dev.c 119.2 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9
/*
 * 	NET3	Protocol independent device support routines.
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or (at your option) any later version.
 *
 *	Derived from the non IP parts of dev.c 1.0.19
10
 * 		Authors:	Ross Biro
L
Linus Torvalds 已提交
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77
 *				Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *				Mark Evans, <evansmp@uhura.aston.ac.uk>
 *
 *	Additional Authors:
 *		Florian la Roche <rzsfl@rz.uni-sb.de>
 *		Alan Cox <gw4pts@gw4pts.ampr.org>
 *		David Hinds <dahinds@users.sourceforge.net>
 *		Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
 *		Adam Sulmicki <adam@cfar.umd.edu>
 *              Pekka Riikonen <priikone@poesidon.pspt.fi>
 *
 *	Changes:
 *              D.J. Barrow     :       Fixed bug where dev->refcnt gets set
 *              			to 2 if register_netdev gets called
 *              			before net_dev_init & also removed a
 *              			few lines of code in the process.
 *		Alan Cox	:	device private ioctl copies fields back.
 *		Alan Cox	:	Transmit queue code does relevant
 *					stunts to keep the queue safe.
 *		Alan Cox	:	Fixed double lock.
 *		Alan Cox	:	Fixed promisc NULL pointer trap
 *		????????	:	Support the full private ioctl range
 *		Alan Cox	:	Moved ioctl permission check into
 *					drivers
 *		Tim Kordas	:	SIOCADDMULTI/SIOCDELMULTI
 *		Alan Cox	:	100 backlog just doesn't cut it when
 *					you start doing multicast video 8)
 *		Alan Cox	:	Rewrote net_bh and list manager.
 *		Alan Cox	: 	Fix ETH_P_ALL echoback lengths.
 *		Alan Cox	:	Took out transmit every packet pass
 *					Saved a few bytes in the ioctl handler
 *		Alan Cox	:	Network driver sets packet type before
 *					calling netif_rx. Saves a function
 *					call a packet.
 *		Alan Cox	:	Hashed net_bh()
 *		Richard Kooijman:	Timestamp fixes.
 *		Alan Cox	:	Wrong field in SIOCGIFDSTADDR
 *		Alan Cox	:	Device lock protection.
 *		Alan Cox	: 	Fixed nasty side effect of device close
 *					changes.
 *		Rudi Cilibrasi	:	Pass the right thing to
 *					set_mac_address()
 *		Dave Miller	:	32bit quantity for the device lock to
 *					make it work out on a Sparc.
 *		Bjorn Ekwall	:	Added KERNELD hack.
 *		Alan Cox	:	Cleaned up the backlog initialise.
 *		Craig Metz	:	SIOCGIFCONF fix if space for under
 *					1 device.
 *	    Thomas Bogendoerfer :	Return ENODEV for dev_open, if there
 *					is no device open function.
 *		Andi Kleen	:	Fix error reporting for SIOCGIFCONF
 *	    Michael Chastain	:	Fix signed/unsigned for SIOCGIFCONF
 *		Cyrus Durgin	:	Cleaned for KMOD
 *		Adam Sulmicki   :	Bug Fix : Network Device Unload
 *					A network device unload needs to purge
 *					the backlog queue.
 *	Paul Rusty Russell	:	SIOCSIFNAME
 *              Pekka Riikonen  :	Netdev boot-time settings code
 *              Andrew Morton   :       Make unregister_netdevice wait
 *              			indefinitely on dev->refcnt
 * 		J Hadi Salim	:	- Backlog queue sampling
 *				        - netif_rx() feedback
 */

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

131 132
#include "net-sysfs.h"

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

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

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

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

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

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

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

EXPORT_SYMBOL(dev_base_lock);

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

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

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

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

	ASSERT_RTNL();

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

		Protocol management and registration routines

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

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

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

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

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

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

	spin_lock_bh(&ptype_lock);

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

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

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

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

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

		      Device Boot-time Settings Routines

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

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

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

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

	return i >= NETDEV_BOOT_SETUP_MAX ? 0 : 1;
}

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

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


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

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

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

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

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

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

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

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

__setup("netdev=", netdev_boot_setup);

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

			    Device Interface Subroutines

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

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

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

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

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

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

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

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

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

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


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

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

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

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

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

	ASSERT_RTNL();

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

	return NULL;
L
Linus Torvalds 已提交
706 707
}

708 709
EXPORT_SYMBOL(dev_getbyhwaddr);

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

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

	return NULL;
720 721 722 723
}

EXPORT_SYMBOL(__dev_getfirstbyhwtype);

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

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

EXPORT_SYMBOL(dev_getfirstbyhwtype);

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

/**
 *	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.
 */
894
int dev_change_name(struct net_device *dev, const char *newname)
L
Linus Torvalds 已提交
895
{
896
	char oldname[IFNAMSIZ];
L
Linus Torvalds 已提交
897
	int err = 0;
898
	int ret;
899
	struct net *net;
L
Linus Torvalds 已提交
900 901

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

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

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

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

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

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

926
rollback:
927 928 929 930 931
	err = device_rename(&dev->dev, dev->name);
	if (err) {
		memcpy(dev->name, oldname, IFNAMSIZ);
		return err;
	}
932 933

	write_lock_bh(&dev_base_lock);
934
	hlist_del(&dev->name_hlist);
935
	hlist_add_head(&dev->name_hlist, dev_name_hash(net, dev->name));
936 937
	write_unlock_bh(&dev_base_lock);

938
	ret = call_netdevice_notifiers(NETDEV_CHANGENAME, dev);
939 940 941 942 943 944 945 946 947 948 949 950 951
	ret = notifier_to_errno(ret);

	if (ret) {
		if (err) {
			printk(KERN_ERR
			       "%s: name change rollback failed: %d.\n",
			       dev->name, ret);
		} else {
			err = ret;
			memcpy(dev->name, oldname, IFNAMSIZ);
			goto rollback;
		}
	}
L
Linus Torvalds 已提交
952 953 954 955

	return err;
}

956 957 958 959
/**
 *	dev_set_alias - change ifalias of a device
 *	@dev: device
 *	@alias: name up to IFALIASZ
960
 *	@len: limit of bytes to copy from info
961 962 963 964 965 966 967 968 969 970
 *
 *	Set ifalias for a device,
 */
int dev_set_alias(struct net_device *dev, const char *alias, size_t len)
{
	ASSERT_RTNL();

	if (len >= IFALIASZ)
		return -EINVAL;

971 972 973 974 975 976 977 978
	if (!len) {
		if (dev->ifalias) {
			kfree(dev->ifalias);
			dev->ifalias = NULL;
		}
		return 0;
	}

979 980 981 982 983 984 985 986 987
	dev->ifalias = krealloc(dev->ifalias, len+1, GFP_KERNEL);
	if (!dev->ifalias)
		return -ENOMEM;

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


988
/**
S
Stephen Hemminger 已提交
989
 *	netdev_features_change - device changes features
990 991 992 993 994 995
 *	@dev: device to cause notification
 *
 *	Called to indicate a device has changed features.
 */
void netdev_features_change(struct net_device *dev)
{
996
	call_netdevice_notifiers(NETDEV_FEAT_CHANGE, dev);
997 998 999
}
EXPORT_SYMBOL(netdev_features_change);

L
Linus Torvalds 已提交
1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010
/**
 *	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) {
1011
		call_netdevice_notifiers(NETDEV_CHANGE, dev);
L
Linus Torvalds 已提交
1012 1013 1014 1015
		rtmsg_ifinfo(RTM_NEWLINK, dev, 0);
	}
}

1016 1017 1018 1019 1020 1021
void netdev_bonding_change(struct net_device *dev)
{
	call_netdevice_notifiers(NETDEV_BONDING_FAILOVER, dev);
}
EXPORT_SYMBOL(netdev_bonding_change);

L
Linus Torvalds 已提交
1022 1023
/**
 *	dev_load 	- load a network module
1024
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
1025 1026 1027 1028 1029 1030 1031
 *	@name: name of interface
 *
 *	If a network interface is not present and the process has suitable
 *	privileges this function loads the module. If module loading is not
 *	available in this kernel then it becomes a nop.
 */

1032
void dev_load(struct net *net, const char *name)
L
Linus Torvalds 已提交
1033
{
1034
	struct net_device *dev;
L
Linus Torvalds 已提交
1035 1036

	read_lock(&dev_base_lock);
1037
	dev = __dev_get_by_name(net, name);
L
Linus Torvalds 已提交
1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059
	read_unlock(&dev_base_lock);

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

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

1060 1061
	ASSERT_RTNL();

L
Linus Torvalds 已提交
1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078
	/*
	 *	Is it already up?
	 */

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

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

	/*
	 *	Call device private open method
	 */
	set_bit(__LINK_STATE_START, &dev->state);
1079 1080 1081 1082 1083

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

	if (!ret && dev->open)
L
Linus Torvalds 已提交
1084 1085
		ret = dev->open(dev);

1086
	/*
L
Linus Torvalds 已提交
1087 1088 1089
	 *	If it went open OK then:
	 */

1090 1091 1092
	if (ret)
		clear_bit(__LINK_STATE_START, &dev->state);
	else {
L
Linus Torvalds 已提交
1093 1094 1095 1096 1097 1098 1099 1100
		/*
		 *	Set the flags.
		 */
		dev->flags |= IFF_UP;

		/*
		 *	Initialize multicasting status
		 */
1101
		dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
1102 1103 1104 1105 1106 1107 1108 1109 1110

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

		/*
		 *	... and announce new interface.
		 */
1111
		call_netdevice_notifiers(NETDEV_UP, dev);
L
Linus Torvalds 已提交
1112
	}
1113

L
Linus Torvalds 已提交
1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127
	return ret;
}

/**
 *	dev_close - shutdown an interface.
 *	@dev: device to shutdown
 *
 *	This function moves an active device into down state. A
 *	%NETDEV_GOING_DOWN is sent to the netdev notifier chain. The device
 *	is then deactivated and finally a %NETDEV_DOWN is sent to the notifier
 *	chain.
 */
int dev_close(struct net_device *dev)
{
1128 1129
	ASSERT_RTNL();

1130 1131
	might_sleep();

L
Linus Torvalds 已提交
1132 1133 1134 1135 1136 1137 1138
	if (!(dev->flags & IFF_UP))
		return 0;

	/*
	 *	Tell people we are going down, so that they can
	 *	prepare to death, when device is still operating.
	 */
1139
	call_netdevice_notifiers(NETDEV_GOING_DOWN, dev);
L
Linus Torvalds 已提交
1140 1141 1142 1143

	clear_bit(__LINK_STATE_START, &dev->state);

	/* Synchronize to scheduled poll. We cannot touch poll list,
1144 1145 1146 1147 1148
	 * it can be even on different cpu. So just clear netif_running().
	 *
	 * dev->stop() will invoke napi_disable() on all of it's
	 * napi_struct instances on this device.
	 */
L
Linus Torvalds 已提交
1149 1150
	smp_mb__after_clear_bit(); /* Commit netif_running(). */

1151 1152
	dev_deactivate(dev);

L
Linus Torvalds 已提交
1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171
	/*
	 *	Call the device specific close. This cannot fail.
	 *	Only if device is UP
	 *
	 *	We allow it to be called even after a DETACH hot-plug
	 *	event.
	 */
	if (dev->stop)
		dev->stop(dev);

	/*
	 *	Device is now down.
	 */

	dev->flags &= ~IFF_UP;

	/*
	 * Tell people we are down
	 */
1172
	call_netdevice_notifiers(NETDEV_DOWN, dev);
L
Linus Torvalds 已提交
1173 1174 1175 1176 1177

	return 0;
}


1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200
/**
 *	dev_disable_lro - disable Large Receive Offload on a device
 *	@dev: device
 *
 *	Disable Large Receive Offload (LRO) on a net device.  Must be
 *	called under RTNL.  This is needed if received packets may be
 *	forwarded to another interface.
 */
void dev_disable_lro(struct net_device *dev)
{
	if (dev->ethtool_ops && dev->ethtool_ops->get_flags &&
	    dev->ethtool_ops->set_flags) {
		u32 flags = dev->ethtool_ops->get_flags(dev);
		if (flags & ETH_FLAG_LRO) {
			flags &= ~ETH_FLAG_LRO;
			dev->ethtool_ops->set_flags(dev, flags);
		}
	}
	WARN_ON(dev->features & NETIF_F_LRO);
}
EXPORT_SYMBOL(dev_disable_lro);


1201 1202
static int dev_boot_phase = 1;

L
Linus Torvalds 已提交
1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217
/*
 *	Device change register/unregister. These are not inline or static
 *	as we export them to the world.
 */

/**
 *	register_netdevice_notifier - register a network notifier block
 *	@nb: notifier
 *
 *	Register a notifier to be called when network device events occur.
 *	The notifier passed is linked into the kernel structures and must
 *	not be reused until it has been unregistered. A negative errno code
 *	is returned on a failure.
 *
 * 	When registered all registration and up events are replayed
1218
 *	to the new notifier to allow device to have a race free
L
Linus Torvalds 已提交
1219 1220 1221 1222 1223 1224
 *	view of the network device list.
 */

int register_netdevice_notifier(struct notifier_block *nb)
{
	struct net_device *dev;
1225
	struct net_device *last;
1226
	struct net *net;
L
Linus Torvalds 已提交
1227 1228 1229
	int err;

	rtnl_lock();
1230
	err = raw_notifier_chain_register(&netdev_chain, nb);
1231 1232
	if (err)
		goto unlock;
1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243
	if (dev_boot_phase)
		goto unlock;
	for_each_net(net) {
		for_each_netdev(net, dev) {
			err = nb->notifier_call(nb, NETDEV_REGISTER, dev);
			err = notifier_to_errno(err);
			if (err)
				goto rollback;

			if (!(dev->flags & IFF_UP))
				continue;
L
Linus Torvalds 已提交
1244

1245 1246
			nb->notifier_call(nb, NETDEV_UP, dev);
		}
L
Linus Torvalds 已提交
1247
	}
1248 1249

unlock:
L
Linus Torvalds 已提交
1250 1251
	rtnl_unlock();
	return err;
1252 1253 1254

rollback:
	last = dev;
1255 1256 1257 1258
	for_each_net(net) {
		for_each_netdev(net, dev) {
			if (dev == last)
				break;
1259

1260 1261 1262 1263 1264
			if (dev->flags & IFF_UP) {
				nb->notifier_call(nb, NETDEV_GOING_DOWN, dev);
				nb->notifier_call(nb, NETDEV_DOWN, dev);
			}
			nb->notifier_call(nb, NETDEV_UNREGISTER, dev);
1265 1266
		}
	}
1267 1268

	raw_notifier_chain_unregister(&netdev_chain, nb);
1269
	goto unlock;
L
Linus Torvalds 已提交
1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283
}

