen 已提交
4572
static gro_result_t napi_skb_finish(gro_result_t ret, struct sk_buff *skb)
H
Herbert Xu 已提交
4573
{
4574 4575
	switch (ret) {
	case GRO_NORMAL:
4576
		if (netif_receive_skb_internal(skb))
4577 4578
			ret = GRO_DROP;
		break;
H
Herbert Xu 已提交
4579

4580
	case GRO_DROP:
H
Herbert Xu 已提交
4581 4582
		kfree_skb(skb);
		break;
4583

4584
	case GRO_MERGED_FREE:
4585 4586
		if (NAPI_GRO_CB(skb)->free == NAPI_GRO_FREE_STOLEN_HEAD) {
			skb_dst_drop(skb);
4587
			kmem_cache_free(skbuff_head_cache, skb);
4588
		} else {
4589
			__kfree_skb(skb);
4590
		}
4591 4592
		break;

4593 4594 4595
	case GRO_HELD:
	case GRO_MERGED:
		break;
H
Herbert Xu 已提交
4596 4597
	}

4598
	return ret;
4599 4600
}

4601
gro_result_t napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
4602
{
4603
	skb_mark_napi_id(skb, napi);
4604
	trace_napi_gro_receive_entry(skb);
4605

4606 4607
	skb_gro_reset_offset(skb);

E
Eric Dumazet 已提交
4608
	return napi_skb_finish(dev_gro_receive(napi, skb), skb);
4609 4610 4611
}
EXPORT_SYMBOL(napi_gro_receive);

S
stephen hemminger 已提交
4612
static void napi_reuse_skb(struct napi_struct *napi, struct sk_buff *skb)
4613
{
4614 4615 4616 4617
	if (unlikely(skb->pfmemalloc)) {
		consume_skb(skb);
		return;
	}
4618
	__skb_pull(skb, skb_headlen(skb));
4619 4620
	/* restore the reserve we had after netdev_alloc_skb_ip_align() */
	skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN - skb_headroom(skb));
4621
	skb->vlan_tci = 0;
H
Herbert Xu 已提交
4622
	skb->dev = napi->dev;
A
Andy Gospodarek 已提交
4623
	skb->skb_iif = 0;
4624 4625
	skb->encapsulation = 0;
	skb_shinfo(skb)->gso_type = 0;
4626
	skb->truesize = SKB_TRUESIZE(skb_end_offset(skb));
4627 4628 4629 4630

	napi->skb = skb;
}

4631
struct sk_buff *napi_get_frags(struct napi_struct *napi)
H
Herbert Xu 已提交
4632 4633 4634 4635
{
	struct sk_buff *skb = napi->skb;

	if (!skb) {
4636
		skb = napi_alloc_skb(napi, GRO_MAX_HEAD);
4637 4638 4639 4640
		if (skb) {
			napi->skb = skb;
			skb_mark_napi_id(skb, napi);
		}
4641
	}
4642 4643
	return skb;
}
4644
EXPORT_SYMBOL(napi_get_frags);
4645

4646 4647 4648
static gro_result_t napi_frags_finish(struct napi_struct *napi,
				      struct sk_buff *skb,
				      gro_result_t ret)
4649
{
4650 4651
	switch (ret) {
	case GRO_NORMAL:
4652 4653 4654 4655
	case GRO_HELD:
		__skb_push(skb, ETH_HLEN);
		skb->protocol = eth_type_trans(skb, skb->dev);
		if (ret == GRO_NORMAL && netif_receive_skb_internal(skb))
4656
			ret = GRO_DROP;
4657
		break;
H
Herbert Xu 已提交
4658

4659 4660 4661 4662
	case GRO_DROP:
	case GRO_MERGED_FREE:
		napi_reuse_skb(napi, skb);
		break;
4663 4664 4665

	case GRO_MERGED:
		break;
4666
	}
H
Herbert Xu 已提交
4667

4668
	return ret;
H
Herbert Xu 已提交
4669
}
4670

4671 4672 4673 4674
/* Upper GRO stack assumes network header starts at gro_offset=0
 * Drivers could call both napi_gro_frags() and napi_gro_receive()
 * We copy ethernet header into skb->data to have a common layout.
 */
E
Eric Dumazet 已提交
4675
static struct sk_buff *napi_frags_skb(struct napi_struct *napi)
4676 4677
{
	struct sk_buff *skb = napi->skb;
4678 4679
	const struct ethhdr *eth;
	unsigned int hlen = sizeof(*eth);
4680 4681 4682

	napi->skb = NULL;

4683 4684 4685 4686 4687 4688 4689
	skb_reset_mac_header(skb);
	skb_gro_reset_offset(skb);

	eth = skb_gro_header_fast(skb, 0);
	if (unlikely(skb_gro_header_hard(skb, hlen))) {
		eth = skb_gro_header_slow(skb, hlen, 0);
		if (unlikely(!eth)) {
4690 4691
			net_warn_ratelimited("%s: dropping impossible skb from %s\n",
					     __func__, napi->dev->name);
4692 4693 4694 4695 4696 4697 4698
			napi_reuse_skb(napi, skb);
			return NULL;
		}
	} else {
		gro_pull_from_frag0(skb, hlen);
		NAPI_GRO_CB(skb)->frag0 += hlen;
		NAPI_GRO_CB(skb)->frag0_len -= hlen;
4699
	}
4700 4701 4702 4703 4704 4705 4706 4707
	__skb_pull(skb, hlen);

	/*
	 * This works because the only protocols we care about don't require
	 * special handling.
	 * We'll fix it up properly in napi_frags_finish()
	 */
	skb->protocol = eth->h_proto;
4708 4709 4710 4711

	return skb;
}

4712
gro_result_t napi_gro_frags(struct napi_struct *napi)
4713
{
4714
	struct sk_buff *skb = napi_frags_skb(napi);
4715 4716

	if (!skb)
4717
		return GRO_DROP;
4718

4719 4720
	trace_napi_gro_frags_entry(skb);

E
Eric Dumazet 已提交
4721
	return napi_frags_finish(napi, skb, dev_gro_receive(napi, skb));
4722
}
H
Herbert Xu 已提交
4723 4724
EXPORT_SYMBOL(napi_gro_frags);

T
Tom Herbert 已提交
4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746 4747 4748 4749
/* Compute the checksum from gro_offset and return the folded value
 * after adding in any pseudo checksum.
 */
__sum16 __skb_gro_checksum_complete(struct sk_buff *skb)
{
	__wsum wsum;
	__sum16 sum;

	wsum = skb_checksum(skb, skb_gro_offset(skb), skb_gro_len(skb), 0);

	/* NAPI_GRO_CB(skb)->csum holds pseudo checksum */
	sum = csum_fold(csum_add(NAPI_GRO_CB(skb)->csum, wsum));
	if (likely(!sum)) {
		if (unlikely(skb->ip_summed == CHECKSUM_COMPLETE) &&
		    !skb->csum_complete_sw)
			netdev_rx_csum_fault(skb->dev);
	}

	NAPI_GRO_CB(skb)->csum = wsum;
	NAPI_GRO_CB(skb)->csum_valid = 1;

	return sum;
}
EXPORT_SYMBOL(__skb_gro_checksum_complete);

4750
/*
4751
 * net_rps_action_and_irq_enable sends any pending IPI's for rps.
4752 4753 4754 4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768
 * 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))
4769
				smp_call_function_single_async(remsd->cpu,
4770
							   &remsd->csd);
4771 4772 4773 4774 4775 4776 4777
			remsd = next;
		}
	} else
#endif
		local_irq_enable();
}

E
Eric Dumazet 已提交
4778 4779 4780 4781 4782 4783 4784 4785 4786
static bool sd_has_rps_ipi_waiting(struct softnet_data *sd)
{
#ifdef CONFIG_RPS
	return sd->rps_ipi_list != NULL;
#else
	return false;
#endif
}

4787
static int process_backlog(struct napi_struct *napi, int quota)
L
Linus Torvalds 已提交
4788 4789
{
	int work = 0;
E
Eric Dumazet 已提交
4790
	struct softnet_data *sd = container_of(napi, struct softnet_data, backlog);
L
Linus Torvalds 已提交
4791

4792 4793 4794
	/* Check if we have pending ipi, its better to send them now,
	 * not waiting net_rx_action() end.
	 */
E
Eric Dumazet 已提交
4795
	if (sd_has_rps_ipi_waiting(sd)) {
4796 4797 4798
		local_irq_disable();
		net_rps_action_and_irq_enable(sd);
	}
E
Eric Dumazet 已提交
4799

4800
	napi->weight = weight_p;
4801
	local_irq_disable();
4802
	while (1) {
L
Linus Torvalds 已提交
4803
		struct sk_buff *skb;
4804 4805

		while ((skb = __skb_dequeue(&sd->process_queue))) {
4806
			rcu_read_lock();
4807 4808
			local_irq_enable();
			__netif_receive_skb(skb);
4809
			rcu_read_unlock();
4810
			local_irq_disable();
4811 4812 4813 4814 4815
			input_queue_head_incr(sd);
			if (++work >= quota) {
				local_irq_enable();
				return work;
			}
4816
		}
L
Linus Torvalds 已提交
4817

E
Eric Dumazet 已提交
4818
		rps_lock(sd);
4819
		if (skb_queue_empty(&sd->input_pkt_queue)) {
E
Eric Dumazet 已提交
4820 4821 4822
			/*
			 * Inline a custom version of __napi_complete().
			 * only current cpu owns and manipulates this napi,
4823 4824 4825
			 * and NAPI_STATE_SCHED is the only possible flag set
			 * on backlog.
			 * We can use a plain write instead of clear_bit(),
E
Eric Dumazet 已提交
4826 4827 4828
			 * and we dont need an smp_mb() memory barrier.
			 */
			napi->state = 0;
4829
			rps_unlock(sd);
E
Eric Dumazet 已提交
4830

4831
			break;
4832
		}
4833 4834 4835

		skb_queue_splice_tail_init(&sd->input_pkt_queue,
					   &sd->process_queue);
E
Eric Dumazet 已提交
4836
		rps_unlock(sd);
4837 4838
	}
	local_irq_enable();
L
Linus Torvalds 已提交
4839

4840 4841
	return work;
}
L
Linus Torvalds 已提交
4842

4843 4844
/**
 * __napi_schedule - schedule for receive
4845
 * @n: entry to schedule
4846
 *
4847 4848
 * The entry's receive function will be scheduled to run.
 * Consider using __napi_schedule_irqoff() if hard irqs are masked.
4849
 */
H
Harvey Harrison 已提交
4850
void __napi_schedule(struct napi_struct *n)
4851 4852
{
	unsigned long flags;
L
Linus Torvalds 已提交
4853

4854
	local_irq_save(flags);
4855
	____napi_schedule(this_cpu_ptr(&softnet_data), n);
4856
	local_irq_restore(flags);
L
Linus Torvalds 已提交
4857
}
4858 4859
EXPORT_SYMBOL(__napi_schedule);

4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871
/**
 * __napi_schedule_irqoff - schedule for receive
 * @n: entry to schedule
 *
 * Variant of __napi_schedule() assuming hard irqs are masked
 */
void __napi_schedule_irqoff(struct napi_struct *n)
{
	____napi_schedule(this_cpu_ptr(&softnet_data), n);
}
EXPORT_SYMBOL(__napi_schedule_irqoff);

4872 4873 4874 4875
void __napi_complete(struct napi_struct *n)
{
	BUG_ON(!test_bit(NAPI_STATE_SCHED, &n->state));

E
Eric Dumazet 已提交
4876
	list_del_init(&n->poll_list);
4877
	smp_mb__before_atomic();
4878 4879 4880 4881
	clear_bit(NAPI_STATE_SCHED, &n->state);
}
EXPORT_SYMBOL(__napi_complete);

4882
void napi_complete_done(struct napi_struct *n, int work_done)
4883 4884 4885 4886 4887 4888 4889 4890 4891 4892
{
	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;

4893 4894
	if (n->gro_list) {
		unsigned long timeout = 0;
E
Eric Dumazet 已提交
4895

4896 4897 4898 4899 4900 4901 4902 4903 4904
		if (work_done)
			timeout = n->dev->gro_flush_timeout;

		if (timeout)
			hrtimer_start(&n->timer, ns_to_ktime(timeout),
				      HRTIMER_MODE_REL_PINNED);
		else
			napi_gro_flush(n, false);
	}
E
Eric Dumazet 已提交
4905 4906 4907 4908 4909 4910 4911 4912
	if (likely(list_empty(&n->poll_list))) {
		WARN_ON_ONCE(!test_and_clear_bit(NAPI_STATE_SCHED, &n->state));
	} else {
		/* If n->poll_list is not empty, we need to mask irqs */
		local_irq_save(flags);
		__napi_complete(n);
		local_irq_restore(flags);
	}
4913
}
4914
EXPORT_SYMBOL(napi_complete_done);
4915

E
Eliezer Tamir 已提交
4916
/* must be called under rcu_read_lock(), as we dont take a reference */
E
Eric Dumazet 已提交
4917
static struct napi_struct *napi_by_id(unsigned int napi_id)
E
Eliezer Tamir 已提交
4918 4919 4920 4921 4922 4923 4924 4925 4926 4927
{
	unsigned int hash = napi_id % HASH_SIZE(napi_hash);
	struct napi_struct *napi;

	hlist_for_each_entry_rcu(napi, &napi_hash[hash], napi_hash_node)
		if (napi->napi_id == napi_id)
			return napi;

	return NULL;
}
E
Eric Dumazet 已提交
4928 4929

#if defined(CONFIG_NET_RX_BUSY_POLL)
4930
#define BUSY_POLL_BUDGET 8
E
Eric Dumazet 已提交
4931 4932 4933
bool sk_busy_loop(struct sock *sk, int nonblock)
{
	unsigned long end_time = !nonblock ? sk_busy_loop_end_time(sk) : 0;
4934
	int (*busy_poll)(struct napi_struct *dev);
E
Eric Dumazet 已提交
4935 4936 4937
	struct napi_struct *napi;
	int rc = false;

4938
	rcu_read_lock();
E
Eric Dumazet 已提交
4939 4940 4941 4942 4943

	napi = napi_by_id(sk->sk_napi_id);
	if (!napi)
		goto out;

4944 4945
	/* Note: ndo_busy_poll method is optional in linux-4.5 */
	busy_poll = napi->dev->netdev_ops->ndo_busy_poll;
E
Eric Dumazet 已提交
4946 4947

	do {
4948
		rc = 0;
4949
		local_bh_disable();
4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964
		if (busy_poll) {
			rc = busy_poll(napi);
		} else if (napi_schedule_prep(napi)) {
			void *have = netpoll_poll_lock(napi);

			if (test_bit(NAPI_STATE_SCHED, &napi->state)) {
				rc = napi->poll(napi, BUSY_POLL_BUDGET);
				trace_napi_poll(napi);
				if (rc == BUSY_POLL_BUDGET) {
					napi_complete_done(napi, rc);
					napi_schedule(napi);
				}
			}
			netpoll_poll_unlock(have);
		}
4965 4966 4967 4968
		if (rc > 0)
			NET_ADD_STATS_BH(sock_net(sk),
					 LINUX_MIB_BUSYPOLLRXPACKETS, rc);
		local_bh_enable();
E
Eric Dumazet 已提交
4969 4970 4971 4972 4973 4974 4975 4976 4977 4978

		if (rc == LL_FLUSH_FAILED)
			break; /* permanent failure */

		cpu_relax();
	} while (!nonblock && skb_queue_empty(&sk->sk_receive_queue) &&
		 !need_resched() && !busy_loop_timeout(end_time));

	rc = !skb_queue_empty(&sk->sk_receive_queue);
out:
4979
	rcu_read_unlock();
E
Eric Dumazet 已提交
4980 4981 4982 4983 4984
	return rc;
}
EXPORT_SYMBOL(sk_busy_loop);

#endif /* CONFIG_NET_RX_BUSY_POLL */
E
Eliezer Tamir 已提交
4985 4986 4987

void napi_hash_add(struct napi_struct *napi)
{
E
Eric Dumazet 已提交
4988 4989
	if (test_bit(NAPI_STATE_NO_BUSY_POLL, &napi->state) ||
	    test_and_set_bit(NAPI_STATE_HASHED, &napi->state))
4990
		return;
E
Eliezer Tamir 已提交
4991