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

int unregister_netdevice_notifier(struct notifier_block *nb)
{
1284 1285 1286
	int err;

	rtnl_lock();
1287
	err = raw_notifier_chain_unregister(&netdev_chain, nb);
1288 1289
	rtnl_unlock();
	return err;
L
Linus Torvalds 已提交
1290 1291 1292 1293 1294
}

/**
 *	call_netdevice_notifiers - call all network notifier blocks
 *      @val: value passed unmodified to notifier function
1295
 *      @dev: net_device pointer passed unmodified to notifier function
L
Linus Torvalds 已提交
1296 1297
 *
 *	Call all network notifier blocks.  Parameters and return value
1298
 *	are as for raw_notifier_call_chain().
L
Linus Torvalds 已提交
1299 1300
 */

1301
int call_netdevice_notifiers(unsigned long val, struct net_device *dev)
L
Linus Torvalds 已提交
1302
{
1303
	return raw_notifier_call_chain(&netdev_chain, val, dev);
L
Linus Torvalds 已提交
1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318
}

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

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

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

1319
static inline void net_timestamp(struct sk_buff *skb)
L
Linus Torvalds 已提交
1320 1321
{
	if (atomic_read(&netstamp_needed))
1322
		__net_timestamp(skb);
1323 1324
	else
		skb->tstamp.tv64 = 0;
L
Linus Torvalds 已提交
1325 1326 1327 1328 1329 1330 1331
}

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

1332
static void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev)
L
Linus Torvalds 已提交
1333 1334
{
	struct packet_type *ptype;
1335 1336

	net_timestamp(skb);
L
Linus Torvalds 已提交
1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353

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

			/* skb->nh should be correctly
			   set by sender, so that the second statement is
			   just protection against buggy protocols.
			 */
1354
			skb_reset_mac_header(skb2);
L
Linus Torvalds 已提交
1355

1356
			if (skb_network_header(skb2) < skb2->data ||
1357
			    skb2->network_header > skb2->tail) {
L
Linus Torvalds 已提交
1358 1359 1360 1361
				if (net_ratelimit())
					printk(KERN_CRIT "protocol %04x is "
					       "buggy, dev %s\n",
					       skb2->protocol, dev->name);
1362
				skb_reset_network_header(skb2);
L
Linus Torvalds 已提交
1363 1364
			}

1365
			skb2->transport_header = skb2->network_header;
L
Linus Torvalds 已提交
1366
			skb2->pkt_type = PACKET_OUTGOING;
D
David S. Miller 已提交
1367
			ptype->func(skb2, skb->dev, ptype, skb->dev);
L
Linus Torvalds 已提交
1368 1369 1370 1371 1372
		}
	}
	rcu_read_unlock();
}

1373

1374
static inline void __netif_reschedule(struct Qdisc *q)
1375
{
1376 1377
	struct softnet_data *sd;
	unsigned long flags;
1378

1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390
	local_irq_save(flags);
	sd = &__get_cpu_var(softnet_data);
	q->next_sched = sd->output_queue;
	sd->output_queue = q;
	raise_softirq_irqoff(NET_TX_SOFTIRQ);
	local_irq_restore(flags);
}

void __netif_schedule(struct Qdisc *q)
{
	if (!test_and_set_bit(__QDISC_STATE_SCHED, &q->state))
		__netif_reschedule(q);
1391 1392 1393
}
EXPORT_SYMBOL(__netif_schedule);

1394
void dev_kfree_skb_irq(struct sk_buff *skb)
1395
{
1396 1397 1398
	if (atomic_dec_and_test(&skb->users)) {
		struct softnet_data *sd;
		unsigned long flags;
1399

1400 1401 1402 1403 1404 1405 1406
		local_irq_save(flags);
		sd = &__get_cpu_var(softnet_data);
		skb->next = sd->completion_queue;
		sd->completion_queue = skb;
		raise_softirq_irqoff(NET_TX_SOFTIRQ);
		local_irq_restore(flags);
	}
1407
}
1408
EXPORT_SYMBOL(dev_kfree_skb_irq);
1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419

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


1420 1421 1422 1423 1424 1425
/**
 * netif_device_detach - mark device as removed
 * @dev: network device
 *
 * Mark device as removed from system and therefore no longer available.
 */
1426 1427 1428 1429 1430 1431 1432 1433 1434
void netif_device_detach(struct net_device *dev)
{
	if (test_and_clear_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
		netif_stop_queue(dev);
	}
}
EXPORT_SYMBOL(netif_device_detach);

1435 1436 1437 1438 1439 1440
/**
 * netif_device_attach - mark device as attached
 * @dev: network device
 *
 * Mark device as attached from system and restart if needed.
 */
1441 1442 1443 1444 1445
void netif_device_attach(struct net_device *dev)
{
	if (!test_and_set_bit(__LINK_STATE_PRESENT, &dev->state) &&
	    netif_running(dev)) {
		netif_wake_queue(dev);
1446
		__netdev_watchdog_up(dev);
1447 1448 1449 1450
	}
}
EXPORT_SYMBOL(netif_device_attach);

1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473
static bool can_checksum_protocol(unsigned long features, __be16 protocol)
{
	return ((features & NETIF_F_GEN_CSUM) ||
		((features & NETIF_F_IP_CSUM) &&
		 protocol == htons(ETH_P_IP)) ||
		((features & NETIF_F_IPV6_CSUM) &&
		 protocol == htons(ETH_P_IPV6)));
}

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

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

	return false;
}
1474

L
Linus Torvalds 已提交
1475 1476 1477 1478
/*
 * Invalidate hardware checksum when packet is to be mangled, and
 * complete checksum manually on outgoing path.
 */
1479
int skb_checksum_help(struct sk_buff *skb)
L
Linus Torvalds 已提交
1480
{
1481
	__wsum csum;
1482
	int ret = 0, offset;
L
Linus Torvalds 已提交
1483

1484
	if (skb->ip_summed == CHECKSUM_COMPLETE)
1485 1486 1487 1488 1489
		goto out_set_summed;

	if (unlikely(skb_shinfo(skb)->gso_size)) {
		/* Let GSO fix up the checksum. */
		goto out_set_summed;
L
Linus Torvalds 已提交
1490 1491
	}

1492 1493 1494 1495 1496 1497 1498 1499 1500
	offset = skb->csum_start - skb_headroom(skb);
	BUG_ON(offset >= skb_headlen(skb));
	csum = skb_checksum(skb, offset, skb->len - offset, 0);

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

	if (skb_cloned(skb) &&
	    !skb_clone_writable(skb, offset + sizeof(__sum16))) {
L
Linus Torvalds 已提交
1501 1502 1503 1504 1505
		ret = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
		if (ret)
			goto out;
	}

1506
	*(__sum16 *)(skb->data + offset) = csum_fold(csum);
1507
out_set_summed:
L
Linus Torvalds 已提交
1508
	skb->ip_summed = CHECKSUM_NONE;
1509
out:
L
Linus Torvalds 已提交
1510 1511 1512
	return ret;
}

1513 1514 1515
/**
 *	skb_gso_segment - Perform segmentation on skb.
 *	@skb: buffer to segment
1516
 *	@features: features for the output path (see dev->features)
1517 1518
 *
 *	This function segments the given skb and returns a list of segments.
1519 1520 1521
 *
 *	It may return NULL if the skb requires no segmentation.  This is
 *	only possible when GSO is used for verifying header integrity.
1522
 */
1523
struct sk_buff *skb_gso_segment(struct sk_buff *skb, int features)
1524 1525 1526
{
	struct sk_buff *segs = ERR_PTR(-EPROTONOSUPPORT);
	struct packet_type *ptype;
A
Al Viro 已提交
1527
	__be16 type = skb->protocol;
1528
	int err;
1529 1530 1531

	BUG_ON(skb_shinfo(skb)->frag_list);

1532
	skb_reset_mac_header(skb);
1533
	skb->mac_len = skb->network_header - skb->mac_header;
1534 1535
	__skb_pull(skb, skb->mac_len);

1536
	if (WARN_ON(skb->ip_summed != CHECKSUM_PARTIAL)) {
1537 1538 1539 1540 1541
		if (skb_header_cloned(skb) &&
		    (err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
			return ERR_PTR(err);
	}

1542
	rcu_read_lock();
1543 1544
	list_for_each_entry_rcu(ptype,
			&ptype_base[ntohs(type) & PTYPE_HASH_MASK], list) {
1545
		if (ptype->type == type && !ptype->dev && ptype->gso_segment) {
1546
			if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) {
1547 1548 1549 1550
				err = ptype->gso_send_check(skb);
				segs = ERR_PTR(err);
				if (err || skb_gso_ok(skb, features))
					break;
1551 1552
				__skb_push(skb, (skb->data -
						 skb_network_header(skb)));
1553
			}
1554
			segs = ptype->gso_segment(skb, features);
1555 1556 1557 1558 1559
			break;
		}
	}
	rcu_read_unlock();

1560
	__skb_push(skb, skb->data - skb_mac_header(skb));
1561

1562 1563 1564 1565 1566
	return segs;
}

EXPORT_SYMBOL(skb_gso_segment);

1567 1568 1569 1570 1571
/* Take action when hardware reception checksum errors are detected. */
#ifdef CONFIG_BUG
void netdev_rx_csum_fault(struct net_device *dev)
{
	if (net_ratelimit()) {
1572
		printk(KERN_ERR "%s: hw csum failure.\n",
1573
			dev ? dev->name : "<unknown>");
1574 1575 1576 1577 1578 1579
		dump_stack();
	}
}
EXPORT_SYMBOL(netdev_rx_csum_fault);
#endif

L
Linus Torvalds 已提交
1580 1581 1582 1583 1584 1585 1586
/* Actually, we should eliminate this check as soon as we know, that:
 * 1. IOMMU is present and allows to map all the memory.
 * 2. No high memory really exists on this machine.
 */

static inline int illegal_highdma(struct net_device *dev, struct sk_buff *skb)
{
1587
#ifdef CONFIG_HIGHMEM
L
Linus Torvalds 已提交
1588 1589 1590 1591 1592 1593 1594 1595 1596
	int i;

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

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

1597
#endif
L
Linus Torvalds 已提交
1598 1599 1600
	return 0;
}

1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634
struct dev_gso_cb {
	void (*destructor)(struct sk_buff *skb);
};

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

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

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

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

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

/**
 *	dev_gso_segment - Perform emulated hardware segmentation on skb.
 *	@skb: buffer to segment
 *
 *	This function segments the given skb and stores the list of segments
 *	in skb->next.
 */
static int dev_gso_segment(struct sk_buff *skb)
{
	struct net_device *dev = skb->dev;
	struct sk_buff *segs;
1635 1636 1637 1638 1639 1640 1641 1642
	int features = dev->features & ~(illegal_highdma(dev, skb) ?
					 NETIF_F_SG : 0);

	segs = skb_gso_segment(skb, features);

	/* Verifying header integrity only. */
	if (!segs)
		return 0;
1643

1644
	if (IS_ERR(segs))
1645 1646 1647 1648 1649 1650 1651 1652 1653
		return PTR_ERR(segs);

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

	return 0;
}

1654 1655
int dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
			struct netdev_queue *txq)
1656 1657
{
	if (likely(!skb->next)) {
S
Stephen Hemminger 已提交
1658
		if (!list_empty(&ptype_all))
1659 1660
			dev_queue_xmit_nit(skb, dev);

1661 1662 1663 1664 1665 1666
		if (netif_needs_gso(dev, skb)) {
			if (unlikely(dev_gso_segment(skb)))
				goto out_kfree_skb;
			if (skb->next)
				goto gso;
		}
1667

1668
		return dev->hard_start_xmit(skb, dev);
1669 1670
	}

1671
gso:
1672 1673 1674 1675 1676 1677 1678 1679
	do {
		struct sk_buff *nskb = skb->next;
		int rc;

		skb->next = nskb->next;
		nskb->next = NULL;
		rc = dev->hard_start_xmit(nskb, dev);
		if (unlikely(rc)) {
1680
			nskb->next = skb->next;
1681 1682 1683
			skb->next = nskb;
			return rc;
		}
1684
		if (unlikely(netif_tx_queue_stopped(txq) && skb->next))
1685
			return NETDEV_TX_BUSY;
1686
	} while (skb->next);
1687

1688 1689 1690 1691 1692 1693 1694
	skb->destructor = DEV_GSO_CB(skb)->destructor;

out_kfree_skb:
	kfree_skb(skb);
	return 0;
}

D
David S. Miller 已提交
1695 1696 1697
static u32 simple_tx_hashrnd;
static int simple_tx_hashrnd_initialized = 0;

1698 1699
static u16 simple_tx_hash(struct net_device *dev, struct sk_buff *skb)
{
D
David S. Miller 已提交
1700 1701
	u32 addr1, addr2, ports;
	u32 hash, ihl;
1702
	u8 ip_proto = 0;
D
David S. Miller 已提交
1703 1704 1705 1706 1707

	if (unlikely(!simple_tx_hashrnd_initialized)) {
		get_random_bytes(&simple_tx_hashrnd, 4);
		simple_tx_hashrnd_initialized = 1;
	}
1708 1709

	switch (skb->protocol) {
1710
	case htons(ETH_P_IP):
1711 1712
		if (!(ip_hdr(skb)->frag_off & htons(IP_MF | IP_OFFSET)))
			ip_proto = ip_hdr(skb)->protocol;
D
David S. Miller 已提交
1713 1714
		addr1 = ip_hdr(skb)->saddr;
		addr2 = ip_hdr(skb)->daddr;
1715 1716
		ihl = ip_hdr(skb)->ihl;
		break;
1717
	case htons(ETH_P_IPV6):
1718
		ip_proto = ipv6_hdr(skb)->nexthdr;
D
David S. Miller 已提交
1719 1720
		addr1 = ipv6_hdr(skb)->saddr.s6_addr32[3];
		addr2 = ipv6_hdr(skb)->daddr.s6_addr32[3];
1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735
		ihl = (40 >> 2);
		break;
	default:
		return 0;
	}


	switch (ip_proto) {
	case IPPROTO_TCP:
	case IPPROTO_UDP:
	case IPPROTO_DCCP:
	case IPPROTO_ESP:
	case IPPROTO_AH:
	case IPPROTO_SCTP:
	case IPPROTO_UDPLITE:
D
David S. Miller 已提交
1736
		ports = *((u32 *) (skb_network_header(skb) + (ihl * 4)));
1737 1738 1739
		break;

	default:
D
David S. Miller 已提交
1740
		ports = 0;
1741 1742 1743
		break;
	}

D
David S. Miller 已提交
1744 1745 1746
	hash = jhash_3words(addr1, addr2, ports, simple_tx_hashrnd);

	return (u16) (((u64) hash * dev->real_num_tx_queues) >> 32);
1747 1748
}

1749 1750 1751
static struct netdev_queue *dev_pick_tx(struct net_device *dev,
					struct sk_buff *skb)
{
1752 1753
	u16 queue_index = 0;