4992
	spin_lock(&napi_hash_lock);
E
Eliezer Tamir 已提交
4993

4994 4995 4996 4997 4998 4999
	/* 0..NR_CPUS+1 range is reserved for sender_cpu use */
	do {
		if (unlikely(++napi_gen_id < NR_CPUS + 1))
			napi_gen_id = NR_CPUS + 1;
	} while (napi_by_id(napi_gen_id));
	napi->napi_id = napi_gen_id;
E
Eliezer Tamir 已提交
5000

5001 5002
	hlist_add_head_rcu(&napi->napi_hash_node,
			   &napi_hash[napi->napi_id % HASH_SIZE(napi_hash)]);
E
Eliezer Tamir 已提交
5003

5004
	spin_unlock(&napi_hash_lock);
E
Eliezer Tamir 已提交
5005 5006 5007 5008 5009 5010
}
EXPORT_SYMBOL_GPL(napi_hash_add);

/* Warning : caller is responsible to make sure rcu grace period
 * is respected before freeing memory containing @napi
 */
5011
bool napi_hash_del(struct napi_struct *napi)
E
Eliezer Tamir 已提交
5012
{
5013 5014
	bool rcu_sync_needed = false;

E
Eliezer Tamir 已提交
5015 5016
	spin_lock(&napi_hash_lock);

5017 5018
	if (test_and_clear_bit(NAPI_STATE_HASHED, &napi->state)) {
		rcu_sync_needed = true;
E
Eliezer Tamir 已提交
5019
		hlist_del_rcu(&napi->napi_hash_node);
5020
	}
E
Eliezer Tamir 已提交
5021
	spin_unlock(&napi_hash_lock);
5022
	return rcu_sync_needed;
E
Eliezer Tamir 已提交
5023 5024 5025
}
EXPORT_SYMBOL_GPL(napi_hash_del);

5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036
static enum hrtimer_restart napi_watchdog(struct hrtimer *timer)
{
	struct napi_struct *napi;

	napi = container_of(timer, struct napi_struct, timer);
	if (napi->gro_list)
		napi_schedule(napi);

	return HRTIMER_NORESTART;
}

5037 5038 5039 5040
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);
5041 5042
	hrtimer_init(&napi->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL_PINNED);
	napi->timer.function = napi_watchdog;
5043
	napi->gro_count = 0;
5044
	napi->gro_list = NULL;
H
Herbert Xu 已提交
5045
	napi->skb = NULL;
5046
	napi->poll = poll;
E
Eric Dumazet 已提交
5047 5048 5049
	if (weight > NAPI_POLL_WEIGHT)
		pr_err_once("netif_napi_add() called with weight %d on device %s\n",
			    weight, dev->name);
5050 5051 5052
	napi->weight = weight;
	list_add(&napi->dev_list, &dev->napi_list);
	napi->dev = dev;
H
Herbert Xu 已提交
5053
#ifdef CONFIG_NETPOLL
5054 5055 5056 5057
	spin_lock_init(&napi->poll_lock);
	napi->poll_owner = -1;
#endif
	set_bit(NAPI_STATE_SCHED, &napi->state);
5058
	napi_hash_add(napi);
5059 5060 5061
}
EXPORT_SYMBOL(netif_napi_add);

5062 5063 5064 5065 5066 5067 5068
void napi_disable(struct napi_struct *n)
{
	might_sleep();
	set_bit(NAPI_STATE_DISABLE, &n->state);

	while (test_and_set_bit(NAPI_STATE_SCHED, &n->state))
		msleep(1);
5069 5070
	while (test_and_set_bit(NAPI_STATE_NPSVC, &n->state))
		msleep(1);
5071 5072 5073 5074 5075 5076 5077

	hrtimer_cancel(&n->timer);

	clear_bit(NAPI_STATE_DISABLE, &n->state);
}
EXPORT_SYMBOL(napi_disable);

5078
/* Must be called in process context */
5079 5080
void netif_napi_del(struct napi_struct *napi)
{
5081 5082 5083
	might_sleep();
	if (napi_hash_del(napi))
		synchronize_net();
5084
	list_del_init(&napi->dev_list);
5085
	napi_free_frags(napi);
5086

E
Eric Dumazet 已提交
5087
	kfree_skb_list(napi->gro_list);
5088
	napi->gro_list = NULL;
5089
	napi->gro_count = 0;
5090 5091 5092
}
EXPORT_SYMBOL(netif_napi_del);

5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118 5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137
static int napi_poll(struct napi_struct *n, struct list_head *repoll)
{
	void *have;
	int work, weight;

	list_del_init(&n->poll_list);

	have = netpoll_poll_lock(n);

	weight = n->weight;

	/* 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
	 * accidentally calling ->poll() when NAPI is not scheduled.
	 */
	work = 0;
	if (test_bit(NAPI_STATE_SCHED, &n->state)) {
		work = n->poll(n, weight);
		trace_napi_poll(n);
	}

	WARN_ON_ONCE(work > weight);

	if (likely(work < weight))
		goto out_unlock;

	/* 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.
	 */
	if (unlikely(napi_disable_pending(n))) {
		napi_complete(n);
		goto out_unlock;
	}

	if (n->gro_list) {
		/* flush too old packets
		 * If HZ < 1000, flush all packets.
		 */
		napi_gro_flush(n, HZ >= 1000);
	}

5138 5139 5140 5141 5142 5143 5144 5145 5146
	/* Some drivers may have called napi_schedule
	 * prior to exhausting their budget.
	 */
	if (unlikely(!list_empty(&n->poll_list))) {
		pr_warn_once("%s: Budget exhausted after napi rescheduled\n",
			     n->dev ? n->dev->name : "backlog");
		goto out_unlock;
	}

5147 5148 5149 5150 5151 5152 5153 5154
	list_add_tail(&n->poll_list, repoll);

out_unlock:
	netpoll_poll_unlock(have);

	return work;
}

L
Linus Torvalds 已提交
5155 5156
static void net_rx_action(struct softirq_action *h)
{
5157
	struct softnet_data *sd = this_cpu_ptr(&softnet_data);
5158
	unsigned long time_limit = jiffies + 2;
5159
	int budget = netdev_budget;
E
Eric Dumazet 已提交
5160 5161
	LIST_HEAD(list);
	LIST_HEAD(repoll);
5162

L
Linus Torvalds 已提交
5163
	local_irq_disable();
E
Eric Dumazet 已提交
5164 5165
	list_splice_init(&sd->poll_list, &list);
	local_irq_enable();
L
Linus Torvalds 已提交
5166

H
Herbert Xu 已提交
5167
	for (;;) {
5168
		struct napi_struct *n;
L
Linus Torvalds 已提交
5169

H
Herbert Xu 已提交
5170 5171 5172 5173 5174 5175
		if (list_empty(&list)) {
			if (!sd_has_rps_ipi_waiting(sd) && list_empty(&repoll))
				return;
			break;
		}

5176 5177 5178
		n = list_first_entry(&list, struct napi_struct, poll_list);
		budget -= napi_poll(n, &repoll);

E
Eric Dumazet 已提交
5179
		/* If softirq window is exhausted then punt.
5180 5181
		 * Allow this to run for 2 jiffies since which will allow
		 * an average latency of 1.5/HZ.
5182
		 */
H
Herbert Xu 已提交
5183 5184 5185 5186 5187
		if (unlikely(budget <= 0 ||
			     time_after_eq(jiffies, time_limit))) {
			sd->time_squeeze++;
			break;
		}
L
Linus Torvalds 已提交
5188
	}
E
Eric Dumazet 已提交
5189

5190
	__kfree_skb_flush();
E
Eric Dumazet 已提交
5191 5192 5193 5194 5195 5196 5197 5198
	local_irq_disable();

	list_splice_tail_init(&sd->poll_list, &list);
	list_splice_tail(&repoll, &list);
	list_splice(&list, &sd->poll_list);
	if (!list_empty(&sd->poll_list))
		__raise_softirq_irqoff(NET_RX_SOFTIRQ);

5199
	net_rps_action_and_irq_enable(sd);
L
Linus Torvalds 已提交
5200 5201
}

5202
struct netdev_adjacent {
J
Jiri Pirko 已提交
5203
	struct net_device *dev;
5204 5205

	/* upper master flag, there can only be one master device per list */
J
Jiri Pirko 已提交
5206
	bool master;
5207 5208 5209 5210

	/* counter for the number of times this device was added to us */
	u16 ref_nr;

5211 5212 5213
	/* private field for the users */
	void *private;

J
Jiri Pirko 已提交
5214 5215 5216 5217
	struct list_head list;
	struct rcu_head rcu;
};

5218
static struct netdev_adjacent *__netdev_find_adj(struct net_device *adj_dev,
5219
						 struct list_head *adj_list)
J
Jiri Pirko 已提交
5220
{
5221 5222
	struct netdev_adjacent *adj;

5223
	list_for_each_entry(adj, adj_list, list) {
5224 5225
		if (adj->dev == adj_dev)
			return adj;
J
Jiri Pirko 已提交
5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243
	}
	return NULL;
}

/**
 * netdev_has_upper_dev - Check if device is linked to an upper device
 * @dev: device
 * @upper_dev: upper device to check
 *
 * Find out if a device is linked to specified upper device and return true
 * in case it is. Note that this checks only immediate upper device,
 * not through a complete stack of devices. The caller must hold the RTNL lock.
 */
bool netdev_has_upper_dev(struct net_device *dev,
			  struct net_device *upper_dev)
{
	ASSERT_RTNL();

5244
	return __netdev_find_adj(upper_dev, &dev->all_adj_list.upper);
J
Jiri Pirko 已提交
5245 5246 5247 5248 5249 5250 5251 5252 5253 5254
}
EXPORT_SYMBOL(netdev_has_upper_dev);

/**
 * netdev_has_any_upper_dev - Check if device is linked to some device
 * @dev: device
 *
 * Find out if a device is linked to an upper device and return true in case
 * it is. The caller must hold the RTNL lock.
 */
S
stephen hemminger 已提交
5255
static bool netdev_has_any_upper_dev(struct net_device *dev)
J
Jiri Pirko 已提交
5256 5257 5258
{
	ASSERT_RTNL();

5259
	return !list_empty(&dev->all_adj_list.upper);
J
Jiri Pirko 已提交
5260 5261 5262 5263 5264 5265 5266 5267 5268 5269 5270
}

/**
 * netdev_master_upper_dev_get - Get master upper device
 * @dev: device
 *
 * Find a master upper device and return pointer to it or NULL in case
 * it's not there. The caller must hold the RTNL lock.
 */
struct net_device *netdev_master_upper_dev_get(struct net_device *dev)
{
5271
	struct netdev_adjacent *upper;
J
Jiri Pirko 已提交
5272 5273 5274

	ASSERT_RTNL();

5275
	if (list_empty(&dev->adj_list.upper))
J
Jiri Pirko 已提交
5276 5277
		return NULL;

5278
	upper = list_first_entry(&dev->adj_list.upper,
5279
				 struct netdev_adjacent, list);
J
Jiri Pirko 已提交
5280 5281 5282 5283 5284 5285
	if (likely(upper->master))
		return upper->dev;
	return NULL;
}
EXPORT_SYMBOL(netdev_master_upper_dev_get);

5286 5287 5288 5289 5290 5291 5292 5293 5294 5295
void *netdev_adjacent_get_private(struct list_head *adj_list)
{
	struct netdev_adjacent *adj;

	adj = list_entry(adj_list, struct netdev_adjacent, list);

	return adj->private;
}
EXPORT_SYMBOL(netdev_adjacent_get_private);

5296 5297 5298 5299 5300 5301 5302 5303 5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320 5321
/**
 * netdev_upper_get_next_dev_rcu - Get the next dev from upper list
 * @dev: device
 * @iter: list_head ** of the current position
 *
 * Gets the next device from the dev's upper list, starting from iter
 * position. The caller must hold RCU read lock.
 */
struct net_device *netdev_upper_get_next_dev_rcu(struct net_device *dev,
						 struct list_head **iter)
{
	struct netdev_adjacent *upper;

	WARN_ON_ONCE(!rcu_read_lock_held() && !lockdep_rtnl_is_held());

	upper = list_entry_rcu((*iter)->next, struct netdev_adjacent, list);

	if (&upper->list == &dev->adj_list.upper)
		return NULL;

	*iter = &upper->list;

	return upper->dev;
}
EXPORT_SYMBOL(netdev_upper_get_next_dev_rcu);

5322 5323
/**
 * netdev_all_upper_get_next_dev_rcu - Get the next dev from upper list
5324 5325 5326 5327 5328 5329
 * @dev: device
 * @iter: list_head ** of the current position
 *
 * Gets the next device from the dev's upper list, starting from iter
 * position. The caller must hold RCU read lock.
 */
5330 5331
struct net_device *netdev_all_upper_get_next_dev_rcu(struct net_device *dev,
						     struct list_head **iter)
5332 5333 5334
{
	struct netdev_adjacent *upper;

5335
	WARN_ON_ONCE(!rcu_read_lock_held() && !lockdep_rtnl_is_held());
5336 5337 5338

	upper = list_entry_rcu((*iter)->next, struct netdev_adjacent, list);

5339
	if (&upper->list == &dev->all_adj_list.upper)
5340 5341 5342 5343 5344 5345
		return NULL;

	*iter = &upper->list;

	return upper->dev;
}
5346
EXPORT_SYMBOL(netdev_all_upper_get_next_dev_rcu);
5347

5348 5349 5350 5351 5352 5353 5354 5355 5356
/**
 * netdev_lower_get_next_private - Get the next ->private from the
 *				   lower neighbour list
 * @dev: device
 * @iter: list_head ** of the current position
 *
 * Gets the next netdev_adjacent->private from the dev's lower neighbour
 * list, starting from iter position. The caller must hold either hold the
 * RTNL lock or its own locking that guarantees that the neighbour lower
5357
 * list will remain unchanged.
5358 5359 5360 5361 5362 5363 5364 5365 5366 5367 5368
 */
void *netdev_lower_get_next_private(struct net_device *dev,
				    struct list_head **iter)
{
	struct netdev_adjacent *lower;

	lower = list_entry(*iter, struct netdev_adjacent, list);

	if (&lower->list == &dev->adj_list.lower)
		return NULL;

5369
	*iter = lower->list.next;
5370 5371 5372 5373 5374 5375 5376 5377 5378 5379 5380 5381 5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392 5393 5394 5395 5396

	return lower->private;
}
EXPORT_SYMBOL(netdev_lower_get_next_private);

/**
 * netdev_lower_get_next_private_rcu - Get the next ->private from the
 *				       lower neighbour list, RCU
 *				       variant
 * @dev: device
 * @iter: list_head ** of the current position
 *
 * Gets the next netdev_adjacent->private from the dev's lower neighbour
 * list, starting from iter position. The caller must hold RCU read lock.
 */
void *netdev_lower_get_next_private_rcu(struct net_device *dev,
					struct list_head **iter)
{
	struct netdev_adjacent *lower;

	WARN_ON_ONCE(!rcu_read_lock_held());

	lower = list_entry_rcu((*iter)->next, struct netdev_adjacent, list);

	if (&lower->list == &dev->adj_list.lower)
		return NULL;

5397
	*iter = &lower->list;
5398 5399 5400 5401 5402

	return lower->private;
}
EXPORT_SYMBOL(netdev_lower_get_next_private_rcu);

5403 5404 5405 5406 5407 5408 5409 5410 5411
/**
 * netdev_lower_get_next - Get the next device from the lower neighbour
 *                         list
 * @dev: device
 * @iter: list_head ** of the current position
 *
 * Gets the next netdev_adjacent from the dev's lower neighbour
 * list, starting from iter position. The caller must hold RTNL lock or
 * its own locking that guarantees that the neighbour lower
5412
 * list will remain unchanged.
5413 5414 5415 5416 5417
 */
void *netdev_lower_get_next(struct net_device *dev, struct list_head **iter)
{
	struct netdev_adjacent *lower;

5418
	lower = list_entry(*iter, struct netdev_adjacent, list);
5419 5420 5421 5422

	if (&lower->list == &dev->adj_list.lower)
		return NULL;

5423
	*iter = lower->list.next;
5424 5425 5426 5427 5428

	return lower->dev;
}
EXPORT_SYMBOL(netdev_lower_get_next);

5429 5430 5431 5432 5433 5434 5435 5436 5437 5438 5439 5440 5441 5442 5443 5444 5445 5446 5447 5448 5449
/**
 * netdev_lower_get_first_private_rcu - Get the first ->private from the
 *				       lower neighbour list, RCU
 *				       variant
 * @dev: device
 *
 * Gets the first netdev_adjacent->private from the dev's lower neighbour
 * list. The caller must hold RCU read lock.
 */
void *netdev_lower_get_first_private_rcu(struct net_device *dev)
{
	struct netdev_adjacent *lower;

	lower = list_first_or_null_rcu(&dev->adj_list.lower,
			struct netdev_adjacent, list);
	if (lower)
		return lower->private;
	return NULL;
}
EXPORT_SYMBOL(netdev_lower_get_first_private_rcu);

J
Jiri Pirko 已提交
5450 5451 5452 5453 5454 5455 5456 5457 5458
/**
 * netdev_master_upper_dev_get_rcu - Get master upper device
 * @dev: device
 *
 * Find a master upper device and return pointer to it or NULL in case
 * it's not there. The caller must hold the RCU read lock.
 */
struct net_device *netdev_master_upper_dev_get_rcu(struct net_device *dev)
{
5459
	struct netdev_adjacent *upper;
J
Jiri Pirko 已提交
5460

5461
	upper = list_first_or_null_rcu(&dev->adj_list.upper,
5462
				       struct netdev_adjacent, list);
J
Jiri Pirko 已提交
5463 5464 5465 5466 5467 5468
	if (upper && likely(upper->master))
		return upper->dev;
	return NULL;
}
EXPORT_SYMBOL(netdev_master_upper_dev_get_rcu);

5469
static int netdev_adjacent_sysfs_add(struct net_device *dev,
5470 5471 5472 5473 5474 5475 5476 5477 5478
			      struct net_device *adj_dev,
			      struct list_head *dev_list)
{
	char linkname[IFNAMSIZ+7];
	sprintf(linkname, dev_list == &dev->adj_list.upper ?
		"upper_%s" : "lower_%s", adj_dev->name);
	return sysfs_create_link(&(dev->dev.kobj), &(adj_dev->dev.kobj),
				 linkname);
}
5479
static void netdev_adjacent_sysfs_del(struct net_device *dev,
5480 5481 5482 5483 5484 5485 5486 5487 5488
			       char *name,
			       struct list_head *dev_list)
{
	char linkname[IFNAMSIZ+7];
	sprintf(linkname, dev_list == &dev->adj_list.upper ?
		"upper_%s" : "lower_%s", name);
	sysfs_remove_link(&(dev->dev.kobj), linkname);
}

5489 5490 5491 5492 5493 5494 5495 5496
static inline bool netdev_adjacent_is_neigh_list(struct net_device *dev,
						 struct net_device *adj_dev,
						 struct list_head *dev_list)
{
	return (dev_list == &dev->adj_list.upper ||
		dev_list == &dev->adj_list.lower) &&
		net_eq(dev_net(dev), dev_net(adj_dev));
}
5497

5498 5499
static int __netdev_adjacent_dev_insert(struct net_device *dev,
					struct net_device *adj_dev,
5500
					struct list_head *dev_list,
5501
					void *private, bool master)
5502 5503
{
	struct netdev_adjacent *adj;
5504
	int ret;
5505

5506
	adj = __netdev_find_adj(adj_dev, dev_list);
5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 5519

	if (adj) {
		adj->ref_nr++;
		return 0;
	}

	adj = kmalloc(sizeof(*adj), GFP_KERNEL);
	if (!adj)
		return -ENOMEM;

	adj->dev = adj_dev;
	adj->master = master;
	adj->ref_nr = 1;
5520
	adj->private = private;
5521
	dev_hold(adj_dev);
5522 5523 5524

	pr_debug("dev_hold for %s, because of link added from %s to %s\n",
		 adj_dev->name, dev->name, adj_dev->name);
5525

5526
	if (netdev_adjacent_is_neigh_list(dev, adj_dev, dev_list)) {
5527
		ret = netdev_adjacent_sysfs_add(dev, adj_dev, dev_list);
5528 5529 5530 5531
		if (ret)
			goto free_adj;
	}

5532
	/* Ensure that master link is always the first item in list. */
5533 5534 5535 5536
	if (master) {
		ret = sysfs_create_link(&(dev->dev.kobj),
					&(adj_dev->dev.kobj), "master");
		if (ret)
5537
			goto remove_symlinks;
5538

5539
		list_add_rcu(&adj->list, dev_list);
5540
	} else {
5541
		list_add_tail_rcu(&adj->list, dev_list);
5542
	}
5543 5544

	return 0;
5545

5546
remove_symlinks:
5547
	if (netdev_adjacent_is_neigh_list(dev, adj_dev, dev_list))
5548
		netdev_adjacent_sysfs_del(dev, adj_dev->name, dev_list);
5549 5550
free_adj:
	kfree(adj);
5551
	dev_put(adj_dev);
5552 5553

	return ret;
5554 5555
}

S
stephen hemminger 已提交
5556 5557 5558
static void __netdev_adjacent_dev_remove(struct net_device *dev,
					 struct net_device *adj_dev,
					 struct list_head *dev_list)
5559 5560 5561
{
	struct netdev_adjacent *adj;

5562
	adj = __netdev_find_adj(adj_dev, dev_list);
5563

5564 5565 5566
	if (!adj) {
		pr_err("tried to remove device %s from %s\n",
		       dev->name, adj_dev->name);
5567
		BUG();
5568
	}
5569 5570

	if (adj->ref_nr > 1) {
5571 5572
		pr_debug("%s to %s ref_nr-- = %d\n", dev->name, adj_dev->name,
			 adj->ref_nr-1);
5573 5574 5575 5576
		adj->ref_nr--;
		return;
	}

5577 5578 5579
	if (adj->master)
		sysfs_remove_link(&(dev->dev.kobj), "master");

5580
	if (netdev_adjacent_is_neigh_list(dev, adj_dev, dev_list))
5581
		netdev_adjacent_sysfs_del(dev, adj_dev->name, dev_list);
5582

5583
	list_del_rcu(&adj->list);
5584 5585
	pr_debug("dev_put for %s, because link removed from %s to %s\n",
		 adj_dev->name, dev->name, adj_dev->name);
5586 5587 5588 5589
	dev_put(adj_dev);
	kfree_rcu(adj, rcu);
}

S
stephen hemminger 已提交
5590 5591 5592 5593 5594
static int __netdev_adjacent_dev_link_lists(struct net_device *dev,
					    struct net_device *upper_dev,
					    struct list_head *up_list,
					    struct list_head *down_list,
					    void *private, bool master)
5595 5596 5597
{
	int ret;

5598 5599
	ret = __netdev_adjacent_dev_insert(dev, upper_dev, up_list, private,
					   master);
5600 5601 5602
	if (ret)
		return ret;

5603 5604
	ret = __netdev_adjacent_dev_insert(upper_dev, dev, down_list, private,
					   false);
5605
	if (ret) {
5606
		__netdev_adjacent_dev_remove(dev, upper_dev, up_list);
5607 5608 5609 5610 5611 5612
		return ret;
	}

	return 0;
}

S
stephen hemminger 已提交
5613 5614
static int __netdev_adjacent_dev_link(struct net_device *dev,
				      struct net_device *upper_dev)
5615
{
5616 5617 5618
	return __netdev_adjacent_dev_link_lists(dev, upper_dev,
						&dev->all_adj_list.upper,
						&upper_dev->all_adj_list.lower,
5619
						NULL, false);
5620 5621
}

S
stephen hemminger 已提交
5622 5623 5624 5625
static void __netdev_adjacent_dev_unlink_lists(struct net_device *dev,
					       struct net_device *upper_dev,
					       struct list_head *up_list,
					       struct list_head *down_list)
5626
{
5627 5628
	__netdev_adjacent_dev_remove(dev, upper_dev, up_list);
	__netdev_adjacent_dev_remove(upper_dev, dev, down_list);
5629 5630
}

S
stephen hemminger 已提交
5631 5632
static void __netdev_adjacent_dev_unlink(struct net_device *dev,
					 struct net_device *upper_dev)
5633
{
5634 5635 5636 5637 5638
	__netdev_adjacent_dev_unlink_lists(dev, upper_dev,
					   &dev->all_adj_list.upper,
					   &upper_dev->all_adj_list.lower);
}

S
stephen hemminger 已提交
5639 5640 5641
static int __netdev_adjacent_dev_link_neighbour(struct net_device *dev,
						struct net_device *upper_dev,
						void *private, bool master)
5642 5643 5644 5645 5646 5647 5648 5649 5650
{
	int ret = __netdev_adjacent_dev_link(dev, upper_dev);

	if (ret)
		return ret;

	ret = __netdev_adjacent_dev_link_lists(dev, upper_dev,
					       &dev->adj_list.upper,
					       &upper_dev->adj_list.lower,
5651
					       private, master);
5652 5653 5654 5655 5656 5657
	if (ret) {
		__netdev_adjacent_dev_unlink(dev, upper_dev);
		return ret;
	}

	return 0;
5658 5659
}

S
stephen hemminger 已提交
5660 5661
static void __netdev_adjacent_dev_unlink_neighbour(struct net_device *dev,
						   struct net_device *upper_dev)
5662 5663 5664 5665 5666 5667
{
	__netdev_adjacent_dev_unlink(dev, upper_dev);
	__netdev_adjacent_dev_unlink_lists(dev, upper_dev,
					   &dev->adj_list.upper,
					   &upper_dev->adj_list.lower);
}
5668

J
Jiri Pirko 已提交
5669
static int __netdev_upper_dev_link(struct net_device *dev,
5670
				   struct net_device *upper_dev, bool master,
5671
				   void *upper_priv, void *upper_info)
J
Jiri Pirko 已提交
5672
{
5673
	struct netdev_notifier_changeupper_info changeupper_info;
5674 5675
	struct netdev_adjacent *i, *j, *to_i, *to_j;
	int ret = 0;
J
Jiri Pirko 已提交
5676 5677 5678 5679 5680 5681 5682

	ASSERT_RTNL();

	if (dev == upper_dev)
		return -EBUSY;

	/* To prevent loops, check if dev is not upper device to upper_dev. */
5683
	if (__netdev_find_adj(dev, &upper_dev->all_adj_list.upper))
J
Jiri Pirko 已提交
5684 5685
		return -EBUSY;

5686
	if (__netdev_find_adj(upper_dev, &dev->adj_list.upper))
J
Jiri Pirko 已提交
5687 5688 5689 5690 5691
		return -EEXIST;

	if (master && netdev_master_upper_dev_get(dev))
		return -EBUSY;

5692 5693 5694
	changeupper_info.upper_dev = upper_dev;
	changeupper_info.master = master;
	changeupper_info.linking = true;
5695
	changeupper_info.upper_info = upper_info;
5696

5697 5698 5699 5700 5701 5702
	ret = call_netdevice_notifiers_info(NETDEV_PRECHANGEUPPER, dev,
					    &changeupper_info.info);
	ret = notifier_to_errno(ret);
	if (ret)
		return ret;

5703
	ret = __netdev_adjacent_dev_link_neighbour(dev, upper_dev, upper_priv,
5704
						   master);
5705 5706
	if (ret)
		return ret;
J
Jiri Pirko 已提交
5707

5708
	/* Now that we linked these devs, make all the upper_dev's
5709
	 * all_adj_list.upper visible to every dev's all_adj_list.lower an
5710 5711 5712
	 * versa, and don't forget the devices itself. All of these
	 * links are non-neighbours.
	 */
5713 5714 5715 5716
	list_for_each_entry(i, &dev->all_adj_list.lower, list) {
		list_for_each_entry(j, &upper_dev->all_adj_list.upper, list) {
			pr_debug("Interlinking %s with %s, non-neighbour\n",
				 i->dev->name, j->dev->name);
5717 5718 5719 5720 5721 5722 5723
			ret = __netdev_adjacent_dev_link(i->dev, j->dev);
			if (ret)
				goto rollback_mesh;
		}
	}

	/* add dev to every upper_dev's upper device */
5724 5725 5726
	list_for_each_entry(i, &upper_dev->all_adj_list.upper, list) {
		pr_debug("linking %s's upper device %s with %s\n",
			 upper_dev->name, i->dev->name, dev->name);
5727 5728 5729 5730 5731 5732
		ret = __netdev_adjacent_dev_link(dev, i->dev);
		if (ret)
			goto rollback_upper_mesh;
	}

	/* add upper_dev to every dev's lower device */
5733 5734 5735
	list_for_each_entry(i, &dev->all_adj_list.lower, list) {
		pr_debug("linking %s's lower device %s with %s\n", dev->name,
			 i->dev->name, upper_dev->name);
5736 5737 5738 5739
		ret = __netdev_adjacent_dev_link(i->dev, upper_dev);
		if (ret)
			goto rollback_lower_mesh;
	}
J
Jiri Pirko 已提交
5740

5741 5742 5743 5744 5745 5746
	ret = call_netdevice_notifiers_info(NETDEV_CHANGEUPPER, dev,
					    &changeupper_info.info);
	ret = notifier_to_errno(ret);
	if (ret)
		goto rollback_lower_mesh;

J
Jiri Pirko 已提交
5747
	return 0;
5748 5749 5750

rollback_lower_mesh:
	to_i = i;
5751
	list_for_each_entry(i, &dev->all_adj_list.lower, list) {
5752 5753 5754 5755 5756 5757 5758 5759 5760
		if (i == to_i)
			break;
		__netdev_adjacent_dev_unlink(i->dev, upper_dev);
	}

	i = NULL;

rollback_upper_mesh:
	to_i = i;
5761
	list_for_each_entry(i, &upper_dev->all_adj_list.upper, list) {
5762 5763 5764 5765 5766 5767 5768 5769 5770 5771
		if (i == to_i)
			break;
		__netdev_adjacent_dev_unlink(dev, i->dev);
	}

	i = j = NULL;

rollback_mesh:
	to_i = i;
	to_j = j;
5772 5773
	list_for_each_entry(i, &dev->all_adj_list.lower, list) {
		list_for_each_entry(j, &upper_dev->all_adj_list.upper, list) {
5774 5775 5776 5777 5778 5779 5780 5781
			if (i == to_i && j == to_j)
				break;
			__netdev_adjacent_dev_unlink(i->dev, j->dev);
		}
		if (i == to_i)
			break;
	}

5782
	__netdev_adjacent_dev_unlink_neighbour(dev, upper_dev);
5783 5784

	return ret;
J
Jiri Pirko 已提交
5785 5786 5787 5788 5789 5790 5791 5792 5793 5794 5795 5796 5797 5798 5799
}

/**
 * netdev_upper_dev_link - Add a link to the upper device
 * @dev: device
 * @upper_dev: new upper device
 *
 * Adds a link to device which is upper to this one. The caller must hold
 * the RTNL lock. On a failure a negative errno code is returned.
 * On success the reference counts are adjusted and the function
 * returns zero.
 */
int netdev_upper_dev_link(struct net_device *dev,
			  struct net_device *upper_dev)
{
5800
	return __netdev_upper_dev_link(dev, upper_dev, false, NULL, NULL);
J
Jiri Pirko 已提交
5801 5802 5803 5804 5805 5806 5807
}
EXPORT_SYMBOL(netdev_upper_dev_link);

/**
 * netdev_master_upper_dev_link - Add a master link to the upper device
 * @dev: device
 * @upper_dev: new upper device
5808
 * @upper_priv: upper device private
5809
 * @upper_info: upper info to be passed down via notifier
J
Jiri Pirko 已提交
5810 5811 5812 5813 5814 5815 5816 5817
 *
 * Adds a link to device which is upper to this one. In this case, only
 * one master upper device can be linked, although other non-master devices
 * might be linked as well. The caller must hold the RTNL lock.
 * On a failure a negative errno code is returned. On success the reference
 * counts are adjusted and the function returns zero.
 */
int netdev_master_upper_dev_link(struct net_device *dev,
5818
				 struct net_device *upper_dev,
5819
				 void *upper_priv, void *upper_info)
J
Jiri Pirko 已提交
5820
{
5821 5822
	return __netdev_upper_dev_link(dev, upper_dev, true,
				       upper_priv, upper_info);
J
Jiri Pirko 已提交
5823 5824 5825 5826 5827 5828 5829 5830 5831 5832 5833 5834 5835 5836
}
EXPORT_SYMBOL(netdev_master_upper_dev_link);