1754 1755
	if (dev->select_queue)
		queue_index = dev->select_queue(dev, skb);
1756 1757
	else if (dev->real_num_tx_queues > 1)
		queue_index = simple_tx_hash(dev, skb);
1758

1759 1760
	skb_set_queue_mapping(skb, queue_index);
	return netdev_get_tx_queue(dev, queue_index);
1761 1762
}

1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787
/**
 *	dev_queue_xmit - transmit a buffer
 *	@skb: buffer to transmit
 *
 *	Queue a buffer for transmission to a network device. The caller must
 *	have set the device and priority and built the buffer before calling
 *	this function. The function can be called from an interrupt.
 *
 *	A negative errno code is returned on a failure. A success does not
 *	guarantee the frame will be transmitted as it may be dropped due
 *	to congestion or traffic shaping.
 *
 * -----------------------------------------------------------------------------------
 *      I notice this method can also return errors from the queue disciplines,
 *      including NET_XMIT_DROP, which is a positive value.  So, errors can also
 *      be positive.
 *
 *      Regardless of the return value, the skb is consumed, so it is currently
 *      difficult to retry a send to this method.  (You can bump the ref count
 *      before sending to hold a reference for retry if you are careful.)
 *
 *      When calling this method, interrupts MUST be enabled.  This is because
 *      the BH enable code must have IRQs enabled so that it will not deadlock.
 *          --BLG
 */
L
Linus Torvalds 已提交
1788 1789 1790
int dev_queue_xmit(struct sk_buff *skb)
{
	struct net_device *dev = skb->dev;
1791
	struct netdev_queue *txq;
L
Linus Torvalds 已提交
1792 1793 1794
	struct Qdisc *q;
	int rc = -ENOMEM;

1795 1796 1797 1798
	/* GSO will handle the following emulations directly. */
	if (netif_needs_gso(dev, skb))
		goto gso;

L
Linus Torvalds 已提交
1799 1800
	if (skb_shinfo(skb)->frag_list &&
	    !(dev->features & NETIF_F_FRAGLIST) &&
H
Herbert Xu 已提交
1801
	    __skb_linearize(skb))
L
Linus Torvalds 已提交
1802 1803 1804 1805 1806 1807 1808 1809
		goto out_kfree_skb;

	/* Fragmented skb is linearized if device does not support SG,
	 * or if at least one of fragments is in highmem and device
	 * does not support DMA from it.
	 */
	if (skb_shinfo(skb)->nr_frags &&
	    (!(dev->features & NETIF_F_SG) || illegal_highdma(dev, skb)) &&
H
Herbert Xu 已提交
1810
	    __skb_linearize(skb))
L
Linus Torvalds 已提交
1811 1812 1813 1814 1815
		goto out_kfree_skb;

	/* If packet is not checksummed and device does not support
	 * checksumming for this protocol, complete checksumming here.
	 */
1816 1817 1818
	if (skb->ip_summed == CHECKSUM_PARTIAL) {
		skb_set_transport_header(skb, skb->csum_start -
					      skb_headroom(skb));
1819 1820
		if (!dev_can_checksum(dev, skb) && skb_checksum_help(skb))
			goto out_kfree_skb;
1821
	}
L
Linus Torvalds 已提交
1822

1823
gso:
1824 1825
	/* Disable soft irqs for various locks below. Also
	 * stops preemption for RCU.
L
Linus Torvalds 已提交
1826
	 */
1827
	rcu_read_lock_bh();
L
Linus Torvalds 已提交
1828

1829
	txq = dev_pick_tx(dev, skb);
1830
	q = rcu_dereference(txq->qdisc);
1831

L
Linus Torvalds 已提交
1832 1833 1834 1835
#ifdef CONFIG_NET_CLS_ACT
	skb->tc_verd = SET_TC_AT(skb->tc_verd,AT_EGRESS);
#endif
	if (q->enqueue) {
1836
		spinlock_t *root_lock = qdisc_lock(q);
1837 1838 1839

		spin_lock(root_lock);

1840
		if (unlikely(test_bit(__QDISC_STATE_DEACTIVATED, &q->state))) {
1841
			kfree_skb(skb);
1842
			rc = NET_XMIT_DROP;
1843 1844 1845
		} else {
			rc = qdisc_enqueue_root(skb, q);
			qdisc_run(q);
1846
		}
1847 1848 1849
		spin_unlock(root_lock);

		goto out;
L
Linus Torvalds 已提交
1850 1851 1852 1853 1854
	}

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

H
Herbert Xu 已提交
1855 1856
	   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
1857 1858 1859 1860 1861 1862 1863 1864 1865 1866
	   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 */

1867
		if (txq->xmit_lock_owner != cpu) {
L
Linus Torvalds 已提交
1868

1869
			HARD_TX_LOCK(dev, txq, cpu);
L
Linus Torvalds 已提交
1870

1871
			if (!netif_tx_queue_stopped(txq)) {
L
Linus Torvalds 已提交
1872
				rc = 0;
1873
				if (!dev_hard_start_xmit(skb, dev, txq)) {
1874
					HARD_TX_UNLOCK(dev, txq);
L
Linus Torvalds 已提交
1875 1876 1877
					goto out;
				}
			}
1878
			HARD_TX_UNLOCK(dev, txq);
L
Linus Torvalds 已提交
1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891
			if (net_ratelimit())
				printk(KERN_CRIT "Virtual device %s asks to "
				       "queue packet!\n", dev->name);
		} else {
			/* Recursion is detected! It is possible,
			 * unfortunately */
			if (net_ratelimit())
				printk(KERN_CRIT "Dead loop on virtual device "
				       "%s, fix it urgently!\n", dev->name);
		}
	}

	rc = -ENETDOWN;
1892
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1893 1894 1895 1896 1897

out_kfree_skb:
	kfree_skb(skb);
	return rc;
out:
1898
	rcu_read_unlock_bh();
L
Linus Torvalds 已提交
1899 1900 1901 1902 1903 1904 1905 1906
	return rc;
}


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

1907 1908 1909
int netdev_max_backlog __read_mostly = 1000;
int netdev_budget __read_mostly = 300;
int weight_p __read_mostly = 64;            /* old backlog weight */
L
Linus Torvalds 已提交
1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937

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


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

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

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

1938
	if (!skb->tstamp.tv64)
1939
		net_timestamp(skb);
L
Linus Torvalds 已提交
1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953

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

	__get_cpu_var(netdev_rx_stat).total++;
	if (queue->input_pkt_queue.qlen <= netdev_max_backlog) {
		if (queue->input_pkt_queue.qlen) {
enqueue:
			__skb_queue_tail(&queue->input_pkt_queue, skb);
			local_irq_restore(flags);
1954
			return NET_RX_SUCCESS;
L
Linus Torvalds 已提交
1955 1956
		}

1957
		napi_schedule(&queue->backlog);
L
Linus Torvalds 已提交
1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
		goto enqueue;
	}

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

	kfree_skb(skb);
	return NET_RX_DROP;
}

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

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

	return err;
}

EXPORT_SYMBOL(netif_rx_ni);

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

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

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

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

1999
			WARN_ON(atomic_read(&skb->users));
L
Linus Torvalds 已提交
2000 2001 2002 2003 2004
			__kfree_skb(skb);
		}
	}

	if (sd->output_queue) {
2005
		struct Qdisc *head;
L
Linus Torvalds 已提交
2006 2007 2008 2009 2010 2011 2012

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

		while (head) {
2013 2014 2015
			struct Qdisc *q = head;
			spinlock_t *root_lock;

L
Linus Torvalds 已提交
2016 2017
			head = head->next_sched;

2018
			root_lock = qdisc_lock(q);
2019
			if (spin_trylock(root_lock)) {
2020 2021 2022
				smp_mb__before_clear_bit();
				clear_bit(__QDISC_STATE_SCHED,
					  &q->state);
2023 2024
				qdisc_run(q);
				spin_unlock(root_lock);
L
Linus Torvalds 已提交
2025
			} else {
2026
				if (!test_bit(__QDISC_STATE_DEACTIVATED,
2027
					      &q->state)) {
2028
					__netif_reschedule(q);
2029 2030 2031 2032 2033
				} else {
					smp_mb__before_clear_bit();
					clear_bit(__QDISC_STATE_SCHED,
						  &q->state);
				}
L
Linus Torvalds 已提交
2034 2035 2036 2037 2038
			}
		}
	}
}

2039 2040 2041
static inline int deliver_skb(struct sk_buff *skb,
			      struct packet_type *pt_prev,
			      struct net_device *orig_dev)
L
Linus Torvalds 已提交
2042 2043
{
	atomic_inc(&skb->users);
D
David S. Miller 已提交
2044
	return pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2045 2046 2047
}

#if defined(CONFIG_BRIDGE) || defined (CONFIG_BRIDGE_MODULE)
2048
/* These hooks defined here for ATM */
L
Linus Torvalds 已提交
2049 2050 2051
struct net_bridge;
struct net_bridge_fdb_entry *(*br_fdb_get_hook)(struct net_bridge *br,
						unsigned char *addr);
2052
void (*br_fdb_put_hook)(struct net_bridge_fdb_entry *ent) __read_mostly;
L
Linus Torvalds 已提交
2053

2054 2055 2056 2057 2058 2059 2060 2061 2062
/*
 * If bridge module is loaded call bridging hook.
 *  returns NULL if packet was consumed.
 */
struct sk_buff *(*br_handle_frame_hook)(struct net_bridge_port *p,
					struct sk_buff *skb) __read_mostly;
static inline struct sk_buff *handle_bridge(struct sk_buff *skb,
					    struct packet_type **pt_prev, int *ret,
					    struct net_device *orig_dev)
L
Linus Torvalds 已提交
2063 2064 2065
{
	struct net_bridge_port *port;

2066 2067 2068
	if (skb->pkt_type == PACKET_LOOPBACK ||
	    (port = rcu_dereference(skb->dev->br_port)) == NULL)
		return skb;
L
Linus Torvalds 已提交
2069 2070

	if (*pt_prev) {
2071
		*ret = deliver_skb(skb, *pt_prev, orig_dev);
L
Linus Torvalds 已提交
2072
		*pt_prev = NULL;
2073 2074
	}

2075
	return br_handle_frame_hook(port, skb);
L
Linus Torvalds 已提交
2076 2077
}
#else
2078
#define handle_bridge(skb, pt_prev, ret, orig_dev)	(skb)
L
Linus Torvalds 已提交
2079 2080
#endif

P
Patrick McHardy 已提交
2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102
#if defined(CONFIG_MACVLAN) || defined(CONFIG_MACVLAN_MODULE)
struct sk_buff *(*macvlan_handle_frame_hook)(struct sk_buff *skb) __read_mostly;
EXPORT_SYMBOL_GPL(macvlan_handle_frame_hook);

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

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

L
Linus Torvalds 已提交
2103 2104 2105 2106 2107
#ifdef CONFIG_NET_CLS_ACT
/* TODO: Maybe we should just force sch_ingress to be compiled in
 * when CONFIG_NET_CLS_ACT is? otherwise some useless instructions
 * a compare and 2 stores extra right now if we dont have it on
 * but have CONFIG_NET_CLS_ACT
2108
 * NOTE: This doesnt stop any functionality; if you dont have
L
Linus Torvalds 已提交
2109 2110 2111
 * the ingress scheduler, you just cant add policies on ingress.
 *
 */
2112
static int ing_filter(struct sk_buff *skb)
L
Linus Torvalds 已提交
2113 2114
{
	struct net_device *dev = skb->dev;
2115
	u32 ttl = G_TC_RTTL(skb->tc_verd);
2116 2117 2118
	struct netdev_queue *rxq;
	int result = TC_ACT_OK;
	struct Qdisc *q;
2119

2120 2121 2122 2123 2124 2125
	if (MAX_RED_LOOP < ttl++) {
		printk(KERN_WARNING
		       "Redir loop detected Dropping packet (%d->%d)\n",
		       skb->iif, dev->ifindex);
		return TC_ACT_SHOT;
	}
L
Linus Torvalds 已提交
2126

2127 2128
	skb->tc_verd = SET_TC_RTTL(skb->tc_verd, ttl);
	skb->tc_verd = SET_TC_AT(skb->tc_verd, AT_INGRESS);
L
Linus Torvalds 已提交
2129

2130 2131
	rxq = &dev->rx_queue;

2132
	q = rxq->qdisc;
2133
	if (q != &noop_qdisc) {
2134
		spin_lock(qdisc_lock(q));
2135 2136
		if (likely(!test_bit(__QDISC_STATE_DEACTIVATED, &q->state)))
			result = qdisc_enqueue_root(skb, q);
2137 2138
		spin_unlock(qdisc_lock(q));
	}
2139 2140 2141

	return result;
}
2142

2143 2144 2145 2146
static inline struct sk_buff *handle_ing(struct sk_buff *skb,
					 struct packet_type **pt_prev,
					 int *ret, struct net_device *orig_dev)
{
2147
	if (skb->dev->rx_queue.qdisc == &noop_qdisc)
2148
		goto out;
L
Linus Torvalds 已提交
2149

2150 2151 2152 2153 2154 2155
	if (*pt_prev) {
		*ret = deliver_skb(skb, *pt_prev, orig_dev);
		*pt_prev = NULL;
	} else {
		/* Huh? Why does turning on AF_PACKET affect this? */
		skb->tc_verd = SET_TC_OK2MUNGE(skb->tc_verd);
L
Linus Torvalds 已提交
2156 2157
	}

2158 2159 2160 2161 2162 2163 2164 2165 2166 2167
	switch (ing_filter(skb)) {
	case TC_ACT_SHOT:
	case TC_ACT_STOLEN:
		kfree_skb(skb);
		return NULL;
	}

out:
	skb->tc_verd = 0;
	return skb;
L
Linus Torvalds 已提交
2168 2169 2170
}
#endif

2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197
/*
 * 	netif_nit_deliver - deliver received packets to network taps
 * 	@skb: buffer
 *
 * 	This function is used to deliver incoming packets to network
 * 	taps. It should be used when the normal netif_receive_skb path
 * 	is bypassed, for example because of VLAN acceleration.
 */
void netif_nit_deliver(struct sk_buff *skb)
{
	struct packet_type *ptype;

	if (list_empty(&ptype_all))
		return;