/**
 * netdev_upper_dev_unlink - Removes a link to upper device
 * @dev: device
 * @upper_dev: new upper device
 *
 * Removes a link to device which is upper to this one. The caller must hold
 * the RTNL lock.
 */
void netdev_upper_dev_unlink(struct net_device *dev,
			     struct net_device *upper_dev)
{
5837
	struct netdev_notifier_changeupper_info changeupper_info;
5838
	struct netdev_adjacent *i, *j;
J
Jiri Pirko 已提交
5839 5840
	ASSERT_RTNL();

5841 5842 5843 5844
	changeupper_info.upper_dev = upper_dev;
	changeupper_info.master = netdev_master_upper_dev_get(dev) == upper_dev;
	changeupper_info.linking = false;

5845 5846 5847
	call_netdevice_notifiers_info(NETDEV_PRECHANGEUPPER, dev,
				      &changeupper_info.info);

5848
	__netdev_adjacent_dev_unlink_neighbour(dev, upper_dev);
5849 5850 5851 5852 5853

	/* Here is the tricky part. We must remove all dev's lower
	 * devices from all upper_dev's upper devices and vice
	 * versa, to maintain the graph relationship.
	 */
5854 5855
	list_for_each_entry(i, &dev->all_adj_list.lower, list)
		list_for_each_entry(j, &upper_dev->all_adj_list.upper, list)
5856 5857 5858 5859 5860
			__netdev_adjacent_dev_unlink(i->dev, j->dev);

	/* remove also the devices itself from lower/upper device
	 * list
	 */
5861
	list_for_each_entry(i, &dev->all_adj_list.lower, list)
5862 5863
		__netdev_adjacent_dev_unlink(i->dev, upper_dev);

5864
	list_for_each_entry(i, &upper_dev->all_adj_list.upper, list)
5865 5866
		__netdev_adjacent_dev_unlink(dev, i->dev);

5867 5868
	call_netdevice_notifiers_info(NETDEV_CHANGEUPPER, dev,
				      &changeupper_info.info);
J
Jiri Pirko 已提交
5869 5870 5871
}
EXPORT_SYMBOL(netdev_upper_dev_unlink);

5872 5873 5874
/**
 * netdev_bonding_info_change - Dispatch event about slave change
 * @dev: device
5875
 * @bonding_info: info to dispatch
5876 5877 5878 5879 5880 5881 5882 5883 5884 5885 5886 5887 5888 5889 5890 5891
 *
 * Send NETDEV_BONDING_INFO to netdev notifiers with info.
 * The caller must hold the RTNL lock.
 */
void netdev_bonding_info_change(struct net_device *dev,
				struct netdev_bonding_info *bonding_info)
{
	struct netdev_notifier_bonding_info	info;

	memcpy(&info.bonding_info, bonding_info,
	       sizeof(struct netdev_bonding_info));
	call_netdevice_notifiers_info(NETDEV_BONDING_INFO, dev,
				      &info.info);
}
EXPORT_SYMBOL(netdev_bonding_info_change);

E
Eric Dumazet 已提交
5892
static void netdev_adjacent_add_links(struct net_device *dev)
5893 5894 5895 5896 5897 5898 5899 5900 5901 5902 5903 5904 5905 5906 5907 5908 5909 5910 5911 5912 5913 5914 5915 5916
{
	struct netdev_adjacent *iter;

	struct net *net = dev_net(dev);

	list_for_each_entry(iter, &dev->adj_list.upper, list) {
		if (!net_eq(net,dev_net(iter->dev)))
			continue;
		netdev_adjacent_sysfs_add(iter->dev, dev,
					  &iter->dev->adj_list.lower);
		netdev_adjacent_sysfs_add(dev, iter->dev,
					  &dev->adj_list.upper);
	}

	list_for_each_entry(iter, &dev->adj_list.lower, list) {
		if (!net_eq(net,dev_net(iter->dev)))
			continue;
		netdev_adjacent_sysfs_add(iter->dev, dev,
					  &iter->dev->adj_list.upper);
		netdev_adjacent_sysfs_add(dev, iter->dev,
					  &dev->adj_list.lower);
	}
}

E
Eric Dumazet 已提交
5917
static void netdev_adjacent_del_links(struct net_device *dev)
5918 5919 5920 5921 5922 5923 5924 5925 5926 5927 5928 5929 5930 5931 5932 5933 5934 5935 5936 5937 5938 5939 5940 5941
{
	struct netdev_adjacent *iter;

	struct net *net = dev_net(dev);

	list_for_each_entry(iter, &dev->adj_list.upper, list) {
		if (!net_eq(net,dev_net(iter->dev)))
			continue;
		netdev_adjacent_sysfs_del(iter->dev, dev->name,
					  &iter->dev->adj_list.lower);
		netdev_adjacent_sysfs_del(dev, iter->dev->name,
					  &dev->adj_list.upper);
	}

	list_for_each_entry(iter, &dev->adj_list.lower, list) {
		if (!net_eq(net,dev_net(iter->dev)))
			continue;
		netdev_adjacent_sysfs_del(iter->dev, dev->name,
					  &iter->dev->adj_list.upper);
		netdev_adjacent_sysfs_del(dev, iter->dev->name,
					  &dev->adj_list.lower);
	}
}

5942
void netdev_adjacent_rename_links(struct net_device *dev, char *oldname)
5943
{
5944
	struct netdev_adjacent *iter;
5945

5946 5947
	struct net *net = dev_net(dev);

5948
	list_for_each_entry(iter, &dev->adj_list.upper, list) {
5949 5950
		if (!net_eq(net,dev_net(iter->dev)))
			continue;
5951 5952 5953 5954 5955
		netdev_adjacent_sysfs_del(iter->dev, oldname,
					  &iter->dev->adj_list.lower);
		netdev_adjacent_sysfs_add(iter->dev, dev,
					  &iter->dev->adj_list.lower);
	}
5956

5957
	list_for_each_entry(iter, &dev->adj_list.lower, list) {
5958 5959
		if (!net_eq(net,dev_net(iter->dev)))
			continue;
5960 5961 5962 5963 5964
		netdev_adjacent_sysfs_del(iter->dev, oldname,
					  &iter->dev->adj_list.upper);
		netdev_adjacent_sysfs_add(iter->dev, dev,
					  &iter->dev->adj_list.upper);
	}
5965 5966 5967 5968 5969 5970 5971 5972 5973
}

void *netdev_lower_dev_get_private(struct net_device *dev,
				   struct net_device *lower_dev)
{
	struct netdev_adjacent *lower;

	if (!lower_dev)
		return NULL;
5974
	lower = __netdev_find_adj(lower_dev, &dev->adj_list.lower);
5975 5976 5977 5978 5979 5980 5981
	if (!lower)
		return NULL;

	return lower->private;
}
EXPORT_SYMBOL(netdev_lower_dev_get_private);

5982 5983

int dev_get_nest_level(struct net_device *dev,
5984
		       bool (*type_check)(const struct net_device *dev))
5985 5986 5987 5988 5989 5990 5991 5992 5993 5994 5995 5996 5997 5998 5999 6000 6001 6002 6003 6004 6005
{
	struct net_device *lower = NULL;
	struct list_head *iter;
	int max_nest = -1;
	int nest;

	ASSERT_RTNL();

	netdev_for_each_lower_dev(dev, lower, iter) {
		nest = dev_get_nest_level(lower, type_check);
		if (max_nest < nest)
			max_nest = nest;
	}

	if (type_check(dev))
		max_nest++;

	return max_nest;
}
EXPORT_SYMBOL(dev_get_nest_level);

6006 6007 6008 6009 6010 6011 6012 6013 6014 6015 6016 6017 6018 6019 6020 6021 6022 6023 6024 6025
/**
 * netdev_lower_change - Dispatch event about lower device state change
 * @lower_dev: device
 * @lower_state_info: state to dispatch
 *
 * Send NETDEV_CHANGELOWERSTATE to netdev notifiers with info.
 * The caller must hold the RTNL lock.
 */
void netdev_lower_state_changed(struct net_device *lower_dev,
				void *lower_state_info)
{
	struct netdev_notifier_changelowerstate_info changelowerstate_info;

	ASSERT_RTNL();
	changelowerstate_info.lower_state_info = lower_state_info;
	call_netdevice_notifiers_info(NETDEV_CHANGELOWERSTATE, lower_dev,
				      &changelowerstate_info.info);
}
EXPORT_SYMBOL(netdev_lower_state_changed);

6026 6027
static void dev_change_rx_flags(struct net_device *dev, int flags)
{
6028 6029
	const struct net_device_ops *ops = dev->netdev_ops;

6030
	if (ops->ndo_change_rx_flags)
6031
		ops->ndo_change_rx_flags(dev, flags);
6032 6033
}

6034
static int __dev_set_promiscuity(struct net_device *dev, int inc, bool notify)
L
Linus Torvalds 已提交
6035
{
6036
	unsigned int old_flags = dev->flags;
6037 6038
	kuid_t uid;
	kgid_t gid;
L
Linus Torvalds 已提交
6039

6040 6041
	ASSERT_RTNL();

6042 6043 6044 6045 6046 6047 6048 6049 6050 6051 6052
	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;
6053 6054
			pr_warn("%s: promiscuity touches roof, set promiscuity failed. promiscuity feature of device might be broken.\n",
				dev->name);
6055 6056 6057
			return -EOVERFLOW;
		}
	}
6058
	if (dev->flags != old_flags) {
6059 6060 6061
		pr_info("device %s %s promiscuous mode\n",
			dev->name,
			dev->flags & IFF_PROMISC ? "entered" : "left");
6062 6063
		if (audit_enabled) {
			current_uid_gid(&uid, &gid);
6064 6065 6066 6067 6068
			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),
6069
				from_kuid(&init_user_ns, audit_get_loginuid(current)),
6070 6071
				from_kuid(&init_user_ns, uid),
				from_kgid(&init_user_ns, gid),
6072
				audit_get_sessionid(current));
6073
		}
6074

6075
		dev_change_rx_flags(dev, IFF_PROMISC);
L
Linus Torvalds 已提交
6076
	}
6077 6078
	if (notify)
		__dev_notify_flags(dev, old_flags, IFF_PROMISC);
6079
	return 0;
L
Linus Torvalds 已提交
6080 6081
}

6082 6083 6084 6085 6086 6087 6088 6089 6090
/**
 *	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.
6091
 *	Return 0 if successful or a negative errno code on error.
6092
 */
6093
int dev_set_promiscuity(struct net_device *dev, int inc)
6094
{
6095
	unsigned int old_flags = dev->flags;
6096
	int err;
6097

6098
	err = __dev_set_promiscuity(dev, inc, true);
6099
	if (err < 0)
6100
		return err;
6101 6102
	if (dev->flags != old_flags)
		dev_set_rx_mode(dev);
6103
	return err;
6104
}
E
Eric Dumazet 已提交
6105
EXPORT_SYMBOL(dev_set_promiscuity);
6106

6107
static int __dev_set_allmulti(struct net_device *dev, int inc, bool notify)
L
Linus Torvalds 已提交
6108
{
6109
	unsigned int old_flags = dev->flags, old_gflags = dev->gflags;
L
Linus Torvalds 已提交
6110

6111 6112
	ASSERT_RTNL();

L
Linus Torvalds 已提交
6113
	dev->flags |= IFF_ALLMULTI;
6114 6115 6116 6117 6118 6119 6120 6121 6122 6123
	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;
6124 6125
			pr_warn("%s: allmulti touches roof, set allmulti failed. allmulti feature of device might be broken.\n",
				dev->name);
6126 6127 6128
			return -EOVERFLOW;
		}
	}
6129
	if (dev->flags ^ old_flags) {
6130
		dev_change_rx_flags(dev, IFF_ALLMULTI);
6131
		dev_set_rx_mode(dev);
6132 6133 6134
		if (notify)
			__dev_notify_flags(dev, old_flags,
					   dev->gflags ^ old_gflags);
6135
	}
6136
	return 0;
6137
}
6138 6139 6140 6141 6142 6143 6144 6145 6146 6147 6148 6149 6150 6151 6152 6153 6154 6155

/**
 *	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.
 *	Return 0 if successful or a negative errno code on error.
 */

int dev_set_allmulti(struct net_device *dev, int inc)
{
	return __dev_set_allmulti(dev, inc, true);
}
E
Eric Dumazet 已提交
6156
EXPORT_SYMBOL(dev_set_allmulti);
6157 6158 6159 6160

/*
 *	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
6161
 *	filtering it is put in promiscuous mode while unicast addresses
6162 6163 6164 6165
 *	are present.
 */
void __dev_set_rx_mode(struct net_device *dev)
{
6166 6167
	const struct net_device_ops *ops = dev->netdev_ops;

6168 6169 6170 6171 6172
	/* dev_open will call this function so the list will stay sane. */
	if (!(dev->flags&IFF_UP))
		return;

	if (!netif_device_present(dev))
6173
		return;
6174

6175
	if (!(dev->priv_flags & IFF_UNICAST_FLT)) {
6176 6177 6178
		/* Unicast addresses changes may only happen under the rtnl,
		 * therefore calling __dev_set_promiscuity here is safe.
		 */
6179
		if (!netdev_uc_empty(dev) && !dev->uc_promisc) {
6180
			__dev_set_promiscuity(dev, 1, false);
6181
			dev->uc_promisc = true;
6182
		} else if (netdev_uc_empty(dev) && dev->uc_promisc) {
6183
			__dev_set_promiscuity(dev, -1, false);
6184
			dev->uc_promisc = false;
6185 6186
		}
	}
6187 6188 6189

	if (ops->ndo_set_rx_mode)
		ops->ndo_set_rx_mode(dev);
6190 6191 6192 6193
}

void dev_set_rx_mode(struct net_device *dev)
{
6194
	netif_addr_lock_bh(dev);
6195
	__dev_set_rx_mode(dev);
6196
	netif_addr_unlock_bh(dev);
L
Linus Torvalds 已提交
6197 6198
}

6199 6200 6201 6202 6203 6204
/**
 *	dev_get_flags - get flags reported to userspace
 *	@dev: device
 *
 *	Get the combination of flag bits exported through APIs to userspace.
 */
6205
unsigned int dev_get_flags(const struct net_device *dev)
L
Linus Torvalds 已提交
6206
{
6207
	unsigned int flags;
L
Linus Torvalds 已提交
6208 6209 6210

	flags = (dev->flags & ~(IFF_PROMISC |
				IFF_ALLMULTI |
S
Stefan Rompf 已提交
6211 6212 6213
				IFF_RUNNING |
				IFF_LOWER_UP |
				IFF_DORMANT)) |
L
Linus Torvalds 已提交
6214 6215 6216
		(dev->gflags & (IFF_PROMISC |
				IFF_ALLMULTI));

S
Stefan Rompf 已提交
6217 6218 6219 6220 6221 6222 6223 6224
	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
6225 6226 6227

	return flags;
}
E
Eric Dumazet 已提交
6228
EXPORT_SYMBOL(dev_get_flags);
L
Linus Torvalds 已提交
6229

6230
int __dev_change_flags(struct net_device *dev, unsigned int flags)
L
Linus Torvalds 已提交
6231
{
6232
	unsigned int old_flags = dev->flags;
6233
	int ret;
L
Linus Torvalds 已提交
6234

6235 6236
	ASSERT_RTNL();

L
Linus Torvalds 已提交
6237 6238 6239 6240 6241 6242 6243 6244 6245 6246 6247 6248 6249 6250
	/*
	 *	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.
	 */

6251 6252
	if ((old_flags ^ flags) & IFF_MULTICAST)
		dev_change_rx_flags(dev, IFF_MULTICAST);
6253