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

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

2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212
/**
 *	netif_receive_skb - process receive buffer from network
 *	@skb: buffer to process
 *
 *	netif_receive_skb() is the main receive data processing function.
 *	It always succeeds. The buffer may be dropped during processing
 *	for congestion control or by the protocol layers.
 *
 *	This function may only be called from softirq context and interrupts
 *	should be enabled.
 *
 *	Return values (usually ignored):
 *	NET_RX_SUCCESS: no congestion
 *	NET_RX_DROP: packet was dropped
 */
L
Linus Torvalds 已提交
2213 2214 2215
int netif_receive_skb(struct sk_buff *skb)
{
	struct packet_type *ptype, *pt_prev;
D
David S. Miller 已提交
2216
	struct net_device *orig_dev;
2217
	struct net_device *null_or_orig;
L
Linus Torvalds 已提交
2218
	int ret = NET_RX_DROP;
A
Al Viro 已提交
2219
	__be16 type;
L
Linus Torvalds 已提交
2220 2221

	/* if we've gotten here through NAPI, check netpoll */
2222
	if (netpoll_receive_skb(skb))
L
Linus Torvalds 已提交
2223 2224
		return NET_RX_DROP;

2225
	if (!skb->tstamp.tv64)
2226
		net_timestamp(skb);
L
Linus Torvalds 已提交
2227

2228 2229
	if (!skb->iif)
		skb->iif = skb->dev->ifindex;
2230

2231
	null_or_orig = NULL;
J
Joe Eykholt 已提交
2232 2233
	orig_dev = skb->dev;
	if (orig_dev->master) {
2234 2235 2236 2237
		if (skb_bond_should_drop(skb))
			null_or_orig = orig_dev; /* deliver only exact match */
		else
			skb->dev = orig_dev->master;
J
Joe Eykholt 已提交
2238
	}
2239

L
Linus Torvalds 已提交
2240 2241
	__get_cpu_var(netdev_rx_stat).total++;

2242
	skb_reset_network_header(skb);
2243
	skb_reset_transport_header(skb);
2244
	skb->mac_len = skb->network_header - skb->mac_header;
L
Linus Torvalds 已提交
2245 2246 2247 2248 2249

	pt_prev = NULL;

	rcu_read_lock();

2250 2251 2252 2253
	/* Don't receive packets in an exiting network namespace */
	if (!net_alive(dev_net(skb->dev)))
		goto out;

L
Linus Torvalds 已提交
2254 2255 2256 2257 2258 2259 2260 2261
#ifdef CONFIG_NET_CLS_ACT
	if (skb->tc_verd & TC_NCLS) {
		skb->tc_verd = CLR_TC_NCLS(skb->tc_verd);
		goto ncls;
	}
#endif

	list_for_each_entry_rcu(ptype, &ptype_all, list) {
2262 2263
		if (ptype->dev == null_or_orig || ptype->dev == skb->dev ||
		    ptype->dev == orig_dev) {
2264
			if (pt_prev)
D
David S. Miller 已提交
2265
				ret = deliver_skb(skb, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2266 2267 2268 2269 2270
			pt_prev = ptype;
		}
	}

#ifdef CONFIG_NET_CLS_ACT
2271 2272
	skb = handle_ing(skb, &pt_prev, &ret, orig_dev);
	if (!skb)
L
Linus Torvalds 已提交
2273 2274 2275 2276
		goto out;
ncls:
#endif

2277
	skb = handle_bridge(skb, &pt_prev, &ret, orig_dev);
P
Patrick McHardy 已提交
2278 2279 2280
	if (!skb)
		goto out;
	skb = handle_macvlan(skb, &pt_prev, &ret, orig_dev);
2281
	if (!skb)
L
Linus Torvalds 已提交
2282 2283 2284
		goto out;

	type = skb->protocol;
2285 2286
	list_for_each_entry_rcu(ptype,
			&ptype_base[ntohs(type) & PTYPE_HASH_MASK], list) {
L
Linus Torvalds 已提交
2287
		if (ptype->type == type &&
2288 2289
		    (ptype->dev == null_or_orig || ptype->dev == skb->dev ||
		     ptype->dev == orig_dev)) {
2290
			if (pt_prev)
D
David S. Miller 已提交
2291
				ret = deliver_skb(skb, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2292 2293 2294 2295 2296
			pt_prev = ptype;
		}
	}

	if (pt_prev) {
D
David S. Miller 已提交
2297
		ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
L
Linus Torvalds 已提交
2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310
	} else {
		kfree_skb(skb);
		/* Jamal, now you will not able to escape explaining
		 * me how you were going to use this. :-)
		 */
		ret = NET_RX_DROP;
	}

out:
	rcu_read_unlock();
	return ret;
}

2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324
/* Network device is going away, flush any packets still pending  */
static void flush_backlog(void *arg)
{
	struct net_device *dev = arg;
	struct softnet_data *queue = &__get_cpu_var(softnet_data);
	struct sk_buff *skb, *tmp;

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

2325
static int process_backlog(struct napi_struct *napi, int quota)
L
Linus Torvalds 已提交
2326 2327 2328 2329 2330
{
	int work = 0;
	struct softnet_data *queue = &__get_cpu_var(softnet_data);
	unsigned long start_time = jiffies;

2331 2332
	napi->weight = weight_p;
	do {
L
Linus Torvalds 已提交
2333 2334 2335 2336
		struct sk_buff *skb;

		local_irq_disable();
		skb = __skb_dequeue(&queue->input_pkt_queue);
2337 2338 2339 2340 2341
		if (!skb) {
			__napi_complete(napi);
			local_irq_enable();
			break;
		}
L
Linus Torvalds 已提交
2342 2343 2344
		local_irq_enable();

		netif_receive_skb(skb);
2345
	} while (++work < quota && jiffies == start_time);
L
Linus Torvalds 已提交
2346

2347 2348
	return work;
}
L
Linus Torvalds 已提交
2349

2350 2351
/**
 * __napi_schedule - schedule for receive
2352
 * @n: entry to schedule
2353 2354 2355
 *
 * The entry's receive function will be scheduled to run
 */
H
Harvey Harrison 已提交
2356
void __napi_schedule(struct napi_struct *n)
2357 2358
{
	unsigned long flags;
L
Linus Torvalds 已提交
2359

2360 2361 2362 2363
	local_irq_save(flags);
	list_add_tail(&n->poll_list, &__get_cpu_var(softnet_data).poll_list);
	__raise_softirq_irqoff(NET_RX_SOFTIRQ);
	local_irq_restore(flags);
L
Linus Torvalds 已提交
2364
}
2365 2366
EXPORT_SYMBOL(__napi_schedule);

L
Linus Torvalds 已提交
2367 2368 2369

static void net_rx_action(struct softirq_action *h)
{
2370
	struct list_head *list = &__get_cpu_var(softnet_data).poll_list;
L
Linus Torvalds 已提交
2371
	unsigned long start_time = jiffies;
2372
	int budget = netdev_budget;
2373 2374
	void *have;

L
Linus Torvalds 已提交
2375 2376
	local_irq_disable();

2377 2378 2379
	while (!list_empty(list)) {
		struct napi_struct *n;
		int work, weight;
L
Linus Torvalds 已提交
2380

2381 2382 2383 2384 2385 2386 2387 2388
		/* If softirq window is exhuasted then punt.
		 *
		 * Note that this is a slight policy change from the
		 * previous NAPI code, which would allow up to 2
		 * jiffies to pass before breaking out.  The test
		 * used to be "jiffies - start_time > 1".
		 */
		if (unlikely(budget <= 0 || jiffies != start_time))
L
Linus Torvalds 已提交
2389 2390 2391 2392
			goto softnet_break;

		local_irq_enable();

2393 2394 2395 2396 2397 2398
		/* Even though interrupts have been re-enabled, this
		 * access is safe because interrupts can only add new
		 * entries to the tail of this list, and only ->poll()
		 * calls can remove this head entry from the list.
		 */
		n = list_entry(list->next, struct napi_struct, poll_list);
L
Linus Torvalds 已提交
2399

2400 2401 2402 2403
		have = netpoll_poll_lock(n);

		weight = n->weight;

2404 2405 2406 2407 2408 2409 2410 2411 2412
		/* This NAPI_STATE_SCHED test is for avoiding a race
		 * with netpoll's poll_napi().  Only the entity which
		 * obtains the lock and sees NAPI_STATE_SCHED set will
		 * actually make the ->poll() call.  Therefore we avoid
		 * accidently calling ->poll() when NAPI is not scheduled.
		 */
		work = 0;
		if (test_bit(NAPI_STATE_SCHED, &n->state))
			work = n->poll(n, weight);
2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424

		WARN_ON_ONCE(work > weight);

		budget -= work;

		local_irq_disable();

		/* Drivers must not modify the NAPI state if they
		 * consume the entire weight.  In such cases this code
		 * still "owns" the NAPI instance and therefore can
		 * move the instance around on the list at-will.
		 */
2425 2426 2427 2428 2429 2430
		if (unlikely(work == weight)) {
			if (unlikely(napi_disable_pending(n)))
				__napi_complete(n);
			else
				list_move_tail(&n->poll_list, list);
		}
2431 2432

		netpoll_poll_unlock(have);
L
Linus Torvalds 已提交
2433 2434
	}
out:
2435
	local_irq_enable();
2436

2437 2438 2439 2440 2441
#ifdef CONFIG_NET_DMA
	/*
	 * There may not be any more sk_buffs coming right now, so push
	 * any pending DMA copies to hardware
	 */
2442 2443
	if (!cpus_empty(net_dma.channel_mask)) {
		int chan_idx;
2444
		for_each_cpu_mask_nr(chan_idx, net_dma.channel_mask) {
2445 2446 2447 2448
			struct dma_chan *chan = net_dma.channels[chan_idx];
			if (chan)
				dma_async_memcpy_issue_pending(chan);
		}
2449 2450
	}
#endif
2451

L
Linus Torvalds 已提交
2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490
	return;

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

static gifconf_func_t * gifconf_list [NPROTO];

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


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

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

2491
static int dev_ifname(struct net *net, struct ifreq __user *arg)
L
Linus Torvalds 已提交
2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503
{
	struct net_device *dev;
	struct ifreq ifr;

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

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

	read_lock(&dev_base_lock);
2504
	dev = __dev_get_by_index(net, ifr.ifr_ifindex);
L
Linus Torvalds 已提交
2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523
	if (!dev) {
		read_unlock(&dev_base_lock);
		return -ENODEV;
	}

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

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

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

2524
static int dev_ifconf(struct net *net, char __user *arg)
L
Linus Torvalds 已提交
2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547
{
	struct ifconf ifc;
	struct net_device *dev;
	char __user *pos;
	int len;
	int total;
	int i;

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

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

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

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

	total = 0;
2548
	for_each_netdev(net, dev) {
L
Linus Torvalds 已提交
2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561
		for (i = 0; i < NPROTO; i++) {
			if (gifconf_list[i]) {
				int done;
				if (!pos)
					done = gifconf_list[i](dev, NULL, 0);
				else
					done = gifconf_list[i](dev, pos + total,
							       len - total);
				if (done < 0)
					return -EFAULT;
				total += done;
			}
		}
2562
	}
L
Linus Torvalds 已提交
2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579

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

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

#ifdef CONFIG_PROC_FS
/*
 *	This is invoked by the /proc filesystem handler to display a device
 *	in detail.
 */
2580
void *dev_seq_start(struct seq_file *seq, loff_t *pos)
2581
	__acquires(dev_base_lock)
L
Linus Torvalds 已提交
2582
{
2583
	struct net *net = seq_file_net(seq);
2584
	loff_t off;
L
Linus Torvalds 已提交
2585 2586
	struct net_device *dev;

2587 2588 2589
	read_lock(&dev_base_lock);
	if (!*pos)
		return SEQ_START_TOKEN;
L
Linus Torvalds 已提交
2590

2591
	off = 1;
2592
	for_each_netdev(net, dev)
2593 2594
		if (off++ == *pos)
			return dev;
L
Linus Torvalds 已提交
2595

2596
	return NULL;
L
Linus Torvalds 已提交
2597 2598 2599 2600
}

void *dev_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
2601
	struct net *net = seq_file_net(seq);
L
Linus Torvalds 已提交
2602
	++*pos;
2603
	return v == SEQ_START_TOKEN ?
2604
		first_net_device(net) : next_net_device((struct net_device *)v);
L
Linus Torvalds 已提交
2605 2606 2607
}

void dev_seq_stop(struct seq_file *seq, void *v)
2608
	__releases(dev_base_lock)
L
Linus Torvalds 已提交
2609 2610 2611 2612 2613 2614
{
	read_unlock(&dev_base_lock);
}

static void dev_seq_printf_stats(struct seq_file *seq, struct net_device *dev)
{
R
Rusty Russell 已提交
2615
	struct net_device_stats *stats = dev->get_stats(dev);
L
Linus Torvalds 已提交
2616

2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633
	seq_printf(seq, "%6s:%8lu %7lu %4lu %4lu %4lu %5lu %10lu %9lu "
		   "%8lu %7lu %4lu %4lu %4lu %5lu %7lu %10lu\n",
		   dev->name, stats->rx_bytes, stats->rx_packets,
		   stats->rx_errors,
		   stats->rx_dropped + stats->rx_missed_errors,
		   stats->rx_fifo_errors,
		   stats->rx_length_errors + stats->rx_over_errors +
		    stats->rx_crc_errors + stats->rx_frame_errors,
		   stats->rx_compressed, stats->multicast,
		   stats->tx_bytes, stats->tx_packets,
		   stats->tx_errors, stats->tx_dropped,
		   stats->tx_fifo_errors, stats->collisions,
		   stats->tx_carrier_errors +
		    stats->tx_aborted_errors +
		    stats->tx_window_errors +
		    stats->tx_heartbeat_errors,
		   stats->tx_compressed);
L
Linus Torvalds 已提交
2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656
}

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

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

2657
	while (*pos < nr_cpu_ids)
2658
		if (cpu_online(*pos)) {
L
Linus Torvalds 已提交
2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685
			rc = &per_cpu(netdev_rx_stat, *pos);
			break;
		} else
			++*pos;
	return rc;
}

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

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

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

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

	seq_printf(seq, "%08x %08x %08x %08x %08x %08x %08x %08x %08x\n",
2686
		   s->total, s->dropped, s->time_squeeze, 0,
2687 2688
		   0, 0, 0, 0, /* was fastroute */
		   s->cpu_collision );
L
Linus Torvalds 已提交
2689 2690 2691
	return 0;
}

2692
static const struct seq_operations dev_seq_ops = {
L
Linus Torvalds 已提交
2693 2694 2695 2696 2697 2698 2699 2700
	.start = dev_seq_start,
	.next  = dev_seq_next,
	.stop  = dev_seq_stop,
	.show  = dev_seq_show,
};

static int dev_seq_open(struct inode *inode, struct file *file)
{
2701 2702
	return seq_open_net(inode, file, &dev_seq_ops,
			    sizeof(struct seq_net_private));
L
Linus Torvalds 已提交
2703 2704
}

2705
static const struct file_operations dev_seq_fops = {
L
Linus Torvalds 已提交
2706 2707 2708 2709
	.owner	 = THIS_MODULE,
	.open    = dev_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
2710
	.release = seq_release_net,
L
Linus Torvalds 已提交
2711 2712
};

2713
static const struct seq_operations softnet_seq_ops = {
L
Linus Torvalds 已提交
2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724
	.start = softnet_seq_start,
	.next  = softnet_seq_next,
	.stop  = softnet_seq_stop,
	.show  = softnet_seq_show,
};

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

2725
static const struct file_operations softnet_seq_fops = {
L
Linus Torvalds 已提交
2726 2727 2728 2729 2730 2731 2732
	.owner	 = THIS_MODULE,
	.open    = softnet_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release,
};

2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744
static void *ptype_get_idx(loff_t pos)
{
	struct packet_type *pt = NULL;
	loff_t i = 0;
	int t;

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

2745
	for (t = 0; t < PTYPE_HASH_SIZE; t++) {
2746 2747 2748 2749 2750 2751 2752 2753 2754 2755
		list_for_each_entry_rcu(pt, &ptype_base[t], list) {
			if (i == pos)
				return pt;
			++i;
		}
	}
	return NULL;
}

static void *ptype_seq_start(struct seq_file *seq, loff_t *pos)
2756
	__acquires(RCU)
2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779
{
	rcu_read_lock();
	return *pos ? ptype_get_idx(*pos - 1) : SEQ_START_TOKEN;
}

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

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

	pt = v;
	nxt = pt->list.next;
	if (pt->type == htons(ETH_P_ALL)) {
		if (nxt != &ptype_all)
			goto found;
		hash = 0;
		nxt = ptype_base[0].next;
	} else
2780
		hash = ntohs(pt->type) & PTYPE_HASH_MASK;
2781 2782

	while (nxt == &ptype_base[hash]) {
2783
		if (++hash >= PTYPE_HASH_SIZE)
2784 2785 2786 2787 2788 2789 2790 2791
			return NULL;
		nxt = ptype_base[hash].next;
	}
found:
	return list_entry(nxt, struct packet_type, list);
}

static void ptype_seq_stop(struct seq_file *seq, void *v)
2792
	__releases(RCU)
2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827
{
	rcu_read_unlock();
}

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

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

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

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

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

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

	if (v == SEQ_START_TOKEN)
		seq_puts(seq, "Type Device      Function\n");
2828
	else if (pt->dev == NULL || dev_net(pt->dev) == seq_file_net(seq)) {
2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851
		if (pt->type == htons(ETH_P_ALL))
			seq_puts(seq, "ALL ");
		else
			seq_printf(seq, "%04x", ntohs(pt->type));

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

	return 0;
}

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

static int ptype_seq_open(struct inode *inode, struct file *file)
{
2852 2853
	return seq_open_net(inode, file, &ptype_seq_ops,
			sizeof(struct seq_net_private));
2854 2855 2856 2857 2858 2859 2860
}

static const struct file_operations ptype_seq_fops = {
	.owner	 = THIS_MODULE,
	.open    = ptype_seq_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
2861
	.release = seq_release_net,
2862 2863 2864
};


2865
static int __net_init dev_proc_net_init(struct net *net)
L
Linus Torvalds 已提交
2866 2867 2868
{
	int rc = -ENOMEM;

2869
	if (!proc_net_fops_create(net, "dev", S_IRUGO, &dev_seq_fops))
L
Linus Torvalds 已提交
2870
		goto out;
2871
	if (!proc_net_fops_create(net, "softnet_stat", S_IRUGO, &softnet_seq_fops))
L
Linus Torvalds 已提交
2872
		goto out_dev;
2873
	if (!proc_net_fops_create(net, "ptype", S_IRUGO, &ptype_seq_fops))
2874
		goto out_softnet;
2875

2876
	if (wext_proc_init(net))
2877
		goto out_ptype;
L
Linus Torvalds 已提交
2878 2879 2880
	rc = 0;
out:
	return rc;
2881
out_ptype:
2882
	proc_net_remove(net, "ptype");
L
Linus Torvalds 已提交
2883
out_softnet:
2884
	proc_net_remove(net, "softnet_stat");
L
Linus Torvalds 已提交
2885
out_dev:
2886
	proc_net_remove(net, "dev");
L
Linus Torvalds 已提交
2887 2888
	goto out;
}
2889

2890
static void __net_exit dev_proc_net_exit(struct net *net)
2891 2892 2893 2894 2895 2896 2897 2898
{
	wext_proc_exit(net);

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

2899
static struct pernet_operations __net_initdata dev_proc_ops = {
2900 2901 2902 2903 2904 2905 2906 2907
	.init = dev_proc_net_init,
	.exit = dev_proc_net_exit,
};

static int __init dev_proc_init(void)
{
	return register_pernet_subsys(&dev_proc_ops);
}
L
Linus Torvalds 已提交
2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936
#else
#define dev_proc_init() 0
#endif	/* CONFIG_PROC_FS */


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

	ASSERT_RTNL();

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

	slave->master = master;
2937

L
Linus Torvalds 已提交
2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951
	synchronize_net();

	if (old)
		dev_put(old);

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

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

2952 2953 2954 2955 2956 2957
static void dev_change_rx_flags(struct net_device *dev, int flags)
{
	if (dev->flags & IFF_UP && dev->change_rx_flags)
		dev->change_rx_flags(dev, flags);
}

2958
static int __dev_set_promiscuity(struct net_device *dev, int inc)
L
Linus Torvalds 已提交
2959 2960 2961
{
	unsigned short old_flags = dev->flags;

2962 2963
	ASSERT_RTNL();

2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980
	dev->flags |= IFF_PROMISC;
	dev->promiscuity += inc;
	if (dev->promiscuity == 0) {
		/*
		 * Avoid overflow.
		 * If inc causes overflow, untouch promisc and return error.
		 */
		if (inc < 0)
			dev->flags &= ~IFF_PROMISC;
		else {
			dev->promiscuity -= inc;
			printk(KERN_WARNING "%s: promiscuity touches roof, "
				"set promiscuity failed, promiscuity feature "
				"of device might be broken.\n", dev->name);
			return -EOVERFLOW;
		}
	}
2981
	if (dev->flags != old_flags) {
L
Linus Torvalds 已提交
2982 2983
		printk(KERN_INFO "device %s %s promiscuous mode\n",
		       dev->name, (dev->flags & IFF_PROMISC) ? "entered" :
2984
							       "left");
2985 2986 2987 2988 2989 2990 2991 2992 2993
		if (audit_enabled)
			audit_log(current->audit_context, GFP_ATOMIC,
				AUDIT_ANOM_PROMISCUOUS,
				"dev=%s prom=%d old_prom=%d auid=%u uid=%u gid=%u ses=%u",
				dev->name, (dev->flags & IFF_PROMISC),
				(old_flags & IFF_PROMISC),
				audit_get_loginuid(current),
				current->uid, current->gid,
				audit_get_sessionid(current));
2994

2995
		dev_change_rx_flags(dev, IFF_PROMISC);
L
Linus Torvalds 已提交
2996
	}
2997
	return 0;
L
Linus Torvalds 已提交
2998 2999
}

3000 3001 3002 3003 3004 3005 3006 3007 3008
/**
 *	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.
3009
 *	Return 0 if successful or a negative errno code on error.
3010
 */
3011
int dev_set_promiscuity(struct net_device *dev, int inc)
3012 3013
{
	unsigned short old_flags = dev->flags;
3014
	int err;
3015

3016
	err = __dev_set_promiscuity(dev, inc);
3017
	if (err < 0)
3018
		return err;
3019 3020
	if (dev->flags != old_flags)
		dev_set_rx_mode(dev);
3021
	return err;
3022 3023
}

L
Linus Torvalds 已提交
3024 3025 3026 3027 3028 3029 3030 3031 3032 3033
/**
 *	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.
3034
 *	Return 0 if successful or a negative errno code on error.
L
Linus Torvalds 已提交
3035 3036
 */

3037
int dev_set_allmulti(struct net_device *dev, int inc)
L
Linus Torvalds 已提交
3038 3039 3040
{
	unsigned short old_flags = dev->flags;

3041 3042
	ASSERT_RTNL();

L
Linus Torvalds 已提交
3043
	dev->flags |= IFF_ALLMULTI;
3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059
	dev->allmulti += inc;
	if (dev->allmulti == 0) {
		/*
		 * Avoid overflow.
		 * If inc causes overflow, untouch allmulti and return error.
		 */
		if (inc < 0)
			dev->flags &= ~IFF_ALLMULTI;
		else {
			dev->allmulti -= inc;
			printk(KERN_WARNING "%s: allmulti touches roof, "
				"set allmulti failed, allmulti feature of "
				"device might be broken.\n", dev->name);
			return -EOVERFLOW;
		}
	}
3060
	if (dev->flags ^ old_flags) {
3061
		dev_change_rx_flags(dev, IFF_ALLMULTI);
3062
		dev_set_rx_mode(dev);
3063
	}
3064
	return 0;
3065 3066 3067 3068 3069
}

/*
 *	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
3070
 *	filtering it is put in promiscuous mode while unicast addresses
3071 3072 3073 3074 3075 3076 3077 3078 3079
 *	are present.
 */
void __dev_set_rx_mode(struct net_device *dev)
{
	/* dev_open will call this function so the list will stay sane. */
	if (!(dev->flags&IFF_UP))
		return;

	if (!netif_device_present(dev))
3080
		return;
3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102

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

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

void dev_set_rx_mode(struct net_device *dev)
{
3103
	netif_addr_lock_bh(dev);
3104
	__dev_set_rx_mode(dev);
3105
	netif_addr_unlock_bh(dev);
L
Linus Torvalds 已提交
3106 3107
}

3108 3109
int __dev_addr_delete(struct dev_addr_list **list, int *count,
		      void *addr, int alen, int glbl)
3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126
{
	struct dev_addr_list *da;

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

			*list = da->next;
			kfree(da);
3127
			(*count)--;
3128 3129 3130 3131 3132 3133
			return 0;
		}
	}
	return -ENOENT;
}

3134 3135
int __dev_addr_add(struct dev_addr_list **list, int *count,
		   void *addr, int alen, int glbl)
3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152
{
	struct dev_addr_list *da;

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

3153
	da = kzalloc(sizeof(*da), GFP_ATOMIC);
3154 3155 3156 3157 3158 3159 3160 3161
	if (da == NULL)
		return -ENOMEM;
	memcpy(da->da_addr, addr, alen);
	da->da_addrlen = alen;
	da->da_users = 1;
	da->da_gusers = glbl ? 1 : 0;
	da->next = *list;
	*list = da;
3162
	(*count)++;
3163 3164 3165
	return 0;
}

3166 3167 3168
/**
 *	dev_unicast_delete	- Release secondary unicast address.
 *	@dev: device
R
Randy Dunlap 已提交
3169 3170
 *	@addr: address to delete
 *	@alen: length of @addr
3171 3172
 *
 *	Release reference to a secondary unicast address and remove it
R
Randy Dunlap 已提交
3173
 *	from the device if the reference count drops to zero.
3174 3175 3176 3177 3178 3179 3180 3181 3182
 *
 * 	The caller must hold the rtnl_mutex.
 */
int dev_unicast_delete(struct net_device *dev, void *addr, int alen)
{
	int err;

	ASSERT_RTNL();

3183
	netif_addr_lock_bh(dev);
3184 3185
	err = __dev_addr_delete(&dev->uc_list, &dev->uc_count, addr, alen, 0);
	if (!err)
3186
		__dev_set_rx_mode(dev);
3187
	netif_addr_unlock_bh(dev);
3188 3189 3190 3191 3192 3193 3194
	return err;
}
EXPORT_SYMBOL(dev_unicast_delete);

/**
 *	dev_unicast_add		- add a secondary unicast address
 *	@dev: device
3195
 *	@addr: address to add
R
Randy Dunlap 已提交
3196
 *	@alen: length of @addr
3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208
 *
 *	Add a secondary unicast address to the device or increase
 *	the reference count if it already exists.
 *
 *	The caller must hold the rtnl_mutex.
 */
int dev_unicast_add(struct net_device *dev, void *addr, int alen)
{
	int err;

	ASSERT_RTNL();

3209
	netif_addr_lock_bh(dev);
3210 3211
	err = __dev_addr_add(&dev->uc_list, &dev->uc_count, addr, alen, 0);
	if (!err)
3212
		__dev_set_rx_mode(dev);
3213
	netif_addr_unlock_bh(dev);
3214 3215 3216 3217
	return err;
}
EXPORT_SYMBOL(dev_unicast_add);

3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279
int __dev_addr_sync(struct dev_addr_list **to, int *to_count,
		    struct dev_addr_list **from, int *from_count)
{
	struct dev_addr_list *da, *next;
	int err = 0;

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

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

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

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

3280
	netif_addr_lock_bh(to);
3281 3282 3283 3284
	err = __dev_addr_sync(&to->uc_list, &to->uc_count,
			      &from->uc_list, &from->uc_count);
	if (!err)
		__dev_set_rx_mode(to);
3285
	netif_addr_unlock_bh(to);
3286 3287 3288 3289 3290
	return err;
}
EXPORT_SYMBOL(dev_unicast_sync);

/**
3291
 *	dev_unicast_unsync - Remove synchronized addresses from the destination device
3292 3293 3294 3295 3296 3297 3298 3299 3300
 *	@to: destination device
 *	@from: source device
 *
 *	Remove all addresses that were added to the destination device by
 *	dev_unicast_sync(). This function is intended to be called from the
 *	dev->stop function of layered software devices.
 */
void dev_unicast_unsync(struct net_device *to, struct net_device *from)
{
3301
	netif_addr_lock_bh(from);
3302
	netif_addr_lock(to);
3303 3304 3305 3306 3307

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

3308
	netif_addr_unlock(to);
3309
	netif_addr_unlock_bh(from);
3310 3311 3312
}
EXPORT_SYMBOL(dev_unicast_unsync);

3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326
static void __dev_addr_discard(struct dev_addr_list **list)
{
	struct dev_addr_list *tmp;

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

3327
static void dev_addr_discard(struct net_device *dev)
3328
{
3329
	netif_addr_lock_bh(dev);
3330

3331 3332 3333
	__dev_addr_discard(&dev->uc_list);
	dev->uc_count = 0;