6254
	dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
6255 6256 6257 6258 6259 6260 6261 6262

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

	ret = 0;
6263
	if ((old_flags ^ flags) & IFF_UP)
6264
		ret = ((old_flags & IFF_UP) ? __dev_close : __dev_open)(dev);
L
Linus Torvalds 已提交
6265 6266

	if ((flags ^ dev->gflags) & IFF_PROMISC) {
E
Eric Dumazet 已提交
6267
		int inc = (flags & IFF_PROMISC) ? 1 : -1;
6268
		unsigned int old_flags = dev->flags;
E
Eric Dumazet 已提交
6269

L
Linus Torvalds 已提交
6270
		dev->gflags ^= IFF_PROMISC;
6271 6272 6273 6274

		if (__dev_set_promiscuity(dev, inc, false) >= 0)
			if (dev->flags != old_flags)
				dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
6275 6276 6277 6278 6279 6280 6281
	}

	/* 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
6282 6283
		int inc = (flags & IFF_ALLMULTI) ? 1 : -1;

L
Linus Torvalds 已提交
6284
		dev->gflags ^= IFF_ALLMULTI;
6285
		__dev_set_allmulti(dev, inc, false);
L
Linus Torvalds 已提交
6286 6287
	}

6288 6289 6290
	return ret;
}

6291 6292
void __dev_notify_flags(struct net_device *dev, unsigned int old_flags,
			unsigned int gchanges)
6293 6294 6295
{
	unsigned int changes = dev->flags ^ old_flags;

6296
	if (gchanges)
6297
		rtmsg_ifinfo(RTM_NEWLINK, dev, gchanges, GFP_ATOMIC);
6298

6299 6300 6301 6302 6303 6304 6305 6306
	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 &&
6307 6308 6309 6310 6311 6312 6313
	    (changes & ~(IFF_UP | IFF_PROMISC | IFF_ALLMULTI | IFF_VOLATILE))) {
		struct netdev_notifier_change_info change_info;

		change_info.flags_changed = changes;
		call_netdevice_notifiers_info(NETDEV_CHANGE, dev,
					      &change_info.info);
	}
6314 6315 6316 6317 6318 6319 6320 6321 6322 6323
}

/**
 *	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.
 */
6324
int dev_change_flags(struct net_device *dev, unsigned int flags)
6325
{
6326
	int ret;
6327
	unsigned int changes, old_flags = dev->flags, old_gflags = dev->gflags;
6328 6329 6330 6331 6332

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

6333
	changes = (old_flags ^ dev->flags) | (old_gflags ^ dev->gflags);
6334
	__dev_notify_flags(dev, old_flags, changes);
L
Linus Torvalds 已提交
6335 6336
	return ret;
}
E
Eric Dumazet 已提交
6337
EXPORT_SYMBOL(dev_change_flags);
L
Linus Torvalds 已提交
6338

6339 6340 6341 6342 6343 6344 6345 6346 6347 6348 6349
static int __dev_set_mtu(struct net_device *dev, int new_mtu)
{
	const struct net_device_ops *ops = dev->netdev_ops;

	if (ops->ndo_change_mtu)
		return ops->ndo_change_mtu(dev, new_mtu);

	dev->mtu = new_mtu;
	return 0;
}

6350 6351 6352 6353 6354 6355 6356
/**
 *	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
6357 6358
int dev_set_mtu(struct net_device *dev, int new_mtu)
{
6359
	int err, orig_mtu;
L
Linus Torvalds 已提交
6360 6361 6362 6363 6364 6365 6366 6367 6368 6369 6370

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

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

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

6371 6372 6373 6374
	err = call_netdevice_notifiers(NETDEV_PRECHANGEMTU, dev);
	err = notifier_to_errno(err);
	if (err)
		return err;
6375

6376 6377
	orig_mtu = dev->mtu;
	err = __dev_set_mtu(dev, new_mtu);
6378

6379 6380 6381 6382 6383 6384 6385 6386 6387 6388 6389
	if (!err) {
		err = call_netdevice_notifiers(NETDEV_CHANGEMTU, dev);
		err = notifier_to_errno(err);
		if (err) {
			/* setting mtu back and notifying everyone again,
			 * so that they have a chance to revert changes.
			 */
			__dev_set_mtu(dev, orig_mtu);
			call_netdevice_notifiers(NETDEV_CHANGEMTU, dev);
		}
	}
L
Linus Torvalds 已提交
6390 6391
	return err;
}
E
Eric Dumazet 已提交
6392
EXPORT_SYMBOL(dev_set_mtu);
L
Linus Torvalds 已提交
6393

6394 6395 6396 6397 6398 6399 6400 6401 6402 6403 6404
/**
 *	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);

6405 6406 6407 6408 6409 6410 6411
/**
 *	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
6412 6413
int dev_set_mac_address(struct net_device *dev, struct sockaddr *sa)
{
6414
	const struct net_device_ops *ops = dev->netdev_ops;
L
Linus Torvalds 已提交
6415 6416
	int err;

6417
	if (!ops->ndo_set_mac_address)
L
Linus Torvalds 已提交
6418 6419 6420 6421 6422
		return -EOPNOTSUPP;
	if (sa->sa_family != dev->type)
		return -EINVAL;
	if (!netif_device_present(dev))
		return -ENODEV;
6423
	err = ops->ndo_set_mac_address(dev, sa);
6424 6425
	if (err)
		return err;
J
Jiri Pirko 已提交
6426
	dev->addr_assign_type = NET_ADDR_SET;
6427
	call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
6428
	add_device_randomness(dev->dev_addr, dev->addr_len);
6429
	return 0;
L
Linus Torvalds 已提交
6430
}
E
Eric Dumazet 已提交
6431
EXPORT_SYMBOL(dev_set_mac_address);
L
Linus Torvalds 已提交
6432

J
Jiri Pirko 已提交
6433 6434 6435
/**
 *	dev_change_carrier - Change device carrier
 *	@dev: device
6436
 *	@new_carrier: new value
J
Jiri Pirko 已提交
6437 6438 6439 6440 6441 6442 6443 6444 6445 6446 6447 6448 6449 6450 6451
 *
 *	Change device carrier
 */
int dev_change_carrier(struct net_device *dev, bool new_carrier)
{
	const struct net_device_ops *ops = dev->netdev_ops;

	if (!ops->ndo_change_carrier)
		return -EOPNOTSUPP;
	if (!netif_device_present(dev))
		return -ENODEV;
	return ops->ndo_change_carrier(dev, new_carrier);
}
EXPORT_SYMBOL(dev_change_carrier);

6452 6453 6454 6455 6456 6457 6458 6459
/**
 *	dev_get_phys_port_id - Get device physical port ID
 *	@dev: device
 *	@ppid: port ID
 *
 *	Get device physical port ID
 */
int dev_get_phys_port_id(struct net_device *dev,
6460
			 struct netdev_phys_item_id *ppid)
6461 6462 6463 6464 6465 6466 6467 6468 6469
{
	const struct net_device_ops *ops = dev->netdev_ops;

	if (!ops->ndo_get_phys_port_id)
		return -EOPNOTSUPP;
	return ops->ndo_get_phys_port_id(dev, ppid);
}
EXPORT_SYMBOL(dev_get_phys_port_id);

6470 6471 6472 6473
/**
 *	dev_get_phys_port_name - Get device physical port name
 *	@dev: device
 *	@name: port name
6474
 *	@len: limit of bytes to copy to name
6475 6476 6477 6478 6479 6480 6481 6482 6483 6484 6485 6486 6487 6488
 *
 *	Get device physical port name
 */
int dev_get_phys_port_name(struct net_device *dev,
			   char *name, size_t len)
{
	const struct net_device_ops *ops = dev->netdev_ops;

	if (!ops->ndo_get_phys_port_name)
		return -EOPNOTSUPP;
	return ops->ndo_get_phys_port_name(dev, name, len);
}
EXPORT_SYMBOL(dev_get_phys_port_name);

6489 6490 6491 6492 6493 6494 6495 6496 6497 6498 6499 6500 6501 6502 6503 6504 6505 6506 6507 6508
/**
 *	dev_change_proto_down - update protocol port state information
 *	@dev: device
 *	@proto_down: new value
 *
 *	This info can be used by switch drivers to set the phys state of the
 *	port.
 */
int dev_change_proto_down(struct net_device *dev, bool proto_down)
{
	const struct net_device_ops *ops = dev->netdev_ops;

	if (!ops->ndo_change_proto_down)
		return -EOPNOTSUPP;
	if (!netif_device_present(dev))
		return -ENODEV;
	return ops->ndo_change_proto_down(dev, proto_down);
}
EXPORT_SYMBOL(dev_change_proto_down);

L
Linus Torvalds 已提交
6509 6510
/**
 *	dev_new_index	-	allocate an ifindex
6511
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
6512 6513 6514 6515 6516
 *
 *	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.
 */
6517
static int dev_new_index(struct net *net)
L
Linus Torvalds 已提交
6518
{
6519
	int ifindex = net->ifindex;
L
Linus Torvalds 已提交
6520 6521 6522
	for (;;) {
		if (++ifindex <= 0)
			ifindex = 1;
6523
		if (!__dev_get_by_index(net, ifindex))
6524
			return net->ifindex = ifindex;
L
Linus Torvalds 已提交
6525 6526 6527 6528
	}
}

/* Delayed registration/unregisteration */
6529
static LIST_HEAD(net_todo_list);
6530
DECLARE_WAIT_QUEUE_HEAD(netdev_unregistering_wq);
L
Linus Torvalds 已提交
6531

6532
static void net_set_todo(struct net_device *dev)
L
Linus Torvalds 已提交
6533 6534
{
	list_add_tail(&dev->todo_list, &net_todo_list);
6535
	dev_net(dev)->dev_unreg_count++;
L
Linus Torvalds 已提交
6536 6537
}

6538
static void rollback_registered_many(struct list_head *head)
6539
{
6540
	struct net_device *dev, *tmp;
6541
	LIST_HEAD(close_head);
6542

6543 6544 6545
	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

6546
	list_for_each_entry_safe(dev, tmp, head, unreg_list) {
6547
		/* Some devices call without registering
6548 6549
		 * for initialization unwind. Remove those
		 * devices and proceed with the remaining.
6550 6551
		 */
		if (dev->reg_state == NETREG_UNINITIALIZED) {
6552 6553
			pr_debug("unregister_netdevice: device %s/%p never was registered\n",
				 dev->name, dev);
6554

6555
			WARN_ON(1);
6556 6557
			list_del(&dev->unreg_list);
			continue;
6558
		}
6559
		dev->dismantle = true;
6560
		BUG_ON(dev->reg_state != NETREG_REGISTERED);
6561
	}
6562

6563
	/* If device is running, close it first. */
6564 6565
	list_for_each_entry(dev, head, unreg_list)
		list_add_tail(&dev->close_list, &close_head);
6566
	dev_close_many(&close_head, true);
6567

6568
	list_for_each_entry(dev, head, unreg_list) {
6569 6570
		/* And unlink it from device chain. */
		unlist_netdevice(dev);
6571

6572
		dev->reg_state = NETREG_UNREGISTERING;
6573
		on_each_cpu(flush_backlog, dev, 1);
6574
	}
6575 6576 6577

	synchronize_net();

6578
	list_for_each_entry(dev, head, unreg_list) {
6579 6580
		struct sk_buff *skb = NULL;

6581 6582
		/* Shutdown queueing discipline. */
		dev_shutdown(dev);
6583 6584


6585 6586 6587 6588
		/* Notify protocols, that we are about to destroy
		   this device. They should clean all the things.
		*/
		call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
6589

6590 6591 6592 6593 6594
		if (!dev->rtnl_link_ops ||
		    dev->rtnl_link_state == RTNL_LINK_INITIALIZED)
			skb = rtmsg_ifinfo_build_skb(RTM_DELLINK, dev, ~0U,
						     GFP_KERNEL);

6595 6596 6597
		/*
		 *	Flush the unicast and multicast chains
		 */
6598
		dev_uc_flush(dev);
6599
		dev_mc_flush(dev);
6600

6601 6602
		if (dev->netdev_ops->ndo_uninit)
			dev->netdev_ops->ndo_uninit(dev);
6603

6604 6605
		if (skb)
			rtmsg_ifinfo_send(skb, dev, GFP_KERNEL);
6606

J
Jiri Pirko 已提交
6607 6608
		/* Notifier chain MUST detach us all upper devices. */
		WARN_ON(netdev_has_any_upper_dev(dev));
6609

6610 6611
		/* Remove entries from kobject tree */
		netdev_unregister_kobject(dev);
6612 6613 6614 6615
#ifdef CONFIG_XPS
		/* Remove XPS queueing entries */
		netif_reset_xps_queues_gt(dev, 0);
#endif
6616
	}
6617

6618
	synchronize_net();
6619

6620
	list_for_each_entry(dev, head, unreg_list)
6621 6622 6623 6624 6625 6626 6627 6628 6629
		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
6630
	list_del(&single);
6631 6632
}

6633 6634 6635 6636 6637
static netdev_features_t netdev_sync_upper_features(struct net_device *lower,
	struct net_device *upper, netdev_features_t features)
{
	netdev_features_t upper_disables = NETIF_F_UPPER_DISABLES;
	netdev_features_t feature;
6638
	int feature_bit;
6639

6640 6641
	for_each_netdev_feature(&upper_disables, feature_bit) {
		feature = __NETIF_F_BIT(feature_bit);
6642 6643 6644 6645 6646 6647 6648 6649 6650 6651 6652 6653 6654 6655 6656 6657
		if (!(upper->wanted_features & feature)
		    && (features & feature)) {
			netdev_dbg(lower, "Dropping feature %pNF, upper dev %s has it off.\n",
				   &feature, upper->name);
			features &= ~feature;
		}
	}

	return features;
}

static void netdev_sync_lower_features(struct net_device *upper,
	struct net_device *lower, netdev_features_t features)
{
	netdev_features_t upper_disables = NETIF_F_UPPER_DISABLES;
	netdev_features_t feature;
6658
	int feature_bit;
6659

6660 6661
	for_each_netdev_feature(&upper_disables, feature_bit) {
		feature = __NETIF_F_BIT(feature_bit);
6662 6663 6664 6665 6666 6667 6668 6669 6670 6671 6672 6673 6674
		if (!(features & feature) && (lower->features & feature)) {
			netdev_dbg(upper, "Disabling feature %pNF on lower dev %s.\n",
				   &feature, lower->name);
			lower->wanted_features &= ~feature;
			netdev_update_features(lower);

			if (unlikely(lower->features & feature))
				netdev_WARN(upper, "failed to disable %pNF on %s!\n",
					    &feature, lower->name);
		}
	}
}

6675 6676
static netdev_features_t netdev_fix_features(struct net_device *dev,
	netdev_features_t features)
6677
{
6678 6679 6680
	/* Fix illegal checksum combinations */
	if ((features & NETIF_F_HW_CSUM) &&
	    (features & (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))) {
6681
		netdev_warn(dev, "mixed HW and IP checksum settings.\n");
6682 6683 6684
		features &= ~(NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM);
	}

6685
	/* TSO requires that SG is present as well. */
6686
	if ((features & NETIF_F_ALL_TSO) && !(features & NETIF_F_SG)) {
6687
		netdev_dbg(dev, "Dropping TSO features since no SG feature.\n");
6688
		features &= ~NETIF_F_ALL_TSO;
6689 6690
	}

6691 6692 6693 6694 6695 6696 6697 6698 6699 6700 6701 6702 6703
	if ((features & NETIF_F_TSO) && !(features & NETIF_F_HW_CSUM) &&
					!(features & NETIF_F_IP_CSUM)) {
		netdev_dbg(dev, "Dropping TSO features since no CSUM feature.\n");
		features &= ~NETIF_F_TSO;
		features &= ~NETIF_F_TSO_ECN;
	}

	if ((features & NETIF_F_TSO6) && !(features & NETIF_F_HW_CSUM) &&
					 !(features & NETIF_F_IPV6_CSUM)) {
		netdev_dbg(dev, "Dropping TSO6 features since no CSUM feature.\n");
		features &= ~NETIF_F_TSO6;
	}

6704 6705 6706 6707
	/* TSO ECN requires that TSO is present as well. */
	if ((features & NETIF_F_ALL_TSO) == NETIF_F_TSO_ECN)
		features &= ~NETIF_F_TSO_ECN;

6708 6709
	/* Software GSO depends on SG. */
	if ((features & NETIF_F_GSO) && !(features & NETIF_F_SG)) {
6710
		netdev_dbg(dev, "Dropping NETIF_F_GSO since no SG feature.\n");
6711 6712 6713
		features &= ~NETIF_F_GSO;
	}

6714
	/* UFO needs SG and checksumming */
6715
	if (features & NETIF_F_UFO) {
6716
		/* maybe split UFO into V4 and V6? */
6717 6718 6719
		if (!(features & NETIF_F_HW_CSUM) &&
		    ((features & (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM)) !=
		     (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM))) {
6720
			netdev_dbg(dev,
6721
				"Dropping NETIF_F_UFO since no checksum offload features.\n");
6722 6723 6724 6725
			features &= ~NETIF_F_UFO;
		}

		if (!(features & NETIF_F_SG)) {
6726
			netdev_dbg(dev,
6727
				"Dropping NETIF_F_UFO since no NETIF_F_SG feature.\n");
6728 6729 6730 6731
			features &= ~NETIF_F_UFO;
		}
	}

J
Jiri Pirko 已提交
6732 6733 6734 6735 6736 6737 6738
#ifdef CONFIG_NET_RX_BUSY_POLL
	if (dev->netdev_ops->ndo_busy_poll)
		features |= NETIF_F_BUSY_POLL;
	else
#endif
		features &= ~NETIF_F_BUSY_POLL;

6739 6740 6741
	return features;
}

6742
int __netdev_update_features(struct net_device *dev)
6743
{
6744
	struct net_device *upper, *lower;
6745
	netdev_features_t features;
6746
	struct list_head *iter;
6747
	int err = -1;
6748

6749 6750
	ASSERT_RTNL();

6751 6752 6753 6754 6755 6756 6757 6758
	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);

6759 6760 6761 6762
	/* some features can't be enabled if they're off an an upper device */
	netdev_for_each_upper_dev_rcu(dev, upper, iter)
		features = netdev_sync_upper_features(dev, upper, features);

6763
	if (dev->features == features)
6764
		goto sync_lower;
6765

6766 6767
	netdev_dbg(dev, "Features changed: %pNF -> %pNF\n",
		&dev->features, &features);
6768 6769 6770

	if (dev->netdev_ops->ndo_set_features)
		err = dev->netdev_ops->ndo_set_features(dev, features);
6771 6772
	else
		err = 0;
6773

6774
	if (unlikely(err < 0)) {
6775
		netdev_err(dev,
6776 6777
			"set_features() failed (%d); wanted %pNF, left %pNF\n",
			err, &features, &dev->features);
6778 6779 6780 6781
		/* return non-0 since some features might have changed and
		 * it's better to fire a spurious notification than miss it
		 */
		return -1;
6782 6783
	}

6784
sync_lower:
6785 6786 6787 6788 6789 6790
	/* some features must be disabled on lower devices when disabled
	 * on an upper device (think: bonding master or bridge)
	 */
	netdev_for_each_lower_dev(dev, lower, iter)
		netdev_sync_lower_features(dev, lower, features);

6791 6792 6793
	if (!err)
		dev->features = features;

6794
	return err < 0 ? 0 : 1;
6795 6796
}

6797 6798 6799 6800 6801 6802 6803 6804
/**
 *	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.
 */
6805 6806 6807 6808
void netdev_update_features(struct net_device *dev)
{
	if (__netdev_update_features(dev))
		netdev_features_change(dev);
6809 6810 6811
}
EXPORT_SYMBOL(netdev_update_features);

6812 6813 6814 6815 6816 6817 6818 6819 6820 6821 6822 6823 6824 6825 6826 6827 6828
/**
 *	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);

6829 6830 6831 6832 6833 6834 6835 6836 6837 6838 6839 6840 6841 6842 6843 6844 6845 6846 6847 6848 6849 6850 6851 6852 6853 6854 6855
/**
 *	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);

6856
#ifdef CONFIG_SYSFS
6857 6858 6859
static int netif_alloc_rx_queues(struct net_device *dev)
{
	unsigned int i, count = dev->num_rx_queues;
T
Tom Herbert 已提交
6860
	struct netdev_rx_queue *rx;
6861
	size_t sz = count * sizeof(*rx);
6862

T
Tom Herbert 已提交
6863
	BUG_ON(count < 1);
6864

6865 6866 6867 6868 6869 6870
	rx = kzalloc(sz, GFP_KERNEL | __GFP_NOWARN | __GFP_REPEAT);
	if (!rx) {
		rx = vzalloc(sz);
		if (!rx)
			return -ENOMEM;
	}
T
Tom Herbert 已提交
6871 6872 6873
	dev->_rx = rx;

	for (i = 0; i < count; i++)
T
Tom Herbert 已提交
6874
		rx[i].dev = dev;
6875 6876
	return 0;
}
T
Tom Herbert 已提交
6877
#endif
6878

C
Changli Gao 已提交
6879 6880 6881 6882 6883 6884 6885
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;
6886
	netdev_queue_numa_node_write(queue, NUMA_NO_NODE);
C
Changli Gao 已提交
6887
	queue->dev = dev;
T
Tom Herbert 已提交
6888 6889 6890
#ifdef CONFIG_BQL
	dql_init(&queue->dql, HZ);
#endif
C
Changli Gao 已提交
6891 6892
}

6893 6894
static void netif_free_tx_queues(struct net_device *dev)
{
W
WANG Cong 已提交
6895
	kvfree(dev->_tx);
6896 6897
}

6898 6899 6900 6901
static int netif_alloc_netdev_queues(struct net_device *dev)
{
	unsigned int count = dev->num_tx_queues;
	struct netdev_queue *tx;
6902
	size_t sz = count * sizeof(*tx);
6903

6904 6905
	if (count < 1 || count > 0xffff)
		return -EINVAL;
6906

6907 6908 6909 6910 6911 6912
	tx = kzalloc(sz, GFP_KERNEL | __GFP_NOWARN | __GFP_REPEAT);
	if (!tx) {
		tx = vzalloc(sz);
		if (!tx)
			return -ENOMEM;
	}
6913
	dev->_tx = tx;
T
Tom Herbert 已提交
6914

6915 6916
	netdev_for_each_tx_queue(dev, netdev_init_one_queue, NULL);
	spin_lock_init(&dev->tx_global_lock);
C
Changli Gao 已提交
6917 6918

	return 0;
6919 6920
}

6921 6922 6923 6924 6925 6926 6927 6928 6929 6930 6931
void netif_tx_stop_all_queues(struct net_device *dev)
{
	unsigned int i;

	for (i = 0; i < dev->num_tx_queues; i++) {
		struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
		netif_tx_stop_queue(txq);
	}
}
EXPORT_SYMBOL(netif_tx_stop_all_queues);

L
Linus Torvalds 已提交
6932 6933 6934 6935 6936 6937 6938 6939 6940 6941 6942 6943 6944 6945 6946 6947 6948 6949 6950 6951
/**
 *	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;
6952
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
6953 6954 6955 6956

	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

6957 6958
	might_sleep();

L
Linus Torvalds 已提交
6959 6960
	/* When net_device's are persistent, this will be fatal. */
	BUG_ON(dev->reg_state != NETREG_UNINITIALIZED);
6961
	BUG_ON(!net);
L
Linus Torvalds 已提交
6962

6963
	spin_lock_init(&dev->addr_list_lock);
6964
	netdev_set_addr_lockdep_class(dev);
L
Linus Torvalds 已提交
6965

6966
	ret = dev_get_valid_name(net, dev, dev->name);
6967 6968 6969
	if (ret < 0)
		goto out;

L
Linus Torvalds 已提交
6970
	/* Init, if this function is available */
6971 6972
	if (dev->netdev_ops->ndo_init) {
		ret = dev->netdev_ops->ndo_init(dev);
L
Linus Torvalds 已提交
6973 6974 6975
		if (ret) {
			if (ret > 0)
				ret = -EIO;
6976
			goto out;
L
Linus Torvalds 已提交
6977 6978
		}
	}
6979

6980 6981
	if (((dev->hw_features | dev->features) &
	     NETIF_F_HW_VLAN_CTAG_FILTER) &&
6982 6983 6984 6985 6986 6987 6988
	    (!dev->netdev_ops->ndo_vlan_rx_add_vid ||
	     !dev->netdev_ops->ndo_vlan_rx_kill_vid)) {
		netdev_WARN(dev, "Buggy VLAN acceleration in driver!\n");
		ret = -EINVAL;
		goto err_uninit;
	}

6989 6990 6991 6992 6993 6994
	ret = -EBUSY;
	if (!dev->ifindex)
		dev->ifindex = dev_new_index(net);
	else if (__dev_get_by_index(net, dev->ifindex))
		goto err_uninit;

6995 6996 6997 6998
	/* Transfer changeable features to wanted_features and enable
	 * software offloads (GSO and GRO).
	 */
	dev->hw_features |= NETIF_F_SOFT_FEATURES;
6999 7000
	dev->features |= NETIF_F_SOFT_FEATURES;
	dev->wanted_features = dev->features & dev->hw_features;
L
Linus Torvalds 已提交
7001

7002
	if (!(dev->flags & IFF_LOOPBACK))
7003
		dev->hw_features |= NETIF_F_NOCACHE_COPY;
7004 7005 7006

	if (dev->hw_features & NETIF_F_TSO)
		dev->hw_features |= NETIF_F_TSO_MANGLEID;
7007

7008
	/* Make NETIF_F_HIGHDMA inheritable to VLAN devices.
7009
	 */
7010
	dev->vlan_features |= NETIF_F_HIGHDMA;
7011

7012 7013 7014 7015
	/* Make NETIF_F_SG inheritable to tunnel devices.
	 */
	dev->hw_enc_features |= NETIF_F_SG;

S
Simon Horman 已提交
7016 7017 7018 7019
	/* Make NETIF_F_SG inheritable to MPLS.
	 */
	dev->mpls_features |= NETIF_F_SG;

7020 7021 7022 7023 7024
	ret = call_netdevice_notifiers(NETDEV_POST_INIT, dev);
	ret = notifier_to_errno(ret);
	if (ret)
		goto err_uninit;

7025
	ret = netdev_register_kobject(dev);
7026
	if (ret)
7027
		goto err_uninit;
7028 7029
	dev->reg_state = NETREG_REGISTERED;

7030
	__netdev_update_features(dev);
7031

L
Linus Torvalds 已提交
7032 7033 7034 7035 7036 7037 7038
	/*
	 *	Default initial state at registry is that the
	 *	device is present.
	 */

	set_bit(__LINK_STATE_PRESENT, &dev->state);

7039 7040
	linkwatch_init_dev(dev);

L
Linus Torvalds 已提交
7041 7042
	dev_init_scheduler(dev);
	dev_hold(dev);
7043
	list_netdevice(dev);
7044
	add_device_randomness(dev->dev_addr, dev->addr_len);
L
Linus Torvalds 已提交
7045

7046 7047 7048 7049 7050 7051 7052
	/* If the device has permanent device address, driver should
	 * set dev_addr and also addr_assign_type should be set to
	 * NET_ADDR_PERM (default value).
	 */
	if (dev->addr_assign_type == NET_ADDR_PERM)
		memcpy(dev->perm_addr, dev->dev_addr, dev->addr_len);

L
Linus Torvalds 已提交
7053
	/* Notify protocols, that a new device appeared. */
7054
	ret = call_netdevice_notifiers(NETDEV_REGISTER, dev);
7055
	ret = notifier_to_errno(ret);
7056 7057 7058 7059
	if (ret) {
		rollback_registered(dev);
		dev->reg_state = NETREG_UNREGISTERED;
	}
7060 7061 7062 7063
	/*
	 *	Prevent userspace races by waiting until the network
	 *	device is fully setup before sending notifications.
	 */
7064 7065
	if (!dev->rtnl_link_ops ||
	    dev->rtnl_link_state == RTNL_LINK_INITIALIZED)
7066
		rtmsg_ifinfo(RTM_NEWLINK, dev, ~0U, GFP_KERNEL);
L
Linus Torvalds 已提交
7067 7068 7069

out:
	return ret;
7070 7071

err_uninit:
7072 7073
	if (dev->netdev_ops->ndo_uninit)
		dev->netdev_ops->ndo_uninit(dev);
7074
	goto out;
L
Linus Torvalds 已提交
7075
}
E
Eric Dumazet 已提交
7076
EXPORT_SYMBOL(register_netdevice);
L
Linus Torvalds 已提交
7077

7078 7079 7080 7081 7082 7083 7084 7085 7086 7087 7088 7089 7090 7091 7092 7093 7094 7095 7096 7097 7098 7099 7100 7101 7102 7103 7104 7105 7106 7107 7108
/**
 *	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
7109 7110 7111 7112 7113
	/* Note : We dont allocate pcpu_refcnt for dummy devices,
	 * because users of this 'device' dont need to change
	 * its refcount.
	 */

7114 7115 7116 7117 7118
	return 0;
}
EXPORT_SYMBOL_GPL(init_dummy_netdev);


L
Linus Torvalds 已提交
7119 7120 7121 7122 7123 7124 7125 7126 7127
/**
 *	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.
 *
7128
 *	This is a wrapper around register_netdevice that takes the rtnl semaphore
L
Linus Torvalds 已提交
7129 7130 7131 7132 7133 7134 7135 7136 7137 7138 7139 7140 7141 7142
 *	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
7143 7144 7145 7146 7147 7148 7149 7150 7151 7152
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);

7153
/**
L
Linus Torvalds 已提交
7154
 * netdev_wait_allrefs - wait until all references are gone.
7155
 * @dev: target net_device
L
Linus Torvalds 已提交
7156 7157 7158 7159 7160 7161 7162
 *
 * 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
7163
 * call dev_put.
L
Linus Torvalds 已提交
7164 7165 7166 7167
 */
static void netdev_wait_allrefs(struct net_device *dev)
{
	unsigned long rebroadcast_time, warning_time;
E
Eric Dumazet 已提交
7168
	int refcnt;
L
Linus Torvalds 已提交
7169

7170 7171
	linkwatch_forget_dev(dev);

L
Linus Torvalds 已提交
7172
	rebroadcast_time = warning_time = jiffies;
E
Eric Dumazet 已提交
7173 7174 7175
	refcnt = netdev_refcnt_read(dev);

	while (refcnt != 0) {
L
Linus Torvalds 已提交
7176
		if (time_after(jiffies, rebroadcast_time + 1 * HZ)) {
7177
			rtnl_lock();
L
Linus Torvalds 已提交
7178 7179

			/* Rebroadcast unregister notification */
7180
			call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
L
Linus Torvalds 已提交
7181

7182
			__rtnl_unlock();
7183
			rcu_barrier();
7184 7185
			rtnl_lock();

7186
			call_netdevice_notifiers(NETDEV_UNREGISTER_FINAL, dev);
L
Linus Torvalds 已提交
7187 7188 7189 7190 7191 7192 7193 7194 7195 7196 7197
			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();
			}

7198
			__rtnl_unlock();
L
Linus Torvalds 已提交
7199 7200 7201 7202 7203 7204

			rebroadcast_time = jiffies;
		}

		msleep(250);

E
Eric Dumazet 已提交
7205 7206
		refcnt = netdev_refcnt_read(dev);

L
Linus Torvalds 已提交
7207
		if (time_after(jiffies, warning_time + 10 * HZ)) {
7208 7209
			pr_emerg("unregister_netdevice: waiting for %s to become free. Usage count = %d\n",
				 dev->name, refcnt);
L
Linus Torvalds 已提交
7210 7211 7212 7213 7214 7215 7216 7217 7218 7219 7220 7221 7222 7223 7224 7225 7226 7227 7228
			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
7229
 * We are invoked by rtnl_unlock().
L
Linus Torvalds 已提交
7230
 * This allows us to deal with problems:
7231
 * 1) We can delete sysfs objects which invoke hotplug
L
Linus Torvalds 已提交
7232 7233 7234
 *    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
7235 7236 7237
 *
 * 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
7238 7239 7240
 */
void netdev_run_todo(void)
{
7241
	struct list_head list;
L
Linus Torvalds 已提交
7242 7243