3334 3335
	__dev_addr_discard(&dev->mc_list);
	dev->mc_count = 0;
3336

3337
	netif_addr_unlock_bh(dev);
3338 3339
}

3340 3341 3342 3343 3344 3345
/**
 *	dev_get_flags - get flags reported to userspace
 *	@dev: device
 *
 *	Get the combination of flag bits exported through APIs to userspace.
 */
L
Linus Torvalds 已提交
3346 3347 3348 3349 3350 3351
unsigned dev_get_flags(const struct net_device *dev)
{
	unsigned flags;

	flags = (dev->flags & ~(IFF_PROMISC |
				IFF_ALLMULTI |
S
Stefan Rompf 已提交
3352 3353 3354
				IFF_RUNNING |
				IFF_LOWER_UP |
				IFF_DORMANT)) |
L
Linus Torvalds 已提交
3355 3356 3357
		(dev->gflags & (IFF_PROMISC |
				IFF_ALLMULTI));

S
Stefan Rompf 已提交
3358 3359 3360 3361 3362 3363 3364 3365
	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
3366 3367 3368 3369

	return flags;
}

3370 3371 3372 3373 3374 3375 3376 3377
/**
 *	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.
 */
L
Linus Torvalds 已提交
3378 3379
int dev_change_flags(struct net_device *dev, unsigned flags)
{
3380
	int ret, changes;
L
Linus Torvalds 已提交
3381 3382
	int old_flags = dev->flags;

3383 3384
	ASSERT_RTNL();

L
Linus Torvalds 已提交
3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398
	/*
	 *	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.
	 */

3399 3400
	if ((old_flags ^ flags) & IFF_MULTICAST)
		dev_change_rx_flags(dev, IFF_MULTICAST);
3401

3402
	dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414

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

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

		if (!ret)
3415
			dev_set_rx_mode(dev);
L
Linus Torvalds 已提交
3416 3417 3418 3419 3420
	}

	if (dev->flags & IFF_UP &&
	    ((old_flags ^ dev->flags) &~ (IFF_UP | IFF_PROMISC | IFF_ALLMULTI |
					  IFF_VOLATILE)))
3421
		call_netdevice_notifiers(NETDEV_CHANGE, dev);
L
Linus Torvalds 已提交
3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438

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

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

3439 3440 3441 3442
	/* Exclude state transition flags, already notified */
	changes = (old_flags ^ dev->flags) & ~(IFF_UP | IFF_RUNNING);
	if (changes)
		rtmsg_ifinfo(RTM_NEWLINK, dev, changes);
L
Linus Torvalds 已提交
3443 3444 3445 3446

	return ret;
}

3447 3448 3449 3450 3451 3452 3453
/**
 *	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
3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473
int dev_set_mtu(struct net_device *dev, int new_mtu)
{
	int err;

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

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

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

	err = 0;
	if (dev->change_mtu)
		err = dev->change_mtu(dev, new_mtu);
	else
		dev->mtu = new_mtu;
	if (!err && dev->flags & IFF_UP)
3474
		call_netdevice_notifiers(NETDEV_CHANGEMTU, dev);
L
Linus Torvalds 已提交
3475 3476 3477
	return err;
}

3478 3479 3480 3481 3482 3483 3484
/**
 *	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
3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496
int dev_set_mac_address(struct net_device *dev, struct sockaddr *sa)
{
	int err;

	if (!dev->set_mac_address)
		return -EOPNOTSUPP;
	if (sa->sa_family != dev->type)
		return -EINVAL;
	if (!netif_device_present(dev))
		return -ENODEV;
	err = dev->set_mac_address(dev, sa);
	if (!err)
3497
		call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
L
Linus Torvalds 已提交
3498 3499 3500 3501
	return err;
}

/*
3502
 *	Perform the SIOCxIFxxx calls, inside read_lock(dev_base_lock)
L
Linus Torvalds 已提交
3503
 */
3504
static int dev_ifsioc_locked(struct net *net, struct ifreq *ifr, unsigned int cmd)
L
Linus Torvalds 已提交
3505 3506
{
	int err;
3507
	struct net_device *dev = __dev_get_by_name(net, ifr->ifr_name);
L
Linus Torvalds 已提交
3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534

	if (!dev)
		return -ENODEV;

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

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

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

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

3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589
		case SIOCGIFSLAVE:
			err = -EINVAL;
			break;

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

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

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

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

	}
	return err;
}

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

	if (!dev)
		return -ENODEV;

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

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

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

L
Linus Torvalds 已提交
3590 3591 3592 3593 3594 3595 3596 3597
		case SIOCSIFHWADDR:
			return dev_set_mac_address(dev, &ifr->ifr_hwaddr);

		case SIOCSIFHWBROADCAST:
			if (ifr->ifr_hwaddr.sa_family != dev->type)
				return -EINVAL;
			memcpy(dev->broadcast, ifr->ifr_hwaddr.sa_data,
			       min(sizeof ifr->ifr_hwaddr.sa_data, (size_t) dev->addr_len));
3598
			call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
L
Linus Torvalds 已提交
3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609
			return 0;

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

		case SIOCADDMULTI:
3610
			if ((!dev->set_multicast_list && !dev->set_rx_mode) ||
L
Linus Torvalds 已提交
3611 3612 3613 3614 3615 3616 3617 3618
			    ifr->ifr_hwaddr.sa_family != AF_UNSPEC)
				return -EINVAL;
			if (!netif_device_present(dev))
				return -ENODEV;
			return dev_mc_add(dev, ifr->ifr_hwaddr.sa_data,
					  dev->addr_len, 1);

		case SIOCDELMULTI:
3619
			if ((!dev->set_multicast_list && !dev->set_rx_mode) ||
L
Linus Torvalds 已提交
3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677
			    ifr->ifr_hwaddr.sa_family != AF_UNSPEC)
				return -EINVAL;
			if (!netif_device_present(dev))
				return -ENODEV;
			return dev_mc_delete(dev, ifr->ifr_hwaddr.sa_data,
					     dev->addr_len, 1);

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

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

		/*
		 *	Unknown or private ioctl
		 */

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

	}
	return err;
}

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

/**
 *	dev_ioctl	-	network device ioctl
3678
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
3679 3680 3681 3682 3683 3684 3685 3686 3687
 *	@cmd: command to issue
 *	@arg: pointer to a struct ifreq in user space
 *
 *	Issue ioctl functions to devices. This is normally called by the
 *	user space syscall interfaces but can sometimes be useful for
 *	other purposes. The return value is the return from the syscall if
 *	positive or a negative errno code on error.
 */

3688
int dev_ioctl(struct net *net, unsigned int cmd, void __user *arg)
L
Linus Torvalds 已提交
3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699
{
	struct ifreq ifr;
	int ret;
	char *colon;

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

	if (cmd == SIOCGIFCONF) {
3700
		rtnl_lock();
3701
		ret = dev_ifconf(net, (char __user *) arg);
3702
		rtnl_unlock();
L
Linus Torvalds 已提交
3703 3704 3705
		return ret;
	}
	if (cmd == SIOCGIFNAME)
3706
		return dev_ifname(net, (struct ifreq __user *)arg);
L
Linus Torvalds 已提交
3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735

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

	ifr.ifr_name[IFNAMSIZ-1] = 0;

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

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

	switch (cmd) {
		/*
		 *	These ioctl calls:
		 *	- can be done by all.
		 *	- atomic and do not require locking.
		 *	- return a value
		 */
		case SIOCGIFFLAGS:
		case SIOCGIFMETRIC:
		case SIOCGIFMTU:
		case SIOCGIFHWADDR:
		case SIOCGIFSLAVE:
		case SIOCGIFMAP:
		case SIOCGIFINDEX:
		case SIOCGIFTXQLEN:
3736
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3737
			read_lock(&dev_base_lock);
3738
			ret = dev_ifsioc_locked(net, &ifr, cmd);
L
Linus Torvalds 已提交
3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749
			read_unlock(&dev_base_lock);
			if (!ret) {
				if (colon)
					*colon = ':';
				if (copy_to_user(arg, &ifr,
						 sizeof(struct ifreq)))
					ret = -EFAULT;
			}
			return ret;

		case SIOCETHTOOL:
3750
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3751
			rtnl_lock();
3752
			ret = dev_ethtool(net, &ifr);
L
Linus Torvalds 已提交
3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773
			rtnl_unlock();
			if (!ret) {
				if (colon)
					*colon = ':';
				if (copy_to_user(arg, &ifr,
						 sizeof(struct ifreq)))
					ret = -EFAULT;
			}
			return ret;

		/*
		 *	These ioctl calls:
		 *	- require superuser power.
		 *	- require strict serialization.
		 *	- return a value
		 */
		case SIOCGMIIPHY:
		case SIOCGMIIREG:
		case SIOCSIFNAME:
			if (!capable(CAP_NET_ADMIN))
				return -EPERM;
3774
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3775
			rtnl_lock();
3776
			ret = dev_ifsioc(net, &ifr, cmd);
L
Linus Torvalds 已提交
3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811
			rtnl_unlock();
			if (!ret) {
				if (colon)
					*colon = ':';
				if (copy_to_user(arg, &ifr,
						 sizeof(struct ifreq)))
					ret = -EFAULT;
			}
			return ret;

		/*
		 *	These ioctl calls:
		 *	- require superuser power.
		 *	- require strict serialization.
		 *	- do not return a value
		 */
		case SIOCSIFFLAGS:
		case SIOCSIFMETRIC:
		case SIOCSIFMTU:
		case SIOCSIFMAP:
		case SIOCSIFHWADDR:
		case SIOCSIFSLAVE:
		case SIOCADDMULTI:
		case SIOCDELMULTI:
		case SIOCSIFHWBROADCAST:
		case SIOCSIFTXQLEN:
		case SIOCSMIIREG:
		case SIOCBONDENSLAVE:
		case SIOCBONDRELEASE:
		case SIOCBONDSETHWADDR:
		case SIOCBONDCHANGEACTIVE:
		case SIOCBRADDIF:
		case SIOCBRDELIF:
			if (!capable(CAP_NET_ADMIN))
				return -EPERM;
3812 3813 3814
			/* fall through */
		case SIOCBONDSLAVEINFOQUERY:
		case SIOCBONDINFOQUERY:
3815
			dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3816
			rtnl_lock();
3817
			ret = dev_ifsioc(net, &ifr, cmd);
L
Linus Torvalds 已提交
3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836
			rtnl_unlock();
			return ret;

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

		/*
		 *	Unknown or private ioctl.
		 */
		default:
			if (cmd == SIOCWANDEV ||
			    (cmd >= SIOCDEVPRIVATE &&
			     cmd <= SIOCDEVPRIVATE + 15)) {
3837
				dev_load(net, ifr.ifr_name);
L
Linus Torvalds 已提交
3838
				rtnl_lock();
3839
				ret = dev_ifsioc(net, &ifr, cmd);
L
Linus Torvalds 已提交
3840 3841 3842 3843 3844 3845 3846
				rtnl_unlock();
				if (!ret && copy_to_user(arg, &ifr,
							 sizeof(struct ifreq)))
					ret = -EFAULT;
				return ret;
			}
			/* Take care of Wireless Extensions */
3847
			if (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST)
3848
				return wext_handle_ioctl(net, &ifr, cmd, arg);
L
Linus Torvalds 已提交
3849 3850 3851 3852 3853 3854 3855
			return -EINVAL;
	}
}


/**
 *	dev_new_index	-	allocate an ifindex
3856
 *	@net: the applicable net namespace
L
Linus Torvalds 已提交
3857 3858 3859 3860 3861
 *
 *	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.
 */
3862
static int dev_new_index(struct net *net)
L
Linus Torvalds 已提交
3863 3864 3865 3866 3867
{
	static int ifindex;
	for (;;) {
		if (++ifindex <= 0)
			ifindex = 1;
3868
		if (!__dev_get_by_index(net, ifindex))
L
Linus Torvalds 已提交
3869 3870 3871 3872 3873
			return ifindex;
	}
}

/* Delayed registration/unregisteration */
3874
static LIST_HEAD(net_todo_list);
L
Linus Torvalds 已提交
3875

3876
static void net_set_todo(struct net_device *dev)
L
Linus Torvalds 已提交
3877 3878 3879 3880
{
	list_add_tail(&dev->todo_list, &net_todo_list);
}

3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924
static void rollback_registered(struct net_device *dev)
{
	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

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

		WARN_ON(1);
		return;
	}

	BUG_ON(dev->reg_state != NETREG_REGISTERED);

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

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

	dev->reg_state = NETREG_UNREGISTERING;

	synchronize_net();

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


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

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

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

	/* Notifier chain MUST detach us from master device. */
3925
	WARN_ON(dev->master);
3926 3927 3928 3929 3930 3931 3932 3933 3934

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

	synchronize_net();

	dev_put(dev);
}

3935 3936 3937
static void __netdev_init_queue_locks_one(struct net_device *dev,
					  struct netdev_queue *dev_queue,
					  void *_unused)
3938 3939
{
	spin_lock_init(&dev_queue->_xmit_lock);
3940
	netdev_set_xmit_lockdep_class(&dev_queue->_xmit_lock, dev->type);
3941 3942 3943 3944 3945
	dev_queue->xmit_lock_owner = -1;
}

static void netdev_init_queue_locks(struct net_device *dev)
{
3946 3947
	netdev_for_each_tx_queue(dev, __netdev_init_queue_locks_one, NULL);
	__netdev_init_queue_locks_one(dev, &dev->rx_queue, NULL);
3948 3949
}

L
Linus Torvalds 已提交
3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971
/**
 *	register_netdevice	- register a network device
 *	@dev: device to register
 *
 *	Take a completed network device structure and add it to the kernel
 *	interfaces. A %NETDEV_REGISTER message is sent to the netdev notifier
 *	chain. 0 is returned on success. A negative errno code is returned
 *	on a failure to set up the device, or if the name is a duplicate.
 *
 *	Callers must hold the rtnl semaphore. You may want
 *	register_netdev() instead of this.
 *
 *	BUGS:
 *	The locking appears insufficient to guarantee two parallel registers
 *	will not get the same name.
 */

int register_netdevice(struct net_device *dev)
{
	struct hlist_head *head;
	struct hlist_node *p;
	int ret;
3972
	struct net *net;
L
Linus Torvalds 已提交
3973 3974 3975 3976

	BUG_ON(dev_boot_phase);
	ASSERT_RTNL();

3977 3978
	might_sleep();

L
Linus Torvalds 已提交
3979 3980
	/* When net_device's are persistent, this will be fatal. */
	BUG_ON(dev->reg_state != NETREG_UNINITIALIZED);
3981 3982
	BUG_ON(!dev_net(dev));
	net = dev_net(dev);
L
Linus Torvalds 已提交
3983

3984
	spin_lock_init(&dev->addr_list_lock);
3985
	netdev_set_addr_lockdep_class(dev);
3986
	netdev_init_queue_locks(dev);
L
Linus Torvalds 已提交
3987 3988 3989 3990 3991 3992 3993 3994 3995

	dev->iflink = -1;

	/* Init, if this function is available */
	if (dev->init) {
		ret = dev->init(dev);
		if (ret) {
			if (ret > 0)
				ret = -EIO;
3996
			goto out;
L
Linus Torvalds 已提交
3997 3998
		}
	}
3999

L
Linus Torvalds 已提交
4000 4001
	if (!dev_valid_name(dev->name)) {
		ret = -EINVAL;
4002
		goto err_uninit;
L
Linus Torvalds 已提交
4003 4004
	}

4005
	dev->ifindex = dev_new_index(net);
L
Linus Torvalds 已提交
4006 4007 4008 4009
	if (dev->iflink == -1)
		dev->iflink = dev->ifindex;

	/* Check for existence of name */
4010
	head = dev_name_hash(net, dev->name);
L
Linus Torvalds 已提交
4011 4012 4013 4014 4015
	hlist_for_each(p, head) {
		struct net_device *d
			= hlist_entry(p, struct net_device, name_hlist);
		if (!strncmp(d->name, dev->name, IFNAMSIZ)) {
			ret = -EEXIST;
4016
			goto err_uninit;
L
Linus Torvalds 已提交
4017
		}
4018
	}
L
Linus Torvalds 已提交
4019

4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035
	/* Fix illegal checksum combinations */
	if ((dev->features & NETIF_F_HW_CSUM) &&
	    (dev->features & (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM))) {
		printk(KERN_NOTICE "%s: mixed HW and IP checksum settings.\n",
		       dev->name);
		dev->features &= ~(NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM);
	}

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


L
Linus Torvalds 已提交
4036 4037
	/* Fix illegal SG+CSUM combinations. */
	if ((dev->features & NETIF_F_SG) &&
4038
	    !(dev->features & NETIF_F_ALL_CSUM)) {
4039
		printk(KERN_NOTICE "%s: Dropping NETIF_F_SG since no checksum feature.\n",
L
Linus Torvalds 已提交
4040 4041 4042 4043 4044 4045 4046
		       dev->name);
		dev->features &= ~NETIF_F_SG;
	}

	/* TSO requires that SG is present as well. */
	if ((dev->features & NETIF_F_TSO) &&
	    !(dev->features & NETIF_F_SG)) {
4047
		printk(KERN_NOTICE "%s: Dropping NETIF_F_TSO since no SG feature.\n",
L
Linus Torvalds 已提交
4048 4049 4050
		       dev->name);
		dev->features &= ~NETIF_F_TSO;
	}
4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064
	if (dev->features & NETIF_F_UFO) {
		if (!(dev->features & NETIF_F_HW_CSUM)) {
			printk(KERN_ERR "%s: Dropping NETIF_F_UFO since no "
					"NETIF_F_HW_CSUM feature.\n",
							dev->name);
			dev->features &= ~NETIF_F_UFO;
		}
		if (!(dev->features & NETIF_F_SG)) {
			printk(KERN_ERR "%s: Dropping NETIF_F_UFO since no "
					"NETIF_F_SG feature.\n",
					dev->name);
			dev->features &= ~NETIF_F_UFO;
		}
	}
L
Linus Torvalds 已提交
4065

4066 4067 4068 4069
	/* Enable software GSO if SG is supported. */
	if (dev->features & NETIF_F_SG)
		dev->features |= NETIF_F_GSO;

4070
	netdev_initialize_kobject(dev);
4071
	ret = netdev_register_kobject(dev);
4072
	if (ret)
4073
		goto err_uninit;
4074 4075
	dev->reg_state = NETREG_REGISTERED;

L
Linus Torvalds 已提交
4076 4077 4078 4079 4080 4081 4082 4083 4084
	/*
	 *	Default initial state at registry is that the
	 *	device is present.
	 */

	set_bit(__LINK_STATE_PRESENT, &dev->state);

	dev_init_scheduler(dev);
	dev_hold(dev);
4085
	list_netdevice(dev);
L
Linus Torvalds 已提交
4086 4087

	/* Notify protocols, that a new device appeared. */
4088
	ret = call_netdevice_notifiers(NETDEV_REGISTER, dev);
4089
	ret = notifier_to_errno(ret);
4090 4091 4092 4093
	if (ret) {
		rollback_registered(dev);
		dev->reg_state = NETREG_UNREGISTERED;
	}
L
Linus Torvalds 已提交
4094 4095 4096

out:
	return ret;
4097 4098 4099 4100 4101

err_uninit:
	if (dev->uninit)
		dev->uninit(dev);
	goto out;
L
Linus Torvalds 已提交
4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112
}

/**
 *	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.
 *
4113
 *	This is a wrapper around register_netdevice that takes the rtnl semaphore
L
Linus Torvalds 已提交
4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131
 *	and expands the device name if you passed a format string to
 *	alloc_netdev.
 */
int register_netdev(struct net_device *dev)
{
	int err;

	rtnl_lock();

	/*
	 * If the name is a format string the caller wants us to do a
	 * name allocation.
	 */
	if (strchr(dev->name, '%')) {
		err = dev_alloc_name(dev, dev->name);
		if (err < 0)
			goto out;
	}
4132

L
Linus Torvalds 已提交
4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148
	err = register_netdevice(dev);
out:
	rtnl_unlock();
	return err;
}
EXPORT_SYMBOL(register_netdev);

/*
 * netdev_wait_allrefs - wait until all references are gone.
 *
 * This is called when unregistering network devices.
 *
 * Any protocol or device that holds a reference should register
 * for netdevice notification, and cleanup and put back the
 * reference if they receive an UNREGISTER event.
 * We can get stuck here if buggy protocols don't correctly
4149
 * call dev_put.
L
Linus Torvalds 已提交
4150 4151 4152 4153 4154 4155 4156 4157
 */
static void netdev_wait_allrefs(struct net_device *dev)
{
	unsigned long rebroadcast_time, warning_time;

	rebroadcast_time = warning_time = jiffies;
	while (atomic_read(&dev->refcnt) != 0) {
		if (time_after(jiffies, rebroadcast_time + 1 * HZ)) {
4158
			rtnl_lock();
L
Linus Torvalds 已提交
4159 4160

			/* Rebroadcast unregister notification */
4161
			call_netdevice_notifiers(NETDEV_UNREGISTER, dev);
L
Linus Torvalds 已提交
4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173

			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();
			}

4174
			__rtnl_unlock();
L
Linus Torvalds 已提交
4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204

			rebroadcast_time = jiffies;
		}

		msleep(250);

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

/* The sequence is:
 *
 *	rtnl_lock();
 *	...
 *	register_netdevice(x1);
 *	register_netdevice(x2);
 *	...
 *	unregister_netdevice(y1);
 *	unregister_netdevice(y2);
 *      ...
 *	rtnl_unlock();
 *	free_netdev(y1);
 *	free_netdev(y2);
 *
H
Herbert Xu 已提交
4205
 * We are invoked by rtnl_unlock().
L
Linus Torvalds 已提交
4206
 * This allows us to deal with problems:
4207
 * 1) We can delete sysfs objects which invoke hotplug
L
Linus Torvalds 已提交
4208 4209 4210
 *    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
4211 4212 4213
 *
 * 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
4214 4215 4216
 */
void netdev_run_todo(void)
{
4217
	struct list_head list;
L
Linus Torvalds 已提交
4218 4219

	/* Snapshot list, allow later requests */
4220
	list_replace_init(&net_todo_list, &list);
H
Herbert Xu 已提交
4221 4222

	__rtnl_unlock();
4223

L
Linus Torvalds 已提交
4224 4225 4226 4227 4228
	while (!list_empty(&list)) {
		struct net_device *dev
			= list_entry(list.next, struct net_device, todo_list);
		list_del(&dev->todo_list);

4229 4230 4231 4232 4233 4234
		if (unlikely(dev->reg_state != NETREG_UNREGISTERING)) {
			printk(KERN_ERR "network todo '%s' but state %d\n",
			       dev->name, dev->reg_state);
			dump_stack();
			continue;
		}
L
Linus Torvalds 已提交
4235

4236
		dev->reg_state = NETREG_UNREGISTERED;
L
Linus Torvalds 已提交
4237

4238 4239
		on_each_cpu(flush_backlog, dev, 1);

4240
		netdev_wait_allrefs(dev);
L
Linus Torvalds 已提交
4241

4242 4243
		/* paranoia */
		BUG_ON(atomic_read(&dev->refcnt));
4244 4245 4246
		WARN_ON(dev->ip_ptr);
		WARN_ON(dev->ip6_ptr);
		WARN_ON(dev->dn_ptr);
L
Linus Torvalds 已提交
4247

4248 4249
		if (dev->destructor)
			dev->destructor(dev);
4250 4251 4252

		/* Free network device */
		kobject_put(&dev->dev.kobj);
L
Linus Torvalds 已提交
4253 4254 4255
	}
}

4256
static struct net_device_stats *internal_stats(struct net_device *dev)
R
Rusty Russell 已提交
4257
{
4258
	return &dev->stats;
R
Rusty Russell 已提交
4259 4260
}

4261
static void netdev_init_one_queue(struct net_device *dev,
4262 4263
				  struct netdev_queue *queue,
				  void *_unused)
4264 4265 4266 4267
{
	queue->dev = dev;
}

4268 4269
static void netdev_init_queues(struct net_device *dev)
{
4270 4271
	netdev_init_one_queue(dev, &dev->rx_queue, NULL);
	netdev_for_each_tx_queue(dev, netdev_init_one_queue, NULL);
4272
	spin_lock_init(&dev->tx_global_lock);
4273 4274
}

L
Linus Torvalds 已提交
4275
/**
4276
 *	alloc_netdev_mq - allocate network device
L
Linus Torvalds 已提交
4277 4278 4279
 *	@sizeof_priv:	size of private data to allocate space for
 *	@name:		device name format string
 *	@setup:		callback to initialize device
4280
 *	@queue_count:	the number of subqueues to allocate
L
Linus Torvalds 已提交
4281 4282
 *
 *	Allocates a struct net_device with private data area for driver use
4283 4284
 *	and performs basic initialization.  Also allocates subquue structs
 *	for each queue on the device at the end of the netdevice.
L
Linus Torvalds 已提交
4285
 */
4286 4287
struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name,
		void (*setup)(struct net_device *), unsigned int queue_count)
L
Linus Torvalds 已提交
4288
{
4289
	struct netdev_queue *tx;
L
Linus Torvalds 已提交
4290
	struct net_device *dev;
4291
	size_t alloc_size;
4292
	void *p;
L
Linus Torvalds 已提交
4293

4294 4295
	BUG_ON(strlen(name) >= sizeof(dev->name));

4296
	alloc_size = sizeof(struct net_device);
4297 4298 4299 4300 4301 4302 4303
	if (sizeof_priv) {
		/* ensure 32-byte alignment of private area */
		alloc_size = (alloc_size + NETDEV_ALIGN_CONST) & ~NETDEV_ALIGN_CONST;
		alloc_size += sizeof_priv;
	}
	/* ensure 32-byte alignment of whole construct */
	alloc_size += NETDEV_ALIGN_CONST;
L
Linus Torvalds 已提交
4304

4305
	p = kzalloc(alloc_size, GFP_KERNEL);
L
Linus Torvalds 已提交
4306
	if (!p) {
4307
		printk(KERN_ERR "alloc_netdev: Unable to allocate device.\n");
L
Linus Torvalds 已提交
4308 4309 4310
		return NULL;
	}

4311
	tx = kcalloc(queue_count, sizeof(struct netdev_queue), GFP_KERNEL);
4312 4313 4314 4315 4316 4317 4318
	if (!tx) {
		printk(KERN_ERR "alloc_netdev: Unable to allocate "
		       "tx qdiscs.\n");
		kfree(p);
		return NULL;
	}

L
Linus Torvalds 已提交
4319 4320 4321
	dev = (struct net_device *)
		(((long)p + NETDEV_ALIGN_CONST) & ~NETDEV_ALIGN_CONST);
	dev->padded = (char *)dev - (char *)p;
4322
	dev_net_set(dev, &init_net);
L
Linus Torvalds 已提交
4323

4324 4325
	dev->_tx = tx;
	dev->num_tx_queues = queue_count;
4326
	dev->real_num_tx_queues = queue_count;
4327

4328 4329
	if (sizeof_priv) {
		dev->priv = ((char *)dev +
4330
			     ((sizeof(struct net_device) + NETDEV_ALIGN_CONST)
4331 4332 4333
			      & ~NETDEV_ALIGN_CONST));
	}

4334
	dev->gso_max_size = GSO_MAX_SIZE;
L
Linus Torvalds 已提交
4335

4336 4337
	netdev_init_queues(dev);

4338
	dev->get_stats = internal_stats;
4339
	netpoll_netdev_init(dev);
L
Linus Torvalds 已提交
4340 4341 4342 4343
	setup(dev);
	strcpy(dev->name, name);
	return dev;
}
4344
EXPORT_SYMBOL(alloc_netdev_mq);
L
Linus Torvalds 已提交
4345 4346 4347 4348 4349

/**
 *	free_netdev - free network device
 *	@dev: device
 *
4350 4351
 *	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
4352 4353 4354 4355
 *	If this is the last reference then it will be freed.
 */
void free_netdev(struct net_device *dev)
{
4356 4357
	release_net(dev_net(dev));

4358 4359
	kfree(dev->_tx);

S
Stephen Hemminger 已提交
4360
	/*  Compatibility with error handling in drivers */
L
Linus Torvalds 已提交
4361 4362 4363 4364 4365 4366 4367 4368
	if (dev->reg_state == NETREG_UNINITIALIZED) {
		kfree((char *)dev - dev->padded);
		return;
	}

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

4369 4370
	/* will free via device release */
	put_device(&dev->dev);
L
Linus Torvalds 已提交
4371
}
4372

4373 4374 4375 4376 4377 4378
/**
 *	synchronize_net -  Synchronize with packet receive processing
 *
 *	Wait for packets currently being received to be done.
 *	Does not block later packets from starting.
 */
4379
void synchronize_net(void)
L
Linus Torvalds 已提交
4380 4381
{
	might_sleep();
4382
	synchronize_rcu();
L
Linus Torvalds 已提交
4383 4384 4385 4386 4387 4388 4389
}

/**
 *	unregister_netdevice - remove device from the kernel
 *	@dev: device
 *
 *	This function shuts down a device interface and removes it
4390
 *	from the kernel tables.
L
Linus Torvalds 已提交
4391 4392 4393 4394 4395
 *
 *	Callers must hold the rtnl semaphore.  You may want
 *	unregister_netdev() instead of this.
 */