	/* Snapshot list, allow later requests */
7244
	list_replace_init(&net_todo_list, &list);
H
Herbert Xu 已提交
7245 7246

	__rtnl_unlock();
7247

7248 7249

	/* Wait for rcu callbacks to finish before next phase */
7250 7251 7252
	if (!list_empty(&list))
		rcu_barrier();

L
Linus Torvalds 已提交
7253 7254
	while (!list_empty(&list)) {
		struct net_device *dev
7255
			= list_first_entry(&list, struct net_device, todo_list);
L
Linus Torvalds 已提交
7256 7257
		list_del(&dev->todo_list);

7258
		rtnl_lock();
7259
		call_netdevice_notifiers(NETDEV_UNREGISTER_FINAL, dev);
7260
		__rtnl_unlock();
7261

7262
		if (unlikely(dev->reg_state != NETREG_UNREGISTERING)) {
7263
			pr_err("network todo '%s' but state %d\n",
7264 7265 7266 7267
			       dev->name, dev->reg_state);
			dump_stack();
			continue;
		}
L
Linus Torvalds 已提交
7268

7269
		dev->reg_state = NETREG_UNREGISTERED;
L
Linus Torvalds 已提交
7270

7271
		netdev_wait_allrefs(dev);
L
Linus Torvalds 已提交
7272

7273
		/* paranoia */
E
Eric Dumazet 已提交
7274
		BUG_ON(netdev_refcnt_read(dev));
7275 7276
		BUG_ON(!list_empty(&dev->ptype_all));
		BUG_ON(!list_empty(&dev->ptype_specific));
7277 7278
		WARN_ON(rcu_access_pointer(dev->ip_ptr));
		WARN_ON(rcu_access_pointer(dev->ip6_ptr));
7279
		WARN_ON(dev->dn_ptr);
L
Linus Torvalds 已提交
7280

7281 7282
		if (dev->destructor)
			dev->destructor(dev);
7283

7284 7285 7286 7287 7288 7289
		/* Report a network device has been unregistered */
		rtnl_lock();
		dev_net(dev)->dev_unreg_count--;
		__rtnl_unlock();
		wake_up(&netdev_unregistering_wq);

7290 7291
		/* Free network device */
		kobject_put(&dev->dev.kobj);
L
Linus Torvalds 已提交
7292 7293 7294
	}
}

7295 7296 7297 7298
/* Convert net_device_stats to rtnl_link_stats64. rtnl_link_stats64 has
 * all the same fields in the same order as net_device_stats, with only
 * the type differing, but rtnl_link_stats64 may have additional fields
 * at the end for newer counters.
7299
 */
7300 7301
void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64,
			     const struct net_device_stats *netdev_stats)
7302 7303
{
#if BITS_PER_LONG == 64
7304
	BUILD_BUG_ON(sizeof(*stats64) < sizeof(*netdev_stats));
7305
	memcpy(stats64, netdev_stats, sizeof(*stats64));
7306 7307 7308
	/* zero out counters that only exist in rtnl_link_stats64 */
	memset((char *)stats64 + sizeof(*netdev_stats), 0,
	       sizeof(*stats64) - sizeof(*netdev_stats));
7309
#else
7310
	size_t i, n = sizeof(*netdev_stats) / sizeof(unsigned long);
7311 7312 7313
	const unsigned long *src = (const unsigned long *)netdev_stats;
	u64 *dst = (u64 *)stats64;

7314
	BUILD_BUG_ON(n > sizeof(*stats64) / sizeof(u64));
7315 7316
	for (i = 0; i < n; i++)
		dst[i] = src[i];
7317 7318 7319
	/* zero out counters that only exist in rtnl_link_stats64 */
	memset((char *)stats64 + n * sizeof(u64), 0,
	       sizeof(*stats64) - n * sizeof(u64));
7320 7321
#endif
}
7322
EXPORT_SYMBOL(netdev_stats_to_stats64);
7323

7324 7325 7326
/**
 *	dev_get_stats	- get network device statistics
 *	@dev: device to get statistics from
7327
 *	@storage: place to store stats
7328
 *
7329 7330 7331 7332
 *	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.
7333
 */
7334 7335
struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev,
					struct rtnl_link_stats64 *storage)
7336
{
7337 7338
	const struct net_device_ops *ops = dev->netdev_ops;

7339 7340
	if (ops->ndo_get_stats64) {
		memset(storage, 0, sizeof(*storage));
7341 7342
		ops->ndo_get_stats64(dev, storage);
	} else if (ops->ndo_get_stats) {
7343
		netdev_stats_to_stats64(storage, ops->ndo_get_stats(dev));
7344 7345
	} else {
		netdev_stats_to_stats64(storage, &dev->stats);
7346
	}
7347
	storage->rx_dropped += atomic_long_read(&dev->rx_dropped);
7348
	storage->tx_dropped += atomic_long_read(&dev->tx_dropped);
7349
	storage->rx_nohandler += atomic_long_read(&dev->rx_nohandler);
7350
	return storage;
R
Rusty Russell 已提交
7351
}
7352
EXPORT_SYMBOL(dev_get_stats);
R
Rusty Russell 已提交
7353

7354
struct netdev_queue *dev_ingress_queue_create(struct net_device *dev)
7355
{
7356
	struct netdev_queue *queue = dev_ingress_queue(dev);
7357

7358 7359 7360 7361 7362 7363 7364
#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);
E
Eric Dumazet 已提交
7365
	RCU_INIT_POINTER(queue->qdisc, &noop_qdisc);
7366 7367 7368 7369
	queue->qdisc_sleeping = &noop_qdisc;
	rcu_assign_pointer(dev->ingress_queue, queue);
#endif
	return queue;
7370 7371
}

7372 7373
static const struct ethtool_ops default_ethtool_ops;

7374 7375 7376 7377 7378 7379 7380 7381
void netdev_set_default_ethtool_ops(struct net_device *dev,
				    const struct ethtool_ops *ops)
{
	if (dev->ethtool_ops == &default_ethtool_ops)
		dev->ethtool_ops = ops;
}
EXPORT_SYMBOL_GPL(netdev_set_default_ethtool_ops);

7382 7383 7384 7385
void netdev_freemem(struct net_device *dev)
{
	char *addr = (char *)dev - dev->padded;

W
WANG Cong 已提交
7386
	kvfree(addr);
7387 7388
}

L
Linus Torvalds 已提交
7389
/**
T
Tom Herbert 已提交
7390
 *	alloc_netdev_mqs - allocate network device
7391 7392 7393 7394 7395 7396
 *	@sizeof_priv:		size of private data to allocate space for
 *	@name:			device name format string
 *	@name_assign_type: 	origin of device name
 *	@setup:			callback to initialize device
 *	@txqs:			the number of TX subqueues to allocate
 *	@rxqs:			the number of RX subqueues to allocate
L
Linus Torvalds 已提交
7397 7398
 *
 *	Allocates a struct net_device with private data area for driver use
L
Li Zhong 已提交
7399
 *	and performs basic initialization.  Also allocates subqueue structs
T
Tom Herbert 已提交
7400
 *	for each queue on the device.
L
Linus Torvalds 已提交
7401
 */
T
Tom Herbert 已提交
7402
struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
7403
		unsigned char name_assign_type,
T
Tom Herbert 已提交
7404 7405
		void (*setup)(struct net_device *),
		unsigned int txqs, unsigned int rxqs)
L
Linus Torvalds 已提交
7406 7407
{
	struct net_device *dev;
7408
	size_t alloc_size;
7409
	struct net_device *p;
L
Linus Torvalds 已提交
7410

7411 7412
	BUG_ON(strlen(name) >= sizeof(dev->name));

T
Tom Herbert 已提交
7413
	if (txqs < 1) {
7414
		pr_err("alloc_netdev: Unable to allocate device with zero queues\n");
7415 7416 7417
		return NULL;
	}

7418
#ifdef CONFIG_SYSFS
T
Tom Herbert 已提交
7419
	if (rxqs < 1) {
7420
		pr_err("alloc_netdev: Unable to allocate device with zero RX queues\n");
T
Tom Herbert 已提交
7421 7422 7423 7424
		return NULL;
	}
#endif

7425
	alloc_size = sizeof(struct net_device);
7426 7427
	if (sizeof_priv) {
		/* ensure 32-byte alignment of private area */
7428
		alloc_size = ALIGN(alloc_size, NETDEV_ALIGN);
7429 7430 7431
		alloc_size += sizeof_priv;
	}
	/* ensure 32-byte alignment of whole construct */
7432
	alloc_size += NETDEV_ALIGN - 1;
L
Linus Torvalds 已提交
7433

7434 7435 7436
	p = kzalloc(alloc_size, GFP_KERNEL | __GFP_NOWARN | __GFP_REPEAT);
	if (!p)
		p = vzalloc(alloc_size);
7437
	if (!p)
L
Linus Torvalds 已提交
7438 7439
		return NULL;

7440
	dev = PTR_ALIGN(p, NETDEV_ALIGN);
L
Linus Torvalds 已提交
7441
	dev->padded = (char *)dev - (char *)p;
7442

E
Eric Dumazet 已提交
7443 7444
	dev->pcpu_refcnt = alloc_percpu(int);
	if (!dev->pcpu_refcnt)
7445
		goto free_dev;
7446 7447

	if (dev_addr_init(dev))
E
Eric Dumazet 已提交
7448
		goto free_pcpu;
7449

7450
	dev_mc_init(dev);
7451
	dev_uc_init(dev);
J
Jiri Pirko 已提交
7452

7453
	dev_net_set(dev, &init_net);
L
Linus Torvalds 已提交
7454

7455
	dev->gso_max_size = GSO_MAX_SIZE;
7456
	dev->gso_max_segs = GSO_MAX_SEGS;
7457 7458 7459

	INIT_LIST_HEAD(&dev->napi_list);
	INIT_LIST_HEAD(&dev->unreg_list);
7460
	INIT_LIST_HEAD(&dev->close_list);
7461
	INIT_LIST_HEAD(&dev->link_watch_list);
7462 7463 7464 7465
	INIT_LIST_HEAD(&dev->adj_list.upper);
	INIT_LIST_HEAD(&dev->adj_list.lower);
	INIT_LIST_HEAD(&dev->all_adj_list.upper);
	INIT_LIST_HEAD(&dev->all_adj_list.lower);
7466 7467
	INIT_LIST_HEAD(&dev->ptype_all);
	INIT_LIST_HEAD(&dev->ptype_specific);
7468
	dev->priv_flags = IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM;
7469 7470
	setup(dev);

7471
	if (!dev->tx_queue_len) {
7472
		dev->priv_flags |= IFF_NO_QUEUE;
7473 7474
		dev->tx_queue_len = 1;
	}
7475

T
Tom Herbert 已提交
7476 7477
	dev->num_tx_queues = txqs;
	dev->real_num_tx_queues = txqs;
7478
	if (netif_alloc_netdev_queues(dev))
7479
		goto free_all;
7480

7481
#ifdef CONFIG_SYSFS
T
Tom Herbert 已提交
7482 7483
	dev->num_rx_queues = rxqs;
	dev->real_num_rx_queues = rxqs;
T
Tom Herbert 已提交
7484
	if (netif_alloc_rx_queues(dev))
7485
		goto free_all;
E
Eric Dumazet 已提交
7486
#endif
T
Tom Herbert 已提交
7487

L
Linus Torvalds 已提交
7488
	strcpy(dev->name, name);
7489
	dev->name_assign_type = name_assign_type;
7490
	dev->group = INIT_NETDEV_GROUP;
7491 7492
	if (!dev->ethtool_ops)
		dev->ethtool_ops = &default_ethtool_ops;
7493 7494 7495

	nf_hook_ingress_init(dev);

L
Linus Torvalds 已提交
7496
	return dev;
7497

7498 7499 7500 7501
free_all:
	free_netdev(dev);
	return NULL;

E
Eric Dumazet 已提交
7502 7503
free_pcpu:
	free_percpu(dev->pcpu_refcnt);
7504 7505
free_dev:
	netdev_freemem(dev);
7506
	return NULL;
L
Linus Torvalds 已提交
7507
}
T
Tom Herbert 已提交
7508
EXPORT_SYMBOL(alloc_netdev_mqs);
L
Linus Torvalds 已提交
7509 7510 7511 7512 7513

/**
 *	free_netdev - free network device
 *	@dev: device
 *
7514 7515
 *	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
7516
 *	If this is the last reference then it will be freed.
7517
 *	Must be called in process context.
L
Linus Torvalds 已提交
7518 7519 7520
 */
void free_netdev(struct net_device *dev)
{
7521 7522
	struct napi_struct *p, *n;

7523
	might_sleep();
7524
	netif_free_tx_queues(dev);
7525
#ifdef CONFIG_SYSFS
7526
	kvfree(dev->_rx);
T
Tom Herbert 已提交
7527
#endif
7528

7529
	kfree(rcu_dereference_protected(dev->ingress_queue, 1));
7530

7531 7532 7533
	/* Flush device addresses */
	dev_addr_flush(dev);

7534 7535 7536
	list_for_each_entry_safe(p, n, &dev->napi_list, dev_list)
		netif_napi_del(p);

E
Eric Dumazet 已提交
7537 7538 7539
	free_percpu(dev->pcpu_refcnt);
	dev->pcpu_refcnt = NULL;

S
Stephen Hemminger 已提交
7540
	/*  Compatibility with error handling in drivers */
L
Linus Torvalds 已提交
7541
	if (dev->reg_state == NETREG_UNINITIALIZED) {
7542
		netdev_freemem(dev);
L
Linus Torvalds 已提交
7543 7544 7545 7546 7547 7548
		return;
	}

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

7549 7550
	/* will free via device release */
	put_device(&dev->dev);
L
Linus Torvalds 已提交
7551
}
E
Eric Dumazet 已提交
7552
EXPORT_SYMBOL(free_netdev);
7553

7554 7555 7556 7557 7558 7559
/**
 *	synchronize_net -  Synchronize with packet receive processing
 *
 *	Wait for packets currently being received to be done.
 *	Does not block later packets from starting.
 */
7560
void synchronize_net(void)
L
Linus Torvalds 已提交
7561 7562
{
	might_sleep();
7563 7564 7565 7566
	if (rtnl_is_locked())
		synchronize_rcu_expedited();
	else
		synchronize_rcu();
L
Linus Torvalds 已提交
7567
}
E
Eric Dumazet 已提交
7568
EXPORT_SYMBOL(synchronize_net);
L
Linus Torvalds 已提交
7569 7570

/**
7571
 *	unregister_netdevice_queue - remove device from the kernel
L
Linus Torvalds 已提交
7572
 *	@dev: device
7573
 *	@head: list
7574
 *
L
Linus Torvalds 已提交
7575
 *	This function shuts down a device interface and removes it
7576
 *	from the kernel tables.
7577
 *	If head not NULL, device is queued to be unregistered later.
L
Linus Torvalds 已提交
7578 7579 7580 7581 7582
 *
 *	Callers must hold the rtnl semaphore.  You may want
 *	unregister_netdev() instead of this.
 */

7583
void unregister_netdevice_queue(struct net_device *dev, struct list_head *head)
L
Linus Torvalds 已提交
7584
{
7585 7586
	ASSERT_RTNL();

7587
	if (head) {
7588
		list_move_tail(&dev->unreg_list, head);
7589 7590 7591 7592 7593
	} else {
		rollback_registered(dev);
		/* Finish processing unregister after unlock */
		net_set_todo(dev);
	}
L
Linus Torvalds 已提交
7594
}
7595
EXPORT_SYMBOL(unregister_netdevice_queue);
L
Linus Torvalds 已提交
7596

7597 7598 7599
/**
 *	unregister_netdevice_many - unregister many devices
 *	@head: list of devices
7600 7601 7602
 *
 *  Note: As most callers use a stack allocated list_head,
 *  we force a list_del() to make sure stack wont be corrupted later.
7603 7604 7605 7606 7607 7608 7609 7610 7611
 */
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);
7612
		list_del(head);
7613 7614
	}
}
7615
EXPORT_SYMBOL(unregister_netdevice_many);
7616