4396
void unregister_netdevice(struct net_device *dev)
L
Linus Torvalds 已提交
4397
{
4398 4399
	ASSERT_RTNL();

4400
	rollback_registered(dev);
L
Linus Torvalds 已提交
4401 4402 4403 4404 4405 4406 4407 4408 4409
	/* Finish processing unregister after unlock */
	net_set_todo(dev);
}

/**
 *	unregister_netdev - remove device from the kernel
 *	@dev: device
 *
 *	This function shuts down a device interface and removes it
4410
 *	from the kernel tables.
L
Linus Torvalds 已提交
4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424
 *
 *	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);

4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458
/**
 *	dev_change_net_namespace - move device to different nethost namespace
 *	@dev: device
 *	@net: network namespace
 *	@pat: If not NULL name pattern to try if the current device name
 *	      is already taken in the destination network namespace.
 *
 *	This function shuts down a device interface and moves it
 *	to a new network namespace. On success 0 is returned, on
 *	a failure a netagive errno code is returned.
 *
 *	Callers must hold the rtnl semaphore.
 */

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

	ASSERT_RTNL();

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

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

	/* Get out if there is nothing todo */
	err = 0;
4459
	if (net_eq(dev_net(dev), net))
4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487
		goto out;

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

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

	/* If device is running close it first. */
4488
	dev_close(dev);
4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509

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

	synchronize_net();

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

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

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

	/* Actually switch the network namespace */
4510
	dev_net_set(dev, net);
4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523

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

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

4524
	/* Fixup kobjects */
4525 4526
	netdev_unregister_kobject(dev);
	err = netdev_register_kobject(dev);
4527
	WARN_ON(err);
4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540

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

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

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

L
Linus Torvalds 已提交
4541 4542 4543 4544 4545
static int dev_cpu_callback(struct notifier_block *nfb,
			    unsigned long action,
			    void *ocpu)
{
	struct sk_buff **list_skb;
4546
	struct Qdisc **list_net;
L
Linus Torvalds 已提交
4547 4548 4549 4550
	struct sk_buff *skb;
	unsigned int cpu, oldcpu = (unsigned long)ocpu;
	struct softnet_data *sd, *oldsd;

4551
	if (action != CPU_DEAD && action != CPU_DEAD_FROZEN)
L
Linus Torvalds 已提交
4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584
		return NOTIFY_OK;

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

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

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

	raise_softirq_irqoff(NET_TX_SOFTIRQ);
	local_irq_enable();

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

	return NOTIFY_OK;
}

4585 4586
#ifdef CONFIG_NET_DMA
/**
R
Randy Dunlap 已提交
4587 4588 4589 4590 4591
 * net_dma_rebalance - try to maintain one DMA channel per CPU
 * @net_dma: DMA client and associated data (lock, channels, channel_mask)
 *
 * This is called when the number of channels allocated to the net_dma client
 * changes.  The net_dma client tries to have one DMA channel per CPU.
4592
 */
4593 4594

static void net_dma_rebalance(struct net_dma *net_dma)
4595
{
4596
	unsigned int cpu, i, n, chan_idx;
4597 4598
	struct dma_chan *chan;

4599
	if (cpus_empty(net_dma->channel_mask)) {
4600
		for_each_online_cpu(cpu)
A
Alexey Dobriyan 已提交
4601
			rcu_assign_pointer(per_cpu(softnet_data, cpu).net_dma, NULL);
4602 4603 4604 4605 4606 4607
		return;
	}

	i = 0;
	cpu = first_cpu(cpu_online_map);

4608
	for_each_cpu_mask_nr(chan_idx, net_dma->channel_mask) {
4609 4610 4611 4612 4613
		chan = net_dma->channels[chan_idx];

		n = ((num_online_cpus() / cpus_weight(net_dma->channel_mask))
		   + (i < (num_online_cpus() %
			cpus_weight(net_dma->channel_mask)) ? 1 : 0));
4614 4615

		while(n) {
A
Alexey Dobriyan 已提交
4616
			per_cpu(softnet_data, cpu).net_dma = chan;
4617 4618 4619 4620 4621 4622 4623 4624 4625 4626
			cpu = next_cpu(cpu, cpu_online_map);
			n--;
		}
		i++;
	}
}

/**
 * netdev_dma_event - event callback for the net_dma_client
 * @client: should always be net_dma_client
R
Randy Dunlap 已提交
4627
 * @chan: DMA channel for the event
R
Randy Dunlap 已提交
4628
 * @state: DMA state to be handled
4629
 */
4630 4631 4632 4633 4634 4635 4636 4637 4638 4639 4640 4641
static enum dma_state_client
netdev_dma_event(struct dma_client *client, struct dma_chan *chan,
	enum dma_state state)
{
	int i, found = 0, pos = -1;
	struct net_dma *net_dma =
		container_of(client, struct net_dma, client);
	enum dma_state_client ack = DMA_DUP; /* default: take no action */

	spin_lock(&net_dma->lock);
	switch (state) {
	case DMA_RESOURCE_AVAILABLE:
4642
		for (i = 0; i < nr_cpu_ids; i++)
4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654
			if (net_dma->channels[i] == chan) {
				found = 1;
				break;
			} else if (net_dma->channels[i] == NULL && pos < 0)
				pos = i;

		if (!found && pos >= 0) {
			ack = DMA_ACK;
			net_dma->channels[pos] = chan;
			cpu_set(pos, net_dma->channel_mask);
			net_dma_rebalance(net_dma);
		}
4655 4656
		break;
	case DMA_RESOURCE_REMOVED:
4657
		for (i = 0; i < nr_cpu_ids; i++)
4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669
			if (net_dma->channels[i] == chan) {
				found = 1;
				pos = i;
				break;
			}

		if (found) {
			ack = DMA_ACK;
			cpu_clear(pos, net_dma->channel_mask);
			net_dma->channels[i] = NULL;
			net_dma_rebalance(net_dma);
		}
4670 4671 4672 4673
		break;
	default:
		break;
	}
4674 4675 4676
	spin_unlock(&net_dma->lock);

	return ack;
4677 4678 4679
}

/**
4680
 * netdev_dma_register - register the networking subsystem as a DMA client
4681 4682 4683
 */
static int __init netdev_dma_register(void)
{
4684 4685 4686 4687 4688 4689 4690
	net_dma.channels = kzalloc(nr_cpu_ids * sizeof(struct net_dma),
								GFP_KERNEL);
	if (unlikely(!net_dma.channels)) {
		printk(KERN_NOTICE
				"netdev_dma: no memory for net_dma.channels\n");
		return -ENOMEM;
	}
4691 4692 4693 4694
	spin_lock_init(&net_dma.lock);
	dma_cap_set(DMA_MEMCPY, net_dma.client.cap_mask);
	dma_async_client_register(&net_dma.client);
	dma_async_client_chan_request(&net_dma.client);
4695 4696 4697 4698 4699 4700
	return 0;
}

#else
static int __init netdev_dma_register(void) { return -ENODEV; }
#endif /* CONFIG_NET_DMA */
L
Linus Torvalds 已提交
4701

4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725
/**
 *	netdev_compute_feature - compute conjunction of two feature sets
 *	@all: first feature set
 *	@one: second feature set
 *
 *	Computes a new feature set after adding a device with feature set
 *	@one to the master device with current feature set @all.  Returns
 *	the new feature set.
 */
int netdev_compute_features(unsigned long all, unsigned long one)
{
	/* if device needs checksumming, downgrade to hw checksumming */
	if (all & NETIF_F_NO_CSUM && !(one & NETIF_F_NO_CSUM))
		all ^= NETIF_F_NO_CSUM | NETIF_F_HW_CSUM;

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

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

4726 4727 4728 4729 4730 4731
	/*
	 * If even one device supports a GSO protocol with software fallback,
	 * enable it for all.
	 */
	all |= one & NETIF_F_GSO_SOFTWARE;

4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746
	/* If even one device supports robust GSO, enable it for all. */
	if (one & NETIF_F_GSO_ROBUST)
		all |= NETIF_F_GSO_ROBUST;

	all &= one | NETIF_F_LLTX;

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

	return all;
}
EXPORT_SYMBOL(netdev_compute_features);

4747 4748 4749 4750 4751 4752 4753 4754 4755 4756 4757 4758 4759
static struct hlist_head *netdev_create_hash(void)
{
	int i;
	struct hlist_head *hash;

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

	return hash;
}

4760
/* Initialize per network namespace state */
4761
static int __net_init netdev_init(struct net *net)
4762 4763 4764
{
	INIT_LIST_HEAD(&net->dev_base_head);

4765 4766 4767
	net->dev_name_head = netdev_create_hash();
	if (net->dev_name_head == NULL)
		goto err_name;
4768

4769 4770 4771
	net->dev_index_head = netdev_create_hash();
	if (net->dev_index_head == NULL)
		goto err_idx;
4772 4773

	return 0;
4774 4775 4776 4777 4778

err_idx:
	kfree(net->dev_name_head);
err_name:
	return -ENOMEM;
4779 4780
}

4781 4782 4783 4784 4785 4786 4787 4788
/**
 *	netdev_drivername - network driver for the device
 *	@dev: network device
 *	@buffer: buffer for resulting name
 *	@len: size of buffer
 *
 *	Determine network driver for device.
 */
4789
char *netdev_drivername(const struct net_device *dev, char *buffer, int len)
4790
{
4791 4792
	const struct device_driver *driver;
	const struct device *parent;
4793 4794 4795 4796 4797 4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808

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

	parent = dev->dev.parent;

	if (!parent)
		return buffer;

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

4809
static void __net_exit netdev_exit(struct net *net)
4810 4811 4812 4813 4814
{
	kfree(net->dev_name_head);
	kfree(net->dev_index_head);
}

4815
static struct pernet_operations __net_initdata netdev_net_ops = {
4816 4817 4818 4819
	.init = netdev_init,
	.exit = netdev_exit,
};

4820
static void __net_exit default_device_exit(struct net *net)
4821 4822 4823 4824 4825 4826 4827 4828 4829
{
	struct net_device *dev, *next;
	/*
	 * Push all migratable of the network devices back to the
	 * initial network namespace
	 */
	rtnl_lock();
	for_each_netdev_safe(net, dev, next) {
		int err;
4830
		char fb_name[IFNAMSIZ];
4831 4832 4833 4834 4835 4836

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

		/* Push remaing network devices to init_net */
4837 4838
		snprintf(fb_name, IFNAMSIZ, "dev%d", dev->ifindex);
		err = dev_change_net_namespace(dev, &init_net, fb_name);
4839
		if (err) {
4840
			printk(KERN_EMERG "%s: failed to move %s to init_net: %d\n",
4841
				__func__, dev->name, err);
4842
			BUG();
4843 4844 4845 4846 4847
		}
	}
	rtnl_unlock();
}

4848
static struct pernet_operations __net_initdata default_device_ops = {
4849 4850 4851
	.exit = default_device_exit,
};

L
Linus Torvalds 已提交
4852 4853 4854 4855 4856 4857 4858 4859 4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871
/*
 *	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;

4872
	if (netdev_kobject_init())
L
Linus Torvalds 已提交
4873 4874 4875
		goto out;

	INIT_LIST_HEAD(&ptype_all);
4876
	for (i = 0; i < PTYPE_HASH_SIZE; i++)
L
Linus Torvalds 已提交
4877 4878
		INIT_LIST_HEAD(&ptype_base[i]);

4879 4880
	if (register_pernet_subsys(&netdev_net_ops))
		goto out;
L
Linus Torvalds 已提交
4881

4882 4883 4884
	if (register_pernet_device(&default_device_ops))
		goto out;

L
Linus Torvalds 已提交
4885 4886 4887 4888
	/*
	 *	Initialise the packet receive queues.
	 */

4889
	for_each_possible_cpu(i) {
L
Linus Torvalds 已提交
4890 4891 4892 4893 4894 4895
		struct softnet_data *queue;

		queue = &per_cpu(softnet_data, i);
		skb_queue_head_init(&queue->input_pkt_queue);
		queue->completion_queue = NULL;
		INIT_LIST_HEAD(&queue->poll_list);
4896 4897 4898

		queue->backlog.poll = process_backlog;
		queue->backlog.weight = weight_p;
L
Linus Torvalds 已提交
4899 4900
	}

4901 4902
	netdev_dma_register();

L
Linus Torvalds 已提交
4903 4904
	dev_boot_phase = 0;

4905 4906
	open_softirq(NET_TX_SOFTIRQ, net_tx_action);
	open_softirq(NET_RX_SOFTIRQ, net_rx_action);
L
Linus Torvalds 已提交
4907 4908 4909 4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920

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

subsys_initcall(net_dev_init);

EXPORT_SYMBOL(__dev_get_by_index);
EXPORT_SYMBOL(__dev_get_by_name);
EXPORT_SYMBOL(__dev_remove_pack);
4921
EXPORT_SYMBOL(dev_valid_name);
L
Linus Torvalds 已提交
4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961
EXPORT_SYMBOL(dev_add_pack);
EXPORT_SYMBOL(dev_alloc_name);
EXPORT_SYMBOL(dev_close);
EXPORT_SYMBOL(dev_get_by_flags);
EXPORT_SYMBOL(dev_get_by_index);
EXPORT_SYMBOL(dev_get_by_name);
EXPORT_SYMBOL(dev_open);
EXPORT_SYMBOL(dev_queue_xmit);
EXPORT_SYMBOL(dev_remove_pack);
EXPORT_SYMBOL(dev_set_allmulti);
EXPORT_SYMBOL(dev_set_promiscuity);
EXPORT_SYMBOL(dev_change_flags);
EXPORT_SYMBOL(dev_set_mtu);
EXPORT_SYMBOL(dev_set_mac_address);
EXPORT_SYMBOL(free_netdev);
EXPORT_SYMBOL(netdev_boot_setup_check);
EXPORT_SYMBOL(netdev_set_master);
EXPORT_SYMBOL(netdev_state_change);
EXPORT_SYMBOL(netif_receive_skb);
EXPORT_SYMBOL(netif_rx);
EXPORT_SYMBOL(register_gifconf);
EXPORT_SYMBOL(register_netdevice);
EXPORT_SYMBOL(register_netdevice_notifier);
EXPORT_SYMBOL(skb_checksum_help);
EXPORT_SYMBOL(synchronize_net);
EXPORT_SYMBOL(unregister_netdevice);
EXPORT_SYMBOL(unregister_netdevice_notifier);
EXPORT_SYMBOL(net_enable_timestamp);
EXPORT_SYMBOL(net_disable_timestamp);
EXPORT_SYMBOL(dev_get_flags);

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

EXPORT_SYMBOL(dev_load);

EXPORT_PER_CPU_SYMBOL(softnet_data);