L
Linus Torvalds 已提交
7617 7618 7619 7620 7621
/**
 *	unregister_netdev - remove device from the kernel
 *	@dev: device
 *
 *	This function shuts down a device interface and removes it
7622
 *	from the kernel tables.
L
Linus Torvalds 已提交
7623 7624 7625 7626 7627 7628 7629 7630 7631 7632 7633 7634 7635
 *
 *	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);

7636 7637 7638 7639 7640 7641 7642 7643 7644 7645 7646 7647 7648 7649 7650 7651 7652 7653 7654 7655 7656 7657 7658 7659 7660 7661 7662 7663 7664 7665 7666
/**
 *	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 */
	if (dev->reg_state != NETREG_REGISTERED)
		goto out;

	/* Get out if there is nothing todo */
	err = 0;
7667
	if (net_eq(dev_net(dev), net))
7668 7669 7670 7671 7672 7673
		goto out;

	/* Pick the destination device name, and ensure
	 * we can use it in the destination network namespace.
	 */
	err = -EEXIST;
7674
	if (__dev_get_by_name(net, dev->name)) {
7675 7676 7677
		/* We get here if we can't use the current device name */
		if (!pat)
			goto out;
7678
		if (dev_get_valid_name(net, dev, pat) < 0)
7679 7680 7681 7682 7683 7684 7685 7686
			goto out;
	}

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

	/* If device is running close it first. */
7687
	dev_close(dev);
7688 7689 7690 7691 7692 7693 7694 7695 7696 7697 7698 7699

	/* 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.
7700 7701 7702 7703

	   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.
7704 7705
	*/
	call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
7706 7707
	rcu_barrier();
	call_netdevice_notifiers(NETDEV_UNREGISTER_FINAL, dev);
7708
	rtmsg_ifinfo(RTM_DELLINK, dev, ~0U, GFP_KERNEL);
7709 7710 7711 7712

	/*
	 *	Flush the unicast and multicast chains
	 */
7713
	dev_uc_flush(dev);
7714
	dev_mc_flush(dev);
7715

7716 7717
	/* Send a netdev-removed uevent to the old namespace */
	kobject_uevent(&dev->dev.kobj, KOBJ_REMOVE);
7718
	netdev_adjacent_del_links(dev);
7719

7720
	/* Actually switch the network namespace */
7721
	dev_net_set(dev, net);
7722 7723

	/* If there is an ifindex conflict assign a new one */
7724
	if (__dev_get_by_index(net, dev->ifindex))
7725 7726
		dev->ifindex = dev_new_index(net);

7727 7728
	/* Send a netdev-add uevent to the new namespace */
	kobject_uevent(&dev->dev.kobj, KOBJ_ADD);
7729
	netdev_adjacent_add_links(dev);
7730

7731
	/* Fixup kobjects */
7732
	err = device_rename(&dev->dev, dev->name);
7733
	WARN_ON(err);
7734 7735 7736 7737 7738 7739 7740

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

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

7741 7742 7743 7744
	/*
	 *	Prevent userspace races by waiting until the network
	 *	device is fully setup before sending notifications.
	 */
7745
	rtmsg_ifinfo(RTM_NEWLINK, dev, ~0U, GFP_KERNEL);
7746

7747 7748 7749 7750 7751
	synchronize_net();
	err = 0;
out:
	return err;
}
7752
EXPORT_SYMBOL_GPL(dev_change_net_namespace);
7753

L
Linus Torvalds 已提交
7754 7755 7756 7757 7758 7759 7760 7761 7762
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;

7763
	if (action != CPU_DEAD && action != CPU_DEAD_FROZEN)
L
Linus Torvalds 已提交
7764 7765 7766 7767 7768 7769 7770 7771 7772 7773 7774 7775 7776 7777 7778 7779
		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. */
7780 7781 7782 7783 7784 7785
	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;
	}
E
Eric Dumazet 已提交
7786 7787 7788 7789 7790 7791 7792 7793 7794 7795 7796 7797 7798 7799
	/* Append NAPI poll list from offline CPU, with one exception :
	 * process_backlog() must be called by cpu owning percpu backlog.
	 * We properly handle process_queue & input_pkt_queue later.
	 */
	while (!list_empty(&oldsd->poll_list)) {
		struct napi_struct *napi = list_first_entry(&oldsd->poll_list,
							    struct napi_struct,
							    poll_list);

		list_del_init(&napi->poll_list);
		if (napi->poll == process_backlog)
			napi->state = 0;
		else
			____napi_schedule(sd, napi);
7800
	}
L
Linus Torvalds 已提交
7801 7802 7803 7804 7805

	raise_softirq_irqoff(NET_TX_SOFTIRQ);
	local_irq_enable();

	/* Process offline CPU's input_pkt_queue */
7806
	while ((skb = __skb_dequeue(&oldsd->process_queue))) {
7807
		netif_rx_ni(skb);
7808
		input_queue_head_incr(oldsd);
T
Tom Herbert 已提交
7809
	}
E
Eric Dumazet 已提交
7810
	while ((skb = skb_dequeue(&oldsd->input_pkt_queue))) {
7811
		netif_rx_ni(skb);
7812 7813
		input_queue_head_incr(oldsd);
	}
L
Linus Torvalds 已提交
7814 7815 7816 7817 7818

	return NOTIFY_OK;
}


7819
/**
7820 7821 7822 7823
 *	netdev_increment_features - increment feature set by one
 *	@all: current feature set
 *	@one: new feature set
 *	@mask: mask feature set
7824 7825
 *
 *	Computes a new feature set after adding a device with feature set
7826 7827
 *	@one to the master device with current feature set @all.  Will not
 *	enable anything that is off in @mask. Returns the new feature set.
7828
 */
7829 7830
netdev_features_t netdev_increment_features(netdev_features_t all,
	netdev_features_t one, netdev_features_t mask)
7831
{
7832
	if (mask & NETIF_F_HW_CSUM)
7833
		mask |= NETIF_F_CSUM_MASK;
7834
	mask |= NETIF_F_VLAN_CHALLENGED;
7835

7836
	all |= one & (NETIF_F_ONE_FOR_ALL | NETIF_F_CSUM_MASK) & mask;
7837
	all &= one | ~NETIF_F_ALL_FOR_ALL;
7838

7839
	/* If one device supports hw checksumming, set for all. */
7840 7841
	if (all & NETIF_F_HW_CSUM)
		all &= ~(NETIF_F_CSUM_MASK & ~NETIF_F_HW_CSUM);
7842 7843 7844

	return all;
}
7845
EXPORT_SYMBOL(netdev_increment_features);
7846

7847
static struct hlist_head * __net_init netdev_create_hash(void)
7848 7849 7850 7851 7852 7853 7854 7855 7856 7857 7858 7859
{
	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;
}

7860
/* Initialize per network namespace state */
7861
static int __net_init netdev_init(struct net *net)
7862
{
7863 7864
	if (net != &init_net)
		INIT_LIST_HEAD(&net->dev_base_head);
7865

7866 7867 7868
	net->dev_name_head = netdev_create_hash();
	if (net->dev_name_head == NULL)
		goto err_name;
7869

7870 7871 7872
	net->dev_index_head = netdev_create_hash();
	if (net->dev_index_head == NULL)
		goto err_idx;
7873 7874

	return 0;
7875 7876 7877 7878 7879

err_idx:
	kfree(net->dev_name_head);
err_name:
	return -ENOMEM;
7880 7881
}

7882 7883 7884 7885 7886 7887
/**
 *	netdev_drivername - network driver for the device
 *	@dev: network device
 *
 *	Determine network driver for device.
 */
7888
const char *netdev_drivername(const struct net_device *dev)
7889
{
7890 7891
	const struct device_driver *driver;
	const struct device *parent;
7892
	const char *empty = "";
7893 7894 7895

	parent = dev->dev.parent;
	if (!parent)
7896
		return empty;
7897 7898 7899

	driver = parent->driver;
	if (driver && driver->name)
7900 7901
		return driver->name;
	return empty;
7902 7903
}

7904 7905
static void __netdev_printk(const char *level, const struct net_device *dev,
			    struct va_format *vaf)
7906
{
7907
	if (dev && dev->dev.parent) {
7908 7909 7910 7911 7912 7913 7914
		dev_printk_emit(level[1] - '0',
				dev->dev.parent,
				"%s %s %s%s: %pV",
				dev_driver_string(dev->dev.parent),
				dev_name(dev->dev.parent),
				netdev_name(dev), netdev_reg_state(dev),
				vaf);
7915
	} else if (dev) {
7916 7917
		printk("%s%s%s: %pV",
		       level, netdev_name(dev), netdev_reg_state(dev), vaf);
7918
	} else {
7919
		printk("%s(NULL net_device): %pV", level, vaf);
7920
	}
7921 7922
}

7923 7924
void netdev_printk(const char *level, const struct net_device *dev,
		   const char *format, ...)
7925 7926 7927 7928 7929 7930 7931 7932 7933
{
	struct va_format vaf;
	va_list args;

	va_start(args, format);

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

7934
	__netdev_printk(level, dev, &vaf);
7935

7936 7937 7938 7939 7940
	va_end(args);
}
EXPORT_SYMBOL(netdev_printk);

#define define_netdev_printk_level(func, level)			\
7941
void func(const struct net_device *dev, const char *fmt, ...)	\
7942 7943 7944 7945 7946 7947 7948 7949 7950
{								\
	struct va_format vaf;					\
	va_list args;						\
								\
	va_start(args, fmt);					\
								\
	vaf.fmt = fmt;						\
	vaf.va = &args;						\
								\
7951
	__netdev_printk(level, dev, &vaf);			\
7952
								\
7953 7954 7955 7956 7957 7958 7959 7960 7961 7962 7963 7964
	va_end(args);						\
}								\
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);

7965
static void __net_exit netdev_exit(struct net *net)
7966 7967 7968 7969 7970
{
	kfree(net->dev_name_head);
	kfree(net->dev_index_head);
}

7971
static struct pernet_operations __net_initdata netdev_net_ops = {
7972 7973 7974 7975
	.init = netdev_init,
	.exit = netdev_exit,
};

7976
static void __net_exit default_device_exit(struct net *net)
7977
{
7978
	struct net_device *dev, *aux;
7979
	/*
7980
	 * Push all migratable network devices back to the
7981 7982 7983
	 * initial network namespace
	 */
	rtnl_lock();
7984
	for_each_netdev_safe(net, dev, aux) {
7985
		int err;
7986
		char fb_name[IFNAMSIZ];
7987 7988 7989 7990 7991

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

7992 7993 7994
		/* Leave virtual devices for the generic cleanup */
		if (dev->rtnl_link_ops)
			continue;
7995

L
Lucas De Marchi 已提交
7996
		/* Push remaining network devices to init_net */
7997 7998
		snprintf(fb_name, IFNAMSIZ, "dev%d", dev->ifindex);
		err = dev_change_net_namespace(dev, &init_net, fb_name);
7999
		if (err) {
8000 8001
			pr_emerg("%s: failed to move %s to init_net: %d\n",
				 __func__, dev->name, err);
8002
			BUG();
8003 8004 8005 8006 8007
		}
	}
	rtnl_unlock();
}

8008 8009 8010 8011 8012 8013 8014
static void __net_exit rtnl_lock_unregistering(struct list_head *net_list)
{
	/* Return with the rtnl_lock held when there are no network
	 * devices unregistering in any network namespace in net_list.
	 */
	struct net *net;
	bool unregistering;
8015
	DEFINE_WAIT_FUNC(wait, woken_wake_function);
8016

8017
	add_wait_queue(&netdev_unregistering_wq, &wait);
8018 8019 8020 8021 8022 8023 8024 8025 8026 8027 8028 8029
	for (;;) {
		unregistering = false;
		rtnl_lock();
		list_for_each_entry(net, net_list, exit_list) {
			if (net->dev_unreg_count > 0) {
				unregistering = true;
				break;
			}
		}
		if (!unregistering)
			break;
		__rtnl_unlock();
8030 8031

		wait_woken(&wait, TASK_UNINTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT);
8032
	}
8033
	remove_wait_queue(&netdev_unregistering_wq, &wait);
8034 8035
}

8036 8037 8038
static void __net_exit default_device_exit_batch(struct list_head *net_list)
{
	/* At exit all network devices most be removed from a network
8039
	 * namespace.  Do this in the reverse order of registration.
8040 8041 8042 8043 8044 8045 8046
	 * 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);

8047 8048 8049 8050 8051 8052 8053 8054 8055 8056 8057 8058
	/* To prevent network device cleanup code from dereferencing
	 * loopback devices or network devices that have been freed
	 * wait here for all pending unregistrations to complete,
	 * before unregistring the loopback device and allowing the
	 * network namespace be freed.
	 *
	 * The netdev todo list containing all network devices
	 * unregistrations that happen in default_device_exit_batch
	 * will run in the rtnl_unlock() at the end of
	 * default_device_exit_batch.
	 */
	rtnl_lock_unregistering(net_list);
8059 8060
	list_for_each_entry(net, net_list, exit_list) {
		for_each_netdev_reverse(net, dev) {
8061
			if (dev->rtnl_link_ops && dev->rtnl_link_ops->dellink)
8062 8063 8064 8065 8066 8067 8068 8069 8070
				dev->rtnl_link_ops->dellink(dev, &dev_kill_list);
			else
				unregister_netdevice_queue(dev, &dev_kill_list);
		}
	}
	unregister_netdevice_many(&dev_kill_list);
	rtnl_unlock();
}

8071
static struct pernet_operations __net_initdata default_device_ops = {
8072
	.exit = default_device_exit,
8073
	.exit_batch = default_device_exit_batch,
8074 8075
};

L
Linus Torvalds 已提交
8076 8077 8078 8079 8080 8081 8082 8083 8084 8085 8086 8087 8088 8089 8090 8091 8092 8093 8094 8095
/*
 *	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;

8096
	if (netdev_kobject_init())
L
Linus Torvalds 已提交
8097 8098 8099
		goto out;

	INIT_LIST_HEAD(&ptype_all);
8100
	for (i = 0; i < PTYPE_HASH_SIZE; i++)
L
Linus Torvalds 已提交
8101 8102
		INIT_LIST_HEAD(&ptype_base[i]);

8103 8104
	INIT_LIST_HEAD(&offload_base);

8105 8106
	if (register_pernet_subsys(&netdev_net_ops))
		goto out;
L
Linus Torvalds 已提交
8107 8108 8109 8110 8111

	/*
	 *	Initialise the packet receive queues.
	 */

8112
	for_each_possible_cpu(i) {
E
Eric Dumazet 已提交
8113
		struct softnet_data *sd = &per_cpu(softnet_data, i);
L
Linus Torvalds 已提交
8114

E
Eric Dumazet 已提交
8115
		skb_queue_head_init(&sd->input_pkt_queue);
8116
		skb_queue_head_init(&sd->process_queue);
E
Eric Dumazet 已提交
8117
		INIT_LIST_HEAD(&sd->poll_list);
8118
		sd->output_queue_tailp = &sd->output_queue;
E
Eric Dumazet 已提交
8119
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
8120 8121 8122
		sd->csd.func = rps_trigger_softirq;
		sd->csd.info = sd;
		sd->cpu = i;
8123
#endif
T
Tom Herbert 已提交
8124

E
Eric Dumazet 已提交
8125 8126
		sd->backlog.poll = process_backlog;
		sd->backlog.weight = weight_p;
L
Linus Torvalds 已提交
8127 8128 8129 8130
	}

	dev_boot_phase = 0;

8131 8132 8133 8134 8135 8136 8137 8138 8139 8140 8141 8142 8143 8144 8145
	/* 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;

8146 8147
	open_softirq(NET_TX_SOFTIRQ, net_tx_action);
	open_softirq(NET_RX_SOFTIRQ, net_rx_action);
L
Linus Torvalds 已提交
8148 8149

	hotcpu_notifier(dev_cpu_callback, 0);
T
Thomas Graf 已提交
8150
	dst_subsys_init();
L
Linus Torvalds 已提交
8151 8152 8153 8154 8155 8156
	rc = 0;
out:
	return rc;
}

subsys_initcall(net_dev_init);